mm: memcontrol: use special workqueue for creating per-memcg caches

debian/changelog

@@ -68,6 +68,7 @@ linux (4.7.6-1) UNRELEASED; urgency=medium
   * of_mdio: Enable fixed PHY support if driver is a module
   * of_mdio: select fixed phy support unconditionally
   * [armhf] Enable driver for SolidRun ClearFog: USB_XHCI_MVEBU as module
+  * mm: memcontrol: use special workqueue for creating per-memcg caches
 
   [ Cyril Brulebois ]
   * Add nls_ascii to the fat-modules udeb, following the change of

debian/patches/bugfix/all/mm-memcontrol-use-special-workqueue-for-creating-per-memcg-caches.patch

@@ -0,0 +1,73 @@
+From: Vladimir Davydov <vdavydov.dev@gmail.com>
+Date: Sat, 1 Oct 2016 16:39:09 +0300
+Subject: mm: memcontrol: use special workqueue for creating per-memcg caches
+Origin: https://patchwork.kernel.org/patch/9361853/
+Bug: https://bugzilla.kernel.org/show_bug.cgi?id=172981
+
+Creating a lot of cgroups at the same time might stall all worker
+threads with kmem cache creation works, because kmem cache creation is
+done with the slab_mutex held. The problem was amplified by commits
+801faf0db894 ("mm/slab: lockless decision to grow cache") in case of
+SLAB and 81ae6d03952c ("mm/slub.c: replace kick_all_cpus_sync() with
+synchronize_sched() in kmem_cache_shrink()") in case of SLUB, which
+increased the maximal time the slab_mutex can be held.
+
+To prevent that from happening, let's use a special ordered single
+threaded workqueue for kmem cache creation. This shouldn't introduce any
+functional changes regarding how kmem caches are created, as the work
+function holds the global slab_mutex during its whole runtime anyway,
+making it impossible to run more than one work at a time. By using a
+single threaded workqueue, we just avoid creating a thread per each
+work. Ordering is required to avoid a situation when a cgroup's work is
+put off indefinitely because there are other cgroups to serve, in other
+words to guarantee fairness.
+
+Link: https://bugzilla.kernel.org/show_bug.cgi?id=172981
+Signed-off-by: Vladimir Davydov <vdavydov.dev@gmail.com>
+Reported-by: Doug Smythies <dsmythies@telus.net>
+Cc: Christoph Lameter <cl@linux.com>
+Cc: David Rientjes <rientjes@google.com>
+Cc: Johannes Weiner <hannes@cmpxchg.org>
+Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com>
+Cc: Michal Hocko <mhocko@kernel.org>
+Cc: Pekka Enberg <penberg@kernel.org>
+Acked-by: Michal Hocko <mhocko@suse.com>
+---
+--- a/mm/memcontrol.c
++++ b/mm/memcontrol.c
+@@ -2232,6 +2232,8 @@ struct memcg_kmem_cache_create_work {
+ 	struct work_struct work;
+ };
+ 
++static struct workqueue_struct *memcg_kmem_cache_create_wq;
++
+ static void memcg_kmem_cache_create_func(struct work_struct *w)
+ {
+ 	struct memcg_kmem_cache_create_work *cw =
+@@ -2263,7 +2265,7 @@ static void __memcg_schedule_kmem_cache_
+ 	cw->cachep = cachep;
+ 	INIT_WORK(&cw->work, memcg_kmem_cache_create_func);
+ 
+-	schedule_work(&cw->work);
++	queue_work(memcg_kmem_cache_create_wq, &cw->work);
+ }
+ 
+ static void memcg_schedule_kmem_cache_create(struct mem_cgroup *memcg,
+@@ -5811,6 +5813,17 @@ static int __init mem_cgroup_init(void)
+ {
+ 	int cpu, node;
+ 
++#ifndef CONFIG_SLOB
++	/*
++	 * Kmem cache creation is mostly done with the slab_mutex held,
++	 * so use a special workqueue to avoid stalling all worker
++	 * threads in case lots of cgroups are created simultaneously.
++	 */
++	memcg_kmem_cache_create_wq =
++		alloc_ordered_workqueue("memcg_kmem_cache_create", 0);
++	BUG_ON(!memcg_kmem_cache_create_wq);
++#endif
++
+ 	hotcpu_notifier(memcg_cpu_hotplug_callback, 0);
+ 
+ 	for_each_possible_cpu(cpu)

debian/patches/series

@@ -77,6 +77,7 @@ bugfix/all/kbuild-do-not-use-hyphen-in-exported-variable-name.patch
 bugfix/all/ext4-fix-bug-838544.patch
 bugfix/all/of_mdio-enable-fixed-phy-support-if-driver-is-a-modu.patch
 bugfix/all/of_mdio-select-fixed-phy-support-unconditionally.patch
+bugfix/all/mm-memcontrol-use-special-workqueue-for-creating-per-memcg-caches.patch
 
 # Miscellaneous features