From 0c8c0f03e3a292e031596484275c14cf39c0ab7a Mon Sep 17 00:00:00 2001
From: Dave Hansen <dave@sr71.net>
Date: Fri, 17 Jul 2015 12:28:11 +0200
Subject: x86/fpu, sched: Dynamically allocate 'struct fpu'
The FPU rewrite removed the dynamic allocations of 'struct fpu'.
But, this potentially wastes massive amounts of memory (2k per
task on systems that do not have AVX-512 for instance).
Instead of having a separate slab, this patch just appends the
space that we need to the 'task_struct' which we dynamically
allocate already. This saves from doing an extra slab
allocation at fork().
The only real downside here is that we have to stick everything
and the end of the task_struct. But, I think the
BUILD_BUG_ON()s I stuck in there should keep that from being too
fragile.
Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Dave Hansen <dave@sr71.net>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/1437128892-9831-2-git-send-email-mingo@kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
include/linux/sched.h | 12 ++++++++++--
1 file changed, 10 insertions(+), 2 deletions(-)
(limited to 'include/linux/sched.h')
diff --git a/include/linux/sched.h b/include/linux/sched.h
index ae21f1591615..e43a41d892b6 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -1522,8 +1522,6 @@ struct task_struct {
/* hung task detection */
unsigned long last_switch_count;
#endif
-/* CPU-specific state of this task */
- struct thread_struct thread;
/* filesystem information */
struct fs_struct *fs;
/* open file information */
@@ -1778,8 +1776,18 @@ struct task_struct {
unsigned long task_state_change;
#endif
int pagefault_disabled;
+/* CPU-specific state of this task */
+ struct thread_struct thread;
+/*
+ * WARNING: on x86, 'thread_struct' contains a variable-sized
+ * structure. It *MUST* be at the end of 'task_struct'.
+ *
+ * Do not put anything below here!
+ */
};
+extern int arch_task_struct_size(void);
+
/* Future-safe accessor for struct task_struct's cpus_allowed. */
#define tsk_cpus_allowed(tsk) (&(tsk)->cpus_allowed)
--
cgit v1.2.3-70-g09d2
From 5aaeb5c01c5b6c0be7b7aadbf3ace9f3a4458c3d Mon Sep 17 00:00:00 2001
From: Ingo Molnar <mingo@kernel.org>
Date: Fri, 17 Jul 2015 12:28:12 +0200
Subject: x86/fpu, sched: Introduce CONFIG_ARCH_WANTS_DYNAMIC_TASK_STRUCT and
use it on x86
Don't burden architectures without dynamic task_struct sizing
with the overhead of dynamic sizing.
Also optimize the x86 code a bit by caching task_struct_size.
Acked-and-Tested-by: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Dave Hansen <dave@sr71.net>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/1437128892-9831-3-git-send-email-mingo@kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
arch/Kconfig | 4 ++++
arch/x86/Kconfig | 1 +
arch/x86/kernel/fpu/init.c | 17 +++++++++--------
arch/x86/kernel/process.c | 2 +-
fs/proc/kcore.c | 4 ++--
include/linux/sched.h | 6 +++++-
kernel/fork.c | 11 +++++------
7 files changed, 27 insertions(+), 18 deletions(-)
(limited to 'include/linux/sched.h')
diff --git a/arch/Kconfig b/arch/Kconfig
index bec6666a3cc4..8a8ea7110de8 100644
--- a/arch/Kconfig
+++ b/arch/Kconfig
@@ -221,6 +221,10 @@ config ARCH_TASK_STRUCT_ALLOCATOR
config ARCH_THREAD_INFO_ALLOCATOR
bool
+# Select if arch wants to size task_struct dynamically via arch_task_struct_size:
+config ARCH_WANTS_DYNAMIC_TASK_STRUCT
+ bool
+
config HAVE_REGS_AND_STACK_ACCESS_API
bool
help
diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig
index 3dbb7e7909ca..b3a1a5d77d92 100644
--- a/arch/x86/Kconfig
+++ b/arch/x86/Kconfig
@@ -41,6 +41,7 @@ config X86
select ARCH_USE_CMPXCHG_LOCKREF if X86_64
select ARCH_USE_QUEUED_RWLOCKS
select ARCH_USE_QUEUED_SPINLOCKS
+ select ARCH_WANTS_DYNAMIC_TASK_STRUCT
select ARCH_WANT_FRAME_POINTERS
select ARCH_WANT_IPC_PARSE_VERSION if X86_32
select ARCH_WANT_OPTIONAL_GPIOLIB
diff --git a/arch/x86/kernel/fpu/init.c b/arch/x86/kernel/fpu/init.c
index deacbfa6b33e..0b39173dd971 100644
--- a/arch/x86/kernel/fpu/init.c
+++ b/arch/x86/kernel/fpu/init.c
@@ -4,6 +4,8 @@
#include <asm/fpu/internal.h>
#include <asm/tlbflush.h>
+#include <linux/sched.h>
+
/*
* Initialize the TS bit in CR0 according to the style of context-switches
* we are using:
@@ -136,16 +138,14 @@ static void __init fpu__init_system_generic(void)
unsigned int xstate_size;
EXPORT_SYMBOL_GPL(xstate_size);
-#define CHECK_MEMBER_AT_END_OF(TYPE, MEMBER) \
- BUILD_BUG_ON((sizeof(TYPE) - \
- offsetof(TYPE, MEMBER) - \
- sizeof(((TYPE *)0)->MEMBER)) > \
- 0) \
+/* Enforce that 'MEMBER' is the last field of 'TYPE': */
+#define CHECK_MEMBER_AT_END_OF(TYPE, MEMBER) \
+ BUILD_BUG_ON(sizeof(TYPE) != offsetofend(TYPE, MEMBER))
/*
- * We append the 'struct fpu' to the task_struct.
+ * We append the 'struct fpu' to the task_struct:
*/
-int __weak arch_task_struct_size(void)
+static void __init fpu__init_task_struct_size(void)
{
int task_size = sizeof(struct task_struct);
@@ -172,7 +172,7 @@ int __weak arch_task_struct_size(void)
CHECK_MEMBER_AT_END_OF(struct thread_struct, fpu);
CHECK_MEMBER_AT_END_OF(struct task_struct, thread);
- return task_size;
+ arch_task_struct_size = task_size;
}
/*
@@ -326,6 +326,7 @@ void __init fpu__init_system(struct cpuinfo_x86 *c)
fpu__init_system_generic();
fpu__init_system_xstate_size_legacy();
fpu__init_system_xstate();
+ fpu__init_task_struct_size();
fpu__init_system_ctx_switch();
}
diff --git a/arch/x86/kernel/process.c b/arch/x86/kernel/process.c
index 975420eac105..397688beed4b 100644
--- a/arch/x86/kernel/process.c
+++ b/arch/x86/kernel/process.c
@@ -81,7 +81,7 @@ EXPORT_SYMBOL_GPL(idle_notifier_unregister);
*/
int arch_dup_task_struct(struct task_struct *dst, struct task_struct *src)
{
- memcpy(dst, src, arch_task_struct_size());
+ memcpy(dst, src, arch_task_struct_size);
return fpu__copy(&dst->thread.fpu, &src->thread.fpu);
}
diff --git a/fs/proc/kcore.c b/fs/proc/kcore.c
index a0fe99485687..92e6726f6e37 100644
--- a/fs/proc/kcore.c
+++ b/fs/proc/kcore.c
@@ -92,7 +92,7 @@ static size_t get_kcore_size(int *nphdr, size_t *elf_buflen)
roundup(sizeof(CORE_STR), 4)) +
roundup(sizeof(struct elf_prstatus), 4) +
roundup(sizeof(struct elf_prpsinfo), 4) +
- roundup(arch_task_struct_size(), 4);
+ roundup(arch_task_struct_size, 4);
*elf_buflen = PAGE_ALIGN(*elf_buflen);
return size + *elf_buflen;
}
@@ -415,7 +415,7 @@ static void elf_kcore_store_hdr(char *bufp, int nphdr, int dataoff)
/* set up the task structure */
notes[2].name = CORE_STR;
notes[2].type = NT_TASKSTRUCT;
- notes[2].datasz = arch_task_struct_size();
+ notes[2].datasz = arch_task_struct_size;
notes[2].data = current;
nhdr->p_filesz += notesize(¬es[2]);
diff --git a/include/linux/sched.h b/include/linux/sched.h
index e43a41d892b6..04b5ada460b4 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -1786,7 +1786,11 @@ struct task_struct {
*/
};
-extern int arch_task_struct_size(void);
+#ifdef CONFIG_ARCH_WANTS_DYNAMIC_TASK_STRUCT
+extern int arch_task_struct_size __read_mostly;
+#else
+# define arch_task_struct_size (sizeof(struct task_struct))
+#endif
/* Future-safe accessor for struct task_struct's cpus_allowed. */
#define tsk_cpus_allowed(tsk) (&(tsk)->cpus_allowed)
diff --git a/kernel/fork.c b/kernel/fork.c
index 431b67a6098c..dbd9b8d7b7cc 100644
--- a/kernel/fork.c
+++ b/kernel/fork.c
@@ -287,21 +287,20 @@ static void set_max_threads(unsigned int max_threads_suggested)
max_threads = clamp_t(u64, threads, MIN_THREADS, MAX_THREADS);
}
-int __weak arch_task_struct_size(void)
-{
- return sizeof(struct task_struct);
-}
+#ifdef CONFIG_ARCH_WANTS_DYNAMIC_TASK_STRUCT
+/* Initialized by the architecture: */
+int arch_task_struct_size __read_mostly;
+#endif
void __init fork_init(void)
{
- int task_struct_size = arch_task_struct_size();
#ifndef CONFIG_ARCH_TASK_STRUCT_ALLOCATOR
#ifndef ARCH_MIN_TASKALIGN
#define ARCH_MIN_TASKALIGN L1_CACHE_BYTES
#endif
/* create a slab on which task_structs can be allocated */
task_struct_cachep =
- kmem_cache_create("task_struct", task_struct_size,
+ kmem_cache_create("task_struct", arch_task_struct_size,
ARCH_MIN_TASKALIGN, SLAB_PANIC | SLAB_NOTRACK, NULL);
#endif
--
cgit v1.2.3-70-g09d2
From 9d7fb04276481c59610983362d8e023d262b58ca Mon Sep 17 00:00:00 2001
From: Peter Zijlstra <peterz@infradead.org>
Date: Tue, 30 Jun 2015 11:30:54 +0200
Subject: sched/cputime: Guarantee stime + utime == rtime
While the current code guarantees monotonicity for stime and utime
independently of one another, it does not guarantee that the sum of
both is equal to the total time we started out with.
This confuses things (and peoples) who look at this sum, like top, and
will report >100% usage followed by a matching period of 0%.
Rework the code to provide both individual monotonicity and a coherent
sum.
Suggested-by: Fredrik Markstrom <fredrik.markstrom@gmail.com>
Reported-by: Fredrik Markstrom <fredrik.markstrom@gmail.com>
Tested-by: Fredrik Markstrom <fredrik.markstrom@gmail.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Stanislaw Gruszka <sgruszka@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: jason.low2@hp.com
Cc: linux-kernel@vger.kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
include/linux/init_task.h | 10 +++++
include/linux/sched.h | 40 ++++++++++--------
kernel/fork.c | 7 ++--
kernel/sched/cputime.c | 101 +++++++++++++++++++++++++++-------------------
4 files changed, 97 insertions(+), 61 deletions(-)
(limited to 'include/linux/sched.h')
diff --git a/include/linux/init_task.h b/include/linux/init_task.h
index e8493fee8160..d0b380ee7d67 100644
--- a/include/linux/init_task.h
+++ b/include/linux/init_task.h
@@ -32,6 +32,14 @@ extern struct fs_struct init_fs;
#define INIT_CPUSET_SEQ(tsk)
#endif
+#ifndef CONFIG_VIRT_CPU_ACCOUNTING_NATIVE
+#define INIT_PREV_CPUTIME(x) .prev_cputime = { \
+ .lock = __RAW_SPIN_LOCK_UNLOCKED(x.prev_cputime.lock), \
+},
+#else
+#define INIT_PREV_CPUTIME(x)
+#endif
+
#define INIT_SIGNALS(sig) { \
.nr_threads = 1, \
.thread_head = LIST_HEAD_INIT(init_task.thread_node), \
@@ -46,6 +54,7 @@ extern struct fs_struct init_fs;
.cputime_atomic = INIT_CPUTIME_ATOMIC, \
.running = 0, \
}, \
+ INIT_PREV_CPUTIME(sig) \
.cred_guard_mutex = \
__MUTEX_INITIALIZER(sig.cred_guard_mutex), \
}
@@ -246,6 +255,7 @@ extern struct task_group root_task_group;
INIT_TASK_RCU_TASKS(tsk) \
INIT_CPUSET_SEQ(tsk) \
INIT_RT_MUTEXES(tsk) \
+ INIT_PREV_CPUTIME(tsk) \
INIT_VTIME(tsk) \
INIT_NUMA_BALANCING(tsk) \
INIT_KASAN(tsk) \
diff --git a/include/linux/sched.h b/include/linux/sched.h
index ae21f1591615..7412070a25cc 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -530,39 +530,49 @@ struct cpu_itimer {
};
/**
- * struct cputime - snaphsot of system and user cputime
+ * struct prev_cputime - snaphsot of system and user cputime
* @utime: time spent in user mode
* @stime: time spent in system mode
+ * @lock: protects the above two fields
*
- * Gathers a generic snapshot of user and system time.
+ * Stores previous user/system time values such that we can guarantee
+ * monotonicity.
*/
-struct cputime {
+struct prev_cputime {
+#ifndef CONFIG_VIRT_CPU_ACCOUNTING_NATIVE
cputime_t utime;
cputime_t stime;
+ raw_spinlock_t lock;
+#endif
};
+static inline void prev_cputime_init(struct prev_cputime *prev)
+{
+#ifndef CONFIG_VIRT_CPU_ACCOUNTING_NATIVE
+ prev->utime = prev->stime = 0;
+ raw_spin_lock_init(&prev->lock);
+#endif
+}
+
/**
* struct task_cputime - collected CPU time counts
* @utime: time spent in user mode, in &cputime_t units
* @stime: time spent in kernel mode, in &cputime_t units
* @sum_exec_runtime: total time spent on the CPU, in nanoseconds
*
- * This is an extension of struct cputime that includes the total runtime
- * spent by the task from the scheduler point of view.
- *
- * As a result, this structure groups together three kinds of CPU time
- * that are tracked for threads and thread groups. Most things considering
- * CPU time want to group these counts together and treat all three
- * of them in parallel.
+ * This structure groups together three kinds of CPU time that are tracked for
+ * threads and thread groups. Most things considering CPU time want to group
+ * these counts together and treat all three of them in parallel.
