2005-04-16 15:20:36 -07:00
|
|
|
/*
|
|
|
|
* Copyright (C) 1995 Linus Torvalds
|
|
|
|
*
|
|
|
|
* Pentium III FXSR, SSE support
|
|
|
|
* Gareth Hughes <gareth@valinux.com>, May 2000
|
|
|
|
*/
|
|
|
|
|
|
|
|
/*
|
|
|
|
* This file handles the architecture-dependent parts of process handling..
|
|
|
|
*/
|
|
|
|
|
|
|
|
#include <stdarg.h>
|
|
|
|
|
2005-06-25 14:54:50 -07:00
|
|
|
#include <linux/cpu.h>
|
2005-04-16 15:20:36 -07:00
|
|
|
#include <linux/errno.h>
|
|
|
|
#include <linux/sched.h>
|
|
|
|
#include <linux/fs.h>
|
|
|
|
#include <linux/kernel.h>
|
|
|
|
#include <linux/mm.h>
|
|
|
|
#include <linux/elfcore.h>
|
|
|
|
#include <linux/smp.h>
|
|
|
|
#include <linux/stddef.h>
|
|
|
|
#include <linux/slab.h>
|
|
|
|
#include <linux/vmalloc.h>
|
|
|
|
#include <linux/user.h>
|
|
|
|
#include <linux/a.out.h>
|
|
|
|
#include <linux/interrupt.h>
|
|
|
|
#include <linux/utsname.h>
|
|
|
|
#include <linux/delay.h>
|
|
|
|
#include <linux/reboot.h>
|
|
|
|
#include <linux/init.h>
|
|
|
|
#include <linux/mc146818rtc.h>
|
|
|
|
#include <linux/module.h>
|
|
|
|
#include <linux/kallsyms.h>
|
|
|
|
#include <linux/ptrace.h>
|
|
|
|
#include <linux/random.h>
|
2006-09-26 01:52:28 -07:00
|
|
|
#include <linux/personality.h>
|
2007-02-16 02:28:07 -07:00
|
|
|
#include <linux/tick.h>
|
2007-05-02 10:27:16 -07:00
|
|
|
#include <linux/percpu.h>
|
2005-04-16 15:20:36 -07:00
|
|
|
|
|
|
|
#include <asm/uaccess.h>
|
|
|
|
#include <asm/pgtable.h>
|
|
|
|
#include <asm/system.h>
|
|
|
|
#include <asm/io.h>
|
|
|
|
#include <asm/ldt.h>
|
|
|
|
#include <asm/processor.h>
|
|
|
|
#include <asm/i387.h>
|
|
|
|
#include <asm/desc.h>
|
2006-01-08 02:05:26 -07:00
|
|
|
#include <asm/vm86.h>
|
2005-04-16 15:20:36 -07:00
|
|
|
#ifdef CONFIG_MATH_EMULATION
|
|
|
|
#include <asm/math_emu.h>
|
|
|
|
#endif
|
|
|
|
|
|
|
|
#include <linux/err.h>
|
|
|
|
|
2005-06-25 14:54:50 -07:00
|
|
|
#include <asm/tlbflush.h>
|
|
|
|
#include <asm/cpu.h>
|
2008-01-30 05:30:17 -07:00
|
|
|
#include <asm/kdebug.h>
|
2005-06-25 14:54:50 -07:00
|
|
|
|
2005-04-16 15:20:36 -07:00
|
|
|
asmlinkage void ret_from_fork(void) __asm__("ret_from_fork");
|
|
|
|
|
|
|
|
static int hlt_counter;
|
|
|
|
|
|
|
|
unsigned long boot_option_idle_override = 0;
|
|
|
|
EXPORT_SYMBOL(boot_option_idle_override);
|
|
|
|
|
2007-05-02 10:27:16 -07:00
|
|
|
DEFINE_PER_CPU(struct task_struct *, current_task) = &init_task;
|
|
|
|
EXPORT_PER_CPU_SYMBOL(current_task);
|
|
|
|
|
|
|
|
DEFINE_PER_CPU(int, cpu_number);
|
|
|
|
EXPORT_PER_CPU_SYMBOL(cpu_number);
|
|
|
|
|
2005-04-16 15:20:36 -07:00
|
|
|
/*
|
|
|
|
* Return saved PC of a blocked thread.
|
|
|
|
*/
|
|
|
|
unsigned long thread_saved_pc(struct task_struct *tsk)
|
|
|
|
{
|
2008-01-30 05:31:02 -07:00
|
|
|
return ((unsigned long *)tsk->thread.sp)[3];
|
2005-04-16 15:20:36 -07:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Powermanagement idle function, if any..
|
|
|
|
*/
|
|
|
|
void (*pm_idle)(void);
|
2005-06-23 00:08:33 -07:00
|
|
|
EXPORT_SYMBOL(pm_idle);
|
2005-04-16 15:20:36 -07:00
|
|
|
static DEFINE_PER_CPU(unsigned int, cpu_idle_state);
|
|
|
|
|
|
|
|
void disable_hlt(void)
|
|
|
|
{
|
|
|
|
hlt_counter++;
|
|
|
|
}
|
|
|
|
|
|
|
|
EXPORT_SYMBOL(disable_hlt);
|
|
|
|
|
|
|
|
void enable_hlt(void)
|
|
|
|
{
|
|
|
|
hlt_counter--;
|
|
|
|
}
|
|
|
|
|
|
|
|
EXPORT_SYMBOL(enable_hlt);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* We use this if we don't have any better
|
|
|
|
* idle routine..
|
|
|
|
*/
|
|
|
|
void default_idle(void)
|
|
|
|
{
|
|
|
|
if (!hlt_counter && boot_cpu_data.hlt_works_ok) {
|
2006-06-26 04:59:11 -07:00
|
|
|
current_thread_info()->status &= ~TS_POLLING;
|
[PATCH] sched: fix bad missed wakeups in the i386, x86_64, ia64, ACPI and APM idle code
Fernando Lopez-Lezcano reported frequent scheduling latencies and audio
xruns starting at the 2.6.18-rt kernel, and those problems persisted all
until current -rt kernels. The latencies were serious and unjustified by
system load, often in the milliseconds range.
After a patient and heroic multi-month effort of Fernando, where he
tested dozens of kernels, tried various configs, boot options,
test-patches of mine and provided latency traces of those incidents, the
following 'smoking gun' trace was captured by him:
_------=> CPU#
/ _-----=> irqs-off
| / _----=> need-resched
|| / _---=> hardirq/softirq
||| / _--=> preempt-depth
|||| /
||||| delay
cmd pid ||||| time | caller
\ / ||||| \ | /
IRQ_19-1479 1D..1 0us : __trace_start_sched_wakeup (try_to_wake_up)
IRQ_19-1479 1D..1 0us : __trace_start_sched_wakeup <<...>-5856> (37 0)
IRQ_19-1479 1D..1 0us : __trace_start_sched_wakeup (c01262ba 0 0)
IRQ_19-1479 1D..1 0us : resched_task (try_to_wake_up)
IRQ_19-1479 1D..1 0us : __spin_unlock_irqrestore (try_to_wake_up)
...
<idle>-0 1...1 11us!: default_idle (cpu_idle)
...
<idle>-0 0Dn.1 602us : smp_apic_timer_interrupt (c0103baf 1 0)
...
<...>-5856 0D..2 618us : __switch_to (__schedule)
<...>-5856 0D..2 618us : __schedule <<idle>-0> (20 162)
<...>-5856 0D..2 619us : __spin_unlock_irq (__schedule)
<...>-5856 0...1 619us : trace_stop_sched_switched (__schedule)
<...>-5856 0D..1 619us : trace_stop_sched_switched <<...>-5856> (37 0)
what is visible in this trace is that CPU#1 ran try_to_wake_up() for
PID:5856, it placed PID:5856 on CPU#0's runqueue and ran resched_task()
for CPU#0. But it decided to not send an IPI that no CPU - due to
TS_POLLING. But CPU#0 never woke up after its NEED_RESCHED bit was set,
and only rescheduled to PID:5856 upon the next lapic timer IRQ. The
result was a 600+ usecs latency and a missed wakeup!
the bug turned out to be an idle-wakeup bug introduced into the mainline
kernel this summer via an optimization in the x86_64 tree:
commit 495ab9c045e1b0e5c82951b762257fe1c9d81564
Author: Andi Kleen <ak@suse.de>
Date: Mon Jun 26 13:59:11 2006 +0200
[PATCH] i386/x86-64/ia64: Move polling flag into thread_info_status
During some profiling I noticed that default_idle causes a lot of
memory traffic. I think that is caused by the atomic operations
to clear/set the polling flag in thread_info. There is actually
no reason to make this atomic - only the idle thread does it
to itself, other CPUs only read it. So I moved it into ti->status.
the problem is this type of change:
if (!hlt_counter && boot_cpu_data.hlt_works_ok) {
- clear_thread_flag(TIF_POLLING_NRFLAG);
+ current_thread_info()->status &= ~TS_POLLING;
smp_mb__after_clear_bit();
while (!need_resched()) {
local_irq_disable();
this changes clear_thread_flag() to an explicit clearing of TS_POLLING.
clear_thread_flag() is defined as:
clear_bit(flag, &ti->flags);
and clear_bit() is a LOCK-ed atomic instruction on all x86 platforms:
static inline void clear_bit(int nr, volatile unsigned long * addr)
{
__asm__ __volatile__( LOCK_PREFIX
"btrl %1,%0"
hence smp_mb__after_clear_bit() is defined as a simple compile barrier:
#define smp_mb__after_clear_bit() barrier()
but the explicit TS_POLLING clearing introduced by the patch:
+ current_thread_info()->status &= ~TS_POLLING;
is not an atomic op! So the clearing of the TS_POLLING bit is freely
reorderable with the reading of the NEED_RESCHED bit - and both now
reside in different memory addresses.
