)]}' { "log": [ { "commit": "e7c3c141240caa84f03ba2100ad552d9e31e8c68", "tree": "3e72deca51bbc2826f6e0161905df5c54e385a78", "parents": [ "f156a7ecbbe30c2f9929cbf1c9cf9b7f89827892" ], "author": { "name": "Mel Gorman", "email": "mgorman@suse.de", "time": "Fri Jan 11 14:32:16 2013 -0800" }, "committer": { "name": "Ethan Chen", "email": "intervigil@gmail.com", "time": "Fri Jan 17 22:12:20 2014 -0800" }, "message": "mm: compaction: partially revert capture of suitable high-order page\n\nEric Wong reported on 3.7 and 3.8-rc2 that ppoll() got stuck when\nwaiting for POLLIN on a local TCP socket. It was easier to trigger if\nthere was disk IO and dirty pages at the same time and he bisected it to\ncommit 1fb3f8ca0e92 (\"mm: compaction: capture a suitable high-order page\nimmediately when it is made available\").\n\nThe intention of that patch was to improve high-order allocations under\nmemory pressure after changes made to reclaim in 3.6 drastically hurt\nTHP allocations but the approach was flawed. For Eric, the problem was\nthat page-\u003epfmemalloc was not being cleared for captured pages leading\nto a poor interaction with swap-over-NFS support causing the packets to\nbe dropped. However, I identified a few more problems with the patch\nincluding the fact that it can increase contention on zone-\u003elock in some\ncases which could result in async direct compaction being aborted early.\n\nIn retrospect the capture patch took the wrong approach. What it should\nhave done is mark the pageblock being migrated as MIGRATE_ISOLATE if it\nwas allocating for THP and avoided races that way. While the patch was\nshowing to improve allocation success rates at the time, the benefit is\nmarginal given the relative complexity and it should be revisited from\nscratch in the context of the other reclaim-related changes that have\ntaken place since the patch was first written and tested. This patch\npartially reverts commit 1fb3f8ca0e92 (\"mm: compaction: capture a\nsuitable high-order page immediately when it is made available\").\n\nChange-Id: I985725a72aac0fdecbf4310c04d176f39e0386dd\nReported-and-tested-by: Eric Wong \u003cnormalperson@yhbt.net\u003e\nTested-by: Eric Dumazet \u003ceric.dumazet@gmail.com\u003e\nCc: \u003cstable@vger.kernel.org\u003e\nSigned-off-by: Mel Gorman \u003cmgorman@suse.de\u003e\nCc: David Miller \u003cdavem@davemloft.net\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\nGit-commit: 8fb74b9fb2b182d54beee592350d9ea1f325917a\nGit-repo: git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git\nSigned-off-by: Laura Abbott \u003clauraa@codeaurora.org\u003e\n" }, { "commit": "b71127a5bec6cb251bde1594cc293ad5d397ee31", "tree": "d9759d210c8df9a2b3fef66a6c5af5b904ddbd16", "parents": [ "f57bc46f5f4a3eae803c41615edd4b366f814c45" ], "author": { "name": "Mel Gorman", "email": "mgorman@suse.de", "time": "Mon Oct 08 16:32:47 2012 -0700" }, "committer": { "name": "Ethan Chen", "email": "intervigil@gmail.com", "time": "Fri Jan 17 22:12:19 2014 -0800" }, "message": "mm: compaction: clear PG_migrate_skip based on compaction and reclaim activity\n\nCompaction caches if a pageblock was scanned and no pages were isolated so\nthat the pageblocks can be skipped in the future to reduce scanning. This\ninformation is not cleared by the page allocator based on activity due to\nthe impact it would have to the page allocator fast paths. Hence there is\na requirement that something clear the cache or pageblocks will be skipped\nforever. Currently the cache is cleared if there were a number of recent\nallocation failures and it has not been cleared within the last 5 seconds.\nTime-based decisions like this are terrible as they have no relationship\nto VM activity and is basically a big hammer.\n\nUnfortunately, accurate heuristics would add cost to some hot paths so\nthis patch implements a rough heuristic. There are two cases where the\ncache is cleared.\n\n1. If a !kswapd process completes a compaction cycle (migrate and free\n scanner meet), the zone is marked compact_blockskip_flush. When kswapd\n goes to sleep, it will clear the cache. This is expected to be the\n common case where the cache is cleared. It does not really matter if\n kswapd happens to be asleep or going to sleep when the flag is set as\n it will be woken on the next allocation request.\n\n2. If there have been multiple failures recently and compaction just\n finished being deferred then a process will clear the cache and start a\n full scan. This situation happens if there are multiple high-order\n allocation requests under heavy memory pressure.\n\nThe clearing of the PG_migrate_skip bits and other scans is inherently\nracy but the race is harmless. For allocations that can fail such as THP,\nthey will simply fail. For requests that cannot fail, they will retry the\nallocation. Tests indicated that scanning rates were roughly similar to\nwhen the time-based heuristic was used and the allocation success rates\nwere similar.\n\nChange-Id: If690ae126badb9f9cc5632e9ffb9d376bf210fb0\nSigned-off-by: Mel Gorman \u003cmgorman@suse.de\u003e\nCc: Rik van Riel \u003criel@redhat.com\u003e\nCc: Richard Davies \u003crichard@arachsys.com\u003e\nCc: Shaohua Li \u003cshli@kernel.org\u003e\nCc: Avi Kivity \u003cavi@redhat.com\u003e\nCc: Rafael Aquini \u003caquini@redhat.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\nGit-commit: 62997027ca5b3d4618198ed8b1aba40b61b1137b\nGit-repo: git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git\nSigned-off-by: Laura Abbott \u003clauraa@codeaurora.org\u003e\n" }, { "commit": "c74068b68d00e36755ce5927bce33d62f1e4edd3", "tree": "a11cd4f023568a2506ab5b26cc0a1014ff88f54b", "parents": [ "bc337a9d96c709c7f4602368eb77401c914ac948" ], "author": { "name": "Mel Gorman", "email": "mgorman@suse.de", "time": "Mon Oct 08 16:29:12 2012 -0700" }, "committer": { "name": "Ethan Chen", "email": "intervigil@gmail.com", "time": "Fri Jan 17 22:12:17 2014 -0800" }, "message": "mm: compaction: capture a suitable high-order page immediately when it is made available\n\nWhile compaction is migrating pages to free up large contiguous blocks\nfor allocation it races with other allocation requests that may steal\nthese blocks or break them up. This patch alters direct compaction to\ncapture a suitable free page as soon as it becomes available to reduce\nthis race. It uses similar logic to split_free_page() to ensure that\nwatermarks are still obeyed.\n\nChange-Id: I46fc38ca67bc50aa7a77a59255caf563f50343a9\nSigned-off-by: Mel Gorman \u003cmgorman@suse.de\u003e\nReviewed-by: Rik van Riel \u003criel@redhat.com\u003e\nReviewed-by: Minchan Kim \u003cminchan@kernel.org\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\nGit-commit: 1fb3f8ca0e9222535a39b884cb67a34628411b9f\nGit-repo: git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git\nSigned-off-by: Laura Abbott \u003clauraa@codeaurora.org\u003e\n" }, { "commit": "0ba338893ca023bbad30546004a776d84a0485d5", "tree": "d6fbd8b2a8ba8a55c8c7c30688463d7b9c6e55a5", "parents": [ "57bb21cbda2ab3f8159059bd8f034ecd48b78a1e" ], "author": { "name": "Mel Gorman", "email": "mgorman@suse.de", "time": "Tue Aug 21 16:16:17 2012 -0700" }, "committer": { "name": "Ethan Chen", "email": "intervigil@gmail.com", "time": "Fri Jan 17 22:12:16 2014 -0800" }, "message": "mm: compaction: Abort async compaction if locks are contended or taking too long\n\nJim Schutt reported a problem that pointed at compaction contending\nheavily on locks. The workload is straight-forward and in his own words;\n\n\tThe systems in question have 24 SAS drives spread across 3 HBAs,\n\trunning 24 Ceph OSD instances, one per drive. FWIW these servers\n\tare dual-socket Intel 5675 Xeons w/48 GB memory. I\u0027ve got ~160\n\tCeph Linux clients doing dd simultaneously to a Ceph file system\n\tbacked by 12 of these servers.