The patch titled
Subject: mm, slub: use migrate_disable() on PREEMPT_RT
has been added to the -mm tree. Its filename is
mm-slub-use-migrate_disable-on-preempt_rt.patch
This patch should soon appear at
https://ozlabs.org/~akpm/mmots/broken-out/mm-slub-use-migrate_disable-on-preempt_rt.patch
and later at
https://ozlabs.org/~akpm/mmotm/broken-out/mm-slub-use-migrate_disable-on-preempt_rt.patch
Before you just go and hit "reply", please:
a) Consider who else should be cc'ed
b) Prefer to cc a suitable mailing list as well
c) Ideally: find the original patch on the mailing list and do a
reply-to-all to that, adding suitable additional cc's
*** Remember to use Documentation/process/submit-checklist.rst when testing your code ***
The -mm tree is included into linux-next and is updated
there every 3-4 working days
------------------------------------------------------
From: Vlastimil Babka <vbabka@xxxxxxx>
Subject: mm, slub: use migrate_disable() on PREEMPT_RT
We currently use preempt_disable() (directly or via get_cpu_ptr()) to
stabilize the pointer to kmem_cache_cpu. On PREEMPT_RT this would be
incompatible with the list_lock spinlock. We can use migrate_disable()
instead, but that increases overhead on !PREEMPT_RT as it's an
unconditional function call.
In order to get the best available mechanism on both PREEMPT_RT and
!PREEMPT_RT, introduce private slub_get_cpu_ptr() and slub_put_cpu_ptr()
wrappers and use them.
Link: https://lkml.kernel.org/r/20210805152000.12817-35-vbabka@xxxxxxx
Signed-off-by: Vlastimil Babka <vbabka@xxxxxxx>
Cc: Christoph Lameter <cl@xxxxxxxxx>
Cc: David Rientjes <rientjes@xxxxxxxxxx>
Cc: Jann Horn <jannh@xxxxxxxxxx>
Cc: Jesper Dangaard Brouer <brouer@xxxxxxxxxx>
Cc: Joonsoo Kim <iamjoonsoo.kim@xxxxxxx>
Cc: Mel Gorman <mgorman@xxxxxxxxxxxxxxxxxxx>
Cc: Mike Galbraith <efault@xxxxxx>
Cc: Pekka Enberg <penberg@xxxxxxxxxx>
Cc: Sebastian Andrzej Siewior <bigeasy@xxxxxxxxxxxxx>
Cc: Thomas Gleixner <tglx@xxxxxxxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
mm/slub.c | 39 ++++++++++++++++++++++++++++++---------
1 file changed, 30 insertions(+), 9 deletions(-)
--- a/mm/slub.c~mm-slub-use-migrate_disable-on-preempt_rt
+++ a/mm/slub.c
@@ -118,6 +118,26 @@
* the fast path and disables lockless freelists.
*/
+/*
+ * We could simply use migrate_disable()/enable() but as long as it's a
+ * function call even on !PREEMPT_RT, use inline preempt_disable() there.
+ */
+#ifndef CONFIG_PREEMPT_RT
+#define slub_get_cpu_ptr(var) get_cpu_ptr(var)
+#define slub_put_cpu_ptr(var) put_cpu_ptr(var)
+#else
+#define slub_get_cpu_ptr(var) \
+({ \
+ migrate_disable(); \
+ this_cpu_ptr(var); \
+})
+#define slub_put_cpu_ptr(var) \
+do { \
+ (void)(var); \
+ migrate_enable(); \
+} while (0)
+#endif
+
#ifdef CONFIG_SLUB_DEBUG
#ifdef CONFIG_SLUB_DEBUG_ON
DEFINE_STATIC_KEY_TRUE(slub_debug_enabled);
@@ -2809,7 +2829,7 @@ redo:
if (unlikely(!pfmemalloc_match(page, gfpflags)))
goto deactivate_slab;
- /* must check again c->page in case IRQ handler changed it */
+ /* must check again c->page in case we got preempted and it changed */
local_irq_save(flags);
if (unlikely(page != c->page)) {
local_irq_restore(flags);
@@ -2868,7 +2888,8 @@ new_slab:
}
if (unlikely(!slub_percpu_partial(c))) {
local_irq_restore(flags);
- goto new_objects; /* stolen by an IRQ handler */
+ /* we were preempted and partial list got empty */
+ goto new_objects;
}
page = c->page = slub_percpu_partial(c);
@@ -2884,9 +2905,9 @@ new_objects:
if (freelist)
goto check_new_page;
- put_cpu_ptr(s->cpu_slab);
+ slub_put_cpu_ptr(s->cpu_slab);
page = new_slab(s, gfpflags, node);
- c = get_cpu_ptr(s->cpu_slab);
+ c = slub_get_cpu_ptr(s->cpu_slab);
if (unlikely(!page)) {
slab_out_of_memory(s, gfpflags, node);
@@ -2969,12 +2990,12 @@ static void *__slab_alloc(struct kmem_ca
* cpu before disabling preemption. Need to reload cpu area
* pointer.
