diff options
Diffstat (limited to 'fs')
338 files changed, 6939 insertions, 4746 deletions
diff --git a/fs/9p/9p.h b/fs/9p/9p.h index 95d72aec1c1..94e2f92ab2e 100644 --- a/fs/9p/9p.h +++ b/fs/9p/9p.h @@ -8,9 +8,8 @@ * Copyright (C) 2002 by Ron Minnich <rminnich@lanl.gov> * * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. + * it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of @@ -235,6 +234,7 @@ struct Tcreate { struct v9fs_str name; u32 perm; u8 mode; + struct v9fs_str extension; }; struct Rcreate { @@ -364,7 +364,7 @@ int v9fs_t_remove(struct v9fs_session_info *v9ses, u32 fid, struct v9fs_fcall **rcall); int v9fs_t_create(struct v9fs_session_info *v9ses, u32 fid, char *name, - u32 perm, u8 mode, struct v9fs_fcall **rcall); + u32 perm, u8 mode, char *extension, struct v9fs_fcall **rcall); int v9fs_t_read(struct v9fs_session_info *v9ses, u32 fid, u64 offset, u32 count, struct v9fs_fcall **rcall); @@ -372,3 +372,4 @@ int v9fs_t_read(struct v9fs_session_info *v9ses, u32 fid, int v9fs_t_write(struct v9fs_session_info *v9ses, u32 fid, u64 offset, u32 count, const char __user * data, struct v9fs_fcall **rcall); +int v9fs_printfcall(char *, int, struct v9fs_fcall *, int); diff --git a/fs/9p/Makefile b/fs/9p/Makefile index 2f4ce43f7b6..87897f84dfb 100644 --- a/fs/9p/Makefile +++ b/fs/9p/Makefile @@ -1,10 +1,9 @@ -obj-$(CONFIG_9P_FS) := 9p2000.o +obj-$(CONFIG_9P_FS) := 9p.o -9p2000-objs := \ +9p-objs := \ trans_fd.o \ - trans_sock.o \ mux.o \ - 9p.o \ + fcall.o \ conv.o \ vfs_super.o \ vfs_inode.o \ @@ -14,5 +13,6 @@ obj-$(CONFIG_9P_FS) := 9p2000.o vfs_dentry.o \ error.o \ v9fs.o \ - fid.o + fid.o \ + fcprint.o diff --git a/fs/9p/conv.c b/fs/9p/conv.c index bba81714246..a767e05b60b 100644 --- a/fs/9p/conv.c +++ b/fs/9p/conv.c @@ -8,9 +8,8 @@ * Copyright (C) 2002 by Ron Minnich <rminnich@lanl.gov> * * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. + * it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of @@ -666,7 +665,8 @@ struct v9fs_fcall *v9fs_create_topen(u32 fid, u8 mode) return fc; } -struct v9fs_fcall *v9fs_create_tcreate(u32 fid, char *name, u32 perm, u8 mode) +struct v9fs_fcall *v9fs_create_tcreate(u32 fid, char *name, u32 perm, u8 mode, + char *extension, int extended) { int size; struct v9fs_fcall *fc; @@ -674,6 +674,9 @@ struct v9fs_fcall *v9fs_create_tcreate(u32 fid, char *name, u32 perm, u8 mode) struct cbuf *bufp = &buffer; size = 4 + 2 + strlen(name) + 4 + 1; /* fid[4] name[s] perm[4] mode[1] */ + if (extended && extension!=NULL) + size += 2 + strlen(extension); /* extension[s] */ + fc = v9fs_create_common(bufp, size, TCREATE); if (IS_ERR(fc)) goto error; @@ -682,6 +685,8 @@ struct v9fs_fcall *v9fs_create_tcreate(u32 fid, char *name, u32 perm, u8 mode) v9fs_put_str(bufp, name, &fc->params.tcreate.name); v9fs_put_int32(bufp, perm, &fc->params.tcreate.perm); v9fs_put_int8(bufp, mode, &fc->params.tcreate.mode); + if (extended) + v9fs_put_str(bufp, extension, &fc->params.tcreate.extension); if (buf_check_overflow(bufp)) { kfree(fc); diff --git a/fs/9p/conv.h b/fs/9p/conv.h index f5896628dae..dd5b6b1b610 100644 --- a/fs/9p/conv.h +++ b/fs/9p/conv.h @@ -8,9 +8,8 @@ * Copyright (C) 2002 by Ron Minnich <rminnich@lanl.gov> * * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. + * it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of @@ -39,7 +38,8 @@ struct v9fs_fcall *v9fs_create_tflush(u16 oldtag); struct v9fs_fcall *v9fs_create_twalk(u32 fid, u32 newfid, u16 nwname, char **wnames); struct v9fs_fcall *v9fs_create_topen(u32 fid, u8 mode); -struct v9fs_fcall *v9fs_create_tcreate(u32 fid, char *name, u32 perm, u8 mode); +struct v9fs_fcall *v9fs_create_tcreate(u32 fid, char *name, u32 perm, u8 mode, + char *extension, int extended); struct v9fs_fcall *v9fs_create_tread(u32 fid, u64 offset, u32 count); struct v9fs_fcall *v9fs_create_twrite(u32 fid, u64 offset, u32 count, const char __user *data); diff --git a/fs/9p/debug.h b/fs/9p/debug.h index fe551032788..4228c0bb3c3 100644 --- a/fs/9p/debug.h +++ b/fs/9p/debug.h @@ -5,9 +5,8 @@ * Copyright (C) 2002 by Ron Minnich <rminnich@lanl.gov> * * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. + * it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of @@ -30,6 +29,7 @@ #define DEBUG_MUX (1<<5) #define DEBUG_TRANS (1<<6) #define DEBUG_SLABS (1<<7) +#define DEBUG_FCALL (1<<8) #define DEBUG_DUMP_PKT 0 diff --git a/fs/9p/error.c b/fs/9p/error.c index e4b6f8f38b6..981fe8ecd78 100644 --- a/fs/9p/error.c +++ b/fs/9p/error.c @@ -11,9 +11,8 @@ * Copyright (C) 2002 by Ron Minnich <rminnich@lanl.gov> * * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. + * it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of diff --git a/fs/9p/error.h b/fs/9p/error.h index a9794e85fe5..5f3ca522b31 100644 --- a/fs/9p/error.h +++ b/fs/9p/error.h @@ -12,9 +12,8 @@ * Copyright (C) 2002 by Ron Minnich <rminnich@lanl.gov> * * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. + * it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of diff --git a/fs/9p/9p.c b/fs/9p/fcall.c index c148e6ba07e..71742ba150c 100644 --- a/fs/9p/9p.c +++ b/fs/9p/fcall.c @@ -1,5 +1,5 @@ /* - * linux/fs/9p/9p.c + * linux/fs/9p/fcall.c * * This file contains functions to perform synchronous 9P calls * @@ -8,9 +8,8 @@ * Copyright (C) 2002 by Ron Minnich <rminnich@lanl.gov> * * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. + * it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of @@ -152,10 +151,9 @@ v9fs_t_clunk(struct v9fs_session_info *v9ses, u32 fid) /** * v9fs_v9fs_t_flush - flush a pending transaction * @v9ses: 9P2000 session information - * @tag: tid to release + * @tag: tag to release * */ - int v9fs_t_flush(struct v9fs_session_info *v9ses, u16 oldtag) { int ret; @@ -172,7 +170,7 @@ int v9fs_t_flush(struct v9fs_session_info *v9ses, u16 oldtag) return ret; } -#endif /* 0 */ +#endif /** * v9fs_t_stat - read a file's meta-data @@ -334,8 +332,8 @@ v9fs_t_remove(struct v9fs_session_info *v9ses, u32 fid, */ int -v9fs_t_create(struct v9fs_session_info *v9ses, u32 fid, char *name, - u32 perm, u8 mode, struct v9fs_fcall **rcp) +v9fs_t_create(struct v9fs_session_info *v9ses, u32 fid, char *name, u32 perm, + u8 mode, char *extension, struct v9fs_fcall **rcp) { int ret; struct v9fs_fcall *tc; @@ -343,7 +341,9 @@ v9fs_t_create(struct v9fs_session_info *v9ses, u32 fid, char *name, dprintk(DEBUG_9P, "fid %d name '%s' perm %x mode %d\n", fid, name, perm, mode); - tc = v9fs_create_tcreate(fid, name, perm, mode); + tc = v9fs_create_tcreate(fid, name, perm, mode, extension, + v9ses->extended); + if (!IS_ERR(tc)) { ret = v9fs_mux_rpc(v9ses->mux, tc, rcp); kfree(tc); diff --git a/fs/9p/fcprint.c b/fs/9p/fcprint.c new file mode 100644 index 00000000000..583e827baeb --- /dev/null +++ b/fs/9p/fcprint.c @@ -0,0 +1,346 @@ +/* + * linux/fs/9p/fcprint.c + * + * Print 9P call. + * + * Copyright (C) 2005 by Latchesar Ionkov <lucho@ionkov.net> + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to: + * Free Software Foundation + * 51 Franklin Street, Fifth Floor + * Boston, MA 02111-1301 USA + * + */ +#include <linux/config.h> +#include <linux/module.h> +#include <linux/errno.h> +#include <linux/fs.h> +#include <linux/idr.h> + +#include "debug.h" +#include "v9fs.h" +#include "9p.h" +#include "mux.h" + +static int +v9fs_printqid(char *buf, int buflen, struct v9fs_qid *q) +{ + int n; + char b[10]; + + n = 0; + if (q->type & V9FS_QTDIR) + b[n++] = 'd'; + if (q->type & V9FS_QTAPPEND) + b[n++] = 'a'; + if (q->type & V9FS_QTAUTH) + b[n++] = 'A'; + if (q->type & V9FS_QTEXCL) + b[n++] = 'l'; + if (q->type & V9FS_QTTMP) + b[n++] = 't'; + if (q->type & V9FS_QTSYMLINK) + b[n++] = 'L'; + b[n] = '\0'; + + return scnprintf(buf, buflen, "(%.16llx %x %s)", (long long int) q->path, + q->version, b); +} + +static int +v9fs_printperm(char *buf, int buflen, int perm) +{ + int n; + char b[15]; + + n = 0; + if (perm & V9FS_DMDIR) + b[n++] = 'd'; + if (perm & V9FS_DMAPPEND) + b[n++] = 'a'; + if (perm & V9FS_DMAUTH) + b[n++] = 'A'; + if (perm & V9FS_DMEXCL) + b[n++] = 'l'; + if (perm & V9FS_DMTMP) + b[n++] = 't'; + if (perm & V9FS_DMDEVICE) + b[n++] = 'D'; + if (perm & V9FS_DMSOCKET) + b[n++] = 'S'; + if (perm & V9FS_DMNAMEDPIPE) + b[n++] = 'P'; + if (perm & V9FS_DMSYMLINK) + b[n++] = 'L'; + b[n] = '\0'; + + return scnprintf(buf, buflen, "%s%03o", b, perm&077); +} + +static int +v9fs_printstat(char *buf, int buflen, struct v9fs_stat *st, int extended) +{ + int n; + + n = scnprintf(buf, buflen, "'%.*s' '%.*s'", st->name.len, + st->name.str, st->uid.len, st->uid.str); + if (extended) + n += scnprintf(buf+n, buflen-n, "(%d)", st->n_uid); + + n += scnprintf(buf+n, buflen-n, " '%.*s'", st->gid.len, st->gid.str); + if (extended) + n += scnprintf(buf+n, buflen-n, "(%d)", st->n_gid); + + n += scnprintf(buf+n, buflen-n, " '%.*s'", st->muid.len, st->muid.str); + if (extended) + n += scnprintf(buf+n, buflen-n, "(%d)", st->n_muid); + + n += scnprintf(buf+n, buflen-n, " q "); + n += v9fs_printqid(buf+n, buflen-n, &st->qid); + n += scnprintf(buf+n, buflen-n, " m "); + n += v9fs_printperm(buf+n, buflen-n, st->mode); + n += scnprintf(buf+n, buflen-n, " at %d mt %d l %lld", + st->atime, st->mtime, (long long int) st->length); + + if (extended) + n += scnprintf(buf+n, buflen-n, " ext '%.*s'", + st->extension.len, st->extension.str); + + return n; +} + +static int +v9fs_dumpdata(char *buf, int buflen, u8 *data, int datalen) +{ + int i, n; + + i = n = 0; + while (i < datalen) { + n += scnprintf(buf + n, buflen - n, "%02x", data[i]); + if (i%4 == 3) + n += scnprintf(buf + n, buflen - n, " "); + if (i%32 == 31) + n += scnprintf(buf + n, buflen - n, "\n"); + + i++; + } + n += scnprintf(buf + n, buflen - n, "\n"); + + return n; +} + +static int +v9fs_printdata(char *buf, int buflen, u8 *data, int datalen) +{ + return v9fs_dumpdata(buf, buflen, data, datalen<16?datalen:16); +} + +int +v9fs_printfcall(char *buf, int buflen, struct v9fs_fcall *fc, int extended) +{ + int i, ret, type, tag; + + if (!fc) + return scnprintf(buf, buflen, "<NULL>"); + + type = fc->id; + tag = fc->tag; + + ret = 0; + switch (type) { + case TVERSION: + ret += scnprintf(buf+ret, buflen-ret, + "Tversion tag %u msize %u version '%.*s'", tag, + fc->params.tversion.msize, fc->params.tversion.version.len, + fc->params.tversion.version.str); + break; + + case RVERSION: + ret += scnprintf(buf+ret, buflen-ret, + "Rversion tag %u msize %u version '%.*s'", tag, + fc->params.rversion.msize, fc->params.rversion.version.len, + fc->params.rversion.version.str); + break; + + case TAUTH: + ret += scnprintf(buf+ret, buflen-ret, + "Tauth tag %u afid %d uname '%.*s' aname '%.*s'", tag, + fc->params.tauth.afid, fc->params.tauth.uname.len, + fc->params.tauth.uname.str, fc->params.tauth.aname.len, + fc->params.tauth.aname.str); + break; + + case RAUTH: + ret += scnprintf(buf+ret, buflen-ret, "Rauth tag %u qid ", tag); + v9fs_printqid(buf+ret, buflen-ret, &fc->params.rauth.qid); + break; + + case TATTACH: + ret += scnprintf(buf+ret, buflen-ret, + "Tattach tag %u fid %d afid %d uname '%.*s' aname '%.*s'", + tag, fc->params.tattach.fid, fc->params.tattach.afid, + fc->params.tattach.uname.len, fc->params.tattach.uname.str, + fc->params.tattach.aname.len, fc->params.tattach.aname.str); + break; + + case RATTACH: + ret += scnprintf(buf+ret, buflen-ret, "Rattach tag %u qid ", tag); + v9fs_printqid(buf+ret, buflen-ret, &fc->params.rattach.qid); + break; + + case RERROR: + ret += scnprintf(buf+ret, buflen-ret, "Rerror tag %u ename '%.*s'", + tag, fc->params.rerror.error.len, + fc->params.rerror.error.str); + if (extended) + ret += scnprintf(buf+ret, buflen-ret, " ecode %d\n", + fc->params.rerror.errno); + break; + + case TFLUSH: + ret += scnprintf(buf+ret, buflen-ret, "Tflush tag %u oldtag %u", + tag, fc->params.tflush.oldtag); + break; + + case RFLUSH: + ret += scnprintf(buf+ret, buflen-ret, "Rflush tag %u", tag); + break; + + case TWALK: + ret += scnprintf(buf+ret, buflen-ret, + "Twalk tag %u fid %d newfid %d nwname %d", tag, + fc->params.twalk.fid, fc->params.twalk.newfid, + fc->params.twalk.nwname); + for(i = 0; i < fc->params.twalk.nwname; i++) + ret += scnprintf(buf+ret, buflen-ret," '%.*s'", + fc->params.twalk.wnames[i].len, + fc->params.twalk.wnames[i].str); + break; + + case RWALK: + ret += scnprintf(buf+ret, buflen-ret, "Rwalk tag %u nwqid %d", + tag, fc->params.rwalk.nwqid); + for(i = 0; i < fc->params.rwalk.nwqid; i++) + ret += v9fs_printqid(buf+ret, buflen-ret, + &fc->params.rwalk.wqids[i]); + break; + + case TOPEN: + ret += scnprintf(buf+ret, buflen-ret, + "Topen tag %u fid %d mode %d", tag, + fc->params.topen.fid, fc->params.topen.mode); + break; + + case ROPEN: + ret += scnprintf(buf+ret, buflen-ret, "Ropen tag %u", tag); + ret += v9fs_printqid(buf+ret, buflen-ret, &fc->params.ropen.qid); + ret += scnprintf(buf+ret, buflen-ret," iounit %d", + fc->params.ropen.iounit); + break; + + case TCREATE: + ret += scnprintf(buf+ret, buflen-ret, + "Tcreate tag %u fid %d name '%.*s' perm ", tag, + fc->params.tcreate.fid, fc->params.tcreate.name.len, + fc->params.tcreate.name.str); + + ret += v9fs_printperm(buf+ret, buflen-ret, fc->params.tcreate.perm); + ret += scnprintf(buf+ret, buflen-ret, " mode %d", + fc->params.tcreate.mode); + break; + + case RCREATE: + ret += scnprintf(buf+ret, buflen-ret, "Rcreate tag %u", tag); + ret += v9fs_printqid(buf+ret, buflen-ret, &fc->params.rcreate.qid); + ret += scnprintf(buf+ret, buflen-ret, " iounit %d", + fc->params.rcreate.iounit); + break; + + case TREAD: + ret += scnprintf(buf+ret, buflen-ret, + "Tread tag %u fid %d offset %lld count %u", tag, + fc->params.tread.fid, + (long long int) fc->params.tread.offset, + fc->params.tread.count); + break; + + case RREAD: + ret += scnprintf(buf+ret, buflen-ret, + "Rread tag %u count %u data ", tag, + fc->params.rread.count); + ret += v9fs_printdata(buf+ret, buflen-ret, fc->params.rread.data, + fc->params.rread.count); + break; + + case TWRITE: + ret += scnprintf(buf+ret, buflen-ret, + "Twrite tag %u fid %d offset %lld count %u data ", + tag, fc->params.twrite.fid, + (long long int) fc->params.twrite.offset, + fc->params.twrite.count); + ret += v9fs_printdata(buf+ret, buflen-ret, fc->params.twrite.data, + fc->params.twrite.count); + break; + + case RWRITE: + ret += scnprintf(buf+ret, buflen-ret, "Rwrite tag %u count %u", + tag, fc->params.rwrite.count); + break; + + case TCLUNK: + ret += scnprintf(buf+ret, buflen-ret, "Tclunk tag %u fid %d", + tag, fc->params.tclunk.fid); + break; + + case RCLUNK: + ret += scnprintf(buf+ret, buflen-ret, "Rclunk tag %u", tag); + break; + + case TREMOVE: + ret += scnprintf(buf+ret, buflen-ret, "Tremove tag %u fid %d", + tag, fc->params.tremove.fid); + break; + + case RREMOVE: + ret += scnprintf(buf+ret, buflen-ret, "Rremove tag %u", tag); + break; + + case TSTAT: + ret += scnprintf(buf+ret, buflen-ret, "Tstat tag %u fid %d", + tag, fc->params.tstat.fid); + break; + + case RSTAT: + ret += scnprintf(buf+ret, buflen-ret, "Rstat tag %u ", tag); + ret += v9fs_printstat(buf+ret, buflen-ret, &fc->params.rstat.stat, + extended); + break; + + case TWSTAT: + ret += scnprintf(buf+ret, buflen-ret, "Twstat tag %u fid %d ", + tag, fc->params.twstat.fid); + ret += v9fs_printstat(buf+ret, buflen-ret, &fc->params.twstat.stat, + extended); + break; + + case RWSTAT: + ret += scnprintf(buf+ret, buflen-ret, "Rwstat tag %u", tag); + break; + + default: + ret += scnprintf(buf+ret, buflen-ret, "unknown type %d", type); + break; + } + + return ret; +} diff --git a/fs/9p/fid.c b/fs/9p/fid.c index c4d13bf904d..b7608af07ce 100644 --- a/fs/9p/fid.c +++ b/fs/9p/fid.c @@ -4,9 +4,8 @@ * Copyright (C) 2005, 2006 by Eric Van Hensbergen <ericvh@gmail.com> * * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. + * it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of diff --git a/fs/9p/fid.h b/fs/9p/fid.h index 1fc2dd08d75..aa974d6875c 100644 --- a/fs/9p/fid.h +++ b/fs/9p/fid.h @@ -4,9 +4,8 @@ * Copyright (C) 2005 by Eric Van Hensbergen <ericvh@gmail.com> * * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. + * it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of diff --git a/fs/9p/mux.c b/fs/9p/mux.c index e2ae60adda9..3e5b124a721 100644 --- a/fs/9p/mux.c +++ b/fs/9p/mux.c @@ -7,9 +7,8 @@ * Copyright (C) 2004-2005 by Latchesar Ionkov <lucho@ionkov.net> * * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. + * it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of @@ -70,7 +69,7 @@ struct v9fs_mux_data { int msize; unsigned char *extended; struct v9fs_transport *trans; - struct v9fs_idpool tidpool; + struct v9fs_idpool tagpool; int err; wait_queue_head_t equeue; struct list_head req_list; @@ -280,8 +279,8 @@ struct v9fs_mux_data *v9fs_mux_init(struct v9fs_transport *trans, int msize, m->msize = msize; m->extended = extended; m->trans = trans; - idr_init(&m->tidpool.pool); - init_MUTEX(&m->tidpool.lock); + idr_init(&m->tagpool.pool); + init_MUTEX(&m->tagpool.lock); m->err = 0; init_waitqueue_head(&m->equeue); INIT_LIST_HEAD(&m->req_list); @@ -635,6 +634,14 @@ static void v9fs_read_work(void *a) goto error; } + if ((v9fs_debug_level&DEBUG_FCALL) == DEBUG_FCALL) { + char buf[150]; + + v9fs_printfcall(buf, sizeof(buf), m->rcall, + *m->extended); + printk(KERN_NOTICE ">>> %p %s\n", m, buf); + } + rcall = m->rcall; rbuf = m->rbuf; if (m->rpos > n) { @@ -740,6 +747,13 @@ static struct v9fs_req *v9fs_send_request(struct v9fs_mux_data *m, v9fs_set_tag(tc, n); + if ((v9fs_debug_level&DEBUG_FCALL) == DEBUG_FCALL) { + char buf[150]; + + v9fs_printfcall(buf, sizeof(buf), tc, *m->extended); + printk(KERN_NOTICE "<<< %p %s\n", m, buf); + } + req->tag = n; req->tcall = tc; req->rcall = NULL; @@ -965,7 +979,7 @@ static u16 v9fs_mux_get_tag(struct v9fs_mux_data *m) { int tag; - tag = v9fs_get_idpool(&m->tidpool); + tag = v9fs_get_idpool(&m->tagpool); if (tag < 0) return V9FS_NOTAG; else @@ -974,6 +988,6 @@ static u16 v9fs_mux_get_tag(struct v9fs_mux_data *m) static void v9fs_mux_put_tag(struct v9fs_mux_data *m, u16 tag) { - if (tag != V9FS_NOTAG && v9fs_check_idpool(tag, &m->tidpool)) - v9fs_put_idpool(tag, &m->tidpool); + if (tag != V9FS_NOTAG && v9fs_check_idpool(tag, &m->tagpool)) + v9fs_put_idpool(tag, &m->tagpool); } diff --git a/fs/9p/mux.h b/fs/9p/mux.h index 17144fdfa11..e90bfd32ea4 100644 --- a/fs/9p/mux.h +++ b/fs/9p/mux.h @@ -7,9 +7,8 @@ * Copyright (C) 2004 by Eric Van Hensbergen <ericvh@gmail.com> * * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. + * it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of diff --git a/fs/9p/trans_fd.c b/fs/9p/trans_fd.c index 5b2ce21b10f..94e0a7fd9fc 100644 --- a/fs/9p/trans_fd.c +++ b/fs/9p/trans_fd.c @@ -1,15 +1,16 @@ /* * linux/fs/9p/trans_fd.c * - * File Descriptor Transport Layer + * Fd transport layer. Includes deprecated socket layer. * - * Copyright (C) 2005 by Latchesar Ionkov <lucho@ionkov.net> - * Copyright (C) 2005 by Eric Van Hensbergen <ericvh@gmail.com> + * Copyright (C) 2006 by Russ Cox <rsc@swtch.com> + * Copyright (C) 2004-2005 by Latchesar Ionkov <lucho@ionkov.net> + * Copyright (C) 2004-2005 by Eric Van Hensbergen <ericvh@gmail.com> + * Copyright (C) 1997-2002 by Ron Minnich <rminnich@sarnoff.com> * * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. + * it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of @@ -25,6 +26,7 @@ */ #include <linux/config.h> +#include <linux/in.h> #include <linux/module.h> #include <linux/net.h> #include <linux/ipv6.h> @@ -40,89 +42,119 @@ #include "v9fs.h" #include "transport.h" +#define V9FS_PORT 564 + struct v9fs_trans_fd { - struct file *in_file; - struct file *out_file; + struct file *rd; + struct file *wr; }; /** - * v9fs_fd_recv - receive from a socket + * v9fs_fd_read- read from a fd * @v9ses: session information * @v: buffer to receive data into * @len: size of receive buffer * */ - -static int v9fs_fd_recv(struct v9fs_transport *trans, void *v, int len) +static int v9fs_fd_read(struct v9fs_transport *trans, void *v, int len) { - struct v9fs_trans_fd *ts = trans ? trans->priv : NULL; + int ret; + struct v9fs_trans_fd *ts; - if (!trans || trans->status != Connected || !ts) - return -EIO; + if (!trans || trans->status == Disconnected || !(ts = trans->priv)) + return -EREMOTEIO; - return kernel_read(ts->in_file, ts->in_file->f_pos, v, len); + if (!(ts->rd->f_flags & O_NONBLOCK)) + dprintk(DEBUG_ERROR, "blocking read ...\n"); + + ret = kernel_read(ts->rd, ts->rd->f_pos, v, len); + if (ret <= 0 && ret != -ERESTARTSYS && ret != -EAGAIN) + trans->status = Disconnected; + return ret; } /** - * v9fs_fd_send - send to a socket + * v9fs_fd_write - write to a socket * @v9ses: session information * @v: buffer to send data from * @len: size of send buffer * */ - -static int v9fs_fd_send(struct v9fs_transport *trans, void *v, int len) +static int v9fs_fd_write(struct v9fs_transport *trans, void *v, int len) { - struct v9fs_trans_fd *ts = trans ? trans->priv : NULL; - mm_segment_t oldfs = get_fs(); - int ret = 0; + int ret; + mm_segment_t oldfs; + struct v9fs_trans_fd *ts; - if (!trans || trans->status != Connected || !ts) - return -EIO; + if (!trans || trans->status == Disconnected || !(ts = trans->priv)) + return -EREMOTEIO; + + if (!(ts->wr->f_flags & O_NONBLOCK)) + dprintk(DEBUG_ERROR, "blocking write ...\n"); oldfs = get_fs(); set_fs(get_ds()); /* The cast to a user pointer is valid due to the set_fs() */ - ret = vfs_write(ts->out_file, (void __user *)v, len, &ts->out_file->f_pos); + ret = vfs_write(ts->wr, (void __user *)v, len, &ts->wr->f_pos); set_fs(oldfs); + if (ret <= 0 && ret != -ERESTARTSYS && ret != -EAGAIN) + trans->status = Disconnected; return ret; } -/** - * v9fs_fd_init - initialize file descriptor transport - * @v9ses: session information - * @addr: address of server to mount - * @data: mount options - * - */ - -static int -v9fs_fd_init(struct v9fs_session_info *v9ses, const char *addr, char *data) +static unsigned int +v9fs_fd_poll(struct v9fs_transport *trans, struct poll_table_struct *pt) { - struct v9fs_trans_fd *ts = NULL; - struct v9fs_transport *trans = v9ses->transport; + int ret, n; + struct v9fs_trans_fd *ts; + mm_segment_t oldfs; - if((v9ses->wfdno == ~0) || (v9ses->rfdno == ~0)) { - printk(KERN_ERR "v9fs: Insufficient options for proto=fd\n"); - return -ENOPROTOOPT; - } + if (!trans || trans->status != Connected || !(ts = trans->priv)) + return -EREMOTEIO; - ts = kmalloc(sizeof(struct v9fs_trans_fd), GFP_KERNEL); + if (!ts->rd->f_op || !ts->rd->f_op->poll) + return -EIO; - if (!ts) - return -ENOMEM; + if (!ts->wr->f_op || !ts->wr->f_op->poll) + return -EIO; - ts->in_file = fget( v9ses->rfdno ); - ts->out_file = fget( v9ses->wfdno ); + oldfs = get_fs(); + set_fs(get_ds()); - if (!ts->in_file || !ts->out_file) { - if (ts->in_file) - fput(ts->in_file); + ret = ts->rd->f_op->poll(ts->rd, pt); + if (ret < 0) + goto end; - if (ts->out_file) - fput(ts->out_file); + if (ts->rd != ts->wr) { + n = ts->wr->f_op->poll(ts->wr, pt); + if (n < 0) { + ret = n; + goto end; + } + ret = (ret & ~POLLOUT) | (n & ~POLLIN); + } + end: + set_fs(oldfs); + return ret; +} + +static int v9fs_fd_open(struct v9fs_session_info *v9ses, int rfd, int wfd) +{ + struct v9fs_transport *trans = v9ses->transport; + struct v9fs_trans_fd *ts = kmalloc(sizeof(struct v9fs_trans_fd), + GFP_KERNEL); + if (!ts) + return -ENOMEM; + + ts->rd = fget(rfd); + ts->wr = fget(wfd); + if (!ts->rd || !ts->wr) { + if (ts->rd) + fput(ts->rd); + if (ts->wr) + fput(ts->wr); kfree(ts); return -EIO; } @@ -133,84 +165,145 @@ v9fs_fd_init(struct v9fs_session_info *v9ses, const char *addr, char *data) return 0; } - -/** - * v9fs_fd_close - shutdown file descriptor - * @trans: private socket structure - * - */ - -static void v9fs_fd_close(struct v9fs_transport *trans) +static int v9fs_fd_init(struct v9fs_session_info *v9ses, const char *addr, + char *data) { - struct v9fs_trans_fd *ts; - - if (!trans) - return; - - ts = xchg(&trans->priv, NULL); + if (v9ses->rfdno == ~0 || v9ses->wfdno == ~0) { + printk(KERN_ERR "v9fs: Insufficient options for proto=fd\n"); + return -ENOPROTOOPT; + } - if (!ts) - return; + return v9fs_fd_open(v9ses, v9ses->rfdno, v9ses->wfdno); +} - trans->status = Disconnected; - if (ts->in_file) - fput(ts->in_file); +static int v9fs_socket_open(struct v9fs_session_info *v9ses, + struct socket *csocket) +{ + int fd, ret; + + csocket->sk->sk_allocation = GFP_NOIO; + if ((fd = sock_map_fd(csocket)) < 0) { + eprintk(KERN_ERR, "v9fs_socket_open: failed to map fd\n"); + ret = fd; + release_csocket: + sock_release(csocket); + return ret; + } - if (ts->out_file) - fput(ts->out_file); + if ((ret = v9fs_fd_open(v9ses, fd, fd)) < 0) { + sockfd_put(csocket); + eprintk(KERN_ERR, "v9fs_socket_open: failed to open fd\n"); + goto release_csocket; + } - kfree(ts); + ((struct v9fs_trans_fd *)v9ses->transport->priv)->rd->f_flags |= + O_NONBLOCK; + return 0; } -static unsigned int -v9fs_fd_poll(struct v9fs_transport *trans, struct poll_table_struct *pt) +static int v9fs_tcp_init(struct v9fs_session_info *v9ses, const char *addr, + char *data) { - int ret, n; - struct v9fs_trans_fd *ts; - mm_segment_t oldfs; + int ret; + struct socket *csocket = NULL; + struct sockaddr_in sin_server; + + sin_server.sin_family = AF_INET; + sin_server.sin_addr.s_addr = in_aton(addr); + sin_server.sin_port = htons(v9ses->port); + sock_create_kern(PF_INET, SOCK_STREAM, IPPROTO_TCP, &csocket); + + if (!csocket) { + eprintk(KERN_ERR, "v9fs_trans_tcp: problem creating socket\n"); + return -1; + } - if (!trans) - return -EIO; + ret = csocket->ops->connect(csocket, + (struct sockaddr *)&sin_server, + sizeof(struct sockaddr_in), 0); + if (ret < 0) { + eprintk(KERN_ERR, + "v9fs_trans_tcp: problem connecting socket to %s\n", + addr); + return ret; + } - ts = trans->priv; - if (trans->status != Connected || !ts) - return -EIO; + return v9fs_socket_open(v9ses, csocket); +} - oldfs = get_fs(); - set_fs(get_ds()); +static int +v9fs_unix_init(struct v9fs_session_info *v9ses, const char *addr, char *data) +{ + int ret; + struct socket *csocket; + struct sockaddr_un sun_server; + + if (strlen(addr) > UNIX_PATH_MAX) { + eprintk(KERN_ERR, "v9fs_trans_unix: address too long: %s\n", + addr); + return -ENAMETOOLONG; + } - if (!ts->in_file->f_op || !ts->in_file->f_op->poll) { - ret = -EIO; - goto end; + sun_server.sun_family = PF_UNIX; + strcpy(sun_server.sun_path, addr); + sock_create_kern(PF_UNIX, SOCK_STREAM, 0, &csocket); + ret = csocket->ops->connect(csocket, (struct sockaddr *)&sun_server, + sizeof(struct sockaddr_un) - 1, 0); + if (ret < 0) { + eprintk(KERN_ERR, + "v9fs_trans_unix: problem connecting socket: %s: %d\n", + addr, ret); + return ret; } - ret = ts->in_file->f_op->poll(ts->in_file, pt); + return v9fs_socket_open(v9ses, csocket); +} - if (ts->out_file != ts->in_file) { - if (!ts->out_file->f_op || !ts->out_file->f_op->poll) { - ret = -EIO; - goto end; - } +/** + * v9fs_sock_close - shutdown socket + * @trans: private socket structure + * + */ +static void v9fs_fd_close(struct v9fs_transport *trans) +{ + struct v9fs_trans_fd *ts; - n = ts->out_file->f_op->poll(ts->out_file, pt); + if (!trans) + return; - ret &= ~POLLOUT; - n &= ~POLLIN; + ts = xchg(&trans->priv, NULL); - ret |= n; - } + if (!ts) + return; -end: - set_fs(oldfs); - return ret; + trans->status = Disconnected; + if (ts->rd) + fput(ts->rd); + if (ts->wr) + fput(ts->wr); + kfree(ts); } - struct v9fs_transport v9fs_trans_fd = { .init = v9fs_fd_init, - .write = v9fs_fd_send, - .read = v9fs_fd_recv, + .write = v9fs_fd_write, + .read = v9fs_fd_read, .close = v9fs_fd_close, .poll = v9fs_fd_poll, }; +struct v9fs_transport v9fs_trans_tcp = { + .init = v9fs_tcp_init, + .write = v9fs_fd_write, + .read = v9fs_fd_read, + .close = v9fs_fd_close, + .poll = v9fs_fd_poll, +}; + +struct v9fs_transport v9fs_trans_unix = { + .init = v9fs_unix_init, + .write = v9fs_fd_write, + .read = v9fs_fd_read, + .close = v9fs_fd_close, + .poll = v9fs_fd_poll, +}; diff --git a/fs/9p/trans_sock.c b/fs/9p/trans_sock.c deleted file mode 100644 index 44e830697ac..00000000000 --- a/fs/9p/trans_sock.c +++ /dev/null @@ -1,334 +0,0 @@ -/* - * linux/fs/9p/trans_socket.c - * - * Socket Transport Layer - * - * Copyright (C) 2004-2005 by Latchesar Ionkov <lucho@ionkov.net> - * Copyright (C) 2004 by Eric Van Hensbergen <ericvh@gmail.com> - * Copyright (C) 1997-2002 by Ron Minnich <rminnich@sarnoff.com> - * Copyright (C) 1995, 1996 by Olaf Kirch <okir@monad.swb.de> - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. - * - * This program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public License - * along with this program; if not, write to: - * Free Software Foundation - * 51 Franklin Street, Fifth Floor - * Boston, MA 02111-1301 USA - * - */ - -#include <linux/config.h> -#include <linux/in.h> -#include <linux/module.h> -#include <linux/net.h> -#include <linux/ipv6.h> -#include <linux/errno.h> -#include <linux/kernel.h> -#include <linux/un.h> -#include <asm/uaccess.h> -#include <linux/inet.h> -#include <linux/idr.h> -#include <linux/file.h> - -#include "debug.h" -#include "v9fs.h" -#include "transport.h" - -#define V9FS_PORT 564 - -struct v9fs_trans_sock { - struct socket *s; - struct file *filp; -}; - -/** - * v9fs_sock_recv - receive from a socket - * @v9ses: session information - * @v: buffer to receive data into - * @len: size of receive buffer - * - */ - -static int v9fs_sock_recv(struct v9fs_transport *trans, void *v, int len) -{ - int ret; - struct v9fs_trans_sock *ts; - - if (!trans || trans->status == Disconnected) { - dprintk(DEBUG_ERROR, "disconnected ...\n"); - return -EREMOTEIO; - } - - ts = trans->priv; - - if (!(ts->filp->f_flags & O_NONBLOCK)) - dprintk(DEBUG_ERROR, "blocking read ...\n"); - - ret = kernel_read(ts->filp, ts->filp->f_pos, v, len); - if (ret <= 0) { - if (ret != -ERESTARTSYS && ret != -EAGAIN) - trans->status = Disconnected; - } - - return ret; -} - -/** - * v9fs_sock_send - send to a socket - * @v9ses: session information - * @v: buffer to send data from - * @len: size of send buffer - * - */ - -static int v9fs_sock_send(struct v9fs_transport *trans, void *v, int len) -{ - int ret; - mm_segment_t oldfs; - struct v9fs_trans_sock *ts; - - if (!trans || trans->status == Disconnected) { - dprintk(DEBUG_ERROR, "disconnected ...\n"); - return -EREMOTEIO; - } - - ts = trans->priv; - if (!ts) { - dprintk(DEBUG_ERROR, "no transport ...\n"); - return -EREMOTEIO; - } - - if (!(ts->filp->f_flags & O_NONBLOCK)) - dprintk(DEBUG_ERROR, "blocking write ...\n"); - - oldfs = get_fs(); - set_fs(get_ds()); - ret = vfs_write(ts->filp, (void __user *)v, len, &ts->filp->f_pos); - set_fs(oldfs); - - if (ret < 0) { - if (ret != -ERESTARTSYS) - trans->status = Disconnected; - } - - return ret; -} - -static unsigned int v9fs_sock_poll(struct v9fs_transport *trans, - struct poll_table_struct *pt) { - - int ret; - struct v9fs_trans_sock *ts; - mm_segment_t oldfs; - - if (!trans) { - dprintk(DEBUG_ERROR, "no transport\n"); - return -EIO; - } - - ts = trans->priv; - if (trans->status != Connected || !ts) { - dprintk(DEBUG_ERROR, "transport disconnected: %d\n", trans->status); - return -EIO; - } - - oldfs = get_fs(); - set_fs(get_ds()); - - if (!ts->filp->f_op || !ts->filp->f_op->poll) { - dprintk(DEBUG_ERROR, "no poll operation\n"); - ret = -EIO; - goto end; - } - - ret = ts->filp->f_op->poll(ts->filp, pt); - -end: - set_fs(oldfs); - return ret; -} - - -/** - * v9fs_tcp_init - initialize TCP socket - * @v9ses: session information - * @addr: address of server to mount - * @data: mount options - * - */ - -static int -v9fs_tcp_init(struct v9fs_session_info *v9ses, const char *addr, char *data) -{ - struct socket *csocket = NULL; - struct sockaddr_in sin_server; - int rc = 0; - struct v9fs_trans_sock *ts = NULL; - struct v9fs_transport *trans = v9ses->transport; - int fd; - - trans->status = Disconnected; - - ts = kmalloc(sizeof(struct v9fs_trans_sock), GFP_KERNEL); - - if (!ts) - return -ENOMEM; - - trans->priv = ts; - ts->s = NULL; - ts->filp = NULL; - - if (!addr) - return -EINVAL; - - dprintk(DEBUG_TRANS, "Connecting to %s\n", addr); - - sin_server.sin_family = AF_INET; - sin_server.sin_addr.s_addr = in_aton(addr); - sin_server.sin_port = htons(v9ses->port); - sock_create_kern(PF_INET, SOCK_STREAM, IPPROTO_TCP, &csocket); - rc = csocket->ops->connect(csocket, - (struct sockaddr *)&sin_server, - sizeof(struct sockaddr_in), 0); - if (rc < 0) { - eprintk(KERN_ERR, - "v9fs_trans_tcp: problem connecting socket to %s\n", - addr); - return rc; - } - csocket->sk->sk_allocation = GFP_NOIO; - - fd = sock_map_fd(csocket); - if (fd < 0) { - sock_release(csocket); - kfree(ts); - trans->priv = NULL; - return fd; - } - - ts->s = csocket; - ts->filp = fget(fd); - ts->filp->f_flags |= O_NONBLOCK; - trans->status = Connected; - - return 0; -} - -/** - * v9fs_unix_init - initialize UNIX domain socket - * @v9ses: session information - * @dev_name: path to named pipe - * @data: mount options - * - */ - -static int -v9fs_unix_init(struct v9fs_session_info *v9ses, const char *dev_name, - char *data) -{ - int rc, fd; - struct socket *csocket; - struct sockaddr_un sun_server; - struct v9fs_transport *trans; - struct v9fs_trans_sock *ts; - - rc = 0; - csocket = NULL; - trans = v9ses->transport; - - trans->status = Disconnected; - - if (strlen(dev_name) > UNIX_PATH_MAX) { - eprintk(KERN_ERR, "v9fs_trans_unix: address too long: %s\n", - dev_name); - return -ENOMEM; - } - - ts = kmalloc(sizeof(struct v9fs_trans_sock), GFP_KERNEL); - if (!ts) - return -ENOMEM; - - trans->priv = ts; - ts->s = NULL; - ts->filp = NULL; - - sun_server.sun_family = PF_UNIX; - strcpy(sun_server.sun_path, dev_name); - sock_create_kern(PF_UNIX, SOCK_STREAM, 0, &csocket); - rc = csocket->ops->connect(csocket, (struct sockaddr *)&sun_server, - sizeof(struct sockaddr_un) - 1, 0); /* -1 *is* important */ - if (rc < 0) { - eprintk(KERN_ERR, - "v9fs_trans_unix: problem connecting socket: %s: %d\n", - dev_name, rc); - return rc; - } - csocket->sk->sk_allocation = GFP_NOIO; - - fd = sock_map_fd(csocket); - if (fd < 0) { - sock_release(csocket); - kfree(ts); - trans->priv = NULL; - return fd; - } - - ts->s = csocket; - ts->filp = fget(fd); - ts->filp->f_flags |= O_NONBLOCK; - trans->status = Connected; - - return 0; -} - -/** - * v9fs_sock_close - shutdown socket - * @trans: private socket structure - * - */ - -static void v9fs_sock_close(struct v9fs_transport *trans) -{ - struct v9fs_trans_sock *ts; - - if (!trans) - return; - - ts = trans->priv; - - if ((ts) && (ts->filp)) { - fput(ts->filp); - ts->filp = NULL; - ts->s = NULL; - trans->status = Disconnected; - } - - kfree(ts); - - trans->priv = NULL; -} - -struct v9fs_transport v9fs_trans_tcp = { - .init = v9fs_tcp_init, - .write = v9fs_sock_send, - .read = v9fs_sock_recv, - .close = v9fs_sock_close, - .poll = v9fs_sock_poll, -}; - -struct v9fs_transport v9fs_trans_unix = { - .init = v9fs_unix_init, - .write = v9fs_sock_send, - .read = v9fs_sock_recv, - .close = v9fs_sock_close, - .poll = v9fs_sock_poll, -}; diff --git a/fs/9p/transport.h b/fs/9p/transport.h index 91fcdb94b36..b38a4b8a41c 100644 --- a/fs/9p/transport.h +++ b/fs/9p/transport.h @@ -7,9 +7,8 @@ * Copyright (C) 2004 by Eric Van Hensbergen <ericvh@gmail.com> * * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. + * it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of diff --git a/fs/9p/v9fs.c b/fs/9p/v9fs.c index 61352491ba3..d37416eb579 100644 --- a/fs/9p/v9fs.c +++ b/fs/9p/v9fs.c @@ -7,9 +7,8 @@ * Copyright (C) 2002 by Ron Minnich <rminnich@lanl.gov> * * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. + * it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of @@ -51,7 +50,7 @@ enum { Opt_port, Opt_msize, Opt_uid, Opt_gid, Opt_afid, Opt_debug, Opt_rfdno, Opt_wfdno, /* String options */ - Opt_name, Opt_remotename, + Opt_uname, Opt_remotename, /* Options that take no arguments */ Opt_legacy, Opt_nodevmap, Opt_unix, Opt_tcp, Opt_fd, /* Error token */ @@ -67,7 +66,7 @@ static match_table_t tokens = { {Opt_rfdno, "rfdno=%u"}, {Opt_wfdno, "wfdno=%u"}, {Opt_debug, "debug=%x"}, - {Opt_name, "name=%s"}, + {Opt_uname, "uname=%s"}, {Opt_remotename, "aname=%s"}, {Opt_unix, "proto=unix"}, {Opt_tcp, "proto=tcp"}, @@ -116,7 +115,7 @@ static void v9fs_parse_options(char *options, struct v9fs_session_info *v9ses) if (!*p) continue; token = match_token(p, tokens, args); - if (token < Opt_name) { + if (token < Opt_uname) { if ((ret = match_int(&args[0], &option)) < 0) { dprintk(DEBUG_ERROR, "integer field, but no integer?\n"); @@ -158,7 +157,7 @@ static void v9fs_parse_options(char *options, struct v9fs_session_info *v9ses) case Opt_fd: v9ses->proto = PROTO_FD; break; - case Opt_name: + case Opt_uname: match_strcpy(v9ses->name, &args[0]); break; case Opt_remotename: @@ -289,7 +288,7 @@ v9fs_session_init(struct v9fs_session_info *v9ses, /* set global debug level */ v9fs_debug_level = v9ses->debug; - /* id pools that are session-dependent: FIDs and TIDs */ + /* id pools that are session-dependent: fids and tags */ idr_init(&v9ses->fidpool.pool); init_MUTEX(&v9ses->fidpool.lock); diff --git a/fs/9p/v9fs.h b/fs/9p/v9fs.h index f337da7a0ee..c134d104cb2 100644 --- a/fs/9p/v9fs.h +++ b/fs/9p/v9fs.h @@ -5,9 +5,8 @@ * Copyright (C) 2002 by Ron Minnich <rminnich@lanl.gov> * * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. + * it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of @@ -91,6 +90,3 @@ void v9fs_session_cancel(struct v9fs_session_info *v9ses); #define V9FS_DEFUSER "nobody" #define V9FS_DEFANAME "" -/* inital pool sizes for fids and tags */ -#define V9FS_START_FIDS 8192 -#define V9FS_START_TIDS 256 diff --git a/fs/9p/v9fs_vfs.h b/fs/9p/v9fs_vfs.h index a759278acaa..f867b8d3e97 100644 --- a/fs/9p/v9fs_vfs.h +++ b/fs/9p/v9fs_vfs.h @@ -5,9 +5,8 @@ * Copyright (C) 2002 by Ron Minnich <rminnich@lanl.gov> * * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. + * it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of @@ -40,8 +39,8 @@ extern struct file_system_type v9fs_fs_type; extern struct address_space_operations v9fs_addr_operations; -extern struct file_operations v9fs_file_operations; -extern struct file_operations v9fs_dir_operations; +extern const struct file_operations v9fs_file_operations; +extern const struct file_operations v9fs_dir_operations; extern struct dentry_operations v9fs_dentry_operations; struct inode *v9fs_get_inode(struct super_block *sb, int mode); diff --git a/fs/9p/vfs_addr.c b/fs/9p/vfs_addr.c index 8100fb5171b..efda46fb64d 100644 --- a/fs/9p/vfs_addr.c +++ b/fs/9p/vfs_addr.c @@ -7,9 +7,8 @@ * Copyright (C) 2002 by Ron Minnich <rminnich@lanl.gov> * * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. + * it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of diff --git a/fs/9p/vfs_dentry.c b/fs/9p/vfs_dentry.c index 12c9cc926b7..062daa6000a 100644 --- a/fs/9p/vfs_dentry.c +++ b/fs/9p/vfs_dentry.c @@ -7,9 +7,8 @@ * Copyright (C) 2002 by Ron Minnich <rminnich@lanl.gov> * * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. + * it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of @@ -51,7 +50,7 @@ * */ -int v9fs_dentry_delete(struct dentry *dentry) +static int v9fs_dentry_delete(struct dentry *dentry) { dprintk(DEBUG_VFS, " dentry: %s (%p)\n", dentry->d_iname, dentry); return 1; diff --git a/fs/9p/vfs_dir.c b/fs/9p/vfs_dir.c index cd5eeb032d6..e32d5971039 100644 --- a/fs/9p/vfs_dir.c +++ b/fs/9p/vfs_dir.c @@ -7,9 +7,8 @@ * Copyright (C) 2002 by Ron Minnich <rminnich@lanl.gov> * * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. + * it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of @@ -205,7 +204,7 @@ int v9fs_dir_release(struct inode *inode, struct file *filp) return 0; } -struct file_operations v9fs_dir_operations = { +const struct file_operations v9fs_dir_operations = { .read = generic_read_dir, .readdir = v9fs_dir_readdir, .open = v9fs_file_open, diff --git a/fs/9p/vfs_file.c b/fs/9p/vfs_file.c index de3a129698d..083dcfcd158 100644 --- a/fs/9p/vfs_file.c +++ b/fs/9p/vfs_file.c @@ -7,9 +7,8 @@ * Copyright (C) 2002 by Ron Minnich <rminnich@lanl.gov> * * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. + * it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of @@ -69,29 +68,30 @@ int v9fs_file_open(struct inode *inode, struct file *file) fid = v9fs_get_idpool(&v9ses->fidpool); if (fid < 0) { - eprintk(KERN_WARNING, "newfid fails!\n"); - return -ENOSPC; - } + eprintk(KERN_WARNING, "newfid fails!\n"); + return -ENOSPC; + } err = v9fs_t_walk(v9ses, vfid->fid, fid, NULL, NULL); if (err < 0) { - dprintk(DEBUG_ERROR, "rewalk didn't work\n"); + dprintk(DEBUG_ERROR, "rewalk didn't work\n"); goto put_fid; } - vfid = kmalloc(sizeof(struct v9fs_fid), GFP_KERNEL); - if (vfid == NULL) { - dprintk(DEBUG_ERROR, "out of memory\n"); - goto clunk_fid; - } - - /* TODO: do special things for O_EXCL, O_NOFOLLOW, O_SYNC */ - /* translate open mode appropriately */ + /* TODO: do special things for O_EXCL, O_NOFOLLOW, O_SYNC */ + /* translate open mode appropriately */ omode = v9fs_uflags2omode(file->f_flags); err = v9fs_t_open(v9ses, fid, omode, &fcall); if (err < 0) { PRINT_FCALL_ERROR("open failed", fcall); - goto destroy_vfid; + goto clunk_fid; + } + + vfid = kmalloc(sizeof(struct v9fs_fid), GFP_KERNEL); + if (vfid == NULL) { + dprintk(DEBUG_ERROR, "out of memory\n"); + err = -ENOMEM; + goto clunk_fid; } file->private_data = vfid; @@ -106,15 +106,12 @@ int v9fs_file_open(struct inode *inode, struct file *file) return 0; -destroy_vfid: - v9fs_fid_destroy(vfid); - clunk_fid: v9fs_t_clunk(v9ses, fid); put_fid: v9fs_put_idpool(fid, &v9ses->fidpool); - kfree(fcall); + kfree(fcall); return err; } @@ -269,7 +266,7 @@ v9fs_file_write(struct file *filp, const char __user * data, return total; } -struct file_operations v9fs_file_operations = { +const struct file_operations v9fs_file_operations = { .llseek = generic_file_llseek, .read = v9fs_file_read, .write = v9fs_file_write, diff --git a/fs/9p/vfs_inode.c b/fs/9p/vfs_inode.c index 651a9e14d9a..133db366d30 100644 --- a/fs/9p/vfs_inode.c +++ b/fs/9p/vfs_inode.c @@ -7,9 +7,8 @@ * Copyright (C) 2002 by Ron Minnich <rminnich@lanl.gov> * * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. + * it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of @@ -255,8 +254,8 @@ struct inode *v9fs_get_inode(struct super_block *sb, int mode) } static int -v9fs_create(struct v9fs_session_info *v9ses, u32 pfid, char *name, - u32 perm, u8 mode, u32 *fidp, struct v9fs_qid *qid, u32 *iounit) +v9fs_create(struct v9fs_session_info *v9ses, u32 pfid, char *name, u32 perm, + u8 mode, char *extension, u32 *fidp, struct v9fs_qid *qid, u32 *iounit) { u32 fid; int err; @@ -271,14 +270,14 @@ v9fs_create(struct v9fs_session_info *v9ses, u32 pfid, char *name, err = v9fs_t_walk(v9ses, pfid, fid, NULL, &fcall); if (err < 0) { PRINT_FCALL_ERROR("clone error", fcall); - goto error; + goto put_fid; } kfree(fcall); - err = v9fs_t_create(v9ses, fid, name, perm, mode, &fcall); + err = v9fs_t_create(v9ses, fid, name, perm, mode, extension, &fcall); if (err < 0) { PRINT_FCALL_ERROR("create fails", fcall); - goto error; + goto clunk_fid; } if (iounit) @@ -293,7 +292,11 @@ v9fs_create(struct v9fs_session_info *v9ses, u32 pfid, char *name, kfree(fcall); return 0; -error: +clunk_fid: + v9fs_t_clunk(v9ses, fid); + fid = V9FS_NOFID; + +put_fid: if (fid >= 0) v9fs_put_idpool(fid, &v9ses->fidpool); @@ -348,7 +351,7 @@ error: return ERR_PTR(err); } -struct inode * +static struct inode * v9fs_inode_from_fid(struct v9fs_session_info *v9ses, u32 fid, struct super_block *sb) { @@ -474,7 +477,7 @@ v9fs_vfs_create(struct inode *dir, struct dentry *dentry, int mode, flags = O_RDWR; err = v9fs_create(v9ses, dfid->fid, (char *) dentry->d_name.name, - perm, v9fs_uflags2omode(flags), &fid, &qid, &iounit); + perm, v9fs_uflags2omode(flags), NULL, &fid, &qid, &iounit); if (err) goto error; @@ -550,7 +553,7 @@ static int v9fs_vfs_mkdir(struct inode *dir, struct dentry *dentry, int mode) perm = unixmode2p9mode(v9ses, mode | S_IFDIR); err = v9fs_create(v9ses, dfid->fid, (char *) dentry->d_name.name, - perm, V9FS_OREAD, &fid, NULL, NULL); + perm, V9FS_OREAD, NULL, &fid, NULL, NULL); if (err) { dprintk(DEBUG_ERROR, "create error %d\n", err); @@ -1008,11 +1011,13 @@ static int v9fs_readlink(struct dentry *dentry, char *buffer, int buflen) /* copy extension buffer into buffer */ if (fcall->params.rstat.stat.extension.len < buflen) - buflen = fcall->params.rstat.stat.extension.len; + buflen = fcall->params.rstat.stat.extension.len + 1; - memcpy(buffer, fcall->params.rstat.stat.extension.str, buflen - 1); + memmove(buffer, fcall->params.rstat.stat.extension.str, buflen - 1); buffer[buflen-1] = 0; + dprintk(DEBUG_ERROR, "%s -> %.*s (%s)\n", dentry->d_name.name, fcall->params.rstat.stat.extension.len, + fcall->params.rstat.stat.extension.str, buffer); retval = buflen; FreeFcall: @@ -1072,7 +1077,7 @@ static void *v9fs_vfs_follow_link(struct dentry *dentry, struct nameidata *nd) if (!link) link = ERR_PTR(-ENOMEM); else { - len = v9fs_readlink(dentry, link, strlen(link)); + len = v9fs_readlink(dentry, link, PATH_MAX); if (len < 0) { __putname(link); @@ -1109,10 +1114,7 @@ static int v9fs_vfs_mkspecial(struct inode *dir, struct dentry *dentry, struct v9fs_session_info *v9ses; struct v9fs_fid *dfid, *vfid; struct inode *inode; - struct v9fs_fcall *fcall; - struct v9fs_wstat wstat; - fcall = NULL; inode = NULL; vfid = NULL; v9ses = v9fs_inode2v9ses(dir); @@ -1125,7 +1127,7 @@ static int v9fs_vfs_mkspecial(struct inode *dir, struct dentry *dentry, } err = v9fs_create(v9ses, dfid->fid, (char *) dentry->d_name.name, - perm, V9FS_OREAD, &fid, NULL, NULL); + perm, V9FS_OREAD, (char *) extension, &fid, NULL, NULL); if (err) goto error; @@ -1148,23 +1150,11 @@ static int v9fs_vfs_mkspecial(struct inode *dir, struct dentry *dentry, goto error; } - /* issue a Twstat */ - v9fs_blank_wstat(&wstat); - wstat.muid = v9ses->name; - wstat.extension = (char *) extension; - err = v9fs_t_wstat(v9ses, vfid->fid, &wstat, &fcall); - if (err < 0) { - PRINT_FCALL_ERROR("wstat error", fcall); - goto error; - } - - kfree(fcall); dentry->d_op = &v9fs_dentry_operations; d_instantiate(dentry, inode); return 0; error: - kfree(fcall); if (vfid) v9fs_fid_destroy(vfid); @@ -1224,7 +1214,7 @@ v9fs_vfs_link(struct dentry *old_dentry, struct inode *dir, } name = __getname(); - sprintf(name, "hardlink(%d)\n", oldfid->fid); + sprintf(name, "%d\n", oldfid->fid); retval = v9fs_vfs_mkspecial(dir, dentry, V9FS_DMLINK, name); __putname(name); @@ -1253,6 +1243,8 @@ v9fs_vfs_mknod(struct inode *dir, struct dentry *dentry, int mode, dev_t rdev) return -EINVAL; name = __getname(); + if (!name) + return -ENOMEM; /* build extension */ if (S_ISBLK(mode)) sprintf(name, "b %u %u", MAJOR(rdev), MINOR(rdev)); diff --git a/fs/9p/vfs_super.c b/fs/9p/vfs_super.c index d05318fa684..b0a0ae509c0 100644 --- a/fs/9p/vfs_super.c +++ b/fs/9p/vfs_super.c @@ -8,9 +8,8 @@ * Copyright (C) 2002 by Ron Minnich <rminnich@lanl.gov> * * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. + * it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of @@ -262,7 +261,7 @@ static struct super_operations v9fs_super_ops = { }; struct file_system_type v9fs_fs_type = { - .name = "9P", + .name = "9p", .get_sb = v9fs_get_sb, .kill_sb = v9fs_kill_super, .owner = THIS_MODULE, diff --git a/fs/Kconfig b/fs/Kconfig index c8d0a209120..e207be68d4c 100644 --- a/fs/Kconfig +++ b/fs/Kconfig @@ -1555,6 +1555,7 @@ config RPCSEC_GSS_SPKM3 select CRYPTO select CRYPTO_MD5 select CRYPTO_DES + select CRYPTO_CAST5 help Provides for secure RPC calls by means of a gss-api mechanism based on the SPKM3 public-key mechanism. diff --git a/fs/adfs/adfs.h b/fs/adfs/adfs.h index f6cd01352cc..29217ff36d4 100644 --- a/fs/adfs/adfs.h +++ b/fs/adfs/adfs.h @@ -85,7 +85,7 @@ void __adfs_error(struct super_block *sb, const char *function, /* dir_*.c */ extern struct inode_operations adfs_dir_inode_operations; -extern struct file_operations adfs_dir_operations; +extern const struct file_operations adfs_dir_operations; extern struct dentry_operations adfs_dentry_operations; extern struct adfs_dir_ops adfs_f_dir_ops; extern struct adfs_dir_ops adfs_fplus_dir_ops; @@ -94,7 +94,7 @@ extern int adfs_dir_update(struct super_block *sb, struct object_info *obj); /* file.c */ extern struct inode_operations adfs_file_inode_operations; -extern struct file_operations adfs_file_operations; +extern const struct file_operations adfs_file_operations; static inline __u32 signed_asl(__u32 val, signed int shift) { diff --git a/fs/adfs/dir.c b/fs/adfs/dir.c index 0b4c3a02807..7b075fc397d 100644 --- a/fs/adfs/dir.c +++ b/fs/adfs/dir.c @@ -196,7 +196,7 @@ out: return ret; } -struct file_operations adfs_dir_operations = { +const struct file_operations adfs_dir_operations = { .read = generic_read_dir, .readdir = adfs_readdir, .fsync = file_fsync, diff --git a/fs/adfs/file.c b/fs/adfs/file.c index 6af10885f9d..1014b9f2117 100644 --- a/fs/adfs/file.c +++ b/fs/adfs/file.c @@ -25,7 +25,7 @@ #include "adfs.h" -struct file_operations adfs_file_operations = { +const struct file_operations adfs_file_operations = { .llseek = generic_file_llseek, .read = generic_file_read, .mmap = generic_file_mmap, diff --git a/fs/affs/affs.h b/fs/affs/affs.h index 0c6799f2137..a43a876742b 100644 --- a/fs/affs/affs.h +++ b/fs/affs/affs.h @@ -192,9 +192,9 @@ extern void affs_dir_truncate(struct inode *); extern struct inode_operations affs_file_inode_operations; extern struct inode_operations affs_dir_inode_operations; extern struct inode_operations affs_symlink_inode_operations; -extern struct file_operations affs_file_operations; -extern struct file_operations affs_file_operations_ofs; -extern struct file_operations affs_dir_operations; +extern const struct file_operations affs_file_operations; +extern const struct file_operations affs_file_operations_ofs; +extern const struct file_operations affs_dir_operations; extern struct address_space_operations affs_symlink_aops; extern struct address_space_operations affs_aops; extern struct address_space_operations affs_aops_ofs; diff --git a/fs/affs/dir.c b/fs/affs/dir.c index 548efd0ee98..5d9649fa181 100644 --- a/fs/affs/dir.c +++ b/fs/affs/dir.c @@ -17,7 +17,7 @@ static int affs_readdir(struct file *, void *, filldir_t); -struct file_operations affs_dir_operations = { +const struct file_operations affs_dir_operations = { .read = generic_read_dir, .readdir = affs_readdir, .fsync = file_fsync, diff --git a/fs/affs/file.c b/fs/affs/file.c index f72fb776ecd..7076262af39 100644 --- a/fs/affs/file.c +++ b/fs/affs/file.c @@ -25,7 +25,7 @@ static struct buffer_head *affs_get_extblock_slow(struct inode *inode, u32 ext); static int affs_file_open(struct inode *inode, struct file *filp); static int affs_file_release(struct inode *inode, struct file *filp); -struct file_operations affs_file_operations = { +const struct file_operations affs_file_operations = { .llseek = generic_file_llseek, .read = generic_file_read, .write = generic_file_write, diff --git a/fs/afs/dir.c b/fs/afs/dir.c index 5c61c24dab2..a6dff6a4f20 100644 --- a/fs/afs/dir.c +++ b/fs/afs/dir.c @@ -32,7 +32,7 @@ static int afs_d_delete(struct dentry *dentry); static int afs_dir_lookup_filldir(void *_cookie, const char *name, int nlen, loff_t fpos, ino_t ino, unsigned dtype); -struct file_operations afs_dir_file_operations = { +const struct file_operations afs_dir_file_operations = { .open = afs_dir_open, .readdir = afs_dir_readdir, }; diff --git a/fs/afs/file.c b/fs/afs/file.c index 150b1922792..7bb716887e2 100644 --- a/fs/afs/file.c +++ b/fs/afs/file.c @@ -28,7 +28,7 @@ static int afs_file_release(struct inode *inode, struct file *file); #endif static int afs_file_readpage(struct file *file, struct page *page); -static int afs_file_invalidatepage(struct page *page, unsigned long offset); +static void afs_file_invalidatepage(struct page *page, unsigned long offset); static int afs_file_releasepage(struct page *page, gfp_t gfp_flags); struct inode_operations afs_file_inode_operations = { @@ -212,7 +212,7 @@ int afs_cache_get_page_cookie(struct page *page, /* * invalidate part or all of a page */ -static int afs_file_invalidatepage(struct page *page, unsigned long offset) +static void afs_file_invalidatepage(struct page *page, unsigned long offset) { int ret = 1; @@ -238,11 +238,11 @@ static int afs_file_invalidatepage(struct page *page, unsigned long offset) if (!PageWriteback(page)) ret = page->mapping->a_ops->releasepage(page, 0); + /* possibly should BUG_ON(!ret); - neilb */ } } _leave(" = %d", ret); - return ret; } /* end afs_file_invalidatepage() */ /*****************************************************************************/ diff --git a/fs/afs/internal.h b/fs/afs/internal.h index ab8f87c6631..72febdf9a35 100644 --- a/fs/afs/internal.h +++ b/fs/afs/internal.h @@ -64,7 +64,7 @@ extern struct cachefs_index_def afs_cache_cell_index_def; * dir.c */ extern struct inode_operations afs_dir_inode_operations; -extern struct file_operations afs_dir_file_operations; +extern const struct file_operations afs_dir_file_operations; /* * file.c @@ -105,7 +105,7 @@ extern struct cachefs_netfs afs_cache_netfs; * mntpt.c */ extern struct inode_operations afs_mntpt_inode_operations; -extern struct file_operations afs_mntpt_file_operations; +extern const struct file_operations afs_mntpt_file_operations; extern struct afs_timer afs_mntpt_expiry_timer; extern struct afs_timer_ops afs_mntpt_expiry_timer_ops; extern unsigned long afs_mntpt_expiry_timeout; diff --git a/fs/afs/mntpt.c b/fs/afs/mntpt.c index 31ee06590de..4e6eeb59b83 100644 --- a/fs/afs/mntpt.c +++ b/fs/afs/mntpt.c @@ -32,7 +32,7 @@ static struct dentry *afs_mntpt_lookup(struct inode *dir, static int afs_mntpt_open(struct inode *inode, struct file *file); static void *afs_mntpt_follow_link(struct dentry *dentry, struct nameidata *nd); -struct file_operations afs_mntpt_file_operations = { +const struct file_operations afs_mntpt_file_operations = { .open = afs_mntpt_open, }; diff --git a/fs/afs/proc.c b/fs/afs/proc.c index 9c81b8f7eef..101d21b6c03 100644 --- a/fs/afs/proc.c +++ b/fs/afs/proc.c @@ -37,7 +37,7 @@ static struct seq_operations afs_proc_cells_ops = { .show = afs_proc_cells_show, }; -static struct file_operations afs_proc_cells_fops = { +static const struct file_operations afs_proc_cells_fops = { .open = afs_proc_cells_open, .read = seq_read, .write = afs_proc_cells_write, @@ -53,7 +53,7 @@ static ssize_t afs_proc_rootcell_write(struct file *file, const char __user *buf, size_t size, loff_t *_pos); -static struct file_operations afs_proc_rootcell_fops = { +static const struct file_operations afs_proc_rootcell_fops = { .open = afs_proc_rootcell_open, .read = afs_proc_rootcell_read, .write = afs_proc_rootcell_write, @@ -77,7 +77,7 @@ static struct seq_operations afs_proc_cell_volumes_ops = { .show = afs_proc_cell_volumes_show, }; -static struct file_operations afs_proc_cell_volumes_fops = { +static const struct file_operations afs_proc_cell_volumes_fops = { .open = afs_proc_cell_volumes_open, .read = seq_read, .llseek = seq_lseek, @@ -101,7 +101,7 @@ static struct seq_operations afs_proc_cell_vlservers_ops = { .show = afs_proc_cell_vlservers_show, }; -static struct file_operations afs_proc_cell_vlservers_fops = { +static const struct file_operations afs_proc_cell_vlservers_fops = { .open = afs_proc_cell_vlservers_open, .read = seq_read, .llseek = seq_lseek, @@ -124,7 +124,7 @@ static struct seq_operations afs_proc_cell_servers_ops = { .show = afs_proc_cell_servers_show, }; -static struct file_operations afs_proc_cell_servers_fops = { +static const struct file_operations afs_proc_cell_servers_fops = { .open = afs_proc_cell_servers_open, .read = seq_read, .llseek = seq_lseek, @@ -122,10 +122,9 @@ static int aio_setup_ring(struct kioctx *ctx) info->nr = 0; info->ring_pages = info->internal_pages; if (nr_pages > AIO_RING_PAGES) { - info->ring_pages = kmalloc(sizeof(struct page *) * nr_pages, GFP_KERNEL); + info->ring_pages = kcalloc(nr_pages, sizeof(struct page *), GFP_KERNEL); if (!info->ring_pages) return -ENOMEM; - memset(info->ring_pages, 0, sizeof(struct page *) * nr_pages); } info->mmap_size = nr_pages * PAGE_SIZE; diff --git a/fs/autofs/autofs_i.h b/fs/autofs/autofs_i.h index 990c28da5ae..a62327f1bdf 100644 --- a/fs/autofs/autofs_i.h +++ b/fs/autofs/autofs_i.h @@ -146,7 +146,7 @@ struct autofs_dir_ent *autofs_expire(struct super_block *,struct autofs_sb_info extern struct inode_operations autofs_root_inode_operations; extern struct inode_operations autofs_symlink_inode_operations; -extern struct file_operations autofs_root_operations; +extern const struct file_operations autofs_root_operations; /* Initializing function */ diff --git a/fs/autofs/dirhash.c b/fs/autofs/dirhash.c index 5ccfcf26310..3fded389d06 100644 --- a/fs/autofs/dirhash.c +++ b/fs/autofs/dirhash.c @@ -92,7 +92,7 @@ struct autofs_dir_ent *autofs_expire(struct super_block *sb, ; dput(dentry); - if ( may_umount(mnt) == 0 ) { + if ( may_umount(mnt) ) { mntput(mnt); DPRINTK(("autofs: signaling expire on %s\n", ent->name)); return ent; /* Expirable! */ diff --git a/fs/autofs/root.c b/fs/autofs/root.c index 870e2cf3301..9cac08d6a87 100644 --- a/fs/autofs/root.c +++ b/fs/autofs/root.c @@ -26,7 +26,7 @@ static int autofs_root_rmdir(struct inode *,struct dentry *); static int autofs_root_mkdir(struct inode *,struct dentry *,int); static int autofs_root_ioctl(struct inode *, struct file *,unsigned int,unsigned long); -struct file_operations autofs_root_operations = { +const struct file_operations autofs_root_operations = { .read = generic_read_dir, .readdir = autofs_root_readdir, .ioctl = autofs_root_ioctl, diff --git a/fs/autofs4/autofs_i.h b/fs/autofs4/autofs_i.h index f54c5b21f87..57c4903614e 100644 --- a/fs/autofs4/autofs_i.h +++ b/fs/autofs4/autofs_i.h @@ -3,6 +3,7 @@ * linux/fs/autofs/autofs_i.h * * Copyright 1997-1998 Transmeta Corporation - All Rights Reserved + * Copyright 2005-2006 Ian Kent <raven@themaw.net> * * This file is part of the Linux kernel and is made available under * the terms of the GNU General Public License, version 2, or at your @@ -41,14 +42,6 @@ #define AUTOFS_SUPER_MAGIC 0x0187 -/* - * If the daemon returns a negative response (AUTOFS_IOC_FAIL) then the - * kernel will keep the negative response cached for up to the time given - * here, although the time can be shorter if the kernel throws the dcache - * entry away. This probably should be settable from user space. - */ -#define AUTOFS_NEGATIVE_TIMEOUT (60*HZ) /* 1 minute */ - /* Unified info structure. This is pointed to by both the dentry and inode structures. Each file in the filesystem has an instance of this structure. It holds a reference to the dentry, so dentries are never @@ -63,6 +56,7 @@ struct autofs_info { struct autofs_sb_info *sbi; unsigned long last_used; + atomic_t count; mode_t mode; size_t size; @@ -83,23 +77,37 @@ struct autofs_wait_queue { int hash; int len; char *name; + u32 dev; + u64 ino; + uid_t uid; + gid_t gid; + pid_t pid; + pid_t tgid; /* This is for status reporting upon return */ int status; - atomic_t notified; + atomic_t notify; atomic_t wait_ctr; }; #define AUTOFS_SBI_MAGIC 0x6d4a556d +#define AUTOFS_TYPE_INDIRECT 0x0001 +#define AUTOFS_TYPE_DIRECT 0x0002 +#define AUTOFS_TYPE_OFFSET 0x0004 + struct autofs_sb_info { u32 magic; struct dentry *root; + int pipefd; struct file *pipe; pid_t oz_pgrp; int catatonic; int version; int sub_version; + int min_proto; + int max_proto; unsigned long exp_timeout; + unsigned int type; int reghost_enabled; int needs_reghost; struct super_block *sb; @@ -166,8 +174,10 @@ int autofs4_expire_multi(struct super_block *, struct vfsmount *, extern struct inode_operations autofs4_symlink_inode_operations; extern struct inode_operations autofs4_dir_inode_operations; extern struct inode_operations autofs4_root_inode_operations; -extern struct file_operations autofs4_dir_operations; -extern struct file_operations autofs4_root_operations; +extern struct inode_operations autofs4_indirect_root_inode_operations; +extern struct inode_operations autofs4_direct_root_inode_operations; +extern const struct file_operations autofs4_dir_operations; +extern const struct file_operations autofs4_root_operations; /* Initializing function */ @@ -176,13 +186,6 @@ struct autofs_info *autofs4_init_ino(struct autofs_info *, struct autofs_sb_info /* Queue management functions */ -enum autofs_notify -{ - NFY_NONE, - NFY_MOUNT, - NFY_EXPIRE -}; - int autofs4_wait(struct autofs_sb_info *,struct dentry *, enum autofs_notify); int autofs4_wait_release(struct autofs_sb_info *,autofs_wqt_t,int); void autofs4_catatonic_mode(struct autofs_sb_info *); @@ -200,12 +203,22 @@ static inline int autofs4_follow_mount(struct vfsmount **mnt, struct dentry **de return res; } +static inline u32 autofs4_get_dev(struct autofs_sb_info *sbi) +{ + return new_encode_dev(sbi->sb->s_dev); +} + +static inline u64 autofs4_get_ino(struct autofs_sb_info *sbi) +{ + return sbi->sb->s_root->d_inode->i_ino; +} + static inline int simple_positive(struct dentry *dentry) { return dentry->d_inode && !d_unhashed(dentry); } -static inline int simple_empty_nolock(struct dentry *dentry) +static inline int __simple_empty(struct dentry *dentry) { struct dentry *child; int ret = 0; @@ -217,3 +230,6 @@ static inline int simple_empty_nolock(struct dentry *dentry) out: return ret; } + +void autofs4_dentry_release(struct dentry *); + diff --git a/fs/autofs4/expire.c b/fs/autofs4/expire.c index dc39589df16..b8ce02607d6 100644 --- a/fs/autofs4/expire.c +++ b/fs/autofs4/expire.c @@ -4,7 +4,7 @@ * * Copyright 1997-1998 Transmeta Corporation -- All Rights Reserved * Copyright 1999-2000 Jeremy Fitzhardinge <jeremy@goop.org> - * Copyright 2001-2003 Ian Kent <raven@themaw.net> + * Copyright 2001-2006 Ian Kent <raven@themaw.net> * * This file is part of the Linux kernel and is made available under * the terms of the GNU General Public License, version 2, or at your @@ -16,7 +16,7 @@ static unsigned long now; -/* Check if a dentry can be expired return 1 if it can else return 0 */ +/* Check if a dentry can be expired */ static inline int autofs4_can_expire(struct dentry *dentry, unsigned long timeout, int do_now) { @@ -41,14 +41,14 @@ static inline int autofs4_can_expire(struct dentry *dentry, attempts if expire fails the first time */ ino->last_used = now; } - return 1; } -/* Check a mount point for busyness return 1 if not busy, otherwise */ -static int autofs4_check_mount(struct vfsmount *mnt, struct dentry *dentry) +/* Check a mount point for busyness */ +static int autofs4_mount_busy(struct vfsmount *mnt, struct dentry *dentry) { - int status = 0; + struct dentry *top = dentry; + int status = 1; DPRINTK("dentry %p %.*s", dentry, (int)dentry->d_name.len, dentry->d_name.name); @@ -63,9 +63,14 @@ static int autofs4_check_mount(struct vfsmount *mnt, struct dentry *dentry) if (is_autofs4_dentry(dentry)) goto done; - /* The big question */ - if (may_umount_tree(mnt) == 0) - status = 1; + /* Update the expiry counter if fs is busy */ + if (!may_umount_tree(mnt)) { + struct autofs_info *ino = autofs4_dentry_ino(top); + ino->last_used = jiffies; + goto done; + } + + status = 0; done: DPRINTK("returning = %d", status); mntput(mnt); @@ -73,78 +78,124 @@ done: return status; } +/* + * Calculate next entry in top down tree traversal. + * From next_mnt in namespace.c - elegant. + */ +static struct dentry *next_dentry(struct dentry *p, struct dentry *root) +{ + struct list_head *next = p->d_subdirs.next; + + if (next == &p->d_subdirs) { + while (1) { + if (p == root) + return NULL; + next = p->d_u.d_child.next; + if (next != &p->d_parent->d_subdirs) + break; + p = p->d_parent; + } + } + return list_entry(next, struct dentry, d_u.d_child); +} + +/* + * Check a direct mount point for busyness. + * Direct mounts have similar expiry semantics to tree mounts. + * The tree is not busy iff no mountpoints are busy and there are no + * autofs submounts. + */ +static int autofs4_direct_busy(struct vfsmount *mnt, + struct dentry *top, + unsigned long timeout, + int do_now) +{ + DPRINTK("top %p %.*s", + top, (int) top->d_name.len, top->d_name.name); + + /* If it's busy update the expiry counters */ + if (!may_umount_tree(mnt)) { + struct autofs_info *ino = autofs4_dentry_ino(top); + if (ino) + ino->last_used = jiffies; + return 1; + } + + /* Timeout of a direct mount is determined by its top dentry */ + if (!autofs4_can_expire(top, timeout, do_now)) + return 1; + + return 0; +} + /* Check a directory tree of mount points for busyness * The tree is not busy iff no mountpoints are busy - * Return 1 if the tree is busy or 0 otherwise */ -static int autofs4_check_tree(struct vfsmount *mnt, - struct dentry *top, - unsigned long timeout, - int do_now) +static int autofs4_tree_busy(struct vfsmount *mnt, + struct dentry *top, + unsigned long timeout, + int do_now) { - struct dentry *this_parent = top; - struct list_head *next; + struct autofs_info *top_ino = autofs4_dentry_ino(top); + struct dentry *p; - DPRINTK("parent %p %.*s", + DPRINTK("top %p %.*s", top, (int)top->d_name.len, top->d_name.name); /* Negative dentry - give up */ if (!simple_positive(top)) - return 0; - - /* Timeout of a tree mount is determined by its top dentry */ - if (!autofs4_can_expire(top, timeout, do_now)) - return 0; - - /* Is someone visiting anywhere in the tree ? */ - if (may_umount_tree(mnt)) - return 0; + return 1; spin_lock(&dcache_lock); -repeat: - next = this_parent->d_subdirs.next; -resume: - while (next != &this_parent->d_subdirs) { - struct dentry *dentry = list_entry(next, struct dentry, d_u.d_child); - + for (p = top; p; p = next_dentry(p, top)) { /* Negative dentry - give up */ - if (!simple_positive(dentry)) { - next = next->next; + if (!simple_positive(p)) continue; - } DPRINTK("dentry %p %.*s", - dentry, (int)dentry->d_name.len, dentry->d_name.name); - - if (!simple_empty_nolock(dentry)) { - this_parent = dentry; - goto repeat; - } + p, (int) p->d_name.len, p->d_name.name); - dentry = dget(dentry); + p = dget(p); spin_unlock(&dcache_lock); - if (d_mountpoint(dentry)) { - /* First busy => tree busy */ - if (!autofs4_check_mount(mnt, dentry)) { - dput(dentry); - return 0; + /* + * Is someone visiting anywhere in the subtree ? + * If there's no mount we need to check the usage + * count for the autofs dentry. + * If the fs is busy update the expiry counter. + */ + if (d_mountpoint(p)) { + if (autofs4_mount_busy(mnt, p)) { + top_ino->last_used = jiffies; + dput(p); + return 1; + } + } else { + struct autofs_info *ino = autofs4_dentry_ino(p); + unsigned int ino_count = atomic_read(&ino->count); + + /* allow for dget above and top is already dgot */ + if (p == top) + ino_count += 2; + else + ino_count++; + + if (atomic_read(&p->d_count) > ino_count) { + top_ino->last_used = jiffies; + dput(p); + return 1; } } - - dput(dentry); + dput(p); spin_lock(&dcache_lock); - next = next->next; - } - - if (this_parent != top) { - next = this_parent->d_u.d_child.next; - this_parent = this_parent->d_parent; - goto resume; } spin_unlock(&dcache_lock); - return 1; + /* Timeout of a tree mount is ultimately determined by its top dentry */ + if (!autofs4_can_expire(top, timeout, do_now)) + return 1; + + return 0; } static struct dentry *autofs4_check_leaves(struct vfsmount *mnt, @@ -152,58 +203,68 @@ static struct dentry *autofs4_check_leaves(struct vfsmount *mnt, unsigned long timeout, int do_now) { - struct dentry *this_parent = parent; - struct list_head *next; + struct dentry *p; DPRINTK("parent %p %.*s", parent, (int)parent->d_name.len, parent->d_name.name); spin_lock(&dcache_lock); -repeat: - next = this_parent->d_subdirs.next; -resume: - while (next != &this_parent->d_subdirs) { - struct dentry *dentry = list_entry(next, struct dentry, d_u.d_child); - + for (p = parent; p; p = next_dentry(p, parent)) { /* Negative dentry - give up */ - if (!simple_positive(dentry)) { - next = next->next; + if (!simple_positive(p)) continue; - } DPRINTK("dentry %p %.*s", - dentry, (int)dentry->d_name.len, dentry->d_name.name); - - if (!list_empty(&dentry->d_subdirs)) { - this_parent = dentry; - goto repeat; - } + p, (int) p->d_name.len, p->d_name.name); - dentry = dget(dentry); + p = dget(p); spin_unlock(&dcache_lock); - if (d_mountpoint(dentry)) { - /* Can we expire this guy */ - if (!autofs4_can_expire(dentry, timeout, do_now)) - goto cont; - + if (d_mountpoint(p)) { /* Can we umount this guy */ - if (autofs4_check_mount(mnt, dentry)) - return dentry; + if (autofs4_mount_busy(mnt, p)) + goto cont; + /* Can we expire this guy */ + if (autofs4_can_expire(p, timeout, do_now)) + return p; } cont: - dput(dentry); + dput(p); spin_lock(&dcache_lock); - next = next->next; } + spin_unlock(&dcache_lock); + return NULL; +} + +/* Check if we can expire a direct mount (possibly a tree) */ +static struct dentry *autofs4_expire_direct(struct super_block *sb, + struct vfsmount *mnt, + struct autofs_sb_info *sbi, + int how) +{ + unsigned long timeout; + struct dentry *root = dget(sb->s_root); + int do_now = how & AUTOFS_EXP_IMMEDIATE; + + if (!sbi->exp_timeout || !root) + return NULL; + + now = jiffies; + timeout = sbi->exp_timeout; + + /* Lock the tree as we must expire as a whole */ + spin_lock(&sbi->fs_lock); + if (!autofs4_direct_busy(mnt, root, timeout, do_now)) { + struct autofs_info *ino = autofs4_dentry_ino(root); - if (this_parent != parent) { - next = this_parent->d_u.d_child.next; - this_parent = this_parent->d_parent; - goto resume; + /* Set this flag early to catch sys_chdir and the like */ + ino->flags |= AUTOFS_INF_EXPIRING; + spin_unlock(&sbi->fs_lock); + return root; } - spin_unlock(&dcache_lock); + spin_unlock(&sbi->fs_lock); + dput(root); return NULL; } @@ -214,10 +275,10 @@ cont: * - it is unused by any user process * - it has been unused for exp_timeout time */ -static struct dentry *autofs4_expire(struct super_block *sb, - struct vfsmount *mnt, - struct autofs_sb_info *sbi, - int how) +static struct dentry *autofs4_expire_indirect(struct super_block *sb, + struct vfsmount *mnt, + struct autofs_sb_info *sbi, + int how) { unsigned long timeout; struct dentry *root = sb->s_root; @@ -241,7 +302,7 @@ static struct dentry *autofs4_expire(struct super_block *sb, struct dentry *dentry = list_entry(next, struct dentry, d_u.d_child); /* Negative dentry - give up */ - if ( !simple_positive(dentry) ) { + if (!simple_positive(dentry)) { next = next->next; continue; } @@ -249,31 +310,36 @@ static struct dentry *autofs4_expire(struct super_block *sb, dentry = dget(dentry); spin_unlock(&dcache_lock); - /* Case 1: indirect mount or top level direct mount */ + /* + * Case 1: (i) indirect mount or top level pseudo direct mount + * (autofs-4.1). + * (ii) indirect mount with offset mount, check the "/" + * offset (autofs-5.0+). + */ if (d_mountpoint(dentry)) { DPRINTK("checking mountpoint %p %.*s", dentry, (int)dentry->d_name.len, dentry->d_name.name); - /* Can we expire this guy */ - if (!autofs4_can_expire(dentry, timeout, do_now)) + /* Can we umount this guy */ + if (autofs4_mount_busy(mnt, dentry)) goto next; - /* Can we umount this guy */ - if (autofs4_check_mount(mnt, dentry)) { + /* Can we expire this guy */ + if (autofs4_can_expire(dentry, timeout, do_now)) { expired = dentry; break; } goto next; } - if ( simple_empty(dentry) ) + if (simple_empty(dentry)) goto next; /* Case 2: tree mount, expire iff entire tree is not busy */ if (!exp_leaves) { /* Lock the tree as we must expire as a whole */ spin_lock(&sbi->fs_lock); - if (autofs4_check_tree(mnt, dentry, timeout, do_now)) { + if (!autofs4_tree_busy(mnt, dentry, timeout, do_now)) { struct autofs_info *inf = autofs4_dentry_ino(dentry); /* Set this flag early to catch sys_chdir and the like */ @@ -283,7 +349,10 @@ static struct dentry *autofs4_expire(struct super_block *sb, break; } spin_unlock(&sbi->fs_lock); - /* Case 3: direct mount, expire individual leaves */ + /* + * Case 3: pseudo direct mount, expire individual leaves + * (autofs-4.1). + */ } else { expired = autofs4_check_leaves(mnt, dentry, timeout, do_now); if (expired) { @@ -297,7 +366,7 @@ next: next = next->next; } - if ( expired ) { + if (expired) { DPRINTK("returning %p %.*s", expired, (int)expired->d_name.len, expired->d_name.name); spin_lock(&dcache_lock); @@ -325,7 +394,7 @@ int autofs4_expire_run(struct super_block *sb, pkt.hdr.proto_version = sbi->version; pkt.hdr.type = autofs_ptype_expire; - if ((dentry = autofs4_expire(sb, mnt, sbi, 0)) == NULL) + if ((dentry = autofs4_expire_indirect(sb, mnt, sbi, 0)) == NULL) return -EAGAIN; pkt.len = dentry->d_name.len; @@ -351,17 +420,22 @@ int autofs4_expire_multi(struct super_block *sb, struct vfsmount *mnt, if (arg && get_user(do_now, arg)) return -EFAULT; - if ((dentry = autofs4_expire(sb, mnt, sbi, do_now)) != NULL) { - struct autofs_info *de_info = autofs4_dentry_ino(dentry); + if (sbi->type & AUTOFS_TYPE_DIRECT) + dentry = autofs4_expire_direct(sb, mnt, sbi, do_now); + else + dentry = autofs4_expire_indirect(sb, mnt, sbi, do_now); + + if (dentry) { + struct autofs_info *ino = autofs4_dentry_ino(dentry); /* This is synchronous because it makes the daemon a little easier */ - de_info->flags |= AUTOFS_INF_EXPIRING; + ino->flags |= AUTOFS_INF_EXPIRING; ret = autofs4_wait(sbi, dentry, NFY_EXPIRE); - de_info->flags &= ~AUTOFS_INF_EXPIRING; + ino->flags &= ~AUTOFS_INF_EXPIRING; dput(dentry); } - + return ret; } diff --git a/fs/autofs4/inode.c b/fs/autofs4/inode.c index 1ad98d48e55..fde78b110dd 100644 --- a/fs/autofs4/inode.c +++ b/fs/autofs4/inode.c @@ -3,6 +3,7 @@ * linux/fs/autofs/inode.c * * Copyright 1997-1998 Transmeta Corporation -- All Rights Reserved + * Copyright 2005-2006 Ian Kent <raven@themaw.net> * * This file is part of the Linux kernel and is made available under * the terms of the GNU General Public License, version 2, or at your @@ -13,6 +14,7 @@ #include <linux/kernel.h> #include <linux/slab.h> #include <linux/file.h> +#include <linux/seq_file.h> #include <linux/pagemap.h> #include <linux/parser.h> #include <linux/bitops.h> @@ -46,6 +48,7 @@ struct autofs_info *autofs4_init_ino(struct autofs_info *ino, ino->size = 0; ino->last_used = jiffies; + atomic_set(&ino->count, 0); ino->sbi = sbi; @@ -64,10 +67,19 @@ struct autofs_info *autofs4_init_ino(struct autofs_info *ino, void autofs4_free_ino(struct autofs_info *ino) { + struct autofs_info *p_ino; + if (ino->dentry) { ino->dentry->d_fsdata = NULL; - if (ino->dentry->d_inode) + if (ino->dentry->d_inode) { + struct dentry *parent = ino->dentry->d_parent; + if (atomic_dec_and_test(&ino->count)) { + p_ino = autofs4_dentry_ino(parent); + if (p_ino && parent != ino->dentry) + atomic_dec(&p_ino->count); + } dput(ino->dentry); + } ino->dentry = NULL; } if (ino->free) @@ -145,20 +157,44 @@ static void autofs4_put_super(struct super_block *sb) autofs4_catatonic_mode(sbi); /* Free wait queues, close pipe */ /* Clean up and release dangling references */ - if (sbi) - autofs4_force_release(sbi); + autofs4_force_release(sbi); kfree(sbi); DPRINTK("shutting down"); } +static int autofs4_show_options(struct seq_file *m, struct vfsmount *mnt) +{ + struct autofs_sb_info *sbi = autofs4_sbi(mnt->mnt_sb); + + if (!sbi) + return 0; + + seq_printf(m, ",fd=%d", sbi->pipefd); + seq_printf(m, ",pgrp=%d", sbi->oz_pgrp); + seq_printf(m, ",timeout=%lu", sbi->exp_timeout/HZ); + seq_printf(m, ",minproto=%d", sbi->min_proto); + seq_printf(m, ",maxproto=%d", sbi->max_proto); + + if (sbi->type & AUTOFS_TYPE_OFFSET) + seq_printf(m, ",offset"); + else if (sbi->type & AUTOFS_TYPE_DIRECT) + seq_printf(m, ",direct"); + else + seq_printf(m, ",indirect"); + + return 0; +} + static struct super_operations autofs4_sops = { .put_super = autofs4_put_super, .statfs = simple_statfs, + .show_options = autofs4_show_options, }; -enum {Opt_err, Opt_fd, Opt_uid, Opt_gid, Opt_pgrp, Opt_minproto, Opt_maxproto}; +enum {Opt_err, Opt_fd, Opt_uid, Opt_gid, Opt_pgrp, Opt_minproto, Opt_maxproto, + Opt_indirect, Opt_direct, Opt_offset}; static match_table_t tokens = { {Opt_fd, "fd=%u"}, @@ -167,11 +203,15 @@ static match_table_t tokens = { {Opt_pgrp, "pgrp=%u"}, {Opt_minproto, "minproto=%u"}, {Opt_maxproto, "maxproto=%u"}, + {Opt_indirect, "indirect"}, + {Opt_direct, "direct"}, + {Opt_offset, "offset"}, {Opt_err, NULL} }; static int parse_options(char *options, int *pipefd, uid_t *uid, gid_t *gid, - pid_t *pgrp, int *minproto, int *maxproto) + pid_t *pgrp, unsigned int *type, + int *minproto, int *maxproto) { char *p; substring_t args[MAX_OPT_ARGS]; @@ -225,6 +265,15 @@ static int parse_options(char *options, int *pipefd, uid_t *uid, gid_t *gid, return 1; *maxproto = option; break; + case Opt_indirect: + *type = AUTOFS_TYPE_INDIRECT; + break; + case Opt_direct: + *type = AUTOFS_TYPE_DIRECT; + break; + case Opt_offset: + *type = AUTOFS_TYPE_DIRECT | AUTOFS_TYPE_OFFSET; + break; default: return 1; } @@ -243,6 +292,10 @@ static struct autofs_info *autofs4_mkroot(struct autofs_sb_info *sbi) return ino; } +static struct dentry_operations autofs4_sb_dentry_operations = { + .d_release = autofs4_dentry_release, +}; + int autofs4_fill_super(struct super_block *s, void *data, int silent) { struct inode * root_inode; @@ -251,7 +304,6 @@ int autofs4_fill_super(struct super_block *s, void *data, int silent) int pipefd; struct autofs_sb_info *sbi; struct autofs_info *ino; - int minproto, maxproto; sbi = (struct autofs_sb_info *) kmalloc(sizeof(*sbi), GFP_KERNEL); if ( !sbi ) @@ -263,12 +315,16 @@ int autofs4_fill_super(struct super_block *s, void *data, int silent) s->s_fs_info = sbi; sbi->magic = AUTOFS_SBI_MAGIC; sbi->root = NULL; + sbi->pipefd = -1; sbi->catatonic = 0; sbi->exp_timeout = 0; sbi->oz_pgrp = process_group(current); sbi->sb = s; sbi->version = 0; sbi->sub_version = 0; + sbi->type = 0; + sbi->min_proto = 0; + sbi->max_proto = 0; mutex_init(&sbi->wq_mutex); spin_lock_init(&sbi->fs_lock); sbi->queues = NULL; @@ -285,38 +341,46 @@ int autofs4_fill_super(struct super_block *s, void *data, int silent) if (!ino) goto fail_free; root_inode = autofs4_get_inode(s, ino); - kfree(ino); if (!root_inode) - goto fail_free; + goto fail_ino; - root_inode->i_op = &autofs4_root_inode_operations; - root_inode->i_fop = &autofs4_root_operations; root = d_alloc_root(root_inode); - pipe = NULL; - if (!root) goto fail_iput; + pipe = NULL; + + root->d_op = &autofs4_sb_dentry_operations; + root->d_fsdata = ino; /* Can this call block? */ if (parse_options(data, &pipefd, &root_inode->i_uid, &root_inode->i_gid, - &sbi->oz_pgrp, - &minproto, &maxproto)) { + &sbi->oz_pgrp, &sbi->type, + &sbi->min_proto, &sbi->max_proto)) { printk("autofs: called with bogus options\n"); goto fail_dput; } + root_inode->i_fop = &autofs4_root_operations; + root_inode->i_op = sbi->type & AUTOFS_TYPE_DIRECT ? + &autofs4_direct_root_inode_operations : + &autofs4_indirect_root_inode_operations; + /* Couldn't this be tested earlier? */ - if (maxproto < AUTOFS_MIN_PROTO_VERSION || - minproto > AUTOFS_MAX_PROTO_VERSION) { + if (sbi->max_proto < AUTOFS_MIN_PROTO_VERSION || + sbi->min_proto > AUTOFS_MAX_PROTO_VERSION) { printk("autofs: kernel does not match daemon version " "daemon (%d, %d) kernel (%d, %d)\n", - minproto, maxproto, + sbi->min_proto, sbi->max_proto, AUTOFS_MIN_PROTO_VERSION, AUTOFS_MAX_PROTO_VERSION); goto fail_dput; } - sbi->version = maxproto > AUTOFS_MAX_PROTO_VERSION ? AUTOFS_MAX_PROTO_VERSION : maxproto; + /* Establish highest kernel protocol version */ + if (sbi->max_proto > AUTOFS_MAX_PROTO_VERSION) + sbi->version = AUTOFS_MAX_PROTO_VERSION; + else + sbi->version = sbi->max_proto; sbi->sub_version = AUTOFS_PROTO_SUBVERSION; DPRINTK("pipe fd = %d, pgrp = %u", pipefd, sbi->oz_pgrp); @@ -329,6 +393,7 @@ int autofs4_fill_super(struct super_block *s, void *data, int silent) if ( !pipe->f_op || !pipe->f_op->write ) goto fail_fput; sbi->pipe = pipe; + sbi->pipefd = pipefd; /* * Take a reference to the root dentry so we get a chance to @@ -356,6 +421,8 @@ fail_dput: fail_iput: printk("autofs: get root dentry failed\n"); iput(root_inode); +fail_ino: + kfree(ino); fail_free: kfree(sbi); fail_unlock: diff --git a/fs/autofs4/root.c b/fs/autofs4/root.c index 62d8d4acb8b..84e030c8ddd 100644 --- a/fs/autofs4/root.c +++ b/fs/autofs4/root.c @@ -4,7 +4,7 @@ * * Copyright 1997-1998 Transmeta Corporation -- All Rights Reserved * Copyright 1999-2000 Jeremy Fitzhardinge <jeremy@goop.org> - * Copyright 2001-2003 Ian Kent <raven@themaw.net> + * Copyright 2001-2006 Ian Kent <raven@themaw.net> * * This file is part of the Linux kernel and is made available under * the terms of the GNU General Public License, version 2, or at your @@ -30,9 +30,9 @@ static int autofs4_dir_close(struct inode *inode, struct file *file); static int autofs4_dir_readdir(struct file * filp, void * dirent, filldir_t filldir); static int autofs4_root_readdir(struct file * filp, void * dirent, filldir_t filldir); static struct dentry *autofs4_lookup(struct inode *,struct dentry *, struct nameidata *); -static int autofs4_dcache_readdir(struct file *, void *, filldir_t); +static void *autofs4_follow_link(struct dentry *, struct nameidata *); -struct file_operations autofs4_root_operations = { +const struct file_operations autofs4_root_operations = { .open = dcache_dir_open, .release = dcache_dir_close, .read = generic_read_dir, @@ -40,14 +40,14 @@ struct file_operations autofs4_root_operations = { .ioctl = autofs4_root_ioctl, }; -struct file_operations autofs4_dir_operations = { +const struct file_operations autofs4_dir_operations = { .open = autofs4_dir_open, .release = autofs4_dir_close, .read = generic_read_dir, .readdir = autofs4_dir_readdir, }; -struct inode_operations autofs4_root_inode_operations = { +struct inode_operations autofs4_indirect_root_inode_operations = { .lookup = autofs4_lookup, .unlink = autofs4_dir_unlink, .symlink = autofs4_dir_symlink, @@ -55,6 +55,14 @@ struct inode_operations autofs4_root_inode_operations = { .rmdir = autofs4_dir_rmdir, }; +struct inode_operations autofs4_direct_root_inode_operations = { + .lookup = autofs4_lookup, + .unlink = autofs4_dir_unlink, + .mkdir = autofs4_dir_mkdir, + .rmdir = autofs4_dir_rmdir, + .follow_link = autofs4_follow_link, +}; + struct inode_operations autofs4_dir_inode_operations = { .lookup = autofs4_lookup, .unlink = autofs4_dir_unlink, @@ -82,87 +90,7 @@ static int autofs4_root_readdir(struct file *file, void *dirent, DPRINTK("needs_reghost = %d", sbi->needs_reghost); - return autofs4_dcache_readdir(file, dirent, filldir); -} - -/* Update usage from here to top of tree, so that scan of - top-level directories will give a useful result */ -static void autofs4_update_usage(struct vfsmount *mnt, struct dentry *dentry) -{ - struct dentry *top = dentry->d_sb->s_root; - - spin_lock(&dcache_lock); - for(; dentry != top; dentry = dentry->d_parent) { - struct autofs_info *ino = autofs4_dentry_ino(dentry); - - if (ino) { - touch_atime(mnt, dentry); - ino->last_used = jiffies; - } - } - spin_unlock(&dcache_lock); -} - -/* - * From 2.4 kernel readdir.c - */ -static int autofs4_dcache_readdir(struct file * filp, void * dirent, filldir_t filldir) -{ - int i; - struct dentry *dentry = filp->f_dentry; - - i = filp->f_pos; - switch (i) { - case 0: - if (filldir(dirent, ".", 1, i, dentry->d_inode->i_ino, DT_DIR) < 0) - break; - i++; - filp->f_pos++; - /* fallthrough */ - case 1: - if (filldir(dirent, "..", 2, i, dentry->d_parent->d_inode->i_ino, DT_DIR) < 0) - break; - i++; - filp->f_pos++; - /* fallthrough */ - default: { - struct list_head *list; - int j = i-2; - - spin_lock(&dcache_lock); - list = dentry->d_subdirs.next; - - for (;;) { - if (list == &dentry->d_subdirs) { - spin_unlock(&dcache_lock); - return 0; - } - if (!j) - break; - j--; - list = list->next; - } - - while(1) { - struct dentry *de = list_entry(list, - struct dentry, d_u.d_child); - - if (!d_unhashed(de) && de->d_inode) { - spin_unlock(&dcache_lock); - if (filldir(dirent, de->d_name.name, de->d_name.len, filp->f_pos, de->d_inode->i_ino, DT_UNKNOWN) < 0) - break; - spin_lock(&dcache_lock); - } - filp->f_pos++; - list = list->next; - if (list != &dentry->d_subdirs) - continue; - spin_unlock(&dcache_lock); - break; - } - } - } - return 0; + return dcache_readdir(file, dirent, filldir); } static int autofs4_dir_open(struct inode *inode, struct file *file) @@ -170,8 +98,16 @@ static int autofs4_dir_open(struct inode *inode, struct file *file) struct dentry *dentry = file->f_dentry; struct vfsmount *mnt = file->f_vfsmnt; struct autofs_sb_info *sbi = autofs4_sbi(dentry->d_sb); + struct dentry *cursor; int status; + status = dcache_dir_open(inode, file); + if (status) + goto out; + + cursor = file->private_data; + cursor->d_fsdata = NULL; + DPRINTK("file=%p dentry=%p %.*s", file, dentry, dentry->d_name.len, dentry->d_name.name); @@ -180,12 +116,15 @@ static int autofs4_dir_open(struct inode *inode, struct file *file) if (autofs4_ispending(dentry)) { DPRINTK("dentry busy"); - return -EBUSY; + dcache_dir_close(inode, file); + status = -EBUSY; + goto out; } + status = -ENOENT; if (!d_mountpoint(dentry) && dentry->d_op && dentry->d_op->d_revalidate) { struct nameidata nd; - int empty; + int empty, ret; /* In case there are stale directory dentrys from a failed mount */ spin_lock(&dcache_lock); @@ -195,13 +134,13 @@ static int autofs4_dir_open(struct inode *inode, struct file *file) if (!empty) d_invalidate(dentry); - nd.dentry = dentry; - nd.mnt = mnt; nd.flags = LOOKUP_DIRECTORY; - status = (dentry->d_op->d_revalidate)(dentry, &nd); + ret = (dentry->d_op->d_revalidate)(dentry, &nd); - if (!status) - return -ENOENT; + if (!ret) { + dcache_dir_close(inode, file); + goto out; + } } if (d_mountpoint(dentry)) { @@ -212,25 +151,29 @@ static int autofs4_dir_open(struct inode *inode, struct file *file) if (!autofs4_follow_mount(&fp_mnt, &fp_dentry)) { dput(fp_dentry); mntput(fp_mnt); - return -ENOENT; + dcache_dir_close(inode, file); + goto out; } fp = dentry_open(fp_dentry, fp_mnt, file->f_flags); status = PTR_ERR(fp); if (IS_ERR(fp)) { - file->private_data = NULL; - return status; + dcache_dir_close(inode, file); + goto out; } - file->private_data = fp; + cursor->d_fsdata = fp; } -out: return 0; +out: + return status; } static int autofs4_dir_close(struct inode *inode, struct file *file) { struct dentry *dentry = file->f_dentry; struct autofs_sb_info *sbi = autofs4_sbi(dentry->d_sb); + struct dentry *cursor = file->private_data; + int status = 0; DPRINTK("file=%p dentry=%p %.*s", file, dentry, dentry->d_name.len, dentry->d_name.name); @@ -240,26 +183,28 @@ static int autofs4_dir_close(struct inode *inode, struct file *file) if (autofs4_ispending(dentry)) { DPRINTK("dentry busy"); - return -EBUSY; + status = -EBUSY; + goto out; } if (d_mountpoint(dentry)) { - struct file *fp = file->private_data; - - if (!fp) - return -ENOENT; - + struct file *fp = cursor->d_fsdata; + if (!fp) { + status = -ENOENT; + goto out; + } filp_close(fp, current->files); - file->private_data = NULL; } out: - return 0; + dcache_dir_close(inode, file); + return status; } static int autofs4_dir_readdir(struct file *file, void *dirent, filldir_t filldir) { struct dentry *dentry = file->f_dentry; struct autofs_sb_info *sbi = autofs4_sbi(dentry->d_sb); + struct dentry *cursor = file->private_data; int status; DPRINTK("file=%p dentry=%p %.*s", @@ -274,7 +219,7 @@ static int autofs4_dir_readdir(struct file *file, void *dirent, filldir_t filldi } if (d_mountpoint(dentry)) { - struct file *fp = file->private_data; + struct file *fp = cursor->d_fsdata; if (!fp) return -ENOENT; @@ -289,27 +234,26 @@ static int autofs4_dir_readdir(struct file *file, void *dirent, filldir_t filldi return status; } out: - return autofs4_dcache_readdir(file, dirent, filldir); + return dcache_readdir(file, dirent, filldir); } -static int try_to_fill_dentry(struct vfsmount *mnt, struct dentry *dentry, int flags) +static int try_to_fill_dentry(struct dentry *dentry, int flags) { - struct super_block *sb = mnt->mnt_sb; - struct autofs_sb_info *sbi = autofs4_sbi(sb); - struct autofs_info *de_info = autofs4_dentry_ino(dentry); + struct autofs_sb_info *sbi = autofs4_sbi(dentry->d_sb); + struct autofs_info *ino = autofs4_dentry_ino(dentry); int status = 0; /* Block on any pending expiry here; invalidate the dentry when expiration is done to trigger mount request with a new dentry */ - if (de_info && (de_info->flags & AUTOFS_INF_EXPIRING)) { + if (ino && (ino->flags & AUTOFS_INF_EXPIRING)) { DPRINTK("waiting for expire %p name=%.*s", dentry, dentry->d_name.len, dentry->d_name.name); status = autofs4_wait(sbi, dentry, NFY_NONE); - + DPRINTK("expire done status=%d", status); - + /* * If the directory still exists the mount request must * continue otherwise it can't be followed at the right @@ -317,34 +261,36 @@ static int try_to_fill_dentry(struct vfsmount *mnt, struct dentry *dentry, int f */ status = d_invalidate(dentry); if (status != -EBUSY) - return 0; + return -ENOENT; } DPRINTK("dentry=%p %.*s ino=%p", dentry, dentry->d_name.len, dentry->d_name.name, dentry->d_inode); - /* Wait for a pending mount, triggering one if there isn't one already */ + /* + * Wait for a pending mount, triggering one if there + * isn't one already + */ if (dentry->d_inode == NULL) { DPRINTK("waiting for mount name=%.*s", dentry->d_name.len, dentry->d_name.name); status = autofs4_wait(sbi, dentry, NFY_MOUNT); - + DPRINTK("mount done status=%d", status); if (status && dentry->d_inode) - return 0; /* Try to get the kernel to invalidate this dentry */ - + return status; /* Try to get the kernel to invalidate this dentry */ + /* Turn this into a real negative dentry? */ if (status == -ENOENT) { - dentry->d_time = jiffies + AUTOFS_NEGATIVE_TIMEOUT; spin_lock(&dentry->d_lock); dentry->d_flags &= ~DCACHE_AUTOFS_PENDING; spin_unlock(&dentry->d_lock); - return 1; + return status; } else if (status) { /* Return a negative dentry, but leave it "pending" */ - return 1; + return status; } /* Trigger mount for path component or follow link */ } else if (flags & (LOOKUP_CONTINUE | LOOKUP_DIRECTORY) || @@ -363,19 +309,87 @@ static int try_to_fill_dentry(struct vfsmount *mnt, struct dentry *dentry, int f spin_lock(&dentry->d_lock); dentry->d_flags &= ~DCACHE_AUTOFS_PENDING; spin_unlock(&dentry->d_lock); - return 0; + return status; } } - /* We don't update the usages for the autofs daemon itself, this - is necessary for recursive autofs mounts */ - if (!autofs4_oz_mode(sbi)) - autofs4_update_usage(mnt, dentry); + /* Initialize expiry counter after successful mount */ + if (ino) + ino->last_used = jiffies; spin_lock(&dentry->d_lock); dentry->d_flags &= ~DCACHE_AUTOFS_PENDING; spin_unlock(&dentry->d_lock); - return 1; + return status; +} + +/* For autofs direct mounts the follow link triggers the mount */ +static void *autofs4_follow_link(struct dentry *dentry, struct nameidata *nd) +{ + struct autofs_sb_info *sbi = autofs4_sbi(dentry->d_sb); + int oz_mode = autofs4_oz_mode(sbi); + unsigned int lookup_type; + int status; + + DPRINTK("dentry=%p %.*s oz_mode=%d nd->flags=%d", + dentry, dentry->d_name.len, dentry->d_name.name, oz_mode, + nd->flags); + + /* If it's our master or we shouldn't trigger a mount we're done */ + lookup_type = nd->flags & (LOOKUP_CONTINUE | LOOKUP_DIRECTORY); + if (oz_mode || !lookup_type) + goto done; + + /* + * If a request is pending wait for it. + * If it's a mount then it won't be expired till at least + * a liitle later and if it's an expire then we might need + * to mount it again. + */ + if (autofs4_ispending(dentry)) { + DPRINTK("waiting for active request %p name=%.*s", + dentry, dentry->d_name.len, dentry->d_name.name); + + status = autofs4_wait(sbi, dentry, NFY_NONE); + + DPRINTK("request done status=%d", status); + } + + /* + * If the dentry contains directories then it is an + * autofs multi-mount with no root mount offset. So + * don't try to mount it again. + */ + spin_lock(&dcache_lock); + if (!d_mountpoint(dentry) && list_empty(&dentry->d_subdirs)) { + spin_unlock(&dcache_lock); + + status = try_to_fill_dentry(dentry, 0); + if (status) + goto out_error; + + /* + * The mount succeeded but if there is no root mount + * it must be an autofs multi-mount with no root offset + * so we don't need to follow the mount. + */ + if (d_mountpoint(dentry)) { + if (!autofs4_follow_mount(&nd->mnt, &nd->dentry)) { + status = -ENOENT; + goto out_error; + } + } + + goto done; + } + spin_unlock(&dcache_lock); + +done: + return NULL; + +out_error: + path_release(nd); + return ERR_PTR(status); } /* @@ -384,47 +398,43 @@ static int try_to_fill_dentry(struct vfsmount *mnt, struct dentry *dentry, int f * yet completely filled in, and revalidate has to delay such * lookups.. */ -static int autofs4_revalidate(struct dentry * dentry, struct nameidata *nd) +static int autofs4_revalidate(struct dentry *dentry, struct nameidata *nd) { - struct inode * dir = dentry->d_parent->d_inode; + struct inode *dir = dentry->d_parent->d_inode; struct autofs_sb_info *sbi = autofs4_sbi(dir->i_sb); int oz_mode = autofs4_oz_mode(sbi); int flags = nd ? nd->flags : 0; - int status = 1; + int status = 0; /* Pending dentry */ if (autofs4_ispending(dentry)) { if (!oz_mode) - status = try_to_fill_dentry(nd->mnt, dentry, flags); - return status; + status = try_to_fill_dentry(dentry, flags); + return !status; } /* Negative dentry.. invalidate if "old" */ if (dentry->d_inode == NULL) - return (dentry->d_time - jiffies <= AUTOFS_NEGATIVE_TIMEOUT); + return 0; /* Check for a non-mountpoint directory with no contents */ spin_lock(&dcache_lock); if (S_ISDIR(dentry->d_inode->i_mode) && !d_mountpoint(dentry) && - list_empty(&dentry->d_subdirs)) { + __simple_empty(dentry)) { DPRINTK("dentry=%p %.*s, emptydir", dentry, dentry->d_name.len, dentry->d_name.name); spin_unlock(&dcache_lock); if (!oz_mode) - status = try_to_fill_dentry(nd->mnt, dentry, flags); - return status; + status = try_to_fill_dentry(dentry, flags); + return !status; } spin_unlock(&dcache_lock); - /* Update the usage list */ - if (!oz_mode) - autofs4_update_usage(nd->mnt, dentry); - return 1; } -static void autofs4_dentry_release(struct dentry *de) +void autofs4_dentry_release(struct dentry *de) { struct autofs_info *inf; @@ -462,12 +472,13 @@ static struct dentry *autofs4_lookup(struct inode *dir, struct dentry *dentry, s DPRINTK("name = %.*s", dentry->d_name.len, dentry->d_name.name); + /* File name too long to exist */ if (dentry->d_name.len > NAME_MAX) - return ERR_PTR(-ENAMETOOLONG);/* File name too long to exist */ + return ERR_PTR(-ENAMETOOLONG); sbi = autofs4_sbi(dir->i_sb); - oz_mode = autofs4_oz_mode(sbi); + DPRINTK("pid = %u, pgrp = %u, catatonic = %d, oz_mode = %d", current->pid, process_group(current), sbi->catatonic, oz_mode); @@ -519,7 +530,7 @@ static struct dentry *autofs4_lookup(struct inode *dir, struct dentry *dentry, s * doesn't do the right thing for all system calls, but it should * be OK for the operations we permit from an autofs. */ - if ( dentry->d_inode && d_unhashed(dentry) ) + if (dentry->d_inode && d_unhashed(dentry)) return ERR_PTR(-ENOENT); return NULL; @@ -531,6 +542,7 @@ static int autofs4_dir_symlink(struct inode *dir, { struct autofs_sb_info *sbi = autofs4_sbi(dir->i_sb); struct autofs_info *ino = autofs4_dentry_ino(dentry); + struct autofs_info *p_ino; struct inode *inode; char *cp; @@ -564,6 +576,10 @@ static int autofs4_dir_symlink(struct inode *dir, dentry->d_fsdata = ino; ino->dentry = dget(dentry); + atomic_inc(&ino->count); + p_ino = autofs4_dentry_ino(dentry->d_parent); + if (p_ino && dentry->d_parent != dentry) + atomic_inc(&p_ino->count); ino->inode = inode; dir->i_mtime = CURRENT_TIME; @@ -590,11 +606,17 @@ static int autofs4_dir_unlink(struct inode *dir, struct dentry *dentry) { struct autofs_sb_info *sbi = autofs4_sbi(dir->i_sb); struct autofs_info *ino = autofs4_dentry_ino(dentry); + struct autofs_info *p_ino; /* This allows root to remove symlinks */ if ( !autofs4_oz_mode(sbi) && !capable(CAP_SYS_ADMIN) ) return -EACCES; + if (atomic_dec_and_test(&ino->count)) { + p_ino = autofs4_dentry_ino(dentry->d_parent); + if (p_ino && dentry->d_parent != dentry) + atomic_dec(&p_ino->count); + } dput(ino->dentry); dentry->d_inode->i_size = 0; @@ -611,6 +633,7 @@ static int autofs4_dir_rmdir(struct inode *dir, struct dentry *dentry) { struct autofs_sb_info *sbi = autofs4_sbi(dir->i_sb); struct autofs_info *ino = autofs4_dentry_ino(dentry); + struct autofs_info *p_ino; if (!autofs4_oz_mode(sbi)) return -EACCES; @@ -625,8 +648,12 @@ static int autofs4_dir_rmdir(struct inode *dir, struct dentry *dentry) spin_unlock(&dentry->d_lock); spin_unlock(&dcache_lock); + if (atomic_dec_and_test(&ino->count)) { + p_ino = autofs4_dentry_ino(dentry->d_parent); + if (p_ino && dentry->d_parent != dentry) + atomic_dec(&p_ino->count); + } dput(ino->dentry); - dentry->d_inode->i_size = 0; dentry->d_inode->i_nlink = 0; @@ -640,6 +667,7 @@ static int autofs4_dir_mkdir(struct inode *dir, struct dentry *dentry, int mode) { struct autofs_sb_info *sbi = autofs4_sbi(dir->i_sb); struct autofs_info *ino = autofs4_dentry_ino(dentry); + struct autofs_info *p_ino; struct inode *inode; if ( !autofs4_oz_mode(sbi) ) @@ -662,6 +690,10 @@ static int autofs4_dir_mkdir(struct inode *dir, struct dentry *dentry, int mode) dentry->d_fsdata = ino; ino->dentry = dget(dentry); + atomic_inc(&ino->count); + p_ino = autofs4_dentry_ino(dentry->d_parent); + if (p_ino && dentry->d_parent != dentry) + atomic_inc(&p_ino->count); ino->inode = inode; dir->i_nlink++; dir->i_mtime = CURRENT_TIME; @@ -745,7 +777,7 @@ static inline int autofs4_ask_umount(struct vfsmount *mnt, int __user *p) { int status = 0; - if (may_umount(mnt) == 0) + if (may_umount(mnt)) status = 1; DPRINTK("returning %d", status); diff --git a/fs/autofs4/waitq.c b/fs/autofs4/waitq.c index be78e9378c0..142ab6aa2aa 100644 --- a/fs/autofs4/waitq.c +++ b/fs/autofs4/waitq.c @@ -3,7 +3,7 @@ * linux/fs/autofs/waitq.c * * Copyright 1997-1998 Transmeta Corporation -- All Rights Reserved - * Copyright 2001-2003 Ian Kent <raven@themaw.net> + * Copyright 2001-2006 Ian Kent <raven@themaw.net> * * This file is part of the Linux kernel and is made available under * the terms of the GNU General Public License, version 2, or at your @@ -33,7 +33,7 @@ void autofs4_catatonic_mode(struct autofs_sb_info *sbi) sbi->catatonic = 1; wq = sbi->queues; sbi->queues = NULL; /* Erase all wait queues */ - while ( wq ) { + while (wq) { nwq = wq->next; wq->status = -ENOENT; /* Magic is gone - report failure */ kfree(wq->name); @@ -45,7 +45,6 @@ void autofs4_catatonic_mode(struct autofs_sb_info *sbi) fput(sbi->pipe); /* Close the pipe */ sbi->pipe = NULL; } - shrink_dcache_sb(sbi->sb); } @@ -98,7 +97,10 @@ static void autofs4_notify_daemon(struct autofs_sb_info *sbi, pkt.hdr.proto_version = sbi->version; pkt.hdr.type = type; - if (type == autofs_ptype_missing) { + switch (type) { + /* Kernel protocol v4 missing and expire packets */ + case autofs_ptype_missing: + { struct autofs_packet_missing *mp = &pkt.missing; pktsz = sizeof(*mp); @@ -107,7 +109,10 @@ static void autofs4_notify_daemon(struct autofs_sb_info *sbi, mp->len = wq->len; memcpy(mp->name, wq->name, wq->len); mp->name[wq->len] = '\0'; - } else if (type == autofs_ptype_expire_multi) { + break; + } + case autofs_ptype_expire_multi: + { struct autofs_packet_expire_multi *ep = &pkt.expire_multi; pktsz = sizeof(*ep); @@ -116,7 +121,34 @@ static void autofs4_notify_daemon(struct autofs_sb_info *sbi, ep->len = wq->len; memcpy(ep->name, wq->name, wq->len); ep->name[wq->len] = '\0'; - } else { + break; + } + /* + * Kernel protocol v5 packet for handling indirect and direct + * mount missing and expire requests + */ + case autofs_ptype_missing_indirect: + case autofs_ptype_expire_indirect: + case autofs_ptype_missing_direct: + case autofs_ptype_expire_direct: + { + struct autofs_v5_packet *packet = &pkt.v5_packet; + + pktsz = sizeof(*packet); + + packet->wait_queue_token = wq->wait_queue_token; + packet->len = wq->len; + memcpy(packet->name, wq->name, wq->len); + packet->name[wq->len] = '\0'; + packet->dev = wq->dev; + packet->ino = wq->ino; + packet->uid = wq->uid; + packet->gid = wq->gid; + packet->pid = wq->pid; + packet->tgid = wq->tgid; + break; + } + default: printk("autofs4_notify_daemon: bad type %d!