CPU idle wakeup very much depends on ordered memory ops, the clearing of
the TS_POLLING flag must always be done before we test need_resched()
and hit the idle instruction(s). [Symmetrically, the wakeup code needs
to set NEED_RESCHED before it tests the TS_POLLING flag, so memory
ordering is paramount.]
Fernando's dual-core Athlon64 system has a sufficiently advanced memory
ordering model so that it triggered this scenario very often.
( And it also turned out that the reason why these latencies never
triggered on my testsystems is that i routinely use idle=poll, which
was the only idle variant not affected by this bug. )
The fix is to change the smp_mb__after_clear_bit() to an smp_mb(), to
act as an absolute barrier between the TS_POLLING write and the
NEED_RESCHED read. This affects almost all idling methods (default,
ACPI, APM), on all 3 x86 architectures: i386, x86_64, ia64.
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Tested-by: Fernando Lopez-Lezcano <nando@ccrma.Stanford.EDU>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-12-22 02:11:56 -07:00
|
|
|
/*
|
|
|
|
* TS_POLLING-cleared state must be visible before we
|
|
|
|
* test NEED_RESCHED:
|
|
|
|
*/
|
|
|
|
smp_mb();
|
|
|
|
|
2006-12-06 18:14:03 -07:00
|
|
|
local_irq_disable();
|
2008-01-30 05:30:06 -07:00
|
|
|
if (!need_resched()) {
|
|
|
|
ktime_t t0, t1;
|
|
|
|
u64 t0n, t1n;
|
|
|
|
|
|
|
|
t0 = ktime_get();
|
|
|
|
t0n = ktime_to_ns(t0);
|
2006-12-06 18:14:03 -07:00
|
|
|
safe_halt(); /* enables interrupts racelessly */
|
2008-01-30 05:30:06 -07:00
|
|
|
local_irq_disable();
|
|
|
|
t1 = ktime_get();
|
|
|
|
t1n = ktime_to_ns(t1);
|
|
|
|
sched_clock_idle_wakeup_event(t1n - t0n);
|
|
|
|
}
|
|
|
|
local_irq_enable();
|
2006-06-26 04:59:11 -07:00
|
|
|
current_thread_info()->status |= TS_POLLING;
|
2005-04-16 15:20:36 -07:00
|
|
|
} else {
|
2006-12-06 18:14:03 -07:00
|
|
|
/* loop is done by the caller */
|
|
|
|
cpu_relax();
|
2005-04-16 15:20:36 -07:00
|
|
|
}
|
|
|
|
}
|
2005-06-23 00:08:33 -07:00
|
|
|
#ifdef CONFIG_APM_MODULE
|
|
|
|
EXPORT_SYMBOL(default_idle);
|
|
|
|
#endif
|
2005-04-16 15:20:36 -07:00
|
|
|
|
|
|
|
/*
|
|
|
|
* On SMP it's slightly faster (but much more power-consuming!)
|
|
|
|
* to poll the ->work.need_resched flag instead of waiting for the
|
|
|
|
* cross-CPU IPI to arrive. Use this option with caution.
|
|
|
|
*/
|
2008-01-30 05:31:03 -07:00
|
|
|
static void poll_idle(void)
|
2005-04-16 15:20:36 -07:00
|
|
|
{
|
2006-12-06 18:14:03 -07:00
|
|
|
cpu_relax();
|
2005-04-16 15:20:36 -07:00
|
|
|
}
|
|
|
|
|
2005-06-25 14:54:50 -07:00
|
|
|
#ifdef CONFIG_HOTPLUG_CPU
|
|
|
|
#include <asm/nmi.h>
|
|
|
|
/* We don't actually take CPU down, just spin without interrupts. */
|
|
|
|
static inline void play_dead(void)
|
|
|
|
{
|
2005-06-25 14:54:56 -07:00
|
|
|
/* This must be done before dead CPU ack */
|
|
|
|
cpu_exit_clear();
|
|
|
|
wbinvd();
|
|
|
|
mb();
|
2005-06-25 14:54:50 -07:00
|
|
|
/* Ack it */
|
|
|
|
__get_cpu_var(cpu_state) = CPU_DEAD;
|
|
|
|
|
2005-06-25 14:54:56 -07:00
|
|
|
/*
|
|
|
|
* With physical CPU hotplug, we should halt the cpu
|
|
|
|
*/
|
2005-06-25 14:54:50 -07:00
|
|
|
local_irq_disable();
|
2005-06-25 14:54:56 -07:00
|
|
|
while (1)
|
2005-09-03 15:56:42 -07:00
|
|
|
halt();
|
2005-06-25 14:54:50 -07:00
|
|
|
}
|
|
|
|
#else
|
|
|
|
static inline void play_dead(void)
|
|
|
|
{
|
|
|
|
BUG();
|
|
|
|
}
|
|
|
|
#endif /* CONFIG_HOTPLUG_CPU */
|
|
|
|
|
2005-04-16 15:20:36 -07:00
|
|
|
/*
|
|
|
|
* The idle thread. There's no useful work to be
|
|
|
|
* done, so just try to conserve power and have a
|
|
|
|
* low exit latency (ie sit in a loop waiting for
|
|
|
|
* somebody to say that they'd like to reschedule)
|
|
|
|
*/
|
2005-06-25 14:54:50 -07:00
|
|
|
void cpu_idle(void)
|
2005-04-16 15:20:36 -07:00
|
|
|
{
|
2005-11-08 22:39:01 -07:00
|
|
|
int cpu = smp_processor_id();
|
2005-06-25 14:54:50 -07:00
|
|
|
|
2006-06-26 04:59:11 -07:00
|
|
|
current_thread_info()->status |= TS_POLLING;
|
[PATCH] sched: resched and cpu_idle rework
Make some changes to the NEED_RESCHED and POLLING_NRFLAG to reduce
confusion, and make their semantics rigid. Improves efficiency of
resched_task and some cpu_idle routines.
* In resched_task:
- TIF_NEED_RESCHED is only cleared with the task's runqueue lock held,
and as we hold it during resched_task, then there is no need for an
atomic test and set there. The only other time this should be set is
when the task's quantum expires, in the timer interrupt - this is
protected against because the rq lock is irq-safe.
- If TIF_NEED_RESCHED is set, then we don't need to do anything. It
won't get unset until the task get's schedule()d off.
- If we are running on the same CPU as the task we resched, then set
TIF_NEED_RESCHED and no further action is required.
- If we are running on another CPU, and TIF_POLLING_NRFLAG is *not* set
after TIF_NEED_RESCHED has been set, then we need to send an IPI.
Using these rules, we are able to remove the test and set operation in
resched_task, and make clear the previously vague semantics of
POLLING_NRFLAG.
* In idle routines:
- Enter cpu_idle with preempt disabled. When the need_resched() condition
becomes true, explicitly call schedule(). This makes things a bit clearer
(IMO), but haven't updated all architectures yet.
- Many do a test and clear of TIF_NEED_RESCHED for some reason. According
to the resched_task rules, this isn't needed (and actually breaks the
assumption that TIF_NEED_RESCHED is only cleared with the runqueue lock
held). So remove that. Generally one less locked memory op when switching
to the idle thread.
- Many idle routines clear TIF_POLLING_NRFLAG, and only set it in the inner
most polling idle loops. The above resched_task semantics allow it to be
set until before the last time need_resched() is checked before going into
a halt requiring interrupt wakeup.
Many idle routines simply never enter such a halt, and so POLLING_NRFLAG
can be always left set, completely eliminating resched IPIs when rescheduling
the idle task.
POLLING_NRFLAG width can be increased, to reduce the chance of resched IPIs.
