- 18 Apr, 2023 24 commits
-
-
Mike Rapoport (IBM) authored
It is enough to keep default values for base and huge pages without letting users to override ARCH_FORCE_MAX_ORDER. Drop the prompt to make the option unvisible in *config. Link: https://lkml.kernel.org/r/20230324052233.2654090-6-rppt@kernel.orgSigned-off-by:
Mike Rapoport (IBM) <rppt@kernel.org> Acked-by:
Kirill A. Shutemov <kirill.shutemov@linux.intel.com> Reviewed-by:
Zi Yan <ziy@nvidia.com> Cc: Arnd Bergmann <arnd@arndb.de> Cc: Catalin Marinas <catalin.marinas@arm.com> Cc: Christophe Leroy <christophe.leroy@csgroup.eu> Cc: David Miller <davem@davemloft.net> Cc: Dinh Nguyen <dinguyen@kernel.org> Cc: Geert Uytterhoeven <geert@linux-m68k.org> Cc: Guo Ren <guoren@kernel.org> Cc: John Paul Adrian Glaubitz <glaubitz@physik.fu-berlin.de> Cc: Max Filippov <jcmvbkbc@gmail.com> Cc: Michael Ellerman <mpe@ellerman.id.au> Cc: Rich Felker <dalias@libc.org> Cc: "Russell King (Oracle)" <linux@armlinux.org.uk> Cc: Will Deacon <will@kernel.org> Cc: Yoshinori Sato <ysato@users.sourceforge.jp> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org>
-
Mike Rapoport (IBM) authored
The default value of ARCH_FORCE_MAX_ORDER matches the generic default defined in the MM code, the architecture does not support huge pages, so there is no need to keep ARCH_FORCE_MAX_ORDER option available. Drop it. Link: https://lkml.kernel.org/r/20230324052233.2654090-5-rppt@kernel.orgSigned-off-by:
-
Mike Rapoport (IBM) authored
The prompt and help text of ARCH_FORCE_MAX_ORDER are not even close to describe this configuration option. Update both to actually describe what this option does. [rppt@kernel.org: change ARCH_FORCE_MAX_ORDER dependencies] Link: https://lkml.kernel.org/r/20230325060828.2662773-4-rppt@kernel.org Link: https://lkml.kernel.org/r/20230324052233.2654090-4-rppt@kernel.orgSigned-off-by:
Kefeng Wang <wangkefeng.wang@huawei.com> Cc: Arnd Bergmann <arnd@arndb.de> Cc: Catalin Marinas <catalin.marinas@arm.com> Cc: Christophe Leroy <christophe.leroy@csgroup.eu> Cc: David Miller <davem@davemloft.net> Cc: Dinh Nguyen <dinguyen@kernel.org> Cc: Geert Uytterhoeven <geert@linux-m68k.org> Cc: Guo Ren <guoren@kernel.org> Cc: John Paul Adrian Glaubitz <glaubitz@physik.fu-berlin.de> Cc: Max Filippov <jcmvbkbc@gmail.com> Cc: Michael Ellerman <mpe@ellerman.id.au> Cc: Rich Felker <dalias@libc.org> Cc: "Russell King (Oracle)" <linux@armlinux.org.uk> Cc: Will Deacon <will@kernel.org> Cc: Yoshinori Sato <ysato@users.sourceforge.jp> Signed-off-by:
-
Mike Rapoport (IBM) authored
It is not a good idea to change fundamental parameters of core memory management. Having predefined ranges suggests that the values within those ranges are sensible, but one has to *really* understand implications of changing MAX_ORDER before actually amending it and ranges don't help here. Drop ranges in definition of ARCH_FORCE_MAX_ORDER and make its prompt visible only if EXPERT=y Link: https://lkml.kernel.org/r/20230324052233.2654090-3-rppt@kernel.orgSigned-off-by:
-
Mike Rapoport (IBM) authored
Patch series "arch,mm: cleanup Kconfig entries for ARCH_FORCE_MAX_ORDER", v3. Several architectures have ARCH_FORCE_MAX_ORDER in their Kconfig and they all have wrong and misleading prompt and help text for this option. Besides, some define insane limits for possible values of ARCH_FORCE_MAX_ORDER, some carefully define ranges only for a subset of possible configurations, some make this option configurable by users for no good reason. This set updates the prompt and help text everywhere and does its best to update actual definitions of ranges where applicable. kbuild generated a bunch of false positives because it assigns -1 to ARCH_FORCE_MAX_ORDER, hopefully this will be fixed soon. This patch (of 14): The prompt and help text of ARCH_FORCE_MAX_ORDER are not even close to describe this configuration option. Update both to actually describe what this option does. Link: https://lkml.kernel.org/r/20230325060828.2662773-1-rppt@kernel.org Link: https://lkml.kernel.org/r/20230324052233.2654090-1-rppt@kernel.org Link: https://lkml.kernel.org/r/20230324052233.2654090-2-rppt@kernel.orgSigned-off-by:
-
Michal Hocko authored
Leonardo Bras has noticed that pcp charge cache draining might be disruptive on workloads relying on 'isolated cpus', a feature commonly used on workloads that are sensitive to interruption and context switching such as vRAN and Industrial Control Systems. There are essentially two ways how to approach the issue. We can either allow the pcp cache to be drained on a different rather than a local cpu or avoid remote flushing on isolated cpus. The current pcp charge cache is really optimized for high performance and it always relies to stick with its cpu. That means it only requires local_lock (preempt_disable on !RT) and draining is handed over to pcp WQ to drain locally again. The former solution (remote draining) would require to add an additional locking to prevent local charges from racing with the draining. This adds an atomic operation to otherwise simple arithmetic fast path in the try_charge path. Another concern is that the remote draining can cause a lock contention for the isolated workloads and therefore interfere with it indirectly via user space interfaces. Another option is to avoid draining scheduling on isolated cpus altogether. That means that those remote cpus would keep their charges even after drain_all_stock returns. This is certainly not optimal either but it shouldn't really cause any major problems. In the worst case (many isolated cpus with charges - each of them with MEMCG_CHARGE_BATCH i.e 64 page) the memory consumption of a memcg would be artificially higher than can be immediately used from other cpus. Theoretically a memcg OOM killer could be triggered pre-maturely. Currently it is not really clear whether this is a practical problem though. Tight memcg limit would be really counter productive to cpu isolated workloads pretty much by definition because any memory reclaimed induced by memcg limit could break user space timing expectations as those usually expect execution in the userspace most of the time. Also charges could be left behind on memcg removal. Any future charge on those isolated cpus will drain that pcp cache so this won't be a permanent leak. Considering cons and pros of both approaches this patch is implementing the second option and simply do not schedule remote draining if the target cpu is isolated. This solution is much more simpler. It doesn't add any new locking and it is more more predictable from the user space POV. Should the pre-mature memcg OOM become a real life problem, we can revisit this decision. [akpm@linux-foundation.org: memcontrol.c needs sched/isolation.h] Link: https://lore.kernel.org/oe-kbuild-all/202303180617.7E3aIlHf-lkp@intel.com/ Link: https://lkml.kernel.org/r/20230317134448.11082-3-mhocko@kernel.orgSigned-off-by:
Michal Hocko <mhocko@suse.com> Suggested-by:
Roman Gushchin <roman.gushchin@linux.dev> Acked-by:
Leonardo Bras <leobras@redhat.com> Acked-by:
Shakeel Butt <shakeelb@google.com> Cc: Marcelo Tosatti <mtosatti@redhat.com> Cc: Muchun Song <muchun.song@linux.dev> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Frederic Weisbecker <frederic@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Ingo Molnar <mingo@elte.hu> Signed-off-by:
-
Frederic Weisbecker authored
Patch series "memcg, cpuisol: do not interfere pcp cache charges draining with cpuisol workloads". Leonardo has reported [1] that pcp memcg charge draining can interfere with cpu isolated workloads. The said draining is done from a WQ context with a pcp worker scheduled on each CPU which holds any cached charges for a specific memcg hierarchy. Operation is not really a common operation [2]. It can be triggered from the userspace though so some care is definitely due. Leonardo has tried to address the issue by allowing remote charge draining [3]. This approach requires an additional locking to synchronize pcp caches sync from a remote cpu from local pcp consumers. Even though the proposed lock was per-cpu there is still potential for contention and less predictable behavior. This patchset addresses the issue from a different angle. Rather than dealing with a potential synchronization, cpus which are isolated are simply never scheduled to be drained. This means that a small amount of charges could be laying around and waiting for a later use or they are flushed when a different memcg is charged from the same cpu. More details are in patch 2. The first patch from Frederic is implementing an abstraction to tell whether a specific cpu has been isolated and therefore require a special treatment. This patch (of 2): Provide this new API to check if a CPU has been isolated either through isolcpus= or nohz_full= kernel parameter. It aims at avoiding kernel load deemed to be safely spared on CPUs running sensitive workload that can't bear any disturbance, such as pcp cache draining. Link: https://lkml.kernel.org/r/20230317134448.11082-1-mhocko@kernel.org Link: https://lkml.kernel.org/r/20230317134448.11082-2-mhocko@kernel.orgSigned-off-by:
Frederic Weisbecker <frederic@kernel.org> Signed-off-by:
-
Ivan Orlov authored
Syzkaller reported the following issue: kernel BUG at mm/khugepaged.c:1823! invalid opcode: 0000 [#1] PREEMPT SMP KASAN CPU: 1 PID: 5097 Comm: syz-executor220 Not tainted 6.2.0-syzkaller-13154-g857f1268 #0 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 02/16/2023 RIP: 0010:collapse_file mm/khugepaged.c:1823 [inline] RIP: 0010:hpage_collapse_scan_file+0x67c8/0x7580 mm/khugepaged.c:2233 Code: 00 00 89 de e8 c9 66 a3 ff 31 ff 89 de e8 c0 66 a3 ff 45 84 f6 0f 85 28 0d 00 00 e8 22 64 a3 ff e9 dc f7 ff ff e8 18 64 a3 ff <0f> 0b f3 0f 1e fa e8 0d 64 a3 ff e9 93 f6 ff ff f3 0f 1e fa 4c 89 RSP: 0018:ffffc90003dff4e0 EFLAGS: 00010093 RAX: ffffffff81e95988 RBX: 00000000000001c1 RCX: ffff8880205b3a80 RDX: 0000000000000000 RSI: 00000000000001c0 RDI: 00000000000001c1 RBP: ffffc90003dff830 R08: ffffffff81e90e67 R09: fffffbfff1a433c3 R10: 0000000000000000 R11: dffffc0000000001 R12: 0000000000000000 R13: ffffc90003dff6c0 R14: 00000000000001c0 R15: 0000000000000000 FS: 00007fdbae5ee700(0000) GS:ffff8880b9900000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 00007fdbae6901e0 CR3: 000000007b2dd000 CR4: 00000000003506e0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: <TASK> madvise_collapse+0x721/0xf50 mm/khugepaged.c:2693 madvise_vma_behavior mm/madvise.c:1086 [inline] madvise_walk_vmas mm/madvise.c:1260 [inline] do_madvise+0x9e5/0x4680 mm/madvise.c:1439 __do_sys_madvise mm/madvise.c:1452 [inline] __se_sys_madvise mm/madvise.c:1450 [inline] __x64_sys_madvise+0xa5/0xb0 mm/madvise.c:1450 do_syscall_x64 arch/x86/entry/common.c:50 [inline] do_syscall_64+0x41/0xc0 arch/x86/entry/common.c:80 entry_SYSCALL_64_after_hwframe+0x63/0xcd The xas_store() call during page cache scanning can potentially translate 'xas' into the error state (with the reproducer provided by the syzkaller the error code is -ENOMEM). However, there are no further checks after the 'xas_store', and the next call of 'xas_next' at the start of the scanning cycle doesn't increase the xa_index, and the issue occurs. This patch will add the xarray state error checking after the xas_store() and the corresponding result error code. Tested via syzbot. [akpm@linux-foundation.org: update include/trace/events/huge_memory.h's SCAN_STATUS] Link: https://lkml.kernel.org/r/20230329145330.23191-1-ivan.orlov0322@gmail.com Link: https://syzkaller.appspot.com/bug?id=7d6bb3760e026ece7524500fe44fb024a0e959fcSigned-off-by:
Ivan Orlov <ivan.orlov0322@gmail.com> Reported-by: syzbot+9578faa5475acb35fa50@syzkaller.appspotmail.com Tested-by:
Zach O'Keefe <zokeefe@google.com> Cc: Yang Shi <shy828301@gmail.com> Cc: Himadri Pandya <himadrispandya@gmail.com> Cc: Ivan Orlov <ivan.orlov0322@gmail.com> Cc: Shuah Khan <skhan@linuxfoundation.org> Cc: Song Liu <songliubraving@fb.com> Cc: Rik van Riel <riel@surriel.com> Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Signed-off-by:
-
Arnd Bergmann authored
Some unknown -mllvm options (i.e. those starting with the letter "h") don't cause an error to be returned by clang, so the cc-option helper adds the unknown hwasan-kernel-mem-intrinsic-prefix=1 flag to CFLAGS with compilers that are new enough for hwasan but too old for this option. This causes a rather unreadable build failure: fixdep: error opening file: scripts/mod/.empty.o.d: No such file or directory make[4]: *** [/home/arnd/arm-soc/scripts/Makefile.build:252: scripts/mod/empty.o] Error 2 fixdep: error opening file: scripts/mod/.devicetable-offsets.s.d: No such file or directory make[4]: *** [/home/arnd/arm-soc/scripts/Makefile.build:114: scripts/mod/devicetable-offsets.s] Error 2 Add a version check to only allow this option with clang-15, gcc-13 or later versions. Link: https://lkml.kernel.org/r/20230418122350.1646391-1-arnd@kernel.org Fixes: 51287dcb ("kasan: emit different calls for instrumentable memintrinsics") Link: https://lore.kernel.org/all/CANpmjNMwYosrvqh4ogDO8rgn+SeDHM2b-shD21wTypm_6MMe=g@mail.gmail.com/Signed-off-by:
Arnd Bergmann <arnd@arndb.de> Reviewed-by:
Marco Elver <elver@google.com> Reviewed-by:
Nathan Chancellor <nathan@kernel.org> Cc: Alexander Potapenko <glider@google.com> Cc: Andrey Konovalov <andreyknvl@gmail.com> Cc: Andrey Ryabinin <ryabinin.a.a@gmail.com> Cc: Dmitry Vyukov <dvyukov@google.com> Cc: Masahiro Yamada <masahiroy@kernel.org> Cc: Michael Ellerman <mpe@ellerman.id.au> Cc: Nick Desaulniers <ndesaulniers@google.com> Cc: Nicolas Schier <nicolas@fjasle.eu> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Tom Rix <trix@redhat.com> Cc: Vincenzo Frascino <Vincenzo.Frascino@arm.com> Signed-off-by:
-
Sergey Senozhatsky authored
The current implementation of the compaction loop fails to set the source zspage pointer to NULL in all cases, leading to a potential issue where __zs_compact() could use a stale zspage pointer. This pointer could even point to a previously freed zspage, causing unexpected behavior in the putback_zspage() and migrate_write_unlock() functions after returning from the compaction loop. Address the issue by ensuring that the source zspage pointer is always set to NULL when it should be. Link: https://lkml.kernel.org/r/20230417130850.1784777-1-senozhatsky@chromium.org Fixes: 5a845e9f ("zsmalloc: rework compaction algorithm") Signed-off-by:
Sergey Senozhatsky <senozhatsky@chromium.org> Reported-by:
Yu Zhao <yuzhao@google.com> Tested-by:
Yosry Ahmed <yosryahmed@google.com> Signed-off-by:
-
Arnd Bergmann authored
clang produces a build failure on x86 for some randconfig builds after a change that moves around code to mm/mm_init.c: Cannot find symbol for section 2: .text. mm/mm_init.o: failed I have not been able to figure out why this happens, but the __weak annotation on arch_has_descending_max_zone_pfns() is the trigger here. Removing the weak function in favor of an open-coded Kconfig option check avoids the problem and becomes clearer as well as better to optimize by the compiler. [arnd@arndb.de: fix logic bug] Link: https://lkml.kernel.org/r/20230415081904.969049-1-arnd@kernel.org Link: https://lkml.kernel.org/r/20230414080418.110236-1-arnd@kernel.org Fixes: 9420f89d ("mm: move most of core MM initialization to mm/mm_init.c") Signed-off-by:
Vlastimil Babka <vbabka@suse.cz> Tested-by:
SeongJae Park <sj@kernel.org> Tested-by:
Geert Uytterhoeven <geert+renesas@glider.be> Acked-by:
-
Kirill A. Shutemov authored
23baf831 ("mm, treewide: redefine MAX_ORDER sanely") results in various boot failures (hang) on arm targets Debug messages reveal the reason. ########### MAX_ORDER=10 start=0 __ffs(start)=-1 min()=10 min_t=-1 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ If start==0, __ffs(start) returns 0xfffffff or (as int) -1, which min_t() interprets as such, while min() apparently uses the returned unsigned long value. Obviously a negative order isn't received well by the rest of the code. [akpm@linux-foundation.org: fix comment, per Mike] Link: https://lkml.kernel.org/r/ZDBa7HWZK69dKKzH@kernel.org Link: https://lkml.kernel.org/r/20230406072529.vupqyrzqnhyozeyh@box.shutemov.name Fixes: 23baf831 ("mm, treewide: redefine MAX_ORDER sanely") Signed-off-by:
"Kirill A. Shutemov" <kirill@shutemov.name> Reported-by:
Guenter Roeck <linux@roeck-us.net> Link: https://lkml.kernel.org/r/9460377a-38aa-4f39-ad57-fb73725f92db@roeck-us.netReviewed-by:
-
Andrew Morton authored
-
Ryusuke Konishi authored
Syzbot still reports uninit-value in nilfs_add_checksums_on_logs() for KMSAN enabled kernels after applying commit 73970316 ("nilfs2: initialize "struct nilfs_binfo_dat"->bi_pad field"). This is because the unused bytes at the end of each block in segment summaries are not initialized. So this fixes the issue by padding the unused bytes with null bytes. Link: https://lkml.kernel.org/r/20230417173513.12598-1-konishi.ryusuke@gmail.comSigned-off-by:
Ryusuke Konishi <konishi.ryusuke@gmail.com> Tested-by:
-
Mel Gorman authored
A bug was reported by Yuanxi Liu where allocating 1G pages at runtime is taking an excessive amount of time for large amounts of memory. Further testing allocating huge pages that the cost is linear i.e. if allocating 1G pages in batches of 10 then the time to allocate nr_hugepages from 10->20->30->etc increases linearly even though 10 pages are allocated at each step. Profiles indicated that much of the time is spent checking the validity within already existing huge pages and then attempting a migration that fails after isolating the range, draining pages and a whole lot of other useless work. Commit eb14d4ee ("mm,page_alloc: drop unnecessary checks from pfn_range_valid_contig") removed two checks, one which ignored huge pages for contiguous allocations as huge pages can sometimes migrate. While there may be value on migrating a 2M page to satisfy a 1G allocation, it's potentially expensive if the 1G allocation fails and it's pointless to try moving a 1G page for a new 1G allocation or scan the tail pages for valid PFNs. Reintroduce the PageHuge check and assume any contiguous region with hugetlbfs pages is unsuitable for a new 1G allocation. The hpagealloc test allocates huge pages in batches and reports the average latency per page over time. This test happens just after boot when fragmentation is not an issue. Units are in milliseconds. hpagealloc 6.3.0-rc6 6.3.0-rc6 6.3.0-rc6 vanilla hugeallocrevert-v1r1 hugeallocsimple-v1r2 Min Latency 26.42 ( 0.00%) 5.07 ( 80.82%) 18.94 ( 28.30%) 1st-qrtle Latency 356.61 ( 0.00%) 5.34 ( 98.50%) 19.85 ( 94.43%) 2nd-qrtle Latency 697.26 ( 0.00%) 5.47 ( 99.22%) 20.44 ( 97.07%) 3rd-qrtle Latency 972.94 ( 0.00%) 5.50 ( 99.43%) 20.81 ( 97.86%) Max-1 Latency 26.42 ( 0.00%) 5.07 ( 80.82%) 18.94 ( 28.30%) Max-5 Latency 82.14 ( 0.00%) 5.11 ( 93.78%) 19.31 ( 76.49%) Max-10 Latency 150.54 ( 0.00%) 5.20 ( 96.55%) 19.43 ( 87.09%) Max-90 Latency 1164.45 ( 0.00%) 5.53 ( 99.52%) 20.97 ( 98.20%) Max-95 Latency 1223.06 ( 0.00%) 5.55 ( 99.55%) 21.06 ( 98.28%) Max-99 Latency 1278.67 ( 0.00%) 5.57 ( 99.56%) 22.56 ( 98.24%) Max Latency 1310.90 ( 0.00%) 8.06 ( 99.39%) 26.62 ( 97.97%) Amean Latency 678.36 ( 0.00%) 5.44 * 99.20%* 20.44 * 96.99%* 6.3.0-rc6 6.3.0-rc6 6.3.0-rc6 vanilla revert-v1 hugeallocfix-v2 Duration User 0.28 0.27 0.30 Duration System 808.66 17.77 35.99 Duration Elapsed 830.87 18.08 36.33 The vanilla kernel is poor, taking up to 1.3 second to allocate a huge page and almost 10 minutes in total to run the test. Reverting the problematic commit reduces it to 8ms at worst and the patch takes 26ms. This patch fixes the main issue with skipping huge pages but leaves the page_count() out because a page with an elevated count potentially can migrate. BugLink: https://bugzilla.kernel.org/show_bug.cgi?id=217022 Link: https://lkml.kernel.org/r/20230414141429.pwgieuwluxwez3rj@techsingularity.net Fixes: eb14d4ee ("mm,page_alloc: drop unnecessary checks from pfn_range_valid_contig") Signed-off-by:
Mel Gorman <mgorman@techsingularity.net> Reported-by:
Yuanxi Liu <y.liu@naruida.com> Acked-by:
David Hildenbrand <david@redhat.com> Acked-by:
Oscar Salvador <osalvador@suse.de> Cc: Matthew Wilcox <willy@infradead.org> Cc: <stable@vger.kernel.org> Signed-off-by:
-
Liam R. Howlett authored
The maple tree limits the gap returned to a window that specifically fits what was asked. This may not be optimal in the case of switching search directions or a gap that does not satisfy the requested space for other reasons. Fix the search by retrying the operation and limiting the search window in the rare occasion that a conflict occurs. Link: https://lkml.kernel.org/r/20230414185919.4175572-1-Liam.Howlett@oracle.com Fixes: 3499a131 ("mm/mmap: use maple tree for unmapped_area{_topdown}") Signed-off-by:
Liam R. Howlett <Liam.Howlett@oracle.com> Reported-by:
Rick Edgecombe <rick.p.edgecombe@intel.com> Cc: <stable@vger.kernel.org> Signed-off-by:
-
Liam R. Howlett authored
The internal function of mas_awalk() was incorrectly skipping the last entry in a node, which could potentially be NULL. This is only a problem for the left-most node in the tree - otherwise that NULL would not exist. Fix mas_awalk() by using the metadata to obtain the end of the node for the loop and the logical pivot as apposed to the raw pivot value. Link: https://lkml.kernel.org/r/20230414145728.4067069-2-Liam.Howlett@oracle.com Fixes: 54a611b6 ("Maple Tree: add new data structure") Signed-off-by:
-
Liam R. Howlett authored
Stop using maple state min/max for the range by passing through pointers for those values. This will allow the maple state to be reused without resetting. Also add some logic to fail out early on searching with invalid arguments. Link: https://lkml.kernel.org/r/20230414145728.4067069-1-Liam.Howlett@oracle.com Fixes: 54a611b6 ("Maple Tree: add new data structure") Signed-off-by:
-
Alexander Potapenko authored
Similarly to kmsan_vmap_pages_range_noflush(), kmsan_ioremap_page_range() must also properly handle allocation/mapping failures. In the case of such, it must clean up the already created metadata mappings and return an error code, so that the error can be propagated to ioremap_page_range(). Without doing so, KMSAN may silently fail to bring the metadata for the page range into a consistent state, which will result in user-visible crashes when trying to access them. Link: https://lkml.kernel.org/r/20230413131223.4135168-2-glider@google.com Fixes: b073d7f8 ("mm: kmsan: maintain KMSAN metadata for page operations") Signed-off-by:
Alexander Potapenko <glider@google.com> Reported-by:
Dipanjan Das <mail.dipanjan.das@gmail.com> Link: https://lore.kernel.org/linux-mm/CANX2M5ZRrRA64k0hOif02TjmY9kbbO2aCBPyq79es34RXZ=cAw@mail.gmail.com/Reviewed-by:
-
Alexander Potapenko authored
As reported by Dipanjan Das, when KMSAN is used together with kernel fault injection (or, generally, even without the latter), calls to kcalloc() or __vmap_pages_range_noflush() may fail, leaving the metadata mappings for the virtual mapping in an inconsistent state. When these metadata mappings are accessed later, the kernel crashes. To address the problem, we return a non-zero error code from kmsan_vmap_pages_range_noflush() in the case of any allocation/mapping failure inside it, and make vmap_pages_range_noflush() return an error if KMSAN fails to allocate the metadata. This patch also removes KMSAN_WARN_ON() from vmap_pages_range_noflush(), as these allocation failures are not fatal anymore. Link: https://lkml.kernel.org/r/20230413131223.4135168-1-glider@google.com Fixes: b073d7f8 ("mm: kmsan: maintain KMSAN metadata for page operations") Signed-off-by:
-
SeongJae Park authored
Commit 799fb82a ("tools/vm: rename tools/vm to tools/mm") missed renaming 'vm' in 'tools/Makefile' to 'mm'. As a result, 'make clean' under 'tools/' directory fails as below: $ make -C tools clean DESCEND vm make[1]: Entering directory '/linux/tools/vm' make[1]: *** No rule to make target 'clean'. Stop. make[1]: Leaving directory '/linux/tools/vm' make: *** [Makefile:173: vm_clean] Error 2 make: Leaving directory '/linux/tools' Do the missed rename. Link: https://lkml.kernel.org/r/20230415203110.13858-1-sj@kernel.org Fixes: 799fb82a ("tools/vm: rename tools/vm to tools/mm") Signed-off-by:
Ricardo Pardini <ricardo@pardini.net> Link: https://lore.kernel.org/linux-mm/20230415202454.13558-1-sj@kernel.org/Tested-by:
-
Mathieu Desnoyers authored
commit f1a79412 ("mm: convert mm's rss stats into percpu_counter") introduces a memory leak by missing a call to destroy_context() when a percpu_counter fails to allocate. Before introducing the per-cpu counter allocations, init_new_context() was the last call that could fail in mm_init(), and thus there was no need to ever invoke destroy_context() in the error paths. Adding the following percpu counter allocations adds error paths after init_new_context(), which means its associated destroy_context() needs to be called when percpu counters fail to allocate. Link: https://lkml.kernel.org/r/20230330133822.66271-1-mathieu.desnoyers@efficios.com Fixes: f1a79412 ("mm: convert mm's rss stats into percpu_counter") Signed-off-by:
Mathieu Desnoyers <mathieu.desnoyers@efficios.com> Acked-by:
-
Tetsuo Handa authored
syzbot is reporting circular locking dependency which involves zonelist_update_seq seqlock [1], for this lock is checked by memory allocation requests which do not need to be retried. One deadlock scenario is kmalloc(GFP_ATOMIC) from an interrupt handler. CPU0 ---- __build_all_zonelists() { write_seqlock(&zonelist_update_seq); // makes zonelist_update_seq.seqcount odd // e.g. timer interrupt handler runs at this moment some_timer_func() { kmalloc(GFP_ATOMIC) { __alloc_pages_slowpath() { read_seqbegin(&zonelist_update_seq) { // spins forever because zonelist_update_seq.seqcount is odd } } } } // e.g. timer interrupt handler finishes write_sequnlock(&zonelist_update_seq); // makes zonelist_update_seq.seqcount even } This deadlock scenario can be easily eliminated by not calling read_seqbegin(&zonelist_update_seq) from !__GFP_DIRECT_RECLAIM allocation requests, for retry is applicable to only __GFP_DIRECT_RECLAIM allocation requests. But Michal Hocko does not know whether we should go with this approach. Another deadlock scenario which syzbot is reporting is a race between kmalloc(GFP_ATOMIC) from tty_insert_flip_string_and_push_buffer() with port->lock held and printk() from __build_all_zonelists() with zonelist_update_seq held. CPU0 CPU1 ---- ---- pty_write() { tty_insert_flip_string_and_push_buffer() { __build_all_zonelists() { write_seqlock(&zonelist_update_seq); build_zonelists() { printk() { vprintk() { vprintk_default() { vprintk_emit() { console_unlock() { console_flush_all() { console_emit_next_record() { con->write() = serial8250_console_write() { spin_lock_irqsave(&port->lock, flags); tty_insert_flip_string() { tty_insert_flip_string_fixed_flag() { __tty_buffer_request_room() { tty_buffer_alloc() { kmalloc(GFP_ATOMIC | __GFP_NOWARN) { __alloc_pages_slowpath() { zonelist_iter_begin() { read_seqbegin(&zonelist_update_seq); // spins forever because zonelist_update_seq.seqcount is odd spin_lock_irqsave(&port->lock, flags); // spins forever because port->lock is held } } } } } } } } spin_unlock_irqrestore(&port->lock, flags); // message is printed to console spin_unlock_irqrestore(&port->lock, flags); } } } } } } } } } write_sequnlock(&zonelist_update_seq); } } } This deadlock scenario can be eliminated by preventing interrupt context from calling kmalloc(GFP_ATOMIC) and preventing printk() from calling console_flush_all() while zonelist_update_seq.seqcount is odd. Since Petr Mladek thinks that __build_all_zonelists() can become a candidate for deferring printk() [2], let's address this problem by disabling local interrupts in order to avoid kmalloc(GFP_ATOMIC) and disabling synchronous printk() in order to avoid console_flush_all() . As a side effect of minimizing duration of zonelist_update_seq.seqcount being odd by disabling synchronous printk(), latency at read_seqbegin(&zonelist_update_seq) for both !__GFP_DIRECT_RECLAIM and __GFP_DIRECT_RECLAIM allocation requests will be reduced. Although, from lockdep perspective, not calling read_seqbegin(&zonelist_update_seq) (i.e. do not record unnecessary locking dependency) from interrupt context is still preferable, even if we don't allow calling kmalloc(GFP_ATOMIC) inside write_seqlock(&zonelist_update_seq)/write_sequnlock(&zonelist_update_seq) section... Link: https://lkml.kernel.org/r/8796b95c-3da3-5885-fddd-6ef55f30e4d3@I-love.SAKURA.ne.jp Fixes: 3d36424b ("mm/page_alloc: fix race condition between build_all_zonelists and page allocation") Link: https://lkml.kernel.org/r/ZCrs+1cDqPWTDFNM@alley [2] Reported-by:syzbot <syzbot+223c7461c58c58a4cb10@syzkaller.appspotmail.com> Link: https://syzkaller.appspot.com/bug?extid=223c7461c58c58a4cb10 [1] Signed-off-by:
Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp> Acked-by:
-
Ondrej Mosnacek authored
Linux Security Modules (LSMs) that implement the "capable" hook will usually emit an access denial message to the audit log whenever they "block" the current task from using the given capability based on their security policy. The occurrence of a denial is used as an indication that the given task has attempted an operation that requires the given access permission, so the callers of functions that perform LSM permission checks must take care to avoid calling them too early (before it is decided if the permission is actually needed to perform the requested operation). The __sys_setres[ug]id() functions violate this convention by first calling ns_capable_setid() and only then checking if the operation requires the capability or not. It means that any caller that has the capability granted by DAC (task's capability set) but not by MAC (LSMs) will generate a "denied" audit record, even if is doing an operation for which the capability is not required. Fix this by reordering the checks such that ns_capable_setid() is checked last and -EPERM is returned immediately if it returns false. While there, also do two small optimizations: * move the capability check before prepare_creds() and * bail out early in case of a no-op. Link: https://lkml.kernel.org/r/20230217162154.837549-1-omosnace@redhat.com Fixes: 1da177e4 ("Linux-2.6.12-rc2") Signed-off-by:
Ondrej Mosnacek <omosnace@redhat.com> Cc: Eric W. Biederman <ebiederm@xmission.com> Cc: <stable@vger.kernel.org> Signed-off-by:
-
- 16 Apr, 2023 12 commits
-
-
Andrew Morton authored
-
Peter Xu authored
This is a proposal to revert commit 914eedcb. I found this when writing a simple UFFDIO_API test to be the first unit test in this set. Two things breaks with the commit: - UFFDIO_API check was lost and missing. According to man page, the kernel should reject ioctl(UFFDIO_API) if uffdio_api.api != 0xaa. This check is needed if the api version will be extended in the future, or user app won't be able to identify which is a new kernel. - Feature flags checks were removed, which means UFFDIO_API with a feature that does not exist will also succeed. According to the man page, we should (and it makes sense) to reject ioctl(UFFDIO_API) if unknown features passed in. Link: https://lore.kernel.org/r/20220722201513.1624158-1-axelrasmussen@google.com Link: https://lkml.kernel.org/r/20230412163922.327282-2-peterx@redhat.com Fixes: 914eedcb ("userfaultfd: don't fail on unrecognized features") Signed-off-by:
Peter Xu <peterx@redhat.com> Acked-by:
-
Baokun Li authored
KASAN report null-ptr-deref: ================================================================== BUG: KASAN: null-ptr-deref in bdi_split_work_to_wbs+0x5c5/0x7b0 Write of size 8 at addr 0000000000000000 by task sync/943 CPU: 5 PID: 943 Comm: sync Tainted: 6.3.0-rc5-next-20230406-dirty #461 Call Trace: <TASK> dump_stack_lvl+0x7f/0xc0 print_report+0x2ba/0x340 kasan_report+0xc4/0x120 kasan_check_range+0x1b7/0x2e0 __kasan_check_write+0x24/0x40 bdi_split_work_to_wbs+0x5c5/0x7b0 sync_inodes_sb+0x195/0x630 sync_inodes_one_sb+0x3a/0x50 iterate_supers+0x106/0x1b0 ksys_sync+0x98/0x160 [...] ================================================================== The race that causes the above issue is as follows: cpu1 cpu2 -------------------------|------------------------- inode_switch_wbs INIT_WORK(&isw->work, inode_switch_wbs_work_fn) queue_rcu_work(isw_wq, &isw->work) // queue_work async inode_switch_wbs_work_fn wb_put_many(old_wb, nr_switched) percpu_ref_put_many ref->data->release(ref) cgwb_release queue_work(cgwb_release_wq, &wb->release_work) // queue_work async &wb->release_work cgwb_release_workfn ksys_sync iterate_supers sync_inodes_one_sb sync_inodes_sb bdi_split_work_to_wbs kmalloc(sizeof(*work), GFP_ATOMIC) // alloc memory failed percpu_ref_exit ref->data = NULL kfree(data) wb_get(wb) percpu_ref_get(&wb->refcnt) percpu_ref_get_many(ref, 1) atomic_long_add(nr, &ref->data->count) atomic64_add(i, v) // trigger null-ptr-deref bdi_split_work_to_wbs() traverses &bdi->wb_list to split work into all wbs. If the allocation of new work fails, the on-stack fallback will be used and the reference count of the current wb is increased afterwards. If cgroup writeback membership switches occur before getting the reference count and the current wb is released as old_wd, then calling wb_get() or wb_put() will trigger the null pointer dereference above. This issue was introduced in v4.3-rc7 (see fix tag1). Both sync_inodes_sb() and __writeback_inodes_sb_nr() calls to bdi_split_work_to_wbs() can trigger this issue. For scenarios called via sync_inodes_sb(), originally commit 7fc5854f ("writeback: synchronize sync(2) against cgroup writeback membership switches") reduced the possibility of the issue by adding wb_switch_rwsem, but in v5.14-rc1 (see fix tag2) removed the "inode_io_list_del_locked(inode, old_wb)" from inode_switch_wbs_work_fn() so that wb->state contains WB_has_dirty_io, thus old_wb is not skipped when traversing wbs in bdi_split_work_to_wbs(), and the issue becomes easily reproducible again. To solve this problem, percpu_ref_exit() is called under RCU protection to avoid race between cgwb_release_workfn() and bdi_split_work_to_wbs(). Moreover, replace wb_get() with wb_tryget() in bdi_split_work_to_wbs(), and skip the current wb if wb_tryget() fails because the wb has already been shutdown. Link: https://lkml.kernel.org/r/20230410130826.1492525-1-libaokun1@huawei.com Fixes: b817525a ("writeback: bdi_writeback iteration must not skip dying ones") Signed-off-by:Baokun Li <libaokun1@huawei.com> Reviewed-by:
Jan Kara <jack@suse.cz> Acked-by:
Tejun Heo <tj@kernel.org> Cc: Alexander Viro <viro@zeniv.linux.org.uk> Cc: Andreas Dilger <adilger.kernel@dilger.ca> Cc: Christian Brauner <brauner@kernel.org> Cc: Dennis Zhou <dennis@kernel.org> Cc: Hou Tao <houtao1@huawei.com> Cc: yangerkun <yangerkun@huawei.com> Cc: Zhang Yi <yi.zhang@huawei.com> Cc: Jens Axboe <axboe@kernel.dk> Cc: <stable@vger.kernel.org> Signed-off-by:
-
Peng Zhang authored
In mas_alloc_nodes(), "node->node_count = 0" means to initialize the node_count field of the new node, but the node may not be a new node. It may be a node that existed before and node_count has a value, setting it to 0 will cause a memory leak. At this time, mas->alloc->total will be greater than the actual number of nodes in the linked list, which may cause many other errors. For example, out-of-bounds access in mas_pop_node(), and mas_pop_node() may return addresses that should not be used. Fix it by initializing node_count only for new nodes. Also, by the way, an if-else statement was removed to simplify the code. Link: https://lkml.kernel.org/r/20230411041005.26205-1-zhangpeng.00@bytedance.com Fixes: 54a611b6 ("Maple Tree: add new data structure") Signed-off-by:
Peng Zhang <zhangpeng.00@bytedance.com> Reviewed-by:
-
Steve Chou authored
When using cull option with 'tg' flag, the fprintf is using pid instead of tgid. It should use tgid instead. Link: https://lkml.kernel.org/r/20230411034929.2071501-1-steve_chou@pesi.com.tw Fixes: 9c8a0a8e ("tools/vm/page_owner_sort.c: support for user-defined culling rules") Signed-off-by:
Steve Chou <steve_chou@pesi.com.tw> Cc: Jiajian Ye <yejiajian2018@email.szu.edu.cn> Cc: <stable@vger.kernel.org> Signed-off-by:
-
Jonathan Toppins authored
Link: https://lkml.kernel.org/r/d79bc6eaf65e68bd1c2a1e1510ab6291ce5926a6.1681162487.git.jtoppins@redhat.comSigned-off-by:
Jonathan Toppins <jtoppins@redhat.com> Cc: Colin Ian King <colin.i.king@gmail.com> Cc: Jakub Kicinski <kuba@kernel.org> Cc: Kirill Tkhai <tkhai@ya.ru> Cc: Konrad Dybcio <konrad.dybcio@linaro.org> Cc: Qais Yousef <qyousef@layalina.io> Cc: Stephen Hemminger <stephen@networkplumber.org> Signed-off-by:
-
Liam R. Howlett authored
set_mempolicy_home_node() iterates over a list of VMAs and calls mbind_range() on each VMA, which also iterates over the singular list of the VMA passed in and potentially splits the VMA. Since the VMA iterator is not passed through, set_mempolicy_home_node() may now point to a stale node in the VMA tree. This can result in a UAF as reported by syzbot. Avoid the stale maple tree node by passing the VMA iterator through to the underlying call to split_vma(). mbind_range() is also overly complicated, since there are two calling functions and one already handles iterating over the VMAs. Simplify mbind_range() to only handle merging and splitting of the VMAs. Align the new loop in do_mbind() and existing loop in set_mempolicy_home_node() to use the reduced mbind_range() function. This allows for a single location of the range calculation and avoids constantly looking up the previous VMA (since this is a loop over the VMAs). Link: https://lore.kernel.org/linux-mm/000000000000c93feb05f87e24ad@google.com/ Fixes: 66850be5 ("mm/mempolicy: use vma iterator & maple state instead of vma linked list") Signed-off-by:
-
Naoya Horiguchi authored
split_huge_page_to_list() WARNs when called for huge zero pages, which sounds to me too harsh because it does not imply a kernel bug, but just notifies the event to admins. On the other hand, this is considered as critical by syzkaller and makes its testing less efficient, which seems to me harmful. So replace the VM_WARN_ON_ONCE_FOLIO with pr_warn_ratelimited. Link: https://lkml.kernel.org/r/20230406082004.2185420-1-naoya.horiguchi@linux.dev Fixes: 478d134e ("mm/huge_memory: do not overkill when splitting huge_zero_page") Signed-off-by:
Naoya Horiguchi <naoya.horiguchi@nec.com> Reported-by: syzbot+07a218429c8d19b1fb25@syzkaller.appspotmail.com Link: https://lore.kernel.org/lkml/000000000000a6f34a05e6efcd01@google.com/Reviewed-by:
Yang Shi <shy828301@gmail.com> Cc: Miaohe Lin <linmiaohe@huawei.com> Cc: Tetsuo Handa <penguin-kernel@i-love.sakura.ne.jp> Cc: Xu Yu <xuyu@linux.alibaba.com> Cc: <stable@vger.kernel.org> Signed-off-by:
-
Liam R. Howlett authored
When the loop over the VMA is terminated early due to an error, the return code could be overwritten with ENOMEM. Fix the return code by only setting the error on early loop termination when the error is not set. User-visible effects include: attempts to run mprotect() against a special mapping or with a poorly-aligned hugetlb address should return -EINVAL, but they presently return -ENOMEM. In other cases an -EACCESS should be returned. Link: https://lkml.kernel.org/r/20230406193050.1363476-1-Liam.Howlett@oracle.com Fixes: 2286a691 ("mm: change mprotect_fixup to vma iterator") Signed-off-by:
-
Peter Xu authored
Khugepaged collapse an anonymous thp in two rounds of scans. The 2nd round done in __collapse_huge_page_isolate() after hpage_collapse_scan_pmd(), during which all the locks will be released temporarily. It means the pgtable can change during this phase before 2nd round starts. It's logically possible some ptes got wr-protected during this phase, and we can errornously collapse a thp without noticing some ptes are wr-protected by userfault. e1e267c7 wanted to avoid it but it only did that for the 1st phase, not the 2nd phase. Since __collapse_huge_page_isolate() happens after a round of small page swapins, we don't need to worry on any !present ptes - if it existed khugepaged will already bail out. So we only need to check present ptes with uffd-wp bit set there. This is something I found only but never had a reproducer, I thought it was one caused a bug in Muhammad's recent pagemap new ioctl work, but it turns out it's not the cause of that but an userspace bug. However this seems to still be a real bug even with a very small race window, still worth to have it fixed and copy stable. Link: https://lkml.kernel.org/r/20230405155120.3608140-1-peterx@redhat.com Fixes: e1e267c7 ("khugepaged: skip collapse if uffd-wp detected") Signed-off-by:
-
David Hildenbrand authored
Looks like what we fixed for hugetlb in commit 44f86392 ("mm/hugetlb: fix uffd-wp handling for migration entries in hugetlb_change_protection()") similarly applies to THP. Setting/clearing uffd-wp on THP migration entries is not implemented properly. Further, while removing migration PMDs considers the uffd-wp bit, inserting migration PMDs does not consider the uffd-wp bit. We have to set/clear independently of the migration entry type in change_huge_pmd() and properly copy the uffd-wp bit in set_pmd_migration_entry(). Verified using a simple reproducer that triggers migration of a THP, that the set_pmd_migration_entry() no longer loses the uffd-wp bit. Link: https://lkml.kernel.org/r/20230405160236.587705-2-david@redhat.com Fixes: f45ec5ff ("userfaultfd: wp: support swap and page migration") Signed-off-by:
-
Qi Zheng authored
The ->percpu_pvec_drained was originally introduced by commit d9ed0d08 ("mm: only drain per-cpu pagevecs once per pagevec usage") to drain per-cpu pagevecs only once per pagevec usage. But after converting the swap code to be more folio-based, the commit c2bc1681 ("mm/swap: add folio_batch_move_lru()") breaks this logic, which would cause ->percpu_pvec_drained to be reset to false, that means per-cpu pagevecs will be drained multiple times per pagevec usage. In theory, there should be no functional changes when converting code to be more folio-based. We should call folio_batch_reinit() in folio_batch_move_lru() instead of folio_batch_init(). And to verify that we still need ->percpu_pvec_drained, I ran mmtests/sparsetruncate-tiny and got the following data: baseline with baseline/ patch/ Min Time 326.00 ( 0.00%) 328.00 ( -0.61%) 1st-qrtle Time 334.00 ( 0.00%) 336.00 ( -0.60%) 2nd-qrtle Time 338.00 ( 0.00%) 341.00 ( -0.89%) 3rd-qrtle Time 343.00 ( 0.00%) 347.00 ( -1.17%) Max-1 Time 326.00 ( 0.00%) 328.00 ( -0.61%) Max-5 Time 327.00 ( 0.00%) 330.00 ( -0.92%) Max-10 Time 328.00 ( 0.00%) 331.00 ( -0.91%) Max-90 Time 350.00 ( 0.00%) 357.00 ( -2.00%) Max-95 Time 395.00 ( 0.00%) 390.00 ( 1.27%) Max-99 Time 508.00 ( 0.00%) 434.00 ( 14.57%) Max Time 547.00 ( 0.00%) 476.00 ( 12.98%) Amean Time 344.61 ( 0.00%) 345.56 * -0.28%* Stddev Time 30.34 ( 0.00%) 19.51 ( 35.69%) CoeffVar Time 8.81 ( 0.00%) 5.65 ( 35.87%) BAmean-99 Time 342.38 ( 0.00%) 344.27 ( -0.55%) BAmean-95 Time 338.58 ( 0.00%) 341.87 ( -0.97%) BAmean-90 Time 336.89 ( 0.00%) 340.26 ( -1.00%) BAmean-75 Time 335.18 ( 0.00%) 338.40 ( -0.96%) BAmean-50 Time 332.54 ( 0.00%) 335.42 ( -0.87%) BAmean-25 Time 329.30 ( 0.00%) 332.00 ( -0.82%) From the above it can be seen that we get similar data to when ->percpu_pvec_drained was introduced, so we still need it. Let's call folio_batch_reinit() in folio_batch_move_lru() to restore the original logic. Link: https://lkml.kernel.org/r/20230405161854.6931-1-zhengqi.arch@bytedance.com Fixes: c2bc1681 ("mm/swap: add folio_batch_move_lru()") Signed-off-by:
Qi Zheng <zhengqi.arch@bytedance.com> Reviewed-by:
Matthew Wilcox (Oracle) <willy@infradead.org> Acked-by:
Mel Gorman <mgorman@suse.de> Cc: Lorenzo Stoakes <lstoakes@gmail.com> Cc: Vlastimil Babka <vbabka@suse.cz> Signed-off-by:
-
- 06 Apr, 2023 4 commits
-
-
Raghavendra K T authored
before: last 6 bits of PID is used as index to store information about tasks accessing VMA's. after: hash_32 is used to take of cases where tasks are created over a period of time, and thus improve collision probability. Result: The patch series overall improves autonuma cost. Kernbench around more than 5% improvement and system time in mmtest autonuma showed more than 80% improvement Link: https://lkml.kernel.org/r/d5a9f75513300caed74e5c8570bba9317b963c2b.1677672277.git.raghavendra.kt@amd.comSigned-off-by:
Raghavendra K T <raghavendra.kt@amd.com> Suggested-by:
Peter Zijlstra <peterz@infradead.org> Cc: Bharata B Rao <bharata@amd.com> Cc: David Hildenbrand <david@redhat.com> Cc: Disha Talreja <dishaa.talreja@amd.com> Cc: Ingo Molnar <mingo@redhat.com> Cc: Mel Gorman <mgorman@techsingularity.net> Cc: Mike Rapoport <rppt@kernel.org> Signed-off-by:
-
Raghavendra K T authored
This helps to ensure that only recently accessed PIDs scan the VMAs. Current implementation: (idea supported by PeterZ) 1. Accessing PID information is maintained in two windows. access_pids[1] being newest. 2. Reset old access PID info i.e. access_pid[0] every (4 * sysctl_numa_balancing_scan_delay) interval after initial scan delay period expires. The above interval seemed to be experimentally optimum since it avoids frequent reset of access info as well as helps clearing the old access info regularly. The reset logic is implemented in scan path. Link: https://lkml.kernel.org/r/f7a675f66d1442d048b4216b2baf94515012c405.1677672277.git.raghavendra.kt@amd.comSigned-off-by: -
Raghavendra K T authored
During Numa scanning make sure only relevant vmas of the tasks are scanned. Before: All the tasks of a process participate in scanning the vma even if they do not access vma in it's lifespan. Now: Except cases of first few unconditional scans, if a process do not touch vma (exluding false positive cases of PID collisions) tasks no longer scan all vma Logic used: 1) 6 bits of PID used to mark active bit in vma numab status during fault to remember PIDs accessing vma. (Thanks Mel) 2) Subsequently in scan path, vma scanning is skipped if current PID had not accessed vma. 3) First two times we do allow unconditional scan to preserve earlier behaviour of scanning. Acknowledgement to Bharata B Rao <bharata@amd.com> for initial patch to store pid information and Peter Zijlstra <peterz@infradead.org> (Usage of test and set bit) Link: https://lkml.kernel.org/r/092f03105c7c1d3450f4636b1ea350407f07640e.1677672277.git.raghavendra.kt@amd.comSigned-off-by:
-
Mel Gorman authored
Pach series "sched/numa: Enhance vma scanning", v3. The patchset proposes one of the enhancements to numa vma scanning suggested by Mel. This is continuation of [3]. Reposting the rebased patchset to akpm mm-unstable tree (March 1) Existing mechanism of scan period involves, scan period derived from per-thread stats. Process Adaptive autoNUMA [1] proposed to gather NUMA fault stats at per-process level to capture aplication behaviour better. During that course of discussion, Mel proposed several ideas to enhance current numa balancing. One of the suggestion was below Track what threads access a VMA. The suggestion was to use an unsigned long pid_mask and use the lower bits to tag approximately what threads access a VMA. Skip VMAs that did not trap a fault. This would be approximate because of PID collisions but would reduce scanning of areas the thread is not interested in. The above suggestion intends not to penalize threads that has no interest in the vma, thus reduce scanning overhead. V3 changes are mostly based on PeterZ comments (details below in changes) Summary of patchset: Current patchset implements: 1. Delay the vma scanning logic for newly created VMA's so that additional overhead of scanning is not incurred for short lived tasks (implementation by Mel) 2. Store the information of tasks accessing VMA in 2 windows. It is regularly cleared in (4*sysctl_numa_balancing_scan_delay) interval. The above time is derived from experimenting (Suggested by PeterZ) to balance between frequent clearing vs obsolete access data 3. hash_32 used to encode task index accessing VMA information 4. VMA's acess information is used to skip scanning for the tasks which had not accessed VMA Changes since V2: patch1: - Renaming of structure, macro to function, - Add explanation to heuristics - Adding more details from result (PeterZ) Patch2: - Usage of test and set bit (PeterZ) - Move storing access PID info to numa_migrate_prep() - Add a note on fainess among tasks allowed to scan (PeterZ) Patch3: - Maintain two windows of access PID information (PeterZ supported implementation and Gave idea to extend to N if needed) Patch4: - Apply hash_32 function to track VMA accessing PIDs (PeterZ) Changes since RFC V1: - Include Mel's vma scan delay patch - Change the accessing pid store logic (Thanks Mel) - Fencing structure / code to NUMA_BALANCING (David, Mel) - Adding clearing access PID logic (Mel) - Descriptive change log ( Mike Rapoport) Things to ponder over: ========================================== - Improvement to clearing accessing PIDs logic (discussed in-detail in patch3 itself (Done in this patchset by implementing 2 window history) - Current scan period is not changed in the patchset, so we do see frequent tries to scan. Relaxing scan period dynamically could improve results further. [1] sched/numa: Process Adaptive autoNUMA Link: https://lore.kernel.org/lkml/20220128052851.17162-1-bharata@amd.com/T/ [2] RFC V1 Link: https://lore.kernel.org/all/cover.1673610485.git.raghavendra.kt@amd.com/ [3] V2 Link: https://lore.kernel.org/lkml/cover.1675159422.git.raghavendra.kt@amd.com/ Results: Summary: Huge autonuma cost reduction seen in mmtest. Kernbench improvement is more than 5% and huge system time (80%+) improvement from mmtest autonuma. (dbench had huge std deviation to post) kernbench =========== 6.2.0-mmunstable-base 6.2.0-mmunstable-patched Amean user-256 22002.51 ( 0.00%) 22649.95 * -2.94%* Amean syst-256 10162.78 ( 0.00%) 8214.13 * 19.17%* Amean elsp-256 160.74 ( 0.00%) 156.92 * 2.38%* Duration User 66017.43 67959.84 Duration System 30503.15 24657.03 Duration Elapsed 504.61 493.12 6.2.0-mmunstable-base 6.2.0-mmunstable-patched Ops NUMA alloc hit 1738835089.00 1738780310.00 Ops NUMA alloc local 1738834448.00 1738779711.00 Ops NUMA base-page range updates 477310.00 392566.00 Ops NUMA PTE updates 477310.00 392566.00 Ops NUMA hint faults 96817.00 87555.00 Ops NUMA hint local faults % 10150.00 2192.00 Ops NUMA hint local percent 10.48 2.50 Ops NUMA pages migrated 86660.00 85363.00 Ops AutoNUMA cost 489.07 442.14 autonumabench =============== 6.2.0-mmunstable-base 6.2.0-mmunstable-patched Amean syst-NUMA01 399.50 ( 0.00%) 52.05 * 86.97%* Amean syst-NUMA01_THREADLOCAL 0.21 ( 0.00%) 0.22 * -5.41%* Amean syst-NUMA02 0.80 ( 0.00%) 0.78 * 2.68%* Amean syst-NUMA02_SMT 0.65 ( 0.00%) 0.68 * -3.95%* Amean elsp-NUMA01 313.26 ( 0.00%) 313.11 * 0.05%* Amean elsp-NUMA01_THREADLOCAL 1.06 ( 0.00%) 1.08 * -1.76%* Amean elsp-NUMA02 3.19 ( 0.00%) 3.24 * -1.52%* Amean elsp-NUMA02_SMT 3.72 ( 0.00%) 3.61 * 2.92%* Duration User 396433.47 324835.96 Duration System 2808.70 376.66 Duration Elapsed 2258.61 2258.12 6.2.0-mmunstable-base 6.2.0-mmunstable-patched Ops NUMA alloc hit 59921806.00 49623489.00 Ops NUMA alloc miss 0.00 0.00 Ops NUMA interleave hit 0.00 0.00 Ops NUMA alloc local 59920880.00 49622594.00 Ops NUMA base-page range updates 152259275.00 50075.00 Ops NUMA PTE updates 152259275.00 50075.00 Ops NUMA PMD updates 0.00 0.00 Ops NUMA hint faults 154660352.00 39014.00 Ops NUMA hint local faults % 138550501.00 23139.00 Ops NUMA hint local percent 89.58 59.31 Ops NUMA pages migrated 8179067.00 14147.00 Ops AutoNUMA cost 774522.98 195.69 This patch (of 4): Currently whenever a new task is created we wait for sysctl_numa_balancing_scan_delay to avoid unnessary scanning overhead. Extend the same logic to new or very short-lived VMAs. [raghavendra.kt@amd.com: add initialization in vm_area_dup())] Link: https://lkml.kernel.org/r/cover.1677672277.git.raghavendra.kt@amd.com Link: https://lkml.kernel.org/r/7a6fbba87c8b51e67efd3e74285bb4cb311a16ca.1677672277.git.raghavendra.kt@amd.comSigned-off-by:
-