*/
struct task_cputime {
cputime_t utime;
cputime_t stime;
unsigned long long sum_exec_runtime;
};
+
/* Alternate field names when used to cache expirations. */
-#define prof_exp stime
#define virt_exp utime
+#define prof_exp stime
#define sched_exp sum_exec_runtime
#define INIT_CPUTIME \
@@ -715,9 +725,7 @@ struct signal_struct {
cputime_t utime, stime, cutime, cstime;
cputime_t gtime;
cputime_t cgtime;
-#ifndef CONFIG_VIRT_CPU_ACCOUNTING_NATIVE
- struct cputime prev_cputime;
-#endif
+ struct prev_cputime prev_cputime;
unsigned long nvcsw, nivcsw, cnvcsw, cnivcsw;
unsigned long min_flt, maj_flt, cmin_flt, cmaj_flt;
unsigned long inblock, oublock, cinblock, coublock;
@@ -1481,9 +1489,7 @@ struct task_struct {
cputime_t utime, stime, utimescaled, stimescaled;
cputime_t gtime;
-#ifndef CONFIG_VIRT_CPU_ACCOUNTING_NATIVE
- struct cputime prev_cputime;
-#endif
+ struct prev_cputime prev_cputime;
#ifdef CONFIG_VIRT_CPU_ACCOUNTING_GEN
seqlock_t vtime_seqlock;
unsigned long long vtime_snap;
diff --git a/kernel/fork.c b/kernel/fork.c
index 1bfefc6f96a4..6e8f807c5716 100644
--- a/kernel/fork.c
+++ b/kernel/fork.c
@@ -1067,6 +1067,7 @@ static int copy_sighand(unsigned long clone_flags, struct task_struct *tsk)
rcu_assign_pointer(tsk->sighand, sig);
if (!sig)
return -ENOMEM;
+
atomic_set(&sig->count, 1);
memcpy(sig->action, current->sighand->action, sizeof(sig->action));
return 0;
@@ -1128,6 +1129,7 @@ static int copy_signal(unsigned long clone_flags, struct task_struct *tsk)
init_sigpending(&sig->shared_pending);
INIT_LIST_HEAD(&sig->posix_timers);
seqlock_init(&sig->stats_lock);
+ prev_cputime_init(&sig->prev_cputime);
hrtimer_init(&sig->real_timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL);
sig->real_timer.function = it_real_fn;
@@ -1335,9 +1337,8 @@ static struct task_struct *copy_process(unsigned long clone_flags,
p->utime = p->stime = p->gtime = 0;
p->utimescaled = p->stimescaled = 0;
-#ifndef CONFIG_VIRT_CPU_ACCOUNTING_NATIVE
- p->prev_cputime.utime = p->prev_cputime.stime = 0;
-#endif
+ prev_cputime_init(&p->prev_cputime);
+
#ifdef CONFIG_VIRT_CPU_ACCOUNTING_GEN
seqlock_init(&p->vtime_seqlock);
p->vtime_snap = 0;
diff --git a/kernel/sched/cputime.c b/kernel/sched/cputime.c
index f5a64ffad176..8cbc3db671df 100644
--- a/kernel/sched/cputime.c
+++ b/kernel/sched/cputime.c
@@ -555,48 +555,43 @@ drop_precision:
}
/*
- * Atomically advance counter to the new value. Interrupts, vcpu
- * scheduling, and scaling inaccuracies can cause cputime_advance
- * to be occasionally called with a new value smaller than counter.
- * Let's enforce atomicity.
+ * Adjust tick based cputime random precision against scheduler runtime
+ * accounting.
*
- * Normally a caller will only go through this loop once, or not
- * at all in case a previous caller updated counter the same jiffy.
- */
-static void cputime_advance(cputime_t *counter, cputime_t new)
-{
- cputime_t old;
-
- while (new > (old = READ_ONCE(*counter)))
- cmpxchg_cputime(counter, old, new);
-}
-
-/*
- * Adjust tick based cputime random precision against scheduler
- * runtime accounting.
+ * Tick based cputime accounting depend on random scheduling timeslices of a
+ * task to be interrupted or not by the timer. Depending on these
+ * circumstances, the number of these interrupts may be over or
+ * under-optimistic, matching the real user and system cputime with a variable
+ * precision.
+ *
+ * Fix this by scaling these tick based values against the total runtime
+ * accounted by the CFS scheduler.
+ *
+ * This code provides the following guarantees:
+ *
+ * stime + utime == rtime
+ * stime_i+1 >= stime_i, utime_i+1 >= utime_i
+ *
+ * Assuming that rtime_i+1 >= rtime_i.
*/
static void cputime_adjust(struct task_cputime *curr,
- struct cputime *prev,
+ struct prev_cputime *prev,
cputime_t *ut, cputime_t *st)
{
cputime_t rtime, stime, utime;
+ unsigned long flags;
- /*
- * Tick based cputime accounting depend on random scheduling
- * timeslices of a task to be interrupted or not by the timer.
- * Depending on these circumstances, the number of these interrupts
- * may be over or under-optimistic, matching the real user and system
- * cputime with a variable precision.
- *
- * Fix this by scaling these tick based values against the total
- * runtime accounted by the CFS scheduler.
- */
+ /* Serialize concurrent callers such that we can honour our guarantees */
+ raw_spin_lock_irqsave(&prev->lock, flags);
rtime = nsecs_to_cputime(curr->sum_exec_runtime);
/*
- * Update userspace visible utime/stime values only if actual execution
- * time is bigger than already exported. Note that can happen, that we
- * provided bigger values due to scaling inaccuracy on big numbers.
+ * This is possible under two circumstances:
+ * - rtime isn't monotonic after all (a bug);
+ * - we got reordered by the lock.
+ *
+ * In both cases this acts as a filter such that the rest of the code
+ * can assume it is monotonic regardless of anything else.
*/
if (prev->stime + prev->utime >= rtime)
goto out;
@@ -606,22 +601,46 @@ static void cputime_adjust(struct task_cputime *curr,
if (utime == 0) {
stime = rtime;
- } else if (stime == 0) {
- utime = rtime;
- } else {
- cputime_t total = stime + utime;
+ goto update;
+ }
- stime = scale_stime((__force u64)stime,
- (__force u64)rtime, (__force u64)total);
- utime = rtime - stime;
+ if (stime == 0) {
+ utime = rtime;
+ goto update;
}
- cputime_advance(&prev->stime, stime);
- cputime_advance(&prev->utime, utime);
+ stime = scale_stime((__force u64)stime, (__force u64)rtime,
+ (__force u64)(stime + utime));
+
+ /*
+ * Make sure stime doesn't go backwards; this preserves monotonicity
+ * for utime because rtime is monotonic.
+ *
+ * utime_i+1 = rtime_i+1 - stime_i
+ * = rtime_i+1 - (rtime_i - utime_i)
+ * = (rtime_i+1 - rtime_i) + utime_i
+ * >= utime_i
+ */
+ if (stime < prev->stime)
+ stime = prev->stime;
+ utime = rtime - stime;
+
+ /*
+ * Make sure utime doesn't go backwards; this still preserves
+ * monotonicity for stime, analogous argument to above.
+ */
+ if (utime < prev->utime) {
+ utime = prev->utime;
+ stime = rtime - utime;
+ }
+update:
+ prev->stime = stime;
+ prev->utime = utime;
out:
*ut = prev->utime;
*st = prev->stime;
+ raw_spin_unlock_irqrestore(&prev->lock, flags);
}
void task_cputime_adjusted(struct task_struct *p, cputime_t *ut, cputime_t *st)
--
cgit v1.2.3-70-g09d2
From 63b0e9edceec10fa41ec33393a1515a5ff444277 Mon Sep 17 00:00:00 2001
From: Mike Galbraith <umgwanakikbuti@gmail.com>
Date: Tue, 14 Jul 2015 17:39:50 +0200
Subject: sched/fair: Beef up wake_wide()
Josef Bacik reported that Facebook sees better performance with their
1:N load (1 dispatch/node, N workers/node) when carrying an old patch
to try very hard to wake to an idle CPU. While looking at wake_wide(),
I noticed that it doesn't pay attention to the wakeup of a many partner
waker, returning 1 only when waking one of its many partners.
Correct that, letting explicit domain flags override the heuristic.
While at it, adjust task_struct bits, we don't need a 64-bit counter.
Tested-by: Josef Bacik <jbacik@fb.com>
Signed-off-by: Mike Galbraith <umgwanakikbuti@gmail.com>
[ Tidy things up. ]
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: kernel-team<Kernel-team@fb.com>
Cc: morten.rasmussen@arm.com
Cc: riel@redhat.com
Link: http://lkml.kernel.org/r/1436888390.7983.49.camel@gmail.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
include/linux/sched.h | 4 +--
kernel/sched/fair.c | 67 +++++++++++++++++++++++++--------------------------
2 files changed, 35 insertions(+), 36 deletions(-)
(limited to 'include/linux/sched.h')
diff --git a/include/linux/sched.h b/include/linux/sched.h
index 7412070a25cc..65a8a8651596 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -1359,9 +1359,9 @@ struct task_struct {
#ifdef CONFIG_SMP
struct llist_node wake_entry;
int on_cpu;
- struct task_struct *last_wakee;
- unsigned long wakee_flips;
+ unsigned int wakee_flips;
unsigned long wakee_flip_decay_ts;
+ struct task_struct *last_wakee;
int wake_cpu;
#endif
diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
index 8b384b8d2f1d..ea23f9f1b51b 100644
--- a/kernel/sched/fair.c
+++ b/kernel/sched/fair.c
@@ -4726,26 +4726,29 @@ static long effective_load(struct task_group *tg, int cpu, long wl, long wg)
#endif
+/*
+ * Detect M:N waker/wakee relationships via a switching-frequency heuristic.
+ * A waker of many should wake a different task than the one last awakened
+ * at a frequency roughly N times higher than one of its wakees. In order
+ * to determine whether we should let the load spread vs consolodating to
+ * shared cache, we look for a minimum 'flip' frequency of llc_size in one
+ * partner, and a factor of lls_size higher frequency in the other. With
+ * both conditions met, we can be relatively sure that the relationship is
+ * non-monogamous, with partner count exceeding socket size. Waker/wakee
+ * being client/server, worker/dispatcher, interrupt source or whatever is
+ * irrelevant, spread criteria is apparent partner count exceeds socket size.
+ */
static int wake_wide(struct task_struct *p)
{
+ unsigned int master = current->wakee_flips;
+ unsigned int slave = p->wakee_flips;
int factor = this_cpu_read(sd_llc_size);
- /*
- * Yeah, it's the switching-frequency, could means many wakee or
- * rapidly switch, use factor here will just help to automatically
- * adjust the loose-degree, so bigger node will lead to more pull.
- */
- if (p->wakee_flips > factor) {
- /*
- * wakee is somewhat hot, it needs certain amount of cpu
- * resource, so if waker is far more hot, prefer to leave
- * it alone.
- */
- if (current->wakee_flips > (factor * p->wakee_flips))
- return 1;
- }
-
- return 0;
+ if (master < slave)
+ swap(master, slave);
+ if (slave < factor || master < slave * factor)
+ return 0;
+ return 1;
}
static int wake_affine(struct sched_domain *sd, struct task_struct *p, int sync)
@@ -4757,13 +4760,6 @@ static int wake_affine(struct sched_domain *sd, struct task_struct *p, int sync)
unsigned long weight;
int balanced;
- /*
- * If we wake multiple tasks be careful to not bounce
- * ourselves around too much.
- */
- if (wake_wide(p))
- return 0;
-
idx = sd->wake_idx;
this_cpu = smp_processor_id();
prev_cpu = task_cpu(p);
@@ -5017,17 +5013,17 @@ select_task_rq_fair(struct task_struct *p, int prev_cpu, int sd_flag, int wake_f
{
struct sched_domain *tmp, *affine_sd = NULL, *sd = NULL;
int cpu = smp_processor_id();
- int new_cpu = cpu;
+ int new_cpu = prev_cpu;
int want_affine = 0;
int sync = wake_flags & WF_SYNC;
if (sd_flag & SD_BALANCE_WAKE)
- want_affine = cpumask_test_cpu(cpu, tsk_cpus_allowed(p));
+ want_affine = !wake_wide(p) && cpumask_test_cpu(cpu, tsk_cpus_allowed(p));
rcu_read_lock();
for_each_domain(cpu, tmp) {
if (!(tmp->flags & SD_LOAD_BALANCE))
- continue;
+ break;
/*
* If both cpu and prev_cpu are part of this domain,
@@ -5041,17 +5037,21 @@ select_task_rq_fair(struct task_struct *p, int prev_cpu, int sd_flag, int wake_f
if (tmp->flags & sd_flag)
sd = tmp;
+ else if (!want_affine)
+ break;
}
- if (affine_sd && cpu != prev_cpu && wake_affine(affine_sd, p, sync))
- prev_cpu = cpu;
-
- if (sd_flag & SD_BALANCE_WAKE) {
- new_cpu = select_idle_sibling(p, prev_cpu);
- goto unlock;
+ if (affine_sd) {
+ sd = NULL; /* Prefer wake_affine over balance flags */
+ if (cpu != prev_cpu && wake_affine(affine_sd, p, sync))
+ new_cpu = cpu;
}
- while (sd) {
+ if (!sd) {
+ if (sd_flag & SD_BALANCE_WAKE) /* XXX always ? */
+ new_cpu = select_idle_sibling(p, new_cpu);
+
+ } else while (sd) {
struct sched_group *group;
int weight;
@@ -5085,7 +5085,6 @@ select_task_rq_fair(struct task_struct *p, int prev_cpu, int sd_flag, int wake_f
}
/* while loop will break here if sd == NULL */
}
-unlock:
rcu_read_unlock();
return new_cpu;
--
cgit v1.2.3-70-g09d2
From fe32d3cd5e8eb0f82e459763374aa80797023403 Mon Sep 17 00:00:00 2001
From: Konstantin Khlebnikov <khlebnikov@yandex-team.ru>
Date: Wed, 15 Jul 2015 12:52:04 +0300
Subject: sched/preempt: Fix cond_resched_lock() and cond_resched_softirq()
These functions check should_resched() before unlocking spinlock/bh-enable:
preempt_count always non-zero => should_resched() always returns false.
cond_resched_lock() worked iff spin_needbreak is set.
This patch adds argument "preempt_offset" to should_resched().
preempt_count offset constants for that:
PREEMPT_DISABLE_OFFSET - offset after preempt_disable()
PREEMPT_LOCK_OFFSET - offset after spin_lock()
SOFTIRQ_DISABLE_OFFSET - offset after local_bh_distable()
SOFTIRQ_LOCK_OFFSET - offset after spin_lock_bh()
Signed-off-by: Konstantin Khlebnikov <khlebnikov@yandex-team.ru>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Alexander Graf <agraf@suse.de>
Cc: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Cc: David Vrabel <david.vrabel@citrix.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Fixes: bdb438065890 ("sched: Extract the basic add/sub preempt_count modifiers")
Link: http://lkml.kernel.org/r/20150715095204.12246.98268.stgit@buzz
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
arch/x86/include/asm/preempt.h | 4 ++--
include/asm-generic/preempt.h | 5 +++--
include/linux/preempt.h | 19 ++++++++++++++-----
include/linux/sched.h | 6 ------
kernel/sched/core.c | 6 +++---
5 files changed, 22 insertions(+), 18 deletions(-)
(limited to 'include/linux/sched.h')
diff --git a/arch/x86/include/asm/preempt.h b/arch/x86/include/asm/preempt.h
index dca71714f860..b12f81022a6b 100644
--- a/arch/x86/include/asm/preempt.h
+++ b/arch/x86/include/asm/preempt.h
@@ -90,9 +90,9 @@ static __always_inline bool __preempt_count_dec_and_test(void)
/*
* Returns true when we need to resched and can (barring IRQ state).
