aboutsummaryrefslogtreecommitdiff
path: root/fs/ceph
diff options
context:
space:
mode:
authorSage Weil <sage@newdream.net>2010-02-09 14:04:02 -0800
committerSage Weil <sage@newdream.net>2010-02-11 11:48:53 -0800
commit6a026589ba333185c466c906376fe022a27a53f9 (patch)
tree7c536163e2192928978b79d5a6bbd79dd5993777 /fs/ceph
parent68c283236a1e0772e1a469dd2ffc17afc300b07b (diff)
ceph: fix sync read eof check deadlock
If a sync read gets a short result from the OSD, it may need to do a getattr to see if it is short due to reaching end-of-file. The getattr was being done while holding a reference to FILE_RD, which can lead to a deadlock if the MDS is revoking that capability bit and can't process the getattr until it does. We fix this by setting a flag if EOF size validation is needed, and doing the getattr in ceph_aio_read, after the RD cap ref is dropped. If the read needs to be continued, we loop and continue traversing the file. Signed-off-by: Sage Weil <sage@newdream.net>
Diffstat (limited to 'fs/ceph')
-rw-r--r--fs/ceph/file.c39
1 files changed, 26 insertions, 13 deletions
diff --git a/fs/ceph/file.c b/fs/ceph/file.c
index bbf1ccf2d56..2c4ae4441ca 100644
--- a/fs/ceph/file.c
+++ b/fs/ceph/file.c
@@ -431,7 +431,8 @@ static void zero_page_vector_range(int off, int len, struct page **pages)
*/
static int striped_read(struct inode *inode,
u64 off, u64 len,
- struct page **pages, int num_pages)
+ struct page **pages, int num_pages,
+ int *checkeof)
{
struct ceph_client *client = ceph_inode_to_client(inode);
struct ceph_inode_info *ci = ceph_inode(inode);
@@ -497,15 +498,7 @@ more:
}
/* check i_size */
- ret = ceph_do_getattr(inode, CEPH_STAT_CAP_SIZE);
- if (ret < 0)
- goto out;
-
- /* hit EOF? */
- if (pos >= inode->i_size)
- goto out;
-
- goto more;
+ *checkeof = 1;
}
out:
@@ -522,7 +515,7 @@ out:
* If the read spans object boundary, just do multiple reads.
*/
static ssize_t ceph_sync_read(struct file *file, char __user *data,
- unsigned len, loff_t *poff)
+ unsigned len, loff_t *poff, int *checkeof)
{
struct inode *inode = file->f_dentry->d_inode;
struct page **pages;
@@ -552,7 +545,7 @@ static ssize_t ceph_sync_read(struct file *file, char __user *data,
if (ret < 0)
goto done;
- ret = striped_read(inode, off, len, pages, num_pages);
+ ret = striped_read(inode, off, len, pages, num_pages, checkeof);
if (ret >= 0 && (file->f_flags & O_DIRECT) == 0)
ret = copy_page_vector_to_user(pages, data, off, ret);
@@ -746,11 +739,14 @@ static ssize_t ceph_aio_read(struct kiocb *iocb, const struct iovec *iov,
size_t len = iov->iov_len;
struct inode *inode = filp->f_dentry->d_inode;
struct ceph_inode_info *ci = ceph_inode(inode);
+ void *base = iov->iov_base;
ssize_t ret;
int got = 0;
+ int checkeof = 0, read = 0;
dout("aio_read %p %llx.%llx %llu~%u trying to get caps on %p\n",
inode, ceph_vinop(inode), pos, (unsigned)len, inode);
+again:
__ceph_do_pending_vmtruncate(inode);
ret = ceph_get_caps(ci, CEPH_CAP_FILE_RD, CEPH_CAP_FILE_CACHE,
&got, -1);
@@ -764,7 +760,7 @@ static ssize_t ceph_aio_read(struct kiocb *iocb, const struct iovec *iov,
(iocb->ki_filp->f_flags & O_DIRECT) ||
(inode->i_sb->s_flags & MS_SYNCHRONOUS))
/* hmm, this isn't really async... */
- ret = ceph_sync_read(filp, iov->iov_base, len, ppos);
+ ret = ceph_sync_read(filp, base, len, ppos, &checkeof);
else
ret = generic_file_aio_read(iocb, iov, nr_segs, pos);
@@ -772,6 +768,23 @@ out:
dout("aio_read %p %llx.%llx dropping cap refs on %s = %d\n",
inode, ceph_vinop(inode), ceph_cap_string(got), (int)ret);
ceph_put_cap_refs(ci, got);
+
+ if (checkeof && ret >= 0) {
+ int statret = ceph_do_getattr(inode, CEPH_STAT_CAP_SIZE);
+
+ /* hit EOF or hole? */
+ if (statret == 0 && *ppos < inode->i_size) {
+ dout("aio_read sync_read hit hole, reading more\n");
+ read += ret;
+ base += ret;
+ len -= ret;
+ checkeof = 0;
+ goto again;
+ }
+ }
+ if (ret >= 0)
+ ret += read;
+
return ret;
}