Signed-off-by: Nick Piggin <npiggin@suse.de>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Con Kolivas <kernel@kolivas.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-11-08 22:39:04 -07:00
|
|
|
|
2005-04-16 15:20:36 -07:00
|
|
|
/* endless idle loop with no priority at all */
|
|
|
|
while (1) {
|
2007-02-16 02:28:07 -07:00
|
|
|
tick_nohz_stop_sched_tick();
|
2005-04-16 15:20:36 -07:00
|
|
|
while (!need_resched()) {
|
|
|
|
void (*idle)(void);
|
|
|
|
|
|
|
|
if (__get_cpu_var(cpu_idle_state))
|
|
|
|
__get_cpu_var(cpu_idle_state) = 0;
|
|
|
|
|
2007-05-12 11:15:24 -07:00
|
|
|
check_pgt_cache();
|
2005-04-16 15:20:36 -07:00
|
|
|
rmb();
|
|
|
|
idle = pm_idle;
|
|
|
|
|
2008-01-30 05:33:13 -07:00
|
|
|
if (rcu_pending(cpu))
|
|
|
|
rcu_check_callbacks(cpu, 0);
|
|
|
|
|
2005-04-16 15:20:36 -07:00
|
|
|
if (!idle)
|
|
|
|
idle = default_idle;
|
|
|
|
|
2005-06-25 14:54:50 -07:00
|
|
|
if (cpu_is_offline(cpu))
|
|
|
|
play_dead();
|
|
|
|
|
2005-04-16 15:20:36 -07:00
|
|
|
__get_cpu_var(irq_stat).idle_timestamp = jiffies;
|
|
|
|
idle();
|
|
|
|
}
|
2007-02-16 02:28:07 -07:00
|
|
|
tick_nohz_restart_sched_tick();
|
2005-11-08 22:39:01 -07:00
|
|
|
preempt_enable_no_resched();
|
2005-04-16 15:20:36 -07:00
|
|
|
schedule();
|
2005-11-08 22:39:01 -07:00
|
|
|
preempt_disable();
|
2005-04-16 15:20:36 -07:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
Kick CPUS that might be sleeping in cpus_idle_wait
Sometimes cpu_idle_wait gets stuck because it might miss CPUS that are
already in idle, have no tasks waiting to run and have no interrupts going
to them. This is common on bootup when switching cpu idle governors.
This patch gives those CPUS that don't check in an IPI kick.
Background:
-----------
I notice this while developing the mcount patches, that every once in a
while the system would hang. Looking deeper, the hang was always at boot
up when registering init_menu of the cpu_idle menu governor. Talking
with Thomas Gliexner, we discovered that one of the CPUS had no timer
events scheduled for it and it was in idle (running with NO_HZ). So the
CPU would not set the cpu_idle_state bit.
Hitting sysrq-t a few times would eventually route the interrupt to the
stuck CPU and the system would continue.
Note, I would have used the PDA isidle but that is set after the
cpu_idle_state bit is cleared, and would leave a window open where we
may miss being kicked.
hmm, looking closer at this, we still have a small race window between
clearing the cpu_idle_state and disabling interrupts (hence the RFC).
CPU0: CPU 1:
--------- ---------
cpu_idle_wait(): cpu_idle():
| __cpu_cpu_var(is_idle) = 1;
| if (__get_cpu_var(cpu_idle_state)) /* == 0 */
per_cpu(cpu_idle_state, 1) = 1; |
if (per_cpu(is_idle, 1)) /* == 1 */ |
smp_call_function(1) |
| receives ipi and runs do_nothing.
wait on map == empty idle();
/* waits forever */
So really we need interrupts off for most of this then. One might think
that we could simply clear the cpu_idle_state from do_nothing, but I'm
assuming that cpu_idle governors can be removed, and this might cause a
race that a governor might be used after the module was removed.
Venki said:
I think your RFC patch is the right solution here. As I see it, there is
no race with your RFC patch. As long as you call a dummy smp_call_function
on all CPUs, we should be OK. We can get rid of cpu_idle_state and the
current wait forever logic altogether with dummy smp_call_function. And so
there wont be any wait forever scenario.
The whole point of cpu_idle_wait() is to make all CPUs come out of idle
loop atleast once. The caller will use cpu_idle_wait something like this.
// Want to change idle handler
- Switch global idle handler to always present default_idle
- call cpu_idle_wait so that all cpus come out of idle for an instant
and stop using old idle pointer and start using default idle
- Change the idle handler to a new handler
- optional cpu_idle_wait if you want all cpus to start using the new
handler immediately.
Maybe the below 1s patch is safe bet for .24. But for .25, I would say we
just replace all complicated logic by simple dummy smp_call_function and
remove cpu_idle_state altogether.
Signed-off-by: Steven Rostedt <srostedt@redhat.com>
Cc: Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>
Acked-by: Ingo Molnar <mingo@elte.hu>
Acked-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Andi Kleen <ak@suse.de>
Cc: Len Brown <lenb@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-01-14 01:55:10 -07:00
|
|
|
static void do_nothing(void *unused)
|
|
|
|
{
|
|
|
|
}
|
|
|
|
|
2005-04-16 15:20:36 -07:00
|
|
|
void cpu_idle_wait(void)
|
|
|
|
{
|
|
|
|
unsigned int cpu, this_cpu = get_cpu();
|
2006-11-17 06:26:18 -07:00
|
|
|
cpumask_t map, tmp = current->cpus_allowed;
|
2005-04-16 15:20:36 -07:00
|
|
|
|
|
|
|
set_cpus_allowed(current, cpumask_of_cpu(this_cpu));
|
|
|
|
put_cpu();
|
|
|
|
|
|
|
|
cpus_clear(map);
|
|
|
|
for_each_online_cpu(cpu) {
|
|
|
|
per_cpu(cpu_idle_state, cpu) = 1;
|
|
|
|
cpu_set(cpu, map);
|
|
|
|
}
|
|
|
|
|
|
|
|
__get_cpu_var(cpu_idle_state) = 0;
|
|
|
|
|
|
|
|
wmb();
|
|
|
|
do {
|
|
|
|
ssleep(1);
|
|
|
|
for_each_online_cpu(cpu) {
|
|
|
|
if (cpu_isset(cpu, map) && !per_cpu(cpu_idle_state, cpu))
|
|
|
|
cpu_clear(cpu, map);
|
|
|
|
}
|
|
|
|
cpus_and(map, map, cpu_online_map);
|
Kick CPUS that might be sleeping in cpus_idle_wait
Sometimes cpu_idle_wait gets stuck because it might miss CPUS that are
already in idle, have no tasks waiting to run and have no interrupts going
to them. This is common on bootup when switching cpu idle governors.
This patch gives those CPUS that don't check in an IPI kick.
Background:
-----------
I notice this while developing the mcount patches, that every once in a
while the system would hang. Looking deeper, the hang was always at boot
up when registering init_menu of the cpu_idle menu governor. Talking
with Thomas Gliexner, we discovered that one of the CPUS had no timer
events scheduled for it and it was in idle (running with NO_HZ). So the
CPU would not set the cpu_idle_state bit.
Hitting sysrq-t a few times would eventually route the interrupt to the
stuck CPU and the system would continue.
Note, I would have used the PDA isidle but that is set after the
cpu_idle_state bit is cleared, and would leave a window open where we
may miss being kicked.
hmm, looking closer at this, we still have a small race window between
clearing the cpu_idle_state and disabling interrupts (hence the RFC).
CPU0: CPU 1:
--------- ---------
cpu_idle_wait(): cpu_idle():
| __cpu_cpu_var(is_idle) = 1;
| if (__get_cpu_var(cpu_idle_state)) /* == 0 */
per_cpu(cpu_idle_state, 1) = 1; |
if (per_cpu(is_idle, 1)) /* == 1 */ |
smp_call_function(1) |
| receives ipi and runs do_nothing.
wait on map == empty idle();
/* waits forever */
So really we need interrupts off for most of this then. One might think
that we could simply clear the cpu_idle_state from do_nothing, but I'm
assuming that cpu_idle governors can be removed, and this might cause a
race that a governor might be used after the module was removed.
Venki said:
I think your RFC patch is the right solution here. As I see it, there is
no race with your RFC patch. As long as you call a dummy smp_call_function
on all CPUs, we should be OK. We can get rid of cpu_idle_state and the
current wait forever logic altogether with dummy smp_call_function. And so
there wont be any wait forever scenario.
The whole point of cpu_idle_wait() is to make all CPUs come out of idle
loop atleast once. The caller will use cpu_idle_wait something like this.
// Want to change idle handler
- Switch global idle handler to always present default_idle
- call cpu_idle_wait so that all cpus come out of idle for an instant
and stop using old idle pointer and start using default idle
- Change the idle handler to a new handler
- optional cpu_idle_wait if you want all cpus to start using the new
handler immediately.
Maybe the below 1s patch is safe bet for .24. But for .25, I would say we
just replace all complicated logic by simple dummy smp_call_function and
remove cpu_idle_state altogether.
Signed-off-by: Steven Rostedt <srostedt@redhat.com>
Cc: Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>
Acked-by: Ingo Molnar <mingo@elte.hu>
Acked-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Andi Kleen <ak@suse.de>
Cc: Len Brown <lenb@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-01-14 01:55:10 -07:00
|
|
|
/*
|
|
|
|
* We waited 1 sec, if a CPU still did not call idle
|
|
|
|
* it may be because it is in idle and not waking up
|
|
|
|
* because it has nothing to do.
|
|
|
|
* Give all the remaining CPUS a kick.
|
|
|
|
*/
|
|
|
|
smp_call_function_mask(map, do_nothing, 0, 0);
|
2005-04-16 15:20:36 -07:00
|
|
|
} while (!cpus_empty(map));
|
2006-11-17 06:26:18 -07:00
|
|
|
|
|
|
|
set_cpus_allowed(current, tmp);
|
2005-04-16 15:20:36 -07:00
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(cpu_idle_wait);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* This uses new MONITOR/MWAIT instructions on P4 processors with PNI,
|
|
|
|
* which can obviate IPI to trigger checking of need_resched.
|
|
|
|
* We execute MONITOR against need_resched and enter optimized wait state
|
|
|
|
* through MWAIT. Whenever someone changes need_resched, we would be woken
|
|
|
|
* up from MWAIT (without an IPI).
|
2006-09-25 16:28:13 -07:00
|
|
|
*
|
|
|
|
* New with Core Duo processors, MWAIT can take some hints based on CPU
|
|
|
|
* capability.
|
2005-04-16 15:20:36 -07:00
|
|
|
*/
|
2008-01-30 05:30:56 -07:00
|
|
|
void mwait_idle_with_hints(unsigned long ax, unsigned long cx)
|
2005-04-16 15:20:36 -07:00
|
|
|
{
|
2006-09-25 16:28:13 -07:00
|
|
|
if (!need_resched()) {
|
[PATCH] sched: resched and cpu_idle rework
Make some changes to the NEED_RESCHED and POLLING_NRFLAG to reduce
confusion, and make their semantics rigid. Improves efficiency of
resched_task and some cpu_idle routines.
* In resched_task:
- TIF_NEED_RESCHED is only cleared with the task's runqueue lock held,
and as we hold it during resched_task, then there is no need for an
atomic test and set there. The only other time this should be set is
when the task's quantum expires, in the timer interrupt - this is
protected against because the rq lock is irq-safe.
- If TIF_NEED_RESCHED is set, then we don't need to do anything. It
won't get unset until the task get's schedule()d off.
- If we are running on the same CPU as the task we resched, then set
TIF_NEED_RESCHED and no further action is required.
- If we are running on another CPU, and TIF_POLLING_NRFLAG is *not* set
after TIF_NEED_RESCHED has been set, then we need to send an IPI.
Using these rules, we are able to remove the test and set operation in
resched_task, and make clear the previously vague semantics of
POLLING_NRFLAG.
* In idle routines:
- Enter cpu_idle with preempt disabled. When the need_resched() condition
becomes true, explicitly call schedule(). This makes things a bit clearer
(IMO), but haven't updated all architectures yet.
- Many do a test and clear of TIF_NEED_RESCHED for some reason. According
to the resched_task rules, this isn't needed (and actually breaks the
assumption that TIF_NEED_RESCHED is only cleared with the runqueue lock
held). So remove that. Generally one less locked memory op when switching
to the idle thread.
- Many idle routines clear TIF_POLLING_NRFLAG, and only set it in the inner
most polling idle loops. The above resched_task semantics allow it to be
set until before the last time need_resched() is checked before going into
a halt requiring interrupt wakeup.
Many idle routines simply never enter such a halt, and so POLLING_NRFLAG
can be always left set, completely eliminating resched IPIs when rescheduling
the idle task.
POLLING_NRFLAG width can be increased, to reduce the chance of resched IPIs.
Signed-off-by: Nick Piggin <npiggin@suse.de>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Con Kolivas <kernel@kolivas.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-11-08 22:39:04 -07:00
|
|
|
__monitor((void *)¤t_thread_info()->flags, 0, 0);
|
|
|
|
smp_mb();
|
2006-09-25 16:28:13 -07:00
|
|
|
if (!need_resched())
|
2008-01-30 05:30:56 -07:00
|
|
|
__mwait(ax, cx);
|
2005-04-16 15:20:36 -07:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2006-09-25 16:28:13 -07:00
|
|
|
/* Default MONITOR/MWAIT with no hints, used for default C1 state */
|
|
|
|
static void mwait_idle(void)
|
|
|
|
{
|
|
|
|
local_irq_enable();
|
2006-12-06 18:14:03 -07:00
|
|
|
mwait_idle_with_hints(0, 0);
|
2006-09-25 16:28:13 -07:00
|
|
|
}
|
|
|
|
|
2008-01-30 05:33:16 -07:00
|
|
|
static int mwait_usable(const struct cpuinfo_x86 *c)
|
|
|
|
{
|
|
|
|
if (force_mwait)
|
|
|
|
return 1;
|
|
|
|
/* Any C1 states supported? */
|
|
|
|
return c->cpuid_level >= 5 && ((cpuid_edx(5) >> 4) & 0xf) > 0;
|
|
|
|
}
|
|
|
|
|
2007-12-19 15:20:18 -07:00
|
|
|
void __cpuinit select_idle_routine(const struct cpuinfo_x86 *c)
|
2005-04-16 15:20:36 -07:00
|
|
|
{
|
2008-01-30 05:33:16 -07:00
|
|
|
if (cpu_has(c, X86_FEATURE_MWAIT) && mwait_usable(c)) {
|
2005-04-16 15:20:36 -07:00
|
|
|
printk("monitor/mwait feature present.\n");
|
|
|
|
/*
|
|
|
|
* Skip, if setup has overridden idle.
|
|
|
|
* One CPU supports mwait => All CPUs supports mwait
|
|
|
|
*/
|
|
|
|
if (!pm_idle) {
|
|
|
|
printk("using mwait in idle threads.\n");
|
|
|
|
pm_idle = mwait_idle;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2007-05-02 10:27:12 -07:00
|
|
|
static int __init idle_setup(char *str)
|
2005-04-16 15:20:36 -07:00
|
|
|
{
|
2007-05-02 10:27:12 -07:00
|
|
|
if (!strcmp(str, "poll")) {
|
2005-04-16 15:20:36 -07:00
|
|
|
printk("using polling idle threads.\n");
|
|
|
|
pm_idle = poll_idle;
|
|
|
|
#ifdef CONFIG_X86_SMP
|
|
|
|
if (smp_num_siblings > 1)
|
|
|
|
printk("WARNING: polling idle and HT enabled, performance may degrade.\n");
|
|
|
|
#endif
|
2007-05-02 10:27:12 -07:00
|
|
|
} else if (!strcmp(str, "mwait"))
|
|
|
|
force_mwait = 1;
|
|
|
|
else
|
|
|
|
return -1;
|
2005-04-16 15:20:36 -07:00
|
|
|
|
|
|
|
boot_option_idle_override = 1;
|
2007-05-02 10:27:12 -07:00
|
|
|
return 0;
|
2005-04-16 15:20:36 -07:00
|
|
|
}
|
2007-05-02 10:27:12 -07:00
|
|
|
early_param("idle", idle_setup);
|
2005-04-16 15:20:36 -07:00
|
|
|
|
2007-10-19 11:35:03 -07:00
|
|
|
void __show_registers(struct pt_regs *regs, int all)
|
2005-04-16 15:20:36 -07:00
