linux/arch/x86/kernel/signal_64.c

496 lines
13 KiB
C
Raw Normal View History

/*
* linux/arch/x86_64/kernel/signal.c
*
* Copyright (C) 1991, 1992 Linus Torvalds
* Copyright (C) 2000, 2001, 2002 Andi Kleen SuSE Labs
*
* 1997-11-28 Modified for POSIX.1b signals by Richard Henderson
* 2000-06-20 Pentium III FXSR, SSE support by Gareth Hughes
* 2000-2002 x86-64 support by Andi Kleen
*/
#include <linux/sched.h>
#include <linux/mm.h>
#include <linux/smp.h>
#include <linux/kernel.h>
#include <linux/signal.h>
#include <linux/errno.h>
#include <linux/wait.h>
#include <linux/ptrace.h>
#include <linux/unistd.h>
#include <linux/stddef.h>
#include <linux/personality.h>
#include <linux/compiler.h>
#include <asm/ucontext.h>
#include <asm/uaccess.h>
#include <asm/i387.h>
#include <asm/proto.h>
#include <asm/ia32_unistd.h>
x86_64: support poll() on /dev/mcelog Background: /dev/mcelog is typically polled manually. This is less than optimal for situations where accurate accounting of MCEs is important. Calling poll() on /dev/mcelog does not work. Description: This patch adds support for poll() to /dev/mcelog. This results in immediate wakeup of user apps whenever the poller finds MCEs. Because the exception handler can not take any locks, it can not call the wakeup itself. Instead, it uses a thread_info flag (TIF_MCE_NOTIFY) which is caught at the next return from interrupt or exit from idle, calling the mce_user_notify() routine. This patch also disables the "fake panic" path of the mce_panic(), because it results in printk()s in the exception handler and crashy systems. This patch also does some small cleanup for essentially unused variables, and moves the user notification into the body of the poller, so it is only called once per poll, rather than once per CPU. Result: Applications can now poll() on /dev/mcelog. When an error is logged (whether through the poller or through an exception) the applications are woken up promptly. This should not affect any previous behaviors. If no MCEs are being logged, there is no overhead. Alternatives: I considered simply supporting poll() through the poller and not using TIF_MCE_NOTIFY at all. However, the time between an uncorrectable error happening and the user application being notified is *the*most* critical window for us. Many uncorrectable errors can be logged to the network if given a chance. I also considered doing the MCE poll directly from the idle notifier, but decided that was overkill. Testing: I used an error-injecting DIMM to create lots of correctable DRAM errors and verified that my user app is woken up in sync with the polling interval. I also used the northbridge to inject uncorrectable ECC errors, and verified (printk() to the rescue) that the notify routine is called and the user app does wake up. I built with PREEMPT on and off, and verified that my machine survives MCEs. [wli@holomorphy.com: build fix] Signed-off-by: Tim Hockin <thockin@google.com> Signed-off-by: William Irwin <bill.irwin@oracle.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Andi Kleen <ak@suse.de> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-07-21 15:10:36 +00:00
#include <asm/mce.h>
/* #define DEBUG_SIG 1 */
#define _BLOCKABLE (~(sigmask(SIGKILL) | sigmask(SIGSTOP)))
int ia32_setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info,
sigset_t *set, struct pt_regs * regs);
int ia32_setup_frame(int sig, struct k_sigaction *ka,
sigset_t *set, struct pt_regs * regs);
asmlinkage long
sys_sigaltstack(const stack_t __user *uss, stack_t __user *uoss,
struct pt_regs *regs)
{
return do_sigaltstack(uss, uoss, regs->rsp);
}
/*
* Do a signal return; undo the signal stack.