\n\nEarly in the test everything looks fine\n\n procs -------------------memory------------------ ---swap-- -----io---- --system-- -----cpu-------\n r b swpd free buff cache si so bi bo in cs us sy id wa st\n 31 15 0 287216 576 38606628 0 0 2 1158 2 14 1 3 95 0 0\n 27 15 0 225288 576 38583384 0 0 18 2222016 203357 134876 11 56 17 15 0\n 28 17 0 219256 576 38544736 0 0 11 2305932 203141 146296 11 49 23 17 0\n 6 18 0 215596 576 38552872 0 0 7 2363207 215264 166502 12 45 22 20 0\n 22 18 0 226984 576 38596404 0 0 3 2445741 223114 179527 12 43 23 22 0\n\nand then it goes to pot\n\n procs -------------------memory------------------ ---swap-- -----io---- --system-- -----cpu-------\n r b swpd free buff cache si so bi bo in cs us sy id wa st\n 163 8 0 464308 576 36791368 0 0 11 22210 866 536 3 13 79 4 0\n 207 14 0 917752 576 36181928 0 0 712 1345376 134598 47367 7 90 1 2 0\n 123 12 0 685516 576 36296148 0 0 429 1386615 158494 60077 8 84 5 3 0\n 123 12 0 598572 576 36333728 0 0 1107 1233281 147542 62351 7 84 5 4 0\n 622 7 0 660768 576 36118264 0 0 557 1345548 151394 59353 7 85 4 3 0\n 223 11 0 283960 576 36463868 0 0 46 1107160 121846 33006 6 93 1 1 0\n\nNote that system CPU usage is very high blocks being written out has\ndropped by 42%. He analysed this with perf and found\n\n perf record -g -a sleep 10\n perf report --sort symbol --call-graph fractal,5\n 34.63% [k] _raw_spin_lock_irqsave\n |\n |--97.30%-- isolate_freepages\n | compaction_alloc\n | unmap_and_move\n | migrate_pages\n | compact_zone\n | compact_zone_order\n | try_to_compact_pages\n | __alloc_pages_direct_compact\n | __alloc_pages_slowpath\n | __alloc_pages_nodemask\n | alloc_pages_vma\n | do_huge_pmd_anonymous_page\n | handle_mm_fault\n | do_page_fault\n | page_fault\n | |\n | |--87.39%-- skb_copy_datagram_iovec\n | | tcp_recvmsg\n | | inet_recvmsg\n | | sock_recvmsg\n | | sys_recvfrom\n | | system_call\n | | __recv\n | | |\n | | --100.00%-- (nil)\n | |\n | --12.61%-- memcpy\n --2.70%-- [...]\n\nThere was other data but primarily it is all showing that compaction is\ncontended heavily on the zone-\u003elock and zone-\u003elru_lock.\n\ncommit [b2eef8c0: mm: compaction: minimise the time IRQs are disabled\nwhile isolating pages for migration] noted that it was possible for\nmigration to hold the lru_lock for an excessive amount of time. Very\nbroadly speaking this patch expands the concept.\n\nThis patch introduces compact_checklock_irqsave() to check if a lock\nis contended or the process needs to be scheduled. If either condition\nis true then async compaction is aborted and the caller is informed.\nThe page allocator will fail a THP allocation if compaction failed due\nto contention. This patch also introduces compact_trylock_irqsave()\nwhich will acquire the lock only if it is not contended and the process\ndoes not need to schedule.