*/
- c = get_cpu_ptr(s->cpu_slab);
+ c = slub_get_cpu_ptr(s->cpu_slab);
#endif
p = ___slab_alloc(s, gfpflags, node, addr, c);
#ifdef CONFIG_PREEMPT_COUNT
- put_cpu_ptr(s->cpu_slab);
+ slub_put_cpu_ptr(s->cpu_slab);
#endif
return p;
}
@@ -3503,7 +3524,7 @@ int kmem_cache_alloc_bulk(struct kmem_ca
* IRQs, which protects against PREEMPT and interrupts
* handlers invoking normal fastpath.
*/
- c = get_cpu_ptr(s->cpu_slab);
+ c = slub_get_cpu_ptr(s->cpu_slab);
local_irq_disable();
for (i = 0; i < size; i++) {
@@ -3549,7 +3570,7 @@ int kmem_cache_alloc_bulk(struct kmem_ca
}
c->tid = next_tid(c->tid);
local_irq_enable();
- put_cpu_ptr(s->cpu_slab);
+ slub_put_cpu_ptr(s->cpu_slab);
/*
* memcg and kmem_cache debug support and memory initialization.
@@ -3559,7 +3580,7 @@ int kmem_cache_alloc_bulk(struct kmem_ca
slab_want_init_on_alloc(flags, s));
return i;
error:
- put_cpu_ptr(s->cpu_slab);
+ slub_put_cpu_ptr(s->cpu_slab);
slab_post_alloc_hook(s, objcg, flags, i, p, false);
__kmem_cache_free_bulk(s, i, p);
return 0;
_
Patches currently in -mm which might be from vbabka@xxxxxxx are
mm-slub-fix-slub_debug-disablement-for-list-of-slabs.patch
mm-slub-dont-call-flush_all-from-slab_debug_trace_open.patch
mm-slub-allocate-private-object-map-for-debugfs-listings.patch
mm-slub-allocate-private-object-map-for-validate_slab_cache.patch
mm-slub-dont-disable-irq-for-debug_check_no_locks_freed.patch
mm-slub-remove-redundant-unfreeze_partials-from-put_cpu_partial.patch
mm-slub-unify-cmpxchg_double_slab-and-__cmpxchg_double_slab.patch
mm-slub-extract-get_partial-from-new_slab_objects.patch
mm-slub-dissolve-new_slab_objects-into-___slab_alloc.patch
mm-slub-return-slab-page-from-get_partial-and-set-c-page-afterwards.patch
mm-slub-restructure-new-page-checks-in-___slab_alloc.patch
mm-slub-simplify-kmem_cache_cpu-and-tid-setup.patch
mm-slub-move-disabling-enabling-irqs-to-___slab_alloc.patch
mm-slub-do-initial-checks-in-___slab_alloc-with-irqs-enabled.patch
mm-slub-move-disabling-irqs-closer-to-get_partial-in-___slab_alloc.patch
mm-slub-restore-irqs-around-calling-new_slab.patch
mm-slub-validate-slab-from-partial-list-or-page-allocator-before-making-it-cpu-slab.patch
mm-slub-check-new-pages-with-restored-irqs.patch
mm-slub-stop-disabling-irqs-around-get_partial.patch
mm-slub-move-reset-of-c-page-and-freelist-out-of-deactivate_slab.patch
mm-slub-make-locking-in-deactivate_slab-irq-safe.patch
mm-slub-call-deactivate_slab-without-disabling-irqs.patch
mm-slub-move-irq-control-into-unfreeze_partials.patch
mm-slub-discard-slabs-in-unfreeze_partials-without-irqs-disabled.patch
mm-slub-detach-whole-partial-list-at-once-in-unfreeze_partials.patch
mm-slub-separate-detaching-of-partial-list-in-unfreeze_partials-from-unfreezing.patch
mm-slub-only-disable-irq-with-spin_lock-in-__unfreeze_partials.patch
mm-slub-dont-disable-irqs-in-slub_cpu_dead.patch
mm-slab-make-flush_slab-possible-to-call-with-irqs-enabled.patch
mm-slub-optionally-save-restore-irqs-in-slab_lock.patch
mm-slub-make-slab_lock-disable-irqs-with-preempt_rt.patch
mm-slub-protect-put_cpu_partial-with-disabled-irqs-instead-of-cmpxchg.patch
mm-slub-use-migrate_disable-on-preempt_rt.patch
mm-slub-convert-kmem_cpu_slab-protection-to-local_lock.patch