*/
struct rt_sigframe
{
char __user *pretcode;
struct ucontext uc;
struct siginfo info;
};
static int
restore_sigcontext(struct pt_regs *regs, struct sigcontext __user *sc, unsigned long *prax)
{
unsigned int err = 0;
/* Always make any pending restarted system calls return -EINTR */
current_thread_info()->restart_block.fn = do_no_restart_syscall;
#define COPY(x) err |= __get_user(regs->x, &sc->x)
COPY(rdi); COPY(rsi); COPY(rbp); COPY(rsp); COPY(rbx);
COPY(rdx); COPY(rcx); COPY(rip);
COPY(r8);
COPY(r9);
COPY(r10);
COPY(r11);
COPY(r12);
COPY(r13);
COPY(r14);
COPY(r15);
/* Kernel saves and restores only the CS segment register on signals,
* which is the bare minimum needed to allow mixed 32/64-bit code.
* App's signal handler can save/restore other segments if needed. */
{
unsigned cs;
err |= __get_user(cs, &sc->cs);
regs->cs = cs | 3; /* Force into user mode */
}
{
unsigned int tmpflags;
err |= __get_user(tmpflags, &sc->eflags);
regs->eflags = (regs->eflags & ~0x40DD5) | (tmpflags & 0x40DD5);
regs->orig_rax = -1; /* disable syscall checks */
}
{
struct _fpstate __user * buf;
err |= __get_user(buf, &sc->fpstate);
if (buf) {
if (!access_ok(VERIFY_READ, buf, sizeof(*buf)))
goto badframe;
err |= restore_i387(buf);
} else {
struct task_struct *me = current;
if (used_math()) {
clear_fpu(me);
clear_used_math();
}
}
}
err |= __get_user(*prax, &sc->rax);
return err;
badframe:
return 1;
}
asmlinkage long sys_rt_sigreturn(struct pt_regs *regs)
{
struct rt_sigframe __user *frame;
sigset_t set;
unsigned long eax;
frame = (struct rt_sigframe __user *)(regs->rsp - 8);
if (!access_ok(VERIFY_READ, frame, sizeof(*frame))) {
goto badframe;
}
if (__copy_from_user(&set, &frame->uc.uc_sigmask, sizeof(set))) {
goto badframe;
}
sigdelsetmask(&set, ~_BLOCKABLE);
spin_lock_irq(&current->sighand->siglock);
current->blocked = set;
recalc_sigpending();
spin_unlock_irq(&current->sighand->siglock);
if (restore_sigcontext(regs, &frame->uc.uc_mcontext, &eax))
goto badframe;
#ifdef DEBUG_SIG
printk("%d sigreturn rip:%lx rsp:%lx frame:%p rax:%lx\n",current->pid,regs->rip,regs->rsp,frame,eax);
#endif
if (do_sigaltstack(&frame->uc.uc_stack, NULL, regs->rsp) == -EFAULT)
goto badframe;
return eax;
badframe:
signal_fault(regs,frame,"sigreturn");
return 0;
}
/*
* Set up a signal frame.
*/
static inline int
setup_sigcontext(struct sigcontext __user *sc, struct pt_regs *regs, unsigned long mask, struct task_struct *me)
{
int err = 0;
err |= __put_user(regs->cs, &sc->cs);
err |= __put_user(0, &sc->gs);
err |= __put_user(0, &sc->fs);
err |= __put_user(regs->rdi, &sc->rdi);
err |= __put_user(regs->rsi, &sc->rsi);
err |= __put_user(regs->rbp, &sc->rbp);
err |= __put_user(regs->rsp, &sc->rsp);
err |= __put_user(regs->rbx, &sc->rbx);
err |= __put_user(regs->rdx, &sc->rdx);
err |= __put_user(regs->rcx, &sc->rcx);
err |= __put_user(regs->rax, &sc->rax);
err |= __put_user(regs->r8, &sc->r8);
err |= __put_user(regs->r9, &sc->r9);
err |= __put_user(regs->r10, &sc->r10);
err |= __put_user(regs->r11, &sc->r11);
err |= __put_user(regs->r12, &sc->r12);
err |= __put_user(regs->r13, &sc->r13);
err |= __put_user(regs->r14, &sc->r14);
err |= __put_user(regs->r15, &sc->r15);
err |= __put_user(me->thread.trap_no, &sc->trapno);
err |= __put_user(me->thread.error_code, &sc->err);
err |= __put_user(regs->rip, &sc->rip);
err |= __put_user(regs->eflags, &sc->eflags);
err |= __put_user(mask, &sc->oldmask);
err |= __put_user(me->thread.cr2, &sc->cr2);
return err;
}
/*
* Determine which stack to use..
