April 2023 - Linux-stable-mirror

[merged mm-hotfixes-stable] mm-vmalloc-avoid-warn_alloc-noise-caused-by-fatal-signal.patch removed from -mm tree

by Andrew Morton

The quilt patch titled Subject: mm: vmalloc: avoid warn_alloc noise caused by fatal signal has been removed from the -mm tree. Its filename was mm-vmalloc-avoid-warn_alloc-noise-caused-by-fatal-signal.patch This patch was dropped because it was merged into the mm-hotfixes-stable branch of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm ------------------------------------------------------ From: Yafang Shao <laoar.shao(a)gmail.com> Subject: mm: vmalloc: avoid warn_alloc noise caused by fatal signal Date: Thu, 30 Mar 2023 16:26:25 +0000 There're some suspicious warn_alloc on my test serer, for example, [13366.518837] warn_alloc: 81 callbacks suppressed [13366.518841] test_verifier: vmalloc error: size 4096, page order 0, failed to allocate pages, mode:0x500dc2(GFP_HIGHUSER|__GFP_ZERO|__GFP_ACCOUNT), nodemask=(null),cpuset=/,mems_allowed=0-1 [13366.522240] CPU: 30 PID: 722463 Comm: test_verifier Kdump: loaded Tainted: G W O 6.2.0+ #638 [13366.524216] Call Trace: [13366.524702] <TASK> [13366.525148] dump_stack_lvl+0x6c/0x80 [13366.525712] dump_stack+0x10/0x20 [13366.526239] warn_alloc+0x119/0x190 [13366.526783] ? alloc_pages_bulk_array_mempolicy+0x9e/0x2a0 [13366.527470] __vmalloc_area_node+0x546/0x5b0 [13366.528066] __vmalloc_node_range+0xc2/0x210 [13366.528660] __vmalloc_node+0x42/0x50 [13366.529186] ? bpf_prog_realloc+0x53/0xc0 [13366.529743] __vmalloc+0x1e/0x30 [13366.530235] bpf_prog_realloc+0x53/0xc0 [13366.530771] bpf_patch_insn_single+0x80/0x1b0 [13366.531351] bpf_jit_blind_constants+0xe9/0x1c0 [13366.531932] ? __free_pages+0xee/0x100 [13366.532457] ? free_large_kmalloc+0x58/0xb0 [13366.533002] bpf_int_jit_compile+0x8c/0x5e0 [13366.533546] bpf_prog_select_runtime+0xb4/0x100 [13366.534108] bpf_prog_load+0x6b1/0xa50 [13366.534610] ? perf_event_task_tick+0x96/0xb0 [13366.535151] ? security_capable+0x3a/0x60 [13366.535663] __sys_bpf+0xb38/0x2190 [13366.536120] ? kvm_clock_get_cycles+0x9/0x10 [13366.536643] __x64_sys_bpf+0x1c/0x30 [13366.537094] do_syscall_64+0x38/0x90 [13366.537554] entry_SYSCALL_64_after_hwframe+0x72/0xdc [13366.538107] RIP: 0033:0x7f78310f8e29 [13366.538561] 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 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 8b 0d 17 e0 2c 00 f7 d8 64 89 01 48 [13366.540286] RSP: 002b:00007ffe2a61fff8 EFLAGS: 00000206 ORIG_RAX: 0000000000000141 [13366.541031] RAX: ffffffffffffffda RBX: 0000000000000000 RCX: 00007f78310f8e29 [13366.541749] RDX: 0000000000000080 RSI: 00007ffe2a6200b0 RDI: 0000000000000005 [13366.542470] RBP: 00007ffe2a620010 R08: 00007ffe2a6202a0 R09: 00007ffe2a6200b0 [13366.543183] R10: 00000000000f423e R11: 0000000000000206 R12: 0000000000407800 [13366.543900] R13: 00007ffe2a620540 R14: 0000000000000000 R15: 0000000000000000 [13366.544623] </TASK> [13366.545260] Mem-Info: [13366.546121] active_anon:81319 inactive_anon:20733 isolated_anon:0 active_file:69450 inactive_file:5624 isolated_file:0 unevictable:0 dirty:10 writeback:0 slab_reclaimable:69649 slab_unreclaimable:48930 mapped:27400 shmem:12868 pagetables:4929 sec_pagetables:0 bounce:0 kernel_misc_reclaimable:0 free:15870308 free_pcp:142935 free_cma:0 [13366.551886] Node 0 active_anon:224836kB inactive_anon:33528kB active_file:175692kB inactive_file:13752kB unevictable:0kB isolated(anon):0kB isolated(file):0kB mapped:59248kB dirty:32kB writeback:0kB shmem:18252kB shmem_thp: 0kB shmem_pmdmapped: 0kB anon_thp: 0kB writeback_tmp:0kB kernel_stack:4616kB pagetables:10664kB sec_pagetables:0kB all_unreclaimable? no [13366.555184] Node 1 active_anon:100440kB inactive_anon:49404kB active_file:102108kB inactive_file:8744kB unevictable:0kB isolated(anon):0kB isolated(file):0kB mapped:50352kB dirty:8kB writeback:0kB shmem:33220kB shmem_thp: 0kB shmem_pmdmapped: 0kB anon_thp: 0kB writeback_tmp:0kB kernel_stack:3896kB pagetables:9052kB sec_pagetables:0kB all_unreclaimable? no [13366.558262] Node 0 DMA free:15360kB boost:0kB min:304kB low:380kB high:456kB reserved_highatomic:0KB active_anon:0kB inactive_anon:0kB active_file:0kB inactive_file:0kB unevictable:0kB writepending:0kB present:15992kB managed:15360kB mlocked:0kB bounce:0kB free_pcp:0kB local_pcp:0kB free_cma:0kB [13366.560821] lowmem_reserve[]: 0 2735 31873 31873 31873 [13366.561981] Node 0 DMA32 free:2790904kB boost:0kB min:56028kB low:70032kB high:84036kB reserved_highatomic:0KB active_anon:1936kB inactive_anon:20kB active_file:396kB inactive_file:344kB unevictable:0kB writepending:0kB present:3129200kB managed:2801520kB mlocked:0kB bounce:0kB free_pcp:5188kB local_pcp:0kB free_cma:0kB [13366.565148] lowmem_reserve[]: 0 0 29137 29137 29137 [13366.566168] Node 0 Normal free:28533824kB boost:0kB min:596740kB low:745924kB high:895108kB reserved_highatomic:28672KB active_anon:222900kB inactive_anon:33508kB active_file:175296kB inactive_file:13408kB unevictable:0kB writepending:32kB present:30408704kB managed:29837172kB mlocked:0kB bounce:0kB free_pcp:295724kB local_pcp:0kB free_cma:0kB [13366.569485] lowmem_reserve[]: 0 0 0 0 0 [13366.570416] Node 1 Normal free:32141144kB boost:0kB min:660504kB low:825628kB high:990752kB reserved_highatomic:69632KB active_anon:100440kB inactive_anon:49404kB active_file:102108kB inactive_file:8744kB unevictable:0kB writepending:8kB present:33554432kB managed:33025372kB mlocked:0kB bounce:0kB free_pcp:270880kB local_pcp:46860kB free_cma:0kB [13366.573403] lowmem_reserve[]: 0 0 0 0 0 [13366.574015] Node 0 DMA: 0*4kB 0*8kB 0*16kB 0*32kB 0*64kB 0*128kB 0*256kB 0*512kB 1*1024kB (U) 1*2048kB (M) 3*4096kB (M) = 15360kB [13366.575474] Node 0 DMA32: 782*4kB (UME) 756*8kB (UME) 736*16kB (UME) 745*32kB (UME) 694*64kB (UME) 653*128kB (UME) 595*256kB (UME) 552*512kB (UME) 454*1024kB (UME) 347*2048kB (UME) 246*4096kB (UME) = 2790904kB [13366.577442] Node 0 Normal: 33856*4kB (UMEH) 51815*8kB (UMEH) 42418*16kB (UMEH) 36272*32kB (UMEH) 22195*64kB (UMEH) 10296*128kB (UMEH) 7238*256kB (UMEH) 5638*512kB (UEH) 5337*1024kB (UMEH) 3506*2048kB (UMEH) 1470*4096kB (UME) = 28533784kB [13366.580460] Node 1 Normal: 15776*4kB (UMEH) 37485*8kB (UMEH) 29509*16kB (UMEH) 21420*32kB (UMEH) 14818*64kB (UMEH) 13051*128kB (UMEH) 9918*256kB (UMEH) 7374*512kB (UMEH) 5397*1024kB (UMEH) 3887*2048kB (UMEH) 2002*4096kB (UME) = 32141240kB [13366.583027] Node 0 hugepages_total=0 hugepages_free=0 hugepages_surp=0 hugepages_size=1048576kB [13366.584380] Node 0 hugepages_total=0 hugepages_free=0 hugepages_surp=0 hugepages_size=2048kB [13366.585702] Node 1 hugepages_total=0 hugepages_free=0 hugepages_surp=0 hugepages_size=1048576kB [13366.587042] Node 1 hugepages_total=0 hugepages_free=0 hugepages_surp=0 hugepages_size=2048kB [13366.588372] 87386 total pagecache pages [13366.589266] 0 pages in swap cache [13366.590327] Free swap = 0kB [13366.591227] Total swap = 0kB [13366.592142] 16777082 pages RAM [13366.593057] 0 pages HighMem/MovableOnly [13366.594037] 357226 pages reserved [13366.594979] 0 pages hwpoisoned This failure really confuse me as there're still lots of available pages. Finally I figured out it was caused by a fatal signal. When a process is allocating memory via vm_area_alloc_pages(), it will break directly even if it hasn't allocated the requested pages when it receives a fatal signal. In that case, we shouldn't show this warn_alloc, as it is useless. We only need to show this warning when there're really no enough pages. Link: https://lkml.kernel.org/r/20230330162625.13604-1-laoar.shao@gmail.com Signed-off-by: Yafang Shao <laoar.shao(a)gmail.com> Reviewed-by: Lorenzo Stoakes <lstoakes(a)gmail.com> Cc: Christoph Hellwig <hch(a)infradead.org> Cc: Uladzislau Rezki (Sony) <urezki(a)gmail.com> Cc: <stable(a)vger.kernel.org> Signed-off-by: Andrew Morton <akpm(a)linux-foundation.org> --- mm/vmalloc.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) --- a/mm/vmalloc.c~mm-vmalloc-avoid-warn_alloc-noise-caused-by-fatal-signal +++ a/mm/vmalloc.c @@ -3042,9 +3042,11 @@ static void *__vmalloc_area_node(struct * allocation request, free them via vfree() if any. */ if (area->nr_pages != nr_small_pages) { - warn_alloc(gfp_mask, NULL, - "vmalloc error: size %lu, page order %u, failed to allocate pages", - area->nr_pages * PAGE_SIZE, page_order); + /* vm_area_alloc_pages() can also fail due to a fatal signal */ + if (!fatal_signal_pending(current)) + warn_alloc(gfp_mask, NULL, + "vmalloc error: size %lu, page order %u, failed to allocate pages", + area->nr_pages * PAGE_SIZE, page_order); goto fail; } _ Patches currently in -mm which might be from laoar.shao(a)gmail.com are

