aboutsummaryrefslogtreecommitdiffstats
path: root/kernel
diff options
context:
space:
mode:
authorPaul Jackson <pj@sgi.com>2006-05-20 18:00:10 -0400
committerLinus Torvalds <torvalds@g5.osdl.org>2006-05-21 15:59:18 -0400
commit36be57ffe39e03aab9fbe857f70c7a6a15bd9e08 (patch)
tree8062e9a9e49fd942fab35e17a2b5afe571a74f63 /kernel
parentbdd804f478a0cc74bf7db8e9f9d5fd379d1b31ca (diff)
[PATCH] cpuset: update cpuset_zones_allowed comment
Update the kernel/cpuset.c:cpuset_zone_allowed() comment. The rule for when mm/page_alloc.c should call cpuset_zone_allowed() was intended to be: Don't call cpuset_zone_allowed() if you can't sleep, unless you pass in the __GFP_HARDWALL flag set in gfp_flag, which disables the code that might scan up ancestor cpusets and sleep. The explanation of this rule in the comment above cpuset_zone_allowed() was stale, as a result of a restructuring of some __alloc_pages() code in November 2005. Rewrite that comment ... Signed-off-by: Paul Jackson <pj@sgi.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Diffstat (limited to 'kernel')
-rw-r--r--kernel/cpuset.c24
1 files changed, 15 insertions, 9 deletions
diff --git a/kernel/cpuset.c b/kernel/cpuset.c
index 72248d1b9e3f..57fd88237afc 100644
--- a/kernel/cpuset.c
+++ b/kernel/cpuset.c
@@ -2231,19 +2231,25 @@ static const struct cpuset *nearest_exclusive_ancestor(const struct cpuset *cs)
2231 * So only GFP_KERNEL allocations, if all nodes in the cpuset are 2231 * So only GFP_KERNEL allocations, if all nodes in the cpuset are
2232 * short of memory, might require taking the callback_mutex mutex. 2232 * short of memory, might require taking the callback_mutex mutex.
2233 * 2233 *
2234 * The first loop over the zonelist in mm/page_alloc.c:__alloc_pages() 2234 * The first call here from mm/page_alloc:get_page_from_freelist()
2235 * calls here with __GFP_HARDWALL always set in gfp_mask, enforcing 2235 * has __GFP_HARDWALL set in gfp_mask, enforcing hardwall cpusets, so
2236 * hardwall cpusets - no allocation on a node outside the cpuset is 2236 * no allocation on a node outside the cpuset is allowed (unless in
2237 * allowed (unless in interrupt, of course). 2237 * interrupt, of course).
2238 * 2238 *
2239 * The second loop doesn't even call here for GFP_ATOMIC requests 2239 * The second pass through get_page_from_freelist() doesn't even call
2240 * (if the __alloc_pages() local variable 'wait' is set). That check 2240 * here for GFP_ATOMIC calls. For those calls, the __alloc_pages()
2241 * and the checks below have the combined affect in the second loop of 2241 * variable 'wait' is not set, and the bit ALLOC_CPUSET is not set
2242 * the __alloc_pages() routine that: 2242 * in alloc_flags. That logic and the checks below have the combined
2243 * affect that:
2243 * in_interrupt - any node ok (current task context irrelevant) 2244 * in_interrupt - any node ok (current task context irrelevant)
2244 * GFP_ATOMIC - any node ok 2245 * GFP_ATOMIC - any node ok
2245 * GFP_KERNEL - any node in enclosing mem_exclusive cpuset ok 2246 * GFP_KERNEL - any node in enclosing mem_exclusive cpuset ok
2246 * GFP_USER - only nodes in current tasks mems allowed ok. 2247 * GFP_USER - only nodes in current tasks mems allowed ok.
2248 *
2249 * Rule:
2250 * Don't call cpuset_zone_allowed() if you can't sleep, unless you
2251 * pass in the __GFP_HARDWALL flag set in gfp_flag, which disables
2252 * the code that might scan up ancestor cpusets and sleep.
2247 **/ 2253 **/
2248 2254
2249int __cpuset_zone_allowed(struct zone *z, gfp_t gfp_mask) 2255int __cpuset_zone_allowed(struct zone *z, gfp_t gfp_mask)