\n\nChange-Id: Ia5318c923b903948072ff279dc9aed698bb6d02d\nReported-by: Jim Schutt \u003cjaschut@sandia.gov\u003e\nTested-by: Jim Schutt \u003cjaschut@sandia.gov\u003e\nSigned-off-by: Mel Gorman \u003cmgorman@suse.de\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\nGit-commit: c67fe3752abe6ab47639e2f9b836900c3dc3da84\nGit-repo: git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git\n[lauraa@codeaurora.org: Minor context fixup in isolate_migratepages_range]\nSigned-off-by: Laura Abbott \u003clauraa@codeaurora.org\u003e\n" }, { "commit": "aff622495c9a0b56148192e53bdec539f5e147f2", "tree": "78f6400d8b6bec3279483006a0e9543e47aa833e", "parents": [ "7be62de99adcab4449d416977b4274985c5fe023" ], "author": { "name": "Rik van Riel", "email": "riel@redhat.com", "time": "Wed Mar 21 16:33:52 2012 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Mar 21 17:54:56 2012 -0700" }, "message": "vmscan: only defer compaction for failed order and higher\n\nCurrently a failed order-9 (transparent hugepage) compaction can lead to\nmemory compaction being temporarily disabled for a memory zone. Even if\nwe only need compaction for an order 2 allocation, eg. for jumbo frames\nnetworking.\n\nThe fix is relatively straightforward: keep track of the highest order at\nwhich compaction is succeeding, and only defer compaction for orders at\nwhich compaction is failing.\n\nSigned-off-by: Rik van Riel \u003criel@redhat.com\u003e\nCc: Andrea Arcangeli \u003caarcange@redhat.com\u003e\nAcked-by: Mel Gorman \u003cmel@csn.ul.ie\u003e\nCc: Johannes Weiner \u003channes@cmpxchg.org\u003e\nCc: Minchan Kim \u003cminchan.kim@gmail.com\u003e\nCc: KOSAKI Motohiro \u003ckosaki.motohiro@jp.fujitsu.com\u003e\nCc: Hillf Danton \u003cdhillf@gmail.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "7be62de99adcab4449d416977b4274985c5fe023", "tree": "20ae021ec9811ad730e6a17a3530d3aa6b5027d0", "parents": [ "fe2c2a106663130a5ab45cb0e3414b52df2fff0c" ], "author": { "name": "Rik van Riel", "email": "riel@redhat.com", "time": "Wed Mar 21 16:33:52 2012 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Mar 21 17:54:56 2012 -0700" }, "message": "vmscan: kswapd carefully call compaction\n\nWith CONFIG_COMPACTION enabled, kswapd does not try to free contiguous\nfree pages, even when it is woken for a higher order request.\n\nThis could be bad for eg. jumbo frame network allocations, which are done\nfrom interrupt context and cannot compact memory themselves. Higher than\nbefore allocation failure rates in the network receive path have been\nobserved in kernels with compaction enabled.\n\nTeach kswapd to defragment the memory zones in a node, but only if\nrequired and compaction is not deferred in a zone.\n\n[akpm@linux-foundation.org: reduce scope of zones_need_compaction]\nSigned-off-by: Rik van Riel \u003criel@redhat.com\u003e\nAcked-by: Mel Gorman \u003cmel@csn.ul.ie\u003e\nCc: Andrea Arcangeli \u003caarcange@redhat.com\u003e\nCc: Johannes Weiner \u003channes@cmpxchg.org\u003e\nCc: Minchan Kim \u003cminchan.kim@gmail.com\u003e\nCc: KOSAKI Motohiro \u003ckosaki.motohiro@jp.fujitsu.com\u003e\nCc: Hillf Danton \u003cdhillf@gmail.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "d43a87e68e9e71d2987a29cc239acec4e8f410c9", "tree": "75b2950b286dcad4c4e7785f991ab0a35146b8d1", "parents": [ "dd73e85f6d8f721d66bcbd2734a5f4bc3d3cd768" ], "author": { "name": "Kyungmin Park", "email": "kyungmin.park@samsung.com", "time": "Mon Oct 31 17:09:08 2011 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Mon Oct 31 17:30:49 2011 -0700" }, "message": "mm: compaction: make compact_zone_order() static\n\nThere\u0027s no compact_zone_order() user outside file scope, so make it static.\n\nSigned-off-by: Kyungmin Park \u003ckyungmin.park@samsung.com\u003e\nAcked-by: David Rientjes \u003crientjes@google.com\u003e\nReviewed-by: Minchan Kim \u003cminchan.kim@gmail.