- 29 Jun, 2021 40 commits
-
-
Mel Gorman authored
This introduces a new sysctl vm.percpu_pagelist_high_fraction. It is similar to the old vm.percpu_pagelist_fraction. The old sysctl increased both pcp->batch and pcp->high with the higher pcp->high potentially reducing zone->lock contention. However, the higher pcp->batch value also potentially increased allocation latency while the PCP was refilled. This sysctl only adjusts pcp->high so that zone->lock contention is potentially reduced but allocation latency during a PCP refill remains the same. # grep -E "high:|batch" /proc/zoneinfo | tail -2 high: 649 batch: 63 # sysctl vm.percpu_pagelist_high_fraction=8 # grep -E "high:|batch" /proc/zoneinfo | tail -2 high: 35071 batch: 63 # sysctl vm.percpu_pagelist_high_fraction=64 high: 4383 batch: 63 # sysctl vm.percpu_pagelist_high_fraction=0 high: 649 batch: 63 [mgorman@techsingularity.net: fix documentation] Link: https://lkml.kernel.org/r/20210528151010.GQ30378@techsingularity.net Link: https://lkml.kernel.org/r/20210525080119.5455-7-mgorman@techsingularity.netSigned-off-by:Mel Gorman <mgorman@techsingularity.net> Acked-by:
Dave Hansen <dave.hansen@linux.intel.com> Acked-by:
Vlastimil Babka <vbabka@suse.cz> Cc: Hillf Danton <hdanton@sina.com> Cc: Michal Hocko <mhocko@kernel.org> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
Mel Gorman authored
When kswapd is active then direct reclaim is potentially active. In either case, it is possible that a zone would be balanced if pages were not trapped on PCP lists. Instead of draining remote pages, simply limit the size of the PCP lists while kswapd is active. Link: https://lkml.kernel.org/r/20210525080119.5455-6-mgorman@techsingularity.netSigned-off-by:
-
Mel Gorman authored
When a task is freeing a large number of order-0 pages, it may acquire the zone->lock multiple times freeing pages in batches. This may unnecessarily contend on the zone lock when freeing very large number of pages. This patch adapts the size of the batch based on the recent pattern to scale the batch size for subsequent frees. As the machines I used were not large enough to test this are not large enough to illustrate a problem, a debugging patch shows patterns like the following (slightly editted for clarity) Baseline vanilla kernel time-unmap-14426 [...] free_pcppages_bulk: free 63 count 378 high 378 time-unmap-14426 [...] free_pcppages_bulk: free 63 count 378 high 378 time-unmap-14426 [...] free_pcppages_bulk: free 63 count 378 high 378 time-unmap-14426 [...] free_pcppages_bulk: free 63 count 378 high 378 time-unmap-14426 [...] free_pcppages_bulk: free 63 count 378 high 378 With patches time-unmap-7724 [...] free_pcppages_bulk: free 126 count 814 high 814 time-unmap-7724 [...] free_pcppages_bulk: free 252 count 814 high 814 time-unmap-7724 [...] free_pcppages_bulk: free 504 count 814 high 814 time-unmap-7724 [...] free_pcppages_bulk: free 751 count 814 high 814 time-unmap-7724 [...] free_pcppages_bulk: free 751 count 814 high 814 Link: https://lkml.kernel.org/r/20210525080119.5455-5-mgorman@techsingularity.netSigned-off-by:
-
Mel Gorman authored
The PCP high watermark is based on the number of online CPUs so the watermarks must be adjusted during CPU hotplug. At the time of hot-remove, the number of online CPUs is already adjusted but during hot-add, a delta needs to be applied to update PCP to the correct value. After this patch is applied, the high watermarks are adjusted correctly. # grep high: /proc/zoneinfo | tail -1 high: 649 # echo 0 > /sys/devices/system/cpu/cpu4/online # grep high: /proc/zoneinfo | tail -1 high: 664 # echo 1 > /sys/devices/system/cpu/cpu4/online # grep high: /proc/zoneinfo | tail -1 high: 649 Link: https://lkml.kernel.org/r/20210525080119.5455-4-mgorman@techsingularity.netSigned-off-by: -
Mel Gorman authored
The pcp high watermark is based on the batch size but there is no relationship between them other than it is convenient to use early in boot. This patch takes the first step and bases pcp->high on the zone low watermark split across the number of CPUs local to a zone while the batch size remains the same to avoid increasing allocation latencies. The intent behind the default pcp->high is "set the number of PCP pages such that if they are all full that background reclaim is not started prematurely". Note that in this patch the pcp->high values are adjusted after memory hotplug events, min_free_kbytes adjustments and watermark scale factor adjustments but not CPU hotplug events which is handled later in the series. On a test KVM instance; Before grep -E "high:|batch" /proc/zoneinfo | tail -2 high: 378 batch: 63 After grep -E "high:|batch" /proc/zoneinfo | tail -2 high: 649 batch: 63 [mgorman@techsingularity.net: fix __setup_per_zone_wmarks for parallel memory hotplug] Link: https://lkml.kernel.org/r/20210528105925.GN30378@techsingularity.net Link: https://lkml.kernel.org/r/20210525080119.5455-3-mgorman@techsingularity.netSigned-off-by: -
Mel Gorman authored
Patch series "Calculate pcp->high based on zone sizes and active CPUs", v2. The per-cpu page allocator (PCP) is meant to reduce contention on the zone lock but the sizing of batch and high is archaic and neither takes the zone size into account or the number of CPUs local to a zone. With larger zones and more CPUs per node, the contention is getting worse. Furthermore, the fact that vm.percpu_pagelist_fraction adjusts both batch and high values means that the sysctl can reduce zone lock contention but also increase allocation latencies. This series disassociates pcp->high from pcp->batch and then scales pcp->high based on the size of the local zone with limited impact to reclaim and accounting for active CPUs but leaves pcp->batch static. It also adapts the number of pages that can be on the pcp list based on recent freeing patterns. The motivation is partially to adjust to larger memory sizes but is also driven by the fact that large batches of page freeing via release_pages() often shows zone contention as a major part of the problem. Another is a bug report based on an older kernel where a multi-terabyte process can takes several minutes to exit. A workaround was to use vm.percpu_pagelist_fraction to increase the pcp->high value but testing indicated that a production workload could not use the same values because of an increase in allocation latencies. Unfortunately, I cannot reproduce this test case myself as the multi-terabyte machines are in active use but it should alleviate the problem. The series aims to address both and partially acts as a pre-requisite. pcp only works with order-0 which is useless for SLUB (when using high orders) and THP (unconditionally). To store high-order pages on PCP, the pcp->high values need to be increased first. This patch (of 6): The vm.percpu_pagelist_fraction is used to increase the batch and high limits for the per-cpu page allocator (PCP). The intent behind the sysctl is to reduce zone lock acquisition when allocating/freeing pages but it has a problem. While it can decrease contention, it can also increase latency on the allocation side due to unreasonably large batch sizes. This leads to games where an administrator adjusts percpu_pagelist_fraction on the fly to work around contention and allocation latency problems. This series aims to alleviate the problems with zone lock contention while avoiding the allocation-side latency problems. For the purposes of review, it's easier to remove this sysctl now and reintroduce a similar sysctl later in the series that deals only with pcp->high. Link: https://lkml.kernel.org/r/20210525080119.5455-1-mgorman@techsingularity.net Link: https://lkml.kernel.org/r/20210525080119.5455-2-mgorman@techsingularity.netSigned-off-by:
-
Minchan Kim authored
alloc_contig_dump_pages() aims for helping debugging page migration failure by elevated page refcount compared to expected_count. (for the detail, please look at migrate_page_move_mapping) However, -ENOMEM is just the case that system is under memory pressure state, not relevant with page refcount at all. Thus, the dumping page list is not helpful for the debugging point of view. Link: https://lkml.kernel.org/r/YKa2Wyo9xqIErpfa@google.comSigned-off-by:
Minchan Kim <minchan@kernel.org> Reviewed-by:
David Hildenbrand <david@redhat.com> Cc: Suren Baghdasaryan <surenb@google.com> Cc: John Dias <joaodias@google.com> Cc: Michal Hocko <mhocko@suse.com> Signed-off-by:
-
Mel Gorman authored
VM events do not need explicit protection by disabling IRQs so update the counter with IRQs enabled in __free_pages_ok. Link: https://lkml.kernel.org/r/20210512095458.30632-10-mgorman@techsingularity.netSigned-off-by:
Peter Zijlstra (Intel) <peterz@infradead.org> Cc: Chuck Lever <chuck.lever@oracle.com> Cc: Ingo Molnar <mingo@kernel.org> Cc: Jesper Dangaard Brouer <brouer@redhat.com> Cc: Michal Hocko <mhocko@kernel.org> Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de> Cc: Thomas Gleixner <tglx@linutronix.de> Signed-off-by:
-
Mel Gorman authored
Historically when freeing pages, free_one_page() assumed that callers had IRQs disabled and the zone->lock could be acquired with spin_lock(). This confuses the scope of what local_lock_irq is protecting and what zone->lock is protecting in free_unref_page_list in particular. This patch uses spin_lock_irqsave() for the zone->lock in free_one_page() instead of relying on callers to have disabled IRQs. free_unref_page_commit() is changed to only deal with PCP pages protected by the local lock. free_unref_page_list() then first frees isolated pages to the buddy lists with free_one_page() and frees the rest of the pages to the PCP via free_unref_page_commit(). The end result is that free_one_page() is no longer depending on side-effects of local_lock to be correct. Note that this may incur a performance penalty while memory hot-remove is running but that is not a common operation. [lkp@intel.com: Ensure CMA pages get addded to correct pcp list] Link: https://lkml.kernel.org/r/20210512095458.30632-9-mgorman@techsingularity.netSigned-off-by:
-
Mel Gorman authored
__free_pages_ok() disables IRQs before calling a common helper free_one_page() that acquires the zone lock. This is not safe according to Documentation/locking/locktypes.rst and in this context, IRQ disabling is not protecting a per_cpu_pages structure either or a local_lock would be used. This patch explicitly acquires the lock with spin_lock_irqsave instead of relying on a helper. This removes the last instance of local_irq_save() in page_alloc.c. Link: https://lkml.kernel.org/r/20210512095458.30632-8-mgorman@techsingularity.netSigned-off-by:
-
Mel Gorman authored
IRQs are left disabled for the zone and node VM event counters. This is unnecessary as the affected counters are allowed to race for preemmption and IRQs. This patch reduces the scope of IRQs being disabled via local_[lock|unlock]_irq on !PREEMPT_RT kernels. One __mod_zone_freepage_state is still called with IRQs disabled. While this could be moved out, it's not free on all architectures as some require IRQs to be disabled for mod_zone_page_state on !PREEMPT_RT kernels. Link: https://lkml.kernel.org/r/20210512095458.30632-7-mgorman@techsingularity.netSigned-off-by:
-
Mel Gorman authored
Now that the zone_statistics are simple counters that do not require special protection, the bulk allocator accounting updates can be batch updated without adding too much complexity with protected RMW updates or using xchg. Link: https://lkml.kernel.org/r/20210512095458.30632-6-mgorman@techsingularity.netSigned-off-by:
-
Mel Gorman authored
__count_numa_event is small enough to be treated similarly to __count_vm_event so inline it. Link: https://lkml.kernel.org/r/20210512095458.30632-5-mgorman@techsingularity.netSigned-off-by:
-
Mel Gorman authored
NUMA statistics are maintained on the zone level for hits, misses, foreign etc but nothing relies on them being perfectly accurate for functional correctness. The counters are used by userspace to get a general overview of a workloads NUMA behaviour but the page allocator incurs a high cost to maintain perfect accuracy similar to what is required for a vmstat like NR_FREE_PAGES. There even is a sysctl vm.numa_stat to allow userspace to turn off the collection of NUMA statistics like NUMA_HIT. This patch converts NUMA_HIT and friends to be NUMA events with similar accuracy to VM events. There is a possibility that slight errors will be introduced but the overall trend as seen by userspace will be similar. The counters are no longer updated from vmstat_refresh context as it is unnecessary overhead for counters that may never be read by userspace. Note that counters could be maintained at the node level to save space but it would have a user-visible impact due to /proc/zoneinfo. [lkp@intel.com: Fix misplaced closing brace for !CONFIG_NUMA] Link: https://lkml.kernel.org/r/20210512095458.30632-4-mgorman@techsingularity.netSigned-off-by:
-
Mel Gorman authored
There is a lack of clarity of what exactly local_irq_save/local_irq_restore protects in page_alloc.c . It conflates the protection of per-cpu page allocation structures with per-cpu vmstat deltas. This patch protects the PCP structure using local_lock which for most configurations is identical to IRQ enabling/disabling. The scope of the lock is still wider than it should be but this is decreased later. It is possible for the local_lock to be embedded safely within struct per_cpu_pages but it adds complexity to free_unref_page_list. [akpm@linux-foundation.org: coding style fixes] [mgorman@techsingularity.net: work around a pahole limitation with zero-sized struct pagesets] Link: https://lkml.kernel.org/r/20210526080741.GW30378@techsingularity.net [lkp@intel.com: Make pagesets static] Link: https://lkml.kernel.org/r/20210512095458.30632-3-mgorman@techsingularity.netSigned-off-by:
-
Mel Gorman authored
The PCP (per-cpu page allocator in page_alloc.c) shares locking requirements with vmstat and the zone lock which is inconvenient and causes some issues. For example, the PCP list and vmstat share the same per-cpu space meaning that it's possible that vmstat updates dirty cache lines holding per-cpu lists across CPUs unless padding is used. Second, PREEMPT_RT does not want to disable IRQs for too long in the page allocator. This series splits the locking requirements and uses locks types more suitable for PREEMPT_RT, reduces the time when special locking is required for stats and reduces the time when IRQs need to be disabled on !PREEMPT_RT kernels. Why local_lock? PREEMPT_RT considers the following sequence to be unsafe as documented in Documentation/locking/locktypes.rst local_irq_disable(); spin_lock(&lock); The pcp allocator has this sequence for rmqueue_pcplist (local_irq_save) -> __rmqueue_pcplist -> rmqueue_bulk (spin_lock). While it's possible to separate this out, it generally means there are points where we enable IRQs and reenable them again immediately. To prevent a migration and the per-cpu pointer going stale, migrate_disable is also needed. That is a custom lock that is similar, but worse, than local_lock. Furthermore, on PREEMPT_RT, it's undesirable to leave IRQs disabled for too long. By converting to local_lock which disables migration on PREEMPT_RT, the locking requirements can be separated and start moving the protections for PCP, stats and the zone lock to PREEMPT_RT-safe equivalent locking. As a bonus, local_lock also means that PROVE_LOCKING does something useful. After that, it's obvious that zone_statistics incurs too much overhead and leaves IRQs disabled for longer than necessary on !PREEMPT_RT kernels. zone_statistics uses perfectly accurate counters requiring IRQs be disabled for parallel RMW sequences when inaccurate ones like vm_events would do. The series makes the NUMA statistics (NUMA_HIT and friends) inaccurate counters that then require no special protection on !PREEMPT_RT. The bulk page allocator can then do stat updates in bulk with IRQs enabled which should improve the efficiency. Technically, this could have been done without the local_lock and vmstat conversion work and the order simply reflects the timing of when different series were implemented. Finally, there are places where we conflate IRQs being disabled for the PCP with the IRQ-safe zone spinlock. The remainder of the series reduces the scope of what is protected by disabled IRQs on !PREEMPT_RT kernels. By the end of the series, page_alloc.c does not call local_irq_save so the locking scope is a bit clearer. The one exception is that modifying NR_FREE_PAGES still happens in places where it's known the IRQs are disabled as it's harmless for PREEMPT_RT and would be expensive to split the locking there. No performance data is included because despite the overhead of the stats, it's within the noise for most workloads on !PREEMPT_RT. However, Jesper Dangaard Brouer ran a page allocation microbenchmark on a E5-1650 v4 @ 3.60GHz CPU on the first version of this series. Focusing on the array variant of the bulk page allocator reveals the following. (CPU: Intel(R) Xeon(R) CPU E5-1650 v4 @ 3.60GHz) ARRAY variant: time_bulk_page_alloc_free_array: step=bulk size Baseline Patched 1 56.383 54.225 (+3.83%) 2 40.047 35.492 (+11.38%) 3 37.339 32.643 (+12.58%) 4 35.578 30.992 (+12.89%) 8 33.592 29.606 (+11.87%) 16 32.362 28.532 (+11.85%) 32 31.476 27.728 (+11.91%) 64 30.633 27.252 (+11.04%) 128 30.596 27.090 (+11.46%) While this is a positive outcome, the series is more likely to be interesting to the RT people in terms of getting parts of the PREEMPT_RT tree into mainline. This patch (of 9): The per-cpu page allocator lists and the per-cpu vmstat deltas are stored in the same struct per_cpu_pages even though vmstats have no direct impact on the per-cpu page lists. This is inconsistent because the vmstats for a node are stored on a dedicated structure. The bigger issue is that the per_cpu_pages structure is not cache-aligned and stat updates either cache conflict with adjacent per-cpu lists incurring a runtime cost or padding is required incurring a memory cost. This patch splits the per-cpu pagelists and the vmstat deltas into separate structures. It's mostly a mechanical conversion but some variable renaming is done to clearly distinguish the per-cpu pages structure (pcp) from the vmstats (pzstats). Superficially, this appears to increase the size of the per_cpu_pages structure but the movement of expire fills a structure hole so there is no impact overall. [mgorman@techsingularity.net: make it W=1 cleaner] Link: https://lkml.kernel.org/r/20210514144622.GA3735@techsingularity.net [mgorman@techsingularity.net: make it W=1 even cleaner] Link: https://lkml.kernel.org/r/20210516140705.GB3735@techsingularity.net [lkp@intel.com: check struct per_cpu_zonestat has a non-zero size] [vbabka@suse.cz: Init zone->per_cpu_zonestats properly] Link: https://lkml.kernel.org/r/20210512095458.30632-1-mgorman@techsingularity.net Link: https://lkml.kernel.org/r/20210512095458.30632-2-mgorman@techsingularity.netSigned-off-by: -
Andrii Nakryiko authored
Commit "mm/page_alloc: convert per-cpu list protection to local_lock" will introduce a zero-sized per-CPU variable, which causes pahole to generate invalid BTF. Only pahole versions 1.18 through 1.21 are impacted, as before 1.18 pahole doesn't know anything about per-CPU variables, and 1.22 contains the proper fix for the issue. Luckily, pahole 1.18 got --skip_encoding_btf_vars option disabling BTF generation for per-CPU variables in anticipation of some unanticipated problems. So use this escape hatch to disable per-CPU var BTF info on those problematic pahole versions. Users relying on availability of per-CPU var BTFs would need to upgrade to pahole 1.22+, but everyone won't notice any regressions. Link: https://lkml.kernel.org/r/20210530002536.3193829-1-andrii@kernel.orgSigned-off-by:
Andrii Nakryiko <andrii@kernel.org> Acked-by:
-
Heiner Kallweit authored
Having such debug messages in the dmesg log may confuse users. Therefore restrict debug output to cases where DEBUG is defined or dynamic debugging is enabled for the respective code piece. Link: https://lkml.kernel.org/r/976adb93-3041-ce63-48fc-55a6096a51c1@gmail.comSigned-off-by:
Heiner Kallweit <hkallweit1@gmail.com> Signed-off-by:
-
Matthew Wilcox (Oracle) authored
If the memmap is virtually contiguous (either because we're using a virtually mapped memmap or because we don't support a discontig memmap at all), then we can implement nth_page() by simple addition. Contrary to popular belief, the compiler is not able to optimise this itself for a vmemmap configuration. This reduces one example user (sg.c) by four instructions: struct page *page = nth_page(rsv_schp->pages[k], offset >> PAGE_SHIFT); before: 49 8b 45 70 mov 0x70(%r13),%rax 48 63 c9 movslq %ecx,%rcx 48 c1 eb 0c shr $0xc,%rbx 48 8b 04 c8 mov (%rax,%rcx,8),%rax 48 2b 05 00 00 00 00 sub 0x0(%rip),%rax R_X86_64_PC32 vmemmap_base-0x4 48 c1 f8 06 sar $0x6,%rax 48 01 d8 add %rbx,%rax 48 c1 e0 06 shl $0x6,%rax 48 03 05 00 00 00 00 add 0x0(%rip),%rax R_X86_64_PC32 vmemmap_base-0x4 after: 49 8b 45 70 mov 0x70(%r13),%rax 48 63 c9 movslq %ecx,%rcx 48 c1 eb 0c shr $0xc,%rbx 48 c1 e3 06 shl $0x6,%rbx 48 03 1c c8 add (%rax,%rcx,8),%rbx Link: https://lkml.kernel.org/r/20210413194625.1472345-1-willy@infradead.orgSigned-off-by:Matthew Wilcox (Oracle) <willy@infradead.org> Reviewed-by:
Christoph Hellwig <hch@lst.de> Reviewed-by:
Zi Yan <ziy@nvidia.com> Reviewed-by:
-
Matthew Wilcox (Oracle) authored
Now that compound_head() accepts a const struct page pointer, these two functions can be marked as not modifying the page pointer they are passed. Link: https://lkml.kernel.org/r/20210416231531.2521383-7-willy@infradead.orgSigned-off-by:
Anshuman Khandual <anshuman.khandual@arm.com> Reviewed-by:
William Kucharski <william.kucharski@oracle.com> Signed-off-by:
-
Matthew Wilcox (Oracle) authored
The struct page is not modified by these routines, so it can be marked const. Link: https://lkml.kernel.org/r/20210416231531.2521383-6-willy@infradead.orgSigned-off-by:
-
Matthew Wilcox (Oracle) authored
If you pass a const pointer to compound_head(), you get a const pointer back; if you pass a mutable pointer, you get a mutable pointer back. Also remove an unnecessary forward definition of struct page; we're about to dereference page->compound_head, so it must already have been defined. Link: https://lkml.kernel.org/r/20210416231531.2521383-5-willy@infradead.orgSigned-off-by:
-
Matthew Wilcox (Oracle) authored
dump_page_owner() only uses struct page to find the page_ext, and lookup_page_ext() already takes a const argument. Link: https://lkml.kernel.org/r/20210416231531.2521383-4-willy@infradead.orgSigned-off-by:
-
Matthew Wilcox (Oracle) authored
Move the PagePoisoned test into dump_page(). Skip the hex print for poisoned pages -- we know they're full of ffffffff. Move the reason printing from __dump_page() to dump_page(). Link: https://lkml.kernel.org/r/20210416231531.2521383-3-willy@infradead.orgSigned-off-by:
-
Aaron Tomlin authored
A customer experienced a low-memory situation and decided to issue a SIGKILL (i.e. a fatal signal). Instead of promptly terminating as one would expect, the aforementioned task remained unresponsive. Further investigation indicated that the task was "stuck" in the reclaim/compaction retry loop. Now, it does not make sense to retry compaction when a fatal signal is pending. In the context of try_to_compact_pages(), indeed COMPACT_SKIPPED can be returned; albeit, not every zone, on the zone list, would be considered in the case a fatal signal is found to be pending. Yet, in should_compact_retry(), given the last known compaction result, each zone, on the zone list, can be considered/or checked (see compaction_zonelist_suitable()). For example, if a zone was found to succeed, then reclaim/compaction would be tried again (notwithstanding the above). This patch ensures that compaction is not needlessly retried irrespective of the last known compaction result e.g. if it was skipped, in the unlikely case a fatal signal is found pending. So, OOM is at least attempted. Link: https://lkml.kernel.org/r/20210520142901.3371299-1-atomlin@redhat.comSigned-off-by:
Aaron Tomlin <atomlin@redhat.com> Reviewed-by:
-
Matthew Wilcox (Oracle) authored
Patch series "Constify struct page arguments". While working on various solutions to the 32-bit struct page size regression, one of the problems I found was the networking stack expects to be able to pass const struct page pointers around, and the mm doesn't provide a lot of const-friendly functions to call. The root tangle of problems is that a lot of functions call VM_BUG_ON_PAGE(), which calls dump_page(), which calls a lot of functions which don't take a const struct page (but could be const). This patch (of 6): The only caller of __dump_page() now opencodes dump_page(), so remove it as an externally visible symbol. Link: https://lkml.kernel.org/r/20210416231531.2521383-1-willy@infradead.org Link: https://lkml.kernel.org/r/20210416231531.2521383-2-willy@infradead.orgSigned-off-by:
-
Mike Rapoport authored
There is a lot of historical ifdefery in is_highmem_idx() and its helper zone_movable_is_highmem() that was required because of two different paths for nodes and zones initialization that were selected at compile time. Until commit 3f08a302 ("mm: remove CONFIG_HAVE_MEMBLOCK_NODE_MAP option") the movable_zone variable was only available for configurations that had CONFIG_HAVE_MEMBLOCK_NODE_MAP enabled so the test in zone_movable_is_highmem() used that variable only for such configurations. For other configurations the test checked if the index of ZONE_MOVABLE was greater by 1 than the index of ZONE_HIGMEM and then movable zone was considered a highmem zone. Needless to say, ZONE_MOVABLE - 1 equals ZONE_HIGHMEM by definition when CONFIG_HIGHMEM=y. Commit 3f08a302 ("mm: remove CONFIG_HAVE_MEMBLOCK_NODE_MAP option") made movable_zone variable always available. Since this variable is set to ZONE_HIGHMEM if CONFIG_HIGHMEM is enabled and highmem zone is populated, it is enough to check whether zone_idx == ZONE_MOVABLE && movable_zone == ZONE_HIGMEM to test if zone index points to a highmem zone. Remove zone_movable_is_highmem() that is not used anywhere except is_highmem_idx() and use the test above in is_highmem_idx() instead. Link: https://lkml.kernel.org/r/20210426141927.1314326-3-rppt@kernel.orgSigned-off-by:
Mike Rapoport <rppt@linux.ibm.com> Reviewed-by:
-
Jungseung Lee authored
Add the details for figuring out which parts of the kernel image is being freed on initmem case. Before: Freeing unused kernel memory: 1024K After: Freeing unused kernel image (initmem) memory: 1024K Link: https://lkml.kernel.org/r/1622706274-4533-1-git-send-email-js07.lee@samsung.comSigned-off-by:
Jungseung Lee <js07.lee@samsung.com> Reviewed-by:
-
Kuan-Ying Lee authored
Add memory corruption identification support for hardware tag-based mode. We store one old free pointer tag and free backtrace instead of five because hardware tag-based kasan only has 16 different tags. If we store as many stacks as SW tag-based kasan does(5 stacks), there is high probability to find the same tag in the stacks when out-of-bound issues happened and we will mistake out-of-bound issue for use-after-free. Link: https://lkml.kernel.org/r/20210626100931.22794-4-Kuan-Ying.Lee@mediatek.comSigned-off-by:
Kuan-Ying Lee <Kuan-Ying.Lee@mediatek.com> Suggested-by:
Marco Elver <elver@google.com> Reviewed-by:
Alexander Potapenko <glider@google.com> Reviewed-by:
Andrey Konovalov <andreyknvl@gmail.com> Reviewed-by:
-
Kuan-Ying Lee authored
1. Move kasan_get_free_track() and kasan_set_free_info() into tags.c and combine these two functions for SW_TAGS and HW_TAGS kasan mode. 2. Move kasan_get_bug_type() to report_tags.c and make this function compatible for SW_TAGS and HW_TAGS kasan mode. Link: https://lkml.kernel.org/r/20210626100931.22794-3-Kuan-Ying.Lee@mediatek.comSigned-off-by:
Greg Kroah-Hartman <gregkh@linuxfoundation.org> Reviewed-by:
-
Kuan-Ying Lee authored
Patch series "kasan: add memory corruption identification support for hw tag-based kasan", v4. Add memory corruption identification for hardware tag-based KASAN mode. This patch (of 3): Rename CONFIG_KASAN_SW_TAGS_IDENTIFY to CONFIG_KASAN_TAGS_IDENTIFY in order to be compatible with hardware tag-based mode. Link: https://lkml.kernel.org/r/20210626100931.22794-1-Kuan-Ying.Lee@mediatek.com Link: https://lkml.kernel.org/r/20210626100931.22794-2-Kuan-Ying.Lee@mediatek.comSigned-off-by:
-
Daniel Axtens authored
powerpc has a variable number of PTRS_PER_*, set at runtime based on the MMU that the kernel is booted under. This means the PTRS_PER_* are no longer constants, and therefore breaks the build. Switch to using MAX_PTRS_PER_*, which are constant. Link: https://lkml.kernel.org/r/20210624034050.511391-5-dja@axtens.netSigned-off-by:
Daniel Axtens <dja@axtens.net> Suggested-by:
Christophe Leroy <christophe.leroy@csgroup.eu> Suggested-by:
Balbir Singh <bsingharora@gmail.com> Reviewed-by:
-
Daniel Axtens authored
Commit c65e774f ("x86/mm: Make PGDIR_SHIFT and PTRS_PER_P4D variable") made PTRS_PER_P4D variable on x86 and introduced MAX_PTRS_PER_P4D as a constant for cases which need a compile-time constant (e.g. fixed-size arrays). powerpc likewise has boot-time selectable MMU features which can cause other mm "constants" to vary. For KASAN, we have some static PTE/PMD/PUD/P4D arrays so we need compile-time maximums for all these constants. Extend the MAX_PTRS_PER_ idiom, and place default definitions in include/pgtable.h. These define MAX_PTRS_PER_x to be PTRS_PER_x unless an architecture has defined MAX_PTRS_PER_x in its arch headers. Clean up pgtable-nop4d.h and s390's MAX_PTRS_PER_P4D definitions while we're at it: both can just pick up the default now. Link: https://lkml.kernel.org/r/20210624034050.511391-4-dja@axtens.netSigned-off-by:
-
Daniel Axtens authored
Allow architectures to define a kasan_arch_is_ready() hook that bails out of any function that's about to touch the shadow unless the arch says that it is ready for the memory to be accessed. This is fairly uninvasive and should have a negligible performance penalty. This will only work in outline mode, so an arch must specify ARCH_DISABLE_KASAN_INLINE if it requires this. Link: https://lkml.kernel.org/r/20210624034050.511391-3-dja@axtens.netSigned-off-by:
-
Daniel Axtens authored
Patch series "KASAN core changes for ppc64 radix KASAN", v16. Building on the work of Christophe, Aneesh and Balbir, I've ported KASAN to 64-bit Book3S kernels running on the Radix MMU. I've been trying this for a while, but we keep having collisions between the kasan code in the mm tree and the code I want to put in to the ppc tree. This series just contains the kasan core changes that we need. There should be no noticeable changes to other platforms. This patch (of 4): For annoying architectural reasons, it's very difficult to support inline instrumentation on powerpc64.* Add a Kconfig flag to allow an arch to disable inline. (It's a bit annoying to be 'backwards', but I'm not aware of any way to have an arch force a symbol to be 'n', rather than 'y'.) We also disable stack instrumentation in this case as it does things that are functionally equivalent to inline instrumentation, namely adding code that touches the shadow directly without going through a C helper. * on ppc64 atm, the shadow lives in virtual memory and isn't accessible in real mode. However, before we turn on virtual memory, we parse the device tree to determine which platform and MMU we're running under. That calls generic DT code, which is instrumented. Inline instrumentation in DT would unconditionally attempt to touch the shadow region, which we won't have set up yet, and would crash. We can make outline mode wait for the arch to be ready, but we can't change what the compiler inserts for inline mode. Link: https://lkml.kernel.org/r/20210624034050.511391-1-dja@axtens.net Link: https://lkml.kernel.org/r/20210624034050.511391-2-dja@axtens.netSigned-off-by:
-
David Gow authored
The KUNIT_EXPECT_KASAN_FAIL() macro currently uses KUNIT_EXPECT_EQ() to compare fail_data.report_expected and fail_data.report_found. This always gave a somewhat useless error message on failure, but the addition of extra compile-time checking with READ_ONCE() has caused it to get much longer, and be truncated before anything useful is displayed. Instead, just check fail_data.report_found by hand (we've just set report_expected to 'true'), and print a better failure message with KUNIT_FAIL(). Because of this, report_expected is no longer used anywhere, and can be removed. Beforehand, a failure in: KUNIT_EXPECT_KASAN_FAIL(test, ((volatile char *)area)[3100]); would have looked like: [22:00:34] [FAILED] vmalloc_oob [22:00:34] # vmalloc_oob: EXPECTATION FAILED at lib/test_kasan.c:991 [22:00:34] Expected ({ do { extern void __compiletime_assert_705(void) __attribute__((__error__("Unsupported access size for {READ,WRITE}_ONCE()."))); if (!((sizeof(fail_data.report_expected) == sizeof(char) || sizeof(fail_data.repp [22:00:34] not ok 45 - vmalloc_oob With this change, it instead looks like: [22:04:04] [FAILED] vmalloc_oob [22:04:04] # vmalloc_oob: EXPECTATION FAILED at lib/test_kasan.c:993 [22:04:04] KASAN failure expected in "((volatile char *)area)[3100]", but none occurred [22:04:04] not ok 45 - vmalloc_oob Also update the example failure in the documentation to reflect this. Link: https://lkml.kernel.org/r/20210606005531.165954-1-davidgow@google.comSigned-off-by:David Gow <davidgow@google.com> Reviewed-by:
Brendan Higgins <brendanhiggins@google.com> Cc: Andrey Ryabinin <ryabinin.a.a@gmail.com> Cc: Dmitry Vyukov <dvyukov@google.com> Cc: Daniel Axtens <dja@axtens.net> Cc: David Gow <davidgow@google.com> Cc: Jonathan Corbet <corbet@lwn.net> Signed-off-by:
-
Alexander Potapenko authored
Most of the contents of KASAN reports are printed with pr_err(), so use a consistent logging level to print the memory access stacks. Link: https://lkml.kernel.org/r/20210506105405.3535023-2-glider@google.comSigned-off-by:
-
Alexander Potapenko authored
dump_stack() is used for many different cases, which may require a log level consistent with other kernel messages surrounding the dump_stack() call. Without that, certain systems that are configured to ignore the default level messages will miss stack traces in critical error reports. This patch introduces dump_stack_lvl() that behaves similarly to dump_stack(), but accepts a custom log level. The old dump_stack() becomes equal to dump_stack_lvl(KERN_DEFAULT). A somewhat similar patch has been proposed in 2012: https://lore.kernel.org/lkml/1332493269.2359.9.camel@hebo/ , but wasn't merged. [elver@google.com: add missing dump_stack_lvl() stub if CONFIG_PRINTK=n] Link: https://lkml.kernel.org/r/YJ0KAM0hQev1AmWe@elver.google.com Link: https://lkml.kernel.org/r/20210506105405.3535023-1-glider@google.comSigned-off-by:
-
Rafael Aquini authored
On non-preemptible kernel builds the watchdog can complain about soft lockups when vfree() is called against large vmalloc areas: [ 210.851798] kvmalloc-test: vmalloc(2199023255552) succeeded [ 238.654842] watchdog: BUG: soft lockup - CPU#181 stuck for 26s! [rmmod:5203] [ 238.662716] Modules linked in: kvmalloc_test(OE-) ... [ 238.772671] CPU: 181 PID: 5203 Comm: rmmod Tainted: G S OE 5.13.0-rc7+ #1 [ 238.781413] Hardware name: Intel Corporation PURLEY/PURLEY, BIOS PLYXCRB1.86B.0553.D01.1809190614 09/19/2018 [ 238.792383] RIP: 0010:free_unref_page+0x52/0x60 [ 238.797447] Code: 48 c1 fd 06 48 89 ee e8 9c d0 ff ff 84 c0 74 19 9c 41 5c fa 48 89 ee 48 89 df e8 b9 ea ff ff 41 f7 c4 00 02 00 00 74 01 fb 5b <5d> 41 5c c3 66 2e 0f 1f 84 00 00 00 00 00 0f 1f 44 00 00 f0 29 77 [ 238.818406] RSP: 0018:ffffb4d87868fe98 EFLAGS: 00000206 [ 238.824236] RAX: 0000000000000000 RBX: 000000001da0c945 RCX: ffffb4d87868fe40 [ 238.832200] RDX: ffffd79d3beed108 RSI: ffffd7998501dc08 RDI: ffff9c6fbffd7010 [ 238.840166] RBP: 000000000d518cbd R08: ffffd7998501dc08 R09: 0000000000000001 [ 238.848131] R10: 0000000000000000 R11: ffffd79d3beee088 R12: 0000000000000202 [ 238.856095] R13: ffff9e5be3eceec0 R14: 0000000000000000 R15: 0000000000000000 [ 238.864059] FS: 00007fe082c2d740(0000) GS:ffff9f4c69b40000(0000) knlGS:0000000000000000 [ 238.873089] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 238.879503] CR2: 000055a000611128 CR3: 000000f6094f6006 CR4: 00000000007706e0 [ 238.887467] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [ 238.895433] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 [ 238.903397] PKRU: 55555554 [ 238.906417] Call Trace: [ 238.909149] __vunmap+0x17c/0x220 [ 238.912851] __x64_sys_delete_module+0x13a/0x250 [ 238.918008] ? syscall_trace_enter.isra.20+0x13c/0x1b0 [ 238.923746] do_syscall_64+0x39/0x80 [ 238.927740] entry_SYSCALL_64_after_hwframe+0x44/0xae Like in other range zapping routines that iterate over a large list, lets just add cond_resched() within __vunmap()'s page-releasing loop in order to avoid the watchdog splats. Link: https://lkml.kernel.org/r/20210622225030.478384-1-aquini@redhat.comSigned-off-by:
Rafael Aquini <aquini@redhat.com> Acked-by:
Nicholas Piggin <npiggin@gmail.com> Reviewed-by:
Uladzislau Rezki (Sony) <urezki@gmail.com> Reviewed-by:
Michal Hocko <mhocko@suse.com> Signed-off-by:
-
Uladzislau Rezki authored
Currently for order-0 pages we use a bulk-page allocator to get set of pages. From the other hand not allocating all pages is something that might occur. In that case we should fallbak to the single-page allocator trying to get missing pages, because it is more permissive(direct reclaim, etc). Introduce a vm_area_alloc_pages() function where the described logic is implemented. Link: https://lkml.kernel.org/r/20210521130718.GA17882@pc638.lanSigned-off-by:
-