\n", type); return; } @@ -162,21 +194,29 @@ int autofs4_wait(struct autofs_sb_info *sbi, struct dentry *dentry, { struct autofs_wait_queue *wq; char *name; - int len, status; + unsigned int len = 0; + unsigned int hash = 0; + int status; /* In catatonic mode, we don't wait for nobody */ - if ( sbi->catatonic ) + if (sbi->catatonic) return -ENOENT; name = kmalloc(NAME_MAX + 1, GFP_KERNEL); if (!name) return -ENOMEM; - len = autofs4_getpath(sbi, dentry, &name); - if (!len) { - kfree(name); - return -ENOENT; + /* If this is a direct mount request create a dummy name */ + if (IS_ROOT(dentry) && (sbi->type & AUTOFS_TYPE_DIRECT)) + len = sprintf(name, "%p", dentry); + else { + len = autofs4_getpath(sbi, dentry, &name); + if (!len) { + kfree(name); + return -ENOENT; + } } + hash = full_name_hash(name, len); if (mutex_lock_interruptible(&sbi->wq_mutex)) { kfree(name); @@ -190,7 +230,7 @@ int autofs4_wait(struct autofs_sb_info *sbi, struct dentry *dentry, break; } - if ( !wq ) { + if (!wq) { /* Can't wait for an expire if there's no mount */ if (notify == NFY_NONE && !d_mountpoint(dentry)) { kfree(name); @@ -200,7 +240,7 @@ int autofs4_wait(struct autofs_sb_info *sbi, struct dentry *dentry, /* Create a new wait queue */ wq = kmalloc(sizeof(struct autofs_wait_queue),GFP_KERNEL); - if ( !wq ) { + if (!wq) { kfree(name); mutex_unlock(&sbi->wq_mutex); return -ENOMEM; @@ -212,12 +252,18 @@ int autofs4_wait(struct autofs_sb_info *sbi, struct dentry *dentry, wq->next = sbi->queues; sbi->queues = wq; init_waitqueue_head(&wq->queue); - wq->hash = dentry->d_name.hash; + wq->hash = hash; wq->name = name; wq->len = len; + wq->dev = autofs4_get_dev(sbi); + wq->ino = autofs4_get_ino(sbi); + wq->uid = current->uid; + wq->gid = current->gid; + wq->pid = current->pid; + wq->tgid = current->tgid; wq->status = -EINTR; /* Status return if interrupted */ atomic_set(&wq->wait_ctr, 2); - atomic_set(&wq->notified, 1); + atomic_set(&wq->notify, 1); mutex_unlock(&sbi->wq_mutex); } else { atomic_inc(&wq->wait_ctr); @@ -227,9 +273,26 @@ int autofs4_wait(struct autofs_sb_info *sbi, struct dentry *dentry, (unsigned long) wq->wait_queue_token, wq->len, wq->name, notify); } - if (notify != NFY_NONE && atomic_dec_and_test(&wq->notified)) { - int type = (notify == NFY_MOUNT ? - autofs_ptype_missing : autofs_ptype_expire_multi); + if (notify != NFY_NONE && atomic_read(&wq->notify)) { + int type; + + atomic_dec(&wq->notify); + + if (sbi->version < 5) { + if (notify == NFY_MOUNT) + type = autofs_ptype_missing; + else + type = autofs_ptype_expire_multi; + } else { + if (notify == NFY_MOUNT) + type = (sbi->type & AUTOFS_TYPE_DIRECT) ? + autofs_ptype_missing_direct : + autofs_ptype_missing_indirect; + else + type = (sbi->type & AUTOFS_TYPE_DIRECT) ? + autofs_ptype_expire_direct : + autofs_ptype_expire_indirect; + } DPRINTK("new wait id = 0x%08lx, name = %.*s, nfy=%d\n", (unsigned long) wq->wait_queue_token, wq->len, wq->name, notify); @@ -240,14 +303,14 @@ int autofs4_wait(struct autofs_sb_info *sbi, struct dentry *dentry, /* wq->name is NULL if and only if the lock is already released */ - if ( sbi->catatonic ) { + if (sbi->catatonic) { /* We might have slept, so check again for catatonic mode */ wq->status = -ENOENT; kfree(wq->name); wq->name = NULL; } - if ( wq->name ) { + if (wq->name) { /* Block all but "shutdown" signals while waiting */ sigset_t oldset; unsigned long irqflags; @@ -283,12 +346,12 @@ int autofs4_wait_release(struct autofs_sb_info *sbi, autofs_wqt_t wait_queue_tok struct autofs_wait_queue *wq, **wql; mutex_lock(&sbi->wq_mutex); - for ( wql = &sbi->queues ; (wq = *wql) != 0 ; wql = &wq->next ) { - if ( wq->wait_queue_token == wait_queue_token ) + for (wql = &sbi->queues ; (wq = *wql) != 0 ; wql = &wq->next) { + if (wq->wait_queue_token == wait_queue_token) break; } - if ( !wq ) { + if (!wq) { mutex_unlock(&sbi->wq_mutex); return -EINVAL; } diff --git a/fs/bad_inode.c b/fs/bad_inode.c index e172180a1d8..80599ae3396 100644 --- a/fs/bad_inode.c +++ b/fs/bad_inode.c @@ -22,7 +22,7 @@ static int return_EIO(void) #define EIO_ERROR ((void *) (return_EIO)) -static struct file_operations bad_file_ops = +static const struct file_operations bad_file_ops = { .llseek = EIO_ERROR, .aio_read = EIO_ERROR, diff --git a/fs/befs/datastream.c b/fs/befs/datastream.c index 785f6b2d5d1..b7d6b920f65 100644 --- a/fs/befs/datastream.c +++ b/fs/befs/datastream.c @@ -118,7 +118,7 @@ befs_fblock2brun(struct super_block *sb, befs_data_stream * data, * befs_read_lsmylink - read long symlink from datastream. * @sb: Filesystem superblock * @ds: Datastrem to read from - * @buf: Buffer in wich to place long symlink data + * @buf: Buffer in which to place long symlink data * @len: Length of the long symlink in bytes * * Returns the number of bytes read diff --git a/fs/befs/linuxvfs.c b/fs/befs/linuxvfs.c index 044a5958782..68ebd10f345 100644 --- a/fs/befs/linuxvfs.c +++ b/fs/befs/linuxvfs.c @@ -64,7 +64,7 @@ static const struct super_operations befs_sops = { /* slab cache for befs_inode_info objects */ static kmem_cache_t *befs_inode_cachep; -static struct file_operations befs_dir_operations = { +static const struct file_operations befs_dir_operations = { .read = generic_read_dir, .readdir = befs_readdir, }; diff --git a/fs/bfs/bfs.h b/fs/bfs/bfs.h index 1fbc53f14ab..9d791004b21 100644 --- a/fs/bfs/bfs.h +++ b/fs/bfs/bfs.h @@ -49,11 +49,11 @@ static inline struct bfs_inode_info *BFS_I(struct inode *inode) /* file.c */ extern struct inode_operations bfs_file_inops; -extern struct file_operations bfs_file_operations; +extern const struct file_operations bfs_file_operations; extern struct address_space_operations bfs_aops; /* dir.c */ extern struct inode_operations bfs_dir_inops; -extern struct file_operations bfs_dir_operations; +extern const struct file_operations bfs_dir_operations; #endif /* _FS_BFS_BFS_H */ diff --git a/fs/bfs/dir.c b/fs/bfs/dir.c index 5af928fa044..26fad962173 100644 --- a/fs/bfs/dir.c +++ b/fs/bfs/dir.c @@ -70,7 +70,7 @@ static int bfs_readdir(struct file * f, void * dirent, filldir_t filldir) return 0; } -struct file_operations bfs_dir_operations = { +const struct file_operations bfs_dir_operations = { .read = generic_read_dir, .readdir = bfs_readdir, .fsync = file_fsync, diff --git a/fs/bfs/file.c b/fs/bfs/file.c index 807723b65da..d83cd74a2e4 100644 --- a/fs/bfs/file.c +++ b/fs/bfs/file.c @@ -17,7 +17,7 @@ #define dprintf(x...) #endif -struct file_operations bfs_file_operations = { +const struct file_operations bfs_file_operations = { .llseek = generic_file_llseek, .read = generic_file_read, .write = generic_file_write, diff --git a/fs/binfmt_elf.c b/fs/binfmt_elf.c index c2eac2a50bd..537893a1601 100644 --- a/fs/binfmt_elf.c +++ b/fs/binfmt_elf.c @@ -500,17 +500,22 @@ out: #define INTERPRETER_AOUT 1 #define INTERPRETER_ELF 2 +#ifndef STACK_RND_MASK +#define STACK_RND_MASK 0x7ff /* with 4K pages 8MB of VA */ +#endif static unsigned long randomize_stack_top(unsigned long stack_top) { unsigned int random_variable = 0; - if (current->flags & PF_RANDOMIZE) - random_variable = get_random_int() % (8*1024*1024); + if (current->flags & PF_RANDOMIZE) { + random_variable = get_random_int() & STACK_RND_MASK; + random_variable <<= PAGE_SHIFT; + } #ifdef CONFIG_STACK_GROWSUP - return PAGE_ALIGN(stack_top + random_variable); + return PAGE_ALIGN(stack_top) + random_variable; #else - return PAGE_ALIGN(stack_top - random_variable); + return PAGE_ALIGN(stack_top) - random_variable; #endif } @@ -1334,7 +1339,7 @@ static int fill_psinfo(struct elf_prpsinfo *psinfo, struct task_struct *p, i = p->state ? ffz(~p->state) + 1 : 0; psinfo->pr_state = i; - psinfo->pr_sname = (i < 0 || i > 5) ? '.' : "RSDTZW"[i]; + psinfo->pr_sname = (i > 5) ? '.' : "RSDTZW"[i]; psinfo->pr_zomb = psinfo->pr_sname == 'Z'; psinfo->pr_nice = task_nice(p); psinfo->pr_flag = p->flags; @@ -1465,12 +1470,11 @@ static int elf_core_dump(long signr, struct pt_regs * regs, struct file * file) read_lock(&tasklist_lock); do_each_thread(g,p) if (current->mm == p->mm && current != p) { - tmp = kmalloc(sizeof(*tmp), GFP_ATOMIC); + tmp = kzalloc(sizeof(*tmp), GFP_ATOMIC); if (!tmp) { read_unlock(&tasklist_lock); goto cleanup; } - memset(tmp, 0, sizeof(*tmp)); INIT_LIST_HEAD(&tmp->list); tmp->thread = p; list_add(&tmp->list, &thread_list); diff --git a/fs/binfmt_elf_fdpic.c b/fs/binfmt_elf_fdpic.c index 5b3076e8ee9..a2e48c999c2 100644 --- a/fs/binfmt_elf_fdpic.c +++ b/fs/binfmt_elf_fdpic.c @@ -572,8 +572,7 @@ static int create_elf_fdpic_tables(struct linux_binprm *bprm, csp -= sizeof(unsigned long); __put_user(bprm->argc, (unsigned long *) csp); - if (csp != sp) - BUG(); + BUG_ON(csp != sp); /* fill in the argv[] array */ #ifdef CONFIG_MMU diff --git a/fs/binfmt_flat.c b/fs/binfmt_flat.c index 108d56bbd0d..69f44dcdb0b 100644 --- a/fs/binfmt_flat.c +++ b/fs/binfmt_flat.c @@ -36,6 +36,7 @@ #include <linux/personality.h> #include <linux/init.h> #include <linux/flat.h> +#include <linux/syscalls.h> #include <asm/byteorder.h> #include <asm/system.h> @@ -426,6 +427,8 @@ static int load_flat_file(struct linux_binprm * bprm, int i, rev, relocs = 0; loff_t fpos; unsigned long start_code, end_code; + int ret; + int exec_fileno; hdr = ((struct flat_hdr *) bprm->buf); /* exec-header */ inode = bprm->file->f_dentry->d_inode; @@ -450,7 +453,8 @@ static int load_flat_file(struct linux_binprm * bprm, */ if (strncmp(hdr->magic, "#!", 2)) printk("BINFMT_FLAT: bad header magic\n"); - return -ENOEXEC; + ret = -ENOEXEC; + goto err; } if (flags & FLAT_FLAG_KTRACE) @@ -458,14 +462,16 @@ static int load_flat_file(struct linux_binprm * bprm, if (rev != FLAT_VERSION && rev != OLD_FLAT_VERSION) { printk("BINFMT_FLAT: bad flat file version 0x%x (supported 0x%x and 0x%x)\n", rev, FLAT_VERSION, OLD_FLAT_VERSION); - return -ENOEXEC; + ret = -ENOEXEC; + goto err; } /* Don't allow old format executables to use shared libraries */ if (rev == OLD_FLAT_VERSION && id != 0) { printk("BINFMT_FLAT: shared libraries are not available before rev 0x%x\n", (int) FLAT_VERSION); - return -ENOEXEC; + ret = -ENOEXEC; + goto err; } /* @@ -478,7 +484,8 @@ static int load_flat_file(struct linux_binprm * bprm, #ifndef CONFIG_BINFMT_ZFLAT if (flags & (FLAT_FLAG_GZIP|FLAT_FLAG_GZDATA)) { printk("Support for ZFLAT executables is not enabled.\n"); - return -ENOEXEC; + ret = -ENOEXEC; + goto err; } #endif @@ -490,14 +497,27 @@ static int load_flat_file(struct linux_binprm * bprm, rlim = current->signal->rlim[RLIMIT_DATA].rlim_cur; if (rlim >= RLIM_INFINITY) rlim = ~0; - if (data_len + bss_len > rlim) - return -ENOMEM; + if (data_len + bss_len > rlim) { + ret = -ENOMEM; + goto err; + } + + /* check file descriptor */ + exec_fileno = get_unused_fd(); + if (exec_fileno < 0) { + ret = -EMFILE; + goto err; + } + get_file(bprm->file); + fd_install(exec_fileno, bprm->file); /* Flush all traces of the currently running executable */ if (id == 0) { result = flush_old_exec(bprm); - if (result) - return result; + if (result) { + ret = result; + goto err_close; + } /* OK, This is the point of no return */ set_personality(PER_LINUX); @@ -527,7 +547,8 @@ static int load_flat_file(struct linux_binprm * bprm, if (!textpos) textpos = (unsigned long) -ENOMEM; printk("Unable to mmap process text, errno %d\n", (int)-textpos); - return(textpos); + ret = textpos; + goto err_close; } down_write(¤t->mm->mmap_sem); @@ -542,7 +563,8 @@ static int load_flat_file(struct linux_binprm * bprm, printk("Unable to allocate RAM for process data, errno %d\n", (int)-datapos); do_munmap(current->mm, textpos, text_len); - return realdatastart; + ret = realdatastart; + goto err_close; } datapos = realdatastart + MAX_SHARED_LIBS * sizeof(unsigned long); @@ -564,7 +586,8 @@ static int load_flat_file(struct linux_binprm * bprm, printk("Unable to read data+bss, errno %d\n", (int)-result); do_munmap(current->mm, textpos, text_len); do_munmap(current->mm, realdatastart, data_len + extra); - return result; + ret = result; + goto err_close; } reloc = (unsigned long *) (datapos+(ntohl(hdr->reloc_start)-text_len)); @@ -582,7 +605,8 @@ static int load_flat_file(struct linux_binprm * bprm, textpos = (unsigned long) -ENOMEM; printk("Unable to allocate RAM for process text/data, errno %d\n", (int)-textpos); - return(textpos); + ret = textpos; + goto err_close; } realdatastart = textpos + ntohl(hdr->data_start); @@ -627,7 +651,8 @@ static int load_flat_file(struct linux_binprm * bprm, printk("Unable to read code+data+bss, errno %d\n",(int)-result); do_munmap(current->mm, textpos, text_len + data_len + extra + MAX_SHARED_LIBS * sizeof(unsigned long)); - return result; + ret = result; + goto err_close; } } @@ -690,8 +715,10 @@ static int load_flat_file(struct linux_binprm * bprm, unsigned long addr; if (*rp) { addr = calc_reloc(*rp, libinfo, id, 0); - if (addr == RELOC_FAILED) - return -ENOEXEC; + if (addr == RELOC_FAILED) { + ret = -ENOEXEC; + goto err_close; + } *rp = addr; } } @@ -718,8 +745,10 @@ static int load_flat_file(struct linux_binprm * bprm, relval = ntohl(reloc[i]); addr = flat_get_relocate_addr(relval); rp = (unsigned long *) calc_reloc(addr, libinfo, id, 1); - if (rp == (unsigned long *)RELOC_FAILED) - return -ENOEXEC; + if (rp == (unsigned long *)RELOC_FAILED) { + ret = -ENOEXEC; + goto err_close; + } /* Get the pointer's value. */ addr = flat_get_addr_from_rp(rp, relval, flags); @@ -731,8 +760,10 @@ static int load_flat_file(struct linux_binprm * bprm, if ((flags & FLAT_FLAG_GOTPIC) == 0) addr = ntohl(addr); addr = calc_reloc(addr, libinfo, id, 0); - if (addr == RELOC_FAILED) - return -ENOEXEC; + if (addr == RELOC_FAILED) { + ret = -ENOEXEC; + goto err_close; + } /* Write back the relocated pointer. */ flat_put_addr_at_rp(rp, addr, relval); @@ -752,6 +783,10 @@ static int load_flat_file(struct linux_binprm * bprm, stack_len); return 0; +err_close: + sys_close(exec_fileno); +err: + return ret; } diff --git a/fs/binfmt_misc.c b/fs/binfmt_misc.c index 6a7b730c206..d73d75591a3 100644 --- a/fs/binfmt_misc.c +++ b/fs/binfmt_misc.c @@ -600,7 +600,7 @@ static ssize_t bm_entry_write(struct file *file, const char __user *buffer, return count; } -static struct file_operations bm_entry_operations = { +static const struct file_operations bm_entry_operations = { .read = bm_entry_read, .write = bm_entry_write, }; @@ -668,7 +668,7 @@ out: return count; } -static struct file_operations bm_register_operations = { +static const struct file_operations bm_register_operations = { .write = bm_register_write, }; @@ -715,7 +715,7 @@ static ssize_t bm_status_write(struct file * file, const char __user * buffer, return count; } -static struct file_operations bm_status_operations = { +static const struct file_operations bm_status_operations = { .read = bm_status_read, .write = bm_status_write, }; @@ -30,7 +30,7 @@ #define BIO_POOL_SIZE 256 -static kmem_cache_t *bio_slab; +static kmem_cache_t *bio_slab __read_mostly; #define BIOVEC_NR_POOLS 6 @@ -39,7 +39,7 @@ static kmem_cache_t *bio_slab; * basically we just need to survive */ #define BIO_SPLIT_ENTRIES 8 -mempool_t *bio_split_pool; +mempool_t *bio_split_pool __read_mostly; struct biovec_slab { int nr_vecs; @@ -636,12 +636,10 @@ static struct bio *__bio_map_user_iov(request_queue_t *q, return ERR_PTR(-ENOMEM); ret = -ENOMEM; - pages = kmalloc(nr_pages * sizeof(struct page *), GFP_KERNEL); + pages = kcalloc(nr_pages, sizeof(struct page *), GFP_KERNEL); if (!pages) goto out; - memset(pages, 0, nr_pages * sizeof(struct page *)); - for (i = 0; i < iov_count; i++) { unsigned long uaddr = (unsigned long)iov[i].iov_base; unsigned long len = iov[i].iov_len; @@ -1127,16 +1125,6 @@ struct bio_pair *bio_split(struct bio *bi, mempool_t *pool, int first_sectors) return bp; } -static void *bio_pair_alloc(gfp_t gfp_flags, void *data) -{ - return kmalloc(sizeof(struct bio_pair), gfp_flags); -} - -static void bio_pair_free(void *bp, void *data) -{ - kfree(bp); -} - /* * create memory pools for biovec's in a bio_set. @@ -1153,8 +1141,7 @@ static int biovec_create_pools(struct bio_set *bs, int pool_entries, int scale) if (i >= scale) pool_entries >>= 1; - *bvp = mempool_create(pool_entries, mempool_alloc_slab, - mempool_free_slab, bp->slab); + *bvp = mempool_create_slab_pool(pool_entries, bp->slab); if (!*bvp) return -ENOMEM; } @@ -1186,15 +1173,12 @@ void bioset_free(struct bio_set *bs) struct bio_set *bioset_create(int bio_pool_size, int bvec_pool_size, int scale) { - struct bio_set *bs = kmalloc(sizeof(*bs), GFP_KERNEL); + struct bio_set *bs = kzalloc(sizeof(*bs), GFP_KERNEL); if (!bs) return NULL; - memset(bs, 0, sizeof(*bs)); - bs->bio_pool = mempool_create(bio_pool_size, mempool_alloc_slab, - mempool_free_slab, bio_slab); - + bs->bio_pool = mempool_create_slab_pool(bio_pool_size, bio_slab); if (!bs->bio_pool) goto bad; @@ -1257,8 +1241,8 @@ static int __init init_bio(void) if (!fs_bio_set) panic("bio: can't allocate bios\n"); - bio_split_pool = mempool_create(BIO_SPLIT_ENTRIES, - bio_pair_alloc, bio_pair_free, NULL); + bio_split_pool = mempool_create_kmalloc_pool(BIO_SPLIT_ENTRIES, + sizeof(struct bio_pair)); if (!bio_split_pool) panic("bio: can't create split pool\n"); diff --git a/fs/block_dev.c b/fs/block_dev.c index 573fc8e0b67..af88c43043d 100644 --- a/fs/block_dev.c +++ b/fs/block_dev.c @@ -131,9 +131,10 @@ blkdev_get_block(struct inode *inode, sector_t iblock, static int blkdev_get_blocks(struct inode *inode, sector_t iblock, - unsigned long max_blocks, struct buffer_head *bh, int create) + struct buffer_head *bh, int create) { sector_t end_block = max_block(I_BDEV(inode)); + unsigned long max_blocks = bh->b_size >> inode->i_blkbits; if ((iblock + max_blocks) > end_block) { max_blocks = end_block - iblock; @@ -234,7 +235,7 @@ static int block_fsync(struct file *filp, struct dentry *dentry, int datasync) */ static __cacheline_aligned_in_smp DEFINE_SPINLOCK(bdev_lock); -static kmem_cache_t * bdev_cachep; +static kmem_cache_t * bdev_cachep __read_mostly; static struct inode *bdev_alloc_inode(struct super_block *sb) { @@ -265,6 +266,9 @@ static void init_once(void * foo, kmem_cache_t * cachep, unsigned long flags) mutex_init(&bdev->bd_mount_mutex); INIT_LIST_HEAD(&bdev->bd_inodes); INIT_LIST_HEAD(&bdev->bd_list); +#ifdef CONFIG_SYSFS + INIT_LIST_HEAD(&bdev->bd_holder_list); +#endif inode_init_once(&ei->vfs_inode); } } @@ -308,7 +312,7 @@ static struct file_system_type bd_type = { .kill_sb = kill_anon_super, }; -static struct vfsmount *bd_mnt; +static struct vfsmount *bd_mnt __read_mostly; struct super_block *blockdev_superblock; void __init bdev_cache_init(void) @@ -489,6 +493,300 @@ void bd_release(struct block_device *bdev) EXPORT_SYMBOL(bd_release); +#ifdef CONFIG_SYSFS +/* + * Functions for bd_claim_by_kobject / bd_release_from_kobject + * + * If a kobject is passed to bd_claim_by_kobject() + * and the kobject has a parent directory, + * following symlinks are created: + * o from the kobject to the claimed bdev + * o from "holders" directory of the bdev to the parent of the kobject + * bd_release_from_kobject() removes these symlinks. + * + * Example: + * If /dev/dm-0 maps to /dev/sda, kobject corresponding to + * /sys/block/dm-0/slaves is passed to bd_claim_by_kobject(), then: + * /sys/block/dm-0/slaves/sda --> /sys/block/sda + * /sys/block/sda/holders/dm-0 --> /sys/block/dm-0 + */ + +static struct kobject *bdev_get_kobj(struct block_device *bdev) +{ + if (bdev->bd_contains != bdev) + return kobject_get(&bdev->bd_part->kobj); + else + return kobject_get(&bdev->bd_disk->kobj); +} + +static struct kobject *bdev_get_holder(struct block_device *bdev) +{ + if (bdev->bd_contains != bdev) + return kobject_get(bdev->bd_part->holder_dir); + else + return kobject_get(bdev->bd_disk->holder_dir); +} + +static void add_symlink(struct kobject *from, struct kobject *to) +{ + if (!from || !to) + return; + sysfs_create_link(from, to, kobject_name(to)); +} + +static void del_symlink(struct kobject *from, struct kobject *to) +{ + if (!from || !to) + return; + sysfs_remove_link(from, kobject_name(to)); +} + +/* + * 'struct bd_holder' contains pointers to kobjects symlinked by + * bd_claim_by_kobject. + * It's connected to bd_holder_list which is protected by bdev->bd_sem. + */ +struct bd_holder { + struct list_head list; /* chain of holders of the bdev */ + int count; /* references from the holder */ + struct kobject *sdir; /* holder object, e.g. "/block/dm-0/slaves" */ + struct kobject *hdev; /* e.g. "/block/dm-0" */ + struct kobject *hdir; /* e.g. "/block/sda/holders" */ + struct kobject *sdev; /* e.g. "/block/sda" */ +}; + +/* + * Get references of related kobjects at once. + * Returns 1 on success. 0 on failure. + * + * Should call bd_holder_release_dirs() after successful use. + */ +static int bd_holder_grab_dirs(struct block_device *bdev, + struct bd_holder *bo) +{ + if (!bdev || !bo) + return 0; + + bo->sdir = kobject_get(bo->sdir); + if (!bo->sdir) + return 0; + + bo->hdev = kobject_get(bo->sdir->parent); + if (!bo->hdev) + goto fail_put_sdir; + + bo->sdev = bdev_get_kobj(bdev); + if (!bo->sdev) + goto fail_put_hdev; + + bo->hdir = bdev_get_holder(bdev); + if (!bo->hdir) + goto fail_put_sdev; + + return 1; + +fail_put_sdev: + kobject_put(bo->sdev); +fail_put_hdev: + kobject_put(bo->hdev); +fail_put_sdir: + kobject_put(bo->sdir); + + return 0; +} + +/* Put references of related kobjects at once. */ +static void bd_holder_release_dirs(struct bd_holder *bo) +{ + kobject_put(bo->hdir); + kobject_put(bo->sdev); + kobject_put(bo->hdev); + kobject_put(bo->sdir); +} + +static struct bd_holder *alloc_bd_holder(struct kobject *kobj) +{ + struct bd_holder *bo; + + bo = kzalloc(sizeof(*bo), GFP_KERNEL); + if (!bo) + return NULL; + + bo->count = 1; + bo->sdir = kobj; + + return bo; +} + +static void free_bd_holder(struct bd_holder *bo) +{ + kfree(bo); +} + +/** + * add_bd_holder - create sysfs symlinks for bd_claim() relationship + * + * @bdev: block device to be bd_claimed + * @bo: preallocated and initialized by alloc_bd_holder() + * + * If there is no matching entry with @bo in @bdev->bd_holder_list, + * add @bo to the list, create symlinks. + * + * Returns 1 if @bo was added to the list. + * Returns 0 if @bo wasn't used by any reason and should be freed. + */ +static int add_bd_holder(struct block_device *bdev, struct bd_holder *bo) +{ + struct bd_holder *tmp; + + if (!bo) + return 0; + + list_for_each_entry(tmp, &bdev->bd_holder_list, list) { + if (tmp->sdir == bo->sdir) { + tmp->count++; + return 0; + } + } + + if (!bd_holder_grab_dirs(bdev, bo)) + return 0; + + add_symlink(bo->sdir, bo->sdev); + add_symlink(bo->hdir, bo->hdev); + list_add_tail(&bo->list, &bdev->bd_holder_list); + return 1; +} + +/** + * del_bd_holder - delete sysfs symlinks for bd_claim() relationship + * + * @bdev: block device to be bd_claimed + * @kobj: holder's kobject + * + * If there is matching entry with @kobj in @bdev->bd_holder_list + * and no other bd_claim() from the same kobject, + * remove the struct bd_holder from the list, delete symlinks for it. + * + * Returns a pointer to the struct bd_holder when it's removed from the list + * and ready to be freed. + * Returns NULL if matching claim isn't found or there is other bd_claim() + * by the same kobject. + */ +static struct bd_holder *del_bd_holder(struct block_device *bdev, + struct kobject *kobj) +{ + struct bd_holder *bo; + + list_for_each_entry(bo, &bdev->bd_holder_list, list) { + if (bo->sdir == kobj) { + bo->count--; + BUG_ON(bo->count < 0); + if (!bo->count) { + list_del(&bo->list); + del_symlink(bo->sdir, bo->sdev); + del_symlink(bo->hdir, bo->hdev); + bd_holder_release_dirs(bo); + return bo; + } + break; + } + } + + return NULL; +} + +/** + * bd_claim_by_kobject - bd_claim() with additional kobject signature + * + * @bdev: block device to be claimed + * @holder: holder's signature + * @kobj: holder's kobject + * + * Do bd_claim() and if it succeeds, create sysfs symlinks between + * the bdev and the holder's kobject. + * Use bd_release_from_kobject() when relesing the claimed bdev. + * + * Returns 0 on success. (same as bd_claim()) + * Returns errno on failure. + */ +static int bd_claim_by_kobject(struct block_device *bdev, void *holder, + struct kobject *kobj) +{ + int res; + struct bd_holder *bo; + + if (!kobj) + return -EINVAL; + + bo = alloc_bd_holder(kobj); + if (!bo) + return -ENOMEM; + + mutex_lock(&bdev->bd_mutex); + res = bd_claim(bdev, holder); + if (res || !add_bd_holder(bdev, bo)) + free_bd_holder(bo); + mutex_unlock(&bdev->bd_mutex); + + return res; +} + +/** + * bd_release_from_kobject - bd_release() with additional kobject signature + * + * @bdev: block device to be released + * @kobj: holder's kobject + * + * Do bd_release() and remove sysfs symlinks created by bd_claim_by_kobject(). + */ +static void bd_release_from_kobject(struct block_device *bdev, + struct kobject *kobj) +{ + struct bd_holder *bo; + + if (!kobj) + return; + + mutex_lock(&bdev->bd_mutex); + bd_release(bdev); + if ((bo = del_bd_holder(bdev, kobj))) + free_bd_holder(bo); + mutex_unlock(&bdev->bd_mutex); +} + +/** + * bd_claim_by_disk - wrapper function for bd_claim_by_kobject() + * + * @bdev: block device to be claimed + * @holder: holder's signature + * @disk: holder's gendisk + * + * Call bd_claim_by_kobject() with getting @disk->slave_dir. + */ +int bd_claim_by_disk(struct block_device *bdev, void *holder, + struct gendisk *disk) +{ + return bd_claim_by_kobject(bdev, holder, kobject_get(disk->slave_dir)); +} +EXPORT_SYMBOL_GPL(bd_claim_by_disk); + +/** + * bd_release_from_disk - wrapper function for bd_release_from_kobject() + * + * @bdev: block device to be claimed + * @disk: holder's gendisk + * + * Call bd_release_from_kobject() and put @disk->slave_dir. + */ +void bd_release_from_disk(struct block_device *bdev, struct gendisk *disk) +{ + bd_release_from_kobject(bdev, disk->slave_dir); + kobject_put(disk->slave_dir); +} +EXPORT_SYMBOL_GPL(bd_release_from_disk); +#endif + /* * Tries to open block device by device number. Use it ONLY if you * really do not have anything better - i.e. when you are behind a @@ -789,7 +1087,7 @@ struct address_space_operations def_blk_aops = { .direct_IO = blkdev_direct_IO, }; -struct file_operations def_blk_fops = { +const struct file_operations def_blk_fops = { .open = blkdev_open, .release = blkdev_close, .llseek = block_llseek, diff --git a/fs/buffer.c b/fs/buffer.c index 6d77ce9f54e..23f1f3a6807 100644 --- a/fs/buffer.c +++ b/fs/buffer.c @@ -160,12 +160,7 @@ int sync_blockdev(struct block_device *bdev) } EXPORT_SYMBOL(sync_blockdev); -/* - * Write out and wait upon all dirty data associated with this - * superblock. Filesystem data as well as the underlying block - * device. Takes the superblock lock. - */ -int fsync_super(struct super_block *sb) +static void __fsync_super(struct super_block *sb) { sync_inodes_sb(sb, 0); DQUOT_SYNC(sb); @@ -177,7 +172,16 @@ int fsync_super(struct super_block *sb) sb->s_op->sync_fs(sb, 1); sync_blockdev(sb->s_bdev); sync_inodes_sb(sb, 1); +} +/* + * Write out and wait upon all dirty data associated with this + * superblock. Filesystem data as well as the underlying block + * device. Takes the superblock lock. + */ +int fsync_super(struct super_block *sb) +{ + __fsync_super(sb); return sync_blockdev(sb->s_bdev); } @@ -216,19 +220,7 @@ struct super_block *freeze_bdev(struct block_device *bdev) sb->s_frozen = SB_FREEZE_WRITE; smp_wmb(); - sync_inodes_sb(sb, 0); - DQUOT_SYNC(sb); - - lock_super(sb); - if (sb->s_dirt && sb->s_op->write_super) - sb->s_op->write_super(sb); - unlock_super(sb); - - if (sb->s_op->sync_fs) - sb->s_op->sync_fs(sb, 1); - - sync_blockdev(sb->s_bdev); - sync_inodes_sb(sb, 1); + __fsync_super(sb); sb->s_frozen = SB_FREEZE_TRANS; smp_wmb(); @@ -434,8 +426,10 @@ __find_get_block_slow(struct block_device *bdev, sector_t block) if (all_mapped) { printk("__find_get_block_slow() failed. " "block=%llu, b_blocknr=%llu\n", - (unsigned long long)block, (unsigned long long)bh->b_blocknr); - printk("b_state=0x%08lx, b_size=%u\n", bh->b_state, bh->b_size); + (unsigned long long)block, + (unsigned long long)bh->b_blocknr); + printk("b_state=0x%08lx, b_size=%zu\n", + bh->b_state, bh->b_size); printk("device blocksize: %d\n", 1 << bd_inode->i_blkbits); } out_unlock: @@ -499,7 +493,7 @@ static void free_more_memory(void) wakeup_pdflush(1024); yield(); - for_each_pgdat(pgdat) { + for_each_online_pgdat(pgdat) { zones = pgdat->node_zonelists[gfp_zone(GFP_NOFS)].zones; if (*zones) try_to_free_pages(zones, GFP_NOFS); @@ -804,8 +798,7 @@ void mark_buffer_dirty_inode(struct buffer_head *bh, struct inode *inode) if (!mapping->assoc_mapping) { mapping->assoc_mapping = buffer_mapping; } else { - if (mapping->assoc_mapping != buffer_mapping) - BUG(); + BUG_ON(mapping->assoc_mapping != buffer_mapping); } if (list_empty(&bh->b_assoc_buffers)) { spin_lock(&buffer_mapping->private_lock); @@ -1122,8 +1115,7 @@ grow_dev_page(struct block_device *bdev, sector_t block, if (!page) return NULL; - if (!PageLocked(page)) - BUG(); + BUG_ON(!PageLocked(page)); if (page_has_buffers(page)) { bh = page_buffers(page); @@ -1530,8 +1522,7 @@ void set_bh_page(struct buffer_head *bh, struct page *page, unsigned long offset) { bh->b_page = page; - if (offset >= PAGE_SIZE) - BUG(); + BUG_ON(offset >= PAGE_SIZE); if (PageHighMem(page)) /* * This catches illegal uses and preserves the offset: @@ -1601,11 +1592,10 @@ EXPORT_SYMBOL(try_to_release_page); * point. Because the caller is about to free (and possibly reuse) those * blocks on-disk. */ -int block_invalidatepage(struct page *page, unsigned long offset) +void block_invalidatepage(struct page *page, unsigned long offset) { struct buffer_head *head, *bh, *next; unsigned int curr_off = 0; - int ret = 1; BUG_ON(!PageLocked(page)); if (!page_has_buffers(page)) @@ -1632,19 +1622,18 @@ int block_invalidatepage(struct page *page, unsigned long offset) * so real IO is not possible anymore. */ if (offset == 0) - ret = try_to_release_page(page, 0); + try_to_release_page(page, 0); out: - return ret; + return; } EXPORT_SYMBOL(block_invalidatepage); -int do_invalidatepage(struct page *page, unsigned long offset) +void do_invalidatepage(struct page *page, unsigned long offset) { - int (*invalidatepage)(struct page *, unsigned long); - invalidatepage = page->mapping->a_ops->invalidatepage; - if (invalidatepage == NULL) - invalidatepage = block_invalidatepage; - return (*invalidatepage)(page, offset); + void (*invalidatepage)(struct page *, unsigned long); + invalidatepage = page->mapping->a_ops->invalidatepage ? : + block_invalidatepage; + (*invalidatepage)(page, offset); } /* @@ -1746,6 +1735,7 @@ static int __block_write_full_page(struct inode *inode, struct page *page, sector_t block; sector_t last_block; struct buffer_head *bh, *head; + const unsigned blocksize = 1 << inode->i_blkbits; int nr_underway = 0; BUG_ON(!PageLocked(page)); @@ -1753,7 +1743,7 @@ static int __block_write_full_page(struct inode *inode, struct page *page, last_block = (i_size_read(inode) - 1) >> inode->i_blkbits; if (!page_has_buffers(page)) { - create_empty_buffers(page, 1 << inode->i_blkbits, + create_empty_buffers(page, blocksize, (1 << BH_Dirty)|(1 << BH_Uptodate)); } @@ -1788,6 +1778,7 @@ static int __block_write_full_page(struct inode *inode, struct page *page, clear_buffer_dirty(bh); set_buffer_uptodate(bh); } else if (!buffer_mapped(bh) && buffer_dirty(bh)) { + WARN_ON(bh->b_size != blocksize); err = get_block(inode, block, bh, 1); if (err) goto recover; @@ -1941,6 +1932,7 @@ static int __block_prepare_write(struct inode *inode, struct page *page, if (buffer_new(bh)) clear_buffer_new(bh); if (!buffer_mapped(bh)) { + WARN_ON(bh->b_size != blocksize); err = get_block(inode, block, bh, 1); if (err) break; @@ -2096,6 +2088,7 @@ int block_read_full_page(struct page *page, get_block_t *get_block) fully_mapped = 0; if (iblock < lblock) { + WARN_ON(bh->b_size != blocksize); err = get_block(inode, iblock, bh, 0); if (err) SetPageError(page); @@ -2417,6 +2410,7 @@ int nobh_prepare_write(struct page *page, unsigned from, unsigned to, create = 1; if (block_start >= to) create = 0; + map_bh.b_size = blocksize; ret = get_block(inode, block_in_file + block_in_page, &map_bh, create); if (ret) @@ -2677,6 +2671,7 @@ int block_truncate_page(struct address_space *mapping, err = 0; if (!buffer_mapped(bh)) { + WARN_ON(bh->b_size != blocksize); err = get_block(inode, iblock, bh, 0); if (err) goto unlock; @@ -2763,6 +2758,7 @@ sector_t generic_block_bmap(struct address_space *mapping, sector_t block, struct inode *inode = mapping->host; tmp.b_state = 0; tmp.b_blocknr = 0; + tmp.b_size = 1 << inode->i_blkbits; get_block(inode, block, &tmp, 0); return tmp.b_blocknr; } @@ -3015,7 +3011,7 @@ out: } EXPORT_SYMBOL(try_to_free_buffers); -int block_sync_page(struct page *page) +void block_sync_page(struct page *page) { struct address_space *mapping; @@ -3023,7 +3019,6 @@ int block_sync_page(struct page *page) mapping = page_mapping(page); if (mapping) blk_run_backing_dev(mapping->backing_dev_info, page); - return 0; } /* diff --git a/fs/char_dev.c b/fs/char_dev.c index 5c36345c9bf..4e1b849f912 100644 --- a/fs/char_dev.c +++ b/fs/char_dev.c @@ -146,12 +146,10 @@ __register_chrdev_region(unsigned int major, unsigned int baseminor, int ret = 0; int i; - cd = kmalloc(sizeof(struct char_device_struct), GFP_KERNEL); + cd = kzalloc(sizeof(struct char_device_struct), GFP_KERNEL); if (cd == NULL) return ERR_PTR(-ENOMEM); - memset(cd, 0, sizeof(struct char_device_struct)); - mutex_lock(&chrdevs_lock); /* temporary */ @@ -252,7 +250,7 @@ int alloc_chrdev_region(dev_t *dev, unsigned baseminor, unsigned count, } int register_chrdev(unsigned int major, const char *name, - struct file_operations *fops) + const struct file_operations *fops) { struct char_device_struct *cd; struct cdev *cdev; @@ -408,7 +406,7 @@ static void cdev_purge(struct cdev *cdev) * is contain the open that then fills in the correct operations * depending on the special file... */ -struct file_operations def_chr_fops = { +const struct file_operations def_chr_fops = { .open = chrdev_open, }; @@ -466,9 +464,8 @@ static struct kobj_type ktype_cdev_dynamic = { struct cdev *cdev_alloc(void) { - struct cdev *p = kmalloc(sizeof(struct cdev), GFP_KERNEL); + struct cdev *p = kzalloc(sizeof(struct cdev), GFP_KERNEL); if (p) { - memset(p, 0, sizeof(struct cdev)); p->kobj.ktype = &ktype_cdev_dynamic; INIT_LIST_HEAD(&p->list); kobject_init(&p->kobj); @@ -476,7 +473,7 @@ struct cdev *cdev_alloc(void) return p; } -void cdev_init(struct cdev *cdev, struct file_operations *fops) +void cdev_init(struct cdev *cdev, const struct file_operations *fops) { memset(cdev, 0, sizeof *cdev); INIT_LIST_HEAD(&cdev->list); diff --git a/fs/cifs/cifsfs.c b/fs/cifs/cifsfs.c index 221b3334b73..4bbc544857b 100644 --- a/fs/cifs/cifsfs.c +++ b/fs/cifs/cifsfs.c @@ -583,7 +583,7 @@ struct inode_operations cifs_symlink_inode_ops = { #endif }; -struct file_operations cifs_file_ops = { +const struct file_operations cifs_file_ops = { .read = do_sync_read, .write = do_sync_write, .readv = generic_file_readv, @@ -607,7 +607,7 @@ struct file_operations cifs_file_ops = { #endif /* CONFIG_CIFS_EXPERIMENTAL */ }; -struct file_operations cifs_file_direct_ops = { +const struct file_operations cifs_file_direct_ops = { /* no mmap, no aio, no readv - BB reevaluate whether they can be done with directio, no cache */ .read = cifs_user_read, @@ -626,7 +626,7 @@ struct file_operations cifs_file_direct_ops = { .dir_notify = cifs_dir_notify, #endif /* CONFIG_CIFS_EXPERIMENTAL */ }; -struct file_operations cifs_file_nobrl_ops = { +const struct file_operations cifs_file_nobrl_ops = { .read = do_sync_read, .write = do_sync_write, .readv = generic_file_readv, @@ -649,7 +649,7 @@ struct file_operations cifs_file_nobrl_ops = { #endif /* CONFIG_CIFS_EXPERIMENTAL */ }; -struct file_operations cifs_file_direct_nobrl_ops = { +const struct file_operations cifs_file_direct_nobrl_ops = { /* no mmap, no aio, no readv - BB reevaluate whether they can be done with directio, no cache */ .read = cifs_user_read, @@ -668,7 +668,7 @@ struct file_operations cifs_file_direct_nobrl_ops = { #endif /* CONFIG_CIFS_EXPERIMENTAL */ }; -struct file_operations cifs_dir_ops = { +const struct file_operations cifs_dir_ops = { .readdir = cifs_readdir, .release = cifs_closedir, .read = generic_read_dir, @@ -738,10 +738,8 @@ cifs_init_request_bufs(void) cERROR(1,("cifs_min_rcv set to maximum (64)")); } - cifs_req_poolp = mempool_create(cifs_min_rcv, - mempool_alloc_slab, - mempool_free_slab, - cifs_req_cachep); + cifs_req_poolp = mempool_create_slab_pool(cifs_min_rcv, + cifs_req_cachep); if(cifs_req_poolp == NULL) { kmem_cache_destroy(cifs_req_cachep); @@ -771,10 +769,8 @@ cifs_init_request_bufs(void) cFYI(1,("cifs_min_small set to maximum (256)")); } - cifs_sm_req_poolp = mempool_create(cifs_min_small, - mempool_alloc_slab, - mempool_free_slab, - cifs_sm_req_cachep); + cifs_sm_req_poolp = mempool_create_slab_pool(cifs_min_small, + cifs_sm_req_cachep); if(cifs_sm_req_poolp == NULL) { mempool_destroy(cifs_req_poolp); @@ -808,10 +804,8 @@ cifs_init_mids(void) if (cifs_mid_cachep == NULL) return -ENOMEM; - cifs_mid_poolp = mempool_create(3 /* a reasonable min simultan opers */, - mempool_alloc_slab, - mempool_free_slab, - cifs_mid_cachep); + /* 3 is a reasonable minimum number of simultaneous operations */ + cifs_mid_poolp = mempool_create_slab_pool(3, cifs_mid_cachep); if(cifs_mid_poolp == NULL) { kmem_cache_destroy(cifs_mid_cachep); return -ENOMEM; diff --git a/fs/cifs/cifsfs.h b/fs/cifs/cifsfs.h index 821a8eb2255..74f405ae4da 100644 --- a/fs/cifs/cifsfs.h +++ b/fs/cifs/cifsfs.h @@ -61,10 +61,10 @@ extern struct inode_operations cifs_file_inode_ops; extern struct inode_operations cifs_symlink_inode_ops; /* Functions related to files and directories */ -extern struct file_operations cifs_file_ops; -extern struct file_operations cifs_file_direct_ops; /* if directio mount */ -extern struct file_operations cifs_file_nobrl_ops; -extern struct file_operations cifs_file_direct_nobrl_ops; /* if directio mount */ +extern const struct file_operations cifs_file_ops; +extern const struct file_operations cifs_file_direct_ops; /* if directio mount */ +extern const struct file_operations cifs_file_nobrl_ops; +extern const struct file_operations cifs_file_direct_nobrl_ops; /* if directio mount */ extern int cifs_open(struct inode *inode, struct file *file); extern int cifs_close(struct inode *inode, struct file *file); extern int cifs_closedir(struct inode *inode, struct file *file); @@ -76,7 +76,7 @@ extern int cifs_lock(struct file *, int, struct file_lock *); extern int cifs_fsync(struct file *, struct dentry *, int); extern int cifs_flush(struct file *); extern int cifs_file_mmap(struct file * , struct vm_area_struct *); -extern struct file_operations cifs_dir_ops; +extern const struct file_operations cifs_dir_ops; extern int cifs_dir_open(struct inode *inode, struct file *file); extern int cifs_readdir(struct file *file, void *direntry, filldir_t filldir); extern int cifs_dir_notify(struct file *, unsigned long arg); diff --git a/fs/cifs/file.c b/fs/cifs/file.c index 165d6742638..fb49aef1f2e 100644 --- a/fs/cifs/file.c +++ b/fs/cifs/file.c @@ -1339,7 +1339,7 @@ int cifs_fsync(struct file *file, struct dentry *dentry, int datasync) return rc; } -/* static int cifs_sync_page(struct page *page) +/* static void cifs_sync_page(struct page *page) { struct address_space *mapping; struct inode *inode; @@ -1353,16 +1353,18 @@ int cifs_fsync(struct file *file, struct dentry *dentry, int datasync) return 0; inode = mapping->host; if (!inode) - return 0; */ + return; */ /* fill in rpages then result = cifs_pagein_inode(inode, index, rpages); */ /* BB finish */ /* cFYI(1, ("rpages is %d for sync page of Index %ld ", rpages, index)); +#if 0 if (rc < 0) return rc; return 0; +#endif } */ /* diff --git a/fs/cifs/inode.c b/fs/cifs/inode.c index ff93a9f81d1..598eec9778f 100644 --- a/fs/cifs/inode.c +++ b/fs/cifs/inode.c @@ -163,9 +163,9 @@ int cifs_get_inode_info_unix(struct inode **pinode, if (num_of_bytes < end_of_file) cFYI(1, ("allocation size less than end of file")); - cFYI(1, - ("Size %ld and blocks %ld", - (unsigned long) inode->i_size, inode->i_blocks)); + cFYI(1, ("Size %ld and blocks %llu", + (unsigned long) inode->i_size, + (unsigned long long)inode->i_blocks)); if (S_ISREG(inode->i_mode)) { cFYI(1, ("File inode")); inode->i_op = &cifs_file_inode_ops; diff --git a/fs/cifs/readdir.c b/fs/cifs/readdir.c index edb3b6eb34b..488bd0d81dc 100644 --- a/fs/cifs/readdir.c +++ b/fs/cifs/readdir.c @@ -197,10 +197,10 @@ static void fill_in_inode(struct inode *tmp_inode, if (allocation_size < end_of_file) cFYI(1, ("May be sparse file, allocation less than file size")); - cFYI(1, - ("File Size %ld and blocks %ld and blocksize %ld", - (unsigned long)tmp_inode->i_size, tmp_inode->i_blocks, - tmp_inode->i_blksize)); + cFYI(1, ("File Size %ld and blocks %llu and blocksize %ld", + (unsigned long)tmp_inode->i_size, + (unsigned long long)tmp_inode->i_blocks, + tmp_inode->i_blksize)); if (S_ISREG(tmp_inode->i_mode)) { cFYI(1, ("File inode")); tmp_inode->i_op = &cifs_file_inode_ops; diff --git a/fs/coda/cache.c b/fs/coda/cache.c index c607d923350..5d052713326 100644 --- a/fs/coda/cache.c +++ b/fs/coda/cache.c @@ -51,7 +51,7 @@ void coda_cache_clear_all(struct super_block *sb) struct coda_sb_info *sbi; sbi = coda_sbp(sb); - if (!sbi) BUG(); + BUG_ON(!sbi); atomic_inc(&permission_epoch); } diff --git a/fs/coda/cnode.c b/fs/coda/cnode.c index 23aeef5aa81..4c9fecbfa91 100644 --- a/fs/coda/cnode.c +++ b/fs/coda/cnode.c @@ -120,8 +120,7 @@ void coda_replace_fid(struct inode *inode, struct CodaFid *oldfid, cii = ITOC(inode); - if (!coda_fideq(&cii->c_fid, oldfid)) - BUG(); + BUG_ON(!coda_fideq(&cii->c_fid, oldfid)); /* replace fid and rehash inode */ /* XXX we probably need to hold some lock here! */ diff --git a/fs/coda/dir.c b/fs/coda/dir.c index 54f76de8a68..71f2ea632e5 100644 --- a/fs/coda/dir.c +++ b/fs/coda/dir.c @@ -82,7 +82,7 @@ struct inode_operations coda_dir_inode_operations = .setattr = coda_setattr, }; -struct file_operations coda_dir_operations = { +const struct file_operations coda_dir_operations = { .llseek = generic_file_llseek, .read = generic_read_dir, .readdir = coda_readdir, diff --git a/fs/coda/file.c b/fs/coda/file.c index 146a991d6eb..7c2642431fa 100644 --- a/fs/coda/file.c +++ b/fs/coda/file.c @@ -288,7 +288,7 @@ int coda_fsync(struct file *coda_file, struct dentry *coda_dentry, int datasync) return err; } -struct file_operations coda_file_operations = { +const struct file_operations coda_file_operations = { .llseek = generic_file_llseek, .read = coda_file_read, .write = coda_file_write, diff --git a/fs/coda/pioctl.c b/fs/coda/pioctl.c index 127714936c6..214822be87b 100644 --- a/fs/coda/pioctl.c +++ b/fs/coda/pioctl.c @@ -36,7 +36,7 @@ struct inode_operations coda_ioctl_inode_operations = .setattr = coda_setattr, }; -struct file_operations coda_ioctl_operations = { +const struct file_operations coda_ioctl_operations = { .owner = THIS_MODULE, .ioctl = coda_pioctl, }; diff --git a/fs/coda/psdev.c b/fs/coda/psdev.c index 98c74fe2e13..6c6771db36d 100644 --- a/fs/coda/psdev.c +++ b/fs/coda/psdev.c @@ -342,7 +342,7 @@ static int coda_psdev_release(struct inode * inode, struct file * file) } -static struct file_operations coda_psdev_fops = { +static const struct file_operations coda_psdev_fops = { .owner = THIS_MODULE, .read = coda_psdev_read, .write = coda_psdev_write, diff --git a/fs/compat.c b/fs/compat.c index 2a88477330f..7f8e26ea427 100644 --- a/fs/compat.c +++ b/fs/compat.c @@ -1476,10 +1476,9 @@ int compat_do_execve(char * filename, int i; retval = -ENOMEM; - bprm = kmalloc(sizeof(*bprm), GFP_KERNEL); + bprm = kzalloc(sizeof(*bprm), GFP_KERNEL); if (!bprm) goto out_ret; - memset(bprm, 0, sizeof(*bprm)); file = open_exec(filename); retval = PTR_ERR(file); @@ -1640,15 +1639,6 @@ void compat_set_fd_set(unsigned long nr, compat_ulong_t __user *ufdset, * This is a virtual copy of sys_select from fs/select.c and probably * should be compared to it from time to time */ -static void *select_bits_alloc(int size) -{ - return kmalloc(6 * size, GFP_KERNEL); -} - -static void select_bits_free(void *bits, int size) -{ - kfree(bits); -} /* * We can actually return ERESTARTSYS instead of EINTR, but I'd @@ -1687,7 +1677,7 @@ int compat_core_sys_select(int n, compat_ulong_t __user *inp, */ ret = -ENOMEM; size = FDS_BYTES(n); - bits = select_bits_alloc(size); + bits = kmalloc(6 * size, GFP_KERNEL); if (!bits) goto out_nofds; fds.in = (unsigned long *) bits; @@ -1721,7 +1711,7 @@ int compat_core_sys_select(int n, compat_ulong_t __user *inp, compat_set_fd_set(n, exp, fds.res_ex); out: - select_bits_free(bits, size); + kfree(bits); out_nofds: return ret; } @@ -2170,9 +2160,12 @@ asmlinkage long compat_sys_nfsservctl(int cmd, struct compat_nfsctl_arg __user * default: err = -EINVAL; - goto done; + break; } + if (err) + goto done; + oldfs = get_fs(); set_fs(KERNEL_DS); /* The __user pointer casts are valid because of the set_fs() */ diff --git a/fs/configfs/configfs_internal.h b/fs/configfs/configfs_internal.h index f70e46951b3..3f4ff7a242b 100644 --- a/fs/configfs/configfs_internal.h +++ b/fs/configfs/configfs_internal.h @@ -72,9 +72,9 @@ extern void configfs_release_fs(void); extern struct rw_semaphore configfs_rename_sem; extern struct super_block * configfs_sb; -extern struct file_operations configfs_dir_operations; -extern struct file_operations configfs_file_operations; -extern struct file_operations bin_fops; +extern const struct file_operations configfs_dir_operations; +extern const struct file_operations configfs_file_operations; +extern const struct file_operations bin_fops; extern struct inode_operations configfs_dir_inode_operations; extern struct inode_operations configfs_symlink_inode_operations; diff --git a/fs/configfs/dir.c b/fs/configfs/dir.c index ca60e3abef4..8ed9b06a982 100644 --- a/fs/configfs/dir.c +++ b/fs/configfs/dir.c @@ -1027,7 +1027,7 @@ static loff_t configfs_dir_lseek(struct file * file, loff_t offset, int origin) return offset; } -struct file_operations configfs_dir_operations = { +const struct file_operations configfs_dir_operations = { .open = configfs_dir_open, .release = configfs_dir_close, .llseek = configfs_dir_lseek, diff --git a/fs/configfs/file.c b/fs/configfs/file.c index 3921920d871..f499803743e 100644 --- a/fs/configfs/file.c +++ b/fs/configfs/file.c @@ -322,7 +322,7 @@ static int configfs_release(struct inode * inode, struct file * filp) return 0; } -struct file_operations configfs_file_operations = { +const struct file_operations configfs_file_operations = { .read = configfs_read_file, .write = configfs_write_file, .llseek = generic_file_llseek, diff --git a/fs/cramfs/inode.c b/fs/cramfs/inode.c index 8ad52f5bf25..9efcc3a164e 100644 --- a/fs/cramfs/inode.c +++ b/fs/cramfs/inode.c @@ -22,16 +22,17 @@ #include <linux/cramfs_fs_sb.h> #include <linux/buffer_head.h> #include <linux/vfs.h> +#include <linux/mutex.h> #include <asm/semaphore.h> #include <asm/uaccess.h> static struct super_operations cramfs_ops; static struct inode_operations cramfs_dir_inode_operations; -static struct file_operations cramfs_directory_operations; +static const struct file_operations cramfs_directory_operations; static struct address_space_operations cramfs_aops; -static DECLARE_MUTEX(read_mutex); +static DEFINE_MUTEX(read_mutex); /* These two macros may change in future, to provide better st_ino @@ -250,20 +251,20 @@ static int cramfs_fill_super(struct super_block *sb, void *data, int silent) memset(sbi, 0, sizeof(struct cramfs_sb_info)); /* Invalidate the read buffers on mount: think disk change.. */ - down(&read_mutex); + mutex_lock(&read_mutex); for (i = 0; i < READ_BUFFERS; i++) buffer_blocknr[i] = -1; /* Read the first block and get the superblock from it */ memcpy(&super, cramfs_read(sb, 0, sizeof(super)), sizeof(super)); - up(&read_mutex); + mutex_unlock(&read_mutex); /* Do sanity checks on the superblock */ if (super.magic != CRAMFS_MAGIC) { /* check at 512 byte offset */ - down(&read_mutex); + mutex_lock(&read_mutex); memcpy(&super, cramfs_read(sb, 512, sizeof(super)), sizeof(super)); - up(&read_mutex); + mutex_unlock(&read_mutex); if (super.magic != CRAMFS_MAGIC) { if (!silent) printk(KERN_ERR "cramfs: wrong magic\n"); @@ -366,7 +367,7 @@ static int cramfs_readdir(struct file *filp, void *dirent, filldir_t filldir) mode_t mode; int namelen, error; - down(&read_mutex); + mutex_lock(&read_mutex); de = cramfs_read(sb, OFFSET(inode) + offset, sizeof(*de)+256); name = (char *)(de+1); @@ -379,7 +380,7 @@ static int cramfs_readdir(struct file *filp, void *dirent, filldir_t filldir) memcpy(buf, name, namelen); ino = CRAMINO(de); mode = de->mode; - up(&read_mutex); + mutex_unlock(&read_mutex); nextoffset = offset + sizeof(*de) + namelen; for (;;) { if (!namelen) { @@ -410,7 +411,7 @@ static struct dentry * cramfs_lookup(struct inode *dir, struct dentry *dentry, s unsigned int offset = 0; int sorted; - down(&read_mutex); + mutex_lock(&read_mutex); sorted = CRAMFS_SB(dir->i_sb)->flags & CRAMFS_FLAG_SORTED_DIRS; while (offset < dir->i_size) { struct cramfs_inode *de; @@ -433,7 +434,7 @@ static struct dentry * cramfs_lookup(struct inode *dir, struct dentry *dentry, s for (;;) { if (!namelen) { - up(&read_mutex); + mutex_unlock(&read_mutex); return ERR_PTR(-EIO); } if (name[namelen-1]) @@ -447,7 +448,7 @@ static struct dentry * cramfs_lookup(struct inode *dir, struct dentry *dentry, s continue; if (!retval) { struct cramfs_inode entry = *de; - up(&read_mutex); + mutex_unlock(&read_mutex); d_add(dentry, get_cramfs_inode(dir->i_sb, &entry)); return NULL; } @@ -455,7 +456,7 @@ static struct dentry * cramfs_lookup(struct inode *dir, struct dentry *dentry, s if (sorted) break; } - up(&read_mutex); + mutex_unlock(&read_mutex); d_add(dentry, NULL); return NULL; } @@ -474,21 +475,21 @@ static int cramfs_readpage(struct file *file, struct page * page) u32 start_offset, compr_len; start_offset = OFFSET(inode) + maxblock*4; - down(&read_mutex); + mutex_lock(&read_mutex); if (page->index) start_offset = *(u32 *) cramfs_read(sb, blkptr_offset-4, 4); compr_len = (*(u32 *) cramfs_read(sb, blkptr_offset, 4) - start_offset); - up(&read_mutex); + mutex_unlock(&read_mutex); pgdata = kmap(page); if (compr_len == 0) ; /* hole */ else { - down(&read_mutex); + mutex_lock(&read_mutex); bytes_filled = cramfs_uncompress_block(pgdata, PAGE_CACHE_SIZE, cramfs_read(sb, start_offset, compr_len), compr_len); - up(&read_mutex); + mutex_unlock(&read_mutex); } } else pgdata = kmap(page); @@ -511,7 +512,7 @@ static struct address_space_operations cramfs_aops = { /* * A directory can only readdir */ -static struct file_operations cramfs_directory_operations = { +static const struct file_operations cramfs_directory_operations = { .llseek = generic_file_llseek, .read = generic_read_dir, .readdir = cramfs_readdir, diff --git a/fs/dcache.c b/fs/dcache.c index 653f64ce98e..19458d39950 100644 --- a/fs/dcache.c +++ b/fs/dcache.c @@ -34,9 +34,8 @@ #include <linux/swap.h> #include <linux/bootmem.h> -/* #define DCACHE_DEBUG 1 */ -int sysctl_vfs_cache_pressure = 100; +int sysctl_vfs_cache_pressure __read_mostly = 100; EXPORT_SYMBOL_GPL(sysctl_vfs_cache_pressure); __cacheline_aligned_in_smp DEFINE_SPINLOCK(dcache_lock); @@ -44,7 +43,7 @@ static seqlock_t rename_lock __cacheline_aligned_in_smp = SEQLOCK_UNLOCKED; EXPORT_SYMBOL(dcache_lock); -static kmem_cache_t *dentry_cache; +static kmem_cache_t *dentry_cache __read_mostly; #define DNAME_INLINE_LEN (sizeof(struct dentry)-offsetof(struct dentry,d_iname)) @@ -59,9 +58,9 @@ static kmem_cache_t *dentry_cache; #define D_HASHBITS d_hash_shift #define D_HASHMASK d_hash_mask -static unsigned int d_hash_mask; -static unsigned int d_hash_shift; -static struct hlist_head *dentry_hashtable; +static unsigned int d_hash_mask __read_mostly; +static unsigned int d_hash_shift __read_mostly; +static struct hlist_head *dentry_hashtable __read_mostly; static LIST_HEAD(dentry_unused); /* Statistics gathering. */ @@ -325,10 +324,13 @@ static struct dentry * __d_find_alias(struct inode *inode, int want_discon) struct dentry * d_find_alias(struct inode *inode) { - struct dentry *de; - spin_lock(&dcache_lock); - de = __d_find_alias(inode, 0); - spin_unlock(&dcache_lock); + struct dentry *de = NULL; + + if (!list_empty(&inode->i_dentry)) { + spin_lock(&dcache_lock); + de = __d_find_alias(inode, 0); + spin_unlock(&dcache_lock); + } return de; } @@ -486,6 +488,7 @@ repeat: continue; } prune_one_dentry(dentry); + cond_resched_lock(&dcache_lock); goto repeat; } spin_unlock(&dcache_lock); @@ -599,10 +602,6 @@ resume: */ if (!list_empty(&dentry->d_subdirs)) { this_parent = dentry; -#ifdef DCACHE_DEBUG -printk(KERN_DEBUG "select_parent: descending to %s/%s, found=%d\n", -dentry->d_parent->d_name.name, dentry->d_name.name, found); -#endif goto repeat; } } @@ -612,10 +611,6 @@ dentry->d_parent->d_name.name, dentry->d_name.name, found); if (this_parent != parent) { next = this_parent->d_u.d_child.next; this_parent = this_parent->d_parent; -#ifdef DCACHE_DEBUG -printk(KERN_DEBUG "select_parent: ascending to %s/%s, found=%d\n", -this_parent->d_parent->d_name.name, this_parent->d_name.name, found); -#endif goto resume; } out: @@ -794,11 +789,12 @@ struct dentry *d_alloc_name(struct dentry *parent, const char *name) void d_instantiate(struct dentry *entry, struct inode * inode) { - if (!list_empty(&entry->d_alias)) BUG(); + BUG_ON(!list_empty(&entry->d_alias)); spin_lock(&dcache_lock); if (inode) list_add(&entry->d_alias, &inode->i_dentry); entry->d_inode = inode; + fsnotify_d_instantiate(entry, inode); spin_unlock(&dcache_lock); security_d_instantiate(entry, inode); } @@ -850,6 +846,7 @@ struct dentry *d_instantiate_unique(struct dentry *entry, struct inode *inode) list_add(&entry->d_alias, &inode->i_dentry); do_negative: entry->d_inode = inode; + fsnotify_d_instantiate(entry, inode); spin_unlock(&dcache_lock); security_d_instantiate(entry, inode); return NULL; @@ -980,6 +977,7 @@ struct dentry *d_splice_alias(struct inode *inode, struct dentry *dentry) new = __d_find_alias(inode, 1); if (new) { BUG_ON(!(new->d_flags & DCACHE_DISCONNECTED)); + fsnotify_d_instantiate(new, inode); spin_unlock(&dcache_lock); security_d_instantiate(new, inode); d_rehash(dentry); @@ -989,6 +987,7 @@ struct dentry *d_splice_alias(struct inode *inode, struct dentry *dentry) /* d_instantiate takes dcache_lock, so we do it by hand */ list_add(&dentry->d_alias, &inode->i_dentry); dentry->d_inode = inode; + fsnotify_d_instantiate(dentry, inode); spin_unlock(&dcache_lock); security_d_instantiate(dentry, inode); d_rehash(dentry); @@ -1173,6 +1172,9 @@ void d_delete(struct dentry * dentry) spin_lock(&dentry->d_lock); isdir = S_ISDIR(dentry->d_inode->i_mode); if (atomic_read(&dentry->d_count) == 1) { + /* remove this and other inotify debug checks after 2.6.18 */ + dentry->d_flags &= ~DCACHE_INOTIFY_PARENT_WATCHED; + dentry_iput(dentry); fsnotify_nameremove(dentry, isdir); return; @@ -1339,6 +1341,7 @@ already_unhashed: list_add(&dentry->d_u.d_child, &dentry->d_parent->d_subdirs); spin_unlock(&target->d_lock); + fsnotify_d_move(dentry); spin_unlock(&dentry->d_lock); write_sequnlock(&rename_lock); spin_unlock(&dcache_lock); @@ -1707,10 +1710,10 @@ static void __init dcache_init(unsigned long mempages) } /* SLAB cache for __getname() consumers */ -kmem_cache_t *names_cachep; +kmem_cache_t *names_cachep __read_mostly; /* SLAB cache for file structures */ -kmem_cache_t *filp_cachep; +kmem_cache_t *filp_cachep __read_mostly; EXPORT_SYMBOL(d_genocide); diff --git a/fs/dcookies.c b/fs/dcookies.c index f8274a8f83b..8749339bf4f 100644 --- a/fs/dcookies.c +++ b/fs/dcookies.c @@ -23,6 +23,7 @@ #include <linux/mm.h> #include <linux/errno.h> #include <linux/dcookies.h> +#include <linux/mutex.h> #include <asm/uaccess.h> /* The dcookies are allocated from a kmem_cache and @@ -36,10 +37,10 @@ struct dcookie_struct { }; static LIST_HEAD(dcookie_users); -static DECLARE_MUTEX(dcookie_sem); -static kmem_cache_t * dcookie_cache; -static struct list_head * dcookie_hashtable; -static size_t hash_size; +static DEFINE_MUTEX(dcookie_mutex); +static kmem_cache_t *dcookie_cache __read_mostly; +static struct list_head *dcookie_hashtable __read_mostly; +static size_t hash_size __read_mostly; static inline int is_live(void) { @@ -114,7 +115,7 @@ int get_dcookie(struct dentry * dentry, struct vfsmount * vfsmnt, int err = 0; struct dcookie_struct * dcs; - down(&dcookie_sem); + mutex_lock(&dcookie_mutex); if (!is_live()) { err = -EINVAL; @@ -134,7 +135,7 @@ int get_dcookie(struct dentry * dentry, struct vfsmount * vfsmnt, *cookie = dcookie_value(dcs); out: - up(&dcookie_sem); + mutex_unlock(&dcookie_mutex); return err; } @@ -157,7 +158,7 @@ asmlinkage long sys_lookup_dcookie(u64 cookie64, char __user * buf, size_t len) if (!capable(CAP_SYS_ADMIN)) return -EPERM; - down(&dcookie_sem); + mutex_lock(&dcookie_mutex); if (!is_live()) { err = -EINVAL; @@ -192,7 +193,7 @@ asmlinkage long sys_lookup_dcookie(u64 cookie64, char __user * buf, size_t len) out_free: kfree(kbuf); out: - up(&dcookie_sem); + mutex_unlock(&dcookie_mutex); return err; } @@ -290,7 +291,7 @@ struct dcookie_user * dcookie_register(void) { struct dcookie_user * user; - down(&dcookie_sem); + mutex_lock(&dcookie_mutex); user = kmalloc(sizeof(struct dcookie_user), GFP_KERNEL); if (!user) @@ -302,7 +303,7 @@ struct dcookie_user * dcookie_register(void) list_add(&user->next, &dcookie_users); out: - up(&dcookie_sem); + mutex_unlock(&dcookie_mutex); return user; out_free: kfree(user); @@ -313,7 +314,7 @@ out_free: void dcookie_unregister(struct dcookie_user * user) { - down(&dcookie_sem); + mutex_lock(&dcookie_mutex); list_del(&user->next); kfree(user); @@ -321,7 +322,7 @@ void dcookie_unregister(struct dcookie_user * user) if (!is_live()) dcookie_exit(); - up(&dcookie_sem); + mutex_unlock(&dcookie_mutex); } EXPORT_SYMBOL_GPL(dcookie_register); diff --git a/fs/debugfs/file.c b/fs/debugfs/file.c index 40c4fc973fa..66a505422e5 100644 --- a/fs/debugfs/file.c +++ b/fs/debugfs/file.c @@ -39,7 +39,7 @@ static int default_open(struct inode *inode, struct file *file) return 0; } -struct file_operations debugfs_file_operations = { +const struct file_operations debugfs_file_operations = { .read = default_read_file, .write = default_write_file, .open = default_open, @@ -213,7 +213,7 @@ static ssize_t write_file_bool(struct file *file, const char __user *user_buf, return count; } -static struct file_operations fops_bool = { +static const struct file_operations fops_bool = { .read = read_file_bool, .write = write_file_bool, .open = default_open, diff --git a/fs/debugfs/inode.c b/fs/debugfs/inode.c index d4f1a2cddd4..85d166cdcae 100644 --- a/fs/debugfs/inode.c +++ b/fs/debugfs/inode.c @@ -191,7 +191,7 @@ static int debugfs_create_by_name(const char *name, mode_t mode, */ struct dentry *debugfs_create_file(const char *name, mode_t mode, struct dentry *parent, void *data, - struct file_operations *fops) + const struct file_operations *fops) { struct dentry *dentry = NULL; int error; diff --git a/fs/devfs/base.c b/fs/devfs/base.c index b621521e09d..52f5059c4f3 100644 --- a/fs/devfs/base.c +++ b/fs/devfs/base.c @@ -856,14 +856,14 @@ static int devfsd_close(struct inode *inode, struct file *file); #ifdef CONFIG_DEVFS_DEBUG static ssize_t stat_read(struct file *file, char __user *buf, size_t len, loff_t * ppos); -static struct file_operations stat_fops = { +static const struct file_operations stat_fops = { .open = nonseekable_open, .read = stat_read, }; #endif /* Devfs daemon file operations */ -static struct file_operations devfsd_fops = { +static const struct file_operations devfsd_fops = { .open = nonseekable_open, .read = devfsd_read, .ioctl = devfsd_ioctl, @@ -1842,8 +1842,8 @@ static int try_modload(struct devfs_entry *parent, struct fs_info *fs_info, static struct inode_operations devfs_iops; static struct inode_operations devfs_dir_iops; -static struct file_operations devfs_fops; -static struct file_operations devfs_dir_fops; +static const struct file_operations devfs_fops; +static const struct file_operations devfs_dir_fops; static struct inode_operations devfs_symlink_iops; static int devfs_notify_change(struct dentry *dentry, struct iattr *iattr) @@ -2061,11 +2061,11 @@ static int devfs_open(struct inode *inode, struct file *file) return err; } /* End Function devfs_open */ -static struct file_operations devfs_fops = { +static const struct file_operations devfs_fops = { .open = devfs_open, }; -static struct file_operations devfs_dir_fops = { +static const struct file_operations devfs_dir_fops = { .read = generic_read_dir, .readdir = devfs_readdir, }; diff --git a/fs/direct-io.c b/fs/direct-io.c index 27f3e787fac..910a8ed74b5 100644 --- a/fs/direct-io.c +++ b/fs/direct-io.c @@ -86,12 +86,12 @@ struct dio { unsigned first_block_in_page; /* doesn't change, Used only once */ int boundary; /* prev block is at a boundary */ int reap_counter; /* rate limit reaping */ - get_blocks_t *get_blocks; /* block mapping function */ + get_block_t *get_block; /* block mapping function */ dio_iodone_t *end_io; /* IO completion function */ sector_t final_block_in_bio; /* current final block in bio + 1 */ sector_t next_block_for_io; /* next block to be put under IO, in dio_blocks units */ - struct buffer_head map_bh; /* last get_blocks() result */ + struct buffer_head map_bh; /* last get_block() result */ /* * Deferred addition of a page to the dio. These variables are @@ -129,6 +129,7 @@ struct dio { /* AIO related stuff */ struct kiocb *iocb; /* kiocb */ int is_async; /* is IO async ? */ + int io_error; /* IO error in completion path */ ssize_t result; /* IO result */ }; @@ -210,9 +211,9 @@ static struct page *dio_get_page(struct dio *dio) /* * Called when all DIO BIO I/O has been completed - let the filesystem - * know, if it registered an interest earlier via get_blocks. Pass the + * know, if it registered an interest earlier via get_block. Pass the * private field of the map buffer_head so that filesystems can use it - * to hold additional state between get_blocks calls and dio_complete. + * to hold additional state between get_block calls and dio_complete. */ static void dio_complete(struct dio *dio, loff_t offset, ssize_t bytes) { @@ -250,6 +251,10 @@ static void finished_one_bio(struct dio *dio) ((offset + transferred) > dio->i_size)) transferred = dio->i_size - offset; + /* check for error in completion path */ + if (dio->io_error) + transferred = dio->io_error; + dio_complete(dio, offset, transferred); /* Complete AIO later if falling back to buffered i/o */ @@ -406,7 +411,7 @@ static int dio_bio_complete(struct dio *dio, struct bio *bio) int page_no; if (!uptodate) - dio->result = -EIO; + dio->io_error = -EIO; if (dio->is_async && dio->rw == READ) { bio_check_pages_dirty(bio); /* transfers ownership */ @@ -488,7 +493,7 @@ static int dio_bio_reap(struct dio *dio) * The fs is allowed to map lots of blocks at once. If it wants to do that, * it uses the passed inode-relative block number as the file offset, as usual. * - * get_blocks() is passed the number of i_blkbits-sized blocks which direct_io + * get_block() is passed the number of i_blkbits-sized blocks which direct_io * has remaining to do. The fs should not map more than this number of blocks. * * If the fs has mapped a lot of blocks, it should populate bh->b_size to @@ -501,7 +506,7 @@ static int dio_bio_reap(struct dio *dio) * In the case of filesystem holes: the fs may return an arbitrarily-large * hole by returning an appropriate value in b_size and by clearing * buffer_mapped(). However the direct-io code will only process holes one - * block at a time - it will repeatedly call get_blocks() as it walks the hole. + * block at a time - it will repeatedly call get_block() as it walks the hole. */ static int get_more_blocks(struct dio *dio) { @@ -519,8 +524,6 @@ static int get_more_blocks(struct dio *dio) */ ret = dio->page_errors; if (ret == 0) { - map_bh->b_state = 0; - map_bh->b_size = 0; BUG_ON(dio->block_in_file >= dio->final_block_in_request); fs_startblk = dio->block_in_file >> dio->blkfactor; dio_count = dio->final_block_in_request - dio->block_in_file; @@ -529,6 +532,9 @@ static int get_more_blocks(struct dio *dio) if (dio_count & blkmask) fs_count++; + map_bh->b_state = 0; + map_bh->b_size = fs_count << dio->inode->i_blkbits; + create = dio->rw == WRITE; if (dio->lock_type == DIO_LOCKING) { if (dio->block_in_file < (i_size_read(dio->inode) >> @@ -537,13 +543,14 @@ static int get_more_blocks(struct dio *dio) } else if (dio->lock_type == DIO_NO_LOCKING) { create = 0; } + /* * For writes inside i_size we forbid block creations: only * overwrites are permitted. We fall back to buffered writes * at a higher level for inside-i_size block-instantiating * writes. */ - ret = (*dio->get_blocks)(dio->inode, fs_startblk, fs_count, + ret = (*dio->get_block)(dio->inode, fs_startblk, map_bh, create); } return ret; @@ -778,11 +785,11 @@ static void dio_zero_block(struct dio *dio, int end) * happily perform page-sized but 512-byte aligned IOs. It is important that * blockdev IO be able to have fine alignment and large sizes. * - * So what we do is to permit the ->get_blocks function to populate bh.b_size + * So what we do is to permit the ->get_block function to populate bh.b_size * with the size of IO which is permitted at this offset and this i_blkbits. * * For best results, the blockdev should be set up with 512-byte i_blkbits and - * it should set b_size to PAGE_SIZE or more inside get_blocks(). This gives + * it should set b_size to PAGE_SIZE or more inside get_block(). This gives * fine alignment but still allows this function to work in PAGE_SIZE units. */ static int do_direct_IO(struct dio *dio) @@ -942,7 +949,7 @@ out: static ssize_t direct_io_worker(int rw, struct kiocb *iocb, struct inode *inode, const struct iovec *iov, loff_t offset, unsigned long nr_segs, - unsigned blkbits, get_blocks_t get_blocks, dio_iodone_t end_io, + unsigned blkbits, get_block_t get_block, dio_iodone_t end_io, struct dio *dio) { unsigned long user_addr; @@ -964,13 +971,14 @@ direct_io_worker(int rw, struct kiocb *iocb, struct inode *inode, dio->boundary = 0; dio->reap_counter = 0; - dio->get_blocks = get_blocks; + dio->get_block = get_block; dio->end_io = end_io; dio->map_bh.b_private = NULL; dio->final_block_in_bio = -1; dio->next_block_for_io = -1; dio->page_errors = 0; + dio->io_error = 0; dio->result = 0; dio->iocb = iocb; dio->i_size = i_size_read(inode); @@ -1171,7 +1179,7 @@ direct_io_worker(int rw, struct kiocb *iocb, struct inode *inode, ssize_t __blockdev_direct_IO(int rw, struct kiocb *iocb, struct inode *inode, struct block_device *bdev, const struct iovec *iov, loff_t offset, - unsigned long nr_segs, get_blocks_t get_blocks, dio_iodone_t end_io, + unsigned long nr_segs, get_block_t get_block, dio_iodone_t end_io, int dio_lock_type) { int seg; @@ -1267,7 +1275,7 @@ __blockdev_direct_IO(int rw, struct kiocb *iocb, struct inode *inode, (end > i_size_read(inode))); retval = direct_io_worker(rw, iocb, inode, iov, offset, - nr_segs, blkbits, get_blocks, end_io, dio); + nr_segs, blkbits, get_block, end_io, dio); if (rw == READ && dio_lock_type == DIO_LOCKING) release_i_mutex = 0; diff --git a/fs/dnotify.c b/fs/dnotify.c index f3b540dd5d1..f932591df5a 100644 --- a/fs/dnotify.c +++ b/fs/dnotify.c @@ -21,9 +21,9 @@ #include <linux/spinlock.h> #include <linux/slab.h> -int dir_notify_enable = 1; +int dir_notify_enable __read_mostly = 1; -static kmem_cache_t *dn_cache; +static kmem_cache_t *dn_cache __read_mostly; static void redo_inode_mask(struct inode *inode) { diff --git a/fs/efs/dir.c b/fs/efs/dir.c index 777c614ff36..17f5b2d3c16 100644 --- a/fs/efs/dir.c +++ b/fs/efs/dir.c @@ -10,7 +10,7 @@ static int efs_readdir(struct file *, void *, filldir_t); -struct file_operations efs_dir_operations = { +const struct file_operations efs_dir_operations = { .read = generic_read_dir, .readdir = efs_readdir, }; diff --git a/fs/eventpoll.c b/fs/eventpoll.c index 1c2b16fda13..242fe1a66ce 100644 --- a/fs/eventpoll.c +++ b/fs/eventpoll.c @@ -281,16 +281,16 @@ static struct mutex epmutex; static struct poll_safewake psw; /* Slab cache used to allocate "struct epitem" */ -static kmem_cache_t *epi_cache; +static kmem_cache_t *epi_cache __read_mostly; /* Slab cache used to allocate "struct eppoll_entry" */ -static kmem_cache_t *pwq_cache; +static kmem_cache_t *pwq_cache __read_mostly; /* Virtual fs used to allocate inodes for eventpoll files */ -static struct vfsmount *eventpoll_mnt; +static struct vfsmount *eventpoll_mnt __read_mostly; /* File callbacks that implement the eventpoll file behaviour */ -static struct file_operations eventpoll_fops = { +static const struct file_operations eventpoll_fops = { .release = ep_eventpoll_close, .poll = ep_eventpoll_poll }; @@ -599,7 +599,7 @@ sys_epoll_ctl(int epfd, int op, int fd, struct epoll_event __user *event) switch (op) { case EPOLL_CTL_ADD: if (!epi) { - epds.events |= POLLERR | POLLHUP; + epds.events |= POLLERR | POLLHUP | POLLRDHUP; error = ep_insert(ep, &epds, tfile, fd); } else @@ -613,7 +613,7 @@ sys_epoll_ctl(int epfd, int op, int fd, struct epoll_event __user *event) break; case EPOLL_CTL_MOD: if (epi) { - epds.events |= POLLERR | POLLHUP; + epds.events |= POLLERR | POLLHUP | POLLRDHUP; error = ep_modify(ep, epi, &epds); } else error = -ENOENT; diff --git a/fs/exec.c b/fs/exec.c index 0b515ac5313..950ebd43cdc 100644 --- a/fs/exec.c +++ b/fs/exec.c @@ -127,7 +127,7 @@ asmlinkage long sys_uselib(const char __user * library) struct nameidata nd; int error; - error = __user_path_lookup_open(library, LOOKUP_FOLLOW, &nd, FMODE_READ); + error = __user_path_lookup_open(library, LOOKUP_FOLLOW, &nd, FMODE_READ|FMODE_EXEC); if (error) goto out; @@ -477,7 +477,7 @@ struct file *open_exec(const char *name) int err; struct file *file; - err = path_lookup_open(AT_FDCWD, name, LOOKUP_FOLLOW, &nd, FMODE_READ); + err = path_lookup_open(AT_FDCWD, name, LOOKUP_FOLLOW, &nd, FMODE_READ|FMODE_EXEC); file = ERR_PTR(err); if (!err) { @@ -616,6 +616,15 @@ static int de_thread(struct task_struct *tsk) kmem_cache_free(sighand_cachep, newsighand); return -EAGAIN; } + + /* + * child_reaper ignores SIGKILL, change it now. + * Reparenting needs write_lock on tasklist_lock, + * so it is safe to do it under read_lock. + */ + if (unlikely(current->group_leader == child_reaper)) + child_reaper = current; + zap_other_threads(current); read_unlock(&tasklist_lock); @@ -632,7 +641,7 @@ static int de_thread(struct task_struct *tsk) * synchronize with any firing (by calling del_timer_sync) * before we can safely let the old group leader die. */ - sig->real_timer.data = current; + sig->tsk = current; spin_unlock_irq(lock); if (hrtimer_cancel(&sig->real_timer)) hrtimer_restart(&sig->real_timer); @@ -699,22 +708,30 @@ static int de_thread(struct task_struct *tsk) remove_parent(current); remove_parent(leader); - switch_exec_pids(leader, current); + + /* Become a process group leader with the old leader's pid. + * Note: The old leader also uses thispid until release_task + * is called. Odd but simple and correct. + */ + detach_pid(current, PIDTYPE_PID); + current->pid = leader->pid; + attach_pid(current, PIDTYPE_PID, current->pid); + attach_pid(current, PIDTYPE_PGID, current->signal->pgrp); + attach_pid(current, PIDTYPE_SID, current->signal->session); + list_add_tail(¤t->tasks, &init_task.tasks); current->parent = current->real_parent = leader->real_parent; leader->parent = leader->real_parent = child_reaper; current->group_leader = current; leader->group_leader = leader; - add_parent(current, current->parent); - add_parent(leader, leader->parent); + add_parent(current); + add_parent(leader); if (ptrace) { current->ptrace = ptrace; __ptrace_link(current, parent); } - list_del(¤t->tasks); - list_add_tail(¤t->tasks, &init_task.tasks); current->exit_signal = SIGCHLD; BUG_ON(leader->exit_state != EXIT_ZOMBIE); @@ -751,7 +768,6 @@ no_thread_group: /* * Move our state over to newsighand and switch it in. */ - spin_lock_init(&newsighand->siglock); atomic_set(&newsighand->count, 1); memcpy(newsighand->action, oldsighand->action, sizeof(newsighand->action)); @@ -768,7 +784,7 @@ no_thread_group: write_unlock_irq(&tasklist_lock); if (atomic_dec_and_test(&oldsighand->count)) - sighand_free(oldsighand); + kmem_cache_free(sighand_cachep, oldsighand); } BUG_ON(!thread_group_leader(current)); @@ -1143,10 +1159,9 @@ int do_execve(char * filename, int i; retval = -ENOMEM; - bprm = kmalloc(sizeof(*bprm), GFP_KERNEL); + bprm = kzalloc(sizeof(*bprm), GFP_KERNEL); if (!bprm) goto out_ret; - memset(bprm, 0, sizeof(*bprm)); file = open_exec(filename); retval = PTR_ERR(file); diff --git a/fs/ext2/dir.c b/fs/ext2/dir.c index b3dbd716cd3..d672aa9f406 100644 --- a/fs/ext2/dir.c +++ b/fs/ext2/dir.c @@ -416,8 +416,7 @@ void ext2_set_link(struct inode *dir, struct ext2_dir_entry_2 *de, lock_page(page); err = page->mapping->a_ops->prepare_write(NULL, page, from, to); - if (err) - BUG(); + BUG_ON(err); de->inode = cpu_to_le32(inode->i_ino); ext2_set_de_type (de, inode); err = ext2_commit_chunk(page, from, to); @@ -554,8 +553,7 @@ int ext2_delete_entry (struct ext2_dir_entry_2 * dir, struct page * page ) from = (char*)pde - (char*)page_address(page); lock_page(page); err = mapping->a_ops->prepare_write(NULL, page, from, to); - if (err) - BUG(); + BUG_ON(err); if (pde) pde->rec_len = cpu_to_le16(to-from); dir->inode = 0; @@ -660,7 +658,7 @@ not_empty: return 0; } -struct file_operations ext2_dir_operations = { +const struct file_operations ext2_dir_operations = { .llseek = generic_file_llseek, .read = generic_read_dir, .readdir = ext2_readdir, diff --git a/fs/ext2/ext2.h b/fs/ext2/ext2.h index 11035ac7986..9f74a62be55 100644 --- a/fs/ext2/ext2.h +++ b/fs/ext2/ext2.h @@ -154,12 +154,12 @@ extern void ext2_write_super (struct super_block *); */ /* dir.c */ -extern struct file_operations ext2_dir_operations; +extern const struct file_operations ext2_dir_operations; /* file.c */ extern struct inode_operations ext2_file_inode_operations; -extern struct file_operations ext2_file_operations; -extern struct file_operations ext2_xip_file_operations; +extern const struct file_operations ext2_file_operations; +extern const struct file_operations ext2_xip_file_operations; /* inode.c */ extern struct address_space_operations ext2_aops; diff --git a/fs/ext2/file.c b/fs/ext2/file.c index a484412fc78..509cceca04d 100644 --- a/fs/ext2/file.c +++ b/fs/ext2/file.c @@ -39,7 +39,7 @@ static int ext2_release_file (struct inode * inode, struct file * filp) * We have mostly NULL's here: the current defaults are ok for * the ext2 filesystem. */ -struct file_operations ext2_file_operations = { +const struct file_operations ext2_file_operations = { .llseek = generic_file_llseek, .read = generic_file_read, .write = generic_file_write, @@ -56,7 +56,7 @@ struct file_operations ext2_file_operations = { }; #ifdef CONFIG_EXT2_FS_XIP -struct file_operations ext2_xip_file_operations = { +const struct file_operations ext2_xip_file_operations = { .llseek = generic_file_llseek, .read = xip_file_read, .write = xip_file_write, diff --git a/fs/ext2/inode.c b/fs/ext2/inode.c index a717837f272..04af9c45dce 100644 --- a/fs/ext2/inode.c +++ b/fs/ext2/inode.c @@ -667,18 +667,6 @@ static sector_t ext2_bmap(struct address_space *mapping, sector_t block) return generic_block_bmap(mapping,block,ext2_get_block); } -static int -ext2_get_blocks(struct inode *inode, sector_t iblock, unsigned long max_blocks, - struct buffer_head *bh_result, int create) -{ - int ret; - - ret = ext2_get_block(inode, iblock, bh_result, create); - if (ret == 0) - bh_result->b_size = (1 << inode->i_blkbits); - return ret; -} - static ssize_t ext2_direct_IO(int rw, struct kiocb *iocb, const struct iovec *iov, loff_t offset, unsigned long nr_segs) @@ -687,7 +675,7 @@ ext2_direct_IO(int rw, struct kiocb *iocb, const struct iovec *iov, struct inode *inode = file->f_mapping->host; return blockdev_direct_IO(rw, iocb, inode, inode->i_sb->s_bdev, iov, - offset, nr_segs, ext2_get_blocks, NULL); + offset, nr_segs, ext2_get_block, NULL); } static int diff --git a/fs/ext2/super.c b/fs/ext2/super.c index 268b73f5847..7e30bae174e 100644 --- a/fs/ext2/super.c +++ b/fs/ext2/super.c @@ -211,8 +211,6 @@ static int ext2_show_options(struct seq_file *seq, struct vfsmount *vfs) if (sbi->s_mount_opt & EXT2_MOUNT_GRPID) seq_puts(seq, ",grpid"); - else - seq_puts(seq, ",nogrpid"); #if defined(CONFIG_QUOTA) if (sbi->s_mount_opt & EXT2_MOUNT_USRQUOTA) diff --git a/fs/ext3/balloc.c b/fs/ext3/balloc.c index 46623f77666..77927d6938f 100644 --- a/fs/ext3/balloc.c +++ b/fs/ext3/balloc.c @@ -653,9 +653,11 @@ claim_block(spinlock_t *lock, int block, struct buffer_head *bh) */ static int ext3_try_to_allocate(struct super_block *sb, handle_t *handle, int group, - struct buffer_head *bitmap_bh, int goal, struct ext3_reserve_window *my_rsv) + struct buffer_head *bitmap_bh, int goal, + unsigned long *count, struct ext3_reserve_window *my_rsv) { int group_first_block, start, end; + unsigned long num = 0; /* we do allocation within the reservation window if we have a window */ if (my_rsv) { @@ -713,8 +715,18 @@ repeat: goto fail_access; goto repeat; } - return goal; + num++; + goal++; + while (num < *count && goal < end + && ext3_test_allocatable(goal, bitmap_bh) + && claim_block(sb_bgl_lock(EXT3_SB(sb), group), goal, bitmap_bh)) { + num++; + goal++; + } + *count = num; + return goal - num; fail_access: + *count = num; return -1; } @@ -999,6 +1011,31 @@ retry: goto retry; } +static void try_to_extend_reservation(struct ext3_reserve_window_node *my_rsv, + struct super_block *sb, int size) +{ + struct ext3_reserve_window_node *next_rsv; + struct rb_node *next; + spinlock_t *rsv_lock = &EXT3_SB(sb)->s_rsv_window_lock; + + if (!spin_trylock(rsv_lock)) + return; + + next = rb_next(&my_rsv->rsv_node); + + if (!next) + my_rsv->rsv_end += size; + else { + next_rsv = list_entry(next, struct ext3_reserve_window_node, rsv_node); + + if ((next_rsv->rsv_start - my_rsv->rsv_end - 1) >= size) + my_rsv->rsv_end += size; + else + my_rsv->rsv_end = next_rsv->rsv_start - 1; + } + spin_unlock(rsv_lock); +} + /* * This is the main function used to allocate a new block and its reservation * window. @@ -1024,11 +1061,12 @@ static int ext3_try_to_allocate_with_rsv(struct super_block *sb, handle_t *handle, unsigned int group, struct buffer_head *bitmap_bh, int goal, struct ext3_reserve_window_node * my_rsv, - int *errp) + unsigned long *count, int *errp) { unsigned long group_first_block; int ret = 0; int fatal; + unsigned long num = *count; *errp = 0; @@ -1051,7 +1089,8 @@ ext3_try_to_allocate_with_rsv(struct super_block *sb, handle_t *handle, * or last attempt to allocate a block with reservation turned on failed */ if (my_rsv == NULL ) { - ret = ext3_try_to_allocate(sb, handle, group, bitmap_bh, goal, NULL); + ret = ext3_try_to_allocate(sb, handle, group, bitmap_bh, + goal, count, NULL); goto out; } /* @@ -1081,6 +1120,8 @@ ext3_try_to_allocate_with_rsv(struct super_block *sb, handle_t *handle, while (1) { if (rsv_is_empty(&my_rsv->rsv_window) || (ret < 0) || !goal_in_my_reservation(&my_rsv->rsv_window, goal, group, sb)) { + if (my_rsv->rsv_goal_size < *count) + my_rsv->rsv_goal_size = *count; ret = alloc_new_reservation(my_rsv, goal, sb, group, bitmap_bh); if (ret < 0) @@ -1088,16 +1129,21 @@ ext3_try_to_allocate_with_rsv(struct super_block *sb, handle_t *handle, if (!goal_in_my_reservation(&my_rsv->rsv_window, goal, group, sb)) goal = -1; - } + } else if (goal > 0 && (my_rsv->rsv_end-goal+1) < *count) + try_to_extend_reservation(my_rsv, sb, + *count-my_rsv->rsv_end + goal - 1); + if ((my_rsv->rsv_start >= group_first_block + EXT3_BLOCKS_PER_GROUP(sb)) || (my_rsv->rsv_end < group_first_block)) BUG(); ret = ext3_try_to_allocate(sb, handle, group, bitmap_bh, goal, - &my_rsv->rsv_window); + &num, &my_rsv->rsv_window); if (ret >= 0) { - my_rsv->rsv_alloc_hit++; + my_rsv->rsv_alloc_hit += num; + *count = num; break; /* succeed */ } + num = *count; } out: if (ret >= 0) { @@ -1154,8 +1200,8 @@ int ext3_should_retry_alloc(struct super_block *sb, int *retries) * bitmap, and then for any free bit if that fails. * This function also updates quota and i_blocks field. */ -int ext3_new_block(handle_t *handle, struct inode *inode, - unsigned long goal, int *errp) +int ext3_new_blocks(handle_t *handle, struct inode *inode, + unsigned long goal, unsigned long *count, int *errp) { struct buffer_head *bitmap_bh = NULL; struct buffer_head *gdp_bh; @@ -1178,6 +1224,7 @@ int ext3_new_block(handle_t *handle, struct inode *inode, static int goal_hits, goal_attempts; #endif unsigned long ngroups; + unsigned long num = *count; *errp = -ENOSPC; sb = inode->i_sb; @@ -1189,7 +1236,7 @@ int ext3_new_block(handle_t *handle, struct inode *inode, /* * Check quota for allocation of this block. */ - if (DQUOT_ALLOC_BLOCK(inode, 1)) { + if (DQUOT_ALLOC_BLOCK(inode, num)) { *errp = -EDQUOT; return 0; } @@ -1244,7 +1291,7 @@ retry: if (!bitmap_bh) goto io_error; ret_block = ext3_try_to_allocate_with_rsv(sb, handle, group_no, - bitmap_bh, ret_block, my_rsv, &fatal); + bitmap_bh, ret_block, my_rsv, &num, &fatal); if (fatal) goto out; if (ret_block >= 0) @@ -1281,7 +1328,7 @@ retry: if (!bitmap_bh) goto io_error; ret_block = ext3_try_to_allocate_with_rsv(sb, handle, group_no, - bitmap_bh, -1, my_rsv, &fatal); + bitmap_bh, -1, my_rsv, &num, &fatal); if (fatal) goto out; if (ret_block >= 0) @@ -1316,13 +1363,15 @@ allocated: target_block = ret_block + group_no * EXT3_BLOCKS_PER_GROUP(sb) + le32_to_cpu(es->s_first_data_block); - if (target_block == le32_to_cpu(gdp->bg_block_bitmap) || - target_block == le32_to_cpu(gdp->bg_inode_bitmap) || + if (in_range(le32_to_cpu(gdp->bg_block_bitmap), target_block, num) || + in_range(le32_to_cpu(gdp->bg_inode_bitmap), target_block, num) || in_range(target_block, le32_to_cpu(gdp->bg_inode_table), + EXT3_SB(sb)->s_itb_per_group) || + in_range(target_block + num - 1, le32_to_cpu(gdp->bg_inode_table), EXT3_SB(sb)->s_itb_per_group)) ext3_error(sb, "ext3_new_block", "Allocating block in system zone - " - "block = %u", target_block); + "blocks from %u, length %lu", target_block, num); performed_allocation = 1; @@ -1341,10 +1390,14 @@ allocated: jbd_lock_bh_state(bitmap_bh); spin_lock(sb_bgl_lock(sbi, group_no)); if (buffer_jbd(bitmap_bh) && bh2jh(bitmap_bh)->b_committed_data) { - if (ext3_test_bit(ret_block, - bh2jh(bitmap_bh)->b_committed_data)) { - printk("%s: block was unexpectedly set in " - "b_committed_data\n", __FUNCTION__); + int i; + + for (i = 0; i < num; i++) { + if (ext3_test_bit(ret_block, + bh2jh(bitmap_bh)->b_committed_data)) { + printk("%s: block was unexpectedly set in " + "b_committed_data\n", __FUNCTION__); + } } } ext3_debug("found bit %d\n", ret_block); @@ -1355,7 +1408,7 @@ allocated: /* ret_block was blockgroup-relative. Now it becomes fs-relative */ ret_block = target_block; - if (ret_block >= le32_to_cpu(es->s_blocks_count)) { + if (ret_block + num - 1 >= le32_to_cpu(es->s_blocks_count)) { ext3_error(sb, "ext3_new_block", "block(%d) >= blocks count(%d) - " "block_group = %d, es == %p ", ret_block, @@ -1373,9 +1426,9 @@ allocated: spin_lock(sb_bgl_lock(sbi, group_no)); gdp->bg_free_blocks_count = - cpu_to_le16(le16_to_cpu(gdp->bg_free_blocks_count) - 1); + cpu_to_le16(le16_to_cpu(gdp->bg_free_blocks_count) - num); spin_unlock(sb_bgl_lock(sbi, group_no)); - percpu_counter_mod(&sbi->s_freeblocks_counter, -1); + percpu_counter_mod(&sbi->s_freeblocks_counter, -num); BUFFER_TRACE(gdp_bh, "journal_dirty_metadata for group descriptor"); err = ext3_journal_dirty_metadata(handle, gdp_bh); @@ -1388,6 +1441,8 @@ allocated: *errp = 0; brelse(bitmap_bh); + DQUOT_FREE_BLOCK(inode, *count-num); + *count = num; return ret_block; io_error: @@ -1401,11 +1456,19 @@ out: * Undo the block allocation */ if (!performed_allocation) - DQUOT_FREE_BLOCK(inode, 1); + DQUOT_FREE_BLOCK(inode, *count); brelse(bitmap_bh); return 0; } +int ext3_new_block(handle_t *handle, struct inode *inode, + unsigned long goal, int *errp) +{ + unsigned long count = 1; + + return ext3_new_blocks(handle, inode, goal, &count, errp); +} + unsigned long ext3_count_free_blocks(struct super_block *sb) { unsigned long desc_count; diff --git a/fs/ext3/bitmap.c b/fs/ext3/bitmap.c index cb16b4c5d5d..ce4f82b9e52 100644 --- a/fs/ext3/bitmap.c +++ b/fs/ext3/bitmap.c @@ -7,11 +7,11 @@ * Universite Pierre et Marie Curie (Paris VI) */ -#ifdef EXT3FS_DEBUG - #include <linux/buffer_head.h> +#include <linux/jbd.h> +#include <linux/ext3_fs.h> -#include "ext3_fs.h" +#ifdef EXT3FS_DEBUG static int nibblemap[] = {4, 3, 3, 2, 3, 2, 2, 1, 3, 2, 2, 1, 2, 1, 1, 0}; diff --git a/fs/ext3/dir.c b/fs/ext3/dir.c index 773459164bb..f37528ed222 100644 --- a/fs/ext3/dir.c +++ b/fs/ext3/dir.c @@ -39,7 +39,7 @@ static int ext3_dx_readdir(struct file * filp, static int ext3_release_dir (struct inode * inode, struct file * filp); -struct file_operations ext3_dir_operations = { +const struct file_operations ext3_dir_operations = { .llseek = generic_file_llseek, .read = generic_read_dir, .readdir = ext3_readdir, /* we take BKL. needed?