|
|
|
{
|
|
|
|
unsigned long cr0 = 0L, cr2 = 0L, cr3 = 0L, cr4 = 0L;
|
2007-07-21 08:10:42 -07:00
|
|
|
unsigned long d0, d1, d2, d3, d6, d7;
|
2008-01-30 05:30:56 -07:00
|
|
|
unsigned long sp;
|
2007-10-19 11:35:03 -07:00
|
|
|
unsigned short ss, gs;
|
|
|
|
|
|
|
|
if (user_mode_vm(regs)) {
|
2008-01-30 05:30:56 -07:00
|
|
|
sp = regs->sp;
|
|
|
|
ss = regs->ss & 0xffff;
|
2007-10-19 11:35:03 -07:00
|
|
|
savesegment(gs, gs);
|
|
|
|
} else {
|
2008-01-30 05:30:56 -07:00
|
|
|
sp = (unsigned long) (®s->sp);
|
2007-10-19 11:35:03 -07:00
|
|
|
savesegment(ss, ss);
|
|
|
|
savesegment(gs, gs);
|
|
|
|
}
|
2005-04-16 15:20:36 -07:00
|
|
|
|
|
|
|
printk("\n");
|
2007-10-19 15:06:00 -07:00
|
|
|
printk("Pid: %d, comm: %s %s (%s %.*s)\n",
|
|
|
|
task_pid_nr(current), current->comm,
|
2007-10-19 11:35:03 -07:00
|
|
|
print_tainted(), init_utsname()->release,
|
|
|
|
(int)strcspn(init_utsname()->version, " "),
|
|
|
|
init_utsname()->version);
|
|
|
|
|
|
|
|
printk("EIP: %04x:[<%08lx>] EFLAGS: %08lx CPU: %d\n",
|
2008-01-30 05:30:56 -07:00
|
|
|
0xffff & regs->cs, regs->ip, regs->flags,
|
2007-10-19 11:35:03 -07:00
|
|
|
smp_processor_id());
|
2008-01-30 05:30:56 -07:00
|
|
|
print_symbol("EIP is at %s\n", regs->ip);
|
2005-04-16 15:20:36 -07:00
|
|
|
|
|
|
|
printk("EAX: %08lx EBX: %08lx ECX: %08lx EDX: %08lx\n",
|
2008-01-30 05:30:56 -07:00
|
|
|
regs->ax, regs->bx, regs->cx, regs->dx);
|
2007-10-19 11:35:03 -07:00
|
|
|
printk("ESI: %08lx EDI: %08lx EBP: %08lx ESP: %08lx\n",
|
2008-01-30 05:30:56 -07:00
|
|
|
regs->si, regs->di, regs->bp, sp);
|
2007-10-19 11:35:03 -07:00
|
|
|
printk(" DS: %04x ES: %04x FS: %04x GS: %04x SS: %04x\n",
|
2008-01-30 05:30:56 -07:00
|
|
|
regs->ds & 0xffff, regs->es & 0xffff,
|
|
|
|
regs->fs & 0xffff, gs, ss);
|
2007-10-19 11:35:03 -07:00
|
|
|
|
|
|
|
if (!all)
|
|
|
|
return;
|
2005-04-16 15:20:36 -07:00
|
|
|
|
2005-09-03 15:56:36 -07:00
|
|
|
cr0 = read_cr0();
|
|
|
|
cr2 = read_cr2();
|
|
|
|
cr3 = read_cr3();
|
2006-01-06 01:11:50 -07:00
|
|
|
cr4 = read_cr4_safe();
|
2007-10-19 11:35:03 -07:00
|
|
|
printk("CR0: %08lx CR2: %08lx CR3: %08lx CR4: %08lx\n",
|
|
|
|
cr0, cr2, cr3, cr4);
|
2007-07-21 08:10:42 -07:00
|
|
|
|
|
|
|
get_debugreg(d0, 0);
|
|
|
|
get_debugreg(d1, 1);
|
|
|
|
get_debugreg(d2, 2);
|
|
|
|
get_debugreg(d3, 3);
|
|
|
|
printk("DR0: %08lx DR1: %08lx DR2: %08lx DR3: %08lx\n",
|
|
|
|
d0, d1, d2, d3);
|
2007-10-19 11:35:03 -07:00
|
|
|
|
2007-07-21 08:10:42 -07:00
|
|
|
get_debugreg(d6, 6);
|
|
|
|
get_debugreg(d7, 7);
|
2007-10-19 11:35:03 -07:00
|
|
|
printk("DR6: %08lx DR7: %08lx\n",
|
|
|
|
d6, d7);
|
|
|
|
}
|
2007-07-21 08:10:42 -07:00
|
|
|
|
2007-10-19 11:35:03 -07:00
|
|
|
void show_regs(struct pt_regs *regs)
|
|
|
|
{
|
|
|
|
__show_registers(regs, 1);
|
2008-01-30 05:33:07 -07:00
|
|
|
show_trace(NULL, regs, ®s->sp, regs->bp);
|
2005-04-16 15:20:36 -07:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
2008-01-30 05:30:56 -07:00
|
|
|
* This gets run with %bx containing the
|
|
|
|
* function to call, and %dx containing
|
2005-04-16 15:20:36 -07:00
|
|
|
* the "args".
|
|
|
|
*/
|
|
|
|
extern void kernel_thread_helper(void);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Create a kernel thread
|
|
|
|
*/
|
|
|
|
int kernel_thread(int (*fn)(void *), void * arg, unsigned long flags)
|
|
|
|
{
|
|
|
|
struct pt_regs regs;
|
|
|
|
|
|
|
|
memset(®s, 0, sizeof(regs));
|
|
|
|
|
2008-01-30 05:30:56 -07:00
|
|
|
regs.bx = (unsigned long) fn;
|
|
|
|
regs.dx = (unsigned long) arg;
|
2005-04-16 15:20:36 -07:00
|
|
|
|
2008-01-30 05:30:56 -07:00
|
|
|
regs.ds = __USER_DS;
|
|
|
|
regs.es = __USER_DS;
|
|
|
|
regs.fs = __KERNEL_PERCPU;
|
|
|
|
regs.orig_ax = -1;
|
|
|
|
regs.ip = (unsigned long) kernel_thread_helper;
|
|
|
|
regs.cs = __KERNEL_CS | get_kernel_rpl();
|
|
|
|
regs.flags = X86_EFLAGS_IF | X86_EFLAGS_SF | X86_EFLAGS_PF | 0x2;
|
2005-04-16 15:20:36 -07:00
|
|
|
|
|
|
|
/* Ok, create the new process.. */
|
2006-10-21 09:37:02 -07:00
|
|
|
return do_fork(flags | CLONE_VM | CLONE_UNTRACED, 0, ®s, 0, NULL, NULL);
|
2005-04-16 15:20:36 -07:00
|
|
|
}
|
2005-06-23 00:08:33 -07:00
|
|
|
EXPORT_SYMBOL(kernel_thread);
|
2005-04-16 15:20:36 -07:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Free current thread data structures etc..
|
|
|
|
*/
|
|
|
|
void exit_thread(void)
|
|
|
|
{
|
|
|
|
/* The process may have allocated an io port bitmap... nuke it. */
|
2006-07-09 18:12:39 -07:00
|
|
|
if (unlikely(test_thread_flag(TIF_IO_BITMAP))) {
|
|
|
|
struct task_struct *tsk = current;
|
|
|
|
struct thread_struct *t = &tsk->thread;
|
2005-04-16 15:20:36 -07:00
|
|
|
int cpu = get_cpu();
|
|
|
|
struct tss_struct *tss = &per_cpu(init_tss, cpu);
|
|
|
|
|
|
|
|
kfree(t->io_bitmap_ptr);
|
|
|
|
t->io_bitmap_ptr = NULL;
|
2006-07-09 18:12:39 -07:00
|
|
|
clear_thread_flag(TIF_IO_BITMAP);
|
2005-04-16 15:20:36 -07:00
|
|
|
/*
|
|
|
|
* Careful, clear this in the TSS too:
|
|
|
|
*/
|
|
|
|
memset(tss->io_bitmap, 0xff, tss->io_bitmap_max);
|
|
|
|
t->io_bitmap_max = 0;
|
|
|
|
tss->io_bitmap_owner = NULL;
|
|
|
|
tss->io_bitmap_max = 0;
|
2007-05-02 10:27:13 -07:00
|
|
|
tss->x86_tss.io_bitmap_base = INVALID_IO_BITMAP_OFFSET;
|
2005-04-16 15:20:36 -07:00
|
|
|
put_cpu();
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
void flush_thread(void)
|
|
|
|
{
|
|
|
|
struct task_struct *tsk = current;
|
|
|
|
|
2008-01-30 05:30:59 -07:00
|
|
|
tsk->thread.debugreg0 = 0;
|
|
|
|
tsk->thread.debugreg1 = 0;
|
|
|
|
tsk->thread.debugreg2 = 0;
|
|
|
|
tsk->thread.debugreg3 = 0;
|
|
|
|
tsk->thread.debugreg6 = 0;
|
|
|
|
tsk->thread.debugreg7 = 0;
|
2005-04-16 15:20:36 -07:00
|
|
|
memset(tsk->thread.tls_array, 0, sizeof(tsk->thread.tls_array));
|
2006-07-09 18:12:39 -07:00
|
|
|
clear_tsk_thread_flag(tsk, TIF_DEBUG);
|
2005-04-16 15:20:36 -07:00
|
|
|
/*
|
|
|
|
* Forget coprocessor state..
|
|
|
|
*/
|
|
|
|
clear_fpu(tsk);
|
|
|
|
clear_used_math();
|
|
|
|
}
|
|
|
|
|
|
|
|
void release_thread(struct task_struct *dead_task)
|
|
|
|
{
|
2006-01-06 01:11:59 -07:00
|
|
|
BUG_ON(dead_task->mm);
|
2005-04-16 15:20:36 -07:00
|
|
|
release_vm86_irqs(dead_task);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* This gets called before we allocate a new thread and copy
|
|
|
|
* the current task into it.