*/
-static __always_inline bool should_resched(void)
+static __always_inline bool should_resched(int preempt_offset)
{
- return unlikely(!raw_cpu_read_4(__preempt_count));
+ return unlikely(raw_cpu_read_4(__preempt_count) == preempt_offset);
}
#ifdef CONFIG_PREEMPT
diff --git a/include/asm-generic/preempt.h b/include/asm-generic/preempt.h
index d0a7a4753db2..0bec580a4885 100644
--- a/include/asm-generic/preempt.h
+++ b/include/asm-generic/preempt.h
@@ -71,9 +71,10 @@ static __always_inline bool __preempt_count_dec_and_test(void)
/*
* Returns true when we need to resched and can (barring IRQ state).
*/
-static __always_inline bool should_resched(void)
+static __always_inline bool should_resched(int preempt_offset)
{
- return unlikely(!preempt_count() && tif_need_resched());
+ return unlikely(preempt_count() == preempt_offset &&
+ tif_need_resched());
}
#ifdef CONFIG_PREEMPT
diff --git a/include/linux/preempt.h b/include/linux/preempt.h
index 84991f185173..bea8dd8ff5e0 100644
--- a/include/linux/preempt.h
+++ b/include/linux/preempt.h
@@ -84,12 +84,20 @@
*/
#define in_nmi() (preempt_count() & NMI_MASK)
+/*
+ * The preempt_count offset after preempt_disable();
+ */
#if defined(CONFIG_PREEMPT_COUNT)
-# define PREEMPT_DISABLE_OFFSET 1
+# define PREEMPT_DISABLE_OFFSET PREEMPT_OFFSET
#else
-# define PREEMPT_DISABLE_OFFSET 0
+# define PREEMPT_DISABLE_OFFSET 0
#endif
+/*
+ * The preempt_count offset after spin_lock()
+ */
+#define PREEMPT_LOCK_OFFSET PREEMPT_DISABLE_OFFSET
+
/*
* The preempt_count offset needed for things like:
*
@@ -103,7 +111,7 @@
*
* Work as expected.
*/
-#define SOFTIRQ_LOCK_OFFSET (SOFTIRQ_DISABLE_OFFSET + PREEMPT_DISABLE_OFFSET)
+#define SOFTIRQ_LOCK_OFFSET (SOFTIRQ_DISABLE_OFFSET + PREEMPT_LOCK_OFFSET)
/*
* Are we running in atomic context? WARNING: this macro cannot
@@ -124,7 +132,8 @@
#if defined(CONFIG_DEBUG_PREEMPT) || defined(CONFIG_PREEMPT_TRACER)
extern void preempt_count_add(int val);
extern void preempt_count_sub(int val);
-#define preempt_count_dec_and_test() ({ preempt_count_sub(1); should_resched(); })
+#define preempt_count_dec_and_test() \
+ ({ preempt_count_sub(1); should_resched(0); })
#else
#define preempt_count_add(val) __preempt_count_add(val)
#define preempt_count_sub(val) __preempt_count_sub(val)
@@ -184,7 +193,7 @@ do { \
#define preempt_check_resched() \
do { \
- if (should_resched()) \
+ if (should_resched(0)) \
__preempt_schedule(); \
} while (0)
diff --git a/include/linux/sched.h b/include/linux/sched.h
index 65a8a8651596..9c144657aace 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -2891,12 +2891,6 @@ extern int _cond_resched(void);
extern int __cond_resched_lock(spinlock_t *lock);
-#ifdef CONFIG_PREEMPT_COUNT
-#define PREEMPT_LOCK_OFFSET PREEMPT_OFFSET
-#else
-#define PREEMPT_LOCK_OFFSET 0
-#endif
-
#define cond_resched_lock(lock) ({ \
___might_sleep(__FILE__, __LINE__, PREEMPT_LOCK_OFFSET);\
__cond_resched_lock(lock); \
diff --git a/kernel/sched/core.c b/kernel/sched/core.c
index fa5826cc612f..f5fad2b12baf 100644
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -4496,7 +4496,7 @@ SYSCALL_DEFINE0(sched_yield)
int __sched _cond_resched(void)
{
- if (should_resched()) {
+ if (should_resched(0)) {
preempt_schedule_common();
return 1;
}
@@ -4514,7 +4514,7 @@ EXPORT_SYMBOL(_cond_resched);
*/
int __cond_resched_lock(spinlock_t *lock)
{
- int resched = should_resched();
+ int resched = should_resched(PREEMPT_LOCK_OFFSET);
int ret = 0;
lockdep_assert_held(lock);
@@ -4536,7 +4536,7 @@ int __sched __cond_resched_softirq(void)
{
BUG_ON(!in_softirq());
- if (should_resched()) {
+ if (should_resched(SOFTIRQ_DISABLE_OFFSET)) {
local_bh_enable();
preempt_schedule_common();
local_bh_disable();
--
cgit v1.2.3-70-g09d2
From 9d89c257dfb9c51a532d69397f6eed75e5168c35 Mon Sep 17 00:00:00 2001
From: Yuyang Du <yuyang.du@intel.com>
Date: Wed, 15 Jul 2015 08:04:37 +0800
Subject: sched/fair: Rewrite runnable load and utilization average tracking
The idea of runnable load average (let runnable time contribute to weight)
was proposed by Paul Turner and Ben Segall, and it is still followed by
this rewrite. This rewrite aims to solve the following issues:
1. cfs_rq's load average (namely runnable_load_avg and blocked_load_avg) is
updated at the granularity of an entity at a time, which results in the
cfs_rq's load average is stale or partially updated: at any time, only
one entity is up to date, all other entities are effectively lagging
behind. This is undesirable.
To illustrate, if we have n runnable entities in the cfs_rq, as time
elapses, they certainly become outdated:
t0: cfs_rq { e1_old, e2_old, ..., en_old }
and when we update:
t1: update e1, then we have cfs_rq { e1_new, e2_old, ..., en_old }
t2: update e2, then we have cfs_rq { e1_old, e2_new, ..., en_old }
...
We solve this by combining all runnable entities' load averages together
in cfs_rq's avg, and update the cfs_rq's avg as a whole. This is based
on the fact that if we regard the update as a function, then:
w * update(e) = update(w * e) and
update(e1) + update(e2) = update(e1 + e2), then
w1 * update(e1) + w2 * update(e2) = update(w1 * e1 + w2 * e2)
therefore, by this rewrite, we have an entirely updated cfs_rq at the
time we update it:
t1: update cfs_rq { e1_new, e2_new, ..., en_new }
t2: update cfs_rq { e1_new, e2_new, ..., en_new }
...
2. cfs_rq's load average is different between top rq->cfs_rq and other
task_group's per CPU cfs_rqs in whether or not blocked_load_average
contributes to the load.
The basic idea behind runnable load average (the same for utilization)
is that the blocked state is taken into account as opposed to only
accounting for the currently runnable state. Therefore, the average
should include both the runnable/running and blocked load averages.
This rewrite does that.
In addition, we also combine runnable/running and blocked averages
of all entities into the cfs_rq's average, and update it together at
once. This is based on the fact that:
update(runnable) + update(blocked) = update(runnable + blocked)
This significantly reduces the code as we don't need to separately
maintain/update runnable/running load and blocked load.
3. How task_group entities' share is calculated is complex and imprecise.
We reduce the complexity in this rewrite to allow a very simple rule:
the task_group's load_avg is aggregated from its per CPU cfs_rqs's
load_avgs. Then group entity's weight is simply proportional to its
own cfs_rq's load_avg / task_group's load_avg. To illustrate,
if a task_group has { cfs_rq1, cfs_rq2, ..., cfs_rqn }, then,
task_group_avg = cfs_rq1_avg + cfs_rq2_avg + ... + cfs_rqn_avg, then
cfs_rqx's entity's share = cfs_rqx_avg / task_group_avg * task_group's share
To sum up, this rewrite in principle is equivalent to the current one, but
fixes the issues described above. Turns out, it significantly reduces the
code complexity and hence increases clarity and efficiency. In addition,
the new averages are more smooth/continuous (no spurious spikes and valleys)
and updated more consistently and quickly to reflect the load dynamics.
As a result, we have less load tracking overhead, better performance,
and especially better power efficiency due to more balanced load.
Signed-off-by: Yuyang Du <yuyang.du@intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: arjan@linux.intel.com
Cc: bsegall@google.com
Cc: dietmar.eggemann@arm.com
Cc: fengguang.wu@intel.com
Cc: len.brown@intel.com
Cc: morten.rasmussen@arm.com
Cc: pjt@google.com
Cc: rafael.j.wysocki@intel.com
Cc: umgwanakikbuti@gmail.com
Cc: vincent.guittot@linaro.org
Link: http://lkml.kernel.org/r/1436918682-4971-3-git-send-email-yuyang.du@intel.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
include/linux/sched.h | 41 ++--
kernel/sched/core.c | 3 -
kernel/sched/debug.c | 41 ++--
kernel/sched/fair.c | 630 ++++++++++++++++----------------------------------
kernel/sched/sched.h | 28 +--
5 files changed, 249 insertions(+), 494 deletions(-)
(limited to 'include/linux/sched.h')
diff --git a/include/linux/sched.h b/include/linux/sched.h
index 9c144657aace..44dca5b35de6 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -1175,29 +1175,24 @@ struct load_weight {
u32 inv_weight;
};
+/*
+ * The load_avg/util_avg accumulates an infinite geometric series.
+ * 1) load_avg factors the amount of time that a sched_entity is
+ * runnable on a rq into its weight. For cfs_rq, it is the aggregated
+ * such weights of all runnable and blocked sched_entities.
+ * 2) util_avg factors frequency scaling into the amount of time
+ * that a sched_entity is running on a CPU, in the range [0..SCHED_LOAD_SCALE].
+ * For cfs_rq, it is the aggregated such times of all runnable and
+ * blocked sched_entities.
+ * The 64 bit load_sum can:
+ * 1) for cfs_rq, afford 4353082796 (=2^64/47742/88761) entities with
+ * the highest weight (=88761) always runnable, we should not overflow
+ * 2) for entity, support any load.weight always runnable
+ */
struct sched_avg {
- u64 last_runnable_update;
- s64 decay_count;
- /*
- * utilization_avg_contrib describes the amount of time that a
- * sched_entity is running on a CPU. It is based on running_avg_sum
- * and is scaled in the range [0..SCHED_LOAD_SCALE].
- * load_avg_contrib described the amount of time that a sched_entity
- * is runnable on a rq. It is based on both runnable_avg_sum and the
- * weight of the task.
- */
- unsigned long load_avg_contrib, utilization_avg_contrib;
- /*
- * These sums represent an infinite geometric series and so are bound
- * above by 1024/(1-y). Thus we only need a u32 to store them for all
- * choices of y < 1-2^(-32)*1024.
- * running_avg_sum reflects the time that the sched_entity is
- * effectively running on the CPU.
- * runnable_avg_sum represents the amount of time a sched_entity is on
- * a runqueue which includes the running time that is monitored by
- * running_avg_sum.
- */
- u32 runnable_avg_sum, avg_period, running_avg_sum;
+ u64 last_update_time, load_sum;
+ u32 util_sum, period_contrib;
+ unsigned long load_avg, util_avg;
};
#ifdef CONFIG_SCHEDSTATS
@@ -1263,7 +1258,7 @@ struct sched_entity {
#endif
#ifdef CONFIG_SMP
- /* Per-entity load-tracking */
+ /* Per entity load average tracking */
struct sched_avg avg;
#endif
};
diff --git a/kernel/sched/core.c b/kernel/sched/core.c
index f5fad2b12baf..3981526539c5 100644
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -2020,9 +2020,6 @@ static void __sched_fork(unsigned long clone_flags, struct task_struct *p)
p->se.prev_sum_exec_runtime = 0;
p->se.nr_migrations = 0;
p->se.vruntime = 0;
-#ifdef CONFIG_SMP
- p->se.avg.decay_count = 0;
-#endif
INIT_LIST_HEAD(&p->se.group_node);
#ifdef CONFIG_SCHEDSTATS
diff --git a/kernel/sched/debug.c b/kernel/sched/debug.c
index 363b7e82554b..74f276f5568c 100644
--- a/kernel/sched/debug.c
+++ b/kernel/sched/debug.c
@@ -88,12 +88,8 @@ static void print_cfs_group_stats(struct seq_file *m, int cpu, struct task_group
#endif
P(se->load.weight);
#ifdef CONFIG_SMP
- P(se->avg.runnable_avg_sum);
- P(se->avg.running_avg_sum);
- P(se->avg.avg_period);
- P(se->avg.load_avg_contrib);
- P(se->avg.utilization_avg_contrib);
- P(se->avg.decay_count);
+ P(se->avg.load_avg);
+ P(se->avg.util_avg);
#endif
#undef PN
#undef P
@@ -209,21 +205,19 @@ void print_cfs_rq(struct seq_file *m, int cpu, struct cfs_rq *cfs_rq)
SEQ_printf(m, " .%-30s: %d\n", "nr_running", cfs_rq->nr_running);
SEQ_printf(m, " .%-30s: %ld\n", "load", cfs_rq->load.weight);
#ifdef CONFIG_SMP
- SEQ_printf(m, " .%-30s: %ld\n", "runnable_load_avg",
- cfs_rq->runnable_load_avg);
- SEQ_printf(m, " .%-30s: %ld\n", "blocked_load_avg",
- cfs_rq->blocked_load_avg);
- SEQ_printf(m, " .%-30s: %ld\n", "utilization_load_avg",
- cfs_rq->utilization_load_avg);
+ SEQ_printf(m, " .%-30s: %lu\n", "load_avg",
+ cfs_rq->avg.load_avg);
+ SEQ_printf(m, " .%-30s: %lu\n", "util_avg",
+ cfs_rq->avg.util_avg);
+ SEQ_printf(m, " .%-30s: %ld\n", "removed_load_avg",
+ atomic_long_read(&cfs_rq->removed_load_avg));
+ SEQ_printf(m, " .%-30s: %ld\n", "removed_util_avg",
+ atomic_long_read(&cfs_rq->removed_util_avg));
#ifdef CONFIG_FAIR_GROUP_SCHED
- SEQ_printf(m, " .%-30s: %ld\n", "tg_load_contrib",
- cfs_rq->tg_load_contrib);
- SEQ_printf(m, " .%-30s: %d\n", "tg_runnable_contrib",
- cfs_rq->tg_runnable_contrib);
+ SEQ_printf(m, " .%-30s: %lu\n", "tg_load_avg_contrib",
+ cfs_rq->tg_load_avg_contrib);
SEQ_printf(m, " .%-30s: %ld\n", "tg_load_avg",
atomic_long_read(&cfs_rq->tg->load_avg));
- SEQ_printf(m, " .%-30s: %d\n", "tg->runnable_avg",
- atomic_read(&cfs_rq->tg->runnable_avg));
#endif
#endif
#ifdef CONFIG_CFS_BANDWIDTH
@@ -631,12 +625,11 @@ void proc_sched_show_task(struct task_struct *p, struct seq_file *m)
P(se.load.weight);
#ifdef CONFIG_SMP
- P(se.avg.runnable_avg_sum);
- P(se.avg.running_avg_sum);
- P(se.avg.avg_period);
- P(se.avg.load_avg_contrib);
- P(se.avg.utilization_avg_contrib);
- P(se.avg.decay_count);
+ P(se.avg.load_sum);
+ P(se.avg.util_sum);
+ P(se.avg.load_avg);
+ P(se.avg.util_avg);
+ P(se.avg.last_update_time);
#endif
P(policy);
P(prio);
diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
index 90292c672a3b..01ffa9509c23 100644
--- a/kernel/sched/fair.c
+++ b/kernel/sched/fair.c
@@ -283,9 +283,6 @@ static inline struct cfs_rq *group_cfs_rq(struct sched_entity *grp)
return grp->my_q;
}
-static void update_cfs_rq_blocked_load(struct cfs_rq *cfs_rq,
- int force_update);
-
static inline void list_add_leaf_cfs_rq(struct cfs_rq *cfs_rq)
{
if (!cfs_rq->on_list) {
@@ -305,8 +302,6 @@ static inline void list_add_leaf_cfs_rq(struct cfs_rq *cfs_rq)
}
cfs_rq->on_list = 1;
- /* We should have no load, but we need to update last_decay. */
- update_cfs_rq_blocked_load(cfs_rq, 0);
}
}
@@ -664,19 +659,31 @@ static u64 sched_vslice(struct cfs_rq *cfs_rq, struct sched_entity *se)
static int select_idle_sibling(struct task_struct *p, int cpu);
static unsigned long task_h_load(struct task_struct *p);
-static inline void __update_task_entity_contrib(struct sched_entity *se);
-static inline void __update_task_entity_utilization(struct sched_entity *se);
+/*
+ * We choose a half-life close to 1 scheduling period.