*/ @@ -131,8 +131,9 @@ static int ext3_readdir(struct file * filp, struct buffer_head *bh = NULL; map_bh.b_state = 0; - err = ext3_get_block_handle(NULL, inode, blk, &map_bh, 0, 0); - if (!err) { + err = ext3_get_blocks_handle(NULL, inode, blk, 1, + &map_bh, 0, 0); + if (err > 0) { page_cache_readahead(sb->s_bdev->bd_inode->i_mapping, &filp->f_ra, filp, diff --git a/fs/ext3/file.c b/fs/ext3/file.c index 59098ea5671..783a796220b 100644 --- a/fs/ext3/file.c +++ b/fs/ext3/file.c @@ -105,7 +105,7 @@ force_commit: return ret; } -struct file_operations ext3_file_operations = { +const struct file_operations ext3_file_operations = { .llseek = generic_file_llseek, .read = do_sync_read, .write = do_sync_write, diff --git a/fs/ext3/inode.c b/fs/ext3/inode.c index 2c361377e0a..48ae0339af1 100644 --- a/fs/ext3/inode.c +++ b/fs/ext3/inode.c @@ -44,16 +44,16 @@ static int ext3_writepage_trans_blocks(struct inode *inode); /* * Test whether an inode is a fast symlink. */ -static inline int ext3_inode_is_fast_symlink(struct inode *inode) +static int ext3_inode_is_fast_symlink(struct inode *inode) { int ea_blocks = EXT3_I(inode)->i_file_acl ? (inode->i_sb->s_blocksize >> 9) : 0; - return (S_ISLNK(inode->i_mode) && - inode->i_blocks - ea_blocks == 0); + return (S_ISLNK(inode->i_mode) && inode->i_blocks - ea_blocks == 0); } -/* The ext3 forget function must perform a revoke if we are freeing data +/* + * The ext3 forget function must perform a revoke if we are freeing data * which has been journaled. Metadata (eg. indirect blocks) must be * revoked in all cases. * @@ -61,10 +61,8 @@ static inline int ext3_inode_is_fast_symlink(struct inode *inode) * but there may still be a record of it in the journal, and that record * still needs to be revoked. */ - -int ext3_forget(handle_t *handle, int is_metadata, - struct inode *inode, struct buffer_head *bh, - int blocknr) +int ext3_forget(handle_t *handle, int is_metadata, struct inode *inode, + struct buffer_head *bh, int blocknr) { int err; @@ -104,10 +102,9 @@ int ext3_forget(handle_t *handle, int is_metadata, } /* - * Work out how many blocks we need to progress with the next chunk of a + * Work out how many blocks we need to proceed with the next chunk of a * truncate transaction. */ - static unsigned long blocks_for_truncate(struct inode *inode) { unsigned long needed; @@ -141,7 +138,6 @@ static unsigned long blocks_for_truncate(struct inode *inode) * extend fails, we need to propagate the failure up and restart the * transaction in the top-level truncate loop. --sct */ - static handle_t *start_transaction(struct inode *inode) { handle_t *result; @@ -194,9 +190,11 @@ void ext3_delete_inode (struct inode * inode) handle = start_transaction(inode); if (IS_ERR(handle)) { - /* If we're going to skip the normal cleanup, we still - * need to make sure that the in-core orphan linked list - * is properly cleaned up. */ + /* + * If we're going to skip the normal cleanup, we still need to + * make sure that the in-core orphan linked list is properly + * cleaned up. + */ ext3_orphan_del(NULL, inode); goto no_delete; } @@ -235,16 +233,6 @@ no_delete: clear_inode(inode); /* We must guarantee clearing of inode... */ } -static int ext3_alloc_block (handle_t *handle, - struct inode * inode, unsigned long goal, int *err) -{ - unsigned long result; - - result = ext3_new_block(handle, inode, goal, err); - return result; -} - - typedef struct { __le32 *p; __le32 key; @@ -257,7 +245,7 @@ static inline void add_chain(Indirect *p, struct buffer_head *bh, __le32 *v) p->bh = bh; } -static inline int verify_chain(Indirect *from, Indirect *to) +static int verify_chain(Indirect *from, Indirect *to) { while (from <= to && from->key == *from->p) from++; @@ -327,10 +315,10 @@ static int ext3_block_to_path(struct inode *inode, offsets[n++] = i_block & (ptrs - 1); final = ptrs; } else { - ext3_warning (inode->i_sb, "ext3_block_to_path", "block > big"); + ext3_warning(inode->i_sb, "ext3_block_to_path", "block > big"); } if (boundary) - *boundary = (i_block & (ptrs - 1)) == (final - 1); + *boundary = final - 1 - (i_block & (ptrs - 1)); return n; } @@ -419,7 +407,6 @@ no_block: * * Caller must make sure that @ind is valid and will stay that way. */ - static unsigned long ext3_find_near(struct inode *inode, Indirect *ind) { struct ext3_inode_info *ei = EXT3_I(inode); @@ -429,17 +416,18 @@ static unsigned long ext3_find_near(struct inode *inode, Indirect *ind) unsigned long colour; /* Try to find previous block */ - for (p = ind->p - 1; p >= start; p--) + for (p = ind->p - 1; p >= start; p--) { if (*p) return le32_to_cpu(*p); + } /* No such thing, so let's try location of indirect block */ if (ind->bh) return ind->bh->b_blocknr; /* - * It is going to be refered from inode itself? OK, just put it into - * the same cylinder group then. + * It is going to be referred to from the inode itself? OK, just put it + * into the same cylinder group then. */ bg_start = (ei->i_block_group * EXT3_BLOCKS_PER_GROUP(inode->i_sb)) + le32_to_cpu(EXT3_SB(inode->i_sb)->s_es->s_first_data_block); @@ -463,7 +451,9 @@ static unsigned long ext3_find_near(struct inode *inode, Indirect *ind) static unsigned long ext3_find_goal(struct inode *inode, long block, Indirect chain[4], Indirect *partial) { - struct ext3_block_alloc_info *block_i = EXT3_I(inode)->i_block_alloc_info; + struct ext3_block_alloc_info *block_i; + + block_i = EXT3_I(inode)->i_block_alloc_info; /* * try the heuristic for sequential allocation, @@ -478,13 +468,113 @@ static unsigned long ext3_find_goal(struct inode *inode, long block, } /** + * ext3_blks_to_allocate: Look up the block map and count the number + * of direct blocks need to be allocated for the given branch. + * + * @branch: chain of indirect blocks + * @k: number of blocks need for indirect blocks + * @blks: number of data blocks to be mapped. + * @blocks_to_boundary: the offset in the indirect block + * + * return the total number of blocks to be allocate, including the + * direct and indirect blocks. + */ +static int ext3_blks_to_allocate(Indirect *branch, int k, unsigned long blks, + int blocks_to_boundary) +{ + unsigned long count = 0; + + /* + * Simple case, [t,d]Indirect block(s) has not allocated yet + * then it's clear blocks on that path have not allocated + */ + if (k > 0) { + /* right now we don't handle cross boundary allocation */ + if (blks < blocks_to_boundary + 1) + count += blks; + else + count += blocks_to_boundary + 1; + return count; + } + + count++; + while (count < blks && count <= blocks_to_boundary && + le32_to_cpu(*(branch[0].p + count)) == 0) { + count++; + } + return count; +} + +/** + * ext3_alloc_blocks: multiple allocate blocks needed for a branch + * @indirect_blks: the number of blocks need to allocate for indirect + * blocks + * + * @new_blocks: on return it will store the new block numbers for + * the indirect blocks(if needed) and the first direct block, + * @blks: on return it will store the total number of allocated + * direct blocks + */ +static int ext3_alloc_blocks(handle_t *handle, struct inode *inode, + unsigned long goal, int indirect_blks, int blks, + unsigned long long new_blocks[4], int *err) +{ + int target, i; + unsigned long count = 0; + int index = 0; + unsigned long current_block = 0; + int ret = 0; + + /* + * Here we try to allocate the requested multiple blocks at once, + * on a best-effort basis. + * To build a branch, we should allocate blocks for + * the indirect blocks(if not allocated yet), and at least + * the first direct block of this branch. That's the + * minimum number of blocks need to allocate(required) + */ + target = blks + indirect_blks; + + while (1) { + count = target; + /* allocating blocks for indirect blocks and direct blocks */ + current_block = ext3_new_blocks(handle,inode,goal,&count,err); + if (*err) + goto failed_out; + + target -= count; + /* allocate blocks for indirect blocks */ + while (index < indirect_blks && count) { + new_blocks[index++] = current_block++; + count--; + } + + if (count > 0) + break; + } + + /* save the new block number for the first direct block */ + new_blocks[index] = current_block; + + /* total number of blocks allocated for direct blocks */ + ret = count; + *err = 0; + return ret; +failed_out: + for (i = 0; i <index; i++) + ext3_free_blocks(handle, inode, new_blocks[i], 1); + return ret; +} + +/** * ext3_alloc_branch - allocate and set up a chain of blocks. * @inode: owner - * @num: depth of the chain (number of blocks to allocate) + * @indirect_blks: number of allocated indirect blocks + * @blks: number of allocated direct blocks * @offsets: offsets (in the blocks) to store the pointers to next. * @branch: place to store the chain in. * - * This function allocates @num blocks, zeroes out all but the last one, + * This function allocates blocks, zeroes out all but the last one, * links them into chain and (if we are synchronous) writes them to disk. * In other words, it prepares a branch that can be spliced onto the * inode. It stores the information about that chain in the branch[], in @@ -501,97 +591,106 @@ static unsigned long ext3_find_goal(struct inode *inode, long block, * ext3_alloc_block() (normally -ENOSPC). Otherwise we set the chain * as described above and return 0. */ - static int ext3_alloc_branch(handle_t *handle, struct inode *inode, - int num, - unsigned long goal, - int *offsets, - Indirect *branch) + int indirect_blks, int *blks, unsigned long goal, + int *offsets, Indirect *branch) { int blocksize = inode->i_sb->s_blocksize; - int n = 0, keys = 0; + int i, n = 0; int err = 0; - int i; - int parent = ext3_alloc_block(handle, inode, goal, &err); - - branch[0].key = cpu_to_le32(parent); - if (parent) { - for (n = 1; n < num; n++) { - struct buffer_head *bh; - /* Allocate the next block */ - int nr = ext3_alloc_block(handle, inode, parent, &err); - if (!nr) - break; - branch[n].key = cpu_to_le32(nr); + struct buffer_head *bh; + int num; + unsigned long long new_blocks[4]; + unsigned long long current_block; - /* - * Get buffer_head for parent block, zero it out - * and set the pointer to new one, then send - * parent to disk. - */ - bh = sb_getblk(inode->i_sb, parent); - if (!bh) - break; - keys = n+1; - branch[n].bh = bh; - lock_buffer(bh); - BUFFER_TRACE(bh, "call get_create_access"); - err = ext3_journal_get_create_access(handle, bh); - if (err) { - unlock_buffer(bh); - brelse(bh); - break; - } + num = ext3_alloc_blocks(handle, inode, goal, indirect_blks, + *blks, new_blocks, &err); + if (err) + return err; - memset(bh->b_data, 0, blocksize); - branch[n].p = (__le32*) bh->b_data + offsets[n]; - *branch[n].p = branch[n].key; - BUFFER_TRACE(bh, "marking uptodate"); - set_buffer_uptodate(bh); + branch[0].key = cpu_to_le32(new_blocks[0]); + /* + * metadata blocks and data blocks are allocated. + */ + for (n = 1; n <= indirect_blks; n++) { + /* + * Get buffer_head for parent block, zero it out + * and set the pointer to new one, then send + * parent to disk. + */ + bh = sb_getblk(inode->i_sb, new_blocks[n-1]); + branch[n].bh = bh; + lock_buffer(bh); + BUFFER_TRACE(bh, "call get_create_access"); + err = ext3_journal_get_create_access(handle, bh); + if (err) { unlock_buffer(bh); + brelse(bh); + goto failed; + } - BUFFER_TRACE(bh, "call ext3_journal_dirty_metadata"); - err = ext3_journal_dirty_metadata(handle, bh); - if (err) - break; - - parent = nr; + memset(bh->b_data, 0, blocksize); + branch[n].p = (__le32 *) bh->b_data + offsets[n]; + branch[n].key = cpu_to_le32(new_blocks[n]); + *branch[n].p = branch[n].key; + if ( n == indirect_blks) { + current_block = new_blocks[n]; + /* + * End of chain, update the last new metablock of + * the chain to point to the new allocated + * data blocks numbers + */ + for (i=1; i < num; i++) + *(branch[n].p + i) = cpu_to_le32(++current_block); } - } - if (n == num) - return 0; + BUFFER_TRACE(bh, "marking uptodate"); + set_buffer_uptodate(bh); + unlock_buffer(bh); + BUFFER_TRACE(bh, "call ext3_journal_dirty_metadata"); + err = ext3_journal_dirty_metadata(handle, bh); + if (err) + goto failed; + } + *blks = num; + return err; +failed: /* Allocation failed, free what we already allocated */ - for (i = 1; i < keys; i++) { + for (i = 1; i <= n ; i++) { BUFFER_TRACE(branch[i].bh, "call journal_forget"); ext3_journal_forget(handle, branch[i].bh); } - for (i = 0; i < keys; i++) - ext3_free_blocks(handle, inode, le32_to_cpu(branch[i].key), 1); + for (i = 0; i <indirect_blks; i++) + ext3_free_blocks(handle, inode, new_blocks[i], 1); + + ext3_free_blocks(handle, inode, new_blocks[i], num); + return err; } /** - * ext3_splice_branch - splice the allocated branch onto inode. - * @inode: owner - * @block: (logical) number of block we are adding - * @chain: chain of indirect blocks (with a missing link - see - * ext3_alloc_branch) - * @where: location of missing link - * @num: number of blocks we are adding - * - * This function fills the missing link and does all housekeeping needed in - * inode (->i_blocks, etc.). In case of success we end up with the full - * chain to new block and return 0. + * ext3_splice_branch - splice the allocated branch onto inode. + * @inode: owner + * @block: (logical) number of block we are adding + * @chain: chain of indirect blocks (with a missing link - see + * ext3_alloc_branch) + * @where: location of missing link + * @num: number of indirect blocks we are adding + * @blks: number of direct blocks we are adding + * + * This function fills the missing link and does all housekeeping needed in + * inode (->i_blocks, etc.). In case of success we end up with the full + * chain to new block and return 0. */ - -static int ext3_splice_branch(handle_t *handle, struct inode *inode, long block, - Indirect chain[4], Indirect *where, int num) +static int ext3_splice_branch(handle_t *handle, struct inode *inode, + long block, Indirect *where, int num, int blks) { int i; int err = 0; - struct ext3_block_alloc_info *block_i = EXT3_I(inode)->i_block_alloc_info; + struct ext3_block_alloc_info *block_i; + unsigned long current_block; + block_i = EXT3_I(inode)->i_block_alloc_info; /* * If we're splicing into a [td]indirect block (as opposed to the * inode) then we need to get write access to the [td]indirect block @@ -608,13 +707,24 @@ static int ext3_splice_branch(handle_t *handle, struct inode *inode, long block, *where->p = where->key; /* + * Update the host buffer_head or inode to point to more just allocated + * direct blocks blocks + */ + if (num == 0 && blks > 1) { + current_block = le32_to_cpu(where->key + 1); + for (i = 1; i < blks; i++) + *(where->p + i ) = cpu_to_le32(current_block++); + } + + /* * update the most recently allocated logical & physical block * in i_block_alloc_info, to assist find the proper goal block for next * allocation */ if (block_i) { - block_i->last_alloc_logical_block = block; - block_i->last_alloc_physical_block = le32_to_cpu(where[num-1].key); + block_i->last_alloc_logical_block = block + blks - 1; + block_i->last_alloc_physical_block = + le32_to_cpu(where[num].key + blks - 1); } /* We are done with atomic stuff, now do the rest of housekeeping */ @@ -625,7 +735,7 @@ static int ext3_splice_branch(handle_t *handle, struct inode *inode, long block, /* had we spliced it onto indirect block? */ if (where->bh) { /* - * akpm: If we spliced it onto an indirect block, we haven't + * If we spliced it onto an indirect block, we haven't * altered the inode. Note however that if it is being spliced * onto an indirect block at the very end of the file (the * file is growing) then we *will* alter the inode to reflect @@ -647,10 +757,13 @@ static int ext3_splice_branch(handle_t *handle, struct inode *inode, long block, return err; err_out: - for (i = 1; i < num; i++) { + for (i = 1; i <= num; i++) { BUFFER_TRACE(where[i].bh, "call journal_forget"); ext3_journal_forget(handle, where[i].bh); + ext3_free_blocks(handle,inode,le32_to_cpu(where[i-1].key),1); } + ext3_free_blocks(handle, inode, le32_to_cpu(where[num].key), blks); + return err; } @@ -666,26 +779,33 @@ err_out: * allocations is needed - we simply release blocks and do not touch anything * reachable from inode. * - * akpm: `handle' can be NULL if create == 0. + * `handle' can be NULL if create == 0. * * The BKL may not be held on entry here. Be sure to take it early. + * return > 0, # of blocks mapped or allocated. + * return = 0, if plain lookup failed. + * return < 0, error case. */ - -int -ext3_get_block_handle(handle_t *handle, struct inode *inode, sector_t iblock, - struct buffer_head *bh_result, int create, int extend_disksize) +int ext3_get_blocks_handle(handle_t *handle, struct inode *inode, + sector_t iblock, unsigned long maxblocks, + struct buffer_head *bh_result, + int create, int extend_disksize) { int err = -EIO; int offsets[4]; Indirect chain[4]; Indirect *partial; unsigned long goal; - int left; - int boundary = 0; - const int depth = ext3_block_to_path(inode, iblock, offsets, &boundary); + int indirect_blks; + int blocks_to_boundary = 0; + int depth; struct ext3_inode_info *ei = EXT3_I(inode); + int count = 0; + unsigned long first_block = 0; + J_ASSERT(handle != NULL || create == 0); + depth = ext3_block_to_path(inode,iblock,offsets,&blocks_to_boundary); if (depth == 0) goto out; @@ -694,8 +814,31 @@ ext3_get_block_handle(handle_t *handle, struct inode *inode, sector_t iblock, /* Simplest case - block found, no allocation needed */ if (!partial) { + first_block = chain[depth - 1].key; clear_buffer_new(bh_result); - goto got_it; + count++; + /*map more blocks*/ + while (count < maxblocks && count <= blocks_to_boundary) { + if (!verify_chain(chain, partial)) { + /* + * Indirect block might be removed by + * truncate while we were reading it. + * Handling of that case: forget what we've + * got now. Flag the err as EAGAIN, so it + * will reread. + */ + err = -EAGAIN; + count = 0; + break; + } + if (le32_to_cpu(*(chain[depth-1].p+count) == + (first_block + count))) + count++; + else + break; + } + if (err != -EAGAIN) + goto got_it; } /* Next simple case - plain lookup or failed read of indirect block */ @@ -723,6 +866,7 @@ ext3_get_block_handle(handle_t *handle, struct inode *inode, sector_t iblock, } partial = ext3_get_branch(inode, depth, offsets, chain, &err); if (!partial) { + count++; mutex_unlock(&ei->truncate_mutex); if (err) goto cleanup; @@ -740,12 +884,19 @@ ext3_get_block_handle(handle_t *handle, struct inode *inode, sector_t iblock, goal = ext3_find_goal(inode, iblock, chain, partial); - left = (chain + depth) - partial; + /* the number of blocks need to allocate for [d,t]indirect blocks */ + indirect_blks = (chain + depth) - partial - 1; /* + * Next look up the indirect map to count the totoal number of + * direct blocks to allocate for this branch. + */ + count = ext3_blks_to_allocate(partial, indirect_blks, + maxblocks, blocks_to_boundary); + /* * Block out ext3_truncate while we alter the tree */ - err = ext3_alloc_branch(handle, inode, left, goal, + err = ext3_alloc_branch(handle, inode, indirect_blks, &count, goal, offsets + (partial - chain), partial); /* @@ -756,8 +907,8 @@ ext3_get_block_handle(handle_t *handle, struct inode *inode, sector_t iblock, * may need to return -EAGAIN upwards in the worst case. --sct */ if (!err) - err = ext3_splice_branch(handle, inode, iblock, chain, - partial, left); + err = ext3_splice_branch(handle, inode, iblock, + partial, indirect_blks, count); /* * i_disksize growing is protected by truncate_mutex. Don't forget to * protect it if you're about to implement concurrent @@ -772,8 +923,9 @@ ext3_get_block_handle(handle_t *handle, struct inode *inode, sector_t iblock, set_buffer_new(bh_result); got_it: map_bh(bh_result, inode->i_sb, le32_to_cpu(chain[depth-1].key)); - if (boundary) + if (blocks_to_boundary == 0) set_buffer_boundary(bh_result); + err = count; /* Clean up and exit */ partial = chain + depth - 1; /* the whole chain */ cleanup: @@ -787,34 +939,21 @@ out: return err; } -static int ext3_get_block(struct inode *inode, sector_t iblock, - struct buffer_head *bh_result, int create) -{ - handle_t *handle = NULL; - int ret; - - if (create) { - handle = ext3_journal_current_handle(); - J_ASSERT(handle != 0); - } - ret = ext3_get_block_handle(handle, inode, iblock, - bh_result, create, 1); - return ret; -} - #define DIO_CREDITS (EXT3_RESERVE_TRANS_BLOCKS + 32) -static int -ext3_direct_io_get_blocks(struct inode *inode, sector_t iblock, - unsigned long max_blocks, struct buffer_head *bh_result, - int create) +static int ext3_get_block(struct inode *inode, sector_t iblock, + struct buffer_head *bh_result, int create) { handle_t *handle = journal_current_handle(); int ret = 0; + unsigned max_blocks = bh_result->b_size >> inode->i_blkbits; - if (!handle) + if (!create) goto get_block; /* A read */ + if (max_blocks == 1) + goto get_block; /* A single block get */ + if (handle->h_transaction->t_state == T_LOCKED) { /* * Huge direct-io writes can hold off commits for long @@ -841,18 +980,22 @@ ext3_direct_io_get_blocks(struct inode *inode, sector_t iblock, } get_block: - if (ret == 0) - ret = ext3_get_block_handle(handle, inode, iblock, - bh_result, create, 0); - bh_result->b_size = (1 << inode->i_blkbits); + if (ret == 0) { + ret = ext3_get_blocks_handle(handle, inode, iblock, + max_blocks, bh_result, create, 0); + if (ret > 0) { + bh_result->b_size = (ret << inode->i_blkbits); + ret = 0; + } + } return ret; } /* * `handle' can be NULL if create is zero */ -struct buffer_head *ext3_getblk(handle_t *handle, struct inode * inode, - long block, int create, int * errp) +struct buffer_head *ext3_getblk(handle_t *handle, struct inode *inode, + long block, int create, int *errp) { struct buffer_head dummy; int fatal = 0, err; @@ -862,8 +1005,16 @@ struct buffer_head *ext3_getblk(handle_t *handle, struct inode * inode, dummy.b_state = 0; dummy.b_blocknr = -1000; buffer_trace_init(&dummy.b_history); - *errp = ext3_get_block_handle(handle, inode, block, &dummy, create, 1); - if (!*errp && buffer_mapped(&dummy)) { + err = ext3_get_blocks_handle(handle, inode, block, 1, + &dummy, create, 1); + if (err == 1) { + err = 0; + } else if (err >= 0) { + WARN_ON(1); + err = -EIO; + } + *errp = err; + if (!err && buffer_mapped(&dummy)) { struct buffer_head *bh; bh = sb_getblk(inode->i_sb, dummy.b_blocknr); if (!bh) { @@ -874,17 +1025,18 @@ struct buffer_head *ext3_getblk(handle_t *handle, struct inode * inode, J_ASSERT(create != 0); J_ASSERT(handle != 0); - /* Now that we do not always journal data, we - should keep in mind whether this should - always journal the new buffer as metadata. - For now, regular file writes use - ext3_get_block instead, so it's not a - problem. */ + /* + * Now that we do not always journal data, we should + * keep in mind whether this should always journal the + * new buffer as metadata. For now, regular file + * writes use ext3_get_block instead, so it's not a + * problem. + */ lock_buffer(bh); BUFFER_TRACE(bh, "call get_create_access"); fatal = ext3_journal_get_create_access(handle, bh); if (!fatal && !buffer_uptodate(bh)) { - memset(bh->b_data, 0, inode->i_sb->s_blocksize); + memset(bh->b_data,0,inode->i_sb->s_blocksize); set_buffer_uptodate(bh); } unlock_buffer(bh); @@ -906,7 +1058,7 @@ err: return NULL; } -struct buffer_head *ext3_bread(handle_t *handle, struct inode * inode, +struct buffer_head *ext3_bread(handle_t *handle, struct inode *inode, int block, int create, int *err) { struct buffer_head * bh; @@ -982,9 +1134,8 @@ static int walk_page_buffers( handle_t *handle, * is elevated. We'll still have enough credits for the tiny quotafile * write. */ - -static int do_journal_get_write_access(handle_t *handle, - struct buffer_head *bh) +static int do_journal_get_write_access(handle_t *handle, + struct buffer_head *bh) { if (!buffer_mapped(bh) || buffer_freed(bh)) return 0; @@ -1025,8 +1176,7 @@ out: return ret; } -int -ext3_journal_dirty_data(handle_t *handle, struct buffer_head *bh) +int ext3_journal_dirty_data(handle_t *handle, struct buffer_head *bh) { int err = journal_dirty_data(handle, bh); if (err) @@ -1051,7 +1201,6 @@ static int commit_write_fn(handle_t *handle, struct buffer_head *bh) * ext3 never places buffers on inode->i_mapping->private_list. metadata * buffers are managed internally. */ - static int ext3_ordered_commit_write(struct file *file, struct page *page, unsigned from, unsigned to) { @@ -1261,7 +1410,7 @@ static int journal_dirty_data_fn(handle_t *handle, struct buffer_head *bh) * we don't need to open a transaction here. */ static int ext3_ordered_writepage(struct page *page, - struct writeback_control *wbc) + struct writeback_control *wbc) { struct inode *inode = page->mapping->host; struct buffer_head *page_bufs; @@ -1430,7 +1579,7 @@ ext3_readpages(struct file *file, struct address_space *mapping, return mpage_readpages(mapping, pages, nr_pages, ext3_get_block); } -static int ext3_invalidatepage(struct page *page, unsigned long offset) +static void ext3_invalidatepage(struct page *page, unsigned long offset) { journal_t *journal = EXT3_JOURNAL(page->mapping->host); @@ -1440,7 +1589,7 @@ static int ext3_invalidatepage(struct page *page, unsigned long offset) if (offset == 0) ClearPageChecked(page); - return journal_invalidatepage(journal, page, offset); + journal_invalidatepage(journal, page, offset); } static int ext3_releasepage(struct page *page, gfp_t wait) @@ -1492,11 +1641,10 @@ static ssize_t ext3_direct_IO(int rw, struct kiocb *iocb, ret = blockdev_direct_IO(rw, iocb, inode, inode->i_sb->s_bdev, iov, offset, nr_segs, - ext3_direct_io_get_blocks, NULL); + ext3_get_block, NULL); /* - * Reacquire the handle: ext3_direct_io_get_block() can restart the - * transaction + * Reacquire the handle: ext3_get_block() can restart the transaction */ handle = journal_current_handle(); @@ -1752,11 +1900,8 @@ static inline int all_zeroes(__le32 *p, __le32 *q) * c) free the subtrees growing from the inode past the @chain[0]. * (no partially truncated stuff there). */ -static Indirect *ext3_find_shared(struct inode *inode, - int depth, - int offsets[4], - Indirect chain[4], - __le32 *top) +static Indirect *ext3_find_shared(struct inode *inode, int depth, + int offsets[4], Indirect chain[4], __le32 *top) { Indirect *partial, *p; int k, err; @@ -1795,8 +1940,7 @@ static Indirect *ext3_find_shared(struct inode *inode, } /* Writer: end */ - while(partial > p) - { + while(partial > p) { brelse(partial->bh); partial--; } @@ -1812,10 +1956,9 @@ no_top: * We release `count' blocks on disk, but (last - first) may be greater * than `count' because there can be holes in there. */ -static void -ext3_clear_blocks(handle_t *handle, struct inode *inode, struct buffer_head *bh, - unsigned long block_to_free, unsigned long count, - __le32 *first, __le32 *last) +static void ext3_clear_blocks(handle_t *handle, struct inode *inode, + struct buffer_head *bh, unsigned long block_to_free, + unsigned long count, __le32 *first, __le32 *last) { __le32 *p; if (try_to_extend_transaction(handle, inode)) { @@ -2076,8 +2219,7 @@ static void ext3_free_branches(handle_t *handle, struct inode *inode, * that's fine - as long as they are linked from the inode, the post-crash * ext3_truncate() run will find them and release them. */ - -void ext3_truncate(struct inode * inode) +void ext3_truncate(struct inode *inode) { handle_t *handle; struct ext3_inode_info *ei = EXT3_I(inode); @@ -2201,29 +2343,26 @@ void ext3_truncate(struct inode * inode) do_indirects: /* Kill the remaining (whole) subtrees */ switch (offsets[0]) { - default: - nr = i_data[EXT3_IND_BLOCK]; - if (nr) { - ext3_free_branches(handle, inode, NULL, - &nr, &nr+1, 1); - i_data[EXT3_IND_BLOCK] = 0; - } - case EXT3_IND_BLOCK: - nr = i_data[EXT3_DIND_BLOCK]; - if (nr) { - ext3_free_branches(handle, inode, NULL, - &nr, &nr+1, 2); - i_data[EXT3_DIND_BLOCK] = 0; - } - case EXT3_DIND_BLOCK: - nr = i_data[EXT3_TIND_BLOCK]; - if (nr) { - ext3_free_branches(handle, inode, NULL, - &nr, &nr+1, 3); - i_data[EXT3_TIND_BLOCK] = 0; - } - case EXT3_TIND_BLOCK: - ; + default: + nr = i_data[EXT3_IND_BLOCK]; + if (nr) { + ext3_free_branches(handle, inode, NULL, &nr, &nr+1, 1); + i_data[EXT3_IND_BLOCK] = 0; + } + case EXT3_IND_BLOCK: + nr = i_data[EXT3_DIND_BLOCK]; + if (nr) { + ext3_free_branches(handle, inode, NULL, &nr, &nr+1, 2); + i_data[EXT3_DIND_BLOCK] = 0; + } + case EXT3_DIND_BLOCK: + nr = i_data[EXT3_TIND_BLOCK]; + if (nr) { + ext3_free_branches(handle, inode, NULL, &nr, &nr+1, 3); + i_data[EXT3_TIND_BLOCK] = 0; + } + case EXT3_TIND_BLOCK: + ; } ext3_discard_reservation(inode); @@ -2232,8 +2371,10 @@ do_indirects: inode->i_mtime = inode->i_ctime = CURRENT_TIME_SEC; ext3_mark_inode_dirty(handle, inode); - /* In a multi-transaction truncate, we only make the final - * transaction synchronous */ + /* + * In a multi-transaction truncate, we only make the final transaction + * synchronous + */ if (IS_SYNC(inode)) handle->h_sync = 1; out_stop: @@ -2259,20 +2400,16 @@ static unsigned long ext3_get_inode_block(struct super_block *sb, struct ext3_group_desc * gdp; - if ((ino != EXT3_ROOT_INO && - ino != EXT3_JOURNAL_INO && - ino != EXT3_RESIZE_INO && - ino < EXT3_FIRST_INO(sb)) || - ino > le32_to_cpu( - EXT3_SB(sb)->s_es->s_inodes_count)) { - ext3_error (sb, "ext3_get_inode_block", + if ((ino != EXT3_ROOT_INO && ino != EXT3_JOURNAL_INO && + ino != EXT3_RESIZE_INO && ino < EXT3_FIRST_INO(sb)) || + ino > le32_to_cpu(EXT3_SB(sb)->s_es->s_inodes_count)) { + ext3_error(sb, "ext3_get_inode_block", "bad inode number: %lu", ino); return 0; } block_group = (ino - 1) / EXT3_INODES_PER_GROUP(sb); if (block_group >= EXT3_SB(sb)->s_groups_count) { - ext3_error (sb, "ext3_get_inode_block", - "group >= groups count"); + ext3_error(sb,"ext3_get_inode_block","group >= groups count"); return 0; } smp_rmb(); @@ -2285,7 +2422,7 @@ static unsigned long ext3_get_inode_block(struct super_block *sb, return 0; } - gdp = (struct ext3_group_desc *) bh->b_data; + gdp = (struct ext3_group_desc *)bh->b_data; /* * Figure out the offset within the block group inode table */ @@ -2834,7 +2971,7 @@ err_out: /* - * akpm: how many blocks doth make a writepage()? + * How many blocks doth make a writepage()? * * With N blocks per page, it may be: * N data blocks @@ -2924,8 +3061,8 @@ ext3_reserve_inode_write(handle_t *handle, struct inode *inode, } /* - * akpm: What we do here is to mark the in-core inode as clean - * with respect to inode dirtiness (it may still be data-dirty). + * What we do here is to mark the in-core inode as clean with respect to inode + * dirtiness (it may still be data-dirty). * This means that the in-core inode may be reaped by prune_icache * without having to perform any I/O. This is a very good thing, * because *any* task may call prune_icache - even ones which @@ -2957,7 +3094,7 @@ int ext3_mark_inode_dirty(handle_t *handle, struct inode *inode) } /* - * akpm: ext3_dirty_inode() is called from __mark_inode_dirty() + * ext3_dirty_inode() is called from __mark_inode_dirty() * * We're really interested in the case where a file is being extended. * i_size has been changed by generic_commit_write() and we thus need @@ -2993,7 +3130,7 @@ out: return; } -#ifdef AKPM +#if 0 /* * Bind an inode's backing buffer_head into this transaction, to prevent * it from being flushed to disk early. Unlike @@ -3001,8 +3138,7 @@ out: * returns no iloc structure, so the caller needs to repeat the iloc * lookup to mark the inode dirty later. */ -static inline int -ext3_pin_inode(handle_t *handle, struct inode *inode) +static int ext3_pin_inode(handle_t *handle, struct inode *inode) { struct ext3_iloc iloc; diff --git a/fs/ext3/super.c b/fs/ext3/super.c index 86e443182de..f8a5266ea1f 100644 --- a/fs/ext3/super.c +++ b/fs/ext3/super.c @@ -1678,12 +1678,6 @@ static int ext3_fill_super (struct super_block *sb, void *data, int silent) } if (test_opt(sb, NOBH)) { - if (sb->s_blocksize_bits != PAGE_CACHE_SHIFT) { - printk(KERN_WARNING "EXT3-fs: Ignoring nobh option " - "since filesystem blocksize doesn't match " - "pagesize\n"); - clear_opt(sbi->s_mount_opt, NOBH); - } if (!(test_opt(sb, DATA_FLAGS) == EXT3_MOUNT_WRITEBACK_DATA)) { printk(KERN_WARNING "EXT3-fs: Ignoring nobh option - " "its supported only with writeback mode\n"); diff --git a/fs/fat/dir.c b/fs/fat/dir.c index 4095bc149eb..698b85bb1dd 100644 --- a/fs/fat/dir.c +++ b/fs/fat/dir.c @@ -741,7 +741,7 @@ static int fat_dir_ioctl(struct inode * inode, struct file * filp, return ret; } -struct file_operations fat_dir_operations = { +const struct file_operations fat_dir_operations = { .read = generic_read_dir, .readdir = fat_readdir, .ioctl = fat_dir_ioctl, diff --git a/fs/fat/file.c b/fs/fat/file.c index 88aa1ae13f9..1ee25232e6a 100644 --- a/fs/fat/file.c +++ b/fs/fat/file.c @@ -112,7 +112,7 @@ int fat_generic_ioctl(struct inode *inode, struct file *filp, } } -struct file_operations fat_file_operations = { +const struct file_operations fat_file_operations = { .llseek = generic_file_llseek, .read = do_sync_read, .write = do_sync_write, diff --git a/fs/fat/inode.c b/fs/fat/inode.c index 297300fe81c..c1ce284f8a9 100644 --- a/fs/fat/inode.c +++ b/fs/fat/inode.c @@ -101,11 +101,11 @@ static int __fat_get_blocks(struct inode *inode, sector_t iblock, } static int fat_get_blocks(struct inode *inode, sector_t iblock, - unsigned long max_blocks, struct buffer_head *bh_result, int create) { struct super_block *sb = inode->i_sb; int err; + unsigned long max_blocks = bh_result->b_size >> inode->i_blkbits; err = __fat_get_blocks(inode, iblock, &max_blocks, bh_result, create); if (err) @@ -1435,9 +1435,6 @@ out_fail: EXPORT_SYMBOL_GPL(fat_fill_super); -int __init fat_cache_init(void); -void fat_cache_destroy(void); - static int __init init_fat_fs(void) { int err; diff --git a/fs/fcntl.c b/fs/fcntl.c index 03c789560fb..2a2479196f9 100644 --- a/fs/fcntl.c +++ b/fs/fcntl.c @@ -412,7 +412,7 @@ out: /* Table to convert sigio signal codes into poll band bitmaps */ -static long band_table[NSIGPOLL] = { +static const long band_table[NSIGPOLL] = { POLLIN | POLLRDNORM, /* POLL_IN */ POLLOUT | POLLWRNORM | POLLWRBAND, /* POLL_OUT */ POLLIN | POLLRDNORM | POLLMSG, /* POLL_MSG */ @@ -531,7 +531,7 @@ int send_sigurg(struct fown_struct *fown) } static DEFINE_RWLOCK(fasync_lock); -static kmem_cache_t *fasync_cache; +static kmem_cache_t *fasync_cache __read_mostly; /* * fasync_helper() is used by some character device drivers (mainly mice) diff --git a/fs/fifo.c b/fs/fifo.c index d13fcd3ec80..889f722ee36 100644 --- a/fs/fifo.c +++ b/fs/fifo.c @@ -145,6 +145,6 @@ err_nocleanup: * is contain the open that then fills in the correct operations * depending on the access mode of the file... */ -struct file_operations def_fifo_fops = { +const struct file_operations def_fifo_fops = { .open = fifo_open, /* will set read or write pipe_fops */ }; diff --git a/fs/file.c b/fs/file.c index bbc74331473..55f4e702256 100644 --- a/fs/file.c +++ b/fs/file.c @@ -373,6 +373,6 @@ static void __devinit fdtable_defer_list_init(int cpu) void __init files_defer_init(void) { int i; - for_each_cpu(i) + for_each_possible_cpu(i) fdtable_defer_list_init(i); } diff --git a/fs/freevxfs/vxfs_extern.h b/fs/freevxfs/vxfs_extern.h index 927acf70c59..1cf1fe8466a 100644 --- a/fs/freevxfs/vxfs_extern.h +++ b/fs/freevxfs/vxfs_extern.h @@ -63,7 +63,7 @@ extern void vxfs_clear_inode(struct inode *); /* vxfs_lookup.c */ extern struct inode_operations vxfs_dir_inode_ops; -extern struct file_operations vxfs_dir_operations; +extern const struct file_operations vxfs_dir_operations; /* vxfs_olt.c */ extern int vxfs_read_olt(struct super_block *, u_long); diff --git a/fs/freevxfs/vxfs_lookup.c b/fs/freevxfs/vxfs_lookup.c index 554eb455722..29cce456c7c 100644 --- a/fs/freevxfs/vxfs_lookup.c +++ b/fs/freevxfs/vxfs_lookup.c @@ -56,7 +56,7 @@ struct inode_operations vxfs_dir_inode_ops = { .lookup = vxfs_lookup, }; -struct file_operations vxfs_dir_operations = { +const struct file_operations vxfs_dir_operations = { .readdir = vxfs_readdir, }; diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c index 785c7213a54..f3fbe2d030f 100644 --- a/fs/fs-writeback.c +++ b/fs/fs-writeback.c @@ -381,8 +381,8 @@ sync_sb_inodes(struct super_block *sb, struct writeback_control *wbc) list_move(&inode->i_list, &sb->s_dirty); } spin_unlock(&inode_lock); - cond_resched(); iput(inode); + cond_resched(); spin_lock(&inode_lock); if (wbc->nr_to_write <= 0) break; diff --git a/fs/fuse/dev.c b/fs/fuse/dev.c index 0c9a2ee54c9..23d1f52eb1b 100644 --- a/fs/fuse/dev.c +++ b/fs/fuse/dev.c @@ -922,7 +922,7 @@ static int fuse_dev_release(struct inode *inode, struct file *file) return 0; } -struct file_operations fuse_dev_operations = { +const struct file_operations fuse_dev_operations = { .owner = THIS_MODULE, .llseek = no_llseek, .read = fuse_dev_read, diff --git a/fs/fuse/dir.c b/fs/fuse/dir.c index c72a8a97935..256355b8025 100644 --- a/fs/fuse/dir.c +++ b/fs/fuse/dir.c @@ -1170,7 +1170,7 @@ static struct inode_operations fuse_dir_inode_operations = { .removexattr = fuse_removexattr, }; -static struct file_operations fuse_dir_operations = { +static const struct file_operations fuse_dir_operations = { .llseek = generic_file_llseek, .read = generic_read_dir, .readdir = fuse_readdir, diff --git a/fs/fuse/file.c b/fs/fuse/file.c index 6f05379b0a0..975f2697e86 100644 --- a/fs/fuse/file.c +++ b/fs/fuse/file.c @@ -12,7 +12,7 @@ #include <linux/slab.h> #include <linux/kernel.h> -static struct file_operations fuse_direct_io_file_operations; +static const struct file_operations fuse_direct_io_file_operations; static int fuse_send_open(struct inode *inode, struct file *file, int isdir, struct fuse_open_out *outargp) @@ -611,7 +611,7 @@ static int fuse_set_page_dirty(struct page *page) return 0; } -static struct file_operations fuse_file_operations = { +static const struct file_operations fuse_file_operations = { .llseek = generic_file_llseek, .read = generic_file_read, .write = generic_file_write, @@ -623,7 +623,7 @@ static struct file_operations fuse_file_operations = { .sendfile = generic_file_sendfile, }; -static struct file_operations fuse_direct_io_file_operations = { +static const struct file_operations fuse_direct_io_file_operations = { .llseek = generic_file_llseek, .read = fuse_direct_read, .write = fuse_direct_write, diff --git a/fs/fuse/fuse_i.h b/fs/fuse/fuse_i.h index 4a83adfec96..a16a04fcf41 100644 --- a/fs/fuse/fuse_i.h +++ b/fs/fuse/fuse_i.h @@ -346,7 +346,7 @@ static inline u64 get_node_id(struct inode *inode) } /** Device operations */ -extern struct file_operations fuse_dev_operations; +extern const struct file_operations fuse_dev_operations; /** * This is the single global spinlock which protects FUSE's structures diff --git a/fs/hfs/bnode.c b/fs/hfs/bnode.c index a7a7d77f3fd..1e44dcfe49c 100644 --- a/fs/hfs/bnode.c +++ b/fs/hfs/bnode.c @@ -306,8 +306,7 @@ void hfs_bnode_unhash(struct hfs_bnode *node) for (p = &node->tree->node_hash[hfs_bnode_hash(node->this)]; *p && *p != node; p = &(*p)->next_hash) ; - if (!*p) - BUG(); + BUG_ON(!*p); *p = node->next_hash; node->tree->node_hash_cnt--; } @@ -415,8 +414,7 @@ struct hfs_bnode *hfs_bnode_create(struct hfs_btree *tree, u32 num) spin_lock(&tree->hash_lock); node = hfs_bnode_findhash(tree, num); spin_unlock(&tree->hash_lock); - if (node) - BUG(); + BUG_ON(node); node = __hfs_bnode_create(tree, num); if (!node) return ERR_PTR(-ENOMEM); @@ -459,8 +457,7 @@ void hfs_bnode_put(struct hfs_bnode *node) dprint(DBG_BNODE_REFS, "put_node(%d:%d): %d\n", node->tree->cnid, node->this, atomic_read(&node->refcnt)); - if (!atomic_read(&node->refcnt)) - BUG(); + BUG_ON(!atomic_read(&node->refcnt)); if (!atomic_dec_and_lock(&node->refcnt, &tree->hash_lock)) return; for (i = 0; i < tree->pages_per_bnode; i++) { diff --git a/fs/hfs/btree.c b/fs/hfs/btree.c index 7bb11edd148..d20131ce4b9 100644 --- a/fs/hfs/btree.c +++ b/fs/hfs/btree.c @@ -36,8 +36,7 @@ struct hfs_btree *hfs_btree_open(struct super_block *sb, u32 id, btree_keycmp ke tree->inode = iget_locked(sb, id); if (!tree->inode) goto free_tree; - if (!(tree->inode->i_state & I_NEW)) - BUG(); + BUG_ON(!(tree->inode->i_state & I_NEW)); { struct hfs_mdb *mdb = HFS_SB(sb)->mdb; HFS_I(tree->inode)->flags = 0; diff --git a/fs/hfs/dir.c b/fs/hfs/dir.c index 534e5a7480e..7cd8cc03aea 100644 --- a/fs/hfs/dir.c +++ b/fs/hfs/dir.c @@ -313,7 +313,7 @@ static int hfs_rename(struct inode *old_dir, struct dentry *old_dentry, return res; } -struct file_operations hfs_dir_operations = { +const struct file_operations hfs_dir_operations = { .read = generic_read_dir, .readdir = hfs_readdir, .llseek = generic_file_llseek, diff --git a/fs/hfs/hfs_fs.h b/fs/hfs/hfs_fs.h index 18ce47ab1b7..3ed8663a8db 100644 --- a/fs/hfs/hfs_fs.h +++ b/fs/hfs/hfs_fs.h @@ -169,7 +169,7 @@ extern int hfs_cat_move(u32, struct inode *, struct qstr *, extern void hfs_cat_build_key(struct super_block *, btree_key *, u32, struct qstr *); /* dir.c */ -extern struct file_operations hfs_dir_operations; +extern const struct file_operations hfs_dir_operations; extern struct inode_operations hfs_dir_inode_operations; /* extent.c */ diff --git a/fs/hfs/inode.c b/fs/hfs/inode.c index 39fd85b9b91..2d4ced22201 100644 --- a/fs/hfs/inode.c +++ b/fs/hfs/inode.c @@ -17,7 +17,7 @@ #include "hfs_fs.h" #include "btree.h" -static struct file_operations hfs_file_operations; +static const struct file_operations hfs_file_operations; static struct inode_operations hfs_file_inode_operations; /*================ Variable-like macros ================*/ @@ -98,17 +98,6 @@ static int hfs_releasepage(struct page *page, gfp_t mask) return res ? try_to_free_buffers(page) : 0; } -static int hfs_get_blocks(struct inode *inode, sector_t iblock, unsigned long max_blocks, - struct buffer_head *bh_result, int create) -{ - int ret; - - ret = hfs_get_block(inode, iblock, bh_result, create); - if (!ret) - bh_result->b_size = (1 << inode->i_blkbits); - return ret; -} - static ssize_t hfs_direct_IO(int rw, struct kiocb *iocb, const struct iovec *iov, loff_t offset, unsigned long nr_segs) { @@ -116,7 +105,7 @@ static ssize_t hfs_direct_IO(int rw, struct kiocb *iocb, struct inode *inode = file->f_dentry->d_inode->i_mapping->host; return blockdev_direct_IO(rw, iocb, inode, inode->i_sb->s_bdev, iov, - offset, nr_segs, hfs_get_blocks, NULL); + offset, nr_segs, hfs_get_block, NULL); } static int hfs_writepages(struct address_space *mapping, @@ -612,7 +601,7 @@ int hfs_inode_setattr(struct dentry *dentry, struct iattr * attr) } -static struct file_operations hfs_file_operations = { +static const struct file_operations hfs_file_operations = { .llseek = generic_file_llseek, .read = generic_file_read, .write = generic_file_write, diff --git a/fs/hfsplus/dir.c b/fs/hfsplus/dir.c index 01a6fe3a395..1f9ece0de32 100644 --- a/fs/hfsplus/dir.c +++ b/fs/hfsplus/dir.c @@ -483,7 +483,7 @@ struct inode_operations hfsplus_dir_inode_operations = { .rename = hfsplus_rename, }; -struct file_operations hfsplus_dir_operations = { +const struct file_operations hfsplus_dir_operations = { .read = generic_read_dir, .readdir = hfsplus_readdir, .ioctl = hfsplus_ioctl, diff --git a/fs/hfsplus/inode.c b/fs/hfsplus/inode.c index 12ed2b7d046..acf66dba3e0 100644 --- a/fs/hfsplus/inode.c +++ b/fs/hfsplus/inode.c @@ -93,17 +93,6 @@ static int hfsplus_releasepage(struct page *page, gfp_t mask) return res ? try_to_free_buffers(page) : 0; } -static int hfsplus_get_blocks(struct inode *inode, sector_t iblock, unsigned long max_blocks, - struct buffer_head *bh_result, int create) -{ - int ret; - - ret = hfsplus_get_block(inode, iblock, bh_result, create); - if (!ret) - bh_result->b_size = (1 << inode->i_blkbits); - return ret; -} - static ssize_t hfsplus_direct_IO(int rw, struct kiocb *iocb, const struct iovec *iov, loff_t offset, unsigned long nr_segs) { @@ -111,7 +100,7 @@ static ssize_t hfsplus_direct_IO(int rw, struct kiocb *iocb, struct inode *inode = file->f_dentry->d_inode->i_mapping->host; return blockdev_direct_IO(rw, iocb, inode, inode->i_sb->s_bdev, iov, - offset, nr_segs, hfsplus_get_blocks, NULL); + offset, nr_segs, hfsplus_get_block, NULL); } static int hfsplus_writepages(struct address_space *mapping, @@ -291,7 +280,7 @@ static struct inode_operations hfsplus_file_inode_operations = { .listxattr = hfsplus_listxattr, }; -static struct file_operations hfsplus_file_operations = { +static const struct file_operations hfsplus_file_operations = { .llseek = generic_file_llseek, .read = generic_file_read, .write = generic_file_write, diff --git a/fs/hostfs/hostfs_kern.c b/fs/hostfs/hostfs_kern.c index b3ad0bd0312..bf0f8e16e43 100644 --- a/fs/hostfs/hostfs_kern.c +++ b/fs/hostfs/hostfs_kern.c @@ -384,7 +384,7 @@ int hostfs_fsync(struct file *file, struct dentry *dentry, int datasync) return fsync_file(HOSTFS_I(dentry->d_inode)->fd, datasync); } -static struct file_operations hostfs_file_fops = { +static const struct file_operations hostfs_file_fops = { .llseek = generic_file_llseek, .read = generic_file_read, .sendfile = generic_file_sendfile, @@ -399,7 +399,7 @@ static struct file_operations hostfs_file_fops = { .fsync = hostfs_fsync, }; -static struct file_operations hostfs_dir_fops = { +static const struct file_operations hostfs_dir_fops = { .llseek = generic_file_llseek, .readdir = hostfs_readdir, .read = generic_read_dir, diff --git a/fs/hostfs/hostfs_user.c b/fs/hostfs/hostfs_user.c index b97809deba6..23b7cee7212 100644 --- a/fs/hostfs/hostfs_user.c +++ b/fs/hostfs/hostfs_user.c @@ -360,7 +360,6 @@ int do_statfs(char *root, long *bsize_out, long long *blocks_out, spare_out[2] = buf.f_spare[2]; spare_out[3] = buf.f_spare[3]; spare_out[4] = buf.f_spare[4]; - spare_out[5] = buf.f_spare[5]; return(0); } diff --git a/fs/hpfs/dir.c b/fs/hpfs/dir.c index 5591f9623aa..ecc9180645a 100644 --- a/fs/hpfs/dir.c +++ b/fs/hpfs/dir.c @@ -310,7 +310,7 @@ struct dentry *hpfs_lookup(struct inode *dir, struct dentry *dentry, struct name return ERR_PTR(-ENOENT); } -struct file_operations hpfs_dir_ops = +const struct file_operations hpfs_dir_ops = { .llseek = hpfs_dir_lseek, .read = generic_read_dir, diff --git a/fs/hpfs/file.c b/fs/hpfs/file.c index 7c995ac4081..d3b9fffe45a 100644 --- a/fs/hpfs/file.c +++ b/fs/hpfs/file.c @@ -119,7 +119,7 @@ static ssize_t hpfs_file_write(struct file *file, const char __user *buf, return retval; } -struct file_operations hpfs_file_ops = +const struct file_operations hpfs_file_ops = { .llseek = generic_file_llseek, .read = generic_file_read, diff --git a/fs/hpfs/hpfs_fn.h b/fs/hpfs/hpfs_fn.h index 4c6473ab3b3..29b7a3e5517 100644 --- a/fs/hpfs/hpfs_fn.h +++ b/fs/hpfs/hpfs_fn.h @@ -240,7 +240,7 @@ void hpfs_set_dentry_operations(struct dentry *); /* dir.c */ struct dentry *hpfs_lookup(struct inode *, struct dentry *, struct nameidata *); -extern struct file_operations hpfs_dir_ops; +extern const struct file_operations hpfs_dir_ops; /* dnode.c */ @@ -266,7 +266,7 @@ void hpfs_set_ea(struct inode *, struct fnode *, char *, char *, int); /* file.c */ int hpfs_file_fsync(struct file *, struct dentry *, int); -extern struct file_operations hpfs_file_ops; +extern const struct file_operations hpfs_file_ops; extern struct inode_operations hpfs_file_iops; extern struct address_space_operations hpfs_aops; diff --git a/fs/hppfs/hppfs_kern.c b/fs/hppfs/hppfs_kern.c index a44dc589739..2ba20cdb5ba 100644 --- a/fs/hppfs/hppfs_kern.c +++ b/fs/hppfs/hppfs_kern.c @@ -558,7 +558,7 @@ static loff_t hppfs_llseek(struct file *file, loff_t off, int where) return(default_llseek(file, off, where)); } -static struct file_operations hppfs_file_fops = { +static const struct file_operations hppfs_file_fops = { .owner = NULL, .llseek = hppfs_llseek, .read = hppfs_read, @@ -609,7 +609,7 @@ static int hppfs_fsync(struct file *file, struct dentry *dentry, int datasync) return(0); } -static struct file_operations hppfs_dir_fops = { +static const struct file_operations hppfs_dir_fops = { .owner = NULL, .readdir = hppfs_readdir, .open = hppfs_dir_open, diff --git a/fs/hugetlbfs/inode.c b/fs/hugetlbfs/inode.c index 25fa8bba8cb..3a5b4e92345 100644 --- a/fs/hugetlbfs/inode.c +++ b/fs/hugetlbfs/inode.c @@ -35,7 +35,7 @@ static struct super_operations hugetlbfs_ops; static struct address_space_operations hugetlbfs_aops; -struct file_operations hugetlbfs_file_operations; +const struct file_operations hugetlbfs_file_operations; static struct inode_operations hugetlbfs_dir_inode_operations; static struct inode_operations hugetlbfs_inode_operations; @@ -566,7 +566,7 @@ static void init_once(void *foo, kmem_cache_t *cachep, unsigned long flags) inode_init_once(&ei->vfs_inode); } -struct file_operations hugetlbfs_file_operations = { +const struct file_operations hugetlbfs_file_operations = { .mmap = hugetlbfs_file_mmap, .fsync = simple_sync_file, .get_unmapped_area = hugetlb_get_unmapped_area, diff --git a/fs/inode.c b/fs/inode.c index a51c671c54c..32b7c337502 100644 --- a/fs/inode.c +++ b/fs/inode.c @@ -56,8 +56,8 @@ #define I_HASHBITS i_hash_shift #define I_HASHMASK i_hash_mask -static unsigned int i_hash_mask; -static unsigned int i_hash_shift; +static unsigned int i_hash_mask __read_mostly; +static unsigned int i_hash_shift __read_mostly; /* * Each inode can be on two separate lists. One is @@ -73,7 +73,7 @@ static unsigned int i_hash_shift; LIST_HEAD(inode_in_use); LIST_HEAD(inode_unused); -static struct hlist_head *inode_hashtable; +static struct hlist_head *inode_hashtable __read_mostly; /* * A simple spinlock to protect the list manipulations. @@ -91,20 +91,20 @@ DEFINE_SPINLOCK(inode_lock); * from its final dispose_list, the struct super_block they refer to * (for inode->i_sb->s_op) may already have been freed and reused. */ -DEFINE_MUTEX(iprune_mutex); +static DEFINE_MUTEX(iprune_mutex); /* * Statistics gathering.. */ struct inodes_stat_t inodes_stat; -static kmem_cache_t * inode_cachep; +static kmem_cache_t * inode_cachep __read_mostly; static struct inode *alloc_inode(struct super_block *sb) { static struct address_space_operations empty_aops; static struct inode_operations empty_iops; - static struct file_operations empty_fops; + static const struct file_operations empty_fops; struct inode *inode; if (sb->s_op->alloc_inode) diff --git a/fs/inotify.c b/fs/inotify.c index 0ee39ef591c..367c487c014 100644 --- a/fs/inotify.c +++ b/fs/inotify.c @@ -38,17 +38,16 @@ #include <asm/ioctls.h> static atomic_t inotify_cookie; -static atomic_t inotify_watches; -static kmem_cache_t *watch_cachep; -static kmem_cache_t *event_cachep; +static kmem_cache_t *watch_cachep __read_mostly; +static kmem_cache_t *event_cachep __read_mostly; -static struct vfsmount *inotify_mnt; +static struct vfsmount *inotify_mnt __read_mostly; /* these are configurable via /proc/sys/fs/inotify/ */ -int inotify_max_user_instances; -int inotify_max_user_watches; -int inotify_max_queued_events; +int inotify_max_user_instances __read_mostly; +int inotify_max_user_watches __read_mostly; +int inotify_max_queued_events __read_mostly; /* * Lock ordering: @@ -381,6 +380,48 @@ static int find_inode(const char __user *dirname, struct nameidata *nd, } /* + * inotify_inode_watched - returns nonzero if there are watches on this inode + * and zero otherwise. We call this lockless, we do not care if we race. + */ +static inline int inotify_inode_watched(struct inode *inode) +{ + return !list_empty(&inode->inotify_watches); +} + +/* + * Get child dentry flag into synch with parent inode. + * Flag should always be clear for negative dentrys. + */ +static void set_dentry_child_flags(struct inode *inode, int watched) +{ + struct dentry *alias; + + spin_lock(&dcache_lock); + list_for_each_entry(alias, &inode->i_dentry, d_alias) { + struct dentry *child; + + list_for_each_entry(child, &alias->d_subdirs, d_u.d_child) { + if (!child->d_inode) { + WARN_ON(child->d_flags & DCACHE_INOTIFY_PARENT_WATCHED); + continue; + } + spin_lock(&child->d_lock); + if (watched) { + WARN_ON(child->d_flags & + DCACHE_INOTIFY_PARENT_WATCHED); + child->d_flags |= DCACHE_INOTIFY_PARENT_WATCHED; + } else { + WARN_ON(!(child->d_flags & + DCACHE_INOTIFY_PARENT_WATCHED)); + child->d_flags&=~DCACHE_INOTIFY_PARENT_WATCHED; + } + spin_unlock(&child->d_lock); + } + } + spin_unlock(&dcache_lock); +} + +/* * create_watch - creates a watch on the given device. * * Callers must hold dev->mutex. Calls inotify_dev_get_wd() so may sleep. @@ -426,7 +467,6 @@ static struct inotify_watch *create_watch(struct inotify_device *dev, get_inotify_watch(watch); atomic_inc(&dev->user->inotify_watches); - atomic_inc(&inotify_watches); return watch; } @@ -458,8 +498,10 @@ static void remove_watch_no_event(struct inotify_watch *watch, list_del(&watch->i_list); list_del(&watch->d_list); + if (!inotify_inode_watched(watch->inode)) + set_dentry_child_flags(watch->inode, 0); + atomic_dec(&dev->user->inotify_watches); - atomic_dec(&inotify_watches); idr_remove(&dev->idr, watch->wd); put_inotify_watch(watch); } @@ -481,16 +523,39 @@ static void remove_watch(struct inotify_watch *watch,struct inotify_device *dev) remove_watch_no_event(watch, dev); } +/* Kernel API */ + /* - * inotify_inode_watched - returns nonzero if there are watches on this inode - * and zero otherwise. We call this lockless, we do not care if we race. + * inotify_d_instantiate - instantiate dcache entry for inode */ -static inline int inotify_inode_watched(struct inode *inode) +void inotify_d_instantiate(struct dentry *entry, struct inode *inode) { - return !list_empty(&inode->inotify_watches); + struct dentry *parent; + + if (!inode) + return; + + WARN_ON(entry->d_flags & DCACHE_INOTIFY_PARENT_WATCHED); + spin_lock(&entry->d_lock); + parent = entry->d_parent; + if (inotify_inode_watched(parent->d_inode)) + entry->d_flags |= DCACHE_INOTIFY_PARENT_WATCHED; + spin_unlock(&entry->d_lock); } -/* Kernel API */ +/* + * inotify_d_move - dcache entry has been moved + */ +void inotify_d_move(struct dentry *entry) +{ + struct dentry *parent; + + parent = entry->d_parent; + if (inotify_inode_watched(parent->d_inode)) + entry->d_flags |= DCACHE_INOTIFY_PARENT_WATCHED; + else + entry->d_flags &= ~DCACHE_INOTIFY_PARENT_WATCHED; +} /** * inotify_inode_queue_event - queue an event to all watches on this inode @@ -538,7 +603,7 @@ void inotify_dentry_parent_queue_event(struct dentry *dentry, u32 mask, struct dentry *parent; struct inode *inode; - if (!atomic_read (&inotify_watches)) + if (!(dentry->d_flags & DCACHE_INOTIFY_PARENT_WATCHED)) return; spin_lock(&dentry->d_lock); @@ -855,7 +920,7 @@ static long inotify_ioctl(struct file *file, unsigned int cmd, return ret; } -static struct file_operations inotify_fops = { +static const struct file_operations inotify_fops = { .poll = inotify_poll, .read = inotify_read, .release = inotify_release, @@ -993,6 +1058,9 @@ asmlinkage long sys_inotify_add_watch(int fd, const char __user *path, u32 mask) goto out; } + if (!inotify_inode_watched(inode)) + set_dentry_child_flags(inode, 1); + /* Add the watch to the device's and the inode's list */ list_add(&watch->d_list, &dev->watches); list_add(&watch->i_list, &inode->inotify_watches); @@ -1065,7 +1133,6 @@ static int __init inotify_setup(void) inotify_max_user_watches = 8192; atomic_set(&inotify_cookie, 0); - atomic_set(&inotify_watches, 0); watch_cachep = kmem_cache_create("inotify_watch_cache", sizeof(struct inotify_watch), diff --git a/fs/isofs/dir.c b/fs/isofs/dir.c index 7901ac9f97a..5440ea292c6 100644 --- a/fs/isofs/dir.c +++ b/fs/isofs/dir.c @@ -16,7 +16,7 @@ static int isofs_readdir(struct file *, void *, filldir_t); -struct file_operations isofs_dir_operations = +const struct file_operations isofs_dir_operations = { .read = generic_read_dir, .readdir = isofs_readdir, diff --git a/fs/isofs/isofs.h b/fs/isofs/isofs.h index 439a19b1bf3..b87ba066f5e 100644 --- a/fs/isofs/isofs.h +++ b/fs/isofs/isofs.h @@ -175,6 +175,6 @@ isofs_normalize_block_and_offset(struct iso_directory_record* de, } extern struct inode_operations isofs_dir_inode_operations; -extern struct file_operations isofs_dir_operations; +extern const struct file_operations isofs_dir_operations; extern struct address_space_operations isofs_symlink_aops; extern struct export_operations isofs_export_ops; diff --git a/fs/jbd/journal.c b/fs/jbd/journal.c index 95a628d8cac..7f96b5cb678 100644 --- a/fs/jbd/journal.c +++ b/fs/jbd/journal.c @@ -33,9 +33,11 @@ #include <linux/mm.h> #include <linux/suspend.h> #include <linux/pagemap.h> +#include <linux/kthread.h> +#include <linux/proc_fs.h> + #include <asm/uaccess.h> #include <asm/page.h> -#include <linux/proc_fs.h> EXPORT_SYMBOL(journal_start); EXPORT_SYMBOL(journal_restart); @@ -111,18 +113,15 @@ static void commit_timeout(unsigned long __data) static int kjournald(void *arg) { - journal_t *journal = (journal_t *) arg; + journal_t *journal = arg; transaction_t *transaction; - struct timer_list timer; - daemonize("kjournald"); - - /* Set up an interval timer which can be used to trigger a - commit wakeup after the commit interval expires */ - init_timer(&timer); - timer.data = (unsigned long) current; - timer.function = commit_timeout; - journal->j_commit_timer = &timer; + /* + * Set up an interval timer which can be used to trigger a commit wakeup + * after the commit interval expires + */ + setup_timer(&journal->j_commit_timer, commit_timeout, + (unsigned long)current); /* Record that the journal thread is running */ journal->j_task = current; @@ -146,7 +145,7 @@ loop: if (journal->j_commit_sequence != journal->j_commit_request) { jbd_debug(1, "OK, requests differ\n"); spin_unlock(&journal->j_state_lock); - del_timer_sync(journal->j_commit_timer); + del_timer_sync(&journal->j_commit_timer); journal_commit_transaction(journal); spin_lock(&journal->j_state_lock); goto loop; @@ -203,7 +202,7 @@ loop: end_loop: spin_unlock(&journal->j_state_lock); - del_timer_sync(journal->j_commit_timer); + del_timer_sync(&journal->j_commit_timer); journal->j_task = NULL; wake_up(&journal->j_wait_done_commit); jbd_debug(1, "Journal thread exiting.\n"); @@ -212,7 +211,7 @@ end_loop: static void journal_start_thread(journal_t *journal) { - kernel_thread(kjournald, journal, CLONE_VM|CLONE_FS|CLONE_FILES); + kthread_run(kjournald, journal, "kjournald"); wait_event(journal->j_wait_done_commit, journal->j_task != 0); } diff --git a/fs/jbd/transaction.c b/fs/jbd/transaction.c index 5fc40888f4c..c609f5034fc 100644 --- a/fs/jbd/transaction.c +++ b/fs/jbd/transaction.c @@ -53,8 +53,8 @@ get_transaction(journal_t *journal, transaction_t *transaction) spin_lock_init(&transaction->t_handle_lock); /* Set up the commit timer for the new transaction. */ - journal->j_commit_timer->expires = transaction->t_expires; - add_timer(journal->j_commit_timer); + journal->j_commit_timer.expires = transaction->t_expires; + add_timer(&journal->j_commit_timer); J_ASSERT(journal->j_running_transaction == NULL); journal->j_running_transaction = transaction; @@ -1873,16 +1873,15 @@ zap_buffer_unlocked: } /** - * int journal_invalidatepage() + * void journal_invalidatepage() * @journal: journal to use for flush... * @page: page to flush * @offset: length of page to invalidate. * * Reap page buffers containing data after offset in page. * - * Return non-zero if the page's buffers were successfully reaped. */ -int journal_invalidatepage(journal_t *journal, +void journal_invalidatepage(journal_t *journal, struct page *page, unsigned long offset) { @@ -1893,7 +1892,7 @@ int journal_invalidatepage(journal_t *journal, if (!PageLocked(page)) BUG(); if (!page_has_buffers(page)) - return 1; + return; /* We will potentially be playing with lists other than just the * data lists (especially for journaled data mode), so be @@ -1916,11 +1915,9 @@ int journal_invalidatepage(journal_t *journal, } while (bh != head); if (!offset) { - if (!may_free || !try_to_free_buffers(page)) - return 0; - J_ASSERT(!page_has_buffers(page)); + if (may_free && try_to_free_buffers(page)) + J_ASSERT(!page_has_buffers(page)); } - return 1; } /* diff --git a/fs/jffs/inode-v23.c b/fs/jffs/inode-v23.c index 5a4519e834d..020cc097c53 100644 --- a/fs/jffs/inode-v23.c +++ b/fs/jffs/inode-v23.c @@ -55,9 +55,9 @@ static int jffs_remove(struct inode *dir, struct dentry *dentry, int type); static struct super_operations jffs_ops; -static struct file_operations jffs_file_operations; +static const struct file_operations jffs_file_operations; static struct inode_operations jffs_file_inode_operations; -static struct file_operations jffs_dir_operations; +static const struct file_operations jffs_dir_operations; static struct inode_operations jffs_dir_inode_operations; static struct address_space_operations jffs_address_operations; @@ -1629,7 +1629,7 @@ static int jffs_fsync(struct file *f, struct dentry *d, int datasync) } -static struct file_operations jffs_file_operations = +static const struct file_operations jffs_file_operations = { .open = generic_file_open, .llseek = generic_file_llseek, @@ -1649,7 +1649,7 @@ static struct inode_operations jffs_file_inode_operations = }; -static struct file_operations jffs_dir_operations = +static const struct file_operations jffs_dir_operations = { .readdir = jffs_readdir, }; diff --git a/fs/jffs2/compr_zlib.c b/fs/jffs2/compr_zlib.c index 4db8be8e90c..5c63e0cdcf4 100644 --- a/fs/jffs2/compr_zlib.c +++ b/fs/jffs2/compr_zlib.c @@ -33,13 +33,14 @@ */ #define STREAM_END_SPACE 12 -static DECLARE_MUTEX(deflate_sem); -static DECLARE_MUTEX(inflate_sem); +static DEFINE_MUTEX(deflate_mutex); +static DEFINE_MUTEX(inflate_mutex); static z_stream inf_strm, def_strm; #ifdef __KERNEL__ /* Linux-only */ #include <linux/vmalloc.h> #include <linux/init.h> +#include <linux/mutex.h> static int __init alloc_workspaces(void) { @@ -79,11 +80,11 @@ static int jffs2_zlib_compress(unsigned char *data_in, if (*dstlen <= STREAM_END_SPACE) return -1; - down(&deflate_sem); + mutex_lock(&deflate_mutex); if (Z_OK != zlib_deflateInit(&def_strm, 3)) { printk(KERN_WARNING "deflateInit failed\n"); - up(&deflate_sem); + mutex_unlock(&deflate_mutex); return -1; } @@ -104,7 +105,7 @@ static int jffs2_zlib_compress(unsigned char *data_in, if (ret != Z_OK) { D1(printk(KERN_DEBUG "deflate in loop returned %d\n", ret)); zlib_deflateEnd(&def_strm); - up(&deflate_sem); + mutex_unlock(&deflate_mutex); return -1; } } @@ -133,7 +134,7 @@ static int jffs2_zlib_compress(unsigned char *data_in, *sourcelen = def_strm.total_in; ret = 0; out: - up(&deflate_sem); + mutex_unlock(&deflate_mutex); return ret; } @@ -145,7 +146,7 @@ static int jffs2_zlib_decompress(unsigned char *data_in, int ret; int wbits = MAX_WBITS; - down(&inflate_sem); + mutex_lock(&inflate_mutex); inf_strm.next_in = data_in; inf_strm.avail_in = srclen; @@ -173,7 +174,7 @@ static int jffs2_zlib_decompress(unsigned char *data_in, if (Z_OK != zlib_inflateInit2(&inf_strm, wbits)) { printk(KERN_WARNING "inflateInit failed\n"); - up(&inflate_sem); + mutex_unlock(&inflate_mutex); return 1; } @@ -183,7 +184,7 @@ static int jffs2_zlib_decompress(unsigned char *data_in, printk(KERN_NOTICE "inflate returned %d\n", ret); } zlib_inflateEnd(&inf_strm); - up(&inflate_sem); + mutex_unlock(&inflate_mutex); return 0; } diff --git a/fs/jffs2/dir.c b/fs/jffs2/dir.c index a7bf9cb2567..8bc7a5018e4 100644 --- a/fs/jffs2/dir.c +++ b/fs/jffs2/dir.c @@ -37,7 +37,7 @@ static int jffs2_mknod (struct inode *,struct dentry *,int,dev_t); static int jffs2_rename (struct inode *, struct dentry *, struct inode *, struct dentry *); -struct file_operations jffs2_dir_operations = +const struct file_operations jffs2_dir_operations = { .read = generic_read_dir, .readdir = jffs2_readdir, diff --git a/fs/jffs2/file.c b/fs/jffs2/file.c index 935f273dc57..9f4171213e5 100644 --- a/fs/jffs2/file.c +++ b/fs/jffs2/file.c @@ -38,7 +38,7 @@ int jffs2_fsync(struct file *filp, struct dentry *dentry, int datasync) return 0; } -struct file_operations jffs2_file_operations = +const struct file_operations jffs2_file_operations = { .llseek = generic_file_llseek, .open = generic_file_open, diff --git a/fs/jffs2/os-linux.h b/fs/jffs2/os-linux.h index 59e7a393200..d307cf54862 100644 --- a/fs/jffs2/os-linux.h +++ b/fs/jffs2/os-linux.h @@ -159,11 +159,11 @@ void jffs2_stop_garbage_collect_thread(struct jffs2_sb_info *c); void jffs2_garbage_collect_trigger(struct jffs2_sb_info *c); /* dir.c */ -extern struct file_operations jffs2_dir_operations; +extern const struct file_operations jffs2_dir_operations; extern struct inode_operations jffs2_dir_inode_operations; /* file.c */ -extern struct file_operations jffs2_file_operations; +extern const struct file_operations jffs2_file_operations; extern struct inode_operations jffs2_file_inode_operations; extern struct address_space_operations jffs2_file_address_operations; int jffs2_fsync(struct file *, struct dentry *, int); diff --git a/fs/jfs/file.c b/fs/jfs/file.c index e1ac6e497e2..1c9745be5ad 100644 --- a/fs/jfs/file.c +++ b/fs/jfs/file.c @@ -100,7 +100,7 @@ struct inode_operations jfs_file_inode_operations = { #endif }; -struct file_operations jfs_file_operations = { +const struct file_operations jfs_file_operations = { .open = jfs_open, .llseek = generic_file_llseek, .write = generic_file_write, diff --git a/fs/jfs/inode.c b/fs/jfs/inode.c index 51a5fed90cc..04eb78f1252 100644 --- a/fs/jfs/inode.c +++ b/fs/jfs/inode.c @@ -258,7 +258,8 @@ jfs_get_blocks(struct inode *ip, sector_t lblock, unsigned long max_blocks, static int jfs_get_block(struct inode *ip, sector_t lblock, struct buffer_head *bh_result, int create) { - return jfs_get_blocks(ip, lblock, 1, bh_result, create); + return jfs_get_blocks(ip, lblock, bh_result->b_size >> ip->i_blkbits, + bh_result, create); } static int jfs_writepage(struct page *page, struct writeback_control *wbc) @@ -301,7 +302,7 @@ static ssize_t jfs_direct_IO(int rw, struct kiocb *iocb, struct inode *inode = file->f_mapping->host; return blockdev_direct_IO(rw, iocb, inode, inode->i_sb->s_bdev, iov, - offset, nr_segs, jfs_get_blocks, NULL); + offset, nr_segs, jfs_get_block, NULL); } struct address_space_operations jfs_aops = { diff --git a/fs/jfs/jfs_inode.h b/fs/jfs/jfs_inode.h index 095d471b9f9..c3007267446 100644 --- a/fs/jfs/jfs_inode.h +++ b/fs/jfs/jfs_inode.h @@ -35,9 +35,9 @@ extern void jfs_set_inode_flags(struct inode *); extern struct address_space_operations jfs_aops; extern struct inode_operations jfs_dir_inode_operations; -extern struct file_operations jfs_dir_operations; +extern const struct file_operations jfs_dir_operations; extern struct inode_operations jfs_file_inode_operations; -extern struct file_operations jfs_file_operations; +extern const struct file_operations jfs_file_operations; extern struct inode_operations jfs_symlink_inode_operations; extern struct dentry_operations jfs_ci_dentry_operations; #endif /* _H_JFS_INODE */ diff --git a/fs/jfs/jfs_logmgr.c b/fs/jfs/jfs_logmgr.c index 0b348b13b55..3315f0b1fbc 100644 --- a/fs/jfs/jfs_logmgr.c +++ b/fs/jfs/jfs_logmgr.c @@ -69,6 +69,7 @@ #include <linux/bio.h> #include <linux/suspend.h> #include <linux/delay.h> +#include <linux/mutex.h> #include "jfs_incore.h" #include "jfs_filsys.h" #include "jfs_metapage.h" @@ -165,7 +166,7 @@ do { \ */ static LIST_HEAD(jfs_external_logs); static struct jfs_log *dummy_log = NULL; -static DECLARE_MUTEX(jfs_log_sem); +static DEFINE_MUTEX(jfs_log_mutex); /* * forward references @@ -1085,20 +1086,20 @@ int lmLogOpen(struct super_block *sb) if (sbi->mntflag & JFS_INLINELOG) return open_inline_log(sb); - down(&jfs_log_sem); + mutex_lock(&jfs_log_mutex); list_for_each_entry(log, &jfs_external_logs, journal_list) { if (log->bdev->bd_dev == sbi->logdev) { if (memcmp(log->uuid, sbi->loguuid, sizeof(log->uuid))) { jfs_warn("wrong uuid on JFS journal\n"); - up(&jfs_log_sem); + mutex_unlock(&jfs_log_mutex); return -EINVAL; } /* * add file system to log active file system list */ if ((rc = lmLogFileSystem(log, sbi, 1))) { - up(&jfs_log_sem); + mutex_unlock(&jfs_log_mutex); return rc; } goto journal_found; @@ -1106,7 +1107,7 @@ int lmLogOpen(struct super_block *sb) } if (!(log = kzalloc(sizeof(struct jfs_log), GFP_KERNEL))) { - up(&jfs_log_sem); + mutex_unlock(&jfs_log_mutex); return -ENOMEM; } INIT_LIST_HEAD(&log->sb_list); @@ -1151,7 +1152,7 @@ journal_found: sbi->log = log; LOG_UNLOCK(log); - up(&jfs_log_sem); + mutex_unlock(&jfs_log_mutex); return 0; /* @@ -1168,7 +1169,7 @@ journal_found: blkdev_put(bdev); free: /* free log descriptor */ - up(&jfs_log_sem); + mutex_unlock(&jfs_log_mutex); kfree(log); jfs_warn("lmLogOpen: exit(%d)", rc); @@ -1212,11 +1213,11 @@ static int open_dummy_log(struct super_block *sb) { int rc; - down(&jfs_log_sem); + mutex_lock(&jfs_log_mutex); if (!dummy_log) { dummy_log = kzalloc(sizeof(struct jfs_log), GFP_KERNEL); if (!dummy_log) { - up(&jfs_log_sem); + mutex_unlock(&jfs_log_mutex); return -ENOMEM; } INIT_LIST_HEAD(&dummy_log->sb_list); @@ -1229,7 +1230,7 @@ static int open_dummy_log(struct super_block *sb) if (rc) { kfree(dummy_log); dummy_log = NULL; - up(&jfs_log_sem); + mutex_unlock(&jfs_log_mutex); return rc; } } @@ -1238,7 +1239,7 @@ static int open_dummy_log(struct super_block *sb) list_add(&JFS_SBI(sb)->log_list, &dummy_log->sb_list); JFS_SBI(sb)->log = dummy_log; LOG_UNLOCK(dummy_log); - up(&jfs_log_sem); + mutex_unlock(&jfs_log_mutex); return 0; } @@ -1466,7 +1467,7 @@ int lmLogClose(struct super_block *sb) jfs_info("lmLogClose: log:0x%p", log); - down(&jfs_log_sem); + mutex_lock(&jfs_log_mutex); LOG_LOCK(log); list_del(&sbi->log_list); LOG_UNLOCK(log); @@ -1516,7 +1517,7 @@ int lmLogClose(struct super_block *sb) kfree(log); out: - up(&jfs_log_sem); + mutex_unlock(&jfs_log_mutex); jfs_info("lmLogClose: exit(%d)", rc); return rc; } diff --git a/fs/jfs/jfs_metapage.c b/fs/jfs/jfs_metapage.c index 5fbaeaadccd..f28696f235c 100644 --- a/fs/jfs/jfs_metapage.c +++ b/fs/jfs/jfs_metapage.c @@ -220,8 +220,8 @@ int __init metapage_init(void) if (metapage_cache == NULL) return -ENOMEM; - metapage_mempool = mempool_create(METAPOOL_MIN_PAGES, mempool_alloc_slab, - mempool_free_slab, metapage_cache); + metapage_mempool = mempool_create_slab_pool(METAPOOL_MIN_PAGES, + metapage_cache); if (metapage_mempool == NULL) { kmem_cache_destroy(metapage_cache); @@ -578,14 +578,13 @@ static int metapage_releasepage(struct page *page, gfp_t gfp_mask) return 0; } -static int metapage_invalidatepage(struct page *page, unsigned long offset) +static void metapage_invalidatepage(struct page *page, unsigned long offset) { BUG_ON(offset); - if (PageWriteback(page)) - return 0; + BUG_ON(PageWriteback(page)); - return metapage_releasepage(page, 0); + metapage_releasepage(page, 0); } struct address_space_operations jfs_metapage_aops = { diff --git a/fs/jfs/namei.c b/fs/jfs/namei.c index 309cee575f7..09ea03f6227 100644 --- a/fs/jfs/namei.c +++ b/fs/jfs/namei.c @@ -1519,7 +1519,7 @@ struct inode_operations jfs_dir_inode_operations = { #endif }; -struct file_operations jfs_dir_operations = { +const struct file_operations jfs_dir_operations = { .read = generic_read_dir, .readdir = jfs_readdir, .fsync = jfs_fsync, diff --git a/fs/libfs.c b/fs/libfs.c index 4fdeaceb892..7145ba7a48d 100644 --- a/fs/libfs.c +++ b/fs/libfs.c @@ -179,7 +179,7 @@ ssize_t generic_read_dir(struct file *filp, char __user *buf, size_t siz, loff_t return -EISDIR; } -struct file_operations simple_dir_operations = { +const struct file_operations simple_dir_operations = { .open = dcache_dir_open, .release = dcache_dir_close, .llseek = dcache_dir_lseek, diff --git a/fs/lockd/clntlock.c b/fs/lockd/clntlock.c index da6354baa0b..bce74446870 100644 --- a/fs/lockd/clntlock.c +++ b/fs/lockd/clntlock.c @@ -44,32 +44,25 @@ static LIST_HEAD(nlm_blocked); /* * Queue up a lock for blocking so that the GRANTED request can see it */ -int nlmclnt_prepare_block(struct nlm_rqst *req, struct nlm_host *host, struct file_lock *fl) +struct nlm_wait *nlmclnt_prepare_block(struct nlm_host *host, struct file_lock *fl) { struct nlm_wait *block; - BUG_ON(req->a_block != NULL); block = kmalloc(sizeof(*block), GFP_KERNEL); - if (block == NULL) - return -ENOMEM; - block->b_host = host; - block->b_lock = fl; - init_waitqueue_head(&block->b_wait); - block->b_status = NLM_LCK_BLOCKED; - - list_add(&block->b_list, &nlm_blocked); - req->a_block = block; - - return 0; + if (block != NULL) { + block->b_host = host; + block->b_lock = fl; + init_waitqueue_head(&block->b_wait); + block->b_status = NLM_LCK_BLOCKED; + list_add(&block->b_list, &nlm_blocked); + } + return block; } -void nlmclnt_finish_block(struct nlm_rqst *req) +void nlmclnt_finish_block(struct nlm_wait *block) { - struct nlm_wait *block = req->a_block; - if (block == NULL) return; - req->a_block = NULL; list_del(&block->b_list); kfree(block); } @@ -77,15 +70,14 @@ void nlmclnt_finish_block(struct nlm_rqst *req) /* * Block on a lock */ -long nlmclnt_block(struct nlm_rqst *req, long timeout) +int nlmclnt_block(struct nlm_wait *block, struct nlm_rqst *req, long timeout) { - struct nlm_wait *block = req->a_block; long ret; /* A borken server might ask us to block even if we didn't * request it. Just say no! */ - if (!req->a_args.block) + if (block == NULL) return -EAGAIN; /* Go to sleep waiting for GRANT callback. Some servers seem @@ -99,13 +91,10 @@ long nlmclnt_block(struct nlm_rqst *req, long timeout) ret = wait_event_interruptible_timeout(block->b_wait, block->b_status != NLM_LCK_BLOCKED, timeout); - - if (block->b_status != NLM_LCK_BLOCKED) { - req->a_res.status = block->b_status; - block->b_status = NLM_LCK_BLOCKED; - } - - return ret; + if (ret < 0) + return -ERESTARTSYS; + req->a_res.status = block->b_status; + return 0; } /* @@ -125,7 +114,15 @@ u32 nlmclnt_grant(const struct sockaddr_in *addr, const struct nlm_lock *lock) list_for_each_entry(block, &nlm_blocked, b_list) { struct file_lock *fl_blocked = block->b_lock; - if (!nlm_compare_locks(fl_blocked, fl)) + if (fl_blocked->fl_start != fl->fl_start) + continue; + if (fl_blocked->fl_end != fl->fl_end) + continue; + /* + * Careful! The NLM server will return the 32-bit "pid" that + * we put on the wire: in this case the lockowner "pid". + */ + if (fl_blocked->fl_u.nfs_fl.owner->pid != lock->svid) continue; if (!nlm_cmp_addr(&block->b_host->h_addr, addr)) continue; @@ -147,34 +144,6 @@ u32 nlmclnt_grant(const struct sockaddr_in *addr, const struct nlm_lock *lock) */ /* - * Mark the locks for reclaiming. - * FIXME: In 2.5 we don't want to iterate through any global file_lock_list. - * Maintain NLM lock reclaiming lists in the nlm_host instead. - */ -static -void nlmclnt_mark_reclaim(struct nlm_host *host) -{ - struct file_lock *fl; - struct inode *inode; - struct list_head *tmp; - - list_for_each(tmp, &file_lock_list) { - fl = list_entry(tmp, struct file_lock, fl_link); - - inode = fl->fl_file->f_dentry->d_inode; - if (inode->i_sb->s_magic != NFS_SUPER_MAGIC) - continue; - if (fl->fl_u.nfs_fl.owner == NULL) - continue; - if (fl->fl_u.nfs_fl.owner->host != host) - continue; - if (!(fl->fl_u.nfs_fl.flags & NFS_LCK_GRANTED)) - continue; - fl->fl_u.nfs_fl.flags |= NFS_LCK_RECLAIM; - } -} - -/* * Someone has sent us an SM_NOTIFY. Ensure we bind to the new port number, * that we mark locks for reclaiming, and that we bump the pseudo NSM state. */ @@ -186,7 +155,12 @@ void nlmclnt_prepare_reclaim(struct nlm_host *host, u32 newstate) host->h_state++; host->h_nextrebind = 0; nlm_rebind_host(host); - nlmclnt_mark_reclaim(host); + + /* + * Mark the locks for reclaiming. + */ + list_splice_init(&host->h_granted, &host->h_reclaim); + dprintk("NLM: reclaiming locks for host %s", host->h_name); } @@ -215,9 +189,7 @@ reclaimer(void *ptr) { struct nlm_host *host = (struct nlm_host *) ptr; struct nlm_wait *block; - struct list_head *tmp; - struct file_lock *fl; - struct inode *inode; + struct file_lock *fl, *next; daemonize("%s-reclaim", host->h_name); allow_signal(SIGKILL); @@ -229,23 +201,13 @@ reclaimer(void *ptr) /* First, reclaim all locks that have been marked. */ restart: - list_for_each(tmp, &file_lock_list) { - fl = list_entry(tmp, struct file_lock, fl_link); + list_for_each_entry_safe(fl, next, &host->h_reclaim, fl_u.nfs_fl.list) { + list_del_init(&fl->fl_u.nfs_fl.list); - inode = fl->fl_file->f_dentry->d_inode; - if (inode->i_sb->s_magic != NFS_SUPER_MAGIC) - continue; - if (fl->fl_u.nfs_fl.owner == NULL) - continue; - if (fl->fl_u.nfs_fl.owner->host != host) - continue; - if (!(fl->fl_u.nfs_fl.flags & NFS_LCK_RECLAIM)) - continue; - - fl->fl_u.nfs_fl.flags &= ~NFS_LCK_RECLAIM; - nlmclnt_reclaim(host, fl); if (signalled()) - break; + continue; + if (nlmclnt_reclaim(host, fl) == 0) + list_add_tail(&fl->fl_u.nfs_fl.list, &host->h_granted); goto restart; } diff --git a/fs/lockd/clntproc.c b/fs/lockd/clntproc.c index 970b6a6aa33..f96e38155b5 100644 --- a/fs/lockd/clntproc.c +++ b/fs/lockd/clntproc.c @@ -132,59 +132,18 @@ static void nlmclnt_setlockargs(struct nlm_rqst *req, struct file_lock *fl) memcpy(&lock->fh, NFS_FH(fl->fl_file->f_dentry->d_inode), sizeof(struct nfs_fh)); lock->caller = system_utsname.nodename; lock->oh.data = req->a_owner; - lock->oh.len = sprintf(req->a_owner, "%d@%s", - current->pid, system_utsname.nodename); - locks_copy_lock(&lock->fl, fl); + lock->oh.len = snprintf(req->a_owner, sizeof(req->a_owner), "%u@%s", + (unsigned int)fl->fl_u.nfs_fl.owner->pid, + system_utsname.nodename); + lock->svid = fl->fl_u.nfs_fl.owner->pid; + lock->fl.fl_start = fl->fl_start; + lock->fl.fl_end = fl->fl_end; + lock->fl.fl_type = fl->fl_type; } static void nlmclnt_release_lockargs(struct nlm_rqst *req) { - struct file_lock *fl = &req->a_args.lock.fl; - - if (fl->fl_ops && fl->fl_ops->fl_release_private) - fl->fl_ops->fl_release_private(fl); -} - -/* - * Initialize arguments for GRANTED call. The nlm_rqst structure - * has been cleared already. - */ -int -nlmclnt_setgrantargs(struct nlm_rqst *call, struct nlm_lock *lock) -{ - locks_copy_lock(&call->a_args.lock.fl, &lock->fl); - memcpy(&call->a_args.lock.fh, &lock->fh, sizeof(call->a_args.lock.fh)); - call->a_args.lock.caller = system_utsname.nodename; - call->a_args.lock.oh.len = lock->oh.len; - - /* set default data area */ - call->a_args.lock.oh.data = call->a_owner; - - if (lock->oh.len > NLMCLNT_OHSIZE) { - void *data = kmalloc(lock->oh.len, GFP_KERNEL); - if (!data) { - nlmclnt_freegrantargs(call); - return 0; - } - call->a_args.lock.oh.data = (u8 *) data; - } - - memcpy(call->a_args.lock.oh.data, lock->oh.data, lock->oh.len); - return 1; -} - -void -nlmclnt_freegrantargs(struct nlm_rqst *call) -{ - struct file_lock *fl = &call->a_args.lock.fl; - /* - * Check whether we allocated memory for the owner. - */ - if (call->a_args.lock.oh.data != (u8 *) call->a_owner) { - kfree(call->a_args.lock.oh.data); - } - if (fl->fl_ops && fl->fl_ops->fl_release_private) - fl->fl_ops->fl_release_private(fl); + BUG_ON(req->a_args.lock.fl.fl_ops != NULL); } /* @@ -193,9 +152,8 @@ nlmclnt_freegrantargs(struct nlm_rqst *call) int nlmclnt_proc(struct inode *inode, int cmd, struct file_lock *fl) { - struct nfs_server *nfssrv = NFS_SERVER(inode); struct nlm_host *host; - struct nlm_rqst reqst, *call = &reqst; + struct nlm_rqst *call; sigset_t oldset; unsigned long flags; int status, proto, vers; @@ -209,23 +167,17 @@ nlmclnt_proc(struct inode *inode, int cmd, struct file_lock *fl) /* Retrieve transport protocol from NFS client */ proto = NFS_CLIENT(inode)->cl_xprt->prot; - if (!(host = nlmclnt_lookup_host(NFS_ADDR(inode), proto, vers))) + host = nlmclnt_lookup_host(NFS_ADDR(inode), proto, vers); + if (host == NULL) return -ENOLCK; - /* Create RPC client handle if not there, and copy soft - * and intr flags from NFS client. */ - if (host->h_rpcclnt == NULL) { - struct rpc_clnt *clnt; + call = nlm_alloc_call(host); + if (call == NULL) + return -ENOMEM; - /* Bind an rpc client to this host handle (does not - * perform a portmapper lookup) */ - if (!(clnt = nlm_bind_host(host))) { - status = -ENOLCK; - goto done; - } - clnt->cl_softrtry = nfssrv->client->cl_softrtry; - clnt->cl_intr = nfssrv->client->cl_intr; - } + nlmclnt_locks_init_private(fl, host); + /* Set up the argument struct */ + nlmclnt_setlockargs(call, fl); /* Keep the old signal mask */ spin_lock_irqsave(¤t->sighand->siglock, flags); @@ -238,26 +190,10 @@ nlmclnt_proc(struct inode *inode, int cmd, struct file_lock *fl) && (current->flags & PF_EXITING)) { sigfillset(¤t->blocked); /* Mask all signals */ recalc_sigpending(); - spin_unlock_irqrestore(¤t->sighand->siglock, flags); - call = nlmclnt_alloc_call(); - if (!call) { - status = -ENOMEM; - goto out_restore; - } call->a_flags = RPC_TASK_ASYNC; - } else { - spin_unlock_irqrestore(¤t->sighand->siglock, flags); - memset(call, 0, sizeof(*call)); - locks_init_lock(&call->a_args.lock.fl); - locks_init_lock(&call->a_res.lock.fl); } - call->a_host = host; - - nlmclnt_locks_init_private(fl, host); - - /* Set up the argument struct */ - nlmclnt_setlockargs(call, fl); + spin_unlock_irqrestore(¤t->sighand->siglock, flags); if (IS_SETLK(cmd) || IS_SETLKW(cmd)) { if (fl->fl_type != F_UNLCK) { @@ -270,41 +206,58 @@ nlmclnt_proc(struct inode *inode, int cmd, struct file_lock *fl) else status = -EINVAL; - out_restore: + fl->fl_ops->fl_release_private(fl); + fl->fl_ops = NULL; + spin_lock_irqsave(¤t->sighand->siglock, flags); current->blocked = oldset; recalc_sigpending(); spin_unlock_irqrestore(¤t->sighand->siglock, flags); -done: dprintk("lockd: clnt proc returns %d\n", status); - nlm_release_host(host); return status; } EXPORT_SYMBOL(nlmclnt_proc); /* * Allocate an NLM RPC call struct + * + * Note: the caller must hold a reference to host. In case of failure, + * this reference will be released. */ -struct nlm_rqst * -nlmclnt_alloc_call(void) +struct nlm_rqst *nlm_alloc_call(struct nlm_host *host) { struct nlm_rqst *call; - while (!signalled()) { - call = (struct nlm_rqst *) kmalloc(sizeof(struct nlm_rqst), GFP_KERNEL); - if (call) { - memset(call, 0, sizeof(*call)); + for(;;) { + call = kzalloc(sizeof(*call), GFP_KERNEL); + if (call != NULL) { locks_init_lock(&call->a_args.lock.fl); locks_init_lock(&call->a_res.lock.fl); + call->a_host = host; return call; } - printk("nlmclnt_alloc_call: failed, waiting for memory\n"); + if (signalled()) + break; + printk("nlm_alloc_call: failed, waiting for memory\n"); schedule_timeout_interruptible(5*HZ); } + nlm_release_host(host); return NULL; } +void nlm_release_call(struct nlm_rqst *call) +{ + nlm_release_host(call->a_host); + nlmclnt_release_lockargs(call); + kfree(call); +} + +static void nlmclnt_rpc_release(void *data) +{ + return nlm_release_call(data); +} + static int nlm_wait_on_grace(wait_queue_head_t *queue) { DEFINE_WAIT(wait); @@ -401,57 +354,45 @@ in_grace_period: /* * Generic NLM call, async version. */ -int nlmsvc_async_call(struct nlm_rqst *req, u32 proc, const struct rpc_call_ops *tk_ops) +static int __nlm_async_call(struct nlm_rqst *req, u32 proc, struct rpc_message *msg, const struct rpc_call_ops *tk_ops) { struct nlm_host *host = req->a_host; struct rpc_clnt *clnt; - struct rpc_message msg = { - .rpc_argp = &req->a_args, - .rpc_resp = &req->a_res, - }; - int status; + int status = -ENOLCK; dprintk("lockd: call procedure %d on %s (async)\n", (int)proc, host->h_name); /* If we have no RPC client yet, create one. */ - if ((clnt = nlm_bind_host(host)) == NULL) - return -ENOLCK; - msg.rpc_proc = &clnt->cl_procinfo[proc]; + clnt = nlm_bind_host(host); + if (clnt == NULL) + goto out_err; + msg->rpc_proc = &clnt->cl_procinfo[proc]; /* bootstrap and kick off the async RPC call */ - status = rpc_call_async(clnt, &msg, RPC_TASK_ASYNC, tk_ops, req); - + status = rpc_call_async(clnt, msg, RPC_TASK_ASYNC, tk_ops, req); + if (status == 0) + return 0; +out_err: + nlm_release_call(req); return status; } -static int nlmclnt_async_call(struct nlm_rqst *req, u32 proc, const struct rpc_call_ops *tk_ops) +int nlm_async_call(struct nlm_rqst *req, u32 proc, const struct rpc_call_ops *tk_ops) { - struct nlm_host *host = req->a_host; - struct rpc_clnt *clnt; - struct nlm_args *argp = &req->a_args; - struct nlm_res *resp = &req->a_res; struct rpc_message msg = { - .rpc_argp = argp, - .rpc_resp = resp, + .rpc_argp = &req->a_args, + .rpc_resp = &req->a_res, }; - int status; - - dprintk("lockd: call procedure %d on %s (async)\n", - (int)proc, host->h_name); - - /* If we have no RPC client yet, create one. */ - if ((clnt = nlm_bind_host(host)) == NULL) - return -ENOLCK; - msg.rpc_proc = &clnt->cl_procinfo[proc]; + return __nlm_async_call(req, proc, &msg, tk_ops); +} - /* Increment host refcount */ - nlm_get_host(host); - /* bootstrap and kick off the async RPC call */ - status = rpc_call_async(clnt, &msg, RPC_TASK_ASYNC, tk_ops, req); - if (status < 0) - nlm_release_host(host); - return status; +int nlm_async_reply(struct nlm_rqst *req, u32 proc, const struct rpc_call_ops *tk_ops) +{ + struct rpc_message msg = { + .rpc_argp = &req->a_res, + }; + return __nlm_async_call(req, proc, &msg, tk_ops); } /* @@ -463,36 +404,41 @@ nlmclnt_test(struct nlm_rqst *req, struct file_lock *fl) int status; status = nlmclnt_call(req, NLMPROC_TEST); - nlmclnt_release_lockargs(req); if (status < 0) - return status; + goto out; - status = req->a_res.status; - if (status == NLM_LCK_GRANTED) { - fl->fl_type = F_UNLCK; - } if (status == NLM_LCK_DENIED) { - /* - * Report the conflicting lock back to the application. - */ - locks_copy_lock(fl, &req->a_res.lock.fl); - fl->fl_pid = 0; - } else { - return nlm_stat_to_errno(req->a_res.status); + switch (req->a_res.status) { + case NLM_LCK_GRANTED: + fl->fl_type = F_UNLCK; + break; + case NLM_LCK_DENIED: + /* + * Report the conflicting lock back to the application. + */ + fl->fl_start = req->a_res.lock.fl.fl_start; + fl->fl_end = req->a_res.lock.fl.fl_start; + fl->fl_type = req->a_res.lock.fl.fl_type; + fl->fl_pid = 0; + break; + default: + status = nlm_stat_to_errno(req->a_res.status); } - - return 0; +out: + nlm_release_call(req); + return status; } static void nlmclnt_locks_copy_lock(struct file_lock *new, struct file_lock *fl) { - memcpy(&new->fl_u.nfs_fl, &fl->fl_u.nfs_fl, sizeof(new->fl_u.nfs_fl)); - nlm_get_lockowner(new->fl_u.nfs_fl.owner); + new->fl_u.nfs_fl.state = fl->fl_u.nfs_fl.state; + new->fl_u.nfs_fl.owner = nlm_get_lockowner(fl->fl_u.nfs_fl.owner); + list_add_tail(&new->fl_u.nfs_fl.list, &fl->fl_u.nfs_fl.owner->host->h_granted); } static void nlmclnt_locks_release_private(struct file_lock *fl) { + list_del(&fl->fl_u.nfs_fl.list); nlm_put_lockowner(fl->fl_u.nfs_fl.owner); - fl->fl_ops = NULL; } static struct file_lock_operations nlmclnt_lock_ops = { @@ -504,8 +450,8 @@ static void nlmclnt_locks_init_private(struct file_lock *fl, struct nlm_host *ho { BUG_ON(fl->fl_ops != NULL); fl->fl_u.nfs_fl.state = 0; - fl->fl_u.nfs_fl.flags = 0; fl->fl_u.nfs_fl.owner = nlm_find_lockowner(host, fl->fl_owner); + INIT_LIST_HEAD(&fl->fl_u.nfs_fl.list); fl->fl_ops = &nlmclnt_lock_ops; } @@ -552,57 +498,52 @@ nlmclnt_lock(struct nlm_rqst *req, struct file_lock *fl) { struct nlm_host *host = req->a_host; struct nlm_res *resp = &req->a_res; - long timeout; - int status; + struct nlm_wait *block = NULL; + int status = -ENOLCK; if (!host->h_monitored && nsm_monitor(host) < 0) { printk(KERN_NOTICE "lockd: failed to monitor %s\n", host->h_name); - status = -ENOLCK; goto out; } - if (req->a_args.block) { - status = nlmclnt_prepare_block(req, host, fl); - if (status < 0) - goto out; - } + block = nlmclnt_prepare_block(host, fl); for(;;) { status = nlmclnt_call(req, NLMPROC_LOCK); if (status < 0) goto out_unblock; - if (resp->status != NLM_LCK_BLOCKED) + if (!req->a_args.block) break; - /* Wait on an NLM blocking lock */ - timeout = nlmclnt_block(req, NLMCLNT_POLL_TIMEOUT); /* Did a reclaimer thread notify us of a server reboot? */ if (resp->status == NLM_LCK_DENIED_GRACE_PERIOD) continue; if (resp->status != NLM_LCK_BLOCKED) break; - if (timeout >= 0) - continue; - /* We were interrupted. Send a CANCEL request to the server + /* Wait on an NLM blocking lock */ + status = nlmclnt_block(block, req, NLMCLNT_POLL_TIMEOUT); + /* if we were interrupted. Send a CANCEL request to the server * and exit */ - status = (int)timeout; - goto out_unblock; + if (status < 0) + goto out_unblock; + if (resp->status != NLM_LCK_BLOCKED) + break; } if (resp->status == NLM_LCK_GRANTED) { fl->fl_u.nfs_fl.state = host->h_state; - fl->fl_u.nfs_fl.flags |= NFS_LCK_GRANTED; fl->fl_flags |= FL_SLEEP; + /* Ensure the resulting lock will get added to granted list */ do_vfs_lock(fl); } status = nlm_stat_to_errno(resp->status); out_unblock: - nlmclnt_finish_block(req); + nlmclnt_finish_block(block); /* Cancel the blocked request if it is still pending */ if (resp->status == NLM_LCK_BLOCKED) nlmclnt_cancel(host, req->a_args.block, fl); out: - nlmclnt_release_lockargs(req); + nlm_release_call(req); return status; } @@ -658,10 +599,6 @@ nlmclnt_unlock(struct nlm_rqst *req, struct file_lock *fl) struct nlm_res *resp = &req->a_res; int status; - /* Clean the GRANTED flag now so the lock doesn't get - * reclaimed while we're stuck in the unlock call. */ - fl->fl_u.nfs_fl.flags &= ~NFS_LCK_GRANTED; - /* * Note: the server is supposed to either grant us the unlock * request, or to deny it with NLM_LCK_DENIED_GRACE_PERIOD. In either @@ -669,32 +606,24 @@ nlmclnt_unlock(struct nlm_rqst *req, struct file_lock *fl) */ do_vfs_lock(fl); - if (req->a_flags & RPC_TASK_ASYNC) { - status = nlmclnt_async_call(req, NLMPROC_UNLOCK, - &nlmclnt_unlock_ops); - /* Hrmf... Do the unlock early since locks_remove_posix() - * really expects us to free the lock synchronously */ - if (status < 0) { - nlmclnt_release_lockargs(req); - kfree(req); - } - return status; - } + if (req->a_flags & RPC_TASK_ASYNC) + return nlm_async_call(req, NLMPROC_UNLOCK, &nlmclnt_unlock_ops); status = nlmclnt_call(req, NLMPROC_UNLOCK); - nlmclnt_release_lockargs(req); if (status < 0) - return status; + goto out; + status = 0; if (resp->status == NLM_LCK_GRANTED) - return 0; + goto out; if (resp->status != NLM_LCK_DENIED_NOLOCKS) printk("lockd: unexpected unlock status: %d\n", resp->status); - /* What to do now? I'm out of my depth... */ - - return -ENOLCK; + status = -ENOLCK; +out: + nlm_release_call(req); + return status; } static void nlmclnt_unlock_callback(struct rpc_task *task, void *data) @@ -716,9 +645,6 @@ static void nlmclnt_unlock_callback(struct rpc_task *task, void *data) if (status != NLM_LCK_GRANTED) printk(KERN_WARNING "lockd: unexpected unlock status: %d\n", status); die: - nlm_release_host(req->a_host); - nlmclnt_release_lockargs(req); - kfree(req); return; retry_rebind: nlm_rebind_host(req->a_host); @@ -728,6 +654,7 @@ die: static const struct rpc_call_ops nlmclnt_unlock_ops = { .rpc_call_done = nlmclnt_unlock_callback, + .rpc_release = nlmclnt_rpc_release, }; /* @@ -749,20 +676,15 @@ static int nlmclnt_cancel(struct nlm_host *host, int block, struct file_lock *fl recalc_sigpending(); spin_unlock_irqrestore(¤t->sighand->siglock, flags); - req = nlmclnt_alloc_call(); + req = nlm_alloc_call(nlm_get_host(host)); if (!req) return -ENOMEM; - req->a_host = host; req->a_flags = RPC_TASK_ASYNC; nlmclnt_setlockargs(req, fl); req->a_args.block = block; - status = nlmclnt_async_call(req, NLMPROC_CANCEL, &nlmclnt_cancel_ops); - if (status < 0) { - nlmclnt_release_lockargs(req); - kfree(req); - } + status = nlm_async_call(req, NLMPROC_CANCEL, &nlmclnt_cancel_ops); spin_lock_irqsave(¤t->sighand->siglock, flags); current->blocked = oldset; @@ -791,6 +713,7 @@ static void nlmclnt_cancel_callback(struct rpc_task *task, void *data) switch (req->a_res.status) { case NLM_LCK_GRANTED: case NLM_LCK_DENIED_GRACE_PERIOD: + case NLM_LCK_DENIED: /* Everything's good */ break; case NLM_LCK_DENIED_NOLOCKS: @@ -802,9 +725,6 @@ static void nlmclnt_cancel_callback(struct rpc_task *task, void *data) } die: - nlm_release_host(req->a_host); - nlmclnt_release_lockargs(req); - kfree(req); return; retry_cancel: @@ -818,6 +738,7 @@ retry_cancel: static const struct rpc_call_ops nlmclnt_cancel_ops = { .rpc_call_done = nlmclnt_cancel_callback, + .rpc_release = nlmclnt_rpc_release, }; /* diff --git a/fs/lockd/host.c b/fs/lockd/host.c index 82f7a0b1d8a..729ac427d35 100644 --- a/fs/lockd/host.c +++ b/fs/lockd/host.c @@ -16,6 +16,7 @@ #include <linux/sunrpc/svc.h> #include <linux/lockd/lockd.h> #include <linux/lockd/sm_inter.h> +#include <linux/mutex.h> #define NLMDBG_FACILITY NLMDBG_HOSTCACHE @@ -30,7 +31,7 @@ static struct nlm_host * nlm_hosts[NLM_HOST_NRHASH]; static unsigned long next_gc; static int nrhosts; -static DECLARE_MUTEX(nlm_host_sema); +static DEFINE_MUTEX(nlm_host_mutex); static void nlm_gc_hosts(void); @@ -71,7 +72,7 @@ nlm_lookup_host(int server, struct sockaddr_in *sin, hash = NLM_ADDRHASH(sin->sin_addr.s_addr); /* Lock hash table */ - down(&nlm_host_sema); + mutex_lock(&nlm_host_mutex); if (time_after_eq(jiffies, next_gc)) nlm_gc_hosts(); @@ -91,7 +92,7 @@ nlm_lookup_host(int server, struct sockaddr_in *sin, nlm_hosts[hash] = host; } nlm_get_host(host); - up(&nlm_host_sema); + mutex_unlock(&nlm_host_mutex); return host; } } @@ -123,12 +124,14 @@ nlm_lookup_host(int server, struct sockaddr_in *sin, nlm_hosts[hash] = host; INIT_LIST_HEAD(&host->h_lockowners); spin_lock_init(&host->h_lock); + INIT_LIST_HEAD(&host->h_granted); + INIT_LIST_HEAD(&host->h_reclaim); if (++nrhosts > NLM_HOST_MAX) next_gc = 0; nohost: - up(&nlm_host_sema); + mutex_unlock(&nlm_host_mutex); return host; } @@ -139,19 +142,19 @@ nlm_find_client(void) * and return it */ int hash; - down(&nlm_host_sema); + mutex_lock(&nlm_host_mutex); for (hash = 0 ; hash < NLM_HOST_NRHASH; hash++) { struct nlm_host *host, **hp; for (hp = &nlm_hosts[hash]; (host = *hp) != 0; hp = &host->h_next) { if (host->h_server && host->h_killed == 0) { nlm_get_host(host); - up(&nlm_host_sema); + mutex_unlock(&nlm_host_mutex); return host; } } } - up(&nlm_host_sema); + mutex_unlock(&nlm_host_mutex); return NULL; } @@ -191,11 +194,12 @@ nlm_bind_host(struct nlm_host *host) xprt->resvport = 1; /* NLM requires a reserved port */ /* Existing NLM servers accept AUTH_UNIX only */ - clnt = rpc_create_client(xprt, host->h_name, &nlm_program, + clnt = rpc_new_client(xprt, host->h_name, &nlm_program, host->h_version, RPC_AUTH_UNIX); if (IS_ERR(clnt)) goto forgetit; clnt->cl_autobind = 1; /* turn on pmap queries */ + clnt->cl_softrtry = 1; /* All queries are soft */ host->h_rpcclnt = clnt; } @@ -242,8 +246,12 @@ void nlm_release_host(struct nlm_host *host) { if (host != NULL) { dprintk("lockd: release host %s\n", host->h_name); - atomic_dec(&host->h_count); BUG_ON(atomic_read(&host->h_count) < 0); + if (atomic_dec_and_test(&host->h_count)) { + BUG_ON(!list_empty(&host->h_lockowners)); + BUG_ON(!list_empty(&host->h_granted)); + BUG_ON(!list_empty(&host->h_reclaim)); + } } } @@ -258,7 +266,7 @@ nlm_shutdown_hosts(void) int i; dprintk("lockd: shutting down host module\n"); - down(&nlm_host_sema); + mutex_lock(&nlm_host_mutex); /* First, make all hosts eligible for gc */ dprintk("lockd: nuking all hosts...\n"); @@ -269,7 +277,7 @@ nlm_shutdown_hosts(void) /* Then, perform a garbage collection pass */ nlm_gc_hosts(); - up(&nlm_host_sema); + mutex_unlock(&nlm_host_mutex); /* complain if any hosts are left */ if (nrhosts) { @@ -331,7 +339,6 @@ nlm_gc_hosts(void) rpc_destroy_client(host->h_rpcclnt); } } - BUG_ON(!list_empty(&host->h_lockowners)); kfree(host); nrhosts--; } diff --git a/fs/lockd/mon.c b/fs/lockd/mon.c index a89cb8aa2c8..3fc683f46b3 100644 --- a/fs/lockd/mon.c +++ b/fs/lockd/mon.c @@ -35,6 +35,10 @@ nsm_mon_unmon(struct nlm_host *host, u32 proc, struct nsm_res *res) struct rpc_clnt *clnt; int status; struct nsm_args args; + struct rpc_message msg = { + .rpc_argp = &args, + .rpc_resp = res, + }; clnt = nsm_create(); if (IS_ERR(clnt)) { @@ -49,7 +53,8 @@ nsm_mon_unmon(struct nlm_host *host, u32 proc, struct nsm_res *res) args.proc = NLMPROC_NSM_NOTIFY; memset(res, 0, sizeof(*res)); - status = rpc_call(clnt, proc, &args, res, 0); + msg.rpc_proc = &clnt->cl_procinfo[proc]; + status = rpc_call_sync(clnt, &msg, 0); if (status < 0) printk(KERN_DEBUG "nsm_mon_unmon: rpc failed, status=%d\n", status); @@ -214,12 +219,16 @@ static struct rpc_procinfo nsm_procedures[] = { .p_encode = (kxdrproc_t) xdr_encode_mon, .p_decode = (kxdrproc_t) xdr_decode_stat_res, .p_bufsiz = MAX(SM_mon_sz, SM_monres_sz) << 2, + .p_statidx = SM_MON, + .p_name = "MONITOR", }, [SM_UNMON] = { .p_proc = SM_UNMON, .p_encode = (kxdrproc_t) xdr_encode_unmon, .p_decode = (kxdrproc_t) xdr_decode_stat, .p_bufsiz = MAX(SM_mon_id_sz, SM_unmonres_sz) << 2, + .p_statidx = SM_UNMON, + .p_name = "UNMONITOR", }, }; diff --git a/fs/lockd/svc.c b/fs/lockd/svc.c index 5e85bde6c12..fd56c8872f3 100644 --- a/fs/lockd/svc.c +++ b/fs/lockd/svc.c @@ -25,6 +25,7 @@ #include <linux/slab.h> #include <linux/smp.h> #include <linux/smp_lock.h> +#include <linux/mutex.h> #include <linux/sunrpc/types.h> #include <linux/sunrpc/stats.h> @@ -43,13 +44,13 @@ static struct svc_program nlmsvc_program; struct nlmsvc_binding * nlmsvc_ops; EXPORT_SYMBOL(nlmsvc_ops); -static DECLARE_MUTEX(nlmsvc_sema); +static DEFINE_MUTEX(nlmsvc_mutex); static unsigned int nlmsvc_users; static pid_t nlmsvc_pid; int nlmsvc_grace_period; unsigned long nlmsvc_timeout; -static DECLARE_MUTEX_LOCKED(lockd_start); +static DECLARE_COMPLETION(lockd_start_done); static DECLARE_WAIT_QUEUE_HEAD(lockd_exit); /* @@ -112,7 +113,7 @@ lockd(struct svc_rqst *rqstp) * Let our maker know we're running. */ nlmsvc_pid = current->pid; - up(&lockd_start); + complete(&lockd_start_done); daemonize("lockd"); @@ -215,7 +216,7 @@ lockd_up(void) struct svc_serv * serv; int error = 0; - down(&nlmsvc_sema); + mutex_lock(&nlmsvc_mutex); /* * Unconditionally increment the user count ... this is * the number of clients who _want_ a lockd process. @@ -263,7 +264,7 @@ lockd_up(void) "lockd_up: create thread failed, error=%d\n", error); goto destroy_and_out; } - down(&lockd_start); + wait_for_completion(&lockd_start_done); /* * Note: svc_serv structures have an initial use count of 1, @@ -272,7 +273,7 @@ lockd_up(void) destroy_and_out: svc_destroy(serv); out: - up(&nlmsvc_sema); + mutex_unlock(&nlmsvc_mutex); return error; } EXPORT_SYMBOL(lockd_up); @@ -285,7 +286,7 @@ lockd_down(void) { static int warned; - down(&nlmsvc_sema); + mutex_lock(&nlmsvc_mutex); if (nlmsvc_users) { if (--nlmsvc_users) goto out; @@ -315,7 +316,7 @@ lockd_down(void) recalc_sigpending(); spin_unlock_irq(¤t->sighand->siglock); out: - up(&nlmsvc_sema); + mutex_unlock(&nlmsvc_mutex); } EXPORT_SYMBOL(lockd_down); diff --git a/fs/lockd/svc4proc.c b/fs/lockd/svc4proc.c index b10f913aa06..a2dd9ccb9b3 100644 --- a/fs/lockd/svc4proc.c +++ b/fs/lockd/svc4proc.c @@ -21,10 +21,6 @@ #define NLMDBG_FACILITY NLMDBG_CLIENT -static u32 nlm4svc_callback(struct svc_rqst *, u32, struct nlm_res *); - -static const struct rpc_call_ops nlm4svc_callback_ops; - /* * Obtain client and file from arguments */ @@ -234,83 +230,89 @@ nlm4svc_proc_granted(struct svc_rqst *rqstp, struct nlm_args *argp, } /* + * This is the generic lockd callback for async RPC calls + */ +static void nlm4svc_callback_exit(struct rpc_task *task, void *data) +{ + dprintk("lockd: %4d callback returned %d\n", task->tk_pid, + -task->tk_status); +} + +static void nlm4svc_callback_release(void *data) +{ + nlm_release_call(data); +} + +static const struct rpc_call_ops nlm4svc_callback_ops = { + .rpc_call_done = nlm4svc_callback_exit, + .rpc_release = nlm4svc_callback_release, +}; + +/* * `Async' versions of the above service routines. They aren't really, * because we send the callback before the reply proper. I hope this * doesn't break any clients. */ -static int -nlm4svc_proc_test_msg(struct svc_rqst *rqstp, struct nlm_args *argp, - void *resp) +static int nlm4svc_callback(struct svc_rqst *rqstp, u32 proc, struct nlm_args *argp, + int (*func)(struct svc_rqst *, struct nlm_args *, struct nlm_res *)) { - struct nlm_res res; - u32 stat; + struct nlm_host *host; + struct nlm_rqst *call; + int stat; - dprintk("lockd: TEST_MSG called\n"); - memset(&res, 0, sizeof(res)); + host = nlmsvc_lookup_host(rqstp); + if (host == NULL) + return rpc_system_err; + + call = nlm_alloc_call(host); + if (call == NULL) + return rpc_system_err; - if ((stat = nlm4svc_proc_test(rqstp, argp, &res)) == 0) - stat = nlm4svc_callback(rqstp, NLMPROC_TEST_RES, &res); - return stat; + stat = func(rqstp, argp, &call->a_res); + if (stat != 0) { + nlm_release_call(call); + return stat; + } + + call->a_flags = RPC_TASK_ASYNC; + if (nlm_async_reply(call, proc, &nlm4svc_callback_ops) < 0) + return rpc_system_err; + return rpc_success; } -static int -nlm4svc_proc_lock_msg(struct svc_rqst *rqstp, struct nlm_args *argp, +static int nlm4svc_proc_test_msg(struct svc_rqst *rqstp, struct nlm_args *argp, void *resp) { - struct nlm_res res; - u32 stat; + dprintk("lockd: TEST_MSG called\n"); + return nlm4svc_callback(rqstp, NLMPROC_TEST_RES, argp, nlm4svc_proc_test); +} +static int nlm4svc_proc_lock_msg(struct svc_rqst *rqstp, struct nlm_args *argp, + void *resp) +{ dprintk("lockd: LOCK_MSG called\n"); - memset(&res, 0, sizeof(res)); - - if ((stat = nlm4svc_proc_lock(rqstp, argp, &res)) == 0) - stat = nlm4svc_callback(rqstp, NLMPROC_LOCK_RES, &res); - return stat; + return nlm4svc_callback(rqstp, NLMPROC_LOCK_RES, argp, nlm4svc_proc_lock); } -static int -nlm4svc_proc_cancel_msg(struct svc_rqst *rqstp, struct nlm_args *argp, +static int nlm4svc_proc_cancel_msg(struct svc_rqst *rqstp, struct nlm_args *argp, void *resp) { - struct nlm_res res; - u32 stat; - dprintk("lockd: CANCEL_MSG called\n"); - memset(&res, 0, sizeof(res)); - - if ((stat = nlm4svc_proc_cancel(rqstp, argp, &res)) == 0) - stat = nlm4svc_callback(rqstp, NLMPROC_CANCEL_RES, &res); - return stat; + return nlm4svc_callback(rqstp, NLMPROC_CANCEL_RES, argp, nlm4svc_proc_cancel); } -static int -nlm4svc_proc_unlock_msg(struct svc_rqst *rqstp, struct nlm_args *argp, +static int nlm4svc_proc_unlock_msg(struct svc_rqst *rqstp, struct nlm_args *argp, void *resp) { - struct nlm_res res; - u32 stat; - dprintk("lockd: UNLOCK_MSG called\n"); - memset(&res, 0, sizeof(res)); - - if ((stat = nlm4svc_proc_unlock(rqstp, argp, &res)) == 0) - stat = nlm4svc_callback(rqstp, NLMPROC_UNLOCK_RES, &res); - return stat; + return nlm4svc_callback(rqstp, NLMPROC_UNLOCK_RES, argp, nlm4svc_proc_unlock); } -static int -nlm4svc_proc_granted_msg(struct svc_rqst *rqstp, struct nlm_args *argp, +static int nlm4svc_proc_granted_msg(struct svc_rqst *rqstp, struct nlm_args *argp, void *resp) { - struct nlm_res res; - u32 stat; - dprintk("lockd: GRANTED_MSG called\n"); - memset(&res, 0, sizeof(res)); - - if ((stat = nlm4svc_proc_granted(rqstp, argp, &res)) == 0) - stat = nlm4svc_callback(rqstp, NLMPROC_GRANTED_RES, &res); - return stat; + return nlm4svc_callback(rqstp, NLMPROC_GRANTED_RES, argp, nlm4svc_proc_granted); } /* @@ -472,55 +474,6 @@ nlm4svc_proc_granted_res(struct svc_rqst *rqstp, struct nlm_res *argp, /* - * This is the generic lockd callback for async RPC calls - */ -static u32 -nlm4svc_callback(struct svc_rqst *rqstp, u32 proc, struct nlm_res *resp) -{ - struct nlm_host *host; - struct nlm_rqst *call; - - if (!(call = nlmclnt_alloc_call())) - return rpc_system_err; - - host = nlmclnt_lookup_host(&rqstp->rq_addr, - rqstp->rq_prot, rqstp->rq_vers); - if (!host) { - kfree(call); - return rpc_system_err; - } - - call->a_flags = RPC_TASK_ASYNC; - call->a_host = host; - memcpy(&call->a_args, resp, sizeof(*resp)); - - if (nlmsvc_async_call(call, proc, &nlm4svc_callback_ops) < 0) - goto error; - - return rpc_success; - error: - kfree(call); - nlm_release_host(host); - return rpc_system_err; -} - -static void nlm4svc_callback_exit(struct rpc_task *task, void *data) -{ - struct nlm_rqst *call = data; - - if (task->tk_status < 0) { - dprintk("lockd: %4d callback failed (errno = %d)\n", - task->tk_pid, -task->tk_status); - } - nlm_release_host(call->a_host); - kfree(call); -} - -static const struct rpc_call_ops nlm4svc_callback_ops = { - .rpc_call_done = nlm4svc_callback_exit, -}; - -/* * NLM Server procedures. */ diff --git a/fs/lockd/svclock.c b/fs/lockd/svclock.c index 9cfced65d4a..d2b66bad7d5 100644 --- a/fs/lockd/svclock.c +++ b/fs/lockd/svclock.c @@ -39,9 +39,12 @@ #define nlm_deadlock nlm_lck_denied #endif +static void nlmsvc_release_block(struct nlm_block *block); static void nlmsvc_insert_block(struct nlm_block *block, unsigned long); static int nlmsvc_remove_block(struct nlm_block *block); +static int nlmsvc_setgrantargs(struct nlm_rqst *call, struct nlm_lock *lock); +static void nlmsvc_freegrantargs(struct nlm_rqst *call); static const struct rpc_call_ops nlmsvc_grant_ops; /* @@ -58,6 +61,7 @@ nlmsvc_insert_block(struct nlm_block *block, unsigned long when) struct nlm_block **bp, *b; dprintk("lockd: nlmsvc_insert_block(%p, %ld)\n", block, when); + kref_get(&block->b_count); if (block->b_queued) nlmsvc_remove_block(block); bp = &nlm_blocked; @@ -90,6 +94,7 @@ nlmsvc_remove_block(struct nlm_block *block) if (b == block) { *bp = block->b_next; block->b_queued = 0; + nlmsvc_release_block(block); return 1; } } @@ -98,11 +103,10 @@ nlmsvc_remove_block(struct nlm_block *block) } /* - * Find a block for a given lock and optionally remove it from - * the list. + * Find a block for a given lock */ static struct nlm_block * -nlmsvc_lookup_block(struct nlm_file *file, struct nlm_lock *lock, int remove) +nlmsvc_lookup_block(struct nlm_file *file, struct nlm_lock *lock) { struct nlm_block **head, *block; struct file_lock *fl; @@ -112,17 +116,14 @@ nlmsvc_lookup_block(struct nlm_file *file, struct nlm_lock *lock, int remove) (long long)lock->fl.fl_start, (long long)lock->fl.fl_end, lock->fl.fl_type); for (head = &nlm_blocked; (block = *head) != 0; head = &block->b_next) { - fl = &block->b_call.a_args.lock.fl; + fl = &block->b_call->a_args.lock.fl; dprintk("lockd: check f=%p pd=%d %Ld-%Ld ty=%d cookie=%s\n", block->b_file, fl->fl_pid, (long long)fl->fl_start, (long long)fl->fl_end, fl->fl_type, - nlmdbg_cookie2a(&block->b_call.a_args.cookie)); + nlmdbg_cookie2a(&block->b_call->a_args.cookie)); if (block->b_file == file && nlm_compare_locks(fl, &lock->fl)) { - if (remove) { - *head = block->b_next; - block->b_queued = 0; - } + kref_get(&block->b_count); return block; } } @@ -150,11 +151,13 @@ nlmsvc_find_block(struct nlm_cookie *cookie, struct sockaddr_in *sin) for (block = nlm_blocked; block; block = block->b_next) { dprintk("cookie: head of blocked queue %p, block %p\n", nlm_blocked, block); - if (nlm_cookie_match(&block->b_call.a_args.cookie,cookie) + if (nlm_cookie_match(&block->b_call->a_args.cookie,cookie) && nlm_cmp_addr(sin, &block->b_host->h_addr)) break; } + if (block != NULL) + kref_get(&block->b_count); return block; } @@ -174,27 +177,30 @@ nlmsvc_create_block(struct svc_rqst *rqstp, struct nlm_file *file, { struct nlm_block *block; struct nlm_host *host; - struct nlm_rqst *call; + struct nlm_rqst *call = NULL; /* Create host handle for callback */ - host = nlmclnt_lookup_host(&rqstp->rq_addr, - rqstp->rq_prot, rqstp->rq_vers); + host = nlmsvc_lookup_host(rqstp); if (host == NULL) return NULL; + call = nlm_alloc_call(host); + if (call == NULL) + return NULL; + /* Allocate memory for block, and initialize arguments */ - if (!(block = (struct nlm_block *) kmalloc(sizeof(*block), GFP_KERNEL))) + block = kzalloc(sizeof(*block), GFP_KERNEL); + if (block == NULL) goto failed; - memset(block, 0, sizeof(*block)); - locks_init_lock(&block->b_call.a_args.lock.fl); - locks_init_lock(&block->b_call.a_res.lock.fl); + kref_init(&block->b_count); - if (!nlmclnt_setgrantargs(&block->b_call, lock)) + if (!nlmsvc_setgrantargs(call, lock)) goto failed_free; /* Set notifier function for VFS, and init args */ - block->b_call.a_args.lock.fl.fl_lmops = &nlmsvc_lock_operations; - block->b_call.a_args.cookie = *cookie; /* see above */ + call->a_args.lock.fl.fl_flags |= FL_SLEEP; + call->a_args.lock.fl.fl_lmops = &nlmsvc_lock_operations; + call->a_args.cookie = *cookie; /* see above */ dprintk("lockd: created block %p...\n", block); @@ -202,22 +208,23 @@ nlmsvc_create_block(struct svc_rqst *rqstp, struct nlm_file *file, block->b_daemon = rqstp->rq_server; block->b_host = host; block->b_file = file; + file->f_count++; /* Add to file's list of blocks */ block->b_fnext = file->f_blocks; file->f_blocks = block; /* Set up RPC arguments for callback */ - call = &block->b_call; - call->a_host = host; + block->b_call = call; call->a_flags = RPC_TASK_ASYNC; + call->a_block = block; return block; failed_free: kfree(block); failed: - nlm_release_host(host); + nlm_release_call(call); return NULL; } @@ -227,29 +234,26 @@ failed: * It is the caller's responsibility to check whether the file * can be closed hereafter. */ -static int -nlmsvc_delete_block(struct nlm_block *block, int unlock) +static int nlmsvc_unlink_block(struct nlm_block *block) { - struct file_lock *fl = &block->b_call.a_args.lock.fl; - struct nlm_file *file = block->b_file; - struct nlm_block **bp; - int status = 0; - - dprintk("lockd: deleting block %p...\n", block); + int status; + dprintk("lockd: unlinking block %p...\n", block); /* Remove block from list */ + status = posix_unblock_lock(block->b_file->f_file, &block->b_call->a_args.lock.fl); nlmsvc_remove_block(block); - if (unlock) - status = posix_unblock_lock(file->f_file, fl); + return status; +} - /* If the block is in the middle of a GRANT callback, - * don't kill it yet. */ - if (block->b_incall) { - nlmsvc_insert_block(block, NLM_NEVER); - block->b_done = 1; - return status; - } +static void nlmsvc_free_block(struct kref *kref) +{ + struct nlm_block *block = container_of(kref, struct nlm_block, b_count); + struct nlm_file *file = block->b_file; + struct nlm_block **bp; + dprintk("lockd: freeing block %p...\n", block); + + down(&file->f_sema); /* Remove block from file's list of blocks */ for (bp = &file->f_blocks; *bp; bp = &(*bp)->b_fnext) { if (*bp == block) { @@ -257,36 +261,93 @@ nlmsvc_delete_block(struct nlm_block *block, int unlock) break; } } + up(&file->f_sema); - if (block->b_host) - nlm_release_host(block->b_host); - nlmclnt_freegrantargs(&block->b_call); + nlmsvc_freegrantargs(block->b_call); + nlm_release_call(block->b_call); + nlm_release_file(block->b_file); kfree(block); - return status; +} + +static void nlmsvc_release_block(struct nlm_block *block) +{ + if (block != NULL) + kref_put(&block->b_count, nlmsvc_free_block); +} + +static void nlmsvc_act_mark(struct nlm_host *host, struct nlm_file *file) +{ + struct nlm_block *block; + + down(&file->f_sema); + for (block = file->f_blocks; block != NULL; block = block->b_fnext) + block->b_host->h_inuse = 1; + up(&file->f_sema); +} + +static void nlmsvc_act_unlock(struct nlm_host *host, struct nlm_file *file) +{ + struct nlm_block *block; + +restart: + down(&file->f_sema); + for (block = file->f_blocks; block != NULL; block = block->b_fnext) { + if (host != NULL && host != block->b_host) + continue; + if (!block->b_queued) + continue; + kref_get(&block->b_count); + up(&file->f_sema); + nlmsvc_unlink_block(block); + nlmsvc_release_block(block); + goto restart; + } + up(&file->f_sema); } /* * Loop over all blocks and perform the action specified. * (NLM_ACT_CHECK handled by nlmsvc_inspect_file). */ -int +void nlmsvc_traverse_blocks(struct nlm_host *host, struct nlm_file *file, int action) { - struct nlm_block *block, *next; - /* XXX: Will everything get cleaned up if we don't unlock here? */ + if (action == NLM_ACT_MARK) + nlmsvc_act_mark(host, file); + else + nlmsvc_act_unlock(host, file); +} - down(&file->f_sema); - for (block = file->f_blocks; block; block = next) { - next = block->b_fnext; - if (action == NLM_ACT_MARK) - block->b_host->h_inuse = 1; - else if (action == NLM_ACT_UNLOCK) { - if (host == NULL || host == block->b_host) - nlmsvc_delete_block(block, 1); - } +/* + * Initialize arguments for GRANTED call. The nlm_rqst structure + * has been cleared already. + */ +static int nlmsvc_setgrantargs(struct nlm_rqst *call, struct nlm_lock *lock) +{ + locks_copy_lock(&call->a_args.lock.fl, &lock->fl); + memcpy(&call->a_args.lock.fh, &lock->fh, sizeof(call->a_args.lock.fh)); + call->a_args.lock.caller = system_utsname.nodename; + call->a_args.lock.oh.len = lock->oh.len; + + /* set default data area */ + call->a_args.lock.oh.data = call->a_owner; + call->a_args.lock.svid = lock->fl.fl_pid; + + if (lock->oh.len > NLMCLNT_OHSIZE) { + void *data = kmalloc(lock->oh.len, GFP_KERNEL); + if (!data) + return 0; + call->a_args.lock.oh.data = (u8 *) data; } - up(&file->f_sema); - return 0; + + memcpy(call->a_args.lock.oh.data, lock->oh.data, lock->oh.len); + return 1; +} + +static void nlmsvc_freegrantargs(struct nlm_rqst *call) +{ + if (call->a_args.lock.oh.data != call->a_owner) + kfree(call->a_args.lock.oh.data); } /* @@ -297,9 +358,9 @@ u32 nlmsvc_lock(struct svc_rqst *rqstp, struct nlm_file *file, struct nlm_lock *lock, int wait, struct nlm_cookie *cookie) { - struct file_lock *conflock; - struct nlm_block *block; + struct nlm_block *block, *newblock = NULL; int error; + u32 ret; dprintk("lockd: nlmsvc_lock(%s/%ld, ty=%d, pi=%d, %Ld-%Ld, bl=%d)\n", file->f_file->f_dentry->d_inode->i_sb->s_id, @@ -310,69 +371,65 @@ nlmsvc_lock(struct svc_rqst *rqstp, struct nlm_file *file, wait); - /* Get existing block (in case client is busy-waiting) */ - block = nlmsvc_lookup_block(file, lock, 0); - - lock->fl.fl_flags |= FL_LOCKD; - + lock->fl.fl_flags &= ~FL_SLEEP; again: /* Lock file against concurrent access */ down(&file->f_sema); + /* Get existing block (in case client is busy-waiting) */ + block = nlmsvc_lookup_block(file, lock); + if (block == NULL) { + if (newblock != NULL) + lock = &newblock->b_call->a_args.lock; + } else + lock = &block->b_call->a_args.lock; - if (!(conflock = posix_test_lock(file->f_file, &lock->fl))) { - error = posix_lock_file(file->f_file, &lock->fl); + error = posix_lock_file(file->f_file, &lock->fl); + lock->fl.fl_flags &= ~FL_SLEEP; - if (block) - nlmsvc_delete_block(block, 0); - up(&file->f_sema); + dprintk("lockd: posix_lock_file returned %d\n", error); - dprintk("lockd: posix_lock_file returned %d\n", -error); - switch(-error) { + switch(error) { case 0: - return nlm_granted; - case EDEADLK: - return nlm_deadlock; - case EAGAIN: - return nlm_lck_denied; + ret = nlm_granted; + goto out; + case -EAGAIN: + break; + case -EDEADLK: + ret = nlm_deadlock; + goto out; default: /* includes ENOLCK */ - return nlm_lck_denied_nolocks; - } + ret = nlm_lck_denied_nolocks; + goto out; } - if (!wait) { - up(&file->f_sema); - return nlm_lck_denied; - } + ret = nlm_lck_denied; + if (!wait) + goto out; - if (posix_locks_deadlock(&lock->fl, conflock)) { - up(&file->f_sema); - return nlm_deadlock; - } + ret = nlm_lck_blocked; + if (block != NULL) + goto out; /* If we don't have a block, create and initialize it. Then * retry because we may have slept in kmalloc. */ /* We have to release f_sema as nlmsvc_create_block may try to * to claim it while doing host garbage collection */ - if (block == NULL) { + if (newblock == NULL) { up(&file->f_sema); dprintk("lockd: blocking on this lock (allocating).\n"); - if (!(block = nlmsvc_create_block(rqstp, file, lock, cookie))) + if (!(newblock = nlmsvc_create_block(rqstp, file, lock, cookie))) return nlm_lck_denied_nolocks; goto again; } /* Append to list of blocked */ - nlmsvc_insert_block(block, NLM_NEVER); - - if (list_empty(&block->b_call.a_args.lock.fl.fl_block)) { - /* Now add block to block list of the conflicting lock - if we haven't done so. */ - dprintk("lockd: blocking on this lock.\n"); - posix_block_lock(conflock, &block->b_call.a_args.lock.fl); - } - + nlmsvc_insert_block(newblock, NLM_NEVER); +out: up(&file->f_sema); - return nlm_lck_blocked; + nlmsvc_release_block(newblock); + nlmsvc_release_block(block); + dprintk("lockd: nlmsvc_lock returned %u\n", ret); + return ret; } /* @@ -382,8 +439,6 @@ u32 nlmsvc_testlock(struct nlm_file *file, struct nlm_lock *lock, struct nlm_lock *conflock) { - struct file_lock *fl; - dprintk("lockd: nlmsvc_testlock(%s/%ld, ty=%d, %Ld-%Ld)\n", file->f_file->f_dentry->d_inode->i_sb->s_id, file->f_file->f_dentry->d_inode->i_ino, @@ -391,13 +446,14 @@ nlmsvc_testlock(struct nlm_file *file, struct nlm_lock *lock, (long long)lock->fl.fl_start, (long long)lock->fl.fl_end); - if ((fl = posix_test_lock(file->f_file, &lock->fl)) != NULL) { + if (posix_test_lock(file->f_file, &lock->fl, &conflock->fl)) { dprintk("lockd: conflicting lock(ty=%d, %Ld-%Ld)\n", - fl->fl_type, (long long)fl->fl_start, - (long long)fl->fl_end); + conflock->fl.fl_type, + (long long)conflock->fl.fl_start, + (long long)conflock->fl.fl_end); conflock->caller = "somehost"; /* FIXME */ conflock->oh.len = 0; /* don't return OH info */ - conflock->fl = *fl; + conflock->svid = conflock->fl.fl_pid; return nlm_lck_denied; } @@ -453,9 +509,12 @@ nlmsvc_cancel_blocked(struct nlm_file *file, struct nlm_lock *lock) (long long)lock->fl.fl_end); down(&file->f_sema); - if ((block = nlmsvc_lookup_block(file, lock, 1)) != NULL) - status = nlmsvc_delete_block(block, 1); + block = nlmsvc_lookup_block(file, lock); up(&file->f_sema); + if (block != NULL) { + status = nlmsvc_unlink_block(block); + nlmsvc_release_block(block); + } return status ? nlm_lck_denied : nlm_granted; } @@ -473,7 +532,7 @@ nlmsvc_notify_blocked(struct file_lock *fl) dprintk("lockd: VFS unblock notification for block %p\n", fl); for (bp = &nlm_blocked; (block = *bp) != 0; bp = &block->b_next) { - if (nlm_compare_locks(&block->b_call.a_args.lock.fl, fl)) { + if (nlm_compare_locks(&block->b_call->a_args.lock.fl, fl)) { nlmsvc_insert_block(block, 0); svc_wake_up(block->b_daemon); return; @@ -508,17 +567,13 @@ static void nlmsvc_grant_blocked(struct nlm_block *block) { struct nlm_file *file = block->b_file; - struct nlm_lock *lock = &block->b_call.a_args.lock; - struct file_lock *conflock; + struct nlm_lock *lock = &block->b_call->a_args.lock; int error; dprintk("lockd: grant blocked lock %p\n", block); - /* First thing is lock the file */ - down(&file->f_sema); - /* Unlink block request from list */ - nlmsvc_remove_block(block); + nlmsvc_unlink_block(block); /* If b_granted is true this means we've been here before. * Just retry the grant callback, possibly refreshing the RPC @@ -529,24 +584,21 @@ nlmsvc_grant_blocked(struct nlm_block *block) } /* Try the lock operation again */ - if ((conflock = posix_test_lock(file->f_file, &lock->fl)) != NULL) { - /* Bummer, we blocked again */ + lock->fl.fl_flags |= FL_SLEEP; + error = posix_lock_file(file->f_file, &lock->fl); + lock->fl.fl_flags &= ~FL_SLEEP; + + switch (error) { + case 0: + break; + case -EAGAIN: dprintk("lockd: lock still blocked\n"); nlmsvc_insert_block(block, NLM_NEVER); - posix_block_lock(conflock, &lock->fl); - up(&file->f_sema); return; - } - - /* Alright, no conflicting lock. Now lock it for real. If the - * following yields an error, this is most probably due to low - * memory. Retry the lock in a few seconds. - */ - if ((error = posix_lock_file(file->f_file, &lock->fl)) < 0) { + default: printk(KERN_WARNING "lockd: unexpected error %d in %s!\n", -error, __FUNCTION__); nlmsvc_insert_block(block, 10 * HZ); - up(&file->f_sema); return; } @@ -554,17 +606,15 @@ callback: /* Lock was granted by VFS. */ dprintk("lockd: GRANTing blocked lock.\n"); block->b_granted = 1; - block->b_incall = 1; /* Schedule next grant callback in 30 seconds */ nlmsvc_insert_block(block, 30 * HZ); /* Call the client */ - nlm_get_host(block->b_call.a_host); - if (nlmsvc_async_call(&block->b_call, NLMPROC_GRANTED_MSG, + kref_get(&block->b_count); + if (nlm_async_call(block->b_call, NLMPROC_GRANTED_MSG, &nlmsvc_grant_ops) < 0) - nlm_release_host(block->b_call.a_host); - up(&file->f_sema); + nlmsvc_release_block(block); } /* @@ -578,20 +628,10 @@ callback: static void nlmsvc_grant_callback(struct rpc_task *task, void *data) { struct nlm_rqst *call = data; - struct nlm_block *block; + struct nlm_block *block = call->a_block; unsigned long timeout; - struct sockaddr_in *peer_addr = RPC_PEERADDR(task->tk_client); dprintk("lockd: GRANT_MSG RPC callback\n"); - dprintk("callback: looking for cookie %s, host (%u.%u.%u.%u)\n", - nlmdbg_cookie2a(&call->a_args.cookie), - NIPQUAD(peer_addr->sin_addr.s_addr)); - if (!(block = nlmsvc_find_block(&call->a_args.cookie, peer_addr))) { - dprintk("lockd: no block for cookie %s, host (%u.%u.%u.%u)\n", - nlmdbg_cookie2a(&call->a_args.cookie), - NIPQUAD(peer_addr->sin_addr.s_addr)); - return; - } /* Technically, we should down the file semaphore here. Since we * move the block towards the head of the queue only, no harm @@ -608,13 +648,18 @@ static void nlmsvc_grant_callback(struct rpc_task *task, void *data) } nlmsvc_insert_block(block, timeout); svc_wake_up(block->b_daemon); - block->b_incall = 0; +} - nlm_release_host(call->a_host); +void nlmsvc_grant_release(void *data) +{ + struct nlm_rqst *call = data; + + nlmsvc_release_block(call->a_block); } static const struct rpc_call_ops nlmsvc_grant_ops = { .rpc_call_done = nlmsvc_grant_callback, + .rpc_release = nlmsvc_grant_release, }; /* @@ -634,25 +679,17 @@ nlmsvc_grant_reply(struct svc_rqst *rqstp, struct nlm_cookie *cookie, u32 status return; file = block->b_file; - file->f_count++; - down(&file->f_sema); - block = nlmsvc_find_block(cookie, &rqstp->rq_addr); if (block) { if (status == NLM_LCK_DENIED_GRACE_PERIOD) { /* Try again in a couple of seconds */ nlmsvc_insert_block(block, 10 * HZ); - up(&file->f_sema); } else { /* Lock is now held by client, or has been rejected. * In both cases, the block should be removed. */ - up(&file->f_sema); - if (status == NLM_LCK_GRANTED) - nlmsvc_delete_block(block, 0); - else - nlmsvc_delete_block(block, 1); + nlmsvc_unlink_block(block); } } - nlm_release_file(file); + nlmsvc_release_block(block); } /* @@ -675,10 +712,12 @@ nlmsvc_retry_blocked(void) break; dprintk("nlmsvc_retry_blocked(%p, when=%ld, done=%d)\n", block, block->b_when, block->b_done); + kref_get(&block->b_count); if (block->b_done) - nlmsvc_delete_block(block, 0); + nlmsvc_unlink_block(block); else nlmsvc_grant_blocked(block); + nlmsvc_release_block(block); } if ((block = nlm_blocked) && block->b_when != NLM_NEVER) diff --git a/fs/lockd/svcproc.c b/fs/lockd/svcproc.c index 35681d9cf1f..d210cf304e9 100644 --- a/fs/lockd/svcproc.c +++ b/fs/lockd/svcproc.c @@ -22,10 +22,6 @@ #define NLMDBG_FACILITY NLMDBG_CLIENT -static u32 nlmsvc_callback(struct svc_rqst *, u32, struct nlm_res *); - -static const struct rpc_call_ops nlmsvc_callback_ops; - #ifdef CONFIG_LOCKD_V4 static u32 cast_to_nlm(u32 status, u32 vers) @@ -262,83 +258,91 @@ nlmsvc_proc_granted(struct svc_rqst *rqstp, struct nlm_args *argp, } /* + * This is the generic lockd callback for async RPC calls + */ +static void nlmsvc_callback_exit(struct rpc_task *task, void *data) +{ + dprintk("lockd: %4d callback returned %d\n", task->tk_pid, + -task->tk_status); +} + +static void nlmsvc_callback_release(void *data) +{ + nlm_release_call(data); +} + +static const struct rpc_call_ops nlmsvc_callback_ops = { + .rpc_call_done = nlmsvc_callback_exit, + .rpc_release = nlmsvc_callback_release, +}; + +/* * `Async' versions of the above service routines. They aren't really, * because we send the callback before the reply proper. I hope this * doesn't break any clients. */ -static int -nlmsvc_proc_test_msg(struct svc_rqst *rqstp, struct nlm_args *argp, - void *resp) +static int nlmsvc_callback(struct svc_rqst *rqstp, u32 proc, struct nlm_args *argp, + int (*func)(struct svc_rqst *, struct nlm_args *, struct nlm_res *)) { - struct nlm_res res; - u32 stat; + struct nlm_host *host; + struct nlm_rqst *call; + int stat; - dprintk("lockd: TEST_MSG called\n"); - memset(&res, 0, sizeof(res)); + host = nlmsvc_lookup_host(rqstp); + if (host == NULL) + return rpc_system_err; - if ((stat = nlmsvc_proc_test(rqstp, argp, &res)) == 0) - stat = nlmsvc_callback(rqstp, NLMPROC_TEST_RES, &res); - return stat; + call = nlm_alloc_call(host); + if (call == NULL) + return rpc_system_err; + + stat = func(rqstp, argp, &call->a_res); + if (stat != 0) { + nlm_release_call(call); + return stat; + } + + call->a_flags = RPC_TASK_ASYNC; + if (nlm_async_reply(call, proc, &nlmsvc_callback_ops) < 0) + return rpc_system_err; + return rpc_success; } -static int -nlmsvc_proc_lock_msg(struct svc_rqst *rqstp, struct nlm_args *argp, +static int nlmsvc_proc_test_msg(struct svc_rqst *rqstp, struct nlm_args *argp, void *resp) { - struct nlm_res res; - u32 stat; + dprintk("lockd: TEST_MSG called\n"); + return nlmsvc_callback(rqstp, NLMPROC_TEST_RES, argp, nlmsvc_proc_test); +} +static int nlmsvc_proc_lock_msg(struct svc_rqst *rqstp, struct nlm_args *argp, + void *resp) +{ dprintk("lockd: LOCK_MSG called\n"); - memset(&res, 0, sizeof(res)); - - if ((stat = nlmsvc_proc_lock(rqstp, argp, &res)) == 0) - stat = nlmsvc_callback(rqstp, NLMPROC_LOCK_RES, &res); - return stat; + return nlmsvc_callback(rqstp, NLMPROC_LOCK_RES, argp, nlmsvc_proc_lock); } -static int -nlmsvc_proc_cancel_msg(struct svc_rqst *rqstp, struct nlm_args *argp, +static int nlmsvc_proc_cancel_msg(struct svc_rqst *rqstp, struct nlm_args *argp, void *resp) { - struct nlm_res res; - u32 stat; - dprintk("lockd: CANCEL_MSG called\n"); - memset(&res, 0, sizeof(res)); - - if ((stat = nlmsvc_proc_cancel(rqstp, argp, &res)) == 0) - stat = nlmsvc_callback(rqstp, NLMPROC_CANCEL_RES, &res); - return stat; + return nlmsvc_callback(rqstp, NLMPROC_CANCEL_RES, argp, nlmsvc_proc_cancel); } static int nlmsvc_proc_unlock_msg(struct svc_rqst *rqstp, struct nlm_args *argp, void *resp) { - struct nlm_res res; - u32 stat; - dprintk("lockd: UNLOCK_MSG called\n"); - memset(&res, 0, sizeof(res)); - - if ((stat = nlmsvc_proc_unlock(rqstp, argp, &res)) == 0) - stat = nlmsvc_callback(rqstp, NLMPROC_UNLOCK_RES, &res); - return stat; + return nlmsvc_callback(rqstp, NLMPROC_UNLOCK_RES, argp, nlmsvc_proc_unlock); } static int nlmsvc_proc_granted_msg(struct svc_rqst *rqstp, struct nlm_args *argp, void *resp) { - struct nlm_res res; - u32 stat; - dprintk("lockd: GRANTED_MSG called\n"); - memset(&res, 0, sizeof(res)); - - if ((stat = nlmsvc_proc_granted(rqstp, argp, &res)) == 0) - stat = nlmsvc_callback(rqstp, NLMPROC_GRANTED_RES, &res); - return stat; + return nlmsvc_callback(rqstp, NLMPROC_GRANTED_RES, argp, nlmsvc_proc_granted); } /* @@ -497,55 +501,6 @@ nlmsvc_proc_granted_res(struct svc_rqst *rqstp, struct nlm_res *argp, } /* - * This is the generic lockd callback for async RPC calls - */ -static u32 -nlmsvc_callback(struct svc_rqst *rqstp, u32 proc, struct nlm_res *resp) -{ - struct nlm_host *host; - struct nlm_rqst *call; - - if (!(call = nlmclnt_alloc_call())) - return rpc_system_err; - - host = nlmclnt_lookup_host(&rqstp->rq_addr, - rqstp->rq_prot, rqstp->rq_vers); - if (!host) { - kfree(call); - return rpc_system_err; - } - - call->a_flags = RPC_TASK_ASYNC; - call->a_host = host; - memcpy(&call->a_args, resp, sizeof(*resp)); - - if (nlmsvc_async_call(call, proc, &nlmsvc_callback_ops) < 0) - goto error; - - return rpc_success; - error: - nlm_release_host(host); - kfree(call); - return rpc_system_err; -} - -static void nlmsvc_callback_exit(struct rpc_task *task, void *data) -{ - struct nlm_rqst *call = data; - - if (task->tk_status < 0) { - dprintk("lockd: %4d callback failed (errno = %d)\n", - task->tk_pid, -task->tk_status); - } - nlm_release_host(call->a_host); - kfree(call); -} - -static const struct rpc_call_ops nlmsvc_callback_ops = { - .rpc_call_done = nlmsvc_callback_exit, -}; - -/* * NLM Server procedures. */ diff --git a/fs/lockd/svcshare.c b/fs/lockd/svcshare.c index 4943fb7836c..27288c83da9 100644 --- a/fs/lockd/svcshare.c +++ b/fs/lockd/svcshare.c @@ -88,7 +88,7 @@ nlmsvc_unshare_file(struct nlm_host *host, struct nlm_file *file, * Traverse all shares for a given file (and host). * NLM_ACT_CHECK is handled by nlmsvc_inspect_file. */ -int +void nlmsvc_traverse_shares(struct nlm_host *host, struct nlm_file *file, int action) { struct nlm_share *share, **shpp; @@ -106,6 +106,4 @@ nlmsvc_traverse_shares(struct nlm_host *host, struct nlm_file *file, int action) } shpp = &share->s_next; } - - return 0; } diff --git a/fs/lockd/svcsubs.c b/fs/lockd/svcsubs.c index 62f4a385177..a570e5c8a93 100644 --- a/fs/lockd/svcsubs.c +++ b/fs/lockd/svcsubs.c @@ -11,6 +11,7 @@ #include <linux/string.h> #include <linux/time.h> #include <linux/in.h> +#include <linux/mutex.h> #include <linux/sunrpc/svc.h> #include <linux/sunrpc/clnt.h> #include <linux/nfsd/nfsfh.h> @@ -28,7 +29,7 @@ #define FILE_HASH_BITS 5 #define FILE_NRHASH (1<<FILE_HASH_BITS) static struct nlm_file * nlm_files[FILE_NRHASH]; -static DECLARE_MUTEX(nlm_file_sema); +static DEFINE_MUTEX(nlm_file_mutex); #ifdef NFSD_DEBUG static inline void nlm_debug_print_fh(char *msg, struct nfs_fh *f) @@ -91,7 +92,7 @@ nlm_lookup_file(struct svc_rqst *rqstp, struct nlm_file **result, hash = file_hash(f); /* Lock file table */ - down(&nlm_file_sema); + mutex_lock(&nlm_file_mutex); for (file = nlm_files[hash]; file; file = file->f_next) if (!nfs_compare_fh(&file->f_handle, f)) @@ -130,7 +131,7 @@ found: nfserr = 0; out_unlock: - up(&nlm_file_sema); + mutex_unlock(&nlm_file_mutex); return nfserr; out_free: @@ -182,7 +183,7 @@ nlm_traverse_locks(struct nlm_host *host, struct nlm_file *file, int action) again: file->f_locks = 0; for (fl = inode->i_flock; fl; fl = fl->fl_next) { - if (!(fl->fl_flags & FL_LOCKD)) + if (fl->fl_lmops != &nlmsvc_lock_operations) continue; /* update current lock count */ @@ -224,9 +225,8 @@ nlm_inspect_file(struct nlm_host *host, struct nlm_file *file, int action) if (file->f_count || file->f_blocks || file->f_shares) return 1; } else { - if (nlmsvc_traverse_blocks(host, file, action) - || nlmsvc_traverse_shares(host, file, action)) - return 1; + nlmsvc_traverse_blocks(host, file, action); + nlmsvc_traverse_shares(host, file, action); } return nlm_traverse_locks(host, file, action); } @@ -240,14 +240,14 @@ nlm_traverse_files(struct nlm_host *host, int action) struct nlm_file *file, **fp; int i; - down(&nlm_file_sema); + mutex_lock(&nlm_file_mutex); for (i = 0; i < FILE_NRHASH; i++) { fp = nlm_files + i; while ((file = *fp) != NULL) { /* Traverse locks, blocks and shares of this file * and update file->f_locks count */ if (nlm_inspect_file(host, file, action)) { - up(&nlm_file_sema); + mutex_unlock(&nlm_file_mutex); return 1; } @@ -262,7 +262,7 @@ nlm_traverse_files(struct nlm_host *host, int action) } } } - up(&nlm_file_sema); + mutex_unlock(&nlm_file_mutex); return 0; } @@ -282,7 +282,7 @@ nlm_release_file(struct nlm_file *file) file, file->f_count); /* Lock file table */ - down(&nlm_file_sema); + mutex_lock(&nlm_file_mutex); /* If there are no more locks etc, delete the file */ if(--file->f_count == 0) { @@ -290,7 +290,7 @@ nlm_release_file(struct nlm_file *file) nlm_delete_file(file); } - up(&nlm_file_sema); + mutex_unlock(&nlm_file_mutex); } /* diff --git a/fs/lockd/xdr.c b/fs/lockd/xdr.c index 1d700a4dd0b..f22a3764461 100644 --- a/fs/lockd/xdr.c +++ b/fs/lockd/xdr.c @@ -131,10 +131,11 @@ nlm_decode_lock(u32 *p, struct nlm_lock *lock) || !(p = nlm_decode_fh(p, &lock->fh)) || !(p = nlm_decode_oh(p, &lock->oh))) return NULL; + lock->svid = ntohl(*p++); locks_init_lock(fl); fl->fl_owner = current->files; - fl->fl_pid = ntohl(*p++); + fl->fl_pid = (pid_t)lock->svid; fl->fl_flags = FL_POSIX; fl->fl_type = F_RDLCK; /* as good as anything else */ start = ntohl(*p++); @@ -174,7 +175,7 @@ nlm_encode_lock(u32 *p, struct nlm_lock *lock) else len = loff_t_to_s32(fl->fl_end - fl->fl_start + 1); - *p++ = htonl(fl->fl_pid); + *p++ = htonl(lock->svid); *p++ = htonl(start); *p++ = htonl(len); @@ -197,7 +198,7 @@ nlm_encode_testres(u32 *p, struct nlm_res *resp) struct file_lock *fl = &resp->lock.fl; *p++ = (fl->fl_type == F_RDLCK)? xdr_zero : xdr_one; - *p++ = htonl(fl->fl_pid); + *p++ = htonl(resp->lock.svid); /* Encode owner handle. */ if (!(p = xdr_encode_netobj(p, &resp->lock.oh))) @@ -298,7 +299,8 @@ nlmsvc_decode_shareargs(struct svc_rqst *rqstp, u32 *p, nlm_args *argp) memset(lock, 0, sizeof(*lock)); locks_init_lock(&lock->fl); - lock->fl.fl_pid = ~(u32) 0; + lock->svid = ~(u32) 0; + lock->fl.fl_pid = (pid_t)lock->svid; if (!(p = nlm_decode_cookie(p, &argp->cookie)) || !(p = xdr_decode_string_inplace(p, &lock->caller, @@ -415,7 +417,8 @@ nlmclt_decode_testres(struct rpc_rqst *req, u32 *p, struct nlm_res *resp) memset(&resp->lock, 0, sizeof(resp->lock)); locks_init_lock(fl); excl = ntohl(*p++); - fl->fl_pid = ntohl(*p++); + resp->lock.svid = ntohl(*p++); + fl->fl_pid = (pid_t)resp->lock.svid; if (!(p = nlm_decode_oh(p, &resp->lock.oh))) return -EIO; @@ -543,7 +546,9 @@ nlmclt_decode_res(struct rpc_rqst *req, u32 *p, struct nlm_res *resp) .p_proc = NLMPROC_##proc, \ .p_encode = (kxdrproc_t) nlmclt_encode_##argtype, \ .p_decode = (kxdrproc_t) nlmclt_decode_##restype, \ - .p_bufsiz = MAX(NLM_##argtype##_sz, NLM_##restype##_sz) << 2 \ + .p_bufsiz = MAX(NLM_##argtype##_sz, NLM_##restype##_sz) << 2, \ + .p_statidx = NLMPROC_##proc, \ + .p_name = #proc, \ } static struct rpc_procinfo nlm_procedures[] = { diff --git a/fs/lockd/xdr4.c b/fs/lockd/xdr4.c index fdcf105a530..36eb175ec33 100644 --- a/fs/lockd/xdr4.c +++ b/fs/lockd/xdr4.c @@ -130,10 +130,11 @@ nlm4_decode_lock(u32 *p, struct nlm_lock *lock) || !(p = nlm4_decode_fh(p, &lock->fh)) || !(p = nlm4_decode_oh(p, &lock->oh))) return NULL; + lock->svid = ntohl(*p++); locks_init_lock(fl); fl->fl_owner = current->files; - fl->fl_pid = ntohl(*p++); + fl->fl_pid = (pid_t)lock->svid; fl->fl_flags = FL_POSIX; fl->fl_type = F_RDLCK; /* as good as anything else */ p = xdr_decode_hyper(p, &start); @@ -167,7 +168,7 @@ nlm4_encode_lock(u32 *p, struct nlm_lock *lock) || (fl->fl_end > NLM4_OFFSET_MAX && fl->fl_end != OFFSET_MAX)) return NULL; - *p++ = htonl(fl->fl_pid); + *p++ = htonl(lock->svid); start = loff_t_to_s64(fl->fl_start); if (fl->fl_end == OFFSET_MAX) @@ -198,7 +199,7 @@ nlm4_encode_testres(u32 *p, struct nlm_res *resp) struct file_lock *fl = &resp->lock.fl; *p++ = (fl->fl_type == F_RDLCK)? xdr_zero : xdr_one; - *p++ = htonl(fl->fl_pid); + *p++ = htonl(resp->lock.svid); /* Encode owner handle. */ if (!(p = xdr_encode_netobj(p, &resp->lock.oh))) @@ -212,8 +213,8 @@ nlm4_encode_testres(u32 *p, struct nlm_res *resp) p = xdr_encode_hyper(p, start); p = xdr_encode_hyper(p, len); - dprintk("xdr: encode_testres (status %d pid %d type %d start %Ld end %Ld)\n", - resp->status, fl->fl_pid, fl->fl_type, + dprintk("xdr: encode_testres (status %u pid %d type %d start %Ld end %Ld)\n", + resp->status, (int)resp->lock.svid, fl->fl_type, (long long)fl->fl_start, (long long)fl->fl_end); } @@ -303,7 +304,8 @@ nlm4svc_decode_shareargs(struct svc_rqst *rqstp, u32 *p, nlm_args *argp) memset(lock, 0, sizeof(*lock)); locks_init_lock(&lock->fl); - lock->fl.fl_pid = ~(u32) 0; + lock->svid = ~(u32) 0; + lock->fl.fl_pid = (pid_t)lock->svid; if (!(p = nlm4_decode_cookie(p, &argp->cookie)) || !(p = xdr_decode_string_inplace(p, &lock->caller, @@ -420,7 +422,8 @@ nlm4clt_decode_testres(struct rpc_rqst *req, u32 *p, struct nlm_res *resp) memset(&resp->lock, 0, sizeof(resp->lock)); locks_init_lock(fl); excl = ntohl(*p++); - fl->fl_pid = ntohl(*p++); + resp->lock.svid = ntohl(*p++); + fl->fl_pid = (pid_t)resp->lock.svid; if (!(p = nlm4_decode_oh(p, &resp->lock.oh))) return -EIO; @@ -548,7 +551,9 @@ nlm4clt_decode_res(struct rpc_rqst *req, u32 *p, struct nlm_res *resp) .p_proc = NLMPROC_##proc, \ .p_encode = (kxdrproc_t) nlm4clt_encode_##argtype, \ .p_decode = (kxdrproc_t) nlm4clt_decode_##restype, \ - .p_bufsiz = MAX(NLM4_##argtype##_sz, NLM4_##restype##_sz) << 2 \ + .p_bufsiz = MAX(NLM4_##argtype##_sz, NLM4_##restype##_sz) << 2, \ + .p_statidx = NLMPROC_##proc, \ + .p_name = #proc, \ } static struct rpc_procinfo nlm4_procedures[] = { diff --git a/fs/locks.c b/fs/locks.c index 909eab8fb1d..4d9e71d43e7 100644 --- a/fs/locks.c +++ b/fs/locks.c @@ -139,13 +139,10 @@ int lease_break_time = 45; #define for_each_lock(inode, lockp) \ for (lockp = &inode->i_flock; *lockp != NULL; lockp = &(*lockp)->fl_next) -LIST_HEAD(file_lock_list); - -EXPORT_SYMBOL(file_lock_list); - +static LIST_HEAD(file_lock_list); static LIST_HEAD(blocked_list); -static kmem_cache_t *filelock_cache; +static kmem_cache_t *filelock_cache __read_mostly; /* Allocate an empty lock structure. */ static struct file_lock *locks_alloc_lock(void) @@ -153,6 +150,21 @@ static struct file_lock *locks_alloc_lock(void) return kmem_cache_alloc(filelock_cache, SLAB_KERNEL); } +static void locks_release_private(struct file_lock *fl) +{ + if (fl->fl_ops) { + if (fl->fl_ops->fl_release_private) + fl->fl_ops->fl_release_private(fl); + fl->fl_ops = NULL; + } + if (fl->fl_lmops) { + if (fl->fl_lmops->fl_release_private) + fl->fl_lmops->fl_release_private(fl); + fl->fl_lmops = NULL; + } + +} + /* Free a lock which is not in use. */ static void locks_free_lock(struct file_lock *fl) { @@ -169,18 +181,7 @@ static void locks_free_lock(struct file_lock *fl) if (!list_empty(&fl->fl_link)) panic("Attempting to free lock on active lock list"); - if (fl->fl_ops) { - if (fl->fl_ops->fl_release_private) - fl->fl_ops->fl_release_private(fl); - fl->fl_ops = NULL; - } - - if (fl->fl_lmops) { - if (fl->fl_lmops->fl_release_private) - fl->fl_lmops->fl_release_private(fl); - fl->fl_lmops = NULL; - } - + locks_release_private(fl); kmem_cache_free(filelock_cache, fl); } @@ -218,24 +219,46 @@ static void init_once(void *foo, kmem_cache_t *cache, unsigned long flags) locks_init_lock(lock); } +static void locks_copy_private(struct file_lock *new, struct file_lock *fl) +{ + if (fl->fl_ops) { + if (fl->fl_ops->fl_copy_lock) + fl->fl_ops->fl_copy_lock(new, fl); + new->fl_ops = fl->fl_ops; + } + if (fl->fl_lmops) { + if (fl->fl_lmops->fl_copy_lock) + fl->fl_lmops->fl_copy_lock(new, fl); + new->fl_lmops = fl->fl_lmops; + } +} + /* * Initialize a new lock from an existing file_lock structure. */ -void locks_copy_lock(struct file_lock *new, struct file_lock *fl) +static void __locks_copy_lock(struct file_lock *new, const struct file_lock *fl) { new->fl_owner = fl->fl_owner; new->fl_pid = fl->fl_pid; - new->fl_file = fl->fl_file; + new->fl_file = NULL; new->fl_flags = fl->fl_flags; new->fl_type = fl->fl_type; new->fl_start = fl->fl_start; new->fl_end = fl->fl_end; + new->fl_ops = NULL; + new->fl_lmops = NULL; +} + +void locks_copy_lock(struct file_lock *new, struct file_lock *fl) +{ + locks_release_private(new); + + __locks_copy_lock(new, fl); + new->fl_file = fl->fl_file; new->fl_ops = fl->fl_ops; new->fl_lmops = fl->fl_lmops; - if (fl->fl_ops && fl->fl_ops->fl_copy_lock) - fl->fl_ops->fl_copy_lock(new, fl); - if (fl->fl_lmops && fl->fl_lmops->fl_copy_lock) - fl->fl_lmops->fl_copy_lock(new, fl); + + locks_copy_private(new, fl); } EXPORT_SYMBOL(locks_copy_lock); @@ -510,12 +533,7 @@ static void locks_delete_block(struct file_lock *waiter) static void locks_insert_block(struct file_lock *blocker, struct file_lock *waiter) { - if (!list_empty(&waiter->fl_block)) { - printk(KERN_ERR "locks_insert_block: removing duplicated lock " - "(pid=%d %Ld-%Ld type=%d)\n", waiter->fl_pid, - waiter->fl_start, waiter->fl_end, waiter->fl_type); - __locks_delete_block(waiter); - } + BUG_ON(!list_empty(&waiter->fl_block)); list_add_tail(&waiter->fl_block, &blocker->fl_block); waiter->fl_next = blocker; if (IS_POSIX(blocker)) @@ -654,8 +672,9 @@ static int locks_block_on_timeout(struct file_lock *blocker, struct file_lock *w return result; } -struct file_lock * -posix_test_lock(struct file *filp, struct file_lock *fl) +int +posix_test_lock(struct file *filp, struct file_lock *fl, + struct file_lock *conflock) { struct file_lock *cfl; @@ -666,9 +685,13 @@ posix_test_lock(struct file *filp, struct file_lock *fl) if (posix_locks_conflict(cfl, fl)) break; } + if (cfl) { + __locks_copy_lock(conflock, cfl); + unlock_kernel(); + return 1; + } unlock_kernel(); - - return (cfl); + return 0; } EXPORT_SYMBOL(posix_test_lock); @@ -769,9 +792,7 @@ out: return error; } -EXPORT_SYMBOL(posix_lock_file); - -static int __posix_lock_file(struct inode *inode, struct file_lock *request) +static int __posix_lock_file_conf(struct inode *inode, struct file_lock *request, struct file_lock *conflock) { struct file_lock *fl; struct file_lock *new_fl, *new_fl2; @@ -795,6 +816,8 @@ static int __posix_lock_file(struct inode *inode, struct file_lock *request) continue; if (!posix_locks_conflict(request, fl)) continue; + if (conflock) + locks_copy_lock(conflock, fl); error = -EAGAIN; if (!(request->fl_flags & FL_SLEEP)) goto out; @@ -904,7 +927,8 @@ static int __posix_lock_file(struct inode *inode, struct file_lock *request) fl->fl_start = request->fl_start; fl->fl_end = request->fl_end; fl->fl_type = request->fl_type; - fl->fl_u = request->fl_u; + locks_release_private(fl); + locks_copy_private(fl, request); request = fl; added = 1; } @@ -963,8 +987,24 @@ static int __posix_lock_file(struct inode *inode, struct file_lock *request) */ int posix_lock_file(struct file *filp, struct file_lock *fl) { - return __posix_lock_file(filp->f_dentry->d_inode, fl); + return __posix_lock_file_conf(filp->f_dentry->d_inode, fl, NULL); } +EXPORT_SYMBOL(posix_lock_file); + +/** + * posix_lock_file_conf - Apply a POSIX-style lock to a file + * @filp: The file to apply the lock to + * @fl: The lock to be applied + * @conflock: Place to return a copy of the conflicting lock, if found. + * + * Except for the conflock parameter, acts just like posix_lock_file. + */ +int posix_lock_file_conf(struct file *filp, struct file_lock *fl, + struct file_lock *conflock) +{ + return __posix_lock_file_conf(filp->f_dentry->d_inode, fl, conflock); +} +EXPORT_SYMBOL(posix_lock_file_conf); /** * posix_lock_file_wait - Apply a POSIX-style lock to a file @@ -980,7 +1020,7 @@ int posix_lock_file_wait(struct file *filp, struct file_lock *fl) int error; might_sleep (); for (;;) { - error = __posix_lock_file(filp->f_dentry->d_inode, fl); + error = posix_lock_file(filp, fl); if ((error != -EAGAIN) || !(fl->fl_flags & FL_SLEEP)) break; error = wait_event_interruptible(fl->fl_wait, !fl->fl_next); @@ -1052,7 +1092,7 @@ int locks_mandatory_area(int read_write, struct inode *inode, fl.fl_end = offset + count - 1; for (;;) { - error = __posix_lock_file(inode, &fl); + error = __posix_lock_file_conf(inode, &fl, NULL); if (error != -EAGAIN) break; if (!(fl.fl_flags & FL_SLEEP)) @@ -1544,7 +1584,7 @@ asmlinkage long sys_flock(unsigned int fd, unsigned int cmd) */ int fcntl_getlk(struct file *filp, struct flock __user *l) { - struct file_lock *fl, file_lock; + struct file_lock *fl, cfl, file_lock; struct flock flock; int error; @@ -1568,7 +1608,7 @@ int fcntl_getlk(struct file *filp, struct flock __user *l) else fl = (file_lock.fl_type == F_UNLCK ? NULL : &file_lock); } else { - fl = posix_test_lock(filp, &file_lock); + fl = (posix_test_lock(filp, &file_lock, &cfl) ? &cfl : NULL); } flock.l_type = F_UNLCK; @@ -1665,7 +1705,7 @@ again: error = filp->f_op->lock(filp, cmd, file_lock); else { for (;;) { - error = __posix_lock_file(inode, file_lock); + error = posix_lock_file(filp, file_lock); if ((error != -EAGAIN) || (cmd == F_SETLK)) break; error = wait_event_interruptible(file_lock->fl_wait, @@ -1698,7 +1738,7 @@ out: */ int fcntl_getlk64(struct file *filp, struct flock64 __user *l) { - struct file_lock *fl, file_lock; + struct file_lock *fl, cfl, file_lock; struct flock64 flock; int error; @@ -1722,7 +1762,7 @@ int fcntl_getlk64(struct file *filp, struct flock64 __user *l) else fl = (file_lock.fl_type == F_UNLCK ? NULL : &file_lock); } else { - fl = posix_test_lock(filp, &file_lock); + fl = (posix_test_lock(filp, &file_lock, &cfl) ? &cfl : NULL); } flock.l_type = F_UNLCK; @@ -1808,7 +1848,7 @@ again: error = filp->f_op->lock(filp, cmd, file_lock); else { for (;;) { - error = __posix_lock_file(inode, file_lock); + error = posix_lock_file(filp, file_lock); if ((error != -EAGAIN) || (cmd == F_SETLK64)) break; error = wait_event_interruptible(file_lock->fl_wait, @@ -1936,21 +1976,6 @@ void locks_remove_flock(struct file *filp) } /** - * posix_block_lock - blocks waiting for a file lock - * @blocker: the lock which is blocking - * @waiter: the lock which conflicts and has to wait - * - * lockd needs to block waiting for locks. - */ -void -posix_block_lock(struct file_lock *blocker, struct file_lock *waiter) -{ - locks_insert_block(blocker, waiter); -} - -EXPORT_SYMBOL(posix_block_lock); - -/** * posix_unblock_lock - stop waiting for a file lock * @filp: how the file was opened * @waiter: the lock which was waiting diff --git a/fs/mbcache.c b/fs/mbcache.c index 73e754fea2d..e4fde1ab22c 100644 --- a/fs/mbcache.c +++ b/fs/mbcache.c @@ -311,7 +311,7 @@ fail: /* * mb_cache_shrink() * - * Removes all cache entires of a device from the cache. All cache entries + * Removes all cache entries of a device from the cache. All cache entries * currently in use cannot be freed, and thus remain in the cache. All others * are freed. * diff --git a/fs/minix/bitmap.c b/fs/minix/bitmap.c index dc6a4e4abcd..4a6abc49418 100644 --- a/fs/minix/bitmap.c +++ b/fs/minix/bitmap.c @@ -56,7 +56,7 @@ void minix_free_block(struct inode * inode, int block) unsigned int bit,zone; if (block < sbi->s_firstdatazone || block >= sbi->s_nzones) { - printk("trying to free block not in datazone\n"); + printk("Trying to free block not in datazone\n"); return; } zone = block - sbi->s_firstdatazone + 1; @@ -124,7 +124,7 @@ minix_V1_raw_inode(struct super_block *sb, ino_t ino, struct buffer_head **bh) ino / MINIX_INODES_PER_BLOCK; *bh = sb_bread(sb, block); if (!*bh) { - printk("unable to read i-node block\n"); + printk("Unable to read inode block\n"); return NULL; } p = (void *)(*bh)->b_data; @@ -149,7 +149,7 @@ minix_V2_raw_inode(struct super_block *sb, ino_t ino, struct buffer_head **bh) ino / MINIX2_INODES_PER_BLOCK; *bh = sb_bread(sb, block); if (!*bh) { - printk("unable to read i-node block\n"); + printk("Unable to read inode block\n"); return NULL; } p = (void *)(*bh)->b_data; @@ -204,7 +204,7 @@ void minix_free_inode(struct inode * inode) bh = sbi->s_imap[ino >> 13]; lock_kernel(); if (!minix_test_and_clear_bit(ino & 8191, bh->b_data)) - printk("minix_free_inode: bit %lu already cleared.\n", ino); + printk("minix_free_inode: bit %lu already cleared\n", ino); unlock_kernel(); mark_buffer_dirty(bh); out: @@ -238,7 +238,7 @@ struct inode * minix_new_inode(const struct inode * dir, int * error) return NULL; } if (minix_test_and_set_bit(j,bh->b_data)) { /* shouldn't happen */ - printk("new_inode: bit already set"); + printk("new_inode: bit already set\n"); unlock_kernel(); iput(inode); return NULL; diff --git a/fs/minix/dir.c b/fs/minix/dir.c index 732502aabc0..69224d1fe04 100644 --- a/fs/minix/dir.c +++ b/fs/minix/dir.c @@ -14,7 +14,7 @@ typedef struct minix_dir_entry minix_dirent; static int minix_readdir(struct file *, void *, filldir_t); -struct file_operations minix_dir_operations = { +const struct file_operations minix_dir_operations = { .read = generic_read_dir, .readdir = minix_readdir, .fsync = minix_sync_file, diff --git a/fs/minix/file.c b/fs/minix/file.c index f1d77acb3f0..420b32882a1 100644 --- a/fs/minix/file.c +++ b/fs/minix/file.c @@ -15,7 +15,7 @@ */ int minix_sync_file(struct file *, struct dentry *, int); -struct file_operations minix_file_operations = { +const struct file_operations minix_file_operations = { .llseek = generic_file_llseek, .read = generic_file_read, .write = generic_file_write, diff --git a/fs/minix/inode.c b/fs/minix/inode.c index d9ffc43fee5..2dcccf1d1b7 100644 --- a/fs/minix/inode.c +++ b/fs/minix/inode.c @@ -127,11 +127,11 @@ static int minix_remount (struct super_block * sb, int * flags, char * data) mark_buffer_dirty(sbi->s_sbh); if (!(sbi->s_mount_state & MINIX_VALID_FS)) - printk ("MINIX-fs warning: remounting unchecked fs, " - "running fsck is recommended.\n"); + printk("MINIX-fs warning: remounting unchecked fs, " + "running fsck is recommended\n"); else if ((sbi->s_mount_state & MINIX_ERROR_FS)) - printk ("MINIX-fs warning: remounting fs with errors, " - "running fsck is recommended.\n"); + printk("MINIX-fs warning: remounting fs with errors, " + "running fsck is recommended\n"); } return 0; } @@ -245,11 +245,11 @@ static int minix_fill_super(struct super_block *s, void *data, int silent) mark_buffer_dirty(bh); } if (!(sbi->s_mount_state & MINIX_VALID_FS)) - printk ("MINIX-fs: mounting unchecked file system, " - "running fsck is recommended.\n"); + printk("MINIX-fs: mounting unchecked file system, " + "running fsck is recommended\n"); else if (sbi->s_mount_state & MINIX_ERROR_FS) - printk ("MINIX-fs: mounting file system with errors, " - "running fsck is recommended.\n"); + printk("MINIX-fs: mounting file system with errors, " + "running fsck is recommended\n"); return 0; out_iput: @@ -273,19 +273,19 @@ out_no_bitmap: out_no_map: if (!silent) - printk ("MINIX-fs: can't allocate map\n"); + printk("MINIX-fs: can't allocate map\n"); goto out_release; out_no_fs: if (!silent) - printk("VFS: Can't find a Minix or Minix V2 filesystem on device " - "%s.\n", s->s_id); + printk("VFS: Can't find a Minix or Minix V2 filesystem " + "on device %s\n", s->s_id); out_release: brelse(bh); goto out; out_bad_hblock: - printk("MINIX-fs: blocksize too small for device.\n"); + printk("MINIX-fs: blocksize too small for device\n"); goto out; out_bad_sb: @@ -524,7 +524,7 @@ int minix_sync_inode(struct inode * inode) sync_dirty_buffer(bh); if (buffer_req(bh) && !buffer_uptodate(bh)) { - printk ("IO error syncing minix inode [%s:%08lx]\n", + printk("IO error syncing minix inode [%s:%08lx]\n", inode->i_sb->s_id, inode->i_ino); err = -1; } diff --git a/fs/minix/itree_v1.c b/fs/minix/itree_v1.c index ba06aef4aca..656b1347a25 100644 --- a/fs/minix/itree_v1.c +++ b/fs/minix/itree_v1.c @@ -25,9 +25,9 @@ static int block_to_path(struct inode * inode, long block, int offsets[DEPTH]) int n = 0; if (block < 0) { - printk("minix_bmap: block<0"); + printk("minix_bmap: block<0\n"); } else if (block >= (minix_sb(inode->i_sb)->s_max_size/BLOCK_SIZE)) { - printk("minix_bmap: block>big"); + printk("minix_bmap: block>big\n"); } else if (block < 7) { offsets[n++] = block; } else if ((block -= 7) < 512) { diff --git a/fs/minix/itree_v2.c b/fs/minix/itree_v2.c index 3adc7675560..9adcdc754e0 100644 --- a/fs/minix/itree_v2.c +++ b/fs/minix/itree_v2.c @@ -25,9 +25,9 @@ static int block_to_path(struct inode * inode, long block, int offsets[DEPTH]) int n = 0; if (block < 0) { - printk("minix_bmap: block<0"); + printk("minix_bmap: block<0\n"); } else if (block >= (minix_sb(inode->i_sb)->s_max_size/BLOCK_SIZE)) { - printk("minix_bmap: block>big"); + printk("minix_bmap: block>big\n"); } else if (block < 7) { offsets[n++] = block; } else if ((block -= 7) < 256) { diff --git a/fs/minix/minix.h b/fs/minix/minix.h index e42a8bb8900..c55b77cdcc8 100644 --- a/fs/minix/minix.h +++ b/fs/minix/minix.h @@ -81,8 +81,8 @@ extern int minix_sync_file(struct file *, struct dentry *, int); extern struct inode_operations minix_file_inode_operations; extern struct inode_operations minix_dir_inode_operations; -extern struct file_operations minix_file_operations; -extern struct file_operations minix_dir_operations; +extern const struct file_operations minix_file_operations; +extern const struct file_operations minix_dir_operations; extern struct dentry_operations minix_dentry_operations; static inline struct minix_sb_info *minix_sb(struct super_block *sb) diff --git a/fs/mpage.c b/fs/mpage.c index e431cb3878d..9bf2eb30e6f 100644 --- a/fs/mpage.c +++ b/fs/mpage.c @@ -163,9 +163,19 @@ map_buffer_to_page(struct page *page, struct buffer_head *bh, int page_block) } while (page_bh != head); } +/* + * This is the worker routine which does all the work of mapping the disk + * blocks and constructs largest possible bios, submits them for IO if the + * blocks are not contiguous on the disk. + * + * We pass a buffer_head back and forth and use its buffer_mapped() flag to + * represent the validity of its disk mapping and to decide when to do the next + * get_block() call. + */ static struct bio * do_mpage_readpage(struct bio *bio, struct page *page, unsigned nr_pages, - sector_t *last_block_in_bio, get_block_t get_block) + sector_t *last_block_in_bio, struct buffer_head *map_bh, + unsigned long *first_logical_block, get_block_t get_block) { struct inode *inode = page->mapping->host; const unsigned blkbits = inode->i_blkbits; @@ -173,33 +183,72 @@ do_mpage_readpage(struct bio *bio, struct page *page, unsigned nr_pages, const unsigned blocksize = 1 << blkbits; sector_t block_in_file; sector_t last_block; + sector_t last_block_in_file; sector_t blocks[MAX_BUF_PER_PAGE]; unsigned page_block; unsigned first_hole = blocks_per_page; struct block_device *bdev = NULL; - struct buffer_head bh; int length; int fully_mapped = 1; + unsigned nblocks; + unsigned relative_block; if (page_has_buffers(page)) goto confused; block_in_file = (sector_t)page->index << (PAGE_CACHE_SHIFT - blkbits); - last_block = (i_size_read(inode) + blocksize - 1) >> blkbits; + last_block = block_in_file + nr_pages * blocks_per_page; + last_block_in_file = (i_size_read(inode) + blocksize - 1) >> blkbits; + if (last_block > last_block_in_file) + last_block = last_block_in_file; + page_block = 0; + + /* + * Map blocks using the result from the previous get_blocks call first. + */ + nblocks = map_bh->b_size >> blkbits; + if (buffer_mapped(map_bh) && block_in_file > *first_logical_block && + block_in_file < (*first_logical_block + nblocks)) { + unsigned map_offset = block_in_file - *first_logical_block; + unsigned last = nblocks - map_offset; + + for (relative_block = 0; ; relative_block++) { + if (relative_block == last) { + clear_buffer_mapped(map_bh); + break; + } + if (page_block == blocks_per_page) + break; + blocks[page_block] = map_bh->b_blocknr + map_offset + + relative_block; + page_block++; + block_in_file++; + } + bdev = map_bh->b_bdev; + } + + /* + * Then do more get_blocks calls until we are done with this page. + */ + map_bh->b_page = page; + while (page_block < blocks_per_page) { + map_bh->b_state = 0; + map_bh->b_size = 0; - bh.b_page = page; - for (page_block = 0; page_block < blocks_per_page; - page_block++, block_in_file++) { - bh.b_state = 0; if (block_in_file < last_block) { - if (get_block(inode, block_in_file, &bh, 0)) + map_bh->b_size = (last_block-block_in_file) << blkbits; + if (get_block(inode, block_in_file, map_bh, 0)) goto confused; + *first_logical_block = block_in_file; } - if (!buffer_mapped(&bh)) { + if (!buffer_mapped(map_bh)) { fully_mapped = 0; if (first_hole == blocks_per_page) first_hole = page_block; + page_block++; + block_in_file++; + clear_buffer_mapped(map_bh); continue; } @@ -209,8 +258,8 @@ do_mpage_readpage(struct bio *bio, struct page *page, unsigned nr_pages, * we just collected from get_block into the page's buffers * so readpage doesn't have to repeat the get_block call */ - if (buffer_uptodate(&bh)) { - map_buffer_to_page(page, &bh, page_block); + if (buffer_uptodate(map_bh)) { + map_buffer_to_page(page, map_bh, page_block); goto confused; } @@ -218,10 +267,20 @@ do_mpage_readpage(struct bio *bio, struct page *page, unsigned nr_pages, goto confused; /* hole -> non-hole */ /* Contiguous blocks? */ - if (page_block && blocks[page_block-1] != bh.b_blocknr-1) + if (page_block && blocks[page_block-1] != map_bh->b_blocknr-1) goto confused; - blocks[page_block] = bh.b_blocknr; - bdev = bh.b_bdev; + nblocks = map_bh->b_size >> blkbits; + for (relative_block = 0; ; relative_block++) { + if (relative_block == nblocks) { + clear_buffer_mapped(map_bh); + break; + } else if (page_block == blocks_per_page) + break; + blocks[page_block] = map_bh->b_blocknr+relative_block; + page_block++; + block_in_file++; + } + bdev = map_bh->b_bdev; } if (first_hole != blocks_per_page) { @@ -260,7 +319,7 @@ alloc_new: goto alloc_new; } - if (buffer_boundary(&bh) || (first_hole != blocks_per_page)) + if (buffer_boundary(map_bh) || (first_hole != blocks_per_page)) bio = mpage_bio_submit(READ, bio); else *last_block_in_bio = blocks[blocks_per_page - 1]; @@ -331,7 +390,10 @@ mpage_readpages(struct address_space *mapping, struct list_head *pages, unsigned page_idx; sector_t last_block_in_bio = 0; struct pagevec lru_pvec; + struct buffer_head map_bh; + unsigned long first_logical_block = 0; + clear_buffer_mapped(&map_bh); pagevec_init(&lru_pvec, 0); for (page_idx = 0; page_idx < nr_pages; page_idx++) { struct page *page = list_entry(pages->prev, struct page, lru); @@ -342,7 +404,9 @@ mpage_readpages(struct address_space *mapping, struct list_head *pages, page->index, GFP_KERNEL)) { bio = do_mpage_readpage(bio, page, nr_pages - page_idx, - &last_block_in_bio, get_block); + &last_block_in_bio, &map_bh, + &first_logical_block, + get_block); if (!pagevec_add(&lru_pvec, page)) __pagevec_lru_add(&lru_pvec); } else { @@ -364,9 +428,12 @@ int mpage_readpage(struct page *page, get_block_t get_block) { struct bio *bio = NULL; sector_t last_block_in_bio = 0; + struct buffer_head map_bh; + unsigned long first_logical_block = 0; - bio = do_mpage_readpage(bio, page, 1, - &last_block_in_bio, get_block); + clear_buffer_mapped(&map_bh); + bio = do_mpage_readpage(bio, page, 1, &last_block_in_bio, + &map_bh, &first_logical_block, get_block); if (bio) mpage_bio_submit(READ, bio); return 0; @@ -472,6 +539,7 @@ __mpage_writepage(struct bio *bio, struct page *page, get_block_t get_block, for (page_block = 0; page_block < blocks_per_page; ) { map_bh.b_state = 0; + map_bh.b_size = 1 << blkbits; if (get_block(inode, block_in_file, &map_bh, 1)) goto confused; if (buffer_new(&map_bh)) diff --git a/fs/namei.c b/fs/namei.c index c72b940797f..22f6e8d16aa 100644 --- a/fs/namei.c +++ b/fs/namei.c @@ -546,6 +546,22 @@ struct path { struct dentry *dentry; }; +static inline void dput_path(struct path *path, struct nameidata *nd) +{ + dput(path->dentry); + if (path->mnt != nd->mnt) + mntput(path->mnt); +} + +static inline void path_to_nameidata(struct path *path, struct nameidata *nd) +{ + dput(nd->dentry); + if (nd->mnt != path->mnt) + mntput(nd->mnt); + nd->mnt = path->mnt; + nd->dentry = path->dentry; +} + static __always_inline int __do_follow_link(struct path *path, struct nameidata *nd) { int error; @@ -555,8 +571,11 @@ static __always_inline int __do_follow_link(struct path *path, struct nameidata touch_atime(path->mnt, dentry); nd_set_link(nd, NULL); - if (path->mnt == nd->mnt) - mntget(path->mnt); + if (path->mnt != nd->mnt) { + path_to_nameidata(path, nd); + dget(dentry); + } + mntget(path->mnt); cookie = dentry->d_inode->i_op->follow_link(dentry, nd); error = PTR_ERR(cookie); if (!IS_ERR(cookie)) { @@ -573,22 +592,6 @@ static __always_inline int __do_follow_link(struct path *path, struct nameidata return error; } -static inline void dput_path(struct path *path, struct nameidata *nd) -{ - dput(path->dentry); - if (path->mnt != nd->mnt) - mntput(path->mnt); -} - -static inline void path_to_nameidata(struct path *path, struct nameidata *nd) -{ - dput(nd->dentry); - if (nd->mnt != path->mnt) - mntput(nd->mnt); - nd->mnt = path->mnt; - nd->dentry = path->dentry; -} - /* * This limits recursive symlink follows to 8, while * limiting consecutive symlinks to 40. @@ -1353,6 +1356,7 @@ static int may_delete(struct inode *dir,struct dentry *victim,int isdir) return -ENOENT; BUG_ON(victim->d_parent->d_inode != dir); + audit_inode_child(victim->d_name.name, victim->d_inode, dir->i_ino); error = permission(dir,MAY_WRITE | MAY_EXEC, NULL); if (error) @@ -1472,7 +1476,7 @@ int vfs_create(struct inode *dir, struct dentry *dentry, int mode, DQUOT_INIT(dir); error = dir->i_op->create(dir, dentry, mode, nd); if (!error) - fsnotify_create(dir, dentry->d_name.name); + fsnotify_create(dir, dentry); return error; } @@ -1628,6 +1632,12 @@ do_last: goto exit; } + if (IS_ERR(nd->intent.open.file)) { + mutex_unlock(&dir->d_inode->i_mutex); + error = PTR_ERR(nd->intent.open.file); + goto exit_dput; + } + /* Negative dentry, just create the file */ if (!path.dentry->d_inode) { if (!IS_POSIXACL(dir->d_inode)) @@ -1793,7 +1803,7 @@ int vfs_mknod(struct inode *dir, struct dentry *dentry, int mode, dev_t dev) DQUOT_INIT(dir); error = dir->i_op->mknod(dir, dentry, mode, dev); if (!error) - fsnotify_create(dir, dentry->d_name.name); + fsnotify_create(dir, dentry); return error; } @@ -1870,7 +1880,7 @@ int vfs_mkdir(struct inode *dir, struct dentry *dentry, int mode) DQUOT_INIT(dir); error = dir->i_op->mkdir(dir, dentry, mode); if (!error) - fsnotify_mkdir(dir, dentry->d_name.name); + fsnotify_mkdir(dir, dentry); return error; } @@ -2133,7 +2143,7 @@ int vfs_symlink(struct inode *dir, struct dentry *dentry, const char *oldname, i DQUOT_INIT(dir); error = dir->i_op->symlink(dir, dentry, oldname); if (!error) - fsnotify_create(dir, dentry->d_name.name); + fsnotify_create(dir, dentry); return error; } @@ -2210,7 +2220,7 @@ int vfs_link(struct dentry *old_dentry, struct inode *dir, struct dentry *new_de error = dir->i_op->link(old_dentry, dir, new_dentry); mutex_unlock(&old_dentry->d_inode->i_mutex); if (!error) - fsnotify_create(dir, new_dentry->d_name.name); + fsnotify_create(dir, new_dentry); return error; } @@ -2621,16 +2631,27 @@ int __page_symlink(struct inode *inode, const char *symname, int len, int err = -ENOMEM; char *kaddr; +retry: page = find_or_create_page(mapping, 0, gfp_mask); if (!page) goto fail; err = mapping->a_ops->prepare_write(NULL, page, 0, len-1); + if (err == AOP_TRUNCATED_PAGE) { + page_cache_release(page); + goto retry; + } if (err) goto fail_map; kaddr = kmap_atomic(page, KM_USER0); memcpy(kaddr, symname, len-1); kunmap_atomic(kaddr, KM_USER0); - mapping->a_ops->commit_write(NULL, page, 0, len-1); + err = mapping->a_ops->commit_write(NULL, page, 0, len-1); + if (err == AOP_TRUNCATED_PAGE) { + page_cache_release(page); + goto retry; + } + if (err) + goto fail_map; /* * Notice that we are _not_ going to block here - end of page is * unmapped, so this will only try to map the rest of page, see @@ -2640,7 +2661,8 @@ int __page_symlink(struct inode *inode, const char *symname, int len, */ if (!PageUptodate(page)) { err = mapping->a_ops->readpage(NULL, page); - wait_on_page_locked(page); + if (err != AOP_TRUNCATED_PAGE) + wait_on_page_locked(page); } else { unlock_page(page); } diff --git a/fs/namespace.c b/fs/namespace.c index 39c81a8d631..bf478addb85 100644 --- a/fs/namespace.c +++ b/fs/namespace.c @@ -43,9 +43,9 @@ __cacheline_aligned_in_smp DEFINE_SPINLOCK(vfsmount_lock); static int event; -static struct list_head *mount_hashtable; +static struct list_head *mount_hashtable __read_mostly; static int hash_mask __read_mostly, hash_bits __read_mostly; -static kmem_cache_t *mnt_cache; +static kmem_cache_t *mnt_cache __read_mostly; static struct rw_semaphore namespace_sem; /* /sys/fs */ @@ -399,6 +399,44 @@ struct seq_operations mounts_op = { .show = show_vfsmnt }; +static int show_vfsstat(struct seq_file *m, void *v) +{ + struct vfsmount *mnt = v; + int err = 0; + + /* device */ + if (mnt->mnt_devname) { + seq_puts(m, "device "); + mangle(m, mnt->mnt_devname); + } else + seq_puts(m, "no device"); + + /* mount point */ + seq_puts(m, " mounted on "); + seq_path(m, mnt, mnt->mnt_root, " \t\n\\"); + seq_putc(m, ' '); + + /* file system type */ + seq_puts(m, "with fstype "); + mangle(m, mnt->mnt_sb->s_type->name); + + /* optional statistics */ + if (mnt->mnt_sb->s_op->show_stats) { + seq_putc(m, ' '); + err = mnt->mnt_sb->s_op->show_stats(m, mnt); + } + + seq_putc(m, '\n'); + return err; +} + +struct seq_operations mountstats_op = { + .start = m_start, + .next = m_next, + .stop = m_stop, + .show = show_vfsstat, +}; + /** * may_umount_tree - check if a mount tree is busy * @mnt: root of mount tree @@ -421,9 +459,9 @@ int may_umount_tree(struct vfsmount *mnt) spin_unlock(&vfsmount_lock); if (actual_refs > minimum_refs) - return -EBUSY; + return 0; - return 0; + return 1; } EXPORT_SYMBOL(may_umount_tree); @@ -443,10 +481,10 @@ EXPORT_SYMBOL(may_umount_tree); */ int may_umount(struct vfsmount *mnt) { - int ret = 0; + int ret = 1; spin_lock(&vfsmount_lock); if (propagate_mount_busy(mnt, 2)) - ret = -EBUSY; + ret = 0; spin_unlock(&vfsmount_lock); return ret; } diff --git a/fs/ncpfs/dir.c b/fs/ncpfs/dir.c index cfd76f431dc..f0860c602d8 100644 --- a/fs/ncpfs/dir.c +++ b/fs/ncpfs/dir.c @@ -49,7 +49,7 @@ extern int ncp_symlink(struct inode *, struct dentry *, const char *); #define ncp_symlink NULL #endif -struct file_operations ncp_dir_operations = +const struct file_operations ncp_dir_operations = { .read = generic_read_dir, .readdir = ncp_readdir, diff --git a/fs/ncpfs/file.c b/fs/ncpfs/file.c index ebdad8f6398..e6b7c67cf05 100644 --- a/fs/ncpfs/file.c +++ b/fs/ncpfs/file.c @@ -283,7 +283,7 @@ static int ncp_release(struct inode *inode, struct file *file) { return 0; } -struct file_operations ncp_file_operations = +const struct file_operations ncp_file_operations = { .llseek = remote_llseek, .read = ncp_file_read, diff --git a/fs/nfs/callback.c b/fs/nfs/callback.c index fcd97406a77..90c95adc8c1 100644 --- a/fs/nfs/callback.c +++ b/fs/nfs/callback.c @@ -14,6 +14,7 @@ #include <linux/sunrpc/svc.h> #include <linux/sunrpc/svcsock.h> #include <linux/nfs_fs.h> +#include <linux/mutex.h> #include <net/inet_sock.h> @@ -31,7 +32,7 @@ struct nfs_callback_data { }; static struct nfs_callback_data nfs_callback_info; -static DECLARE_MUTEX(nfs_callback_sema); +static DEFINE_MUTEX(nfs_callback_mutex); static struct svc_program nfs4_callback_program; unsigned int nfs_callback_set_tcpport; @@ -55,7 +56,12 @@ static void nfs_callback_svc(struct svc_rqst *rqstp) complete(&nfs_callback_info.started); - while (nfs_callback_info.users != 0 || !signalled()) { + for(;;) { + if (signalled()) { + if (nfs_callback_info.users == 0) + break; + flush_signals(current); + } /* * Listen for a request on the socket */ @@ -73,6 +79,7 @@ static void nfs_callback_svc(struct svc_rqst *rqstp) svc_process(serv, rqstp); } + svc_exit_thread(rqstp); nfs_callback_info.pid = 0; complete(&nfs_callback_info.stopped); unlock_kernel(); @@ -89,7 +96,7 @@ int nfs_callback_up(void) int ret = 0; lock_kernel(); - down(&nfs_callback_sema); + mutex_lock(&nfs_callback_mutex); if (nfs_callback_info.users++ || nfs_callback_info.pid != 0) goto out; init_completion(&nfs_callback_info.started); @@ -115,7 +122,7 @@ int nfs_callback_up(void) nfs_callback_info.serv = serv; wait_for_completion(&nfs_callback_info.started); out: - up(&nfs_callback_sema); + mutex_unlock(&nfs_callback_mutex); unlock_kernel(); return ret; out_destroy: @@ -133,13 +140,15 @@ int nfs_callback_down(void) int ret = 0; lock_kernel(); - down(&nfs_callback_sema); - if (--nfs_callback_info.users || nfs_callback_info.pid == 0) - goto out; - kill_proc(nfs_callback_info.pid, SIGKILL, 1); - wait_for_completion(&nfs_callback_info.stopped); -out: - up(&nfs_callback_sema); + mutex_lock(&nfs_callback_mutex); + nfs_callback_info.users--; + do { + if (nfs_callback_info.users != 0 || nfs_callback_info.pid == 0) + break; + if (kill_proc(nfs_callback_info.pid, SIGKILL, 1) < 0) + break; + } while (wait_for_completion_timeout(&nfs_callback_info.stopped, 5*HZ) == 0); + mutex_unlock(&nfs_callback_mutex); unlock_kernel(); return ret; } diff --git a/fs/nfs/callback_xdr.c b/fs/nfs/callback_xdr.c index 7c33b9a81a9..05c38cf40b6 100644 --- a/fs/nfs/callback_xdr.c +++ b/fs/nfs/callback_xdr.c @@ -330,7 +330,7 @@ static unsigned encode_op_hdr(struct xdr_stream *xdr, uint32_t op, uint32_t res) static unsigned encode_getattr_res(struct svc_rqst *rqstp, struct xdr_stream *xdr, const struct cb_getattrres *res) { - uint32_t *savep; + uint32_t *savep = NULL; unsigned status = res->status; if (unlikely(status != 0)) @@ -358,23 +358,26 @@ static unsigned process_op(struct svc_rqst *rqstp, struct xdr_stream *xdr_in, void *argp, struct xdr_stream *xdr_out, void *resp) { - struct callback_op *op; - unsigned int op_nr; + struct callback_op *op = &callback_ops[0]; + unsigned int op_nr = OP_CB_ILLEGAL; unsigned int status = 0; long maxlen; unsigned res; dprintk("%s: start\n", __FUNCTION__); status = decode_op_hdr(xdr_in, &op_nr); - if (unlikely(status != 0)) { - op_nr = OP_CB_ILLEGAL; - op = &callback_ops[0]; - } else if (unlikely(op_nr != OP_CB_GETATTR && op_nr != OP_CB_RECALL)) { - op_nr = OP_CB_ILLEGAL; - op = &callback_ops[0]; - status = htonl(NFS4ERR_OP_ILLEGAL); - } else - op = &callback_ops[op_nr]; + if (likely(status == 0)) { + switch (op_nr) { + case OP_CB_GETATTR: + case OP_CB_RECALL: + op = &callback_ops[op_nr]; + break; + default: + op_nr = OP_CB_ILLEGAL; + op = &callback_ops[0]; + status = htonl(NFS4ERR_OP_ILLEGAL); + } + } maxlen = xdr_out->end - xdr_out->p; if (maxlen > 0 && maxlen < PAGE_SIZE) { @@ -416,6 +419,7 @@ static int nfs4_callback_compound(struct svc_rqst *rqstp, void *argp, void *resp decode_compound_hdr_arg(&xdr_in, &hdr_arg); hdr_res.taglen = hdr_arg.taglen; hdr_res.tag = hdr_arg.tag; + hdr_res.nops = NULL; encode_compound_hdr_res(&xdr_out, &hdr_res); for (;;) { diff --git a/fs/nfs/delegation.c b/fs/nfs/delegation.c index c6f07c1c71e..d3be923d4e4 100644 --- a/fs/nfs/delegation.c +++ b/fs/nfs/delegation.c @@ -421,3 +421,22 @@ void nfs_delegation_reap_unclaimed(struct nfs4_client *clp) nfs_free_delegation(delegation); } } + +int nfs4_copy_delegation_stateid(nfs4_stateid *dst, struct inode *inode) +{ + struct nfs4_client *clp = NFS_SERVER(inode)->nfs4_state; + struct nfs_inode *nfsi = NFS_I(inode); + struct nfs_delegation *delegation; + int res = 0; + + if (nfsi->delegation_state == 0) + return 0; + spin_lock(&clp->cl_lock); + delegation = nfsi->delegation; + if (delegation != NULL) { + memcpy(dst->data, delegation->stateid.data, sizeof(dst->data)); + res = 1; + } + spin_unlock(&clp->cl_lock); + return res; +} diff --git a/fs/nfs/delegation.h b/fs/nfs/delegation.h index 7a0b2bfce77..3858694652f 100644 --- a/fs/nfs/delegation.h +++ b/fs/nfs/delegation.h @@ -41,6 +41,7 @@ void nfs_delegation_reap_unclaimed(struct nfs4_client *clp); int nfs4_proc_delegreturn(struct inode *inode, struct rpc_cred *cred, const nfs4_stateid *stateid); int nfs4_open_delegation_recall(struct dentry *dentry, struct nfs4_state *state); int nfs4_lock_delegation_recall(struct nfs4_state *state, struct file_lock *fl); +int nfs4_copy_delegation_stateid(nfs4_stateid *dst, struct inode *inode); static inline int nfs_have_delegation(struct inode *inode, int flags) { diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c index a1554bead69..a23f3489416 100644 --- a/fs/nfs/dir.c +++ b/fs/nfs/dir.c @@ -34,6 +34,7 @@ #include "nfs4_fs.h" #include "delegation.h" +#include "iostat.h" #define NFS_PARANOIA 1 /* #define NFS_DEBUG_VERBOSE 1 */ @@ -53,7 +54,7 @@ static int nfs_rename(struct inode *, struct dentry *, static int nfs_fsync_dir(struct file *, struct dentry *, int); static loff_t nfs_llseek_dir(struct file *, loff_t, int); -struct file_operations nfs_dir_operations = { +const struct file_operations nfs_dir_operations = { .llseek = nfs_llseek_dir, .read = generic_read_dir, .readdir = nfs_readdir, @@ -129,6 +130,9 @@ nfs_opendir(struct inode *inode, struct file *filp) { int res = 0; + dfprintk(VFS, "NFS: opendir(%s/%ld)\n", + inode->i_sb->s_id, inode->i_ino); + lock_kernel(); /* Call generic open code in order to cache credentials */ if (!res) @@ -172,7 +176,9 @@ int nfs_readdir_filler(nfs_readdir_descriptor_t *desc, struct page *page) unsigned long timestamp; int error; - dfprintk(VFS, "NFS: nfs_readdir_filler() reading cookie %Lu into page %lu.\n", (long long)desc->entry->cookie, page->index); + dfprintk(DIRCACHE, "NFS: %s: reading cookie %Lu into page %lu\n", + __FUNCTION__, (long long)desc->entry->cookie, + page->index); again: timestamp = jiffies; @@ -244,7 +250,8 @@ int find_dirent(nfs_readdir_descriptor_t *desc) status; while((status = dir_decode(desc)) == 0) { - dfprintk(VFS, "NFS: found cookie %Lu\n", (unsigned long long)entry->cookie); + dfprintk(DIRCACHE, "NFS: %s: examining cookie %Lu\n", + __FUNCTION__, (unsigned long long)entry->cookie); if (entry->prev_cookie == *desc->dir_cookie) break; if (loop_count++ > 200) { @@ -252,7 +259,6 @@ int find_dirent(nfs_readdir_descriptor_t *desc) schedule(); } } - dfprintk(VFS, "NFS: find_dirent() returns %d\n", status); return status; } @@ -276,7 +282,8 @@ int find_dirent_index(nfs_readdir_descriptor_t *desc) if (status) break; - dfprintk(VFS, "NFS: found cookie %Lu at index %Ld\n", (unsigned long long)entry->cookie, desc->current_index); + dfprintk(DIRCACHE, "NFS: found cookie %Lu at index %Ld\n", + (unsigned long long)entry->cookie, desc->current_index); if (desc->file->f_pos == desc->current_index) { *desc->dir_cookie = entry->cookie; @@ -288,7 +295,6 @@ int find_dirent_index(nfs_readdir_descriptor_t *desc) schedule(); } } - dfprintk(VFS, "NFS: find_dirent_index() returns %d\n", status); return status; } @@ -303,7 +309,9 @@ int find_dirent_page(nfs_readdir_descriptor_t *desc) struct page *page; int status; - dfprintk(VFS, "NFS: find_dirent_page() searching directory page %ld\n", desc->page_index); + dfprintk(DIRCACHE, "NFS: %s: searching page %ld for target %Lu\n", + __FUNCTION__, desc->page_index, + (long long) *desc->dir_cookie); page = read_cache_page(inode->i_mapping, desc->page_index, (filler_t *)nfs_readdir_filler, desc); @@ -324,7 +332,7 @@ int find_dirent_page(nfs_readdir_descriptor_t *desc) if (status < 0) dir_page_release(desc); out: - dfprintk(VFS, "NFS: find_dirent_page() returns %d\n", status); + dfprintk(DIRCACHE, "NFS: %s: returns %d\n", __FUNCTION__, status); return status; read_error: page_cache_release(page); @@ -346,13 +354,15 @@ int readdir_search_pagecache(nfs_readdir_descriptor_t *desc) /* Always search-by-index from the beginning of the cache */ if (*desc->dir_cookie == 0) { - dfprintk(VFS, "NFS: readdir_search_pagecache() searching for offset %Ld\n", (long long)desc->file->f_pos); + dfprintk(DIRCACHE, "NFS: readdir_search_pagecache() searching for offset %Ld\n", + (long long)desc->file->f_pos); desc->page_index = 0; desc->entry->cookie = desc->entry->prev_cookie = 0; desc->entry->eof = 0; desc->current_index = 0; } else - dfprintk(VFS, "NFS: readdir_search_pagecache() searching for cookie %Lu\n", (unsigned long long)*desc->dir_cookie); + dfprintk(DIRCACHE, "NFS: readdir_search_pagecache() searching for cookie %Lu\n", + (unsigned long long)*desc->dir_cookie); for (;;) { res = find_dirent_page(desc); @@ -365,7 +375,8 @@ int readdir_search_pagecache(nfs_readdir_descriptor_t *desc) schedule(); } } - dfprintk(VFS, "NFS: readdir_search_pagecache() returned %d\n", res); + + dfprintk(DIRCACHE, "NFS: %s: returns %d\n", __FUNCTION__, res); return res; } @@ -390,7 +401,8 @@ int nfs_do_filldir(nfs_readdir_descriptor_t *desc, void *dirent, int loop_count = 0, res; - dfprintk(VFS, "NFS: nfs_do_filldir() filling starting @ cookie %Lu\n", (long long)entry->cookie); + dfprintk(DIRCACHE, "NFS: nfs_do_filldir() filling starting @ cookie %Lu\n", + (unsigned long long)entry->cookie); for(;;) { unsigned d_type = DT_UNKNOWN; @@ -427,7 +439,8 @@ int nfs_do_filldir(nfs_readdir_descriptor_t *desc, void *dirent, dir_page_release(desc); if (dentry != NULL) dput(dentry); - dfprintk(VFS, "NFS: nfs_do_filldir() filling ended @ cookie %Lu; returning = %d\n", (unsigned long long)*desc->dir_cookie, res); + dfprintk(DIRCACHE, "NFS: nfs_do_filldir() filling ended @ cookie %Lu; returning = %d\n", + (unsigned long long)*desc->dir_cookie, res); return res; } @@ -453,7 +466,8 @@ int uncached_readdir(nfs_readdir_descriptor_t *desc, void *dirent, struct page *page = NULL; int status; - dfprintk(VFS, "NFS: uncached_readdir() searching for cookie %Lu\n", (unsigned long long)*desc->dir_cookie); + dfprintk(DIRCACHE, "NFS: uncached_readdir() searching for cookie %Lu\n", + (unsigned long long)*desc->dir_cookie); page = alloc_page(GFP_HIGHUSER); if (!page) { @@ -485,7 +499,8 @@ int uncached_readdir(nfs_readdir_descriptor_t *desc, void *dirent, desc->entry->cookie = desc->entry->prev_cookie = 0; desc->entry->eof = 0; out: - dfprintk(VFS, "NFS: uncached_readdir() returns %d\n", status); + dfprintk(DIRCACHE, "NFS: %s: returns %d\n", + __FUNCTION__, status); return status; out_release: dir_page_release(desc); @@ -507,6 +522,11 @@ static int nfs_readdir(struct file *filp, void *dirent, filldir_t filldir) struct nfs_fattr fattr; long res; + dfprintk(VFS, "NFS: readdir(%s/%s) starting at cookie %Lu\n", + dentry->d_parent->d_name.name, dentry->d_name.name, + (long long)filp->f_pos); + nfs_inc_stats(inode, NFSIOS_VFSGETDENTS); + lock_kernel(); res = nfs_revalidate_inode(NFS_SERVER(inode), inode); @@ -566,9 +586,12 @@ static int nfs_readdir(struct file *filp, void *dirent, filldir_t filldir) } } unlock_kernel(); - if (res < 0) - return res; - return 0; + if (res > 0) + res = 0; + dfprintk(VFS, "NFS: readdir(%s/%s) returns %ld\n", + dentry->d_parent->d_name.name, dentry->d_name.name, + res); + return res; } loff_t nfs_llseek_dir(struct file *filp, loff_t offset, int origin) @@ -599,6 +622,10 @@ out: */ int nfs_fsync_dir(struct file *filp, struct dentry *dentry, int datasync) { + dfprintk(VFS, "NFS: fsync_dir(%s/%s) datasync %d\n", + dentry->d_parent->d_name.name, dentry->d_name.name, + datasync); + return 0; } @@ -713,6 +740,7 @@ static int nfs_lookup_revalidate(struct dentry * dentry, struct nameidata *nd) parent = dget_parent(dentry); lock_kernel(); dir = parent->d_inode; + nfs_inc_stats(dir, NFSIOS_DENTRYREVALIDATE); inode = dentry->d_inode; if (!inode) { @@ -722,8 +750,9 @@ static int nfs_lookup_revalidate(struct dentry * dentry, struct nameidata *nd) } if (is_bad_inode(inode)) { - dfprintk(VFS, "nfs_lookup_validate: %s/%s has dud inode\n", - dentry->d_parent->d_name.name, dentry->d_name.name); + dfprintk(LOOKUPCACHE, "%s: %s/%s has dud inode\n", + __FUNCTION__, dentry->d_parent->d_name.name, + dentry->d_name.name); goto out_bad; } @@ -755,6 +784,9 @@ static int nfs_lookup_revalidate(struct dentry * dentry, struct nameidata *nd) out_valid: unlock_kernel(); dput(parent); + dfprintk(LOOKUPCACHE, "NFS: %s(%s/%s) is valid\n", + __FUNCTION__, dentry->d_parent->d_name.name, + dentry->d_name.name); return 1; out_zap_parent: nfs_zap_caches(dir); @@ -771,6 +803,9 @@ out_zap_parent: d_drop(dentry); unlock_kernel(); dput(parent); + dfprintk(LOOKUPCACHE, "NFS: %s(%s/%s) is invalid\n", + __FUNCTION__, dentry->d_parent->d_name.name, + dentry->d_name.name); return 0; } @@ -844,6 +879,7 @@ static struct dentry *nfs_lookup(struct inode *dir, struct dentry * dentry, stru dfprintk(VFS, "NFS: lookup(%s/%s)\n", dentry->d_parent->d_name.name, dentry->d_name.name); + nfs_inc_stats(dir, NFSIOS_VFSLOOKUP); res = ERR_PTR(-ENAMETOOLONG); if (dentry->d_name.len > NFS_SERVER(dir)->namelen) @@ -865,9 +901,9 @@ static struct dentry *nfs_lookup(struct inode *dir, struct dentry * dentry, stru res = ERR_PTR(error); goto out_unlock; } - res = ERR_PTR(-EACCES); inode = nfs_fhget(dentry->d_sb, &fhandle, &fattr); - if (!inode) + res = (struct dentry *)inode; + if (IS_ERR(res)) goto out_unlock; no_entry: res = d_add_unique(dentry, inode); @@ -912,6 +948,9 @@ static struct dentry *nfs_atomic_lookup(struct inode *dir, struct dentry *dentry struct dentry *res = NULL; int error; + dfprintk(VFS, "NFS: atomic_lookup(%s/%ld), %s\n", + dir->i_sb->s_id, dir->i_ino, dentry->d_name.name); + /* Check that we are indeed trying to open this file */ if (!is_atomic_open(dir, nd)) goto no_open; @@ -1057,7 +1096,7 @@ static struct dentry *nfs_readdir_lookup(nfs_readdir_descriptor_t *desc) return NULL; dentry->d_op = NFS_PROTO(dir)->dentry_ops; inode = nfs_fhget(dentry->d_sb, entry->fh, entry->fattr); - if (!inode) { + if (IS_ERR(inode)) { dput(dentry); return NULL; } @@ -1095,9 +1134,9 @@ int nfs_instantiate(struct dentry *dentry, struct nfs_fh *fhandle, if (error < 0) goto out_err; } - error = -ENOMEM; inode = nfs_fhget(dentry->d_sb, fhandle, fattr); - if (inode == NULL) + error = PTR_ERR(inode); + if (IS_ERR(inode)) goto out_err; d_instantiate(dentry, inode); return 0; @@ -1119,8 +1158,8 @@ static int nfs_create(struct inode *dir, struct dentry *dentry, int mode, int error; int open_flags = 0; - dfprintk(VFS, "NFS: create(%s/%ld, %s\n", dir->i_sb->s_id, - dir->i_ino, dentry->d_name.name); + dfprintk(VFS, "NFS: create(%s/%ld), %s\n", + dir->i_sb->s_id, dir->i_ino, dentry->d_name.name); attr.ia_mode = mode; attr.ia_valid = ATTR_MODE; @@ -1153,8 +1192,8 @@ nfs_mknod(struct inode *dir, struct dentry *dentry, int mode, dev_t rdev) struct iattr attr; int status; - dfprintk(VFS, "NFS: mknod(%s/%ld, %s\n", dir->i_sb->s_id, - dir->i_ino, dentry->d_name.name); + dfprintk(VFS, "NFS: mknod(%s/%ld), %s\n", + dir->i_sb->s_id, dir->i_ino, dentry->d_name.name); if (!new_valid_dev(rdev)) return -EINVAL; @@ -1186,8 +1225,8 @@ static int nfs_mkdir(struct inode *dir, struct dentry *dentry, int mode) struct iattr attr; int error; - dfprintk(VFS, "NFS: mkdir(%s/%ld, %s\n", dir->i_sb->s_id, - dir->i_ino, dentry->d_name.name); + dfprintk(VFS, "NFS: mkdir(%s/%ld), %s\n", + dir->i_sb->s_id, dir->i_ino, dentry->d_name.name); attr.ia_valid = ATTR_MODE; attr.ia_mode = mode | S_IFDIR; @@ -1212,8 +1251,8 @@ static int nfs_rmdir(struct inode *dir, struct dentry *dentry) { int error; - dfprintk(VFS, "NFS: rmdir(%s/%ld, %s\n", dir->i_sb->s_id, - dir->i_ino, dentry->d_name.name); + dfprintk(VFS, "NFS: rmdir(%s/%ld), %s\n", + dir->i_sb->s_id, dir->i_ino, dentry->d_name.name); lock_kernel(); nfs_begin_data_update(dir); @@ -1241,6 +1280,7 @@ static int nfs_sillyrename(struct inode *dir, struct dentry *dentry) dfprintk(VFS, "NFS: silly-rename(%s/%s, ct=%d)\n", dentry->d_parent->d_name.name, dentry->d_name.name, atomic_read(&dentry->d_count)); + nfs_inc_stats(dir, NFSIOS_SILLYRENAME); #ifdef NFS_PARANOIA if (!dentry->d_inode) @@ -1268,8 +1308,8 @@ dentry->d_parent->d_name.name, dentry->d_name.name); sillycounter++; sprintf(suffix, "%*.*x", countersize, countersize, sillycounter); - dfprintk(VFS, "trying to rename %s to %s\n", - dentry->d_name.name, silly); + dfprintk(VFS, "NFS: trying to rename %s to %s\n", + dentry->d_name.name, silly); sdentry = lookup_one_len(silly, dentry->d_parent, slen); /* @@ -1640,6 +1680,8 @@ int nfs_permission(struct inode *inode, int mask, struct nameidata *nd) struct rpc_cred *cred; int res = 0; + nfs_inc_stats(inode, NFSIOS_VFSACCESS); + if (mask == 0) goto out; /* Is this sys_access() ? */ @@ -1679,13 +1721,15 @@ force_lookup: res = PTR_ERR(cred); unlock_kernel(); out: + dfprintk(VFS, "NFS: permission(%s/%ld), mask=0x%x, res=%d\n", + inode->i_sb->s_id, inode->i_ino, mask, res); return res; out_notsup: res = nfs_revalidate_inode(NFS_SERVER(inode), inode); if (res == 0) res = generic_permission(inode, mask, NULL); unlock_kernel(); - return res; + goto out; } /* diff --git a/fs/nfs/direct.c b/fs/nfs/direct.c index 4ae2f3b33fe..0f583cb16dd 100644 --- a/fs/nfs/direct.c +++ b/fs/nfs/direct.c @@ -7,11 +7,11 @@ * * There are important applications whose performance or correctness * depends on uncached access to file data. Database clusters - * (multiple copies of the same instance running on separate hosts) + * (multiple copies of the same instance running on separate hosts) * implement their own cache coherency protocol that subsumes file - * system cache protocols. Applications that process datasets - * considerably larger than the client's memory do not always benefit - * from a local cache. A streaming video server, for instance, has no + * system cache protocols. Applications that process datasets + * considerably larger than the client's memory do not always benefit + * from a local cache. A streaming video server, for instance, has no * need to cache the contents of a file. * * When an application requests uncached I/O, all read and write requests @@ -34,6 +34,7 @@ * 08 Jun 2003 Port to 2.5 APIs --cel * 31 Mar 2004 Handle direct I/O without VFS support --cel * 15 Sep 2004 Parallel async reads --cel + * 04 May 2005 support O_DIRECT with aio --cel * */ @@ -54,10 +55,10 @@ #include <asm/uaccess.h> #include <asm/atomic.h> +#include "iostat.h" + #define NFSDBG_FACILITY NFSDBG_VFS -#define MAX_DIRECTIO_SIZE (4096UL << PAGE_SHIFT) -static void nfs_free_user_pages(struct page **pages, int npages, int do_dirty); static kmem_cache_t *nfs_direct_cachep; /* @@ -65,38 +66,78 @@ static kmem_cache_t *nfs_direct_cachep; */ struct nfs_direct_req { struct kref kref; /* release manager */ - struct list_head list; /* nfs_read_data structs */ - wait_queue_head_t wait; /* wait for i/o completion */ + + /* I/O parameters */ + struct list_head list, /* nfs_read/write_data structs */ + rewrite_list; /* saved nfs_write_data structs */ + struct nfs_open_context *ctx; /* file open context info */ + struct kiocb * iocb; /* controlling i/o request */ + struct inode * inode; /* target file of i/o */ + unsigned long user_addr; /* location of user's buffer */ + size_t user_count; /* total bytes to move */ + loff_t pos; /* starting offset in file */ struct page ** pages; /* pages in our buffer */ unsigned int npages; /* count of pages */ - atomic_t complete, /* i/os we're waiting for */ - count, /* bytes actually processed */ + + /* completion state */ + spinlock_t lock; /* protect completion state */ + int outstanding; /* i/os we're waiting for */ + ssize_t count, /* bytes actually processed */ error; /* any reported error */ + struct completion completion; /* wait for i/o completion */ + + /* commit state */ + struct nfs_write_data * commit_data; /* special write_data for commits */ + int flags; +#define NFS_ODIRECT_DO_COMMIT (1) /* an unstable reply was received */ +#define NFS_ODIRECT_RESCHED_WRITES (2) /* write verification failed */ + struct nfs_writeverf verf; /* unstable write verifier */ }; +static void nfs_direct_write_schedule(struct nfs_direct_req *dreq, int sync); +static void nfs_direct_write_complete(struct nfs_direct_req *dreq, struct inode *inode); /** - * nfs_get_user_pages - find and set up pages underlying user's buffer - * rw: direction (read or write) - * user_addr: starting address of this segment of user's buffer - * count: size of this segment - * @pages: returned array of page struct pointers underlying user's buffer + * nfs_direct_IO - NFS address space operation for direct I/O + * @rw: direction (read or write) + * @iocb: target I/O control block + * @iov: array of vectors that define I/O buffer + * @pos: offset in file to begin the operation + * @nr_segs: size of iovec array + * + * The presence of this routine in the address space ops vector means + * the NFS client supports direct I/O. However, we shunt off direct + * read and write requests before the VFS gets them, so this method + * should never be called. */ -static inline int -nfs_get_user_pages(int rw, unsigned long user_addr, size_t size, - struct page ***pages) +ssize_t nfs_direct_IO(int rw, struct kiocb *iocb, const struct iovec *iov, loff_t pos, unsigned long nr_segs) +{ + struct dentry *dentry = iocb->ki_filp->f_dentry; + + dprintk("NFS: nfs_direct_IO (%s) off/no(%Ld/%lu) EINVAL\n", + dentry->d_name.name, (long long) pos, nr_segs); + + return -EINVAL; +} + +static void nfs_free_user_pages(struct page **pages, int npages, int do_dirty) +{ + int i; + for (i = 0; i < npages; i++) { + struct page *page = pages[i]; + if (do_dirty && !PageCompound(page)) + set_page_dirty_lock(page); + page_cache_release(page); + } + kfree(pages); +} + +static inline int nfs_get_user_pages(int rw, unsigned long user_addr, size_t size, struct page ***pages) { int result = -ENOMEM; unsigned long page_count; size_t array_size; - /* set an arbitrary limit to prevent type overflow */ - /* XXX: this can probably be as large as INT_MAX */ - if (size > MAX_DIRECTIO_SIZE) { - *pages = NULL; - return -EFBIG; - } - page_count = (user_addr + size + PAGE_SIZE - 1) >> PAGE_SHIFT; page_count -= user_addr >> PAGE_SHIFT; @@ -108,75 +149,117 @@ nfs_get_user_pages(int rw, unsigned long user_addr, size_t size, page_count, (rw == READ), 0, *pages, NULL); up_read(¤t->mm->mmap_sem); - /* - * If we got fewer pages than expected from get_user_pages(), - * the user buffer runs off the end of a mapping; return EFAULT. - */ - if (result >= 0 && result < page_count) { - nfs_free_user_pages(*pages, result, 0); + if (result != page_count) { + /* + * If we got fewer pages than expected from + * get_user_pages(), the user buffer runs off the + * end of a mapping; return EFAULT. + */ + if (result >= 0) { + nfs_free_user_pages(*pages, result, 0); + result = -EFAULT; + } else + kfree(*pages); *pages = NULL; - result = -EFAULT; } } return result; } -/** - * nfs_free_user_pages - tear down page struct array - * @pages: array of page struct pointers underlying target buffer - * @npages: number of pages in the array - * @do_dirty: dirty the pages as we release them - */ -static void -nfs_free_user_pages(struct page **pages, int npages, int do_dirty) +static inline struct nfs_direct_req *nfs_direct_req_alloc(void) { - int i; - for (i = 0; i < npages; i++) { - struct page *page = pages[i]; - if (do_dirty && !PageCompound(page)) - set_page_dirty_lock(page); - page_cache_release(page); - } - kfree(pages); + struct nfs_direct_req *dreq; + + dreq = kmem_cache_alloc(nfs_direct_cachep, SLAB_KERNEL); + if (!dreq) + return NULL; + + kref_init(&dreq->kref); + init_completion(&dreq->completion); + INIT_LIST_HEAD(&dreq->list); + INIT_LIST_HEAD(&dreq->rewrite_list); + dreq->iocb = NULL; + dreq->ctx = NULL; + spin_lock_init(&dreq->lock); + dreq->outstanding = 0; + dreq->count = 0; + dreq->error = 0; + dreq->flags = 0; + + return dreq; } -/** - * nfs_direct_req_release - release nfs_direct_req structure for direct read - * @kref: kref object embedded in an nfs_direct_req structure - * - */ static void nfs_direct_req_release(struct kref *kref) { struct nfs_direct_req *dreq = container_of(kref, struct nfs_direct_req, kref); + + if (dreq->ctx != NULL) + put_nfs_open_context(dreq->ctx); kmem_cache_free(nfs_direct_cachep, dreq); } -/** - * nfs_direct_read_alloc - allocate nfs_read_data structures for direct read - * @count: count of bytes for the read request - * @rsize: local rsize setting +/* + * Collects and returns the final error value/byte-count. + */ +static ssize_t nfs_direct_wait(struct nfs_direct_req *dreq) +{ + ssize_t result = -EIOCBQUEUED; + + /* Async requests don't wait here */ + if (dreq->iocb) + goto out; + + result = wait_for_completion_interruptible(&dreq->completion); + + if (!result) + result = dreq->error; + if (!result) + result = dreq->count; + +out: + kref_put(&dreq->kref, nfs_direct_req_release); + return (ssize_t) result; +} + +/* + * We must hold a reference to all the pages in this direct read request + * until the RPCs complete. This could be long *after* we are woken up in + * nfs_direct_wait (for instance, if someone hits ^C on a slow server). * + * In addition, synchronous I/O uses a stack-allocated iocb. Thus we + * can't trust the iocb is still valid here if this is a synchronous + * request. If the waiter is woken prematurely, the iocb is long gone. + */ +static void nfs_direct_complete(struct nfs_direct_req *dreq) +{ + nfs_free_user_pages(dreq->pages, dreq->npages, 1); + + if (dreq->iocb) { + long res = (long) dreq->error; + if (!res) + res = (long) dreq->count; + aio_complete(dreq->iocb, res, 0); + } + complete_all(&dreq->completion); + + kref_put(&dreq->kref, nfs_direct_req_release); +} + +/* * Note we also set the number of requests we have in the dreq when we are * done. This prevents races with I/O completion so we will always wait * until all requests have been dispatched and completed. */ -static struct nfs_direct_req *nfs_direct_read_alloc(size_t nbytes, unsigned int rsize) +static struct nfs_direct_req *nfs_direct_read_alloc(size_t nbytes, size_t rsize) { struct list_head *list; struct nfs_direct_req *dreq; - unsigned int reads = 0; unsigned int rpages = (rsize + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT; - dreq = kmem_cache_alloc(nfs_direct_cachep, SLAB_KERNEL); + dreq = nfs_direct_req_alloc(); if (!dreq) return NULL; - kref_init(&dreq->kref); - init_waitqueue_head(&dreq->wait); - INIT_LIST_HEAD(&dreq->list); - atomic_set(&dreq->count, 0); - atomic_set(&dreq->error, 0); - list = &dreq->list; for(;;) { struct nfs_read_data *data = nfs_readdata_alloc(rpages); @@ -196,72 +279,70 @@ static struct nfs_direct_req *nfs_direct_read_alloc(size_t nbytes, unsigned int list_add(&data->pages, list); data->req = (struct nfs_page *) dreq; - reads++; + dreq->outstanding++; if (nbytes <= rsize) break; nbytes -= rsize; } kref_get(&dreq->kref); - atomic_set(&dreq->complete, reads); return dreq; } -/** - * nfs_direct_read_result - handle a read reply for a direct read request - * @data: address of NFS READ operation control block - * @status: status of this NFS READ operation - * - * We must hold a reference to all the pages in this direct read request - * until the RPCs complete. This could be long *after* we are woken up in - * nfs_direct_read_wait (for instance, if someone hits ^C on a slow server). - */ -static void nfs_direct_read_result(struct nfs_read_data *data, int status) +static void nfs_direct_read_result(struct rpc_task *task, void *calldata) { + struct nfs_read_data *data = calldata; struct nfs_direct_req *dreq = (struct nfs_direct_req *) data->req; - if (likely(status >= 0)) - atomic_add(data->res.count, &dreq->count); + if (nfs_readpage_result(task, data) != 0) + return; + + spin_lock(&dreq->lock); + + if (likely(task->tk_status >= 0)) + dreq->count += data->res.count; else - atomic_set(&dreq->error, status); + dreq->error = task->tk_status; - if (unlikely(atomic_dec_and_test(&dreq->complete))) { - nfs_free_user_pages(dreq->pages, dreq->npages, 1); - wake_up(&dreq->wait); - kref_put(&dreq->kref, nfs_direct_req_release); + if (--dreq->outstanding) { + spin_unlock(&dreq->lock); + return; } + + spin_unlock(&dreq->lock); + nfs_direct_complete(dreq); } -/** - * nfs_direct_read_schedule - dispatch NFS READ operations for a direct read - * @dreq: address of nfs_direct_req struct for this request - * @inode: target inode - * @ctx: target file open context - * @user_addr: starting address of this segment of user's buffer - * @count: size of this segment - * @file_offset: offset in file to begin the operation - * +static const struct rpc_call_ops nfs_read_direct_ops = { + .rpc_call_done = nfs_direct_read_result, + .rpc_release = nfs_readdata_release, +}; + +/* * For each nfs_read_data struct that was allocated on the list, dispatch * an NFS READ operation */ -static void nfs_direct_read_schedule(struct nfs_direct_req *dreq, - struct inode *inode, struct nfs_open_context *ctx, - unsigned long user_addr, size_t count, loff_t file_offset) +static void nfs_direct_read_schedule(struct nfs_direct_req *dreq) { + struct nfs_open_context *ctx = dreq->ctx; + struct inode *inode = ctx->dentry->d_inode; struct list_head *list = &dreq->list; struct page **pages = dreq->pages; + size_t count = dreq->user_count; + loff_t pos = dreq->pos; + size_t rsize = NFS_SERVER(inode)->rsize; unsigned int curpage, pgbase; - unsigned int rsize = NFS_SERVER(inode)->rsize; curpage = 0; - pgbase = user_addr & ~PAGE_MASK; + pgbase = dreq->user_addr & ~PAGE_MASK; do { struct nfs_read_data *data; - unsigned int bytes; + size_t bytes; bytes = rsize; if (count < rsize) bytes = count; + BUG_ON(list_empty(list)); data = list_entry(list->next, struct nfs_read_data, pages); list_del_init(&data->pages); @@ -269,7 +350,7 @@ static void nfs_direct_read_schedule(struct nfs_direct_req *dreq, data->cred = ctx->cred; data->args.fh = NFS_FH(inode); data->args.context = ctx; - data->args.offset = file_offset; + data->args.offset = pos; data->args.pgbase = pgbase; data->args.pages = &pages[curpage]; data->args.count = bytes; @@ -277,77 +358,38 @@ static void nfs_direct_read_schedule(struct nfs_direct_req *dreq, data->res.eof = 0; data->res.count = bytes; + rpc_init_task(&data->task, NFS_CLIENT(inode), RPC_TASK_ASYNC, + &nfs_read_direct_ops, data); NFS_PROTO(inode)->read_setup(data); data->task.tk_cookie = (unsigned long) inode; - data->complete = nfs_direct_read_result; lock_kernel(); rpc_execute(&data->task); unlock_kernel(); - dfprintk(VFS, "NFS: %4d initiated direct read call (req %s/%Ld, %u bytes @ offset %Lu)\n", + dfprintk(VFS, "NFS: %5u initiated direct read call (req %s/%Ld, %zu bytes @ offset %Lu)\n", data->task.tk_pid, inode->i_sb->s_id, (long long)NFS_FILEID(inode), bytes, (unsigned long long)data->args.offset); - file_offset += bytes; + pos += bytes; pgbase += bytes; curpage += pgbase >> PAGE_SHIFT; pgbase &= ~PAGE_MASK; count -= bytes; } while (count != 0); + BUG_ON(!list_empty(list)); } -/** - * nfs_direct_read_wait - wait for I/O completion for direct reads - * @dreq: request on which we are to wait - * @intr: whether or not this wait can be interrupted - * - * Collects and returns the final error value/byte-count. - */ -static ssize_t nfs_direct_read_wait(struct nfs_direct_req *dreq, int intr) -{ - int result = 0; - - if (intr) { - result = wait_event_interruptible(dreq->wait, - (atomic_read(&dreq->complete) == 0)); - } else { - wait_event(dreq->wait, (atomic_read(&dreq->complete) == 0)); - } - - if (!result) - result = atomic_read(&dreq->error); - if (!result) - result = atomic_read(&dreq->count); - - kref_put(&dreq->kref, nfs_direct_req_release); - return (ssize_t) result; -} - -/** - * nfs_direct_read_seg - Read in one iov segment. Generate separate - * read RPCs for each "rsize" bytes. - * @inode: target inode - * @ctx: target file open context - * @user_addr: starting address of this segment of user's buffer - * @count: size of this segment - * @file_offset: offset in file to begin the operation - * @pages: array of addresses of page structs defining user's buffer - * @nr_pages: number of pages in the array - * - */ -static ssize_t nfs_direct_read_seg(struct inode *inode, - struct nfs_open_context *ctx, unsigned long user_addr, - size_t count, loff_t file_offset, struct page **pages, - unsigned int nr_pages) +static ssize_t nfs_direct_read(struct kiocb *iocb, unsigned long user_addr, size_t count, loff_t pos, struct page **pages, unsigned int nr_pages) { ssize_t result; sigset_t oldset; + struct inode *inode = iocb->ki_filp->f_mapping->host; struct rpc_clnt *clnt = NFS_CLIENT(inode); struct nfs_direct_req *dreq; @@ -355,284 +397,350 @@ static ssize_t nfs_direct_read_seg(struct inode *inode, if (!dreq) return -ENOMEM; + dreq->user_addr = user_addr; + dreq->user_count = count; + dreq->pos = pos; dreq->pages = pages; dreq->npages = nr_pages; + dreq->inode = inode; + dreq->ctx = get_nfs_open_context((struct nfs_open_context *)iocb->ki_filp->private_data); + if (!is_sync_kiocb(iocb)) + dreq->iocb = iocb; + nfs_add_stats(inode, NFSIOS_DIRECTREADBYTES, count); rpc_clnt_sigmask(clnt, &oldset); - nfs_direct_read_schedule(dreq, inode, ctx, user_addr, count, - file_offset); - result = nfs_direct_read_wait(dreq, clnt->cl_intr); + nfs_direct_read_schedule(dreq); + result = nfs_direct_wait(dreq); rpc_clnt_sigunmask(clnt, &oldset); return result; } -/** - * nfs_direct_read - For each iov segment, map the user's buffer - * then generate read RPCs. - * @inode: target inode - * @ctx: target file open context - * @iov: array of vectors that define I/O buffer - * file_offset: offset in file to begin the operation - * nr_segs: size of iovec array - * - * We've already pushed out any non-direct writes so that this read - * will see them when we read from the server. - */ -static ssize_t -nfs_direct_read(struct inode *inode, struct nfs_open_context *ctx, - const struct iovec *iov, loff_t file_offset, - unsigned long nr_segs) +static void nfs_direct_free_writedata(struct nfs_direct_req *dreq) { - ssize_t tot_bytes = 0; - unsigned long seg = 0; - - while ((seg < nr_segs) && (tot_bytes >= 0)) { - ssize_t result; - int page_count; - struct page **pages; - const struct iovec *vec = &iov[seg++]; - unsigned long user_addr = (unsigned long) vec->iov_base; - size_t size = vec->iov_len; - - page_count = nfs_get_user_pages(READ, user_addr, size, &pages); - if (page_count < 0) { - nfs_free_user_pages(pages, 0, 0); - if (tot_bytes > 0) - break; - return page_count; - } + list_splice_init(&dreq->rewrite_list, &dreq->list); + while (!list_empty(&dreq->list)) { + struct nfs_write_data *data = list_entry(dreq->list.next, struct nfs_write_data, pages); + list_del(&data->pages); + nfs_writedata_release(data); + } +} - result = nfs_direct_read_seg(inode, ctx, user_addr, size, - file_offset, pages, page_count); +#if defined(CONFIG_NFS_V3) || defined(CONFIG_NFS_V4) +static void nfs_direct_write_reschedule(struct nfs_direct_req *dreq) +{ + struct list_head *pos; - if (result <= 0) { - if (tot_bytes > 0) - break; - return result; - } - tot_bytes += result; - file_offset += result; - if (result < size) - break; + list_splice_init(&dreq->rewrite_list, &dreq->list); + list_for_each(pos, &dreq->list) + dreq->outstanding++; + dreq->count = 0; + + nfs_direct_write_schedule(dreq, FLUSH_STABLE); +} + +static void nfs_direct_commit_result(struct rpc_task *task, void *calldata) +{ + struct nfs_write_data *data = calldata; + struct nfs_direct_req *dreq = (struct nfs_direct_req *) data->req; + + /* Call the NFS version-specific code */ + if (NFS_PROTO(data->inode)->commit_done(task, data) != 0) + return; + if (unlikely(task->tk_status < 0)) { + dreq->error = task->tk_status; + dreq->flags = NFS_ODIRECT_RESCHED_WRITES; + } + if (memcmp(&dreq->verf, &data->verf, sizeof(data->verf))) { + dprintk("NFS: %5u commit verify failed\n", task->tk_pid); + dreq->flags = NFS_ODIRECT_RESCHED_WRITES; } - return tot_bytes; + dprintk("NFS: %5u commit returned %d\n", task->tk_pid, task->tk_status); + nfs_direct_write_complete(dreq, data->inode); } -/** - * nfs_direct_write_seg - Write out one iov segment. Generate separate - * write RPCs for each "wsize" bytes, then commit. - * @inode: target inode - * @ctx: target file open context - * user_addr: starting address of this segment of user's buffer - * count: size of this segment - * file_offset: offset in file to begin the operation - * @pages: array of addresses of page structs defining user's buffer - * nr_pages: size of pages array - */ -static ssize_t nfs_direct_write_seg(struct inode *inode, - struct nfs_open_context *ctx, unsigned long user_addr, - size_t count, loff_t file_offset, struct page **pages, - int nr_pages) +static const struct rpc_call_ops nfs_commit_direct_ops = { + .rpc_call_done = nfs_direct_commit_result, + .rpc_release = nfs_commit_release, +}; + +static void nfs_direct_commit_schedule(struct nfs_direct_req *dreq) { - const unsigned int wsize = NFS_SERVER(inode)->wsize; - size_t request; - int curpage, need_commit; - ssize_t result, tot_bytes; - struct nfs_writeverf first_verf; - struct nfs_write_data *wdata; - - wdata = nfs_writedata_alloc(NFS_SERVER(inode)->wpages); - if (!wdata) - return -ENOMEM; + struct nfs_write_data *data = dreq->commit_data; + struct rpc_task *task = &data->task; - wdata->inode = inode; - wdata->cred = ctx->cred; - wdata->args.fh = NFS_FH(inode); - wdata->args.context = ctx; - wdata->args.stable = NFS_UNSTABLE; - if (IS_SYNC(inode) || NFS_PROTO(inode)->version == 2 || count <= wsize) - wdata->args.stable = NFS_FILE_SYNC; - wdata->res.fattr = &wdata->fattr; - wdata->res.verf = &wdata->verf; + data->inode = dreq->inode; + data->cred = dreq->ctx->cred; - nfs_begin_data_update(inode); -retry: - need_commit = 0; - tot_bytes = 0; - curpage = 0; - request = count; - wdata->args.pgbase = user_addr & ~PAGE_MASK; - wdata->args.offset = file_offset; - do { - wdata->args.count = request; - if (wdata->args.count > wsize) - wdata->args.count = wsize; - wdata->args.pages = &pages[curpage]; + data->args.fh = NFS_FH(data->inode); + data->args.offset = dreq->pos; + data->args.count = dreq->user_count; + data->res.count = 0; + data->res.fattr = &data->fattr; + data->res.verf = &data->verf; - dprintk("NFS: direct write: c=%u o=%Ld ua=%lu, pb=%u, cp=%u\n", - wdata->args.count, (long long) wdata->args.offset, - user_addr + tot_bytes, wdata->args.pgbase, curpage); + rpc_init_task(&data->task, NFS_CLIENT(dreq->inode), RPC_TASK_ASYNC, + &nfs_commit_direct_ops, data); + NFS_PROTO(data->inode)->commit_setup(data, 0); - lock_kernel(); - result = NFS_PROTO(inode)->write(wdata); - unlock_kernel(); + data->task.tk_priority = RPC_PRIORITY_NORMAL; + data->task.tk_cookie = (unsigned long)data->inode; + /* Note: task.tk_ops->rpc_release will free dreq->commit_data */ + dreq->commit_data = NULL; - if (result <= 0) { - if (tot_bytes > 0) - break; - goto out; - } + dprintk("NFS: %5u initiated commit call\n", task->tk_pid); - if (tot_bytes == 0) - memcpy(&first_verf.verifier, &wdata->verf.verifier, - sizeof(first_verf.verifier)); - if (wdata->verf.committed != NFS_FILE_SYNC) { - need_commit = 1; - if (memcmp(&first_verf.verifier, &wdata->verf.verifier, - sizeof(first_verf.verifier))) - goto sync_retry; - } + lock_kernel(); + rpc_execute(&data->task); + unlock_kernel(); +} - tot_bytes += result; +static void nfs_direct_write_complete(struct nfs_direct_req *dreq, struct inode *inode) +{ + int flags = dreq->flags; - /* in case of a short write: stop now, let the app recover */ - if (result < wdata->args.count) + dreq->flags = 0; + switch (flags) { + case NFS_ODIRECT_DO_COMMIT: + nfs_direct_commit_schedule(dreq); break; + case NFS_ODIRECT_RESCHED_WRITES: + nfs_direct_write_reschedule(dreq); + break; + default: + nfs_end_data_update(inode); + if (dreq->commit_data != NULL) + nfs_commit_free(dreq->commit_data); + nfs_direct_free_writedata(dreq); + nfs_direct_complete(dreq); + } +} - wdata->args.offset += result; - wdata->args.pgbase += result; - curpage += wdata->args.pgbase >> PAGE_SHIFT; - wdata->args.pgbase &= ~PAGE_MASK; - request -= result; - } while (request != 0); +static void nfs_alloc_commit_data(struct nfs_direct_req *dreq) +{ + dreq->commit_data = nfs_commit_alloc(0); + if (dreq->commit_data != NULL) + dreq->commit_data->req = (struct nfs_page *) dreq; +} +#else +static inline void nfs_alloc_commit_data(struct nfs_direct_req *dreq) +{ + dreq->commit_data = NULL; +} - /* - * Commit data written so far, even in the event of an error - */ - if (need_commit) { - wdata->args.count = tot_bytes; - wdata->args.offset = file_offset; +static void nfs_direct_write_complete(struct nfs_direct_req *dreq, struct inode *inode) +{ + nfs_end_data_update(inode); + nfs_direct_free_writedata(dreq); + nfs_direct_complete(dreq); +} +#endif - lock_kernel(); - result = NFS_PROTO(inode)->commit(wdata); - unlock_kernel(); +static struct nfs_direct_req *nfs_direct_write_alloc(size_t nbytes, size_t wsize) +{ + struct list_head *list; + struct nfs_direct_req *dreq; + unsigned int wpages = (wsize + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT; + + dreq = nfs_direct_req_alloc(); + if (!dreq) + return NULL; + + list = &dreq->list; + for(;;) { + struct nfs_write_data *data = nfs_writedata_alloc(wpages); - if (result < 0 || memcmp(&first_verf.verifier, - &wdata->verf.verifier, - sizeof(first_verf.verifier)) != 0) - goto sync_retry; + if (unlikely(!data)) { + while (!list_empty(list)) { + data = list_entry(list->next, + struct nfs_write_data, pages); + list_del(&data->pages); + nfs_writedata_free(data); + } + kref_put(&dreq->kref, nfs_direct_req_release); + return NULL; + } + + INIT_LIST_HEAD(&data->pages); + list_add(&data->pages, list); + + data->req = (struct nfs_page *) dreq; + dreq->outstanding++; + if (nbytes <= wsize) + break; + nbytes -= wsize; } - result = tot_bytes; -out: - nfs_end_data_update(inode); - nfs_writedata_free(wdata); - return result; + nfs_alloc_commit_data(dreq); -sync_retry: - wdata->args.stable = NFS_FILE_SYNC; - goto retry; + kref_get(&dreq->kref); + return dreq; } -/** - * nfs_direct_write - For each iov segment, map the user's buffer - * then generate write and commit RPCs. - * @inode: target inode - * @ctx: target file open context - * @iov: array of vectors that define I/O buffer - * file_offset: offset in file to begin the operation - * nr_segs: size of iovec array - * - * Upon return, generic_file_direct_IO invalidates any cached pages - * that non-direct readers might access, so they will pick up these - * writes immediately. - */ -static ssize_t nfs_direct_write(struct inode *inode, - struct nfs_open_context *ctx, const struct iovec *iov, - loff_t file_offset, unsigned long nr_segs) +static void nfs_direct_write_result(struct rpc_task *task, void *calldata) { - ssize_t tot_bytes = 0; - unsigned long seg = 0; - - while ((seg < nr_segs) && (tot_bytes >= 0)) { - ssize_t result; - int page_count; - struct page **pages; - const struct iovec *vec = &iov[seg++]; - unsigned long user_addr = (unsigned long) vec->iov_base; - size_t size = vec->iov_len; - - page_count = nfs_get_user_pages(WRITE, user_addr, size, &pages); - if (page_count < 0) { - nfs_free_user_pages(pages, 0, 0); - if (tot_bytes > 0) - break; - return page_count; - } + struct nfs_write_data *data = calldata; + struct nfs_direct_req *dreq = (struct nfs_direct_req *) data->req; + int status = task->tk_status; + + if (nfs_writeback_done(task, data) != 0) + return; + + spin_lock(&dreq->lock); - result = nfs_direct_write_seg(inode, ctx, user_addr, size, - file_offset, pages, page_count); - nfs_free_user_pages(pages, page_count, 0); + if (likely(status >= 0)) + dreq->count += data->res.count; + else + dreq->error = task->tk_status; - if (result <= 0) { - if (tot_bytes > 0) + if (data->res.verf->committed != NFS_FILE_SYNC) { + switch (dreq->flags) { + case 0: + memcpy(&dreq->verf, &data->verf, sizeof(dreq->verf)); + dreq->flags = NFS_ODIRECT_DO_COMMIT; break; - return result; + case NFS_ODIRECT_DO_COMMIT: + if (memcmp(&dreq->verf, &data->verf, sizeof(dreq->verf))) { + dprintk("NFS: %5u write verify failed\n", task->tk_pid); + dreq->flags = NFS_ODIRECT_RESCHED_WRITES; + } } - tot_bytes += result; - file_offset += result; - if (result < size) - break; } - return tot_bytes; + /* In case we have to resend */ + data->args.stable = NFS_FILE_SYNC; + + spin_unlock(&dreq->lock); } -/** - * nfs_direct_IO - NFS address space operation for direct I/O - * rw: direction (read or write) - * @iocb: target I/O control block - * @iov: array of vectors that define I/O buffer - * file_offset: offset in file to begin the operation - * nr_segs: size of iovec array - * +/* + * NB: Return the value of the first error return code. Subsequent + * errors after the first one are ignored. */ -ssize_t -nfs_direct_IO(int rw, struct kiocb *iocb, const struct iovec *iov, - loff_t file_offset, unsigned long nr_segs) +static void nfs_direct_write_release(void *calldata) { - ssize_t result = -EINVAL; - struct file *file = iocb->ki_filp; - struct nfs_open_context *ctx; - struct dentry *dentry = file->f_dentry; - struct inode *inode = dentry->d_inode; + struct nfs_write_data *data = calldata; + struct nfs_direct_req *dreq = (struct nfs_direct_req *) data->req; - /* - * No support for async yet - */ - if (!is_sync_kiocb(iocb)) - return result; - - ctx = (struct nfs_open_context *)file->private_data; - switch (rw) { - case READ: - dprintk("NFS: direct_IO(read) (%s) off/no(%Lu/%lu)\n", - dentry->d_name.name, file_offset, nr_segs); - - result = nfs_direct_read(inode, ctx, iov, - file_offset, nr_segs); - break; - case WRITE: - dprintk("NFS: direct_IO(write) (%s) off/no(%Lu/%lu)\n", - dentry->d_name.name, file_offset, nr_segs); - - result = nfs_direct_write(inode, ctx, iov, - file_offset, nr_segs); - break; - default: - break; + spin_lock(&dreq->lock); + if (--dreq->outstanding) { + spin_unlock(&dreq->lock); + return; } + spin_unlock(&dreq->lock); + + nfs_direct_write_complete(dreq, data->inode); +} + +static const struct rpc_call_ops nfs_write_direct_ops = { + .rpc_call_done = nfs_direct_write_result, + .rpc_release = nfs_direct_write_release, +}; + +/* + * For each nfs_write_data struct that was allocated on the list, dispatch + * an NFS WRITE operation + */ +static void nfs_direct_write_schedule(struct nfs_direct_req *dreq, int sync) +{ + struct nfs_open_context *ctx = dreq->ctx; + struct inode *inode = ctx->dentry->d_inode; + struct list_head *list = &dreq->list; + struct page **pages = dreq->pages; + size_t count = dreq->user_count; + loff_t pos = dreq->pos; + size_t wsize = NFS_SERVER(inode)->wsize; + unsigned int curpage, pgbase; + + curpage = 0; + pgbase = dreq->user_addr & ~PAGE_MASK; + do { + struct nfs_write_data *data; + size_t bytes; + + bytes = wsize; + if (count < wsize) + bytes = count; + + BUG_ON(list_empty(list)); + data = list_entry(list->next, struct nfs_write_data, pages); + list_move_tail(&data->pages, &dreq->rewrite_list); + + data->inode = inode; + data->cred = ctx->cred; + data->args.fh = NFS_FH(inode); + data->args.context = ctx; + data->args.offset = pos; + data->args.pgbase = pgbase; + data->args.pages = &pages[curpage]; + data->args.count = bytes; + data->res.fattr = &data->fattr; + data->res.count = bytes; + data->res.verf = &data->verf; + + rpc_init_task(&data->task, NFS_CLIENT(inode), RPC_TASK_ASYNC, + &nfs_write_direct_ops, data); + NFS_PROTO(inode)->write_setup(data, sync); + + data->task.tk_priority = RPC_PRIORITY_NORMAL; + data->task.tk_cookie = (unsigned long) inode; + + lock_kernel(); + rpc_execute(&data->task); + unlock_kernel(); + + dfprintk(VFS, "NFS: %5u initiated direct write call (req %s/%Ld, %zu bytes @ offset %Lu)\n", + data->task.tk_pid, + inode->i_sb->s_id, + (long long)NFS_FILEID(inode), + bytes, + (unsigned long long)data->args.offset); + + pos += bytes; + pgbase += bytes; + curpage += pgbase >> PAGE_SHIFT; + pgbase &= ~PAGE_MASK; + + count -= bytes; + } while (count != 0); + BUG_ON(!list_empty(list)); +} + +static ssize_t nfs_direct_write(struct kiocb *iocb, unsigned long user_addr, size_t count, loff_t pos, struct page **pages, int nr_pages) +{ + ssize_t result; + sigset_t oldset; + struct inode *inode = iocb->ki_filp->f_mapping->host; + struct rpc_clnt *clnt = NFS_CLIENT(inode); + struct nfs_direct_req *dreq; + size_t wsize = NFS_SERVER(inode)->wsize; + int sync = 0; + + dreq = nfs_direct_write_alloc(count, wsize); + if (!dreq) + return -ENOMEM; + if (dreq->commit_data == NULL || count < wsize) + sync = FLUSH_STABLE; + + dreq->user_addr = user_addr; + dreq->user_count = count; + dreq->pos = pos; + dreq->pages = pages; + dreq->npages = nr_pages; + dreq->inode = inode; + dreq->ctx = get_nfs_open_context((struct nfs_open_context *)iocb->ki_filp->private_data); + if (!is_sync_kiocb(iocb)) + dreq->iocb = iocb; + + nfs_add_stats(inode, NFSIOS_DIRECTWRITTENBYTES, count); + + nfs_begin_data_update(inode); + + rpc_clnt_sigmask(clnt, &oldset); + nfs_direct_write_schedule(dreq, sync); + result = nfs_direct_wait(dreq); + rpc_clnt_sigunmask(clnt, &oldset); + return result; } @@ -640,49 +748,40 @@ nfs_direct_IO(int rw, struct kiocb *iocb, const struct iovec *iov, * nfs_file_direct_read - file direct read operation for NFS files * @iocb: target I/O control block * @buf: user's buffer into which to read data - * count: number of bytes to read - * pos: byte offset in file where reading starts + * @count: number of bytes to read + * @pos: byte offset in file where reading starts * * We use this function for direct reads instead of calling * generic_file_aio_read() in order to avoid gfar's check to see if * the request starts before the end of the file. For that check * to work, we must generate a GETATTR before each direct read, and * even then there is a window between the GETATTR and the subsequent - * READ where the file size could change. So our preference is simply + * READ where the file size could change. Our preference is simply * to do all reads the application wants, and the server will take * care of managing the end of file boundary. - * + * * This function also eliminates unnecessarily updating the file's * atime locally, as the NFS server sets the file's atime, and this * client must read the updated atime from the server back into its * cache. */ -ssize_t -nfs_file_direct_read(struct kiocb *iocb, char __user *buf, size_t count, loff_t pos) +ssize_t nfs_file_direct_read(struct kiocb *iocb, char __user *buf, size_t count, loff_t pos) { ssize_t retval = -EINVAL; - loff_t *ppos = &iocb->ki_pos; + int page_count; + struct page **pages; struct file *file = iocb->ki_filp; - struct nfs_open_context *ctx = - (struct nfs_open_context *) file->private_data; struct address_space *mapping = file->f_mapping; - struct inode *inode = mapping->host; - struct iovec iov = { - .iov_base = buf, - .iov_len = count, - }; dprintk("nfs: direct read(%s/%s, %lu@%Ld)\n", file->f_dentry->d_parent->d_name.name, file->f_dentry->d_name.name, (unsigned long) count, (long long) pos); - if (!is_sync_kiocb(iocb)) - goto out; if (count < 0) goto out; retval = -EFAULT; - if (!access_ok(VERIFY_WRITE, iov.iov_base, iov.iov_len)) + if (!access_ok(VERIFY_WRITE, buf, count)) goto out; retval = 0; if (!count) @@ -692,9 +791,16 @@ nfs_file_direct_read(struct kiocb *iocb, char __user *buf, size_t count, loff_t if (retval) goto out; - retval = nfs_direct_read(inode, ctx, &iov, pos, 1); + retval = nfs_get_user_pages(READ, (unsigned long) buf, + count, &pages); + if (retval < 0) + goto out; + page_count = retval; + + retval = nfs_direct_read(iocb, (unsigned long) buf, count, pos, + pages, page_count); if (retval > 0) - *ppos = pos + retval; + iocb->ki_pos = pos + retval; out: return retval; @@ -704,8 +810,8 @@ out: * nfs_file_direct_write - file direct write operation for NFS files * @iocb: target I/O control block * @buf: user's buffer from which to write data - * count: number of bytes to write - * pos: byte offset in file where writing starts + * @count: number of bytes to write + * @pos: byte offset in file where writing starts * * We use this function for direct writes instead of calling * generic_file_aio_write() in order to avoid taking the inode @@ -725,28 +831,19 @@ out: * Note that O_APPEND is not supported for NFS direct writes, as there * is no atomic O_APPEND write facility in the NFS protocol. */ -ssize_t -nfs_file_direct_write(struct kiocb *iocb, const char __user *buf, size_t count, loff_t pos) +ssize_t nfs_file_direct_write(struct kiocb *iocb, const char __user *buf, size_t count, loff_t pos) { ssize_t retval; + int page_count; + struct page **pages; struct file *file = iocb->ki_filp; - struct nfs_open_context *ctx = - (struct nfs_open_context *) file->private_data; struct address_space *mapping = file->f_mapping; - struct inode *inode = mapping->host; - struct iovec iov = { - .iov_base = (char __user *)buf, - }; dfprintk(VFS, "nfs: direct write(%s/%s, %lu@%Ld)\n", file->f_dentry->d_parent->d_name.name, file->f_dentry->d_name.name, (unsigned long) count, (long long) pos); - retval = -EINVAL; - if (!is_sync_kiocb(iocb)) - goto out; - retval = generic_write_checks(file, &pos, &count, 0); if (retval) goto out; @@ -757,19 +854,35 @@ nfs_file_direct_write(struct kiocb *iocb, const char __user *buf, size_t count, retval = 0; if (!count) goto out; - iov.iov_len = count, retval = -EFAULT; - if (!access_ok(VERIFY_READ, iov.iov_base, iov.iov_len)) + if (!access_ok(VERIFY_READ, buf, count)) goto out; retval = nfs_sync_mapping(mapping); if (retval) goto out; - retval = nfs_direct_write(inode, ctx, &iov, pos, 1); + retval = nfs_get_user_pages(WRITE, (unsigned long) buf, + count, &pages); + if (retval < 0) + goto out; + page_count = retval; + + retval = nfs_direct_write(iocb, (unsigned long) buf, count, + pos, pages, page_count); + + /* + * XXX: nfs_end_data_update() already ensures this file's + * cached data is subsequently invalidated. Do we really + * need to call invalidate_inode_pages2() again here? + * + * For aio writes, this invalidation will almost certainly + * occur before the writes complete. Kind of racey. + */ if (mapping->nrpages) invalidate_inode_pages2(mapping); + if (retval > 0) iocb->ki_pos = pos + retval; @@ -777,6 +890,10 @@ out: return retval; } +/** + * nfs_init_directcache - create a slab cache for nfs_direct_req structures + * + */ int nfs_init_directcache(void) { nfs_direct_cachep = kmem_cache_create("nfs_direct_cache", @@ -790,6 +907,10 @@ int nfs_init_directcache(void) return 0; } +/** + * nfs_init_directcache - destroy the slab cache for nfs_direct_req structures + * + */ void nfs_destroy_directcache(void) { if (kmem_cache_destroy(nfs_direct_cachep)) diff --git a/fs/nfs/file.c b/fs/nfs/file.c index 7a79fbe9f53..f1df2c8d925 100644 --- a/fs/nfs/file.c +++ b/fs/nfs/file.c @@ -32,6 +32,7 @@ #include <asm/system.h> #include "delegation.h" +#include "iostat.h" #define NFSDBG_FACILITY NFSDBG_FILE @@ -48,7 +49,7 @@ static int nfs_check_flags(int flags); static int nfs_lock(struct file *filp, int cmd, struct file_lock *fl); static int nfs_flock(struct file *filp, int cmd, struct file_lock *fl); -struct file_operations nfs_file_operations = { +const struct file_operations nfs_file_operations = { .llseek = nfs_file_llseek, .read = do_sync_read, .write = do_sync_write, @@ -102,18 +103,15 @@ static int nfs_check_flags(int flags) static int nfs_file_open(struct inode *inode, struct file *filp) { - struct nfs_server *server = NFS_SERVER(inode); - int (*open)(struct inode *, struct file *); int res; res = nfs_check_flags(filp->f_flags); if (res) return res; + nfs_inc_stats(inode, NFSIOS_VFSOPEN); lock_kernel(); - /* Do NFSv4 open() call */ - if ((open = server->rpc_ops->file_open) != NULL) - res = open(inode, filp); + res = NFS_SERVER(inode)->rpc_ops->file_open(inode, filp); unlock_kernel(); return res; } @@ -124,6 +122,7 @@ nfs_file_release(struct inode *inode, struct file *filp) /* Ensure that dirty pages are flushed out with the right creds */ if (filp->f_mode & FMODE_WRITE) filemap_fdatawrite(filp->f_mapping); + nfs_inc_stats(inode, NFSIOS_VFSRELEASE); return NFS_PROTO(inode)->file_release(inode, filp); } @@ -199,6 +198,7 @@ nfs_file_flush(struct file *file) if ((file->f_mode & FMODE_WRITE) == 0) return 0; + nfs_inc_stats(inode, NFSIOS_VFSFLUSH); lock_kernel(); /* Ensure that data+attribute caches are up to date after close() */ status = nfs_wb_all(inode); @@ -229,6 +229,7 @@ nfs_file_read(struct kiocb *iocb, char __user * buf, size_t count, loff_t pos) (unsigned long) count, (unsigned long) pos); result = nfs_revalidate_file(inode, iocb->ki_filp); + nfs_add_stats(inode, NFSIOS_NORMALREADBYTES, count); if (!result) result = generic_file_aio_read(iocb, buf, count, pos); return result; @@ -282,6 +283,7 @@ nfs_fsync(struct file *file, struct dentry *dentry, int datasync) dfprintk(VFS, "nfs: fsync(%s/%ld)\n", inode->i_sb->s_id, inode->i_ino); + nfs_inc_stats(inode, NFSIOS_VFSFSYNC); lock_kernel(); status = nfs_wb_all(inode); if (!status) { @@ -316,6 +318,16 @@ static int nfs_commit_write(struct file *file, struct page *page, unsigned offse return status; } +static void nfs_invalidate_page(struct page *page, unsigned long offset) +{ + /* FIXME: we really should cancel any unstarted writes on this page */ +} + +static int nfs_release_page(struct page *page, gfp_t gfp) +{ + return !nfs_wb_page(page->mapping->host, page); +} + struct address_space_operations nfs_file_aops = { .readpage = nfs_readpage, .readpages = nfs_readpages, @@ -324,6 +336,8 @@ struct address_space_operations nfs_file_aops = { .writepages = nfs_writepages, .prepare_write = nfs_prepare_write, .commit_write = nfs_commit_write, + .invalidatepage = nfs_invalidate_page, + .releasepage = nfs_release_page, #ifdef CONFIG_NFS_DIRECTIO .direct_IO = nfs_direct_IO, #endif @@ -365,6 +379,7 @@ nfs_file_write(struct kiocb *iocb, const char __user *buf, size_t count, loff_t if (!count) goto out; + nfs_add_stats(inode, NFSIOS_NORMALWRITTENBYTES, count); result = generic_file_aio_write(iocb, buf, count, pos); out: return result; @@ -376,15 +391,17 @@ out_swapfile: static int do_getlk(struct file *filp, int cmd, struct file_lock *fl) { - struct file_lock *cfl; + struct file_lock cfl; struct inode *inode = filp->f_mapping->host; int status = 0; lock_kernel(); /* Try local locking first */ - cfl = posix_test_lock(filp, fl); - if (cfl != NULL) { - locks_copy_lock(fl, cfl); + if (posix_test_lock(filp, fl, &cfl)) { + fl->fl_start = cfl.fl_start; + fl->fl_end = cfl.fl_end; + fl->fl_type = cfl.fl_type; + fl->fl_pid = cfl.fl_pid; goto out; } @@ -425,10 +442,8 @@ static int do_vfs_lock(struct file *file, struct file_lock *fl) static int do_unlk(struct file *filp, int cmd, struct file_lock *fl) { struct inode *inode = filp->f_mapping->host; - sigset_t oldset; int status; - rpc_clnt_sigmask(NFS_CLIENT(inode), &oldset); /* * Flush all pending writes before doing anything * with locks.. @@ -446,17 +461,14 @@ static int do_unlk(struct file *filp, int cmd, struct file_lock *fl) else status = do_vfs_lock(filp, fl); unlock_kernel(); - rpc_clnt_sigunmask(NFS_CLIENT(inode), &oldset); return status; } static int do_setlk(struct file *filp, int cmd, struct file_lock *fl) { struct inode *inode = filp->f_mapping->host; - sigset_t oldset; int status; - rpc_clnt_sigmask(NFS_CLIENT(inode), &oldset); /* * Flush all pending writes before doing anything * with locks.. @@ -489,7 +501,6 @@ static int do_setlk(struct file *filp, int cmd, struct file_lock *fl) nfs_sync_mapping(filp->f_mapping); nfs_zap_caches(inode); out: - rpc_clnt_sigunmask(NFS_CLIENT(inode), &oldset); return status; } @@ -504,9 +515,7 @@ static int nfs_lock(struct file *filp, int cmd, struct file_lock *fl) inode->i_sb->s_id, inode->i_ino, fl->fl_type, fl->fl_flags, (long long)fl->fl_start, (long long)fl->fl_end); - - if (!inode) - return -EINVAL; + nfs_inc_stats(inode, NFSIOS_VFSLOCK); /* No mandatory locks over NFS */ if ((inode->i_mode & (S_ISGID | S_IXGRP)) == S_ISGID && @@ -531,9 +540,6 @@ static int nfs_flock(struct file *filp, int cmd, struct file_lock *fl) inode->i_sb->s_id, inode->i_ino, fl->fl_type, fl->fl_flags); - if (!inode) - return -EINVAL; - /* * No BSD flocks over NFS allowed. * Note: we could try to fake a POSIX lock request here by diff --git a/fs/nfs/idmap.c b/fs/nfs/idmap.c index 821edd30333..3fab5b0cfc5 100644 --- a/fs/nfs/idmap.c +++ b/fs/nfs/idmap.c @@ -35,6 +35,7 @@ */ #include <linux/module.h> +#include <linux/mutex.h> #include <linux/init.h> #include <linux/types.h> #include <linux/slab.h> @@ -74,8 +75,8 @@ struct idmap { struct dentry *idmap_dentry; wait_queue_head_t idmap_wq; struct idmap_msg idmap_im; - struct semaphore idmap_lock; /* Serializes upcalls */ - struct semaphore idmap_im_lock; /* Protects the hashtable */ + struct mutex idmap_lock; /* Serializes upcalls */ + struct mutex idmap_im_lock; /* Protects the hashtable */ struct idmap_hashtable idmap_user_hash; struct idmap_hashtable idmap_group_hash; }; @@ -101,11 +102,9 @@ nfs_idmap_new(struct nfs4_client *clp) if (clp->cl_idmap != NULL) return; - if ((idmap = kmalloc(sizeof(*idmap), GFP_KERNEL)) == NULL) + if ((idmap = kzalloc(sizeof(*idmap), GFP_KERNEL)) == NULL) return; - memset(idmap, 0, sizeof(*idmap)); - snprintf(idmap->idmap_path, sizeof(idmap->idmap_path), "%s/idmap", clp->cl_rpcclient->cl_pathname); @@ -116,8 +115,8 @@ nfs_idmap_new(struct nfs4_client *clp) return; } - init_MUTEX(&idmap->idmap_lock); - init_MUTEX(&idmap->idmap_im_lock); + mutex_init(&idmap->idmap_lock); + mutex_init(&idmap->idmap_im_lock); init_waitqueue_head(&idmap->idmap_wq); idmap->idmap_user_hash.h_type = IDMAP_TYPE_USER; idmap->idmap_group_hash.h_type = IDMAP_TYPE_GROUP; @@ -132,6 +131,8 @@ nfs_idmap_delete(struct nfs4_client *clp) if (!idmap) return; + dput(idmap->idmap_dentry); + idmap->idmap_dentry = NULL; rpc_unlink(idmap->idmap_path); clp->cl_idmap = NULL; kfree(idmap); @@ -232,8 +233,8 @@ nfs_idmap_id(struct idmap *idmap, struct idmap_hashtable *h, if (namelen >= IDMAP_NAMESZ) return -EINVAL; - down(&idmap->idmap_lock); - down(&idmap->idmap_im_lock); + mutex_lock(&idmap->idmap_lock); + mutex_lock(&idmap->idmap_im_lock); he = idmap_lookup_name(h, name, namelen); if (he != NULL) { @@ -259,11 +260,11 @@ nfs_idmap_id(struct idmap *idmap, struct idmap_hashtable *h, } set_current_state(TASK_UNINTERRUPTIBLE); - up(&idmap->idmap_im_lock); + mutex_unlock(&idmap->idmap_im_lock); schedule(); current->state = TASK_RUNNING; remove_wait_queue(&idmap->idmap_wq, &wq); - down(&idmap->idmap_im_lock); + mutex_lock(&idmap->idmap_im_lock); if (im->im_status & IDMAP_STATUS_SUCCESS) { *id = im->im_id; @@ -272,8 +273,8 @@ nfs_idmap_id(struct idmap *idmap, struct idmap_hashtable *h, out: memset(im, 0, sizeof(*im)); - up(&idmap->idmap_im_lock); - up(&idmap->idmap_lock); + mutex_unlock(&idmap->idmap_im_lock); + mutex_unlock(&idmap->idmap_lock); return (ret); } @@ -293,8 +294,8 @@ nfs_idmap_name(struct idmap *idmap, struct idmap_hashtable *h, im = &idmap->idmap_im; - down(&idmap->idmap_lock); - down(&idmap->idmap_im_lock); + mutex_lock(&idmap->idmap_lock); + mutex_lock(&idmap->idmap_im_lock); he = idmap_lookup_id(h, id); if (he != 0) { @@ -320,11 +321,11 @@ nfs_idmap_name(struct idmap *idmap, struct idmap_hashtable *h, } set_current_state(TASK_UNINTERRUPTIBLE); - up(&idmap->idmap_im_lock); + mutex_unlock(&idmap->idmap_im_lock); schedule(); current->state = TASK_RUNNING; remove_wait_queue(&idmap->idmap_wq, &wq); - down(&idmap->idmap_im_lock); + mutex_lock(&idmap->idmap_im_lock); if (im->im_status & IDMAP_STATUS_SUCCESS) { if ((len = strnlen(im->im_name, IDMAP_NAMESZ)) == 0) @@ -335,8 +336,8 @@ nfs_idmap_name(struct idmap *idmap, struct idmap_hashtable *h, out: memset(im, 0, sizeof(*im)); - up(&idmap->idmap_im_lock); - up(&idmap->idmap_lock); + mutex_unlock(&idmap->idmap_im_lock); + mutex_unlock(&idmap->idmap_lock); return ret; } @@ -380,7 +381,7 @@ idmap_pipe_downcall(struct file *filp, const char __user *src, size_t mlen) if (copy_from_user(&im_in, src, mlen) != 0) return (-EFAULT); - down(&idmap->idmap_im_lock); + mutex_lock(&idmap->idmap_im_lock); ret = mlen; im->im_status = im_in.im_status; @@ -440,7 +441,7 @@ idmap_pipe_downcall(struct file *filp, const char __user *src, size_t mlen) idmap_update_entry(he, im_in.im_name, namelen_in, im_in.im_id); ret = mlen; out: - up(&idmap->idmap_im_lock); + mutex_unlock(&idmap->idmap_im_lock); return ret; } @@ -452,10 +453,10 @@ idmap_pipe_destroy_msg(struct rpc_pipe_msg *msg) if (msg->errno >= 0) return; - down(&idmap->idmap_im_lock); + mutex_lock(&idmap->idmap_im_lock); im->im_status = IDMAP_STATUS_LOOKUPFAIL; wake_up(&idmap->idmap_wq); - up(&idmap->idmap_im_lock); + mutex_unlock(&idmap->idmap_im_lock); } /* diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c index 3413996f9a8..2f7656b911b 100644 --- a/fs/nfs/inode.c +++ b/fs/nfs/inode.c @@ -26,6 +26,7 @@ #include <linux/unistd.h> #include <linux/sunrpc/clnt.h> #include <linux/sunrpc/stats.h> +#include <linux/sunrpc/metrics.h> #include <linux/nfs_fs.h> #include <linux/nfs_mount.h> #include <linux/nfs4_mount.h> @@ -42,6 +43,7 @@ #include "nfs4_fs.h" #include "callback.h" #include "delegation.h" +#include "iostat.h" #define NFSDBG_FACILITY NFSDBG_VFS #define NFS_PARANOIA 1 @@ -65,6 +67,7 @@ static void nfs_clear_inode(struct inode *); static void nfs_umount_begin(struct super_block *); static int nfs_statfs(struct super_block *, struct kstatfs *); static int nfs_show_options(struct seq_file *, struct vfsmount *); +static int nfs_show_stats(struct seq_file *, struct vfsmount *); static void nfs_zap_acl_cache(struct inode *); static struct rpc_program nfs_program; @@ -78,6 +81,7 @@ static struct super_operations nfs_sops = { .clear_inode = nfs_clear_inode, .umount_begin = nfs_umount_begin, .show_options = nfs_show_options, + .show_stats = nfs_show_stats, }; /* @@ -133,7 +137,7 @@ nfs_fattr_to_ino_t(struct nfs_fattr *fattr) static int nfs_write_inode(struct inode *inode, int sync) { - int flags = sync ? FLUSH_WAIT : 0; + int flags = sync ? FLUSH_SYNC : 0; int ret; ret = nfs_commit_inode(inode, flags); @@ -237,7 +241,6 @@ static struct inode * nfs_get_root(struct super_block *sb, struct nfs_fh *rootfh, struct nfs_fsinfo *fsinfo) { struct nfs_server *server = NFS_SB(sb); - struct inode *rooti; int error; error = server->rpc_ops->getroot(server, rootfh, fsinfo); @@ -246,10 +249,7 @@ nfs_get_root(struct super_block *sb, struct nfs_fh *rootfh, struct nfs_fsinfo *f return ERR_PTR(error); } - rooti = nfs_fhget(sb, rootfh, fsinfo->fattr); - if (!rooti) - return ERR_PTR(-ENOMEM); - return rooti; + return nfs_fhget(sb, rootfh, fsinfo->fattr); } /* @@ -277,6 +277,10 @@ nfs_sb_init(struct super_block *sb, rpc_authflavor_t authflavor) sb->s_magic = NFS_SUPER_MAGIC; + server->io_stats = nfs_alloc_iostats(); + if (server->io_stats == NULL) + return -ENOMEM; + root_inode = nfs_get_root(sb, &server->fh, &fsinfo); /* Did getting the root inode fail? */ if (IS_ERR(root_inode)) { @@ -290,6 +294,9 @@ nfs_sb_init(struct super_block *sb, rpc_authflavor_t authflavor) } sb->s_root->d_op = server->rpc_ops->dentry_ops; + /* mount time stamp, in seconds */ + server->mount_time = jiffies; + /* Get some general file system info */ if (server->namelen == 0 && server->rpc_ops->pathconf(server, &server->fh, &pathinfo) >= 0) @@ -396,6 +403,9 @@ nfs_create_client(struct nfs_server *server, const struct nfs_mount_data *data) nfs_init_timeout_values(&timeparms, proto, data->timeo, data->retrans); + server->retrans_timeo = timeparms.to_initval; + server->retrans_count = timeparms.to_retries; + /* create transport and client */ xprt = xprt_create_proto(proto, &server->addr, &timeparms); if (IS_ERR(xprt)) { @@ -579,7 +589,7 @@ nfs_statfs(struct super_block *sb, struct kstatfs *buf) } -static int nfs_show_options(struct seq_file *m, struct vfsmount *mnt) +static void nfs_show_mount_options(struct seq_file *m, struct nfs_server *nfss, int showdefaults) { static struct proc_nfs_info { int flag; @@ -588,28 +598,26 @@ static int nfs_show_options(struct seq_file *m, struct vfsmount *mnt) } nfs_info[] = { { NFS_MOUNT_SOFT, ",soft", ",hard" }, { NFS_MOUNT_INTR, ",intr", "" }, - { NFS_MOUNT_POSIX, ",posix", "" }, { NFS_MOUNT_NOCTO, ",nocto", "" }, { NFS_MOUNT_NOAC, ",noac", "" }, - { NFS_MOUNT_NONLM, ",nolock", ",lock" }, + { NFS_MOUNT_NONLM, ",nolock", "" }, { NFS_MOUNT_NOACL, ",noacl", "" }, { 0, NULL, NULL } }; struct proc_nfs_info *nfs_infop; - struct nfs_server *nfss = NFS_SB(mnt->mnt_sb); char buf[12]; char *proto; - seq_printf(m, ",v%d", nfss->rpc_ops->version); + seq_printf(m, ",vers=%d", nfss->rpc_ops->version); seq_printf(m, ",rsize=%d", nfss->rsize); seq_printf(m, ",wsize=%d", nfss->wsize); - if (nfss->acregmin != 3*HZ) + if (nfss->acregmin != 3*HZ || showdefaults) seq_printf(m, ",acregmin=%d", nfss->acregmin/HZ); - if (nfss->acregmax != 60*HZ) + if (nfss->acregmax != 60*HZ || showdefaults) seq_printf(m, ",acregmax=%d", nfss->acregmax/HZ); - if (nfss->acdirmin != 30*HZ) + if (nfss->acdirmin != 30*HZ || showdefaults) seq_printf(m, ",acdirmin=%d", nfss->acdirmin/HZ); - if (nfss->acdirmax != 60*HZ) + if (nfss->acdirmax != 60*HZ || showdefaults) seq_printf(m, ",acdirmax=%d", nfss->acdirmax/HZ); for (nfs_infop = nfs_info; nfs_infop->flag; nfs_infop++) { if (nfss->flags & nfs_infop->flag) @@ -629,8 +637,96 @@ static int nfs_show_options(struct seq_file *m, struct vfsmount *mnt) proto = buf; } seq_printf(m, ",proto=%s", proto); + seq_printf(m, ",timeo=%lu", 10U * nfss->retrans_timeo / HZ); + seq_printf(m, ",retrans=%u", nfss->retrans_count); +} + +static int nfs_show_options(struct seq_file *m, struct vfsmount *mnt) +{ + struct nfs_server *nfss = NFS_SB(mnt->mnt_sb); + + nfs_show_mount_options(m, nfss, 0); + seq_puts(m, ",addr="); seq_escape(m, nfss->hostname, " \t\n\\"); + + return 0; +} + +static int nfs_show_stats(struct seq_file *m, struct vfsmount *mnt) +{ + int i, cpu; + struct nfs_server *nfss = NFS_SB(mnt->mnt_sb); + struct rpc_auth *auth = nfss->client->cl_auth; + struct nfs_iostats totals = { }; + + seq_printf(m, "statvers=%s", NFS_IOSTAT_VERS); + + /* + * Display all mount option settings + */ + seq_printf(m, "\n\topts:\t"); + seq_puts(m, mnt->mnt_sb->s_flags & MS_RDONLY ? "ro" : "rw"); + seq_puts(m, mnt->mnt_sb->s_flags & MS_SYNCHRONOUS ? ",sync" : ""); + seq_puts(m, mnt->mnt_sb->s_flags & MS_NOATIME ? ",noatime" : ""); + seq_puts(m, mnt->mnt_sb->s_flags & MS_NODIRATIME ? ",nodiratime" : ""); + nfs_show_mount_options(m, nfss, 1); + + seq_printf(m, "\n\tage:\t%lu", (jiffies - nfss->mount_time) / HZ); + + seq_printf(m, "\n\tcaps:\t"); + seq_printf(m, "caps=0x%x", nfss->caps); + seq_printf(m, ",wtmult=%d", nfss->wtmult); + seq_printf(m, ",dtsize=%d", nfss->dtsize); + seq_printf(m, ",bsize=%d", nfss->bsize); + seq_printf(m, ",namelen=%d", nfss->namelen); + +#ifdef CONFIG_NFS_V4 + if (nfss->rpc_ops->version == 4) { + seq_printf(m, "\n\tnfsv4:\t"); + seq_printf(m, "bm0=0x%x", nfss->attr_bitmask[0]); + seq_printf(m, ",bm1=0x%x", nfss->attr_bitmask[1]); + seq_printf(m, ",acl=0x%x", nfss->acl_bitmask); + } +#endif + + /* + * Display security flavor in effect for this mount + */ + seq_printf(m, "\n\tsec:\tflavor=%d", auth->au_ops->au_flavor); + if (auth->au_flavor) + seq_printf(m, ",pseudoflavor=%d", auth->au_flavor); + + /* + * Display superblock I/O counters + */ + for (cpu = 0; cpu < NR_CPUS; cpu++) { + struct nfs_iostats *stats; + + if (!cpu_possible(cpu)) + continue; + + preempt_disable(); + stats = per_cpu_ptr(nfss->io_stats, cpu); + + for (i = 0; i < __NFSIOS_COUNTSMAX; i++) + totals.events[i] += stats->events[i]; + for (i = 0; i < __NFSIOS_BYTESMAX; i++) + totals.bytes[i] += stats->bytes[i]; + + preempt_enable(); + } + + seq_printf(m, "\n\tevents:\t"); + for (i = 0; i < __NFSIOS_COUNTSMAX; i++) + seq_printf(m, "%lu ", totals.events[i]); + seq_printf(m, "\n\tbytes:\t"); + for (i = 0; i < __NFSIOS_BYTESMAX; i++) + seq_printf(m, "%Lu ", totals.bytes[i]); + seq_printf(m, "\n"); + + rpc_print_iostats(m, nfss->client); + return 0; } @@ -660,6 +756,8 @@ static void nfs_zap_caches_locked(struct inode *inode) struct nfs_inode *nfsi = NFS_I(inode); int mode = inode->i_mode; + nfs_inc_stats(inode, NFSIOS_ATTRINVALIDATE); + NFS_ATTRTIMEO(inode) = NFS_MINATTRTIMEO(inode); NFS_ATTRTIMEO_UPDATE(inode) = jiffies; @@ -751,7 +849,7 @@ nfs_fhget(struct super_block *sb, struct nfs_fh *fh, struct nfs_fattr *fattr) .fh = fh, .fattr = fattr }; - struct inode *inode = NULL; + struct inode *inode = ERR_PTR(-ENOENT); unsigned long hash; if ((fattr->valid & NFS_ATTR_FATTR) == 0) @@ -764,8 +862,11 @@ nfs_fhget(struct super_block *sb, struct nfs_fh *fh, struct nfs_fattr *fattr) hash = nfs_fattr_to_ino_t(fattr); - if (!(inode = iget5_locked(sb, hash, nfs_find_actor, nfs_init_locked, &desc))) + inode = iget5_locked(sb, hash, nfs_find_actor, nfs_init_locked, &desc); + if (inode == NULL) { + inode = ERR_PTR(-ENOMEM); goto out_no_inode; + } if (inode->i_state & I_NEW) { struct nfs_inode *nfsi = NFS_I(inode); @@ -834,7 +935,7 @@ out: return inode; out_no_inode: - printk("nfs_fhget: iget failed\n"); + dprintk("nfs_fhget: iget failed with error %ld\n", PTR_ERR(inode)); goto out; } @@ -847,6 +948,8 @@ nfs_setattr(struct dentry *dentry, struct iattr *attr) struct nfs_fattr fattr; int error; + nfs_inc_stats(inode, NFSIOS_VFSSETATTR); + if (attr->ia_valid & ATTR_SIZE) { if (!S_ISREG(inode->i_mode) || attr->ia_size == i_size_read(inode)) attr->ia_valid &= ~ATTR_SIZE; @@ -859,11 +962,9 @@ nfs_setattr(struct dentry *dentry, struct iattr *attr) lock_kernel(); nfs_begin_data_update(inode); - /* Write all dirty data if we're changing file permissions or size */ - if ((attr->ia_valid & (ATTR_MODE|ATTR_UID|ATTR_GID|ATTR_SIZE)) != 0) { - filemap_write_and_wait(inode->i_mapping); - nfs_wb_all(inode); - } + /* Write all dirty data */ + filemap_write_and_wait(inode->i_mapping); + nfs_wb_all(inode); /* * Return any delegations if we're going to change ACLs */ @@ -902,6 +1003,7 @@ void nfs_setattr_update_inode(struct inode *inode, struct iattr *attr) spin_unlock(&inode->i_lock); } if ((attr->ia_valid & ATTR_SIZE) != 0) { + nfs_inc_stats(inode, NFSIOS_SETATTRTRUNC); inode->i_size = attr->ia_size; vmtruncate(inode, attr->ia_size); } @@ -949,7 +1051,7 @@ int nfs_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat *stat) int err; /* Flush out writes to the server in order to update c/mtime */ - nfs_sync_inode(inode, 0, 0, FLUSH_WAIT|FLUSH_NOCOMMIT); + nfs_sync_inode_wait(inode, 0, 0, FLUSH_NOCOMMIT); /* * We may force a getattr if the user cares about atime. @@ -973,7 +1075,7 @@ int nfs_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat *stat) return err; } -struct nfs_open_context *alloc_nfs_open_context(struct dentry *dentry, struct rpc_cred *cred) +static struct nfs_open_context *alloc_nfs_open_context(struct vfsmount *mnt, struct dentry *dentry, struct rpc_cred *cred) { struct nfs_open_context *ctx; @@ -981,6 +1083,7 @@ struct nfs_open_context *alloc_nfs_open_context(struct dentry *dentry, struct rp if (ctx != NULL) { atomic_set(&ctx->count, 1); ctx->dentry = dget(dentry); + ctx->vfsmnt = mntget(mnt); ctx->cred = get_rpccred(cred); ctx->state = NULL; ctx->lockowner = current->files; @@ -1011,6 +1114,7 @@ void put_nfs_open_context(struct nfs_open_context *ctx) if (ctx->cred != NULL) put_rpccred(ctx->cred); dput(ctx->dentry); + mntput(ctx->vfsmnt); kfree(ctx); } } @@ -1019,7 +1123,7 @@ void put_nfs_open_context(struct nfs_open_context *ctx) * Ensure that mmap has a recent RPC credential for use when writing out * shared pages */ -void nfs_file_set_open_context(struct file *filp, struct nfs_open_context *ctx) +static void nfs_file_set_open_context(struct file *filp, struct nfs_open_context *ctx) { struct inode *inode = filp->f_dentry->d_inode; struct nfs_inode *nfsi = NFS_I(inode); @@ -1051,7 +1155,7 @@ struct nfs_open_context *nfs_find_open_context(struct inode *inode, struct rpc_c return ctx; } -void nfs_file_clear_open_context(struct file *filp) +static void nfs_file_clear_open_context(struct file *filp) { struct inode *inode = filp->f_dentry->d_inode; struct nfs_open_context *ctx = (struct nfs_open_context *)filp->private_data; @@ -1076,7 +1180,7 @@ int nfs_open(struct inode *inode, struct file *filp) cred = rpcauth_lookupcred(NFS_CLIENT(inode)->cl_auth, 0); if (IS_ERR(cred)) return PTR_ERR(cred); - ctx = alloc_nfs_open_context(filp->f_dentry, cred); + ctx = alloc_nfs_open_context(filp->f_vfsmnt, filp->f_dentry, cred); put_rpccred(cred); if (ctx == NULL) return -ENOMEM; @@ -1185,6 +1289,7 @@ int nfs_attribute_timeout(struct inode *inode) */ int nfs_revalidate_inode(struct nfs_server *server, struct inode *inode) { + nfs_inc_stats(inode, NFSIOS_INODEREVALIDATE); if (!(NFS_I(inode)->cache_validity & (NFS_INO_INVALID_ATTR|NFS_INO_INVALID_DATA)) && !nfs_attribute_timeout(inode)) return NFS_STALE(inode) ? -ESTALE : 0; @@ -1201,6 +1306,7 @@ void nfs_revalidate_mapping(struct inode *inode, struct address_space *mapping) struct nfs_inode *nfsi = NFS_I(inode); if (nfsi->cache_validity & NFS_INO_INVALID_DATA) { + nfs_inc_stats(inode, NFSIOS_DATAINVALIDATE); if (S_ISREG(inode->i_mode)) nfs_sync_mapping(mapping); invalidate_inode_pages2(mapping); @@ -1299,39 +1405,37 @@ static int nfs_check_inode_attributes(struct inode *inode, struct nfs_fattr *fat if ((fattr->valid & NFS_ATTR_FATTR) == 0) return 0; + /* Has the inode gone and changed behind our back? */ + if (nfsi->fileid != fattr->fileid + || (inode->i_mode & S_IFMT) != (fattr->mode & S_IFMT)) { + return -EIO; + } + /* Are we in the process of updating data on the server? */ data_unstable = nfs_caches_unstable(inode); /* Do atomic weak cache consistency updates */ nfs_wcc_update_inode(inode, fattr); - if ((fattr->valid & NFS_ATTR_FATTR_V4) != 0 && - nfsi->change_attr != fattr->change_attr) { + if ((fattr->valid & NFS_ATTR_FATTR_V4) != 0) { + if (nfsi->change_attr == fattr->change_attr) + goto out; nfsi->cache_validity |= NFS_INO_INVALID_ATTR; if (!data_unstable) nfsi->cache_validity |= NFS_INO_REVAL_PAGECACHE; } - /* Has the inode gone and changed behind our back? */ - if (nfsi->fileid != fattr->fileid - || (inode->i_mode & S_IFMT) != (fattr->mode & S_IFMT)) { - return -EIO; - } - - cur_size = i_size_read(inode); - new_isize = nfs_size_to_loff_t(fattr->size); - /* Verify a few of the more important attributes */ if (!timespec_equal(&inode->i_mtime, &fattr->mtime)) { nfsi->cache_validity |= NFS_INO_INVALID_ATTR; if (!data_unstable) nfsi->cache_validity |= NFS_INO_REVAL_PAGECACHE; } - if (cur_size != new_isize) { - nfsi->cache_validity |= NFS_INO_INVALID_ATTR; - if (nfsi->npages == 0) - nfsi->cache_validity |= NFS_INO_REVAL_PAGECACHE; - } + + cur_size = i_size_read(inode); + new_isize = nfs_size_to_loff_t(fattr->size); + if (cur_size != new_isize && nfsi->npages == 0) + nfsi->cache_validity |= NFS_INO_INVALID_ATTR|NFS_INO_REVAL_PAGECACHE; /* Have any file permissions changed? */ if ((inode->i_mode & S_IALLUGO) != (fattr->mode & S_IALLUGO) @@ -1343,6 +1447,7 @@ static int nfs_check_inode_attributes(struct inode *inode, struct nfs_fattr *fat if (inode->i_nlink != fattr->nlink) nfsi->cache_validity |= NFS_INO_INVALID_ATTR; +out: if (!timespec_equal(&inode->i_atime, &fattr->atime)) nfsi->cache_validity |= NFS_INO_INVALID_ATIME; @@ -1481,15 +1586,6 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr) nfsi->cache_change_attribute = jiffies; } - if ((fattr->valid & NFS_ATTR_FATTR_V4) - && nfsi->change_attr != fattr->change_attr) { - dprintk("NFS: change_attr change on server for file %s/%ld\n", - inode->i_sb->s_id, inode->i_ino); - nfsi->change_attr = fattr->change_attr; - invalid |= NFS_INO_INVALID_ATTR|NFS_INO_INVALID_DATA|NFS_INO_INVALID_ACCESS|NFS_INO_INVALID_ACL; - nfsi->cache_change_attribute = jiffies; - } - /* If ctime has changed we should definitely clear access+acl caches */ if (!timespec_equal(&inode->i_ctime, &fattr->ctime)) { invalid |= NFS_INO_INVALID_ACCESS|NFS_INO_INVALID_ACL; @@ -1519,8 +1615,20 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr) inode->i_blksize = fattr->du.nfs2.blocksize; } + if ((fattr->valid & NFS_ATTR_FATTR_V4)) { + if (nfsi->change_attr != fattr->change_attr) { + dprintk("NFS: change_attr change on server for file %s/%ld\n", + inode->i_sb->s_id, inode->i_ino); + nfsi->change_attr = fattr->change_attr; + invalid |= NFS_INO_INVALID_ATTR|NFS_INO_INVALID_DATA|NFS_INO_INVALID_ACCESS|NFS_INO_INVALID_ACL; + nfsi->cache_change_attribute = jiffies; + } else + invalid &= ~(NFS_INO_INVALID_ATTR|NFS_INO_INVALID_DATA); + } + /* Update attrtimeo value if we're out of the unstable period */ if (invalid & NFS_INO_INVALID_ATTR) { + nfs_inc_stats(inode, NFSIOS_ATTRINVALIDATE); nfsi->attrtimeo = NFS_MINATTRTIMEO(inode); nfsi->attrtimeo_timestamp = jiffies; } else if (time_after(jiffies, nfsi->attrtimeo_timestamp+nfsi->attrtimeo)) { @@ -1637,10 +1745,9 @@ static struct super_block *nfs_get_sb(struct file_system_type *fs_type, #endif /* CONFIG_NFS_V3 */ s = ERR_PTR(-ENOMEM); - server = kmalloc(sizeof(struct nfs_server), GFP_KERNEL); + server = kzalloc(sizeof(struct nfs_server), GFP_KERNEL); if (!server) goto out_err; - memset(server, 0, sizeof(struct nfs_server)); /* Zero out the NFS state stuff */ init_nfsv4_state(server); server->client = server->client_sys = server->client_acl = ERR_PTR(-EINVAL); @@ -1712,6 +1819,7 @@ static void nfs_kill_super(struct super_block *s) rpciod_down(); /* release rpciod */ + nfs_free_iostats(server->io_stats); kfree(server->hostname); kfree(server); } @@ -1738,6 +1846,7 @@ static struct super_operations nfs4_sops = { .clear_inode = nfs4_clear_inode, .umount_begin = nfs_umount_begin, .show_options = nfs_show_options, + .show_stats = nfs_show_stats, }; /* @@ -1800,6 +1909,9 @@ static int nfs4_fill_super(struct super_block *sb, struct nfs4_mount_data *data, nfs_init_timeout_values(&timeparms, data->proto, data->timeo, data->retrans); + server->retrans_timeo = timeparms.to_initval; + server->retrans_count = timeparms.to_retries; + clp = nfs4_get_client(&server->addr.sin_addr); if (!clp) { dprintk("%s: failed to create NFS4 client.\n", __FUNCTION__); @@ -1941,10 +2053,9 @@ static struct super_block *nfs4_get_sb(struct file_system_type *fs_type, return ERR_PTR(-EINVAL); } - server = kmalloc(sizeof(struct nfs_server), GFP_KERNEL); + server = kzalloc(sizeof(struct nfs_server), GFP_KERNEL); if (!server) return ERR_PTR(-ENOMEM); - memset(server, 0, sizeof(struct nfs_server)); /* Zero out the NFS state stuff */ init_nfsv4_state(server); server->client = server->client_sys = server->client_acl = ERR_PTR(-EINVAL); @@ -2024,10 +2135,12 @@ static void nfs4_kill_super(struct super_block *sb) if (server->client != NULL && !IS_ERR(server->client)) rpc_shutdown_client(server->client); - rpciod_down(); /* release rpciod */ destroy_nfsv4_state(server); + rpciod_down(); + + nfs_free_iostats(server->io_stats); kfree(server->hostname); kfree(server); } diff --git a/fs/nfs/iostat.h b/fs/nfs/iostat.h new file mode 100644 index 00000000000..6350ecbde58 --- /dev/null +++ b/fs/nfs/iostat.h @@ -0,0 +1,164 @@ +/* + * linux/fs/nfs/iostat.h + * + * Declarations for NFS client per-mount statistics + * + * Copyright (C) 2005, 2006 Chuck Lever <cel@netapp.com> + * + * NFS client per-mount statistics provide information about the health of + * the NFS client and the health of each NFS mount point. Generally these + * are not for detailed problem diagnosis, but simply to indicate that there + * is a problem. + * + * These counters are not meant to be human-readable, but are meant to be + * integrated into system monitoring tools such as "sar" and "iostat". As + * such, the counters are sampled by the tools over time, and are never + * zeroed after a file system is mounted. Moving averages can be computed + * by the tools by taking the difference between two instantaneous samples + * and dividing that by the time between the samples. + */ + +#ifndef _NFS_IOSTAT +#define _NFS_IOSTAT + +#define NFS_IOSTAT_VERS "1.0" + +/* + * NFS byte counters + * + * 1. SERVER - the number of payload bytes read from or written to the + * server by the NFS client via an NFS READ or WRITE request. + * + * 2. NORMAL - the number of bytes read or written by applications via + * the read(2) and write(2) system call interfaces. + * + * 3. DIRECT - the number of bytes read or written from files opened + * with the O_DIRECT flag. + * + * These counters give a view of the data throughput into and out of the NFS + * client. Comparing the number of bytes requested by an application with the + * number of bytes the client requests from the server can provide an + * indication of client efficiency (per-op, cache hits, etc). + * + * These counters can also help characterize which access methods are in + * use. DIRECT by itself shows whether there is any O_DIRECT traffic. + * NORMAL + DIRECT shows how much data is going through the system call + * interface. A large amount of SERVER traffic without much NORMAL or + * DIRECT traffic shows that applications are using mapped files. + * + * NFS page counters + * + * These count the number of pages read or written via nfs_readpage(), + * nfs_readpages(), or their write equivalents. + */ +enum nfs_stat_bytecounters { + NFSIOS_NORMALREADBYTES = 0, + NFSIOS_NORMALWRITTENBYTES, + NFSIOS_DIRECTREADBYTES, + NFSIOS_DIRECTWRITTENBYTES, + NFSIOS_SERVERREADBYTES, + NFSIOS_SERVERWRITTENBYTES, + NFSIOS_READPAGES, + NFSIOS_WRITEPAGES, + __NFSIOS_BYTESMAX, +}; + +/* + * NFS event counters + * + * These counters provide a low-overhead way of monitoring client activity + * without enabling NFS trace debugging. The counters show the rate at + * which VFS requests are made, and how often the client invalidates its + * data and attribute caches. This allows system administrators to monitor + * such things as how close-to-open is working, and answer questions such + * as "why are there so many GETATTR requests on the wire?" + * + * They also count anamolous events such as short reads and writes, silly + * renames due to close-after-delete, and operations that change the size + * of a file (such operations can often be the source of data corruption + * if applications aren't using file locking properly). + */ +enum nfs_stat_eventcounters { + NFSIOS_INODEREVALIDATE = 0, + NFSIOS_DENTRYREVALIDATE, + NFSIOS_DATAINVALIDATE, + NFSIOS_ATTRINVALIDATE, + NFSIOS_VFSOPEN, + NFSIOS_VFSLOOKUP, + NFSIOS_VFSACCESS, + NFSIOS_VFSUPDATEPAGE, + NFSIOS_VFSREADPAGE, + NFSIOS_VFSREADPAGES, + NFSIOS_VFSWRITEPAGE, + NFSIOS_VFSWRITEPAGES, + NFSIOS_VFSGETDENTS, + NFSIOS_VFSSETATTR, + NFSIOS_VFSFLUSH, + NFSIOS_VFSFSYNC, + NFSIOS_VFSLOCK, + NFSIOS_VFSRELEASE, + NFSIOS_CONGESTIONWAIT, + NFSIOS_SETATTRTRUNC, + NFSIOS_EXTENDWRITE, + NFSIOS_SILLYRENAME, + NFSIOS_SHORTREAD, + NFSIOS_SHORTWRITE, + NFSIOS_DELAY, + __NFSIOS_COUNTSMAX, +}; + +#ifdef __KERNEL__ + +#include <linux/percpu.h> +#include <linux/cache.h> + +struct nfs_iostats { + unsigned long long bytes[__NFSIOS_BYTESMAX]; + unsigned long events[__NFSIOS_COUNTSMAX]; +} ____cacheline_aligned; + +static inline void nfs_inc_server_stats(struct nfs_server *server, enum nfs_stat_eventcounters stat) +{ + struct nfs_iostats *iostats; + int cpu; + + cpu = get_cpu(); + iostats = per_cpu_ptr(server->io_stats, cpu); + iostats->events[stat] ++; + put_cpu_no_resched(); +} + +static inline void nfs_inc_stats(struct inode *inode, enum nfs_stat_eventcounters stat) +{ + nfs_inc_server_stats(NFS_SERVER(inode), stat); +} + +static inline void nfs_add_server_stats(struct nfs_server *server, enum nfs_stat_bytecounters stat, unsigned long addend) +{ + struct nfs_iostats *iostats; + int cpu; + + cpu = get_cpu(); + iostats = per_cpu_ptr(server->io_stats, cpu); + iostats->bytes[stat] += addend; + put_cpu_no_resched(); +} + +static inline void nfs_add_stats(struct inode *inode, enum nfs_stat_bytecounters stat, unsigned long addend) +{ + nfs_add_server_stats(NFS_SERVER(inode), stat, addend); +} + +static inline struct nfs_iostats *nfs_alloc_iostats(void) +{ + return alloc_percpu(struct nfs_iostats); +} + +static inline void nfs_free_iostats(struct nfs_iostats *stats) +{ + if (stats != NULL) + free_percpu(stats); +} + +#endif +#endif diff --git a/fs/nfs/mount_clnt.c b/fs/nfs/mount_clnt.c index 0b9a78353d6..445abb4d421 100644 --- a/fs/nfs/mount_clnt.c +++ b/fs/nfs/mount_clnt.c @@ -49,9 +49,12 @@ nfsroot_mount(struct sockaddr_in *addr, char *path, struct nfs_fh *fh, struct mnt_fhstatus result = { .fh = fh }; + struct rpc_message msg = { + .rpc_argp = path, + .rpc_resp = &result, + }; char hostname[32]; int status; - int call; dprintk("NFS: nfs_mount(%08x:%s)\n", (unsigned)ntohl(addr->sin_addr.s_addr), path); @@ -61,8 +64,12 @@ nfsroot_mount(struct sockaddr_in *addr, char *path, struct nfs_fh *fh, if (IS_ERR(mnt_clnt)) return PTR_ERR(mnt_clnt); - call = (version == NFS_MNT3_VERSION) ? MOUNTPROC3_MNT : MNTPROC_MNT; - status = rpc_call(mnt_clnt, call, path, &result, 0); + if (version == NFS_MNT3_VERSION) + msg.rpc_proc = &mnt_clnt->cl_procinfo[MOUNTPROC3_MNT]; + else + msg.rpc_proc = &mnt_clnt->cl_procinfo[MNTPROC_MNT]; + + status = rpc_call_sync(mnt_clnt, &msg, 0); return status < 0? status : (result.status? -EACCES : 0); } @@ -137,6 +144,8 @@ static struct rpc_procinfo mnt_procedures[] = { .p_encode = (kxdrproc_t) xdr_encode_dirpath, .p_decode = (kxdrproc_t) xdr_decode_fhstatus, .p_bufsiz = MNT_dirpath_sz << 2, + .p_statidx = MNTPROC_MNT, + .p_name = "MOUNT", }, }; @@ -146,6 +155,8 @@ static struct rpc_procinfo mnt3_procedures[] = { .p_encode = (kxdrproc_t) xdr_encode_dirpath, .p_decode = (kxdrproc_t) xdr_decode_fhstatus3, .p_bufsiz = MNT_dirpath_sz << 2, + .p_statidx = MOUNTPROC3_MNT, + .p_name = "MOUNT", }, }; diff --git a/fs/nfs/nfs2xdr.c b/fs/nfs/nfs2xdr.c index 6548a65de94..f0015fa876e 100644 --- a/fs/nfs/nfs2xdr.c +++ b/fs/nfs/nfs2xdr.c @@ -682,7 +682,9 @@ nfs_stat_to_errno(int stat) .p_encode = (kxdrproc_t) nfs_xdr_##argtype, \ .p_decode = (kxdrproc_t) nfs_xdr_##restype, \ .p_bufsiz = MAX(NFS_##argtype##_sz,NFS_##restype##_sz) << 2, \ - .p_timer = timer \ + .p_timer = timer, \ + .p_statidx = NFSPROC_##proc, \ + .p_name = #proc, \ } struct rpc_procinfo nfs_procedures[] = { PROC(GETATTR, fhandle, attrstat, 1), diff --git a/fs/nfs/nfs3acl.c b/fs/nfs/nfs3acl.c index 6a5bbc0ae94..33287879bd2 100644 --- a/fs/nfs/nfs3acl.c +++ b/fs/nfs/nfs3acl.c @@ -190,6 +190,10 @@ struct posix_acl *nfs3_proc_getacl(struct inode *inode, int type) struct nfs3_getaclres res = { .fattr = &fattr, }; + struct rpc_message msg = { + .rpc_argp = &args, + .rpc_resp = &res, + }; struct posix_acl *acl; int status, count; @@ -218,8 +222,8 @@ struct posix_acl *nfs3_proc_getacl(struct inode *inode, int type) return NULL; dprintk("NFS call getacl\n"); - status = rpc_call(server->client_acl, ACLPROC3_GETACL, - &args, &res, 0); + msg.rpc_proc = &server->client_acl->cl_procinfo[ACLPROC3_GETACL]; + status = rpc_call_sync(server->client_acl, &msg, 0); dprintk("NFS reply getacl: %d\n", status); /* pages may have been allocated at the xdr layer. */ @@ -286,6 +290,10 @@ static int nfs3_proc_setacls(struct inode *inode, struct posix_acl *acl, .acl_access = acl, .pages = pages, }; + struct rpc_message msg = { + .rpc_argp = &args, + .rpc_resp = &fattr, + }; int status, count; status = -EOPNOTSUPP; @@ -306,8 +314,8 @@ static int nfs3_proc_setacls(struct inode *inode, struct posix_acl *acl, dprintk("NFS call setacl\n"); nfs_begin_data_update(inode); - status = rpc_call(server->client_acl, ACLPROC3_SETACL, - &args, &fattr, 0); + msg.rpc_proc = &server->client_acl->cl_procinfo[ACLPROC3_SETACL]; + status = rpc_call_sync(server->client_acl, &msg, 0); spin_lock(&inode->i_lock); NFS_I(inode)->cache_validity |= NFS_INO_INVALID_ACCESS; spin_unlock(&inode->i_lock); diff --git a/fs/nfs/nfs3proc.c b/fs/nfs/nfs3proc.c index ed67567f055..cf186f0d2b3 100644 --- a/fs/nfs/nfs3proc.c +++ b/fs/nfs/nfs3proc.c @@ -19,6 +19,8 @@ #include <linux/smp_lock.h> #include <linux/nfs_mount.h> +#include "iostat.h" + #define NFSDBG_FACILITY NFSDBG_PROC extern struct rpc_procinfo nfs3_procedures[]; @@ -41,27 +43,14 @@ nfs3_rpc_wrapper(struct rpc_clnt *clnt, struct rpc_message *msg, int flags) return res; } -static inline int -nfs3_rpc_call_wrapper(struct rpc_clnt *clnt, u32 proc, void *argp, void *resp, int flags) -{ - struct rpc_message msg = { - .rpc_proc = &clnt->cl_procinfo[proc], - .rpc_argp = argp, - .rpc_resp = resp, - }; - return nfs3_rpc_wrapper(clnt, &msg, flags); -} - -#define rpc_call(clnt, proc, argp, resp, flags) \ - nfs3_rpc_call_wrapper(clnt, proc, argp, resp, flags) -#define rpc_call_sync(clnt, msg, flags) \ - nfs3_rpc_wrapper(clnt, msg, flags) +#define rpc_call_sync(clnt, msg, flags) nfs3_rpc_wrapper(clnt, msg, flags) static int -nfs3_async_handle_jukebox(struct rpc_task *task) +nfs3_async_handle_jukebox(struct rpc_task *task, struct inode *inode) { if (task->tk_status != -EJUKEBOX) return 0; + nfs_inc_stats(inode, NFSIOS_DELAY); task->tk_status = 0; rpc_restart_call(task); rpc_delay(task, NFS_JUKEBOX_RETRY_TIME); @@ -72,14 +61,21 @@ static int do_proc_get_root(struct rpc_clnt *client, struct nfs_fh *fhandle, struct nfs_fsinfo *info) { + struct rpc_message msg = { + .rpc_proc = &nfs3_procedures[NFS3PROC_FSINFO], + .rpc_argp = fhandle, + .rpc_resp = info, + }; int status; dprintk("%s: call fsinfo\n", __FUNCTION__); nfs_fattr_init(info->fattr); - status = rpc_call(client, NFS3PROC_FSINFO, fhandle, info, 0); + status = rpc_call_sync(client, &msg, 0); dprintk("%s: reply fsinfo: %d\n", __FUNCTION__, status); if (!(info->fattr->valid & NFS_ATTR_FATTR)) { - status = rpc_call(client, NFS3PROC_GETATTR, fhandle, info->fattr, 0); + msg.rpc_proc = &nfs3_procedures[NFS3PROC_GETATTR]; + msg.rpc_resp = info->fattr; + status = rpc_call_sync(client, &msg, 0); dprintk("%s: reply getattr: %d\n", __FUNCTION__, status); } return status; @@ -107,12 +103,16 @@ static int nfs3_proc_getattr(struct nfs_server *server, struct nfs_fh *fhandle, struct nfs_fattr *fattr) { + struct rpc_message msg = { + .rpc_proc = &nfs3_procedures[NFS3PROC_GETATTR], + .rpc_argp = fhandle, + .rpc_resp = fattr, + }; int status; dprintk("NFS call getattr\n"); nfs_fattr_init(fattr); - status = rpc_call(server->client, NFS3PROC_GETATTR, - fhandle, fattr, 0); + status = rpc_call_sync(server->client, &msg, 0); dprintk("NFS reply getattr: %d\n", status); return status; } @@ -126,11 +126,16 @@ nfs3_proc_setattr(struct dentry *dentry, struct nfs_fattr *fattr, .fh = NFS_FH(inode), .sattr = sattr, }; + struct rpc_message msg = { + .rpc_proc = &nfs3_procedures[NFS3PROC_SETATTR], + .rpc_argp = &arg, + .rpc_resp = fattr, + }; int status; dprintk("NFS call setattr\n"); nfs_fattr_init(fattr); - status = rpc_call(NFS_CLIENT(inode), NFS3PROC_SETATTR, &arg, fattr, 0); + status = rpc_call_sync(NFS_CLIENT(inode), &msg, 0); if (status == 0) nfs_setattr_update_inode(inode, sattr); dprintk("NFS reply setattr: %d\n", status); @@ -152,15 +157,23 @@ nfs3_proc_lookup(struct inode *dir, struct qstr *name, .fh = fhandle, .fattr = fattr }; + struct rpc_message msg = { + .rpc_proc = &nfs3_procedures[NFS3PROC_LOOKUP], + .rpc_argp = &arg, + .rpc_resp = &res, + }; int status; dprintk("NFS call lookup %s\n", name->name); nfs_fattr_init(&dir_attr); nfs_fattr_init(fattr); - status = rpc_call(NFS_CLIENT(dir), NFS3PROC_LOOKUP, &arg, &res, 0); - if (status >= 0 && !(fattr->valid & NFS_ATTR_FATTR)) - status = rpc_call(NFS_CLIENT(dir), NFS3PROC_GETATTR, - fhandle, fattr, 0); + status = rpc_call_sync(NFS_CLIENT(dir), &msg, 0); + if (status >= 0 && !(fattr->valid & NFS_ATTR_FATTR)) { + msg.rpc_proc = &nfs3_procedures[NFS3PROC_GETATTR]; + msg.rpc_argp = fhandle; + msg.rpc_resp = fattr; + status = rpc_call_sync(NFS_CLIENT(dir), &msg, 0); + } dprintk("NFS reply lookup: %d\n", status); if (status >= 0) status = nfs_refresh_inode(dir, &dir_attr); @@ -180,7 +193,7 @@ static int nfs3_proc_access(struct inode *inode, struct nfs_access_entry *entry) .rpc_proc = &nfs3_procedures[NFS3PROC_ACCESS], .rpc_argp = &arg, .rpc_resp = &res, - .rpc_cred = entry->cred + .rpc_cred = entry->cred, }; int mode = entry->mask; int status; @@ -226,12 +239,16 @@ static int nfs3_proc_readlink(struct inode *inode, struct page *page, .pglen = pglen, .pages = &page }; + struct rpc_message msg = { + .rpc_proc = &nfs3_procedures[NFS3PROC_READLINK], + .rpc_argp = &args, + .rpc_resp = &fattr, + }; int status; dprintk("NFS call readlink\n"); nfs_fattr_init(&fattr); - status = rpc_call(NFS_CLIENT(inode), NFS3PROC_READLINK, - &args, &fattr, 0); + status = rpc_call_sync(NFS_CLIENT(inode), &msg, 0); nfs_refresh_inode(inode, &fattr); dprintk("NFS reply readlink: %d\n", status); return status; @@ -327,6 +344,11 @@ nfs3_proc_create(struct inode *dir, struct dentry *dentry, struct iattr *sattr, .fh = &fhandle, .fattr = &fattr }; + struct rpc_message msg = { + .rpc_proc = &nfs3_procedures[NFS3PROC_CREATE], + .rpc_argp = &arg, + .rpc_resp = &res, + }; mode_t mode = sattr->ia_mode; int status; @@ -343,8 +365,8 @@ nfs3_proc_create(struct inode *dir, struct dentry *dentry, struct iattr *sattr, again: nfs_fattr_init(&dir_attr); nfs_fattr_init(&fattr); - status = rpc_call(NFS_CLIENT(dir), NFS3PROC_CREATE, &arg, &res, 0); - nfs_post_op_update_inode(dir, &dir_attr); + status = rpc_call_sync(NFS_CLIENT(dir), &msg, 0); + nfs_refresh_inode(dir, &dir_attr); /* If the server doesn't support the exclusive creation semantics, * try again with simple 'guarded' mode. */ @@ -447,7 +469,7 @@ nfs3_proc_unlink_done(struct dentry *dir, struct rpc_task *task) struct rpc_message *msg = &task->tk_msg; struct nfs_fattr *dir_attr; - if (nfs3_async_handle_jukebox(task)) + if (nfs3_async_handle_jukebox(task, dir->d_inode)) return 1; if (msg->rpc_argp) { dir_attr = (struct nfs_fattr*)msg->rpc_resp; @@ -474,12 +496,17 @@ nfs3_proc_rename(struct inode *old_dir, struct qstr *old_name, .fromattr = &old_dir_attr, .toattr = &new_dir_attr }; + struct rpc_message msg = { + .rpc_proc = &nfs3_procedures[NFS3PROC_RENAME], + .rpc_argp = &arg, + .rpc_resp = &res, + }; int status; dprintk("NFS call rename %s -> %s\n", old_name->name, new_name->name); nfs_fattr_init(&old_dir_attr); nfs_fattr_init(&new_dir_attr); - status = rpc_call(NFS_CLIENT(old_dir), NFS3PROC_RENAME, &arg, &res, 0); + status = rpc_call_sync(NFS_CLIENT(old_dir), &msg, 0); nfs_post_op_update_inode(old_dir, &old_dir_attr); nfs_post_op_update_inode(new_dir, &new_dir_attr); dprintk("NFS reply rename: %d\n", status); @@ -500,12 +527,17 @@ nfs3_proc_link(struct inode *inode, struct inode *dir, struct qstr *name) .dir_attr = &dir_attr, .fattr = &fattr }; + struct rpc_message msg = { + .rpc_proc = &nfs3_procedures[NFS3PROC_LINK], + .rpc_argp = &arg, + .rpc_resp = &res, + }; int status; dprintk("NFS call link %s\n", name->name); nfs_fattr_init(&dir_attr); nfs_fattr_init(&fattr); - status = rpc_call(NFS_CLIENT(inode), NFS3PROC_LINK, &arg, &res, 0); + status = rpc_call_sync(NFS_CLIENT(inode), &msg, 0); nfs_post_op_update_inode(dir, &dir_attr); nfs_post_op_update_inode(inode, &fattr); dprintk("NFS reply link: %d\n", status); @@ -531,6 +563,11 @@ nfs3_proc_symlink(struct inode *dir, struct qstr *name, struct qstr *path, .fh = fhandle, .fattr = fattr }; + struct rpc_message msg = { + .rpc_proc = &nfs3_procedures[NFS3PROC_SYMLINK], + .rpc_argp = &arg, + .rpc_resp = &res, + }; int status; if (path->len > NFS3_MAXPATHLEN) @@ -538,7 +575,7 @@ nfs3_proc_symlink(struct inode *dir, struct qstr *name, struct qstr *path, dprintk("NFS call symlink %s -> %s\n", name->name, path->name); nfs_fattr_init(&dir_attr); nfs_fattr_init(fattr); - status = rpc_call(NFS_CLIENT(dir), NFS3PROC_SYMLINK, &arg, &res, 0); + status = rpc_call_sync(NFS_CLIENT(dir), &msg, 0); nfs_post_op_update_inode(dir, &dir_attr); dprintk("NFS reply symlink: %d\n", status); return status; @@ -560,6 +597,11 @@ nfs3_proc_mkdir(struct inode *dir, struct dentry *dentry, struct iattr *sattr) .fh = &fhandle, .fattr = &fattr }; + struct rpc_message msg = { + .rpc_proc = &nfs3_procedures[NFS3PROC_MKDIR], + .rpc_argp = &arg, + .rpc_resp = &res, + }; int mode = sattr->ia_mode; int status; @@ -569,7 +611,7 @@ nfs3_proc_mkdir(struct inode *dir, struct dentry *dentry, struct iattr *sattr) nfs_fattr_init(&dir_attr); nfs_fattr_init(&fattr); - status = rpc_call(NFS_CLIENT(dir), NFS3PROC_MKDIR, &arg, &res, 0); + status = rpc_call_sync(NFS_CLIENT(dir), &msg, 0); nfs_post_op_update_inode(dir, &dir_attr); if (status != 0) goto out; @@ -591,11 +633,16 @@ nfs3_proc_rmdir(struct inode *dir, struct qstr *name) .name = name->name, .len = name->len }; + struct rpc_message msg = { + .rpc_proc = &nfs3_procedures[NFS3PROC_RMDIR], + .rpc_argp = &arg, + .rpc_resp = &dir_attr, + }; int status; dprintk("NFS call rmdir %s\n", name->name); nfs_fattr_init(&dir_attr); - status = rpc_call(NFS_CLIENT(dir), NFS3PROC_RMDIR, &arg, &dir_attr, 0); + status = rpc_call_sync(NFS_CLIENT(dir), &msg, 0); nfs_post_op_update_inode(dir, &dir_attr); dprintk("NFS reply rmdir: %d\n", status); return status; @@ -672,6 +719,11 @@ nfs3_proc_mknod(struct inode *dir, struct dentry *dentry, struct iattr *sattr, .fh = &fh, .fattr = &fattr }; + struct rpc_message msg = { + .rpc_proc = &nfs3_procedures[NFS3PROC_MKNOD], + .rpc_argp = &arg, + .rpc_resp = &res, + }; mode_t mode = sattr->ia_mode; int status; @@ -690,7 +742,7 @@ nfs3_proc_mknod(struct inode *dir, struct dentry *dentry, struct iattr *sattr, nfs_fattr_init(&dir_attr); nfs_fattr_init(&fattr); - status = rpc_call(NFS_CLIENT(dir), NFS3PROC_MKNOD, &arg, &res, 0); + status = rpc_call_sync(NFS_CLIENT(dir), &msg, 0); nfs_post_op_update_inode(dir, &dir_attr); if (status != 0) goto out; @@ -707,11 +759,16 @@ static int nfs3_proc_statfs(struct nfs_server *server, struct nfs_fh *fhandle, struct nfs_fsstat *stat) { + struct rpc_message msg = { + .rpc_proc = &nfs3_procedures[NFS3PROC_FSSTAT], + .rpc_argp = fhandle, + .rpc_resp = stat, + }; int status; dprintk("NFS call fsstat\n"); nfs_fattr_init(stat->fattr); - status = rpc_call(server->client, NFS3PROC_FSSTAT, fhandle, stat, 0); + status = rpc_call_sync(server->client, &msg, 0); dprintk("NFS reply statfs: %d\n", status); return status; } @@ -720,11 +777,16 @@ static int nfs3_proc_fsinfo(struct nfs_server *server, struct nfs_fh *fhandle, struct nfs_fsinfo *info) { + struct rpc_message msg = { + .rpc_proc = &nfs3_procedures[NFS3PROC_FSINFO], + .rpc_argp = fhandle, + .rpc_resp = info, + }; int status; dprintk("NFS call fsinfo\n"); nfs_fattr_init(info->fattr); - status = rpc_call(server->client_sys, NFS3PROC_FSINFO, fhandle, info, 0); + status = rpc_call_sync(server->client_sys, &msg, 0); dprintk("NFS reply fsinfo: %d\n", status); return status; } @@ -733,40 +795,34 @@ static int nfs3_proc_pathconf(struct nfs_server *server, struct nfs_fh *fhandle, struct nfs_pathconf *info) { + struct rpc_message msg = { + .rpc_proc = &nfs3_procedures[NFS3PROC_PATHCONF], + .rpc_argp = fhandle, + .rpc_resp = info, + }; int status; dprintk("NFS call pathconf\n"); nfs_fattr_init(info->fattr); - status = rpc_call(server->client, NFS3PROC_PATHCONF, fhandle, info, 0); + status = rpc_call_sync(server->client, &msg, 0); dprintk("NFS reply pathconf: %d\n", status); return status; } extern u32 *nfs3_decode_dirent(u32 *, struct nfs_entry *, int); -static void nfs3_read_done(struct rpc_task *task, void *calldata) +static int nfs3_read_done(struct rpc_task *task, struct nfs_read_data *data) { - struct nfs_read_data *data = calldata; - - if (nfs3_async_handle_jukebox(task)) - return; + if (nfs3_async_handle_jukebox(task, data->inode)) + return -EAGAIN; /* Call back common NFS readpage processing */ if (task->tk_status >= 0) nfs_refresh_inode(data->inode, &data->fattr); - nfs_readpage_result(task, calldata); + return 0; } -static const struct rpc_call_ops nfs3_read_ops = { - .rpc_call_done = nfs3_read_done, - .rpc_release = nfs_readdata_release, -}; - -static void -nfs3_proc_read_setup(struct nfs_read_data *data) +static void nfs3_proc_read_setup(struct nfs_read_data *data) { - struct rpc_task *task = &data->task; - struct inode *inode = data->inode; - int flags; struct rpc_message msg = { .rpc_proc = &nfs3_procedures[NFS3PROC_READ], .rpc_argp = &data->args, @@ -774,37 +830,20 @@ nfs3_proc_read_setup(struct nfs_read_data *data) .rpc_cred = data->cred, }; - /* N.B. Do we need to test? Never called for swapfile inode */ - flags = RPC_TASK_ASYNC | (IS_SWAPFILE(inode)? NFS_RPC_SWAPFLAGS : 0); - - /* Finalize the task. */ - rpc_init_task(task, NFS_CLIENT(inode), flags, &nfs3_read_ops, data); - rpc_call_setup(task, &msg, 0); + rpc_call_setup(&data->task, &msg, 0); } -static void nfs3_write_done(struct rpc_task *task, void *calldata) +static int nfs3_write_done(struct rpc_task *task, struct nfs_write_data *data) { - struct nfs_write_data *data = calldata; - - if (nfs3_async_handle_jukebox(task)) - return; + if (nfs3_async_handle_jukebox(task, data->inode)) + return -EAGAIN; if (task->tk_status >= 0) nfs_post_op_update_inode(data->inode, data->res.fattr); - nfs_writeback_done(task, calldata); + return 0; } -static const struct rpc_call_ops nfs3_write_ops = { - .rpc_call_done = nfs3_write_done, - .rpc_release = nfs_writedata_release, -}; - -static void -nfs3_proc_write_setup(struct nfs_write_data *data, int how) +static void nfs3_proc_write_setup(struct nfs_write_data *data, int how) { - struct rpc_task *task = &data->task; - struct inode *inode = data->inode; - int stable; - int flags; struct rpc_message msg = { .rpc_proc = &nfs3_procedures[NFS3PROC_WRITE], .rpc_argp = &data->args, @@ -812,45 +851,28 @@ nfs3_proc_write_setup(struct nfs_write_data *data, int how) .rpc_cred = data->cred, }; + data->args.stable = NFS_UNSTABLE; if (how & FLUSH_STABLE) { - if (!NFS_I(inode)->ncommit) - stable = NFS_FILE_SYNC; - else - stable = NFS_DATA_SYNC; - } else - stable = NFS_UNSTABLE; - data->args.stable = stable; - - /* Set the initial flags for the task. */ - flags = (how & FLUSH_SYNC) ? 0 : RPC_TASK_ASYNC; + data->args.stable = NFS_FILE_SYNC; + if (NFS_I(data->inode)->ncommit) + data->args.stable = NFS_DATA_SYNC; + } /* Finalize the task. */ - rpc_init_task(task, NFS_CLIENT(inode), flags, &nfs3_write_ops, data); - rpc_call_setup(task, &msg, 0); + rpc_call_setup(&data->task, &msg, 0); } -static void nfs3_commit_done(struct rpc_task *task, void *calldata) +static int nfs3_commit_done(struct rpc_task *task, struct nfs_write_data *data) { - struct nfs_write_data *data = calldata; - - if (nfs3_async_handle_jukebox(task)) - return; + if (nfs3_async_handle_jukebox(task, data->inode)) + return -EAGAIN; if (task->tk_status >= 0) nfs_post_op_update_inode(data->inode, data->res.fattr); - nfs_commit_done(task, calldata); + return 0; } -static const struct rpc_call_ops nfs3_commit_ops = { - .rpc_call_done = nfs3_commit_done, - .rpc_release = nfs_commit_release, -}; - -static void -nfs3_proc_commit_setup(struct nfs_write_data *data, int how) +static void nfs3_proc_commit_setup(struct nfs_write_data *data, int how) { - struct rpc_task *task = &data->task; - struct inode *inode = data->inode; - int flags; struct rpc_message msg = { .rpc_proc = &nfs3_procedures[NFS3PROC_COMMIT], .rpc_argp = &data->args, @@ -858,12 +880,7 @@ nfs3_proc_commit_setup(struct nfs_write_data *data, int how) .rpc_cred = data->cred, }; - /* Set the initial flags for the task. */ - flags = (how & FLUSH_SYNC) ? 0 : RPC_TASK_ASYNC; - - /* Finalize the task. */ - rpc_init_task(task, NFS_CLIENT(inode), flags, &nfs3_commit_ops, data); - rpc_call_setup(task, &msg, 0); + rpc_call_setup(&data->task, &msg, 0); } static int @@ -902,8 +919,11 @@ struct nfs_rpc_ops nfs_v3_clientops = { .pathconf = nfs3_proc_pathconf, .decode_dirent = nfs3_decode_dirent, .read_setup = nfs3_proc_read_setup, + .read_done = nfs3_read_done, .write_setup = nfs3_proc_write_setup, + .write_done = nfs3_write_done, .commit_setup = nfs3_proc_commit_setup, + .commit_done = nfs3_commit_done, .file_open = nfs_open, .file_release = nfs_release, .lock = nfs3_proc_lock, diff --git a/fs/nfs/nfs3xdr.c b/fs/nfs/nfs3xdr.c index 5224a191efb..ec233619687 100644 --- a/fs/nfs/nfs3xdr.c +++ b/fs/nfs/nfs3xdr.c @@ -1109,7 +1109,9 @@ nfs3_xdr_setaclres(struct rpc_rqst *req, u32 *p, struct nfs_fattr *fattr) .p_encode = (kxdrproc_t) nfs3_xdr_##argtype, \ .p_decode = (kxdrproc_t) nfs3_xdr_##restype, \ .p_bufsiz = MAX(NFS3_##argtype##_sz,NFS3_##restype##_sz) << 2, \ - .p_timer = timer \ + .p_timer = timer, \ + .p_statidx = NFS3PROC_##proc, \ + .p_name = #proc, \ } struct rpc_procinfo nfs3_procedures[] = { @@ -1150,6 +1152,7 @@ static struct rpc_procinfo nfs3_acl_procedures[] = { .p_decode = (kxdrproc_t) nfs3_xdr_getaclres, .p_bufsiz = MAX(ACL3_getaclargs_sz, ACL3_getaclres_sz) << 2, .p_timer = 1, + .p_name = "GETACL", }, [ACLPROC3_SETACL] = { .p_proc = ACLPROC3_SETACL, @@ -1157,6 +1160,7 @@ static struct rpc_procinfo nfs3_acl_procedures[] = { .p_decode = (kxdrproc_t) nfs3_xdr_setaclres, .p_bufsiz = MAX(ACL3_setaclargs_sz, ACL3_setaclres_sz) << 2, .p_timer = 0, + .p_name = "SETACL", }, }; diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index f8c0066e02e..47ece1dd3c6 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -51,6 +51,7 @@ #include "nfs4_fs.h" #include "delegation.h" +#include "iostat.h" #define NFSDBG_FACILITY NFSDBG_PROC @@ -335,7 +336,7 @@ static struct nfs4_state *nfs4_opendata_to_nfs4_state(struct nfs4_opendata *data if (!(data->f_attr.valid & NFS_ATTR_FATTR)) goto out; inode = nfs_fhget(data->dir->d_sb, &data->o_res.fh, &data->f_attr); - if (inode == NULL) + if (IS_ERR(inode)) goto out; state = nfs4_get_open_state(inode, data->owner); if (state == NULL) @@ -604,11 +605,14 @@ static int _nfs4_proc_open_confirm(struct nfs4_opendata *data) int status; atomic_inc(&data->count); + /* + * If rpc_run_task() ends up calling ->rpc_release(), we + * want to ensure that it takes the 'error' code path. + */ + data->rpc_status = -ENOMEM; task = rpc_run_task(server->client, RPC_TASK_ASYNC, &nfs4_open_confirm_ops, data); - if (IS_ERR(task)) { - nfs4_opendata_free(data); + if (IS_ERR(task)) return PTR_ERR(task); - } status = nfs4_wait_for_completion_rpc_task(task); if (status != 0) { data->cancelled = 1; @@ -707,11 +711,14 @@ static int _nfs4_proc_open(struct nfs4_opendata *data) int status; atomic_inc(&data->count); + /* + * If rpc_run_task() ends up calling ->rpc_release(), we + * want to ensure that it takes the 'error' code path. + */ + data->rpc_status = -ENOMEM; task = rpc_run_task(server->client, RPC_TASK_ASYNC, &nfs4_open_ops, data); - if (IS_ERR(task)) { - nfs4_opendata_free(data); + if (IS_ERR(task)) return PTR_ERR(task); - } status = nfs4_wait_for_completion_rpc_task(task); if (status != 0) { data->cancelled = 1; @@ -908,7 +915,7 @@ out_put_state_owner: static struct nfs4_state *nfs4_open_delegated(struct inode *inode, int flags, struct rpc_cred *cred) { struct nfs4_exception exception = { }; - struct nfs4_state *res; + struct nfs4_state *res = ERR_PTR(-EIO); int err; do { @@ -1017,12 +1024,12 @@ static struct nfs4_state *nfs4_do_open(struct inode *dir, struct dentry *dentry, return res; } -static int _nfs4_do_setattr(struct nfs_server *server, struct nfs_fattr *fattr, - struct nfs_fh *fhandle, struct iattr *sattr, - struct nfs4_state *state) +static int _nfs4_do_setattr(struct inode *inode, struct nfs_fattr *fattr, + struct iattr *sattr, struct nfs4_state *state) { + struct nfs_server *server = NFS_SERVER(inode); struct nfs_setattrargs arg = { - .fh = fhandle, + .fh = NFS_FH(inode), .iap = sattr, .server = server, .bitmask = server->attr_bitmask, @@ -1041,7 +1048,9 @@ static int _nfs4_do_setattr(struct nfs_server *server, struct nfs_fattr *fattr, nfs_fattr_init(fattr); - if (state != NULL) { + if (nfs4_copy_delegation_stateid(&arg.stateid, inode)) { + /* Use that stateid */ + } else if (state != NULL) { msg.rpc_cred = state->owner->so_cred; nfs4_copy_stateid(&arg.stateid, state, current->files); } else @@ -1053,16 +1062,15 @@ static int _nfs4_do_setattr(struct nfs_server *server, struct nfs_fattr *fattr, return status; } -static int nfs4_do_setattr(struct nfs_server *server, struct nfs_fattr *fattr, - struct nfs_fh *fhandle, struct iattr *sattr, - struct nfs4_state *state) +static int nfs4_do_setattr(struct inode *inode, struct nfs_fattr *fattr, + struct iattr *sattr, struct nfs4_state *state) { + struct nfs_server *server = NFS_SERVER(inode); struct nfs4_exception exception = { }; int err; do { err = nfs4_handle_exception(server, - _nfs4_do_setattr(server, fattr, fhandle, sattr, - state), + _nfs4_do_setattr(inode, fattr, sattr, state), &exception); } while (exception.retry); return err; @@ -1503,8 +1511,7 @@ nfs4_proc_setattr(struct dentry *dentry, struct nfs_fattr *fattr, if (ctx != NULL) state = ctx->state; - status = nfs4_do_setattr(NFS_SERVER(inode), fattr, - NFS_FH(inode), sattr, state); + status = nfs4_do_setattr(inode, fattr, sattr, state); if (status == 0) nfs_setattr_update_inode(inode, sattr); if (ctx != NULL) @@ -1823,8 +1830,7 @@ nfs4_proc_create(struct inode *dir, struct dentry *dentry, struct iattr *sattr, d_instantiate(dentry, igrab(state->inode)); if (flags & O_EXCL) { struct nfs_fattr fattr; - status = nfs4_do_setattr(NFS_SERVER(dir), &fattr, - NFS_FH(state->inode), sattr, state); + status = nfs4_do_setattr(state->inode, &fattr, sattr, state); if (status == 0) nfs_setattr_update_inode(state->inode, sattr); } @@ -2344,75 +2350,50 @@ static int nfs4_proc_pathconf(struct nfs_server *server, struct nfs_fh *fhandle, return err; } -static void nfs4_read_done(struct rpc_task *task, void *calldata) +static int nfs4_read_done(struct rpc_task *task, struct nfs_read_data *data) { - struct nfs_read_data *data = calldata; - struct inode *inode = data->inode; + struct nfs_server *server = NFS_SERVER(data->inode); - if (nfs4_async_handle_error(task, NFS_SERVER(inode)) == -EAGAIN) { + if (nfs4_async_handle_error(task, server) == -EAGAIN) { rpc_restart_call(task); - return; + return -EAGAIN; } if (task->tk_status > 0) - renew_lease(NFS_SERVER(inode), data->timestamp); - /* Call back common NFS readpage processing */ - nfs_readpage_result(task, calldata); + renew_lease(server, data->timestamp); + return 0; } -static const struct rpc_call_ops nfs4_read_ops = { - .rpc_call_done = nfs4_read_done, - .rpc_release = nfs_readdata_release, -}; - -static void -nfs4_proc_read_setup(struct nfs_read_data *data) +static void nfs4_proc_read_setup(struct nfs_read_data *data) { - struct rpc_task *task = &data->task; struct rpc_message msg = { .rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_READ], .rpc_argp = &data->args, .rpc_resp = &data->res, .rpc_cred = data->cred, }; - struct inode *inode = data->inode; - int flags; data->timestamp = jiffies; - /* N.B. Do we need to test? Never called for swapfile inode */ - flags = RPC_TASK_ASYNC | (IS_SWAPFILE(inode)? NFS_RPC_SWAPFLAGS : 0); - - /* Finalize the task. */ - rpc_init_task(task, NFS_CLIENT(inode), flags, &nfs4_read_ops, data); - rpc_call_setup(task, &msg, 0); + rpc_call_setup(&data->task, &msg, 0); } -static void nfs4_write_done(struct rpc_task *task, void *calldata) +static int nfs4_write_done(struct rpc_task *task, struct nfs_write_data *data) { - struct nfs_write_data *data = calldata; struct inode *inode = data->inode; if (nfs4_async_handle_error(task, NFS_SERVER(inode)) == -EAGAIN) { rpc_restart_call(task); - return; + return -EAGAIN; } if (task->tk_status >= 0) { renew_lease(NFS_SERVER(inode), data->timestamp); nfs_post_op_update_inode(inode, data->res.fattr); } - /* Call back common NFS writeback processing */ - nfs_writeback_done(task, calldata); + return 0; } -static const struct rpc_call_ops nfs4_write_ops = { - .rpc_call_done = nfs4_write_done, - .rpc_release = nfs_writedata_release, -}; - -static void -nfs4_proc_write_setup(struct nfs_write_data *data, int how) +static void nfs4_proc_write_setup(struct nfs_write_data *data, int how) { - struct rpc_task *task = &data->task; struct rpc_message msg = { .rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_WRITE], .rpc_argp = &data->args, @@ -2422,7 +2403,6 @@ nfs4_proc_write_setup(struct nfs_write_data *data, int how) struct inode *inode = data->inode; struct nfs_server *server = NFS_SERVER(inode); int stable; - int flags; if (how & FLUSH_STABLE) { if (!NFS_I(inode)->ncommit) @@ -2437,57 +2417,37 @@ nfs4_proc_write_setup(struct nfs_write_data *data, int how) data->timestamp = jiffies; - /* Set the initial flags for the task. */ - flags = (how & FLUSH_SYNC) ? 0 : RPC_TASK_ASYNC; - /* Finalize the task. */ - rpc_init_task(task, NFS_CLIENT(inode), flags, &nfs4_write_ops, data); - rpc_call_setup(task, &msg, 0); + rpc_call_setup(&data->task, &msg, 0); } -static void nfs4_commit_done(struct rpc_task *task, void *calldata) +static int nfs4_commit_done(struct rpc_task *task, struct nfs_write_data *data) { - struct nfs_write_data *data = calldata; struct inode *inode = data->inode; if (nfs4_async_handle_error(task, NFS_SERVER(inode)) == -EAGAIN) { rpc_restart_call(task); - return; + return -EAGAIN; } if (task->tk_status >= 0) nfs_post_op_update_inode(inode, data->res.fattr); - /* Call back common NFS writeback processing */ - nfs_commit_done(task, calldata); + return 0; } -static const struct rpc_call_ops nfs4_commit_ops = { - .rpc_call_done = nfs4_commit_done, - .rpc_release = nfs_commit_release, -}; - -static void -nfs4_proc_commit_setup(struct nfs_write_data *data, int how) +static void nfs4_proc_commit_setup(struct nfs_write_data *data, int how) { - struct rpc_task *task = &data->task; struct rpc_message msg = { .rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_COMMIT], .rpc_argp = &data->args, .rpc_resp = &data->res, .rpc_cred = data->cred, }; - struct inode *inode = data->inode; - struct nfs_server *server = NFS_SERVER(inode); - int flags; + struct nfs_server *server = NFS_SERVER(data->inode); data->args.bitmask = server->attr_bitmask; data->res.server = server; - /* Set the initial flags for the task. */ - flags = (how & FLUSH_SYNC) ? 0 : RPC_TASK_ASYNC; - - /* Finalize the task. */ - rpc_init_task(task, NFS_CLIENT(inode), flags, &nfs4_commit_ops, data); - rpc_call_setup(task, &msg, 0); + rpc_call_setup(&data->task, &msg, 0); } /* @@ -2755,8 +2715,10 @@ nfs4_async_handle_error(struct rpc_task *task, const struct nfs_server *server) rpc_wake_up_task(task); task->tk_status = 0; return -EAGAIN; - case -NFS4ERR_GRACE: case -NFS4ERR_DELAY: + nfs_inc_server_stats((struct nfs_server *) server, + NFSIOS_DELAY); + case -NFS4ERR_GRACE: rpc_delay(task, NFS4_POLL_RETRY_MAX); task->tk_status = 0; return -EAGAIN; @@ -2893,8 +2855,7 @@ int nfs4_proc_setclientid(struct nfs4_client *clp, u32 program, unsigned short p return status; } -int -nfs4_proc_setclientid_confirm(struct nfs4_client *clp, struct rpc_cred *cred) +static int _nfs4_proc_setclientid_confirm(struct nfs4_client *clp, struct rpc_cred *cred) { struct nfs_fsinfo fsinfo; struct rpc_message msg = { @@ -2918,6 +2879,24 @@ nfs4_proc_setclientid_confirm(struct nfs4_client *clp, struct rpc_cred *cred) return status; } +int nfs4_proc_setclientid_confirm(struct nfs4_client *clp, struct rpc_cred *cred) +{ + long timeout; + int err; + do { + err = _nfs4_proc_setclientid_confirm(clp, cred); + switch (err) { + case 0: + return err; + case -NFS4ERR_RESOURCE: + /* The IBM lawyers misread another document! */ + case -NFS4ERR_DELAY: + err = nfs4_delay(clp->cl_rpcclient, &timeout); + } + } while (err == 0); + return err; +} + struct nfs4_delegreturndata { struct nfs4_delegreturnargs args; struct nfs4_delegreturnres res; @@ -2958,7 +2937,7 @@ static void nfs4_delegreturn_release(void *calldata) kfree(calldata); } -const static struct rpc_call_ops nfs4_delegreturn_ops = { +static const struct rpc_call_ops nfs4_delegreturn_ops = { .rpc_call_prepare = nfs4_delegreturn_prepare, .rpc_call_done = nfs4_delegreturn_done, .rpc_release = nfs4_delegreturn_release, @@ -2986,10 +2965,8 @@ static int _nfs4_proc_delegreturn(struct inode *inode, struct rpc_cred *cred, co data->rpc_status = 0; task = rpc_run_task(NFS_CLIENT(inode), RPC_TASK_ASYNC, &nfs4_delegreturn_ops, data); - if (IS_ERR(task)) { - nfs4_delegreturn_release(data); + if (IS_ERR(task)) return PTR_ERR(task); - } status = nfs4_wait_for_completion_rpc_task(task); if (status == 0) { status = data->rpc_status; @@ -3209,7 +3186,6 @@ static struct rpc_task *nfs4_do_unlck(struct file_lock *fl, struct nfs_seqid *seqid) { struct nfs4_unlockdata *data; - struct rpc_task *task; data = nfs4_alloc_unlockdata(fl, ctx, lsp, seqid); if (data == NULL) { @@ -3219,10 +3195,7 @@ static struct rpc_task *nfs4_do_unlck(struct file_lock *fl, /* Unlock _before_ we do the RPC call */ do_vfs_lock(fl->fl_file, fl); - task = rpc_run_task(NFS_CLIENT(lsp->ls_state->inode), RPC_TASK_ASYNC, &nfs4_locku_ops, data); - if (IS_ERR(task)) - nfs4_locku_release_calldata(data); - return task; + return rpc_run_task(NFS_CLIENT(lsp->ls_state->inode), RPC_TASK_ASYNC, &nfs4_locku_ops, data); } static int nfs4_proc_unlck(struct nfs4_state *state, int cmd, struct file_lock *request) @@ -3403,10 +3376,8 @@ static int _nfs4_do_setlk(struct nfs4_state *state, int cmd, struct file_lock *f data->arg.reclaim = 1; task = rpc_run_task(NFS_CLIENT(state->inode), RPC_TASK_ASYNC, &nfs4_lock_ops, data); - if (IS_ERR(task)) { - nfs4_lock_release(data); + if (IS_ERR(task)) return PTR_ERR(task); - } ret = nfs4_wait_for_completion_rpc_task(task); if (ret == 0) { ret = data->rpc_status; @@ -3588,6 +3559,8 @@ ssize_t nfs4_listxattr(struct dentry *dentry, char *buf, size_t buflen) { size_t len = strlen(XATTR_NAME_NFSV4_ACL) + 1; + if (!nfs4_server_supports_acls(NFS_SERVER(dentry->d_inode))) + return 0; if (buf && buflen < len) return -ERANGE; if (buf) @@ -3644,8 +3617,11 @@ struct nfs_rpc_ops nfs_v4_clientops = { .pathconf = nfs4_proc_pathconf, .decode_dirent = nfs4_decode_dirent, .read_setup = nfs4_proc_read_setup, + .read_done = nfs4_read_done, .write_setup = nfs4_proc_write_setup, + .write_done = nfs4_write_done, .commit_setup = nfs4_proc_commit_setup, + .commit_done = nfs4_commit_done, .file_open = nfs_open, .file_release = nfs_release, .lock = nfs4_proc_lock, diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c index afad0255e7d..96e5b82c153 100644 --- a/fs/nfs/nfs4state.c +++ b/fs/nfs/nfs4state.c @@ -977,6 +977,7 @@ out: out_error: printk(KERN_WARNING "Error: state recovery failed on NFSv4 server %u.%u.%u.%u with error %d\n", NIPQUAD(clp->cl_addr.s_addr), -status); + set_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state); goto out; } diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c index 0a1bd36a483..7c5d70efe72 100644 --- a/fs/nfs/nfs4xdr.c +++ b/fs/nfs/nfs4xdr.c @@ -4344,6 +4344,8 @@ nfs_stat_to_errno(int stat) .p_encode = (kxdrproc_t) nfs4_xdr_##argtype, \ .p_decode = (kxdrproc_t) nfs4_xdr_##restype, \ .p_bufsiz = MAX(NFS4_##argtype##_sz,NFS4_##restype##_sz) << 2, \ + .p_statidx = NFSPROC4_CLNT_##proc, \ + .p_name = #proc, \ } struct rpc_procinfo nfs4_procedures[] = { diff --git a/fs/nfs/pagelist.c b/fs/nfs/pagelist.c index d53857b148e..106aca388eb 100644 --- a/fs/nfs/pagelist.c +++ b/fs/nfs/pagelist.c @@ -85,6 +85,9 @@ nfs_create_request(struct nfs_open_context *ctx, struct inode *inode, atomic_set(&req->wb_complete, 0); req->wb_index = page->index; page_cache_get(page); + BUG_ON(PagePrivate(page)); + BUG_ON(!PageLocked(page)); + BUG_ON(page->mapping->host != inode); req->wb_offset = offset; req->wb_pgbase = offset; req->wb_bytes = count; @@ -132,9 +135,11 @@ void nfs_clear_page_writeback(struct nfs_page *req) { struct nfs_inode *nfsi = NFS_I(req->wb_context->dentry->d_inode); - spin_lock(&nfsi->req_lock); - radix_tree_tag_clear(&nfsi->nfs_page_tree, req->wb_index, NFS_PAGE_TAG_WRITEBACK); - spin_unlock(&nfsi->req_lock); + if (req->wb_page != NULL) { + spin_lock(&nfsi->req_lock); + radix_tree_tag_clear(&nfsi->nfs_page_tree, req->wb_index, NFS_PAGE_TAG_WRITEBACK); + spin_unlock(&nfsi->req_lock); + } nfs_unlock_request(req); } @@ -147,8 +152,9 @@ void nfs_clear_page_writeback(struct nfs_page *req) */ void nfs_clear_request(struct nfs_page *req) { - if (req->wb_page) { - page_cache_release(req->wb_page); + struct page *page = req->wb_page; + if (page != NULL) { + page_cache_release(page); req->wb_page = NULL; } } diff --git a/fs/nfs/proc.c b/fs/nfs/proc.c index f5150d71c03..9dd85cac2df 100644 --- a/fs/nfs/proc.c +++ b/fs/nfs/proc.c @@ -58,16 +58,23 @@ nfs_proc_get_root(struct nfs_server *server, struct nfs_fh *fhandle, { struct nfs_fattr *fattr = info->fattr; struct nfs2_fsstat fsinfo; + struct rpc_message msg = { + .rpc_proc = &nfs_procedures[NFSPROC_GETATTR], + .rpc_argp = fhandle, + .rpc_resp = fattr, + }; int status; dprintk("%s: call getattr\n", __FUNCTION__); nfs_fattr_init(fattr); - status = rpc_call(server->client_sys, NFSPROC_GETATTR, fhandle, fattr, 0); + status = rpc_call_sync(server->client_sys, &msg, 0); dprintk("%s: reply getattr: %d\n", __FUNCTION__, status); if (status) return status; dprintk("%s: call statfs\n", __FUNCTION__); - status = rpc_call(server->client_sys, NFSPROC_STATFS, fhandle, &fsinfo, 0); + msg.rpc_proc = &nfs_procedures[NFSPROC_STATFS]; + msg.rpc_resp = &fsinfo; + status = rpc_call_sync(server->client_sys, &msg, 0); dprintk("%s: reply statfs: %d\n", __FUNCTION__, status); if (status) return status; @@ -90,12 +97,16 @@ static int nfs_proc_getattr(struct nfs_server *server, struct nfs_fh *fhandle, struct nfs_fattr *fattr) { + struct rpc_message msg = { + .rpc_proc = &nfs_procedures[NFSPROC_GETATTR], + .rpc_argp = fhandle, + .rpc_resp = fattr, + }; int status; dprintk("NFS call getattr\n"); nfs_fattr_init(fattr); - status = rpc_call(server->client, NFSPROC_GETATTR, - fhandle, fattr, 0); + status = rpc_call_sync(server->client, &msg, 0); dprintk("NFS reply getattr: %d\n", status); return status; } @@ -109,6 +120,11 @@ nfs_proc_setattr(struct dentry *dentry, struct nfs_fattr *fattr, .fh = NFS_FH(inode), .sattr = sattr }; + struct rpc_message msg = { + .rpc_proc = &nfs_procedures[NFSPROC_SETATTR], + .rpc_argp = &arg, + .rpc_resp = fattr, + }; int status; /* Mask out the non-modebit related stuff from attr->ia_mode */ @@ -116,7 +132,7 @@ nfs_proc_setattr(struct dentry *dentry, struct nfs_fattr *fattr, dprintk("NFS call setattr\n"); nfs_fattr_init(fattr); - status = rpc_call(NFS_CLIENT(inode), NFSPROC_SETATTR, &arg, fattr, 0); + status = rpc_call_sync(NFS_CLIENT(inode), &msg, 0); if (status == 0) nfs_setattr_update_inode(inode, sattr); dprintk("NFS reply setattr: %d\n", status); @@ -136,11 +152,16 @@ nfs_proc_lookup(struct inode *dir, struct qstr *name, .fh = fhandle, .fattr = fattr }; + struct rpc_message msg = { + .rpc_proc = &nfs_procedures[NFSPROC_LOOKUP], + .rpc_argp = &arg, + .rpc_resp = &res, + }; int status; dprintk("NFS call lookup %s\n", name->name); nfs_fattr_init(fattr); - status = rpc_call(NFS_CLIENT(dir), NFSPROC_LOOKUP, &arg, &res, 0); + status = rpc_call_sync(NFS_CLIENT(dir), &msg, 0); dprintk("NFS reply lookup: %d\n", status); return status; } @@ -154,10 +175,14 @@ static int nfs_proc_readlink(struct inode *inode, struct page *page, .pglen = pglen, .pages = &page }; + struct rpc_message msg = { + .rpc_proc = &nfs_procedures[NFSPROC_READLINK], + .rpc_argp = &args, + }; int status; dprintk("NFS call readlink\n"); - status = rpc_call(NFS_CLIENT(inode), NFSPROC_READLINK, &args, NULL, 0); + status = rpc_call_sync(NFS_CLIENT(inode), &msg, 0); dprintk("NFS reply readlink: %d\n", status); return status; } @@ -233,11 +258,16 @@ nfs_proc_create(struct inode *dir, struct dentry *dentry, struct iattr *sattr, .fh = &fhandle, .fattr = &fattr }; + struct rpc_message msg = { + .rpc_proc = &nfs_procedures[NFSPROC_CREATE], + .rpc_argp = &arg, + .rpc_resp = &res, + }; int status; nfs_fattr_init(&fattr); dprintk("NFS call create %s\n", dentry->d_name.name); - status = rpc_call(NFS_CLIENT(dir), NFSPROC_CREATE, &arg, &res, 0); + status = rpc_call_sync(NFS_CLIENT(dir), &msg, 0); if (status == 0) status = nfs_instantiate(dentry, &fhandle, &fattr); dprintk("NFS reply create: %d\n", status); @@ -263,6 +293,11 @@ nfs_proc_mknod(struct inode *dir, struct dentry *dentry, struct iattr *sattr, .fh = &fhandle, .fattr = &fattr }; + struct rpc_message msg = { + .rpc_proc = &nfs_procedures[NFSPROC_CREATE], + .rpc_argp = &arg, + .rpc_resp = &res, + }; int status, mode; dprintk("NFS call mknod %s\n", dentry->d_name.name); @@ -277,13 +312,13 @@ nfs_proc_mknod(struct inode *dir, struct dentry *dentry, struct iattr *sattr, } nfs_fattr_init(&fattr); - status = rpc_call(NFS_CLIENT(dir), NFSPROC_CREATE, &arg, &res, 0); + status = rpc_call_sync(NFS_CLIENT(dir), &msg, 0); nfs_mark_for_revalidate(dir); if (status == -EINVAL && S_ISFIFO(mode)) { sattr->ia_mode = mode; nfs_fattr_init(&fattr); - status = rpc_call(NFS_CLIENT(dir), NFSPROC_CREATE, &arg, &res, 0); + status = rpc_call_sync(NFS_CLIENT(dir), &msg, 0); } if (status == 0) status = nfs_instantiate(dentry, &fhandle, &fattr); @@ -302,8 +337,6 @@ nfs_proc_remove(struct inode *dir, struct qstr *name) struct rpc_message msg = { .rpc_proc = &nfs_procedures[NFSPROC_REMOVE], .rpc_argp = &arg, - .rpc_resp = NULL, - .rpc_cred = NULL }; int status; @@ -355,10 +388,14 @@ nfs_proc_rename(struct inode *old_dir, struct qstr *old_name, .toname = new_name->name, .tolen = new_name->len }; + struct rpc_message msg = { + .rpc_proc = &nfs_procedures[NFSPROC_RENAME], + .rpc_argp = &arg, + }; int status; dprintk("NFS call rename %s -> %s\n", old_name->name, new_name->name); - status = rpc_call(NFS_CLIENT(old_dir), NFSPROC_RENAME, &arg, NULL, 0); + status = rpc_call_sync(NFS_CLIENT(old_dir), &msg, 0); nfs_mark_for_revalidate(old_dir); nfs_mark_for_revalidate(new_dir); dprintk("NFS reply rename: %d\n", status); @@ -374,10 +411,14 @@ nfs_proc_link(struct inode *inode, struct inode *dir, struct qstr *name) .toname = name->name, .tolen = name->len }; + struct rpc_message msg = { + .rpc_proc = &nfs_procedures[NFSPROC_LINK], + .rpc_argp = &arg, + }; int status; dprintk("NFS call link %s\n", name->name); - status = rpc_call(NFS_CLIENT(inode), NFSPROC_LINK, &arg, NULL, 0); + status = rpc_call_sync(NFS_CLIENT(inode), &msg, 0); nfs_mark_for_revalidate(inode); nfs_mark_for_revalidate(dir); dprintk("NFS reply link: %d\n", status); @@ -397,6 +438,10 @@ nfs_proc_symlink(struct inode *dir, struct qstr *name, struct qstr *path, .tolen = path->len, .sattr = sattr }; + struct rpc_message msg = { + .rpc_proc = &nfs_procedures[NFSPROC_SYMLINK], + .rpc_argp = &arg, + }; int status; if (path->len > NFS2_MAXPATHLEN) @@ -404,7 +449,7 @@ nfs_proc_symlink(struct inode *dir, struct qstr *name, struct qstr *path, dprintk("NFS call symlink %s -> %s\n", name->name, path->name); nfs_fattr_init(fattr); fhandle->size = 0; - status = rpc_call(NFS_CLIENT(dir), NFSPROC_SYMLINK, &arg, NULL, 0); + status = rpc_call_sync(NFS_CLIENT(dir), &msg, 0); nfs_mark_for_revalidate(dir); dprintk("NFS reply symlink: %d\n", status); return status; @@ -425,11 +470,16 @@ nfs_proc_mkdir(struct inode *dir, struct dentry *dentry, struct iattr *sattr) .fh = &fhandle, .fattr = &fattr }; + struct rpc_message msg = { + .rpc_proc = &nfs_procedures[NFSPROC_MKDIR], + .rpc_argp = &arg, + .rpc_resp = &res, + }; int status; dprintk("NFS call mkdir %s\n", dentry->d_name.name); nfs_fattr_init(&fattr); - status = rpc_call(NFS_CLIENT(dir), NFSPROC_MKDIR, &arg, &res, 0); + status = rpc_call_sync(NFS_CLIENT(dir), &msg, 0); nfs_mark_for_revalidate(dir); if (status == 0) status = nfs_instantiate(dentry, &fhandle, &fattr); @@ -445,10 +495,14 @@ nfs_proc_rmdir(struct inode *dir, struct qstr *name) .name = name->name, .len = name->len }; + struct rpc_message msg = { + .rpc_proc = &nfs_procedures[NFSPROC_RMDIR], + .rpc_argp = &arg, + }; int status; dprintk("NFS call rmdir %s\n", name->name); - status = rpc_call(NFS_CLIENT(dir), NFSPROC_RMDIR, &arg, NULL, 0); + status = rpc_call_sync(NFS_CLIENT(dir), &msg, 0); nfs_mark_for_revalidate(dir); dprintk("NFS reply rmdir: %d\n", status); return status; @@ -470,13 +524,12 @@ nfs_proc_readdir(struct dentry *dentry, struct rpc_cred *cred, .fh = NFS_FH(dir), .cookie = cookie, .count = count, - .pages = &page + .pages = &page, }; struct rpc_message msg = { .rpc_proc = &nfs_procedures[NFSPROC_READDIR], .rpc_argp = &arg, - .rpc_resp = NULL, - .rpc_cred = cred + .rpc_cred = cred, }; int status; @@ -495,11 +548,16 @@ nfs_proc_statfs(struct nfs_server *server, struct nfs_fh *fhandle, struct nfs_fsstat *stat) { struct nfs2_fsstat fsinfo; + struct rpc_message msg = { + .rpc_proc = &nfs_procedures[NFSPROC_STATFS], + .rpc_argp = fhandle, + .rpc_resp = &fsinfo, + }; int status; dprintk("NFS call statfs\n"); nfs_fattr_init(stat->fattr); - status = rpc_call(server->client, NFSPROC_STATFS, fhandle, &fsinfo, 0); + status = rpc_call_sync(server->client, &msg, 0); dprintk("NFS reply statfs: %d\n", status); if (status) goto out; @@ -518,11 +576,16 @@ nfs_proc_fsinfo(struct nfs_server *server, struct nfs_fh *fhandle, struct nfs_fsinfo *info) { struct nfs2_fsstat fsinfo; + struct rpc_message msg = { + .rpc_proc = &nfs_procedures[NFSPROC_STATFS], + .rpc_argp = fhandle, + .rpc_resp = &fsinfo, + }; int status; dprintk("NFS call fsinfo\n"); nfs_fattr_init(info->fattr); - status = rpc_call(server->client, NFSPROC_STATFS, fhandle, &fsinfo, 0); + status = rpc_call_sync(server->client, &msg, 0); dprintk("NFS reply fsinfo: %d\n", status); if (status) goto out; @@ -550,10 +613,8 @@ nfs_proc_pathconf(struct nfs_server *server, struct nfs_fh *fhandle, extern u32 * nfs_decode_dirent(u32 *, struct nfs_entry *, int); -static void nfs_read_done(struct rpc_task *task, void *calldata) +static int nfs_read_done(struct rpc_task *task, struct nfs_read_data *data) { - struct nfs_read_data *data = calldata; - if (task->tk_status >= 0) { nfs_refresh_inode(data->inode, data->res.fattr); /* Emulate the eof flag, which isn't normally needed in NFSv2 @@ -562,20 +623,11 @@ static void nfs_read_done(struct rpc_task *task, void *calldata) if (data->args.offset + data->args.count >= data->res.fattr->size) data->res.eof = 1; } - nfs_readpage_result(task, calldata); + return 0; } -static const struct rpc_call_ops nfs_read_ops = { - .rpc_call_done = nfs_read_done, - .rpc_release = nfs_readdata_release, -}; - -static void -nfs_proc_read_setup(struct nfs_read_data *data) +static void nfs_proc_read_setup(struct nfs_read_data *data) { - struct rpc_task *task = &data->task; - struct inode *inode = data->inode; - int flags; struct rpc_message msg = { .rpc_proc = &nfs_procedures[NFSPROC_READ], .rpc_argp = &data->args, @@ -583,34 +635,18 @@ nfs_proc_read_setup(struct nfs_read_data *data) .rpc_cred = data->cred, }; - /* N.B. Do we need to test? Never called for swapfile inode */ - flags = RPC_TASK_ASYNC | (IS_SWAPFILE(inode)? NFS_RPC_SWAPFLAGS : 0); - - /* Finalize the task. */ - rpc_init_task(task, NFS_CLIENT(inode), flags, &nfs_read_ops, data); - rpc_call_setup(task, &msg, 0); + rpc_call_setup(&data->task, &msg, 0); } -static void nfs_write_done(struct rpc_task *task, void *calldata) +static int nfs_write_done(struct rpc_task *task, struct nfs_write_data *data) { - struct nfs_write_data *data = calldata; - if (task->tk_status >= 0) nfs_post_op_update_inode(data->inode, data->res.fattr); - nfs_writeback_done(task, calldata); + return 0; } -static const struct rpc_call_ops nfs_write_ops = { - .rpc_call_done = nfs_write_done, - .rpc_release = nfs_writedata_release, -}; - -static void -nfs_proc_write_setup(struct nfs_write_data *data, int how) +static void nfs_proc_write_setup(struct nfs_write_data *data, int how) { - struct rpc_task *task = &data->task; - struct inode *inode = data->inode; - int flags; struct rpc_message msg = { .rpc_proc = &nfs_procedures[NFSPROC_WRITE], .rpc_argp = &data->args, @@ -621,12 +657,8 @@ nfs_proc_write_setup(struct nfs_write_data *data, int how) /* Note: NFSv2 ignores @stable and always uses NFS_FILE_SYNC */ data->args.stable = NFS_FILE_SYNC; - /* Set the initial flags for the task. */ - flags = (how & FLUSH_SYNC) ? 