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "d527caf22e48480b102c7c6ee5b9ba12170148f7", "tree": "7d53a2c430f8c020b6fa8390396dd2d1ce480b9a", "parents": [ "89699605fe7cfd8611900346f61cb6cbf179b10a" ], "author": { "name": "Andrea Arcangeli", "email": "aarcange@redhat.com", "time": "Tue Mar 22 16:30:38 2011 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue Mar 22 17:44:00 2011 -0700" }, "message": "mm: compaction: prevent kswapd compacting memory to reduce CPU usage\n\nThis patch reverts 5a03b051 (\"thp: use compaction in kswapd for GFP_ATOMIC\norder \u003e 0\") due to reports stating that kswapd CPU usage was higher and\nIRQs were being disabled more frequently. This was reported at\nhttp://www.spinics.net/linux/fedora/alsa-user/msg09885.html.\n\nWithout this patch applied, CPU usage by kswapd hovers around the 20% mark\naccording to the tester (Arthur Marsh:\nhttp://www.spinics.net/linux/fedora/alsa-user/msg09899.html). With this\npatch applied, it\u0027s around 2%.\n\nThe problem is not related to THP which specifies __GFP_NO_KSWAPD but is\ntriggered by high-order allocations hitting the low watermark for their\norder and waking kswapd on kernels with CONFIG_COMPACTION set. The most\ncommon trigger for this is network cards configured for jumbo frames but\nit\u0027s also possible it\u0027ll be triggered by fork-heavy workloads (order-1)\nand some wireless cards which depend on order-1 allocations.\n\nThe symptoms for the user will be high CPU usage by kswapd in low-memory\nsituations which could be confused with another writeback problem. While\na patch like 5a03b051 may be reintroduced in the future, this patch plays\nit safe for now and reverts it.\n\n[mel@csn.ul.ie: Beefed up the changelog]\nSigned-off-by: Andrea Arcangeli \u003caarcange@redhat.com\u003e\nSigned-off-by: Mel Gorman \u003cmel@csn.ul.ie\u003e\nReported-by: Arthur Marsh \u003carthur.marsh@internode.on.net\u003e\nTested-by: Arthur Marsh \u003carthur.marsh@internode.on.net\u003e\nCc: \u003cstable@kernel.org\u003e\t\t[2.6.38.1]\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "5a03b051ed87e72b959f32a86054e1142ac4cf55", "tree": "31f0e8efb86d48b0292f8a7ea4bd9cf7c930a0ab", "parents": [ "878aee7d6b5504e01b9caffce080e792b6b8d090" ], "author": { "name": "Andrea Arcangeli", "email": "aarcange@redhat.com", "time": "Thu Jan 13 15:47:11 2011 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Jan 13 17:32:46 2011 -0800" }, "message": "thp: use compaction in kswapd for GFP_ATOMIC order \u003e 0\n\nThis takes advantage of memory compaction to properly generate pages of\norder \u003e 0 if regular page reclaim fails and priority level becomes more\nsevere and we don\u0027t reach the proper watermarks.\n\nSigned-off-by: Andrea Arcangeli \u003caarcange@redhat.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "77f1fe6b08b13a87391549c8a820ddc817b6f50e", "tree": "720865bd0994da3787b6f37d33b2ee4c26a2de6c", "parents": [ "3e7d344970673c5334cf7b5bb27c8c0942b06126" ], "author": { "name": "Mel Gorman", "email": "mel@csn.ul.ie", "time": "Thu Jan 13 15:45:57 2011 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Jan 13 17:32:34 2011 -0800" }, "message": "mm: migration: allow migration to operate asynchronously and avoid synchronous compaction in the faster path\n\nMigration synchronously waits for writeback if the initial passes fails.\nCallers of memory compaction do not necessarily want this behaviour if the\ncaller is latency sensitive or expects that synchronous migration is not\ngoing to have a significantly better success rate.\n\nThis patch adds a sync parameter to migrate_pages() allowing the caller to\nindicate if wait_on_page_writeback() is allowed within migration or not.