cpuset: mm: reduce large amounts of memory barrier related damage v3

Commit c0ff7453 ("cpuset,mm: fix no node to alloc memory when
changing cpuset's mems") wins a super prize for the largest number of
memory barriers entered into fast paths for one commit.

[get|put]_mems_allowed is incredibly heavy with pairs of full memory
barriers inserted into a number of hot paths.  This was detected while
investigating at large page allocator slowdown introduced some time
after 2.6.32.  The largest portion of this overhead was shown by
oprofile to be at an mfence introduced by this commit into the page
allocator hot path.

For extra style points, the commit introduced the use of yield() in an
implementation of what looks like a spinning mutex.

This patch replaces the full memory barriers on both read and write
sides with a sequence counter with just read barriers on the fast path
side.  This is much cheaper on some architectures, including x86.  The
main bulk of the patch is the retry logic if the nodemask changes in a
manner that can cause a false failure.

While updating the nodemask, a check is made to see if a false failure
is a risk.  If it is, the sequence number gets bumped and parallel
allocators will briefly stall while the nodemask update takes place.

In a page fault test microbenchmark, oprofile samples from
__alloc_pages_nodemask went from 4.53% of all samples to 1.15%.  The
actual results were

                             3.3.0-rc3          3.3.0-rc3
                             rc3-vanilla        nobarrier-v2r1
    Clients   1 UserTime       0.07 (  0.00%)   0.08 (-14.19%)
    Clients   2 UserTime       0.07 (  0.00%)   0.07 (  2.72%)
    Clients   4 UserTime       0.08 (  0.00%)   0.07 (  3.29%)
    Clients   1 SysTime        0.70 (  0.00%)   0.65 (  6.65%)
    Clients   2 SysTime        0.85 (  0.00%)   0.82 (  3.65%)
    Clients   4 SysTime        1.41 (  0.00%)   1.41 (  0.32%)
    Clients   1 WallTime       0.77 (  0.00%)   0.74 (  4.19%)
    Clients   2 WallTime       0.47 (  0.00%)   0.45 (  3.73%)
    Clients   4 WallTime       0.38 (  0.00%)   0.37 (  1.58%)
    Clients   1 Flt/sec/cpu  497620.28 (  0.00%) 520294.53 (  4.56%)
    Clients   2 Flt/sec/cpu  414639.05 (  0.00%) 429882.01 (  3.68%)
    Clients   4 Flt/sec/cpu  257959.16 (  0.00%) 258761.48 (  0.31%)
    Clients   1 Flt/sec      495161.39 (  0.00%) 517292.87 (  4.47%)
    Clients   2 Flt/sec      820325.95 (  0.00%) 850289.77 (  3.65%)
    Clients   4 Flt/sec      1020068.93 (  0.00%) 1022674.06 (  0.26%)
    MMTests Statistics: duration
    Sys Time Running Test (seconds)             135.68    132.17
    User+Sys Time Running Test (seconds)         164.2    160.13
    Total Elapsed Time (seconds)                123.46    120.87

The overall improvement is small but the System CPU time is much
improved and roughly in correlation to what oprofile reported (these
performance figures are without profiling so skew is expected).  The
actual number of page faults is noticeably improved.

For benchmarks like kernel builds, the overall benefit is marginal but
the system CPU time is slightly reduced.

To test the actual bug the commit fixed I opened two terminals.  The
first ran within a cpuset and continually ran a small program that
faulted 100M of anonymous data.  In a second window, the nodemask of the
cpuset was continually randomised in a loop.