1 year, 8 months

1
0
0 0

[merged mm-hotfixes-stable] nilfs2-fix-potential-uaf-of-struct-nilfs_sc_info-in-nilfs_segctor_thread.patch removed from -mm tree

by Andrew Morton

The quilt patch titled Subject: nilfs2: fix potential UAF of struct nilfs_sc_info in nilfs_segctor_thread() has been removed from the -mm tree. Its filename was nilfs2-fix-potential-uaf-of-struct-nilfs_sc_info-in-nilfs_segctor_thread.patch This patch was dropped because it was merged into the mm-hotfixes-stable branch of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm ------------------------------------------------------ From: Ryusuke Konishi <konishi.ryusuke(a)gmail.com> Subject: nilfs2: fix potential UAF of struct nilfs_sc_info in nilfs_segctor_thread() Date: Tue, 28 Mar 2023 02:53:18 +0900 The finalization of nilfs_segctor_thread() can race with nilfs_segctor_kill_thread() which terminates that thread, potentially causing a use-after-free BUG as KASAN detected. At the end of nilfs_segctor_thread(), it assigns NULL to "sc_task" member of "struct nilfs_sc_info" to indicate the thread has finished, and then notifies nilfs_segctor_kill_thread() of this using waitqueue "sc_wait_task" on the struct nilfs_sc_info. However, here, immediately after the NULL assignment to "sc_task", it is possible that nilfs_segctor_kill_thread() will detect it and return to continue the deallocation, freeing the nilfs_sc_info structure before the thread does the notification. This fixes the issue by protecting the NULL assignment to "sc_task" and its notification, with spinlock "sc_state_lock" of the struct nilfs_sc_info. Since nilfs_segctor_kill_thread() does a final check to see if "sc_task" is NULL with "sc_state_lock" locked, this can eliminate the race. Link: https://lkml.kernel.org/r/20230327175318.8060-1-konishi.ryusuke@gmail.com Reported-by: syzbot+b08ebcc22f8f3e6be43a(a)syzkaller.appspotmail.com Link: https://lkml.kernel.org/r/00000000000000660d05f7dfa877@google.com Signed-off-by: Ryusuke Konishi <konishi.ryusuke(a)gmail.com> Cc: <stable(a)vger.kernel.org> Signed-off-by: Andrew Morton <akpm(a)linux-foundation.org> --- fs/nilfs2/segment.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) --- a/fs/nilfs2/segment.c~nilfs2-fix-potential-uaf-of-struct-nilfs_sc_info-in-nilfs_segctor_thread +++ a/fs/nilfs2/segment.c @@ -2609,11 +2609,10 @@ static int nilfs_segctor_thread(void *ar goto loop; end_thread: - spin_unlock(&sci->sc_state_lock); - /* end sync. */ sci->sc_task = NULL; wake_up(&sci->sc_wait_task); /* for nilfs_segctor_kill_thread() */ + spin_unlock(&sci->sc_state_lock); return 0; } _ Patches currently in -mm which might be from konishi.ryusuke(a)gmail.com are

1 year, 8 months

1
0
0 0

[merged mm-hotfixes-stable] zsmalloc-document-freeable-stats.patch removed from -mm tree

by Andrew Morton

The quilt patch titled Subject: zsmalloc: document freeable stats has been removed from the -mm tree. Its filename was zsmalloc-document-freeable-stats.patch This patch was dropped because it was merged into the mm-hotfixes-stable branch of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm ------------------------------------------------------ From: Sergey Senozhatsky <senozhatsky(a)chromium.org> Subject: zsmalloc: document freeable stats Date: Sat, 25 Mar 2023 11:46:31 +0900 When freeable class stat was added to classes file (back in 2016) we forgot to update zsmalloc documentation. Fix that. Link: https://lkml.kernel.org/r/20230325024631.2817153-3-senozhatsky@chromium.org Fixes: 1120ed548394 ("mm/zsmalloc: add `freeable' column to pool stat") Signed-off-by: Sergey Senozhatsky <senozhatsky(a)chromium.org> Cc: Minchan Kim <minchan(a)kernel.org> Cc: <stable(a)vger.kernel.org> Signed-off-by: Andrew Morton <akpm(a)linux-foundation.org> --- Documentation/mm/zsmalloc.rst | 2 ++ 1 file changed, 2 insertions(+) --- a/Documentation/mm/zsmalloc.rst~zsmalloc-document-freeable-stats +++ a/Documentation/mm/zsmalloc.rst @@ -83,6 +83,8 @@ pages_used the number of pages allocated for the class pages_per_zspage the number of 0-order pages to make a zspage +freeable + the approximate number of pages class compaction can free Each zspage maintains inuse counter which keeps track of the number of objects stored in the zspage. The inuse counter determines the zspage's _ Patches currently in -mm which might be from senozhatsky(a)chromium.org are