0 : RPC_TASK_ASYNC; - /* Finalize the task. */ - rpc_init_task(task, NFS_CLIENT(inode), flags, &nfs_write_ops, data); - rpc_call_setup(task, &msg, 0); + rpc_call_setup(&data->task, &msg, 0); } static void @@ -672,7 +704,9 @@ struct nfs_rpc_ops nfs_v2_clientops = { .pathconf = nfs_proc_pathconf, .decode_dirent = nfs_decode_dirent, .read_setup = nfs_proc_read_setup, + .read_done = nfs_read_done, .write_setup = nfs_proc_write_setup, + .write_done = nfs_write_done, .commit_setup = nfs_proc_commit_setup, .file_open = nfs_open, .file_release = nfs_release, diff --git a/fs/nfs/read.c b/fs/nfs/read.c index 05eb43fadf8..624ca7146b6 100644 --- a/fs/nfs/read.c +++ b/fs/nfs/read.c @@ -31,17 +31,49 @@ #include <asm/system.h> +#include "iostat.h" + #define NFSDBG_FACILITY NFSDBG_PAGECACHE static int nfs_pagein_one(struct list_head *, struct inode *); -static void nfs_readpage_result_partial(struct nfs_read_data *, int); -static void nfs_readpage_result_full(struct nfs_read_data *, int); +static const struct rpc_call_ops nfs_read_partial_ops; +static const struct rpc_call_ops nfs_read_full_ops; static kmem_cache_t *nfs_rdata_cachep; -mempool_t *nfs_rdata_mempool; +static mempool_t *nfs_rdata_mempool; #define MIN_POOL_READ (32) +struct nfs_read_data *nfs_readdata_alloc(unsigned int pagecount) +{ + struct nfs_read_data *p = mempool_alloc(nfs_rdata_mempool, SLAB_NOFS); + + if (p) { + memset(p, 0, sizeof(*p)); + INIT_LIST_HEAD(&p->pages); + if (pagecount < NFS_PAGEVEC_SIZE) + p->pagevec = &p->page_array[0]; + else { + size_t size = ++pagecount * sizeof(struct page *); + p->pagevec = kmalloc(size, GFP_NOFS); + if (p->pagevec) { + memset(p->pagevec, 0, size); + } else { + mempool_free(p, nfs_rdata_mempool); + p = NULL; + } + } + } + return p; +} + +void nfs_readdata_free(struct nfs_read_data *p) +{ + if (p && (p->pagevec != &p->page_array[0])) + kfree(p->pagevec); + mempool_free(p, nfs_rdata_mempool); +} + void nfs_readdata_release(void *data) { nfs_readdata_free(data); @@ -133,6 +165,8 @@ static int nfs_readpage_sync(struct nfs_open_context *ctx, struct inode *inode, } count -= result; rdata->args.pgbase += result; + nfs_add_stats(inode, NFSIOS_SERVERREADBYTES, result); + /* Note: result == 0 should only happen if we're caching * a write that extends the file and punches a hole. */ @@ -196,9 +230,11 @@ static void nfs_readpage_release(struct nfs_page *req) * Set up the NFS read request struct */ static void nfs_read_rpcsetup(struct nfs_page *req, struct nfs_read_data *data, + const struct rpc_call_ops *call_ops, unsigned int count, unsigned int offset) { struct inode *inode; + int flags; data->req = req; data->inode = inode = req->wb_context->dentry->d_inode; @@ -216,6 +252,9 @@ static void nfs_read_rpcsetup(struct nfs_page *req, struct nfs_read_data *data, data->res.eof = 0; nfs_fattr_init(&data->fattr); + /* Set up the initial task struct. */ + flags = RPC_TASK_ASYNC | (IS_SWAPFILE(inode)? NFS_RPC_SWAPFLAGS : 0); + rpc_init_task(&data->task, NFS_CLIENT(inode), flags, call_ops, data); NFS_PROTO(inode)->read_setup(data); data->task.tk_cookie = (unsigned long)inode; @@ -303,14 +342,15 @@ static int nfs_pagein_multi(struct list_head *head, struct inode *inode) list_del_init(&data->pages); data->pagevec[0] = page; - data->complete = nfs_readpage_result_partial; if (nbytes > rsize) { - nfs_read_rpcsetup(req, data, rsize, offset); + nfs_read_rpcsetup(req, data, &nfs_read_partial_ops, + rsize, offset); offset += rsize; nbytes -= rsize; } else { - nfs_read_rpcsetup(req, data, nbytes, offset); + nfs_read_rpcsetup(req, data, &nfs_read_partial_ops, + nbytes, offset); nbytes = 0; } nfs_execute_read(data); @@ -356,8 +396,7 @@ static int nfs_pagein_one(struct list_head *head, struct inode *inode) } req = nfs_list_entry(data->pages.next); - data->complete = nfs_readpage_result_full; - nfs_read_rpcsetup(req, data, count, 0); + nfs_read_rpcsetup(req, data, &nfs_read_full_ops, count, 0); nfs_execute_read(data); return 0; @@ -391,12 +430,15 @@ nfs_pagein_list(struct list_head *head, int rpages) /* * Handle a read reply that fills part of a page. */ -static void nfs_readpage_result_partial(struct nfs_read_data *data, int status) +static void nfs_readpage_result_partial(struct rpc_task *task, void *calldata) { + struct nfs_read_data *data = calldata; struct nfs_page *req = data->req; struct page *page = req->wb_page; - if (status >= 0) { + if (nfs_readpage_result(task, data) != 0) + return; + if (task->tk_status >= 0) { unsigned int request = data->args.count; unsigned int result = data->res.count; @@ -415,20 +457,28 @@ static void nfs_readpage_result_partial(struct nfs_read_data *data, int status) } } +static const struct rpc_call_ops nfs_read_partial_ops = { + .rpc_call_done = nfs_readpage_result_partial, + .rpc_release = nfs_readdata_release, +}; + /* * This is the callback from RPC telling us whether a reply was * received or some error occurred (timeout or socket shutdown). */ -static void nfs_readpage_result_full(struct nfs_read_data *data, int status) +static void nfs_readpage_result_full(struct rpc_task *task, void *calldata) { + struct nfs_read_data *data = calldata; unsigned int count = data->res.count; + if (nfs_readpage_result(task, data) != 0) + return; while (!list_empty(&data->pages)) { struct nfs_page *req = nfs_list_entry(data->pages.next); struct page *page = req->wb_page; nfs_list_remove_request(req); - if (status >= 0) { + if (task->tk_status >= 0) { if (count < PAGE_CACHE_SIZE) { if (count < req->wb_bytes) memclear_highpage_flush(page, @@ -444,22 +494,33 @@ static void nfs_readpage_result_full(struct nfs_read_data *data, int status) } } +static const struct rpc_call_ops nfs_read_full_ops = { + .rpc_call_done = nfs_readpage_result_full, + .rpc_release = nfs_readdata_release, +}; + /* * This is the callback from RPC telling us whether a reply was * received or some error occurred (timeout or socket shutdown). */ -void nfs_readpage_result(struct rpc_task *task, void *calldata) +int nfs_readpage_result(struct rpc_task *task, struct nfs_read_data *data) { - struct nfs_read_data *data = calldata; struct nfs_readargs *argp = &data->args; struct nfs_readres *resp = &data->res; - int status = task->tk_status; + int status; dprintk("NFS: %4d nfs_readpage_result, (status %d)\n", - task->tk_pid, status); + task->tk_pid, task->tk_status); + + status = NFS_PROTO(data->inode)->read_done(task, data); + if (status != 0) + return status; + + nfs_add_stats(data->inode, NFSIOS_SERVERREADBYTES, resp->count); /* Is this a short read? */ if (task->tk_status >= 0 && resp->count < argp->count && !resp->eof) { + nfs_inc_stats(data->inode, NFSIOS_SHORTREAD); /* Has the server at least made some progress? */ if (resp->count != 0) { /* Yes, so retry the read at the end of the data */ @@ -467,14 +528,14 @@ void nfs_readpage_result(struct rpc_task *task, void *calldata) argp->pgbase += resp->count; argp->count -= resp->count; rpc_restart_call(task); - return; + return -EAGAIN; } task->tk_status = -EIO; } spin_lock(&data->inode->i_lock); NFS_I(data->inode)->cache_validity |= NFS_INO_INVALID_ATIME; spin_unlock(&data->inode->i_lock); - data->complete(data, status); + return 0; } /* @@ -491,6 +552,9 @@ int nfs_readpage(struct file *file, struct page *page) dprintk("NFS: nfs_readpage (%p %ld@%lu)\n", page, PAGE_CACHE_SIZE, page->index); + nfs_inc_stats(inode, NFSIOS_VFSREADPAGE); + nfs_add_stats(inode, NFSIOS_READPAGES, 1); + /* * Try to flush any pending writes to the file.. * @@ -570,6 +634,7 @@ int nfs_readpages(struct file *filp, struct address_space *mapping, inode->i_sb->s_id, (long long)NFS_FILEID(inode), nr_pages); + nfs_inc_stats(inode, NFSIOS_VFSREADPAGES); if (filp == NULL) { desc.ctx = nfs_find_open_context(inode, NULL, FMODE_READ); @@ -582,6 +647,7 @@ int nfs_readpages(struct file *filp, struct address_space *mapping, if (!list_empty(&head)) { int err = nfs_pagein_list(&head, server->rpages); if (!ret) + nfs_add_stats(inode, NFSIOS_READPAGES, err); ret = err; } put_nfs_open_context(desc.ctx); @@ -597,10 +663,8 @@ int nfs_init_readpagecache(void) if (nfs_rdata_cachep == NULL) return -ENOMEM; - nfs_rdata_mempool = mempool_create(MIN_POOL_READ, - mempool_alloc_slab, - mempool_free_slab, - nfs_rdata_cachep); + nfs_rdata_mempool = mempool_create_slab_pool(MIN_POOL_READ, + nfs_rdata_cachep); if (nfs_rdata_mempool == NULL) return -ENOMEM; diff --git a/fs/nfs/unlink.c b/fs/nfs/unlink.c index a65c7b53d55..0e28189c215 100644 --- a/fs/nfs/unlink.c +++ b/fs/nfs/unlink.c @@ -163,10 +163,9 @@ nfs_async_unlink(struct dentry *dentry) struct rpc_clnt *clnt = NFS_CLIENT(dir->d_inode); int status = -ENOMEM; - data = kmalloc(sizeof(*data), GFP_KERNEL); + data = kzalloc(sizeof(*data), GFP_KERNEL); if (!data) goto out; - memset(data, 0, sizeof(*data)); data->cred = rpcauth_lookupcred(clnt->cl_auth, 0); if (IS_ERR(data->cred)) { diff --git a/fs/nfs/write.c b/fs/nfs/write.c index 9449b683550..4cfada2cc09 100644 --- a/fs/nfs/write.c +++ b/fs/nfs/write.c @@ -63,6 +63,7 @@ #include <linux/smp_lock.h> #include "delegation.h" +#include "iostat.h" #define NFSDBG_FACILITY NFSDBG_PAGECACHE @@ -76,20 +77,21 @@ static struct nfs_page * nfs_update_request(struct nfs_open_context*, struct inode *, struct page *, unsigned int, unsigned int); -static void nfs_writeback_done_partial(struct nfs_write_data *, int); -static void nfs_writeback_done_full(struct nfs_write_data *, int); static int nfs_wait_on_write_congestion(struct address_space *, int); static int nfs_wait_on_requests(struct inode *, unsigned long, unsigned int); static int nfs_flush_inode(struct inode *inode, unsigned long idx_start, unsigned int npages, int how); +static const struct rpc_call_ops nfs_write_partial_ops; +static const struct rpc_call_ops nfs_write_full_ops; +static const struct rpc_call_ops nfs_commit_ops; static kmem_cache_t *nfs_wdata_cachep; -mempool_t *nfs_wdata_mempool; +static mempool_t *nfs_wdata_mempool; static mempool_t *nfs_commit_mempool; static DECLARE_WAIT_QUEUE_HEAD(nfs_write_congestion); -static inline struct nfs_write_data *nfs_commit_alloc(unsigned int pagecount) +struct nfs_write_data *nfs_commit_alloc(unsigned int pagecount) { struct nfs_write_data *p = mempool_alloc(nfs_commit_mempool, SLAB_NOFS); @@ -100,11 +102,39 @@ static inline struct nfs_write_data *nfs_commit_alloc(unsigned int pagecount) p->pagevec = &p->page_array[0]; else { size_t size = ++pagecount * sizeof(struct page *); + p->pagevec = kzalloc(size, GFP_NOFS); + if (!p->pagevec) { + mempool_free(p, nfs_commit_mempool); + p = NULL; + } + } + } + return p; +} + +void nfs_commit_free(struct nfs_write_data *p) +{ + if (p && (p->pagevec != &p->page_array[0])) + kfree(p->pagevec); + mempool_free(p, nfs_commit_mempool); +} + +struct nfs_write_data *nfs_writedata_alloc(unsigned int pagecount) +{ + struct nfs_write_data *p = mempool_alloc(nfs_wdata_mempool, SLAB_NOFS); + + if (p) { + memset(p, 0, sizeof(*p)); + INIT_LIST_HEAD(&p->pages); + if (pagecount < NFS_PAGEVEC_SIZE) + p->pagevec = &p->page_array[0]; + else { + size_t size = ++pagecount * sizeof(struct page *); p->pagevec = kmalloc(size, GFP_NOFS); if (p->pagevec) { memset(p->pagevec, 0, size); } else { - mempool_free(p, nfs_commit_mempool); + mempool_free(p, nfs_wdata_mempool); p = NULL; } } @@ -112,11 +142,11 @@ static inline struct nfs_write_data *nfs_commit_alloc(unsigned int pagecount) return p; } -static inline void nfs_commit_free(struct nfs_write_data *p) +void nfs_writedata_free(struct nfs_write_data *p) { if (p && (p->pagevec != &p->page_array[0])) kfree(p->pagevec); - mempool_free(p, nfs_commit_mempool); + mempool_free(p, nfs_wdata_mempool); } void nfs_writedata_release(void *wdata) @@ -136,6 +166,7 @@ static void nfs_grow_file(struct page *page, unsigned int offset, unsigned int c end = ((loff_t)page->index << PAGE_CACHE_SHIFT) + ((loff_t)offset+count); if (i_size >= end) return; + nfs_inc_stats(inode, NFSIOS_EXTENDWRITE); i_size_write(inode, end); } @@ -225,6 +256,7 @@ static int nfs_writepage_sync(struct nfs_open_context *ctx, struct inode *inode, wdata->args.pgbase += result; written += result; count -= result; + nfs_add_stats(inode, NFSIOS_SERVERWRITTENBYTES, result); } while (count); /* Update file length */ nfs_grow_file(page, offset, written); @@ -281,6 +313,9 @@ int nfs_writepage(struct page *page, struct writeback_control *wbc) int priority = wb_priority(wbc); int err; + nfs_inc_stats(inode, NFSIOS_VFSWRITEPAGE); + nfs_add_stats(inode, NFSIOS_WRITEPAGES, 1); + /* * Note: We need to ensure that we have a reference to the inode * if we are to do asynchronous writes. If not, waiting @@ -345,6 +380,8 @@ int nfs_writepages(struct address_space *mapping, struct writeback_control *wbc) struct inode *inode = mapping->host; int err; + nfs_inc_stats(inode, NFSIOS_VFSWRITEPAGES); + err = generic_writepages(mapping, wbc); if (err) return err; @@ -356,6 +393,7 @@ int nfs_writepages(struct address_space *mapping, struct writeback_control *wbc) err = nfs_flush_inode(inode, 0, 0, wb_priority(wbc)); if (err < 0) goto out; + nfs_add_stats(inode, NFSIOS_WRITEPAGES, err); wbc->nr_to_write -= err; if (!wbc->nonblocking && wbc->sync_mode == WB_SYNC_ALL) { err = nfs_wait_on_requests(inode, 0, 0); @@ -391,6 +429,7 @@ static int nfs_inode_add_request(struct inode *inode, struct nfs_page *req) if (nfs_have_delegation(inode, FMODE_WRITE)) nfsi->change_attr++; } + SetPagePrivate(req->wb_page); nfsi->npages++; atomic_inc(&req->wb_count); return 0; @@ -407,6 +446,7 @@ static void nfs_inode_remove_request(struct nfs_page *req) BUG_ON (!NFS_WBACK_BUSY(req)); spin_lock(&nfsi->req_lock); + ClearPagePrivate(req->wb_page); radix_tree_delete(&nfsi->nfs_page_tree, req->wb_index); nfsi->npages--; if (!nfsi->npages) { @@ -499,8 +539,7 @@ nfs_mark_request_commit(struct nfs_page *req) * * Interruptible by signals only if mounted with intr flag. */ -static int -nfs_wait_on_requests(struct inode *inode, unsigned long idx_start, unsigned int npages) +static int nfs_wait_on_requests_locked(struct inode *inode, unsigned long idx_start, unsigned int npages) { struct nfs_inode *nfsi = NFS_I(inode); struct nfs_page *req; @@ -513,7 +552,6 @@ nfs_wait_on_requests(struct inode *inode, unsigned long idx_start, unsigned int else idx_end = idx_start + npages - 1; - spin_lock(&nfsi->req_lock); next = idx_start; while (radix_tree_gang_lookup_tag(&nfsi->nfs_page_tree, (void **)&req, next, 1, NFS_PAGE_TAG_WRITEBACK)) { if (req->wb_index > idx_end) @@ -526,15 +564,25 @@ nfs_wait_on_requests(struct inode *inode, unsigned long idx_start, unsigned int spin_unlock(&nfsi->req_lock); error = nfs_wait_on_request(req); nfs_release_request(req); + spin_lock(&nfsi->req_lock); if (error < 0) return error; - spin_lock(&nfsi->req_lock); res++; } - spin_unlock(&nfsi->req_lock); return res; } +static int nfs_wait_on_requests(struct inode *inode, unsigned long idx_start, unsigned int npages) +{ + struct nfs_inode *nfsi = NFS_I(inode); + int ret; + + spin_lock(&nfsi->req_lock); + ret = nfs_wait_on_requests_locked(inode, idx_start, npages); + spin_unlock(&nfsi->req_lock); + return ret; +} + /* * nfs_scan_dirty - Scan an inode for dirty requests * @inode: NFS inode to scan @@ -586,6 +634,11 @@ nfs_scan_commit(struct inode *inode, struct list_head *dst, unsigned long idx_st } return res; } +#else +static inline int nfs_scan_commit(struct inode *inode, struct list_head *dst, unsigned long idx_start, unsigned int npages) +{ + return 0; +} #endif static int nfs_wait_on_write_congestion(struct address_space *mapping, int intr) @@ -598,6 +651,9 @@ static int nfs_wait_on_write_congestion(struct address_space *mapping, int intr) if (!bdi_write_congested(bdi)) return 0; + + nfs_inc_stats(mapping->host, NFSIOS_CONGESTIONWAIT); + if (intr) { struct rpc_clnt *clnt = NFS_CLIENT(mapping->host); sigset_t oldset; @@ -653,8 +709,11 @@ static struct nfs_page * nfs_update_request(struct nfs_open_context* ctx, spin_unlock(&nfsi->req_lock); error = nfs_wait_on_request(req); nfs_release_request(req); - if (error < 0) + if (error < 0) { + if (new) + nfs_release_request(new); return ERR_PTR(error); + } continue; } spin_unlock(&nfsi->req_lock); @@ -748,6 +807,8 @@ int nfs_updatepage(struct file *file, struct page *page, struct nfs_page *req; int status = 0; + nfs_inc_stats(inode, NFSIOS_VFSUPDATEPAGE); + dprintk("NFS: nfs_updatepage(%s/%s %d@%Ld)\n", file->f_dentry->d_parent->d_name.name, file->f_dentry->d_name.name, count, @@ -857,10 +918,12 @@ static inline int flush_task_priority(int how) */ static void nfs_write_rpcsetup(struct nfs_page *req, struct nfs_write_data *data, + const struct rpc_call_ops *call_ops, unsigned int count, unsigned int offset, int how) { struct inode *inode; + int flags; /* Set up the RPC argument and reply structs * NB: take care not to mess about with data->commit et al. */ @@ -881,6 +944,9 @@ static void nfs_write_rpcsetup(struct nfs_page *req, data->res.verf = &data->verf; nfs_fattr_init(&data->fattr); + /* Set up the initial task struct. */ + flags = (how & FLUSH_SYNC) ? 0 : RPC_TASK_ASYNC; + rpc_init_task(&data->task, NFS_CLIENT(inode), flags, call_ops, data); NFS_PROTO(inode)->write_setup(data, how); data->task.tk_priority = flush_task_priority(how); @@ -910,7 +976,7 @@ static void nfs_execute_write(struct nfs_write_data *data) * Generate multiple small requests to write out a single * contiguous dirty area on one page. */ -static int nfs_flush_multi(struct list_head *head, struct inode *inode, int how) +static int nfs_flush_multi(struct inode *inode, struct list_head *head, int how) { struct nfs_page *req = nfs_list_entry(head->next); struct page *page = req->wb_page; @@ -944,14 +1010,15 @@ static int nfs_flush_multi(struct list_head *head, struct inode *inode, int how) list_del_init(&data->pages); data->pagevec[0] = page; - data->complete = nfs_writeback_done_partial; if (nbytes > wsize) { - nfs_write_rpcsetup(req, data, wsize, offset, how); + nfs_write_rpcsetup(req, data, &nfs_write_partial_ops, + wsize, offset, how); offset += wsize; nbytes -= wsize; } else { - nfs_write_rpcsetup(req, data, nbytes, offset, how); + nfs_write_rpcsetup(req, data, &nfs_write_partial_ops, + nbytes, offset, how); nbytes = 0; } nfs_execute_write(data); @@ -978,16 +1045,13 @@ out_bad: * This is the case if nfs_updatepage detects a conflicting request * that has been written but not committed. */ -static int nfs_flush_one(struct list_head *head, struct inode *inode, int how) +static int nfs_flush_one(struct inode *inode, struct list_head *head, int how) { struct nfs_page *req; struct page **pages; struct nfs_write_data *data; unsigned int count; - if (NFS_SERVER(inode)->wsize < PAGE_CACHE_SIZE) - return nfs_flush_multi(head, inode, how); - data = nfs_writedata_alloc(NFS_SERVER(inode)->wpages); if (!data) goto out_bad; @@ -1005,9 +1069,8 @@ static int nfs_flush_one(struct list_head *head, struct inode *inode, int how) } req = nfs_list_entry(data->pages.next); - data->complete = nfs_writeback_done_full; /* Set up the argument struct */ - nfs_write_rpcsetup(req, data, count, 0, how); + nfs_write_rpcsetup(req, data, &nfs_write_full_ops, count, 0, how); nfs_execute_write(data); return 0; @@ -1021,24 +1084,32 @@ static int nfs_flush_one(struct list_head *head, struct inode *inode, int how) return -ENOMEM; } -static int -nfs_flush_list(struct list_head *head, int wpages, int how) +static int nfs_flush_list(struct inode *inode, struct list_head *head, int npages, int how) { LIST_HEAD(one_request); - struct nfs_page *req; - int error = 0; - unsigned int pages = 0; + int (*flush_one)(struct inode *, struct list_head *, int); + struct nfs_page *req; + int wpages = NFS_SERVER(inode)->wpages; + int wsize = NFS_SERVER(inode)->wsize; + int error; - while (!list_empty(head)) { - pages += nfs_coalesce_requests(head, &one_request, wpages); + flush_one = nfs_flush_one; + if (wsize < PAGE_CACHE_SIZE) + flush_one = nfs_flush_multi; + /* For single writes, FLUSH_STABLE is more efficient */ + if (npages <= wpages && npages == NFS_I(inode)->npages + && nfs_list_entry(head->next)->wb_bytes <= wsize) + how |= FLUSH_STABLE; + + do { + nfs_coalesce_requests(head, &one_request, wpages); req = nfs_list_entry(one_request.next); - error = nfs_flush_one(&one_request, req->wb_context->dentry->d_inode, how); + error = flush_one(inode, &one_request, how); if (error < 0) - break; - } - if (error >= 0) - return pages; - + goto out_err; + } while (!list_empty(head)); + return 0; +out_err: while (!list_empty(head)) { req = nfs_list_entry(head->next); nfs_list_remove_request(req); @@ -1051,8 +1122,9 @@ nfs_flush_list(struct list_head *head, int wpages, int how) /* * Handle a write reply that flushed part of a page. */ -static void nfs_writeback_done_partial(struct nfs_write_data *data, int status) +static void nfs_writeback_done_partial(struct rpc_task *task, void *calldata) { + struct nfs_write_data *data = calldata; struct nfs_page *req = data->req; struct page *page = req->wb_page; @@ -1062,11 +1134,14 @@ static void nfs_writeback_done_partial(struct nfs_write_data *data, int status) req->wb_bytes, (long long)req_offset(req)); - if (status < 0) { + if (nfs_writeback_done(task, data) != 0) + return; + + if (task->tk_status < 0) { ClearPageUptodate(page); SetPageError(page); - req->wb_context->error = status; - dprintk(", error = %d\n", status); + req->wb_context->error = task->tk_status; + dprintk(", error = %d\n", task->tk_status); } else { #if defined(CONFIG_NFS_V3) || defined(CONFIG_NFS_V4) if (data->verf.committed < NFS_FILE_SYNC) { @@ -1087,6 +1162,11 @@ static void nfs_writeback_done_partial(struct nfs_write_data *data, int status) nfs_writepage_release(req); } +static const struct rpc_call_ops nfs_write_partial_ops = { + .rpc_call_done = nfs_writeback_done_partial, + .rpc_release = nfs_writedata_release, +}; + /* * Handle a write reply that flushes a whole page. * @@ -1094,11 +1174,15 @@ static void nfs_writeback_done_partial(struct nfs_write_data *data, int status) * writebacks since the page->count is kept > 1 for as long * as the page has a write request pending. */ -static void nfs_writeback_done_full(struct nfs_write_data *data, int status) +static void nfs_writeback_done_full(struct rpc_task *task, void *calldata) { + struct nfs_write_data *data = calldata; struct nfs_page *req; struct page *page; + if (nfs_writeback_done(task, data) != 0) + return; + /* Update attributes as result of writeback. */ while (!list_empty(&data->pages)) { req = nfs_list_entry(data->pages.next); @@ -1111,13 +1195,13 @@ static void nfs_writeback_done_full(struct nfs_write_data *data, int status) req->wb_bytes, (long long)req_offset(req)); - if (status < 0) { + if (task->tk_status < 0) { ClearPageUptodate(page); SetPageError(page); - req->wb_context->error = status; + req->wb_context->error = task->tk_status; end_page_writeback(page); nfs_inode_remove_request(req); - dprintk(", error = %d\n", status); + dprintk(", error = %d\n", task->tk_status); goto next; } end_page_writeback(page); @@ -1139,18 +1223,30 @@ static void nfs_writeback_done_full(struct nfs_write_data *data, int status) } } +static const struct rpc_call_ops nfs_write_full_ops = { + .rpc_call_done = nfs_writeback_done_full, + .rpc_release = nfs_writedata_release, +}; + + /* * This function is called when the WRITE call is complete. */ -void nfs_writeback_done(struct rpc_task *task, void *calldata) +int nfs_writeback_done(struct rpc_task *task, struct nfs_write_data *data) { - struct nfs_write_data *data = calldata; struct nfs_writeargs *argp = &data->args; struct nfs_writeres *resp = &data->res; + int status; dprintk("NFS: %4d nfs_writeback_done (status %d)\n", task->tk_pid, task->tk_status); + /* Call the NFS version-specific code */ + status = NFS_PROTO(data->inode)->write_done(task, data); + if (status != 0) + return status; + nfs_add_stats(data->inode, NFSIOS_SERVERWRITTENBYTES, resp->count); + #if defined(CONFIG_NFS_V3) || defined(CONFIG_NFS_V4) if (resp->verf->committed < argp->stable && task->tk_status >= 0) { /* We tried a write call, but the server did not @@ -1176,6 +1272,8 @@ void nfs_writeback_done(struct rpc_task *task, void *calldata) if (task->tk_status >= 0 && resp->count < argp->count) { static unsigned long complain; + nfs_inc_stats(data->inode, NFSIOS_SHORTWRITE); + /* Has the server at least made some progress? */ if (resp->count != 0) { /* Was this an NFSv2 write or an NFSv3 stable write? */ @@ -1191,7 +1289,7 @@ void nfs_writeback_done(struct rpc_task *task, void *calldata) argp->stable = NFS_FILE_SYNC; } rpc_restart_call(task); - return; + return -EAGAIN; } if (time_before(complain, jiffies)) { printk(KERN_WARNING @@ -1202,11 +1300,7 @@ void nfs_writeback_done(struct rpc_task *task, void *calldata) /* Can't do anything about it except throw an error. */ task->tk_status = -EIO; } - - /* - * Process the nfs_page list - */ - data->complete(data, task->tk_status); + return 0; } @@ -1220,10 +1314,12 @@ void nfs_commit_release(void *wdata) * Set up the argument/result storage required for the RPC call. */ static void nfs_commit_rpcsetup(struct list_head *head, - struct nfs_write_data *data, int how) + struct nfs_write_data *data, + int how) { struct nfs_page *first; struct inode *inode; + int flags; /* Set up the RPC argument and reply structs * NB: take care not to mess about with data->commit et al. */ @@ -1243,7 +1339,10 @@ static void nfs_commit_rpcsetup(struct list_head *head, data->res.fattr = &data->fattr; data->res.verf = &data->verf; nfs_fattr_init(&data->fattr); - + + /* Set up the initial task struct. */ + flags = (how & FLUSH_SYNC) ? 0 : RPC_TASK_ASYNC; + rpc_init_task(&data->task, NFS_CLIENT(inode), flags, &nfs_commit_ops, data); NFS_PROTO(inode)->commit_setup(data, how); data->task.tk_priority = flush_task_priority(how); @@ -1284,7 +1383,7 @@ nfs_commit_list(struct inode *inode, struct list_head *head, int how) /* * COMMIT call returned */ -void nfs_commit_done(struct rpc_task *task, void *calldata) +static void nfs_commit_done(struct rpc_task *task, void *calldata) { struct nfs_write_data *data = calldata; struct nfs_page *req; @@ -1293,6 +1392,10 @@ void nfs_commit_done(struct rpc_task *task, void *calldata) dprintk("NFS: %4d nfs_commit_done (status %d)\n", task->tk_pid, task->tk_status); + /* Call the NFS version-specific code */ + if (NFS_PROTO(data->inode)->commit_done(task, data) != 0) + return; + while (!list_empty(&data->pages)) { req = nfs_list_entry(data->pages.next); nfs_list_remove_request(req); @@ -1326,6 +1429,16 @@ void nfs_commit_done(struct rpc_task *task, void *calldata) } sub_page_state(nr_unstable,res); } + +static const struct rpc_call_ops nfs_commit_ops = { + .rpc_call_done = nfs_commit_done, + .rpc_release = nfs_commit_release, +}; +#else +static inline int nfs_commit_list(struct inode *inode, struct list_head *head, int how) +{ + return 0; +} #endif static int nfs_flush_inode(struct inode *inode, unsigned long idx_start, @@ -1333,24 +1446,16 @@ static int nfs_flush_inode(struct inode *inode, unsigned long idx_start, { struct nfs_inode *nfsi = NFS_I(inode); LIST_HEAD(head); - int res, - error = 0; + int res; spin_lock(&nfsi->req_lock); res = nfs_scan_dirty(inode, &head, idx_start, npages); spin_unlock(&nfsi->req_lock); if (res) { - struct nfs_server *server = NFS_SERVER(inode); - - /* For single writes, FLUSH_STABLE is more efficient */ - if (res == nfsi->npages && nfsi->npages <= server->wpages) { - if (res > 1 || nfs_list_entry(head.next)->wb_bytes <= server->wsize) - how |= FLUSH_STABLE; - } - error = nfs_flush_list(&head, server->wpages, how); + int error = nfs_flush_list(inode, &head, res, how); + if (error < 0) + return error; } - if (error < 0) - return error; return res; } @@ -1359,14 +1464,13 @@ int nfs_commit_inode(struct inode *inode, int how) { struct nfs_inode *nfsi = NFS_I(inode); LIST_HEAD(head); - int res, - error = 0; + int res; spin_lock(&nfsi->req_lock); res = nfs_scan_commit(inode, &head, 0, 0); spin_unlock(&nfsi->req_lock); if (res) { - error = nfs_commit_list(inode, &head, how); + int error = nfs_commit_list(inode, &head, how); if (error < 0) return error; } @@ -1374,28 +1478,38 @@ int nfs_commit_inode(struct inode *inode, int how) } #endif -int nfs_sync_inode(struct inode *inode, unsigned long idx_start, - unsigned int npages, int how) +int nfs_sync_inode_wait(struct inode *inode, unsigned long idx_start, + unsigned int npages, int how) { + struct nfs_inode *nfsi = NFS_I(inode); + LIST_HEAD(head); int nocommit = how & FLUSH_NOCOMMIT; - int wait = how & FLUSH_WAIT; - int error; - - how &= ~(FLUSH_WAIT|FLUSH_NOCOMMIT); + int pages, ret; + how &= ~FLUSH_NOCOMMIT; + spin_lock(&nfsi->req_lock); do { - if (wait) { - error = nfs_wait_on_requests(inode, idx_start, npages); - if (error != 0) - continue; - } - error = nfs_flush_inode(inode, idx_start, npages, how); - if (error != 0) + ret = nfs_wait_on_requests_locked(inode, idx_start, npages); + if (ret != 0) continue; - if (!nocommit) - error = nfs_commit_inode(inode, how); - } while (error > 0); - return error; + pages = nfs_scan_dirty(inode, &head, idx_start, npages); + if (pages != 0) { + spin_unlock(&nfsi->req_lock); + ret = nfs_flush_list(inode, &head, pages, how); + spin_lock(&nfsi->req_lock); + continue; + } + if (nocommit) + break; + pages = nfs_scan_commit(inode, &head, 0, 0); + if (pages == 0) + break; + spin_unlock(&nfsi->req_lock); + ret = nfs_commit_list(inode, &head, how); + spin_lock(&nfsi->req_lock); + } while (ret >= 0); + spin_unlock(&nfsi->req_lock); + return ret; } int nfs_init_writepagecache(void) @@ -1407,17 +1521,13 @@ int nfs_init_writepagecache(void) if (nfs_wdata_cachep == NULL) return -ENOMEM; - nfs_wdata_mempool = mempool_create(MIN_POOL_WRITE, - mempool_alloc_slab, - mempool_free_slab, - nfs_wdata_cachep); + nfs_wdata_mempool = mempool_create_slab_pool(MIN_POOL_WRITE, + nfs_wdata_cachep); if (nfs_wdata_mempool == NULL) return -ENOMEM; - nfs_commit_mempool = mempool_create(MIN_POOL_COMMIT, - mempool_alloc_slab, - mempool_free_slab, - nfs_wdata_cachep); + nfs_commit_mempool = mempool_create_slab_pool(MIN_POOL_COMMIT, + nfs_wdata_cachep); if (nfs_commit_mempool == NULL) return -ENOMEM; diff --git a/fs/nfsd/export.c b/fs/nfsd/export.c index 417ec02df44..c340be0a3f5 100644 --- a/fs/nfsd/export.c +++ b/fs/nfsd/export.c @@ -57,27 +57,17 @@ static int exp_verify_string(char *cp, int max); #define EXPKEY_HASHMASK (EXPKEY_HASHMAX -1) static struct cache_head *expkey_table[EXPKEY_HASHMAX]; -static inline int svc_expkey_hash(struct svc_expkey *item) +static void expkey_put(struct kref *ref) { - int hash = item->ek_fsidtype; - char * cp = (char*)item->ek_fsid; - int len = key_len(item->ek_fsidtype); + struct svc_expkey *key = container_of(ref, struct svc_expkey, h.ref); - hash ^= hash_mem(cp, len, EXPKEY_HASHBITS); - hash ^= hash_ptr(item->ek_client, EXPKEY_HASHBITS); - return hash & EXPKEY_HASHMASK; -} - -void expkey_put(struct cache_head *item, struct cache_detail *cd) -{ - if (cache_put(item, cd)) { - struct svc_expkey *key = container_of(item, struct svc_expkey, h); - if (test_bit(CACHE_VALID, &item->flags) && - !test_bit(CACHE_NEGATIVE, &item->flags)) - exp_put(key->ek_export); - auth_domain_put(key->ek_client); - kfree(key); + if (test_bit(CACHE_VALID, &key->h.flags) && + !test_bit(CACHE_NEGATIVE, &key->h.flags)) { + dput(key->ek_dentry); + mntput(key->ek_mnt); } + auth_domain_put(key->ek_client); + kfree(key); } static void expkey_request(struct cache_detail *cd, @@ -95,7 +85,10 @@ static void expkey_request(struct cache_detail *cd, (*bpp)[-1] = '\n'; } -static struct svc_expkey *svc_expkey_lookup(struct svc_expkey *, int); +static struct svc_expkey *svc_expkey_update(struct svc_expkey *new, struct svc_expkey *old); +static struct svc_expkey *svc_expkey_lookup(struct svc_expkey *); +static struct cache_detail svc_expkey_cache; + static int expkey_parse(struct cache_detail *cd, char *mesg, int mlen) { /* client fsidtype fsid [path] */ @@ -106,6 +99,7 @@ static int expkey_parse(struct cache_detail *cd, char *mesg, int mlen) int fsidtype; char *ep; struct svc_expkey key; + struct svc_expkey *ek; if (mesg[mlen-1] != '\n') return -EINVAL; @@ -150,40 +144,38 @@ static int expkey_parse(struct cache_detail *cd, char *mesg, int mlen) key.ek_fsidtype = fsidtype; memcpy(key.ek_fsid, buf, len); + ek = svc_expkey_lookup(&key); + err = -ENOMEM; + if (!ek) + goto out; + /* now we want a pathname, or empty meaning NEGATIVE */ + err = -EINVAL; if ((len=qword_get(&mesg, buf, PAGE_SIZE)) < 0) goto out; dprintk("Path seems to be <%s>\n", buf); err = 0; if (len == 0) { - struct svc_expkey *ek; set_bit(CACHE_NEGATIVE, &key.h.flags); - ek = svc_expkey_lookup(&key, 1); + ek = svc_expkey_update(&key, ek); if (ek) - expkey_put(&ek->h, &svc_expkey_cache); + cache_put(&ek->h, &svc_expkey_cache); + else err = -ENOMEM; } else { struct nameidata nd; - struct svc_expkey *ek; - struct svc_export *exp; err = path_lookup(buf, 0, &nd); if (err) goto out; dprintk("Found the path %s\n", buf); - exp = exp_get_by_name(dom, nd.mnt, nd.dentry, NULL); - - err = -ENOENT; - if (!exp) - goto out_nd; - key.ek_export = exp; - dprintk("And found export\n"); + key.ek_mnt = nd.mnt; + key.ek_dentry = nd.dentry; - ek = svc_expkey_lookup(&key, 1); + ek = svc_expkey_update(&key, ek); if (ek) - expkey_put(&ek->h, &svc_expkey_cache); - exp_put(exp); - err = 0; - out_nd: + cache_put(&ek->h, &svc_expkey_cache); + else + err = -ENOMEM; path_release(&nd); } cache_flush(); @@ -214,35 +206,31 @@ static int expkey_show(struct seq_file *m, if (test_bit(CACHE_VALID, &h->flags) && !test_bit(CACHE_NEGATIVE, &h->flags)) { seq_printf(m, " "); - seq_path(m, ek->ek_export->ex_mnt, ek->ek_export->ex_dentry, "\\ \t\n"); + seq_path(m, ek->ek_mnt, ek->ek_dentry, "\\ \t\n"); } seq_printf(m, "\n"); return 0; } - -struct cache_detail svc_expkey_cache = { - .owner = THIS_MODULE, - .hash_size = EXPKEY_HASHMAX, - .hash_table = expkey_table, - .name = "nfsd.fh", - .cache_put = expkey_put, - .cache_request = expkey_request, - .cache_parse = expkey_parse, - .cache_show = expkey_show, -}; -static inline int svc_expkey_match (struct svc_expkey *a, struct svc_expkey *b) +static inline int expkey_match (struct cache_head *a, struct cache_head *b) { - if (a->ek_fsidtype != b->ek_fsidtype || - a->ek_client != b->ek_client || - memcmp(a->ek_fsid, b->ek_fsid, key_len(a->ek_fsidtype)) != 0) + struct svc_expkey *orig = container_of(a, struct svc_expkey, h); + struct svc_expkey *new = container_of(b, struct svc_expkey, h); + + if (orig->ek_fsidtype != new->ek_fsidtype || + orig->ek_client != new->ek_client || + memcmp(orig->ek_fsid, new->ek_fsid, key_len(orig->ek_fsidtype)) != 0) return 0; return 1; } -static inline void svc_expkey_init(struct svc_expkey *new, struct svc_expkey *item) +static inline void expkey_init(struct cache_head *cnew, + struct cache_head *citem) { - cache_get(&item->ek_client->h); + struct svc_expkey *new = container_of(cnew, struct svc_expkey, h); + struct svc_expkey *item = container_of(citem, struct svc_expkey, h); + + kref_get(&item->ek_client->ref); new->ek_client = item->ek_client; new->ek_fsidtype = item->ek_fsidtype; new->ek_fsid[0] = item->ek_fsid[0]; @@ -250,39 +238,94 @@ static inline void svc_expkey_init(struct svc_expkey *new, struct svc_expkey *it new->ek_fsid[2] = item->ek_fsid[2]; } -static inline void svc_expkey_update(struct svc_expkey *new, struct svc_expkey *item) +static inline void expkey_update(struct cache_head *cnew, + struct cache_head *citem) +{ + struct svc_expkey *new = container_of(cnew, struct svc_expkey, h); + struct svc_expkey *item = container_of(citem, struct svc_expkey, h); + + new->ek_mnt = mntget(item->ek_mnt); + new->ek_dentry = dget(item->ek_dentry); +} + +static struct cache_head *expkey_alloc(void) { - cache_get(&item->ek_export->h); - new->ek_export = item->ek_export; + struct svc_expkey *i = kmalloc(sizeof(*i), GFP_KERNEL); + if (i) + return &i->h; + else + return NULL; } -static DefineSimpleCacheLookup(svc_expkey,0) /* no inplace updates */ +static struct cache_detail svc_expkey_cache = { + .owner = THIS_MODULE, + .hash_size = EXPKEY_HASHMAX, + .hash_table = expkey_table, + .name = "nfsd.fh", + .cache_put = expkey_put, + .cache_request = expkey_request, + .cache_parse = expkey_parse, + .cache_show = expkey_show, + .match = expkey_match, + .init = expkey_init, + .update = expkey_update, + .alloc = expkey_alloc, +}; -#define EXPORT_HASHBITS 8 -#define EXPORT_HASHMAX (1<< EXPORT_HASHBITS) -#define EXPORT_HASHMASK (EXPORT_HASHMAX -1) +static struct svc_expkey * +svc_expkey_lookup(struct svc_expkey *item) +{ + struct cache_head *ch; + int hash = item->ek_fsidtype; + char * cp = (char*)item->ek_fsid; + int len = key_len(item->ek_fsidtype); -static struct cache_head *export_table[EXPORT_HASHMAX]; + hash ^= hash_mem(cp, len, EXPKEY_HASHBITS); + hash ^= hash_ptr(item->ek_client, EXPKEY_HASHBITS); + hash &= EXPKEY_HASHMASK; -static inline int svc_export_hash(struct svc_export *item) + ch = sunrpc_cache_lookup(&svc_expkey_cache, &item->h, + hash); + if (ch) + return container_of(ch, struct svc_expkey, h); + else + return NULL; +} + +static struct svc_expkey * +svc_expkey_update(struct svc_expkey *new, struct svc_expkey *old) { - int rv; + struct cache_head *ch; + int hash = new->ek_fsidtype; + char * cp = (char*)new->ek_fsid; + int len = key_len(new->ek_fsidtype); - rv = hash_ptr(item->ex_client, EXPORT_HASHBITS); - rv ^= hash_ptr(item->ex_dentry, EXPORT_HASHBITS); - rv ^= hash_ptr(item->ex_mnt, EXPORT_HASHBITS); - return rv; + hash ^= hash_mem(cp, len, EXPKEY_HASHBITS); + hash ^= hash_ptr(new->ek_client, EXPKEY_HASHBITS); + hash &= EXPKEY_HASHMASK; + + ch = sunrpc_cache_update(&svc_expkey_cache, &new->h, + &old->h, hash); + if (ch) + return container_of(ch, struct svc_expkey, h); + else + return NULL; } -void svc_export_put(struct cache_head *item, struct cache_detail *cd) + +#define EXPORT_HASHBITS 8 +#define EXPORT_HASHMAX (1<< EXPORT_HASHBITS) +#define EXPORT_HASHMASK (EXPORT_HASHMAX -1) + +static struct cache_head *export_table[EXPORT_HASHMAX]; + +static void svc_export_put(struct kref *ref) { - if (cache_put(item, cd)) { - struct svc_export *exp = container_of(item, struct svc_export, h); - dput(exp->ex_dentry); - mntput(exp->ex_mnt); - auth_domain_put(exp->ex_client); - kfree(exp); - } + struct svc_export *exp = container_of(ref, struct svc_export, h.ref); + dput(exp->ex_dentry); + mntput(exp->ex_mnt); + auth_domain_put(exp->ex_client); + kfree(exp); } static void svc_export_request(struct cache_detail *cd, @@ -304,7 +347,9 @@ static void svc_export_request(struct cache_detail *cd, (*bpp)[-1] = '\n'; } -static struct svc_export *svc_export_lookup(struct svc_export *, int); +static struct svc_export *svc_export_update(struct svc_export *new, + struct svc_export *old); +static struct svc_export *svc_export_lookup(struct svc_export *); static int check_export(struct inode *inode, int flags) { @@ -417,11 +462,16 @@ static int svc_export_parse(struct cache_detail *cd, char *mesg, int mlen) if (err) goto out; } - expp = svc_export_lookup(&exp, 1); + expp = svc_export_lookup(&exp); if (expp) - exp_put(expp); - err = 0; + expp = svc_export_update(&exp, expp); + else + err = -ENOMEM; cache_flush(); + if (expp == NULL) + err = -ENOMEM; + else + exp_put(expp); out: if (nd.dentry) path_release(&nd); @@ -455,6 +505,46 @@ static int svc_export_show(struct seq_file *m, seq_puts(m, ")\n"); return 0; } +static int svc_export_match(struct cache_head *a, struct cache_head *b) +{ + struct svc_export *orig = container_of(a, struct svc_export, h); + struct svc_export *new = container_of(b, struct svc_export, h); + return orig->ex_client == new->ex_client && + orig->ex_dentry == new->ex_dentry && + orig->ex_mnt == new->ex_mnt; +} + +static void svc_export_init(struct cache_head *cnew, struct cache_head *citem) +{ + struct svc_export *new = container_of(cnew, struct svc_export, h); + struct svc_export *item = container_of(citem, struct svc_export, h); + + kref_get(&item->ex_client->ref); + new->ex_client = item->ex_client; + new->ex_dentry = dget(item->ex_dentry); + new->ex_mnt = mntget(item->ex_mnt); +} + +static void export_update(struct cache_head *cnew, struct cache_head *citem) +{ + struct svc_export *new = container_of(cnew, struct svc_export, h); + struct svc_export *item = container_of(citem, struct svc_export, h); + + new->ex_flags = item->ex_flags; + new->ex_anon_uid = item->ex_anon_uid; + new->ex_anon_gid = item->ex_anon_gid; + new->ex_fsid = item->ex_fsid; +} + +static struct cache_head *svc_export_alloc(void) +{ + struct svc_export *i = kmalloc(sizeof(*i), GFP_KERNEL); + if (i) + return &i->h; + else + return NULL; +} + struct cache_detail svc_export_cache = { .owner = THIS_MODULE, .hash_size = EXPORT_HASHMAX, @@ -464,34 +554,49 @@ struct cache_detail svc_export_cache = { .cache_request = svc_export_request, .cache_parse = svc_export_parse, .cache_show = svc_export_show, + .match = svc_export_match, + .init = svc_export_init, + .update = export_update, + .alloc = svc_export_alloc, }; -static inline int svc_export_match(struct svc_export *a, struct svc_export *b) +static struct svc_export * +svc_export_lookup(struct svc_export *exp) { - return a->ex_client == b->ex_client && - a->ex_dentry == b->ex_dentry && - a->ex_mnt == b->ex_mnt; -} -static inline void svc_export_init(struct svc_export *new, struct svc_export *item) -{ - cache_get(&item->ex_client->h); - new->ex_client = item->ex_client; - new->ex_dentry = dget(item->ex_dentry); - new->ex_mnt = mntget(item->ex_mnt); + struct cache_head *ch; + int hash; + hash = hash_ptr(exp->ex_client, EXPORT_HASHBITS); + hash ^= hash_ptr(exp->ex_dentry, EXPORT_HASHBITS); + hash ^= hash_ptr(exp->ex_mnt, EXPORT_HASHBITS); + + ch = sunrpc_cache_lookup(&svc_export_cache, &exp->h, + hash); + if (ch) + return container_of(ch, struct svc_export, h); + else + return NULL; } -static inline void svc_export_update(struct svc_export *new, struct svc_export *item) +static struct svc_export * +svc_export_update(struct svc_export *new, struct svc_export *old) { - new->ex_flags = item->ex_flags; - new->ex_anon_uid = item->ex_anon_uid; - new->ex_anon_gid = item->ex_anon_gid; - new->ex_fsid = item->ex_fsid; + struct cache_head *ch; + int hash; + hash = hash_ptr(old->ex_client, EXPORT_HASHBITS); + hash ^= hash_ptr(old->ex_dentry, EXPORT_HASHBITS); + hash ^= hash_ptr(old->ex_mnt, EXPORT_HASHBITS); + + ch = sunrpc_cache_update(&svc_export_cache, &new->h, + &old->h, + hash); + if (ch) + return container_of(ch, struct svc_export, h); + else + return NULL; } -static DefineSimpleCacheLookup(svc_export,1) /* allow inplace updates */ - -struct svc_expkey * +static struct svc_expkey * exp_find_key(svc_client *clp, int fsid_type, u32 *fsidv, struct cache_req *reqp) { struct svc_expkey key, *ek; @@ -504,7 +609,7 @@ exp_find_key(svc_client *clp, int fsid_type, u32 *fsidv, struct cache_req *reqp) key.ek_fsidtype = fsid_type; memcpy(key.ek_fsid, fsidv, key_len(fsid_type)); - ek = svc_expkey_lookup(&key, 0); + ek = svc_expkey_lookup(&key); if (ek != NULL) if ((err = cache_check(&svc_expkey_cache, &ek->h, reqp))) ek = ERR_PTR(err); @@ -519,13 +624,16 @@ static int exp_set_key(svc_client *clp, int fsid_type, u32 *fsidv, key.ek_client = clp; key.ek_fsidtype = fsid_type; memcpy(key.ek_fsid, fsidv, key_len(fsid_type)); - key.ek_export = exp; + key.ek_mnt = exp->ex_mnt; + key.ek_dentry = exp->ex_dentry; key.h.expiry_time = NEVER; key.h.flags = 0; - ek = svc_expkey_lookup(&key, 1); + ek = svc_expkey_lookup(&key); + if (ek) + ek = svc_expkey_update(&key,ek); if (ek) { - expkey_put(&ek->h, &svc_expkey_cache); + cache_put(&ek->h, &svc_expkey_cache); return 0; } return -ENOMEM; @@ -573,7 +681,7 @@ exp_get_by_name(svc_client *clp, struct vfsmount *mnt, struct dentry *dentry, key.ex_mnt = mnt; key.ex_dentry = dentry; - exp = svc_export_lookup(&key, 0); + exp = svc_export_lookup(&key); if (exp != NULL) switch (cache_check(&svc_export_cache, &exp->h, reqp)) { case 0: break; @@ -654,7 +762,7 @@ static void exp_fsid_unhash(struct svc_export *exp) ek = exp_get_fsid_key(exp->ex_client, exp->ex_fsid); if (ek && !IS_ERR(ek)) { ek->h.expiry_time = get_seconds()-1; - expkey_put(&ek->h, &svc_expkey_cache); + cache_put(&ek->h, &svc_expkey_cache); } svc_expkey_cache.nextcheck = get_seconds(); } @@ -692,7 +800,7 @@ static void exp_unhash(struct svc_export *exp) ek = exp_get_key(exp->ex_client, inode->i_sb->s_dev, inode->i_ino); if (ek && !IS_ERR(ek)) { ek->h.expiry_time = get_seconds()-1; - expkey_put(&ek->h, &svc_expkey_cache); + cache_put(&ek->h, &svc_expkey_cache); } svc_expkey_cache.nextcheck = get_seconds(); } @@ -741,8 +849,8 @@ exp_export(struct nfsctl_export *nxp) if ((nxp->ex_flags & NFSEXP_FSID) && (fsid_key = exp_get_fsid_key(clp, nxp->ex_dev)) && !IS_ERR(fsid_key) && - fsid_key->ek_export && - fsid_key->ek_export != exp) + fsid_key->ek_mnt && + (fsid_key->ek_mnt != nd.mnt || fsid_key->ek_dentry != nd.dentry) ) goto finish; if (exp) { @@ -775,13 +883,13 @@ exp_export(struct nfsctl_export *nxp) new.ex_anon_gid = nxp->ex_anon_gid; new.ex_fsid = nxp->ex_dev; - exp = svc_export_lookup(&new, 1); + exp = svc_export_lookup(&new); + if (exp) + exp = svc_export_update(&new, exp); - if (exp == NULL) + if (!exp) goto finish; - err = 0; - if (exp_hash(clp, exp) || exp_fsid_hash(clp, exp)) { /* failed to create at least one index */ @@ -794,7 +902,7 @@ finish: if (exp) exp_put(exp); if (fsid_key && !IS_ERR(fsid_key)) - expkey_put(&fsid_key->h, &svc_expkey_cache); + cache_put(&fsid_key->h, &svc_expkey_cache); if (clp) auth_domain_put(clp); path_release(&nd); @@ -912,6 +1020,24 @@ out: return err; } +struct svc_export * +exp_find(struct auth_domain *clp, int fsid_type, u32 *fsidv, + struct cache_req *reqp) +{ + struct svc_export *exp; + struct svc_expkey *ek = exp_find_key(clp, fsid_type, fsidv, reqp); + if (!ek || IS_ERR(ek)) + return ERR_PTR(PTR_ERR(ek)); + + exp = exp_get_by_name(clp, ek->ek_mnt, ek->ek_dentry, reqp); + cache_put(&ek->h, &svc_expkey_cache); + + if (!exp || IS_ERR(exp)) + return ERR_PTR(PTR_ERR(exp)); + return exp; +} + + /* * Called when we need the filehandle for the root of the pseudofs, * for a given NFSv4 client. The root is defined to be the @@ -922,6 +1048,7 @@ exp_pseudoroot(struct auth_domain *clp, struct svc_fh *fhp, struct cache_req *creq) { struct svc_expkey *fsid_key; + struct svc_export *exp; int rv; u32 fsidv[2]; @@ -933,9 +1060,15 @@ exp_pseudoroot(struct auth_domain *clp, struct svc_fh *fhp, if (!fsid_key || IS_ERR(fsid_key)) return nfserr_perm; - rv = fh_compose(fhp, fsid_key->ek_export, - fsid_key->ek_export->ex_dentry, NULL); - expkey_put(&fsid_key->h, &svc_expkey_cache); + exp = exp_get_by_name(clp, fsid_key->ek_mnt, fsid_key->ek_dentry, creq); + if (exp == NULL) + rv = nfserr_perm; + else if (IS_ERR(exp)) + rv = nfserrno(PTR_ERR(exp)); + else + rv = fh_compose(fhp, exp, + fsid_key->ek_dentry, NULL); + cache_put(&fsid_key->h, &svc_expkey_cache); return rv; } @@ -1054,7 +1187,7 @@ static int e_show(struct seq_file *m, void *p) cache_get(&exp->h); if (cache_check(&svc_export_cache, &exp->h, NULL)) return 0; - if (cache_put(&exp->h, &svc_export_cache)) BUG(); + cache_put(&exp->h, &svc_export_cache); return svc_export_show(m, &svc_export_cache, cp); } @@ -1129,7 +1262,6 @@ exp_delclient(struct nfsctl_client *ncp) */ if (dom) { err = auth_unix_forget_old(dom); - dom->h.expiry_time = get_seconds(); auth_domain_put(dom); } diff --git a/fs/nfsd/nfs4callback.c b/fs/nfsd/nfs4callback.c index 8d3d23c8a4d..c872bd07fc1 100644 --- a/fs/nfsd/nfs4callback.c +++ b/fs/nfsd/nfs4callback.c @@ -326,6 +326,8 @@ out: .p_encode = (kxdrproc_t) nfs4_xdr_##argtype, \ .p_decode = (kxdrproc_t) nfs4_xdr_##restype, \ .p_bufsiz = MAX(NFS4_##argtype##_sz,NFS4_##restype##_sz) << 2, \ + .p_statidx = NFSPROC4_CB_##call, \ + .p_name = #proc, \ } static struct rpc_procinfo nfs4_cb_procedures[] = { diff --git a/fs/nfsd/nfs4idmap.c b/fs/nfsd/nfs4idmap.c index 13369650cdf..4b6aa60dfce 100644 --- a/fs/nfsd/nfs4idmap.c +++ b/fs/nfsd/nfs4idmap.c @@ -76,21 +76,18 @@ struct ent { char authname[IDMAP_NAMESZ]; }; -#define DefineSimpleCacheLookupMap(STRUCT, FUNC) \ - DefineCacheLookup(struct STRUCT, h, FUNC##_lookup, \ - (struct STRUCT *item, int set), /*no setup */, \ - & FUNC##_cache, FUNC##_hash(item), FUNC##_match(item, tmp), \ - STRUCT##_init(new, item), STRUCT##_update(tmp, item), 0) - /* Common entry handling */ #define ENT_HASHBITS 8 #define ENT_HASHMAX (1 << ENT_HASHBITS) #define ENT_HASHMASK (ENT_HASHMAX - 1) -static inline void -ent_init(struct ent *new, struct ent *itm) +static void +ent_init(struct cache_head *cnew, struct cache_head *citm) { + struct ent *new = container_of(cnew, struct ent, h); + struct ent *itm = container_of(citm, struct ent, h); + new->id = itm->id; new->type = itm->type; @@ -98,19 +95,21 @@ ent_init(struct ent *new, struct ent *itm) strlcpy(new->authname, itm->authname, sizeof(new->name)); } -static inline void -ent_update(struct ent *new, struct ent *itm) +static void +ent_put(struct kref *ref) { - ent_init(new, itm); + struct ent *map = container_of(ref, struct ent, h.ref); + kfree(map); } -static void -ent_put(struct cache_head *ch, struct cache_detail *cd) +static struct cache_head * +ent_alloc(void) { - if (cache_put(ch, cd)) { - struct ent *map = container_of(ch, struct ent, h); - kfree(map); - } + struct ent *e = kmalloc(sizeof(*e), GFP_KERNEL); + if (e) + return &e->h; + else + return NULL; } /* @@ -149,9 +148,12 @@ idtoname_request(struct cache_detail *cd, struct cache_head *ch, char **bpp, (*bpp)[-1] = '\n'; } -static inline int -idtoname_match(struct ent *a, struct ent *b) +static int +idtoname_match(struct cache_head *ca, struct cache_head *cb) { + struct ent *a = container_of(ca, struct ent, h); + struct ent *b = container_of(cb, struct ent, h); + return (a->id == b->id && a->type == b->type && strcmp(a->authname, b->authname) == 0); } @@ -184,7 +186,8 @@ warn_no_idmapd(struct cache_detail *detail) static int idtoname_parse(struct cache_detail *, char *, int); -static struct ent *idtoname_lookup(struct ent *, int); +static struct ent *idtoname_lookup(struct ent *); +static struct ent *idtoname_update(struct ent *, struct ent *); static struct cache_detail idtoname_cache = { .owner = THIS_MODULE, @@ -196,6 +199,10 @@ static struct cache_detail idtoname_cache = { .cache_parse = idtoname_parse, .cache_show = idtoname_show, .warn_no_listener = warn_no_idmapd, + .match = idtoname_match, + .init = ent_init, + .update = ent_init, + .alloc = ent_alloc, }; int @@ -238,6 +245,11 @@ idtoname_parse(struct cache_detail *cd, char *buf, int buflen) if (ent.h.expiry_time == 0) goto out; + error = -ENOMEM; + res = idtoname_lookup(&ent); + if (!res) + goto out; + /* Name */ error = qword_get(&buf, buf1, PAGE_SIZE); if (error == -EINVAL) @@ -252,10 +264,11 @@ idtoname_parse(struct cache_detail *cd, char *buf, int buflen) memcpy(ent.name, buf1, sizeof(ent.name)); } error = -ENOMEM; - if ((res = idtoname_lookup(&ent, 1)) == NULL) + res = idtoname_update(&ent, res); + if (res == NULL) goto out; - ent_put(&res->h, &idtoname_cache); + cache_put(&res->h, &idtoname_cache); error = 0; out: @@ -264,7 +277,31 @@ out: return error; } -static DefineSimpleCacheLookupMap(ent, idtoname); + +static struct ent * +idtoname_lookup(struct ent *item) +{ + struct cache_head *ch = sunrpc_cache_lookup(&idtoname_cache, + &item->h, + idtoname_hash(item)); + if (ch) + return container_of(ch, struct ent, h); + else + return NULL; +} + +static struct ent * +idtoname_update(struct ent *new, struct ent *old) +{ + struct cache_head *ch = sunrpc_cache_update(&idtoname_cache, + &new->h, &old->h, + idtoname_hash(new)); + if (ch) + return container_of(ch, struct ent, h); + else + return NULL; +} + /* * Name -> ID cache @@ -291,9 +328,12 @@ nametoid_request(struct cache_detail *cd, struct cache_head *ch, char **bpp, (*bpp)[-1] = '\n'; } -static inline int -nametoid_match(struct ent *a, struct ent *b) +static int +nametoid_match(struct cache_head *ca, struct cache_head *cb) { + struct ent *a = container_of(ca, struct ent, h); + struct ent *b = container_of(cb, struct ent, h); + return (a->type == b->type && strcmp(a->name, b->name) == 0 && strcmp(a->authname, b->authname) == 0); } @@ -317,7 +357,8 @@ nametoid_show(struct seq_file *m, struct cache_detail *cd, struct cache_head *h) return 0; } -static struct ent *nametoid_lookup(struct ent *, int); +static struct ent *nametoid_lookup(struct ent *); +static struct ent *nametoid_update(struct ent *, struct ent *); static int nametoid_parse(struct cache_detail *, char *, int); static struct cache_detail nametoid_cache = { @@ -330,6 +371,10 @@ static struct cache_detail nametoid_cache = { .cache_parse = nametoid_parse, .cache_show = nametoid_show, .warn_no_listener = warn_no_idmapd, + .match = nametoid_match, + .init = ent_init, + .update = ent_init, + .alloc = ent_alloc, }; static int @@ -379,10 +424,14 @@ nametoid_parse(struct cache_detail *cd, char *buf, int buflen) set_bit(CACHE_NEGATIVE, &ent.h.flags); error = -ENOMEM; - if ((res = nametoid_lookup(&ent, 1)) == NULL) + res = nametoid_lookup(&ent); + if (res == NULL) + goto out; + res = nametoid_update(&ent, res); + if (res == NULL) goto out; - ent_put(&res->h, &nametoid_cache); + cache_put(&res->h, &nametoid_cache); error = 0; out: kfree(buf1); @@ -390,7 +439,30 @@ out: return (error); } -static DefineSimpleCacheLookupMap(ent, nametoid); + +static struct ent * +nametoid_lookup(struct ent *item) +{ + struct cache_head *ch = sunrpc_cache_lookup(&nametoid_cache, + &item->h, + nametoid_hash(item)); + if (ch) + return container_of(ch, struct ent, h); + else + return NULL; +} + +static struct ent * +nametoid_update(struct ent *new, struct ent *old) +{ + struct cache_head *ch = sunrpc_cache_update(&nametoid_cache, + &new->h, &old->h, + nametoid_hash(new)); + if (ch) + return container_of(ch, struct ent, h); + else + return NULL; +} /* * Exported API @@ -458,24 +530,24 @@ idmap_defer(struct cache_req *req) } static inline int -do_idmap_lookup(struct ent *(*lookup_fn)(struct ent *, int), struct ent *key, +do_idmap_lookup(struct ent *(*lookup_fn)(struct ent *), struct ent *key, struct cache_detail *detail, struct ent **item, struct idmap_defer_req *mdr) { - *item = lookup_fn(key, 0); + *item = lookup_fn(key); if (!*item) return -ENOMEM; return cache_check(detail, &(*item)->h, &mdr->req); } static inline int -do_idmap_lookup_nowait(struct ent *(*lookup_fn)(struct ent *, int), +do_idmap_lookup_nowait(struct ent *(*lookup_fn)(struct ent *), struct ent *key, struct cache_detail *detail, struct ent **item) { int ret = -ENOMEM; - *item = lookup_fn(key, 0); + *item = lookup_fn(key); if (!*item) goto out_err; ret = -ETIMEDOUT; @@ -488,7 +560,7 @@ do_idmap_lookup_nowait(struct ent *(*lookup_fn)(struct ent *, int), goto out_put; return 0; out_put: - ent_put(&(*item)->h, detail); + cache_put(&(*item)->h, detail); out_err: *item = NULL; return ret; @@ -496,7 +568,7 @@ out_err: static int idmap_lookup(struct svc_rqst *rqstp, - struct ent *(*lookup_fn)(struct ent *, int), struct ent *key, + struct ent *(*lookup_fn)(struct ent *), struct ent *key, struct cache_detail *detail, struct ent **item) { struct idmap_defer_req *mdr; @@ -539,7 +611,7 @@ idmap_name_to_id(struct svc_rqst *rqstp, int type, const char *name, u32 namelen if (ret) return ret; *id = item->id; - ent_put(&item->h, &nametoid_cache); + cache_put(&item->h, &nametoid_cache); return 0; } @@ -561,7 +633,7 @@ idmap_id_to_name(struct svc_rqst *rqstp, int type, uid_t id, char *name) ret = strlen(item->name); BUG_ON(ret > IDMAP_NAMESZ); memcpy(name, item->name, ret); - ent_put(&item->h, &idtoname_cache); + cache_put(&item->h, &idtoname_cache); return ret; } diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 1143cfb6454..47ec112b266 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -49,6 +49,7 @@ #include <linux/nfsd/state.h> #include <linux/nfsd/xdr4.h> #include <linux/namei.h> +#include <linux/mutex.h> #define NFSDDBG_FACILITY NFSDDBG_PROC @@ -77,11 +78,11 @@ static void nfs4_set_recdir(char *recdir); /* Locking: * - * client_sema: + * client_mutex: * protects clientid_hashtbl[], clientstr_hashtbl[], * unconfstr_hashtbl[], uncofid_hashtbl[]. */ -static DECLARE_MUTEX(client_sema); +static DEFINE_MUTEX(client_mutex); static kmem_cache_t *stateowner_slab = NULL; static kmem_cache_t *file_slab = NULL; @@ -91,13 +92,13 @@ static kmem_cache_t *deleg_slab = NULL; void nfs4_lock_state(void) { - down(&client_sema); + mutex_lock(&client_mutex); } void nfs4_unlock_state(void) { - up(&client_sema); + mutex_unlock(&client_mutex); } static inline u32 @@ -2639,7 +2640,7 @@ nfsd4_lock(struct svc_rqst *rqstp, struct svc_fh *current_fh, struct nfsd4_lock struct nfs4_stateid *lock_stp; struct file *filp; struct file_lock file_lock; - struct file_lock *conflock; + struct file_lock conflock; int status = 0; unsigned int strhashval; @@ -2749,37 +2750,31 @@ nfsd4_lock(struct svc_rqst *rqstp, struct svc_fh *current_fh, struct nfsd4_lock * Note: locks.c uses the BKL to protect the inode's lock list. */ - status = posix_lock_file(filp, &file_lock); - dprintk("NFSD: nfsd4_lock: posix_lock_file status %d\n",status); + /* XXX?: Just to divert the locks_release_private at the start of + * locks_copy_lock: */ + conflock.fl_ops = NULL; + conflock.fl_lmops = NULL; + status = posix_lock_file_conf(filp, &file_lock, &conflock); + dprintk("NFSD: nfsd4_lock: posix_lock_file_conf status %d\n",status); switch (-status) { case 0: /* success! */ update_stateid(&lock_stp->st_stateid); memcpy(&lock->lk_resp_stateid, &lock_stp->st_stateid, sizeof(stateid_t)); - goto out; - case (EAGAIN): - goto conflicting_lock; + break; + case (EAGAIN): /* conflock holds conflicting lock */ + status = nfserr_denied; + dprintk("NFSD: nfsd4_lock: conflicting lock found!\n"); + nfs4_set_lock_denied(&conflock, &lock->lk_denied); + break; case (EDEADLK): status = nfserr_deadlock; - dprintk("NFSD: nfsd4_lock: posix_lock_file() failed! status %d\n",status); - goto out; + break; default: - status = nfserrno(status); - dprintk("NFSD: nfsd4_lock: posix_lock_file() failed! status %d\n",status); - goto out; - } - -conflicting_lock: - dprintk("NFSD: nfsd4_lock: conflicting lock found!\n"); - status = nfserr_denied; - /* XXX There is a race here. Future patch needed to provide - * an atomic posix_lock_and_test_file - */ - if (!(conflock = posix_test_lock(filp, &file_lock))) { - status = nfserr_serverfault; - goto out; + dprintk("NFSD: nfsd4_lock: posix_lock_file_conf() failed! status %d\n",status); + status = nfserr_resource; + break; } - nfs4_set_lock_denied(conflock, &lock->lk_denied); out: if (status && lock->lk_is_new && lock_sop) release_stateowner(lock_sop); @@ -2800,7 +2795,7 @@ nfsd4_lockt(struct svc_rqst *rqstp, struct svc_fh *current_fh, struct nfsd4_lock struct inode *inode; struct file file; struct file_lock file_lock; - struct file_lock *conflicting_lock; + struct file_lock conflock; int status; if (nfs4_in_grace()) @@ -2864,10 +2859,9 @@ nfsd4_lockt(struct svc_rqst *rqstp, struct svc_fh *current_fh, struct nfsd4_lock file.f_dentry = current_fh->fh_dentry; status = nfs_ok; - conflicting_lock = posix_test_lock(&file, &file_lock); - if (conflicting_lock) { + if (posix_test_lock(&file, &file_lock, &conflock)) { status = nfserr_denied; - nfs4_set_lock_denied(conflicting_lock, &lockt->lt_denied); + nfs4_set_lock_denied(&conflock, &lockt->lt_denied); } out: nfs4_unlock_state(); diff --git a/fs/nfsd/nfsctl.c b/fs/nfsd/nfsctl.c index c8960aff096..3ef017b3b5b 100644 --- a/fs/nfsd/nfsctl.c +++ b/fs/nfsd/nfsctl.c @@ -134,7 +134,7 @@ static ssize_t nfsctl_transaction_read(struct file *file, char __user *buf, size return simple_transaction_read(file, buf, size, pos); } -static struct file_operations transaction_ops = { +static const struct file_operations transaction_ops = { .write = nfsctl_transaction_write, .read = nfsctl_transaction_read, .release = simple_transaction_release, @@ -146,7 +146,7 @@ static int exports_open(struct inode *inode, struct file *file) return seq_open(file, &nfs_exports_op); } -static struct file_operations exports_operations = { +static const struct file_operations exports_operations = { .open = exports_open, .read = seq_read, .llseek = seq_lseek, diff --git a/fs/nfsd/nfsfh.c b/fs/nfsd/nfsfh.c index 7a3e397b4ed..3f2ec2e6d06 100644 --- a/fs/nfsd/nfsfh.c +++ b/fs/nfsd/nfsfh.c @@ -506,7 +506,7 @@ fh_put(struct svc_fh *fhp) nfsd_nr_put++; } if (exp) { - svc_export_put(&exp->h, &svc_export_cache); + cache_put(&exp->h, &svc_export_cache); fhp->fh_export = NULL; } return; diff --git a/fs/nfsd/stats.c b/fs/nfsd/stats.c index 1cf955bcc52..57265d56380 100644 --- a/fs/nfsd/stats.c +++ b/fs/nfsd/stats.c @@ -80,7 +80,7 @@ static int nfsd_proc_open(struct inode *inode, struct file *file) return single_open(file, nfsd_proc_show, NULL); } -static struct file_operations nfsd_proc_fops = { +static const struct file_operations nfsd_proc_fops = { .owner = THIS_MODULE, .open = nfsd_proc_open, .read = seq_read, diff --git a/fs/nfsd/vfs.c b/fs/nfsd/vfs.c index 5320e5afadd..31018333dc3 100644 --- a/fs/nfsd/vfs.c +++ b/fs/nfsd/vfs.c @@ -706,7 +706,7 @@ nfsd_close(struct file *filp) * after it. */ static inline int nfsd_dosync(struct file *filp, struct dentry *dp, - struct file_operations *fop) + const struct file_operations *fop) { struct inode *inode = dp->d_inode; int (*fsync) (struct file *, struct dentry *, int); diff --git a/fs/ntfs/dir.c b/fs/ntfs/dir.c index 9d9ed3fe371..d1e2c6f9f05 100644 --- a/fs/ntfs/dir.c +++ b/fs/ntfs/dir.c @@ -1553,7 +1553,7 @@ static int ntfs_dir_fsync(struct file *filp, struct dentry *dentry, #endif /* NTFS_RW */ -struct file_operations ntfs_dir_ops = { +const struct file_operations ntfs_dir_ops = { .llseek = generic_file_llseek, /* Seek inside directory. */ .read = generic_read_dir, /* Return -EISDIR. */ .readdir = ntfs_readdir, /* Read directory contents. */ diff --git a/fs/ntfs/file.c b/fs/ntfs/file.c index f5d057e4acc..c63a83e8da9 100644 --- a/fs/ntfs/file.c +++ b/fs/ntfs/file.c @@ -2294,7 +2294,7 @@ static int ntfs_file_fsync(struct file *filp, struct dentry *dentry, #endif /* NTFS_RW */ -struct file_operations ntfs_file_ops = { +const struct file_operations ntfs_file_ops = { .llseek = generic_file_llseek, /* Seek inside file. */ .read = generic_file_read, /* Read from file. */ .aio_read = generic_file_aio_read, /* Async read from file. */ @@ -2337,6 +2337,6 @@ struct inode_operations ntfs_file_inode_ops = { #endif /* NTFS_RW */ }; -struct file_operations ntfs_empty_file_ops = {}; +const struct file_operations ntfs_empty_file_ops = {}; struct inode_operations ntfs_empty_inode_ops = {}; diff --git a/fs/ntfs/logfile.c b/fs/ntfs/logfile.c index 0fd70295cca..4af2ad1193e 100644 --- a/fs/ntfs/logfile.c +++ b/fs/ntfs/logfile.c @@ -515,10 +515,10 @@ BOOL ntfs_check_logfile(struct inode *log_vi, RESTART_PAGE_HEADER **rp) log_page_size = PAGE_CACHE_SIZE; log_page_mask = log_page_size - 1; /* - * Use generic_ffs() instead of ffs() to enable the compiler to + * Use ntfs_ffs() instead of ffs() to enable the compiler to * optimize log_page_size and log_page_bits into constants. */ - log_page_bits = generic_ffs(log_page_size) - 1; + log_page_bits = ntfs_ffs(log_page_size) - 1; size &= ~(s64)(log_page_size - 1); /* * Ensure the log file is big enough to store at least the two restart diff --git a/fs/ntfs/mft.c b/fs/ntfs/mft.c index 4e72bc7afdf..2438c00ec0c 100644 --- a/fs/ntfs/mft.c +++ b/fs/ntfs/mft.c @@ -2670,7 +2670,7 @@ mft_rec_already_initialized: ni->name_len = 4; ni->itype.index.block_size = 4096; - ni->itype.index.block_size_bits = generic_ffs(4096) - 1; + ni->itype.index.block_size_bits = ntfs_ffs(4096) - 1; ni->itype.index.collation_rule = COLLATION_FILE_NAME; if (vol->cluster_size <= ni->itype.index.block_size) { ni->itype.index.vcn_size = vol->cluster_size; diff --git a/fs/ntfs/ntfs.h b/fs/ntfs/ntfs.h index 0624c8ef4d9..bf7b3d7c093 100644 --- a/fs/ntfs/ntfs.h +++ b/fs/ntfs/ntfs.h @@ -60,13 +60,13 @@ extern struct kmem_cache *ntfs_index_ctx_cache; extern struct address_space_operations ntfs_aops; extern struct address_space_operations ntfs_mst_aops; -extern struct file_operations ntfs_file_ops; +extern const struct file_operations ntfs_file_ops; extern struct inode_operations ntfs_file_inode_ops; -extern struct file_operations ntfs_dir_ops; +extern const struct file_operations ntfs_dir_ops; extern struct inode_operations ntfs_dir_inode_ops; -extern struct file_operations ntfs_empty_file_ops; +extern const struct file_operations ntfs_empty_file_ops; extern struct inode_operations ntfs_empty_inode_ops; extern struct export_operations ntfs_export_ops; @@ -132,4 +132,33 @@ extern int ntfs_ucstonls(const ntfs_volume *vol, const ntfschar *ins, /* From fs/ntfs/upcase.c */ extern ntfschar *generate_default_upcase(void); +static inline int ntfs_ffs(int x) +{ + int r = 1; + + if (!x) + return 0; + if (!(x & 0xffff)) { + x >>= 16; + r += 16; + } + if (!(x & 0xff)) { + x >>= 8; + r += 8; + } + if (!(x & 0xf)) { + x >>= 4; + r += 4; + } + if (!(x & 3)) { + x >>= 2; + r += 2; + } + if (!(x & 1)) { + x >>= 1; + r += 1; + } + return r; +} + #endif /* _LINUX_NTFS_H */ diff --git a/fs/ocfs2/alloc.c b/fs/ocfs2/alloc.c index 6b9812db377..edaab05a93e 100644 --- a/fs/ocfs2/alloc.c +++ b/fs/ocfs2/alloc.c @@ -566,9 +566,8 @@ static int ocfs2_do_insert_extent(struct ocfs2_super *osb, next_free = le16_to_cpu(el->l_next_free_rec); if (next_free == 0) { ocfs2_error(inode->i_sb, - "Dinode %"MLFu64" has a bad " - "extent list", - OCFS2_I(inode)->ip_blkno); + "Dinode %llu has a bad extent list", + (unsigned long long)OCFS2_I(inode)->ip_blkno); status = -EIO; goto bail; } @@ -611,9 +610,8 @@ static int ocfs2_do_insert_extent(struct ocfs2_super *osb, next_free = le16_to_cpu(el->l_next_free_rec); if (next_free == 0) { ocfs2_error(inode->i_sb, - "Dinode %"MLFu64" has a bad " - "extent list", - OCFS2_I(inode)->ip_blkno); + "Dinode %llu has a bad extent list", + (unsigned long long)OCFS2_I(inode)->ip_blkno); status = -EIO; goto bail; } @@ -652,8 +650,9 @@ static int ocfs2_do_insert_extent(struct ocfs2_super *osb, /* having an empty extent at eof is legal. */ if (el->l_recs[i].e_cpos != fe->i_clusters) { ocfs2_error(inode->i_sb, - "Dinode %"MLFu64" trailing extent is bad: " + "Dinode %llu trailing extent is bad: " "cpos (%u) != number of clusters (%u)", + (unsigned long long)OCFS2_I(inode)->ip_blkno, le32_to_cpu(el->l_recs[i].e_cpos), le32_to_cpu(fe->i_clusters)); status = -EIO; @@ -747,19 +746,19 @@ static int ocfs2_find_branch_target(struct ocfs2_super *osb, while(le16_to_cpu(el->l_tree_depth) > 1) { if (le16_to_cpu(el->l_next_free_rec) == 0) { - ocfs2_error(inode->i_sb, "Dinode %"MLFu64" has empty " + ocfs2_error(inode->i_sb, "Dinode %llu has empty " "extent list (next_free_rec == 0)", - OCFS2_I(inode)->ip_blkno); + (unsigned long long)OCFS2_I(inode)->ip_blkno); status = -EIO; goto bail; } i = le16_to_cpu(el->l_next_free_rec) - 1; blkno = le64_to_cpu(el->l_recs[i].e_blkno); if (!blkno) { - ocfs2_error(inode->i_sb, "Dinode %"MLFu64" has extent " + ocfs2_error(inode->i_sb, "Dinode %llu has extent " "list where extent # %d has no physical " "block start", - OCFS2_I(inode)->ip_blkno, i); + (unsigned long long)OCFS2_I(inode)->ip_blkno, i); status = -EIO; goto bail; } @@ -826,9 +825,9 @@ int ocfs2_insert_extent(struct ocfs2_super *osb, mlog_entry_void(); - mlog(0, "add %u clusters starting at block %"MLFu64" to " - "inode %"MLFu64"\n", - new_clusters, start_blk, OCFS2_I(inode)->ip_blkno); + mlog(0, "add %u clusters starting at block %llu to inode %llu\n", + new_clusters, (unsigned long long)start_blk, + (unsigned long long)OCFS2_I(inode)->ip_blkno); fe = (struct ocfs2_dinode *) fe_bh->b_data; el = &fe->id2.i_list; @@ -963,8 +962,8 @@ static int ocfs2_truncate_log_append(struct ocfs2_super *osb, struct ocfs2_dinode *di; struct ocfs2_truncate_log *tl; - mlog_entry("start_blk = %"MLFu64", num_clusters = %u\n", start_blk, - num_clusters); + mlog_entry("start_blk = %llu, num_clusters = %u\n", + (unsigned long long)start_blk, num_clusters); BUG_ON(mutex_trylock(&tl_inode->i_mutex)); @@ -981,8 +980,9 @@ static int ocfs2_truncate_log_append(struct ocfs2_super *osb, tl_count = le16_to_cpu(tl->tl_count); mlog_bug_on_msg(tl_count > ocfs2_truncate_recs_per_inode(osb->sb) || tl_count == 0, - "Truncate record count on #%"MLFu64" invalid (" - "wanted %u, actual %u\n", OCFS2_I(tl_inode)->ip_blkno, + "Truncate record count on #%llu invalid " + "wanted %u, actual %u\n", + (unsigned long long)OCFS2_I(tl_inode)->ip_blkno, ocfs2_truncate_recs_per_inode(osb->sb), le16_to_cpu(tl->tl_count)); @@ -1002,8 +1002,8 @@ static int ocfs2_truncate_log_append(struct ocfs2_super *osb, } mlog(0, "Log truncate of %u clusters starting at cluster %u to " - "%"MLFu64" (index = %d)\n", num_clusters, start_cluster, - OCFS2_I(tl_inode)->ip_blkno, index); + "%llu (index = %d)\n", num_clusters, start_cluster, + (unsigned long long)OCFS2_I(tl_inode)->ip_blkno, index); if (ocfs2_truncate_log_can_coalesce(tl, start_cluster)) { /* @@ -1134,8 +1134,8 @@ static int __ocfs2_flush_truncate_log(struct ocfs2_super *osb) } num_to_flush = le16_to_cpu(tl->tl_used); - mlog(0, "Flush %u records from truncate log #%"MLFu64"\n", - num_to_flush, OCFS2_I(tl_inode)->ip_blkno); + mlog(0, "Flush %u records from truncate log #%llu\n", + num_to_flush, (unsigned long long)OCFS2_I(tl_inode)->ip_blkno); if (!num_to_flush) { status = 0; goto bail; @@ -1360,8 +1360,8 @@ int ocfs2_complete_truncate_log_recovery(struct ocfs2_super *osb, tl = &tl_copy->id2.i_dealloc; num_recs = le16_to_cpu(tl->tl_used); - mlog(0, "cleanup %u records from %"MLFu64"\n", num_recs, - tl_copy->i_blkno); + mlog(0, "cleanup %u records from %llu\n", num_recs, + (unsigned long long)tl_copy->i_blkno); mutex_lock(&tl_inode->i_mutex); for(i = 0; i < num_recs; i++) { @@ -1529,7 +1529,8 @@ static int ocfs2_find_new_last_ext_blk(struct ocfs2_super *osb, *new_last_eb = bh; get_bh(*new_last_eb); - mlog(0, "returning block %"MLFu64"\n", le64_to_cpu(eb->h_blkno)); + mlog(0, "returning block %llu\n", + (unsigned long long)le64_to_cpu(eb->h_blkno)); bail: if (bh) brelse(bh); @@ -1646,8 +1647,8 @@ static int ocfs2_do_truncate(struct ocfs2_super *osb, /* if our tree depth > 0, update all the tree blocks below us. */ while (depth) { - mlog(0, "traveling tree (depth = %d, next_eb = %"MLFu64")\n", - depth, next_eb); + mlog(0, "traveling tree (depth = %d, next_eb = %llu)\n", + depth, (unsigned long long)next_eb); status = ocfs2_read_block(osb, next_eb, &eb_bh, OCFS2_BH_CACHED, inode); if (status < 0) { @@ -1674,12 +1675,12 @@ static int ocfs2_do_truncate(struct ocfs2_super *osb, i = le16_to_cpu(el->l_next_free_rec) - 1; - mlog(0, "extent block %"MLFu64", before: record %d: " - "(%u, %u, %"MLFu64"), next = %u\n", - le64_to_cpu(eb->h_blkno), i, + mlog(0, "extent block %llu, before: record %d: " + "(%u, %u, %llu), next = %u\n", + (unsigned long long)le64_to_cpu(eb->h_blkno), i, le32_to_cpu(el->l_recs[i].e_cpos), le32_to_cpu(el->l_recs[i].e_clusters), - le64_to_cpu(el->l_recs[i].e_blkno), + (unsigned long long)le64_to_cpu(el->l_recs[i].e_blkno), le16_to_cpu(el->l_next_free_rec)); BUG_ON(le32_to_cpu(el->l_recs[i].e_clusters) < clusters_to_del); @@ -1697,12 +1698,12 @@ static int ocfs2_do_truncate(struct ocfs2_super *osb, BUG_ON(!el->l_next_free_rec); le16_add_cpu(&el->l_next_free_rec, -1); } - mlog(0, "extent block %"MLFu64", after: record %d: " - "(%u, %u, %"MLFu64"), next = %u\n", - le64_to_cpu(eb->h_blkno), i, + mlog(0, "extent block %llu, after: record %d: " + "(%u, %u, %llu), next = %u\n", + (unsigned long long)le64_to_cpu(eb->h_blkno), i, le32_to_cpu(el->l_recs[i].e_cpos), le32_to_cpu(el->l_recs[i].e_clusters), - le64_to_cpu(el->l_recs[i].e_blkno), + (unsigned long long)le64_to_cpu(el->l_recs[i].e_blkno), le16_to_cpu(el->l_next_free_rec)); status = ocfs2_journal_dirty(handle, eb_bh); @@ -1792,10 +1793,10 @@ int ocfs2_commit_truncate(struct ocfs2_super *osb, last_eb = le64_to_cpu(fe->i_last_eb_blk); start: mlog(0, "ocfs2_commit_truncate: fe->i_clusters = %u, " - "last_eb = %"MLFu64", fe->i_last_eb_blk = %"MLFu64", " + "last_eb = %llu, fe->i_last_eb_blk = %llu, " "fe->id2.i_list.l_tree_depth = %u last_eb_bh = %p\n", - le32_to_cpu(fe->i_clusters), last_eb, - le64_to_cpu(fe->i_last_eb_blk), + le32_to_cpu(fe->i_clusters), (unsigned long long)last_eb, + (unsigned long long)le64_to_cpu(fe->i_last_eb_blk), le16_to_cpu(fe->id2.i_list.l_tree_depth), last_eb_bh); if (last_eb != le64_to_cpu(fe->i_last_eb_blk)) { @@ -1934,16 +1935,17 @@ int ocfs2_prepare_truncate(struct ocfs2_super *osb, fe = (struct ocfs2_dinode *) fe_bh->b_data; mlog(0, "fe->i_clusters = %u, new_i_clusters = %u, fe->i_size =" - "%"MLFu64"\n", fe->i_clusters, new_i_clusters, fe->i_size); + "%llu\n", fe->i_clusters, new_i_clusters, + (unsigned long long)fe->i_size); if (le32_to_cpu(fe->i_clusters) <= new_i_clusters) { - ocfs2_error(inode->i_sb, "Dinode %"MLFu64" has cluster count " - "%u and size %"MLFu64" whereas struct inode has " + ocfs2_error(inode->i_sb, "Dinode %llu has cluster count " + "%u and size %llu whereas struct inode has " "cluster count %u and size %llu which caused an " "invalid truncate to %u clusters.", - le64_to_cpu(fe->i_blkno), + (unsigned long long)le64_to_cpu(fe->i_blkno), le32_to_cpu(fe->i_clusters), - le64_to_cpu(fe->i_size), + (unsigned long long)le64_to_cpu(fe->i_size), OCFS2_I(inode)->ip_clusters, i_size_read(inode), new_i_clusters); mlog_meta_lvb(ML_ERROR, &OCFS2_I(inode)->ip_meta_lockres); diff --git a/fs/ocfs2/aops.c b/fs/ocfs2/aops.c index 8f4467a930a..0d858d0b25b 100644 --- a/fs/ocfs2/aops.c +++ b/fs/ocfs2/aops.c @@ -74,8 +74,8 @@ static int ocfs2_symlink_get_block(struct inode *inode, sector_t iblock, fe = (struct ocfs2_dinode *) bh->b_data; if (!OCFS2_IS_VALID_DINODE(fe)) { - mlog(ML_ERROR, "Invalid dinode #%"MLFu64": signature = %.*s\n", - fe->i_blkno, 7, fe->i_signature); + mlog(ML_ERROR, "Invalid dinode #%llu: signature = %.*s\n", + (unsigned long long)fe->i_blkno, 7, fe->i_signature); goto bail; } @@ -162,8 +162,8 @@ static int ocfs2_get_block(struct inode *inode, sector_t iblock, NULL); if (err) { mlog(ML_ERROR, "Error %d from get_blocks(0x%p, %llu, 1, " - "%"MLFu64", NULL)\n", err, inode, - (unsigned long long)iblock, p_blkno); + "%llu, NULL)\n", err, inode, (unsigned long long)iblock, + (unsigned long long)p_blkno); goto bail; } @@ -171,13 +171,15 @@ static int ocfs2_get_block(struct inode *inode, sector_t iblock, if (bh_result->b_blocknr == 0) { err = -EIO; - mlog(ML_ERROR, "iblock = %llu p_blkno = %"MLFu64" " - "blkno=(%"MLFu64")\n", (unsigned long long)iblock, - p_blkno, OCFS2_I(inode)->ip_blkno); + mlog(ML_ERROR, "iblock = %llu p_blkno = %llu blkno=(%llu)\n", + (unsigned long long)iblock, + (unsigned long long)p_blkno, + (unsigned long long)OCFS2_I(inode)->ip_blkno); } past_eof = ocfs2_blocks_for_bytes(inode->i_sb, i_size_read(inode)); - mlog(0, "Inode %lu, past_eof = %"MLFu64"\n", inode->i_ino, past_eof); + mlog(0, "Inode %lu, past_eof = %llu\n", inode->i_ino, + (unsigned long long)past_eof); if (create && (iblock >= past_eof)) set_buffer_new(bh_result); @@ -538,7 +540,6 @@ bail: * fs_count, map_bh, dio->rw == WRITE); */ static int ocfs2_direct_IO_get_blocks(struct inode *inode, sector_t iblock, - unsigned long max_blocks, struct buffer_head *bh_result, int create) { int ret; @@ -546,6 +547,7 @@ static int ocfs2_direct_IO_get_blocks(struct inode *inode, sector_t iblock, u64 p_blkno; int contig_blocks; unsigned char blocksize_bits; + unsigned long max_blocks = bh_result->b_size >> inode->i_blkbits; if (!inode || !bh_result) { mlog(ML_ERROR, "inode or bh_result is null\n"); diff --git a/fs/ocfs2/buffer_head_io.c b/fs/ocfs2/buffer_head_io.c index bae3d7548be..9a24adf9be6 100644 --- a/fs/ocfs2/buffer_head_io.c +++ b/fs/ocfs2/buffer_head_io.c @@ -97,8 +97,8 @@ int ocfs2_read_blocks(struct ocfs2_super *osb, u64 block, int nr, int i, ignore_cache = 0; struct buffer_head *bh; - mlog_entry("(block=(%"MLFu64"), nr=(%d), flags=%d, inode=%p)\n", - block, nr, flags, inode); + mlog_entry("(block=(%llu), nr=(%d), flags=%d, inode=%p)\n", + (unsigned long long)block, nr, flags, inode); if (osb == NULL || osb->sb == NULL || bhs == NULL) { status = -EINVAL; @@ -143,9 +143,9 @@ int ocfs2_read_blocks(struct ocfs2_super *osb, u64 block, int nr, if (flags & OCFS2_BH_CACHED && !ocfs2_buffer_uptodate(inode, bh)) { mlog(ML_UPTODATE, - "bh (%llu), inode %"MLFu64" not uptodate\n", + "bh (%llu), inode %llu not uptodate\n", (unsigned long long)bh->b_blocknr, - OCFS2_I(inode)->ip_blkno); + (unsigned long long)OCFS2_I(inode)->ip_blkno); ignore_cache = 1; } @@ -222,7 +222,8 @@ int ocfs2_read_blocks(struct ocfs2_super *osb, u64 block, int nr, if (inode) mutex_unlock(&OCFS2_I(inode)->ip_io_mutex); - mlog(ML_BH_IO, "block=(%"MLFu64"), nr=(%d), cached=%s\n", block, nr, + mlog(ML_BH_IO, "block=(%llu), nr=(%d), cached=%s\n", + (unsigned long long)block, nr, (!(flags & OCFS2_BH_CACHED) || ignore_cache) ? "no" : "yes"); bail: diff --git a/fs/ocfs2/cluster/heartbeat.c b/fs/ocfs2/cluster/heartbeat.c index d08971d29b6..bff0f0d0686 100644 --- a/fs/ocfs2/cluster/heartbeat.c +++ b/fs/ocfs2/cluster/heartbeat.c @@ -449,11 +449,11 @@ static u32 o2hb_compute_block_crc_le(struct o2hb_region *reg, static void o2hb_dump_slot(struct o2hb_disk_heartbeat_block *hb_block) { - mlog(ML_ERROR, "Dump slot information: seq = 0x%"MLFx64", node = %u, " - "cksum = 0x%x, generation 0x%"MLFx64"\n", - le64_to_cpu(hb_block->hb_seq), hb_block->hb_node, - le32_to_cpu(hb_block->hb_cksum), - le64_to_cpu(hb_block->hb_generation)); + mlog(ML_ERROR, "Dump slot information: seq = 0x%llx, node = %u, " + "cksum = 0x%x, generation 0x%llx\n", + (long long)le64_to_cpu(hb_block->hb_seq), + hb_block->hb_node, le32_to_cpu(hb_block->hb_cksum), + (long long)le64_to_cpu(hb_block->hb_generation)); } static int o2hb_verify_crc(struct o2hb_region *reg, @@ -516,8 +516,9 @@ static inline void o2hb_prepare_block(struct o2hb_region *reg, hb_block->hb_cksum = cpu_to_le32(o2hb_compute_block_crc_le(reg, hb_block)); - mlog(ML_HB_BIO, "our node generation = 0x%"MLFx64", cksum = 0x%x\n", - cpu_to_le64(generation), le32_to_cpu(hb_block->hb_cksum)); + mlog(ML_HB_BIO, "our node generation = 0x%llx, cksum = 0x%x\n", + (long long)cpu_to_le64(generation), + le32_to_cpu(hb_block->hb_cksum)); } static void o2hb_fire_callbacks(struct o2hb_callback *hbcall, @@ -686,19 +687,20 @@ static int o2hb_check_slot(struct o2hb_region *reg, if (slot->ds_last_generation != le64_to_cpu(hb_block->hb_generation)) { gen_changed = 1; slot->ds_equal_samples = 0; - mlog(ML_HEARTBEAT, "Node %d changed generation (0x%"MLFx64" " - "to 0x%"MLFx64")\n", slot->ds_node_num, - slot->ds_last_generation, - le64_to_cpu(hb_block->hb_generation)); + mlog(ML_HEARTBEAT, "Node %d changed generation (0x%llx " + "to 0x%llx)\n", slot->ds_node_num, + (long long)slot->ds_last_generation, + (long long)le64_to_cpu(hb_block->hb_generation)); } slot->ds_last_generation = le64_to_cpu(hb_block->hb_generation); - mlog(ML_HEARTBEAT, "Slot %d gen 0x%"MLFx64" cksum 0x%x " - "seq %"MLFu64" last %"MLFu64" changed %u equal %u\n", - slot->ds_node_num, slot->ds_last_generation, - le32_to_cpu(hb_block->hb_cksum), le64_to_cpu(hb_block->hb_seq), - slot->ds_last_time, slot->ds_changed_samples, + mlog(ML_HEARTBEAT, "Slot %d gen 0x%llx cksum 0x%x " + "seq %llu last %llu changed %u equal %u\n", + slot->ds_node_num, (long long)slot->ds_last_generation, + le32_to_cpu(hb_block->hb_cksum), + (unsigned long long)le64_to_cpu(hb_block->hb_seq), + (unsigned long long)slot->ds_last_time, slot->ds_changed_samples, slot->ds_equal_samples); spin_lock(&o2hb_live_lock); @@ -708,8 +710,8 @@ fire_callbacks: * changes at any time during their dead time */ if (list_empty(&slot->ds_live_item) && slot->ds_changed_samples >= O2HB_LIVE_THRESHOLD) { - mlog(ML_HEARTBEAT, "Node %d (id 0x%"MLFx64") joined my " - "region\n", slot->ds_node_num, slot->ds_last_generation); + mlog(ML_HEARTBEAT, "Node %d (id 0x%llx) joined my region\n", + slot->ds_node_num, (long long)slot->ds_last_generation); /* first on the list generates a callback */ if (list_empty(&o2hb_live_slots[slot->ds_node_num])) { diff --git a/fs/ocfs2/cluster/masklog.h b/fs/ocfs2/cluster/masklog.h index 2cadc3009c8..73edad78253 100644 --- a/fs/ocfs2/cluster/masklog.h +++ b/fs/ocfs2/cluster/masklog.h @@ -256,16 +256,6 @@ extern struct mlog_bits mlog_and_bits, mlog_not_bits; } \ } while (0) -#if (BITS_PER_LONG == 32) || defined(CONFIG_X86_64) || (defined(CONFIG_UML_X86) && defined(CONFIG_64BIT)) -#define MLFi64 "lld" -#define MLFu64 "llu" -#define MLFx64 "llx" -#else -#define MLFi64 "ld" -#define MLFu64 "lu" -#define MLFx64 "lx" -#endif - #include <linux/kobject.h> #include <linux/sysfs.h> int mlog_sys_init(struct subsystem *o2cb_subsys); diff --git a/fs/ocfs2/dcache.c b/fs/ocfs2/dcache.c index bd85182e97b..1a01380e387 100644 --- a/fs/ocfs2/dcache.c +++ b/fs/ocfs2/dcache.c @@ -64,15 +64,16 @@ static int ocfs2_dentry_revalidate(struct dentry *dentry, /* did we or someone else delete this inode? */ if (OCFS2_I(inode)->ip_flags & OCFS2_INODE_DELETED) { spin_unlock(&OCFS2_I(inode)->ip_lock); - mlog(0, "inode (%"MLFu64") deleted, returning false\n", - OCFS2_I(inode)->ip_blkno); + mlog(0, "inode (%llu) deleted, returning false\n", + (unsigned long long)OCFS2_I(inode)->ip_blkno); goto bail; } spin_unlock(&OCFS2_I(inode)->ip_lock); if (!inode->i_nlink) { - mlog(0, "Inode %"MLFu64" orphaned, returning false " - "dir = %d\n", OCFS2_I(inode)->ip_blkno, + mlog(0, "Inode %llu orphaned, returning false " + "dir = %d\n", + (unsigned long long)OCFS2_I(inode)->ip_blkno, S_ISDIR(inode->i_mode)); goto bail; } diff --git a/fs/ocfs2/dir.c b/fs/ocfs2/dir.c index 57158fa75d9..ae47f450792 100644 --- a/fs/ocfs2/dir.c +++ b/fs/ocfs2/dir.c @@ -83,7 +83,8 @@ int ocfs2_readdir(struct file * filp, void * dirent, filldir_t filldir) struct super_block * sb = inode->i_sb; int have_disk_lock = 0; - mlog_entry("dirino=%"MLFu64"\n", OCFS2_I(inode)->ip_blkno); + mlog_entry("dirino=%llu\n", + (unsigned long long)OCFS2_I(inode)->ip_blkno); stored = 0; bh = NULL; @@ -104,9 +105,9 @@ int ocfs2_readdir(struct file * filp, void * dirent, filldir_t filldir) blk = (filp->f_pos) >> sb->s_blocksize_bits; bh = ocfs2_bread(inode, blk, &err, 0); if (!bh) { - mlog(ML_ERROR, "directory #%"MLFu64" contains a hole " - "at offset %lld\n", - OCFS2_I(inode)->ip_blkno, + mlog(ML_ERROR, + "directory #%llu contains a hole at offset %lld\n", + (unsigned long long)OCFS2_I(inode)->ip_blkno, filp->f_pos); filp->f_pos += sb->s_blocksize - offset; continue; @@ -214,9 +215,9 @@ int ocfs2_find_files_on_disk(const char *name, int status = -ENOENT; struct ocfs2_super *osb = OCFS2_SB(inode->i_sb); - mlog_entry("(osb=%p, parent=%"MLFu64", name='%.*s', blkno=%p, " - "inode=%p)\n", - osb, OCFS2_I(inode)->ip_blkno, namelen, name, blkno, inode); + mlog_entry("(osb=%p, parent=%llu, name='%.*s', blkno=%p, inode=%p)\n", + osb, (unsigned long long)OCFS2_I(inode)->ip_blkno, + namelen, name, blkno, inode); *dirent_bh = ocfs2_find_entry(name, namelen, inode, dirent); if (!*dirent_bh || !*dirent) { @@ -255,8 +256,8 @@ int ocfs2_check_dir_for_entry(struct inode *dir, struct buffer_head *dirent_bh = NULL; struct ocfs2_dir_entry *dirent = NULL; - mlog_entry("dir %"MLFu64", name '%.*s'\n", OCFS2_I(dir)->ip_blkno, - namelen, name); + mlog_entry("dir %llu, name '%.*s'\n", + (unsigned long long)OCFS2_I(dir)->ip_blkno, namelen, name); ret = -EEXIST; dirent_bh = ocfs2_find_entry(name, namelen, dir, &dirent); @@ -287,9 +288,8 @@ int ocfs2_empty_dir(struct inode *inode) if ((i_size_read(inode) < (OCFS2_DIR_REC_LEN(1) + OCFS2_DIR_REC_LEN(2))) || !(bh = ocfs2_bread(inode, 0, &err, 0))) { - mlog(ML_ERROR, "bad directory (dir #%"MLFu64") - " - "no data block\n", - OCFS2_I(inode)->ip_blkno); + mlog(ML_ERROR, "bad directory (dir #%llu) - no data block\n", + (unsigned long long)OCFS2_I(inode)->ip_blkno); return 1; } @@ -300,9 +300,8 @@ int ocfs2_empty_dir(struct inode *inode) !le64_to_cpu(de1->inode) || strcmp(".", de->name) || strcmp("..", de1->name)) { - mlog(ML_ERROR, "bad directory (dir #%"MLFu64") - " - "no `.' or `..'\n", - OCFS2_I(inode)->ip_blkno); + mlog(ML_ERROR, "bad directory (dir #%llu) - no `.' or `..'\n", + (unsigned long long)OCFS2_I(inode)->ip_blkno); brelse(bh); return 1; } @@ -314,9 +313,8 @@ int ocfs2_empty_dir(struct inode *inode) bh = ocfs2_bread(inode, offset >> sb->s_blocksize_bits, &err, 0); if (!bh) { - mlog(ML_ERROR, "directory #%"MLFu64" contains " - "a hole at offset %lu\n", - OCFS2_I(inode)->ip_blkno, offset); + mlog(ML_ERROR, "dir %llu has a hole at %lu\n", + (unsigned long long)OCFS2_I(inode)->ip_blkno, offset); offset += sb->s_blocksize; continue; } @@ -406,8 +404,8 @@ static int ocfs2_extend_dir(struct ocfs2_super *osb, mlog_entry_void(); dir_i_size = i_size_read(dir); - mlog(0, "extending dir %"MLFu64" (i_size = %lld)\n", - OCFS2_I(dir)->ip_blkno, dir_i_size); + mlog(0, "extending dir %llu (i_size = %lld)\n", + (unsigned long long)OCFS2_I(dir)->ip_blkno, dir_i_size); handle = ocfs2_alloc_handle(osb); if (handle == NULL) { @@ -531,8 +529,8 @@ int ocfs2_prepare_dir_for_insert(struct ocfs2_super *osb, mlog_entry_void(); - mlog(0, "getting ready to insert namelen %d into dir %"MLFu64"\n", - namelen, OCFS2_I(dir)->ip_blkno); + mlog(0, "getting ready to insert namelen %d into dir %llu\n", + namelen, (unsigned long long)OCFS2_I(dir)->ip_blkno); BUG_ON(!S_ISDIR(dir->i_mode)); fe = (struct ocfs2_dinode *) parent_fe_bh->b_data; diff --git a/fs/ocfs2/dlm/dlmast.c b/fs/ocfs2/dlm/dlmast.c index 8d17d28ef91..355593dd8ef 100644 --- a/fs/ocfs2/dlm/dlmast.c +++ b/fs/ocfs2/dlm/dlmast.c @@ -307,8 +307,11 @@ int dlm_proxy_ast_handler(struct o2net_msg *msg, u32 len, void *data) if (past->type != DLM_AST && past->type != DLM_BAST) { - mlog(ML_ERROR, "Unknown ast type! %d, cookie=%"MLFu64", " - "name=%.*s\n", past->type, cookie, locklen, name); + mlog(ML_ERROR, "Unknown ast type! %d, cookie=%u:%llu" + "name=%.*s\n", past->type, + dlm_get_lock_cookie_node(cookie), + dlm_get_lock_cookie_seq(cookie), + locklen, name); ret = DLM_IVLOCKID; goto leave; } @@ -316,9 +319,11 @@ int dlm_proxy_ast_handler(struct o2net_msg *msg, u32 len, void *data) res = dlm_lookup_lockres(dlm, name, locklen); if (!res) { mlog(ML_ERROR, "got %sast for unknown lockres! " - "cookie=%"MLFu64", name=%.*s, namelen=%u\n", + "cookie=%u:%llu, name=%.*s, namelen=%u\n", past->type == DLM_AST ? "" : "b", - cookie, locklen, name, locklen); + dlm_get_lock_cookie_node(cookie), + dlm_get_lock_cookie_seq(cookie), + locklen, name, locklen); ret = DLM_IVLOCKID; goto leave; } @@ -360,9 +365,12 @@ int dlm_proxy_ast_handler(struct o2net_msg *msg, u32 len, void *data) goto do_ast; } - mlog(ML_ERROR, "got %sast for unknown lock! cookie=%"MLFu64", " - "name=%.*s, namelen=%u\n", - past->type == DLM_AST ? "" : "b", cookie, locklen, name, locklen); + mlog(ML_ERROR, "got %sast for unknown lock! cookie=%u:%llu, " + "name=%.*s, namelen=%u\n", + past->type == DLM_AST ? "" : "b", + dlm_get_lock_cookie_node(cookie), + dlm_get_lock_cookie_seq(cookie), + locklen, name, locklen); ret = DLM_NORMAL; unlock_out: diff --git a/fs/ocfs2/dlm/dlmcommon.h b/fs/ocfs2/dlm/dlmcommon.h index 9c772583744..88cc43df18f 100644 --- a/fs/ocfs2/dlm/dlmcommon.h +++ b/fs/ocfs2/dlm/dlmcommon.h @@ -630,6 +630,21 @@ __dlm_lockres_state_to_status(struct dlm_lock_resource *res) return status; } +static inline u8 dlm_get_lock_cookie_node(u64 cookie) +{ + u8 ret; + cookie >>= 56; + ret = (u8)(cookie & 0xffULL); + return ret; +} + +static inline unsigned long long dlm_get_lock_cookie_seq(u64 cookie) +{ + unsigned long long ret; + ret = ((unsigned long long)cookie) & 0x00ffffffffffffffULL; + return ret; +} + struct dlm_lock * dlm_new_lock(int type, u8 node, u64 cookie, struct dlm_lockstatus *lksb); void dlm_lock_get(struct dlm_lock *lock); @@ -658,6 +673,7 @@ void dlm_complete_thread(struct dlm_ctxt *dlm); int dlm_launch_recovery_thread(struct dlm_ctxt *dlm); void dlm_complete_recovery_thread(struct dlm_ctxt *dlm); void dlm_wait_for_recovery(struct dlm_ctxt *dlm); +void dlm_kick_recovery_thread(struct dlm_ctxt *dlm); int dlm_is_node_dead(struct dlm_ctxt *dlm, u8 node); int dlm_wait_for_node_death(struct dlm_ctxt *dlm, u8 node, int timeout); @@ -762,6 +778,11 @@ int dlm_request_all_locks_handler(struct o2net_msg *msg, u32 len, void *data); int dlm_reco_data_done_handler(struct o2net_msg *msg, u32 len, void *data); int dlm_begin_reco_handler(struct o2net_msg *msg, u32 len, void *data); int dlm_finalize_reco_handler(struct o2net_msg *msg, u32 len, void *data); +int dlm_do_master_requery(struct dlm_ctxt *dlm, struct dlm_lock_resource *res, + u8 nodenum, u8 *real_master); +int dlm_lockres_master_requery(struct dlm_ctxt *dlm, + struct dlm_lock_resource *res, u8 *real_master); + int dlm_dispatch_assert_master(struct dlm_ctxt *dlm, struct dlm_lock_resource *res, diff --git a/fs/ocfs2/dlm/dlmconvert.c b/fs/ocfs2/dlm/dlmconvert.c index f66e2d818cc..8285228d9e3 100644 --- a/fs/ocfs2/dlm/dlmconvert.c +++ b/fs/ocfs2/dlm/dlmconvert.c @@ -284,8 +284,10 @@ enum dlm_status dlmconvert_remote(struct dlm_ctxt *dlm, if (lock->ml.convert_type != LKM_IVMODE) { __dlm_print_one_lock_resource(res); mlog(ML_ERROR, "converting a remote lock that is already " - "converting! (cookie=%"MLFu64", conv=%d)\n", - lock->ml.cookie, lock->ml.convert_type); + "converting! (cookie=%u:%llu, conv=%d)\n", + dlm_get_lock_cookie_node(lock->ml.cookie), + dlm_get_lock_cookie_seq(lock->ml.cookie), + lock->ml.convert_type); status = DLM_DENIED; goto bail; } @@ -513,8 +515,9 @@ int dlm_convert_lock_handler(struct o2net_msg *msg, u32 len, void *data) leave: if (!lock) mlog(ML_ERROR, "did not find lock to convert on grant queue! " - "cookie=%"MLFu64"\n", - cnv->cookie); + "cookie=%u:%llu\n", + dlm_get_lock_cookie_node(cnv->cookie), + dlm_get_lock_cookie_seq(cnv->cookie)); else dlm_lock_put(lock); diff --git a/fs/ocfs2/dlm/dlmdebug.c b/fs/ocfs2/dlm/dlmdebug.c index 54f61b76ab5..c7eae5d3324 100644 --- a/fs/ocfs2/dlm/dlmdebug.c +++ b/fs/ocfs2/dlm/dlmdebug.c @@ -72,8 +72,10 @@ void __dlm_print_one_lock_resource(struct dlm_lock_resource *res) lock = list_entry(iter2, struct dlm_lock, list); spin_lock(&lock->spinlock); mlog(ML_NOTICE, " type=%d, conv=%d, node=%u, " - "cookie=%"MLFu64", ast=(empty=%c,pend=%c), bast=(empty=%c,pend=%c)\n", - lock->ml.type, lock->ml.convert_type, lock->ml.node, lock->ml.cookie, + "cookie=%u:%llu, ast=(empty=%c,pend=%c), bast=(empty=%c,pend=%c)\n", + lock->ml.type, lock->ml.convert_type, lock->ml.node, + dlm_get_lock_cookie_node(lock->ml.cookie), + dlm_get_lock_cookie_seq(lock->ml.cookie), list_empty(&lock->ast_list) ? 'y' : 'n', lock->ast_pending ? 'y' : 'n', list_empty(&lock->bast_list) ? 'y' : 'n', @@ -85,8 +87,10 @@ void __dlm_print_one_lock_resource(struct dlm_lock_resource *res) lock = list_entry(iter2, struct dlm_lock, list); spin_lock(&lock->spinlock); mlog(ML_NOTICE, " type=%d, conv=%d, node=%u, " - "cookie=%"MLFu64", ast=(empty=%c,pend=%c), bast=(empty=%c,pend=%c)\n", - lock->ml.type, lock->ml.convert_type, lock->ml.node, lock->ml.cookie, + "cookie=%u:%llu, ast=(empty=%c,pend=%c), bast=(empty=%c,pend=%c)\n", + lock->ml.type, lock->ml.convert_type, lock->ml.node, + dlm_get_lock_cookie_node(lock->ml.cookie), + dlm_get_lock_cookie_seq(lock->ml.cookie), list_empty(&lock->ast_list) ? 'y' : 'n', lock->ast_pending ? 'y' : 'n', list_empty(&lock->bast_list) ? 'y' : 'n', @@ -98,8 +102,10 @@ void __dlm_print_one_lock_resource(struct dlm_lock_resource *res) lock = list_entry(iter2, struct dlm_lock, list); spin_lock(&lock->spinlock); mlog(ML_NOTICE, " type=%d, conv=%d, node=%u, " - "cookie=%"MLFu64", ast=(empty=%c,pend=%c), bast=(empty=%c,pend=%c)\n", - lock->ml.type, lock->ml.convert_type, lock->ml.node, lock->ml.cookie, + "cookie=%u:%llu, ast=(empty=%c,pend=%c), bast=(empty=%c,pend=%c)\n", + lock->ml.type, lock->ml.convert_type, lock->ml.node, + dlm_get_lock_cookie_node(lock->ml.cookie), + dlm_get_lock_cookie_seq(lock->ml.cookie), list_empty(&lock->ast_list) ? 'y' : 'n', lock->ast_pending ? 'y' : 'n', list_empty(&lock->bast_list) ? 'y' : 'n', diff --git a/fs/ocfs2/dlm/dlmlock.c b/fs/ocfs2/dlm/dlmlock.c index 671d4ff222c..6fea28318d6 100644 --- a/fs/ocfs2/dlm/dlmlock.c +++ b/fs/ocfs2/dlm/dlmlock.c @@ -141,13 +141,23 @@ static enum dlm_status dlmlock_master(struct dlm_ctxt *dlm, res->lockname.len)) { kick_thread = 1; call_ast = 1; + } else { + mlog(0, "%s: returning DLM_NORMAL to " + "node %u for reco lock\n", dlm->name, + lock->ml.node); } } else { /* for NOQUEUE request, unless we get the * lock right away, return DLM_NOTQUEUED */ - if (flags & LKM_NOQUEUE) + if (flags & LKM_NOQUEUE) { status = DLM_NOTQUEUED; - else { + if (dlm_is_recovery_lock(res->lockname.name, + res->lockname.len)) { + mlog(0, "%s: returning NOTQUEUED to " + "node %u for reco lock\n", dlm->name, + lock->ml.node); + } + } else { dlm_lock_get(lock); list_add_tail(&lock->list, &res->blocked); kick_thread = 1; diff --git a/fs/ocfs2/dlm/dlmmaster.c b/fs/ocfs2/dlm/dlmmaster.c index 847dd3cc4cf..940be4c13b1 100644 --- a/fs/ocfs2/dlm/dlmmaster.c +++ b/fs/ocfs2/dlm/dlmmaster.c @@ -239,6 +239,8 @@ static void dlm_remove_nonlocal_locks(struct dlm_ctxt *dlm, static int dlm_mark_lockres_migrating(struct dlm_ctxt *dlm, struct dlm_lock_resource *res, u8 target); +static int dlm_pre_master_reco_lockres(struct dlm_ctxt *dlm, + struct dlm_lock_resource *res); int dlm_is_host_down(int errno) @@ -677,6 +679,7 @@ struct dlm_lock_resource * dlm_get_lock_resource(struct dlm_ctxt *dlm, struct dlm_node_iter iter; unsigned int namelen; int tries = 0; + int bit, wait_on_recovery = 0; BUG_ON(!lockid); @@ -762,6 +765,18 @@ lookup: dlm_init_mle(mle, DLM_MLE_MASTER, dlm, res, NULL, 0); set_bit(dlm->node_num, mle->maybe_map); list_add(&mle->list, &dlm->master_list); + + /* still holding the dlm spinlock, check the recovery map + * to see if there are any nodes that still need to be + * considered. these will not appear in the mle nodemap + * but they might own this lockres. wait on them. */ + bit = find_next_bit(dlm->recovery_map, O2NM_MAX_NODES, 0); + if (bit < O2NM_MAX_NODES) { + mlog(ML_NOTICE, "%s:%.*s: at least one node (%d) to" + "recover before lock mastery can begin\n", + dlm->name, namelen, (char *)lockid, bit); + wait_on_recovery = 1; + } } /* at this point there is either a DLM_MLE_BLOCK or a @@ -779,6 +794,39 @@ lookup: spin_unlock(&dlm->master_lock); spin_unlock(&dlm->spinlock); + while (wait_on_recovery) { + /* any cluster changes that occurred after dropping the + * dlm spinlock would be detectable be a change on the mle, + * so we only need to clear out the recovery map once. */ + if (dlm_is_recovery_lock(lockid, namelen)) { + mlog(ML_NOTICE, "%s: recovery map is not empty, but " + "must master $RECOVERY lock now\n", dlm->name); + if (!dlm_pre_master_reco_lockres(dlm, res)) + wait_on_recovery = 0; + else { + mlog(0, "%s: waiting 500ms for heartbeat state " + "change\n", dlm->name); + msleep(500); + } + continue; + } + + dlm_kick_recovery_thread(dlm); + msleep(100); + dlm_wait_for_recovery(dlm); + + spin_lock(&dlm->spinlock); + bit = find_next_bit(dlm->recovery_map, O2NM_MAX_NODES, 0); + if (bit < O2NM_MAX_NODES) { + mlog(ML_NOTICE, "%s:%.*s: at least one node (%d) to" + "recover before lock mastery can begin\n", + dlm->name, namelen, (char *)lockid, bit); + wait_on_recovery = 1; + } else + wait_on_recovery = 0; + spin_unlock(&dlm->spinlock); + } + /* must wait for lock to be mastered elsewhere */ if (blocked) goto wait; @@ -792,7 +840,15 @@ redo_request: mlog_errno(ret); if (mle->master != O2NM_MAX_NODES) { /* found a master ! */ - break; + if (mle->master <= nodenum) + break; + /* if our master request has not reached the master + * yet, keep going until it does. this is how the + * master will know that asserts are needed back to + * the lower nodes. */ + mlog(0, "%s:%.*s: requests only up to %u but master " + "is %u, keep going\n", dlm->name, namelen, + lockid, nodenum, mle->master); } } @@ -860,7 +916,19 @@ recheck: /* check if another node has already become the owner */ spin_lock(&res->spinlock); if (res->owner != DLM_LOCK_RES_OWNER_UNKNOWN) { + mlog(0, "%s:%.*s: owner is suddenly %u\n", dlm->name, + res->lockname.len, res->lockname.name, res->owner); spin_unlock(&res->spinlock); + /* this will cause the master to re-assert across + * the whole cluster, freeing up mles */ + ret = dlm_do_master_request(mle, res->owner); + if (ret < 0) { + /* give recovery a chance to run */ + mlog(ML_ERROR, "link to %u went down?: %d\n", res->owner, ret); + msleep(500); + goto recheck; + } + ret = 0; goto leave; } spin_unlock(&res->spinlock); @@ -1244,13 +1312,14 @@ int dlm_master_request_handler(struct o2net_msg *msg, u32 len, void *data) { u8 response = DLM_MASTER_RESP_MAYBE; struct dlm_ctxt *dlm = data; - struct dlm_lock_resource *res; + struct dlm_lock_resource *res = NULL; struct dlm_master_request *request = (struct dlm_master_request *) msg->buf; struct dlm_master_list_entry *mle = NULL, *tmpmle = NULL; char *name; unsigned int namelen; int found, ret; int set_maybe; + int dispatch_assert = 0; if (!dlm_grab(dlm)) return DLM_MASTER_RESP_NO; @@ -1287,7 +1356,6 @@ way_up_top: } if (res->owner == dlm->node_num) { - u32 flags = DLM_ASSERT_MASTER_MLE_CLEANUP; spin_unlock(&res->spinlock); // mlog(0, "this node is the master\n"); response = DLM_MASTER_RESP_YES; @@ -1300,16 +1368,7 @@ way_up_top: * caused all nodes up to this one to * create mles. this node now needs to * go back and clean those up. */ - mlog(0, "%u is the owner of %.