[PATCH v4 3/3] cpuset: PF_SPREAD_PAGE and PF_SPREAD_SLAB should be atomic flags
From: Zefan Li <hidden>
Date: 2014-09-25 01:41:23
Also in:
lkml
Subsystem:
control group - cpuset, documentation, memory management, scheduler, the rest · Maintainers:
Waiman Long, Jonathan Corbet, Andrew Morton, Ingo Molnar, Peter Zijlstra, Juri Lelli, Vincent Guittot, Linus Torvalds
When we change cpuset.memory_spread_{page,slab}, cpuset will flip
PF_SPREAD_{PAGE,SLAB} bit of tsk->flags for each task in that cpuset.
This should be done using atomic bitops, but currently we don't,
which is broken.
Tetsuo reported a hard-to-reproduce kernel crash on RHEL6, which happened
when one thread tried to clear PF_USED_MATH while at the same time another
thread tried to flip PF_SPREAD_PAGE/PF_SPREAD_SLAB. They both operate on
the same task.
Here's the full report:
https://lkml.org/lkml/2014/9/19/230
To fix this, we make PF_SPREAD_PAGE and PF_SPREAD_SLAB atomic flags.
v4:
- updated mm/slab.c. (Fengguang Wu)
- updated Documentation.
Cc: Peter Zijlstra <redacted>
Cc: Ingo Molnar <redacted>
Cc: Miao Xie <redacted>
Cc: Kees Cook <redacted>
Fixes: 950592f7b991 ("cpusets: update tasks' page/slab spread flags in time")
Cc: <redacted> # 2.6.31+
Reported-by: Tetsuo Handa <penguin-kernel-JPay3/Yim36HaxMnTkn67Xf5DAMn2ifp@public.gmane.org>
Signed-off-by: Zefan Li <redacted>
---
Documentation/cgroups/cpusets.txt | 6 +++---
include/linux/cpuset.h | 4 ++--
include/linux/sched.h | 13 +++++++++++--
kernel/cpuset.c | 9 +++++----
mm/slab.c | 4 ++--
5 files changed, 23 insertions(+), 13 deletions(-)
diff --git a/Documentation/cgroups/cpusets.txt b/Documentation/cgroups/cpusets.txt
index 7740038..3c94ff3 100644
--- a/Documentation/cgroups/cpusets.txt
+++ b/Documentation/cgroups/cpusets.txt@@ -345,14 +345,14 @@ the named feature on. The implementation is simple. Setting the flag 'cpuset.memory_spread_page' turns on a per-process flag -PF_SPREAD_PAGE for each task that is in that cpuset or subsequently +PFA_SPREAD_PAGE for each task that is in that cpuset or subsequently joins that cpuset. The page allocation calls for the page cache -is modified to perform an inline check for this PF_SPREAD_PAGE task +is modified to perform an inline check for this PFA_SPREAD_PAGE task flag, and if set, a call to a new routine cpuset_mem_spread_node() returns the node to prefer for the allocation. Similarly, setting 'cpuset.memory_spread_slab' turns on the flag -PF_SPREAD_SLAB, and appropriately marked slab caches will allocate +PFA_SPREAD_SLAB, and appropriately marked slab caches will allocate pages from the node returned by cpuset_mem_spread_node(). The cpuset_mem_spread_node() routine is also simple. It uses the
diff --git a/include/linux/cpuset.h b/include/linux/cpuset.h
index 0d4e067..2f073db 100644
--- a/include/linux/cpuset.h
+++ b/include/linux/cpuset.h@@ -94,12 +94,12 @@ extern int cpuset_slab_spread_node(void); static inline int cpuset_do_page_mem_spread(void) { - return current->flags & PF_SPREAD_PAGE; + return task_spread_page(current); } static inline int cpuset_do_slab_mem_spread(void) { - return current->flags & PF_SPREAD_SLAB; + return task_spread_slab(current); } extern int current_cpuset_is_being_rebound(void);