*/
static void __user *
get_stack(struct k_sigaction *ka, struct pt_regs *regs, unsigned long size)
{
unsigned long rsp;
/* Default to using normal stack - redzone*/
rsp = regs->rsp - 128;
/* This is the X/Open sanctioned signal stack switching. */
if (ka->sa.sa_flags & SA_ONSTACK) {
if (sas_ss_flags(rsp) == 0)
rsp = current->sas_ss_sp + current->sas_ss_size;
}
return (void __user *)round_down(rsp - size, 16);
}
static int setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info,
sigset_t *set, struct pt_regs * regs)
{
struct rt_sigframe __user *frame;
struct _fpstate __user *fp = NULL;
int err = 0;
struct task_struct *me = current;
if (used_math()) {
fp = get_stack(ka, regs, sizeof(struct _fpstate));
frame = (void __user *)round_down(
(unsigned long)fp - sizeof(struct rt_sigframe), 16) - 8;
if (!access_ok(VERIFY_WRITE, fp, sizeof(struct _fpstate)))
goto give_sigsegv;
if (save_i387(fp) < 0)
err |= -1;
} else
frame = get_stack(ka, regs, sizeof(struct rt_sigframe)) - 8;
if (!access_ok(VERIFY_WRITE, frame, sizeof(*frame)))
goto give_sigsegv;
if (ka->sa.sa_flags & SA_SIGINFO) {
err |= copy_siginfo_to_user(&frame->info, info);
if (err)
goto give_sigsegv;
}
/* Create the ucontext. */
err |= __put_user(0, &frame->uc.uc_flags);
err |= __put_user(0, &frame->uc.uc_link);
err |= __put_user(me->sas_ss_sp, &frame->uc.uc_stack.ss_sp);
err |= __put_user(sas_ss_flags(regs->rsp),
&frame->uc.uc_stack.ss_flags);
err |= __put_user(me->sas_ss_size, &frame->uc.uc_stack.ss_size);
err |= setup_sigcontext(&frame->uc.uc_mcontext, regs, set->sig[0], me);
err |= __put_user(fp, &frame->uc.uc_mcontext.fpstate);
if (sizeof(*set) == 16) {
__put_user(set->sig[0], &frame->uc.uc_sigmask.sig[0]);
__put_user(set->sig[1], &frame->uc.uc_sigmask.sig[1]);
} else
err |= __copy_to_user(&frame->uc.uc_sigmask, set, sizeof(*set));
/* Set up to return from userspace. If provided, use a stub
already in userspace. */
/* x86-64 should always use SA_RESTORER. */
if (ka->sa.sa_flags & SA_RESTORER) {
err |= __put_user(ka->sa.sa_restorer, &frame->pretcode);
} else {
/* could use a vstub here */
goto give_sigsegv;
}
if (err)
goto give_sigsegv;
#ifdef DEBUG_SIG
printk("%d old rip %lx old rsp %lx old rax %lx\n", current->pid,regs->rip,regs->rsp,regs->rax);
#endif
/* Set up registers for signal handler */
regs->rdi = sig;
/* In case the signal handler was declared without prototypes */
regs->rax = 0;
/* This also works for non SA_SIGINFO handlers because they expect the
next argument after the signal number on the stack. */
regs->rsi = (unsigned long)&frame->info;
regs->rdx = (unsigned long)&frame->uc;
regs->rip = (unsigned long) ka->sa.sa_handler;
regs->rsp = (unsigned long)frame;
/* Set up the CS register to run signal handlers in 64-bit mode,
even if the handler happens to be interrupting 32-bit code. */
regs->cs = __USER_CS;
/* This, by contrast, has nothing to do with segment registers -
see include/asm-x86_64/uaccess.h for details. */
set_fs(USER_DS);
regs->eflags &= ~TF_MASK;
if (test_thread_flag(TIF_SINGLESTEP))
ptrace_notify(SIGTRAP);
#ifdef DEBUG_SIG
printk("SIG deliver (%s:%d): sp=%p pc=%lx ra=%p\n",
current->comm, current->pid, frame, regs->rip, frame->pretcode);
#endif
return 0;
give_sigsegv:
force_sigsegv(sig, current);
return -EFAULT;
}
/*
* OK, we're invoking a handler
*/
static int
handle_signal(unsigned long sig, siginfo_t *info, struct k_sigaction *ka,
sigset_t *oldset, struct pt_regs *regs)
{
int ret;
#ifdef DEBUG_SIG
printk("handle_signal pid:%d sig:%lu rip:%lx rsp:%lx regs=%p\n",
current->pid, sig,
regs->rip, regs->rsp, regs);
#endif
/* Are we from a system call? */
if ((long)regs->orig_rax >= 0) {
/* If so, check system call restarting.. */
switch (regs->rax) {
case -ERESTART_RESTARTBLOCK:
case -ERESTARTNOHAND:
regs->rax = -EINTR;
break;
case -ERESTARTSYS:
if (!(ka->sa.sa_flags & SA_RESTART)) {
regs->rax = -EINTR;
break;
}
/* fallthrough */
case -ERESTARTNOINTR:
regs->rax = regs->orig_rax;
regs->rip -= 2;
break;
}
}
/*
* If TF is set due to a debugger (PT_DTRACE), clear the TF
* flag so that register information in the sigcontext is
* correct.
*/
if (unlikely(regs->eflags & TF_MASK)) {
if (likely(current->ptrace & PT_DTRACE)) {
current->ptrace &= ~PT_DTRACE;
regs->eflags &= ~TF_MASK;
}
}
#ifdef CONFIG_IA32_EMULATION
if (test_thread_flag(TIF_IA32)) {
if (ka->sa.sa_flags & SA_SIGINFO)
ret = ia32_setup_rt_frame(sig, ka, info, oldset, regs);
else
ret = ia32_setup_frame(sig, ka, oldset, regs);
} else
#endif
ret = setup_rt_frame(sig, ka, info, oldset, regs);
if (ret == 0) {
spin_lock_irq(&current->sighand->siglock);
sigorsets(&current->blocked,&current->blocked,&ka->sa.sa_mask);
2005-08-29 15:44:09 +00:00
if (!(ka->sa.sa_flags & SA_NODEFER))
sigaddset(&current->blocked,sig);
recalc_sigpending();
spin_unlock_irq(&current->sighand->siglock);
}
return ret;
}
/*
* Note that 'init' is a special process: it doesn't get signals it doesn't
* want to handle. Thus you cannot kill init even with a SIGKILL even by
* mistake.
*/
static void do_signal(struct pt_regs *regs)
{
struct k_sigaction ka;
siginfo_t info;
int signr;
sigset_t *oldset;
/*
* We want the common case to go fast, which
* is why we may in certain cases get here from
* kernel mode. Just return without doing anything
* if so.
*/
if (!user_mode(regs))
return;
if (test_thread_flag(TIF_RESTORE_SIGMASK))
oldset = &current->saved_sigmask;
else
oldset = &current->blocked;
signr = get_signal_to_deliver(&info, &ka, regs, NULL);
if (signr > 0) {
/* Reenable any watchpoints before delivering the
* signal to user space. The processor register will
* have been cleared if the watchpoint triggered
* inside the kernel.
*/
if (current->thread.debugreg7)
set_debugreg(current->thread.debugreg7, 7);
/* Whee! Actually deliver the signal. */
if (handle_signal(signr, &info, &ka, oldset, regs) == 0) {
/* a signal was successfully delivered; the saved
* sigmask will have been stored in the signal frame,
* and will be restored by sigreturn, so we can simply
* clear the TIF_RESTORE_SIGMASK flag */
clear_thread_flag(TIF_RESTORE_SIGMASK);
}
return;
}
/* Did we come from a system call? */
if ((long)regs->orig_rax >= 0) {
/* Restart the system call - no handlers present */
long res = regs->rax;
switch (res) {
case -ERESTARTNOHAND:
case -ERESTARTSYS:
case -ERESTARTNOINTR:
regs->rax = regs->orig_rax;
regs->rip -= 2;
break;
case -ERESTART_RESTARTBLOCK:
regs->rax = test_thread_flag(TIF_IA32) ?
