From: Mel Gorman <mgorman@suse.de> To: Johannes Weiner <hannes@cmpxchg.org> Cc: Andrew Morton <akpm@linux-foundation.org>, Dave Hansen <dave.hansen@intel.com>, Rik van Riel <riel@redhat.com>, Linux-MM <linux-mm@kvack.org>, LKML <linux-kernel@vger.kernel.org>, Mel Gorman <mgorman@suse.de> Subject: [PATCH 1/7] mm: page_alloc: exclude unreclaimable allocations from zone fairness policy Date: Fri, 13 Dec 2013 14:10:01 +0000 [thread overview] Message-ID: <1386943807-29601-2-git-send-email-mgorman@suse.de> (raw) In-Reply-To: <1386943807-29601-1-git-send-email-mgorman@suse.de> From: Johannes Weiner <hannes@cmpxchg.org> Dave Hansen noted a regression in a microbenchmark that loops around open() and close() on an 8-node NUMA machine and bisected it down to 81c0a2bb515f ("mm: page_alloc: fair zone allocator policy"). That change forces the slab allocations of the file descriptor to spread out to all 8 nodes, causing remote references in the page allocator and slab. The round-robin policy is only there to provide fairness among memory allocations that are reclaimed involuntarily based on pressure in each zone. It does not make sense to apply it to unreclaimable kernel allocations that are freed manually, in this case instantly after the allocation, and incur the remote reference costs twice for no reason. Only round-robin allocations that are usually freed through page reclaim or slab shrinking. Cc: <stable@kernel.org> Bisected-by: Dave Hansen <dave.hansen@intel.com> Signed-off-by: Johannes Weiner <hannes@cmpxchg.org> Signed-off-by: Mel Gorman <mgorman@suse.de> --- mm/page_alloc.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/mm/page_alloc.c b/mm/page_alloc.c index 580a5f0..f861d02 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -1920,7 +1920,8 @@ zonelist_scan: * back to remote zones that do not partake in the * fairness round-robin cycle of this zonelist. */ - if (alloc_flags & ALLOC_WMARK_LOW) { + if ((alloc_flags & ALLOC_WMARK_LOW) && + (gfp_mask & GFP_MOVABLE_MASK)) { if (zone_page_state(zone, NR_ALLOC_BATCH) <= 0) continue; if (zone_reclaim_mode && -- 1.8.4
WARNING: multiple messages have this Message-ID (diff)
From: Mel Gorman <mgorman@suse.de> To: Johannes Weiner <hannes@cmpxchg.org> Cc: Andrew Morton <akpm@linux-foundation.org>, Dave Hansen <dave.hansen@intel.com>, Rik van Riel <riel@redhat.com>, Linux-MM <linux-mm@kvack.org>, LKML <linux-kernel@vger.kernel.org>, Mel Gorman <mgorman@suse.de> Subject: [PATCH 1/7] mm: page_alloc: exclude unreclaimable allocations from zone fairness policy Date: Fri, 13 Dec 2013 14:10:01 +0000 [thread overview] Message-ID: <1386943807-29601-2-git-send-email-mgorman@suse.de> (raw) In-Reply-To: <1386943807-29601-1-git-send-email-mgorman@suse.de> From: Johannes Weiner <hannes@cmpxchg.org> Dave Hansen noted a regression in a microbenchmark that loops around open() and close() on an 8-node NUMA machine and bisected it down to 81c0a2bb515f ("mm: page_alloc: fair zone allocator policy"). That change forces the slab allocations of the file descriptor to spread out to all 8 nodes, causing remote references in the page allocator and slab. The round-robin policy is only there to provide fairness among memory allocations that are reclaimed involuntarily based on pressure in each zone. It does not make sense to apply it to unreclaimable kernel allocations that are freed manually, in this case instantly after the allocation, and incur the remote reference costs twice for no reason. Only round-robin allocations that are usually freed through page reclaim or slab shrinking. Cc: <stable@kernel.org> Bisected-by: Dave Hansen <dave.hansen@intel.com> Signed-off-by: Johannes Weiner <hannes@cmpxchg.org> Signed-off-by: Mel Gorman <mgorman@suse.de> --- mm/page_alloc.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/mm/page_alloc.c b/mm/page_alloc.c index 580a5f0..f861d02 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -1920,7 +1920,8 @@ zonelist_scan: * back to remote zones that do not partake in the * fairness round-robin cycle of this zonelist. */ - if (alloc_flags & ALLOC_WMARK_LOW) { + if ((alloc_flags & ALLOC_WMARK_LOW) && + (gfp_mask & GFP_MOVABLE_MASK)) { if (zone_page_state(zone, NR_ALLOC_BATCH) <= 0) continue; if (zone_reclaim_mode && -- 1.8.4 -- 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>