diff --git a/include/linux/sched.h b/include/linux/sched.h
index 5630763..7b1cafe 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h@@ -1903,8 +1903,6 @@ extern void thread_group_cputime_adjusted(struct task_struct *p, cputime_t *ut, #define PF_KTHREAD 0x00200000 /* I am a kernel thread */ #define PF_RANDOMIZE 0x00400000 /* randomize virtual address space */ #define PF_SWAPWRITE 0x00800000 /* Allowed to write to swap */ -#define PF_SPREAD_PAGE 0x01000000 /* Spread page cache over cpuset */ -#define PF_SPREAD_SLAB 0x02000000 /* Spread some slab caches over cpuset */ #define PF_NO_SETAFFINITY 0x04000000 /* Userland is not allowed to meddle with cpus_allowed */ #define PF_MCE_EARLY 0x08000000 /* Early kill for mce process policy */ #define PF_MUTEX_TESTER 0x20000000 /* Thread belongs to the rt mutex tester */
@@ -1958,6 +1956,9 @@ static inline void memalloc_noio_restore(unsigned int flags) /* Per-process atomic flags. */ #define PFA_NO_NEW_PRIVS 0 /* May not gain new privileges. */ +#define PFA_SPREAD_PAGE 1 /* Spread page cache over cpuset */ +#define PFA_SPREAD_SLAB 2 /* Spread some slab caches over cpuset */ + #define TASK_PFA_TEST(name, func) \ static inline bool task_##func(struct task_struct *p) \
@@ -1972,6 +1973,14 @@ static inline void memalloc_noio_restore(unsigned int flags) TASK_PFA_TEST(NO_NEW_PRIVS, no_new_privs) TASK_PFA_SET(NO_NEW_PRIVS, no_new_privs) +TASK_PFA_TEST(SPREAD_PAGE, spread_page) +TASK_PFA_SET(SPREAD_PAGE, spread_page) +TASK_PFA_CLEAR(SPREAD_PAGE, spread_page) + +TASK_PFA_TEST(SPREAD_SLAB, spread_slab) +TASK_PFA_SET(SPREAD_SLAB, spread_slab) +TASK_PFA_CLEAR(SPREAD_SLAB, spread_slab) + /* * task->jobctl flags */
diff --git a/kernel/cpuset.c b/kernel/cpuset.c
index a37f4ed..1f107c7 100644
--- a/kernel/cpuset.c
+++ b/kernel/cpuset.c@@ -365,13 +365,14 @@ static void cpuset_update_task_spread_flag(struct cpuset *cs, struct task_struct *tsk) { if (is_spread_page(cs)) - tsk->flags |= PF_SPREAD_PAGE; + task_set_spread_page(tsk); else - tsk->flags &= ~PF_SPREAD_PAGE; + task_clear_spread_page(tsk); + if (is_spread_slab(cs)) - tsk->flags |= PF_SPREAD_SLAB; + task_set_spread_slab(tsk); else - tsk->flags &= ~PF_SPREAD_SLAB; + task_clear_spread_slab(tsk); } /*
diff --git a/mm/slab.c b/mm/slab.c
index a467b30..881951e 100644
--- a/mm/slab.c
+++ b/mm/slab.c@@ -2994,7 +2994,7 @@ out: #ifdef CONFIG_NUMA /* - * Try allocating on another node if PF_SPREAD_SLAB is a mempolicy is set. + * Try allocating on another node if PFA_SPREAD_SLAB is a mempolicy is set. * * If we are in_interrupt, then process context, including cpusets and * mempolicy, may not apply and should not be used for allocation policy.
@@ -3226,7 +3226,7 @@ __do_cache_alloc(struct kmem_cache *cache, gfp_t flags) { void *objp; - if (current->mempolicy || unlikely(current->flags & PF_SPREAD_SLAB)) { + if (current->mempolicy || cpuset_do_slab_mem_spread()) { objp = alternate_node_alloc(cache, flags); if (objp) goto out;
--
1.8.0.2