linux-mm.kvack.org archive mirror
 help / color / mirror / Atom feed
From: Jesper Dangaard Brouer <brouer@redhat.com>
To: linux-mm@kvack.org
Cc: Christoph Lameter <cl@linux.com>,
	Vladimir Davydov <vdavydov@virtuozzo.com>,
	Jesper Dangaard Brouer <brouer@redhat.com>,
	Joonsoo Kim <iamjoonsoo.kim@lge.com>,
	Linus Torvalds <torvalds@linux-foundation.org>,
	Andrew Morton <akpm@linux-foundation.org>
Subject: [RFC PATCH V2 2/9] mm: generalize avoid fault-inject on bootstrap kmem_cache
Date: Tue, 08 Dec 2015 17:18:32 +0100	[thread overview]
Message-ID: <20151208161832.21945.55076.stgit@firesoul> (raw)
In-Reply-To: <20151208161751.21945.53936.stgit@firesoul>

Move slab_should_failslab() check from SLAB allocator to generic
slab_pre_alloc_hook().  The check guards against slab alloc
fault-injects failures for the bootstrap slab that is used for
allocating "kmem_cache" objects to the allocator itself.

I'm not really happy with this code...
---
 mm/failslab.c |    2 ++
 mm/slab.c     |    8 --------
 mm/slab.h     |   23 ++++++++++++++++++++++-
 3 files changed, 24 insertions(+), 9 deletions(-)

diff --git a/mm/failslab.c b/mm/failslab.c
index 79171b4a5826..a2ad28ba696c 100644
--- a/mm/failslab.c
+++ b/mm/failslab.c
@@ -13,6 +13,8 @@ static struct {
 
 bool should_failslab(size_t size, gfp_t gfpflags, unsigned long cache_flags)
 {
+	// Should we place bootstrap kmem_cache check here???
+
 	if (gfpflags & __GFP_NOFAIL)
 		return false;
 
diff --git a/mm/slab.c b/mm/slab.c
index 4765c97ce690..4684c2496982 100644
--- a/mm/slab.c
+++ b/mm/slab.c
@@ -2917,14 +2917,6 @@ static void *cache_alloc_debugcheck_after(struct kmem_cache *cachep,
 #define cache_alloc_debugcheck_after(a,b,objp,d) (objp)
 #endif
 
-static bool slab_should_failslab(struct kmem_cache *cachep, gfp_t flags)
-{
-	if (unlikely(cachep == kmem_cache))
-		return false;
-
-	return should_failslab(cachep->object_size, flags, cachep->flags);
-}
-
 static inline void *____cache_alloc(struct kmem_cache *cachep, gfp_t flags)
 {
 	void *objp;
diff --git a/mm/slab.h b/mm/slab.h
index 588bc5281fc8..4e7b0e62f3f4 100644
--- a/mm/slab.h
+++ b/mm/slab.h
@@ -360,6 +360,27 @@ static inline size_t slab_ksize(const struct kmem_cache *s)
 }
 #endif
 
+/* FIXME: This construct sucks, because this compare+branch needs to
+ * get removed by compiler then !CONFIG_FAILSLAB (maybe compiler is
+ * smart enough to realize only "false" can be generated).
+ *
+ * Comments please: Pulling out CONFIG_FAILSLAB here looks ugly...
+ *  should we instead change API of should_failslab() ??
+ *
+ * Next question: is the bootstrap cache check okay to add for all
+ * allocators? (this would be the easiest, else need more ugly ifdef's)
+ */
+static inline bool slab_should_failslab(struct kmem_cache *cachep, gfp_t flags)
+{
+	/* No fault-injection for bootstrap cache */
+#ifdef CONFIG_FAILSLAB
+	if (unlikely(cachep == kmem_cache))
+		return false;
+#endif
+
+	return should_failslab(cachep->object_size, flags, cachep->flags);
+}
+
 static inline struct kmem_cache *slab_pre_alloc_hook(struct kmem_cache *s,
 						     gfp_t flags)
 {
@@ -367,7 +388,7 @@ static inline struct kmem_cache *slab_pre_alloc_hook(struct kmem_cache *s,
 	lockdep_trace_alloc(flags);
 	might_sleep_if(gfpflags_allow_blocking(flags));
 