|
|
|
|
*/
|
|
|
|
void prepare_to_copy(struct task_struct *tsk)
|
|
|
|
{
|
|
|
|
unlazy_fpu(tsk);
|
|
|
|
}
|
|
|
|
|
2008-01-30 05:30:56 -07:00
|
|
|
int copy_thread(int nr, unsigned long clone_flags, unsigned long sp,
|
2005-04-16 15:20:36 -07:00
|
|
|
unsigned long unused,
|
|
|
|
struct task_struct * p, struct pt_regs * regs)
|
|
|
|
{
|
|
|
|
struct pt_regs * childregs;
|
|
|
|
struct task_struct *tsk;
|
|
|
|
int err;
|
|
|
|
|
2006-01-12 02:05:41 -07:00
|
|
|
childregs = task_pt_regs(p);
|
2005-05-05 16:15:03 -07:00
|
|
|
*childregs = *regs;
|
2008-01-30 05:30:56 -07:00
|
|
|
childregs->ax = 0;
|
|
|
|
childregs->sp = sp;
|
2005-05-05 16:15:03 -07:00
|
|
|
|
2008-01-30 05:31:02 -07:00
|
|
|
p->thread.sp = (unsigned long) childregs;
|
|
|
|
p->thread.sp0 = (unsigned long) (childregs+1);
|
2005-04-16 15:20:36 -07:00
|
|
|
|
2008-01-30 05:31:02 -07:00
|
|
|
p->thread.ip = (unsigned long) ret_from_fork;
|
2005-04-16 15:20:36 -07:00
|
|
|
|
2008-01-30 05:31:03 -07:00
|
|
|
savesegment(gs, p->thread.gs);
|
2005-04-16 15:20:36 -07:00
|
|
|
|
|
|
|
tsk = current;
|
2006-07-09 18:12:39 -07:00
|
|
|
if (unlikely(test_tsk_thread_flag(tsk, TIF_IO_BITMAP))) {
|
2006-09-30 23:27:21 -07:00
|
|
|
p->thread.io_bitmap_ptr = kmemdup(tsk->thread.io_bitmap_ptr,
|
|
|
|
IO_BITMAP_BYTES, GFP_KERNEL);
|
2005-04-16 15:20:36 -07:00
|
|
|
if (!p->thread.io_bitmap_ptr) {
|
|
|
|
p->thread.io_bitmap_max = 0;
|
|
|
|
return -ENOMEM;
|
|
|
|
}
|
2006-07-09 18:12:39 -07:00
|
|
|
set_tsk_thread_flag(p, TIF_IO_BITMAP);
|
2005-04-16 15:20:36 -07:00
|
|
|
}
|
|
|
|
|
2008-01-30 05:30:46 -07:00
|
|
|
err = 0;
|
|
|
|
|
2005-04-16 15:20:36 -07:00
|
|
|
/*
|
|
|
|
* Set a new TLS for the child thread?
|
|
|
|
*/
|
2008-01-30 05:30:46 -07:00
|
|
|
if (clone_flags & CLONE_SETTLS)
|
|
|
|
err = do_set_thread_area(p, -1,
|
2008-01-30 05:30:56 -07:00
|
|
|
(struct user_desc __user *)childregs->si, 0);
|
2005-04-16 15:20:36 -07:00
|
|
|
|
|
|
|
if (err && p->thread.io_bitmap_ptr) {
|
|
|
|
kfree(p->thread.io_bitmap_ptr);
|
|
|
|
p->thread.io_bitmap_max = 0;
|
|
|
|
}
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* fill in the user structure for a core dump..
|
|
|
|
*/
|
|
|
|
void dump_thread(struct pt_regs * regs, struct user * dump)
|
|
|
|
{
|
2008-01-30 05:30:56 -07:00
|
|
|
u16 gs;
|
2005-04-16 15:20:36 -07:00
|
|
|
|
|
|
|
/* changed the size calculations - should hopefully work better. lbt */
|
|
|
|
dump->magic = CMAGIC;
|
|
|
|
dump->start_code = 0;
|
2008-01-30 05:30:56 -07:00
|
|
|
dump->start_stack = regs->sp & ~(PAGE_SIZE - 1);
|
2005-04-16 15:20:36 -07:00
|
|
|
dump->u_tsize = ((unsigned long) current->mm->end_code) >> PAGE_SHIFT;
|
|
|
|
dump->u_dsize = ((unsigned long) (current->mm->brk + (PAGE_SIZE-1))) >> PAGE_SHIFT;
|
|
|
|
dump->u_dsize -= dump->u_tsize;
|
|
|
|
dump->u_ssize = 0;
|
2008-01-30 05:30:59 -07:00
|
|
|
dump->u_debugreg[0] = current->thread.debugreg0;
|
|
|
|
dump->u_debugreg[1] = current->thread.debugreg1;
|
|
|
|
dump->u_debugreg[2] = current->thread.debugreg2;
|
|
|
|
dump->u_debugreg[3] = current->thread.debugreg3;
|
|
|
|
dump->u_debugreg[4] = 0;
|
|
|
|
dump->u_debugreg[5] = 0;
|
|
|
|
dump->u_debugreg[6] = current->thread.debugreg6;
|
|
|
|
dump->u_debugreg[7] = current->thread.debugreg7;
|
2005-04-16 15:20:36 -07:00
|
|
|
|
|
|
|
if (dump->start_stack < TASK_SIZE)
|
|
|
|
dump->u_ssize = ((unsigned long) (TASK_SIZE - dump->start_stack)) >> PAGE_SHIFT;
|
|
|
|
|
2008-01-30 05:30:56 -07:00
|
|
|
dump->regs.bx = regs->bx;
|
|
|
|
dump->regs.cx = regs->cx;
|
|
|
|
dump->regs.dx = regs->dx;
|
|
|
|
dump->regs.si = regs->si;
|
|
|
|
dump->regs.di = regs->di;
|
|
|
|
dump->regs.bp = regs->bp;
|
|
|
|
dump->regs.ax = regs->ax;
|
|
|
|
dump->regs.ds = (u16)regs->ds;
|
|
|
|
dump->regs.es = (u16)regs->es;
|
|
|
|
dump->regs.fs = (u16)regs->fs;
|
|
|
|
savesegment(gs,gs);
|
|
|
|
dump->regs.orig_ax = regs->orig_ax;
|
|
|
|
dump->regs.ip = regs->ip;
|
|
|
|
dump->regs.cs = (u16)regs->cs;
|
|
|
|
dump->regs.flags = regs->flags;
|
|
|
|
dump->regs.sp = regs->sp;
|
|
|
|
dump->regs.ss = (u16)regs->ss;
|
2005-04-16 15:20:36 -07:00
|
|
|
|
|
|
|
dump->u_fpvalid = dump_fpu (regs, &dump->i387);
|
|
|
|
}
|
2005-06-23 00:08:33 -07:00
|
|
|
EXPORT_SYMBOL(dump_thread);
|
2005-04-16 15:20:36 -07:00
|
|
|
|
2007-07-15 23:41:33 -07:00
|
|
|
#ifdef CONFIG_SECCOMP
|
2008-01-30 05:31:21 -07:00
|
|
|
static void hard_disable_TSC(void)
|
2007-07-15 23:41:33 -07:00
|
|
|
{
|
|
|
|
write_cr4(read_cr4() | X86_CR4_TSD);
|
|
|
|
}
|
|
|
|
void disable_TSC(void)
|
|
|
|
{
|
|
|
|
preempt_disable();
|
|
|
|
if (!test_and_set_thread_flag(TIF_NOTSC))
|
|
|
|
/*
|
|
|
|
* Must flip the CPU state synchronously with
|
|
|
|
* TIF_NOTSC in the current running context.