__NR_ia32_restart_syscall :
__NR_restart_syscall;
regs->rip -= 2;
break;
}
}
/* if there's no signal to deliver, we just put the saved sigmask
back. */
if (test_thread_flag(TIF_RESTORE_SIGMASK)) {
clear_thread_flag(TIF_RESTORE_SIGMASK);
sigprocmask(SIG_SETMASK, &current->saved_sigmask, NULL);
}
}
void
do_notify_resume(struct pt_regs *regs, void *unused, __u32 thread_info_flags)
{
#ifdef DEBUG_SIG
printk("do_notify_resume flags:%x rip:%lx rsp:%lx caller:%p pending:%x\n",
thread_info_flags, regs->rip, regs->rsp, __builtin_return_address(0),signal_pending(current));
#endif
/* Pending single-step? */
if (thread_info_flags & _TIF_SINGLESTEP) {
regs->eflags |= TF_MASK;
clear_thread_flag(TIF_SINGLESTEP);
}
x86_64: support poll() on /dev/mcelog Background: /dev/mcelog is typically polled manually. This is less than optimal for situations where accurate accounting of MCEs is important. Calling poll() on /dev/mcelog does not work. Description: This patch adds support for poll() to /dev/mcelog. This results in immediate wakeup of user apps whenever the poller finds MCEs. Because the exception handler can not take any locks, it can not call the wakeup itself. Instead, it uses a thread_info flag (TIF_MCE_NOTIFY) which is caught at the next return from interrupt or exit from idle, calling the mce_user_notify() routine. This patch also disables the "fake panic" path of the mce_panic(), because it results in printk()s in the exception handler and crashy systems. This patch also does some small cleanup for essentially unused variables, and moves the user notification into the body of the poller, so it is only called once per poll, rather than once per CPU. Result: Applications can now poll() on /dev/mcelog. When an error is logged (whether through the poller or through an exception) the applications are woken up promptly. This should not affect any previous behaviors. If no MCEs are being logged, there is no overhead. Alternatives: I considered simply supporting poll() through the poller and not using TIF_MCE_NOTIFY at all. However, the time between an uncorrectable error happening and the user application being notified is *the*most* critical window for us. Many uncorrectable errors can be logged to the network if given a chance. I also considered doing the MCE poll directly from the idle notifier, but decided that was overkill. Testing: I used an error-injecting DIMM to create lots of correctable DRAM errors and verified that my user app is woken up in sync with the polling interval. I also used the northbridge to inject uncorrectable ECC errors, and verified (printk() to the rescue) that the notify routine is called and the user app does wake up. I built with PREEMPT on and off, and verified that my machine survives MCEs. [wli@holomorphy.com: build fix] Signed-off-by: Tim Hockin <thockin@google.com> Signed-off-by: William Irwin <bill.irwin@oracle.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Andi Kleen <ak@suse.de> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-07-21 15:10:36 +00:00
#ifdef CONFIG_X86_MCE
/* notify userspace of pending MCEs */
if (thread_info_flags & _TIF_MCE_NOTIFY)
mce_notify_user();
#endif /* CONFIG_X86_MCE */
/* deal with pending signal delivery */
if (thread_info_flags & (_TIF_SIGPENDING|_TIF_RESTORE_SIGMASK))
do_signal(regs);
}
void signal_fault(struct pt_regs *regs, void __user *frame, char *where)
{
struct task_struct *me = current;
if (show_unhandled_signals && printk_ratelimit())
printk("%s[%d] bad frame in %s frame:%p rip:%lx rsp:%lx orax:%lx\n",
me->comm,me->pid,where,frame,regs->rip,regs->rsp,regs->orig_rax);
force_sig(SIGSEGV, me);
}