+ * Note: The tables below are dependent on this value.
+ */
+#define LOAD_AVG_PERIOD 32
+#define LOAD_AVG_MAX 47742 /* maximum possible load avg */
+#define LOAD_AVG_MAX_N 345 /* number of full periods to produce LOAD_MAX_AVG */
/* Give new task start runnable values to heavy its load in infant time */
void init_task_runnable_average(struct task_struct *p)
{
- u32 slice;
+ struct sched_avg *sa = &p->se.avg;
- slice = sched_slice(task_cfs_rq(p), &p->se) >> 10;
- p->se.avg.runnable_avg_sum = p->se.avg.running_avg_sum = slice;
- p->se.avg.avg_period = slice;
- __update_task_entity_contrib(&p->se);
- __update_task_entity_utilization(&p->se);
+ sa->last_update_time = 0;
+ /*
+ * sched_avg's period_contrib should be strictly less then 1024, so
+ * we give it 1023 to make sure it is almost a period (1024us), and
+ * will definitely be update (after enqueue).
+ */
+ sa->period_contrib = 1023;
+ sa->load_avg = scale_load_down(p->se.load.weight);
+ sa->load_sum = sa->load_avg * LOAD_AVG_MAX;
+ sa->util_avg = scale_load_down(SCHED_LOAD_SCALE);
+ sa->util_sum = LOAD_AVG_MAX;
+ /* when this task enqueue'ed, it will contribute to its cfs_rq's load_avg */
}
#else
void init_task_runnable_average(struct task_struct *p)
@@ -1698,8 +1705,8 @@ static u64 numa_get_avg_runtime(struct task_struct *p, u64 *period)
delta = runtime - p->last_sum_exec_runtime;
*period = now - p->last_task_numa_placement;
} else {
- delta = p->se.avg.runnable_avg_sum;
- *period = p->se.avg.avg_period;
+ delta = p->se.avg.load_sum / p->se.load.weight;
+ *period = LOAD_AVG_MAX;
}
p->last_sum_exec_runtime = runtime;
@@ -2347,13 +2354,13 @@ static inline long calc_tg_weight(struct task_group *tg, struct cfs_rq *cfs_rq)
long tg_weight;
/*
- * Use this CPU's actual weight instead of the last load_contribution
- * to gain a more accurate current total weight. See
- * __update_cfs_rq_tg_load_contrib().
+ * Use this CPU's real-time load instead of the last load contribution
+ * as the updating of the contribution is delayed, and we will use the
+ * the real-time load to calc the share. See update_tg_load_avg().
*/
tg_weight = atomic_long_read(&tg->load_avg);
- tg_weight -= cfs_rq->tg_load_contrib;
- tg_weight += cfs_rq->load.weight;
+ tg_weight -= cfs_rq->tg_load_avg_contrib;
+ tg_weight += cfs_rq->avg.load_avg;
return tg_weight;
}
@@ -2363,7 +2370,7 @@ static long calc_cfs_shares(struct cfs_rq *cfs_rq, struct task_group *tg)
long tg_weight, load, shares;
tg_weight = calc_tg_weight(tg, cfs_rq);
- load = cfs_rq->load.weight;
+ load = cfs_rq->avg.load_avg;
shares = (tg->shares * load);
if (tg_weight)
@@ -2425,14 +2432,6 @@ static inline void update_cfs_shares(struct cfs_rq *cfs_rq)
#endif /* CONFIG_FAIR_GROUP_SCHED */
#ifdef CONFIG_SMP
-/*
- * We choose a half-life close to 1 scheduling period.
- * Note: The tables below are dependent on this value.
- */
-#define LOAD_AVG_PERIOD 32
-#define LOAD_AVG_MAX 47742 /* maximum possible load avg */
-#define LOAD_AVG_MAX_N 345 /* number of full periods to produce LOAD_MAX_AVG */
-
/* Precomputed fixed inverse multiplies for multiplication by y^n */
static const u32 runnable_avg_yN_inv[] = {
0xffffffff, 0xfa83b2da, 0xf5257d14, 0xefe4b99a, 0xeac0c6e6, 0xe5b906e6,
@@ -2481,9 +2480,8 @@ static __always_inline u64 decay_load(u64 val, u64 n)
local_n %= LOAD_AVG_PERIOD;
}
- val *= runnable_avg_yN_inv[local_n];
- /* We don't use SRR here since we always want to round down. */
- return val >> 32;
+ val = mul_u64_u32_shr(val, runnable_avg_yN_inv[local_n], 32);
+ return val;
}
/*
@@ -2542,23 +2540,22 @@ static u32 __compute_runnable_contrib(u64 n)
* load_avg = u_0` + y*(u_0 + u_1*y + u_2*y^2 + ... )
* = u_0 + u_1*y + u_2*y^2 + ... [re-labeling u_i --> u_{i+1}]
*/
-static __always_inline int __update_entity_runnable_avg(u64 now, int cpu,
- struct sched_avg *sa,
- int runnable,
- int running)
+static __always_inline int
+__update_load_avg(u64 now, int cpu, struct sched_avg *sa,
+ unsigned long weight, int running)
{
u64 delta, periods;
- u32 runnable_contrib;
+ u32 contrib;
int delta_w, decayed = 0;
unsigned long scale_freq = arch_scale_freq_capacity(NULL, cpu);
- delta = now - sa->last_runnable_update;
+ delta = now - sa->last_update_time;
/*
* This should only happen when time goes backwards, which it
* unfortunately does during sched clock init when we swap over to TSC.
*/
if ((s64)delta < 0) {
- sa->last_runnable_update = now;
+ sa->last_update_time = now;
return 0;
}
@@ -2569,26 +2566,26 @@ static __always_inline int __update_entity_runnable_avg(u64 now, int cpu,
delta >>= 10;
if (!delta)
return 0;
- sa->last_runnable_update = now;
+ sa->last_update_time = now;
/* delta_w is the amount already accumulated against our next period */
- delta_w = sa->avg_period % 1024;
+ delta_w = sa->period_contrib;
if (delta + delta_w >= 1024) {
- /* period roll-over */
decayed = 1;
+ /* how much left for next period will start over, we don't know yet */
+ sa->period_contrib = 0;
+
/*
* Now that we know we're crossing a period boundary, figure
* out how much from delta we need to complete the current
* period and accrue it.
*/
delta_w = 1024 - delta_w;
- if (runnable)
- sa->runnable_avg_sum += delta_w;
+ if (weight)
+ sa->load_sum += weight * delta_w;
if (running)
- sa->running_avg_sum += delta_w * scale_freq
- >> SCHED_CAPACITY_SHIFT;
- sa->avg_period += delta_w;
+ sa->util_sum += delta_w * scale_freq >> SCHED_CAPACITY_SHIFT;
delta -= delta_w;
@@ -2596,334 +2593,156 @@ static __always_inline int __update_entity_runnable_avg(u64 now, int cpu,
periods = delta / 1024;
delta %= 1024;
- sa->runnable_avg_sum = decay_load(sa->runnable_avg_sum,
- periods + 1);
- sa->running_avg_sum = decay_load(sa->running_avg_sum,
- periods + 1);
- sa->avg_period = decay_load(sa->avg_period,
- periods + 1);
+ sa->load_sum = decay_load(sa->load_sum, periods + 1);
+ sa->util_sum = decay_load((u64)(sa->util_sum), periods + 1);
/* Efficiently calculate \sum (1..n_period) 1024*y^i */
- runnable_contrib = __compute_runnable_contrib(periods);
- if (runnable)
- sa->runnable_avg_sum += runnable_contrib;
+ contrib = __compute_runnable_contrib(periods);
+ if (weight)
+ sa->load_sum += weight * contrib;
if (running)
- sa->running_avg_sum += runnable_contrib * scale_freq
- >> SCHED_CAPACITY_SHIFT;
- sa->avg_period += runnable_contrib;
+ sa->util_sum += contrib * scale_freq >> SCHED_CAPACITY_SHIFT;
}
/* Remainder of delta accrued against u_0` */
- if (runnable)
- sa->runnable_avg_sum += delta;
+ if (weight)
+ sa->load_sum += weight * delta;
if (running)
- sa->running_avg_sum += delta * scale_freq
- >> SCHED_CAPACITY_SHIFT;
- sa->avg_period += delta;
-
- return decayed;
-}
-
-/* Synchronize an entity's decay with its parenting cfs_rq.*/
-static inline u64 __synchronize_entity_decay(struct sched_entity *se)
-{
- struct cfs_rq *cfs_rq = cfs_rq_of(se);
- u64 decays = atomic64_read(&cfs_rq->decay_counter);
+ sa->util_sum += delta * scale_freq >> SCHED_CAPACITY_SHIFT;
- decays -= se->avg.decay_count;
- se->avg.decay_count = 0;
- if (!decays)
- return 0;
+ sa->period_contrib += delta;
- se->avg.load_avg_contrib = decay_load(se->avg.load_avg_contrib, decays);
- se->avg.utilization_avg_contrib =
- decay_load(se->avg.utilization_avg_contrib, decays);
+ if (decayed) {
+ sa->load_avg = div_u64(sa->load_sum, LOAD_AVG_MAX);
+ sa->util_avg = (sa->util_sum << SCHED_LOAD_SHIFT) / LOAD_AVG_MAX;
+ }
- return decays;
+ return decayed;
}
#ifdef CONFIG_FAIR_GROUP_SCHED
-static inline void __update_cfs_rq_tg_load_contrib(struct cfs_rq *cfs_rq,
- int force_update)
-{
- struct task_group *tg = cfs_rq->tg;
- long tg_contrib;
-
- tg_contrib = cfs_rq->runnable_load_avg + cfs_rq->blocked_load_avg;
- tg_contrib -= cfs_rq->tg_load_contrib;
-
- if (!tg_contrib)
- return;
-
- if (force_update || abs(tg_contrib) > cfs_rq->tg_load_contrib / 8) {
- atomic_long_add(tg_contrib, &tg->load_avg);
- cfs_rq->tg_load_contrib += tg_contrib;
- }
-}
-
/*
- * Aggregate cfs_rq runnable averages into an equivalent task_group
- * representation for computing load contributions.
+ * Updating tg's load_avg is necessary before update_cfs_share (which is done)
+ * and effective_load (which is not done because it is too costly).
*/
-static inline void __update_tg_runnable_avg(struct sched_avg *sa,
- struct cfs_rq *cfs_rq)
+static inline void update_tg_load_avg(struct cfs_rq *cfs_rq, int force)
{
- struct task_group *tg = cfs_rq->tg;
- long contrib;
-
- /* The fraction of a cpu used by this cfs_rq */
- contrib = div_u64((u64)sa->runnable_avg_sum << NICE_0_SHIFT,
- sa->avg_period + 1);
- contrib -= cfs_rq->tg_runnable_contrib;
+ long delta = cfs_rq->avg.load_avg - cfs_rq->tg_load_avg_contrib;
- if (abs(contrib) > cfs_rq->tg_runnable_contrib / 64) {
- atomic_add(contrib, &tg->runnable_avg);
- cfs_rq->tg_runnable_contrib += contrib;
- }
-}
-
-static inline void __update_group_entity_contrib(struct sched_entity *se)
-{
- struct cfs_rq *cfs_rq = group_cfs_rq(se);
- struct task_group *tg = cfs_rq->tg;
- int runnable_avg;
-
- u64 contrib;
-
- contrib = cfs_rq->tg_load_contrib * tg->shares;
- se->avg.load_avg_contrib = div_u64(contrib,
- atomic_long_read(&tg->load_avg) + 1);
-
- /*
- * For group entities we need to compute a correction term in the case
- * that they are consuming <1 cpu so that we would contribute the same
- * load as a task of equal weight.
- *
- * Explicitly co-ordinating this measurement would be expensive, but
- * fortunately the sum of each cpus contribution forms a usable
- * lower-bound on the true value.
- *
- * Consider the aggregate of 2 contributions. Either they are disjoint
- * (and the sum represents true value) or they are disjoint and we are
- * understating by the aggregate of their overlap.
- *
- * Extending this to N cpus, for a given overlap, the maximum amount we
- * understand is then n_i(n_i+1)/2 * w_i where n_i is the number of
- * cpus that overlap for this interval and w_i is the interval width.
- *
- * On a small machine; the first term is well-bounded which bounds the
- * total error since w_i is a subset of the period. Whereas on a
- * larger machine, while this first term can be larger, if w_i is the
- * of consequential size guaranteed to see n_i*w_i quickly converge to
- * our upper bound of 1-cpu.
- */
- runnable_avg = atomic_read(&tg->runnable_avg);
- if (runnable_avg < NICE_0_LOAD) {
- se->avg.load_avg_contrib *= runnable_avg;
- se->avg.load_avg_contrib >>= NICE_0_SHIFT;
+ if (force || abs(delta) > cfs_rq->tg_load_avg_contrib / 64) {
+ atomic_long_add(delta, &cfs_rq->tg->load_avg);
+ cfs_rq->tg_load_avg_contrib = cfs_rq->avg.load_avg;
}
}
#else /* CONFIG_FAIR_GROUP_SCHED */
-static inline void __update_cfs_rq_tg_load_contrib(struct cfs_rq *cfs_rq,
- int force_update) {}
-static inline void __update_tg_runnable_avg(struct sched_avg *sa,
- struct cfs_rq *cfs_rq) {}
-static inline void __update_group_entity_contrib(struct sched_entity *se) {}
+static inline void update_tg_load_avg(struct cfs_rq *cfs_rq, int force) {}
#endif /* CONFIG_FAIR_GROUP_SCHED */
-static inline void __update_task_entity_contrib(struct sched_entity *se)
-{
- u32 contrib;
-
- /* avoid overflowing a 32-bit type w/ SCHED_LOAD_SCALE */
- contrib = se->avg.runnable_avg_sum * scale_load_down(se->load.weight);
- contrib /= (se->avg.avg_period + 1);
- se->avg.load_avg_contrib = scale_load(contrib);
-}
+static inline u64 cfs_rq_clock_task(struct cfs_rq *cfs_rq);
-/* Compute the current contribution to load_avg by se, return any delta */
-static long __update_entity_load_avg_contrib(struct sched_entity *se)
+/* Group cfs_rq's load_avg is used for task_h_load and update_cfs_share */
+static inline int update_cfs_rq_load_avg(u64 now, struct cfs_rq *cfs_rq)
{
- long old_contrib = se->avg.load_avg_contrib;
+ int decayed;
+ struct sched_avg *sa = &cfs_rq->avg;
- if (entity_is_task(se)) {
- __update_task_entity_contrib(se);
- } else {
- __update_tg_runnable_avg(&se->avg, group_cfs_rq(se));
- __update_group_entity_contrib(se);
+ if (atomic_long_read(&cfs_rq->removed_load_avg)) {
+ long r = atomic_long_xchg(&cfs_rq->removed_load_avg, 0);
+ sa->load_avg = max_t(long, sa->load_avg - r, 0);
+ sa->load_sum = max_t(s64, sa->load_sum - r * LOAD_AVG_MAX, 0);
}
- return se->avg.load_avg_contrib - old_contrib;
-}
-
-
-static inline void __update_task_entity_utilization(struct sched_entity *se)
-{
- u32 contrib;
-
- /* avoid overflowing a 32-bit type w/ SCHED_LOAD_SCALE */
- contrib = se->avg.running_avg_sum * scale_load_down(SCHED_LOAD_SCALE);
- contrib /= (se->avg.avg_period + 1);
- se->avg.utilization_avg_contrib = scale_load(contrib);
-}
+ if (atomic_long_read(&cfs_rq->removed_util_avg)) {
+ long r = atomic_long_xchg(&cfs_rq->removed_util_avg, 0);
+ sa->util_avg = max_t(long, sa->util_avg - r, 0);
+ sa->util_sum = max_t(s32, sa->util_sum -
+ ((r * LOAD_AVG_MAX) >> SCHED_LOAD_SHIFT), 0);
+ }
-static long __update_entity_utilization_avg_contrib(struct sched_entity *se)
-{
- long old_contrib = se->avg.utilization_avg_contrib;
+ decayed = __update_load_avg(now, cpu_of(rq_of(cfs_rq)), sa,
+ scale_load_down(cfs_rq->load.weight), cfs_rq->curr != NULL);
- if (entity_is_task(se))
- __update_task_entity_utilization(se);
- else
- se->avg.utilization_avg_contrib =
- group_cfs_rq(se)->utilization_load_avg;
-
- return se->avg.utilization_avg_contrib - old_contrib;
-}
+#ifndef CONFIG_64BIT
+ smp_wmb();
+ cfs_rq->load_last_update_time_copy = sa->last_update_time;
+#endif
-static inline void subtract_blocked_load_contrib(struct cfs_rq *cfs_rq,
- long load_contrib)
-{
- if (likely(load_contrib < cfs_rq->blocked_load_avg))
- cfs_rq->blocked_load_avg -= load_contrib;
- else
- cfs_rq->blocked_load_avg = 0;
+ return decayed;
}
-static inline u64 cfs_rq_clock_task(struct cfs_rq *cfs_rq);
-
-/* Update a sched_entity's runnable average */
-static inline void update_entity_load_avg(struct sched_entity *se,
- int update_cfs_rq)
+/* Update task and its cfs_rq load average */
+static inline void update_load_avg(struct sched_entity *se, int update_tg)
{
struct cfs_rq *cfs_rq = cfs_rq_of(se);
- long contrib_delta, utilization_delta;
int cpu = cpu_of(rq_of(cfs_rq));
- u64 now;
+ u64 now = cfs_rq_clock_task(cfs_rq);
/*
- * For a group entity we need to use their owned cfs_rq_clock_task() in
- * case they are the parent of a throttled hierarchy.
+ * Track task load average for carrying it to new CPU after migrated, and
+ * track group sched_entity load average for task_h_load calc in migration
*/
- if (entity_is_task(se))
- now = cfs_rq_clock_task(cfs_rq);
- else
- now = cfs_rq_clock_task(group_cfs_rq(se));
+ __update_load_avg(now, cpu, &se->avg,
+ se->on_rq * scale_load_down(se->load.weight), cfs_rq->curr == se);
- if (!__update_entity_runnable_avg(now, cpu, &se->avg, se->on_rq,
- cfs_rq->curr == se))
- return;
-
- contrib_delta = __update_entity_load_avg_contrib(se);
- utilization_delta = __update_entity_utilization_avg_contrib(se);
-
- if (!update_cfs_rq)
- return;
-
- if (se->on_rq) {
- cfs_rq->runnable_load_avg += contrib_delta;
- cfs_rq->utilization_load_avg += utilization_delta;
- } else {
- subtract_blocked_load_contrib(cfs_rq, -contrib_delta);
- }
+ if (update_cfs_rq_load_avg(now, cfs_rq) && update_tg)
+ update_tg_load_avg(cfs_rq, 0);
}
-/*
- * Decay the load contributed by all blocked children and account this so that
- * their contribution may appropriately discounted when they wake up.
- */
-static void update_cfs_rq_blocked_load(struct cfs_rq *cfs_rq, int force_update)
+/* Add the load generated by se into cfs_rq's load average */
+static inline void
+enqueue_entity_load_avg(struct cfs_rq *cfs_rq, struct sched_entity *se)
{
- u64 now = cfs_rq_clock_task(cfs_rq) >> 20;
- u64 decays;
-
- decays = now - cfs_rq->last_decay;
- if (!decays && !force_update)
- return;
+ struct sched_avg *sa = &se->avg;
+ u64 now = cfs_rq_clock_task(cfs_rq);
+ int migrated = 0, decayed;
- if (atomic_long_read(&cfs_rq->removed_load)) {
- unsigned long removed_load;
- removed_load = atomic_long_xchg(&cfs_rq->removed_load, 0);
- subtract_blocked_load_contrib(cfs_rq, removed_load);
+ if (sa->last_update_time == 0) {
+ sa->last_update_time = now;
+ migrated = 1;
}
-
- if (decays) {
- cfs_rq->blocked_load_avg = decay_load(cfs_rq->blocked_load_avg,
- decays);
- atomic64_add(decays, &cfs_rq->decay_counter);
- cfs_rq->last_decay = now;
+ else {
+ __update_load_avg(now, cpu_of(rq_of(cfs_rq)), sa,
+ se->on_rq * scale_load_down(se->load.weight), cfs_rq->curr == se);
}
- __update_cfs_rq_tg_load_contrib(cfs_rq, force_update);
-}
+ decayed = update_cfs_rq_load_avg(now, cfs_rq);
-/* Add the load generated by se into cfs_rq's child load-average */
-static inline void enqueue_entity_load_avg(struct cfs_rq *cfs_rq,
- struct sched_entity *se,
- int wakeup)
-{
- /*
- * We track migrations using entity decay_count <= 0, on a wake-up
- * migration we use a negative decay count to track the remote decays
- * accumulated while sleeping.
- *
- * Newly forked tasks are enqueued with se->avg.decay_count == 0, they
- * are seen by enqueue_entity_load_avg() as a migration with an already
- * constructed load_avg_contrib.
- */
- if (unlikely(se->avg.decay_count <= 0)) {
- se->avg.last_runnable_update = rq_clock_task(rq_of(cfs_rq));
- if (se->avg.decay_count) {
- /*
- * In a wake-up migration we have to approximate the
- * time sleeping. This is because we can't synchronize
- * clock_task between the two cpus, and it is not
- * guaranteed to be read-safe. Instead, we can
- * approximate this using our carried decays, which are
- * explicitly atomically readable.
- */
- se->avg.last_runnable_update -= (-se->avg.decay_count)
- << 20;
- update_entity_load_avg(se, 0);
- /* Indicate that we're now synchronized and on-rq */
- se->avg.decay_count = 0;
- }
- wakeup = 0;
- } else {
- __synchronize_entity_decay(se);
+ if (migrated) {
+ cfs_rq->avg.load_avg += sa->load_avg;
+ cfs_rq->avg.load_sum += sa->load_sum;
+ cfs_rq->avg.util_avg += sa->util_avg;
+ cfs_rq->avg.util_sum += sa->util_sum;
}
- /* migrated tasks did not contribute to our blocked load */
- if (wakeup) {
- subtract_blocked_load_contrib(cfs_rq, se->avg.load_avg_contrib);
- update_entity_load_avg(se, 0);
- }
-
- cfs_rq->runnable_load_avg += se->avg.load_avg_contrib;
- cfs_rq->utilization_load_avg += se->avg.utilization_avg_contrib;
- /* we force update consideration on load-balancer moves */
- update_cfs_rq_blocked_load(cfs_rq, !wakeup);
+ if (decayed || migrated)
+ update_tg_load_avg(cfs_rq, 0);
}
/*
- * Remove se's load from this cfs_rq child load-average, if the entity is
- * transitioning to a blocked state we track its projected decay using
- * blocked_load_avg.
+ * Task first catches up with cfs_rq, and then subtract
+ * itself from the cfs_rq (task must be off the queue now).
*/
-static inline void dequeue_entity_load_avg(struct cfs_rq *cfs_rq,
- struct sched_entity *se,
- int sleep)
+void remove_entity_load_avg(struct sched_entity *se)
{
- update_entity_load_avg(se, 1);
- /* we force update consideration on load-balancer moves */
- update_cfs_rq_blocked_load(cfs_rq, !sleep);
+ struct cfs_rq *cfs_rq = cfs_rq_of(se);
+ u64 last_update_time;
+
+#ifndef CONFIG_64BIT
+ u64 last_update_time_copy;
- cfs_rq->runnable_load_avg -= se->avg.load_avg_contrib;
- cfs_rq->utilization_load_avg -= se->avg.utilization_avg_contrib;
- if (sleep) {
- cfs_rq->blocked_load_avg += se->avg.load_avg_contrib;
- se->avg.decay_count = atomic64_read(&cfs_rq->decay_counter);
- } /* migrations, e.g. sleep=0 leave decay_count == 0 */
+ do {
+ last_update_time_copy = cfs_rq->load_last_update_time_copy;
+ smp_rmb();
+ last_update_time = cfs_rq->avg.last_update_time;
+ } while (last_update_time != last_update_time_copy);
+#else
+ last_update_time = cfs_rq->avg.last_update_time;
+#endif
+
+ __update_load_avg(last_update_time, cpu_of(rq_of(cfs_rq)), &se->avg, 0, 0);
+ atomic_long_add(se->avg.load_avg, &cfs_rq->removed_load_avg);
+ atomic_long_add(se->avg.util_avg, &cfs_rq->removed_util_avg);
}
/*
@@ -2948,16 +2767,10 @@ static int idle_balance(struct rq *this_rq);
#else /* CONFIG_SMP */
-static inline void update_entity_load_avg(struct sched_entity *se,
- int update_cfs_rq) {}
-static inline void enqueue_entity_load_avg(struct cfs_rq *cfs_rq,
- struct sched_entity *se,
- int wakeup) {}
-static inline void dequeue_entity_load_avg(struct cfs_rq *cfs_rq,
- struct sched_entity *se,
- int sleep) {}
-static inline void update_cfs_rq_blocked_load(struct cfs_rq *cfs_rq,
- int force_update) {}
+static inline void update_load_avg(struct sched_entity *se, int update_tg) {}
+static inline void
+enqueue_entity_load_avg(struct cfs_rq *cfs_rq, struct sched_entity *se) {}
+static inline void remove_entity_load_avg(struct sched_entity *se) {}
static inline int idle_balance(struct rq *rq)
{
@@ -3089,7 +2902,7 @@ enqueue_entity(struct cfs_rq *cfs_rq, struct sched_entity *se, int flags)
* Update run-time statistics of the 'current'.
*/
update_curr(cfs_rq);
- enqueue_entity_load_avg(cfs_rq, se, flags & ENQUEUE_WAKEUP);
+ enqueue_entity_load_avg(cfs_rq, se);
account_entity_enqueue(cfs_rq, se);
update_cfs_shares(cfs_rq);
@@ -3164,7 +2977,7 @@ dequeue_entity(struct cfs_rq *cfs_rq, struct sched_entity *se, int flags)
* Update run-time statistics of the 'current'.
*/
update_curr(cfs_rq);
- dequeue_entity_load_avg(cfs_rq, se, flags & DEQUEUE_SLEEP);
+ update_load_avg(se, 1);
update_stats_dequeue(cfs_rq, se);
if (flags & DEQUEUE_SLEEP) {
@@ -3254,7 +3067,7 @@ set_next_entity(struct cfs_rq *cfs_rq, struct sched_entity *se)
*/
update_stats_wait_end(cfs_rq, se);
__dequeue_entity(cfs_rq, se);
- update_entity_load_avg(se, 1);
+ update_load_avg(se, 1);
}
update_stats_curr_start(cfs_rq, se);
@@ -3354,7 +3167,7 @@ static void put_prev_entity(struct cfs_rq *cfs_rq, struct sched_entity *prev)
/* Put 'current' back into the tree. */
__enqueue_entity(cfs_rq, prev);
/* in !on_rq case, update occurred at dequeue */
- update_entity_load_avg(prev, 1);
+ update_load_avg(prev, 0);
}
cfs_rq->curr = NULL;
}
@@ -3370,8 +3183,7 @@ entity_tick(struct cfs_rq *cfs_rq, struct sched_entity *curr, int queued)
/*
* Ensure that runnable average is periodically updated.
*/
- update_entity_load_avg(curr, 1);
- update_cfs_rq_blocked_load(cfs_rq, 1);
+ update_load_avg(curr, 1);
update_cfs_shares(cfs_rq);
#ifdef CONFIG_SCHED_HRTICK
@@ -4244,8 +4056,8 @@ enqueue_task_fair(struct rq *rq, struct task_struct *p, int flags)
if (cfs_rq_throttled(cfs_rq))
break;
+ update_load_avg(se, 1);
update_cfs_shares(cfs_rq);
- update_entity_load_avg(se, 1);
}
if (!se)
@@ -4304,8 +4116,8 @@ static void dequeue_task_fair(struct rq *rq, struct task_struct *p, int flags)
if (cfs_rq_throttled(cfs_rq))
break;
+ update_load_avg(se, 1);
update_cfs_shares(cfs_rq);
- update_entity_load_avg(se, 1);
}
if (!se)
@@ -4444,7 +4256,7 @@ static void __update_cpu_load(struct rq *this_rq, unsigned long this_load,
static void update_idle_cpu_load(struct rq *this_rq)
{
unsigned long curr_jiffies = READ_ONCE(jiffies);
- unsigned long load = this_rq->cfs.runnable_load_avg;
+ unsigned long load = this_rq->cfs.avg.load_avg;
unsigned long pending_updates;
/*
@@ -4490,7 +4302,7 @@ void update_cpu_load_nohz(void)
*/
void update_cpu_load_active(struct rq *this_rq)
{
- unsigned long load = this_rq->cfs.runnable_load_avg;
+ unsigned long load = this_rq->cfs.avg.load_avg;
/*
* See the mess around update_idle_cpu_load() / update_cpu_load_nohz().
*/
@@ -4501,7 +4313,7 @@ void update_cpu_load_active(struct rq *this_rq)
/* Used instead of source_load when we know the type == 0 */
static unsigned long weighted_cpuload(const int cpu)
{
- return cpu_rq(cpu)->cfs.runnable_load_avg;
+ return cpu_rq(cpu)->cfs.avg.load_avg;
}
/*
@@ -4551,7 +4363,7 @@ static unsigned long cpu_avg_load_per_task(int cpu)
{
struct rq *rq = cpu_rq(cpu);
unsigned long nr_running = READ_ONCE(rq->cfs.h_nr_running);
- unsigned long load_avg = rq->cfs.runnable_load_avg;
+ unsigned long load_avg = rq->cfs.avg.load_avg;
if (nr_running)
return load_avg / nr_running;
@@ -4670,7 +4482,7 @@ static long effective_load(struct task_group *tg, int cpu, long wl, long wg)
/*
* w = rw_i + @wl
*/
- w = se->my_q->load.weight + wl;
+ w = se->my_q->avg.load_avg + wl;
/*
* wl = S * s'_i; see (2)
@@ -4691,7 +4503,7 @@ static long effective_load(struct task_group *tg, int cpu, long wl, long wg)
/*
* wl = dw_i = S * (s'_i - s_i); see (3)
*/
- wl -= se->load.weight;
+ wl -= se->avg.load_avg;
/*
* Recursively apply this logic to all parent groups to compute
@@ -4761,14 +4573,14 @@ static int wake_affine(struct sched_domain *sd, struct task_struct *p, int sync)
*/
if (sync) {
tg = task_group(current);
- weight = current->se.load.weight;
+ weight = current->se.avg.load_avg;
this_load += effective_load(tg, this_cpu, -weight, -weight);
load += effective_load(tg, prev_cpu, 0, -weight);
}
tg = task_group(p);
- weight = p->se.load.weight;
+ weight = p->se.avg.load_avg;
/*
* In low-load situations, where prev_cpu is idle and this_cpu is idle
@@ -4961,12 +4773,12 @@ done:
* tasks. The unit of the return value must be the one of capacity so we can
* compare the usage with the capacity of the CPU that is available for CFS
* task (ie cpu_capacity).
- * cfs.utilization_load_avg is the sum of running time of runnable tasks on a
+ * cfs.avg.util_avg is the sum of running time of runnable tasks on a
* CPU. It represents the amount of utilization of a CPU in the range
* [0..SCHED_LOAD_SCALE]. The usage of a CPU can't be higher than the full
* capacity of the CPU because it's about the running time on this CPU.
- * Nevertheless, cfs.utilization_load_avg can be higher than SCHED_LOAD_SCALE
- * because of unfortunate rounding in avg_period and running_load_avg or just
+ * Nevertheless, cfs.avg.util_avg can be higher than SCHED_LOAD_SCALE
+ * because of unfortunate rounding in util_avg or just
* after migrating tasks until the average stabilizes with the new running
* time. So we need to check that the usage stays into the range
* [0..cpu_capacity_orig] and cap if necessary.
@@ -4975,7 +4787,7 @@ done:
*/
static int get_cpu_usage(int cpu)
{
- unsigned long usage = cpu_rq(cpu)->cfs.utilization_load_avg;
+ unsigned long usage = cpu_rq(cpu)->cfs.avg.util_avg;
unsigned long capacity = capacity_orig_of(cpu);
if (usage >= SCHED_LOAD_SCALE)
@@ -5084,26 +4896,22 @@ select_task_rq_fair(struct task_struct *p, int prev_cpu, int sd_flag, int wake_f
* previous cpu. However, the caller only guarantees p->pi_lock is held; no
* other assumptions, including the state of rq->lock, should be made.
*/
-static void
-migrate_task_rq_fair(struct task_struct *p, int next_cpu)
+static void migrate_task_rq_fair(struct task_struct *p, int next_cpu)
{
- struct sched_entity *se = &p->se;
- struct cfs_rq *cfs_rq = cfs_rq_of(se);
-
/*
- * Load tracking: accumulate removed load so that it can be processed
- * when we next update owning cfs_rq under rq->lock. Tasks contribute
- * to blocked load iff they have a positive decay-count. It can never
- * be negative here since on-rq tasks have decay-count == 0.
+ * We are supposed to update the task to "current" time, then its up to date
+ * and ready to go to new CPU/cfs_rq. But we have difficulty in getting
+ * what current time is, so simply throw away the out-of-date time. This
+ * will result in the wakee task is less decayed, but giving the wakee more
+ * load sounds not bad.
*/
- if (se->avg.decay_count) {
- se->avg.decay_count = -__synchronize_entity_decay(se);
- atomic_long_add(se->avg.load_avg_contrib,
- &cfs_rq->removed_load);
- }
+ remove_entity_load_avg(&p->se);
+
+ /* Tell new CPU we are migrated */
+ p->se.avg.last_update_time = 0;
/* We have migrated, no longer consider this task hot */
- se->exec_start = 0;
+ p->se.exec_start = 0;
}
#endif /* CONFIG_SMP */
@@ -5966,36 +5774,6 @@ static void attach_tasks(struct lb_env *env)
}
#ifdef CONFIG_FAIR_GROUP_SCHED
-/*
- * update tg->load_weight by folding this cpu's load_avg
- */
-static void __update_blocked_averages_cpu(struct task_group *tg, int cpu)
-{
- struct sched_entity *se = tg->se[cpu];
- struct cfs_rq *cfs_rq = tg->cfs_rq[cpu];
-
- /* throttled entities do not contribute to load */
- if (throttled_hierarchy(cfs_rq))
- return;
-
- update_cfs_rq_blocked_load(cfs_rq, 1);
-
- if (se) {
- update_entity_load_avg(se, 1);
- /*
- * We pivot on our runnable average having decayed to zero for
- * list removal. This generally implies that all our children
- * have also been removed (modulo rounding error or bandwidth
- * control); however, such cases are rare and we can fix these
- * at enqueue.
- *
- * TODO: fix up out-of-order children on enqueue.
- */
- if (!se->avg.runnable_avg_sum && !cfs_rq->nr_running)
- list_del_leaf_cfs_rq(cfs_rq);
- }
-}
-
static void update_blocked_averages(int cpu)
{
struct rq *rq = cpu_rq(cpu);
@@ -6004,19 +5782,19 @@ static void update_blocked_averages(int cpu)
raw_spin_lock_irqsave(&rq->lock, flags);
update_rq_clock(rq);
+
/*
* Iterates the task_group tree in a bottom up fashion, see
* list_add_leaf_cfs_rq() for details.
*/
for_each_leaf_cfs_rq(rq, cfs_rq) {
- /*
- * Note: We may want to consider periodically releasing
- * rq->lock about these updates so that creating many task
- * groups does not result in continually extending hold time.
- */
- __update_blocked_averages_cpu(cfs_rq->tg, rq->cpu);
- }
+ /* throttled entities do not contribute to load */
+ if (throttled_hierarchy(cfs_rq))
+ continue;
+ if (update_cfs_rq_load_avg(cfs_rq_clock_task(cfs_rq), cfs_rq))
+ update_tg_load_avg(cfs_rq, 0);
+ }
raw_spin_unlock_irqrestore(&rq->lock, flags);
}
@@ -6044,14 +5822,13 @@ static void update_cfs_rq_h_load(struct cfs_rq *cfs_rq)
}
if (!se) {
- cfs_rq->h_load = cfs_rq->runnable_load_avg;
+ cfs_rq->h_load = cfs_rq->avg.load_avg;
cfs_rq->last_h_load_update = now;
}
while ((se = cfs_rq->h_load_next) != NULL) {
load = cfs_rq->h_load;
- load = div64_ul(load * se->avg.load_avg_contrib,
- cfs_rq->runnable_load_avg + 1);
+ load = div64_ul(load * se->avg.load_avg, cfs_rq->avg.load_avg + 1);
cfs_rq = group_cfs_rq(se);
cfs_rq->h_load = load;
cfs_rq->last_h_load_update = now;
@@ -6063,8 +5840,8 @@ static unsigned long task_h_load(struct task_struct *p)
struct cfs_rq *cfs_rq = task_cfs_rq(p);
update_cfs_rq_h_load(cfs_rq);
- return div64_ul(p->se.avg.load_avg_contrib * cfs_rq->h_load,
- cfs_rq->runnable_load_avg + 1);
+ return div64_ul(p->se.avg.load_avg * cfs_rq->h_load,
+ cfs_rq->avg.load_avg + 1);
}
#else
static inline void update_blocked_averages(int cpu)
@@ -6073,7 +5850,7 @@ static inline void update_blocked_averages(int cpu)
static unsigned long task_h_load(struct task_struct *p)
{
- return p->se.avg.load_avg_contrib;
+ return p->se.avg.load_avg;
}
#endif
@@ -8071,15 +7848,18 @@ static void switched_from_fair(struct rq *rq, struct task_struct *p)
}
#ifdef CONFIG_SMP
- /*
- * Remove our load from contribution when we leave sched_fair
- * and ensure we don't carry in an old decay_count if we
- * switch back.
- */
- if (se->avg.decay_count) {
- __synchronize_entity_decay(se);
- subtract_blocked_load_contrib(cfs_rq, se->avg.load_avg_contrib);
- }
+ /* Catch up with the cfs_rq and remove our load when we leave */
+ __update_load_avg(cfs_rq->avg.last_update_time, cpu_of(rq), &se->avg,
+ se->on_rq * scale_load_down(se->load.weight), cfs_rq->curr == se);
+
+ cfs_rq->avg.load_avg =
+ max_t(long, cfs_rq->avg.load_avg - se->avg.load_avg, 0);
+ cfs_rq->avg.load_sum =
+ max_t(s64, cfs_rq->avg.load_sum - se->avg.load_sum, 0);
+ cfs_rq->avg.util_avg =
+ max_t(long, cfs_rq->avg.util_avg - se->avg.util_avg, 0);
+ cfs_rq->avg.util_sum =
+ max_t(s32, cfs_rq->avg.util_sum - se->avg.util_sum, 0);
#endif
}
@@ -8136,8 +7916,8 @@ void init_cfs_rq(struct cfs_rq *cfs_rq)
cfs_rq->min_vruntime_copy = cfs_rq->min_vruntime;
#endif
#ifdef CONFIG_SMP
- atomic64_set(&cfs_rq->decay_counter, 1);
- atomic_long_set(&cfs_rq->removed_load, 0);
+ atomic_long_set(&cfs_rq->removed_load_avg, 0);
+ atomic_long_set(&cfs_rq->removed_util_avg, 0);
#endif
}
@@ -8182,14 +7962,14 @@ static void task_move_group_fair(struct task_struct *p, int queued)
if (!queued) {
cfs_rq = cfs_rq_of(se);
se->vruntime += cfs_rq->min_vruntime;
+
#ifdef CONFIG_SMP
- /*
- * migrate_task_rq_fair() will have removed our previous
- * contribution, but we must synchronize for ongoing future
- * decay.
- */
- se->avg.decay_count = atomic64_read(&cfs_rq->decay_counter);
- cfs_rq->blocked_load_avg += se->avg.load_avg_contrib;
+ /* Virtually synchronize task with its new cfs_rq */
+ p->se.avg.last_update_time = cfs_rq->avg.last_update_time;
+ cfs_rq->avg.load_avg += p->se.avg.load_avg;
+ cfs_rq->avg.load_sum += p->se.avg.load_sum;
+ cfs_rq->avg.util_avg += p->se.avg.util_avg;
+ cfs_rq->avg.util_sum += p->se.avg.util_sum;
#endif
}
}
diff --git a/kernel/sched/sched.h b/kernel/sched/sched.h
index e13210cce7e8..dcde941a585b 100644
--- a/kernel/sched/sched.h
+++ b/kernel/sched/sched.h
@@ -245,7 +245,6 @@ struct task_group {
#ifdef CONFIG_SMP
atomic_long_t load_avg;
- atomic_t runnable_avg;
#endif
#endif
@@ -366,27 +365,18 @@ struct cfs_rq {
#ifdef CONFIG_SMP
/*
- * CFS Load tracking
- * Under CFS, load is tracked on a per-entity basis and aggregated up.
- * This allows for the description of both thread and group usage (in
- * the FAIR_GROUP_SCHED case).
- * runnable_load_avg is the sum of the load_avg_contrib of the
- * sched_entities on the rq.
- * blocked_load_avg is similar to runnable_load_avg except that its
- * the blocked sched_entities on the rq.
- * utilization_load_avg is the sum of the average running time of the
- * sched_entities on the rq.
+ * CFS load tracking
*/
- unsigned long runnable_load_avg, blocked_load_avg, utilization_load_avg;
- atomic64_t decay_counter;
- u64 last_decay;
- atomic_long_t removed_load;
-
+ struct sched_avg avg;
#ifdef CONFIG_FAIR_GROUP_SCHED
- /* Required to track per-cpu representation of a task_group */
- u32 tg_runnable_contrib;
- unsigned long tg_load_contrib;
+ unsigned long tg_load_avg_contrib;
+#endif
+ atomic_long_t removed_load_avg, removed_util_avg;
+#ifndef CONFIG_64BIT
+ u64 load_last_update_time_copy;
+#endif
+#ifdef CONFIG_FAIR_GROUP_SCHED
/*
* h_load = weight * f(tg)
*
--
cgit v1.2.3-70-g09d2
From 25834c73f93af7f0712c98ca4593691592e6b360 Mon Sep 17 00:00:00 2001
From: Peter Zijlstra <peterz@infradead.org>
Date: Fri, 15 May 2015 17:43:34 +0200
Subject: sched: Fix a race between __kthread_bind() and sched_setaffinity()
Because sched_setscheduler() checks p->flags & PF_NO_SETAFFINITY
without locks, a caller might observe an old value and race with the
set_cpus_allowed_ptr() call from __kthread_bind() and effectively undo
it:
__kthread_bind()
do_set_cpus_allowed()
<SYSCALL>
sched_setaffinity()
if (p->flags & PF_NO_SETAFFINITIY)
set_cpus_allowed_ptr()
p->flags |= PF_NO_SETAFFINITY
Fix the bug by putting everything under the regular scheduler locks.
This also closes a hole in the serialization of task_struct::{nr_,}cpus_allowed.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Tejun Heo <tj@kernel.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: dedekind1@gmail.com
Cc: juri.lelli@arm.com
Cc: mgorman@suse.de
Cc: riel@redhat.com
Cc: rostedt@goodmis.org
Link: http://lkml.kernel.org/r/20150515154833.545640346@infradead.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
include/linux/kthread.h | 1 +
include/linux/sched.h | 7 -------
kernel/kthread.c | 20 +++++++++++++++++---
kernel/sched/core.c | 36 ++++++++++++++++++++++++++++++++----
kernel/workqueue.c | 6 ++----
5 files changed, 52 insertions(+), 18 deletions(-)
(limited to 'include/linux/sched.h')
diff --git a/include/linux/kthread.h b/include/linux/kthread.h
index 13d55206ccf6..869b21dcf503 100644
--- a/include/linux/kthread.h
+++ b/include/linux/kthread.h
@@ -38,6 +38,7 @@ struct task_struct *kthread_create_on_cpu(int (*threadfn)(void *data),
})
void kthread_bind(struct task_struct *k, unsigned int cpu);
+void kthread_bind_mask(struct task_struct *k, const struct cpumask *mask);
int kthread_stop(struct task_struct *k);
bool kthread_should_stop(void);
bool kthread_should_park(void);
diff --git a/include/linux/sched.h b/include/linux/sched.h
index 44dca5b35de6..81bb4577274b 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -2203,13 +2203,6 @@ static inline void calc_load_enter_idle(void) { }
static inline void calc_load_exit_idle(void) { }
#endif /* CONFIG_NO_HZ_COMMON */
-#ifndef CONFIG_CPUMASK_OFFSTACK
-static inline int set_cpus_allowed(struct task_struct *p, cpumask_t new_mask)
-{
- return set_cpus_allowed_ptr(p, &new_mask);
-}
-#endif
-
/*
* Do not use outside of architecture code which knows its limitations.
*
diff --git a/kernel/kthread.c b/kernel/kthread.c
index 10e489c448fe..7c40a189becc 100644
--- a/kernel/kthread.c
+++ b/kernel/kthread.c
@@ -325,16 +325,30 @@ struct task_struct *kthread_create_on_node(int (*threadfn)(void *data),
}
EXPORT_SYMBOL(kthread_create_on_node);
-static void __kthread_bind(struct task_struct *p, unsigned int cpu, long state)
+static void __kthread_bind_mask(struct task_struct *p, const struct cpumask *mask, long state)
{
- /* Must have done schedule() in kthread() before we set_task_cpu */
+ unsigned long flags;
+
if (!wait_task_inactive(p, state)) {
WARN_ON(1);
return;
}
+
/* It's safe because the task is inactive. */
- do_set_cpus_allowed(p, cpumask_of(cpu));
+ raw_spin_lock_irqsave(&p->pi_lock, flags);
+ do_set_cpus_allowed(p, mask);
p->flags |= PF_NO_SETAFFINITY;
+ raw_spin_unlock_irqrestore(&p->pi_lock, flags);
+}
+
+static void __kthread_bind(struct task_struct *p, unsigned int cpu, long state)
+{
+ __kthread_bind_mask(p, cpumask_of(cpu), state);
+}
+
+void kthread_bind_mask(struct task_struct *p, const struct cpumask *mask)
+{
+ __kthread_bind_mask(p, mask, TASK_UNINTERRUPTIBLE);
}
/**
diff --git a/kernel/sched/core.c b/kernel/sched/core.c
index ea6d74345e60..2e3b983da836 100644
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -1153,6 +1153,8 @@ static int migration_cpu_stop(void *data)
void do_set_cpus_allowed(struct task_struct *p, const struct cpumask *new_mask)
{
+ lockdep_assert_held(&p->pi_lock);
+
if (p->sched_class->set_cpus_allowed)
p->sched_class->set_cpus_allowed(p, new_mask);
@@ -1169,7 +1171,8 @@ void do_set_cpus_allowed(struct task_struct *p, const struct cpumask *new_mask)
* task must not exit() & deallocate itself prematurely. The
* call is not atomic; no spinlocks may be held.
*/
-int set_cpus_allowed_ptr(struct task_struct *p, const struct cpumask *new_mask)
+static int __set_cpus_allowed_ptr(struct task_struct *p,
+ const struct cpumask *new_mask, bool check)
{
unsigned long flags;
struct rq *rq;
@@ -1178,6 +1181,15 @@ int set_cpus_allowed_ptr(struct task_struct *p, const struct cpumask *new_mask)
rq = task_rq_lock(p, &flags);
+ /*
+ * Must re-check here, to close a race against __kthread_bind(),
+ * sched_setaffinity() is not guaranteed to observe the flag.
+ */
+ if (check && (p->flags & PF_NO_SETAFFINITY)) {
+ ret = -EINVAL;
+ goto out;
+ }
+
if (cpumask_equal(&p->cpus_allowed, new_mask))
goto out;
@@ -1214,6 +1226,11 @@ out:
return ret;
}
+
+int set_cpus_allowed_ptr(struct task_struct *p, const struct cpumask *new_mask)
+{
+ return __set_cpus_allowed_ptr(p, new_mask, false);
+}
EXPORT_SYMBOL_GPL(set_cpus_allowed_ptr);
void set_task_cpu(struct task_struct *p, unsigned int new_cpu)
@@ -1595,6 +1612,15 @@ static void update_avg(u64 *avg, u64 sample)
s64 diff = sample - *avg;
*avg += diff >> 3;
}
+
+#else
+
+static inline int __set_cpus_allowed_ptr(struct task_struct *p,
+ const struct cpumask *new_mask, bool check)
+{
+ return set_cpus_allowed_ptr(p, new_mask);
+}
+
#endif /* CONFIG_SMP */
static void
@@ -4340,7 +4366,7 @@ long sched_setaffinity(pid_t pid, const struct cpumask *in_mask)
}
#endif
again:
- retval = set_cpus_allowed_ptr(p, new_mask);
+ retval = __set_cpus_allowed_ptr(p, new_mask, true);
if (!retval) {
cpuset_cpus_allowed(p, cpus_allowed);
@@ -4865,7 +4891,8 @@ void init_idle(struct task_struct *idle, int cpu)
struct rq *rq = cpu_rq(cpu);
unsigned long flags;
- raw_spin_lock_irqsave(&rq->lock, flags);
+ raw_spin_lock_irqsave(&idle->pi_lock, flags);
+ raw_spin_lock(&rq->lock);
__sched_fork(0, idle);
idle->state = TASK_RUNNING;
@@ -4891,7 +4918,8 @@ void init_idle(struct task_struct *idle, int cpu)
#if defined(CONFIG_SMP)
idle->on_cpu = 1;
#endif
- raw_spin_unlock_irqrestore(&rq->lock, flags);
+ raw_spin_unlock(&rq->lock);
+ raw_spin_unlock_irqrestore(&idle->pi_lock, flags);
/* Set the preempt count _outside_ the spinlocks! */
init_idle_preempt_count(idle, cpu);
diff --git a/kernel/workqueue.c b/kernel/workqueue.c
index 4c4f06176f74..f5782d5fd196 100644
--- a/kernel/workqueue.c
+++ b/kernel/workqueue.c
@@ -1714,9 +1714,7 @@ static struct worker *create_worker(struct worker_pool *pool)
goto fail;
set_user_nice(worker->task, pool->attrs->nice);
-
- /* prevent userland from meddling with cpumask of workqueue workers */
- worker->task->flags |= PF_NO_SETAFFINITY;
+ kthread_bind_mask(worker->task, pool->attrs->cpumask);
/* successful, attach the worker to the pool */
worker_attach_to_pool(worker, pool);
@@ -3856,7 +3854,7 @@ struct workqueue_struct *__alloc_workqueue_key(const char *fmt,
}
wq->rescuer = rescuer;
- rescuer->task->flags |= PF_NO_SETAFFINITY;
+ kthread_bind_mask(rescuer->task, cpu_possible_mask);
wake_up_process(rescuer->task);
}
--
cgit v1.2.3-70-g09d2
From 72b252aed506b8f1a03f7abd29caef4cdf6a043b Mon Sep 17 00:00:00 2001
From: Mel Gorman <mgorman@suse.de>
Date: Fri, 4 Sep 2015 15:47:32 -0700
Subject: mm: send one IPI per CPU to TLB flush all entries after unmapping
pages
An IPI is sent to flush remote TLBs when a page is unmapped that was
potentially accesssed by other CPUs. There are many circumstances where
this happens but the obvious one is kswapd reclaiming pages belonging to a
running process as kswapd and the task are likely running on separate
CPUs.
On small machines, this is not a significant problem but as machine gets
larger with more cores and more memory, the cost of these IPIs can be
high. This patch uses a simple structure that tracks CPUs that
potentially have TLB entries for pages being unmapped. When the unmapping
is complete, the full TLB is flushed on the assumption that a refill cost
is lower than flushing individual entries.
Architectures wishing to do this must give the following guarantee.
If a clean page is unmapped and not immediately flushed, the
architecture must guarantee that a write to that linear address
from a CPU with a cached TLB entry will trap a page fault.
This is essentially what the kernel already depends on but the window is
much larger with this patch applied and is worth highlighting. The
architecture should consider whether the cost of the full TLB flush is
higher than sending an IPI to flush each individual entry. An additional
architecture helper called flush_tlb_local is required. It's a trivial
wrapper with some accounting in the x86 case.
The impact of this patch depends on the workload as measuring any benefit
requires both mapped pages co-located on the LRU and memory pressure. The
case with the biggest impact is multiple processes reading mapped pages
taken from the vm-scalability test suite. The test case uses NR_CPU
readers of mapped files that consume 10*RAM.
Linear mapped reader on a 4-node machine with 64G RAM and 48 CPUs
4.2.0-rc1 4.2.0-rc1
vanilla flushfull-v7
Ops lru-file-mmap-read-elapsed 159.62 ( 0.00%) 120.68 ( 24.40%)
Ops lru-file-mmap-read-time_range 30.59 ( 0.00%) 2.80 ( 90.85%)
Ops lru-file-mmap-read-time_stddv 6.70 ( 0.00%) 0.64 ( 90.38%)
4.2.0-rc1 4.2.0-rc1
vanilla flushfull-v7
User 581.00 611.43
System 5804.93 4111.76
Elapsed 161.03 122.12
This is showing that the readers completed 24.40% faster with 29% less
system CPU time. From vmstats, it is known that the vanilla kernel was
interrupted roughly 900K times per second during the steady phase of the
test and the patched kernel was interrupts 180K times per second.
The impact is lower on a single socket machine.
4.2.0-rc1 4.2.0-rc1
vanilla flushfull-v7
Ops lru-file-mmap-read-elapsed 25.33 ( 0.00%) 20.38 ( 19.54%)
Ops lru-file-mmap-read-time_range 0.91 ( 0.00%) 1.44 (-58.24%)
Ops lru-file-mmap-read-time_stddv 0.28 ( 0.00%) 0.47 (-65.34%)
4.2.0-rc1 4.2.0-rc1
vanilla flushfull-v7
User 58.09 57.64
System 111.82 76.56
Elapsed 27.29 22.55
It's still a noticeable improvement with vmstat showing interrupts went
from roughly 500K per second to 45K per second.
The patch will have no impact on workloads with no memory pressure or have
relatively few mapped pages. It will have an unpredictable impact on the
workload running on the CPU being flushed as it'll depend on how many TLB
entries need to be refilled and how long that takes. Worst case, the TLB
will be completely cleared of active entries when the target PFNs were not
resident at all.
[sasha.levin@oracle.com: trace tlb flush after disabling preemption in try_to_unmap_flush]
Signed-off-by: Mel Gorman <mgorman@suse.de>
Reviewed-by: Rik van Riel <riel@redhat.com>
Cc: Dave Hansen <dave.hansen@intel.com>
Acked-by: Ingo Molnar <mingo@kernel.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
Cc: Michal Hocko <mhocko@suse.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
arch/x86/Kconfig | 1 +
arch/x86/include/asm/tlbflush.h | 6 +++
include/linux/rmap.h | 3 ++
include/linux/sched.h | 16 +++++++
init/Kconfig | 10 ++++
mm/internal.h | 11 +++++
mm/rmap.c | 104 +++++++++++++++++++++++++++++++++++++++-
mm/vmscan.c | 23 ++++++++-
8 files changed, 172 insertions(+), 2 deletions(-)
(limited to 'include/linux/sched.h')
diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig
index 48f7433dac6f..117e2f373e50 100644
--- a/arch/x86/Kconfig
+++ b/arch/x86/Kconfig
@@ -41,6 +41,7 @@ config X86
select ARCH_USE_CMPXCHG_LOCKREF if X86_64
select ARCH_USE_QUEUED_RWLOCKS
select ARCH_USE_QUEUED_SPINLOCKS
+ select ARCH_WANT_BATCHED_UNMAP_TLB_FLUSH if SMP
select ARCH_WANTS_DYNAMIC_TASK_STRUCT
select ARCH_WANT_FRAME_POINTERS
select ARCH_WANT_IPC_PARSE_VERSION if X86_32
diff --git a/arch/x86/include/asm/tlbflush.h b/arch/x86/include/asm/tlbflush.h
index cd791948b286..6df2029405a3 100644
--- a/arch/x86/include/asm/tlbflush.h
+++ b/arch/x86/include/asm/tlbflush.h
@@ -261,6 +261,12 @@ static inline void reset_lazy_tlbstate(void)
#endif /* SMP */
+/* Not inlined due to inc_irq_stat not being defined yet */
+#define flush_tlb_local() { \
+ inc_irq_stat(irq_tlb_count); \
+ local_flush_tlb(); \
+}
+
#ifndef CONFIG_PARAVIRT
#define flush_tlb_others(mask, mm, start, end) \
native_flush_tlb_others(mask, mm, start, end)
diff --git a/include/linux/rmap.h b/include/linux/rmap.h
index c89c53a113a8..29446aeef36e 100644
--- a/include/linux/rmap.h
+++ b/include/linux/rmap.h
@@ -89,6 +89,9 @@ enum ttu_flags {
TTU_IGNORE_MLOCK = (1 << 8), /* ignore mlock */
TTU_IGNORE_ACCESS = (1 << 9), /* don't age */
TTU_IGNORE_HWPOISON = (1 << 10),/* corrupted page is recoverable */
+ TTU_BATCH_FLUSH = (1 << 11), /* Batch TLB flushes where possible
+ * and caller guarantees they will
+ * do a final flush if necessary */
};
#ifdef CONFIG_MMU
diff --git a/include/linux/sched.h b/include/linux/sched.h
index 119823decc46..3c602c20c717 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -1344,6 +1344,18 @@ enum perf_event_task_context {
perf_nr_task_contexts,
};
+/* Track pages that require TLB flushes */
+struct tlbflush_unmap_batch {
+ /*
+ * Each bit set is a CPU that potentially has a TLB entry for one of
+ * the PFNs being flushed. See set_tlb_ubc_flush_pending().