1 year, 8 months

1
0
0 0

[merged mm-hotfixes-stable] zsmalloc-document-new-fullness-grouping.patch removed from -mm tree

by Andrew Morton

The quilt patch titled Subject: zsmalloc: document new fullness grouping has been removed from the -mm tree. Its filename was zsmalloc-document-new-fullness-grouping.patch This patch was dropped because it was merged into the mm-hotfixes-stable branch of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm ------------------------------------------------------ From: Sergey Senozhatsky <senozhatsky(a)chromium.org> Subject: zsmalloc: document new fullness grouping Date: Sat, 25 Mar 2023 11:46:30 +0900 Patch series "zsmalloc: minor documentation updates". Two minor patches that bring zsmalloc documentation up to date. This patch (of 2): Update documentation and reflect new zspages fullness grouping (we don't use almost_empty and almost_full anymore). Link: https://lkml.kernel.org/r/20230325024631.2817153-1-senozhatsky@chromium.org Link: https://lkml.kernel.org/r/20230325024631.2817153-2-senozhatsky@chromium.org Signed-off-by: Sergey Senozhatsky <senozhatsky(a)chromium.org> Fixes: 67e157eb3639 ("zsmalloc: show per fullness group class stats") Cc: Minchan Kim <minchan(a)kernel.org> Cc: <stable(a)vger.kernel.org> Signed-off-by: Andrew Morton <akpm(a)linux-foundation.org> --- Documentation/mm/zsmalloc.rst | 133 +++++++++++++++++--------------- 1 file changed, 74 insertions(+), 59 deletions(-) --- a/Documentation/mm/zsmalloc.rst~zsmalloc-document-new-fullness-grouping +++ a/Documentation/mm/zsmalloc.rst @@ -39,13 +39,12 @@ With CONFIG_ZSMALLOC_STAT, we could see # cat /sys/kernel/debug/zsmalloc/zram0/classes - class size almost_full almost_empty obj_allocated obj_used pages_used pages_per_zspage + class size 10% 20% 30% 40% 50% 60% 70% 80% 90% 99% 100% obj_allocated obj_used pages_used pages_per_zspage freeable ... ... - 9 176 0 1 186 129 8 4 - 10 192 1 0 2880 2872 135 3 - 11 208 0 1 819 795 42 2 - 12 224 0 1 219 159 12 4 + 30 512 0 12 4 1 0 1 0 0 1 0 414 3464 3346 433 1 14 + 31 528 2 7 2 2 1 0 1 0 0 2 117 4154 3793 536 4 44 + 32 544 6 3 4 1 2 1 0 0 0 1 260 4170 3965 556 2 26 ... ... @@ -54,10 +53,28 @@ class index size object size zspage stores -almost_empty - the number of ZS_ALMOST_EMPTY zspages(see below) -almost_full - the number of ZS_ALMOST_FULL zspages(see below) +10% + the number of zspages with usage ratio less than 10% (see below) +20% + the number of zspages with usage ratio between 10% and 20% +30% + the number of zspages with usage ratio between 20% and 30% +40% + the number of zspages with usage ratio between 30% and 40% +50% + the number of zspages with usage ratio between 40% and 50% +60% + the number of zspages with usage ratio between 50% and 60% +70% + the number of zspages with usage ratio between 60% and 70% +80% + the number of zspages with usage ratio between 70% and 80% +90% + the number of zspages with usage ratio between 80% and 90% +99% + the number of zspages with usage ratio between 90% and 99% +100% + the number of zspages with usage ratio 100% obj_allocated the number of objects allocated obj_used @@ -67,18 +84,11 @@ pages_used pages_per_zspage the number of 0-order pages to make a zspage -We assign a zspage to ZS_ALMOST_EMPTY fullness group when n <= N / f, where - -* n = number of allocated objects -* N = total number of objects zspage can store -* f = fullness_threshold_frac(ie, 4 at the moment) - -Similarly, we assign zspage to: - -* ZS_ALMOST_FULL when n > N / f -* ZS_EMPTY when n == 0 -* ZS_FULL when n == N - +Each zspage maintains inuse counter which keeps track of the number of +objects stored in the zspage. The inuse counter determines the zspage's +"fullness group" which is calculated as the ratio of the "inuse" objects to +the total number of objects the zspage can hold (objs_per_zspage). The +closer the inuse counter is to objs_per_zspage, the better. Internals ========= @@ -94,10 +104,10 @@ of objects that each zspage can store. For instance, consider the following size classes::: - class size almost_full almost_empty obj_allocated obj_used pages_used pages_per_zspage freeable + class size 10% .... 100% obj_allocated obj_used pages_used pages_per_zspage freeable ... - 94 1536 0 0 0 0 0 3 0 - 100 1632 0 0 0 0 0 2 0 + 94 1536 0 .... 0 0 0 0 3 0 + 100 1632 0 .... 0 0 0 0 2 0 ... @@ -134,10 +144,11 @@ reduces memory wastage. Let's take a closer look at the bottom of `/sys/kernel/debug/zsmalloc/zramX/classes`::: - class size almost_full almost_empty obj_allocated obj_used pages_used pages_per_zspage freeable + class size 10% .... 100% obj_allocated obj_used pages_used pages_per_zspage freeable + ... - 202 3264 0 0 0 0 0 4 0 - 254 4096 0 0 0 0 0 1 0 + 202 3264 0 .. 0 0 0 0 4 0 + 254 4096 0 .. 0 0 0 0 1 0 ... Size class #202 stores objects of size 3264 bytes and has a maximum of 4 pages @@ -151,40 +162,42 @@ efficient storage of large objects. For zspage chain size of 8, huge class watermark becomes 3632 bytes::: - class size almost_full almost_empty obj_allocated obj_used pages_used pages_per_zspage freeable + class size 10% .... 100% obj_allocated obj_used pages_used pages_per_zspage freeable + ... - 202 3264 0 0 0 0 0 4 0 - 211 3408 0 0 0 0 0 5 0 - 217 3504 0 0 0 0 0 6 0 - 222 3584 0 0 0 0 0 7 0 - 225 3632 0 0 0 0 0 8 0 - 254 4096 0 0 0 0 0 1 0 + 202 3264 0 .. 0 0 0 0 4 0 + 211 3408 0 .. 0 0 0 0 5 0 + 217 3504 0 .. 0 0 0 0 6 0 + 222 3584 0 .. 0 0 0 0 7 0 + 225 3632 0 .. 0 0 0 0 8 0 + 254 4096 0 .. 0 0 0 0 1 0 ... For zspage chain size of 16, huge class watermark becomes 3840 bytes::: - class size almost_full almost_empty obj_allocated obj_used pages_used pages_per_zspage freeable + class size 10% .... 100% obj_allocated obj_used pages_used pages_per_zspage freeable + ... - 202 3264 0 0 0 0 0 4 0 - 206 3328 0 0 0 0 0 13 0 - 207 3344 0 0 0 0 0 9 0 - 208 3360 0 0 0 0 0 14 0 - 211 3408 0 0 0 0 0 5 0 - 212 3424 0 0 0 0 0 16 0 - 214 3456 0 0 0 0 0 11 0 - 217 3504 0 0 0 0 0 6 0 - 219 3536 0 0 0 0 0 13 0 - 222 3584 0 0 0 0 0 7 0 - 223 3600 0 0 0 0 0 15 0 - 225 3632 0 0 0 0 0 8 0 - 228 3680 0 0 0 0 0 9 0 - 230 3712 0 0 0 0 0 10 0 - 232 3744 0 0 0 0 0 11 0 - 234 3776 0 0 0 0 0 12 0 - 235 3792 0 0 0 0 0 13 0 - 236 3808 0 0 0 0 0 14 0 - 238 3840 0 0 0 0 0 15 0 - 254 4096 0 0 0 0 0 1 0 + 202 3264 0 .. 0 0 0 0 4 0 + 206 3328 0 .. 0 0 0 0 13 0 + 207 3344 0 .. 0 0 0 0 9 0 + 208 3360 0 .. 0 0 0 0 14 0 + 211 3408 0 .. 0 0 0 0 5 0 + 212 3424 0 .. 0 0 0 0 16 0 + 214 3456 0 .. 0 0 0 0 11 0 + 217 3504 0 .. 0 0 0 0 6 0 + 219 3536 0 .. 0 0 0 0 13 0 + 222 3584 0 .. 0 0 0 0 7 0 + 223 3600 0 .. 0 0 0 0 15 0 + 225 3632 0 .. 0 0 0 0 8 0 + 228 3680 0 .. 0 0 0 0 9 0 + 230 3712 0 .. 0 0 0 0 10 0 + 232 3744 0 .. 0 0 0 0 11 0 + 234 3776 0 .. 0 0 0 0 12 0 + 235 3792 0 .. 0 0 0 0 13 0 + 236 3808 0 .. 0 0 0 0 14 0 + 238 3840 0 .. 0 0 0 0 15 0 + 254 4096 0 .. 0 0 0 0 1 0 ... Overall the combined zspage chain size effect on zsmalloc pool configuration::: @@ -214,9 +227,10 @@ zram as a build artifacts storage (Linux zsmalloc classes stats::: - class size almost_full almost_empty obj_allocated obj_used pages_used pages_per_zspage freeable + class size 10% .... 100% obj_allocated obj_used pages_used pages_per_zspage freeable + ... - Total 13 51 413836 412973 159955 3 + Total 13 .. 51 413836 412973 159955 3 zram mm_stat::: @@ -227,9 +241,10 @@ zram as a build artifacts storage (Linux zsmalloc classes stats::: - class size almost_full almost_empty obj_allocated obj_used pages_used pages_per_zspage freeable + class size 10% .... 100% obj_allocated obj_used pages_used pages_per_zspage freeable + ... - Total 18 87 414852 412978 156666 0 + Total 18 .. 87 414852 412978 156666 0 zram mm_stat::: _ Patches currently in -mm which might be from senozhatsky(a)chromium.org are

