2bbb6817c0
It is assumed that rcu won't be used once we switch to tickless mode and until we restart the tick. However this is not always true, as in x86-64 where we dereference the idle notifiers after the tick is stopped. To prepare for fixing this, add two new APIs: tick_nohz_idle_enter_norcu() and tick_nohz_idle_exit_norcu(). If no use of RCU is made in the idle loop between tick_nohz_enter_idle() and tick_nohz_exit_idle() calls, the arch must instead call the new *_norcu() version such that the arch doesn't need to call rcu_idle_enter() and rcu_idle_exit(). Otherwise the arch must call tick_nohz_enter_idle() and tick_nohz_exit_idle() and also call explicitly: - rcu_idle_enter() after its last use of RCU before the CPU is put to sleep. - rcu_idle_exit() before the first use of RCU after the CPU is woken up. Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Mike Frysinger <vapier@gentoo.org> Cc: Guan Xuetao <gxt@mprc.pku.edu.cn> Cc: David Miller <davem@davemloft.net> Cc: Chris Metcalf <cmetcalf@tilera.com> Cc: Hans-Christian Egtvedt <hans-christian.egtvedt@atmel.com> Cc: Ralf Baechle <ralf@linux-mips.org> Cc: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: Ingo Molnar <mingo@redhat.com> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: H. Peter Anvin <hpa@zytor.com> Cc: Russell King <linux@arm.linux.org.uk> Cc: Paul Mackerras <paulus@samba.org> Cc: Heiko Carstens <heiko.carstens@de.ibm.com> Cc: Paul Mundt <lethal@linux-sh.org> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
259 lines
6.8 KiB
C
259 lines
6.8 KiB
C
/*
|
|
* Copyright (C) 2008-2009 Michal Simek <monstr@monstr.eu>
|
|
* Copyright (C) 2008-2009 PetaLogix
|
|
* Copyright (C) 2006 Atmark Techno, Inc.
|
|
*
|
|
* This file is subject to the terms and conditions of the GNU General Public
|
|
* License. See the file "COPYING" in the main directory of this archive
|
|
* for more details.
|
|
*/
|
|
|
|
#include <linux/module.h>
|
|
#include <linux/sched.h>
|
|
#include <linux/pm.h>
|
|
#include <linux/tick.h>
|
|
#include <linux/bitops.h>
|
|
#include <asm/system.h>
|
|
#include <asm/pgalloc.h>
|
|
#include <asm/uaccess.h> /* for USER_DS macros */
|
|
#include <asm/cacheflush.h>
|
|
|
|
void show_regs(struct pt_regs *regs)
|
|
{
|
|
printk(KERN_INFO " Registers dump: mode=%X\r\n", regs->pt_mode);
|
|
printk(KERN_INFO " r1=%08lX, r2=%08lX, r3=%08lX, r4=%08lX\n",
|
|
regs->r1, regs->r2, regs->r3, regs->r4);
|
|
printk(KERN_INFO " r5=%08lX, r6=%08lX, r7=%08lX, r8=%08lX\n",
|
|
regs->r5, regs->r6, regs->r7, regs->r8);
|
|
printk(KERN_INFO " r9=%08lX, r10=%08lX, r11=%08lX, r12=%08lX\n",
|
|
regs->r9, regs->r10, regs->r11, regs->r12);
|
|
printk(KERN_INFO " r13=%08lX, r14=%08lX, r15=%08lX, r16=%08lX\n",
|
|
regs->r13, regs->r14, regs->r15, regs->r16);
|
|
printk(KERN_INFO " r17=%08lX, r18=%08lX, r19=%08lX, r20=%08lX\n",
|
|
regs->r17, regs->r18, regs->r19, regs->r20);
|
|
printk(KERN_INFO " r21=%08lX, r22=%08lX, r23=%08lX, r24=%08lX\n",
|
|
regs->r21, regs->r22, regs->r23, regs->r24);
|
|
printk(KERN_INFO " r25=%08lX, r26=%08lX, r27=%08lX, r28=%08lX\n",
|
|