|
|
|
|
*/
|
|
|
|
hard_disable_TSC();
|
|
|
|
preempt_enable();
|
|
|
|
}
|
2008-01-30 05:31:21 -07:00
|
|
|
static void hard_enable_TSC(void)
|
2007-07-15 23:41:33 -07:00
|
|
|
{
|
|
|
|
write_cr4(read_cr4() & ~X86_CR4_TSD);
|
|
|
|
}
|
|
|
|
#endif /* CONFIG_SECCOMP */
|
|
|
|
|
|
|
|
static noinline void
|
|
|
|
__switch_to_xtra(struct task_struct *prev_p, struct task_struct *next_p,
|
|
|
|
struct tss_struct *tss)
|
2005-04-16 15:20:36 -07:00
|
|
|
{
|
2008-01-30 05:30:54 -07:00
|
|
|
struct thread_struct *prev, *next;
|
2008-01-30 05:31:09 -07:00
|
|
|
unsigned long debugctl;
|
2006-07-09 18:12:39 -07:00
|
|
|
|
2008-01-30 05:30:54 -07:00
|
|
|
prev = &prev_p->thread;
|
2006-07-09 18:12:39 -07:00
|
|
|
next = &next_p->thread;
|
|
|
|
|
2008-01-30 05:31:09 -07:00
|
|
|
debugctl = prev->debugctlmsr;
|
|
|
|
if (next->ds_area_msr != prev->ds_area_msr) {
|
|
|
|
/* we clear debugctl to make sure DS
|
|
|
|
* is not in use when we change it */
|
|
|
|
debugctl = 0;
|
|
|
|
wrmsrl(MSR_IA32_DEBUGCTLMSR, 0);
|
|
|
|
wrmsr(MSR_IA32_DS_AREA, next->ds_area_msr, 0);
|
|
|
|
}
|
|
|
|
|
|
|
|
if (next->debugctlmsr != debugctl)
|
2008-01-30 05:30:54 -07:00
|
|
|
wrmsr(MSR_IA32_DEBUGCTLMSR, next->debugctlmsr, 0);
|
|
|
|
|
2006-07-09 18:12:39 -07:00
|
|
|
if (test_tsk_thread_flag(next_p, TIF_DEBUG)) {
|
2008-01-30 05:30:59 -07:00
|
|
|
set_debugreg(next->debugreg0, 0);
|
|
|
|
set_debugreg(next->debugreg1, 1);
|
|
|
|
set_debugreg(next->debugreg2, 2);
|
|
|
|
set_debugreg(next->debugreg3, 3);
|
2006-07-09 18:12:39 -07:00
|
|
|
/* no 4 and 5 */
|
2008-01-30 05:30:59 -07:00
|
|
|
set_debugreg(next->debugreg6, 6);
|
|
|
|
set_debugreg(next->debugreg7, 7);
|
2006-07-09 18:12:39 -07:00
|
|
|
}
|
|
|
|
|
2007-07-15 23:41:33 -07:00
|
|
|
#ifdef CONFIG_SECCOMP
|
|
|
|
if (test_tsk_thread_flag(prev_p, TIF_NOTSC) ^
|
|
|
|
test_tsk_thread_flag(next_p, TIF_NOTSC)) {
|
|
|
|
/* prev and next are different */
|
|
|
|
if (test_tsk_thread_flag(next_p, TIF_NOTSC))
|
|
|
|
hard_disable_TSC();
|
|
|
|
else
|
|
|
|
hard_enable_TSC();
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
|
2008-01-30 05:31:09 -07:00
|
|
|
if (test_tsk_thread_flag(prev_p, TIF_BTS_TRACE_TS))
|
|
|
|
ptrace_bts_take_timestamp(prev_p, BTS_TASK_DEPARTS);
|
|
|
|
|
|
|
|
if (test_tsk_thread_flag(next_p, TIF_BTS_TRACE_TS))
|
|
|
|
ptrace_bts_take_timestamp(next_p, BTS_TASK_ARRIVES);
|
|
|
|
|
|
|
|
|
2006-07-09 18:12:39 -07:00
|
|
|
if (!test_tsk_thread_flag(next_p, TIF_IO_BITMAP)) {
|
2005-04-16 15:20:36 -07:00
|
|
|
/*
|
|
|
|
* Disable the bitmap via an invalid offset. We still cache
|
|
|
|
* the previous bitmap owner and the IO bitmap contents:
|
|
|
|
*/
|
2007-05-02 10:27:13 -07:00
|
|
|
tss->x86_tss.io_bitmap_base = INVALID_IO_BITMAP_OFFSET;
|
2005-04-16 15:20:36 -07:00
|
|
|
return;
|
|
|
|
}
|
2006-07-09 18:12:39 -07:00
|
|
|
|
2005-04-16 15:20:36 -07:00
|
|
|
if (likely(next == tss->io_bitmap_owner)) {
|
|
|
|
/*
|
|
|
|
* Previous owner of the bitmap (hence the bitmap content)
|
|
|
|
* matches the next task, we dont have to do anything but
|
|
|
|
* to set a valid offset in the TSS:
|
|
|
|
*/
|
2007-05-02 10:27:13 -07:00
|
|
|
tss->x86_tss.io_bitmap_base = IO_BITMAP_OFFSET;
|
2005-04-16 15:20:36 -07:00
|
|
|
return;
|
|
|
|
}
|
|
|
|
/*
|
|
|
|
* Lazy TSS's I/O bitmap copy. We set an invalid offset here
|
|
|
|
* and we let the task to get a GPF in case an I/O instruction
|
|
|
|
* is performed. The handler of the GPF will verify that the
|
|
|
|
* faulting task has a valid I/O bitmap and, it true, does the
|
|
|
|
* real copy and restart the instruction. This will save us
|
|
|
|
* redundant copies when the currently switched task does not
|
|
|
|
* perform any I/O during its timeslice.
|
|
|
|
*/
|
2007-05-02 10:27:13 -07:00
|
|
|
tss->x86_tss.io_bitmap_base = INVALID_IO_BITMAP_OFFSET_LAZY;
|
2005-04-16 15:20:36 -07:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* switch_to(x,yn) should switch tasks from x to y.
|
|
|
|
*
|
|
|
|
* We fsave/fwait so that an exception goes off at the right time
|
|
|
|
* (as a call from the fsave or fwait in effect) rather than to
|
|
|
|
* the wrong process. Lazy FP saving no longer makes any sense
|
|
|
|
* with modern CPU's, and this simplifies a lot of things (SMP
|
|
|
|
* and UP become the same).
|
|
|
|
*
|
|
|
|
* NOTE! We used to use the x86 hardware context switching. The
|
|
|
|
* reason for not using it any more becomes apparent when you
|
|
|
|
* try to recover gracefully from saved state that is no longer
|
|
|
|
* valid (stale segment register values in particular). With the
|
|
|
|
* hardware task-switch, there is no way to fix up bad state in
|
|
|
|
* a reasonable manner.
|
|
|
|
*
|
|
|
|
* The fact that Intel documents the hardware task-switching to
|
|
|
|
* be slow is a fairly red herring - this code is not noticeably
|
|
|
|
* faster. However, there _is_ some room for improvement here,
|
|
|
|
* so the performance issues may eventually be a valid point.
|
|
|
|
* More important, however, is the fact that this allows us much
|
|
|
|
* more flexibility.
|
|
|
|
*
|
2008-01-30 05:30:56 -07:00
|
|
|
* The return value (in %ax) will be the "prev" task after
|
2005-04-16 15:20:36 -07:00
|
|
|
* the task-switch, and shows up in ret_from_fork in entry.S,
|
|
|
|
* for example.
|
|
|
|
*/
|
2008-01-30 05:31:17 -07:00
|
|
|
struct task_struct * __switch_to(struct task_struct *prev_p, struct task_struct *next_p)
|
2005-04-16 15:20:36 -07:00
|
|
|
{
|
|
|
|
struct thread_struct *prev = &prev_p->thread,
|
|
|
|
*next = &next_p->thread;
|
|
|
|
int cpu = smp_processor_id();
|
|
|
|
struct tss_struct *tss = &per_cpu(init_tss, cpu);
|
|
|
|
|
|
|
|
/* never put a printk in __switch_to... printk() calls wake_up*() indirectly */
|
|
|
|
|
|
|
|
__unlazy_fpu(prev_p);
|
|
|
|
|
2006-12-06 18:14:01 -07:00
|
|
|
|
|
|
|
/* we're going to use this soon, after a few expensive things */
|
|
|
|
if (next_p->fpu_counter > 5)
|
|
|
|
prefetch(&next->i387.fxsave);
|
|
|
|
|
2005-04-16 15:20:36 -07:00
|
|
|
/*
|
2005-09-03 15:56:39 -07:00
|
|
|
* Reload esp0.
|
2005-04-16 15:20:36 -07:00
|
|
|
*/
|
2008-01-30 05:31:02 -07:00
|
|
|
load_sp0(tss, next);
|
2005-04-16 15:20:36 -07:00
|
|
|
|
|
|
|
/*
|
2007-02-13 05:26:20 -07:00
|
|
|
* Save away %gs. No need to save %fs, as it was saved on the
|
[PATCH] i386: Use %gs as the PDA base-segment in the kernel
This patch is the meat of the PDA change. This patch makes several related
changes:
1: Most significantly, %gs is now used in the kernel. This means that on
entry, the old value of %gs is saved away, and it is reloaded with
__KERNEL_PDA.
2: entry.S constructs the stack in the shape of struct pt_regs, and this
is passed around the kernel so that the process's saved register
state can be accessed.
Unfortunately struct pt_regs doesn't currently have space for %gs
(or %fs). This patch extends pt_regs to add space for gs (no space
is allocated for %fs, since it won't be used, and it would just
complicate the code in entry.S to work around the space).
3: Because %gs is now saved on the stack like %ds, %es and the integer
registers, there are a number of places where it no longer needs to
be handled specially; namely context switch, and saving/restoring the
register state in a signal context.
4: And since kernel threads run in kernel space and call normal kernel
code, they need to be created with their %gs == __KERNEL_PDA.
