From 4ba069b802c29eee066385f9826e2d83716626b4 Mon Sep 17 00:00:00 2001 From: Michael Grundy Date: Wed, 20 Sep 2006 15:58:39 +0200 Subject: [S390] add kprobes support. Signed-off-by: Michael Grundy Signed-off-by: David Wilder Signed-off-by: Martin Schwidefsky --- arch/s390/Kconfig | 14 + arch/s390/kernel/Makefile | 1 + arch/s390/kernel/entry.S | 12 + arch/s390/kernel/entry64.S | 12 + arch/s390/kernel/kprobes.c | 657 +++++++++++++++++++++++++++++++++++++++++ arch/s390/kernel/traps.c | 31 +- arch/s390/kernel/vmlinux.lds.S | 1 + arch/s390/mm/fault.c | 40 ++- 8 files changed, 764 insertions(+), 4 deletions(-) create mode 100644 arch/s390/kernel/kprobes.c (limited to 'arch') diff --git a/arch/s390/Kconfig b/arch/s390/Kconfig index 2f4f70c4dbb..76122ce1e6c 100644 --- a/arch/s390/Kconfig +++ b/arch/s390/Kconfig @@ -487,8 +487,22 @@ source "drivers/net/Kconfig" source "fs/Kconfig" +menu "Instrumentation Support" + source "arch/s390/oprofile/Kconfig" +config KPROBES + bool "Kprobes (EXPERIMENTAL)" + depends on EXPERIMENTAL && MODULES + help + Kprobes allows you to trap at almost any kernel address and + execute a callback function. register_kprobe() establishes + a probepoint and specifies the callback. Kprobes is useful + for kernel debugging, non-intrusive instrumentation and testing. + If in doubt, say "N". + +endmenu + source "arch/s390/Kconfig.debug" source "security/Kconfig" diff --git a/arch/s390/kernel/Makefile b/arch/s390/kernel/Makefile index 9a33ed6ca69..33a5069c0e1 100644 --- a/arch/s390/kernel/Makefile +++ b/arch/s390/kernel/Makefile @@ -24,6 +24,7 @@ obj-$(CONFIG_COMPAT) += compat_linux.o compat_signal.o \ obj-$(CONFIG_VIRT_TIMER) += vtime.o obj-$(CONFIG_STACKTRACE) += stacktrace.o +obj-$(CONFIG_KPROBES) += kprobes.o # Kexec part S390_KEXEC_OBJS := machine_kexec.o crash.o diff --git a/arch/s390/kernel/entry.S b/arch/s390/kernel/entry.S index 5b5799ac8f8..0c712b78a7e 100644 --- a/arch/s390/kernel/entry.S +++ b/arch/s390/kernel/entry.S @@ -505,6 +505,8 @@ pgm_no_vtime2: mvc __THREAD_per+__PER_address(4,%r1),__LC_PER_ADDRESS mvc __THREAD_per+__PER_access_id(1,%r1),__LC_PER_ACCESS_ID oi __TI_flags+3(%r9),_TIF_SINGLE_STEP # set TIF_SINGLE_STEP + tm SP_PSW+1(%r15),0x01 # kernel per event ? + bz BASED(kernel_per) l %r3,__LC_PGM_ILC # load program interruption code la %r8,0x7f nr %r8,%r3 # clear per-event-bit and ilc @@ -536,6 +538,16 @@ pgm_no_vtime3: stosm __SF_EMPTY(%r15),0x03 # reenable interrupts b BASED(sysc_do_svc) +# +# per was called from kernel, must be kprobes +# +kernel_per: + mvi SP_TRAP+1(%r15),0x28 # set trap indication to pgm check + la %r2,SP_PTREGS(%r15) # address of register-save area + l %r1,BASED(.Lhandle_per) # load adr. of per handler + la %r14,BASED(sysc_leave) # load adr. of system return + br %r1 # branch to do_single_step + /* * IO interrupt handler routine */ diff --git a/arch/s390/kernel/entry64.S b/arch/s390/kernel/entry64.S index 56f5f613b86..8b956d1538f 100644 --- a/arch/s390/kernel/entry64.S +++ b/arch/s390/kernel/entry64.S @@ -518,6 +518,8 @@ pgm_no_vtime2: #endif lg %r9,__LC_THREAD_INFO # load pointer to thread_info struct lg %r1,__TI_task(%r9) + tm SP_PSW+1(%r15),0x01 # kernel per event ? + jz kernel_per mvc __THREAD_per+__PER_atmid(2,%r1),__LC_PER_ATMID mvc __THREAD_per+__PER_address(8,%r1),__LC_PER_ADDRESS mvc __THREAD_per+__PER_access_id(1,%r1),__LC_PER_ACCESS_ID @@ -553,6 +555,16 @@ pgm_no_vtime3: stosm __SF_EMPTY(%r15),0x03 # reenable interrupts j sysc_do_svc +# +# per was called from kernel, must be kprobes +# +kernel_per: + lhi %r0,__LC_PGM_OLD_PSW + sth %r0,SP_TRAP(%r15) # set trap indication to pgm check + la %r2,SP_PTREGS(%r15) # address of register-save area + larl %r14,sysc_leave # load adr. of system ret, no work + jg do_single_step # branch to do_single_step + /* * IO interrupt handler routine */ diff --git a/arch/s390/kernel/kprobes.c b/arch/s390/kernel/kprobes.c new file mode 100644 index 00000000000..ca28fb0b379 --- /dev/null +++ b/arch/s390/kernel/kprobes.c @@ -0,0 +1,657 @@ +/* + * Kernel Probes (KProbes) + * + * 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 the Free Software + * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. + * + * Copyright (C) IBM Corporation, 2002, 2006 + * + * s390 port, used ppc64 as template. Mike Grundy + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +DEFINE_PER_CPU(struct kprobe *, current_kprobe) = NULL; +DEFINE_PER_CPU(struct kprobe_ctlblk, kprobe_ctlblk); + +int __kprobes arch_prepare_kprobe(struct kprobe *p) +{ + /* Make sure the probe isn't going on a difficult instruction */ + if (is_prohibited_opcode((kprobe_opcode_t *) p->addr)) + return -EINVAL; + + if ((unsigned long)p->addr & 0x01) { + printk("Attempt to register kprobe at an unaligned address\n"); + return -EINVAL; + } + + /* Use the get_insn_slot() facility for correctness */ + if (!(p->ainsn.insn = get_insn_slot())) + return -ENOMEM; + + memcpy(p->ainsn.insn, p->addr, MAX_INSN_SIZE * sizeof(kprobe_opcode_t)); + + get_instruction_type(&p->ainsn); + p->opcode = *p->addr; + return 0; +} + +int __kprobes is_prohibited_opcode(kprobe_opcode_t *instruction) +{ + switch (*(__u8 *) instruction) { + case 0x0c: /* bassm */ + case 0x0b: /* bsm */ + case 0x83: /* diag */ + case 0x44: /* ex */ + return -EINVAL; + } + switch (*(__u16 *) instruction) { + case 0x0101: /* pr */ + case 0xb25a: /* bsa */ + case 0xb240: /* bakr */ + case 0xb258: /* bsg */ + case 0xb218: /* pc */ + case 0xb228: /* pt */ + return -EINVAL; + } + return 0; +} + +void __kprobes get_instruction_type(struct arch_specific_insn *ainsn) +{ + /* default fixup method */ + ainsn->fixup = FIXUP_PSW_NORMAL; + + /* save r1 operand */ + ainsn->reg = (*ainsn->insn & 0xf0) >> 4; + + /* save the instruction length (pop 5-5) in bytes */ + switch (*(__u8 *) (ainsn->insn) >> 4) { + case 0: + ainsn->ilen = 2; + break; + case 1: + case 2: + ainsn->ilen = 4; + break; + case 3: + ainsn->ilen = 6; + break; + } + + switch (*(__u8 *) ainsn->insn) { + case 0x05: /* balr */ + case 0x0d: /* basr */ + ainsn->fixup = FIXUP_RETURN_REGISTER; + /* if r2 = 0, no branch will be taken */ + if ((*ainsn->insn & 0x0f) == 0) + ainsn->fixup |= FIXUP_BRANCH_NOT_TAKEN; + break; + case 0x06: /* bctr */ + case 0x07: /* bcr */ + ainsn->fixup = FIXUP_BRANCH_NOT_TAKEN; + break; + case 0x45: /* bal */ + case 0x4d: /* bas */ + ainsn->fixup = FIXUP_RETURN_REGISTER; + break; + case 0x47: /* bc */ + case 0x46: /* bct */ + case 0x86: /* bxh */ + case 0x87: /* bxle */ + ainsn->fixup = FIXUP_BRANCH_NOT_TAKEN; + break; + case 0x82: /* lpsw */ + ainsn->fixup = FIXUP_NOT_REQUIRED; + break; + case 0xb2: /* lpswe */ + if (*(((__u8 *) ainsn->insn) + 1) == 0xb2) { + ainsn->fixup = FIXUP_NOT_REQUIRED; + } + break; + case 0xa7: /* bras */ + if ((*ainsn->insn & 0x0f) == 0x05) { + ainsn->fixup |= FIXUP_RETURN_REGISTER; + } + break; + case 0xc0: + if ((*ainsn->insn & 0x0f) == 0x00 /* larl */ + || (*ainsn->insn & 0x0f) == 0x05) /* brasl */ + ainsn->fixup |= FIXUP_RETURN_REGISTER; + break; + case 0xeb: + if (*(((__u8 *) ainsn->insn) + 5 ) == 0x44 || /* bxhg */ + *(((__u8 *) ainsn->insn) + 5) == 0x45) {/* bxleg */ + ainsn->fixup = FIXUP_BRANCH_NOT_TAKEN; + } + break; + case 0xe3: /* bctg */ + if (*(((__u8 *) ainsn->insn) + 5) == 0x46) { + ainsn->fixup = FIXUP_BRANCH_NOT_TAKEN; + } + break; + } +} + +static int __kprobes swap_instruction(void *aref) +{ + struct ins_replace_args *args = aref; + int err = -EFAULT; + + asm volatile( + "0: mvc 0(2,%2),0(%3)\n" + "1: la %0,0\n" + "2:\n" + EX_TABLE(0b,2b) + : "+d" (err), "=m" (*args->ptr) + : "a" (args->ptr), "a" (&args->new), "m" (args->new)); + return err; +} + +void __kprobes arch_arm_kprobe(struct kprobe *p) +{ + struct kprobe_ctlblk *kcb = get_kprobe_ctlblk(); + unsigned long status = kcb->kprobe_status; + struct ins_replace_args args; + + args.ptr = p->addr; + args.old = p->opcode; + args.new = BREAKPOINT_INSTRUCTION; + + kcb->kprobe_status = KPROBE_SWAP_INST; + stop_machine_run(swap_instruction, &args, NR_CPUS); + kcb->kprobe_status = status; +} + +void __kprobes arch_disarm_kprobe(struct kprobe *p) +{ + struct kprobe_ctlblk *kcb = get_kprobe_ctlblk(); + unsigned long status = kcb->kprobe_status; + struct ins_replace_args args; + + args.ptr = p->addr; + args.old = BREAKPOINT_INSTRUCTION; + args.new = p->opcode; + + kcb->kprobe_status = KPROBE_SWAP_INST; + stop_machine_run(swap_instruction, &args, NR_CPUS); + kcb->kprobe_status = status; +} + +void __kprobes arch_remove_kprobe(struct kprobe *p) +{ + mutex_lock(&kprobe_mutex); + free_insn_slot(p->ainsn.insn); + mutex_unlock(&kprobe_mutex); +} + +static void __kprobes prepare_singlestep(struct kprobe *p, struct pt_regs *regs) +{ + per_cr_bits kprobe_per_regs[1]; + + memset(kprobe_per_regs, 0, sizeof(per_cr_bits)); + regs->psw.addr = (unsigned long)p->ainsn.insn | PSW_ADDR_AMODE; + + /* Set up the per control reg info, will pass to lctl */ + kprobe_per_regs[0].em_instruction_fetch = 1; + kprobe_per_regs[0].starting_addr = (unsigned long)p->ainsn.insn; + kprobe_per_regs[0].ending_addr = (unsigned long)p->ainsn.insn + 1; + + /* Set the PER control regs, turns on single step for this address */ + __ctl_load(kprobe_per_regs, 9, 11); + regs->psw.mask |= PSW_MASK_PER; + regs->psw.mask &= ~(PSW_MASK_IO | PSW_MASK_EXT | PSW_MASK_MCHECK); +} + +static void __kprobes save_previous_kprobe(struct kprobe_ctlblk *kcb) +{ + kcb->prev_kprobe.kp = kprobe_running(); + kcb->prev_kprobe.status = kcb->kprobe_status; + kcb->prev_kprobe.kprobe_saved_imask = kcb->kprobe_saved_imask; + memcpy(kcb->prev_kprobe.kprobe_saved_ctl, kcb->kprobe_saved_ctl, + sizeof(kcb->kprobe_saved_ctl)); +} + +static void __kprobes restore_previous_kprobe(struct kprobe_ctlblk *kcb) +{ + __get_cpu_var(current_kprobe) = kcb->prev_kprobe.kp; + kcb->kprobe_status = kcb->prev_kprobe.status; + kcb->kprobe_saved_imask = kcb->prev_kprobe.kprobe_saved_imask; + memcpy(kcb->kprobe_saved_ctl, kcb->prev_kprobe.kprobe_saved_ctl, + sizeof(kcb->kprobe_saved_ctl)); +} + +static void __kprobes set_current_kprobe(struct kprobe *p, struct pt_regs *regs, + struct kprobe_ctlblk *kcb) +{ + __get_cpu_var(current_kprobe) = p; + /* Save the interrupt and per flags */ + kcb->kprobe_saved_imask = regs->psw.mask & + (PSW_MASK_PER | PSW_MASK_IO | PSW_MASK_EXT | PSW_MASK_MCHECK); + /* Save the control regs that govern PER */ + __ctl_store(kcb->kprobe_saved_ctl, 9, 11); +} + +/* Called with kretprobe_lock held */ +void __kprobes arch_prepare_kretprobe(struct kretprobe *rp, + struct pt_regs *regs) +{ + struct kretprobe_instance *ri; + + if ((ri = get_free_rp_inst(rp)) != NULL) { + ri->rp = rp; + ri->task = current; + ri->ret_addr = (kprobe_opcode_t *) regs->gprs[14]; + + /* Replace the return addr with trampoline addr */ + regs->gprs[14] = (unsigned long)&kretprobe_trampoline; + + add_rp_inst(ri); + } else { + rp->nmissed++; + } +} + +static int __kprobes kprobe_handler(struct pt_regs *regs) +{ + struct kprobe *p; + int ret = 0; + unsigned long *addr = (unsigned long *) + ((regs->psw.addr & PSW_ADDR_INSN) - 2); + struct kprobe_ctlblk *kcb; + + /* + * We don't want to be preempted for the entire + * duration of kprobe processing + */ + preempt_disable(); + kcb = get_kprobe_ctlblk(); + + /* Check we're not actually recursing */ + if (kprobe_running()) { + p = get_kprobe(addr); + if (p) { + if (kcb->kprobe_status == KPROBE_HIT_SS && + *p->ainsn.insn == BREAKPOINT_INSTRUCTION) { + regs->psw.mask &= ~PSW_MASK_PER; + regs->psw.mask |= kcb->kprobe_saved_imask; + goto no_kprobe; + } + /* We have reentered the kprobe_handler(), since + * another probe was hit while within the handler. + * We here save the original kprobes variables and + * just single step on the instruction of the new probe + * without calling any user handlers. + */ + save_previous_kprobe(kcb); + set_current_kprobe(p, regs, kcb); + kprobes_inc_nmissed_count(p); + prepare_singlestep(p, regs); + kcb->kprobe_status = KPROBE_REENTER; + return 1; + } else { + p = __get_cpu_var(current_kprobe); + if (p->break_handler && p->break_handler(p, regs)) { + goto ss_probe; + } + } + goto no_kprobe; + } + + p = get_kprobe(addr); + if (!p) { + if (*addr != BREAKPOINT_INSTRUCTION) { + /* + * The breakpoint instruction was removed right + * after we hit it. Another cpu has removed + * either a probepoint or a debugger breakpoint + * at this address. In either case, no further + * handling of this interrupt is appropriate. + * + */ + ret = 1; + } + /* Not one of ours: let kernel handle it */ + goto no_kprobe; + } + + kcb->kprobe_status = KPROBE_HIT_ACTIVE; + set_current_kprobe(p, regs, kcb); + if (p->pre_handler && p->pre_handler(p, regs)) + /* handler has already set things up, so skip ss setup */ + return 1; + +ss_probe: + prepare_singlestep(p, regs); + kcb->kprobe_status = KPROBE_HIT_SS; + return 1; + +no_kprobe: + preempt_enable_no_resched(); + return ret; +} + +/* + * Function return probe trampoline: + * - init_kprobes() establishes a probepoint here + * - When the probed function returns, this probe + * causes the handlers to fire + */ +void __kprobes kretprobe_trampoline_holder(void) +{ + asm volatile(".global kretprobe_trampoline\n" + "kretprobe_trampoline: bcr 0,0\n"); +} + +/* + * Called when the probe at kretprobe trampoline is hit + */ +int __kprobes trampoline_probe_handler(struct kprobe *p, struct pt_regs *regs) +{ + struct kretprobe_instance *ri = NULL; + struct hlist_head *head; + struct hlist_node *node, *tmp; + unsigned long flags, orig_ret_address = 0; + unsigned long trampoline_address = (unsigned long)&kretprobe_trampoline; + + spin_lock_irqsave(&kretprobe_lock, flags); + head = kretprobe_inst_table_head(current); + + /* + * It is possible to have multiple instances associated with a given + * task either because an multiple functions in the call path + * have a return probe installed on them, and/or more then one return + * return probe was registered for a target function. + * + * We can handle this because: + * - instances are always inserted at the head of the list + * - when multiple return probes are registered for the same + * function, the first instance's ret_addr will point to the + * real return address, and all the rest will point to + * kretprobe_trampoline + */ + hlist_for_each_entry_safe(ri, node, tmp, head, hlist) { + if (ri->task != current) + /* another task is sharing our hash bucket */ + continue; + + if (ri->rp && ri->rp->handler) + ri->rp->handler(ri, regs); + + orig_ret_address = (unsigned long)ri->ret_addr; + recycle_rp_inst(ri); + + if (orig_ret_address != trampoline_address) { + /* + * This is the real return address. Any other + * instances associated with this task are for + * other calls deeper on the call stack + */ + break; + } + } + BUG_ON(!orig_ret_address || (orig_ret_address == trampoline_address)); + regs->psw.addr = orig_ret_address | PSW_ADDR_AMODE; + + reset_current_kprobe(); + spin_unlock_irqrestore(&kretprobe_lock, flags); + preempt_enable_no_resched(); + + /* + * By returning a non-zero value, we are telling + * kprobe_handler() that we don't want the post_handler + * to run (and have re-enabled preemption) + */ + return 1; +} + +/* + * Called after single-stepping. p->addr is the address of the + * instruction whose first byte has been replaced by the "breakpoint" + * instruction. To avoid the SMP problems that can occur when we + * temporarily put back the original opcode to single-step, we + * single-stepped a copy of the instruction. The address