Without the commit, the program would fail every so often (usually
within 10 seconds) and obviously with the commit everything worked fine.
With this patch applied, it also worked fine so the fix should be
functionally equivalent.
Signed-off-by: Mel Gorman <mgorman@suse.de>
Cc: Miao Xie <miaox@cn.fujitsu.com>
Cc: David Rientjes <rientjes@google.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Christoph Lameter <cl@linux.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

include/linux/cpuset.h

@@ -89,42 +89,33 @@ extern void rebuild_sched_domains(void);
 extern void cpuset_print_task_mems_allowed(struct task_struct *p);
 
 /*
- * reading current mems_allowed and mempolicy in the fastpath must protected
- * by get_mems_allowed()
+ * get_mems_allowed is required when making decisions involving mems_allowed
+ * such as during page allocation. mems_allowed can be updated in parallel
+ * and depending on the new value an operation can fail potentially causing
+ * process failure. A retry loop with get_mems_allowed and put_mems_allowed
+ * prevents these artificial failures.
  */
-static inline void get_mems_allowed(void)
+static inline unsigned int get_mems_allowed(void)
 {
-	current->mems_allowed_change_disable++;
-
-	/*
-	 * ensure that reading mems_allowed and mempolicy happens after the
-	 * update of ->mems_allowed_change_disable.
-	 *
-	 * the write-side task finds ->mems_allowed_change_disable is not 0,
-	 * and knows the read-side task is reading mems_allowed or mempolicy,
-	 * so it will clear old bits lazily.
-	 */
-	smp_mb();
+	return read_seqcount_begin(&current->mems_allowed_seq);
 }
 
-static inline void put_mems_allowed(void)
-{
-	/*
-	 * ensure that reading mems_allowed and mempolicy before reducing
-	 * mems_allowed_change_disable.
-	 *
-	 * the write-side task will know that the read-side task is still
-	 * reading mems_allowed or mempolicy, don't clears old bits in the
-	 * nodemask.
+/*
+ * If this returns false, the operation that took place after get_mems_allowed
+ * may have failed. It is up to the caller to retry the operation if
+ * appropriate.
  */
-	smp_mb();
-	--ACCESS_ONCE(current->mems_allowed_change_disable);
+static inline bool put_mems_allowed(unsigned int seq)
+{
+	return !read_seqcount_retry(&current->mems_allowed_seq, seq);
 }
 
 static inline void set_mems_allowed(nodemask_t nodemask)
 {
 	task_lock(current);
+	write_seqcount_begin(&current->mems_allowed_seq);
 	current->mems_allowed = nodemask;
+	write_seqcount_end(&current->mems_allowed_seq);
 	task_unlock(current);
 }
 
@@ -234,12 +225,14 @@ static inline void set_mems_allowed(nodemask_t nodemask)
 {
 }
 
-static inline void get_mems_allowed(void)
+static inline unsigned int get_mems_allowed(void)
 {
+	return 0;
 }
 
-static inline void put_mems_allowed(void)
+static inline bool put_mems_allowed(unsigned int seq)
 {
+	return true;
 }
 
 #endif /* !CONFIG_CPUSETS */

include/linux/init_task.h

@@ -29,6 +29,13 @@ extern struct fs_struct init_fs;
 #define INIT_GROUP_RWSEM(sig)
 #endif
 
+#ifdef CONFIG_CPUSETS
+#define INIT_CPUSET_SEQ							\
+	.mems_allowed_seq = SEQCNT_ZERO,
+#else
+#define INIT_CPUSET_SEQ
+#endif
+
 #define INIT_SIGNALS(sig) {						\
 	.nr_threads	= 1,						\
 	.wait_chldexit	= __WAIT_QUEUE_HEAD_INITIALIZER(sig.wait_chldexit),\
@@ -192,6 +199,7 @@ extern struct cred init_cred;
 	INIT_FTRACE_GRAPH						\
 	INIT_TRACE_RECURSION						\
 	INIT_TASK_RCU_PREEMPT(tsk)					\
+	INIT_CPUSET_SEQ							\
 }
 