regs->r25, regs->r26, regs->r27, regs->r28);
|
|
printk(KERN_INFO " r29=%08lX, r30=%08lX, r31=%08lX, rPC=%08lX\n",
|
|
regs->r29, regs->r30, regs->r31, regs->pc);
|
|
printk(KERN_INFO " msr=%08lX, ear=%08lX, esr=%08lX, fsr=%08lX\n",
|
|
regs->msr, regs->ear, regs->esr, regs->fsr);
|
|
}
|
|
|
|
void (*pm_idle)(void);
|
|
void (*pm_power_off)(void) = NULL;
|
|
EXPORT_SYMBOL(pm_power_off);
|
|
|
|
static int hlt_counter = 1;
|
|
|
|
void disable_hlt(void)
|
|
{
|
|
hlt_counter++;
|
|
}
|
|
EXPORT_SYMBOL(disable_hlt);
|
|
|
|
void enable_hlt(void)
|
|
{
|
|
hlt_counter--;
|
|
}
|
|
EXPORT_SYMBOL(enable_hlt);
|
|
|
|
static int __init nohlt_setup(char *__unused)
|
|
{
|
|
hlt_counter = 1;
|
|
return 1;
|
|
}
|
|
__setup("nohlt", nohlt_setup);
|
|
|
|
static int __init hlt_setup(char *__unused)
|
|
{
|
|
hlt_counter = 0;
|
|
return 1;
|
|
}
|
|
__setup("hlt", hlt_setup);
|
|
|
|
void default_idle(void)
|
|
{
|
|
if (likely(hlt_counter)) {
|
|
local_irq_disable();
|
|
stop_critical_timings();
|
|
cpu_relax();
|
|
start_critical_timings();
|
|
local_irq_enable();
|
|
} else {
|
|
clear_thread_flag(TIF_POLLING_NRFLAG);
|
|
smp_mb__after_clear_bit();
|
|
local_irq_disable();
|
|
while (!need_resched())
|
|
cpu_sleep();
|
|
local_irq_enable();
|
|
set_thread_flag(TIF_POLLING_NRFLAG);
|
|
}
|
|
}
|
|
|
|
void cpu_idle(void)
|
|
{
|
|
set_thread_flag(TIF_POLLING_NRFLAG);
|
|
|
|
/* endless idle loop with no priority at all */
|
|
while (1) {
|
|
void (*idle)(void) = pm_idle;
|
|
|
|
if (!idle)
|
|
idle = default_idle;
|
|
|
|
tick_nohz_idle_enter_norcu();
|
|
while (!need_resched())
|
|
idle();
|
|
tick_nohz_idle_exit_norcu();
|
|
|
|
preempt_enable_no_resched();
|
|
schedule();
|
|
preempt_disable();
|
|
check_pgt_cache();
|
|
}
|
|
}
|
|
|
|
void flush_thread(void)
|
|
{
|
|
}
|
|
|
|
int copy_thread(unsigned long clone_flags, unsigned long usp,
|
|
unsigned long unused,
|
|
struct task_struct *p, struct pt_regs *regs)
|
|
{
|
|
struct pt_regs *childregs = task_pt_regs(p);
|
|
struct thread_info *ti = task_thread_info(p);
|
|
|
|
*childregs = *regs;
|
|
if (user_mode(regs))
|
|
childregs->r1 = usp;
|
|
else
|
|
childregs->r1 = ((unsigned long) ti) + THREAD_SIZE;
|
|
|
|
#ifndef CONFIG_MMU
|
|
memset(&ti->cpu_context, 0, sizeof(struct cpu_context));
|
|
ti->cpu_context.r1 = (unsigned long)childregs;
|
|
ti->cpu_context.msr = (unsigned long)childregs->msr;
|
|
#else
|
|
|
|
/* if creating a kernel thread then update the current reg (we don't
|
|
* want to use the parent's value when restoring by POP_STATE) */
|
|
if (kernel_mode(regs))
|
|
/* save new current on stack to use POP_STATE */
|
|
childregs->CURRENT_TASK = (unsigned long)p;
|
|
/* if returning to user then use the parent's value of this register */
|
|
|
|
/* if we're creating a new kernel thread then just zeroing all
|
|
* the registers. That's OK for a brand new thread.*/
|
|
/* Pls. note that some of them will be restored in POP_STATE */
|
|
if (kernel_mode(regs))
|
|
memset(&ti->cpu_context, 0, sizeof(struct cpu_context));
|
|
/* if this thread is created for fork/vfork/clone, then we want to
|
|
* restore all the parent's context */
|
|
/* in addition to the registers which will be restored by POP_STATE */
|
|
else {
|
|
ti->cpu_context = *(struct cpu_context *)regs;
|
|
childregs->msr |= MSR_UMS;
|
|
}
|
|
|
|
/* FIXME STATE_SAVE_PT_OFFSET; */
|
|
ti->cpu_context.r1 = (unsigned long)childregs;
|
|
/* we should consider the fact that childregs is a copy of the parent
|
|
* regs which were saved immediately after entering the kernel state
|
|
* before enabling VM. This MSR will be restored in switch_to and
|
|
* RETURN() and we want to have the right machine state there
|
|
* specifically this state must have INTs disabled before and enabled
|
|
* after performing rtbd
|
|
* compose the right MSR for RETURN(). It will work for switch_to also
|
|
* excepting for VM and UMS
|
|
* don't touch UMS , CARRY and cache bits
|
|
* right now MSR is a copy of parent one */
|
|
childregs->msr |= MSR_BIP;
|
|
childregs->msr &= ~MSR_EIP;
|
|
childregs->msr |= MSR_IE;
|
|
childregs->msr &= ~MSR_VM;
|
|
childregs->msr |= MSR_VMS;
|
|
childregs->msr |= MSR_EE; /* exceptions will be enabled*/
|
|
|
|
ti->cpu_context.msr = (childregs->msr|MSR_VM);
|
|
ti->cpu_context.msr &= ~MSR_UMS; /* switch_to to kernel mode */
|
|
ti->cpu_context.msr &= ~MSR_IE;
|
|
#endif
|
|
ti->cpu_context.r15 = (unsigned long)ret_from_fork - 8;
|
|
|
|
if (clone_flags & CLONE_SETTLS)
|
|
;
|
|
|
|
return 0;
|
|
}
|
|
|
|
#ifndef CONFIG_MMU
|
|
/*
|
|
* Return saved PC of a blocked thread.
|
|
*/
|
|
unsigned long thread_saved_pc(struct task_struct *tsk)
|
|
{
|
|
struct cpu_context *ctx =
|
|
&(((struct thread_info *)(tsk->stack))->cpu_context);
|
|
|
|
/* Check whether the thread is blocked in resume() */
|
|
if (in_sched_functions(ctx->r15))
|
|
return (unsigned long)ctx->r15;
|
|
else
|
|
return ctx->r14;
|
|
}
|
|
#endif
|
|
|
|
static void kernel_thread_helper(int (*fn)(void *), void *arg)
|
|
{
|
|
fn(arg);
|
|
do_exit(-1);
|
|
}
|
|
|
|
int kernel_thread(int (*fn)(void *), void *arg, unsigned long flags)
|
|
{
|
|
struct pt_regs regs;
|
|
|
|
memset(®s, 0, sizeof(regs));
|
|
/* store them in non-volatile registers */
|
|
regs.r5 = (unsigned long)fn;
|
|
regs.r6 = (unsigned long)arg;
|
|
local_save_flags(regs.msr);
|
|
regs.pc = (unsigned long)kernel_thread_helper;
|
|
regs.pt_mode = 1;
|
|
|
|
return do_fork(flags | CLONE_VM | CLONE_UNTRACED, 0,
|
|
®s, 0, NULL, NULL);
|
|
}
|
|
EXPORT_SYMBOL_GPL(kernel_thread);
|
|
|
|
unsigned long get_wchan(struct task_struct *p)
|
|
{
|
|
/* TBD (used by procfs) */
|
|
return 0;
|
|
}
|
|
|
|
/* Set up a thread for executing a new program */
|
|
void start_thread(struct pt_regs *regs, unsigned long pc, unsigned long usp)
|
|
{
|
|
regs->pc = pc;
|
|
regs->r1 = usp;
|
|
regs->pt_mode = 0;
|
|
#ifdef CONFIG_MMU
|
|
regs->msr |= MSR_UMS;
|
|
#endif
|
|
}
|
|
|
|
#ifdef CONFIG_MMU
|
|
#include <linux/elfcore.h>
|
|
/*
|
|
* Set up a thread for executing a new program
|
|
*/
|
|
int dump_fpu(struct pt_regs *regs, elf_fpregset_t *fpregs)
|
|
{
|
|
return 0; /* MicroBlaze has no separate FPU registers */
|
|
}
|
|
#endif /* CONFIG_MMU */
|