next prev parent reply other threads:[~2013-12-13 14:10 UTC|newest] Thread overview: 84+ messages / expand[flat|nested] mbox.gz Atom feed top 2013-12-13 14:10 [RFC PATCH 0/7] Configurable fair allocation zone policy v2r6 Mel Gorman 2013-12-13 14:10 ` Mel Gorman 2013-12-13 14:10 ` Mel Gorman [this message] 2013-12-13 14:10 ` [PATCH 1/7] mm: page_alloc: exclude unreclaimable allocations from zone fairness policy Mel Gorman 2013-12-13 15:45 ` Rik van Riel 2013-12-13 15:45 ` Rik van Riel 2013-12-13 14:10 ` [PATCH 2/7] mm: page_alloc: Break out zone page aging distribution into its own helper Mel Gorman 2013-12-13 14:10 ` Mel Gorman 2013-12-13 15:46 ` Rik van Riel 2013-12-13 15:46 ` Rik van Riel 2013-12-16 20:16 ` Johannes Weiner 2013-12-16 20:16 ` Johannes Weiner 2013-12-13 14:10 ` [PATCH 3/7] mm: page_alloc: Use zone node IDs to approximate locality Mel Gorman 2013-12-13 14:10 ` Mel Gorman 2013-12-16 13:20 ` Rik van Riel 2013-12-16 13:20 ` Rik van Riel 2013-12-16 20:25 ` Johannes Weiner 2013-12-16 20:25 ` Johannes Weiner 2013-12-17 11:13 ` Mel Gorman 2013-12-17 11:13 ` Mel Gorman 2013-12-17 15:38 ` Johannes Weiner 2013-12-17 15:38 ` Johannes Weiner 2013-12-17 16:08 ` Mel Gorman 2013-12-17 16:08 ` Mel Gorman 2013-12-17 20:11 ` Johannes Weiner 2013-12-17 20:11 ` Johannes Weiner 2013-12-17 21:03 ` Mel Gorman 2013-12-17 21:03 ` Mel Gorman 2013-12-17 22:31 ` Johannes Weiner 2013-12-17 22:31 ` Johannes Weiner 2013-12-13 14:10 ` [PATCH 4/7] mm: Annotate page cache allocations Mel Gorman 2013-12-13 14:10 ` Mel Gorman 2013-12-16 15:20 ` Rik van Riel 2013-12-16 15:20 ` Rik van Riel 2013-12-13 14:10 ` [PATCH 5/7] mm: page_alloc: Make zone distribution page aging policy configurable Mel Gorman 2013-12-13 14:10 ` Mel Gorman 2013-12-16 19:25 ` Rik van Riel 2013-12-16 19:25 ` Rik van Riel 2013-12-16 20:42 ` Johannes Weiner 2013-12-16 20:42 ` Johannes Weiner 2013-12-17 15:29 ` Mel Gorman 2013-12-17 15:29 ` Mel Gorman 2013-12-17 15:54 ` Johannes Weiner 2013-12-17 15:54 ` Johannes Weiner 2013-12-17 16:14 ` Mel Gorman 2013-12-17 16:14 ` Mel Gorman 2013-12-17 17:43 ` Johannes Weiner 2013-12-17 17:43 ` Johannes Weiner 2013-12-17 21:22 ` Mel Gorman 2013-12-17 21:22 ` Mel Gorman 2013-12-17 22:57 ` Johannes Weiner 2013-12-17 22:57 ` Johannes Weiner 2013-12-17 23:24 ` Mel Gorman 2013-12-17 23:24 ` Mel Gorman 2013-12-13 14:10 ` [PATCH 6/7] mm: page_alloc: Only account batch allocations requests that are eligible Mel Gorman 2013-12-13 14:10 ` Mel Gorman 2013-12-16 20:52 ` Johannes Weiner 2013-12-16 20:52 ` Johannes Weiner 2013-12-17 11:20 ` Mel Gorman 2013-12-17 11:20 ` Mel Gorman 2013-12-17 15:43 ` Johannes Weiner 2013-12-17 15:43 ` Johannes Weiner 2013-12-17 16:06 ` Mel Gorman 2013-12-17 16:06 ` Mel Gorman 2013-12-13 14:10 ` [PATCH 7/7] mm: page_alloc: Default allow file pages to use remote nodes for fair allocation policy Mel Gorman 2013-12-13 14:10 ` Mel Gorman 2013-12-13 17:04 ` Johannes Weiner 2013-12-13 17:04 ` Johannes Weiner 2013-12-13 19:20 ` Mel Gorman 2013-12-13 19:20 ` Mel Gorman 2013-12-13 22:15 ` Johannes Weiner 2013-12-13 22:15 ` Johannes Weiner 2013-12-17 16:04 ` Mel Gorman 2013-12-17 16:04 ` Mel Gorman 2013-12-16 19:26 ` Rik van Riel 2013-12-16 19:26 ` Rik van Riel 2013-12-17 15:07 ` [RFC PATCH 0/7] Configurable fair allocation zone policy v2r6 Zlatko Calusic 2013-12-17 15:07 ` Zlatko Calusic 2013-12-17 21:23 ` Mel Gorman 2013-12-17 21:23 ` Mel Gorman 2013-12-21 16:03 ` Zlatko Calusic 2013-12-21 16:03 ` Zlatko Calusic 2013-12-23 10:26 ` Mel Gorman 2013-12-23 10:26 ` Mel Gorman
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=1386943807-29601-2-git-send-email-mgorman@suse.de \ --to=mgorman@suse.de \ --cc=akpm@linux-foundation.org \ --cc=dave.hansen@intel.com \ --cc=hannes@cmpxchg.org \ --cc=linux-kernel@vger.kernel.org \ --cc=linux-mm@kvack.org \ --cc=riel@redhat.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: linkBe sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes, see mirroring instructions on how to clone and mirror all data and code used by this external index.