 

include/linux/sched.h

@@ -1514,7 +1514,7 @@ struct task_struct {
 #endif
 #ifdef CONFIG_CPUSETS
 	nodemask_t mems_allowed;	/* Protected by alloc_lock */
-	int mems_allowed_change_disable;
+	seqcount_t mems_allowed_seq;	/* Seqence no to catch updates */
 	int cpuset_mem_spread_rotor;
 	int cpuset_slab_spread_rotor;
 #endif

kernel/cpuset.c

@@ -964,7 +964,6 @@ static void cpuset_change_task_nodemask(struct task_struct *tsk,
 {
 	bool need_loop;
 
-repeat:
 	/*
 	 * Allow tasks that have access to memory reserves because they have
 	 * been OOM killed to get memory anywhere.
@@ -983,45 +982,19 @@ static void cpuset_change_task_nodemask(struct task_struct *tsk,
 	 */
 	need_loop = task_has_mempolicy(tsk) ||
 			!nodes_intersects(*newmems, tsk->mems_allowed);
-	nodes_or(tsk->mems_allowed, tsk->mems_allowed, *newmems);
-	mpol_rebind_task(tsk, newmems, MPOL_REBIND_STEP1);
 
-	/*
-	 * ensure checking ->mems_allowed_change_disable after setting all new
-	 * allowed nodes.
-	 *
-	 * the read-side task can see an nodemask with new allowed nodes and
-	 * old allowed nodes. and if it allocates page when cpuset clears newly
-	 * disallowed ones continuous, it can see the new allowed bits.
-	 *
-	 * And if setting all new allowed nodes is after the checking, setting
-	 * all new allowed nodes and clearing newly disallowed ones will be done
-	 * continuous, and the read-side task may find no node to alloc page.
-	 */
-	smp_mb();
-
-	/*
-	 * Allocation of memory is very fast, we needn't sleep when waiting
-	 * for the read-side.
-	 */
-	while (need_loop && ACCESS_ONCE(tsk->mems_allowed_change_disable)) {
-		task_unlock(tsk);
-		if (!task_curr(tsk))
-			yield();
-		goto repeat;
-	}
+	if (need_loop)
+		write_seqcount_begin(&tsk->mems_allowed_seq);
 
-	/*
-	 * ensure checking ->mems_allowed_change_disable before clearing all new
-	 * disallowed nodes.
-	 *
-	 * if clearing newly disallowed bits before the checking, the read-side
-	 * task may find no node to alloc page.
-	 */
-	smp_mb();
+	nodes_or(tsk->mems_allowed, tsk->mems_allowed, *newmems);
+	mpol_rebind_task(tsk, newmems, MPOL_REBIND_STEP1);
 
 	mpol_rebind_task(tsk, newmems, MPOL_REBIND_STEP2);
 	tsk->mems_allowed = *newmems;
+
+	if (need_loop)
+		write_seqcount_end(&tsk->mems_allowed_seq);
+
 	task_unlock(tsk);
 }
 

kernel/fork.c

@@ -1237,6 +1237,7 @@ static struct task_struct *copy_process(unsigned long clone_flags,
 #ifdef CONFIG_CPUSETS
 	p->cpuset_mem_spread_rotor = NUMA_NO_NODE;
 	p->cpuset_slab_spread_rotor = NUMA_NO_NODE;
+	seqcount_init(&p->mems_allowed_seq);
 #endif
 #ifdef CONFIG_TRACE_IRQFLAGS
 	p->irq_events = 0;

mm/filemap.c

@@ -499,10 +499,13 @@ struct page *__page_cache_alloc(gfp_t gfp)
 	struct page *page;
 
 	if (cpuset_do_page_mem_spread()) {
-		get_mems_allowed();
+		unsigned int cpuset_mems_cookie;
+		do {
+			cpuset_mems_cookie = get_mems_allowed();
 			n = cpuset_mem_spread_node();
 			page = alloc_pages_exact_node(n, gfp, 0);
-		put_mems_allowed();
+		} while (!put_mems_allowed(cpuset_mems_cookie) && !page);
+
 		return page;
 	}
 	return alloc_pages(gfp, 0);

mm/hugetlb.c

@@ -454,14 +454,16 @@ static struct page *dequeue_huge_page_vma(struct hstate *h,
 				struct vm_area_struct *vma,
 				unsigned long address, int avoid_reserve)
 {
-	struct page *page = NULL;
+	struct page *page;
 	struct mempolicy *mpol;
 	nodemask_t *nodemask;
 	struct zonelist *zonelist;
 	struct zone *zone;
 	struct zoneref *z;
+	unsigned int cpuset_mems_cookie;
 