+ */
+ struct cpumask cpumask;
+
+ /* True if any bit in cpumask is set */
+ bool flush_required;
+};
+
struct task_struct {
volatile long state; /* -1 unrunnable, 0 runnable, >0 stopped */
void *stack;
@@ -1700,6 +1712,10 @@ struct task_struct {
unsigned long numa_pages_migrated;
#endif /* CONFIG_NUMA_BALANCING */
+#ifdef CONFIG_ARCH_WANT_BATCHED_UNMAP_TLB_FLUSH
+ struct tlbflush_unmap_batch tlb_ubc;
+#endif
+
struct rcu_head rcu;
/*
diff --git a/init/Kconfig b/init/Kconfig
index 161acd8bc56f..cf7e4824c8d0 100644
--- a/init/Kconfig
+++ b/init/Kconfig
@@ -882,6 +882,16 @@ config GENERIC_SCHED_CLOCK
config ARCH_SUPPORTS_NUMA_BALANCING
bool
+#
+# For architectures that prefer to flush all TLBs after a number of pages
+# are unmapped instead of sending one IPI per page to flush. The architecture
+# must provide guarantees on what happens if a clean TLB cache entry is
+# written after the unmap. Details are in mm/rmap.c near the check for
+# should_defer_flush. The architecture should also consider if the full flush
+# and the refill costs are offset by the savings of sending fewer IPIs.
+config ARCH_WANT_BATCHED_UNMAP_TLB_FLUSH
+ bool
+
#
# For architectures that know their GCC __int128 support is sound
#
diff --git a/mm/internal.h b/mm/internal.h
index 36b23f1e2ca6..bd6372ac5f7f 100644
--- a/mm/internal.h
+++ b/mm/internal.h
@@ -426,4 +426,15 @@ unsigned long reclaim_clean_pages_from_list(struct zone *zone,
#define ALLOC_CMA 0x80 /* allow allocations from CMA areas */
#define ALLOC_FAIR 0x100 /* fair zone allocation */
+enum ttu_flags;
+struct tlbflush_unmap_batch;
+
+#ifdef CONFIG_ARCH_WANT_BATCHED_UNMAP_TLB_FLUSH
+void try_to_unmap_flush(void);
+#else
+static inline void try_to_unmap_flush(void)
+{
+}
+
+#endif /* CONFIG_ARCH_WANT_BATCHED_UNMAP_TLB_FLUSH */
#endif /* __MM_INTERNAL_H */
diff --git a/mm/rmap.c b/mm/rmap.c
index 171b68768df1..326d5d89e45c 100644
--- a/mm/rmap.c
+++ b/mm/rmap.c
@@ -62,6 +62,8 @@
#include <asm/tlbflush.h>
+#include <trace/events/tlb.h>
+
#include "internal.h"
static struct kmem_cache *anon_vma_cachep;
@@ -583,6 +585,89 @@ vma_address(struct page *page, struct vm_area_struct *vma)
return address;
}
+#ifdef CONFIG_ARCH_WANT_BATCHED_UNMAP_TLB_FLUSH
+static void percpu_flush_tlb_batch_pages(void *data)
+{
+ /*
+ * All TLB entries are flushed on the assumption that it is
+ * cheaper to flush all TLBs and let them be refilled than
+ * flushing individual PFNs. Note that we do not track mm's
+ * to flush as that might simply be multiple full TLB flushes
+ * for no gain.
+ */
+ count_vm_tlb_event(NR_TLB_REMOTE_FLUSH_RECEIVED);
+ flush_tlb_local();
+}
+
+/*
+ * Flush TLB entries for recently unmapped pages from remote CPUs. It is
+ * important if a PTE was dirty when it was unmapped that it's flushed
+ * before any IO is initiated on the page to prevent lost writes. Similarly,
+ * it must be flushed before freeing to prevent data leakage.
+ */
+void try_to_unmap_flush(void)
+{
+ struct tlbflush_unmap_batch *tlb_ubc = ¤t->tlb_ubc;
+ int cpu;
+
+ if (!tlb_ubc->flush_required)
+ return;
+
+ cpu = get_cpu();
+
+ trace_tlb_flush(TLB_REMOTE_SHOOTDOWN, -1UL);
+
+ if (cpumask_test_cpu(cpu, &tlb_ubc->cpumask))
+ percpu_flush_tlb_batch_pages(&tlb_ubc->cpumask);
+
+ if (cpumask_any_but(&tlb_ubc->cpumask, cpu) < nr_cpu_ids) {
+ smp_call_function_many(&tlb_ubc->cpumask,
+ percpu_flush_tlb_batch_pages, (void *)tlb_ubc, true);
+ }
+ cpumask_clear(&tlb_ubc->cpumask);
+ tlb_ubc->flush_required = false;
+ put_cpu();
+}
+
+static void set_tlb_ubc_flush_pending(struct mm_struct *mm,
+ struct page *page)
+{
+ struct tlbflush_unmap_batch *tlb_ubc = ¤t->tlb_ubc;
+
+ cpumask_or(&tlb_ubc->cpumask, &tlb_ubc->cpumask, mm_cpumask(mm));
+ tlb_ubc->flush_required = true;
+}
+
+/*
+ * Returns true if the TLB flush should be deferred to the end of a batch of
+ * unmap operations to reduce IPIs.
+ */
+static bool should_defer_flush(struct mm_struct *mm, enum ttu_flags flags)
+{
+ bool should_defer = false;
+
+ if (!(flags & TTU_BATCH_FLUSH))
+ return false;
+
+ /* If remote CPUs need to be flushed then defer batch the flush */
+ if (cpumask_any_but(mm_cpumask(mm), get_cpu()) < nr_cpu_ids)
+ should_defer = true;
+ put_cpu();
+
+ return should_defer;
+}
+#else
+static void set_tlb_ubc_flush_pending(struct mm_struct *mm,
+ struct page *page)
+{
+}
+
+static bool should_defer_flush(struct mm_struct *mm, enum ttu_flags flags)
+{
+ return false;
+}
+#endif /* CONFIG_ARCH_WANT_BATCHED_UNMAP_TLB_FLUSH */
+
/*
* At what user virtual address is page expected in vma?
* Caller should check the page is actually part of the vma.
@@ -1220,7 +1305,24 @@ static int try_to_unmap_one(struct page *page, struct vm_area_struct *vma,
/* Nuke the page table entry. */
flush_cache_page(vma, address, page_to_pfn(page));
- pteval = ptep_clear_flush(vma, address, pte);
+ if (should_defer_flush(mm, flags)) {
+ /*
+ * We clear the PTE but do not flush so potentially a remote
+ * CPU could still be writing to the page. If the entry was
+ * previously clean then the architecture must guarantee that
+ * a clear->dirty transition on a cached TLB entry is written
+ * through and traps if the PTE is unmapped.
+ */
+ pteval = ptep_get_and_clear(mm, address, pte);
+
+ /* Potentially writable TLBs must be flushed before IO */
+ if (pte_dirty(pteval))
+ flush_tlb_page(vma, address);
+ else
+ set_tlb_ubc_flush_pending(mm, page);
+ } else {
+ pteval = ptep_clear_flush(vma, address, pte);
+ }
/* Move the dirty bit to the physical page now the pte is gone. */
if (pte_dirty(pteval))
diff --git a/mm/vmscan.c b/mm/vmscan.c
index 8286938c70de..99ec00d6a5dd 100644
--- a/mm/vmscan.c
+++ b/mm/vmscan.c
@@ -1057,7 +1057,8 @@ static unsigned long shrink_page_list(struct list_head *page_list,
* processes. Try to unmap it here.
*/
if (page_mapped(page) && mapping) {
- switch (try_to_unmap(page, ttu_flags)) {
+ switch (try_to_unmap(page,
+ ttu_flags|TTU_BATCH_FLUSH)) {
case SWAP_FAIL:
goto activate_locked;
case SWAP_AGAIN:
@@ -1208,6 +1209,7 @@ keep:
}
mem_cgroup_uncharge_list(&free_pages);
+ try_to_unmap_flush();
free_hot_cold_page_list(&free_pages, true);
list_splice(&ret_pages, page_list);
@@ -2151,6 +2153,23 @@ out:
}
}
+#ifdef CONFIG_ARCH_WANT_BATCHED_UNMAP_TLB_FLUSH
+static void init_tlb_ubc(void)
+{
+ /*
+ * This deliberately does not clear the cpumask as it's expensive
+ * and unnecessary. If there happens to be data in there then the
+ * first SWAP_CLUSTER_MAX pages will send an unnecessary IPI and
+ * then will be cleared.
+ */
+ current->tlb_ubc.flush_required = false;
+}
+#else
+static inline void init_tlb_ubc(void)
+{
+}
+#endif /* CONFIG_ARCH_WANT_BATCHED_UNMAP_TLB_FLUSH */
+
/*
* This is a basic per-zone page freer. Used by both kswapd and direct reclaim.
*/
@@ -2185,6 +2204,8 @@ static void shrink_lruvec(struct lruvec *lruvec, int swappiness,
scan_adjusted = (global_reclaim(sc) && !current_is_kswapd() &&
sc->priority == DEF_PRIORITY);
+ init_tlb_ubc();
+
blk_start_plug(&plug);
while (nr[LRU_INACTIVE_ANON] || nr[LRU_ACTIVE_FILE] ||
nr[LRU_INACTIVE_FILE]) {
--
cgit v1.2.3-70-g09d2
From d950c9477d51f0cefc2ed3cf76e695d46af0d9c1 Mon Sep 17 00:00:00 2001
From: Mel Gorman <mgorman@suse.de>
Date: Fri, 4 Sep 2015 15:47:35 -0700
Subject: mm: defer flush of writable TLB entries
If a PTE is unmapped and it's dirty then it was writable recently. Due to
deferred TLB flushing, it's best to assume a writable TLB cache entry
exists. With that assumption, the TLB must be flushed before any IO can
start or the page is freed to avoid lost writes or data corruption. This
patch defers flushing of potentially writable TLBs as long as possible.
Signed-off-by: Mel Gorman <mgorman@suse.de>
Reviewed-by: Rik van Riel <riel@redhat.com>
Cc: Dave Hansen <dave.hansen@intel.com>
Acked-by: Ingo Molnar <mingo@kernel.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
include/linux/sched.h | 7 +++++++
mm/internal.h | 4 ++++
mm/rmap.c | 28 +++++++++++++++++++++-------
mm/vmscan.c | 7 ++++++-
4 files changed, 38 insertions(+), 8 deletions(-)
(limited to 'include/linux/sched.h')
diff --git a/include/linux/sched.h b/include/linux/sched.h
index 3c602c20c717..a4ab9daa387c 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -1354,6 +1354,13 @@ struct tlbflush_unmap_batch {
/* True if any bit in cpumask is set */
bool flush_required;
+
+ /*
+ * If true then the PTE was dirty when unmapped. The entry must be
+ * flushed before IO is initiated or a stale TLB entry potentially
+ * allows an update without redirtying the page.
+ */
+ bool writable;
};
struct task_struct {
diff --git a/mm/internal.h b/mm/internal.h
index bd6372ac5f7f..1195dd2d6a2b 100644
--- a/mm/internal.h
+++ b/mm/internal.h
@@ -431,10 +431,14 @@ struct tlbflush_unmap_batch;
#ifdef CONFIG_ARCH_WANT_BATCHED_UNMAP_TLB_FLUSH
void try_to_unmap_flush(void);
+void try_to_unmap_flush_dirty(void);
#else
static inline void try_to_unmap_flush(void)
{
}
+static inline void try_to_unmap_flush_dirty(void)
+{
+}
#endif /* CONFIG_ARCH_WANT_BATCHED_UNMAP_TLB_FLUSH */
#endif /* __MM_INTERNAL_H */
diff --git a/mm/rmap.c b/mm/rmap.c
index 326d5d89e45c..0db38e7d0a72 100644
--- a/mm/rmap.c
+++ b/mm/rmap.c
@@ -626,16 +626,34 @@ void try_to_unmap_flush(void)
}
cpumask_clear(&tlb_ubc->cpumask);
tlb_ubc->flush_required = false;
+ tlb_ubc->writable = false;
put_cpu();
}
+/* Flush iff there are potentially writable TLB entries that can race with IO */
+void try_to_unmap_flush_dirty(void)
+{
+ struct tlbflush_unmap_batch *tlb_ubc = ¤t->tlb_ubc;
+
+ if (tlb_ubc->writable)
+ try_to_unmap_flush();
+}
+
static void set_tlb_ubc_flush_pending(struct mm_struct *mm,
- struct page *page)
+ struct page *page, bool writable)
{
struct tlbflush_unmap_batch *tlb_ubc = ¤t->tlb_ubc;
cpumask_or(&tlb_ubc->cpumask, &tlb_ubc->cpumask, mm_cpumask(mm));
tlb_ubc->flush_required = true;
+
+ /*
+ * If the PTE was dirty then it's best to assume it's writable. The
+ * caller must use try_to_unmap_flush_dirty() or try_to_unmap_flush()
+ * before the page is queued for IO.
+ */
+ if (writable)
+ tlb_ubc->writable = true;
}
/*
@@ -658,7 +676,7 @@ static bool should_defer_flush(struct mm_struct *mm, enum ttu_flags flags)
}
#else
static void set_tlb_ubc_flush_pending(struct mm_struct *mm,
- struct page *page)
+ struct page *page, bool writable)
{
}
@@ -1315,11 +1333,7 @@ static int try_to_unmap_one(struct page *page, struct vm_area_struct *vma,
*/
pteval = ptep_get_and_clear(mm, address, pte);
- /* Potentially writable TLBs must be flushed before IO */
- if (pte_dirty(pteval))
- flush_tlb_page(vma, address);
- else
- set_tlb_ubc_flush_pending(mm, page);
+ set_tlb_ubc_flush_pending(mm, page, pte_dirty(pteval));
} else {
pteval = ptep_clear_flush(vma, address, pte);
}
diff --git a/mm/vmscan.c b/mm/vmscan.c
index 99ec00d6a5dd..b1139039122a 100644
--- a/mm/vmscan.c
+++ b/mm/vmscan.c
@@ -1098,7 +1098,12 @@ static unsigned long shrink_page_list(struct list_head *page_list,
if (!sc->may_writepage)
goto keep_locked;
- /* Page is dirty, try to write it out here */
+ /*
+ * Page is dirty. Flush the TLB if a writable entry
+ * potentially exists to avoid CPU writes after IO
+ * starts and then write it out here.
+ */
+ try_to_unmap_flush_dirty();
switch (pageout(page, mapping, sc)) {
case PAGE_KEEP:
goto keep_locked;
--
cgit v1.2.3-70-g09d2