1 year, 8 months

1
0
0 0

[merged mm-hotfixes-stable] fsdax-force-clear-dirty-mark-if-cow.patch removed from -mm tree

by Andrew Morton

The quilt patch titled Subject: fsdax: force clear dirty mark if CoW has been removed from the -mm tree. Its filename was fsdax-force-clear-dirty-mark-if-cow.patch This patch was dropped because it was merged into the mm-hotfixes-stable branch of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm ------------------------------------------------------ From: Shiyang Ruan <ruansy.fnst(a)fujitsu.com> Subject: fsdax: force clear dirty mark if CoW Date: Fri, 24 Mar 2023 10:28:00 +0000 XFS allows CoW on non-shared extents to combat fragmentation[1]. The old non-shared extent could be mwrited before, its dax entry is marked dirty. This results in a WARNing: [ 28.512349] ------------[ cut here ]------------ [ 28.512622] WARNING: CPU: 2 PID: 5255 at fs/dax.c:390 dax_insert_entry+0x342/0x390 [ 28.513050] Modules linked in: rpcsec_gss_krb5 auth_rpcgss nfsv4 nfs lockd grace fscache netfs nft_fib_inet nft_fib_ipv4 nft_fib_ipv6 nft_fib nft_reject_inet nf_reject_ipv4 nf_reject_ipv6 nft_reject nft_ct nf_conntrack nf_defrag_ipv6 nf_defrag_ipv4 ip_set nf_tables [ 28.515462] CPU: 2 PID: 5255 Comm: fsstress Kdump: loaded Not tainted 6.3.0-rc1-00001-g85e1481e19c1-dirty #117 [ 28.515902] Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS Arch Linux 1.16.1-1-1 04/01/2014 [ 28.516307] RIP: 0010:dax_insert_entry+0x342/0x390 [ 28.516536] Code: 30 5b 5d 41 5c 41 5d 41 5e 41 5f c3 cc cc cc cc 48 8b 45 20 48 83 c0 01 e9 e2 fe ff ff 48 8b 45 20 48 83 c0 01 e9 cd fe ff ff <0f> 0b e9 53 ff ff ff 48 8b 7c 24 08 31 f6 e8 1b 61 a1 00 eb 8c 48 [ 28.517417] RSP: 0000:ffffc9000845fb18 EFLAGS: 00010086 [ 28.517721] RAX: 0000000000000053 RBX: 0000000000000155 RCX: 000000000018824b [ 28.518113] RDX: 0000000000000000 RSI: ffffffff827525a6 RDI: 00000000ffffffff [ 28.518515] RBP: ffffea00062092c0 R08: 0000000000000000 R09: ffffc9000845f9c8 [ 28.518905] R10: 0000000000000003 R11: ffffffff82ddb7e8 R12: 0000000000000155 [ 28.519301] R13: 0000000000000000 R14: 000000000018824b R15: ffff88810cfa76b8 [ 28.519703] FS: 00007f14a0c94740(0000) GS:ffff88817bd00000(0000) knlGS:0000000000000000 [ 28.520148] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 28.520472] CR2: 00007f14a0c8d000 CR3: 000000010321c004 CR4: 0000000000770ee0 [ 28.520863] PKRU: 55555554 [ 28.521043] Call Trace: [ 28.521219] <TASK> [ 28.521368] dax_fault_iter+0x196/0x390 [ 28.521595] dax_iomap_pte_fault+0x19b/0x3d0 [ 28.521852] __xfs_filemap_fault+0x234/0x2b0 [ 28.522116] __do_fault+0x30/0x130 [ 28.522334] do_fault+0x193/0x340 [ 28.522586] __handle_mm_fault+0x2d3/0x690 [ 28.522975] handle_mm_fault+0xe6/0x2c0 [ 28.523259] do_user_addr_fault+0x1bc/0x6f0 [ 28.523521] exc_page_fault+0x60/0x140 [ 28.523763] asm_exc_page_fault+0x22/0x30 [ 28.524001] RIP: 0033:0x7f14a0b589ca [ 28.524225] Code: c5 fe 7f 07 c5 fe 7f 47 20 c5 fe 7f 47 40 c5 fe 7f 47 60 c5 f8 77 c3 66 0f 1f 84 00 00 00 00 00 40 0f b6 c6 48 89 d1 48 89 fa <f3> aa 48 89 d0 c5 f8 77 c3 66 66 2e 0f 1f 84 00 00 00 00 00 66 90 [ 28.525198] RSP: 002b:00007fff1dea1c98 EFLAGS: 00010202 [ 28.525505] RAX: 000000000000001e RBX: 000000000014a000 RCX: 0000000000006046 [ 28.525895] RDX: 00007f14a0c82000 RSI: 000000000000001e RDI: 00007f14a0c8d000 [ 28.526290] RBP: 000000000000006f R08: 0000000000000004 R09: 000000000014a000 [ 28.526681] R10: 0000000000000008 R11: 0000000000000246 R12: 028f5c28f5c28f5c [ 28.527067] R13: 8f5c28f5c28f5c29 R14: 0000000000011046 R15: 00007f14a0c946c0 [ 28.527449] </TASK> [ 28.527600] ---[ end trace 0000000000000000 ]--- To be able to delete this entry, clear its dirty mark before invalidate_inode_pages2_range(). [1] https://lore.kernel.org/linux-xfs/20230321151339.GA11376@frogsfrogsfrogs/ Link: https://lkml.kernel.org/r/1679653680-2-1-git-send-email-ruansy.fnst@fujitsu… Fixes: f80e1668888f3 ("fsdax: invalidate pages when CoW") Signed-off-by: Shiyang Ruan <ruansy.fnst(a)fujitsu.com> Cc: Dan Williams <dan.j.williams(a)intel.com> Cc: Darrick J. Wong <djwong(a)kernel.org> Cc: Jan Kara <jack(a)suse.cz> Cc: Matthew Wilcox (Oracle) <willy(a)infradead.org> Cc: <stable(a)vger.kernel.org> Signed-off-by: Andrew Morton <akpm(a)linux-foundation.org> --- fs/dax.c | 37 +++++++++++++++++++++++++++++++++++++ 1 file changed, 37 insertions(+) --- a/fs/dax.c~fsdax-force-clear-dirty-mark-if-cow +++ a/fs/dax.c @@ -781,6 +781,33 @@ out: return ret; } +static int __dax_clear_dirty_range(struct address_space *mapping, + pgoff_t start, pgoff_t end) +{ + XA_STATE(xas, &mapping->i_pages, start); + unsigned int scanned = 0; + void *entry; + + xas_lock_irq(&xas); + xas_for_each(&xas, entry, end) { + entry = get_unlocked_entry(&xas, 0); + xas_clear_mark(&xas, PAGECACHE_TAG_DIRTY); + xas_clear_mark(&xas, PAGECACHE_TAG_TOWRITE); + put_unlocked_entry(&xas, entry, WAKE_NEXT); + + if (++scanned % XA_CHECK_SCHED) + continue; + + xas_pause(&xas); + xas_unlock_irq(&xas); + cond_resched(); + xas_lock_irq(&xas); + } + xas_unlock_irq(&xas); + + return 0; +} + /* * Delete DAX entry at @index from @mapping. Wait for it * to be unlocked before deleting it. @@ -1440,6 +1467,16 @@ static loff_t dax_iomap_iter(const struc * written by write(2) is visible in mmap. */ if (iomap->flags & IOMAP_F_NEW || cow) { + /* + * Filesystem allows CoW on non-shared extents. The src extents + * may have been mmapped with dirty mark before. To be able to + * invalidate its dax entries, we need to clear the dirty mark + * in advance. + */ + if (cow) + __dax_clear_dirty_range(iomi->inode->i_mapping, + pos >> PAGE_SHIFT, + (end - 1) >> PAGE_SHIFT); invalidate_inode_pages2_range(iomi->inode->i_mapping, pos >> PAGE_SHIFT, (end - 1) >> PAGE_SHIFT); _ Patches currently in -mm which might be from ruansy.fnst(a)fujitsu.com are

1 year, 8 months

1
0
0 0

[merged mm-hotfixes-stable] mm-hugetlb-fix-uffd-wr-protection-for-cow-optimization-path.patch removed from -mm tree

by Andrew Morton

The quilt patch titled Subject: mm/hugetlb: fix uffd wr-protection for CoW optimization path has been removed from the -mm tree. Its filename was mm-hugetlb-fix-uffd-wr-protection-for-cow-optimization-path.patch This patch was dropped because it was merged into the mm-hotfixes-stable branch of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm ------------------------------------------------------ From: Peter Xu <peterx(a)redhat.com> Subject: mm/hugetlb: fix uffd wr-protection for CoW optimization path Date: Tue, 21 Mar 2023 15:18:40 -0400 This patch fixes an issue that a hugetlb uffd-wr-protected mapping can be writable even with uffd-wp bit set. It only happens with hugetlb private mappings, when someone firstly wr-protects a missing pte (which will install a pte marker), then a write to the same page without any prior access to the page. Userfaultfd-wp trap for hugetlb was implemented in hugetlb_fault() before reaching hugetlb_wp() to avoid taking more locks that userfault won't need. However there's one CoW optimization path that can trigger hugetlb_wp() inside hugetlb_no_page(), which will bypass the trap. This patch skips hugetlb_wp() for CoW and retries the fault if uffd-wp bit is detected. The new path will only trigger in the CoW optimization path because generic hugetlb_fault() (e.g. when a present pte was wr-protected) will resolve the uffd-wp bit already. Also make sure anonymous UNSHARE won't be affected and can still be resolved, IOW only skip CoW not CoR. This patch will be needed for v5.19+ hence copy stable. [peterx(a)redhat.com: v2] Link: https://lkml.kernel.org/r/ZBzOqwF2wrHgBVZb@x1n [peterx(a)redhat.com: v3] Link: https://lkml.kernel.org/r/20230324142620.2344140-1-peterx@redhat.com Link: https://lkml.kernel.org/r/20230321191840.1897940-1-peterx@redhat.com Fixes: 166f3ecc0daf ("mm/hugetlb: hook page faults for uffd write protection") Signed-off-by: Peter Xu <peterx(a)redhat.com> Reported-by: Muhammad Usama Anjum <usama.anjum(a)collabora.com> Tested-by: Muhammad Usama Anjum <usama.anjum(a)collabora.com> Acked-by: David Hildenbrand <david(a)redhat.com> Reviewed-by: Mike Kravetz <mike.kravetz(a)oracle.com> Cc: Andrea Arcangeli <aarcange(a)redhat.com> Cc: Axel Rasmussen <axelrasmussen(a)google.com> Cc: Mike Rapoport <rppt(a)linux.vnet.ibm.com> Cc: Nadav Amit <nadav.amit(a)gmail.com> Cc: <stable(a)vger.kernel.org> Signed-off-by: Andrew Morton <akpm(a)linux-foundation.org> --- mm/hugetlb.c | 14 ++++++++++++-- 1 file changed, 12 insertions(+), 2 deletions(-) --- a/mm/hugetlb.c~mm-hugetlb-fix-uffd-wr-protection-for-cow-optimization-path +++ a/mm/hugetlb.c @@ -5478,7 +5478,7 @@ static vm_fault_t hugetlb_wp(struct mm_s struct folio *pagecache_folio, spinlock_t *ptl) { const bool unshare = flags & FAULT_FLAG_UNSHARE; - pte_t pte; + pte_t pte = huge_ptep_get(ptep); struct hstate *h = hstate_vma(vma); struct page *old_page; struct folio *new_folio; @@ -5488,6 +5488,17 @@ static vm_fault_t hugetlb_wp(struct mm_s struct mmu_notifier_range range; /* + * Never handle CoW for uffd-wp protected pages. It should be only + * handled when the uffd-wp protection is removed. + * + * Note that only the CoW optimization path (in hugetlb_no_page()) + * can trigger this, because hugetlb_fault() will always resolve + * uffd-wp bit first. + */ + if (!unshare && huge_pte_uffd_wp(pte)) + return 0; + + /* * hugetlb does not support FOLL_FORCE-style write faults that keep the * PTE mapped R/O such as maybe_mkwrite() would do. */ @@ -5500,7 +5511,6 @@ static vm_fault_t hugetlb_wp(struct mm_s return 0; } - pte = huge_ptep_get(ptep); old_page = pte_page(pte); delayacct_wpcopy_start(); _ Patches currently in -mm which might be from peterx(a)redhat.com are mm-khugepaged-check-again-on-anon-uffd-wp-during-isolation.patch mm-uffd-uffd_feature_wp_unpopulated.patch mm-uffd-uffd_feature_wp_unpopulated-fix.patch selftests-mm-smoke-test-uffd_feature_wp_unpopulated.patch

1 year, 8 months

1
0
0 0

[merged mm-hotfixes-stable] mm-enable-maple-tree-rcu-mode-by-default.patch removed from -mm tree

by Andrew Morton

The quilt patch titled Subject: mm: enable maple tree RCU mode by default has been removed from the -mm tree. Its filename was mm-enable-maple-tree-rcu-mode-by-default.patch This patch was dropped because it was merged into the mm-hotfixes-stable branch of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm ------------------------------------------------------ From: "Liam R. Howlett" <Liam.Howlett(a)oracle.com> Subject: mm: enable maple tree RCU mode by default Date: Mon, 27 Feb 2023 09:36:07 -0800 Use the maple tree in RCU mode for VMA tracking. The maple tree tracks the stack and is able to update the pivot (lower/upper boundary) in-place to allow the page fault handler to write to the tree while holding just the mmap read lock. This is safe as the writes to the stack have a guard VMA which ensures there will always be a NULL in the direction of the growth and thus will only update a pivot. It is possible, but not recommended, to have VMAs that grow up/down without guard VMAs. syzbot has constructed a testcase which sets up a VMA to grow and consume the empty space. Overwriting the entire NULL entry causes the tree to be altered in a way that is not safe for concurrent readers; the readers may see a node being rewritten or one that does not match the maple state they are using. Enabling RCU mode allows the concurrent readers to see a stable node and will return the expected result. [Liam.Howlett(a)Oracle.com: we don't need to free the nodes with RCU[ Link: https://lore.kernel.org/linux-mm/000000000000b0a65805f663ace6@google.com/ Link: https://lkml.kernel.org/r/20230227173632.3292573-9-surenb@google.com Fixes: d4af56c5c7c6 ("mm: start tracking VMAs with maple tree") Signed-off-by: Liam R. Howlett <Liam.Howlett(a)oracle.com> Signed-off-by: Suren Baghdasaryan <surenb(a)google.com> Reported-by: syzbot+8d95422d3537159ca390(a)syzkaller.appspotmail.com Cc: <stable(a)vger.kernel.org> Signed-off-by: Andrew Morton <akpm(a)linux-foundation.org> --- include/linux/mm_types.h | 3 ++- kernel/fork.c | 3 +++ mm/mmap.c | 3 ++- 3 files changed, 7 insertions(+), 2 deletions(-) --- a/include/linux/mm_types.h~mm-enable-maple-tree-rcu-mode-by-default +++ a/include/linux/mm_types.h @@ -774,7 +774,8 @@ struct mm_struct { unsigned long cpu_bitmap[]; }; -#define MM_MT_FLAGS (MT_FLAGS_ALLOC_RANGE | MT_FLAGS_LOCK_EXTERN) +#define MM_MT_FLAGS (MT_FLAGS_ALLOC_RANGE | MT_FLAGS_LOCK_EXTERN | \ + MT_FLAGS_USE_RCU) extern struct mm_struct init_mm; /* Pointer magic because the dynamic array size confuses some compilers. */ --- a/kernel/fork.c~mm-enable-maple-tree-rcu-mode-by-default +++ a/kernel/fork.c @@ -617,6 +617,7 @@ static __latent_entropy int dup_mmap(str if (retval) goto out; + mt_clear_in_rcu(vmi.mas.tree); for_each_vma(old_vmi, mpnt) { struct file *file; @@ -700,6 +701,8 @@ static __latent_entropy int dup_mmap(str retval = arch_dup_mmap(oldmm, mm); loop_out: vma_iter_free(&vmi); + if (!retval) + mt_set_in_rcu(vmi.mas.tree); out: mmap_write_unlock(mm); flush_tlb_mm(oldmm); --- a/mm/mmap.c~mm-enable-maple-tree-rcu-mode-by-default +++ a/mm/mmap.c @@ -2277,7 +2277,7 @@ do_vmi_align_munmap(struct vma_iterator int count = 0; int error = -ENOMEM; MA_STATE(mas_detach, &mt_detach, 0, 0); - mt_init_flags(&mt_detach, MT_FLAGS_LOCK_EXTERN); + mt_init_flags(&mt_detach, vmi->mas.tree->ma_flags & MT_FLAGS_LOCK_MASK); mt_set_external_lock(&mt_detach, &mm->mmap_lock); /* @@ -3037,6 +3037,7 @@ void exit_mmap(struct mm_struct *mm) */ set_bit(MMF_OOM_SKIP, &mm->flags); mmap_write_lock(mm); + mt_clear_in_rcu(&mm->mm_mt); free_pgtables(&tlb, &mm->mm_mt, vma, FIRST_USER_ADDRESS, USER_PGTABLES_CEILING); tlb_finish_mmu(&tlb); _ Patches currently in -mm which might be from Liam.Howlett(a)oracle.com are

1 year, 8 months

1
0
0 0

[merged mm-hotfixes-stable] maple_tree-add-rcu-lock-checking-to-rcu-callback-functions.patch removed from -mm tree

by Andrew Morton

The quilt patch titled Subject: maple_tree: add RCU lock checking to rcu callback functions has been removed from the -mm tree. Its filename was maple_tree-add-rcu-lock-checking-to-rcu-callback-functions.patch This patch was dropped because it was merged into the mm-hotfixes-stable branch of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm ------------------------------------------------------ From: "Liam R. Howlett" <Liam.Howlett(a)oracle.com> Subject: maple_tree: add RCU lock checking to rcu callback functions Date: Mon, 27 Feb 2023 09:36:06 -0800 Dereferencing RCU objects within the RCU callback without the RCU check has caused lockdep to complain. Fix the RCU dereferencing by using the RCU callback lock to ensure the operation is safe. Also stop creating a new lock to use for dereferencing during destruction of the tree or subtree. Instead, pass through a pointer to the tree that has the lock that is held for RCU dereferencing checking. It also does not make sense to use the maple state in the freeing scenario as the tree walk is a special case where the tree no longer has the normal encodings and parent pointers. Link: https://lkml.kernel.org/r/20230227173632.3292573-8-surenb@google.com Fixes: 54a611b60590 ("Maple Tree: add new data structure") Signed-off-by: Liam R. Howlett <Liam.Howlett(a)oracle.com> Reported-by: Suren Baghdasaryan <surenb(a)google.com> Cc: <stable(a)vger.kernel.org> Signed-off-by: Andrew Morton <akpm(a)linux-foundation.org> --- lib/maple_tree.c | 188 ++++++++++++++++++++++----------------------- 1 file changed, 96 insertions(+), 92 deletions(-) --- a/lib/maple_tree.c~maple_tree-add-rcu-lock-checking-to-rcu-callback-functions +++ a/lib/maple_tree.c @@ -824,6 +824,11 @@ static inline void *mt_slot(const struct return rcu_dereference_check(slots[offset], mt_locked(mt)); } +static inline void *mt_slot_locked(struct maple_tree *mt, void __rcu **slots, + unsigned char offset) +{ + return rcu_dereference_protected(slots[offset], mt_locked(mt)); +} /* * mas_slot_locked() - Get the slot value when holding the maple tree lock. * @mas: The maple state @@ -835,7 +840,7 @@ static inline void *mt_slot(const struct static inline void *mas_slot_locked(struct ma_state *mas, void __rcu **slots, unsigned char offset) { - return rcu_dereference_protected(slots[offset], mt_locked(mas->tree)); + return mt_slot_locked(mas->tree, slots, offset); } /* @@ -907,34 +912,35 @@ static inline void ma_set_meta(struct ma } /* - * mas_clear_meta() - clear the metadata information of a node, if it exists - * @mas: The maple state + * mt_clear_meta() - clear the metadata information of a node, if it exists + * @mt: The maple tree * @mn: The maple node - * @mt: The maple node type + * @type: The maple node type * @offset: The offset of the highest sub-gap in this node. * @end: The end of the data in this node. */ -static inline void mas_clear_meta(struct ma_state *mas, struct maple_node *mn, - enum maple_type mt) +static inline void mt_clear_meta(struct maple_tree *mt, struct maple_node *mn, + enum maple_type type) { struct maple_metadata *meta; unsigned long *pivots; void __rcu **slots; void *next; - switch (mt) { + switch (type) { case maple_range_64: pivots = mn->mr64.pivot; if (unlikely(pivots[MAPLE_RANGE64_SLOTS - 2])) { slots = mn->mr64.slot; - next = mas_slot_locked(mas, slots, - MAPLE_RANGE64_SLOTS - 1); - if (unlikely((mte_to_node(next) && mte_node_type(next)))) - return; /* The last slot is a node, no metadata */ + next = mt_slot_locked(mt, slots, + MAPLE_RANGE64_SLOTS - 1); + if (unlikely((mte_to_node(next) && + mte_node_type(next)))) + return; /* no metadata, could be node */ } fallthrough; case maple_arange_64: - meta = ma_meta(mn, mt); + meta = ma_meta(mn, type); break; default: return; @@ -5483,7 +5489,7 @@ no_gap: } /* - * mas_dead_leaves() - Mark all leaves of a node as dead. + * mte_dead_leaves() - Mark all leaves of a node as dead. * @mas: The maple state * @slots: Pointer to the slot array * @type: The maple node type @@ -5493,16 +5499,16 @@ no_gap: * Return: The number of leaves marked as dead. */ static inline -unsigned char mas_dead_leaves(struct ma_state *mas, void __rcu **slots, - enum maple_type mt) +unsigned char mte_dead_leaves(struct maple_enode *enode, struct maple_tree *mt, + void __rcu **slots) { struct maple_node *node; enum maple_type type; void *entry; int offset; - for (offset = 0; offset < mt_slots[mt]; offset++) { - entry = mas_slot_locked(mas, slots, offset); + for (offset = 0; offset < mt_slot_count(enode); offset++) { + entry = mt_slot(mt, slots, offset); type = mte_node_type(entry); node = mte_to_node(entry); /* Use both node and type to catch LE & BE metadata */ @@ -5517,162 +5523,160 @@ unsigned char mas_dead_leaves(struct ma_ return offset; } -static void __rcu **mas_dead_walk(struct ma_state *mas, unsigned char offset) +/** + * mte_dead_walk() - Walk down a dead tree to just before the leaves + * @enode: The maple encoded node + * @offset: The starting offset + * + * Note: This can only be used from the RCU callback context. + */ +static void __rcu **mte_dead_walk(struct maple_enode **enode, unsigned char offset) { - struct maple_node *next; + struct maple_node *node, *next; void __rcu **slots = NULL; - next = mas_mn(mas); + next = mte_to_node(*enode); do { - mas->node = mt_mk_node(next, next->type); - slots = ma_slots(next, next->type); - next = mas_slot_locked(mas, slots, offset); + *enode = ma_enode_ptr(next); + node = mte_to_node(*enode); + slots = ma_slots(node, node->type); + next = rcu_dereference_protected(slots[offset], + lock_is_held(&rcu_callback_map)); offset = 0; } while (!ma_is_leaf(next->type)); return slots; } +/** + * mt_free_walk() - Walk & free a tree in the RCU callback context + * @head: The RCU head that's within the node. + * + * Note: This can only be used from the RCU callback context. + */ static void mt_free_walk(struct rcu_head *head) { void __rcu **slots; struct maple_node *node, *start; - struct maple_tree mt; + struct maple_enode *enode; unsigned char offset; enum maple_type type; - MA_STATE(mas, &mt, 0, 0); node = container_of(head, struct maple_node, rcu); if (ma_is_leaf(node->type)) goto free_leaf; - mt_init_flags(&mt, node->ma_flags); - mas_lock(&mas); start = node; - mas.node = mt_mk_node(node, node->type); - slots = mas_dead_walk(&mas, 0); - node = mas_mn(&mas); + enode = mt_mk_node(node, node->type); + slots = mte_dead_walk(&enode, 0); + node = mte_to_node(enode); do { mt_free_bulk(node->slot_len, slots); offset = node->parent_slot + 1; - mas.node = node->piv_parent; - if (mas_mn(&mas) == node) - goto start_slots_free; - - type = mte_node_type(mas.node); - slots = ma_slots(mte_to_node(mas.node), type); - if ((offset < mt_slots[type]) && (slots[offset])) - slots = mas_dead_walk(&mas, offset); - - node = mas_mn(&mas); + enode = node->piv_parent; + if (mte_to_node(enode) == node) + goto free_leaf; + + type = mte_node_type(enode); + slots = ma_slots(mte_to_node(enode), type); + if ((offset < mt_slots[type]) && + rcu_dereference_protected(slots[offset], + lock_is_held(&rcu_callback_map))) + slots = mte_dead_walk(&enode, offset); + node = mte_to_node(enode); } while ((node != start) || (node->slot_len < offset)); slots = ma_slots(node, node->type); mt_free_bulk(node->slot_len, slots); -start_slots_free: - mas_unlock(&mas); free_leaf: mt_free_rcu(&node->rcu); } -static inline void __rcu **mas_destroy_descend(struct ma_state *mas, - struct maple_enode *prev, unsigned char offset) +static inline void __rcu **mte_destroy_descend(struct maple_enode **enode, + struct maple_tree *mt, struct maple_enode *prev, unsigned char offset) { struct maple_node *node; - struct maple_enode *next = mas->node; + struct maple_enode *next = *enode; void __rcu **slots = NULL; + enum maple_type type; + unsigned char next_offset = 0; do { - mas->node = next; - node = mas_mn(mas); - slots = ma_slots(node, mte_node_type(mas->node)); - next = mas_slot_locked(mas, slots, 0); - if ((mte_dead_node(next))) { - mte_to_node(next)->type = mte_node_type(next); - next = mas_slot_locked(mas, slots, 1); - } + *enode = next; + node = mte_to_node(*enode); + type = mte_node_type(*enode); + slots = ma_slots(node, type); + next = mt_slot_locked(mt, slots, next_offset); + if ((mte_dead_node(next))) + next = mt_slot_locked(mt, slots, ++next_offset); - mte_set_node_dead(mas->node); - node->type = mte_node_type(mas->node); - mas_clear_meta(mas, node, node->type); + mte_set_node_dead(*enode); + node->type = type; node->piv_parent = prev; node->parent_slot = offset; - offset = 0; - prev = mas->node; + offset = next_offset; + next_offset = 0; + prev = *enode; } while (!mte_is_leaf(next)); return slots; } -static void mt_destroy_walk(struct maple_enode *enode, unsigned char ma_flags, +static void mt_destroy_walk(struct maple_enode *enode, struct maple_tree *mt, bool free) { void __rcu **slots; struct maple_node *node = mte_to_node(enode); struct maple_enode *start; - struct maple_tree mt; - MA_STATE(mas, &mt, 0, 0); - - mas.node = enode; if (mte_is_leaf(enode)) { node->type = mte_node_type(enode); goto free_leaf; } - ma_flags &= ~MT_FLAGS_LOCK_MASK; - mt_init_flags(&mt, ma_flags); - mas_lock(&mas); - - mte_to_node(enode)->ma_flags = ma_flags; start = enode; - slots = mas_destroy_descend(&mas, start, 0); - node = mas_mn(&mas); + slots = mte_destroy_descend(&enode, mt, start, 0); + node = mte_to_node(enode); // Updated in the above call. do { enum maple_type type; unsigned char offset; struct maple_enode *parent, *tmp; - node->type = mte_node_type(mas.node); - node->slot_len = mas_dead_leaves(&mas, slots, node->type); + node->slot_len = mte_dead_leaves(enode, mt, slots); if (free) mt_free_bulk(node->slot_len, slots); offset = node->parent_slot + 1; - mas.node = node->piv_parent; - if (mas_mn(&mas) == node) - goto start_slots_free; + enode = node->piv_parent; + if (mte_to_node(enode) == node) + goto free_leaf; - type = mte_node_type(mas.node); - slots = ma_slots(mte_to_node(mas.node), type); + type = mte_node_type(enode); + slots = ma_slots(mte_to_node(enode), type); if (offset >= mt_slots[type]) goto next; - tmp = mas_slot_locked(&mas, slots, offset); + tmp = mt_slot_locked(mt, slots, offset); if (mte_node_type(tmp) && mte_to_node(tmp)) { - parent = mas.node; - mas.node = tmp; - slots = mas_destroy_descend(&mas, parent, offset); + parent = enode; + enode = tmp; + slots = mte_destroy_descend(&enode, mt, parent, offset); } next: - node = mas_mn(&mas); - } while (start != mas.node); + node = mte_to_node(enode); + } while (start != enode); - node = mas_mn(&mas); - node->type = mte_node_type(mas.node); - node->slot_len = mas_dead_leaves(&mas, slots, node->type); + node = mte_to_node(enode); + node->slot_len = mte_dead_leaves(enode, mt, slots); if (free) mt_free_bulk(node->slot_len, slots); -start_slots_free: - mas_unlock(&mas); - free_leaf: if (free) mt_free_rcu(&node->rcu); else - mas_clear_meta(&mas, node, node->type); + mt_clear_meta(mt, node, node->type); } /* @@ -5688,10 +5692,10 @@ static inline void mte_destroy_walk(stru struct maple_node *node = mte_to_node(enode); if (mt_in_rcu(mt)) { - mt_destroy_walk(enode, mt->ma_flags, false); + mt_destroy_walk(enode, mt, false); call_rcu(&node->rcu, mt_free_walk); } else { - mt_destroy_walk(enode, mt->ma_flags, true); + mt_destroy_walk(enode, mt, true); } } _ Patches currently in -mm which might be from Liam.Howlett(a)oracle.com are

1 year, 8 months

1
0
0 0

[merged mm-hotfixes-stable] maple_tree-add-smp_rmb-to-dead-node-detection.patch removed from -mm tree

by Andrew Morton

The quilt patch titled Subject: maple_tree: add smp_rmb() to dead node detection has been removed from the -mm tree. Its filename was maple_tree-add-smp_rmb-to-dead-node-detection.patch This patch was dropped because it was merged into the mm-hotfixes-stable branch of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm ------------------------------------------------------ From: "Liam R. Howlett" <Liam.Howlett(a)oracle.com> Subject: maple_tree: add smp_rmb() to dead node detection Date: Mon, 27 Feb 2023 09:36:05 -0800 Add an smp_rmb() before reading the parent pointer to ensure that anything read from the node prior to the parent pointer hasn't been reordered ahead of this check. The is necessary for RCU mode. Link: https://lkml.kernel.org/r/20230227173632.3292573-7-surenb@google.com Fixes: 54a611b60590 ("Maple Tree: add new data structure") Signed-off-by: Liam R. Howlett <Liam.Howlett(a)oracle.com> Signed-off-by: Suren Baghdasaryan <surenb(a)google.com> Cc: <stable(a)vger.kernel.org> Signed-off-by: Andrew Morton <akpm(a)linux-foundation.org> --- lib/maple_tree.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) --- a/lib/maple_tree.c~maple_tree-add-smp_rmb-to-dead-node-detection +++ a/lib/maple_tree.c @@ -539,9 +539,11 @@ static inline struct maple_node *mte_par */ static inline bool ma_dead_node(const struct maple_node *node) { - struct maple_node *parent = (void *)((unsigned long) - node->parent & ~MAPLE_NODE_MASK); + struct maple_node *parent; + /* Do not reorder reads from the node prior to the parent check */ + smp_rmb(); + parent = (void *)((unsigned long) node->parent & ~MAPLE_NODE_MASK); return (parent == node); } @@ -556,6 +558,8 @@ static inline bool mte_dead_node(const s struct maple_node *parent, *node; node = mte_to_node(enode); + /* Do not reorder reads from the node prior to the parent check */ + smp_rmb(); parent = mte_parent(enode); return (parent == node); } _ Patches currently in -mm which might be from Liam.Howlett(a)oracle.com are

1 year, 8 months

1
0
0 0

[merged mm-hotfixes-stable] maple_tree-fix-write-memory-barrier-of-nodes-once-dead-for-rcu-mode.patch removed from -mm tree

by Andrew Morton

The quilt patch titled Subject: maple_tree: fix write memory barrier of nodes once dead for RCU mode has been removed from the -mm tree. Its filename was maple_tree-fix-write-memory-barrier-of-nodes-once-dead-for-rcu-mode.patch This patch was dropped because it was merged into the mm-hotfixes-stable branch of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm ------------------------------------------------------ From: "Liam R. Howlett" <Liam.Howlett(a)oracle.com> Subject: maple_tree: fix write memory barrier of nodes once dead for RCU mode Date: Mon, 27 Feb 2023 09:36:04 -0800 During the development of the maple tree, the strategy of freeing multiple nodes changed and, in the process, the pivots were reused to store pointers to dead nodes. To ensure the readers see accurate pivots, the writers need to mark the nodes as dead and call smp_wmb() to ensure any readers can identify the node as dead before using the pivot values. There were two places where the old method of marking the node as dead without smp_wmb() were being used, which resulted in RCU readers seeing the wrong pivot value before seeing the node was dead. Fix this race condition by using mte_set_node_dead() which has the smp_wmb() call to ensure the race is closed. Add a WARN_ON() to the ma_free_rcu() call to ensure all nodes being freed are marked as dead to ensure there are no other call paths besides the two updated paths. This is necessary for the RCU mode of the maple tree. Link: https://lkml.kernel.org/r/20230227173632.3292573-6-surenb@google.com Fixes: 54a611b60590 ("Maple Tree: add new data structure") Signed-off-by: Liam R. Howlett <Liam.Howlett(a)oracle.com> Signed-off-by: Suren Baghdasaryan <surenb(a)google.com> Cc: <stable(a)vger.kernel.org> Signed-off-by: Andrew Morton <akpm(a)linux-foundation.org> --- lib/maple_tree.c | 7 +++++-- tools/testing/radix-tree/maple.c | 16 ++++++++++++++++ 2 files changed, 21 insertions(+), 2 deletions(-) --- a/lib/maple_tree.c~maple_tree-fix-write-memory-barrier-of-nodes-once-dead-for-rcu-mode +++ a/lib/maple_tree.c @@ -185,7 +185,7 @@ static void mt_free_rcu(struct rcu_head */ static void ma_free_rcu(struct maple_node *node) { - node->parent = ma_parent_ptr(node); + WARN_ON(node->parent != ma_parent_ptr(node)); call_rcu(&node->rcu, mt_free_rcu); } @@ -1778,8 +1778,10 @@ static inline void mas_replace(struct ma rcu_assign_pointer(slots[offset], mas->node); } - if (!advanced) + if (!advanced) { + mte_set_node_dead(old_enode); mas_free(mas, old_enode); + } } /* @@ -4218,6 +4220,7 @@ static inline bool mas_wr_node_store(str done: mas_leaf_set_meta(mas, newnode, dst_pivots, maple_leaf_64, new_end); if (in_rcu) { + mte_set_node_dead(mas->node); mas->node = mt_mk_node(newnode, wr_mas->type); mas_replace(mas, false); } else { --- a/tools/testing/radix-tree/maple.c~maple_tree-fix-write-memory-barrier-of-nodes-once-dead-for-rcu-mode +++ a/tools/testing/radix-tree/maple.c @@ -108,6 +108,7 @@ static noinline void check_new_node(stru MT_BUG_ON(mt, mn->slot[1] != NULL); MT_BUG_ON(mt, mas_allocated(&mas) != 0); + mn->parent = ma_parent_ptr(mn); ma_free_rcu(mn); mas.node = MAS_START; mas_nomem(&mas, GFP_KERNEL); @@ -160,6 +161,7 @@ static noinline void check_new_node(stru MT_BUG_ON(mt, mas_allocated(&mas) != i); MT_BUG_ON(mt, !mn); MT_BUG_ON(mt, not_empty(mn)); + mn->parent = ma_parent_ptr(mn); ma_free_rcu(mn); } @@ -192,6 +194,7 @@ static noinline void check_new_node(stru MT_BUG_ON(mt, not_empty(mn)); MT_BUG_ON(mt, mas_allocated(&mas) != i - 1); MT_BUG_ON(mt, !mn); + mn->parent = ma_parent_ptr(mn); ma_free_rcu(mn); } @@ -210,6 +213,7 @@ static noinline void check_new_node(stru mn = mas_pop_node(&mas); MT_BUG_ON(mt, not_empty(mn)); MT_BUG_ON(mt, mas_allocated(&mas) != j - 1); + mn->parent = ma_parent_ptr(mn); ma_free_rcu(mn); } MT_BUG_ON(mt, mas_allocated(&mas) != 0); @@ -233,6 +237,7 @@ static noinline void check_new_node(stru MT_BUG_ON(mt, mas_allocated(&mas) != i - j); mn = mas_pop_node(&mas); MT_BUG_ON(mt, not_empty(mn)); + mn->parent = ma_parent_ptr(mn); ma_free_rcu(mn); MT_BUG_ON(mt, mas_allocated(&mas) != i - j - 1); } @@ -269,6 +274,7 @@ static noinline void check_new_node(stru mn = mas_pop_node(&mas); /* get the next node. */ MT_BUG_ON(mt, mn == NULL); MT_BUG_ON(mt, not_empty(mn)); + mn->parent = ma_parent_ptr(mn); ma_free_rcu(mn); } MT_BUG_ON(mt, mas_allocated(&mas) != 0); @@ -294,6 +300,7 @@ static noinline void check_new_node(stru mn = mas_pop_node(&mas2); /* get the next node. */ MT_BUG_ON(mt, mn == NULL); MT_BUG_ON(mt, not_empty(mn)); + mn->parent = ma_parent_ptr(mn); ma_free_rcu(mn); } MT_BUG_ON(mt, mas_allocated(&mas2) != 0); @@ -334,10 +341,12 @@ static noinline void check_new_node(stru MT_BUG_ON(mt, mas_allocated(&mas) != MAPLE_ALLOC_SLOTS + 2); mn = mas_pop_node(&mas); MT_BUG_ON(mt, not_empty(mn)); + mn->parent = ma_parent_ptr(mn); ma_free_rcu(mn); for (i = 1; i <= MAPLE_ALLOC_SLOTS + 1; i++) { mn = mas_pop_node(&mas); MT_BUG_ON(mt, not_empty(mn)); + mn->parent = ma_parent_ptr(mn); ma_free_rcu(mn); } MT_BUG_ON(mt, mas_allocated(&mas) != 0); @@ -375,6 +384,7 @@ static noinline void check_new_node(stru mas_node_count(&mas, i); /* Request */ mas_nomem(&mas, GFP_KERNEL); /* Fill request */ mn = mas_pop_node(&mas); /* get the next node. */ + mn->parent = ma_parent_ptr(mn); ma_free_rcu(mn); mas_destroy(&mas); @@ -382,10 +392,13 @@ static noinline void check_new_node(stru mas_node_count(&mas, i); /* Request */ mas_nomem(&mas, GFP_KERNEL); /* Fill request */ mn = mas_pop_node(&mas); /* get the next node. */ + mn->parent = ma_parent_ptr(mn); ma_free_rcu(mn); mn = mas_pop_node(&mas); /* get the next node. */ + mn->parent = ma_parent_ptr(mn); ma_free_rcu(mn); mn = mas_pop_node(&mas); /* get the next node. */ + mn->parent = ma_parent_ptr(mn); ma_free_rcu(mn); mas_destroy(&mas); } @@ -35369,6 +35382,7 @@ static noinline void check_prealloc(stru MT_BUG_ON(mt, allocated != 1 + height * 3); mn = mas_pop_node(&mas); MT_BUG_ON(mt, mas_allocated(&mas) != allocated - 1); + mn->parent = ma_parent_ptr(mn); ma_free_rcu(mn); MT_BUG_ON(mt, mas_preallocate(&mas, GFP_KERNEL) != 0); mas_destroy(&mas); @@ -35386,6 +35400,7 @@ static noinline void check_prealloc(stru mas_destroy(&mas); allocated = mas_allocated(&mas); MT_BUG_ON(mt, allocated != 0); + mn->parent = ma_parent_ptr(mn); ma_free_rcu(mn); MT_BUG_ON(mt, mas_preallocate(&mas, GFP_KERNEL) != 0); @@ -35756,6 +35771,7 @@ void farmer_tests(void) tree.ma_root = mt_mk_node(node, maple_leaf_64); mt_dump(&tree); + node->parent = ma_parent_ptr(node); ma_free_rcu(node); /* Check things that will make lockdep angry */ _ Patches currently in -mm which might be from Liam.Howlett(a)oracle.com are

1 year, 8 months

1
0
0 0

2024

2023

2022

2021

2020

2019

2018

2017

Linux-stable-mirror April 2023