-	get_mems_allowed();
+retry_cpuset:
+	cpuset_mems_cookie = get_mems_allowed();
 	zonelist = huge_zonelist(vma, address,
 					htlb_alloc_mask, &mpol, &nodemask);
 	/*
@@ -488,10 +490,15 @@ static struct page *dequeue_huge_page_vma(struct hstate *h,
 			}
 		}
 	}
-err:
+
 	mpol_cond_put(mpol);
-	put_mems_allowed();
+	if (unlikely(!put_mems_allowed(cpuset_mems_cookie) && !page))
+		goto retry_cpuset;
 	return page;
+
+err:
+	mpol_cond_put(mpol);
+	return NULL;
 }
 
 static void update_and_free_page(struct hstate *h, struct page *page)

mm/mempolicy.c

@@ -1850,18 +1850,24 @@ struct page *
 alloc_pages_vma(gfp_t gfp, int order, struct vm_area_struct *vma,
 		unsigned long addr, int node)
 {
-	struct mempolicy *pol = get_vma_policy(current, vma, addr);
+	struct mempolicy *pol;
 	struct zonelist *zl;
 	struct page *page;
+	unsigned int cpuset_mems_cookie;
+
+retry_cpuset:
+	pol = get_vma_policy(current, vma, addr);
+	cpuset_mems_cookie = get_mems_allowed();
 
-	get_mems_allowed();
 	if (unlikely(pol->mode == MPOL_INTERLEAVE)) {
 		unsigned nid;
 
 		nid = interleave_nid(pol, vma, addr, PAGE_SHIFT + order);
 		mpol_cond_put(pol);
 		page = alloc_page_interleave(gfp, order, nid);
-		put_mems_allowed();
+		if (unlikely(!put_mems_allowed(cpuset_mems_cookie) && !page))
+			goto retry_cpuset;
+
 		return page;
 	}
 	zl = policy_zonelist(gfp, pol, node);
@@ -1872,7 +1878,8 @@ alloc_pages_vma(gfp_t gfp, int order, struct vm_area_struct *vma,
 		struct page *page =  __alloc_pages_nodemask(gfp, order,
 						zl, policy_nodemask(gfp, pol));
 		__mpol_put(pol);
-		put_mems_allowed();
+		if (unlikely(!put_mems_allowed(cpuset_mems_cookie) && !page))
+			goto retry_cpuset;
 		return page;
 	}
 	/*
@@ -1880,7 +1887,8 @@ alloc_pages_vma(gfp_t gfp, int order, struct vm_area_struct *vma,
 	 */
 	page = __alloc_pages_nodemask(gfp, order, zl,
 				      policy_nodemask(gfp, pol));
-	put_mems_allowed();
+	if (unlikely(!put_mems_allowed(cpuset_mems_cookie) && !page))
+		goto retry_cpuset;
 	return page;
 }
 
@@ -1907,11 +1915,14 @@ struct page *alloc_pages_current(gfp_t gfp, unsigned order)
 {
 	struct mempolicy *pol = current->mempolicy;
 	struct page *page;
+	unsigned int cpuset_mems_cookie;
 
 	if (!pol || in_interrupt() || (gfp & __GFP_THISNODE))
 		pol = &default_policy;
 
-	get_mems_allowed();
+retry_cpuset:
+	cpuset_mems_cookie = get_mems_allowed();
+
 	/*
 	 * No reference counting needed for current->mempolicy
 	 * nor system default_policy
@@ -1922,7 +1933,10 @@ struct page *alloc_pages_current(gfp_t gfp, unsigned order)
 		page = __alloc_pages_nodemask(gfp, order,
 				policy_zonelist(gfp, pol, numa_node_id()),
 				policy_nodemask(gfp, pol));
-	put_mems_allowed();
+
+	if (unlikely(!put_mems_allowed(cpuset_mems_cookie) && !page))
+		goto retry_cpuset;
+
 	return page;
 }
 EXPORT_SYMBOL(alloc_pages_current);

mm/page_alloc.c

@@ -2380,8 +2380,9 @@ __alloc_pages_nodemask(gfp_t gfp_mask, unsigned int order,
 {
 	enum zone_type high_zoneidx = gfp_zone(gfp_mask);
 	struct zone *preferred_zone;
-	struct page *page;
+	struct page *page = NULL;
 	int migratetype = allocflags_to_migratetype(gfp_mask);
+	unsigned int cpuset_mems_cookie;
 
 	gfp_mask &= gfp_allowed_mask;
 
@@ -2400,15 +2401,15 @@ __alloc_pages_nodemask(gfp_t gfp_mask, unsigned int order,
 	if (unlikely(!zonelist->_zonerefs->zone))
 		return NULL;
 
-	get_mems_allowed();
+retry_cpuset:
+	cpuset_mems_cookie = get_mems_allowed();
+
 	/* The preferred zone is used for statistics later */
 	first_zones_zonelist(zonelist, high_zoneidx,
 				nodemask ? : &cpuset_current_mems_allowed,
 				&preferred_zone);
-	if (!preferred_zone) {
-		put_mems_allowed();
-		return NULL;
-	}
+	if (!preferred_zone)
+		goto out;
 
 	/* First allocation attempt */
 	page = get_page_from_freelist(gfp_mask|__GFP_HARDWALL, nodemask, order,
@@ -2418,9 +2419,19 @@ __alloc_pages_nodemask(gfp_t gfp_mask, unsigned int order,
 		page = __alloc_pages_slowpath(gfp_mask, order,
 				zonelist, high_zoneidx, nodemask,
 				preferred_zone, migratetype);
-	put_mems_allowed();
 
 	trace_mm_page_alloc(page, order, gfp_mask, migratetype);
+
+out:
+	/*
+	 * When updating a task's mems_allowed, it is possible to race with
+	 * parallel threads in such a way that an allocation can fail while
+	 * the mask is being updated. If a page allocation is about to fail,
+	 * check if the cpuset changed during allocation and if so, retry.
+	 */
+	if (unlikely(!put_mems_allowed(cpuset_mems_cookie) && !page))
+		goto retry_cpuset;
+
 	return page;
 }
 EXPORT_SYMBOL(__alloc_pages_nodemask);
@@ -2634,13 +2645,15 @@ void si_meminfo_node(struct sysinfo *val, int nid)
 bool skip_free_areas_node(unsigned int flags, int nid)
 {
 	bool ret = false;
+	unsigned int cpuset_mems_cookie;
 
 	if (!(flags & SHOW_MEM_FILTER_NODES))
 		goto out;
 
-	get_mems_allowed();
+	do {
+		cpuset_mems_cookie = get_mems_allowed();
 		ret = !node_isset(nid, cpuset_current_mems_allowed);
-	put_mems_allowed();
+	} while (!put_mems_allowed(cpuset_mems_cookie));
 out:
 	return ret;
 }

mm/slab.c

@@ -3284,12 +3284,10 @@ static void *alternate_node_alloc(struct kmem_cache *cachep, gfp_t flags)
 	if (in_interrupt() || (flags & __GFP_THISNODE))
 		return NULL;
 	nid_alloc = nid_here = numa_mem_id();
-	get_mems_allowed();
 	if (cpuset_do_slab_mem_spread() && (cachep->flags & SLAB_MEM_SPREAD))
 		nid_alloc = cpuset_slab_spread_node();
 	else if (current->mempolicy)
 		nid_alloc = slab_node(current->mempolicy);
-	put_mems_allowed();
 	if (nid_alloc != nid_here)
 		return ____cache_alloc_node(cachep, flags, nid_alloc);
 	return NULL;
@@ -3312,14 +3310,17 @@ static void *fallback_alloc(struct kmem_cache *cache, gfp_t flags)
 	enum zone_type high_zoneidx = gfp_zone(flags);
 	void *obj = NULL;
 	int nid;
+	unsigned int cpuset_mems_cookie;
 
 	if (flags & __GFP_THISNODE)
 		return NULL;
 
-	get_mems_allowed();
-	zonelist = node_zonelist(slab_node(current->mempolicy), flags);
 	local_flags = flags & (GFP_CONSTRAINT_MASK|GFP_RECLAIM_MASK);
 
+retry_cpuset:
+	cpuset_mems_cookie = get_mems_allowed();
+	zonelist = node_zonelist(slab_node(current->mempolicy), flags);
+
 retry:
 	/*
 	 * Look through allowed nodes for objects available
@@ -3372,7 +3373,9 @@ static void *fallback_alloc(struct kmem_cache *cache, gfp_t flags)
 			}
 		}
 	}
-	put_mems_allowed();
+
+	if (unlikely(!put_mems_allowed(cpuset_mems_cookie) && !obj))
+		goto retry_cpuset;
 	return obj;
 }
 

mm/slub.c

@@ -1581,6 +1581,7 @@ static struct page *get_any_partial(struct kmem_cache *s, gfp_t flags,
 	struct zone *zone;
 	enum zone_type high_zoneidx = gfp_zone(flags);
 	void *object;
+	unsigned int cpuset_mems_cookie;
 
 	/*
 	 * The defrag ratio allows a configuration of the tradeoffs between
@@ -1604,7 +1605,8 @@ static struct page *get_any_partial(struct kmem_cache *s, gfp_t flags,
 			get_cycles() % 1024 > s->remote_node_defrag_ratio)
 		return NULL;
 
-	get_mems_allowed();
+	do {
+		cpuset_mems_cookie = get_mems_allowed();
 		zonelist = node_zonelist(slab_node(current->mempolicy), flags);
 		for_each_zone_zonelist(zone, z, zonelist, high_zoneidx) {
 			struct kmem_cache_node *n;
@@ -1615,12 +1617,20 @@ static struct page *get_any_partial(struct kmem_cache *s, gfp_t flags,
 					n->nr_partial > s->min_partial) {
 				object = get_partial_node(s, n, c);
 				if (object) {
-				put_mems_allowed();
+					/*
+					 * Return the object even if
+					 * put_mems_allowed indicated that
+					 * the cpuset mems_allowed was
+					 * updated in parallel. It's a
+					 * harmless race between the alloc
+					 * and the cpuset update.
+					 */
+					put_mems_allowed(cpuset_mems_cookie);
 					return object;
 				}
 			}
 		}
-	put_mems_allowed();
+	} while (!put_mems_allowed(cpuset_mems_cookie));
 #endif
 	return NULL;
 }

mm/vmscan.c

@@ -2343,7 +2343,6 @@ static unsigned long do_try_to_free_pages(struct zonelist *zonelist,
 	unsigned long writeback_threshold;
 	bool aborted_reclaim;
 
-	get_mems_allowed();
 	delayacct_freepages_start();
 
 	if (global_reclaim(sc))
@@ -2407,7 +2406,6 @@ static unsigned long do_try_to_free_pages(struct zonelist *zonelist,
 
 out:
 	delayacct_freepages_end();
-	put_mems_allowed();
 
 	if (sc->nr_reclaimed)
 		return sc->nr_reclaimed;