BMC/Intel-BMC/linux.git/mm, branch dev-5.15-intel

mm/mmu_notifier.c: fix race in mmu_interval_notifier_remove()

2022-04-27T12:38:58+00:00

commit 319561669a59d8e9206ab311ae5433ef92fd79d1 upstream. In some cases it is possible for mmu_interval_notifier_remove() to race with mn_tree_inv_end() allowing it to return while the notifier data structure is still in use. Consider the following sequence: CPU0 - mn_tree_inv_end() CPU1 - mmu_interval_notifier_remove() ----------------------------------- ------------------------------------ spin_lock(subscriptions->lock); seq = subscriptions->invalidate_seq; spin_lock(subscriptions->lock); spin_unlock(subscriptions->lock); subscriptions->invalidate_seq++; wait_event(invalidate_seq != seq); return; interval_tree_remove(interval_sub); kfree(interval_sub); spin_unlock(subscriptions->lock); wake_up_all(); As the wait_event() condition is true it will return immediately. This can lead to use-after-free type errors if the caller frees the data structure containing the interval notifier subscription while it is still on a deferred list. Fix this by taking the appropriate lock when reading invalidate_seq to ensure proper synchronisation. I observed this whilst running stress testing during some development. You do have to be pretty unlucky, but it leads to the usual problems of use-after-free (memory corruption, kernel crash, difficult to diagnose WARN_ON, etc). Link: https://lkml.kernel.org/r/20220420043734.476348-1-apopple@nvidia.com Fixes: 99cb252f5e68 ("mm/mmu_notifier: add an interval tree notifier") Signed-off-by: Alistair Popple Signed-off-by: Jason Gunthorpe Cc: Christian König Cc: John Hubbard Cc: Ralph Campbell Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds Signed-off-by: Greg Kroah-Hartman

oom_kill.c: futex: delay the OOM reaper to allow time for proper futex cleanup

2022-04-27T12:38:58+00:00

commit e4a38402c36e42df28eb1a5394be87e6571fb48a upstream. The pthread struct is allocated on PRIVATE|ANONYMOUS memory [1] which can be targeted by the oom reaper. This mapping is used to store the futex robust list head; the kernel does not keep a copy of the robust list and instead references a userspace address to maintain the robustness during a process death. A race can occur between exit_mm and the oom reaper that allows the oom reaper to free the memory of the futex robust list before the exit path has handled the futex death: CPU1 CPU2 -------------------------------------------------------------------- page_fault do_exit "signal" wake_oom_reaper oom_reaper oom_reap_task_mm (invalidates mm) exit_mm exit_mm_release futex_exit_release futex_cleanup exit_robust_list get_user (EFAULT- can't access memory) If the get_user EFAULT's, the kernel will be unable to recover the waiters on the robust_list, leaving userspace mutexes hung indefinitely. Delay the OOM reaper, allowing more time for the exit path to perform the futex cleanup. Reproducer: https://gitlab.com/jsavitz/oom_futex_reproducer Based on a patch by Michal Hocko. Link: https://elixir.bootlin.com/glibc/glibc-2.35/source/nptl/allocatestack.c#L370 [1] Link: https://lkml.kernel.org/r/20220414144042.677008-1-npache@redhat.com Fixes: 212925802454 ("mm: oom: let oom_reap_task and exit_mmap run concurrently") Signed-off-by: Joel Savitz Signed-off-by: Nico Pache Co-developed-by: Joel Savitz Suggested-by: Thomas Gleixner Acked-by: Thomas Gleixner Acked-by: Michal Hocko Cc: Rafael Aquini Cc: Waiman Long Cc: Herton R. Krzesinski Cc: Juri Lelli Cc: Vincent Guittot Cc: Dietmar Eggemann Cc: Steven Rostedt Cc: Ben Segall Cc: Mel Gorman Cc: Daniel Bristot de Oliveira Cc: David Rientjes Cc: Andrea Arcangeli Cc: Davidlohr Bueso Cc: Peter Zijlstra Cc: Ingo Molnar Cc: Joel Savitz Cc: Darren Hart Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds Signed-off-by: Greg Kroah-Hartman

mm, hugetlb: allow for "high" userspace addresses

2022-04-27T12:38:57+00:00

commit 5f24d5a579d1eace79d505b148808a850b417d4c upstream. This is a fix for commit f6795053dac8 ("mm: mmap: Allow for "high" userspace addresses") for hugetlb. This patch adds support for "high" userspace addresses that are optionally supported on the system and have to be requested via a hint mechanism ("high" addr parameter to mmap). Architectures such as powerpc and x86 achieve this by making changes to their architectural versions of hugetlb_get_unmapped_area() function. However, arm64 uses the generic version of that function. So take into account arch_get_mmap_base() and arch_get_mmap_end() in hugetlb_get_unmapped_area(). To allow that, move those two macros out of mm/mmap.c into include/linux/sched/mm.h If these macros are not defined in architectural code then they default to (TASK_SIZE) and (base) so should not introduce any behavioural changes to architectures that do not define them. For the time being, only ARM64 is affected by this change. Catalin (ARM64) said "We should have fixed hugetlb_get_unmapped_area() as well when we added support for 52-bit VA. The reason for commit f6795053dac8 was to prevent normal mmap() from returning addresses above 48-bit by default as some user-space had hard assumptions about this. It's a slight ABI change if you do this for hugetlb_get_unmapped_area() but I doubt anyone would notice. It's more likely that the current behaviour would cause issues, so I'd rather have them consistent. Basically when arm64 gained support for 52-bit addresses we did not want user-space calling mmap() to suddenly get such high addresses, otherwise we could have inadvertently broken some programs (similar behaviour to x86 here). Hence we added commit f6795053dac8. But we missed hugetlbfs which could still get such high mmap() addresses. So in theory that's a potential regression that should have bee addressed at the same time as commit f6795053dac8 (and before arm64 enabled 52-bit addresses)" Link: https://lkml.kernel.org/r/ab847b6edb197bffdfe189e70fb4ac76bfe79e0d.1650033747.git.christophe.leroy@csgroup.eu Fixes: f6795053dac8 ("mm: mmap: Allow for "high" userspace addresses") Signed-off-by: Christophe Leroy Reviewed-by: Catalin Marinas Cc: Steve Capper Cc: Will Deacon Cc: [5.0.x] Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds Signed-off-by: Greg Kroah-Hartman

memcg: sync flush only if periodic flush is delayed

2022-04-27T12:38:57+00:00

commit 9b3016154c913b2e7ec5ae5c9a42eb9e732d86aa upstream. Daniel Dao has reported [1] a regression on workloads that may trigger a lot of refaults (anon and file). The underlying issue is that flushing rstat is expensive. Although rstat flush are batched with (nr_cpus * MEMCG_BATCH) stat updates, it seems like there are workloads which genuinely do stat updates larger than batch value within short amount of time. Since the rstat flush can happen in the performance critical codepaths like page faults, such workload can suffer greatly. This patch fixes this regression by making the rstat flushing conditional in the performance critical codepaths. More specifically, the kernel relies on the async periodic rstat flusher to flush the stats and only if the periodic flusher is delayed by more than twice the amount of its normal time window then the kernel allows rstat flushing from the performance critical codepaths. Now the question: what are the side-effects of this change? The worst that can happen is the refault codepath will see 4sec old lruvec stats and may cause false (or missed) activations of the refaulted page which may under-or-overestimate the workingset size. Though that is not very concerning as the kernel can already miss or do false activations. There are two more codepaths whose flushing behavior is not changed by this patch and we may need to come to them in future. One is the writeback stats used by dirty throttling and second is the deactivation heuristic in the reclaim. For now keeping an eye on them and if there is report of regression due to these codepaths, we will reevaluate then. Link: https://lore.kernel.org/all/CA+wXwBSyO87ZX5PVwdHm-=dBjZYECGmfnydUicUyrQqndgX2MQ@mail.gmail.com [1] Link: https://lkml.kernel.org/r/20220304184040.1304781-1-shakeelb@google.com Fixes: 1f828223b799 ("memcg: flush lruvec stats in the refault") Signed-off-by: Shakeel Butt Reported-by: Daniel Dao Tested-by: Ivan Babrou Cc: Michal Hocko Cc: Roman Gushchin Cc: Johannes Weiner Cc: Michal Koutný Cc: Frank Hofmann Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds Signed-off-by: Greg Kroah-Hartman

mm/memory-failure.c: skip huge_zero_page in memory_failure()

2022-04-27T12:38:57+00:00

commit d173d5417fb67411e623d394aab986d847e47dad upstream. Kernel panic when injecting memory_failure for the global huge_zero_page, when CONFIG_DEBUG_VM is enabled, as follows. Injecting memory failure for pfn 0x109ff9 at process virtual address 0x20ff9000 page:00000000fb053fc3 refcount:2 mapcount:0 mapping:0000000000000000 index:0x0 pfn:0x109e00 head:00000000fb053fc3 order:9 compound_mapcount:0 compound_pincount:0 flags: 0x17fffc000010001(locked|head|node=0|zone=2|lastcpupid=0x1ffff) raw: 017fffc000010001 0000000000000000 dead000000000122 0000000000000000 raw: 0000000000000000 0000000000000000 00000002ffffffff 0000000000000000 page dumped because: VM_BUG_ON_PAGE(is_huge_zero_page(head)) ------------[ cut here ]------------ kernel BUG at mm/huge_memory.c:2499! invalid opcode: 0000 [#1] PREEMPT SMP PTI CPU: 6 PID: 553 Comm: split_bug Not tainted 5.18.0-rc1+ #11 Hardware name: Alibaba Cloud Alibaba Cloud ECS, BIOS 3288b3c 04/01/2014 RIP: 0010:split_huge_page_to_list+0x66a/0x880 Code: 84 9b fb ff ff 48 8b 7c 24 08 31 f6 e8 9f 5d 2a 00 b8 b8 02 00 00 e9 e8 fb ff ff 48 c7 c6 e8 47 3c 82 4c b RSP: 0018:ffffc90000dcbdf8 EFLAGS: 00010246 RAX: 000000000000003c RBX: 0000000000000001 RCX: 0000000000000000 RDX: 0000000000000000 RSI: ffffffff823e4c4f RDI: 00000000ffffffff RBP: ffff88843fffdb40 R08: 0000000000000000 R09: 00000000fffeffff R10: ffffc90000dcbc48 R11: ffffffff82d68448 R12: ffffea0004278000 R13: ffffffff823c6203 R14: 0000000000109ff9 R15: ffffea000427fe40 FS: 00007fc375a26740(0000) GS:ffff88842fd80000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 00007fc3757c9290 CR3: 0000000102174006 CR4: 00000000003706e0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: try_to_split_thp_page+0x3a/0x130 memory_failure+0x128/0x800 madvise_inject_error.cold+0x8b/0xa1 __x64_sys_madvise+0x54/0x60 do_syscall_64+0x35/0x80 entry_SYSCALL_64_after_hwframe+0x44/0xae RIP: 0033:0x7fc3754f8bf9 Code: 01 00 48 81 c4 80 00 00 00 e9 f1 fe ff ff 0f 1f 00 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 8 RSP: 002b:00007ffeda93a1d8 EFLAGS: 00000217 ORIG_RAX: 000000000000001c RAX: ffffffffffffffda RBX: 0000000000000000 RCX: 00007fc3754f8bf9 RDX: 0000000000000064 RSI: 0000000000003000 RDI: 0000000020ff9000 RBP: 00007ffeda93a200 R08: 0000000000000000 R09: 0000000000000000 R10: 00000000ffffffff R11: 0000000000000217 R12: 0000000000400490 R13: 00007ffeda93a2e0 R14: 0000000000000000 R15: 0000000000000000 This makes huge_zero_page bail out explicitly before split in memory_failure(), thus the panic above won't happen again. Link: https://lkml.kernel.org/r/497d3835612610e370c74e697ea3c721d1d55b9c.1649775850.git.xuyu@linux.alibaba.com Fixes: 6a46079cf57a ("HWPOISON: The high level memory error handler in the VM v7") Signed-off-by: Xu Yu Reported-by: Abaci Suggested-by: Naoya Horiguchi Acked-by: Naoya Horiguchi Reviewed-by: Miaohe Lin Cc: Anshuman Khandual Cc: Oscar Salvador Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds Signed-off-by: Greg Kroah-Hartman

mm, kfence: support kmem_dump_obj() for KFENCE objects

2022-04-27T12:38:51+00:00

commit 2dfe63e61cc31ee59ce951672b0850b5229cd5b0 upstream. Calling kmem_obj_info() via kmem_dump_obj() on KFENCE objects has been producing garbage data due to the object not actually being maintained by SLAB or SLUB. Fix this by implementing __kfence_obj_info() that copies relevant information to struct kmem_obj_info when the object was allocated by KFENCE; this is called by a common kmem_obj_info(), which also calls the slab/slub/slob specific variant now called __kmem_obj_info(). For completeness, kmem_dump_obj() now displays if the object was allocated by KFENCE. Link: https://lore.kernel.org/all/20220323090520.GG16885@xsang-OptiPlex-9020/ Link: https://lkml.kernel.org/r/20220406131558.3558585-1-elver@google.com Fixes: b89fb5ef0ce6 ("mm, kfence: insert KFENCE hooks for SLUB") Fixes: d3fb45f370d9 ("mm, kfence: insert KFENCE hooks for SLAB") Signed-off-by: Marco Elver Reviewed-by: Hyeonggon Yoo <42.hyeyoo@gmail.com> Reported-by: kernel test robot Acked-by: Vlastimil Babka [slab] Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds Signed-off-by: Greg Kroah-Hartman

mm: page_alloc: fix building error on -Werror=array-compare

2022-04-27T12:38:50+00:00

commit ca831f29f8f25c97182e726429b38c0802200c8f upstream. Arthur Marsh reported we would hit the error below when building kernel with gcc-12: CC mm/page_alloc.o mm/page_alloc.c: In function `mem_init_print_info': mm/page_alloc.c:8173:27: error: comparison between two arrays [-Werror=array-compare] 8173 | if (start <= pos && pos < end && size > adj) \ | In C++20, the comparision between arrays should be warned. Link: https://lkml.kernel.org/r/20211125130928.32465-1-sxwjean@me.com Signed-off-by: Xiongwei Song Reported-by: Arthur Marsh Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds Cc: Khem Raj Signed-off-by: Greg Kroah-Hartman

mm: kmemleak: take a full lowmem check in kmemleak_*_phys()

2022-04-20T07:34:18+00:00

commit 23c2d497de21f25898fbea70aeb292ab8acc8c94 upstream. The kmemleak_*_phys() apis do not check the address for lowmem's min boundary, while the caller may pass an address below lowmem, which will trigger an oops: # echo scan > /sys/kernel/debug/kmemleak Unable to handle kernel paging request at virtual address ff5fffffffe00000 Oops [#1] Modules linked in: CPU: 2 PID: 134 Comm: bash Not tainted 5.18.0-rc1-next-20220407 #33 Hardware name: riscv-virtio,qemu (DT) epc : scan_block+0x74/0x15c ra : scan_block+0x72/0x15c epc : ffffffff801e5806 ra : ffffffff801e5804 sp : ff200000104abc30 gp : ffffffff815cd4e8 tp : ff60000004cfa340 t0 : 0000000000000200 t1 : 00aaaaaac23954cc t2 : 00000000000003ff s0 : ff200000104abc90 s1 : ffffffff81b0ff28 a0 : 0000000000000000 a1 : ff5fffffffe01000 a2 : ffffffff81b0ff28 a3 : 0000000000000002 a4 : 0000000000000001 a5 : 0000000000000000 a6 : ff200000104abd7c a7 : 0000000000000005 s2 : ff5fffffffe00ff9 s3 : ffffffff815cd998 s4 : ffffffff815d0e90 s5 : ffffffff81b0ff28 s6 : 0000000000000020 s7 : ffffffff815d0eb0 s8 : ffffffffffffffff s9 : ff5fffffffe00000 s10: ff5fffffffe01000 s11: 0000000000000022 t3 : 00ffffffaa17db4c t4 : 000000000000000f t5 : 0000000000000001 t6 : 0000000000000000 status: 0000000000000100 badaddr: ff5fffffffe00000 cause: 000000000000000d scan_gray_list+0x12e/0x1a6 kmemleak_scan+0x2aa/0x57e kmemleak_write+0x32a/0x40c full_proxy_write+0x56/0x82 vfs_write+0xa6/0x2a6 ksys_write+0x6c/0xe2 sys_write+0x22/0x2a ret_from_syscall+0x0/0x2 The callers may not quite know the actual address they pass(e.g. from devicetree). So the kmemleak_*_phys() apis should guarantee the address they finally use is in lowmem range, so check the address for lowmem's min boundary. Link: https://lkml.kernel.org/r/20220413122925.33856-1-patrick.wang.shcn@gmail.com Signed-off-by: Patrick Wang Acked-by: Catalin Marinas Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds Signed-off-by: Greg Kroah-Hartman

mm: fix unexpected zeroed page mapping with zram swap

2022-04-20T07:34:18+00:00

commit e914d8f00391520ecc4495dd0ca0124538ab7119 upstream. Two processes under CLONE_VM cloning, user process can be corrupted by seeing zeroed page unexpectedly. CPU A CPU B do_swap_page do_swap_page SWP_SYNCHRONOUS_IO path SWP_SYNCHRONOUS_IO path swap_readpage valid data swap_slot_free_notify delete zram entry swap_readpage zeroed(invalid) data pte_lock map the *zero data* to userspace pte_unlock pte_lock if (!pte_same) goto out_nomap; pte_unlock return and next refault will read zeroed data The swap_slot_free_notify is bogus for CLONE_VM case since it doesn't increase the refcount of swap slot at copy_mm so it couldn't catch up whether it's safe or not to discard data from backing device. In the case, only the lock it could rely on to synchronize swap slot freeing is page table lock. Thus, this patch gets rid of the swap_slot_free_notify function. With this patch, CPU A will see correct data. CPU A CPU B do_swap_page do_swap_page SWP_SYNCHRONOUS_IO path SWP_SYNCHRONOUS_IO path swap_readpage original data pte_lock map the original data swap_free swap_range_free bd_disk->fops->swap_slot_free_notify swap_readpage read zeroed data pte_unlock pte_lock if (!pte_same) goto out_nomap; pte_unlock return on next refault will see mapped data by CPU B The concern of the patch would increase memory consumption since it could keep wasted memory with compressed form in zram as well as uncompressed form in address space. However, most of cases of zram uses no readahead and do_swap_page is followed by swap_free so it will free the compressed form from in zram quickly. Link: https://lkml.kernel.org/r/YjTVVxIAsnKAXjTd@google.com Fixes: 0bcac06f27d7 ("mm, swap: skip swapcache for swapin of synchronous device") Reported-by: Ivan Babrou Tested-by: Ivan Babrou Signed-off-by: Minchan Kim Cc: Nitin Gupta Cc: Sergey Senozhatsky Cc: Jens Axboe Cc: David Hildenbrand Cc: [4.14+] Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds Signed-off-by: Greg Kroah-Hartman

mm, page_alloc: fix build_zonerefs_node()

2022-04-20T07:34:18+00:00

commit e553f62f10d93551eb883eca227ac54d1a4fad84 upstream. Since commit 6aa303defb74 ("mm, vmscan: only allocate and reclaim from zones with pages managed by the buddy allocator") only zones with free memory are included in a built zonelist. This is problematic when e.g. all memory of a zone has been ballooned out when zonelists are being rebuilt. The decision whether to rebuild the zonelists when onlining new memory is done based on populated_zone() returning 0 for the zone the memory will be added to. The new zone is added to the zonelists only, if it has free memory pages (managed_zone() returns a non-zero value) after the memory has been onlined. This implies, that onlining memory will always free the added pages to the allocator immediately, but this is not true in all cases: when e.g. running as a Xen guest the onlined new memory will be added only to the ballooned memory list, it will be freed only when the guest is being ballooned up afterwards. Another problem with using managed_zone() for the decision whether a zone is being added to the zonelists is, that a zone with all memory used will in fact be removed from all zonelists in case the zonelists happen to be rebuilt. Use populated_zone() when building a zonelist as it has been done before that commit. There was a report that QubesOS (based on Xen) is hitting this problem. Xen has switched to use the zone device functionality in kernel 5.9 and QubesOS wants to use memory hotplugging for guests in order to be able to start a guest with minimal memory and expand it as needed. This was the report leading to the patch. Link: https://lkml.kernel.org/r/20220407120637.9035-1-jgross@suse.com Fixes: 6aa303defb74 ("mm, vmscan: only allocate and reclaim from zones with pages managed by the buddy allocator") Signed-off-by: Juergen Gross Reported-by: Marek Marczykowski-Górecki Acked-by: Michal Hocko Acked-by: David Hildenbrand Cc: Marek Marczykowski-Górecki Reviewed-by: Wei Yang Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds Signed-off-by: Greg Kroah-Hartman