of this + * copy is p->ainsn.insn. + */ +static void __kprobes resume_execution(struct kprobe *p, struct pt_regs *regs) +{ + struct kprobe_ctlblk *kcb = get_kprobe_ctlblk(); + + regs->psw.addr &= PSW_ADDR_INSN; + + if (p->ainsn.fixup & FIXUP_PSW_NORMAL) + regs->psw.addr = (unsigned long)p->addr + + ((unsigned long)regs->psw.addr - + (unsigned long)p->ainsn.insn); + + if (p->ainsn.fixup & FIXUP_BRANCH_NOT_TAKEN) + if ((unsigned long)regs->psw.addr - + (unsigned long)p->ainsn.insn == p->ainsn.ilen) + regs->psw.addr = (unsigned long)p->addr + p->ainsn.ilen; + + if (p->ainsn.fixup & FIXUP_RETURN_REGISTER) + regs->gprs[p->ainsn.reg] = ((unsigned long)p->addr + + (regs->gprs[p->ainsn.reg] - + (unsigned long)p->ainsn.insn)) + | PSW_ADDR_AMODE; + + regs->psw.addr |= PSW_ADDR_AMODE; + /* turn off PER mode */ + regs->psw.mask &= ~PSW_MASK_PER; + /* Restore the original per control regs */ + __ctl_load(kcb->kprobe_saved_ctl, 9, 11); + regs->psw.mask |= kcb->kprobe_saved_imask; +} + +static int __kprobes post_kprobe_handler(struct pt_regs *regs) +{ + struct kprobe *cur = kprobe_running(); + struct kprobe_ctlblk *kcb = get_kprobe_ctlblk(); + + if (!cur) + return 0; + + if ((kcb->kprobe_status != KPROBE_REENTER) && cur->post_handler) { + kcb->kprobe_status = KPROBE_HIT_SSDONE; + cur->post_handler(cur, regs, 0); + } + + resume_execution(cur, regs); + + /*Restore back the original saved kprobes variables and continue. */ + if (kcb->kprobe_status == KPROBE_REENTER) { + restore_previous_kprobe(kcb); + goto out; + } + reset_current_kprobe(); +out: + preempt_enable_no_resched(); + + /* + * if somebody else is singlestepping across a probe point, psw mask + * will have PER set, in which case, continue the remaining processing + * of do_single_step, as if this is not a probe hit. + */ + if (regs->psw.mask & PSW_MASK_PER) { + return 0; + } + + return 1; +} + +static int __kprobes kprobe_fault_handler(struct pt_regs *regs, int trapnr) +{ + struct kprobe *cur = kprobe_running(); + struct kprobe_ctlblk *kcb = get_kprobe_ctlblk(); + const struct exception_table_entry *entry; + + switch(kcb->kprobe_status) { + case KPROBE_SWAP_INST: + /* We are here because the instruction replacement failed */ + return 0; + case KPROBE_HIT_SS: + case KPROBE_REENTER: + /* + * We are here because the instruction being single + * stepped caused a page fault. We reset the current + * kprobe and the nip points back to the probe address + * and allow the page fault handler to continue as a + * normal page fault. + */ + regs->psw.addr = (unsigned long)cur->addr | PSW_ADDR_AMODE; + regs->psw.mask &= ~PSW_MASK_PER; + regs->psw.mask |= kcb->kprobe_saved_imask; + if (kcb->kprobe_status == KPROBE_REENTER) + restore_previous_kprobe(kcb); + else + reset_current_kprobe(); + preempt_enable_no_resched(); + break; + case KPROBE_HIT_ACTIVE: + case KPROBE_HIT_SSDONE: + /* + * We increment the nmissed count for accounting, + * we can also use npre/npostfault count for accouting + * these specific fault cases. + */ + kprobes_inc_nmissed_count(cur); + + /* + * We come here because instructions in the pre/post + * handler caused the page_fault, this could happen + * if handler tries to access user space by + * copy_from_user(), get_user() etc. Let the + * user-specified handler try to fix it first. + */ + if (cur->fault_handler && cur->fault_handler(cur, regs, trapnr)) + return 1; + + /* + * In case the user-specified fault handler returned + * zero, try to fix up. + */ + entry = search_exception_tables(regs->psw.addr & PSW_ADDR_INSN); + if (entry) { + regs->psw.addr = entry->fixup | PSW_ADDR_AMODE; + return 1; + } + + /* + * fixup_exception() could not handle it, + * Let do_page_fault() fix it. + */ + break; + default: + break; + } + return 0; +} + +/* + * Wrapper routine to for handling exceptions. + */ +int __kprobes kprobe_exceptions_notify(struct notifier_block *self, + unsigned long val, void *data) +{ + struct die_args *args = (struct die_args *)data; + int ret = NOTIFY_DONE; + + switch (val) { + case DIE_BPT: + if (kprobe_handler(args->regs)) + ret = NOTIFY_STOP; + break; + case DIE_SSTEP: + if (post_kprobe_handler(args->regs)) + ret = NOTIFY_STOP; + break; + case DIE_TRAP: + case DIE_PAGE_FAULT: + /* kprobe_running() needs smp_processor_id() */ + preempt_disable(); + if (kprobe_running() && + kprobe_fault_handler(args->regs, args->trapnr)) + ret = NOTIFY_STOP; + preempt_enable(); + break; + default: + break; + } + return ret; +} + +int __kprobes setjmp_pre_handler(struct kprobe *p, struct pt_regs *regs) +{ + struct jprobe *jp = container_of(p, struct jprobe, kp); + unsigned long addr; + struct kprobe_ctlblk *kcb = get_kprobe_ctlblk(); + + memcpy(&kcb->jprobe_saved_regs, regs, sizeof(struct pt_regs)); + + /* setup return addr to the jprobe handler routine */ + regs->psw.addr = (unsigned long)(jp->entry) | PSW_ADDR_AMODE; + + /* r14 is the function return address */ + kcb->jprobe_saved_r14 = (unsigned long)regs->gprs[14]; + /* r15 is the stack pointer */ + kcb->jprobe_saved_r15 = (unsigned long)regs->gprs[15]; + addr = (unsigned long)kcb->jprobe_saved_r15; + + memcpy(kcb->jprobes_stack, (kprobe_opcode_t *) addr, + MIN_STACK_SIZE(addr)); + return 1; +} + +void __kprobes jprobe_return(void) +{ + asm volatile(".word 0x0002"); +} + +void __kprobes jprobe_return_end(void) +{ + asm volatile("bcr 0,0"); +} + +int __kprobes longjmp_break_handler(struct kprobe *p, struct pt_regs *regs) +{ + struct kprobe_ctlblk *kcb = get_kprobe_ctlblk(); + unsigned long stack_addr = (unsigned long)(kcb->jprobe_saved_r15); + + /* Put the regs back */ + memcpy(regs, &kcb->jprobe_saved_regs, sizeof(struct pt_regs)); + /* put the stack back */ + memcpy((kprobe_opcode_t *) stack_addr, kcb->jprobes_stack, + MIN_STACK_SIZE(stack_addr)); + preempt_enable_no_resched(); + return 1; +} + +static struct kprobe trampoline_p = { + .addr = (kprobe_opcode_t *) & kretprobe_trampoline, + .pre_handler = trampoline_probe_handler +}; + +int __init arch_init_kprobes(void) +{ + return register_kprobe(&trampoline_p); +} diff --git a/arch/s390/kernel/traps.c b/arch/s390/kernel/traps.c index bde1d1d5985..c4982c96342 100644 --- a/arch/s390/kernel/traps.c +++ b/arch/s390/kernel/traps.c @@ -29,6 +29,7 @@ #include #include #include +#include #include #include @@ -39,6 +40,7 @@ #include #include #include +#include /* Called from entry.S only */ extern void handle_per_exception(struct pt_regs *regs); @@ -74,6 +76,20 @@ static int kstack_depth_to_print = 12; static int kstack_depth_to_print = 20; #endif /* CONFIG_64BIT */ +ATOMIC_NOTIFIER_HEAD(s390die_chain); + +int register_die_notifier(struct notifier_block *nb) +{ + return atomic_notifier_chain_register(&s390die_chain, nb); +} +EXPORT_SYMBOL(register_die_notifier); + +int unregister_die_notifier(struct notifier_block *nb) +{ + return atomic_notifier_chain_unregister(&s390die_chain, nb); +} +EXPORT_SYMBOL(unregister_die_notifier); + /* * For show_trace we have tree different stack to consider: * - the panic stack which is used if the kernel stack has overflown @@ -305,8 +321,9 @@ report_user_fault(long interruption_code, struct pt_regs *regs) #endif } -static void inline do_trap(long interruption_code, int signr, char *str, - struct pt_regs *regs, siginfo_t *info) +static void __kprobes inline do_trap(long interruption_code, int signr, + char *str, struct pt_regs *regs, + siginfo_t *info) { /* * We got all needed information from the lowcore and can @@ -315,6 +332,10 @@ static void inline do_trap(long interruption_code, int signr, char *str, if (regs->psw.mask & PSW_MASK_PSTATE) local_irq_enable(); + if (notify_die(DIE_TRAP, str, regs, interruption_code, + interruption_code, signr) == NOTIFY_STOP) + return; + if (regs->psw.mask & PSW_MASK_PSTATE) { struct task_struct *tsk = current; @@ -336,8 +357,12 @@ static inline void __user *get_check_address(struct pt_regs *regs) return (void __user *)((regs->psw.addr-S390_lowcore.pgm_ilc) & PSW_ADDR_INSN); } -void do_single_step(struct pt_regs *regs) +void __kprobes do_single_step(struct pt_regs *regs) { + if (notify_die(DIE_SSTEP, "sstep", regs, 0, 0, + SIGTRAP) == NOTIFY_STOP){ + return; + } if ((current->ptrace & PT_PTRACED) != 0) force_sig(SIGTRAP, current); } diff --git a/arch/s390/kernel/vmlinux.lds.S b/arch/s390/kernel/vmlinux.lds.S index ff5f7bb34f7..df0c16ab8e9 100644 --- a/arch/s390/kernel/vmlinux.lds.S +++ b/arch/s390/kernel/vmlinux.lds.S @@ -24,6 +24,7 @@ SECTIONS *(.text) SCHED_TEXT LOCK_TEXT + KPROBES_TEXT *(.fixup) *(.gnu.warning) } = 0x0700 diff --git a/arch/s390/mm/fault.c b/arch/s390/mm/fault.c index 7cd82575813..44f0cda7e72 100644 --- a/arch/s390/mm/fault.c +++ b/arch/s390/mm/fault.c @@ -25,10 +25,12 @@ #include #include #include +#include #include #include #include +#include #ifndef CONFIG_64BIT #define __FAIL_ADDR_MASK 0x7ffff000 @@ -48,6 +50,38 @@ extern int sysctl_userprocess_debug; extern void die(const char *,struct pt_regs *,long); +#ifdef CONFIG_KPROBES +ATOMIC_NOTIFIER_HEAD(notify_page_fault_chain); +int register_page_fault_notifier(struct notifier_block *nb) +{ + return atomic_notifier_chain_register(¬ify_page_fault_chain, nb); +} + +int unregister_page_fault_notifier(struct notifier_block *nb) +{ + return atomic_notifier_chain_unregister(¬ify_page_fault_chain, nb); +} + +static inline int notify_page_fault(enum die_val val, const char *str, + struct pt_regs *regs, long err, int trap, int sig) +{ + struct die_args args = { + .regs = regs, + .str = str, + .err = err, + .trapnr = trap, + .signr = sig + }; + return atomic_notifier_call_chain(¬ify_page_fault_chain, val, &args); +} +#else +static inline int notify_page_fault(enum die_val val, const char *str, + struct pt_regs *regs, long err, int trap, int sig) +{ + return NOTIFY_DONE; +} +#endif + extern spinlock_t timerlist_lock; /* @@ -159,7 +193,7 @@ static void do_sigsegv(struct pt_regs *regs, unsigned long error_code, * 11 Page translation -> Not present (nullification) * 3b Region third trans. -> Not present (nullification) */ -static inline void +static inline void __kprobes do_exception(struct pt_regs *regs, unsigned long error_code, int is_protection) { struct task_struct *tsk; @@ -173,6 +207,10 @@ do_exception(struct pt_regs *regs, unsigned long error_code, int is_protection) tsk = current; mm = tsk->mm; + if (notify_page_fault(DIE_PAGE_FAULT, "page fault", regs, error_code, 14, + SIGSEGV) == NOTIFY_STOP) + return; + /* * Check for low-address protection. This needs to be treated * as a special case because the translation exception code -- cgit v1.2.3 From 65912a84c0f33304fa5ea004c7b6ee58d5f5572e Mon Sep 17 00:00:00 2001 From: Heiko Carstens Date: Wed, 20 Sep 2006 15:58:41 +0200 Subject: [S390] initrd vs. bootmem bitmap. Move initrd if the bitmap of the bootmem allocator would overwrite it. In addition this patch sets the default size and address of the initrd to 0. Therefore all boot loaders must set the initrd size and address correctly. This is especially relevant for ftp boot via HMC/SE, where this change requires a special patch file entry in the .ins file which sets these two values contained at address 0x10408 and 0x10410. Signed-off-by: Heiko Carstens Signed-off-by: Martin Schwidefsky --- arch/s390/kernel/head.S | 10 ++++++---- arch/s390/kernel/head31.S | 4 ++-- arch/s390/kernel/head64.S | 4 ++-- arch/s390/kernel/setup.c | 41 ++++++++++++++++++++++++++++++++++++++--- 4 files changed, 48 insertions(+), 11 deletions(-) (limited to 'arch') diff --git a/arch/s390/kernel/head.S b/arch/s390/kernel/head.S index adad8863ee2..a6e9bdb5359 100644 --- a/arch/s390/kernel/head.S +++ b/arch/s390/kernel/head.S @@ -272,7 +272,7 @@ iplstart: # load parameter file from ipl device # .Lagain1: - l %r2,INITRD_START+ARCH_OFFSET-PARMAREA(%r12) # ramdisk loc. is temp + l %r2,.Linitrd # ramdisk loc. is temp bas %r14,.Lloader # load parameter file ltr %r2,%r2 # got anything ? bz .Lnopf @@ -280,7 +280,7 @@ iplstart: bnh .Lnotrunc la %r2,895 .Lnotrunc: - l %r4,INITRD_START+ARCH_OFFSET-PARMAREA(%r12) + l %r4,.Linitrd clc 0(3,%r4),.L_hdr # if it is HDRx bz .Lagain1 # skip dataset header clc 0(3,%r4),.L_eof # if it is EOFx @@ -323,14 +323,15 @@ iplstart: # load ramdisk from ipl device # .Lagain2: - l %r2,INITRD_START+ARCH_OFFSET-PARMAREA(%r12) # addr of ramdisk + l %r2,.Linitrd # addr of ramdisk + st %r2,INITRD_START+ARCH_OFFSET-PARMAREA(%r12) bas %r14,.Lloader # load ramdisk st %r2,INITRD_SIZE+ARCH_OFFSET-PARMAREA(%r12) # store size of ramdisk ltr %r2,%r2 bnz .Lrdcont st %r2,INITRD_START+ARCH_OFFSET-PARMAREA(%r12) # no ramdisk found .Lrdcont: - l %r2,INITRD_START+ARCH_OFFSET-PARMAREA(%r12) + l %r2,.Linitrd clc 0(3,%r2),.L_hdr # skip HDRx and EOFx bz .Lagain2 @@ -379,6 +380,7 @@ iplstart: l %r1,.Lstartup br %r1 +.Linitrd:.long _end + 0x400000 # default address of initrd .Lparm: .long PARMAREA .Lstartup: .long startup .Lcvtab:.long _ebcasc # ebcdic to ascii table diff --git a/arch/s390/kernel/head31.S b/arch/s390/kernel/head31.S index a4dc61f3285..0e46077d714 100644 --- a/arch/s390/kernel/head31.S +++ b/arch/s390/kernel/head31.S @@ -26,8 +26,8 @@ startup:basr %r13,0 # get base # .org PARMAREA .long 0,0 # IPL_DEVICE - .long 0,RAMDISK_ORIGIN # INITRD_START - .long 0,RAMDISK_SIZE # INITRD_SIZE + .long 0,0 # INITRD_START + .long 0,0 # INITRD_SIZE .org COMMAND_LINE .byte "root=/dev/ram0 ro" diff --git a/arch/s390/kernel/head64.S b/arch/s390/kernel/head64.S index 9d80c5b1ef9..3e0341acd04 100644 --- a/arch/s390/kernel/head64.S +++ b/arch/s390/kernel/head64.S @@ -26,8 +26,8 @@ startup:basr %r13,0 # get base # .org PARMAREA .quad 0 # IPL_DEVICE - .quad RAMDISK_ORIGIN # INITRD_START - .quad RAMDISK_SIZE # INITRD_SIZE + .quad 0 # INITRD_START + .quad 0 # INITRD_SIZE .org COMMAND_LINE .byte "root=/dev/ram0 ro" diff --git a/arch/s390/kernel/setup.c b/arch/s390/kernel/setup.c index c902f059c7a..89051e8a5d8 100644 --- a/arch/s390/kernel/setup.c +++ b/arch/s390/kernel/setup.c @@ -37,6 +37,7 @@ #include #include #include +#include #include #include @@ -501,13 +502,47 @@ setup_memory(void) * partially used pages are not usable - thus * we are rounding upwards: */ - start_pfn = (__pa(&_end) + PAGE_SIZE - 1) >> PAGE_SHIFT; - end_pfn = max_pfn = memory_end >> PAGE_SHIFT; + start_pfn = PFN_UP(__pa(&_end)); + end_pfn = max_pfn = PFN_DOWN(memory_end); /* Initialize storage key for kernel pages */ for (init_pfn = 0 ; init_pfn < start_pfn; init_pfn++) page_set_storage_key(init_pfn << PAGE_SHIFT, PAGE_DEFAULT_KEY); +#ifdef CONFIG_BLK_DEV_INITRD + /* + * Move the initrd in case the bitmap of the bootmem allocater + * would overwrite it. + */ + + if (INITRD_START && INITRD_SIZE) { + unsigned long bmap_size; + unsigned long start; + + bmap_size = bootmem_bootmap_pages(end_pfn - start_pfn + 1); + bmap_size = PFN_PHYS(bmap_size); + + if (PFN_PHYS(start_pfn) + bmap_size > INITRD_START) { + start = PFN_PHYS(start_pfn) + bmap_size + PAGE_SIZE; + + if (start + INITRD_SIZE > memory_end) { + printk("initrd extends beyond end of memory " + "(0x%08lx > 0x%08lx)\n" + "disabling initrd\n", + start + INITRD_SIZE, memory_end); + INITRD_START = INITRD_SIZE = 0; + } else { + printk("Moving initrd (0x%08lx -> 0x%08lx, " + "size: %ld)\n", + INITRD_START, start, INITRD_SIZE); + memmove((void *) start, (void *) INITRD_START, + INITRD_SIZE); + INITRD_START = start; + } + } + } +#endif + /* * Initialize the boot-time allocator (with low memory only): */ @@ -559,7 +594,7 @@ setup_memory(void) reserve_bootmem(start_pfn << PAGE_SHIFT, bootmap_size); #ifdef CONFIG_BLK_DEV_INITRD - if (INITRD_START) { + if (INITRD_START && INITRD_SIZE) { if (INITRD_START + INITRD_SIZE <= memory_end) { reserve_bootmem(INITRD_START, INITRD_SIZE); initrd_start = INITRD_START; -- cgit v1.2.3 From f19bfb2c9b8675590fbecb43e5ce3b34ee321185 Mon Sep 17 00:00:00 2001 From: Michael Holzheu Date: Wed, 20 Sep 2006 15:58:44 +0200 Subject: [S390] hypfs comment cleanup. Correct some comments in the hypervisor filesystem. Signed-off-by: Michael Holzheu Signed-off-by: Martin Schwidefsky --- arch/s390/hypfs/hypfs.h | 2 +- arch/s390/hypfs/hypfs_diag.c | 2 +- arch/s390/hypfs/hypfs_diag.h | 2 +- arch/s390/hypfs/inode.c | 2 +- 4 files changed, 4 insertions(+), 4 deletions(-) (limited to 'arch') diff --git a/arch/s390/hypfs/hypfs.h b/arch/s390/hypfs/hypfs.h index ea5567be00f..f3dbd91965c 100644 --- a/arch/s390/hypfs/hypfs.h +++ b/arch/s390/hypfs/hypfs.h @@ -1,5 +1,5 @@ /* - * fs/hypfs/hypfs.h + * arch/s390/hypfs/hypfs.h * Hypervisor filesystem for Linux on s390. * * Copyright (C) IBM Corp. 2006 diff --git a/arch/s390/hypfs/hypfs_diag.c b/arch/s390/hypfs/hypfs_diag.c index 1785bce2b91..874d761c981 100644 --- a/arch/s390/hypfs/hypfs_diag.c +++ b/arch/s390/hypfs/hypfs_diag.c @@ -1,5 +1,5 @@ /* - * fs/hypfs/hypfs_diag.c + * arch/s390/hypfs/hypfs_diag.c * Hypervisor filesystem for Linux on s390. Diag 204 and 224 * implementation. * diff --git a/arch/s390/hypfs/hypfs_diag.h b/arch/s390/hypfs/hypfs_diag.h index 793dea6b9bb..256b384aebe 100644 --- a/arch/s390/hypfs/hypfs_diag.h +++ b/arch/s390/hypfs/hypfs_diag.h @@ -1,5 +1,5 @@ /* - * fs/hypfs/hypfs_diag.h + * arch/s390/hypfs_diag.h * Hypervisor filesystem for Linux on s390. * * Copyright (C) IBM Corp. 2006 diff --git a/arch/s390/hypfs/inode.c b/arch/s390/hypfs/inode.c index 18c091925ea..bdcad2ea1ff 100644 --- a/arch/s390/hypfs/inode.c +++ b/arch/s390/hypfs/inode.c @@ -1,5 +1,5 @@ /* - * fs/hypfs/inode.c + * arch/s390/hypfs/inode.c * Hypervisor filesystem for Linux on s390. * * Copyright (C) IBM Corp. 2006 -- cgit v1.2.3 From 331c982d4a6b43cdc0d056956a1cae8a7d6237bf Mon Sep 17 00:00:00 2001 From: Michael Holzheu Date: Wed, 20 Sep 2006 15:58:47 +0200 Subject: [S390] hypfs compiler warnings. Add casts to avoid compiler warnings. Signed-off-by: Michael Holzheu Signed-off-by: Martin Schwidefsky --- arch/s390/hypfs/hypfs_diag.c | 12 ++++++++---- 1 file changed, 8 insertions(+), 4 deletions(-) (limited to 'arch') diff --git a/arch/s390/hypfs/hypfs_diag.c b/arch/s390/hypfs/hypfs_diag.c index 874d761c981..fee5aee605f 100644 --- a/arch/s390/hypfs/hypfs_diag.c +++ b/arch/s390/hypfs/hypfs_diag.c @@ -432,12 +432,14 @@ static int diag204_probe(void) buf = diag204_get_buffer(INFO_EXT, &pages); if (!IS_ERR(buf)) { - if (diag204(SUBC_STIB7 | INFO_EXT, pages, buf) >= 0) { + if (diag204((unsigned long)SUBC_STIB7 | + (unsigned long)INFO_EXT, pages, buf) >= 0) { diag204_store_sc = SUBC_STIB7; diag204_info_type = INFO_EXT; goto out; } - if (diag204(SUBC_STIB6 | INFO_EXT, pages, buf) >= 0) { + if (diag204((unsigned long)SUBC_STIB6 | + (unsigned long)INFO_EXT, pages, buf) >= 0) { diag204_store_sc = SUBC_STIB7; diag204_info_type = INFO_EXT; goto out; @@ -452,7 +454,8 @@ static int diag204_probe(void) rc = PTR_ERR(buf); goto fail_alloc; } - if (diag204(SUBC_STIB4 | INFO_SIMPLE, pages, buf) >= 0) { + if (diag204((unsigned long)SUBC_STIB4 | + (unsigned long)INFO_SIMPLE, pages, buf) >= 0) { diag204_store_sc = SUBC_STIB4; diag204_info_type = INFO_SIMPLE; goto out; @@ -476,7 +479,8 @@ static void *diag204_store(void) buf = diag204_get_buffer(diag204_info_type, &pages); if (IS_ERR(buf)) goto out; - if (diag204(diag204_store_sc | diag204_info_type, pages, buf) < 0) + if (diag204((unsigned long)diag204_store_sc | + (unsigned long)diag204_info_type, pages, buf) < 0) return ERR_PTR(-ENOSYS); out: return buf; -- cgit v1.2.3 From ff6b8ea68f4b7353f88b97024f28127e2148aa00 Mon Sep 17 00:00:00 2001 From: Michael Holzheu Date: Wed, 20 Sep 2006 15:58:49 +0200 Subject: [S390] ipl/dump on panic. It is now possible to specify a ccw/fcp dump device which is used to automatically create a system dump in case of a kernel panic. The dump device can be configured under /sys/firmware/dump. In addition it is now possible to specify a ccw/fcp device which is used for the next reboot of Linux. The reipl device can be configured under /sys/firmware/reipl. Signed-off-by: Michael Holzheu Signed-off-by: Martin Schwidefsky --- arch/s390/kernel/Makefile | 2 +- arch/s390/kernel/head31.S | 3 + arch/s390/kernel/head64.S | 3 + arch/s390/kernel/ipl.c | 942 ++++++++++++++++++++++++++++++++++++++++++ arch/s390/kernel/reipl.S | 33 +- arch/s390/kernel/reipl64.S | 34 +- arch/s390/kernel/reipl_diag.c | 39 -- arch/s390/kernel/setup.c | 220 +--------- arch/s390/kernel/smp.c | 10 +- 9 files changed, 1002 insertions(+), 284 deletions(-) create mode 100644 arch/s390/kernel/ipl.c delete mode 100644 arch/s390/kernel/reipl_diag.c (limited to 'arch') diff --git a/arch/s390/kernel/Makefile b/arch/s390/kernel/Makefile index 33a5069c0e1..aa978978d3d 100644 --- a/arch/s390/kernel/Makefile +++ b/arch/s390/kernel/Makefile @@ -6,7 +6,7 @@ EXTRA_AFLAGS := -traditional obj-y := bitmap.o traps.o time.o process.o \ setup.o sys_s390.o ptrace.o signal.o cpcmd.o ebcdic.o \ - semaphore.o s390_ext.o debug.o profile.o irq.o reipl_diag.o + semaphore.o s390_ext.o debug.o profile.o irq.o ipl.o obj-y += $(if $(CONFIG_64BIT),entry64.o,entry.o) obj-y += $(if $(CONFIG_64BIT),reipl64.o,reipl.o) diff --git a/arch/s390/kernel/head31.S b/arch/s390/kernel/head31.S index 0e46077d714..d8bb68a7252 100644 --- a/arch/s390/kernel/head31.S +++ b/arch/s390/kernel/head31.S @@ -38,6 +38,7 @@ startup:basr %r13,0 # get base startup_continue: basr %r13,0 # get base .LPG1: GET_IPL_DEVICE + mvi __LC_AR_MODE_ID,0 # set ESA flag (mode 0) lctl %c0,%c15,.Lctl-.LPG1(%r13) # load control registers l %r12,.Lparmaddr-.LPG1(%r13) # pointer to parameter area # move IPL device to lowcore @@ -274,6 +275,8 @@ startup_continue: .Lparmaddr: .long PARMAREA .Lsccbaddr: .long .Lsccb .org 0x12000 +.globl s390_readinfo_sccb +s390_readinfo_sccb: .Lsccb: .hword 0x1000 # length, one page .byte 0x00,0x00,0x00 diff --git a/arch/s390/kernel/head64.S b/arch/s390/kernel/head64.S index 3e0341acd04..c2005101fee 100644 --- a/arch/s390/kernel/head64.S +++ b/arch/s390/kernel/head64.S @@ -41,6 +41,7 @@ startup_continue: srl %r13,1 GET_IPL_DEVICE lhi %r1,1 # mode 1 = esame + mvi __LC_AR_MODE_ID,1 # set esame flag slr %r0,%r0 # set cpuid to zero sigp %r1,%r0,0x12 # switch to esame mode sam64 # switch to 64 bit mode @@ -269,6 +270,8 @@ startup_continue: .quad PARMAREA .org 0x12000 +.globl s390_readinfo_sccb +s390_readinfo_sccb: .Lsccb: .hword 0x1000 # length, one page .byte 0x00,0x00,0x00 diff --git a/arch/s390/kernel/ipl.c b/arch/s390/kernel/ipl.c new file mode 100644 index 00000000000..105ee15a2b3 --- /dev/null +++ b/arch/s390/kernel/ipl.c @@ -0,0 +1,942 @@ +/* + * arch/s390/kernel/ipl.c + * ipl/reipl/dump support for Linux on s390. + * + * Copyright (C) IBM Corp. 2005,2006 + * Author(s): Michael Holzheu + * Heiko Carstens + * Volker Sameske + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#define IPL_PARM_BLOCK_VERSION 0 + +enum ipl_type { + IPL_TYPE_NONE = 1, + IPL_TYPE_UNKNOWN = 2, + IPL_TYPE_CCW = 4, + IPL_TYPE_FCP = 8, +}; + +#define IPL_NONE_STR "none" +#define IPL_UNKNOWN_STR "unknown" +#define IPL_CCW_STR "ccw" +#define IPL_FCP_STR "fcp" + +static char *ipl_type_str(enum ipl_type type) +{ + switch (type) { + case IPL_TYPE_NONE: + return IPL_NONE_STR; + case IPL_TYPE_CCW: + return IPL_CCW_STR; + case IPL_TYPE_FCP: + return IPL_FCP_STR; + case IPL_TYPE_UNKNOWN: + default: + return IPL_UNKNOWN_STR; + } +} + +enum ipl_method { + IPL_METHOD_NONE, + IPL_METHOD_CCW_CIO, + IPL_METHOD_CCW_DIAG, + IPL_METHOD_CCW_VM, + IPL_METHOD_FCP_RO_DIAG, + IPL_METHOD_FCP_RW_DIAG, + IPL_METHOD_FCP_RO_VM, +}; + +enum shutdown_action { + SHUTDOWN_REIPL, + SHUTDOWN_DUMP, + SHUTDOWN_STOP, +}; + +#define SHUTDOWN_REIPL_STR "reipl" +#define SHUTDOWN_DUMP_STR "dump" +#define SHUTDOWN_STOP_STR "stop" + +static char *shutdown_action_str(enum shutdown_action action) +{ + switch (action) { + case SHUTDOWN_REIPL: + return SHUTDOWN_REIPL_STR; + case SHUTDOWN_DUMP: + return SHUTDOWN_DUMP_STR; + case SHUTDOWN_STOP: + return SHUTDOWN_STOP_STR; + default: + BUG(); + } +} + +enum diag308_subcode { + DIAG308_IPL = 3, + DIAG308_DUMP = 4, + DIAG308_SET = 5, + DIAG308_STORE = 6, +}; + +enum diag308_ipl_type { + DIAG308_IPL_TYPE_FCP = 0, + DIAG308_IPL_TYPE_CCW = 2, +}; + +enum diag308_opt { + DIAG308_IPL_OPT_IPL = 0x10, + DIAG308_IPL_OPT_DUMP = 0x20, +}; + +enum diag308_rc { + DIAG308_RC_OK = 1, +}; + +static int diag308_set_works = 0; + +static int reipl_capabilities = IPL_TYPE_UNKNOWN; +static enum ipl_type reipl_type = IPL_TYPE_UNKNOWN; +static enum ipl_method reipl_method = IPL_METHOD_NONE; +static struct ipl_parameter_block *reipl_block_fcp; +static struct ipl_parameter_block *reipl_block_ccw; + +static int dump_capabilities = IPL_TYPE_NONE; +static enum ipl_type dump_type = IPL_TYPE_NONE; +static enum ipl_method dump_method = IPL_METHOD_NONE; +static struct ipl_parameter_block *dump_block_fcp; +static struct ipl_parameter_block *dump_block_ccw; + +static enum shutdown_action on_panic_action = SHUTDOWN_STOP; + +static int diag308(unsigned long subcode, void *addr) +{ + register unsigned long _addr asm("0") = (unsigned long)addr; + register unsigned long _rc asm("1") = 0; + + asm volatile ( + " diag %0,%2,0x308\n" + "0: \n" + ".section __ex_table,\"a\"\n" +#ifdef CONFIG_64BIT + " .align 8\n" + " .quad 0b, 0b\n" +#else + " .align 4\n" + " .long 0b, 0b\n" +#endif + ".previous\n" + : "+d" (_addr), "+d" (_rc) + : "d" (subcode) : "cc", "memory" ); + + return _rc; +} + +/* SYSFS */ + +#define DEFINE_IPL_ATTR_RO(_prefix, _name, _format, _value) \ +static ssize_t sys_##_prefix##_##_name##_show(struct subsystem *subsys, \ + char *page) \ +{ \ + return sprintf(page, _format, _value); \ +} \ +static struct subsys_attribute sys_##_prefix##_##_name##_attr = \ + __ATTR(_name, S_IRUGO, sys_##_prefix##_##_name##_show, NULL); + +#define DEFINE_IPL_ATTR_RW(_prefix, _name, _fmt_out, _fmt_in, _value) \ +static ssize_t sys_##_prefix##_##_name##_show(struct subsystem *subsys, \ + char *page) \ +{ \ + return sprintf(page, _fmt_out, \ + (unsigned long long) _value); \ +} \ +static ssize_t sys_##_prefix##_##_name##_store(struct subsystem *subsys,\ + const char *buf, size_t len) \ +{ \ + unsigned long long value; \ + if (sscanf(buf, _fmt_in, &value) != 1) \ + return -EINVAL; \ + _value = value; \ + return len; \ +} \ +static struct subsys_attribute sys_##_prefix##_##_name##_attr = \ + __ATTR(_name,(S_IRUGO | S_IWUSR), \ + sys_##_prefix##_##_name##_show, \ + sys_##_prefix##_##_name##_store); + +static void make_attrs_ro(struct attribute **attrs) +{ + while (*attrs) { + (*attrs)->mode = S_IRUGO; + attrs++; + } +} + +/* + * ipl section + */ + +static enum ipl_type ipl_get_type(void) +{ + struct ipl_parameter_block *ipl = IPL_PARMBLOCK_START; + + if (!IPL_DEVNO_VALID) + return IPL_TYPE_UNKNOWN; + if (!IPL_PARMBLOCK_VALID) + return IPL_TYPE_CCW; + if (ipl->hdr.version > IPL_MAX_SUPPORTED_VERSION) + return IPL_TYPE_UNKNOWN; + if (ipl->hdr.pbt != DIAG308_IPL_TYPE_FCP) + return IPL_TYPE_UNKNOWN; + return IPL_TYPE_FCP; +} + +static ssize_t ipl_type_show(struct subsystem *subsys, char *page) +{ + return sprintf(page, "%s\n", ipl_type_str(ipl_get_type())); +} + +static struct subsys_attribute sys_ipl_type_attr = __ATTR_RO(ipl_type); + +static ssize_t sys_ipl_device_show(struct subsystem *subsys, char *page) +{ + struct ipl_parameter_block *ipl = IPL_PARMBLOCK_START; + + switch (ipl_get_type()) { + case IPL_TYPE_CCW: + return sprintf(page, "0.0.%04x\n", ipl_devno); + case IPL_TYPE_FCP: + return sprintf(page, "0.0.%04x\n", ipl->ipl_info.fcp.devno); + default: + return 0; + } +} + +static struct subsys_attribute sys_ipl_device_attr = + __ATTR(device, S_IRUGO, sys_ipl_device_show, NULL); + +static ssize_t ipl_parameter_read(struct kobject *kobj, char *buf, loff_t off, + size_t count) +{ + unsigned int size = IPL_PARMBLOCK_SIZE; + + if (off > size) + return 0; + if (off + count > size) + count = size - off; + memcpy(buf, (void *)IPL_PARMBLOCK_START + off, count); + return count; +} + +static struct bin_attribute ipl_parameter_attr = { + .attr = { + .name = "binary_parameter", + .mode = S_IRUGO, + .owner = THIS_MODULE, + }, + .size = PAGE_SIZE, + .read = &ipl_parameter_read, +}; + +static ssize_t ipl_scp_data_read(struct kobject *kobj, char *buf, loff_t off, + size_t count) +{ + unsigned int size = IPL_PARMBLOCK_START->ipl_info.fcp.scp_data_len; + void *scp_data = &IPL_PARMBLOCK_START->ipl_info.fcp.scp_data; + + if (off > size) + return 0; + if (off + count > size) + count = size - off; + memcpy(buf, scp_data + off, count); + return count; +} + +static struct bin_attribute ipl_scp_data_attr = { + .attr = { + .name = "scp_data", + .mode = S_IRUGO, + .owner = THIS_MODULE, + }, + .size = PAGE_SIZE, + .read = &ipl_scp_data_read, +}; + +/* FCP ipl device attributes */ + +DEFINE_IPL_ATTR_RO(ipl_fcp, wwpn, "0x%016llx\n", (unsigned long long) + IPL_PARMBLOCK_START->ipl_info.fcp.wwpn); +DEFINE_IPL_ATTR_RO(ipl_fcp, lun, "0x%016llx\n", (unsigned long long) + IPL_PARMBLOCK_START->ipl_info.fcp.lun); +DEFINE_IPL_ATTR_RO(ipl_fcp, bootprog, "%lld\n", (unsigned long long) + IPL_PARMBLOCK_START->ipl_info.fcp.bootprog); +DEFINE_IPL_ATTR_RO(ipl_fcp, br_lba, "%lld\n", (unsigned long long) + IPL_PARMBLOCK_START->ipl_info.fcp.br_lba); + +static struct attribute *ipl_fcp_attrs[] = { + &sys_ipl_type_attr.attr, + &sys_ipl_device_attr.attr, + &sys_ipl_fcp_wwpn_attr.attr, + &sys_ipl_fcp_lun_attr.attr, + &sys_ipl_fcp_bootprog_attr.attr, + &sys_ipl_fcp_br_lba_attr.attr, + NULL, +}; + +static struct attribute_group ipl_fcp_attr_group = { + .attrs = ipl_fcp_attrs, +}; + +/* CCW ipl device attributes */ + +static struct attribute *ipl_ccw_attrs[] = { + &sys_ipl_type_attr.attr, + &sys_ipl_device_attr.attr, + NULL, +}; + +static struct attribute_group ipl_ccw_attr_group = { + .attrs = ipl_ccw_attrs, +}; + +/* UNKNOWN ipl device attributes */ + +static struct attribute *ipl_unknown_attrs[] = { + &sys_ipl_type_attr.attr, + NULL, +}; + +static struct attribute_group ipl_unknown_attr_group = { + .attrs = ipl_unknown_attrs, +}; + +static decl_subsys(ipl, NULL, NULL); + +/* + * reipl section + */ + +/* FCP reipl device attributes */ + +DEFINE_IPL_ATTR_RW(reipl_fcp, wwpn, "0x%016llx\n", "%016llx\n", + reipl_block_fcp->ipl_info.fcp.wwpn); +DEFINE_IPL_ATTR_RW(reipl_fcp, lun, "0x%016llx\n", "%016llx\n", + reipl_block_fcp->ipl_info.fcp.lun); +DEFINE_IPL_ATTR_RW(reipl_fcp, bootprog, "%lld\n", "%lld\n", + reipl_block_fcp->ipl_info.fcp.bootprog); +DEFINE_IPL_ATTR_RW(reipl_fcp, br_lba, "%lld\n", "%lld\n", + reipl_block_fcp->ipl_info.fcp.br_lba); +DEFINE_IPL_ATTR_RW(reipl_fcp, device, "0.0.%04llx\n", "0.0.%llx\n", + reipl_block_fcp->ipl_info.fcp.devno); + +static struct attribute *reipl_fcp_attrs[] = { + &sys_reipl_fcp_device_attr.attr, + &sys_reipl_fcp_wwpn_attr.attr, + &sys_reipl_fcp_lun_attr.attr, + &sys_reipl_fcp_bootprog_attr.attr, + &sys_reipl_fcp_br_lba_attr.attr, + NULL, +}; + +static struct attribute_group reipl_fcp_attr_group = { + .name = IPL_FCP_STR, + .attrs = reipl_fcp_attrs, +}; + +/* CCW reipl device attributes */ + +DEFINE_IPL_ATTR_RW(reipl_ccw, device, "0.0.%04llx\n", "0.0.