\nFor reclaim/compaction, try_to_compact_pages() is first called\nasynchronously, direct reclaim runs and then try_to_compact_pages() is\ncalled synchronously as there is a greater expectation that it\u0027ll succeed.\n\n[akpm@linux-foundation.org: build/merge fix]\nSigned-off-by: Mel Gorman \u003cmel@csn.ul.ie\u003e\nCc: Andrea Arcangeli \u003caarcange@redhat.com\u003e\nCc: KOSAKI Motohiro \u003ckosaki.motohiro@jp.fujitsu.com\u003e\nCc: Rik van Riel \u003criel@redhat.com\u003e\nAcked-by: Johannes Weiner \u003channes@cmpxchg.org\u003e\nCc: Andy Whitcroft \u003capw@shadowen.org\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "3e7d344970673c5334cf7b5bb27c8c0942b06126", "tree": "832ecb4da5fd27efa5a503df5b96bfdee2a52ffd", "parents": [ "ee64fc9354e515a79c7232cfde65c88ec627308b" ], "author": { "name": "Mel Gorman", "email": "mel@csn.ul.ie", "time": "Thu Jan 13 15:45:56 2011 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Jan 13 17:32:33 2011 -0800" }, "message": "mm: vmscan: reclaim order-0 and use compaction instead of lumpy reclaim\n\nLumpy reclaim is disruptive. It reclaims a large number of pages and\nignores the age of the pages it reclaims. This can incur significant\nstalls and potentially increase the number of major faults.\n\nCompaction has reached the point where it is considered reasonably stable\n(meaning it has passed a lot of testing) and is a potential candidate for\ndisplacing lumpy reclaim. This patch introduces an alternative to lumpy\nreclaim whe compaction is available called reclaim/compaction. The basic\noperation is very simple - instead of selecting a contiguous range of\npages to reclaim, a number of order-0 pages are reclaimed and then\ncompaction is later by either kswapd (compact_zone_order()) or direct\ncompaction (__alloc_pages_direct_compact()).\n\n[akpm@linux-foundation.org: fix build]\n[akpm@linux-foundation.org: use conventional task_struct naming]\nSigned-off-by: Mel Gorman \u003cmel@csn.ul.ie\u003e\nCc: Andrea Arcangeli \u003caarcange@redhat.com\u003e\nCc: KOSAKI Motohiro \u003ckosaki.motohiro@jp.fujitsu.com\u003e\nCc: Rik van Riel \u003criel@redhat.com\u003e\nAcked-by: Johannes Weiner \u003channes@cmpxchg.org\u003e\nCc: Andy Whitcroft \u003capw@shadowen.org\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "4f92e2586b43a2402e116055d4edda704f911b5b", "tree": "6a765ebeba951c02a7878bcea52a4769ad2e45c2", "parents": [ "5e7719058079a1423ccce56148b0aaa56b2df821" ], "author": { "name": "Mel Gorman", "email": "mel@csn.ul.ie", "time": "Mon May 24 14:32:32 2010 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue May 25 08:07:00 2010 -0700" }, "message": "mm: compaction: defer compaction using an exponential backoff when compaction fails\n\nThe fragmentation index may indicate that a failure is due to external\nfragmentation but after a compaction run completes, it is still possible\nfor an allocation to fail. There are two obvious reasons as to why\n\n o Page migration cannot move all pages so fragmentation remains\n o A suitable page may exist but watermarks are not met\n\nIn the event of compaction followed by an allocation failure, this patch\ndefers further compaction in the zone (1 \u003c\u003c compact_defer_shift) times.\nIf the next compaction attempt also fails, compact_defer_shift is\nincreased up to a maximum of 6. If compaction succeeds, the defer\ncounters are reset again.\n\nThe zone that is deferred is the first zone in the zonelist - i.e. the\npreferred zone. To defer compaction in the other zones, the information\nwould need to be stored in the zonelist or implemented similar to the\nzonelist_cache. This would impact the fast-paths and is not justified at\nthis time.