-	if (should_failslab(s->object_size, flags, s->flags))
+	if (slab_should_failslab(s, flags))
 		return NULL;
 
 	return memcg_kmem_get_cache(s, flags);

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

  parent reply	other threads:[~2015-12-08 16:19 UTC|newest]

Thread overview: 41+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-12-03 15:56 [RFC PATCH 0/2] slab: implement bulking for SLAB allocator Jesper Dangaard Brouer
2015-12-03 15:57 ` [RFC PATCH 1/2] slab: implement bulk alloc in " Jesper Dangaard Brouer
2015-12-04 10:16   ` Jesper Dangaard Brouer
2015-12-04 17:10   ` Christoph Lameter
2015-12-07 10:20     ` Jesper Dangaard Brouer
2015-12-07 14:57       ` Christoph Lameter
2015-12-03 15:57 ` [RFC PATCH 2/2] slab: implement bulk free " Jesper Dangaard Brouer
2015-12-04 17:17   ` Christoph Lameter
2015-12-07 11:25     ` Jesper Dangaard Brouer
2015-12-07 14:59       ` Christoph Lameter
2015-12-08 13:39         ` Jesper Dangaard Brouer
2015-12-08 14:11           ` Christoph Lameter
2015-12-08 14:12         ` Vladimir Davydov
2015-12-08 14:15           ` Christoph Lameter
2015-12-08 14:56             ` Vladimir Davydov
2015-12-08 15:13               ` Jesper Dangaard Brouer
2015-12-08 15:32               ` Christoph Lameter
2015-12-04  9:01 ` [RFC PATCH 0/2] slab: implement bulking for " Jesper Dangaard Brouer
2015-12-08 16:18 ` [RFC PATCH V2 0/9] slab: cleanup and bulk API for SLAB Jesper Dangaard Brouer
2015-12-08 16:18   ` [RFC PATCH V2 1/9] mm/slab: move SLUB alloc hooks to common mm/slab.h Jesper Dangaard Brouer
2015-12-09 15:43     ` Christoph Lameter
2015-12-08 16:18   ` Jesper Dangaard Brouer [this message]
2015-12-09  2:36     ` [RFC PATCH V2 2/9] mm: generalize avoid fault-inject on bootstrap kmem_cache Joonsoo Kim
2015-12-08 16:18   ` [RFC PATCH V2 3/9] slab: use slab_pre_alloc_hook in SLAB allocator Jesper Dangaard Brouer
2015-12-08 16:18   ` [RFC PATCH V2 4/9] mm: kmemcheck skip object if slab allocation failed Jesper Dangaard Brouer
2015-12-08 16:18   ` [RFC PATCH V2 5/9] slab: use slab_post_alloc_hook in SLAB allocator Jesper Dangaard Brouer
2015-12-08 16:18   ` [RFC PATCH V2 6/9] slab: implement bulk alloc " Jesper Dangaard Brouer
2015-12-08 16:18   ` [RFC PATCH V2 7/9] slab: avoid running debug SLAB code with IRQs disabled for alloc_bulk Jesper Dangaard Brouer
2015-12-08 16:19   ` [RFC PATCH V2 8/9] slab: implement bulk free in SLAB allocator Jesper Dangaard Brouer
2015-12-09 16:06     ` Christoph Lameter
2015-12-09 18:53       ` Jesper Dangaard Brouer
2015-12-09 19:41         ` Christoph Lameter
2015-12-09 20:50           ` Jesper Dangaard Brouer
2015-12-10 15:15             ` Christoph Lameter
2015-12-10 15:10           ` Jesper Dangaard Brouer
2015-12-10 15:18             ` Christoph Lameter
2015-12-10 15:26               ` Vladimir Davydov
2015-12-10 17:24                 ` Christoph Lameter
2015-12-14 15:19       ` Jesper Dangaard Brouer
2015-12-15 12:02         ` Jesper Dangaard Brouer
2015-12-08 16:19   ` [RFC PATCH V2 9/9] slab: annotate code to generate more compact asm code Jesper Dangaard Brouer

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20151208161832.21945.55076.stgit@firesoul \
    --to=brouer@redhat.com \
    --cc=akpm@linux-foundation.org \
    --cc=cl@linux.com \
    --cc=iamjoonsoo.kim@lge.com \
    --cc=linux-mm@kvack.org \
    --cc=torvalds@linux-foundation.org \
    --cc=vdavydov@virtuozzo.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox