projects/linux·commit b2455396

[PATCH] cpuset: memory_spread_slab drop useless PF_SPREAD_PAGE check

b2455396 · Paul Jackson <[email protected]> · 2006-03-24 03:16:12 -08:00

The hook in the slab cache allocation path to handle cpuset memory
spreading for tasks in cpusets with 'memory_spread_slab' enabled has a
modest performance bug.  The hook calls into the memory spreading handler
alternate_node_alloc() if either of 'memory_spread_slab' or
'memory_spread_page' is enabled, even though the handler does nothing
(albeit harmlessly) for the page case

Fix - drop PF_SPREAD_PAGE from the set of flag bits that are used to
trigger a call to alternate_node_alloc().

The page case is handled by separate hooks -- see the calls conditioned on
cpuset_do_page_mem_spread() in mm/filemap.c

Signed-off-by: Paul Jackson <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Commit
b2455396be35383c4eebc6745cc718b1dd9e23df
Parents
151a4420

Changes

1 files changed · +2 −3

FileStatusAddDelContent
mm/slab.cmodified+2−3View file

mm/slab.c

View file
diff --git a/mm/slab.c b/mm/slab.c
index f80b523..26138c9 100644
--- a/mm/slab.c
+++ b/mm/slab.c
@@ -2809,8 +2809,7 @@ static inline void *____cache_alloc(struct kmem_cache *cachep, gfp_t flags)
struct array_cache *ac;
#ifdef CONFIG_NUMA
- if (unlikely(current->flags & (PF_SPREAD_PAGE | PF_SPREAD_SLAB |
- PF_MEMPOLICY))) {
+ if (unlikely(current->flags & (PF_SPREAD_SLAB | PF_MEMPOLICY))) {
objp = alternate_node_alloc(cachep, flags);
if (objp != NULL)
return objp;
@@ -2849,7 +2848,7 @@ static __always_inline void *__cache_alloc(struct kmem_cache *cachep,
#ifdef CONFIG_NUMA
/*
- * Try allocating on another node if PF_SPREAD_PAGE|PF_SPREAD_SLAB|PF_MEMPOLICY.
+ * Try allocating on another node if PF_SPREAD_SLAB|PF_MEMPOLICY.
*
* If we are in_interrupt, then process context, including cpusets and
* mempolicy, may not apply and should not be used for allocation policy.