%llx\n", + reipl_block_ccw->ipl_info.ccw.devno); + +static struct attribute *reipl_ccw_attrs[] = { + &sys_reipl_ccw_device_attr.attr, + NULL, +}; + +static struct attribute_group reipl_ccw_attr_group = { + .name = IPL_CCW_STR, + .attrs = reipl_ccw_attrs, +}; + +/* reipl type */ + +static int reipl_set_type(enum ipl_type type) +{ + if (!(reipl_capabilities & type)) + return -EINVAL; + + switch(type) { + case IPL_TYPE_CCW: + if (MACHINE_IS_VM) + reipl_method = IPL_METHOD_CCW_VM; + else + reipl_method = IPL_METHOD_CCW_CIO; + break; + case IPL_TYPE_FCP: + if (diag308_set_works) + reipl_method = IPL_METHOD_FCP_RW_DIAG; + else if (MACHINE_IS_VM) + reipl_method = IPL_METHOD_FCP_RO_VM; + else + reipl_method = IPL_METHOD_FCP_RO_DIAG; + break; + default: + reipl_method = IPL_METHOD_NONE; + } + reipl_type = type; + return 0; +} + +static ssize_t reipl_type_show(struct subsystem *subsys, char *page) +{ + return sprintf(page, "%s\n", ipl_type_str(reipl_type)); +} + +static ssize_t reipl_type_store(struct subsystem *subsys, const char *buf, + size_t len) +{ + int rc = -EINVAL; + + if (strncmp(buf, IPL_CCW_STR, strlen(IPL_CCW_STR)) == 0) + rc = reipl_set_type(IPL_TYPE_CCW); + else if (strncmp(buf, IPL_FCP_STR, strlen(IPL_FCP_STR)) == 0) + rc = reipl_set_type(IPL_TYPE_FCP); + return (rc != 0) ? rc : len; +} + +static struct subsys_attribute reipl_type_attr = + __ATTR(reipl_type, 0644, reipl_type_show, reipl_type_store); + +static decl_subsys(reipl, NULL, NULL); + +/* + * dump section + */ + +/* FCP dump device attributes */ + +DEFINE_IPL_ATTR_RW(dump_fcp, wwpn, "0x%016llx\n", "%016llx\n", + dump_block_fcp->ipl_info.fcp.wwpn); +DEFINE_IPL_ATTR_RW(dump_fcp, lun, "0x%016llx\n", "%016llx\n", + dump_block_fcp->ipl_info.fcp.lun); +DEFINE_IPL_ATTR_RW(dump_fcp, bootprog, "%lld\n", "%lld\n", + dump_block_fcp->ipl_info.fcp.bootprog); +DEFINE_IPL_ATTR_RW(dump_fcp, br_lba, "%lld\n", "%lld\n", + dump_block_fcp->ipl_info.fcp.br_lba); +DEFINE_IPL_ATTR_RW(dump_fcp, device, "0.0.%04llx\n", "0.0.%llx\n", + dump_block_fcp->ipl_info.fcp.devno); + +static struct attribute *dump_fcp_attrs[] = { + &sys_dump_fcp_device_attr.attr, + &sys_dump_fcp_wwpn_attr.attr, + &sys_dump_fcp_lun_attr.attr, + &sys_dump_fcp_bootprog_attr.attr, + &sys_dump_fcp_br_lba_attr.attr, + NULL, +}; + +static struct attribute_group dump_fcp_attr_group = { + .name = IPL_FCP_STR, + .attrs = dump_fcp_attrs, +}; + +/* CCW dump device attributes */ + +DEFINE_IPL_ATTR_RW(dump_ccw, device, "0.0.%04llx\n", "0.0.%llx\n", + dump_block_ccw->ipl_info.ccw.devno); + +static struct attribute *dump_ccw_attrs[] = { + &sys_dump_ccw_device_attr.attr, + NULL, +}; + +static struct attribute_group dump_ccw_attr_group = { + .name = IPL_CCW_STR, + .attrs = dump_ccw_attrs, +}; + +/* dump type */ + +static int dump_set_type(enum ipl_type type) +{ + if (!(dump_capabilities & type)) + return -EINVAL; + switch(type) { + case IPL_TYPE_CCW: + if (MACHINE_IS_VM) + dump_method = IPL_METHOD_CCW_VM; + else + dump_method = IPL_METHOD_CCW_CIO; + break; + case IPL_TYPE_FCP: + dump_method = IPL_METHOD_FCP_RW_DIAG; + break; + default: + dump_method = IPL_METHOD_NONE; + } + dump_type = type; + return 0; +} + +static ssize_t dump_type_show(struct subsystem *subsys, char *page) +{ + return sprintf(page, "%s\n", ipl_type_str(dump_type)); +} + +static ssize_t dump_type_store(struct subsystem *subsys, const char *buf, + size_t len) +{ + int rc = -EINVAL; + + if (strncmp(buf, IPL_NONE_STR, strlen(IPL_NONE_STR)) == 0) + rc = dump_set_type(IPL_TYPE_NONE); + else if (strncmp(buf, IPL_CCW_STR, strlen(IPL_CCW_STR)) == 0) + rc = dump_set_type(IPL_TYPE_CCW); + else if (strncmp(buf, IPL_FCP_STR, strlen(IPL_FCP_STR)) == 0) + rc = dump_set_type(IPL_TYPE_FCP); + return (rc != 0) ? rc : len; +} + +static struct subsys_attribute dump_type_attr = + __ATTR(dump_type, 0644, dump_type_show, dump_type_store); + +static decl_subsys(dump, NULL, NULL); + +#ifdef CONFIG_SMP +static void dump_smp_stop_all(void) +{ + int cpu; + preempt_disable(); + for_each_online_cpu(cpu) { + if (cpu == smp_processor_id()) + continue; + while (signal_processor(cpu, sigp_stop) == sigp_busy) + udelay(10); + } + preempt_enable(); +} +#else +#define dump_smp_stop_all() do { } while (0) +#endif + +/* + * Shutdown actions section + */ + +static decl_subsys(shutdown_actions, NULL, NULL); + +/* on panic */ + +static ssize_t on_panic_show(struct subsystem *subsys, char *page) +{ + return sprintf(page, "%s\n", shutdown_action_str(on_panic_action)); +} + +static ssize_t on_panic_store(struct subsystem *subsys, const char *buf, + size_t len) +{ + if (strncmp(buf, SHUTDOWN_REIPL_STR, strlen(SHUTDOWN_REIPL_STR)) == 0) + on_panic_action = SHUTDOWN_REIPL; + else if (strncmp(buf, SHUTDOWN_DUMP_STR, + strlen(SHUTDOWN_DUMP_STR)) == 0) + on_panic_action = SHUTDOWN_DUMP; + else if (strncmp(buf, SHUTDOWN_STOP_STR, + strlen(SHUTDOWN_STOP_STR)) == 0) + on_panic_action = SHUTDOWN_STOP; + else + return -EINVAL; + + return len; +} + +static struct subsys_attribute on_panic_attr = + __ATTR(on_panic, 0644, on_panic_show, on_panic_store); + +static void print_fcp_block(struct ipl_parameter_block *fcp_block) +{ + printk(KERN_EMERG "wwpn: %016llx\n", + (unsigned long long)fcp_block->ipl_info.fcp.wwpn); + printk(KERN_EMERG "lun: %016llx\n", + (unsigned long long)fcp_block->ipl_info.fcp.lun); + printk(KERN_EMERG "bootprog: %lld\n", + (unsigned long long)fcp_block->ipl_info.fcp.bootprog); + printk(KERN_EMERG "br_lba: %lld\n", + (unsigned long long)fcp_block->ipl_info.fcp.br_lba); + printk(KERN_EMERG "device: %llx\n", + (unsigned long long)fcp_block->ipl_info.fcp.devno); + printk(KERN_EMERG "opt: %x\n", fcp_block->ipl_info.fcp.opt); +} + +void do_reipl(void) +{ + struct ccw_dev_id devid; + static char buf[100]; + + switch (reipl_type) { + case IPL_TYPE_CCW: + printk(KERN_EMERG "reboot on ccw device: 0.0.%04x\n", + reipl_block_ccw->ipl_info.ccw.devno); + break; + case IPL_TYPE_FCP: + printk(KERN_EMERG "reboot on fcp device:\n"); + print_fcp_block(reipl_block_fcp); + break; + default: + break; + } + + switch (reipl_method) { + case IPL_METHOD_CCW_CIO: + devid.devno = reipl_block_ccw->ipl_info.ccw.devno; + devid.ssid = 0; + reipl_ccw_dev(&devid); + break; + case IPL_METHOD_CCW_VM: + sprintf(buf, "IPL %X", reipl_block_ccw->ipl_info.ccw.devno); + cpcmd(buf, NULL, 0, NULL); + break; + case IPL_METHOD_CCW_DIAG: + diag308(DIAG308_SET, reipl_block_ccw); + diag308(DIAG308_IPL, NULL); + break; + case IPL_METHOD_FCP_RW_DIAG: + diag308(DIAG308_SET, reipl_block_fcp); + diag308(DIAG308_IPL, NULL); + break; + case IPL_METHOD_FCP_RO_DIAG: + diag308(DIAG308_IPL, NULL); + break; + case IPL_METHOD_FCP_RO_VM: + cpcmd("IPL", NULL, 0, NULL); + break; + case IPL_METHOD_NONE: + default: + if (MACHINE_IS_VM) + cpcmd("IPL", NULL, 0, NULL); + diag308(DIAG308_IPL, NULL); + break; + } + panic("reipl failed!\n"); +} + +static void do_dump(void) +{ + struct ccw_dev_id devid; + static char buf[100]; + + switch (dump_type) { + case IPL_TYPE_CCW: + printk(KERN_EMERG "Automatic dump on ccw device: 0.0.%04x\n", + dump_block_ccw->ipl_info.ccw.devno); + break; + case IPL_TYPE_FCP: + printk(KERN_EMERG "Automatic dump on fcp device:\n"); + print_fcp_block(dump_block_fcp); + break; + default: + return; + } + + switch (dump_method) { + case IPL_METHOD_CCW_CIO: + dump_smp_stop_all(); + devid.devno = dump_block_ccw->ipl_info.ccw.devno; + devid.ssid = 0; + reipl_ccw_dev(&devid); + break; + case IPL_METHOD_CCW_VM: + dump_smp_stop_all(); + sprintf(buf, "STORE STATUS"); + cpcmd(buf, NULL, 0, NULL); + sprintf(buf, "IPL %X", dump_block_ccw->ipl_info.ccw.devno); + cpcmd(buf, NULL, 0, NULL); + break; + case IPL_METHOD_CCW_DIAG: + diag308(DIAG308_SET, dump_block_ccw); + diag308(DIAG308_DUMP, NULL); + break; + case IPL_METHOD_FCP_RW_DIAG: + diag308(DIAG308_SET, dump_block_fcp); + diag308(DIAG308_DUMP, NULL); + break; + case IPL_METHOD_NONE: + default: + return; + } + printk(KERN_EMERG "Dump failed!\n"); +} + +/* init functions */ + +static int __init ipl_register_fcp_files(void) +{ + int rc; + + rc = sysfs_create_group(&ipl_subsys.kset.kobj, + &ipl_fcp_attr_group); + if (rc) + goto out; + rc = sysfs_create_bin_file(&ipl_subsys.kset.kobj, + &ipl_parameter_attr); + if (rc) + goto out_ipl_parm; + rc = sysfs_create_bin_file(&ipl_subsys.kset.kobj, + &ipl_scp_data_attr); + if (!rc) + goto out; + + sysfs_remove_bin_file(&ipl_subsys.kset.kobj, &ipl_parameter_attr); + +out_ipl_parm: + sysfs_remove_group(&ipl_subsys.kset.kobj, &ipl_fcp_attr_group); +out: + return rc; +} + +static int __init ipl_init(void) +{ + int rc; + + rc = firmware_register(&ipl_subsys); + if (rc) + return rc; + switch (ipl_get_type()) { + case IPL_TYPE_CCW: + rc = sysfs_create_group(&ipl_subsys.kset.kobj, + &ipl_ccw_attr_group); + break; + case IPL_TYPE_FCP: + rc = ipl_register_fcp_files(); + break; + default: + rc = sysfs_create_group(&ipl_subsys.kset.kobj, + &ipl_unknown_attr_group); + break; + } + if (rc) + firmware_unregister(&ipl_subsys); + return rc; +} + +static void __init reipl_probe(void) +{ + void *buffer; + + buffer = (void *) get_zeroed_page(GFP_KERNEL); + if (!buffer) + return; + if (diag308(DIAG308_STORE, buffer) == DIAG308_RC_OK) + diag308_set_works = 1; + free_page((unsigned long)buffer); +} + +static int __init reipl_ccw_init(void) +{ + int rc; + + reipl_block_ccw = (void *) get_zeroed_page(GFP_KERNEL); + if (!reipl_block_ccw) + return -ENOMEM; + rc = sysfs_create_group(&reipl_subsys.kset.kobj, &reipl_ccw_attr_group); + if (rc) { + free_page((unsigned long)reipl_block_ccw); + return rc; + } + reipl_block_ccw->hdr.len = IPL_PARM_BLK_CCW_LEN; + reipl_block_ccw->hdr.version = IPL_PARM_BLOCK_VERSION; + reipl_block_ccw->hdr.blk0_len = sizeof(reipl_block_ccw->ipl_info.ccw); + reipl_block_ccw->hdr.pbt = DIAG308_IPL_TYPE_CCW; + if (ipl_get_type() == IPL_TYPE_CCW) + reipl_block_ccw->ipl_info.ccw.devno = ipl_devno; + reipl_capabilities |= IPL_TYPE_CCW; + return 0; +} + +static int __init reipl_fcp_init(void) +{ + int rc; + + if ((!diag308_set_works) && (ipl_get_type() != IPL_TYPE_FCP)) + return 0; + if ((!diag308_set_works) && (ipl_get_type() == IPL_TYPE_FCP)) + make_attrs_ro(reipl_fcp_attrs); + + reipl_block_fcp = (void *) get_zeroed_page(GFP_KERNEL); + if (!reipl_block_fcp) + return -ENOMEM; + rc = sysfs_create_group(&reipl_subsys.kset.kobj, &reipl_fcp_attr_group); + if (rc) { + free_page((unsigned long)reipl_block_fcp); + return rc; + } + if (ipl_get_type() == IPL_TYPE_FCP) { + memcpy(reipl_block_fcp, IPL_PARMBLOCK_START, PAGE_SIZE); + } else { + reipl_block_fcp->hdr.len = IPL_PARM_BLK_FCP_LEN; + reipl_block_fcp->hdr.version = IPL_PARM_BLOCK_VERSION; + reipl_block_fcp->hdr.blk0_len = + sizeof(reipl_block_fcp->ipl_info.fcp); + reipl_block_fcp->hdr.pbt = DIAG308_IPL_TYPE_FCP; + reipl_block_fcp->ipl_info.fcp.opt = DIAG308_IPL_OPT_IPL; + } + reipl_capabilities |= IPL_TYPE_FCP; + return 0; +} + +static int __init reipl_init(void) +{ + int rc; + + rc = firmware_register(&reipl_subsys); + if (rc) + return rc; + rc = subsys_create_file(&reipl_subsys, &reipl_type_attr); + if (rc) { + firmware_unregister(&reipl_subsys); + return rc; + } + rc = reipl_ccw_init(); + if (rc) + return rc; + rc = reipl_fcp_init(); + if (rc) + return rc; + rc = reipl_set_type(ipl_get_type()); + if (rc) + return rc; + return 0; +} + +static int __init dump_ccw_init(void) +{ + int rc; + + dump_block_ccw = (void *) get_zeroed_page(GFP_KERNEL); + if (!dump_block_ccw) + return -ENOMEM; + rc = sysfs_create_group(&dump_subsys.kset.kobj, &dump_ccw_attr_group); + if (rc) { + free_page((unsigned long)dump_block_ccw); + return rc; + } + dump_block_ccw->hdr.len = IPL_PARM_BLK_CCW_LEN; + dump_block_ccw->hdr.version = IPL_PARM_BLOCK_VERSION; + dump_block_ccw->hdr.blk0_len = sizeof(reipl_block_ccw->ipl_info.ccw); + dump_block_ccw->hdr.pbt = DIAG308_IPL_TYPE_CCW; + dump_capabilities |= IPL_TYPE_CCW; + return 0; +} + +extern char s390_readinfo_sccb[]; + +static int __init dump_fcp_init(void) +{ + int rc; + + if(!(s390_readinfo_sccb[91] & 0x2)) + return 0; /* LDIPL DUMP is not installed */ + if (!diag308_set_works) + return 0; + dump_block_fcp = (void *) get_zeroed_page(GFP_KERNEL); + if (!dump_block_fcp) + return -ENOMEM; + rc = sysfs_create_group(&dump_subsys.kset.kobj, &dump_fcp_attr_group); + if (rc) { + free_page((unsigned long)dump_block_fcp); + return rc; + } + dump_block_fcp->hdr.len = IPL_PARM_BLK_FCP_LEN; + dump_block_fcp->hdr.version = IPL_PARM_BLOCK_VERSION; + dump_block_fcp->hdr.blk0_len = sizeof(dump_block_fcp->ipl_info.fcp); + dump_block_fcp->hdr.pbt = DIAG308_IPL_TYPE_FCP; + dump_block_fcp->ipl_info.fcp.opt = DIAG308_IPL_OPT_DUMP; + dump_capabilities |= IPL_TYPE_FCP; + return 0; +} + +#define SHUTDOWN_ON_PANIC_PRIO 0 + +static int shutdown_on_panic_notify(struct notifier_block *self, + unsigned long event, void *data) +{ + if (on_panic_action == SHUTDOWN_DUMP) + do_dump(); + else if (on_panic_action == SHUTDOWN_REIPL) + do_reipl(); + return NOTIFY_OK; +} + +static struct notifier_block shutdown_on_panic_nb = { + .notifier_call = shutdown_on_panic_notify, + .priority = SHUTDOWN_ON_PANIC_PRIO +}; + +static int __init dump_init(void) +{ + int rc; + + rc = firmware_register(&dump_subsys); + if (rc) + return rc; + rc = subsys_create_file(&dump_subsys, &dump_type_attr); + if (rc) { + firmware_unregister(&dump_subsys); + return rc; + } + rc = dump_ccw_init(); + if (rc) + return rc; + rc = dump_fcp_init(); + if (rc) + return rc; + dump_set_type(IPL_TYPE_NONE); + return 0; +} + +static int __init shutdown_actions_init(void) +{ + int rc; + + rc = firmware_register(&shutdown_actions_subsys); + if (rc) + return rc; + rc = subsys_create_file(&shutdown_actions_subsys, &on_panic_attr); + if (rc) { + firmware_unregister(&shutdown_actions_subsys); + return rc; + } + atomic_notifier_chain_register(&panic_notifier_list, + &shutdown_on_panic_nb); + return 0; +} + +static int __init s390_ipl_init(void) +{ + int rc; + + reipl_probe(); + rc = ipl_init(); + if (rc) + return rc; + rc = reipl_init(); + if (rc) + return rc; + rc = dump_init(); + if (rc) + return rc; + rc = shutdown_actions_init(); + if (rc) + return rc; + return 0; +} + +__initcall(s390_ipl_init); diff --git a/arch/s390/kernel/reipl.S b/arch/s390/kernel/reipl.S index 658e5ac484f..4562cdbce8e 100644 --- a/arch/s390/kernel/reipl.S +++ b/arch/s390/kernel/reipl.S @@ -8,13 +8,30 @@ #include - .globl do_reipl -do_reipl: basr %r13,0 + .globl do_reipl_asm +do_reipl_asm: basr %r13,0 .Lpg0: lpsw .Lnewpsw-.Lpg0(%r13) -.Lpg1: lctl %c6,%c6,.Lall-.Lpg0(%r13) - stctl %c0,%c0,.Lctlsave-.Lpg0(%r13) - ni .Lctlsave-.Lpg0(%r13),0xef - lctl %c0,%c0,.Lctlsave-.Lpg0(%r13) + + # switch off lowcore protection + +.Lpg1: stctl %c0,%c0,.Lctlsave1-.Lpg0(%r13) + stctl %c0,%c0,.Lctlsave2-.Lpg0(%r13) + ni .Lctlsave1-.Lpg0(%r13),0xef + lctl %c0,%c0,.Lctlsave1-.Lpg0(%r13) + + # do store status of all registers + + stm %r0,%r15,__LC_GPREGS_SAVE_AREA + stctl %c0,%c15,__LC_CREGS_SAVE_AREA + mvc __LC_CREGS_SAVE_AREA(4),.Lctlsave2-.Lpg0(%r13) + stam %a0,%a15,__LC_AREGS_SAVE_AREA + stpx __LC_PREFIX_SAVE_AREA + stckc .Lclkcmp-.Lpg0(%r13) + mvc __LC_CLOCK_COMP_SAVE_AREA(8),.Lclkcmp-.Lpg0(%r13) + stpt __LC_CPU_TIMER_SAVE_AREA + st %r13, __LC_PSW_SAVE_AREA+4 + + lctl %c6,%c6,.Lall-.Lpg0(%r13) lr %r1,%r2 mvc __LC_PGM_NEW_PSW(8),.Lpcnew-.Lpg0(%r13) stsch .Lschib-.Lpg0(%r13) @@ -46,9 +63,11 @@ do_reipl: basr %r13,0 .Ldisab: st %r14,.Ldispsw+4-.Lpg0(%r13) lpsw .Ldispsw-.Lpg0(%r13) .align 8 +.Lclkcmp: .quad 0x0000000000000000 .Lall: .long 0xff000000 .Lnull: .long 0x00000000 -.Lctlsave: .long 0x00000000 +.Lctlsave1: .long 0x00000000 +.Lctlsave2: .long 0x00000000 .align 8 .Lnewpsw: .long 0x00080000,0x80000000+.Lpg1 .Lpcnew: .long 0x00080000,0x80000000+.Lecs diff --git a/arch/s390/kernel/reipl64.S b/arch/s390/kernel/reipl64.S index 4d090d60f3e..95bd1e234f6 100644 --- a/arch/s390/kernel/reipl64.S +++ b/arch/s390/kernel/reipl64.S @@ -8,13 +8,30 @@ */ #include - .globl do_reipl -do_reipl: basr %r13,0 -.Lpg0: lpswe .Lnewpsw-.Lpg0(%r13) + .globl do_reipl_asm +do_reipl_asm: basr %r13,0 + + # do store status of all registers + +.Lpg0: stg %r1,.Lregsave-.Lpg0(%r13) + lghi %r1,0x1000 + stmg %r0,%r15,__LC_GPREGS_SAVE_AREA-0x1000(%r1) + lg %r0,.Lregsave-.Lpg0(%r13) + stg %r0,__LC_GPREGS_SAVE_AREA-0x1000+8(%r1) + stctg %c0,%c15,__LC_CREGS_SAVE_AREA-0x1000(%r1) + stam %a0,%a15,__LC_AREGS_SAVE_AREA-0x1000(%r1) + stpx __LC_PREFIX_SAVE_AREA-0x1000(%r1) + stfpc __LC_FP_CREG_SAVE_AREA-0x1000(%r1) + stckc .Lclkcmp-.Lpg0(%r13) + mvc __LC_CLOCK_COMP_SAVE_AREA-0x1000(8,%r1),.Lclkcmp-.Lpg0(%r13) + stpt __LC_CPU_TIMER_SAVE_AREA-0x1000(%r1) + stg %r13, __LC_PSW_SAVE_AREA-0x1000+8(%r1) + + lpswe .Lnewpsw-.Lpg0(%r13) .Lpg1: lctlg %c6,%c6,.Lall-.Lpg0(%r13) - stctg %c0,%c0,.Lctlsave-.Lpg0(%r13) - ni .Lctlsave+4-.Lpg0(%r13),0xef - lctlg %c0,%c0,.Lctlsave-.Lpg0(%r13) + stctg %c0,%c0,.Lregsave-.Lpg0(%r13) + ni .Lregsave+4-.Lpg0(%r13),0xef + lctlg %c0,%c0,.Lregsave-.Lpg0(%r13) lgr %r1,%r2 mvc __LC_PGM_NEW_PSW(16),.Lpcnew-.Lpg0(%r13) stsch .Lschib-.Lpg0(%r13) @@ -50,8 +67,9 @@ do_reipl: basr %r13,0 st %r14,.Ldispsw+12-.Lpg0(%r13) lpswe .Ldispsw-.Lpg0(%r13) .align 8 +.Lclkcmp: .quad 0x0000000000000000 .Lall: .quad 0x00000000ff000000 -.Lctlsave: .quad 0x0000000000000000 +.Lregsave: .quad 0x0000000000000000 .Lnull: .long 0x0000000000000000 .align 16 /* @@ -92,5 +110,3 @@ do_reipl: basr %r13,0 .long 0x00000000,0x00000000 .long 0x00000000,0x00000000 - - diff --git a/arch/s390/kernel/reipl_diag.c b/arch/s390/kernel/reipl_diag.c deleted file mode 100644 index 1f33951ba43..00000000000 --- a/arch/s390/kernel/reipl_diag.c +++ /dev/null @@ -1,39 +0,0 @@ -/* - * This file contains the implementation of the - * Linux re-IPL support - * - * (C) Copyright IBM Corp. 2005 - * - * Author(s): Volker Sameske (sameske@de.ibm.com) - * - */ - -#include - -static unsigned int reipl_diag_rc1; -static unsigned int reipl_diag_rc2; - -/* - * re-IPL the system using the last used IPL parameters - */ -void reipl_diag(void) -{ - asm volatile ( - " la %%r4,0\n" - " la %%r5,0\n" - " diag %%r4,%2,0x308\n" - "0:\n" - " st %%r4,%0\n" - " st %%r5,%1\n" - ".section __ex_table,\"a\"\n" -#ifdef CONFIG_64BIT - " .align 8\n" - " .quad 0b, 0b\n" -#else - " .align 4\n" - " .long 0b, 0b\n" -#endif - ".previous\n" - : "=m" (reipl_diag_rc1), "=m" (reipl_diag_rc2) - : "d" (3) : "cc", "4", "5" ); -} diff --git a/arch/s390/kernel/setup.c b/arch/s390/kernel/setup.c index 89051e8a5d8..f2a9165ca4f 100644 --- a/arch/s390/kernel/setup.c +++ b/arch/s390/kernel/setup.c @@ -285,16 +285,9 @@ void (*_machine_power_off)(void) = machine_power_off_smp; /* * Reboot, halt and power_off routines for non SMP. */ -extern void reipl(unsigned long devno); -extern void reipl_diag(void); static void do_machine_restart_nonsmp(char * __unused) { - reipl_diag(); - - if (MACHINE_IS_VM) - cpcmd ("IPL", NULL, 0, NULL); - else - reipl (0x10000 | S390_lowcore.ipl_device); + do_reipl(); } static void do_machine_halt_nonsmp(void) @@ -755,214 +748,3 @@ struct seq_operations cpuinfo_op = { .show = show_cpuinfo, }; -#define DEFINE_IPL_ATTR(_name, _format, _value) \ -static ssize_t ipl_##_name##_show(struct subsystem *subsys, \ - char *page) \ -{ \ - return sprintf(page, _format, _value); \ -} \ -static struct subsys_attribute ipl_##_name##_attr = \ - __ATTR(_name, S_IRUGO, ipl_##_name##_show, NULL); - -DEFINE_IPL_ATTR(wwpn, "0x%016llx\n", (unsigned long long) - IPL_PARMBLOCK_START->fcp.wwpn); -DEFINE_IPL_ATTR(lun, "0x%016llx\n", (unsigned long long) - IPL_PARMBLOCK_START->fcp.lun); -DEFINE_IPL_ATTR(bootprog, "%lld\n", (unsigned long long) - IPL_PARMBLOCK_START->fcp.bootprog); -DEFINE_IPL_ATTR(br_lba, "%lld\n", (unsigned long long) - IPL_PARMBLOCK_START->fcp.br_lba); - -enum ipl_type_type { - ipl_type_unknown, - ipl_type_ccw, - ipl_type_fcp, -}; - -static enum ipl_type_type -get_ipl_type(void) -{ - struct ipl_parameter_block *ipl = IPL_PARMBLOCK_START; - - if (!IPL_DEVNO_VALID) - return ipl_type_unknown; - if (!IPL_PARMBLOCK_VALID) - return ipl_type_ccw; - if (ipl->hdr.header.version > IPL_MAX_SUPPORTED_VERSION) - return ipl_type_unknown; - if (ipl->fcp.pbt != IPL_TYPE_FCP) - return ipl_type_unknown; - return ipl_type_fcp; -} - -static ssize_t -ipl_type_show(struct subsystem *subsys, char *page) -{ - switch (get_ipl_type()) { - case ipl_type_ccw: - return sprintf(page, "ccw\n"); - case ipl_type_fcp: - return sprintf(page, "fcp\n"); - default: - return sprintf(page, "unknown\n"); - } -} - -static struct subsys_attribute ipl_type_attr = __ATTR_RO(ipl_type); - -static ssize_t -ipl_device_show(struct subsystem *subsys, char *page) -{ - struct ipl_parameter_block *ipl = IPL_PARMBLOCK_START; - - switch (get_ipl_type()) { - case ipl_type_ccw: - return sprintf(page, "0.0.%04x\n", ipl_devno); - case ipl_type_fcp: - return sprintf(page, "0.0.%04x\n", ipl->fcp.devno); - default: - return 0; - } -} - -static struct subsys_attribute ipl_device_attr = - __ATTR(device, S_IRUGO, ipl_device_show, NULL); - -static struct attribute *ipl_fcp_attrs[] = { - &ipl_type_attr.attr, - &ipl_device_attr.attr, - &ipl_wwpn_attr.attr, - &ipl_lun_attr.attr, - &ipl_bootprog_attr.attr, - &ipl_br_lba_attr.attr, - NULL, -}; - -static struct attribute_group ipl_fcp_attr_group = { - .attrs = ipl_fcp_attrs, -}; - -static struct attribute *ipl_ccw_attrs[] = { - &ipl_type_attr.attr, - &ipl_device_attr.attr, - NULL, -}; - -static struct attribute_group ipl_ccw_attr_group = { - .attrs = ipl_ccw_attrs, -}; - -static struct attribute *ipl_unknown_attrs[] = { - &ipl_type_attr.attr, - NULL, -}; - -static struct attribute_group ipl_unknown_attr_group = { - .attrs = ipl_unknown_attrs, -}; - -static ssize_t -ipl_parameter_read(struct kobject *kobj, char *buf, loff_t off, size_t count) -{ - unsigned int size = IPL_PARMBLOCK_SIZE; - - if (off > size) - return 0; - if (off + count > size) - count = size - off; - - memcpy(buf, (void *) IPL_PARMBLOCK_START + off, count); - return count; -} - -static struct bin_attribute ipl_parameter_attr = { - .attr = { - .name = "binary_parameter", - .mode = S_IRUGO, - .owner = THIS_MODULE, - }, - .size = PAGE_SIZE, - .read = &ipl_parameter_read, -}; - -static ssize_t -ipl_scp_data_read(struct kobject *kobj, char *buf, loff_t off, size_t count) -{ - unsigned int size = IPL_PARMBLOCK_START->fcp.scp_data_len; - void *scp_data = &IPL_PARMBLOCK_START->fcp.scp_data; - - if (off > size) - return 0; - if (off + count > size) - count = size - off; - - memcpy(buf, scp_data + off, count); - return count; -} - -static struct bin_attribute ipl_scp_data_attr = { - .attr = { - .name = "scp_data", - .mode = S_IRUGO, - .owner = THIS_MODULE, - }, - .size = PAGE_SIZE, - .read = &ipl_scp_data_read, -}; - -static decl_subsys(ipl, NULL, NULL); - -static int ipl_register_fcp_files(void) -{ - int rc; - - rc = sysfs_create_group(&ipl_subsys.kset.kobj, - &ipl_fcp_attr_group); - if (rc) - goto out; - rc = sysfs_create_bin_file(&ipl_subsys.kset.kobj, - &ipl_parameter_attr); - if (rc) - goto out_ipl_parm; - rc = sysfs_create_bin_file(&ipl_subsys.kset.kobj, - &ipl_scp_data_attr); - if (!rc) - goto out; - - sysfs_remove_bin_file(&ipl_subsys.kset.kobj, &ipl_parameter_attr); - -out_ipl_parm: - sysfs_remove_group(&ipl_subsys.kset.kobj, &ipl_fcp_attr_group); -out: - return rc; -} - -static int __init -ipl_device_sysfs_register(void) { - int rc; - - rc = firmware_register(&ipl_subsys); - if (rc) - goto out; - - switch (get_ipl_type()) { - case ipl_type_ccw: - rc = sysfs_create_group(&ipl_subsys.kset.kobj, - &ipl_ccw_attr_group); - break; - case ipl_type_fcp: - rc = ipl_register_fcp_files(); - break; - default: - rc = sysfs_create_group(&ipl_subsys.kset.kobj, - &ipl_unknown_attr_group); - break; - } - - if (rc) - firmware_unregister(&ipl_subsys); -out: - return rc; -} - -__initcall(ipl_device_sysfs_register); diff --git a/arch/s390/kernel/smp.c b/arch/s390/kernel/smp.c index 8e03219eea7..b2e6f4c8d38 100644 --- a/arch/s390/kernel/smp.c +++ b/arch/s390/kernel/smp.c @@ -59,9 +59,6 @@ static struct task_struct *current_set[NR_CPUS]; extern char vmhalt_cmd[]; extern char vmpoff_cmd[]; -extern void reipl(unsigned long devno); -extern void reipl_diag(void); - static void smp_ext_bitcall(int, ec_bit_sig); static void smp_ext_bitcall_others(ec_bit_sig); @@ -279,12 +276,7 @@ static void do_machine_restart(void * __unused) * interrupted by an external interrupt and s390irq * locks are always held disabled). */ - reipl_diag(); - - if (MACHINE_IS_VM) - cpcmd ("IPL", NULL, 0, NULL); - else - reipl (0x10000 | S390_lowcore.ipl_device); + do_reipl(); } void machine_restart_smp(char * __unused) -- cgit v1.2.3 From 8427082a506f7ae0abf82ce0047a045ec4309e59 Mon Sep 17 00:00:00 2001 From: Heiko Carstens Date: Wed, 20 Sep 2006 15:58:54 +0200 Subject: [S390] fix syscall restart handling. If do_signal() gets called several times before returning to user space and no signal is pending (e.g. cancelled by a debugger) syscall restart handling could be done several times. This would change the user space PSW to an address prior to the syscall instruction. Fix this by making sure that syscall restart handling is only done once. Signed-off-by: Heiko Carstens Signed-off-by: Martin Schwidefsky --- arch/s390/kernel/signal.c | 1 + 1 file changed, 1 insertion(+) (limited to 'arch') diff --git a/arch/s390/kernel/signal.c b/arch/s390/kernel/signal.c index a887b686f27..dd05423f87a 100644 --- a/arch/s390/kernel/signal.c +++ b/arch/s390/kernel/signal.c @@ -457,6 +457,7 @@ void do_signal(struct pt_regs *regs) case -ERESTART_RESTARTBLOCK: regs->gprs[2] = -EINTR; } + regs->trap = -1; /* Don't deal with this again. */ } /* Get signal to deliver. When running under ptrace, at this point -- cgit v1.2.3 From 7d5d688f724dd5a651d1ce7bc3ea7c03d28137a1 Mon Sep 17 00:00:00 2001 From: Heiko Carstens Date: Wed, 20 Sep 2006 15:59:00 +0200 Subject: [S390] Use simple_strtoul instead of own cmm_strtoul wrapper. Fix compile warning with some configurations: arch/s390/mm/cmm.c:58: warning: 'cmm_strtoul' defined but not used Originally cmm_strtoul was introduced because simple_strtoul couldn't handle strings with hexadecimal numbers that contained a capital 'X'. Since this is no longer true cmm_strtoul can be removed. Signed-off-by: Heiko Carstens Signed-off-by: Martin Schwidefsky --- arch/s390/mm/cmm.c | 30 +++++++----------------------- 1 file changed, 7 insertions(+), 23 deletions(-) (limited to 'arch') diff --git a/arch/s390/mm/cmm.c b/arch/s390/mm/cmm.c index ceea51cff03..786a44dba5b 100644 --- a/arch/s390/mm/cmm.c +++ b/arch/s390/mm/cmm.c @@ -52,22 +52,6 @@ static struct timer_list cmm_timer; static void cmm_timer_fn(unsigned long); static void cmm_set_timer(void); -static long -cmm_strtoul(const char *cp, char **endp) -{ - unsigned int base = 10; - - if (*cp == '0') { - base = 8; - cp++; - if ((*cp == 'x' || *cp == 'X') && isxdigit(cp[1])) { - base = 16; - cp++; - } - } - return simple_strtoul(cp, endp, base); -} - static long cmm_alloc_pages(long pages, long *counter, struct cmm_page_array **list) { @@ -276,7 +260,7 @@ cmm_pages_handler(ctl_table *ctl, int write, struct file *filp, return -EFAULT; buf[sizeof(buf) - 1] = '\0'; cmm_skip_blanks(buf, &p); - pages = cmm_strtoul(p, &p); + pages = simple_strtoul(p, &p, 0); if (ctl == &cmm_table[0]) cmm_set_pages(pages); else @@ -317,9 +301,9 @@ cmm_timeout_handler(ctl_table *ctl, int write, struct file *filp, return -EFAULT; buf[sizeof(buf) - 1] = '\0'; cmm_skip_blanks(buf, &p); - pages = cmm_strtoul(p, &p); + pages = simple_strtoul(p, &p, 0); cmm_skip_blanks(p, &p); - seconds = cmm_strtoul(p, &p); + seconds = simple_strtoul(p, &p, 0); cmm_set_timeout(pages, seconds); } else { len = sprintf(buf, "%ld %ld\n", @@ -382,24 +366,24 @@ cmm_smsg_target(char *from, char *msg) if (strncmp(msg, "SHRINK", 6) == 0) { if (!cmm_skip_blanks(msg + 6, &msg)) return; - pages = cmm_strtoul(msg, &msg); + pages = simple_strtoul(msg, &msg, 0); cmm_skip_blanks(msg, &msg); if (*msg == '\0') cmm_set_pages(pages); } else if (strncmp(msg, "RELEASE", 7) == 0) { if (!cmm_skip_blanks(msg + 7, &msg)) return; - pages = cmm_strtoul(msg, &msg); + pages = simple_strtoul(msg, &msg, 0); cmm_skip_blanks(msg, &msg); if (*msg == '\0') cmm_add_timed_pages(pages); } else if (strncmp(msg, "REUSE", 5) == 0) { if (!cmm_skip_blanks(msg + 5, &msg)) return; - pages = cmm_strtoul(msg, &msg); + pages = simple_strtoul(msg, &msg, 0); if (!cmm_skip_blanks(msg, &msg)) return; - seconds = cmm_strtoul(msg, &msg); + seconds = simple_strtoul(msg, &msg, 0); cmm_skip_blanks(msg, &msg); if (*msg == '\0') cmm_set_timeout(pages, seconds); -- cgit v1.2.3 From 1375fc1fb0434a26f93c59b1b9f3fdb8bf90bba5 Mon Sep 17 00:00:00 2001 From: Heiko Carstens Date: Wed, 20 Sep 2006 15:59:12 +0200 Subject: [S390] __exit cleanup. Signed-off-by: Heiko Carstens Signed-off-by: Martin Schwidefsky --- arch/s390/hypfs/hypfs_diag.c | 2 +- arch/s390/kernel/vmlinux.lds.S | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) (limited to 'arch') diff --git a/arch/s390/hypfs/hypfs_diag.c b/arch/s390/hypfs/hypfs_diag.c index fee5aee605f..75144efbb92 100644 --- a/arch/s390/hypfs/hypfs_diag.c +++ b/arch/s390/hypfs/hypfs_diag.c @@ -535,7 +535,7 @@ __init int hypfs_diag_init(void) return rc; } -__exit void hypfs_diag_exit(void) +void hypfs_diag_exit(void) { diag224_delete_name_table(); diag204_free_buffer(); diff --git a/arch/s390/kernel/vmlinux.lds.S b/arch/s390/kernel/vmlinux.lds.S index df0c16ab8e9..af9e69a0301 100644 --- a/arch/s390/kernel/vmlinux.lds.S +++ b/arch/s390/kernel/vmlinux.lds.S @@ -118,7 +118,7 @@ SECTIONS /* Sections to be discarded */ /DISCARD/ : { - *(.exitcall.exit) + *(.exit.text) *(.exit.data) *(.exitcall.exit) } /* Stabs debugging sections. */ -- cgit v1.2.3 From e87bfe51b5ca2db99dd680bbb1e8fe3c94b607df Mon Sep 17 00:00:00 2001 From: Heiko Carstens Date: Wed, 20 Sep 2006 15:59:15 +0200 Subject: [S390] convert some assembler to C. Convert GET_IPL_DEVICE assembler macro to C function. Signed-off-by: Heiko Carstens Signed-off-by: Martin Schwidefsky --- arch/s390/kernel/head.S | 59 ----------------------------------------------- arch/s390/kernel/head31.S | 43 +++++++++++++++++++++++----------- arch/s390/kernel/head64.S | 39 ++++++++++++++++++++----------- arch/s390/kernel/ipl.c | 4 ++-- 4 files changed, 57 insertions(+), 88 deletions(-) (limited to 'arch') diff --git a/arch/s390/kernel/head.S b/arch/s390/kernel/head.S index a6e9bdb5359..0f1db268a8a 100644 --- a/arch/s390/kernel/head.S +++ b/arch/s390/kernel/head.S @@ -481,65 +481,6 @@ start: .byte 0xf0,0xf1,0xf2,0xf3,0xf4,0xf5,0xf6,0xf7 .byte 0xf8,0xf9,0xfa,0xfb,0xfc,0xfd,0xfe,0xff -.macro GET_IPL_DEVICE -.Lget_ipl_device: - l %r1,0xb8 # get sid - sll %r1,15 # test if subchannel is enabled - srl %r1,31 - ltr %r1,%r1 - bz 2f-.LPG1(%r13) # subchannel disabled - l %r1,0xb8 - la %r5,.Lipl_schib-.LPG1(%r13) - stsch 0(%r5) # get schib of subchannel - bnz 2f-.LPG1(%r13) # schib not available - tm 5(%r5),0x01 # devno valid? - bno 2f-.LPG1(%r13) - la %r6,ipl_parameter_flags-.LPG1(%r13) - oi 3(%r6),0x01 # set flag - la %r2,ipl_devno-.LPG1(%r13) - mvc 0(2,%r2),6(%r5) # store devno - tm 4(%r5),0x80 # qdio capable device? - bno 2f-.LPG1(%r13) - oi 3(%r6),0x02 # set flag - - # copy ipl parameters - - lhi %r0,4096 - l %r2,20(%r0) # get address of parameter list - lhi %r3,IPL_PARMBLOCK_ORIGIN - st %r3,20(%r0) - lhi %r4,1 - cr %r2,%r3 # start parameters < destination ? - jl 0f - lhi %r1,1 # copy direction is upwards - j 1f -0: lhi %r1,-1 # copy direction is downwards - ar %r2,%r0 - ar %r3,%r0 - ar %r2,%r1 - ar %r3,%r1 -1: mvc 0(1,%r3),0(%r2) # finally copy ipl parameters - ar %r3,%r1 - ar %r2,%r1 - sr %r0,%r4 - jne 1b - b 2f-.LPG1(%r13) - - .align 4 -.Lipl_schib: - .rept 13 - .long 0 - .endr - - .globl ipl_parameter_flags -ipl_parameter_flags: - .long 0 - .globl ipl_devno -ipl_devno: - .word 0 -2: -.endm - #ifdef CONFIG_64BIT #include "head64.S" #else diff --git a/arch/s390/kernel/head31.S b/arch/s390/kernel/head31.S index d8bb68a7252..1fa9fa1ca74 100644 --- a/arch/s390/kernel/head31.S +++ b/arch/s390/kernel/head31.S @@ -37,13 +37,23 @@ startup:basr %r13,0 # get base startup_continue: basr %r13,0 # get base -.LPG1: GET_IPL_DEVICE - mvi __LC_AR_MODE_ID,0 # set ESA flag (mode 0) +.LPG1: mvi __LC_AR_MODE_ID,0 # set ESA flag (mode 0) lctl %c0,%c15,.Lctl-.LPG1(%r13) # load control registers l %r12,.Lparmaddr-.LPG1(%r13) # pointer to parameter area # move IPL device to lowcore mvc __LC_IPLDEV(4),IPL_DEVICE-PARMAREA(%r12) +# +# Setup stack +# + l %r15,.Linittu-.LPG1(%r13) + mvc __LC_CURRENT(4),__TI_task(%r15) + ahi %r15,1<<(PAGE_SHIFT+THREAD_ORDER) # init_task_union+THREAD_SIZE + st %r15,__LC_KERNEL_STACK # set end of kernel stack + ahi %r15,-96 + xc __SF_BACKCHAIN(4,%r15),__SF_BACKCHAIN(%r15) # clear backchain + l %r14,.Lipl_save_parameters-.LPG1(%r13) + basr %r14,%r14 # # clear bss memory # @@ -115,6 +125,10 @@ startup_continue: b .Lfchunk-.LPG1(%r13) .align 4 +.Lipl_save_parameters: + .long ipl_save_parameters +.Linittu: + .long init_thread_union .Lpmask: .byte 0 .align 8 @@ -274,6 +288,20 @@ startup_continue: .Lbss_end: .long _end .Lparmaddr: .long PARMAREA .Lsccbaddr: .long .Lsccb + + .globl ipl_schib +ipl_schib: + .rept 13 + .long 0 + .endr + + .globl ipl_flags +ipl_flags: + .long 0 + .globl ipl_devno +ipl_devno: + .word 0 + .org 0x12000 .globl s390_readinfo_sccb s390_readinfo_sccb: @@ -305,16 +333,6 @@ s390_readinfo_sccb: .globl _stext _stext: basr %r13,0 # get base .LPG3: -# -# Setup stack -# - l %r15,.Linittu-.LPG3(%r13) - mvc __LC_CURRENT(4),__TI_task(%r15) - ahi %r15,1<<(PAGE_SHIFT+THREAD_ORDER) # init_task_union+THREAD_SIZE - st %r15,__LC_KERNEL_STACK # set end of kernel stack - ahi %r15,-96 - xc __SF_BACKCHAIN(4,%r15),__SF_BACKCHAIN(%r15) # clear backchain - # check control registers stctl %c0,%c15,0(%r15) oi 2(%r15),0x40 # enable sigp emergency signal @@ -333,6 +351,5 @@ _stext: basr %r13,0 # get base # .align 8 .Ldw: .long 0x000a0000,0x00000000 -.Linittu:.long init_thread_union .Lstart:.long start_kernel .Laregs:.long 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 diff --git a/arch/s390/kernel/head64.S b/arch/s390/kernel/head64.S index c2005101fee..1ebaa338aa7 100644 --- a/arch/s390/kernel/head64.S +++ b/arch/s390/kernel/head64.S @@ -39,7 +39,6 @@ startup_continue: basr %r13,0 # get base .LPG1: sll %r13,1 # remove high order bit srl %r13,1 - GET_IPL_DEVICE lhi %r1,1 # mode 1 = esame mvi __LC_AR_MODE_ID,1 # set esame flag slr %r0,%r0 # set cpuid to zero @@ -49,7 +48,18 @@ startup_continue: lg %r12,.Lparmaddr-.LPG1(%r13)# pointer to parameter area # move IPL device to lowcore mvc __LC_IPLDEV(4),IPL_DEVICE+4-PARMAREA(%r12) +# +# Setup stack +# + larl %r15,init_thread_union + lg %r14,__TI_task(%r15) # cache current in lowcore + stg %r14,__LC_CURRENT + aghi %r15,1<<(PAGE_SHIFT+THREAD_ORDER) # init_task_union + THREAD_SIZE + stg %r15,__LC_KERNEL_STACK # set end of kernel stack + aghi %r15,-160 + xc __SF_BACKCHAIN(4,%r15),__SF_BACKCHAIN(%r15) # clear backchain + brasl %r14,ipl_save_parameters # # clear bss memory # @@ -269,6 +279,19 @@ startup_continue: .Lparmaddr: .quad PARMAREA + .globl ipl_schib +ipl_schib: + .rept 13 + .long 0 + .endr + + .globl ipl_flags +ipl_flags: + .long 0 + .globl ipl_devno +ipl_devno: + .word 0 + .org 0x12000 .globl s390_readinfo_sccb s390_readinfo_sccb: @@ -300,24 +323,12 @@ s390_readinfo_sccb: .globl _stext _stext: basr %r13,0 # get base .LPG3: -# -# Setup stack -# - larl %r15,init_thread_union - lg %r14,__TI_task(%r15) # cache current in lowcore - stg %r14,__LC_CURRENT - aghi %r15,1<<(PAGE_SHIFT+THREAD_ORDER) # init_task_union + THREAD_SIZE - stg %r15,__LC_KERNEL_STACK # set end of kernel stack - aghi %r15,-160 - xc __SF_BACKCHAIN(4,%r15),__SF_BACKCHAIN(%r15) # clear backchain - # check control registers stctg %c0,%c15,0(%r15) oi 6(%r15),0x40 # enable sigp emergency signal oi 4(%r15),0x10 # switch on low address proctection lctlg %c0,%c15,0(%r15) -# lam 0,15,.Laregs-.LPG3(%r13) # load access regs needed by uaccess brasl %r14,start_kernel # go to C code # @@ -325,7 +336,7 @@ _stext: basr %r13,0 # get base # basr %r13,0 lpswe .Ldw-.(%r13) # load disabled wait psw -# + .align 8 .Ldw: .quad 0x0002000180000000,0x0000000000000000 .Laregs: .long 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 diff --git a/arch/s390/kernel/ipl.c b/arch/s390/kernel/ipl.c index 105ee15a2b3..6555cc48e28 100644 --- a/arch/s390/kernel/ipl.c +++ b/arch/s390/kernel/ipl.c @@ -189,9 +189,9 @@ static enum ipl_type ipl_get_type(void) { struct ipl_parameter_block *ipl = IPL_PARMBLOCK_START; - if (!IPL_DEVNO_VALID) + if (!(ipl_flags & IPL_DEVNO_VALID)) return IPL_TYPE_UNKNOWN; - if (!IPL_PARMBLOCK_VALID) + if (!(ipl_flags & IPL_PARMBLOCK_VALID)) return IPL_TYPE_CCW; if (ipl->hdr.version > IPL_MAX_SUPPORTED_VERSION) return IPL_TYPE_UNKNOWN; -- cgit v1.2.3 From 9514e2311be97a01e8669c4de78e9fea37489f09 Mon Sep 17 00:00:00 2001 From: Heiko Carstens Date: Wed, 20 Sep 2006 15:59:22 +0200 Subject: [S390] Kernel stack overflow handling. Substract the size of the initial stack frame from the correct register. Otherwise we will end up in a program check loop. Fix the offset into the save area as well. Signed-off-by: Heiko Carstens Signed-off-by: Martin Schwidefsky --- arch/s390/kernel/entry64.S | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'arch') diff --git a/arch/s390/kernel/entry64.S b/arch/s390/kernel/entry64.S index 8b956d1538f..29bbfbab733 100644 --- a/arch/s390/kernel/entry64.S +++ b/arch/s390/kernel/entry64.S @@ -827,7 +827,7 @@ restart_go: */ stack_overflow: lg %r15,__LC_PANIC_STACK # change to panic stack - aghi %r1,-SP_SIZE + aghi %r15,-SP_SIZE mvc SP_PSW(16,%r15),0(%r12) # move user PSW to stack stmg %r0,%r11,SP_R0(%r15) # store gprs %r0-%r11 to kernel stack la %r1,__LC_SAVE_AREA @@ -835,7 +835,7 @@ stack_overflow: je 0f chi %r12,__LC_PGM_OLD_PSW je 0f - la %r1,__LC_SAVE_AREA+16 + la %r1,__LC_SAVE_AREA+32 0: mvc SP_R12(32,%r15),0(%r1) # move %r12-%r15 to stack xc __SF_BACKCHAIN(8,%r15),__SF_BACKCHAIN(%r15) # clear back chain la %r2,SP_PTREGS(%r15) # load pt_regs -- cgit v1.2.3 From 1f38d61347203055b55e34083cce7a9cd8c529a9 Mon Sep 17 00:00:00 2001 From: Gerald Schaefer Date: Wed, 20 Sep 2006 15:59:26 +0200 Subject: [S390] cleanup appldata. Introduce asm header that contains the appldata data structures and the diag inline assembly. Signed-off-by: Gerald Schaefer Signed-off-by: Martin Schwidefsky --- arch/s390/appldata/appldata.h | 16 -------- arch/s390/appldata/appldata_base.c | 81 ++++++-------------------------------- arch/s390/appldata/appldata_os.c | 1 + 3 files changed, 13 insertions(+), 85 deletions(-) (limited to 'arch') diff --git a/arch/s390/appldata/appldata.h b/arch/s390/appldata/appldata.h index 71d65eb3065..0429481dea6 100644 --- a/arch/s390/appldata/appldata.h +++ b/arch/s390/appldata/appldata.h @@ -29,22 +29,6 @@ #define CTL_APPLDATA_NET_SUM 2125 #define CTL_APPLDATA_PROC 2126 -#ifndef CONFIG_64BIT - -#define APPLDATA_START_INTERVAL_REC 0x00 /* Function codes for */ -#define APPLDATA_STOP_REC 0x01 /* DIAG 0xDC */ -#define APPLDATA_GEN_EVENT_RECORD 0x02 -#define APPLDATA_START_CONFIG_REC 0x03 - -#else - -#define APPLDATA_START_INTERVAL_REC 0x80 -#define APPLDATA_STOP_REC 0x81 -#define APPLDATA_GEN_EVENT_RECORD 0x82 -#define APPLDATA_START_CONFIG_REC 0x83 - -#endif /* CONFIG_64BIT */ - #define P_INFO(x...) printk(KERN_INFO MY_PRINT_NAME " info: " x) #define P_ERROR(x...) printk(KERN_ERR MY_PRINT_NAME " error: " x) #define P_WARNING(x...) printk(KERN_WARNING MY_PRINT_NAME " status: " x) diff --git a/arch/s390/appldata/appldata_base.c b/arch/s390/appldata/appldata_base.c index a0a94e0ef8d..b69ed742f98 100644 --- a/arch/s390/appldata/appldata_base.c +++ b/arch/s390/appldata/appldata_base.c @@ -14,20 +14,20 @@ #include #include #include -#include -#include -#include #include #include #include #include #include #include -#include -//#include #include #include #include +#include +#include +#include +#include +#include #include "appldata.h" @@ -39,34 +39,6 @@ #define TOD_MICRO 0x01000 /* nr. of TOD clock units for 1 microsecond */ - -/* - * Parameter list for DIAGNOSE X'DC' - */ -#ifndef CONFIG_64BIT -struct appldata_parameter_list { - u16 diag; /* The DIAGNOSE code X'00DC' */ - u8 function; /* The function code for the DIAGNOSE */ - u8 parlist_length; /* Length of the parameter list */ - u32 product_id_addr; /* Address of the 16-byte product ID */ - u16 reserved; - u16 buffer_length; /* Length of the application data buffer */ - u32 buffer_addr; /* Address of the application data buffer */ -}; -#else -struct appldata_parameter_list { - u16 diag; - u8 function; - u8 parlist_length; - u32 unused01; - u16 reserved; - u16 buffer_length; - u32 unused02; - u64 product_id_addr; - u64 buffer_addr; -}; -#endif /* CONFIG_64BIT */ - /* * /proc entries (sysctl) */ @@ -181,46 +153,17 @@ static void appldata_work_fn(void *data) int appldata_diag(char record_nr, u16 function, unsigned long buffer, u16 length, char *mod_lvl) { - unsigned long ry; - struct appldata_product_id { - char prod_nr[7]; /* product nr. */ - char prod_fn[2]; /* product function */ - char record_nr; /* record nr. */ - char version_nr[2]; /* version */ - char release_nr[2]; /* release */ - char mod_lvl[2]; /* modification lvl. */ - } appldata_product_id = { - /* all strings are EBCDIC, record_nr is byte */ + struct appldata_product_id id = { .prod_nr = {0xD3, 0xC9, 0xD5, 0xE4, - 0xE7, 0xD2, 0xD9}, /* "LINUXKR" */ - .prod_fn = {0xD5, 0xD3}, /* "NL" */ + 0xE7, 0xD2, 0xD9}, /* "LINUXKR" */ + .prod_fn = 0xD5D3, /* "NL" */ .record_nr = record_nr, - .version_nr = {0xF2, 0xF6}, /* "26" */ - .release_nr = {0xF0, 0xF1}, /* "01" */ - .mod_lvl = {mod_lvl[0], mod_lvl[1]}, - }; - struct appldata_parameter_list appldata_parameter_list = { - .diag = 0xDC, - .function = function, - .parlist_length = - sizeof(appldata_parameter_list), - .buffer_length = length, - .product_id_addr = - (unsigned long) &appldata_product_id, - .buffer_addr = virt_to_phys((void *) buffer) + .version_nr = 0xF2F6, /* "26" */ + .release_nr = 0xF0F1, /* "01" */ + .mod_lvl = (mod_lvl[0]) << 8 | mod_lvl[1], }; - if (!MACHINE_IS_VM) - return -ENOSYS; - ry = -1; - asm volatile( - "diag %1,%0,0xDC\n\t" - : "=d" (ry) - : "d" (&appldata_parameter_list), - "m" (appldata_parameter_list), - "m" (appldata_product_id) - : "cc"); - return (int) ry; + return appldata_asm(&id, function, (void *) buffer, length); } /************************ timer, work, DIAG ****************************/ diff --git a/arch/s390/appldata/appldata_os.c b/arch/s390/appldata/appldata_os.c index 161acc5c8a1..76a15523ae9 100644 --- a/arch/s390/appldata/appldata_os.c +++ b/arch/s390/appldata/appldata_os.c @@ -16,6 +16,7 @@ #include #include #include +#include #include #include "appldata.h" -- cgit v1.2.3 From 07d43ce6a2ba0bb914078c3b066a7a3bab57599d Mon Sep 17 00:00:00 2001 From: Heiko Carstens Date: Wed, 20 Sep 2006 15:59:29 +0200 Subject: [S390] Remove kexec experimental flag. Follow other architectures and remove kexec experimental flag. Signed-off-by: Heiko Carstens Signed-off-by: Martin Schwidefsky --- arch/s390/Kconfig | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) (limited to 'arch') diff --git a/arch/s390/Kconfig b/arch/s390/Kconfig index 76122ce1e6c..b216ca659cd 100644 --- a/arch/s390/Kconfig +++ b/arch/s390/Kconfig @@ -460,8 +460,7 @@ config S390_HYPFS_FS information in an s390 hypervisor environment. config KEXEC - bool "kexec system call (EXPERIMENTAL)" - depends on EXPERIMENTAL + bool "kexec system call" help kexec is a system call that implements the ability to shutdown your current kernel, and to start another kernel. It is like a reboot -- cgit v1.2.3 From 31b58088292c7f00f0b81088bfb557285b0b6247 Mon Sep 17 00:00:00 2001 From: Melissa Howland Date: Wed, 20 Sep 2006 15:59:34 +0200 Subject: [S390] Linux API for writing z/VM APPLDATA Monitor records. This patch delivers a new Linux API in the form of a misc char device that is useable from user space and allows write access to the z/VM APPLDATA Monitor Records collected by the *MONITOR System Service of z/VM. Signed-off-by: Melissa Howland Signed-off-by: Martin Schwidefsky --- arch/s390/defconfig | 1 + 1 file changed, 1 insertion(+) (limited to 'arch') diff --git a/arch/s390/defconfig b/arch/s390/defconfig index f1d4591eddb..35da53986b1 100644 --- a/arch/s390/defconfig +++ b/arch/s390/defconfig @@ -428,6 +428,7 @@ CONFIG_S390_TAPE_34XX=m # CONFIG_VMLOGRDR is not set # CONFIG_VMCP is not set # CONFIG_MONREADER is not set +CONFIG_MONWRITER=m # # Cryptographic devices -- cgit v1.2.3 From 9282ed929758b82f448a40d3c17319d794970624 Mon Sep 17 00:00:00 2001 From: Gerald Schaefer Date: Wed, 20 Sep 2006 15:59:37 +0200 Subject: [S390] Cleanup in page table related code. Changed and simplified some page table related #defines and code. Signed-off-by: Gerald Schaefer Signed-off-by: Martin Schwidefsky --- arch/s390/mm/init.c | 36 ++++++++++++++---------------------- 1 file changed, 14 insertions(+), 22 deletions(-) (limited to 'arch') diff --git a/arch/s390/mm/init.c b/arch/s390/mm/init.c index 6e6b6de7777..cfd9b8f7a52 100644 --- a/arch/s390/mm/init.c +++ b/arch/s390/mm/init.c @@ -108,16 +108,23 @@ void __init paging_init(void) unsigned long pgdir_k = (__pa(swapper_pg_dir) & PAGE_MASK) | _KERNSEG_TABLE; static const int ssm_mask = 0x04000000L; unsigned long ro_start_pfn, ro_end_pfn; + unsigned long zones_size[MAX_NR_ZONES]; ro_start_pfn = PFN_DOWN((unsigned long)&__start_rodata); ro_end_pfn = PFN_UP((unsigned long)&__end_rodata); + memset(zones_size, 0, sizeof(zones_size)); + zones_size[ZONE_DMA] = max_low_pfn; + free_area_init_node(0, &contig_page_data, zones_size, + __pa(PAGE_OFFSET) >> PAGE_SHIFT, + zholes_size); + /* unmap whole virtual address space */ pg_dir = swapper_pg_dir; - for (i=0;ipgd0 = (_PAGE_TABLE | __pa(pg_table)); - pg_dir->pgd1 = (_PAGE_TABLE | (__pa(pg_table)+1024)); - pg_dir->pgd2 = (_PAGE_TABLE | (__pa(pg_table)+2048)); - pg_dir->pgd3 = (_PAGE_TABLE | (__pa(pg_table)+3072)); + pmd_populate_kernel(&init_mm, (pmd_t *) pg_dir, pg_table); pg_dir++; for (tmp = 0 ; tmp < PTRS_PER_PTE ; tmp++,pg_table++) { @@ -143,8 +147,8 @@ void __init paging_init(void) else pte = pfn_pte(pfn, PAGE_KERNEL); if (pfn >= max_low_pfn) - pte_clear(&init_mm, 0, &pte); - set_pte(pg_table, pte); + pte_val(pte) = _PAGE_TYPE_EMPTY; + set_pte(pg_table, pte); pfn++; } } @@ -159,16 +163,6 @@ void __init paging_init(void) : : "m" (pgdir_k), "m" (ssm_mask)); local_flush_tlb(); - - { - unsigned long zones_size[MAX_NR_ZONES]; - - memset(zones_size, 0, sizeof(zones_size)); - zones_size[ZONE_DMA] = max_low_pfn; - free_area_init_node(0, &contig_page_data, zones_size, - __pa(PAGE_OFFSET) >> PAGE_SHIFT, - zholes_size); - } return; } @@ -236,10 +230,8 @@ void __init paging_init(void) pte = pfn_pte(pfn, __pgprot(_PAGE_RO)); else pte = pfn_pte(pfn, PAGE_KERNEL); - if (pfn >= max_low_pfn) { - pte_clear(&init_mm, 0, &pte); - continue; - } + if (pfn >= max_low_pfn) + pte_val(pte) = _PAGE_TYPE_EMPTY; set_pte(pt_dir, pte); pfn++; } -- cgit v1.2.3 From 6837a8c352efcc5efc70424e9bfd94ff9bfa9a47 Mon Sep 17 00:00:00 2001 From: Gerald Schaefer Date: Wed, 20 Sep 2006 15:59:39 +0200 Subject: [S390] Cleanup in signal handling code. Signed-off-by: Gerald Schaefer Signed-off-by: Martin Schwidefsky --- arch/s390/kernel/signal.c | 39 +++++++++++++++++---------------------- 1 file changed, 17 insertions(+), 22 deletions(-) (limited to 'arch') diff --git a/arch/s390/kernel/signal.c b/arch/s390/kernel/signal.c index dd05423f87a..642095ec7c0 100644 --- a/arch/s390/kernel/signal.c +++ b/arch/s390/kernel/signal.c @@ -114,29 +114,26 @@ sys_sigaltstack(const stack_t __user *uss, stack_t __user *uoss, static int save_sigregs(struct pt_regs *regs, _sigregs __user *sregs) { unsigned long old_mask = regs->psw.mask; - int err; - + _sigregs user_sregs; + save_access_regs(current->thread.acrs); /* Copy a 'clean' PSW mask to the user to avoid leaking information about whether PER is currently on. */ regs->psw.mask = PSW_MASK_MERGE(PSW_USER_BITS, regs->psw.mask); - err = __copy_to_user(&sregs->regs.psw, ®s->psw, - sizeof(sregs->regs.psw)+sizeof(sregs->regs.gprs)); + memcpy(&user_sregs.regs.psw, ®s->psw, sizeof(sregs->regs.psw) + + sizeof(sregs->regs.gprs)); regs->psw.mask = old_mask; - if (err != 0) - return err; - err = __copy_to_user(&sregs->regs.acrs, current->thread.acrs, - sizeof(sregs->regs.acrs)); - if (err != 0) - return err; + memcpy(&user_sregs.regs.acrs, current->thread.acrs, + sizeof(sregs->regs.acrs)); /* * We have to store the fp registers to current->thread.fp_regs * to merge them with the emulated registers. */ save_fp_regs(¤t->thread.fp_regs); - return __copy_to_user(&sregs->fpregs, ¤t->thread.fp_regs, - sizeof(s390_fp_regs)); + memcpy(&user_sregs.fpregs, ¤t->thread.fp_regs, + sizeof(s390_fp_regs)); + return __copy_to_user(sregs, &user_sregs, sizeof(_sigregs)); } /* Returns positive number on error */ @@ -144,27 +141,25 @@ static int restore_sigregs(struct pt_regs *regs, _sigregs __user *sregs) { unsigned long old_mask = regs->psw.mask; int err; + _sigregs user_sregs; /* Alwys make any pending restarted system call return -EINTR */ current_thread_info()->restart_block.fn = do_no_restart_syscall; - err = __copy_from_user(®s->psw, &sregs->regs.psw, - sizeof(sregs->regs.psw)+sizeof(sregs->regs.gprs)); + err = __copy_from_user(&user_sregs, sregs, sizeof(_sigregs)); regs->psw.mask = PSW_MASK_MERGE(old_mask, regs->psw.mask); regs->psw.addr |= PSW_ADDR_AMODE; if (err) return err; - err = __copy_from_user(¤t->thread.acrs, &sregs->regs.acrs, - sizeof(sregs->regs.acrs)); - if (err) - return err; + memcpy(®s->psw, &user_sregs.regs.psw, sizeof(sregs->regs.psw) + + sizeof(sregs->regs.gprs)); + memcpy(¤t->thread.acrs, &user_sregs.regs.acrs, + sizeof(sregs->regs.acrs)); restore_access_regs(current->thread.acrs); - err = __copy_from_user(¤t->thread.fp_regs, &sregs->fpregs, - sizeof(s390_fp_regs)); + memcpy(¤t->thread.fp_regs, &user_sregs.fpregs, + sizeof(s390_fp_regs)); current->thread.fp_regs.fpc &= FPC_VALID_MASK; - if (err) - return err; restore_fp_regs(¤t->thread.fp_regs); regs->trap = -1; /* disable syscall checks */ -- cgit v1.2.3 From d02765d1af743567398eb6d523dea0ba5e5e7e8e Mon Sep 17 00:00:00 2001 From: Gerald Schaefer Date: Wed, 20 Sep 2006 15:59:42 +0200 Subject: [S390] Make user-copy operations run-time configurable. Introduces a struct uaccess_ops which allows setting user-copy operations at run-time. Signed-off-by: Gerald Schaefer Signed-off-by: Martin Schwidefsky --- arch/s390/kernel/s390_ksyms.c | 6 - arch/s390/kernel/setup.c | 7 + arch/s390/lib/Makefile | 3 +- arch/s390/lib/uaccess.S | 211 -------------------------- arch/s390/lib/uaccess64.S | 207 ------------------------- arch/s390/lib/uaccess_std.c | 340 ++++++++++++++++++++++++++++++++++++++++++ 6 files changed, 348 insertions(+), 426 deletions(-) delete mode 100644 arch/s390/lib/uaccess.S delete mode 100644 arch/s390/lib/uaccess64.S create mode 100644 arch/s390/lib/uaccess_std.c (limited to 'arch') diff --git a/arch/s390/kernel/s390_ksyms.c b/arch/s390/kernel/s390_ksyms.c index c73a45467fa..9f19e833a56 100644 --- a/arch/s390/kernel/s390_ksyms.c +++ b/arch/s390/kernel/s390_ksyms.c @@ -25,12 +25,6 @@ EXPORT_SYMBOL(_oi_bitmap); EXPORT_SYMBOL(_ni_bitmap); EXPORT_SYMBOL(_zb_findmap); EXPORT_SYMBOL(_sb_findmap); -EXPORT_SYMBOL(__copy_from_user_asm); -EXPORT_SYMBOL(__copy_to_user_asm); -EXPORT_SYMBOL(__copy_in_user_asm); -EXPORT_SYMBOL(__clear_user_asm); -EXPORT_SYMBOL(__strncpy_from_user_asm); -EXPORT_SYMBOL(__strnlen_user_asm); EXPORT_SYMBOL(diag10); /* diff --git a/arch/s390/kernel/setup.c b/arch/s390/kernel/setup.c index f2a9165ca4f..e229af59976 100644 --- a/arch/s390/kernel/setup.c +++ b/arch/s390/kernel/setup.c @@ -50,6 +50,12 @@ #include #include +/* + * User copy operations. + */ +struct uaccess_ops uaccess; +EXPORT_SYMBOL_GPL(uaccess); + /* * Machine setup.. */ @@ -641,6 +647,7 @@ setup_arch(char **cmdline_p) memory_end = memory_size; + memcpy(&uaccess, &uaccess_std, sizeof(uaccess)); parse_early_param(); #ifndef CONFIG_64BIT diff --git a/arch/s390/lib/Makefile b/arch/s390/lib/Makefile index e05d087a6ea..96c82424d88 100644 --- a/arch/s390/lib/Makefile +++ b/arch/s390/lib/Makefile @@ -4,6 +4,5 @@ EXTRA_AFLAGS := -traditional -lib-y += delay.o string.o -lib-y += $(if $(CONFIG_64BIT),uaccess64.o,uaccess.o) +lib-y += delay.o string.o uaccess_std.o lib-$(CONFIG_SMP) += spinlock.o diff --git a/arch/s390/lib/uaccess.S b/arch/s390/lib/uaccess.S deleted file mode 100644 index 837275284d9..00000000000 --- a/arch/s390/lib/uaccess.S +++ /dev/null @@ -1,211 +0,0 @@ -/* - * arch/s390/lib/uaccess.S - * __copy_{from|to}_user functions. - * - * s390 - * Copyright (C) 2000,2002 IBM Deutschland Entwicklung GmbH, IBM Corporation - * Authors(s): Martin Schwidefsky (schwidefsky@de.ibm.com) - * - * These functions have standard call interface - */ - -#include -#include -#include - - .text - .align 4 - .globl __copy_from_user_asm - # %r2 = to, %r3 = n, %r4 = from -__copy_from_user_asm: - slr %r0,%r0 -0: mvcp 0(%r3,%r2),0(%r4),%r0 - jnz 1f - slr %r2,%r2 - br %r14 -1: la %r2,256(%r2) - la %r4,256(%r4) - ahi %r3,-256 -2: mvcp 0(%r3,%r2),0(%r4),%r0 - jnz 1b -3: slr %r2,%r2 - br %r14 -4: lhi %r0,-4096 - lr %r5,%r4 - slr %r5,%r0 - nr %r5,%r0 # %r5 = (%r4 + 4096) & -4096 - slr %r5,%r4 # %r5 = #bytes to next user page boundary - clr %r3,%r5 # copy crosses next page boundary ? - jnh 6f # no, the current page faulted - # move with the reduced length which is < 256 -5: mvcp 0(%r5,%r2),0(%r4),%r0 - slr %r3,%r5 -6: lr %r2,%r3 - br %r14 - .section __ex_table,"a" - .long 0b,4b - .long 2b,4b - .long 5b,6b - .previous - - .align 4 - .text - .globl __copy_to_user_asm - # %r2 = from, %r3 = n, %r4 = to -__copy_to_user_asm: - slr %r0,%r0 -0: mvcs 0(%r3,%r4),0(%r2),%r0 - jnz 1f - slr %r2,%r2 - br %r14 -1: la %r2,256(%r2) - la %r4,256(%r4) - ahi %r3,-256 -2: mvcs 0(%r3,%r4),0(%r2),%r0 - jnz 1b -3: slr %r2,%r2 - br %r14 -4: lhi %r0,-4096 - lr %r5,%r4 - slr %r5,%r0 - nr %r5,%r0 # %r5 = (%r4 + 4096) & -4096 - slr %r5,%r4 # %r5 = #bytes to next user page boundary - clr %r3,%r5 # copy crosses next page boundary ? - jnh 6f # no, the current page faulted - # move with the reduced length which is < 256 -5: mvcs 0(%r5,%r4),0(%r2),%r0 - slr %r3,%r5 -6: lr %r2,%r3 - br %r14 - .section __ex_table,"a" - .long 0b,4b - .long 2b,4b - .long 5b,6b - .previous - - .align 4 - .text - .globl __copy_in_user_asm - # %r2 = from, %r3 = n, %r4 = to -__copy_in_user_asm: - ahi %r3,-1 - jo 6f - sacf 256 - bras %r1,4f -0: ahi %r3,257 -1: mvc 0(1,%r4),0(%r2) - la %r2,1(%r2) - la %r4,1(%r4) - ahi %r3,-1 - jnz 1b -2: lr %r2,%r3 - br %r14 -3: mvc 0(256,%r4),0(%r2) - la %r2,256(%r2) - la %r4,256(%r4) -4: ahi %r3,-256 - jnm 3b -5: ex %r3,4(%r1) - sacf 0 -6: slr %r2,%r2 - br %r14 - .section __ex_table,"a" - .long 1b,2b - .long 3b,0b - .long 5b,0b - .previous - - .align 4 - .text - .globl __clear_user_asm - # %r2 = to, %r3 = n -__clear_user_asm: - bras %r5,0f - .long empty_zero_page -0: l %r5,0(%r5) - slr %r0,%r0 -1: mvcs 0(%r3,%r2),0(%r5),%r0 - jnz 2f - slr %r2,%r2 - br %r14 -2: la %r2,256(%r2) - ahi %r3,-256 -3: mvcs 0(%r3,%r2),0(%r5),%r0 - jnz 2b -4: slr %r2,%r2 - br %r14 -5: lhi %r0,-4096 - lr %r4,%r2 - slr %r4,%r0 - nr %r4,%r0 # %r4 = (%r2 + 4096) & -4096 - slr %r4,%r2 # %r4 = #bytes to next user page boundary - clr %r3,%r4 # clear crosses next page boundary ? - jnh 7f # no, the current page faulted - # clear with the reduced length which is < 256 -6: mvcs 0(%r4,%r2),0(%r5),%r0 - slr %r3,%r4 -7: lr %r2,%r3 - br %r14 - .section __ex_table,"a" - .long 1b,5b - .long 3b,5b - .long 6b,7b - .previous - - .align 4 - .text - .globl __strncpy_from_user_asm - # %r2 = count, %r3 = dst, %r4 = src -__strncpy_from_user_asm: - lhi %r0,0 - lr %r1,%r4 - la %r4,0(%r4) # clear high order bit from %r4 - la %r2,0(%r2,%r4) # %r2 points to first byte after string - sacf 256 -0: srst %r2,%r1 - jo 0b - sacf 0 - lr %r1,%r2 - jh 1f # \0 found in string ? - ahi %r1,1 # include \0 in copy -1: slr %r1,%r4 # %r1 = copy length (without \0) - slr %r2,%r4 # %r2 = return length (including \0) -2: mvcp 0(%r1,%r3),0(%r4),%r0 - jnz 3f - br %r14 -3: la %r3,256(%r3) - la %r4,256(%r4) - ahi %r1,-256 - mvcp 0(%r1,%r3),0(%r4),%r0 - jnz 3b - br %r14 -4: sacf 0 - lhi %r2,-EFAULT - br %r14 - .section __ex_table,"a" - .long 0b,4b - .previous - - .align 4 - .text - .globl __strnlen_user_asm - # %r2 = count, %r3 = src -__strnlen_user_asm: - lhi %r0,0 - lr %r1,%r3 - la %r3,0(%r3) # clear high order bit from %r4 - la %r2,0(%r2,%r3) # %r2 points to first byte after string - sacf 256 -0: srst %r2,%r1 - jo 0b - sacf 0 - ahi %r2,1 # strnlen_user result includes the \0 - # or return count+1 if \0 not found - slr %r2,%r3 - br %r14 -2: sacf 0 - slr %r2,%r2 # return 0 on exception - br %r14 - .section __ex_table,"a" - .long 0b,2b - .previous diff --git a/arch/s390/lib/uaccess64.S b/arch/s390/lib/uaccess64.S deleted file mode 100644 index 1f755be22f9..00000000000 --- a/arch/s390/lib/uaccess64.S +++ /dev/null @@ -1,207 +0,0 @@ -/* - * arch/s390x/lib/uaccess.S - * __copy_{from|to}_user functions. - * - * s390 - * Copyright (C) 2000,2002 IBM Deutschland Entwicklung GmbH, IBM Corporation - * Authors(s): Martin Schwidefsky (schwidefsky@de.ibm.com) - * - * These functions have standard call interface - */ - -#include -#include -#include - - .text - .align 4 - .globl __copy_from_user_asm - # %r2 = to, %r3 = n, %r4 = from -__copy_from_user_asm: - slgr %r0,%r0 -0: mvcp 0(%r3,%r2),0(%r4),%r0 - jnz 1f - slgr %r2,%r2 - br %r14 -1: la %r2,256(%r2) - la %r4,256(%r4) - aghi %r3,-256 -2: mvcp 0(%r3,%r2),0(%r4),%r0 - jnz 1b -3: slgr %r2,%r2 - br %r14 -4: lghi %r0,-4096 - lgr %r5,%r4 - slgr %r5,%r0 - ngr %r5,%r0 # %r5 = (%r4 + 4096) & -4096 - slgr %r5,%r4 # %r5 = #bytes to next user page boundary - clgr %r3,%r5 # copy crosses next page boundary ? - jnh 6f # no, the current page faulted - # move with the reduced length which is < 256 -5: mvcp 0(%r5,%r2),0(%r4),%r0 - slgr %r3,%r5 -6: lgr %r2,%r3 - br %r14 - .section __ex_table,"a" - .quad 0b,4b - .quad 2b,4b - .quad 5b,6b - .previous - - .align 4 - .text - .globl __copy_to_user_asm - # %r2 = from, %r3 = n, %r4 = to -__copy_to_user_asm: - slgr %r0,%r0 -0: mvcs 0(%r3,%r4),0(%r2),%r0 - jnz 1f - slgr %r2,%r2 - br %r14 -1: la %r2,256(%r2) - la %r4,256(%r4) - aghi %r3,-256 -2: mvcs 0(%r3,%r4),0(%r2),%r0 - jnz 1b -3: slgr %r2,%r2 - br %r14 -4: lghi %r0,-4096 - lgr %r5,%r4 - slgr %r5,%r0 - ngr %r5,%r0 # %r5 = (%r4 + 4096) & -4096 - slgr %r5,%r4 # %r5 = #bytes to next user page boundary - clgr %r3,%r5 # copy crosses next page boundary ? - jnh 6f # no, the current page faulted - # move with the reduced length which is < 256 -5: mvcs 0(%r5,%r4),0(%r2),%r0 - slgr %r3,%r5 -6: lgr %r2,%r3 - br %r14 - .section __ex_table,"a" - .quad 0b,4b - .quad 2b,4b - .quad 5b,6b - .previous - - .align 4 - .text - .globl __copy_in_user_asm - # %r2 = from, %r3 = n, %r4 = to -__copy_in_user_asm: - aghi %r3,-1 - jo 6f - sacf 256 - bras %r1,4f -0: aghi %r3,257 -1: mvc 0(1,%r4),0(%r2) - la %r2,1(%r2) - la %r4,1(%r4) - aghi %r3,-1 - jnz 1b -2: lgr %r2,%r3 - br %r14 -3: mvc 0(256,%r4),0(%r2) - la %r2,256(%r2) - la %r4,256(%r4) -4: aghi %r3,-256 - jnm 3b -5: ex %r3,4(%r1) - sacf 0 -6: slgr %r2,%r2 - br 14 - .section __ex_table,"a" - .quad 1b,2b - .quad 3b,0b - .quad 5b,0b - .previous - - .align 4 - .text - .globl __clear_user_asm - # %r2 = to, %r3 = n -__clear_user_asm: - slgr %r0,%r0 - larl %r5,empty_zero_page -1: mvcs 0(%r3,%r2),0(%r5),%r0 - jnz 2f - slgr %r2,%r2 - br %r14 -2: la %r2,256(%r2) - aghi %r3,-256 -3: mvcs 0(%r3,%r2),0(%r5),%r0 - jnz 2b -4: slgr %r2,%r2 - br %r14 -5: lghi %r0,-4096 - lgr %r4,%r2 - slgr %r4,%r0 - ngr %r4,%r0 # %r4 = (%r2 + 4096) & -4096 - slgr %r4,%r2 # %r4 = #bytes to next user page boundary - clgr %r3,%r4 # clear crosses next page boundary ? - jnh 7f # no, the current page faulted - # clear with the reduced length which is < 256 -6: mvcs 0(%r4,%r2),0(%r5),%r0 - slgr %r3,%r4 -7: lgr %r2,%r3 - br %r14 - .section __ex_table,"a" - .quad 1b,5b - .quad 3b,5b - .quad 6b,7b - .previous - - .align 4 - .text - .globl __strncpy_from_user_asm - # %r2 = count, %r3 = dst, %r4 = src -__strncpy_from_user_asm: - lghi %r0,0 - lgr %r1,%r4 - la %r2,0(%r2,%r4) # %r2 points to first byte after string - sacf 256 -0: srst %r2,%r1 - jo 0b - sacf 0 - lgr %r1,%r2 - jh 1f # \0 found in string ? - aghi %r1,1 # include \0 in copy -1: slgr %r1,%r4 # %r1 = copy length (without \0) - slgr %r2,%r4 # %r2 = return length (including \0) -2: mvcp 0(%r1,%r3),0(%r4),%r0 - jnz 3f - br %r14 -3: la %r3,256(%r3) - la %r4,256(%r4) - aghi %r1,-256 - mvcp 0(%r1,%r3),0(%r4),%r0 - jnz 3b - br %r14 -4: sacf 0 - lghi %r2,-EFAULT - br %r14 - .section __ex_table,"a" - .quad 0b,4b - .previous - - .align 4 - .text - .globl __strnlen_user_asm - # %r2 = count, %r3 = src -__strnlen_user_asm: - lghi %r0,0 - lgr %r1,%r3 - la %r2,0(%r2,%r3) # %r2 points to first byte after string - sacf 256 -0: srst %r2,%r1 - jo 0b - sacf 0 - aghi %r2,1 # strnlen_user result includes the \0 - # or return count+1 if \0 not found - slgr %r2,%r3 - br %r14 -2: sacf 0 - slgr %r2,%r2 # return 0 on exception - br %r14 - .section __ex_table,"a" - .quad 0b,2b - .previous diff --git a/arch/s390/lib/uaccess_std.c b/arch/s390/lib/uaccess_std.c new file mode 100644 index 00000000000..9a4d4a29ea7 --- /dev/null +++ b/arch/s390/lib/uaccess_std.c @@ -0,0 +1,340 @@ +/* + * arch/s390/lib/uaccess_std.c + * + * Standard user space access functions based on mvcp/mvcs and doing + * interesting things in the secondary space mode. + * + * Copyright (C) IBM Corp. 2006 + * Author(s): Martin Schwidefsky (schwidefsky@de.ibm.com), + * Gerald Schaefer (gerald.schaefer@de.ibm.com) + */ + +#include +#include +#include +#include + +#ifndef __s390x__ +#define AHI "ahi" +#define ALR "alr" +#define CLR "clr" +#define LHI "lhi" +#define SLR "slr" +#else +#define AHI "aghi" +#define ALR "algr" +#define CLR "clgr" +#define LHI "lghi" +#define SLR "slgr" +#endif + +size_t copy_from_user_std(size_t size, const void __user *ptr, void *x) +{ + unsigned long tmp1, tmp2; + + tmp1 = -256UL; + asm volatile( + "0: mvcp 0(%0,%2),0(%1),%3\n" + " jz 5f\n" + "1:"ALR" %0,%3\n" + " la %1,256(%1)\n" + " la %2,256(%2)\n" + "2: mvcp 0(%0,%2),0(%1),%3\n" + " jnz 1b\n" + " j 5f\n" + "3: la %4,255(%1)\n" /* %4 = ptr + 255 */ + " "LHI" %3,-4096\n" + " nr %4,%3\n" /* %4 = (ptr + 255) & -4096 */ + " "SLR" %4,%1\n" + " "CLR" %0,%4\n" /* copy crosses next page boundary? */ + " jnh 6f\n" + "4: mvcp 0(%4,%2),0(%1),%3\n" + " "SLR" %0,%4\n" + " j 6f\n" + "5:"SLR" %0,%0\n" + "6: \n" + EX_TABLE(0b,3b) EX_TABLE(2b,3b) EX_TABLE(4b,6b) + : "+a" (size), "+a" (ptr), "+a" (x), "+a" (tmp1), "=a" (tmp2) + : : "cc", "memory"); + return size; +} + +size_t copy_from_user_std_small(size_t size, const void __user *ptr, void *x) +{ + unsigned long tmp1, tmp2; + + tmp1 = 0UL; + asm volatile( + "0: mvcp 0(%0,%2),0(%1),%3\n" + " "SLR" %0,%0\n" + " j 3f\n" + "1: la %4,255(%1)\n" /* %4 = ptr + 255 */ + " "LHI" %3,-4096\n" + " nr %4,%3\n" /* %4 = (ptr + 255) & -4096 */ + " "SLR" %4,%1\n" + " "CLR" %0,%4\n" /* copy crosses next page boundary? */ + " jnh 3f\n" + "2: mvcp 0(%4,%2),0(%1),%3\n" + " "SLR" %0,%4\n" + "3:\n" + EX_TABLE(0b,1b) EX_TABLE(2b,3b) + : "+a" (size), "+a" (ptr), "+a" (x), "+a" (tmp1), "=a" (tmp2) + : : "cc", "memory"); + return size; +} + +size_t copy_to_user_std(size_t size, void __user *ptr, const void *x) +{ + unsigned long tmp1, tmp2; + + tmp1 = -256UL; + asm volatile( + "0: mvcs 0(%0,%1),0(%2),%3\n" + " jz 5f\n" + "1:"ALR" %0,%3\n" + " la %1,256(%1)\n" + " la %2,256(%2)\n" + "2: mvcs 0(%0,%1),0(%2),%3\n" + " jnz 1b\n" + " j 5f\n" + "3: la %4,255(%1)\n" /* %4 = ptr + 255 */ + " "LHI" %3,-4096\n" + " nr %4,%3\n" /* %4 = (ptr + 255) & -4096 */ + " "SLR" %4,%1\n" + " "CLR" %0,%4\n" /* copy crosses next page boundary? */ + " jnh 6f\n" + "4: mvcs 0(%4,%1),0(%2),%3\n" + " "SLR" %0,%4\n" + " j 6f\n" + "5:"SLR" %0,%0\n" + "6: \n" + EX_TABLE(0b,3b) EX_TABLE(2b,3b) EX_TABLE(4b,6b) + : "+a" (size), "+a" (ptr), "+a" (x), "+a" (tmp1), "=a" (tmp2) + : : "cc", "memory"); + return size; +} + +size_t copy_to_user_std_small(size_t size, void __user *ptr, const void *x) +{ + unsigned long tmp1, tmp2; + + tmp1 = 0UL; + asm volatile( + "0: mvcs 0(%0,%1),0(%2),%3\n" + " "SLR" %0,%0\n" + " j 3f\n" + "1: la %4,255(%1)\n" /* ptr + 255 */ + " "LHI" %3,-4096\n" + " nr %4,%3\n" /* (ptr + 255) & -4096UL */ + " "SLR" %4,%1\n" + " "CLR" %0,%4\n" /* copy crosses next page boundary? */ + " jnh 3f\n" + "2: mvcs 0(%4,%1),0(%2),%3\n" + " "SLR" %0,%4\n" + "3:\n" + EX_TABLE(0b,1b) EX_TABLE(2b,3b) + : "+a" (size), "+a" (ptr), "+a" (x), "+a" (tmp1), "=a" (tmp2) + : : "cc", "memory"); + return size; +} + +size_t copy_in_user_std(size_t size, void __user *to, const void __user *from) +{ + unsigned long tmp1; + + asm volatile( + " "AHI" %0,-1\n" + " jo 5f\n" + " sacf 256\n" + " bras %3,3f\n" + "0:"AHI" %0,257\n" + "1: mvc 0(1,%1),0(%2)\n" + " la %1,1(%1)\n" + " la %2,1(%2)\n" + " "AHI" %0,-1\n" + " jnz 1b\n" + " j 5f\n" + "2: mvc 0(256,%1),0(%2)\n" + " la %1,256(%1)\n" + " la %2,256(%2)\n" + "3:"AHI" %0,-256\n" + " jnm 2b\n" + "4: ex %0,1b-0b(%3)\n" + " sacf 0\n" + "5: "SLR" %0,%0\n" + "6:\n" + EX_TABLE(1b,6b) EX_TABLE(2b,0b) EX_TABLE(4b,0b) + : "+a" (size), "+a" (to), "+a" (from), "=a" (tmp1) + : : "cc", "memory"); + return size; +} + +size_t clear_user_std(size_t size, void __user *to) +{ + unsigned long tmp1, tmp2; + + asm volatile( + " "AHI" %0,-1\n" + " jo 5f\n" + " sacf 256\n" + " bras %3,3f\n" + " xc 0(1,%1),0(%1)\n" + "0:"AHI" %0,257\n" + " la %2,255(%1)\n" /* %2 = ptr + 255 */ + " srl %2,12\n" + " sll %2,12\n" /* %2 = (ptr + 255) & -4096 */ + " "SLR" %2,%1\n" + " "CLR" %0,%2\n" /* clear crosses next page boundary? */ + " jnh 5f\n" + " "AHI" %2,-1\n" + "1: ex %2,0(%3)\n" + " "AHI" %2,1\n" + " "SLR" %0,%2\n" + " j 5f\n" + "2: xc 0(256,%1),0(%1)\n" + " la %1,256(%1)\n" + "3:"AHI" %0,-256\n" + " jnm 2b\n" + "4: ex %0,0(%3)\n" + " sacf 0\n" + "5: "SLR" %0,%0\n" + "6:\n" + EX_TABLE(1b,6b) EX_TABLE(2b,0b) EX_TABLE(4b,0b) + : "+a" (size), "+a" (to), "=a" (tmp1), "=a" (tmp2) + : : "cc", "memory"); + return size; +} + +size_t strnlen_user_std(size_t size, const char __user *src) +{ + register unsigned long reg0 asm("0") = 0UL; + unsigned long tmp1, tmp2; + + asm volatile( + " la %2,0(%1)\n" + " la %3,0(%0,%1)\n" + " "SLR" %0,%0\n" + " sacf 256\n" + "0: srst %3,%2\n" + " jo 0b\n" + " la %0,1(%3)\n" /* strnlen_user results includes \0 */ + " "SLR" %0,%1\n" + "1: sacf 0\n" + EX_TABLE(0b,1b) + : "+a" (size), "+a" (src), "=a" (tmp1), "=a" (tmp2) + : "d" (reg0) : "cc", "memory"); + return size; +} + +size_t strncpy_from_user_std(size_t size, const char __user *src, char *dst) +{ + register unsigned long reg0 asm("0") = 0UL; + unsigned long tmp1, tmp2; + + asm volatile( + " la %3,0(%1)\n" + " la %4,0(%0,%1)\n" + " sacf 256\n" + "0: srst %4,%3\n" + " jo 0b\n" + " sacf 0\n" + " la %0,0(%4)\n" + " jh 1f\n" /* found \0 in string ? */ + " "AHI" %4,1\n" /* include \0 in copy */ + "1:"SLR" %0,%1\n" /* %0 = return length (without \0) */ + " "SLR" %4,%1\n" /* %4 = copy length (including \0) */ + "2: mvcp 0(%4,%2),0(%1),%5\n" + " jz 9f\n" + "3:"AHI" %4,-256\n" + " la %1,256(%1)\n" + " la %2,256(%2)\n" + "4: mvcp 0(%4,%2),0(%1),%5\n" + " jnz 3b\n" + " j 9f\n" + "7: sacf 0\n" + "8:"LHI" %0,%6\n" + "9:\n" + EX_TABLE(0b,7b) EX_TABLE(2b,8b) EX_TABLE(4b,8b) + : "+a" (size), "+a" (src), "+d" (dst), "=a" (tmp1), "=a" (tmp2) + : "d" (reg0), "K" (-EFAULT) : "cc", "memory"); + return size; +} + +#define __futex_atomic_op(insn, ret, oldval, newval, uaddr, oparg) \ + asm volatile( \ + " sacf 256\n" \ + "0: l %1,0(%6)\n" \ + "1:"insn \ + "2: cs %1,%2,0(%6)\n" \ + "3: jl 1b\n" \ + " lhi %0,0\n" \ + "4: sacf 0\n" \ + EX_TABLE(0b,4b) EX_TABLE(2b,4b) EX_TABLE(3b,4b) \ + : "=d" (ret), "=&d" (oldval), "=&d" (newval), \ + "=m" (*uaddr) \ + : "0" (-EFAULT), "d" (oparg), "a" (uaddr), \ + "m" (*uaddr) : "cc"); + +int futex_atomic_op(int op, int __user *uaddr, int oparg, int *old) +{ + int oldval = 0, newval, ret; + + inc_preempt_count(); + + switch (op) { + case FUTEX_OP_SET: + __futex_atomic_op("lr %2,%5\n", + ret, oldval, newval, uaddr, oparg); + break; + case FUTEX_OP_ADD: + __futex_atomic_op("lr %2,%1\nar %2,%5\n", + ret, oldval, newval, uaddr, oparg); + break; + case FUTEX_OP_OR: + __futex_atomic_op("lr %2,%1\nor %2,%5\n", + ret, oldval, newval, uaddr, oparg); + break; + case FUTEX_OP_ANDN: + __futex_atomic_op("lr %2,%1\nnr %2,%5\n", + ret, oldval, newval, uaddr, oparg); + break; + case FUTEX_OP_XOR: + __futex_atomic_op("lr %2,%1\nxr %2,%5\n", + ret, oldval, newval, uaddr, oparg); + break; + default: + ret = -ENOSYS; + } + dec_preempt_count(); + *old = oldval; + return ret; +} + +int futex_atomic_cmpxchg(int __user *uaddr, int oldval, int newval) +{ + int ret; + + asm volatile( + " sacf 256\n" + " cs %1,%4,0(%5)\n" + "0: lr %0,%1\n" + "1: sacf 0\n" + EX_TABLE(0b,1b) + : "=d" (ret), "+d" (oldval), "=m" (*uaddr) + : "0" (-EFAULT), "d" (newval), "a" (uaddr), "m" (*uaddr) + : "cc", "memory" ); + return ret; +} + +struct uaccess_ops uaccess_std = { + .copy_from_user = copy_from_user_std, + .copy_from_user_small = copy_from_user_std_small, + .copy_to_user = copy_to_user_std, + .copy_to_user_small = copy_to_user_std_small, + .copy_in_user = copy_in_user_std, + .clear_user = clear_user_std, + .strnlen_user = strnlen_user_std, + .strncpy_from_user = strncpy_from_user_std, + .futex_atomic_op = futex_atomic_op, + .futex_atomic_cmpxchg = futex_atomic_cmpxchg, +}; -- cgit v1.2.3 From 6c2a9e6df60478e712f3c3d98b5047778a82a3d7 Mon Sep 17 00:00:00 2001 From: Gerald Schaefer Date: Wed, 20 Sep 2006 15:59:44 +0200 Subject: [S390] Use alternative user-copy operations for new hardware. This introduces new user-copy operations which are optimized for copying more than 256 Bytes on new hardware. Signed-off-by: Gerald Schaefer Signed-off-by: Martin Schwidefsky --- arch/s390/kernel/head64.S | 13 ++++ arch/s390/kernel/setup.c | 6 +- arch/s390/lib/Makefile | 1 + arch/s390/lib/uaccess_mvcos.c | 156 ++++++++++++++++++++++++++++++++++++++++++ 4 files changed, 175 insertions(+), 1 deletion(-) create mode 100644 arch/s390/lib/uaccess_mvcos.c (limited to 'arch') diff --git a/arch/s390/kernel/head64.S b/arch/s390/kernel/head64.S index 1ebaa338aa7..a8bdd96494c 100644 --- a/arch/s390/kernel/head64.S +++ b/arch/s390/kernel/head64.S @@ -250,6 +250,19 @@ startup_continue: oi 7(%r12),0x80 # set IDTE flag 0: +# +# find out if we have the MVCOS instruction +# + la %r1,0f-.LPG1(%r13) # set program check address + stg %r1,__LC_PGM_NEW_PSW+8 + .short 0xc800 # mvcos 0(%r0),0(%r0),%r0 + .short 0x0000 + .short 0x0000 +0: tm 0x8f,0x13 # special-operation exception? + bno 1f-.LPG1(%r13) # if yes, MVCOS is present + oi 6(%r12),2 # set MVCOS flag +1: + lpswe .Lentry-.LPG1(13) # jump to _stext in primary-space, # virtual and never return ... .align 16 diff --git a/arch/s390/kernel/setup.c b/arch/s390/kernel/setup.c index e229af59976..e3d9325f602 100644 --- a/arch/s390/kernel/setup.c +++ b/arch/s390/kernel/setup.c @@ -647,7 +647,11 @@ setup_arch(char **cmdline_p) memory_end = memory_size; - memcpy(&uaccess, &uaccess_std, sizeof(uaccess)); + if (MACHINE_HAS_MVCOS) + memcpy(&uaccess, &uaccess_mvcos, sizeof(uaccess)); + else + memcpy(&uaccess, &uaccess_std, sizeof(uaccess)); + parse_early_param(); #ifndef CONFIG_64BIT diff --git a/arch/s390/lib/Makefile b/arch/s390/lib/Makefile index 96c82424d88..c42ffedfdb4 100644 --- a/arch/s390/lib/Makefile +++ b/arch/s390/lib/Makefile @@ -5,4 +5,5 @@ EXTRA_AFLAGS := -traditional lib-y += delay.o string.o uaccess_std.o +lib-$(CONFIG_64BIT) += uaccess_mvcos.o lib-$(CONFIG_SMP) += spinlock.o diff --git a/arch/s390/lib/uaccess_mvcos.c b/arch/s390/lib/uaccess_mvcos.c new file mode 100644 index 00000000000..86c96d6c191 --- /dev/null +++ b/arch/s390/lib/uaccess_mvcos.c @@ -0,0 +1,156 @@ +/* + * arch/s390/lib/uaccess_mvcos.c + * + * Optimized user space space access functions based on mvcos. + * + * Copyright (C) IBM Corp. 2006 + * Author(s): Martin Schwidefsky (schwidefsky@de.ibm.com), + * Gerald Schaefer (gerald.schaefer@de.ibm.com) + */ + +#include +#include +#include +#include + +#ifndef __s390x__ +#define AHI "ahi" +#define ALR "alr" +#define CLR "clr" +#define LHI "lhi" +#define SLR "slr" +#else +#define AHI "aghi" +#define ALR "algr" +#define CLR "clgr" +#define LHI "lghi" +#define SLR "slgr" +#endif + +size_t copy_from_user_mvcos(size_t size, const void __user *ptr, void *x) +{ + register unsigned long reg0 asm("0") = 0x81UL; + unsigned long tmp1, tmp2; + + tmp1 = -4096UL; + asm volatile( + "0: .insn ss,0xc80000000000,0(%0,%2),0(%1),0\n" + " jz 4f\n" + "1:"ALR" %0,%3\n" + " "SLR" %1,%3\n" + " "SLR" %2,%3\n" + " j 0b\n" + "2: la %4,4095(%1)\n"/* %4 = ptr + 4095 */ + " nr %4,%3\n" /* %4 = (ptr + 4095) & -4096 */ + " "SLR" %4,%1\n" + " "CLR" %0,%4\n" /* copy crosses next page boundary? */ + " jnh 5f\n" + "3: .insn ss,0xc80000000000,0(%4,%2),0(%1),0\n" + " "SLR" %0,%4\n" + " j 5f\n" + "4:"SLR" %0,%0\n" + "5: \n" + EX_TABLE(0b,2b) EX_TABLE(3b,5b) + : "+a" (size), "+a" (ptr), "+a" (x), "+a" (tmp1), "=a" (tmp2) + : "d" (reg0) : "cc", "memory"); + return size; +} + +size_t copy_to_user_mvcos(size_t size, void __user *ptr, const void *x) +{ + register unsigned long reg0 asm("0") = 0x810000UL; + unsigned long tmp1, tmp2; + + tmp1 = -4096UL; + asm volatile( + "0: .insn ss,0xc80000000000,0(%0,%1),0(%2),0\n" + " jz 4f\n" + "1:"ALR" %0,%3\n" + " "SLR" %1,%3\n" + " "SLR" %2,%3\n" + " j 0b\n" + "2: la %4,4095(%1)\n"/* %4 = ptr + 4095 */ + " nr %4,%3\n" /* %4 = (ptr + 4095) & -4096 */ + " "SLR" %4,%1\n" + " "CLR" %0,%4\n" /* copy crosses next page boundary? */ + " jnh 5f\n" + "3: .insn ss,0xc80000000000,0(%4,%1),0(%2),0\n" + " "SLR" %0,%4\n" + " j 5f\n" + "4:"SLR" %0,%0\n" + "5: \n" + EX_TABLE(0b,2b) EX_TABLE(3b,5b) + : "+a" (size), "+a" (ptr), "+a" (x), "+a" (tmp1), "=a" (tmp2) + : "d" (reg0) : "cc", "memory"); + return size; +} + +size_t copy_in_user_mvcos(size_t size, void __user *to, const void __user *from) +{ + register unsigned long reg0 asm("0") = 0x810081UL; + unsigned long tmp1, tmp2; + + tmp1 = -4096UL; + /* FIXME: copy with reduced length. */ + asm volatile( + "0: .insn ss,0xc80000000000,0(%0,%1),0(%2),0\n" + " jz 2f\n" + "1:"ALR" %0,%3\n" + " "SLR" %1,%3\n" + " "SLR" %2,%3\n" + " j 0b\n" + "2:"SLR" %0,%0\n" + "3: \n" + EX_TABLE(0b,3b) + : "+a" (size), "+a" (to), "+a" (from), "+a" (tmp1), "=a" (tmp2) + : "d" (reg0) : "cc", "memory"); + return size; +} + +size_t clear_user_mvcos(size_t size, void __user *to) +{ + register unsigned long reg0 asm("0") = 0x810000UL; + unsigned long tmp1, tmp2; + + tmp1 = -4096UL; + asm volatile( + "0: .insn ss,0xc80000000000,0(%0,%1),0(%4),0\n" + " jz 4f\n" + "1:"ALR" %0,%2\n" + " "SLR" %1,%2\n" + " j 0b\n" + "2: la %3,4095(%1)\n"/* %4 = to + 4095 */ + " nr %3,%2\n" /* %4 = (to + 4095) & -4096 */ + " "SLR" %3,%1\n" + " "CLR" %0,%3\n" /* copy crosses next page boundary? */ + " jnh 5f\n" + "3: .insn ss,0xc80000000000,0(%3,%1),0(%4),0\n" + " "SLR" %0,%3\n" + " j 5f\n" + "4:"SLR" %0,%0\n" + "5: \n" + EX_TABLE(0b,2b) EX_TABLE(3b,5b) + : "+a" (size), "+a" (to), "+a" (tmp1), "=a" (tmp2) + : "a" (empty_zero_page), "d" (reg0) : "cc", "memory"); + return size; +} + +extern size_t copy_from_user_std_small(size_t, const void __user *, void *); +extern size_t copy_to_user_std_small(size_t, void __user *, const void *); +extern size_t strnlen_user_std(size_t, const char __user *); +extern size_t strncpy_from_user_std(size_t, const char __user *, char *); +extern int futex_atomic_op(int, int __user *, int, int *); +extern int futex_atomic_cmpxchg(int __user *, int, int); + +struct uaccess_ops uaccess_mvcos = { + .copy_from_user = copy_from_user_mvcos, + .copy_from_user_small = copy_from_user_std_small, + .copy_to_user = copy_to_user_mvcos, + .copy_to_user_small = copy_to_user_std_small, + .copy_in_user = copy_in_user_mvcos, + .clear_user = clear_user_mvcos, + .strnlen_user = strnlen_user_std, + .strncpy_from_user = strncpy_from_user_std, + .futex_atomic_op = futex_atomic_op, + .futex_atomic_cmpxchg = futex_atomic_cmpxchg, +}; -- cgit v1.2.3 From 388c571cffc4ae4e64f0786333e811308acbbc10 Mon Sep 17 00:00:00 2001 From: Michael Holzheu Date: Wed, 20 Sep 2006 16:00:04 +0200 Subject: [S390] hypfs crashes with invalid mount option. When an invalid mount option is specified, no root inode is created for hypfs, hypfs_fill_super() returns with -EINVAL and then hypfs_kill_super() is called. hypfs_kill_super() does not check if the root inode has been initialized. This patch adds this check. Signed-off-by: Michael Holzheu Signed-off-by: Martin Schwidefsky --- arch/s390/hypfs/inode.c | 10 ++++++---- 1 file changed, 6 insertions(+), 4 deletions(-) (limited to 'arch') diff --git a/arch/s390/hypfs/inode.c b/arch/s390/hypfs/inode.c index bdcad2ea1ff..bdade5f2e32 100644 --- a/arch/s390/hypfs/inode.c +++ b/arch/s390/hypfs/inode.c @@ -312,10 +312,12 @@ static void hypfs_kill_super(struct super_block *sb) { struct hypfs_sb_info *sb_info = sb->s_fs_info; - hypfs_delete_tree(sb->s_root); - hypfs_remove(sb_info->update_file); - kfree(sb->s_fs_info); - sb->s_fs_info = NULL; + if (sb->s_root) { + hypfs_delete_tree(sb->s_root); + hypfs_remove(sb_info->update_file); + kfree(sb->s_fs_info); + sb->s_fs_info = NULL; + } kill_litter_super(sb); } -- cgit v1.2.3