\n\nSigned-off-by: Mel Gorman \u003cmel@csn.ul.ie\u003e\nCc: Rik van Riel \u003criel@redhat.com\u003e\nCc: Minchan Kim \u003cminchan.kim@gmail.com\u003e\nCc: KOSAKI Motohiro \u003ckosaki.motohiro@jp.fujitsu.com\u003e\nCc: Christoph Lameter \u003ccl@linux-foundation.org\u003e\nCc: KAMEZAWA Hiroyuki \u003ckamezawa.hiroyu@jp.fujitsu.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "5e7719058079a1423ccce56148b0aaa56b2df821", "tree": "3666cb3d5540dcaa3d8e7df8c293a0ad603a181c", "parents": [ "56de7263fcf3eb10c8dcdf8d59a9cec831795f3f" ], "author": { "name": "Mel Gorman", "email": "mel@csn.ul.ie", "time": "Mon May 24 14:32:31 2010 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue May 25 08:06:59 2010 -0700" }, "message": "mm: compaction: add a tunable that decides when memory should be compacted and when it should be reclaimed\n\nThe kernel applies some heuristics when deciding if memory should be\ncompacted or reclaimed to satisfy a high-order allocation. One of these\nis based on the fragmentation. If the index is below 500, memory will not\nbe compacted. This choice is arbitrary and not based on data. To help\noptimise the system and set a sensible default for this value, this patch\nadds a sysctl extfrag_threshold. The kernel will only compact memory if\nthe fragmentation index is above the extfrag_threshold.\n\n[randy.dunlap@oracle.com: Fix build errors when proc fs is not configured]\nSigned-off-by: Mel Gorman \u003cmel@csn.ul.ie\u003e\nSigned-off-by: Randy Dunlap \u003crandy.dunlap@oracle.com\u003e\nCc: Rik van Riel \u003criel@redhat.com\u003e\nCc: Minchan Kim \u003cminchan.kim@gmail.com\u003e\nCc: KOSAKI Motohiro \u003ckosaki.motohiro@jp.fujitsu.com\u003e\nCc: Christoph Lameter \u003ccl@linux-foundation.org\u003e\nCc: KAMEZAWA Hiroyuki \u003ckamezawa.hiroyu@jp.fujitsu.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "56de7263fcf3eb10c8dcdf8d59a9cec831795f3f", "tree": "164637c0b678e20adfdcec4129563d9234faf405", "parents": [ "ed4a6d7f0676db50b5023cc01f6cda82a2f2a307" ], "author": { "name": "Mel Gorman", "email": "mel@csn.ul.ie", "time": "Mon May 24 14:32:30 2010 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue May 25 08:06:59 2010 -0700" }, "message": "mm: compaction: direct compact when a high-order allocation fails\n\nOrdinarily when a high-order allocation fails, direct reclaim is entered\nto free pages to satisfy the allocation. With this patch, it is\ndetermined if an allocation failed due to external fragmentation instead\nof low memory and if so, the calling process will compact until a suitable\npage is freed. Compaction by moving pages in memory is considerably\ncheaper than paging out to disk and works where there are locked pages or\nno swap. If compaction fails to free a page of a suitable size, then\nreclaim will still occur.\n\nDirect compaction returns as soon as possible. As each block is\ncompacted, it is checked if a suitable page has been freed and if so, it\nreturns.\n\n[akpm@linux-foundation.org: Fix build errors]\n[aarcange@redhat.com: fix count_vm_event preempt in memory compaction direct reclaim]\nSigned-off-by: Mel Gorman \u003cmel@csn.ul.ie\u003e\nAcked-by: Rik van Riel \u003criel@redhat.com\u003e\nReviewed-by: Minchan Kim \u003cminchan.kim@gmail.com\u003e\nCc: KOSAKI Motohiro \u003ckosaki.motohiro@jp.fujitsu.com\u003e\nCc: Christoph Lameter \u003ccl@linux-foundation.org\u003e\nCc: KAMEZAWA Hiroyuki \u003ckamezawa.hiroyu@jp.fujitsu.com\u003e\nSigned-off-by: Andrea Arcangeli \u003caarcange@redhat.