Signed-off-by: Jeremy Fitzhardinge <jeremy@xensource.com>
Signed-off-by: Andi Kleen <ak@suse.de>
Cc: Chuck Ebbert <76306.1226@compuserve.com>
Cc: Zachary Amsden <zach@vmware.com>
Cc: Jan Beulich <jbeulich@novell.com>
Cc: Andi Kleen <ak@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
2006-12-06 18:14:02 -07:00
|
|
|
* stack on entry. No need to save %es and %ds, as those are
|
|
|
|
* always kernel segments while inside the kernel. Doing this
|
|
|
|
* before setting the new TLS descriptors avoids the situation
|
|
|
|
* where we temporarily have non-reloadable segments in %fs
|
|
|
|
* and %gs. This could be an issue if the NMI handler ever
|
|
|
|
* used %fs or %gs (it does not today), or if the kernel is
|
|
|
|
* running inside of a hypervisor layer.
|
2005-04-16 15:20:36 -07:00
|
|
|
*/
|
2007-02-13 05:26:20 -07:00
|
|
|
savesegment(gs, prev->gs);
|
2005-04-16 15:20:36 -07:00
|
|
|
|
|
|
|
/*
|
2005-09-03 15:56:39 -07:00
|
|
|
* Load the per-thread Thread-Local Storage descriptor.
|
2005-04-16 15:20:36 -07:00
|
|
|
*/
|
2005-09-03 15:56:39 -07:00
|
|
|
load_TLS(next, cpu);
|
2005-04-16 15:20:36 -07:00
|
|
|
|
2007-02-13 05:26:21 -07:00
|
|
|
/*
|
|
|
|
* Restore IOPL if needed. In normal use, the flags restore
|
|
|
|
* in the switch assembly will handle this. But if the kernel
|
|
|
|
* is running virtualized at a non-zero CPL, the popf will
|
|
|
|
* not restore flags, so it must be done in a separate step.
|
|
|
|
*/
|
|
|
|
if (get_kernel_rpl() && unlikely(prev->iopl != next->iopl))
|
|
|
|
set_iopl_mask(next->iopl);
|
|
|
|
|
2005-04-16 15:20:36 -07:00
|
|
|
/*
|
2006-07-09 18:12:39 -07:00
|
|
|
* Now maybe handle debug registers and/or IO bitmaps
|
2005-04-16 15:20:36 -07:00
|
|
|
*/
|
2007-07-15 23:41:33 -07:00
|
|
|
if (unlikely(task_thread_info(prev_p)->flags & _TIF_WORK_CTXSW_PREV ||
|
|
|
|
task_thread_info(next_p)->flags & _TIF_WORK_CTXSW_NEXT))
|
|
|
|
__switch_to_xtra(prev_p, next_p, tss);
|
2005-06-27 14:36:36 -07:00
|
|
|
|
2007-02-13 05:26:21 -07:00
|
|
|
/*
|
|
|
|
* Leave lazy mode, flushing any hypercalls made here.
|
|
|
|
* This must be done before restoring TLS segments so
|
|
|
|
* the GDT and LDT are properly updated, and must be
|
|
|
|
* done before math_state_restore, so the TS bit is up
|
|
|
|
* to date.
|
|
|
|
*/
|
|
|
|
arch_leave_lazy_cpu_mode();
|
|
|
|
|
2006-12-06 18:14:01 -07:00
|
|
|
/* If the task has used fpu the last 5 timeslices, just do a full
|
|
|
|
* restore of the math state immediately to avoid the trap; the
|
|
|
|
* chances of needing FPU soon are obviously high now
|
|
|
|
*/
|
|
|
|
if (next_p->fpu_counter > 5)
|
|
|
|
math_state_restore();
|
|
|
|
|
2007-02-13 05:26:21 -07:00
|
|
|
/*
|
|
|
|
* Restore %gs if needed (which is common)
|
|
|
|
*/
|
|
|
|
if (prev->gs | next->gs)
|
|
|
|
loadsegment(gs, next->gs);
|
|
|
|
|
2007-05-02 10:27:16 -07:00
|
|
|
x86_write_percpu(current_task, next_p);
|
2007-02-13 05:26:21 -07:00
|
|
|
|
2005-04-16 15:20:36 -07:00
|
|
|
return prev_p;
|
|
|
|
}
|
|
|
|
|
|
|
|
asmlinkage int sys_fork(struct pt_regs regs)
|
|
|
|
{
|
2008-01-30 05:30:56 -07:00
|
|
|
return do_fork(SIGCHLD, regs.sp, ®s, 0, NULL, NULL);
|
2005-04-16 15:20:36 -07:00
|
|
|
}
|
|
|
|
|
|
|
|
asmlinkage int sys_clone(struct pt_regs regs)
|
|
|
|
{
|
|
|
|
unsigned long clone_flags;
|
|
|
|
unsigned long newsp;
|
|
|
|
int __user *parent_tidptr, *child_tidptr;
|
|
|
|
|
2008-01-30 05:30:56 -07:00
|
|
|
clone_flags = regs.bx;
|
|
|
|
newsp = regs.cx;
|
|
|
|
parent_tidptr = (int __user *)regs.dx;
|
|
|
|
child_tidptr = (int __user *)regs.di;
|
2005-04-16 15:20:36 -07:00
|
|
|
if (!newsp)
|
2008-01-30 05:30:56 -07:00
|
|
|
newsp = regs.sp;
|
2005-04-16 15:20:36 -07:00
|
|
|
return do_fork(clone_flags, newsp, ®s, 0, parent_tidptr, child_tidptr);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* This is trivial, and on the face of it looks like it
|
|
|
|
* could equally well be done in user mode.
|
|
|
|
*
|
|
|
|
* Not so, for quite unobvious reasons - register pressure.
|
|
|
|
* In user mode vfork() cannot have a stack frame, and if
|
|
|
|
* done by calling the "clone()" system call directly, you
|
|
|
|
* do not have enough call-clobbered registers to hold all
|
|
|
|
* the information you need.
|
|
|
|
*/
|
|
|
|
asmlinkage int sys_vfork(struct pt_regs regs)
|
|
|
|
{
|
2008-01-30 05:30:56 -07:00
|
|
|
return do_fork(CLONE_VFORK | CLONE_VM | SIGCHLD, regs.sp, ®s, 0, NULL, NULL);
|
2005-04-16 15:20:36 -07:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* sys_execve() executes a new program.
|
|
|
|
*/
|
|
|
|
asmlinkage int sys_execve(struct pt_regs regs)
|
|
|
|
{
|
|
|
|
int error;
|
|
|
|
char * filename;
|
|
|
|
|
2008-01-30 05:30:56 -07:00
|
|
|
filename = getname((char __user *) regs.bx);
|
2005-04-16 15:20:36 -07:00
|
|
|
error = PTR_ERR(filename);
|
|
|
|
if (IS_ERR(filename))
|
|
|
|
goto out;
|
|
|
|
error = do_execve(filename,
|
2008-01-30 05:30:56 -07:00
|
|
|
(char __user * __user *) regs.cx,
|
|
|
|
(char __user * __user *) regs.dx,
|
2005-04-16 15:20:36 -07:00
|
|
|
®s);
|
|
|
|
if (error == 0) {
|
|
|
|
/* Make sure we don't return using sysenter.. */
|
|
|
|
set_thread_flag(TIF_IRET);
|
|
|
|
}
|
|
|
|
putname(filename);
|
|
|
|
out:
|
|
|
|
return error;
|
|
|
|
}
|
|
|
|
|
|
|
|
#define top_esp (THREAD_SIZE - sizeof(unsigned long))
|
|
|
|
#define top_ebp (THREAD_SIZE - 2*sizeof(unsigned long))
|
|
|
|
|
|
|
|
unsigned long get_wchan(struct task_struct *p)
|
|
|
|
{
|
2008-01-30 05:30:56 -07:00
|
|
|
unsigned long bp, sp, ip;
|
2005-04-16 15:20:36 -07:00
|
|
|
unsigned long stack_page;
|
|
|
|
int count = 0;
|
|
|
|
if (!p || p == current || p->state == TASK_RUNNING)
|
|
|
|
return 0;
|
2006-01-12 02:05:41 -07:00
|
|
|
stack_page = (unsigned long)task_stack_page(p);
|
2008-01-30 05:31:02 -07:00
|
|
|
sp = p->thread.sp;
|
2008-01-30 05:30:56 -07:00
|
|
|
if (!stack_page || sp < stack_page || sp > top_esp+stack_page)
|
2005-04-16 15:20:36 -07:00
|
|
|
return 0;
|
2008-01-30 05:30:56 -07:00
|
|
|
/* include/asm-i386/system.h:switch_to() pushes bp last. */
|
|
|
|
bp = *(unsigned long *) sp;
|
2005-04-16 15:20:36 -07:00
|
|
|
do {
|
2008-01-30 05:30:56 -07:00
|
|
|
if (bp < stack_page || bp > top_ebp+stack_page)
|
2005-04-16 15:20:36 -07:00
|
|
|
return 0;
|
2008-01-30 05:30:56 -07:00
|
|
|
ip = *(unsigned long *) (bp+4);
|
|
|
|
if (!in_sched_functions(ip))
|
|
|
|
return ip;
|
|
|
|
bp = *(unsigned long *) bp;
|
2005-04-16 15:20:36 -07:00
|
|
|
} while (count++ < 16);
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
unsigned long arch_align_stack(unsigned long sp)
|
|
|
|
{
|
2006-09-26 01:52:28 -07:00
|
|
|
if (!(current->personality & ADDR_NO_RANDOMIZE) && randomize_va_space)
|
2005-04-16 15:20:36 -07:00
|
|
|
sp -= get_random_int() % 8192;
|
|
|
|
return sp & ~0xf;
|
|
|
|
}
|
2008-01-30 05:30:40 -07:00
|
|
|
|
|
|
|
unsigned long arch_randomize_brk(struct mm_struct *mm)
|
|
|
|
{
|
|
|
|
unsigned long range_end = mm->brk + 0x02000000;
|
|
|
|
return randomize_range(mm->brk, range_end, 0) ? : mm->brk;
|
|
|
|
}
|