*s, cleaning everyone else\n", - dlm->node_num, res->lockname.len, res->lockname.name); - ret = dlm_dispatch_assert_master(dlm, res, 1, - request->node_idx, - flags); - if (ret < 0) { - mlog(ML_ERROR, "failed to dispatch assert " - "master work\n"); - response = DLM_MASTER_RESP_ERROR; - } + dispatch_assert = 1; goto send_response; } else if (res->owner != DLM_LOCK_RES_OWNER_UNKNOWN) { spin_unlock(&res->spinlock); @@ -1357,9 +1416,13 @@ way_up_top: } } else if (tmpmle->master != DLM_LOCK_RES_OWNER_UNKNOWN) { set_maybe = 0; - if (tmpmle->master == dlm->node_num) + if (tmpmle->master == dlm->node_num) { response = DLM_MASTER_RESP_YES; - else + /* this node will be the owner. + * go back and clean the mles on any + * other nodes */ + dispatch_assert = 1; + } else response = DLM_MASTER_RESP_NO; } else { // mlog(0, "this node is attempting to " @@ -1398,8 +1461,8 @@ way_up_top: mle = (struct dlm_master_list_entry *) kmem_cache_alloc(dlm_mle_cache, GFP_KERNEL); if (!mle) { - // bad bad bad... this sucks. response = DLM_MASTER_RESP_ERROR; + mlog_errno(-ENOMEM); goto send_response; } spin_lock(&dlm->spinlock); @@ -1418,25 +1481,19 @@ way_up_top: // mlog(0, "mle was found\n"); set_maybe = 1; spin_lock(&tmpmle->spinlock); + if (tmpmle->master == dlm->node_num) { + mlog(ML_ERROR, "no lockres, but an mle with this node as master!\n"); + BUG(); + } if (tmpmle->type == DLM_MLE_BLOCK) response = DLM_MASTER_RESP_NO; else if (tmpmle->type == DLM_MLE_MIGRATION) { mlog(0, "migration mle was found (%u->%u)\n", tmpmle->master, tmpmle->new_master); - if (tmpmle->master == dlm->node_num) { - mlog(ML_ERROR, "no lockres, but migration mle " - "says that this node is master!\n"); - BUG(); - } /* real master can respond on its own */ response = DLM_MASTER_RESP_NO; - } else { - if (tmpmle->master == dlm->node_num) { - response = DLM_MASTER_RESP_YES; - set_maybe = 0; - } else - response = DLM_MASTER_RESP_MAYBE; - } + } else + response = DLM_MASTER_RESP_MAYBE; if (set_maybe) set_bit(request->node_idx, tmpmle->maybe_map); spin_unlock(&tmpmle->spinlock); @@ -1449,6 +1506,24 @@ way_up_top: dlm_put_mle(tmpmle); } send_response: + + if (dispatch_assert) { + if (response != DLM_MASTER_RESP_YES) + mlog(ML_ERROR, "invalid response %d\n", response); + if (!res) { + mlog(ML_ERROR, "bad lockres while trying to assert!\n"); + BUG(); + } + mlog(0, "%u is the owner of %.*s, cleaning everyone else\n", + dlm->node_num, res->lockname.len, res->lockname.name); + ret = dlm_dispatch_assert_master(dlm, res, 0, request->node_idx, + DLM_ASSERT_MASTER_MLE_CLEANUP); + if (ret < 0) { + mlog(ML_ERROR, "failed to dispatch assert master work\n"); + response = DLM_MASTER_RESP_ERROR; + } + } + dlm_put(dlm); return response; } @@ -1471,8 +1546,11 @@ static int dlm_do_assert_master(struct dlm_ctxt *dlm, const char *lockname, int to, tmpret; struct dlm_node_iter iter; int ret = 0; + int reassert; BUG_ON(namelen > O2NM_MAX_NAME_LEN); +again: + reassert = 0; /* note that if this nodemap is empty, it returns 0 */ dlm_node_iter_init(nodemap, &iter); @@ -1504,9 +1582,17 @@ static int dlm_do_assert_master(struct dlm_ctxt *dlm, const char *lockname, "got %d.\n", namelen, lockname, to, r); dlm_dump_lock_resources(dlm); BUG(); + } else if (r == EAGAIN) { + mlog(0, "%.*s: node %u create mles on other " + "nodes and requests a re-assert\n", + namelen, lockname, to); + reassert = 1; } } + if (reassert) + goto again; + return ret; } @@ -1528,6 +1614,8 @@ int dlm_assert_master_handler(struct o2net_msg *msg, u32 len, void *data) char *name; unsigned int namelen; u32 flags; + int master_request = 0; + int ret = 0; if (!dlm_grab(dlm)) return 0; @@ -1642,11 +1730,22 @@ ok: // mlog(0, "woo! got an assert_master from node %u!\n", // assert->node_idx); if (mle) { - int extra_ref; + int extra_ref = 0; + int nn = -1; spin_lock(&mle->spinlock); - extra_ref = !!(mle->type == DLM_MLE_BLOCK - || mle->type == DLM_MLE_MIGRATION); + if (mle->type == DLM_MLE_BLOCK || mle->type == DLM_MLE_MIGRATION) + extra_ref = 1; + else { + /* MASTER mle: if any bits set in the response map + * then the calling node needs to re-assert to clear + * up nodes that this node contacted */ + while ((nn = find_next_bit (mle->response_map, O2NM_MAX_NODES, + nn+1)) < O2NM_MAX_NODES) { + if (nn != dlm->node_num && nn != assert->node_idx) + master_request = 1; + } + } mle->master = assert->node_idx; atomic_set(&mle->woken, 1); wake_up(&mle->wq); @@ -1677,10 +1776,15 @@ ok: } done: + ret = 0; if (res) dlm_lockres_put(res); dlm_put(dlm); - return 0; + if (master_request) { + mlog(0, "need to tell master to reassert\n"); + ret = EAGAIN; // positive. negative would shoot down the node. + } + return ret; kill: /* kill the caller! */ @@ -1713,6 +1817,10 @@ int dlm_dispatch_assert_master(struct dlm_ctxt *dlm, item->u.am.request_from = request_from; item->u.am.flags = flags; + if (ignore_higher) + mlog(0, "IGNORE HIGHER: %.*s\n", res->lockname.len, + res->lockname.name); + spin_lock(&dlm->work_lock); list_add_tail(&item->list, &dlm->work_list); spin_unlock(&dlm->work_lock); @@ -1775,6 +1883,61 @@ static void dlm_assert_master_worker(struct dlm_work_item *item, void *data) mlog(0, "finished with dlm_assert_master_worker\n"); } +/* SPECIAL CASE for the $RECOVERY lock used by the recovery thread. + * We cannot wait for node recovery to complete to begin mastering this + * lockres because this lockres is used to kick off recovery! ;-) + * So, do a pre-check on all living nodes to see if any of those nodes + * think that $RECOVERY is currently mastered by a dead node. If so, + * we wait a short time to allow that node to get notified by its own + * heartbeat stack, then check again. All $RECOVERY lock resources + * mastered by dead nodes are purged when the hearbeat callback is + * fired, so we can know for sure that it is safe to continue once + * the node returns a live node or no node. */ +static int dlm_pre_master_reco_lockres(struct dlm_ctxt *dlm, + struct dlm_lock_resource *res) +{ + struct dlm_node_iter iter; + int nodenum; + int ret = 0; + u8 master = DLM_LOCK_RES_OWNER_UNKNOWN; + + spin_lock(&dlm->spinlock); + dlm_node_iter_init(dlm->domain_map, &iter); + spin_unlock(&dlm->spinlock); + + while ((nodenum = dlm_node_iter_next(&iter)) >= 0) { + /* do not send to self */ + if (nodenum == dlm->node_num) + continue; + ret = dlm_do_master_requery(dlm, res, nodenum, &master); + if (ret < 0) { + mlog_errno(ret); + if (!dlm_is_host_down(ret)) + BUG(); + /* host is down, so answer for that node would be + * DLM_LOCK_RES_OWNER_UNKNOWN. continue. */ + } + + if (master != DLM_LOCK_RES_OWNER_UNKNOWN) { + /* check to see if this master is in the recovery map */ + spin_lock(&dlm->spinlock); + if (test_bit(master, dlm->recovery_map)) { + mlog(ML_NOTICE, "%s: node %u has not seen " + "node %u go down yet, and thinks the " + "dead node is mastering the recovery " + "lock. must wait.\n", dlm->name, + nodenum, master); + ret = -EAGAIN; + } + spin_unlock(&dlm->spinlock); + mlog(0, "%s: reco lock master is %u\n", dlm->name, + master); + break; + } + } + return ret; +} + /* * DLM_MIGRATE_LOCKRES diff --git a/fs/ocfs2/dlm/dlmrecovery.c b/fs/ocfs2/dlm/dlmrecovery.c index 1e232000f3f..805cbabac05 100644 --- a/fs/ocfs2/dlm/dlmrecovery.c +++ b/fs/ocfs2/dlm/dlmrecovery.c @@ -58,7 +58,7 @@ static void dlm_do_local_recovery_cleanup(struct dlm_ctxt *dlm, u8 dead_node); static int dlm_recovery_thread(void *data); void dlm_complete_recovery_thread(struct dlm_ctxt *dlm); int dlm_launch_recovery_thread(struct dlm_ctxt *dlm); -static void dlm_kick_recovery_thread(struct dlm_ctxt *dlm); +void dlm_kick_recovery_thread(struct dlm_ctxt *dlm); static int dlm_do_recovery(struct dlm_ctxt *dlm); static int dlm_pick_recovery_master(struct dlm_ctxt *dlm); @@ -78,15 +78,9 @@ static int dlm_send_mig_lockres_msg(struct dlm_ctxt *dlm, u8 send_to, struct dlm_lock_resource *res, int total_locks); -static int dlm_lockres_master_requery(struct dlm_ctxt *dlm, - struct dlm_lock_resource *res, - u8 *real_master); static int dlm_process_recovery_data(struct dlm_ctxt *dlm, struct dlm_lock_resource *res, struct dlm_migratable_lockres *mres); -static int dlm_do_master_requery(struct dlm_ctxt *dlm, - struct dlm_lock_resource *res, - u8 nodenum, u8 *real_master); static int dlm_send_finalize_reco_message(struct dlm_ctxt *dlm); static int dlm_send_all_done_msg(struct dlm_ctxt *dlm, u8 dead_node, u8 send_to); @@ -165,7 +159,7 @@ void dlm_dispatch_work(void *data) * RECOVERY THREAD */ -static void dlm_kick_recovery_thread(struct dlm_ctxt *dlm) +void dlm_kick_recovery_thread(struct dlm_ctxt *dlm) { /* wake the recovery thread * this will wake the reco thread in one of three places @@ -750,10 +744,12 @@ static void dlm_request_all_locks_worker(struct dlm_work_item *item, void *data) dlm->name, dlm->reco.dead_node, dlm->reco.new_master, dead_node, reco_master); mlog(ML_ERROR, "%s: name=%.*s master=%u locks=%u/%u flags=%u " - "entry[0]={c=%"MLFu64",l=%u,f=%u,t=%d,ct=%d,hb=%d,n=%u}\n", + "entry[0]={c=%u:%llu,l=%u,f=%u,t=%d,ct=%d,hb=%d,n=%u}\n", dlm->name, mres->lockname_len, mres->lockname, mres->master, mres->num_locks, mres->total_locks, mres->flags, - mres->ml[0].cookie, mres->ml[0].list, mres->ml[0].flags, + dlm_get_lock_cookie_node(mres->ml[0].cookie), + dlm_get_lock_cookie_seq(mres->ml[0].cookie), + mres->ml[0].list, mres->ml[0].flags, mres->ml[0].type, mres->ml[0].convert_type, mres->ml[0].highest_blocked, mres->ml[0].node); BUG(); @@ -1316,9 +1312,8 @@ leave: -static int dlm_lockres_master_requery(struct dlm_ctxt *dlm, - struct dlm_lock_resource *res, - u8 *real_master) +int dlm_lockres_master_requery(struct dlm_ctxt *dlm, + struct dlm_lock_resource *res, u8 *real_master) { struct dlm_node_iter iter; int nodenum; @@ -1360,8 +1355,10 @@ static int dlm_lockres_master_requery(struct dlm_ctxt *dlm, ret = dlm_do_master_requery(dlm, res, nodenum, real_master); if (ret < 0) { mlog_errno(ret); - BUG(); - /* TODO: need to figure a way to restart this */ + if (!dlm_is_host_down(ret)) + BUG(); + /* host is down, so answer for that node would be + * DLM_LOCK_RES_OWNER_UNKNOWN. continue. */ } if (*real_master != DLM_LOCK_RES_OWNER_UNKNOWN) { mlog(0, "lock master is %u\n", *real_master); @@ -1372,9 +1369,8 @@ static int dlm_lockres_master_requery(struct dlm_ctxt *dlm, } -static int dlm_do_master_requery(struct dlm_ctxt *dlm, - struct dlm_lock_resource *res, - u8 nodenum, u8 *real_master) +int dlm_do_master_requery(struct dlm_ctxt *dlm, struct dlm_lock_resource *res, + u8 nodenum, u8 *real_master) { int ret = -EINVAL; struct dlm_master_requery req; @@ -1519,9 +1515,11 @@ static int dlm_process_recovery_data(struct dlm_ctxt *dlm, /* lock is always created locally first, and * destroyed locally last. it must be on the list */ if (!lock) { + u64 c = ml->cookie; mlog(ML_ERROR, "could not find local lock " - "with cookie %"MLFu64"!\n", - ml->cookie); + "with cookie %u:%llu!\n", + dlm_get_lock_cookie_node(c), + dlm_get_lock_cookie_seq(c)); BUG(); } BUG_ON(lock->ml.node != ml->node); @@ -1739,6 +1737,13 @@ static void dlm_finish_local_lockres_recovery(struct dlm_ctxt *dlm, } else continue; + if (!list_empty(&res->recovering)) { + mlog(0, "%s:%.*s: lockres was " + "marked RECOVERING, owner=%u\n", + dlm->name, res->lockname.len, + res->lockname.name, res->owner); + list_del_init(&res->recovering); + } spin_lock(&res->spinlock); dlm_change_lockres_owner(dlm, res, new_master); res->state &= ~DLM_LOCK_RES_RECOVERING; @@ -2258,7 +2263,10 @@ int dlm_begin_reco_handler(struct o2net_msg *msg, u32 len, void *data) mlog(0, "%u not in domain/live_nodes map " "so setting it in reco map manually\n", br->dead_node); - set_bit(br->dead_node, dlm->recovery_map); + /* force the recovery cleanup in __dlm_hb_node_down + * both of these will be cleared in a moment */ + set_bit(br->dead_node, dlm->domain_map); + set_bit(br->dead_node, dlm->live_nodes_map); __dlm_hb_node_down(dlm, br->dead_node); } spin_unlock(&dlm->spinlock); diff --git a/fs/ocfs2/dlm/dlmunlock.c b/fs/ocfs2/dlm/dlmunlock.c index c95f08d2e92..7b1a2754267 100644 --- a/fs/ocfs2/dlm/dlmunlock.c +++ b/fs/ocfs2/dlm/dlmunlock.c @@ -244,8 +244,10 @@ leave: if (actions & DLM_UNLOCK_FREE_LOCK) { /* this should always be coupled with list removal */ BUG_ON(!(actions & DLM_UNLOCK_REMOVE_LOCK)); - mlog(0, "lock %"MLFu64" should be gone now! refs=%d\n", - lock->ml.cookie, atomic_read(&lock->lock_refs.refcount)-1); + mlog(0, "lock %u:%llu should be gone now! refs=%d\n", + dlm_get_lock_cookie_node(lock->ml.cookie), + dlm_get_lock_cookie_seq(lock->ml.cookie), + atomic_read(&lock->lock_refs.refcount)-1); dlm_lock_put(lock); } if (actions & DLM_UNLOCK_CALL_AST) @@ -493,8 +495,9 @@ int dlm_unlock_lock_handler(struct o2net_msg *msg, u32 len, void *data) not_found: if (!found) mlog(ML_ERROR, "failed to find lock to unlock! " - "cookie=%"MLFu64"\n", - unlock->cookie); + "cookie=%u:%llu\n", + dlm_get_lock_cookie_node(unlock->cookie), + dlm_get_lock_cookie_seq(unlock->cookie)); else { /* send the lksb->status back to the other node */ status = lksb->status; diff --git a/fs/ocfs2/dlmglue.c b/fs/ocfs2/dlmglue.c index e971ec2f840..64cd52860c8 100644 --- a/fs/ocfs2/dlmglue.c +++ b/fs/ocfs2/dlmglue.c @@ -231,9 +231,9 @@ static void ocfs2_build_lock_name(enum ocfs2_lock_type type, BUG_ON(type >= OCFS2_NUM_LOCK_TYPES); - len = snprintf(name, OCFS2_LOCK_ID_MAX_LEN, "%c%s%016"MLFx64"%08x", - ocfs2_lock_type_char(type), OCFS2_LOCK_ID_PAD, blkno, - generation); + len = snprintf(name, OCFS2_LOCK_ID_MAX_LEN, "%c%s%016llx%08x", + ocfs2_lock_type_char(type), OCFS2_LOCK_ID_PAD, + (long long)blkno, generation); BUG_ON(len != (OCFS2_LOCK_ID_MAX_LEN - 1)); @@ -533,8 +533,8 @@ static void ocfs2_inode_ast_func(void *opaque) inode = ocfs2_lock_res_inode(lockres); - mlog(0, "AST fired for inode %"MLFu64", l_action = %u, type = %s\n", - OCFS2_I(inode)->ip_blkno, lockres->l_action, + mlog(0, "AST fired for inode %llu, l_action = %u, type = %s\n", + (unsigned long long)OCFS2_I(inode)->ip_blkno, lockres->l_action, ocfs2_lock_type_string(lockres->l_type)); BUG_ON(!ocfs2_is_inode_lock(lockres)); @@ -544,8 +544,8 @@ static void ocfs2_inode_ast_func(void *opaque) lksb = &(lockres->l_lksb); if (lksb->status != DLM_NORMAL) { mlog(ML_ERROR, "ocfs2_inode_ast_func: lksb status value of %u " - "on inode %"MLFu64"\n", lksb->status, - OCFS2_I(inode)->ip_blkno); + "on inode %llu\n", lksb->status, + (unsigned long long)OCFS2_I(inode)->ip_blkno); spin_unlock_irqrestore(&lockres->l_lock, flags); mlog_exit_void(); return; @@ -646,10 +646,9 @@ static void ocfs2_inode_bast_func(void *opaque, int level) inode = ocfs2_lock_res_inode(lockres); osb = OCFS2_SB(inode->i_sb); - mlog(0, "BAST fired for inode %"MLFu64", blocking = %d, level = %d " - "type = %s\n", OCFS2_I(inode)->ip_blkno, level, - lockres->l_level, - ocfs2_lock_type_string(lockres->l_type)); + mlog(0, "BAST fired for inode %llu, blocking %d, level %d type %s\n", + (unsigned long long)OCFS2_I(inode)->ip_blkno, level, + lockres->l_level, ocfs2_lock_type_string(lockres->l_type)); ocfs2_generic_bast_func(osb, lockres, level); @@ -1104,7 +1103,7 @@ int ocfs2_create_new_inode_locks(struct inode *inode) mlog_entry_void(); - mlog(0, "Inode %"MLFu64"\n", OCFS2_I(inode)->ip_blkno); + mlog(0, "Inode %llu\n", (unsigned long long)OCFS2_I(inode)->ip_blkno); /* NOTE: That we don't increment any of the holder counts, nor * do we add anything to a journal handle. Since this is @@ -1149,8 +1148,8 @@ int ocfs2_rw_lock(struct inode *inode, int write) mlog_entry_void(); - mlog(0, "inode %"MLFu64" take %s RW lock\n", - OCFS2_I(inode)->ip_blkno, + mlog(0, "inode %llu take %s RW lock\n", + (unsigned long long)OCFS2_I(inode)->ip_blkno, write ? "EXMODE" : "PRMODE"); lockres = &OCFS2_I(inode)->ip_rw_lockres; @@ -1173,8 +1172,8 @@ void ocfs2_rw_unlock(struct inode *inode, int write) mlog_entry_void(); - mlog(0, "inode %"MLFu64" drop %s RW lock\n", - OCFS2_I(inode)->ip_blkno, + mlog(0, "inode %llu drop %s RW lock\n", + (unsigned long long)OCFS2_I(inode)->ip_blkno, write ? "EXMODE" : "PRMODE"); ocfs2_cluster_unlock(OCFS2_SB(inode->i_sb), lockres, level); @@ -1193,8 +1192,8 @@ int ocfs2_data_lock_full(struct inode *inode, mlog_entry_void(); - mlog(0, "inode %"MLFu64" take %s DATA lock\n", - OCFS2_I(inode)->ip_blkno, + mlog(0, "inode %llu take %s DATA lock\n", + (unsigned long long)OCFS2_I(inode)->ip_blkno, write ? "EXMODE" : "PRMODE"); /* We'll allow faking a readonly data lock for @@ -1278,8 +1277,8 @@ void ocfs2_data_unlock(struct inode *inode, mlog_entry_void(); - mlog(0, "inode %"MLFu64" drop %s DATA lock\n", - OCFS2_I(inode)->ip_blkno, + mlog(0, "inode %llu drop %s DATA lock\n", + (unsigned long long)OCFS2_I(inode)->ip_blkno, write ? "EXMODE" : "PRMODE"); if (!ocfs2_is_hard_readonly(OCFS2_SB(inode->i_sb))) @@ -1462,9 +1461,9 @@ static int ocfs2_meta_lock_update(struct inode *inode, spin_lock(&oi->ip_lock); if (oi->ip_flags & OCFS2_INODE_DELETED) { - mlog(0, "Orphaned inode %"MLFu64" was deleted while we " + mlog(0, "Orphaned inode %llu was deleted while we " "were waiting on a lock. ip_flags = 0x%x\n", - oi->ip_blkno, oi->ip_flags); + (unsigned long long)oi->ip_blkno, oi->ip_flags); spin_unlock(&oi->ip_lock); status = -ENOENT; goto bail; @@ -1485,8 +1484,8 @@ static int ocfs2_meta_lock_update(struct inode *inode, ocfs2_extent_map_trunc(inode, 0); if (ocfs2_meta_lvb_is_trustable(lockres)) { - mlog(0, "Trusting LVB on inode %"MLFu64"\n", - oi->ip_blkno); + mlog(0, "Trusting LVB on inode %llu\n", + (unsigned long long)oi->ip_blkno); ocfs2_refresh_inode_from_lvb(inode); } else { /* Boo, we have to go to disk. */ @@ -1514,15 +1513,16 @@ static int ocfs2_meta_lock_update(struct inode *inode, } mlog_bug_on_msg(inode->i_generation != le32_to_cpu(fe->i_generation), - "Invalid dinode %"MLFu64" disk generation: %u " + "Invalid dinode %llu disk generation: %u " "inode->i_generation: %u\n", - oi->ip_blkno, le32_to_cpu(fe->i_generation), + (unsigned long long)oi->ip_blkno, + le32_to_cpu(fe->i_generation), inode->i_generation); mlog_bug_on_msg(le64_to_cpu(fe->i_dtime) || !(fe->i_flags & cpu_to_le32(OCFS2_VALID_FL)), - "Stale dinode %"MLFu64" dtime: %"MLFu64" " - "flags: 0x%x\n", oi->ip_blkno, - le64_to_cpu(fe->i_dtime), + "Stale dinode %llu dtime: %llu flags: 0x%x\n", + (unsigned long long)oi->ip_blkno, + (unsigned long long)le64_to_cpu(fe->i_dtime), le32_to_cpu(fe->i_flags)); ocfs2_refresh_inode(inode, fe); @@ -1581,8 +1581,8 @@ int ocfs2_meta_lock_full(struct inode *inode, mlog_entry_void(); - mlog(0, "inode %"MLFu64", take %s META lock\n", - OCFS2_I(inode)->ip_blkno, + mlog(0, "inode %llu, take %s META lock\n", + (unsigned long long)OCFS2_I(inode)->ip_blkno, ex ? "EXMODE" : "PRMODE"); status = 0; @@ -1716,8 +1716,8 @@ void ocfs2_meta_unlock(struct inode *inode, mlog_entry_void(); - mlog(0, "inode %"MLFu64" drop %s META lock\n", - OCFS2_I(inode)->ip_blkno, + mlog(0, "inode %llu drop %s META lock\n", + (unsigned long long)OCFS2_I(inode)->ip_blkno, ex ? "EXMODE" : "PRMODE"); if (!ocfs2_is_hard_readonly(OCFS2_SB(inode->i_sb))) @@ -2017,7 +2017,7 @@ out: return ret; } -static struct file_operations ocfs2_dlm_debug_fops = { +static const struct file_operations ocfs2_dlm_debug_fops = { .open = ocfs2_dlm_debug_open, .release = ocfs2_dlm_debug_release, .read = seq_read, @@ -2686,8 +2686,8 @@ static void ocfs2_data_convert_worker(struct ocfs2_lock_res *lockres, mapping = inode->i_mapping; if (filemap_fdatawrite(mapping)) { - mlog(ML_ERROR, "Could not sync inode %"MLFu64" for downconvert!", - OCFS2_I(inode)->ip_blkno); + mlog(ML_ERROR, "Could not sync inode %llu for downconvert!", + (unsigned long long)OCFS2_I(inode)->ip_blkno); } sync_mapping_buffers(mapping); if (blocking == LKM_EXMODE) { @@ -2717,7 +2717,8 @@ int ocfs2_unblock_data(struct ocfs2_lock_res *lockres, inode = ocfs2_lock_res_inode(lockres); osb = OCFS2_SB(inode->i_sb); - mlog(0, "unblock inode %"MLFu64"\n", OCFS2_I(inode)->ip_blkno); + mlog(0, "unblock inode %llu\n", + (unsigned long long)OCFS2_I(inode)->ip_blkno); status = ocfs2_generic_unblock_lock(osb, lockres, @@ -2726,8 +2727,8 @@ int ocfs2_unblock_data(struct ocfs2_lock_res *lockres, if (status < 0) mlog_errno(status); - mlog(0, "inode %"MLFu64", requeue = %d\n", - OCFS2_I(inode)->ip_blkno, *requeue); + mlog(0, "inode %llu, requeue = %d\n", + (unsigned long long)OCFS2_I(inode)->ip_blkno, *requeue); mlog_exit(status); return status; @@ -2767,14 +2768,15 @@ int ocfs2_unblock_meta(struct ocfs2_lock_res *lockres, inode = ocfs2_lock_res_inode(lockres); - mlog(0, "unblock inode %"MLFu64"\n", OCFS2_I(inode)->ip_blkno); + mlog(0, "unblock inode %llu\n", + (unsigned long long)OCFS2_I(inode)->ip_blkno); status = ocfs2_do_unblock_meta(inode, requeue); if (status < 0) mlog_errno(status); - mlog(0, "inode %"MLFu64", requeue = %d\n", - OCFS2_I(inode)->ip_blkno, *requeue); + mlog(0, "inode %llu, requeue = %d\n", + (unsigned long long)OCFS2_I(inode)->ip_blkno, *requeue); mlog_exit(status); return status; @@ -2893,12 +2895,13 @@ void ocfs2_dump_meta_lvb_info(u64 level, lockres->l_name, function, line); mlog(level, "version: %u, clusters: %u\n", be32_to_cpu(lvb->lvb_version), be32_to_cpu(lvb->lvb_iclusters)); - mlog(level, "size: %"MLFu64", uid %u, gid %u, mode 0x%x\n", - be64_to_cpu(lvb->lvb_isize), be32_to_cpu(lvb->lvb_iuid), - be32_to_cpu(lvb->lvb_igid), be16_to_cpu(lvb->lvb_imode)); - mlog(level, "nlink %u, atime_packed 0x%"MLFx64", " - "ctime_packed 0x%"MLFx64", mtime_packed 0x%"MLFx64"\n", - be16_to_cpu(lvb->lvb_inlink), be64_to_cpu(lvb->lvb_iatime_packed), - be64_to_cpu(lvb->lvb_ictime_packed), - be64_to_cpu(lvb->lvb_imtime_packed)); + mlog(level, "size: %llu, uid %u, gid %u, mode 0x%x\n", + (unsigned long long)be64_to_cpu(lvb->lvb_isize), + be32_to_cpu(lvb->lvb_iuid), be32_to_cpu(lvb->lvb_igid), + be16_to_cpu(lvb->lvb_imode)); + mlog(level, "nlink %u, atime_packed 0x%llx, ctime_packed 0x%llx, " + "mtime_packed 0x%llx\n", be16_to_cpu(lvb->lvb_inlink), + (long long)be64_to_cpu(lvb->lvb_iatime_packed), + (long long)be64_to_cpu(lvb->lvb_ictime_packed), + (long long)be64_to_cpu(lvb->lvb_imtime_packed)); } diff --git a/fs/ocfs2/export.c b/fs/ocfs2/export.c index 5810160d92a..ec55ab3c121 100644 --- a/fs/ocfs2/export.c +++ b/fs/ocfs2/export.c @@ -95,8 +95,8 @@ static struct dentry *ocfs2_get_parent(struct dentry *child) mlog_entry("(0x%p, '%.*s')\n", child, child->d_name.len, child->d_name.name); - mlog(0, "find parent of directory %"MLFu64"\n", - OCFS2_I(dir)->ip_blkno); + mlog(0, "find parent of directory %llu\n", + (unsigned long long)OCFS2_I(dir)->ip_blkno); status = ocfs2_meta_lock(dir, NULL, NULL, 0); if (status < 0) { @@ -115,7 +115,8 @@ static struct dentry *ocfs2_get_parent(struct dentry *child) inode = ocfs2_iget(OCFS2_SB(dir->i_sb), blkno); if (IS_ERR(inode)) { - mlog(ML_ERROR, "Unable to create inode %"MLFu64"\n", blkno); + mlog(ML_ERROR, "Unable to create inode %llu\n", + (unsigned long long)blkno); parent = ERR_PTR(-EACCES); goto bail_unlock; } @@ -160,8 +161,8 @@ static int ocfs2_encode_fh(struct dentry *dentry, __be32 *fh, int *max_len, blkno = OCFS2_I(inode)->ip_blkno; generation = inode->i_generation; - mlog(0, "Encoding fh: blkno: %"MLFu64", generation: %u\n", - blkno, generation); + mlog(0, "Encoding fh: blkno: %llu, generation: %u\n", + (unsigned long long)blkno, generation); len = 3; fh[0] = cpu_to_le32((u32)(blkno >> 32)); @@ -186,8 +187,8 @@ static int ocfs2_encode_fh(struct dentry *dentry, __be32 *fh, int *max_len, len = 6; type = 2; - mlog(0, "Encoding parent: blkno: %"MLFu64", generation: %u\n", - blkno, generation); + mlog(0, "Encoding parent: blkno: %llu, generation: %u\n", + (unsigned long long)blkno, generation); } *max_len = len; @@ -220,16 +221,17 @@ static struct dentry *ocfs2_decode_fh(struct super_block *sb, __be32 *fh, parent.ih_blkno |= (u64)le32_to_cpu(fh[4]); parent.ih_generation = le32_to_cpu(fh[5]); - mlog(0, "Decoding parent: blkno: %"MLFu64", generation: %u\n", - parent.ih_blkno, parent.ih_generation); + mlog(0, "Decoding parent: blkno: %llu, generation: %u\n", + (unsigned long long)parent.ih_blkno, + parent.ih_generation); } handle.ih_blkno = (u64)le32_to_cpu(fh[0]) << 32; handle.ih_blkno |= (u64)le32_to_cpu(fh[1]); handle.ih_generation = le32_to_cpu(fh[2]); - mlog(0, "Encoding fh: blkno: %"MLFu64", generation: %u\n", - handle.ih_blkno, handle.ih_generation); + mlog(0, "Encoding fh: blkno: %llu, generation: %u\n", + (unsigned long long)handle.ih_blkno, handle.ih_generation); ret = ocfs2_export_ops.find_exported_dentry(sb, &handle, &parent, acceptable, context); diff --git a/fs/ocfs2/extent_map.c b/fs/ocfs2/extent_map.c index e6f207eebab..4601fc256f1 100644 --- a/fs/ocfs2/extent_map.c +++ b/fs/ocfs2/extent_map.c @@ -182,10 +182,10 @@ static int ocfs2_extent_map_find_leaf(struct inode *inode, if (rec_end > OCFS2_I(inode)->ip_clusters) { mlog_errno(ret); ocfs2_error(inode->i_sb, - "Extent %d at e_blkno %"MLFu64" of inode %"MLFu64" goes past ip_clusters of %u\n", + "Extent %d at e_blkno %llu of inode %llu goes past ip_clusters of %u\n", i, - le64_to_cpu(rec->e_blkno), - OCFS2_I(inode)->ip_blkno, + (unsigned long long)le64_to_cpu(rec->e_blkno), + (unsigned long long)OCFS2_I(inode)->ip_blkno, OCFS2_I(inode)->ip_clusters); goto out_free; } @@ -233,11 +233,11 @@ static int ocfs2_extent_map_find_leaf(struct inode *inode, if (blkno) { mlog_errno(ret); ocfs2_error(inode->i_sb, - "Multiple extents for (cpos = %u, clusters = %u) on inode %"MLFu64"; e_blkno %"MLFu64" and rec %d at e_blkno %"MLFu64"\n", + "Multiple extents for (cpos = %u, clusters = %u) on inode %llu; e_blkno %llu and rec %d at e_blkno %llu\n", cpos, clusters, - OCFS2_I(inode)->ip_blkno, - blkno, i, - le64_to_cpu(rec->e_blkno)); + (unsigned long long)OCFS2_I(inode)->ip_blkno, + (unsigned long long)blkno, i, + (unsigned long long)le64_to_cpu(rec->e_blkno)); goto out_free; } @@ -251,9 +251,9 @@ static int ocfs2_extent_map_find_leaf(struct inode *inode, ret = -EBADR; if (!blkno) { ocfs2_error(inode->i_sb, - "No record found for (cpos = %u, clusters = %u) on inode %"MLFu64"\n", + "No record found for (cpos = %u, clusters = %u) on inode %llu\n", cpos, clusters, - OCFS2_I(inode)->ip_blkno); + (unsigned long long)OCFS2_I(inode)->ip_blkno); mlog_errno(ret); goto out_free; } @@ -288,10 +288,10 @@ static int ocfs2_extent_map_find_leaf(struct inode *inode, ret = -EBADR; mlog_errno(ret); ocfs2_error(inode->i_sb, - "Extent %d at e_blkno %"MLFu64" of inode %"MLFu64" goes past ip_clusters of %u\n", + "Extent %d at e_blkno %llu of inode %llu goes past ip_clusters of %u\n", i, - le64_to_cpu(rec->e_blkno), - OCFS2_I(inode)->ip_blkno, + (unsigned long long)le64_to_cpu(rec->e_blkno), + (unsigned long long)OCFS2_I(inode)->ip_blkno, OCFS2_I(inode)->ip_clusters); return ret; } @@ -557,9 +557,9 @@ static int ocfs2_extent_map_insert(struct inode *inode, ret = -EBADR; mlog_errno(ret); ocfs2_error(inode->i_sb, - "Zero e_clusters on non-tail extent record at e_blkno %"MLFu64" on inode %"MLFu64"\n", - le64_to_cpu(rec->e_blkno), - OCFS2_I(inode)->ip_blkno); + "Zero e_clusters on non-tail extent record at e_blkno %llu on inode %llu\n", + (unsigned long long)le64_to_cpu(rec->e_blkno), + (unsigned long long)OCFS2_I(inode)->ip_blkno); return ret; } @@ -660,10 +660,10 @@ int ocfs2_extent_map_append(struct inode *inode, mlog_bug_on_msg((le32_to_cpu(rec->e_cpos) + le32_to_cpu(rec->e_clusters)) != (em->em_clusters + new_clusters), - "Inode %"MLFu64":\n" + "Inode %llu:\n" "rec->e_cpos = %u + rec->e_clusters = %u = %u\n" "em->em_clusters = %u + new_clusters = %u = %u\n", - OCFS2_I(inode)->ip_blkno, + (unsigned long long)OCFS2_I(inode)->ip_blkno, le32_to_cpu(rec->e_cpos), le32_to_cpu(rec->e_clusters), le32_to_cpu(rec->e_cpos) + le32_to_cpu(rec->e_clusters), em->em_clusters, new_clusters, diff --git a/fs/ocfs2/file.c b/fs/ocfs2/file.c index 8a4048b55fd..34e903a6a46 100644 --- a/fs/ocfs2/file.c +++ b/fs/ocfs2/file.c @@ -220,8 +220,9 @@ static int ocfs2_truncate_file(struct inode *inode, struct ocfs2_super *osb = OCFS2_SB(inode->i_sb); struct ocfs2_truncate_context *tc = NULL; - mlog_entry("(inode = %"MLFu64", new_i_size = %"MLFu64"\n", - OCFS2_I(inode)->ip_blkno, new_i_size); + mlog_entry("(inode = %llu, new_i_size = %llu\n", + (unsigned long long)OCFS2_I(inode)->ip_blkno, + (unsigned long long)new_i_size); truncate_inode_pages(inode->i_mapping, new_i_size); @@ -233,23 +234,26 @@ static int ocfs2_truncate_file(struct inode *inode, } mlog_bug_on_msg(le64_to_cpu(fe->i_size) != i_size_read(inode), - "Inode %"MLFu64", inode i_size = %lld != di " - "i_size = %"MLFu64", i_flags = 0x%x\n", - OCFS2_I(inode)->ip_blkno, + "Inode %llu, inode i_size = %lld != di " + "i_size = %llu, i_flags = 0x%x\n", + (unsigned long long)OCFS2_I(inode)->ip_blkno, i_size_read(inode), - le64_to_cpu(fe->i_size), le32_to_cpu(fe->i_flags)); + (unsigned long long)le64_to_cpu(fe->i_size), + le32_to_cpu(fe->i_flags)); if (new_i_size > le64_to_cpu(fe->i_size)) { - mlog(0, "asked to truncate file with size (%"MLFu64") " - "to size (%"MLFu64")!\n", - le64_to_cpu(fe->i_size), new_i_size); + mlog(0, "asked to truncate file with size (%llu) to size (%llu)!\n", + (unsigned long long)le64_to_cpu(fe->i_size), + (unsigned long long)new_i_size); status = -EINVAL; mlog_errno(status); goto bail; } - mlog(0, "inode %"MLFu64", i_size = %"MLFu64", new_i_size = %"MLFu64"\n", - le64_to_cpu(fe->i_blkno), le64_to_cpu(fe->i_size), new_i_size); + mlog(0, "inode %llu, i_size = %llu, new_i_size = %llu\n", + (unsigned long long)le64_to_cpu(fe->i_blkno), + (unsigned long long)le64_to_cpu(fe->i_size), + (unsigned long long)new_i_size); /* lets handle the simple truncate cases before doing any more * cluster locking. */ @@ -378,8 +382,8 @@ int ocfs2_do_extend_allocation(struct ocfs2_super *osb, } block = ocfs2_clusters_to_blocks(osb->sb, bit_off); - mlog(0, "Allocating %u clusters at block %u for inode %"MLFu64"\n", - num_bits, bit_off, OCFS2_I(inode)->ip_blkno); + mlog(0, "Allocating %u clusters at block %u for inode %llu\n", + num_bits, bit_off, (unsigned long long)OCFS2_I(inode)->ip_blkno); status = ocfs2_insert_extent(osb, handle, inode, fe_bh, block, num_bits, meta_ac); if (status < 0) { @@ -449,9 +453,9 @@ static int ocfs2_extend_allocation(struct inode *inode, restart_all: BUG_ON(le32_to_cpu(fe->i_clusters) != OCFS2_I(inode)->ip_clusters); - mlog(0, "extend inode %"MLFu64", i_size = %lld, fe->i_clusters = %u, " + mlog(0, "extend inode %llu, i_size = %lld, fe->i_clusters = %u, " "clusters_to_add = %u\n", - OCFS2_I(inode)->ip_blkno, i_size_read(inode), + (unsigned long long)OCFS2_I(inode)->ip_blkno, i_size_read(inode), fe->i_clusters, clusters_to_add); handle = ocfs2_alloc_handle(osb); @@ -569,8 +573,8 @@ restarted_transaction: } } - mlog(0, "fe: i_clusters = %u, i_size=%"MLFu64"\n", - fe->i_clusters, fe->i_size); + mlog(0, "fe: i_clusters = %u, i_size=%llu\n", + fe->i_clusters, (unsigned long long)fe->i_size); mlog(0, "inode: ip_clusters=%u, i_size=%lld\n", OCFS2_I(inode)->ip_clusters, i_size_read(inode)); @@ -865,8 +869,8 @@ static int ocfs2_write_remove_suid(struct inode *inode) struct ocfs2_super *osb = OCFS2_SB(inode->i_sb); struct ocfs2_dinode *di; - mlog_entry("(Inode %"MLFu64", mode 0%o)\n", oi->ip_blkno, - inode->i_mode); + mlog_entry("(Inode %llu, mode 0%o)\n", + (unsigned long long)oi->ip_blkno, inode->i_mode); handle = ocfs2_start_trans(osb, NULL, OCFS2_INODE_UPDATE_CREDITS); if (handle == NULL) { @@ -1172,7 +1176,7 @@ struct inode_operations ocfs2_special_file_iops = { .getattr = ocfs2_getattr, }; -struct file_operations ocfs2_fops = { +const struct file_operations ocfs2_fops = { .read = do_sync_read, .write = do_sync_write, .sendfile = generic_file_sendfile, @@ -1184,7 +1188,7 @@ struct file_operations ocfs2_fops = { .aio_write = ocfs2_file_aio_write, }; -struct file_operations ocfs2_dops = { +const struct file_operations ocfs2_dops = { .read = generic_read_dir, .readdir = ocfs2_readdir, .fsync = ocfs2_sync_file, diff --git a/fs/ocfs2/file.h b/fs/ocfs2/file.h index a5ea33b2406..740c9e7ca59 100644 --- a/fs/ocfs2/file.h +++ b/fs/ocfs2/file.h @@ -26,8 +26,8 @@ #ifndef OCFS2_FILE_H #define OCFS2_FILE_H -extern struct file_operations ocfs2_fops; -extern struct file_operations ocfs2_dops; +extern const struct file_operations ocfs2_fops; +extern const struct file_operations ocfs2_dops; extern struct inode_operations ocfs2_file_iops; extern struct inode_operations ocfs2_special_file_iops; struct ocfs2_alloc_context; diff --git a/fs/ocfs2/inode.c b/fs/ocfs2/inode.c index 315472a5c19..327a5b7b86e 100644 --- a/fs/ocfs2/inode.c +++ b/fs/ocfs2/inode.c @@ -95,7 +95,7 @@ struct inode *ocfs2_iget(struct ocfs2_super *osb, u64 blkno) struct super_block *sb = osb->sb; struct ocfs2_find_inode_args args; - mlog_entry("(blkno = %"MLFu64")\n", blkno); + mlog_entry("(blkno = %llu)\n", (unsigned long long)blkno); /* Ok. By now we've either got the offsets passed to us by the * caller, or we just pulled them off the bh. Lets do some @@ -134,8 +134,8 @@ struct inode *ocfs2_iget(struct ocfs2_super *osb, u64 blkno) bail: if (!IS_ERR(inode)) { - mlog(0, "returning inode with number %"MLFu64"\n", - OCFS2_I(inode)->ip_blkno); + mlog(0, "returning inode with number %llu\n", + (unsigned long long)OCFS2_I(inode)->ip_blkno); mlog_exit_ptr(inode); } else mlog_errno(PTR_ERR(inode)); @@ -219,7 +219,8 @@ int ocfs2_populate_inode(struct inode *inode, struct ocfs2_dinode *fe, struct ocfs2_super *osb; int status = -EINVAL; - mlog_entry("(0x%p, size:%"MLFu64")\n", inode, fe->i_size); + mlog_entry("(0x%p, size:%llu)\n", inode, + (unsigned long long)fe->i_size); sb = inode->i_sb; osb = OCFS2_SB(sb); @@ -228,9 +229,10 @@ int ocfs2_populate_inode(struct inode *inode, struct ocfs2_dinode *fe, * today. change if needed. */ if (!OCFS2_IS_VALID_DINODE(fe) || !(fe->i_flags & cpu_to_le32(OCFS2_VALID_FL))) { - mlog(ML_ERROR, "Invalid dinode: i_ino=%lu, i_blkno=%"MLFu64", " + mlog(ML_ERROR, "Invalid dinode: i_ino=%lu, i_blkno=%llu, " "signature = %.*s, flags = 0x%x\n", - inode->i_ino, le64_to_cpu(fe->i_blkno), 7, + inode->i_ino, + (unsigned long long)le64_to_cpu(fe->i_blkno), 7, fe->i_signature, le32_to_cpu(fe->i_flags)); goto bail; } @@ -268,8 +270,9 @@ int ocfs2_populate_inode(struct inode *inode, struct ocfs2_dinode *fe, if (OCFS2_I(inode)->ip_blkno != le64_to_cpu(fe->i_blkno)) mlog(ML_ERROR, - "ip_blkno %"MLFu64" != i_blkno %"MLFu64"!\n", - OCFS2_I(inode)->ip_blkno, fe->i_blkno); + "ip_blkno %llu != i_blkno %llu!\n", + (unsigned long long)OCFS2_I(inode)->ip_blkno, + (unsigned long long)fe->i_blkno); OCFS2_I(inode)->ip_clusters = le32_to_cpu(fe->i_clusters); OCFS2_I(inode)->ip_orphaned_slot = OCFS2_INVALID_SLOT; @@ -278,8 +281,8 @@ int ocfs2_populate_inode(struct inode *inode, struct ocfs2_dinode *fe, inode->i_ino = ino_from_blkno(inode->i_sb, le64_to_cpu(fe->i_blkno)); - mlog(0, "blkno = %"MLFu64", ino = %lu, create_ino = %s\n", - fe->i_blkno, inode->i_ino, create_ino ? "true" : "false"); + mlog(0, "blkno = %llu, ino = %lu, create_ino = %s\n", + (unsigned long long)fe->i_blkno, inode->i_ino, create_ino ? "true" : "false"); inode->i_nlink = le16_to_cpu(fe->i_links_count); @@ -371,8 +374,8 @@ static int ocfs2_read_locked_inode(struct inode *inode, fe = (struct ocfs2_dinode *) bh->b_data; if (!OCFS2_IS_VALID_DINODE(fe)) { - mlog(ML_ERROR, "Invalid dinode #%"MLFu64": signature = %.*s\n", - fe->i_blkno, 7, fe->i_signature); + mlog(ML_ERROR, "Invalid dinode #%llu: signature = %.*s\n", + (unsigned long long)fe->i_blkno, 7, fe->i_signature); make_bad_inode(inode); goto bail; } @@ -386,8 +389,8 @@ static int ocfs2_read_locked_inode(struct inode *inode, status = -EINVAL; if (ocfs2_populate_inode(inode, fe, 0) < 0) { - mlog(ML_ERROR, "populate inode failed! i_blkno=%"MLFu64", " - "i_ino=%lu\n", fe->i_blkno, inode->i_ino); + mlog(ML_ERROR, "populate failed! i_blkno=%llu, i_ino=%lu\n", + (unsigned long long)fe->i_blkno, inode->i_ino); make_bad_inode(inode); goto bail; } @@ -675,8 +678,8 @@ static int ocfs2_inode_is_valid_to_delete(struct inode *inode) * never get here as system file inodes should always have a * positive link count. */ if (oi->ip_flags & OCFS2_INODE_SYSTEM_FILE) { - mlog(ML_ERROR, "Skipping delete of system file %"MLFu64".\n", - oi->ip_blkno); + mlog(ML_ERROR, "Skipping delete of system file %llu\n", + (unsigned long long)oi->ip_blkno); goto bail_unlock; } @@ -715,16 +718,16 @@ static int ocfs2_query_inode_wipe(struct inode *inode, * ocfs2_delete_inode, another node might have asked to delete * the inode. Recheck our flags to catch this. */ if (!ocfs2_inode_is_valid_to_delete(inode)) { - mlog(0, "Skipping delete of %"MLFu64" because flags changed\n", - oi->ip_blkno); + mlog(0, "Skipping delete of %llu because flags changed\n", + (unsigned long long)oi->ip_blkno); goto bail; } /* Now that we have an up to date inode, we can double check * the link count. */ if (inode->i_nlink) { - mlog(0, "Skipping delete of %"MLFu64" because nlink = %u\n", - oi->ip_blkno, inode->i_nlink); + mlog(0, "Skipping delete of %llu because nlink = %u\n", + (unsigned long long)oi->ip_blkno, inode->i_nlink); goto bail; } @@ -734,9 +737,11 @@ static int ocfs2_query_inode_wipe(struct inode *inode, /* for lack of a better error? */ status = -EEXIST; mlog(ML_ERROR, - "Inode %"MLFu64" (on-disk %"MLFu64") not orphaned! " + "Inode %llu (on-disk %llu) not orphaned! " "Disk flags 0x%x, inode flags 0x%x\n", - oi->ip_blkno, di->i_blkno, di->i_flags, oi->ip_flags); + (unsigned long long)oi->ip_blkno, + (unsigned long long)di->i_blkno, di->i_flags, + oi->ip_flags); goto bail; } @@ -753,8 +758,8 @@ static int ocfs2_query_inode_wipe(struct inode *inode, * disk and let them worry about deleting it. */ if (status == -EBUSY) { status = 0; - mlog(0, "Skipping delete of %"MLFu64" because it is in use on" - "other nodes\n", oi->ip_blkno); + mlog(0, "Skipping delete of %llu because it is in use on" + "other nodes\n", (unsigned long long)oi->ip_blkno); goto bail; } if (status < 0) { @@ -768,13 +773,13 @@ static int ocfs2_query_inode_wipe(struct inode *inode, * into. This may happen during node death and * recovery knows how to clean it up so we can safely * ignore this inode for now on. */ - mlog(0, "Nobody knew where inode %"MLFu64" was orphaned!\n", - oi->ip_blkno); + mlog(0, "Nobody knew where inode %llu was orphaned!\n", + (unsigned long long)oi->ip_blkno); } else { *wipe = 1; - mlog(0, "Inode %"MLFu64" is ok to wipe from orphan dir %d\n", - oi->ip_blkno, oi->ip_orphaned_slot); + mlog(0, "Inode %llu is ok to wipe from orphan dir %d\n", + (unsigned long long)oi->ip_blkno, oi->ip_orphaned_slot); } spin_unlock(&oi->ip_lock); @@ -788,8 +793,8 @@ bail: static void ocfs2_cleanup_delete_inode(struct inode *inode, int sync_data) { - mlog(0, "Cleanup inode %"MLFu64", sync = %d\n", - OCFS2_I(inode)->ip_blkno, sync_data); + mlog(0, "Cleanup inode %llu, sync = %d\n", + (unsigned long long)OCFS2_I(inode)->ip_blkno, sync_data); if (sync_data) write_inode_now(inode, 1); truncate_inode_pages(&inode->i_data, 0); @@ -897,8 +902,8 @@ void ocfs2_clear_inode(struct inode *inode) if (!inode) goto bail; - mlog(0, "Clearing inode: %"MLFu64", nlink = %u\n", - OCFS2_I(inode)->ip_blkno, inode->i_nlink); + mlog(0, "Clearing inode: %llu, nlink = %u\n", + (unsigned long long)OCFS2_I(inode)->ip_blkno, inode->i_nlink); mlog_bug_on_msg(OCFS2_SB(inode->i_sb) == NULL, "Inode=%lu\n", inode->i_ino); @@ -919,8 +924,8 @@ void ocfs2_clear_inode(struct inode *inode) ocfs2_checkpoint_inode(inode); mlog_bug_on_msg(!list_empty(&oi->ip_io_markers), - "Clear inode of %"MLFu64", inode has io markers\n", - oi->ip_blkno); + "Clear inode of %llu, inode has io markers\n", + (unsigned long long)oi->ip_blkno); ocfs2_extent_map_drop(inode, 0); ocfs2_extent_map_init(inode); @@ -936,20 +941,20 @@ void ocfs2_clear_inode(struct inode *inode) ocfs2_metadata_cache_purge(inode); mlog_bug_on_msg(oi->ip_metadata_cache.ci_num_cached, - "Clear inode of %"MLFu64", inode has %u cache items\n", - oi->ip_blkno, oi->ip_metadata_cache.ci_num_cached); + "Clear inode of %llu, inode has %u cache items\n", + (unsigned long long)oi->ip_blkno, oi->ip_metadata_cache.ci_num_cached); mlog_bug_on_msg(!(oi->ip_flags & OCFS2_INODE_CACHE_INLINE), - "Clear inode of %"MLFu64", inode has a bad flag\n", - oi->ip_blkno); + "Clear inode of %llu, inode has a bad flag\n", + (unsigned long long)oi->ip_blkno); mlog_bug_on_msg(spin_is_locked(&oi->ip_lock), - "Clear inode of %"MLFu64", inode is locked\n", - oi->ip_blkno); + "Clear inode of %llu, inode is locked\n", + (unsigned long long)oi->ip_blkno); mlog_bug_on_msg(!mutex_trylock(&oi->ip_io_mutex), - "Clear inode of %"MLFu64", io_mutex is locked\n", - oi->ip_blkno); + "Clear inode of %llu, io_mutex is locked\n", + (unsigned long long)oi->ip_blkno); mutex_unlock(&oi->ip_io_mutex); /* @@ -957,19 +962,19 @@ void ocfs2_clear_inode(struct inode *inode) * kernel 1, world 0 */ mlog_bug_on_msg(!down_write_trylock(&oi->ip_alloc_sem), - "Clear inode of %"MLFu64", alloc_sem is locked\n", - oi->ip_blkno); + "Clear inode of %llu, alloc_sem is locked\n", + (unsigned long long)oi->ip_blkno); up_write(&oi->ip_alloc_sem); mlog_bug_on_msg(oi->ip_open_count, - "Clear inode of %"MLFu64" has open count %d\n", - oi->ip_blkno, oi->ip_open_count); + "Clear inode of %llu has open count %d\n", + (unsigned long long)oi->ip_blkno, oi->ip_open_count); mlog_bug_on_msg(!list_empty(&oi->ip_handle_list), - "Clear inode of %"MLFu64" has non empty handle list\n", - oi->ip_blkno); + "Clear inode of %llu has non empty handle list\n", + (unsigned long long)oi->ip_blkno); mlog_bug_on_msg(oi->ip_handle, - "Clear inode of %"MLFu64" has non empty handle pointer\n", - oi->ip_blkno); + "Clear inode of %llu has non empty handle pointer\n", + (unsigned long long)oi->ip_blkno); /* Clear all other flags. */ oi->ip_flags = OCFS2_INODE_CACHE_INLINE; @@ -991,8 +996,8 @@ void ocfs2_drop_inode(struct inode *inode) mlog_entry_void(); - mlog(0, "Drop inode %"MLFu64", nlink = %u, ip_flags = 0x%x\n", - oi->ip_blkno, inode->i_nlink, oi->ip_flags); + mlog(0, "Drop inode %llu, nlink = %u, ip_flags = 0x%x\n", + (unsigned long long)oi->ip_blkno, inode->i_nlink, oi->ip_flags); /* Testing ip_orphaned_slot here wouldn't work because we may * not have gotten a delete_inode vote from any other nodes @@ -1069,8 +1074,8 @@ int ocfs2_inode_revalidate(struct dentry *dentry) struct inode *inode = dentry->d_inode; int status = 0; - mlog_entry("(inode = 0x%p, ino = %"MLFu64")\n", inode, - inode ? OCFS2_I(inode)->ip_blkno : 0ULL); + mlog_entry("(inode = 0x%p, ino = %llu)\n", inode, + inode ? (unsigned long long)OCFS2_I(inode)->ip_blkno : 0ULL); if (!inode) { mlog(0, "eep, no inode!\n"); @@ -1114,7 +1119,8 @@ int ocfs2_mark_inode_dirty(struct ocfs2_journal_handle *handle, int status; struct ocfs2_dinode *fe = (struct ocfs2_dinode *) bh->b_data; - mlog_entry("(inode %"MLFu64")\n", OCFS2_I(inode)->ip_blkno); + mlog_entry("(inode %llu)\n", + (unsigned long long)OCFS2_I(inode)->ip_blkno); status = ocfs2_journal_access(handle, inode, bh, OCFS2_JOURNAL_ACCESS_WRITE); diff --git a/fs/ocfs2/journal.c b/fs/ocfs2/journal.c index 4be801f4559..6a610ae5358 100644 --- a/fs/ocfs2/journal.c +++ b/fs/ocfs2/journal.c @@ -377,7 +377,7 @@ int ocfs2_journal_access(struct ocfs2_journal_handle *handle, BUG_ON(!bh); BUG_ON(!(handle->flags & OCFS2_HANDLE_STARTED)); - mlog_entry("bh->b_blocknr=%llu, type=%d (\"%s\"), bh->b_size = %hu\n", + mlog_entry("bh->b_blocknr=%llu, type=%d (\"%s\"), bh->b_size = %zu\n", (unsigned long long)bh->b_blocknr, type, (type == OCFS2_JOURNAL_ACCESS_CREATE) ? "OCFS2_JOURNAL_ACCESS_CREATE" : @@ -503,8 +503,8 @@ static void ocfs2_handle_cleanup_locks(struct ocfs2_journal *journal, ocfs2_meta_unlock(inode, 1); if (atomic_read(&inode->i_count) == 1) mlog(ML_ERROR, - "Inode %"MLFu64", I'm doing a last iput for!", - OCFS2_I(inode)->ip_blkno); + "Inode %llu, I'm doing a last iput for!", + (unsigned long long)OCFS2_I(inode)->ip_blkno); iput(inode); kmem_cache_free(ocfs2_lock_cache, lock); } @@ -582,7 +582,8 @@ int ocfs2_journal_init(struct ocfs2_journal *journal, int *dirty) } mlog(0, "inode->i_size = %lld\n", inode->i_size); - mlog(0, "inode->i_blocks = %lu\n", inode->i_blocks); + mlog(0, "inode->i_blocks = %llu\n", + (unsigned long long)inode->i_blocks); mlog(0, "inode->ip_clusters = %u\n", OCFS2_I(inode)->ip_clusters); /* call the kernels journal init function now */ @@ -640,8 +641,9 @@ static int ocfs2_journal_toggle_dirty(struct ocfs2_super *osb, /* This is called from startup/shutdown which will * handle the errors in a specific manner, so no need * to call ocfs2_error() here. */ - mlog(ML_ERROR, "Journal dinode %"MLFu64" has invalid " - "signature: %.*s", fe->i_blkno, 7, fe->i_signature); + mlog(ML_ERROR, "Journal dinode %llu has invalid " + "signature: %.*s", (unsigned long long)fe->i_blkno, 7, + fe->i_signature); status = -EIO; goto out; } @@ -849,8 +851,9 @@ static int ocfs2_force_read_journal(struct inode *inode) memset(bhs, 0, sizeof(struct buffer_head *) * CONCURRENT_JOURNAL_FILL); - mlog(0, "Force reading %lu blocks\n", - (inode->i_blocks >> (inode->i_sb->s_blocksize_bits - 9))); + mlog(0, "Force reading %llu blocks\n", + (unsigned long long)(inode->i_blocks >> + (inode->i_sb->s_blocksize_bits - 9))); v_blkno = 0; while (v_blkno < @@ -934,8 +937,8 @@ void ocfs2_complete_recovery(void *data) la_dinode = item->lri_la_dinode; if (la_dinode) { - mlog(0, "Clean up local alloc %"MLFu64"\n", - la_dinode->i_blkno); + mlog(0, "Clean up local alloc %llu\n", + (unsigned long long)la_dinode->i_blkno); ret = ocfs2_complete_local_alloc_recovery(osb, la_dinode); @@ -947,8 +950,8 @@ void ocfs2_complete_recovery(void *data) tl_dinode = item->lri_tl_dinode; if (tl_dinode) { - mlog(0, "Clean up truncate log %"MLFu64"\n", - tl_dinode->i_blkno); + mlog(0, "Clean up truncate log %llu\n", + (unsigned long long)tl_dinode->i_blkno); ret = ocfs2_complete_truncate_log_recovery(osb, tl_dinode); @@ -1473,11 +1476,11 @@ static int ocfs2_queue_orphans(struct ocfs2_super *osb, if (de->file_type > OCFS2_FT_MAX) { mlog(ML_ERROR, "block %llu contains invalid de: " - "inode = %"MLFu64", rec_len = %u, " + "inode = %llu, rec_len = %u, " "name_len = %u, file_type = %u, " "name='%.*s'\n", (unsigned long long)bh->b_blocknr, - le64_to_cpu(de->inode), + (unsigned long long)le64_to_cpu(de->inode), le16_to_cpu(de->rec_len), de->name_len, de->file_type, @@ -1494,8 +1497,8 @@ static int ocfs2_queue_orphans(struct ocfs2_super *osb, if (IS_ERR(iter)) continue; - mlog(0, "queue orphan %"MLFu64"\n", - OCFS2_I(iter)->ip_blkno); + mlog(0, "queue orphan %llu\n", + (unsigned long long)OCFS2_I(iter)->ip_blkno); /* No locking is required for the next_orphan * queue as there is only ever a single * process doing orphan recovery. */ @@ -1588,7 +1591,7 @@ static int ocfs2_recover_orphans(struct ocfs2_super *osb, while (inode) { oi = OCFS2_I(inode); - mlog(0, "iput orphan %"MLFu64"\n", oi->ip_blkno); + mlog(0, "iput orphan %llu\n", (unsigned long long)oi->ip_blkno); iter = oi->ip_next_orphan; diff --git a/fs/ocfs2/localalloc.c b/fs/ocfs2/localalloc.c index 149b3518166..0d1973ea32b 100644 --- a/fs/ocfs2/localalloc.c +++ b/fs/ocfs2/localalloc.c @@ -143,8 +143,8 @@ int ocfs2_load_local_alloc(struct ocfs2_super *osb) if (!(le32_to_cpu(alloc->i_flags) & (OCFS2_LOCAL_ALLOC_FL|OCFS2_BITMAP_FL))) { - mlog(ML_ERROR, "Invalid local alloc inode, %"MLFu64"\n", - OCFS2_I(inode)->ip_blkno); + mlog(ML_ERROR, "Invalid local alloc inode, %llu\n", + (unsigned long long)OCFS2_I(inode)->ip_blkno); status = -EINVAL; goto bail; } @@ -493,9 +493,9 @@ int ocfs2_reserve_local_alloc_bits(struct ocfs2_super *osb, if (le32_to_cpu(alloc->id1.bitmap1.i_used) != ocfs2_local_alloc_count_bits(alloc)) { - ocfs2_error(osb->sb, "local alloc inode %"MLFu64" says it has " + ocfs2_error(osb->sb, "local alloc inode %llu says it has " "%u free bits, but a count shows %u", - le64_to_cpu(alloc->i_blkno), + (unsigned long long)le64_to_cpu(alloc->i_blkno), le32_to_cpu(alloc->id1.bitmap1.i_used), ocfs2_local_alloc_count_bits(alloc)); status = -EIO; @@ -753,10 +753,11 @@ static int ocfs2_sync_local_to_main(struct ocfs2_super *osb, ocfs2_clusters_to_blocks(osb->sb, start - count); - mlog(0, "freeing %u bits starting at local " - "alloc bit %u (la_start_blk = %"MLFu64", " - "blkno = %"MLFu64")\n", count, start - count, - la_start_blk, blkno); + mlog(0, "freeing %u bits starting at local alloc bit " + "%u (la_start_blk = %llu, blkno = %llu)\n", + count, start - count, + (unsigned long long)la_start_blk, + (unsigned long long)blkno); status = ocfs2_free_clusters(handle, main_bm_inode, main_bm_bh, blkno, count); diff --git a/fs/ocfs2/namei.c b/fs/ocfs2/namei.c index f6b77ff1d2b..0673862c8bd 100644 --- a/fs/ocfs2/namei.c +++ b/fs/ocfs2/namei.c @@ -161,8 +161,8 @@ static struct dentry *ocfs2_lookup(struct inode *dir, struct dentry *dentry, goto bail; } - mlog(0, "find name %.*s in directory %"MLFu64"\n", dentry->d_name.len, - dentry->d_name.name, OCFS2_I(dir)->ip_blkno); + mlog(0, "find name %.*s in directory %llu\n", dentry->d_name.len, + dentry->d_name.name, (unsigned long long)OCFS2_I(dir)->ip_blkno); status = ocfs2_meta_lock(dir, NULL, NULL, 0); if (status < 0) { @@ -180,7 +180,8 @@ static struct dentry *ocfs2_lookup(struct inode *dir, struct dentry *dentry, inode = ocfs2_iget(OCFS2_SB(dir->i_sb), blkno); if (IS_ERR(inode)) { - mlog(ML_ERROR, "Unable to create inode %"MLFu64"\n", blkno); + mlog(ML_ERROR, "Unable to create inode %llu\n", + (unsigned long long)blkno); ret = ERR_PTR(-EACCES); goto bail_unlock; } @@ -310,8 +311,8 @@ static int ocfs2_mknod(struct inode *dir, osb = OCFS2_SB(dir->i_sb); if (S_ISDIR(mode) && (dir->i_nlink >= OCFS2_LINK_MAX)) { - mlog(ML_ERROR, "inode %"MLFu64" has i_nlink of %u\n", - OCFS2_I(dir)->ip_blkno, dir->i_nlink); + mlog(ML_ERROR, "inode %llu has i_nlink of %u\n", + (unsigned long long)OCFS2_I(dir)->ip_blkno, dir->i_nlink); status = -EMLINK; goto leave; } @@ -562,9 +563,9 @@ static int ocfs2_mknod_locked(struct ocfs2_super *osb, if (ocfs2_populate_inode(inode, fe, 1) < 0) { mlog(ML_ERROR, "populate inode failed! bh->b_blocknr=%llu, " - "i_blkno=%"MLFu64", i_ino=%lu\n", + "i_blkno=%llu, i_ino=%lu\n", (unsigned long long) (*new_fe_bh)->b_blocknr, - fe->i_blkno, inode->i_ino); + (unsigned long long)fe->i_blkno, inode->i_ino); BUG(); } @@ -765,7 +766,7 @@ static int ocfs2_unlink(struct inode *dir, BUG_ON(dentry->d_parent->d_inode != dir); - mlog(0, "ino = %"MLFu64"\n", OCFS2_I(inode)->ip_blkno); + mlog(0, "ino = %llu\n", (unsigned long long)OCFS2_I(inode)->ip_blkno); if (inode == osb->root_inode) { mlog(0, "Cannot delete the root directory\n"); @@ -799,9 +800,9 @@ static int ocfs2_unlink(struct inode *dir, if (OCFS2_I(inode)->ip_blkno != blkno) { status = -ENOENT; - mlog(0, "ip_blkno (%"MLFu64") != dirent blkno (%"MLFu64") " - "ip_flags = %x\n", OCFS2_I(inode)->ip_blkno, blkno, - OCFS2_I(inode)->ip_flags); + mlog(0, "ip_blkno %llu != dirent blkno %llu ip_flags = %x\n", + (unsigned long long)OCFS2_I(inode)->ip_blkno, + (unsigned long long)blkno, OCFS2_I(inode)->ip_flags); goto leave; } @@ -946,8 +947,9 @@ static int ocfs2_double_lock(struct ocfs2_super *osb, struct buffer_head **tmpbh; struct inode *tmpinode; - mlog_entry("(inode1 = %"MLFu64", inode2 = %"MLFu64")\n", - oi1->ip_blkno, oi2->ip_blkno); + mlog_entry("(inode1 = %llu, inode2 = %llu)\n", + (unsigned long long)oi1->ip_blkno, + (unsigned long long)oi2->ip_blkno); BUG_ON(!handle); @@ -1187,9 +1189,9 @@ static int ocfs2_rename(struct inode *old_dir, if (OCFS2_I(new_inode)->ip_blkno != newfe_blkno) { status = -EACCES; - mlog(0, "Inode blkno (%"MLFu64") and dir (%"MLFu64") " - "disagree. ip_flags = %x\n", - OCFS2_I(new_inode)->ip_blkno, newfe_blkno, + mlog(0, "Inode %llu and dir %llu disagree. flags = %x\n", + (unsigned long long)OCFS2_I(new_inode)->ip_blkno, + (unsigned long long)newfe_blkno, OCFS2_I(new_inode)->ip_flags); goto bail; } @@ -1215,9 +1217,9 @@ static int ocfs2_rename(struct inode *old_dir, newfe = (struct ocfs2_dinode *) newfe_bh->b_data; - mlog(0, "aha rename over existing... new_de=%p " - "new_blkno=%"MLFu64" newfebh=%p bhblocknr=%llu\n", - new_de, newfe_blkno, newfe_bh, newfe_bh ? + mlog(0, "aha rename over existing... new_de=%p new_blkno=%llu " + "newfebh=%p bhblocknr=%llu\n", new_de, + (unsigned long long)newfe_blkno, newfe_bh, newfe_bh ? (unsigned long long)newfe_bh->b_blocknr : 0ULL); if (S_ISDIR(new_inode->i_mode) || (new_inode->i_nlink == 1)) { @@ -1354,8 +1356,8 @@ static int ocfs2_rename(struct inode *old_dir, if (new_dir_nlink != new_dir->i_nlink) { if (!new_dir_bh) { mlog(ML_ERROR, "need to change nlink for new " - "dir %"MLFu64" from %d to %d but bh is " - "NULL\n", OCFS2_I(new_dir)->ip_blkno, + "dir %llu from %d to %d but bh is NULL\n", + (unsigned long long)OCFS2_I(new_dir)->ip_blkno, (int)new_dir_nlink, new_dir->i_nlink); } else { struct ocfs2_dinode *fe; @@ -1372,10 +1374,9 @@ static int ocfs2_rename(struct inode *old_dir, if (old_dir_nlink != old_dir->i_nlink) { if (!old_dir_bh) { mlog(ML_ERROR, "need to change nlink for old dir " - "%"MLFu64" from %d to %d but bh is NULL!\n", - OCFS2_I(old_dir)->ip_blkno, - (int)old_dir_nlink, - old_dir->i_nlink); + "%llu from %d to %d but bh is NULL!\n", + (unsigned long long)OCFS2_I(old_dir)->ip_blkno, + (int)old_dir_nlink, old_dir->i_nlink); } else { struct ocfs2_dinode *fe; status = ocfs2_journal_access(handle, old_dir, @@ -1443,8 +1444,9 @@ static int ocfs2_create_symlink_data(struct ocfs2_super *osb, * write i_size + 1 bytes. */ blocks = (bytes_left + sb->s_blocksize - 1) >> sb->s_blocksize_bits; - mlog_entry("i_blocks = %lu, i_size = %llu, blocks = %d\n", - inode->i_blocks, i_size_read(inode), blocks); + mlog_entry("i_blocks = %llu, i_size = %llu, blocks = %d\n", + (unsigned long long)inode->i_blocks, + i_size_read(inode), blocks); /* Sanity check -- make sure we're going to fit. */ if (bytes_left > @@ -1634,9 +1636,9 @@ static int ocfs2_symlink(struct inode *dir, NULL); if (status < 0) { if (status != -ENOSPC && status != -EINTR) { - mlog(ML_ERROR, "Failed to extend file to " - "%"MLFu64"\n", - newsize); + mlog(ML_ERROR, + "Failed to extend file to %llu\n", + (unsigned long long)newsize); mlog_errno(status); status = -ENOSPC; } @@ -1716,10 +1718,11 @@ int ocfs2_check_dir_entry(struct inode * dir, error_msg = "directory entry across blocks"; if (error_msg != NULL) - mlog(ML_ERROR, "bad entry in directory #%"MLFu64": %s - " - "offset=%lu, inode=%"MLFu64", rec_len=%d, name_len=%d\n", - OCFS2_I(dir)->ip_blkno, error_msg, offset, - le64_to_cpu(de->inode), rlen, de->name_len); + mlog(ML_ERROR, "bad entry in directory #%llu: %s - " + "offset=%lu, inode=%llu, rec_len=%d, name_len=%d\n", + (unsigned long long)OCFS2_I(dir)->ip_blkno, error_msg, + offset, (unsigned long long)le64_to_cpu(de->inode), rlen, + de->name_len); return error_msg == NULL ? 1 : 0; } @@ -2021,8 +2024,8 @@ static int ocfs2_blkno_stringify(u64 blkno, char *name) mlog_entry_void(); - namelen = snprintf(name, OCFS2_ORPHAN_NAMELEN + 1, "%016"MLFx64, - blkno); + namelen = snprintf(name, OCFS2_ORPHAN_NAMELEN + 1, "%016llx", + (long long)blkno); if (namelen <= 0) { if (namelen) status = namelen; @@ -2167,8 +2170,8 @@ static int ocfs2_orphan_add(struct ocfs2_super *osb, OCFS2_I(inode)->ip_orphaned_slot = osb->slot_num; spin_unlock(&OCFS2_I(inode)->ip_lock); - mlog(0, "Inode %"MLFu64" orphaned in slot %d\n", - OCFS2_I(inode)->ip_blkno, osb->slot_num); + mlog(0, "Inode %llu orphaned in slot %d\n", + (unsigned long long)OCFS2_I(inode)->ip_blkno, osb->slot_num); leave: if (orphan_dir_inode) @@ -2202,8 +2205,9 @@ int ocfs2_orphan_del(struct ocfs2_super *osb, goto leave; } - mlog(0, "removing '%s' from orphan dir %"MLFu64" (namelen=%d)\n", - name, OCFS2_I(orphan_dir_inode)->ip_blkno, OCFS2_ORPHAN_NAMELEN); + mlog(0, "removing '%s' from orphan dir %llu (namelen=%d)\n", + name, (unsigned long long)OCFS2_I(orphan_dir_inode)->ip_blkno, + OCFS2_ORPHAN_NAMELEN); /* find it's spot in the orphan directory */ target_de_bh = ocfs2_find_entry(name, OCFS2_ORPHAN_NAMELEN, diff --git a/fs/ocfs2/ocfs2.h b/fs/ocfs2/ocfs2.h index e89de9b6e49..da1093039c0 100644 --- a/fs/ocfs2/ocfs2.h +++ b/fs/ocfs2/ocfs2.h @@ -357,8 +357,8 @@ static inline int ocfs2_is_soft_readonly(struct ocfs2_super *osb) #define OCFS2_RO_ON_INVALID_DINODE(__sb, __di) do { \ typeof(__di) ____di = (__di); \ ocfs2_error((__sb), \ - "Dinode # %"MLFu64" has bad signature %.*s", \ - (____di)->i_blkno, 7, \ + "Dinode # %llu has bad signature %.*s", \ + (unsigned long long)(____di)->i_blkno, 7, \ (____di)->i_signature); \ } while (0); @@ -368,8 +368,8 @@ static inline int ocfs2_is_soft_readonly(struct ocfs2_super *osb) #define OCFS2_RO_ON_INVALID_EXTENT_BLOCK(__sb, __eb) do { \ typeof(__eb) ____eb = (__eb); \ ocfs2_error((__sb), \ - "Extent Block # %"MLFu64" has bad signature %.*s", \ - (____eb)->h_blkno, 7, \ + "Extent Block # %llu has bad signature %.*s", \ + (unsigned long long)(____eb)->h_blkno, 7, \ (____eb)->h_signature); \ } while (0); @@ -379,8 +379,8 @@ static inline int ocfs2_is_soft_readonly(struct ocfs2_super *osb) #define OCFS2_RO_ON_INVALID_GROUP_DESC(__sb, __gd) do { \ typeof(__gd) ____gd = (__gd); \ ocfs2_error((__sb), \ - "Group Descriptor # %"MLFu64" has bad signature %.*s", \ - (____gd)->bg_blkno, 7, \ + "Group Descriptor # %llu has bad signature %.*s", \ + (unsigned long long)(____gd)->bg_blkno, 7, \ (____gd)->bg_signature); \ } while (0); diff --git a/fs/ocfs2/suballoc.c b/fs/ocfs2/suballoc.c index c46c164aefb..195523090c8 100644 --- a/fs/ocfs2/suballoc.c +++ b/fs/ocfs2/suballoc.c @@ -157,8 +157,9 @@ static int ocfs2_block_group_fill(struct ocfs2_journal_handle *handle, mlog_entry_void(); if (((unsigned long long) bg_bh->b_blocknr) != group_blkno) { - ocfs2_error(alloc_inode->i_sb, "group block (%"MLFu64") " - "!= b_blocknr (%llu)", group_blkno, + ocfs2_error(alloc_inode->i_sb, "group block (%llu) != " + "b_blocknr (%llu)", + (unsigned long long)group_blkno, (unsigned long long) bg_bh->b_blocknr); status = -EIO; goto bail; @@ -280,8 +281,8 @@ static int ocfs2_block_group_alloc(struct ocfs2_super *osb, /* setup the group */ bg_blkno = ocfs2_clusters_to_blocks(osb->sb, bit_off); - mlog(0, "new descriptor, record %u, at block %"MLFu64"\n", - alloc_rec, bg_blkno); + mlog(0, "new descriptor, record %u, at block %llu\n", + alloc_rec, (unsigned long long)bg_blkno); bg_bh = sb_getblk(osb->sb, bg_blkno); if (!bg_bh) { @@ -382,8 +383,8 @@ static int ocfs2_reserve_suballoc_bits(struct ocfs2_super *osb, goto bail; } if (!(fe->i_flags & cpu_to_le32(OCFS2_CHAIN_FL))) { - ocfs2_error(alloc_inode->i_sb, "Invalid chain allocator " - "# %"MLFu64, le64_to_cpu(fe->i_blkno)); + ocfs2_error(alloc_inode->i_sb, "Invalid chain allocator %llu", + (unsigned long long)le64_to_cpu(fe->i_blkno)); status = -EIO; goto bail; } @@ -829,9 +830,10 @@ static int ocfs2_relink_block_group(struct ocfs2_journal_handle *handle, goto out; } - mlog(0, "In suballoc %"MLFu64", chain %u, move group %"MLFu64" to " - "top, prev = %"MLFu64"\n", - fe->i_blkno, chain, bg->bg_blkno, prev_bg->bg_blkno); + mlog(0, "Suballoc %llu, chain %u, move group %llu to top, prev = %llu\n", + (unsigned long long)fe->i_blkno, chain, + (unsigned long long)bg->bg_blkno, + (unsigned long long)prev_bg->bg_blkno); fe_ptr = le64_to_cpu(fe->id2.i_chain.cl_recs[chain].c_blkno); bg_ptr = le64_to_cpu(bg->bg_next_group); @@ -974,8 +976,9 @@ static int ocfs2_search_chain(struct ocfs2_alloc_context *ac, struct ocfs2_group_desc *bg; chain = ac->ac_chain; - mlog(0, "trying to alloc %u bits from chain %u, inode %"MLFu64"\n", - bits_wanted, chain, OCFS2_I(alloc_inode)->ip_blkno); + mlog(0, "trying to alloc %u bits from chain %u, inode %llu\n", + bits_wanted, chain, + (unsigned long long)OCFS2_I(alloc_inode)->ip_blkno); status = ocfs2_read_block(OCFS2_SB(alloc_inode->i_sb), le64_to_cpu(cl->cl_recs[chain].c_blkno), @@ -1027,8 +1030,8 @@ static int ocfs2_search_chain(struct ocfs2_alloc_context *ac, goto bail; } - mlog(0, "alloc succeeds: we give %u bits from block group %"MLFu64"\n", - tmp_bits, bg->bg_blkno); + mlog(0, "alloc succeeds: we give %u bits from block group %llu\n", + tmp_bits, (unsigned long long)bg->bg_blkno); *num_bits = tmp_bits; @@ -1092,8 +1095,8 @@ static int ocfs2_search_chain(struct ocfs2_alloc_context *ac, goto bail; } - mlog(0, "Allocated %u bits from suballocator %"MLFu64"\n", - *num_bits, fe->i_blkno); + mlog(0, "Allocated %u bits from suballocator %llu\n", *num_bits, + (unsigned long long)fe->i_blkno); *bg_blkno = le64_to_cpu(bg->bg_blkno); bail: @@ -1134,9 +1137,9 @@ static int ocfs2_claim_suballoc_bits(struct ocfs2_super *osb, } if (le32_to_cpu(fe->id1.bitmap1.i_used) >= le32_to_cpu(fe->id1.bitmap1.i_total)) { - ocfs2_error(osb->sb, "Chain allocator dinode %"MLFu64" has %u" - "used bits but only %u total.", - le64_to_cpu(fe->i_blkno), + ocfs2_error(osb->sb, "Chain allocator dinode %llu has %u used " + "bits but only %u total.", + (unsigned long long)le64_to_cpu(fe->i_blkno), le32_to_cpu(fe->id1.bitmap1.i_used), le32_to_cpu(fe->id1.bitmap1.i_total)); status = -EIO; @@ -1479,10 +1482,9 @@ static int ocfs2_free_suballoc_bits(struct ocfs2_journal_handle *handle, } BUG_ON((count + start_bit) > ocfs2_bits_per_group(cl)); - mlog(0, "suballocator %"MLFu64": freeing %u bits from group %"MLFu64 - ", starting at %u\n", - OCFS2_I(alloc_inode)->ip_blkno, count, bg_blkno, - start_bit); + mlog(0, "%llu: freeing %u bits from group %llu, starting at %u\n", + (unsigned long long)OCFS2_I(alloc_inode)->ip_blkno, count, + (unsigned long long)bg_blkno, start_bit); status = ocfs2_read_block(osb, bg_blkno, &group_bh, OCFS2_BH_CACHED, alloc_inode); @@ -1592,10 +1594,10 @@ int ocfs2_free_clusters(struct ocfs2_journal_handle *handle, ocfs2_block_to_cluster_group(bitmap_inode, start_blk, &bg_blkno, &bg_start_bit); - mlog(0, "want to free %u clusters starting at block %"MLFu64"\n", - num_clusters, start_blk); - mlog(0, "bg_blkno = %"MLFu64", bg_start_bit = %u\n", - bg_blkno, bg_start_bit); + mlog(0, "want to free %u clusters starting at block %llu\n", + num_clusters, (unsigned long long)start_blk); + mlog(0, "bg_blkno = %llu, bg_start_bit = %u\n", + (unsigned long long)bg_blkno, bg_start_bit); status = ocfs2_free_suballoc_bits(handle, bitmap_inode, bitmap_bh, bg_start_bit, bg_blkno, @@ -1616,18 +1618,22 @@ static inline void ocfs2_debug_bg(struct ocfs2_group_desc *bg) printk("bg_free_bits_count: %u\n", bg->bg_free_bits_count); printk("bg_chain: %u\n", bg->bg_chain); printk("bg_generation: %u\n", le32_to_cpu(bg->bg_generation)); - printk("bg_next_group: %"MLFu64"\n", bg->bg_next_group); - printk("bg_parent_dinode: %"MLFu64"\n", bg->bg_parent_dinode); - printk("bg_blkno: %"MLFu64"\n", bg->bg_blkno); + printk("bg_next_group: %llu\n", + (unsigned long long)bg->bg_next_group); + printk("bg_parent_dinode: %llu\n", + (unsigned long long)bg->bg_parent_dinode); + printk("bg_blkno: %llu\n", + (unsigned long long)bg->bg_blkno); } static inline void ocfs2_debug_suballoc_inode(struct ocfs2_dinode *fe) { int i; - printk("Suballoc Inode %"MLFu64":\n", fe->i_blkno); + printk("Suballoc Inode %llu:\n", (unsigned long long)fe->i_blkno); printk("i_signature: %s\n", fe->i_signature); - printk("i_size: %"MLFu64"\n", fe->i_size); + printk("i_size: %llu\n", + (unsigned long long)fe->i_size); printk("i_clusters: %u\n", fe->i_clusters); printk("i_generation: %u\n", le32_to_cpu(fe->i_generation)); @@ -1645,7 +1651,7 @@ static inline void ocfs2_debug_suballoc_inode(struct ocfs2_dinode *fe) fe->id2.i_chain.cl_recs[i].c_free); printk("fe->id2.i_chain.cl_recs[%d].c_total: %u\n", i, fe->id2.i_chain.cl_recs[i].c_total); - printk("fe->id2.i_chain.cl_recs[%d].c_blkno: %"MLFu64"\n", i, - fe->id2.i_chain.cl_recs[i].c_blkno); + printk("fe->id2.i_chain.cl_recs[%d].c_blkno: %llu\n", i, + (unsigned long long)fe->id2.i_chain.cl_recs[i].c_blkno); } } diff --git a/fs/ocfs2/super.c b/fs/ocfs2/super.c index 44d8b524823..949b3dac30f 100644 --- a/fs/ocfs2/super.c +++ b/fs/ocfs2/super.c @@ -1428,8 +1428,9 @@ static int ocfs2_initialize_super(struct super_block *sb, osb->fs_generation = le32_to_cpu(di->i_fs_generation); mlog(0, "vol_label: %s\n", osb->vol_label); mlog(0, "uuid: %s\n", osb->uuid_str); - mlog(0, "root_blkno=%"MLFu64", system_dir_blkno=%"MLFu64"\n", - osb->root_blkno, osb->system_dir_blkno); + mlog(0, "root_blkno=%llu, system_dir_blkno=%llu\n", + (unsigned long long)osb->root_blkno, + (unsigned long long)osb->system_dir_blkno); osb->osb_dlm_debug = ocfs2_new_dlm_debug(); if (!osb->osb_dlm_debug) { @@ -1472,8 +1473,8 @@ static int ocfs2_initialize_super(struct super_block *sb, osb->bitmap_cpg = le16_to_cpu(di->id2.i_chain.cl_cpg); osb->num_clusters = le32_to_cpu(di->id1.bitmap1.i_total); brelse(bitmap_bh); - mlog(0, "cluster bitmap inode: %"MLFu64", clusters per group: %u\n", - osb->bitmap_blkno, osb->bitmap_cpg); + mlog(0, "cluster bitmap inode: %llu, clusters per group: %u\n", + (unsigned long long)osb->bitmap_blkno, osb->bitmap_cpg); status = ocfs2_init_slot_info(osb); if (status < 0) { @@ -1531,8 +1532,9 @@ static int ocfs2_verify_volume(struct ocfs2_dinode *di, OCFS2_MINOR_REV_LEVEL); } else if (bh->b_blocknr != le64_to_cpu(di->i_blkno)) { mlog(ML_ERROR, "bad block number on superblock: " - "found %"MLFu64", should be %llu\n", - di->i_blkno, (unsigned long long)bh->b_blocknr); + "found %llu, should be %llu\n", + (unsigned long long)di->i_blkno, + (unsigned long long)bh->b_blocknr); } else if (le32_to_cpu(di->id2.i_super.s_clustersize_bits) < 12 || le32_to_cpu(di->id2.i_super.s_clustersize_bits) > 20) { mlog(ML_ERROR, "bad cluster size found: %u\n", diff --git a/fs/ocfs2/super.h b/fs/ocfs2/super.h index c564177dfbd..783f5270f2a 100644 --- a/fs/ocfs2/super.h +++ b/fs/ocfs2/super.h @@ -33,12 +33,16 @@ int ocfs2_publish_get_mount_state(struct ocfs2_super *osb, void __ocfs2_error(struct super_block *sb, const char *function, - const char *fmt, ...); + const char *fmt, ...) + __attribute__ ((format (printf, 3, 4))); + #define ocfs2_error(sb, fmt, args...) __ocfs2_error(sb, __PRETTY_FUNCTION__, fmt, ##args) void __ocfs2_abort(struct super_block *sb, const char *function, - const char *fmt, ...); + const char *fmt, ...) + __attribute__ ((format (printf, 3, 4))); + #define ocfs2_abort(sb, fmt, args...) __ocfs2_abort(sb, __PRETTY_FUNCTION__, fmt, ##args) #endif /* OCFS2_SUPER_H */ diff --git a/fs/ocfs2/uptodate.c b/fs/ocfs2/uptodate.c index 300b5bedfb2..04a684dfdd9 100644 --- a/fs/ocfs2/uptodate.c +++ b/fs/ocfs2/uptodate.c @@ -119,8 +119,8 @@ void ocfs2_metadata_cache_purge(struct inode *inode) tree = !(oi->ip_flags & OCFS2_INODE_CACHE_INLINE); to_purge = ci->ci_num_cached; - mlog(0, "Purge %u %s items from Inode %"MLFu64"\n", to_purge, - tree ? "array" : "tree", oi->ip_blkno); + mlog(0, "Purge %u %s items from Inode %llu\n", to_purge, + tree ? "array" : "tree", (unsigned long long)oi->ip_blkno); /* If we're a tree, save off the root so that we can safely * initialize the cache. We do the work to free tree members @@ -136,8 +136,8 @@ void ocfs2_metadata_cache_purge(struct inode *inode) * easily detect counting errors. Unfortunately, this is only * meaningful for trees. */ if (tree && purged != to_purge) - mlog(ML_ERROR, "Inode %"MLFu64", count = %u, purged = %u\n", - oi->ip_blkno, to_purge, purged); + mlog(ML_ERROR, "Inode %llu, count = %u, purged = %u\n", + (unsigned long long)oi->ip_blkno, to_purge, purged); } /* Returns the index in the cache array, -1 if not found. @@ -186,8 +186,9 @@ static int ocfs2_buffer_cached(struct ocfs2_inode_info *oi, spin_lock(&oi->ip_lock); - mlog(0, "Inode %"MLFu64", query block %llu (inline = %u)\n", - oi->ip_blkno, (unsigned long long) bh->b_blocknr, + mlog(0, "Inode %llu, query block %llu (inline = %u)\n", + (unsigned long long)oi->ip_blkno, + (unsigned long long) bh->b_blocknr, !!(oi->ip_flags & OCFS2_INODE_CACHE_INLINE)); if (oi->ip_flags & OCFS2_INODE_CACHE_INLINE) @@ -293,12 +294,12 @@ static void ocfs2_expand_cache(struct ocfs2_inode_info *oi, struct ocfs2_caching_info *ci = &oi->ip_metadata_cache; mlog_bug_on_msg(ci->ci_num_cached != OCFS2_INODE_MAX_CACHE_ARRAY, - "Inode %"MLFu64", num cached = %u, should be %u\n", - oi->ip_blkno, ci->ci_num_cached, + "Inode %llu, num cached = %u, should be %u\n", + (unsigned long long)oi->ip_blkno, ci->ci_num_cached, OCFS2_INODE_MAX_CACHE_ARRAY); mlog_bug_on_msg(!(oi->ip_flags & OCFS2_INODE_CACHE_INLINE), - "Inode %"MLFu64" not marked as inline anymore!\n", - oi->ip_blkno); + "Inode %llu not marked as inline anymore!\n", + (unsigned long long)oi->ip_blkno); assert_spin_locked(&oi->ip_lock); /* Be careful to initialize the tree members *first* because @@ -316,8 +317,8 @@ static void ocfs2_expand_cache(struct ocfs2_inode_info *oi, tree[i] = NULL; } - mlog(0, "Expanded %"MLFu64" to a tree cache: flags 0x%x, num = %u\n", - oi->ip_blkno, oi->ip_flags, ci->ci_num_cached); + mlog(0, "Expanded %llu to a tree cache: flags 0x%x, num = %u\n", + (unsigned long long)oi->ip_blkno, oi->ip_flags, ci->ci_num_cached); } /* Slow path function - memory allocation is necessary. See the @@ -332,8 +333,9 @@ static void __ocfs2_set_buffer_uptodate(struct ocfs2_inode_info *oi, struct ocfs2_meta_cache_item *tree[OCFS2_INODE_MAX_CACHE_ARRAY] = { NULL, }; - mlog(0, "Inode %"MLFu64", block %llu, expand = %d\n", - oi->ip_blkno, (unsigned long long) block, expand_tree); + mlog(0, "Inode %llu, block %llu, expand = %d\n", + (unsigned long long)oi->ip_blkno, + (unsigned long long)block, expand_tree); new = kmem_cache_alloc(ocfs2_uptodate_cachep, GFP_KERNEL); if (!new) { @@ -414,8 +416,9 @@ void ocfs2_set_buffer_uptodate(struct inode *inode, if (ocfs2_buffer_cached(oi, bh)) return; - mlog(0, "Inode %"MLFu64", inserting block %llu\n", oi->ip_blkno, - (unsigned long long) bh->b_blocknr); + mlog(0, "Inode %llu, inserting block %llu\n", + (unsigned long long)oi->ip_blkno, + (unsigned long long)bh->b_blocknr); /* No need to recheck under spinlock - insertion is guarded by * ip_io_mutex */ @@ -504,8 +507,9 @@ void ocfs2_remove_from_cache(struct inode *inode, struct ocfs2_caching_info *ci = &oi->ip_metadata_cache; spin_lock(&oi->ip_lock); - mlog(0, "Inode %"MLFu64", remove %llu, items = %u, array = %u\n", - oi->ip_blkno, (unsigned long long) block, ci->ci_num_cached, + mlog(0, "Inode %llu, remove %llu, items = %u, array = %u\n", + (unsigned long long)oi->ip_blkno, + (unsigned long long) block, ci->ci_num_cached, oi->ip_flags & OCFS2_INODE_CACHE_INLINE); if (oi->ip_flags & OCFS2_INODE_CACHE_INLINE) { diff --git a/fs/ocfs2/vote.c b/fs/ocfs2/vote.c index 021978e0576..53049a20419 100644 --- a/fs/ocfs2/vote.c +++ b/fs/ocfs2/vote.c @@ -190,20 +190,21 @@ static int ocfs2_process_delete_request(struct inode *inode, OCFS2_INVALID_SLOT && OCFS2_I(inode)->ip_orphaned_slot != (*orphaned_slot), - "Inode %"MLFu64": This node thinks it's " + "Inode %llu: This node thinks it's " "orphaned in slot %d, messaged it's in %d\n", - OCFS2_I(inode)->ip_blkno, + (unsigned long long)OCFS2_I(inode)->ip_blkno, OCFS2_I(inode)->ip_orphaned_slot, *orphaned_slot); - mlog(0, "Setting orphaned slot for inode %"MLFu64" to %d\n", - OCFS2_I(inode)->ip_blkno, *orphaned_slot); + mlog(0, "Setting orphaned slot for inode %llu to %d\n", + (unsigned long long)OCFS2_I(inode)->ip_blkno, + *orphaned_slot); OCFS2_I(inode)->ip_orphaned_slot = *orphaned_slot; } else { - mlog(0, "Sending back orphaned slot %d for inode %"MLFu64"\n", + mlog(0, "Sending back orphaned slot %d for inode %llu\n", OCFS2_I(inode)->ip_orphaned_slot, - OCFS2_I(inode)->ip_blkno); + (unsigned long long)OCFS2_I(inode)->ip_blkno); *orphaned_slot = OCFS2_I(inode)->ip_orphaned_slot; } @@ -226,8 +227,8 @@ static int ocfs2_process_delete_request(struct inode *inode, } if (filemap_fdatawrite(inode->i_mapping)) { - mlog(ML_ERROR, "Could not sync inode %"MLFu64" for delete!\n", - OCFS2_I(inode)->ip_blkno); + mlog(ML_ERROR, "Could not sync inode %llu for delete!\n", + (unsigned long long)OCFS2_I(inode)->ip_blkno); goto done; } sync_mapping_buffers(inode->i_mapping); @@ -302,8 +303,8 @@ static void ocfs2_process_dentry_request(struct inode *inode, struct list_head *p; struct ocfs2_inode_info *oi = OCFS2_I(inode); - mlog(0, "parent %"MLFu64", namelen = %u, name = %.*s\n", parent_blkno, - namelen, namelen, name); + mlog(0, "parent %llu, namelen = %u, name = %.*s\n", + (unsigned long long)parent_blkno, namelen, namelen, name); spin_lock(&dcache_lock); @@ -370,9 +371,10 @@ static void ocfs2_process_vote(struct ocfs2_super *osb, if (request == OCFS2_VOTE_REQ_DELETE) orphaned_slot = be32_to_cpu(msg->md1.v_orphaned_slot); - mlog(0, "processing vote: request = %u, blkno = %"MLFu64", " + mlog(0, "processing vote: request = %u, blkno = %llu, " "generation = %u, node_num = %u, priv1 = %u\n", request, - blkno, generation, node_num, be32_to_cpu(msg->md1.v_generic1)); + (unsigned long long)blkno, generation, node_num, + be32_to_cpu(msg->md1.v_generic1)); if (!ocfs2_is_valid_vote_request(request)) { mlog(ML_ERROR, "Invalid vote request %d from node %u\n", @@ -419,11 +421,12 @@ static void ocfs2_process_vote(struct ocfs2_super *osb, * we had not found an inode in the first place. */ if (inode->i_generation != generation) { mlog(0, "generation passed %u != inode generation = %u, " - "ip_flags = %x, ip_blkno = %"MLFu64", msg %"MLFu64", " - "i_count = %u, message type = %u\n", - generation, inode->i_generation, OCFS2_I(inode)->ip_flags, - OCFS2_I(inode)->ip_blkno, blkno, - atomic_read(&inode->i_count), request); + "ip_flags = %x, ip_blkno = %llu, msg %llu, i_count = %u, " + "message type = %u\n", generation, inode->i_generation, + OCFS2_I(inode)->ip_flags, + (unsigned long long)OCFS2_I(inode)->ip_blkno, + (unsigned long long)blkno, atomic_read(&inode->i_count), + request); iput(inode); inode = NULL; goto respond; @@ -830,8 +833,9 @@ static void ocfs2_delete_response_cb(void *priv, orphaned_slot = be32_to_cpu(resp->r_orphaned_slot); node = be32_to_cpu(resp->r_hdr.h_node_num); - mlog(0, "node %d tells us that inode %"MLFu64" is orphaned in slot " - "%d\n", node, OCFS2_I(inode)->ip_blkno, orphaned_slot); + mlog(0, "node %d tells us that inode %llu is orphaned in slot %d\n", + node, (unsigned long long)OCFS2_I(inode)->ip_blkno, + orphaned_slot); /* The other node may not actually know which slot the inode * is orphaned in. */ @@ -845,9 +849,9 @@ static void ocfs2_delete_response_cb(void *priv, spin_lock(&OCFS2_I(inode)->ip_lock); mlog_bug_on_msg(OCFS2_I(inode)->ip_orphaned_slot != orphaned_slot && OCFS2_I(inode)->ip_orphaned_slot - != OCFS2_INVALID_SLOT, "Inode %"MLFu64": Node %d " - "says it's orphaned in slot %d, we think it's in %d\n", - OCFS2_I(inode)->ip_blkno, + != OCFS2_INVALID_SLOT, "Inode %llu: Node %d says it's " + "orphaned in slot %d, we think it's in %d\n", + (unsigned long long)OCFS2_I(inode)->ip_blkno, be32_to_cpu(resp->r_hdr.h_node_num), orphaned_slot, OCFS2_I(inode)->ip_orphaned_slot); @@ -869,8 +873,8 @@ int ocfs2_request_delete_vote(struct inode *inode) delete_cb.rc_cb = ocfs2_delete_response_cb; delete_cb.rc_priv = inode; - mlog(0, "Inode %"MLFu64", we start thinking orphaned slot is %d\n", - OCFS2_I(inode)->ip_blkno, orphaned_slot); + mlog(0, "Inode %llu, we start thinking orphaned slot is %d\n", + (unsigned long long)OCFS2_I(inode)->ip_blkno, orphaned_slot); status = -ENOMEM; request = ocfs2_new_vote_request(osb, OCFS2_I(inode)->ip_blkno, @@ -895,8 +899,8 @@ static void ocfs2_setup_unlink_vote(struct ocfs2_vote_msg *request, * d_delete against it. Parent directory block and full name * should suffice. */ - mlog(0, "unlink/rename request: parent: %"MLFu64" name: %.*s\n", - OCFS2_I(parent)->ip_blkno, dentry->d_name.len, + mlog(0, "unlink/rename request: parent: %llu name: %.*s\n", + (unsigned long long)OCFS2_I(parent)->ip_blkno, dentry->d_name.len, dentry->d_name.name); request->v_unlink_parent = cpu_to_be64(OCFS2_I(parent)->ip_blkno); @@ -1082,7 +1086,8 @@ static int ocfs2_handle_response_message(struct o2net_msg *msg, mlog(0, "received response message:\n"); mlog(0, "h_response_id = %u\n", response_id); mlog(0, "h_request = %u\n", be32_to_cpu(resp->r_hdr.h_request)); - mlog(0, "h_blkno = %"MLFu64"\n", be64_to_cpu(resp->r_hdr.h_blkno)); + mlog(0, "h_blkno = %llu\n", + (unsigned long long)be64_to_cpu(resp->r_hdr.h_blkno)); mlog(0, "h_generation = %u\n", be32_to_cpu(resp->r_hdr.h_generation)); mlog(0, "h_node_num = %u\n", node_num); mlog(0, "r_response = %d\n", response_status); @@ -1138,8 +1143,8 @@ static int ocfs2_handle_vote_message(struct o2net_msg *msg, mlog(0, "h_response_id = %u\n", be32_to_cpu(work->w_msg.v_hdr.h_response_id)); mlog(0, "h_request = %u\n", be32_to_cpu(work->w_msg.v_hdr.h_request)); - mlog(0, "h_blkno = %"MLFu64"\n", - be64_to_cpu(work->w_msg.v_hdr.h_blkno)); + mlog(0, "h_blkno = %llu\n", + (unsigned long long)be64_to_cpu(work->w_msg.v_hdr.h_blkno)); mlog(0, "h_generation = %u\n", be32_to_cpu(work->w_msg.v_hdr.h_generation)); mlog(0, "h_node_num = %u\n", diff --git a/fs/open.c b/fs/open.c index 1091dadd6c3..c32c89d6d8d 100644 --- a/fs/open.c +++ b/fs/open.c @@ -27,6 +27,7 @@ #include <linux/pagemap.h> #include <linux/syscalls.h> #include <linux/rcupdate.h> +#include <linux/audit.h> #include <asm/unistd.h> @@ -626,6 +627,8 @@ asmlinkage long sys_fchmod(unsigned int fd, mode_t mode) dentry = file->f_dentry; inode = dentry->d_inode; + audit_inode(NULL, inode, 0); + err = -EROFS; if (IS_RDONLY(inode)) goto out_putf; @@ -775,7 +778,10 @@ asmlinkage long sys_fchown(unsigned int fd, uid_t user, gid_t group) file = fget(fd); if (file) { - error = chown_common(file->f_dentry, user, group); + struct dentry * dentry; + dentry = file->f_dentry; + audit_inode(NULL, dentry->d_inode, 0); + error = chown_common(dentry, user, group); fput(file); } return error; @@ -890,6 +896,10 @@ EXPORT_SYMBOL(filp_open); * a fully instantiated struct file to the caller. * This function is meant to be called from within a filesystem's * lookup method. + * Beware of calling it for non-regular files! Those ->open methods might block + * (e.g. in fifo_open), leaving you with parent locked (and in case of fifo, + * leading to a deadlock, as nobody can open that fifo anymore, because + * another process to open fifo will block on locked parent when doing lookup). * Note that in case of error, nd->intent.open.file is destroyed, but the * path information remains valid. * If the open callback is set to NULL, then the standard f_op->open() diff --git a/fs/openpromfs/inode.c b/fs/openpromfs/inode.c index aeb0106890e..0f14276a2e5 100644 --- a/fs/openpromfs/inode.c +++ b/fs/openpromfs/inode.c @@ -581,17 +581,17 @@ int property_release (struct inode *inode, struct file *filp) return 0; } -static struct file_operations openpromfs_prop_ops = { +static const struct file_operations openpromfs_prop_ops = { .read = property_read, .write = property_write, .release = property_release, }; -static struct file_operations openpromfs_nodenum_ops = { +static const struct file_operations openpromfs_nodenum_ops = { .read = nodenum_read, }; -static struct file_operations openprom_operations = { +static const struct file_operations openprom_operations = { .read = generic_read_dir, .readdir = openpromfs_readdir, }; diff --git a/fs/partitions/check.c b/fs/partitions/check.c index f924f459bdb..af0cb4b9e78 100644 --- a/fs/partitions/check.c +++ b/fs/partitions/check.c @@ -297,6 +297,25 @@ struct kobj_type ktype_part = { .sysfs_ops = &part_sysfs_ops, }; +static inline void partition_sysfs_add_subdir(struct hd_struct *p) +{ + struct kobject *k; + + k = kobject_get(&p->kobj); + p->holder_dir = kobject_add_dir(k, "holders"); + kobject_put(k); +} + +static inline void disk_sysfs_add_subdirs(struct gendisk *disk) +{ + struct kobject *k; + + k = kobject_get(&disk->kobj); + disk->holder_dir = kobject_add_dir(k, "holders"); + disk->slave_dir = kobject_add_dir(k, "slaves"); + kobject_put(k); +} + void delete_partition(struct gendisk *disk, int part) { struct hd_struct *p = disk->part[part-1]; @@ -310,6 +329,8 @@ void delete_partition(struct gendisk *disk, int part) p->ios[0] = p->ios[1] = 0; p->sectors[0] = p->sectors[1] = 0; devfs_remove("%s/part%d", disk->devfs_name, part); + if (p->holder_dir) + kobject_unregister(p->holder_dir); kobject_unregister(&p->kobj); } @@ -337,6 +358,7 @@ void add_partition(struct gendisk *disk, int part, sector_t start, sector_t len) p->kobj.parent = &disk->kobj; p->kobj.ktype = &ktype_part; kobject_register(&p->kobj); + partition_sysfs_add_subdir(p); disk->part[part-1] = p; } @@ -383,6 +405,7 @@ void register_disk(struct gendisk *disk) if ((err = kobject_add(&disk->kobj))) return; disk_sysfs_symlinks(disk); + disk_sysfs_add_subdirs(disk); kobject_uevent(&disk->kobj, KOBJ_ADD); /* No minors to use for partitions */ @@ -483,6 +506,10 @@ void del_gendisk(struct gendisk *disk) devfs_remove_disk(disk); + if (disk->holder_dir) + kobject_unregister(disk->holder_dir); + if (disk->slave_dir) + kobject_unregister(disk->slave_dir); if (disk->driverfs_dev) { char *disk_name = make_block_name(disk); sysfs_remove_link(&disk->kobj, "device"); diff --git a/fs/partitions/devfs.c b/fs/partitions/devfs.c index 87f50444fd3..3f0a780c9ce 100644 --- a/fs/partitions/devfs.c +++ b/fs/partitions/devfs.c @@ -6,7 +6,7 @@ #include <linux/vmalloc.h> #include <linux/genhd.h> #include <linux/bitops.h> -#include <asm/semaphore.h> +#include <linux/mutex.h> struct unique_numspace { @@ -16,7 +16,7 @@ struct unique_numspace { struct semaphore mutex; }; -static DECLARE_MUTEX(numspace_mutex); +static DEFINE_MUTEX(numspace_mutex); static int expand_numspace(struct unique_numspace *s) { @@ -48,7 +48,7 @@ static int alloc_unique_number(struct unique_numspace *s) { int rval = 0; - down(&numspace_mutex); + mutex_lock(&numspace_mutex); if (s->num_free < 1) rval = expand_numspace(s); if (!rval) { @@ -56,7 +56,7 @@ static int alloc_unique_number(struct unique_numspace *s) --s->num_free; __set_bit(rval, s->bits); } - up(&numspace_mutex); + mutex_unlock(&numspace_mutex); return rval; } @@ -66,11 +66,11 @@ static void dealloc_unique_number(struct unique_numspace *s, int number) int old_val; if (number >= 0) { - down(&numspace_mutex); + mutex_lock(&numspace_mutex); old_val = __test_and_clear_bit(number, s->bits); if (old_val) ++s->num_free; - up(&numspace_mutex); + mutex_unlock(&numspace_mutex); } } diff --git a/fs/partitions/mac.c b/fs/partitions/mac.c index bb22cdd0cb1..813292f2121 100644 --- a/fs/partitions/mac.c +++ b/fs/partitions/mac.c @@ -12,6 +12,7 @@ #include "mac.h" #ifdef CONFIG_PPC_PMAC +#include <asm/machdep.h> extern void note_bootable_part(dev_t dev, int part, int goodness); #endif @@ -79,7 +80,7 @@ int mac_partition(struct parsed_partitions *state, struct block_device *bdev) * If this is the first bootable partition, tell the * setup code, in case it wants to make this the root. */ - if (_machine == _MACH_Pmac) { + if (machine_is(powermac)) { int goodness = 0; mac_fix_string(part->processor, 16); diff --git a/fs/pipe.c b/fs/pipe.c index 8aada8e426f..e2f4f1d9ffc 100644 --- a/fs/pipe.c +++ b/fs/pipe.c @@ -568,7 +568,7 @@ pipe_rdwr_open(struct inode *inode, struct file *filp) * The file_operations structs are not static because they * are also used in linux/fs/fifo.c to do operations on FIFOs. */ -struct file_operations read_fifo_fops = { +const struct file_operations read_fifo_fops = { .llseek = no_llseek, .read = pipe_read, .readv = pipe_readv, @@ -580,7 +580,7 @@ struct file_operations read_fifo_fops = { .fasync = pipe_read_fasync, }; -struct file_operations write_fifo_fops = { +const struct file_operations write_fifo_fops = { .llseek = no_llseek, .read = bad_pipe_r, .write = pipe_write, @@ -592,7 +592,7 @@ struct file_operations write_fifo_fops = { .fasync = pipe_write_fasync, }; -struct file_operations rdwr_fifo_fops = { +const struct file_operations rdwr_fifo_fops = { .llseek = no_llseek, .read = pipe_read, .readv = pipe_readv, @@ -662,10 +662,9 @@ struct inode* pipe_new(struct inode* inode) { struct pipe_inode_info *info; - info = kmalloc(sizeof(struct pipe_inode_info), GFP_KERNEL); + info = kzalloc(sizeof(struct pipe_inode_info), GFP_KERNEL); if (!info) goto fail_page; - memset(info, 0, sizeof(*info)); inode->i_pipe = info; init_waitqueue_head(PIPE_WAIT(*inode)); @@ -676,7 +675,7 @@ fail_page: return NULL; } -static struct vfsmount *pipe_mnt; +static struct vfsmount *pipe_mnt __read_mostly; static int pipefs_delete_dentry(struct dentry *dentry) { return 1; diff --git a/fs/proc/array.c b/fs/proc/array.c index 7eb1bd7f800..7a76ad57023 100644 --- a/fs/proc/array.c +++ b/fs/proc/array.c @@ -330,7 +330,6 @@ static int do_task_stat(struct task_struct *task, char * buffer, int whole) unsigned long min_flt = 0, maj_flt = 0; cputime_t cutime, cstime, utime, stime; unsigned long rsslim = 0; - DEFINE_KTIME(it_real_value); struct task_struct *t; char tcomm[sizeof(task->comm)]; @@ -386,7 +385,6 @@ static int do_task_stat(struct task_struct *task, char * buffer, int whole) utime = cputime_add(utime, task->signal->utime); stime = cputime_add(stime, task->signal->stime); } - it_real_value = task->signal->real_timer.expires; } ppid = pid_alive(task) ? task->group_leader->real_parent->tgid : 0; read_unlock(&tasklist_lock); @@ -413,7 +411,7 @@ static int do_task_stat(struct task_struct *task, char * buffer, int whole) start_time = nsec_to_clock_t(start_time); res = sprintf(buffer,"%d (%s) %c %d %d %d %d %d %lu %lu \ -%lu %lu %lu %lu %lu %ld %ld %ld %ld %d %ld %llu %lu %ld %lu %lu %lu %lu %lu \ +%lu %lu %lu %lu %lu %ld %ld %ld %ld %d 0 %llu %lu %ld %lu %lu %lu %lu %lu \ %lu %lu %lu %lu %lu %lu %lu %lu %d %d %lu %lu\n", task->pid, tcomm, @@ -435,7 +433,6 @@ static int do_task_stat(struct task_struct *task, char * buffer, int whole) priority, nice, num_threads, - (long) ktime_to_clock_t(it_real_value), start_time, vsize, mm ? get_mm_rss(mm) : 0, diff --git a/fs/proc/base.c b/fs/proc/base.c index 20feb7568de..8f1f49ceebe 100644 --- a/fs/proc/base.c +++ b/fs/proc/base.c @@ -104,6 +104,7 @@ enum pid_directory_inos { PROC_TGID_MAPS, PROC_TGID_NUMA_MAPS, PROC_TGID_MOUNTS, + PROC_TGID_MOUNTSTATS, PROC_TGID_WCHAN, #ifdef CONFIG_MMU PROC_TGID_SMAPS, @@ -144,6 +145,7 @@ enum pid_directory_inos { PROC_TID_MAPS, PROC_TID_NUMA_MAPS, PROC_TID_MOUNTS, + PROC_TID_MOUNTSTATS, PROC_TID_WCHAN, #ifdef CONFIG_MMU PROC_TID_SMAPS, @@ -201,6 +203,7 @@ static struct pid_entry tgid_base_stuff[] = { E(PROC_TGID_ROOT, "root", S_IFLNK|S_IRWXUGO), E(PROC_TGID_EXE, "exe", S_IFLNK|S_IRWXUGO), E(PROC_TGID_MOUNTS, "mounts", S_IFREG|S_IRUGO), + E(PROC_TGID_MOUNTSTATS, "mountstats", S_IFREG|S_IRUSR), #ifdef CONFIG_MMU E(PROC_TGID_SMAPS, "smaps", S_IFREG|S_IRUGO), #endif @@ -732,6 +735,38 @@ static struct file_operations proc_mounts_operations = { .poll = mounts_poll, }; +extern struct seq_operations mountstats_op; +static int mountstats_open(struct inode *inode, struct file *file) +{ + struct task_struct *task = proc_task(inode); + int ret = seq_open(file, &mountstats_op); + + if (!ret) { + struct seq_file *m = file->private_data; + struct namespace *namespace; + task_lock(task); + namespace = task->namespace; + if (namespace) + get_namespace(namespace); + task_unlock(task); + + if (namespace) + m->private = namespace; + else { + seq_release(inode, file); + ret = -EINVAL; + } + } + return ret; +} + +static struct file_operations proc_mountstats_operations = { + .open = mountstats_open, + .read = seq_read, + .llseek = seq_lseek, + .release = mounts_release, +}; + #define PROC_BLOCK_SIZE (3*1024) /* 4K page size but our output routines use some slack for overruns */ static ssize_t proc_info_read(struct file * file, char __user * buf, @@ -1730,6 +1765,10 @@ static struct dentry *proc_pident_lookup(struct inode *dir, inode->i_fop = &proc_smaps_operations; break; #endif + case PROC_TID_MOUNTSTATS: + case PROC_TGID_MOUNTSTATS: + inode->i_fop = &proc_mountstats_operations; + break; #ifdef CONFIG_SECURITY case PROC_TID_ATTR: inode->i_nlink = 2; diff --git a/fs/proc/generic.c b/fs/proc/generic.c index 20e5c4509a4..4ba03009cf7 100644 --- a/fs/proc/generic.c +++ b/fs/proc/generic.c @@ -19,6 +19,7 @@ #include <linux/idr.h> #include <linux/namei.h> #include <linux/bitops.h> +#include <linux/spinlock.h> #include <asm/uaccess.h> #include "internal.h" @@ -29,6 +30,8 @@ static ssize_t proc_file_write(struct file *file, const char __user *buffer, size_t count, loff_t *ppos); static loff_t proc_file_lseek(struct file *, loff_t, int); +DEFINE_SPINLOCK(proc_subdir_lock); + int proc_match(int len, const char *name, struct proc_dir_entry *de) { if (de->namelen != len) @@ -277,7 +280,9 @@ static int xlate_proc_name(const char *name, const char *cp = name, *next; struct proc_dir_entry *de; int len; + int rtn = 0; + spin_lock(&proc_subdir_lock); de = &proc_root; while (1) { next = strchr(cp, '/'); @@ -289,13 +294,17 @@ static int xlate_proc_name(const char *name, if (proc_match(len, cp, de)) break; } - if (!de) - return -ENOENT; + if (!de) { + rtn = -ENOENT; + goto out; + } cp += len + 1; } *residual = cp; *ret = de; - return 0; +out: + spin_unlock(&proc_subdir_lock); + return rtn; } static DEFINE_IDR(proc_inum_idr); @@ -380,6 +389,7 @@ struct dentry *proc_lookup(struct inode * dir, struct dentry *dentry, struct nam int error = -ENOENT; lock_kernel(); + spin_lock(&proc_subdir_lock); de = PDE(dir); if (de) { for (de = de->subdir; de ; de = de->next) { @@ -388,12 +398,15 @@ struct dentry *proc_lookup(struct inode * dir, struct dentry *dentry, struct nam if (!memcmp(dentry->d_name.name, de->name, de->namelen)) { unsigned int ino = de->low_ino; + spin_unlock(&proc_subdir_lock); error = -EINVAL; inode = proc_get_inode(dir->i_sb, ino, de); + spin_lock(&proc_subdir_lock); break; } } } + spin_unlock(&proc_subdir_lock); unlock_kernel(); if (inode) { @@ -447,11 +460,13 @@ int proc_readdir(struct file * filp, filp->f_pos++; /* fall through */ default: + spin_lock(&proc_subdir_lock); de = de->subdir; i -= 2; for (;;) { if (!de) { ret = 1; + spin_unlock(&proc_subdir_lock); goto out; } if (!i) @@ -461,12 +476,16 @@ int proc_readdir(struct file * filp, } do { + /* filldir passes info to user space */ + spin_unlock(&proc_subdir_lock); if (filldir(dirent, de->name, de->namelen, filp->f_pos, de->low_ino, de->mode >> 12) < 0) goto out; + spin_lock(&proc_subdir_lock); filp->f_pos++; de = de->next; } while (de); + spin_unlock(&proc_subdir_lock); } ret = 1; out: unlock_kernel(); @@ -500,9 +519,13 @@ static int proc_register(struct proc_dir_entry * dir, struct proc_dir_entry * dp if (i == 0) return -EAGAIN; dp->low_ino = i; + + spin_lock(&proc_subdir_lock); dp->next = dir->subdir; dp->parent = dir; dir->subdir = dp; + spin_unlock(&proc_subdir_lock); + if (S_ISDIR(dp->mode)) { if (dp->proc_iops == NULL) { dp->proc_fops = &proc_dir_operations; @@ -537,7 +560,7 @@ static void proc_kill_inodes(struct proc_dir_entry *de) struct file * filp = list_entry(p, struct file, f_u.fu_list); struct dentry * dentry = filp->f_dentry; struct inode * inode; - struct file_operations *fops; + const struct file_operations *fops; if (dentry->d_op != &proc_dentry_operations) continue; @@ -694,6 +717,8 @@ void remove_proc_entry(const char *name, struct proc_dir_entry *parent) if (!parent && xlate_proc_name(name, &parent, &fn) != 0) goto out; len = strlen(fn); + + spin_lock(&proc_subdir_lock); for (p = &parent->subdir; *p; p=&(*p)->next ) { if (!proc_match(len, fn, *p)) continue; @@ -714,6 +739,7 @@ void remove_proc_entry(const char *name, struct proc_dir_entry *parent) } break; } + spin_unlock(&proc_subdir_lock); out: return; } diff --git a/fs/proc/internal.h b/fs/proc/internal.h index 95a1cf32b83..0502f17b860 100644 --- a/fs/proc/internal.h +++ b/fs/proc/internal.h @@ -30,7 +30,7 @@ do { \ #endif -extern void create_seq_entry(char *name, mode_t mode, struct file_operations *f); +extern void create_seq_entry(char *name, mode_t mode, const struct file_operations *f); extern int proc_exe_link(struct inode *, struct dentry **, struct vfsmount **); extern int proc_tid_stat(struct task_struct *, char *); extern int proc_tgid_stat(struct task_struct *, char *); diff --git a/fs/proc/kcore.c b/fs/proc/kcore.c index adc2cd95169..17f6e8fa139 100644 --- a/fs/proc/kcore.c +++ b/fs/proc/kcore.c @@ -31,7 +31,7 @@ static int open_kcore(struct inode * inode, struct file * filp) static ssize_t read_kcore(struct file *, char __user *, size_t, loff_t *); -struct file_operations proc_kcore_operations = { +const struct file_operations proc_kcore_operations = { .read = read_kcore, .open = open_kcore, }; diff --git a/fs/proc/kmsg.c b/fs/proc/kmsg.c index 10d37bf2520..ff3b90b56e9 100644 --- a/fs/proc/kmsg.c +++ b/fs/proc/kmsg.c @@ -47,7 +47,7 @@ static unsigned int kmsg_poll(struct file *file, poll_table *wait) } -struct file_operations proc_kmsg_operations = { +const struct file_operations proc_kmsg_operations = { .read = kmsg_read, .poll = kmsg_poll, .open = kmsg_open, diff --git a/fs/proc/proc_devtree.c b/fs/proc/proc_devtree.c index 9bdd077d6f5..abdf068bc27 100644 --- a/fs/proc/proc_devtree.c +++ b/fs/proc/proc_devtree.c @@ -52,7 +52,8 @@ static int property_read_proc(char *page, char **start, off_t off, * Add a property to a node */ static struct proc_dir_entry * -__proc_device_tree_add_prop(struct proc_dir_entry *de, struct property *pp) +__proc_device_tree_add_prop(struct proc_dir_entry *de, struct property *pp, + const char *name) { struct proc_dir_entry *ent; @@ -60,14 +61,14 @@ __proc_device_tree_add_prop(struct proc_dir_entry *de, struct property *pp) * Unfortunately proc_register puts each new entry * at the beginning of the list. So we rearrange them. */ - ent = create_proc_read_entry(pp->name, - strncmp(pp->name, "security-", 9) + ent = create_proc_read_entry(name, + strncmp(name, "security-", 9) ? S_IRUGO : S_IRUSR, de, property_read_proc, pp); if (ent == NULL) return NULL; - if (!strncmp(pp->name, "security-", 9)) + if (!strncmp(name, "security-", 9)) ent->size = 0; /* don't leak number of password chars */ else ent->size = pp->length; @@ -78,7 +79,7 @@ __proc_device_tree_add_prop(struct proc_dir_entry *de, struct property *pp) void proc_device_tree_add_prop(struct proc_dir_entry *pde, struct property *prop) { - __proc_device_tree_add_prop(pde, prop); + __proc_device_tree_add_prop(pde, prop, prop->name); } void proc_device_tree_remove_prop(struct proc_dir_entry *pde, @@ -106,6 +107,69 @@ void proc_device_tree_update_prop(struct proc_dir_entry *pde, } /* + * Various dodgy firmware might give us nodes and/or properties with + * conflicting names. That's generally ok, except for exporting via /proc, + * so munge names here to ensure they're unique. + */ + +static int duplicate_name(struct proc_dir_entry *de, const char *name) +{ + struct proc_dir_entry *ent; + int found = 0; + + spin_lock(&proc_subdir_lock); + + for (ent = de->subdir; ent != NULL; ent = ent->next) { + if (strcmp(ent->name, name) == 0) { + found = 1; + break; + } + } + + spin_unlock(&proc_subdir_lock); + + return found; +} + +static const char *fixup_name(struct device_node *np, struct proc_dir_entry *de, + const char *name) +{ + char *fixed_name; + int fixup_len = strlen(name) + 2 + 1; /* name + #x + \0 */ + int i = 1, size; + +realloc: + fixed_name = kmalloc(fixup_len, GFP_KERNEL); + if (fixed_name == NULL) { + printk(KERN_ERR "device-tree: Out of memory trying to fixup " + "name \"%s\"\n", name); + return name; + } + +retry: + size = snprintf(fixed_name, fixup_len, "%s#%d", name, i); + size++; /* account for NULL */ + + if (size > fixup_len) { + /* We ran out of space, free and reallocate. */ + kfree(fixed_name); + fixup_len = size; + goto realloc; + } + + if (duplicate_name(de, fixed_name)) { + /* Multiple duplicates. Retry with a different offset. */ + i++; + goto retry; + } + + printk(KERN_WARNING "device-tree: Duplicate name in %s, " + "renamed to \"%s\"\n", np->full_name, fixed_name); + + return fixed_name; +} + +/* * Process a node, adding entries for its children and its properties. */ void proc_device_tree_add_node(struct device_node *np, @@ -118,35 +182,30 @@ void proc_device_tree_add_node(struct device_node *np, set_node_proc_entry(np, de); for (child = NULL; (child = of_get_next_child(np, child));) { + /* Use everything after the last slash, or the full name */ p = strrchr(child->full_name, '/'); if (!p) p = child->full_name; else ++p; + + if (duplicate_name(de, p)) + p = fixup_name(np, de, p); + ent = proc_mkdir(p, de); if (ent == 0) break; proc_device_tree_add_node(child, ent); } of_node_put(child); + for (pp = np->properties; pp != 0; pp = pp->next) { - /* - * Yet another Apple device-tree bogosity: on some machines, - * they have properties & nodes with the same name. Those - * properties are quite unimportant for us though, thus we - * simply "skip" them here, but we do have to check. - */ - for (ent = de->subdir; ent != NULL; ent = ent->next) - if (!strcmp(ent->name, pp->name)) - break; - if (ent != NULL) { - printk(KERN_WARNING "device-tree: property \"%s\" name" - " conflicts with node in %s\n", pp->name, - np->full_name); - continue; - } + p = pp->name; + + if (duplicate_name(de, p)) + p = fixup_name(np, de, p); - ent = __proc_device_tree_add_prop(de, pp); + ent = __proc_device_tree_add_prop(de, pp, p); if (ent == 0) break; } diff --git a/fs/proc/proc_misc.c b/fs/proc/proc_misc.c index 826c131994c..ef5a3323f4b 100644 --- a/fs/proc/proc_misc.c +++ b/fs/proc/proc_misc.c @@ -485,6 +485,40 @@ static struct file_operations proc_slabinfo_operations = { .llseek = seq_lseek, .release = seq_release, }; + +#ifdef CONFIG_DEBUG_SLAB_LEAK +extern struct seq_operations slabstats_op; +static int slabstats_open(struct inode *inode, struct file *file) +{ + unsigned long *n = kzalloc(PAGE_SIZE, GFP_KERNEL); + int ret = -ENOMEM; + if (n) { + ret = seq_open(file, &slabstats_op); + if (!ret) { + struct seq_file *m = file->private_data; + *n = PAGE_SIZE / (2 * sizeof(unsigned long)); + m->private = n; + n = NULL; + } + kfree(n); + } + return ret; +} + +static int slabstats_release(struct inode *inode, struct file *file) +{ + struct seq_file *m = file->private_data; + kfree(m->private); + return seq_release(inode, file); +} + +static struct file_operations proc_slabstats_operations = { + .open = slabstats_open, + .read = seq_read, + .llseek = seq_lseek, + .release = slabstats_release, +}; +#endif #endif static int show_stat(struct seq_file *p, void *v) @@ -500,7 +534,7 @@ static int show_stat(struct seq_file *p, void *v) if (wall_to_monotonic.tv_nsec) --jif; - for_each_cpu(i) { + for_each_possible_cpu(i) { int j; user = cputime64_add(user, kstat_cpu(i).cpustat.user); @@ -697,7 +731,7 @@ static struct file_operations proc_sysrq_trigger_operations = { struct proc_dir_entry *proc_root_kcore; -void create_seq_entry(char *name, mode_t mode, struct file_operations *f) +void create_seq_entry(char *name, mode_t mode, const struct file_operations *f) { struct proc_dir_entry *entry; entry = create_proc_entry(name, mode, NULL); @@ -744,6 +778,9 @@ void __init proc_misc_init(void) create_seq_entry("interrupts", 0, &proc_interrupts_operations); #ifdef CONFIG_SLAB create_seq_entry("slabinfo",S_IWUSR|S_IRUGO,&proc_slabinfo_operations); +#ifdef CONFIG_DEBUG_SLAB_LEAK + create_seq_entry("slab_allocators", 0 ,&proc_slabstats_operations); +#endif #endif create_seq_entry("buddyinfo",S_IRUGO, &fragmentation_file_operations); create_seq_entry("vmstat",S_IRUGO, &proc_vmstat_file_operations); diff --git a/fs/proc/vmcore.c b/fs/proc/vmcore.c index 4063fb32f78..7efa73d44c9 100644 --- a/fs/proc/vmcore.c +++ b/fs/proc/vmcore.c @@ -172,7 +172,7 @@ static int open_vmcore(struct inode *inode, struct file *filp) return 0; } -struct file_operations proc_vmcore_operations = { +const struct file_operations proc_vmcore_operations = { .read = read_vmcore, .open = open_vmcore, }; diff --git a/fs/qnx4/dir.c b/fs/qnx4/dir.c index 7a8f5595c26..9031948fefd 100644 --- a/fs/qnx4/dir.c +++ b/fs/qnx4/dir.c @@ -81,7 +81,7 @@ out: return 0; } -struct file_operations qnx4_dir_operations = +const struct file_operations qnx4_dir_operations = { .read = generic_read_dir, .readdir = qnx4_readdir, diff --git a/fs/qnx4/file.c b/fs/qnx4/file.c index c33963fded9..62af4b1348b 100644 --- a/fs/qnx4/file.c +++ b/fs/qnx4/file.c @@ -19,7 +19,7 @@ * We have mostly NULL's here: the current defaults are ok for * the qnx4 filesystem. */ -struct file_operations qnx4_file_operations = +const struct file_operations qnx4_file_operations = { .llseek = generic_file_llseek, .read = generic_file_read, diff --git a/fs/ramfs/file-mmu.c b/fs/ramfs/file-mmu.c index 6ada2095b9a..00a933eb820 100644 --- a/fs/ramfs/file-mmu.c +++ b/fs/ramfs/file-mmu.c @@ -32,7 +32,7 @@ struct address_space_operations ramfs_aops = { .commit_write = simple_commit_write }; -struct file_operations ramfs_file_operations = { +const struct file_operations ramfs_file_operations = { .read = generic_file_read, .write = generic_file_write, .mmap = generic_file_mmap, diff --git a/fs/ramfs/file-nommu.c b/fs/ramfs/file-nommu.c index b1ca234068f..f443a84b98a 100644 --- a/fs/ramfs/file-nommu.c +++ b/fs/ramfs/file-nommu.c @@ -33,7 +33,7 @@ struct address_space_operations ramfs_aops = { .commit_write = simple_commit_write }; -struct file_operations ramfs_file_operations = { +const struct file_operations ramfs_file_operations = { .mmap = ramfs_nommu_mmap, .get_unmapped_area = ramfs_nommu_get_unmapped_area, .read = generic_file_read, diff --git a/fs/ramfs/internal.h b/fs/ramfs/internal.h index 272c8a7120b..313237631b4 100644 --- a/fs/ramfs/internal.h +++ b/fs/ramfs/internal.h @@ -11,5 +11,5 @@ extern struct address_space_operations ramfs_aops; -extern struct file_operations ramfs_file_operations; +extern const struct file_operations ramfs_file_operations; extern struct inode_operations ramfs_file_inode_operations; diff --git a/fs/read_write.c b/fs/read_write.c index 3f7a1a62165..6256ca81a71 100644 --- a/fs/read_write.c +++ b/fs/read_write.c @@ -19,7 +19,7 @@ #include <asm/uaccess.h> #include <asm/unistd.h> -struct file_operations generic_ro_fops = { +const struct file_operations generic_ro_fops = { .llseek = generic_file_llseek, .read = generic_file_read, .mmap = generic_file_readonly_mmap, @@ -470,7 +470,7 @@ static ssize_t do_readv_writev(int type, struct file *file, * verify all the pointers */ ret = -EINVAL; - if ((nr_segs > UIO_MAXIOV) || (nr_segs <= 0)) + if (nr_segs > UIO_MAXIOV) goto out; if (!file->f_op) goto out; diff --git a/fs/reiserfs/dir.c b/fs/reiserfs/dir.c index d71ac657928..973c819f803 100644 --- a/fs/reiserfs/dir.c +++ b/fs/reiserfs/dir.c @@ -18,7 +18,7 @@ static int reiserfs_readdir(struct file *, void *, filldir_t); static int reiserfs_dir_fsync(struct file *filp, struct dentry *dentry, int datasync); -struct file_operations reiserfs_dir_operations = { +const struct file_operations reiserfs_dir_operations = { .read = generic_read_dir, .readdir = reiserfs_readdir, .fsync = reiserfs_dir_fsync, diff --git a/fs/reiserfs/file.c b/fs/reiserfs/file.c index be12879bb17..010094d14da 100644 --- a/fs/reiserfs/file.c +++ b/fs/reiserfs/file.c @@ -1532,7 +1532,7 @@ static ssize_t reiserfs_file_write(struct file *file, /* the file we are going t buf += write_bytes; *ppos = pos += write_bytes; count -= write_bytes; - balance_dirty_pages_ratelimited(inode->i_mapping); + balance_dirty_pages_ratelimited_nr(inode->i_mapping, num_pages); } /* this is only true on error */ @@ -1546,10 +1546,10 @@ static ssize_t reiserfs_file_write(struct file *file, /* the file we are going t } } - if ((file->f_flags & O_SYNC) || IS_SYNC(inode)) - res = - generic_osync_inode(inode, file->f_mapping, - OSYNC_METADATA | OSYNC_DATA); + if (likely(res >= 0) && + (unlikely((file->f_flags & O_SYNC) || IS_SYNC(inode)))) + res = generic_osync_inode(inode, file->f_mapping, + OSYNC_METADATA | OSYNC_DATA); mutex_unlock(&inode->i_mutex); reiserfs_async_progress_wait(inode->i_sb); @@ -1566,7 +1566,7 @@ static ssize_t reiserfs_aio_write(struct kiocb *iocb, const char __user * buf, return generic_file_aio_write(iocb, buf, count, pos); } -struct file_operations reiserfs_file_operations = { +const struct file_operations reiserfs_file_operations = { .read = generic_file_read, .write = reiserfs_file_write, .ioctl = reiserfs_ioctl, diff --git a/fs/reiserfs/fix_node.c b/fs/reiserfs/fix_node.c index aa22588019e..5600d3d60cf 100644 --- a/fs/reiserfs/fix_node.c +++ b/fs/reiserfs/fix_node.c @@ -191,9 +191,7 @@ static void create_virtual_node(struct tree_balance *tb, int h) "vs-8045: create_virtual_node: rdkey %k, affected item==%d (mode==%c) Must be %c", key, vn->vn_affected_item_num, vn->vn_mode, M_DELETE); - } else - /* we can delete directory item, that has only one directory entry in it */ - ; + } } #endif diff --git a/fs/reiserfs/inode.c b/fs/reiserfs/inode.c index d60f6238c66..9857e50f85e 100644 --- a/fs/reiserfs/inode.c +++ b/fs/reiserfs/inode.c @@ -466,7 +466,6 @@ static int reiserfs_get_block_create_0(struct inode *inode, sector_t block, direct_IO request. */ static int reiserfs_get_blocks_direct_io(struct inode *inode, sector_t iblock, - unsigned long max_blocks, struct buffer_head *bh_result, int create) { @@ -2793,7 +2792,7 @@ static int invalidatepage_can_drop(struct inode *inode, struct buffer_head *bh) } /* clm -- taken from fs/buffer.c:block_invalidate_page */ -static int reiserfs_invalidatepage(struct page *page, unsigned long offset) +static void reiserfs_invalidatepage(struct page *page, unsigned long offset) { struct buffer_head *head, *bh, *next; struct inode *inode = page->mapping->host; @@ -2832,10 +2831,12 @@ static int reiserfs_invalidatepage(struct page *page, unsigned long offset) * The get_block cached value has been unconditionally invalidated, * so real IO is not possible anymore. */ - if (!offset && ret) + if (!offset && ret) { ret = try_to_release_page(page, 0); + /* maybe should BUG_ON(!ret); - neilb */ + } out: - return ret; + return; } static int reiserfs_set_page_dirty(struct page *page) diff --git a/fs/reiserfs/item_ops.c b/fs/reiserfs/item_ops.c index e237cd668e5..7a88adbceef 100644 --- a/fs/reiserfs/item_ops.c +++ b/fs/reiserfs/item_ops.c @@ -275,7 +275,7 @@ static void indirect_print_item(struct item_head *ih, char *item) int j; __le32 *unp; __u32 prev = INT_MAX; - int num; + int num = 0; unp = (__le32 *) item; diff --git a/fs/reiserfs/journal.c b/fs/reiserfs/journal.c index 5a9d2722fa0..1b73529b809 100644 --- a/fs/reiserfs/journal.c +++ b/fs/reiserfs/journal.c @@ -2227,6 +2227,9 @@ static int journal_read_transaction(struct super_block *p_s_sb, journal->j_start = cur_dblock - SB_ONDISK_JOURNAL_1st_BLOCK(p_s_sb); journal->j_last_flush_trans_id = trans_id; journal->j_trans_id = trans_id + 1; + /* check for trans_id overflow */ + if (journal->j_trans_id == 0) + journal->j_trans_id = 10; brelse(c_bh); brelse(d_bh); kfree(log_blocks); @@ -2450,6 +2453,9 @@ static int journal_read(struct super_block *p_s_sb) journal->j_start = le32_to_cpu(jh->j_first_unflushed_offset); journal->j_trans_id = le32_to_cpu(jh->j_last_flush_trans_id) + 1; + /* check for trans_id overflow */ + if (journal->j_trans_id == 0) + journal->j_trans_id = 10; journal->j_last_flush_trans_id = le32_to_cpu(jh->j_last_flush_trans_id); journal->j_mount_id = le32_to_cpu(jh->j_mount_id) + 1; @@ -3873,8 +3879,8 @@ static int do_journal_end(struct reiserfs_transaction_handle *th, int cur_write_start = 0; /* start index of current log write */ int old_start; int i; - int flush = flags & FLUSH_ALL; - int wait_on_commit = flags & WAIT; + int flush; + int wait_on_commit; struct reiserfs_journal_list *jl, *temp_jl; struct list_head *entry, *safe; unsigned long jindex; @@ -3884,6 +3890,13 @@ static int do_journal_end(struct reiserfs_transaction_handle *th, BUG_ON(th->t_refcount > 1); BUG_ON(!th->t_trans_id); + /* protect flush_older_commits from doing mistakes if the + transaction ID counter gets overflowed. */ + if (th->t_trans_id == ~0UL) + flags |= FLUSH_ALL | COMMIT_NOW | WAIT; + flush = flags & FLUSH_ALL; + wait_on_commit = flags & WAIT; + put_fs_excl(); current->journal_info = th->t_handle_save; reiserfs_check_lock_depth(p_s_sb, "journal end"); @@ -4105,7 +4118,9 @@ static int do_journal_end(struct reiserfs_transaction_handle *th, journal->j_first = NULL; journal->j_len = 0; journal->j_trans_start_time = 0; - journal->j_trans_id++; + /* check for trans_id overflow */ + if (++journal->j_trans_id == 0) + journal->j_trans_id = 10; journal->j_current_jl->j_trans_id = journal->j_trans_id; journal->j_must_wait = 0; journal->j_len_alloc = 0; diff --git a/fs/reiserfs/prints.c b/fs/reiserfs/prints.c index 78b40621b88..27bd3a1df2a 100644 --- a/fs/reiserfs/prints.c +++ b/fs/reiserfs/prints.c @@ -143,7 +143,7 @@ static void sprintf_buffer_head(char *buf, struct buffer_head *bh) char b[BDEVNAME_SIZE]; sprintf(buf, - "dev %s, size %d, blocknr %llu, count %d, state 0x%lx, page %p, (%s, %s, %s)", + "dev %s, size %zd, blocknr %llu, count %d, state 0x%lx, page %p, (%s, %s, %s)", bdevname(bh->b_bdev, b), bh->b_size, (unsigned long long)bh->b_blocknr, atomic_read(&(bh->b_count)), bh->b_state, bh->b_page, diff --git a/fs/reiserfs/procfs.c b/fs/reiserfs/procfs.c index ef6caed9336..731688e1cfe 100644 --- a/fs/reiserfs/procfs.c +++ b/fs/reiserfs/procfs.c @@ -470,7 +470,7 @@ static int r_open(struct inode *inode, struct file *file) return ret; } -static struct file_operations r_file_operations = { +static const struct file_operations r_file_operations = { .open = r_open, .read = seq_read, .llseek = seq_lseek, diff --git a/fs/reiserfs/stree.c b/fs/reiserfs/stree.c index e2d08d7bcff..d2b25e1ba6e 100644 --- a/fs/reiserfs/stree.c +++ b/fs/reiserfs/stree.c @@ -981,6 +981,8 @@ static inline int prepare_for_direntry_item(struct path *path, return M_CUT; } +#define JOURNAL_FOR_FREE_BLOCK_AND_UPDATE_SD (2 * JOURNAL_PER_BALANCE_CNT + 1) + /* If the path points to a directory or direct item, calculate mode and the size cut, for balance. If the path points to an indirect item, remove some number of its unformatted nodes. In case of file truncate calculate whether this item must be deleted/truncated or last @@ -1020,148 +1022,79 @@ static char prepare_for_delete_or_cut(struct reiserfs_transaction_handle *th, st /* Case of an indirect item. */ { - int n_unfm_number, /* Number of the item unformatted nodes. */ - n_counter, n_blk_size; - __le32 *p_n_unfm_pointer; /* Pointer to the unformatted node number. */ - __u32 tmp; - struct item_head s_ih; /* Item header. */ - char c_mode; /* Returned mode of the balance. */ - int need_research; - - n_blk_size = p_s_sb->s_blocksize; - - /* Search for the needed object indirect item until there are no unformatted nodes to be removed. */ - do { - need_research = 0; - p_s_bh = PATH_PLAST_BUFFER(p_s_path); - /* Copy indirect item header to a temp variable. */ - copy_item_head(&s_ih, PATH_PITEM_HEAD(p_s_path)); - /* Calculate number of unformatted nodes in this item. */ - n_unfm_number = I_UNFM_NUM(&s_ih); - - RFALSE(!is_indirect_le_ih(&s_ih) || !n_unfm_number || - pos_in_item(p_s_path) + 1 != n_unfm_number, - "PAP-5240: invalid item %h " - "n_unfm_number = %d *p_n_pos_in_item = %d", - &s_ih, n_unfm_number, pos_in_item(p_s_path)); - - /* Calculate balance mode and position in the item to remove unformatted nodes. */ - if (n_new_file_length == max_reiserfs_offset(inode)) { /* Case of delete. */ - pos_in_item(p_s_path) = 0; - *p_n_cut_size = -(IH_SIZE + ih_item_len(&s_ih)); - c_mode = M_DELETE; - } else { /* Case of truncate. */ - if (n_new_file_length < le_ih_k_offset(&s_ih)) { - pos_in_item(p_s_path) = 0; - *p_n_cut_size = - -(IH_SIZE + ih_item_len(&s_ih)); - c_mode = M_DELETE; /* Delete this item. */ - } else { - /* indirect item must be truncated starting from *p_n_pos_in_item-th position */ - pos_in_item(p_s_path) = - (n_new_file_length + n_blk_size - - le_ih_k_offset(&s_ih)) >> p_s_sb-> - s_blocksize_bits; - - RFALSE(pos_in_item(p_s_path) > - n_unfm_number, - "PAP-5250: invalid position in the item"); - - /* Either convert last unformatted node of indirect item to direct item or increase - its free space. */ - if (pos_in_item(p_s_path) == - n_unfm_number) { - *p_n_cut_size = 0; /* Nothing to cut. */ - return M_CONVERT; /* Maybe convert last unformatted node to the direct item. */ - } - /* Calculate size to cut. */ - *p_n_cut_size = - -(ih_item_len(&s_ih) - - pos_in_item(p_s_path) * - UNFM_P_SIZE); - - c_mode = M_CUT; /* Cut from this indirect item. */ - } - } + int blk_size = p_s_sb->s_blocksize; + struct item_head s_ih; + int need_re_search; + int delete = 0; + int result = M_CUT; + int pos = 0; + + if ( n_new_file_length == max_reiserfs_offset (inode) ) { + /* prepare_for_delete_or_cut() is called by + * reiserfs_delete_item() */ + n_new_file_length = 0; + delete = 1; + } + + do { + need_re_search = 0; + *p_n_cut_size = 0; + p_s_bh = PATH_PLAST_BUFFER(p_s_path); + copy_item_head(&s_ih, PATH_PITEM_HEAD(p_s_path)); + pos = I_UNFM_NUM(&s_ih); - RFALSE(n_unfm_number <= pos_in_item(p_s_path), - "PAP-5260: invalid position in the indirect item"); - - /* pointers to be cut */ - n_unfm_number -= pos_in_item(p_s_path); - /* Set pointer to the last unformatted node pointer that is to be cut. */ - p_n_unfm_pointer = - (__le32 *) B_I_PITEM(p_s_bh, - &s_ih) + I_UNFM_NUM(&s_ih) - - 1 - *p_n_removed; - - /* We go through the unformatted nodes pointers of the indirect - item and look for the unformatted nodes in the cache. If we - found some of them we free it, zero corresponding indirect item - entry and log buffer containing that indirect item. For this we - need to prepare last path element for logging. If some - unformatted node has b_count > 1 we must not free this - unformatted node since it is in use. */ - reiserfs_prepare_for_journal(p_s_sb, p_s_bh, 1); - // note: path could be changed, first line in for loop takes care - // of it + while (le_ih_k_offset (&s_ih) + (pos - 1) * blk_size > n_new_file_length) { + __u32 *unfm, block; - for (n_counter = *p_n_removed; - n_counter < n_unfm_number; - n_counter++, p_n_unfm_pointer--) { + /* Each unformatted block deletion may involve one additional + * bitmap block into the transaction, thereby the initial + * journal space reservation might not be enough. */ + if (!delete && (*p_n_cut_size) != 0 && + reiserfs_transaction_free_space(th) < JOURNAL_FOR_FREE_BLOCK_AND_UPDATE_SD) { + break; + } - cond_resched(); - if (item_moved(&s_ih, p_s_path)) { - need_research = 1; - break; - } - RFALSE(p_n_unfm_pointer < - (__le32 *) B_I_PITEM(p_s_bh, &s_ih) - || p_n_unfm_pointer > - (__le32 *) B_I_PITEM(p_s_bh, - &s_ih) + - I_UNFM_NUM(&s_ih) - 1, - "vs-5265: pointer out of range"); - - /* Hole, nothing to remove. */ - if (!get_block_num(p_n_unfm_pointer, 0)) { - (*p_n_removed)++; - continue; - } + unfm = (__u32 *)B_I_PITEM(p_s_bh, &s_ih) + pos - 1; + block = get_block_num(unfm, 0); - (*p_n_removed)++; + if (block != 0) { + reiserfs_prepare_for_journal(p_s_sb, p_s_bh, 1); + put_block_num(unfm, 0, 0); + journal_mark_dirty (th, p_s_sb, p_s_bh); + reiserfs_free_block(th, inode, block, 1); + } - tmp = get_block_num(p_n_unfm_pointer, 0); - put_block_num(p_n_unfm_pointer, 0, 0); - journal_mark_dirty(th, p_s_sb, p_s_bh); - reiserfs_free_block(th, inode, tmp, 1); - if (item_moved(&s_ih, p_s_path)) { - need_research = 1; - break; - } - } + cond_resched(); - /* a trick. If the buffer has been logged, this - ** will do nothing. If we've broken the loop without - ** logging it, it will restore the buffer - ** - */ - reiserfs_restore_prepared_buffer(p_s_sb, p_s_bh); - - /* This loop can be optimized. */ - } while ((*p_n_removed < n_unfm_number || need_research) && - search_for_position_by_key(p_s_sb, p_s_item_key, - p_s_path) == - POSITION_FOUND); - - RFALSE(*p_n_removed < n_unfm_number, - "PAP-5310: indirect item is not found"); - RFALSE(item_moved(&s_ih, p_s_path), - "after while, comp failed, retry"); - - if (c_mode == M_CUT) - pos_in_item(p_s_path) *= UNFM_P_SIZE; - return c_mode; + if (item_moved (&s_ih, p_s_path)) { + need_re_search = 1; + break; + } + + pos --; + (*p_n_removed) ++; + (*p_n_cut_size) -= UNFM_P_SIZE; + + if (pos == 0) { + (*p_n_cut_size) -= IH_SIZE; + result = M_DELETE; + break; + } + } + /* a trick. If the buffer has been logged, this will do nothing. If + ** we've broken the loop without logging it, it will restore the + ** buffer */ + reiserfs_restore_prepared_buffer(p_s_sb, p_s_bh); + } while (need_re_search && + search_for_position_by_key(p_s_sb, p_s_item_key, p_s_path) == POSITION_FOUND); + pos_in_item(p_s_path) = pos * UNFM_P_SIZE; + + if (*p_n_cut_size == 0) { + /* Nothing were cut. maybe convert last unformatted node to the + * direct item? */ + result = M_CONVERT; + } + return result; } } @@ -1948,7 +1881,8 @@ int reiserfs_do_truncate(struct reiserfs_transaction_handle *th, struct inode *p ** sure the file is consistent before ending the current trans ** and starting a new one */ - if (journal_transaction_should_end(th, th->t_blocks_allocated)) { + if (journal_transaction_should_end(th, 0) || + reiserfs_transaction_free_space(th) <= JOURNAL_FOR_FREE_BLOCK_AND_UPDATE_SD) { int orig_len_alloc = th->t_blocks_allocated; decrement_counters_in_path(&s_search_path); @@ -1962,7 +1896,7 @@ int reiserfs_do_truncate(struct reiserfs_transaction_handle *th, struct inode *p if (err) goto out; err = journal_begin(th, p_s_inode->i_sb, - JOURNAL_PER_BALANCE_CNT * 6); + JOURNAL_FOR_FREE_BLOCK_AND_UPDATE_SD + JOURNAL_PER_BALANCE_CNT * 4) ; if (err) goto out; reiserfs_update_inode_transaction(p_s_inode); diff --git a/fs/reiserfs/super.c b/fs/reiserfs/super.c index 93e6ef9360e..cae2abbc0c7 100644 --- a/fs/reiserfs/super.c +++ b/fs/reiserfs/super.c @@ -685,14 +685,14 @@ static const arg_desc_t logging_mode[] = { (1 << REISERFS_DATA_ORDERED | 1 << REISERFS_DATA_WRITEBACK)}, {"writeback", 1 << REISERFS_DATA_WRITEBACK, (1 << REISERFS_DATA_ORDERED | 1 << REISERFS_DATA_LOG)}, - {NULL, 0} + {.value = NULL} }; /* possible values for -o barrier= */ static const arg_desc_t barrier_mode[] = { {"none", 1 << REISERFS_BARRIER_NONE, 1 << REISERFS_BARRIER_FLUSH}, {"flush", 1 << REISERFS_BARRIER_FLUSH, 1 << REISERFS_BARRIER_NONE}, - {NULL, 0} + {.value = NULL} }; /* possible values for "-o block-allocator=" and bits which are to be set in @@ -890,7 +890,7 @@ static int reiserfs_parse_options(struct super_block *s, char *options, /* strin {"acl",.setmask = 1 << REISERFS_UNSUPPORTED_OPT}, {"noacl",.clrmask = 1 << REISERFS_UNSUPPORTED_OPT}, #endif - {"nolog",}, /* This is unsupported */ + {.option_name = "nolog"}, {"replayonly",.setmask = 1 << REPLAYONLY}, {"block-allocator",.arg_required = 'a',.values = balloc}, {"data",.arg_required = 'd',.values = logging_mode}, @@ -908,7 +908,7 @@ static int reiserfs_parse_options(struct super_block *s, char *options, /* strin {"grpjquota",.arg_required = 'g' | (1 << REISERFS_OPT_ALLOWEMPTY),.values = NULL}, {"jqfmt",.arg_required = 'f',.values = NULL}, - {NULL,} + {.option_name = NULL} }; *blocks = 0; diff --git a/fs/reiserfs/xattr_acl.c b/fs/reiserfs/xattr_acl.c index ab8894c3b9e..58c418fbca2 100644 --- a/fs/reiserfs/xattr_acl.c +++ b/fs/reiserfs/xattr_acl.c @@ -182,7 +182,7 @@ struct posix_acl *reiserfs_get_acl(struct inode *inode, int type) { char *name, *value; struct posix_acl *acl, **p_acl; - size_t size; + int size; int retval; struct reiserfs_inode_info *reiserfs_i = REISERFS_I(inode); @@ -206,7 +206,7 @@ struct posix_acl *reiserfs_get_acl(struct inode *inode, int type) return posix_acl_dup(*p_acl); size = reiserfs_xattr_get(inode, name, NULL, 0); - if ((int)size < 0) { + if (size < 0) { if (size == -ENODATA || size == -ENOSYS) { *p_acl = ERR_PTR(-ENODATA); return NULL; diff --git a/fs/romfs/inode.c b/fs/romfs/inode.c index c2fc424d7d5..9b9eda7b335 100644 --- a/fs/romfs/inode.c +++ b/fs/romfs/inode.c @@ -463,7 +463,7 @@ static struct address_space_operations romfs_aops = { .readpage = romfs_readpage }; -static struct file_operations romfs_dir_operations = { +static const struct file_operations romfs_dir_operations = { .read = generic_read_dir, .readdir = romfs_readdir, }; diff --git a/fs/select.c b/fs/select.c index 1815a57d225..b3a3a1326af 100644 --- a/fs/select.c +++ b/fs/select.c @@ -29,12 +29,6 @@ #define ROUND_UP(x,y) (((x)+(y)-1)/(y)) #define DEFAULT_POLLMASK (POLLIN | POLLOUT | POLLRDNORM | POLLWRNORM) -struct poll_table_entry { - struct file * filp; - wait_queue_t wait; - wait_queue_head_t * wait_address; -}; - struct poll_table_page { struct poll_table_page * next; struct poll_table_entry * entry; @@ -64,13 +58,23 @@ void poll_initwait(struct poll_wqueues *pwq) init_poll_funcptr(&pwq->pt, __pollwait); pwq->error = 0; pwq->table = NULL; + pwq->inline_index = 0; } EXPORT_SYMBOL(poll_initwait); +static void free_poll_entry(struct poll_table_entry *entry) +{ + remove_wait_queue(entry->wait_address,&entry->wait); + fput(entry->filp); +} + void poll_freewait(struct poll_wqueues *pwq) { struct poll_table_page * p = pwq->table; + int i; + for (i = 0; i < pwq->inline_index; i++) + free_poll_entry(pwq->inline_entries + i); while (p) { struct poll_table_entry * entry; struct poll_table_page *old; @@ -78,8 +82,7 @@ void poll_freewait(struct poll_wqueues *pwq) entry = p->entry; do { entry--; - remove_wait_queue(entry->wait_address,&entry->wait); - fput(entry->filp); + free_poll_entry(entry); } while (entry > p->entries); old = p; p = p->next; @@ -89,12 +92,14 @@ void poll_freewait(struct poll_wqueues *pwq) EXPORT_SYMBOL(poll_freewait); -static void __pollwait(struct file *filp, wait_queue_head_t *wait_address, - poll_table *_p) +static struct poll_table_entry *poll_get_entry(poll_table *_p) { struct poll_wqueues *p = container_of(_p, struct poll_wqueues, pt); struct poll_table_page *table = p->table; + if (p->inline_index < N_INLINE_POLL_ENTRIES) + return p->inline_entries + p->inline_index++; + if (!table || POLL_TABLE_FULL(table)) { struct poll_table_page *new_table; @@ -102,7 +107,7 @@ static void __pollwait(struct file *filp, wait_queue_head_t *wait_address, if (!new_table) { p->error = -ENOMEM; __set_current_state(TASK_RUNNING); - return; + return NULL; } new_table->entry = new_table->entries; new_table->next = table; @@ -110,16 +115,21 @@ static void __pollwait(struct file *filp, wait_queue_head_t *wait_address, table = new_table; } - /* Add a new entry */ - { - struct poll_table_entry * entry = table->entry; - table->entry = entry+1; - get_file(filp); - entry->filp = filp; - entry->wait_address = wait_address; - init_waitqueue_entry(&entry->wait, current); - add_wait_queue(wait_address,&entry->wait); - } + return table->entry++; +} + +/* Add a new entry */ +static void __pollwait(struct file *filp, wait_queue_head_t *wait_address, + poll_table *p) +{ + struct poll_table_entry *entry = poll_get_entry(p); + if (!entry) + return; + get_file(filp); + entry->filp = filp; + entry->wait_address = wait_address; + init_waitqueue_entry(&entry->wait, current); + add_wait_queue(wait_address,&entry->wait); } #define FDS_IN(fds, n) (fds->in + n) @@ -210,7 +220,7 @@ int do_select(int n, fd_set_bits *fds, s64 *timeout) for (i = 0; i < n; ++rinp, ++routp, ++rexp) { unsigned long in, out, ex, all_bits, bit = 1, mask, j; unsigned long res_in = 0, res_out = 0, res_ex = 0; - struct file_operations *f_op = NULL; + const struct file_operations *f_op = NULL; struct file *file = NULL; in = *inp++; out = *outp++; ex = *exp++; @@ -221,17 +231,18 @@ int do_select(int n, fd_set_bits *fds, s64 *timeout) } for (j = 0; j < __NFDBITS; ++j, ++i, bit <<= 1) { + int fput_needed; if (i >= n) break; if (!(bit & all_bits)) continue; - file = fget(i); + file = fget_light(i, &fput_needed); if (file) { f_op = file->f_op; mask = DEFAULT_POLLMASK; if (f_op && f_op->poll) mask = (*f_op->poll)(file, retval ? NULL : wait); - fput(file); + fput_light(file, fput_needed); if ((mask & POLLIN_SET) && (in & bit)) { res_in |= bit; retval++; @@ -284,16 +295,6 @@ int do_select(int n, fd_set_bits *fds, s64 *timeout) return retval; } -static void *select_bits_alloc(int size) -{ - return kmalloc(6 * size, GFP_KERNEL); -} - -static void select_bits_free(void *bits, int size) -{ - kfree(bits); -} - /* * We can actually return ERESTARTSYS instead of EINTR, but I'd * like to be certain this leads to no problems. So I return @@ -312,6 +313,8 @@ static int core_sys_select(int n, fd_set __user *inp, fd_set __user *outp, char *bits; int ret, size, max_fdset; struct fdtable *fdt; + /* Allocate small arguments on the stack to save memory and be faster */ + char stack_fds[SELECT_STACK_ALLOC]; ret = -EINVAL; if (n < 0) @@ -332,7 +335,10 @@ static int core_sys_select(int n, fd_set __user *inp, fd_set __user *outp, */ ret = -ENOMEM; size = FDS_BYTES(n); - bits = select_bits_alloc(size); + if (6*size < SELECT_STACK_ALLOC) + bits = stack_fds; + else + bits = kmalloc(6 * size, GFP_KERNEL); if (!bits) goto out_nofds; fds.in = (unsigned long *) bits; @@ -367,7 +373,8 @@ static int core_sys_select(int n, fd_set __user *inp, fd_set __user *outp, ret = -EFAULT; out: - select_bits_free(bits, size); + if (bits != stack_fds) + kfree(bits); out_nofds: return ret; } @@ -551,14 +558,15 @@ static void do_pollfd(unsigned int num, struct pollfd * fdpage, fdp = fdpage+i; fd = fdp->fd; if (fd >= 0) { - struct file * file = fget(fd); + int fput_needed; + struct file * file = fget_light(fd, &fput_needed); mask = POLLNVAL; if (file != NULL) { mask = DEFAULT_POLLMASK; if (file->f_op && file->f_op->poll) mask = file->f_op->poll(file, *pwait); mask &= fdp->events | POLLERR | POLLHUP; - fput(file); + fput_light(file, fput_needed); } if (mask) { *pwait = NULL; @@ -619,6 +627,9 @@ static int do_poll(unsigned int nfds, struct poll_list *list, return count; } +#define N_STACK_PPS ((sizeof(stack_pps) - sizeof(struct poll_list)) / \ + sizeof(struct pollfd)) + int do_sys_poll(struct pollfd __user *ufds, unsigned int nfds, s64 *timeout) { struct poll_wqueues table; @@ -628,6 +639,9 @@ int do_sys_poll(struct pollfd __user *ufds, unsigned int nfds, s64 *timeout) struct poll_list *walk; struct fdtable *fdt; int max_fdset; + /* Allocate small arguments on the stack to save memory and be faster */ + char stack_pps[POLL_STACK_ALLOC]; + struct poll_list *stack_pp = NULL; /* Do a sanity check on nfds ... */ rcu_read_lock(); @@ -645,14 +659,23 @@ int do_sys_poll(struct pollfd __user *ufds, unsigned int nfds, s64 *timeout) err = -ENOMEM; while(i!=0) { struct poll_list *pp; - pp = kmalloc(sizeof(struct poll_list)+ - sizeof(struct pollfd)* - (i>POLLFD_PER_PAGE?POLLFD_PER_PAGE:i), - GFP_KERNEL); - if(pp==NULL) - goto out_fds; + int num, size; + if (stack_pp == NULL) + num = N_STACK_PPS; + else + num = POLLFD_PER_PAGE; + if (num > i) + num = i; + size = sizeof(struct poll_list) + sizeof(struct pollfd)*num; + if (!stack_pp) + stack_pp = pp = (struct poll_list *)stack_pps; + else { + pp = kmalloc(size, GFP_KERNEL); + if (!pp) + goto out_fds; + } pp->next=NULL; - pp->len = (i>POLLFD_PER_PAGE?POLLFD_PER_PAGE:i); + pp->len = num; if (head == NULL) head = pp; else @@ -660,7 +683,7 @@ int do_sys_poll(struct pollfd __user *ufds, unsigned int nfds, s64 *timeout) walk = pp; if (copy_from_user(pp->entries, ufds + nfds-i, - sizeof(struct pollfd)*pp->len)) { + sizeof(struct pollfd)*num)) { err = -EFAULT; goto out_fds; } @@ -689,7 +712,8 @@ out_fds: walk = head; while(walk!=NULL) { struct poll_list *pp = walk->next; - kfree(walk); + if (walk != stack_pp) + kfree(walk); walk = pp; } poll_freewait(&table); diff --git a/fs/smbfs/dir.c b/fs/smbfs/dir.c index 0424d06b147..34c7a11d91f 100644 --- a/fs/smbfs/dir.c +++ b/fs/smbfs/dir.c @@ -34,7 +34,7 @@ static int smb_rename(struct inode *, struct dentry *, static int smb_make_node(struct inode *,struct dentry *,int,dev_t); static int smb_link(struct dentry *, struct inode *, struct dentry *); -struct file_operations smb_dir_operations = +const struct file_operations smb_dir_operations = { .read = generic_read_dir, .readdir = smb_readdir, diff --git a/fs/smbfs/file.c b/fs/smbfs/file.c index 7042e62726a..c56bd99a970 100644 --- a/fs/smbfs/file.c +++ b/fs/smbfs/file.c @@ -401,7 +401,7 @@ smb_file_permission(struct inode *inode, int mask, struct nameidata *nd) return error; } -struct file_operations smb_file_operations = +const struct file_operations smb_file_operations = { .llseek = remote_llseek, .read = smb_file_read, diff --git a/fs/smbfs/inode.c b/fs/smbfs/inode.c index 44ed1d418b4..fdeabc0a34f 100644 --- a/fs/smbfs/inode.c +++ b/fs/smbfs/inode.c @@ -217,7 +217,7 @@ smb_set_inode_attr(struct inode *inode, struct smb_fattr *fattr) if (inode->i_mtime.tv_sec != last_time || inode->i_size != last_sz) { VERBOSE("%ld changed, old=%ld, new=%ld, oz=%ld, nz=%ld\n", inode->i_ino, - (long) last_time, (long) inode->i_mtime, + (long) last_time, (long) inode->i_mtime.tv_sec, (long) last_sz, (long) inode->i_size); if (!S_ISDIR(inode->i_mode)) diff --git a/fs/smbfs/proto.h b/fs/smbfs/proto.h index e866ec8660d..47664597e6b 100644 --- a/fs/smbfs/proto.h +++ b/fs/smbfs/proto.h @@ -35,7 +35,7 @@ extern int smb_proc_symlink(struct smb_sb_info *server, struct dentry *d, const extern int smb_proc_link(struct smb_sb_info *server, struct dentry *dentry, struct dentry *new_dentry); extern void smb_install_null_ops(struct smb_ops *ops); /* dir.c */ -extern struct file_operations smb_dir_operations; +extern const struct file_operations smb_dir_operations; extern struct inode_operations smb_dir_inode_operations; extern struct inode_operations smb_dir_inode_operations_unix; extern void smb_new_dentry(struct dentry *dentry); @@ -64,7 +64,7 @@ extern int smb_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat extern int smb_notify_change(struct dentry *dentry, struct iattr *attr); /* file.c */ extern struct address_space_operations smb_file_aops; -extern struct file_operations smb_file_operations; +extern const struct file_operations smb_file_operations; extern struct inode_operations smb_file_inode_operations; /* ioctl.c */ extern int smb_ioctl(struct inode *inode, struct file *filp, unsigned int cmd, unsigned long arg); diff --git a/fs/super.c b/fs/super.c index 37554b87618..a66f66bb804 100644 --- a/fs/super.c +++ b/fs/super.c @@ -37,6 +37,7 @@ #include <linux/writeback.h> /* for the emergency remount stuff */ #include <linux/idr.h> #include <linux/kobject.h> +#include <linux/mutex.h> #include <asm/uaccess.h> @@ -55,11 +56,10 @@ DEFINE_SPINLOCK(sb_lock); */ static struct super_block *alloc_super(void) { - struct super_block *s = kmalloc(sizeof(struct super_block), GFP_USER); + struct super_block *s = kzalloc(sizeof(struct super_block), GFP_USER); static struct super_operations default_op; if (s) { - memset(s, 0, sizeof(struct super_block)); if (security_sb_alloc(s)) { kfree(s); s = NULL; @@ -381,9 +381,9 @@ restart: void sync_filesystems(int wait) { struct super_block *sb; - static DECLARE_MUTEX(mutex); + static DEFINE_MUTEX(mutex); - down(&mutex); /* Could be down_interruptible */ + mutex_lock(&mutex); /* Could be down_interruptible */ spin_lock(&sb_lock); list_for_each_entry(sb, &super_blocks, s_list) { if (!sb->s_op->sync_fs) @@ -412,7 +412,7 @@ restart: goto restart; } spin_unlock(&sb_lock); - up(&mutex); + mutex_unlock(&mutex); } /** diff --git a/fs/sysfs/bin.c b/fs/sysfs/bin.c index 78899eeab97..c16a93c353c 100644 --- a/fs/sysfs/bin.c +++ b/fs/sysfs/bin.c @@ -163,7 +163,7 @@ static int release(struct inode * inode, struct file * file) return 0; } -struct file_operations bin_fops = { +const struct file_operations bin_fops = { .read = read, .write = write, .mmap = mmap, diff --git a/fs/sysfs/dir.c b/fs/sysfs/dir.c index 9ee95686444..f26880a4785 100644 --- a/fs/sysfs/dir.c +++ b/fs/sysfs/dir.c @@ -503,7 +503,7 @@ static loff_t sysfs_dir_lseek(struct file * file, loff_t offset, int origin) return offset; } -struct file_operations sysfs_dir_operations = { +const struct file_operations sysfs_dir_operations = { .open = sysfs_dir_open, .release = sysfs_dir_close, .llseek = sysfs_dir_lseek, diff --git a/fs/sysfs/file.c b/fs/sysfs/file.c index 5e83e724678..830f76fa098 100644 --- a/fs/sysfs/file.c +++ b/fs/sysfs/file.c @@ -348,7 +348,7 @@ static int sysfs_release(struct inode * inode, struct file * filp) return 0; } -struct file_operations sysfs_file_operations = { +const struct file_operations sysfs_file_operations = { .read = sysfs_read_file, .write = sysfs_write_file, .llseek = generic_file_llseek, diff --git a/fs/sysfs/sysfs.h b/fs/sysfs/sysfs.h index cf11d5b789d..32958a7c50e 100644 --- a/fs/sysfs/sysfs.h +++ b/fs/sysfs/sysfs.h @@ -21,9 +21,9 @@ extern int sysfs_setattr(struct dentry *dentry, struct iattr *iattr); extern struct rw_semaphore sysfs_rename_sem; extern struct super_block * sysfs_sb; -extern struct file_operations sysfs_dir_operations; -extern struct file_operations sysfs_file_operations; -extern struct file_operations bin_fops; +extern const struct file_operations sysfs_dir_operations; +extern const struct file_operations sysfs_file_operations; +extern const struct file_operations bin_fops; extern struct inode_operations sysfs_dir_inode_operations; extern struct inode_operations sysfs_symlink_inode_operations; diff --git a/fs/sysv/dir.c b/fs/sysv/dir.c index cce8b05cba5..8c66e9270dd 100644 --- a/fs/sysv/dir.c +++ b/fs/sysv/dir.c @@ -20,7 +20,7 @@ static int sysv_readdir(struct file *, void *, filldir_t); -struct file_operations sysv_dir_operations = { +const struct file_operations sysv_dir_operations = { .read = generic_read_dir, .readdir = sysv_readdir, .fsync = sysv_sync_file, diff --git a/fs/sysv/file.c b/fs/sysv/file.c index da69abc0624..a59e303135f 100644 --- a/fs/sysv/file.c +++ b/fs/sysv/file.c @@ -19,7 +19,7 @@ * We have mostly NULLs here: the current defaults are OK for * the coh filesystem. */ -struct file_operations sysv_file_operations = { +const struct file_operations sysv_file_operations = { .llseek = generic_file_llseek, .read = generic_file_read, .write = generic_file_write, diff --git a/fs/sysv/sysv.h b/fs/sysv/sysv.h index b7f9b4a42aa..393a480e4de 100644 --- a/fs/sysv/sysv.h +++ b/fs/sysv/sysv.h @@ -159,8 +159,8 @@ extern ino_t sysv_inode_by_name(struct dentry *); extern struct inode_operations sysv_file_inode_operations; extern struct inode_operations sysv_dir_inode_operations; extern struct inode_operations sysv_fast_symlink_inode_operations; -extern struct file_operations sysv_file_operations; -extern struct file_operations sysv_dir_operations; +extern const struct file_operations sysv_file_operations; +extern const struct file_operations sysv_dir_operations; extern struct address_space_operations sysv_aops; extern struct super_operations sysv_sops; extern struct dentry_operations sysv_dentry_operations; diff --git a/fs/udf/dir.c b/fs/udf/dir.c index f5222527fe3..8c28efa3b8f 100644 --- a/fs/udf/dir.c +++ b/fs/udf/dir.c @@ -42,7 +42,7 @@ static int do_udf_readdir(struct inode *, struct file *, filldir_t, void *); /* readdir and lookup functions */ -struct file_operations udf_dir_operations = { +const struct file_operations udf_dir_operations = { .read = generic_read_dir, .readdir = udf_readdir, .ioctl = udf_ioctl, diff --git a/fs/udf/file.c b/fs/udf/file.c index a6f2acc1f15..e34b00e303f 100644 --- a/fs/udf/file.c +++ b/fs/udf/file.c @@ -248,7 +248,7 @@ static int udf_release_file(struct inode * inode, struct file * filp) return 0; } -struct file_operations udf_file_operations = { +const struct file_operations udf_file_operations = { .read = generic_file_read, .ioctl = udf_ioctl, .open = generic_file_open, diff --git a/fs/udf/inode.c b/fs/udf/inode.c index d04cff2273b..81e0e8459af 100644 --- a/fs/udf/inode.c +++ b/fs/udf/inode.c @@ -1341,13 +1341,11 @@ udf_update_inode(struct inode *inode, int do_sync) if (UDF_QUERY_FLAG(inode->i_sb, UDF_FLAG_UID_FORGET)) fe->uid = cpu_to_le32(-1); - else if (inode->i_uid != UDF_SB(inode->i_sb)->s_uid) - fe->uid = cpu_to_le32(inode->i_uid); + else fe->uid = cpu_to_le32(inode->i_uid); if (UDF_QUERY_FLAG(inode->i_sb, UDF_FLAG_GID_FORGET)) fe->gid = cpu_to_le32(-1); - else if (inode->i_gid != UDF_SB(inode->i_sb)->s_gid) - fe->gid = cpu_to_le32(inode->i_gid); + else fe->gid = cpu_to_le32(inode->i_gid); udfperms = ((inode->i_mode & S_IRWXO) ) | ((inode->i_mode & S_IRWXG) << 2) | diff --git a/fs/udf/udfdecl.h b/fs/udf/udfdecl.h index 1d5800e0cbe..023e19ba5a2 100644 --- a/fs/udf/udfdecl.h +++ b/fs/udf/udfdecl.h @@ -44,9 +44,9 @@ struct buffer_head; struct super_block; extern struct inode_operations udf_dir_inode_operations; -extern struct file_operations udf_dir_operations; +extern const struct file_operations udf_dir_operations; extern struct inode_operations udf_file_inode_operations; -extern struct file_operations udf_file_operations; +extern const struct file_operations udf_file_operations; extern struct address_space_operations udf_aops; extern struct address_space_operations udf_adinicb_aops; extern struct address_space_operations udf_symlink_aops; diff --git a/fs/ufs/dir.c b/fs/ufs/dir.c index 7c10c68902a..1a561202d3f 100644 --- a/fs/ufs/dir.c +++ b/fs/ufs/dir.c @@ -620,7 +620,7 @@ int ufs_empty_dir (struct inode * inode) return 1; } -struct file_operations ufs_dir_operations = { +const struct file_operations ufs_dir_operations = { .read = generic_read_dir, .readdir = ufs_readdir, .fsync = file_fsync, diff --git a/fs/ufs/file.c b/fs/ufs/file.c index 62ad481810e..312fd3f8631 100644 --- a/fs/ufs/file.c +++ b/fs/ufs/file.c @@ -31,7 +31,7 @@ * the ufs filesystem. */ -struct file_operations ufs_file_operations = { +const struct file_operations ufs_file_operations = { .llseek = generic_file_llseek, .read = generic_file_read, .write = generic_file_write, diff --git a/fs/ufs/super.c b/fs/ufs/super.c index d257644a1ae..db98a4c71e6 100644 --- a/fs/ufs/super.c +++ b/fs/ufs/super.c @@ -575,7 +575,7 @@ static int ufs_fill_super(struct super_block *sb, void *data, int silent) if (!silent) printk("You didn't specify the type of your ufs filesystem\n\n" "mount -t ufs -o ufstype=" - "sun|sunx86|44bsd|ufs2|5xbsd|old|hp|nextstep|netxstep-cd|openstep ...\n\n" + "sun|sunx86|44bsd|ufs2|5xbsd|old|hp|nextstep|nextstep-cd|openstep ...\n\n" ">>>WARNING<<< Wrong ufstype may corrupt your filesystem, " "default is ufstype=old\n"); ufs_set_opt (sbi->s_mount_opt, UFSTYPE_OLD); diff --git a/fs/xattr.c b/fs/xattr.c index 80eca7d3d69..e416190f5e9 100644 --- a/fs/xattr.c +++ b/fs/xattr.c @@ -17,6 +17,7 @@ #include <linux/syscalls.h> #include <linux/module.h> #include <linux/fsnotify.h> +#include <linux/audit.h> #include <asm/uaccess.h> @@ -234,12 +235,15 @@ sys_fsetxattr(int fd, char __user *name, void __user *value, size_t size, int flags) { struct file *f; + struct dentry *dentry; int error = -EBADF; f = fget(fd); if (!f) return error; - error = setxattr(f->f_dentry, name, value, size, flags); + dentry = f->f_dentry; + audit_inode(NULL, dentry->d_inode, 0); + error = setxattr(dentry, name, value, size, flags); fput(f); return error; } @@ -458,12 +462,15 @@ asmlinkage long sys_fremovexattr(int fd, char __user *name) { struct file *f; + struct dentry *dentry; int error = -EBADF; f = fget(fd); if (!f) return error; - error = removexattr(f->f_dentry, name); + dentry = f->f_dentry; + audit_inode(NULL, dentry->d_inode, 0); + error = removexattr(dentry, name); fput(f); return error; } diff --git a/fs/xfs/linux-2.6/mrlock.h b/fs/xfs/linux-2.6/mrlock.h index 16b44c3c236..1b262b790d9 100644 --- a/fs/xfs/linux-2.6/mrlock.h +++ b/fs/xfs/linux-2.6/mrlock.h @@ -79,7 +79,7 @@ static inline void mrdemote(mrlock_t *mrp) * Debug-only routine, without some platform-specific asm code, we can * now only answer requests regarding whether we hold the lock for write * (reader state is outside our visibility, we only track writer state). - * Note: means !ismrlocked would give false positivies, so don't do that. + * Note: means !ismrlocked would give false positives, so don't do that. */ static inline int ismrlocked(mrlock_t *mrp, int type) { diff --git a/fs/xfs/linux-2.6/xfs_aops.c b/fs/xfs/linux-2.6/xfs_aops.c index 97fc056130e..6cbbd165c60 100644 --- a/fs/xfs/linux-2.6/xfs_aops.c +++ b/fs/xfs/linux-2.6/xfs_aops.c @@ -372,7 +372,7 @@ static inline int bio_add_buffer(struct bio *bio, struct buffer_head *bh) * assumes that all buffers on the page are started at the same time. * * The fix is two passes across the ioend list - one to start writeback on the - * bufferheads, and then the second one submit them for I/O. + * buffer_heads, and then submit them for I/O on the second pass. */ STATIC void xfs_submit_ioend( @@ -699,7 +699,7 @@ xfs_convert_page( /* * page_dirty is initially a count of buffers on the page before - * EOF and is decrememted as we move each into a cleanable state. + * EOF and is decremented as we move each into a cleanable state. * * Derivation: * @@ -842,7 +842,7 @@ xfs_cluster_write( * page if possible. * The bh->b_state's cannot know if any of the blocks or which block for * that matter are dirty due to mmap writes, and therefore bh uptodate is - * only vaild if the page itself isn't completely uptodate. Some layers + * only valid if the page itself isn't completely uptodate. Some layers * may clear the page dirty flag prior to calling write page, under the * assumption the entire page will be written out; by not writing out the * whole page the page can be reused before all valid dirty data is @@ -892,7 +892,7 @@ xfs_page_state_convert( /* * page_dirty is initially a count of buffers on the page before - * EOF and is decrememted as we move each into a cleanable state. + * EOF and is decremented as we move each into a cleanable state. * * Derivation: * @@ -1223,10 +1223,9 @@ free_buffers: } STATIC int -__xfs_get_block( +__xfs_get_blocks( struct inode *inode, sector_t iblock, - unsigned long blocks, struct buffer_head *bh_result, int create, int direct, @@ -1236,22 +1235,17 @@ __xfs_get_block( xfs_iomap_t iomap; xfs_off_t offset; ssize_t size; - int retpbbm = 1; + int niomap = 1; int error; offset = (xfs_off_t)iblock << inode->i_blkbits; - if (blocks) - size = (ssize_t) min_t(xfs_off_t, LONG_MAX, - (xfs_off_t)blocks << inode->i_blkbits); - else - size = 1 << inode->i_blkbits; - + ASSERT(bh_result->b_size >= (1 << inode->i_blkbits)); + size = bh_result->b_size; VOP_BMAP(vp, offset, size, - create ? flags : BMAPI_READ, &iomap, &retpbbm, error); + create ? flags : BMAPI_READ, &iomap, &niomap, error); if (error) return -error; - - if (retpbbm == 0) + if (niomap == 0) return 0; if (iomap.iomap_bn != IOMAP_DADDR_NULL) { @@ -1271,12 +1265,16 @@ __xfs_get_block( } } - /* If this is a realtime file, data might be on a new device */ + /* + * If this is a realtime file, data may be on a different device. + * to that pointed to from the buffer_head b_bdev currently. + */ bh_result->b_bdev = iomap.iomap_target->bt_bdev; - /* If we previously allocated a block out beyond eof and - * we are now coming back to use it then we will need to - * flag it as new even if it has a disk address. + /* + * If we previously allocated a block out beyond eof and we are + * now coming back to use it then we will need to flag it as new + * even if it has a disk address. */ if (create && ((!buffer_mapped(bh_result) && !buffer_uptodate(bh_result)) || @@ -1292,38 +1290,36 @@ __xfs_get_block( } } - if (blocks) { + if (direct || size > (1 << inode->i_blkbits)) { ASSERT(iomap.iomap_bsize - iomap.iomap_delta > 0); offset = min_t(xfs_off_t, - iomap.iomap_bsize - iomap.iomap_delta, - (xfs_off_t)blocks << inode->i_blkbits); - bh_result->b_size = (u32) min_t(xfs_off_t, UINT_MAX, offset); + iomap.iomap_bsize - iomap.iomap_delta, size); + bh_result->b_size = (ssize_t)min_t(xfs_off_t, LONG_MAX, offset); } return 0; } int -xfs_get_block( +xfs_get_blocks( struct inode *inode, sector_t iblock, struct buffer_head *bh_result, int create) { - return __xfs_get_block(inode, iblock, 0, bh_result, - create, 0, BMAPI_WRITE); + return __xfs_get_blocks(inode, iblock, + bh_result, create, 0, BMAPI_WRITE); } STATIC int xfs_get_blocks_direct( struct inode *inode, sector_t iblock, - unsigned long max_blocks, struct buffer_head *bh_result, int create) { - return __xfs_get_block(inode, iblock, max_blocks, bh_result, - create, 1, BMAPI_WRITE|BMAPI_DIRECT); + return __xfs_get_blocks(inode, iblock, + bh_result, create, 1, BMAPI_WRITE|BMAPI_DIRECT); } STATIC void @@ -1338,9 +1334,9 @@ xfs_end_io_direct( /* * Non-NULL private data means we need to issue a transaction to * convert a range from unwritten to written extents. This needs - * to happen from process contect but aio+dio I/O completion + * to happen from process context but aio+dio I/O completion * happens from irq context so we need to defer it to a workqueue. - * This is not nessecary for synchronous direct I/O, but we do + * This is not necessary for synchronous direct I/O, but we do * it anyway to keep the code uniform and simpler. * * The core direct I/O code might be changed to always call the @@ -1357,7 +1353,7 @@ xfs_end_io_direct( } /* - * blockdev_direct_IO can return an error even afer the I/O + * blockdev_direct_IO can return an error even after the I/O * completion handler was called. Thus we need to protect * against double-freeing. */ @@ -1404,7 +1400,7 @@ xfs_vm_prepare_write( unsigned int from, unsigned int to) { - return block_prepare_write(page, from, to, xfs_get_block); + return block_prepare_write(page, from, to, xfs_get_blocks); } STATIC sector_t @@ -1421,7 +1417,7 @@ xfs_vm_bmap( VOP_RWLOCK(vp, VRWLOCK_READ); VOP_FLUSH_PAGES(vp, (xfs_off_t)0, -1, 0, FI_REMAPF, error); VOP_RWUNLOCK(vp, VRWLOCK_READ); - return generic_block_bmap(mapping, block, xfs_get_block); + return generic_block_bmap(mapping, block, xfs_get_blocks); } STATIC int @@ -1429,7 +1425,7 @@ xfs_vm_readpage( struct file *unused, struct page *page) { - return mpage_readpage(page, xfs_get_block); + return mpage_readpage(page, xfs_get_blocks); } STATIC int @@ -1439,17 +1435,17 @@ xfs_vm_readpages( struct list_head *pages, unsigned nr_pages) { - return mpage_readpages(mapping, pages, nr_pages, xfs_get_block); + return mpage_readpages(mapping, pages, nr_pages, xfs_get_blocks); } -STATIC int +STATIC void xfs_vm_invalidatepage( struct page *page, unsigned long offset) { xfs_page_trace(XFS_INVALIDPAGE_ENTER, page->mapping->host, page, offset); - return block_invalidatepage(page, offset); + block_invalidatepage(page, offset); } struct address_space_operations xfs_address_space_operations = { diff --git a/fs/xfs/linux-2.6/xfs_aops.h b/fs/xfs/linux-2.6/xfs_aops.h index 795699f121d..60716543c68 100644 --- a/fs/xfs/linux-2.6/xfs_aops.h +++ b/fs/xfs/linux-2.6/xfs_aops.h @@ -41,6 +41,6 @@ typedef struct xfs_ioend { } xfs_ioend_t; extern struct address_space_operations xfs_address_space_operations; -extern int xfs_get_block(struct inode *, sector_t, struct buffer_head *, int); +extern int xfs_get_blocks(struct inode *, sector_t, struct buffer_head *, int); #endif /* __XFS_IOPS_H__ */ diff --git a/fs/xfs/linux-2.6/xfs_export.h b/fs/xfs/linux-2.6/xfs_export.h index e5b0559700a..e794ca4efc7 100644 --- a/fs/xfs/linux-2.6/xfs_export.h +++ b/fs/xfs/linux-2.6/xfs_export.h @@ -54,7 +54,7 @@ * Note, the NFS filehandle also includes an fsid portion which * may have an inode number in it. That number is hardcoded to * 32bits and there is no way for XFS to intercept it. In - * practice this means when exporting an XFS filesytem with 64bit + * practice this means when exporting an XFS filesystem with 64bit * inodes you should either export the mountpoint (rather than * a subdirectory) or use the "fsid" export option. */ diff --git a/fs/xfs/linux-2.6/xfs_file.c b/fs/xfs/linux-2.6/xfs_file.c index 185567a6a56..85997b1205f 100644 --- a/fs/xfs/linux-2.6/xfs_file.c +++ b/fs/xfs/linux-2.6/xfs_file.c @@ -528,7 +528,7 @@ open_exec_out: } #endif /* HAVE_FOP_OPEN_EXEC */ -struct file_operations xfs_file_operations = { +const struct file_operations xfs_file_operations = { .llseek = generic_file_llseek, .read = do_sync_read, .write = do_sync_write, @@ -550,7 +550,7 @@ struct file_operations xfs_file_operations = { #endif }; -struct file_operations xfs_invis_file_operations = { +const struct file_operations xfs_invis_file_operations = { .llseek = generic_file_llseek, .read = do_sync_read, .write = do_sync_write, @@ -570,7 +570,7 @@ struct file_operations xfs_invis_file_operations = { }; -struct file_operations xfs_dir_file_operations = { +const struct file_operations xfs_dir_file_operations = { .read = generic_read_dir, .readdir = xfs_file_readdir, .unlocked_ioctl = xfs_file_ioctl, diff --git a/fs/xfs/linux-2.6/xfs_ioctl32.c b/fs/xfs/linux-2.6/xfs_ioctl32.c index b6321abd9a8..251bfe451a3 100644 --- a/fs/xfs/linux-2.6/xfs_ioctl32.c +++ b/fs/xfs/linux-2.6/xfs_ioctl32.c @@ -72,7 +72,7 @@ xfs_ioctl32_flock( copy_in_user(&p->l_pid, &p32->l_pid, sizeof(u32)) || copy_in_user(&p->l_pad, &p32->l_pad, 4*sizeof(u32))) return -EFAULT; - + return (unsigned long)p; } @@ -107,11 +107,15 @@ xfs_ioctl32_bulkstat( #endif STATIC long -xfs_compat_ioctl(int mode, struct file *f, unsigned cmd, unsigned long arg) +xfs_compat_ioctl( + int mode, + struct file *file, + unsigned cmd, + unsigned long arg) { + struct inode *inode = file->f_dentry->d_inode; + vnode_t *vp = vn_from_inode(inode); int error; - struct inode *inode = f->f_dentry->d_inode; - vnode_t *vp = vn_to_inode(inode); switch (cmd) { case XFS_IOC_DIOINFO: @@ -189,7 +193,7 @@ xfs_compat_ioctl(int mode, struct file *f, unsigned cmd, unsigned long arg) return -ENOIOCTLCMD; } - VOP_IOCTL(vp, inode, f, mode, cmd, (void __user *)arg, error); + VOP_IOCTL(vp, inode, file, mode, cmd, (void __user *)arg, error); VMODIFY(vp); return error; @@ -197,18 +201,18 @@ xfs_compat_ioctl(int mode, struct file *f, unsigned cmd, unsigned long arg) long xfs_file_compat_ioctl( - struct file *f, + struct file *file, unsigned cmd, unsigned long arg) { - return xfs_compat_ioctl(0, f, cmd, arg); + return xfs_compat_ioctl(0, file, cmd, arg); } long xfs_file_compat_invis_ioctl( - struct file *f, + struct file *file, unsigned cmd, unsigned long arg) { - return xfs_compat_ioctl(IO_INVIS, f, cmd, arg); + return xfs_compat_ioctl(IO_INVIS, file, cmd, arg); } diff --git a/fs/xfs/linux-2.6/xfs_iops.c b/fs/xfs/linux-2.6/xfs_iops.c index af487437bd7..149237304fb 100644 --- a/fs/xfs/linux-2.6/xfs_iops.c +++ b/fs/xfs/linux-2.6/xfs_iops.c @@ -708,7 +708,7 @@ STATIC void xfs_vn_truncate( struct inode *inode) { - block_truncate_page(inode->i_mapping, inode->i_size, xfs_get_block); + block_truncate_page(inode->i_mapping, inode->i_size, xfs_get_blocks); } STATIC int diff --git a/fs/xfs/linux-2.6/xfs_iops.h b/fs/xfs/linux-2.6/xfs_iops.h index a8417d7af5f..ad6173da567 100644 --- a/fs/xfs/linux-2.6/xfs_iops.h +++ b/fs/xfs/linux-2.6/xfs_iops.h @@ -22,9 +22,9 @@ extern struct inode_operations xfs_inode_operations; extern struct inode_operations xfs_dir_inode_operations; extern struct inode_operations xfs_symlink_inode_operations; -extern struct file_operations xfs_file_operations; -extern struct file_operations xfs_dir_file_operations; -extern struct file_operations xfs_invis_file_operations; +extern const struct file_operations xfs_file_operations; +extern const struct file_operations xfs_dir_file_operations; +extern const struct file_operations xfs_invis_file_operations; extern int xfs_ioctl(struct bhv_desc *, struct inode *, struct file *, int, unsigned int, void __user *); diff --git a/fs/xfs/linux-2.6/xfs_lrw.c b/fs/xfs/linux-2.6/xfs_lrw.c index 0169360475c..84ddf189389 100644 --- a/fs/xfs/linux-2.6/xfs_lrw.c +++ b/fs/xfs/linux-2.6/xfs_lrw.c @@ -681,7 +681,7 @@ start: eventsent = 1; /* - * The iolock was dropped and reaquired in XFS_SEND_DATA + * The iolock was dropped and reacquired in XFS_SEND_DATA * so we have to recheck the size when appending. * We will only "goto start;" once, since having sent the * event prevents another call to XFS_SEND_DATA, which is diff --git a/fs/xfs/linux-2.6/xfs_super.c b/fs/xfs/linux-2.6/xfs_super.c index 8355faf8ffd..1884300417e 100644 --- a/fs/xfs/linux-2.6/xfs_super.c +++ b/fs/xfs/linux-2.6/xfs_super.c @@ -375,9 +375,8 @@ xfs_init_zones(void) if (!xfs_ioend_zone) goto out_destroy_vnode_zone; - xfs_ioend_pool = mempool_create(4 * MAX_BUF_PER_PAGE, - mempool_alloc_slab, mempool_free_slab, - xfs_ioend_zone); + xfs_ioend_pool = mempool_create_slab_pool(4 * MAX_BUF_PER_PAGE, + xfs_ioend_zone); if (!xfs_ioend_pool) goto out_free_ioend_zone; return 0; diff --git a/fs/xfs/linux-2.6/xfs_vfs.h b/fs/xfs/linux-2.6/xfs_vfs.h index 8fed356db05..841200c0309 100644 --- a/fs/xfs/linux-2.6/xfs_vfs.h +++ b/fs/xfs/linux-2.6/xfs_vfs.h @@ -92,7 +92,7 @@ typedef enum { #define SYNC_FSDATA 0x0020 /* flush fs data (e.g. superblocks) */ #define SYNC_REFCACHE 0x0040 /* prune some of the nfs ref cache */ #define SYNC_REMOUNT 0x0080 /* remount readonly, no dummy LRs */ -#define SYNC_QUIESCE 0x0100 /* quiesce fileystem for a snapshot */ +#define SYNC_QUIESCE 0x0100 /* quiesce filesystem for a snapshot */ typedef int (*vfs_mount_t)(bhv_desc_t *, struct xfs_mount_args *, struct cred *); diff --git a/fs/xfs/quota/xfs_dquot_item.c b/fs/xfs/quota/xfs_dquot_item.c index e4e5f05b841..546f48af882 100644 --- a/fs/xfs/quota/xfs_dquot_item.c +++ b/fs/xfs/quota/xfs_dquot_item.c @@ -221,7 +221,7 @@ xfs_qm_dqunpin_wait( * as possible. * * We must not be holding the AIL_LOCK at this point. Calling incore() to - * search the buffercache can be a time consuming thing, and AIL_LOCK is a + * search the buffer cache can be a time consuming thing, and AIL_LOCK is a * spinlock. */ STATIC void diff --git a/fs/xfs/quota/xfs_qm.c b/fs/xfs/quota/xfs_qm.c index 1fb757ef3f4..73c1e5e80c0 100644 --- a/fs/xfs/quota/xfs_qm.c +++ b/fs/xfs/quota/xfs_qm.c @@ -289,7 +289,7 @@ xfs_qm_rele_quotafs_ref( /* * This is called at mount time from xfs_mountfs to initialize the quotainfo - * structure and start the global quotamanager (xfs_Gqm) if it hasn't done + * structure and start the global quota manager (xfs_Gqm) if it hasn't done * so already. Note that the superblock has not been read in yet. */ void @@ -807,7 +807,7 @@ xfs_qm_dqattach_one( * Given a udquot and gdquot, attach a ptr to the group dquot in the * udquot as a hint for future lookups. The idea sounds simple, but the * execution isn't, because the udquot might have a group dquot attached - * already and getting rid of that gets us into lock ordering contraints. + * already and getting rid of that gets us into lock ordering constraints. * The process is complicated more by the fact that the dquots may or may not * be locked on entry. */ @@ -1094,10 +1094,10 @@ xfs_qm_sync( } /* * If we can't grab the flush lock then if the caller - * really wanted us to give this our best shot, + * really wanted us to give this our best shot, so * see if we can give a push to the buffer before we wait * on the flush lock. At this point, we know that - * eventhough the dquot is being flushed, + * even though the dquot is being flushed, * it has (new) dirty data. */ xfs_qm_dqflock_pushbuf_wait(dqp); @@ -1491,7 +1491,7 @@ xfs_qm_reset_dqcounts( /* * Do a sanity check, and if needed, repair the dqblk. Don't * output any warnings because it's perfectly possible to - * find unitialized dquot blks. See comment in xfs_qm_dqcheck. + * find uninitialised dquot blks. See comment in xfs_qm_dqcheck. */ (void) xfs_qm_dqcheck(ddq, id+j, type, XFS_QMOPT_DQREPAIR, "xfs_quotacheck"); @@ -1580,7 +1580,7 @@ xfs_qm_dqiterate( error = 0; /* - * This looks racey, but we can't keep an inode lock across a + * This looks racy, but we can't keep an inode lock across a * trans_reserve. But, this gets called during quotacheck, and that * happens only at mount time which is single threaded. */ @@ -1824,7 +1824,7 @@ xfs_qm_dqusage_adjust( * we have to start from the beginning anyway. * Once we're done, we'll log all the dquot bufs. * - * The *QUOTA_ON checks below may look pretty racey, but quotachecks + * The *QUOTA_ON checks below may look pretty racy, but quotachecks * and quotaoffs don't race. (Quotachecks happen at mount time only). */ if (XFS_IS_UQUOTA_ON(mp)) { diff --git a/fs/xfs/quota/xfs_qm_syscalls.c b/fs/xfs/quota/xfs_qm_syscalls.c index 676884394aa..c55db463bbf 100644 --- a/fs/xfs/quota/xfs_qm_syscalls.c +++ b/fs/xfs/quota/xfs_qm_syscalls.c @@ -912,7 +912,7 @@ xfs_qm_export_dquot( /* * Internally, we don't reset all the timers when quota enforcement - * gets turned off. No need to confuse the userlevel code, + * gets turned off. No need to confuse the user level code, * so return zeroes in that case. */ if (! XFS_IS_QUOTA_ENFORCED(mp)) { diff --git a/fs/xfs/quota/xfs_trans_dquot.c b/fs/xfs/quota/xfs_trans_dquot.c index 3290975d31f..d8e131ec0aa 100644 --- a/fs/xfs/quota/xfs_trans_dquot.c +++ b/fs/xfs/quota/xfs_trans_dquot.c @@ -804,7 +804,7 @@ xfs_trans_reserve_quota_bydquots( } /* - * Didnt change anything critical, so, no need to log + * Didn't change anything critical, so, no need to log */ return (0); } diff --git a/fs/xfs/xfs_acl.c b/fs/xfs/xfs_acl.c index 4ff0f4e41c6..2539af34eb6 100644 --- a/fs/xfs/xfs_acl.c +++ b/fs/xfs/xfs_acl.c @@ -395,7 +395,7 @@ xfs_acl_allow_set( * The access control process to determine the access permission: * if uid == file owner id, use the file owner bits. * if gid == file owner group id, use the file group bits. - * scan ACL for a maching user or group, and use matched entry + * scan ACL for a matching user or group, and use matched entry * permission. Use total permissions of all matching group entries, * until all acl entries are exhausted. The final permission produced * by matching acl entry or entries needs to be & with group permission. diff --git a/fs/xfs/xfs_ag.h b/fs/xfs/xfs_ag.h index a96e2ffce0c..dc2361dd740 100644 --- a/fs/xfs/xfs_ag.h +++ b/fs/xfs/xfs_ag.h @@ -179,7 +179,7 @@ typedef struct xfs_perag { char pagf_init; /* this agf's entry is initialized */ char pagi_init; /* this agi's entry is initialized */ - char pagf_metadata; /* the agf is prefered to be metadata */ + char pagf_metadata; /* the agf is preferred to be metadata */ char pagi_inodeok; /* The agi is ok for inodes */ __uint8_t pagf_levels[XFS_BTNUM_AGF]; /* # of levels in bno & cnt btree */ diff --git a/fs/xfs/xfs_alloc.c b/fs/xfs/xfs_alloc.c index f4328e1e2a7..64ee07db0d5 100644 --- a/fs/xfs/xfs_alloc.c +++ b/fs/xfs/xfs_alloc.c @@ -511,7 +511,7 @@ STATIC void xfs_alloc_trace_busy( char *name, /* function tag string */ char *str, /* additional string */ - xfs_mount_t *mp, /* file system mount poing */ + xfs_mount_t *mp, /* file system mount point */ xfs_agnumber_t agno, /* allocation group number */ xfs_agblock_t agbno, /* a.g. relative block number */ xfs_extlen_t len, /* length of extent */ @@ -1843,7 +1843,7 @@ xfs_alloc_fix_freelist( } else agbp = NULL; - /* If this is a metadata prefered pag and we are user data + /* If this is a metadata preferred pag and we are user data * then try somewhere else if we are not being asked to * try harder at this point */ @@ -2458,7 +2458,7 @@ error0: /* * AG Busy list management * The busy list contains block ranges that have been freed but whose - * transacations have not yet hit disk. If any block listed in a busy + * transactions have not yet hit disk. If any block listed in a busy * list is reused, the transaction that freed it must be forced to disk * before continuing to use the block. * diff --git a/fs/xfs/xfs_alloc.h b/fs/xfs/xfs_alloc.h index 3546dea27b7..2d1f8928b26 100644 --- a/fs/xfs/xfs_alloc.h +++ b/fs/xfs/xfs_alloc.h @@ -68,7 +68,7 @@ typedef struct xfs_alloc_arg { xfs_alloctype_t otype; /* original allocation type */ char wasdel; /* set if allocation was prev delayed */ char wasfromfl; /* set if allocation is from freelist */ - char isfl; /* set if is freelist blocks - !actg */ + char isfl; /* set if is freelist blocks - !acctg */ char userdata; /* set if this is user data */ } xfs_alloc_arg_t; diff --git a/fs/xfs/xfs_attr.c b/fs/xfs/xfs_attr.c index 093fac476bd..b6e1e02bbb2 100644 --- a/fs/xfs/xfs_attr.c +++ b/fs/xfs/xfs_attr.c @@ -294,7 +294,7 @@ xfs_attr_set_int(xfs_inode_t *dp, const char *name, int namelen, xfs_trans_ihold(args.trans, dp); /* - * If the attribute list is non-existant or a shortform list, + * If the attribute list is non-existent or a shortform list, * upgrade it to a single-leaf-block attribute list. */ if ((dp->i_d.di_aformat == XFS_DINODE_FMT_LOCAL) || @@ -1584,7 +1584,7 @@ out: * Fill in the disk block numbers in the state structure for the buffers * that are attached to the state structure. * This is done so that we can quickly reattach ourselves to those buffers - * after some set of transaction commit's has released these buffers. + * after some set of transaction commits have released these buffers. */ STATIC int xfs_attr_fillstate(xfs_da_state_t *state) @@ -1631,7 +1631,7 @@ xfs_attr_fillstate(xfs_da_state_t *state) /* * Reattach the buffers to the state structure based on the disk block * numbers stored in the state structure. - * This is done after some set of transaction commit's has released those + * This is done after some set of transaction commits have released those * buffers from our grip. */ STATIC int diff --git a/fs/xfs/xfs_attr_leaf.c b/fs/xfs/xfs_attr_leaf.c index 717682747bd..9462be86aa1 100644 --- a/fs/xfs/xfs_attr_leaf.c +++ b/fs/xfs/xfs_attr_leaf.c @@ -524,7 +524,7 @@ xfs_attr_shortform_compare(const void *a, const void *b) /* * Copy out entries of shortform attribute lists for attr_list(). - * Shortform atrtribute lists are not stored in hashval sorted order. + * Shortform attribute lists are not stored in hashval sorted order. * If the output buffer is not large enough to hold them all, then we * we have to calculate each entries' hashvalue and sort them before * we can begin returning them to the user. @@ -1541,7 +1541,7 @@ xfs_attr_leaf_toosmall(xfs_da_state_t *state, int *action) /* * Check for the degenerate case of the block being empty. * If the block is empty, we'll simply delete it, no need to - * coalesce it with a sibling block. We choose (aribtrarily) + * coalesce it with a sibling block. We choose (arbitrarily) * to merge with the forward block unless it is NULL. */ if (count == 0) { diff --git a/fs/xfs/xfs_behavior.c b/fs/xfs/xfs_behavior.c index 9880adae393..f4fe3715a80 100644 --- a/fs/xfs/xfs_behavior.c +++ b/fs/xfs/xfs_behavior.c @@ -31,7 +31,7 @@ * The behavior chain is ordered based on the 'position' number which * lives in the first field of the ops vector (higher numbers first). * - * Attemps to insert duplicate ops result in an EINVAL return code. + * Attempts to insert duplicate ops result in an EINVAL return code. * Otherwise, return 0 to indicate success. */ int @@ -84,7 +84,7 @@ bhv_insert(bhv_head_t *bhp, bhv_desc_t *bdp) /* * Remove a behavior descriptor from a position in a behavior chain; - * the postition is guaranteed not to be the first position. + * the position is guaranteed not to be the first position. * Should only be called by the bhv_remove() macro. */ void diff --git a/fs/xfs/xfs_behavior.h b/fs/xfs/xfs_behavior.h index 2cd89bb5ab1..1d8ff103201 100644 --- a/fs/xfs/xfs_behavior.h +++ b/fs/xfs/xfs_behavior.h @@ -39,7 +39,7 @@ * behaviors is synchronized with operations-in-progress (oip's) so that * the oip's always see a consistent view of the chain. * - * The term "interpostion" is used to refer to the act of inserting + * The term "interposition" is used to refer to the act of inserting * a behavior such that it interposes on (i.e., is inserted in front * of) a particular other behavior. A key example of this is when a * system implementing distributed single system image wishes to @@ -51,7 +51,7 @@ * * Behavior synchronization is logic which is necessary under certain * circumstances that there is no conflict between ongoing operations - * traversing the behavior chain and those dunamically modifying the + * traversing the behavior chain and those dynamically modifying the * behavior chain. Because behavior synchronization adds extra overhead * to virtual operation invocation, we want to restrict, as much as * we can, the requirement for this extra code, to those situations diff --git a/fs/xfs/xfs_bmap.c b/fs/xfs/xfs_bmap.c index 2d702e4a74a..d384e489705 100644 --- a/fs/xfs/xfs_bmap.c +++ b/fs/xfs/xfs_bmap.c @@ -3467,113 +3467,6 @@ done: return error; } -xfs_bmbt_rec_t * /* pointer to found extent entry */ -xfs_bmap_do_search_extents( - xfs_bmbt_rec_t *base, /* base of extent list */ - xfs_extnum_t lastx, /* last extent index used */ - xfs_extnum_t nextents, /* number of file extents */ - xfs_fileoff_t bno, /* block number searched for */ - int *eofp, /* out: end of file found */ - xfs_extnum_t *lastxp, /* out: last extent index */ - xfs_bmbt_irec_t *gotp, /* out: extent entry found */ - xfs_bmbt_irec_t *prevp) /* out: previous extent entry found */ -{ - xfs_bmbt_rec_t *ep; /* extent list entry pointer */ - xfs_bmbt_irec_t got; /* extent list entry, decoded */ - int high; /* high index of binary search */ - int low; /* low index of binary search */ - - /* - * Initialize the extent entry structure to catch access to - * uninitialized br_startblock field. - */ - got.br_startoff = 0xffa5a5a5a5a5a5a5LL; - got.br_blockcount = 0xa55a5a5a5a5a5a5aLL; - got.br_state = XFS_EXT_INVALID; - -#if XFS_BIG_BLKNOS - got.br_startblock = 0xffffa5a5a5a5a5a5LL; -#else - got.br_startblock = 0xffffa5a5; -#endif - - if (lastx != NULLEXTNUM && lastx < nextents) - ep = base + lastx; - else - ep = NULL; - prevp->br_startoff = NULLFILEOFF; - if (ep && bno >= (got.br_startoff = xfs_bmbt_get_startoff(ep)) && - bno < got.br_startoff + - (got.br_blockcount = xfs_bmbt_get_blockcount(ep))) - *eofp = 0; - else if (ep && lastx < nextents - 1 && - bno >= (got.br_startoff = xfs_bmbt_get_startoff(ep + 1)) && - bno < got.br_startoff + - (got.br_blockcount = xfs_bmbt_get_blockcount(ep + 1))) { - lastx++; - ep++; - *eofp = 0; - } else if (nextents == 0) - *eofp = 1; - else if (bno == 0 && - (got.br_startoff = xfs_bmbt_get_startoff(base)) == 0) { - ep = base; - lastx = 0; - got.br_blockcount = xfs_bmbt_get_blockcount(ep); - *eofp = 0; - } else { - low = 0; - high = nextents - 1; - /* binary search the extents array */ - while (low <= high) { - XFS_STATS_INC(xs_cmp_exlist); - lastx = (low + high) >> 1; - ep = base + lastx; - got.br_startoff = xfs_bmbt_get_startoff(ep); - got.br_blockcount = xfs_bmbt_get_blockcount(ep); - if (bno < got.br_startoff) - high = lastx - 1; - else if (bno >= got.br_startoff + got.br_blockcount) - low = lastx + 1; - else { - got.br_startblock = xfs_bmbt_get_startblock(ep); - got.br_state = xfs_bmbt_get_state(ep); - *eofp = 0; - *lastxp = lastx; - *gotp = got; - return ep; - } - } - if (bno >= got.br_startoff + got.br_blockcount) { - lastx++; - if (lastx == nextents) { - *eofp = 1; - got.br_startblock = xfs_bmbt_get_startblock(ep); - got.br_state = xfs_bmbt_get_state(ep); - *prevp = got; - ep = NULL; - } else { - *eofp = 0; - xfs_bmbt_get_all(ep, prevp); - ep++; - got.br_startoff = xfs_bmbt_get_startoff(ep); - got.br_blockcount = xfs_bmbt_get_blockcount(ep); - } - } else { - *eofp = 0; - if (ep > base) - xfs_bmbt_get_all(ep - 1, prevp); - } - } - if (ep) { - got.br_startblock = xfs_bmbt_get_startblock(ep); - got.br_state = xfs_bmbt_get_state(ep); - } - *lastxp = lastx; - *gotp = got; - return ep; -} - /* * Search the extent records for the entry containing block bno. * If bno lies in a hole, point to the next entry. If bno lies diff --git a/fs/xfs/xfs_bmap.h b/fs/xfs/xfs_bmap.h index 011ccaa9a1c..f83399c89ce 100644 --- a/fs/xfs/xfs_bmap.h +++ b/fs/xfs/xfs_bmap.h @@ -362,14 +362,6 @@ xfs_bmbt_rec_t * xfs_bmap_search_multi_extents(struct xfs_ifork *, xfs_fileoff_t, int *, xfs_extnum_t *, xfs_bmbt_irec_t *, xfs_bmbt_irec_t *); -/* - * Search an extent list for the extent which includes block - * bno. - */ -xfs_bmbt_rec_t *xfs_bmap_do_search_extents(xfs_bmbt_rec_t *, - xfs_extnum_t, xfs_extnum_t, xfs_fileoff_t, int *, - xfs_extnum_t *, xfs_bmbt_irec_t *, xfs_bmbt_irec_t *); - #endif /* __KERNEL__ */ #endif /* __XFS_BMAP_H__ */ diff --git a/fs/xfs/xfs_buf_item.c b/fs/xfs/xfs_buf_item.c index 07e2324152b..5fed15682dd 100644 --- a/fs/xfs/xfs_buf_item.c +++ b/fs/xfs/xfs_buf_item.c @@ -98,12 +98,12 @@ xfs_buf_item_flush_log_debug( } /* - * This function is called to verify that our caller's have logged + * This function is called to verify that our callers have logged * all the bytes that they changed. * * It does this by comparing the original copy of the buffer stored in * the buf log item's bli_orig array to the current copy of the buffer - * and ensuring that all bytes which miscompare are set in the bli_logged + * and ensuring that all bytes which mismatch are set in the bli_logged * array of the buf log item. */ STATIC void diff --git a/fs/xfs/xfs_cap.h b/fs/xfs/xfs_cap.h index 433ec537f9b..d0035c6e951 100644 --- a/fs/xfs/xfs_cap.h +++ b/fs/xfs/xfs_cap.h @@ -38,7 +38,7 @@ typedef struct xfs_cap_set { /* * For Linux, we take the bitfields directly from capability.h * and no longer attempt to keep this attribute ondisk compatible - * with IRIX. Since this attribute is only set on exectuables, + * with IRIX. Since this attribute is only set on executables, * it just doesn't make much sense to try. We do use a different * named attribute though, to avoid confusion. */ diff --git a/fs/xfs/xfs_da_btree.c b/fs/xfs/xfs_da_btree.c index 4bae3a76c67..8988b905117 100644 --- a/fs/xfs/xfs_da_btree.c +++ b/fs/xfs/xfs_da_btree.c @@ -840,7 +840,7 @@ xfs_da_node_toosmall(xfs_da_state_t *state, int *action) /* * Check for the degenerate case of the block being empty. * If the block is empty, we'll simply delete it, no need to - * coalesce it with a sibling block. We choose (aribtrarily) + * coalesce it with a sibling block. We choose (arbitrarily) * to merge with the forward block unless it is NULL. */ if (count == 0) { diff --git a/fs/xfs/xfs_dir2_block.c b/fs/xfs/xfs_dir2_block.c index bd5cee6aa51..972ded59547 100644 --- a/fs/xfs/xfs_dir2_block.c +++ b/fs/xfs/xfs_dir2_block.c @@ -533,7 +533,7 @@ xfs_dir2_block_getdents( /* * Reached the end of the block. - * Set the offset to a nonexistent block 1 and return. + * Set the offset to a non-existent block 1 and return. */ *eofp = 1; diff --git a/fs/xfs/xfs_dir2_leaf.c b/fs/xfs/xfs_dir2_leaf.c index 08648b18265..0f5e2f2ce6e 100644 --- a/fs/xfs/xfs_dir2_leaf.c +++ b/fs/xfs/xfs_dir2_leaf.c @@ -515,7 +515,7 @@ xfs_dir2_leaf_addname( ASSERT(be32_to_cpu(leaf->ents[highstale].address) == XFS_DIR2_NULL_DATAPTR); /* - * Copy entries down to copver the stale entry + * Copy entries down to cover the stale entry * and make room for the new entry. */ if (highstale - index > 0) diff --git a/fs/xfs/xfs_dir2_node.c b/fs/xfs/xfs_dir2_node.c index af556f16a0c..ac511ab9c52 100644 --- a/fs/xfs/xfs_dir2_node.c +++ b/fs/xfs/xfs_dir2_node.c @@ -830,7 +830,7 @@ xfs_dir2_leafn_rebalance( state->inleaf = 1; blk2->index = 0; cmn_err(CE_ALERT, - "xfs_dir2_leafn_rebalance: picked the wrong leaf? reverting orignal leaf: " + "xfs_dir2_leafn_rebalance: picked the wrong leaf? reverting original leaf: " "blk1->index %d\n", blk1->index); } diff --git a/fs/xfs/xfs_dir_leaf.c b/fs/xfs/xfs_dir_leaf.c index ee88751c3be..6d711869262 100644 --- a/fs/xfs/xfs_dir_leaf.c +++ b/fs/xfs/xfs_dir_leaf.c @@ -1341,7 +1341,7 @@ xfs_dir_leaf_toosmall(xfs_da_state_t *state, int *action) /* * Check for the degenerate case of the block being empty. * If the block is empty, we'll simply delete it, no need to - * coalesce it with a sibling block. We choose (aribtrarily) + * coalesce it with a sibling block. We choose (arbitrarily) * to merge with the forward block unless it is NULL. */ if (count == 0) { diff --git a/fs/xfs/xfs_fsops.c b/fs/xfs/xfs_fsops.c index 56caa88713a..dfa3527b20a 100644 --- a/fs/xfs/xfs_fsops.c +++ b/fs/xfs/xfs_fsops.c @@ -477,7 +477,7 @@ xfs_fs_counts( * * xfs_reserve_blocks is called to set m_resblks * in the in-core mount table. The number of unused reserved blocks - * is kept in m_resbls_avail. + * is kept in m_resblks_avail. * * Reserve the requested number of blocks if available. Otherwise return * as many as possible to satisfy the request. The actual number diff --git a/fs/xfs/xfs_ialloc.c b/fs/xfs/xfs_ialloc.c index 0024892841a..4eeb856183b 100644 --- a/fs/xfs/xfs_ialloc.c +++ b/fs/xfs/xfs_ialloc.c @@ -136,7 +136,7 @@ xfs_ialloc_ag_alloc( int ninodes; /* num inodes per buf */ xfs_agino_t thisino; /* current inode number, for loop */ int version; /* inode version number to use */ - int isaligned; /* inode allocation at stripe unit */ + int isaligned = 0; /* inode allocation at stripe unit */ /* boundary */ args.tp = tp; @@ -152,47 +152,75 @@ xfs_ialloc_ag_alloc( return XFS_ERROR(ENOSPC); args.minlen = args.maxlen = XFS_IALLOC_BLOCKS(args.mp); /* - * Set the alignment for the allocation. - * If stripe alignment is turned on then align at stripe unit - * boundary. - * If the cluster size is smaller than a filesystem block - * then we're doing I/O for inodes in filesystem block size pieces, - * so don't need alignment anyway. - */ - isaligned = 0; - if (args.mp->m_sinoalign) { - ASSERT(!(args.mp->m_flags & XFS_MOUNT_NOALIGN)); - args.alignment = args.mp->m_dalign; - isaligned = 1; - } else if (XFS_SB_VERSION_HASALIGN(&args.mp->m_sb) && - args.mp->m_sb.sb_inoalignmt >= - XFS_B_TO_FSBT(args.mp, XFS_INODE_CLUSTER_SIZE(args.mp))) - args.alignment = args.mp->m_sb.sb_inoalignmt; - else - args.alignment = 1; + * First try to allocate inodes contiguous with the last-allocated + * chunk of inodes. If the filesystem is striped, this will fill + * an entire stripe unit with inodes. + */ agi = XFS_BUF_TO_AGI(agbp); - /* - * Need to figure out where to allocate the inode blocks. - * Ideally they should be spaced out through the a.g. - * For now, just allocate blocks up front. - */ - args.agbno = be32_to_cpu(agi->agi_root); - args.fsbno = XFS_AGB_TO_FSB(args.mp, be32_to_cpu(agi->agi_seqno), - args.agbno); - /* - * Allocate a fixed-size extent of inodes. - */ - args.type = XFS_ALLOCTYPE_NEAR_BNO; - args.mod = args.total = args.wasdel = args.isfl = args.userdata = - args.minalignslop = 0; - args.prod = 1; - /* - * Allow space for the inode btree to split. - */ - args.minleft = XFS_IN_MAXLEVELS(args.mp) - 1; - if ((error = xfs_alloc_vextent(&args))) - return error; + newino = be32_to_cpu(agi->agi_newino); + if(likely(newino != NULLAGINO)) { + args.agbno = XFS_AGINO_TO_AGBNO(args.mp, newino) + + XFS_IALLOC_BLOCKS(args.mp); + args.fsbno = XFS_AGB_TO_FSB(args.mp, + be32_to_cpu(agi->agi_seqno), args.agbno); + args.type = XFS_ALLOCTYPE_THIS_BNO; + args.mod = args.total = args.wasdel = args.isfl = + args.userdata = args.minalignslop = 0; + args.prod = 1; + args.alignment = 1; + /* + * Allow space for the inode btree to split. + */ + args.minleft = XFS_IN_MAXLEVELS(args.mp) - 1; + if ((error = xfs_alloc_vextent(&args))) + return error; + } else + args.fsbno = NULLFSBLOCK; + if (unlikely(args.fsbno == NULLFSBLOCK)) { + /* + * Set the alignment for the allocation. + * If stripe alignment is turned on then align at stripe unit + * boundary. + * If the cluster size is smaller than a filesystem block + * then we're doing I/O for inodes in filesystem block size + * pieces, so don't need alignment anyway. + */ + isaligned = 0; + if (args.mp->m_sinoalign) { + ASSERT(!(args.mp->m_flags & XFS_MOUNT_NOALIGN)); + args.alignment = args.mp->m_dalign; + isaligned = 1; + } else if (XFS_SB_VERSION_HASALIGN(&args.mp->m_sb) && + args.mp->m_sb.sb_inoalignmt >= + XFS_B_TO_FSBT(args.mp, + XFS_INODE_CLUSTER_SIZE(args.mp))) + args.alignment = args.mp->m_sb.sb_inoalignmt; + else + args.alignment = 1; + /* + * Need to figure out where to allocate the inode blocks. + * Ideally they should be spaced out through the a.g. + * For now, just allocate blocks up front. + */ + args.agbno = be32_to_cpu(agi->agi_root); + args.fsbno = XFS_AGB_TO_FSB(args.mp, + be32_to_cpu(agi->agi_seqno), args.agbno); + /* + * Allocate a fixed-size extent of inodes. + */ + args.type = XFS_ALLOCTYPE_NEAR_BNO; + args.mod = args.total = args.wasdel = args.isfl = + args.userdata = args.minalignslop = 0; + args.prod = 1; + /* + * Allow space for the inode btree to split. + */ + args.minleft = XFS_IN_MAXLEVELS(args.mp) - 1; + if ((error = xfs_alloc_vextent(&args))) + return error; + } + /* * If stripe alignment is turned on, then try again with cluster * alignment. @@ -1023,7 +1051,7 @@ xfs_difree( rec.ir_freecount++; /* - * When an inode cluster is free, it becomes elgible for removal + * When an inode cluster is free, it becomes eligible for removal */ if ((mp->m_flags & XFS_MOUNT_IDELETE) && (rec.ir_freecount == XFS_IALLOC_INODES(mp))) { diff --git a/fs/xfs/xfs_iget.c b/fs/xfs/xfs_iget.c index 3ce35a6f700..bb33113eef9 100644 --- a/fs/xfs/xfs_iget.c +++ b/fs/xfs/xfs_iget.c @@ -509,7 +509,7 @@ retry: } else { /* * If the inode is not fully constructed due to - * filehandle mistmatches wait for the inode to go + * filehandle mismatches wait for the inode to go * away and try again. * * iget_locked will call __wait_on_freeing_inode diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c index 88a517fad07..48146bdc6bd 100644 --- a/fs/xfs/xfs_inode.c +++ b/fs/xfs/xfs_inode.c @@ -160,7 +160,7 @@ xfs_inotobp( xfs_dinode_t *dip; /* - * Call the space managment code to find the location of the + * Call the space management code to find the location of the * inode on disk. */ imap.im_blkno = 0; @@ -837,7 +837,7 @@ xfs_dic2xflags( /* * Given a mount structure and an inode number, return a pointer - * to a newly allocated in-core inode coresponding to the given + * to a newly allocated in-core inode corresponding to the given * inode number. * * Initialize the inode's attributes and extent pointers if it @@ -2723,7 +2723,7 @@ xfs_ipin( /* * Decrement the pin count of the given inode, and wake up * anyone in xfs_iwait_unpin() if the count goes to 0. The - * inode must have been previoulsy pinned with a call to xfs_ipin(). + * inode must have been previously pinned with a call to xfs_ipin(). */ void xfs_iunpin( @@ -3690,7 +3690,7 @@ void xfs_iext_add( xfs_ifork_t *ifp, /* inode fork pointer */ xfs_extnum_t idx, /* index to begin adding exts */ - int ext_diff) /* nubmer of extents to add */ + int ext_diff) /* number of extents to add */ { int byte_diff; /* new bytes being added */ int new_size; /* size of extents after adding */ @@ -4038,7 +4038,7 @@ xfs_iext_remove_indirect( xfs_extnum_t ext_diff; /* extents to remove in current list */ xfs_extnum_t nex1; /* number of extents before idx */ xfs_extnum_t nex2; /* extents after idx + count */ - int nlists; /* entries in indirecton array */ + int nlists; /* entries in indirection array */ int page_idx = idx; /* index in target extent list */ ASSERT(ifp->if_flags & XFS_IFEXTIREC); @@ -4291,9 +4291,9 @@ xfs_iext_bno_to_ext( xfs_filblks_t blockcount = 0; /* number of blocks in extent */ xfs_bmbt_rec_t *ep = NULL; /* pointer to target extent */ xfs_ext_irec_t *erp = NULL; /* indirection array pointer */ - int high; /* upper boundry in search */ + int high; /* upper boundary in search */ xfs_extnum_t idx = 0; /* index of target extent */ - int low; /* lower boundry in search */ + int low; /* lower boundary in search */ xfs_extnum_t nextents; /* number of file extents */ xfs_fileoff_t startoff = 0; /* start offset of extent */ diff --git a/fs/xfs/xfs_inode_item.c b/fs/xfs/xfs_inode_item.c index 36aa1fcb90a..7497a481b2f 100644 --- a/fs/xfs/xfs_inode_item.c +++ b/fs/xfs/xfs_inode_item.c @@ -580,7 +580,7 @@ xfs_inode_item_unpin_remove( * been or is in the process of being flushed, then (ideally) we'd like to * see if the inode's buffer is still incore, and if so give it a nudge. * We delay doing so until the pushbuf routine, though, to avoid holding - * the AIL lock across a call to the blackhole which is the buffercache. + * the AIL lock across a call to the blackhole which is the buffer cache. * Also we don't want to sleep in any device strategy routines, which can happen * if we do the subsequent bawrite in here. */ diff --git a/fs/xfs/xfs_itable.c b/fs/xfs/xfs_itable.c index 32247b6bfee..94068d014f2 100644 --- a/fs/xfs/xfs_itable.c +++ b/fs/xfs/xfs_itable.c @@ -272,7 +272,7 @@ xfs_bulkstat( size_t statstruct_size, /* sizeof struct filling */ char __user *ubuffer, /* buffer with inode stats */ int flags, /* defined in xfs_itable.h */ - int *done) /* 1 if there're more stats to get */ + int *done) /* 1 if there are more stats to get */ { xfs_agblock_t agbno=0;/* allocation group block number */ xfs_buf_t *agbp; /* agi header buffer */ @@ -676,7 +676,7 @@ xfs_bulkstat_single( xfs_mount_t *mp, /* mount point for filesystem */ xfs_ino_t *lastinop, /* inode to return */ char __user *buffer, /* buffer with inode stats */ - int *done) /* 1 if there're more stats to get */ + int *done) /* 1 if there are more stats to get */ { int count; /* count value for bulkstat call */ int error; /* return value */ diff --git a/fs/xfs/xfs_itable.h b/fs/xfs/xfs_itable.h index 047d834ed21..11eb4e1b18c 100644 --- a/fs/xfs/xfs_itable.h +++ b/fs/xfs/xfs_itable.h @@ -60,7 +60,7 @@ xfs_bulkstat( size_t statstruct_size,/* sizeof struct that we're filling */ char __user *ubuffer,/* buffer with inode stats */ int flags, /* flag to control access method */ - int *done); /* 1 if there're more stats to get */ + int *done); /* 1 if there are more stats to get */ int xfs_bulkstat_single( diff --git a/fs/xfs/xfs_log.c b/fs/xfs/xfs_log.c index 9176995160e..32e841d2f26 100644 --- a/fs/xfs/xfs_log.c +++ b/fs/xfs/xfs_log.c @@ -59,7 +59,7 @@ STATIC xlog_t * xlog_alloc_log(xfs_mount_t *mp, int num_bblks); STATIC int xlog_space_left(xlog_t *log, int cycle, int bytes); STATIC int xlog_sync(xlog_t *log, xlog_in_core_t *iclog); -STATIC void xlog_unalloc_log(xlog_t *log); +STATIC void xlog_dealloc_log(xlog_t *log); STATIC int xlog_write(xfs_mount_t *mp, xfs_log_iovec_t region[], int nentries, xfs_log_ticket_t tic, xfs_lsn_t *start_lsn, @@ -304,7 +304,7 @@ xfs_log_done(xfs_mount_t *mp, if ((ticket->t_flags & XLOG_TIC_PERM_RESERV) == 0 || (flags & XFS_LOG_REL_PERM_RESERV)) { /* - * Release ticket if not permanent reservation or a specifc + * Release ticket if not permanent reservation or a specific * request has been made to release a permanent reservation. */ xlog_trace_loggrant(log, ticket, "xfs_log_done: (non-permanent)"); @@ -511,7 +511,7 @@ xfs_log_mount(xfs_mount_t *mp, vfsp->vfs_flag |= VFS_RDONLY; if (error) { cmn_err(CE_WARN, "XFS: log mount/recovery failed: error %d", error); - xlog_unalloc_log(mp->m_log); + xlog_dealloc_log(mp->m_log); return error; } } @@ -667,7 +667,7 @@ xfs_log_unmount_write(xfs_mount_t *mp) * * Go through the motions of sync'ing and releasing * the iclog, even though no I/O will actually happen, - * we need to wait for other log I/O's that may already + * we need to wait for other log I/Os that may already * be in progress. Do this as a separate section of * code so we'll know if we ever get stuck here that * we're in this odd situation of trying to unmount @@ -704,7 +704,7 @@ xfs_log_unmount_write(xfs_mount_t *mp) void xfs_log_unmount_dealloc(xfs_mount_t *mp) { - xlog_unalloc_log(mp->m_log); + xlog_dealloc_log(mp->m_log); } /* @@ -1492,7 +1492,7 @@ xlog_sync(xlog_t *log, ASSERT(XFS_BUF_ADDR(bp) <= log->l_logBBsize-1); ASSERT(XFS_BUF_ADDR(bp) + BTOBB(count) <= log->l_logBBsize); - /* account for internal log which does't start at block #0 */ + /* account for internal log which doesn't start at block #0 */ XFS_BUF_SET_ADDR(bp, XFS_BUF_ADDR(bp) + log->l_logBBstart); XFS_BUF_WRITE(bp); if ((error = XFS_bwrite(bp))) { @@ -1506,10 +1506,10 @@ xlog_sync(xlog_t *log, /* - * Unallocate a log structure + * Deallocate a log structure */ void -xlog_unalloc_log(xlog_t *log) +xlog_dealloc_log(xlog_t *log) { xlog_in_core_t *iclog, *next_iclog; xlog_ticket_t *tic, *next_tic; @@ -1539,7 +1539,7 @@ xlog_unalloc_log(xlog_t *log) if ((log->l_ticket_cnt != log->l_ticket_tcnt) && !XLOG_FORCED_SHUTDOWN(log)) { xfs_fs_cmn_err(CE_WARN, log->l_mp, - "xlog_unalloc_log: (cnt: %d, total: %d)", + "xlog_dealloc_log: (cnt: %d, total: %d)", log->l_ticket_cnt, log->l_ticket_tcnt); /* ASSERT(log->l_ticket_cnt == log->l_ticket_tcnt); */ @@ -1562,7 +1562,7 @@ xlog_unalloc_log(xlog_t *log) #endif log->l_mp->m_log = NULL; kmem_free(log, sizeof(xlog_t)); -} /* xlog_unalloc_log */ +} /* xlog_dealloc_log */ /* * Update counters atomically now that memcpy is done. @@ -2829,7 +2829,7 @@ xlog_state_release_iclog(xlog_t *log, /* * We let the log lock go, so it's possible that we hit a log I/O - * error or someother SHUTDOWN condition that marks the iclog + * error or some other SHUTDOWN condition that marks the iclog * as XLOG_STATE_IOERROR before the bwrite. However, we know that * this iclog has consistent data, so we ignore IOERROR * flags after this point. diff --git a/fs/xfs/xfs_log.h b/fs/xfs/xfs_log.h index 4b2ac88dbb8..eacb3d4987f 100644 --- a/fs/xfs/xfs_log.h +++ b/fs/xfs/xfs_log.h @@ -27,7 +27,7 @@ #ifdef __KERNEL__ /* - * By comparing each compnent, we don't have to worry about extra + * By comparing each component, we don't have to worry about extra * endian issues in treating two 32 bit numbers as one 64 bit number */ static inline xfs_lsn_t _lsn_cmp(xfs_lsn_t lsn1, xfs_lsn_t lsn2) diff --git a/fs/xfs/xfs_log_recover.c b/fs/xfs/xfs_log_recover.c index add13f507ed..1f0016b0b4e 100644 --- a/fs/xfs/xfs_log_recover.c +++ b/fs/xfs/xfs_log_recover.c @@ -583,7 +583,7 @@ xlog_find_head( * x | x ... | x - 1 | x * Another case that fits this picture would be * x | x + 1 | x ... | x - * In this case the head really is somwhere at the end of the + * In this case the head really is somewhere at the end of the * log, as one of the latest writes at the beginning was * incomplete. * One more case is @@ -2799,7 +2799,7 @@ xlog_recover_do_trans( * we don't need to worry about the block number being * truncated in > 1 TB buffers because in user-land, * we're now n32 or 64-bit so xfs_daddr_t is 64-bits so - * the blkno's will get through the user-mode buffer + * the blknos will get through the user-mode buffer * cache properly. The only bad case is o32 kernels * where xfs_daddr_t is 32-bits but mount will warn us * off a > 1 TB filesystem before we get here. diff --git a/fs/xfs/xfs_mount.c b/fs/xfs/xfs_mount.c index 20e8abc16d1..72e7e78bfff 100644 --- a/fs/xfs/xfs_mount.c +++ b/fs/xfs/xfs_mount.c @@ -393,7 +393,7 @@ xfs_initialize_perag( break; } - /* This ag is prefered for inodes */ + /* This ag is preferred for inodes */ pag = &mp->m_perag[index]; pag->pagi_inodeok = 1; if (index < max_metadata) @@ -1728,7 +1728,7 @@ xfs_mount_log_sbunit( * We cannot use the hotcpu_register() function because it does * not allow notifier instances. We need a notifier per filesystem * as we need to be able to identify the filesystem to balance - * the counters out. This is acheived by having a notifier block + * the counters out. This is achieved by having a notifier block * embedded in the xfs_mount_t and doing pointer magic to get the * mount pointer from the notifier block address. */ diff --git a/fs/xfs/xfs_mount.h b/fs/xfs/xfs_mount.h index ebd73960e9d..66cbee79864 100644 --- a/fs/xfs/xfs_mount.h +++ b/fs/xfs/xfs_mount.h @@ -379,7 +379,7 @@ typedef struct xfs_mount { #endif int m_dalign; /* stripe unit */ int m_swidth; /* stripe width */ - int m_sinoalign; /* stripe unit inode alignmnt */ + int m_sinoalign; /* stripe unit inode alignment */ int m_attr_magicpct;/* 37% of the blocksize */ int m_dir_magicpct; /* 37% of the dir blocksize */ __uint8_t m_mk_sharedro; /* mark shared ro on unmount */ diff --git a/fs/xfs/xfs_quota.h b/fs/xfs/xfs_quota.h index 82a08baf437..4f6a034de7f 100644 --- a/fs/xfs/xfs_quota.h +++ b/fs/xfs/xfs_quota.h @@ -31,7 +31,7 @@ typedef __uint32_t xfs_dqid_t; /* - * Eventhough users may not have quota limits occupying all 64-bits, + * Even though users may not have quota limits occupying all 64-bits, * they may need 64-bit accounting. Hence, 64-bit quota-counters, * and quota-limits. This is a waste in the common case, but hey ... */ @@ -246,7 +246,7 @@ typedef struct xfs_qoff_logformat { #ifdef __KERNEL__ /* * This check is done typically without holding the inode lock; - * that may seem racey, but it is harmless in the context that it is used. + * that may seem racy, but it is harmless in the context that it is used. * The inode cannot go inactive as long a reference is kept, and * therefore if dquot(s) were attached, they'll stay consistent. * If, for example, the ownership of the inode changes while diff --git a/fs/xfs/xfs_trans.c b/fs/xfs/xfs_trans.c index 2918956553a..8d056cef5d1 100644 --- a/fs/xfs/xfs_trans.c +++ b/fs/xfs/xfs_trans.c @@ -490,7 +490,7 @@ xfs_trans_mod_sb( case XFS_TRANS_SB_RES_FREXTENTS: /* * The allocation has already been applied to the - * in-core superblocks's counter. This should only + * in-core superblock's counter. This should only * be applied to the on-disk superblock. */ ASSERT(delta < 0); @@ -611,7 +611,7 @@ xfs_trans_apply_sb_deltas( if (whole) /* - * Log the whole thing, the fields are discontiguous. + * Log the whole thing, the fields are noncontiguous. */ xfs_trans_log_buf(tp, bp, 0, sizeof(xfs_sb_t) - 1); else @@ -669,7 +669,7 @@ xfs_trans_unreserve_and_mod_sb( /* * Apply any superblock modifications to the in-core version. * The t_res_fdblocks_delta and t_res_frextents_delta fields are - * explicity NOT applied to the in-core superblock. + * explicitly NOT applied to the in-core superblock. * The idea is that that has already been done. */ if (tp->t_flags & XFS_TRANS_SB_DIRTY) { diff --git a/fs/xfs/xfs_trans.h b/fs/xfs/xfs_trans.h index e48befa4e33..100d9a4b38e 100644 --- a/fs/xfs/xfs_trans.h +++ b/fs/xfs/xfs_trans.h @@ -354,7 +354,7 @@ typedef struct xfs_trans { xfs_lsn_t t_commit_lsn; /* log seq num of end of * transaction. */ struct xfs_mount *t_mountp; /* ptr to fs mount struct */ - struct xfs_dquot_acct *t_dqinfo; /* accting info for dquots */ + struct xfs_dquot_acct *t_dqinfo; /* acctg info for dquots */ xfs_trans_callback_t t_callback; /* transaction callback */ void *t_callarg; /* callback arg */ unsigned int t_flags; /* misc flags */ diff --git a/fs/xfs/xfs_trans_inode.c b/fs/xfs/xfs_trans_inode.c index e341409172d..7c5894d59f8 100644 --- a/fs/xfs/xfs_trans_inode.c +++ b/fs/xfs/xfs_trans_inode.c @@ -272,7 +272,7 @@ xfs_trans_log_inode( * This is to coordinate with the xfs_iflush() and xfs_iflush_done() * routines in the eventual clearing of the ilf_fields bits. * See the big comment in xfs_iflush() for an explanation of - * this coorination mechanism. + * this coordination mechanism. */ flags |= ip->i_itemp->ili_last_fields; ip->i_itemp->ili_format.ilf_fields |= flags; diff --git a/fs/xfs/xfs_vfsops.c b/fs/xfs/xfs_vfsops.c index d4ec4dfaf19..504d2a80747 100644 --- a/fs/xfs/xfs_vfsops.c +++ b/fs/xfs/xfs_vfsops.c @@ -880,10 +880,10 @@ xfs_statvfs( * determine if they should be flushed sync, async, or * delwri. * SYNC_CLOSE - This flag is passed when the system is being - * unmounted. We should sync and invalidate everthing. + * unmounted. We should sync and invalidate everything. * SYNC_FSDATA - This indicates that the caller would like to make * sure the superblock is safe on disk. We can ensure - * this by simply makeing sure the log gets flushed + * this by simply making sure the log gets flushed * if SYNC_BDFLUSH is set, and by actually writing it * out otherwise. * @@ -908,7 +908,7 @@ xfs_sync( * * This routine supports all of the flags defined for the generic VFS_SYNC * interface as explained above under xfs_sync. In the interests of not - * changing interfaces within the 6.5 family, additional internallly- + * changing interfaces within the 6.5 family, additional internally- * required functions are specified within a separate xflags parameter, * only available by calling this routine. * @@ -1090,7 +1090,7 @@ xfs_sync_inodes( * If this is just vfs_sync() or pflushd() calling * then we can skip inodes for which it looks like * there is nothing to do. Since we don't have the - * inode locked this is racey, but these are periodic + * inode locked this is racy, but these are periodic * calls so it doesn't matter. For the others we want * to know for sure, so we at least try to lock them. */ @@ -1429,7 +1429,7 @@ xfs_sync_inodes( * * This routine supports all of the flags defined for the generic VFS_SYNC * interface as explained above under xfs_sync. In the interests of not - * changing interfaces within the 6.5 family, additional internallly- + * changing interfaces within the 6.5 family, additional internally- * required functions are specified within a separate xflags parameter, * only available by calling this routine. * diff --git a/fs/xfs/xfs_vnodeops.c b/fs/xfs/xfs_vnodeops.c index 0f0a64e81db..de49601919c 100644 --- a/fs/xfs/xfs_vnodeops.c +++ b/fs/xfs/xfs_vnodeops.c @@ -848,7 +848,7 @@ xfs_setattr( * If this is a synchronous mount, make sure that the * transaction goes to disk before returning to the user. * This is slightly sub-optimal in that truncates require - * two sync transactions instead of one for wsync filesytems. + * two sync transactions instead of one for wsync filesystems. * One for the truncate and one for the timestamps since we * don't want to change the timestamps unless we're sure the * truncate worked. Truncates are less than 1% of the laddis @@ -1170,7 +1170,7 @@ xfs_fsync( /* * If this inode is on the RT dev we need to flush that - * cache aswell. + * cache as well. */ if (ip->i_d.di_flags & XFS_DIFLAG_REALTIME) xfs_blkdev_issue_flush(ip->i_mount->m_rtdev_targp); @@ -1380,7 +1380,7 @@ xfs_inactive_symlink_rmt( */ ntp = xfs_trans_dup(tp); /* - * Commit the transaction containing extent freeing and EFD's. + * Commit the transaction containing extent freeing and EFDs. * If we get an error on the commit here or on the reserve below, * we need to unlock the inode since the new transaction doesn't * have the inode attached. @@ -2023,7 +2023,7 @@ xfs_create( XFS_QM_DQRELE(mp, gdqp); /* - * Propogate the fact that the vnode changed after the + * Propagate the fact that the vnode changed after the * xfs_inode locks have been released. */ VOP_VNODE_CHANGE(vp, VCHANGE_FLAGS_TRUNCATED, 3); @@ -2370,7 +2370,7 @@ xfs_remove( * for a log reservation. Since we'll have to wait for the * inactive code to complete before returning from xfs_iget, * we need to make sure that we don't have log space reserved - * when we call xfs_iget. Instead we get an unlocked referece + * when we call xfs_iget. Instead we get an unlocked reference * to the inode before getting our log reservation. */ error = xfs_get_dir_entry(dentry, &ip); @@ -3020,7 +3020,7 @@ xfs_rmdir( * for a log reservation. Since we'll have to wait for the * inactive code to complete before returning from xfs_iget, * we need to make sure that we don't have log space reserved - * when we call xfs_iget. Instead we get an unlocked referece + * when we call xfs_iget. Instead we get an unlocked reference * to the inode before getting our log reservation. */ error = xfs_get_dir_entry(dentry, &cdp); |