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "ed4a6d7f0676db50b5023cc01f6cda82a2f2a307", "tree": "6496343ba964e105cf30f8703714471f22edc806", "parents": [ "76ab0f530e4a01d4dc20cdc1d5e87753c579dc18" ], "author": { "name": "Mel Gorman", "email": "mel@csn.ul.ie", "time": "Mon May 24 14:32:29 2010 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue May 25 08:06:59 2010 -0700" }, "message": "mm: compaction: add /sys trigger for per-node memory compaction\n\nAdd a per-node sysfs file called compact. When the file is written to,\neach zone in that node is compacted. The intention that this would be\nused by something like a job scheduler in a batch system before a job\nstarts so that the job can allocate the maximum number of hugepages\nwithout significant start-up cost.\n\nSigned-off-by: Mel Gorman \u003cmel@csn.ul.ie\u003e\nAcked-by: Rik van Riel \u003criel@redhat.com\u003e\nReviewed-by: KOSAKI Motohiro \u003ckosaki.motohiro@jp.fujitsu.com\u003e\nReviewed-by: Christoph Lameter \u003ccl@linux-foundation.org\u003e\nReviewed-by: Minchan Kim \u003cminchan.kim@gmail.com\u003e\nReviewed-by: KAMEZAWA Hiroyuki \u003ckamezawa.hiroyu@jp.fujitsu.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "76ab0f530e4a01d4dc20cdc1d5e87753c579dc18", "tree": "8e1566df85e02f67876685c36d217fec4845f79f", "parents": [ "748446bb6b5a9390b546af38ec899c868a9dbcf0" ], "author": { "name": "Mel Gorman", "email": "mel@csn.ul.ie", "time": "Mon May 24 14:32:28 2010 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue May 25 08:06:59 2010 -0700" }, "message": "mm: compaction: add /proc trigger for memory compaction\n\nAdd a proc file /proc/sys/vm/compact_memory. When an arbitrary value is\nwritten to the file, all zones are compacted. The expected user of such a\ntrigger is a job scheduler that prepares the system before the target\napplication runs.\n\nSigned-off-by: Mel Gorman \u003cmel@csn.ul.ie\u003e\nAcked-by: Rik van Riel \u003criel@redhat.com\u003e\nReviewed-by: KAMEZAWA Hiroyuki \u003ckamezawa.hiroyu@jp.fujitsu.com\u003e\nReviewed-by: Minchan Kim \u003cminchan.kim@gmail.com\u003e\nReviewed-by: KOSAKI Motohiro \u003ckosaki.motohiro@jp.fujitsu.com\u003e\nReviewed-by: Christoph Lameter \u003ccl@linux-foundation.org\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "748446bb6b5a9390b546af38ec899c868a9dbcf0", "tree": "4c27d0805a5e094b39ff938ad60dd270b953a79f", "parents": [ "c175a0ce7584e5b498fff8cbdb9aa7912aa9fbba" ], "author": { "name": "Mel Gorman", "email": "mel@csn.ul.ie", "time": "Mon May 24 14:32:27 2010 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue May 25 08:06:59 2010 -0700" }, "message": "mm: compaction: memory compaction core\n\nThis patch is the core of a mechanism which compacts memory in a zone by\nrelocating movable pages towards the end of the zone.\n\nA single compaction run involves a migration scanner and a free scanner.\nBoth scanners operate on pageblock-sized areas in the zone. The migration\nscanner starts at the bottom of the zone and searches for all movable\npages within each area, isolating them onto a private list called\nmigratelist. The free scanner starts at the top of the zone and searches\nfor suitable areas and consumes the free pages within making them\navailable for the migration scanner. The pages isolated for migration are\nthen migrated to the newly isolated free pages.\n\n[aarcange@redhat.com: Fix unsafe optimisation]\n[mel@csn.ul.ie: do not schedule work on other CPUs for compaction]\nSigned-off-by: Mel Gorman \u003cmel@csn.ul.ie\u003e\nAcked-by: Rik van Riel \u003criel@redhat.com\u003e\nReviewed-by: Minchan Kim \u003cminchan.kim@gmail.com\u003e\nCc: KOSAKI Motohiro \u003ckosaki.motohiro@jp.fujitsu.com\u003e\nCc: Christoph Lameter \u003ccl@linux-foundation.org\u003e\nCc: KAMEZAWA Hiroyuki \u003ckamezawa.hiroyu@jp.fujitsu.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" } ] }