From c288983dddf714216428774e022ad78f48dd8cb1 Mon Sep 17 00:00:00 2001 From: Tetsuo Handa Date: Fri, 2 Jun 2017 14:46:31 -0700 Subject: mm/page_alloc.c: make sure OOM victim can try allocations with no watermarks once Roman Gushchin has reported that the OOM killer can trivially selects next OOM victim when a thread doing memory allocation from page fault path was selected as first OOM victim. allocate invoked oom-killer: gfp_mask=0x14280ca(GFP_HIGHUSER_MOVABLE|__GFP_ZERO), nodemask=(null), order=0, oom_score_adj=0 allocate cpuset=/ mems_allowed=0 CPU: 1 PID: 492 Comm: allocate Not tainted 4.12.0-rc1-mm1+ #181 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS Ubuntu-1.8.2-1ubuntu1 04/01/2014 Call Trace: oom_kill_process+0x219/0x3e0 out_of_memory+0x11d/0x480 __alloc_pages_slowpath+0xc84/0xd40 __alloc_pages_nodemask+0x245/0x260 alloc_pages_vma+0xa2/0x270 __handle_mm_fault+0xca9/0x10c0 handle_mm_fault+0xf3/0x210 __do_page_fault+0x240/0x4e0 trace_do_page_fault+0x37/0xe0 do_async_page_fault+0x19/0x70 async_page_fault+0x28/0x30 ... Out of memory: Kill process 492 (allocate) score 899 or sacrifice child Killed process 492 (allocate) total-vm:2052368kB, anon-rss:1894576kB, file-rss:4kB, shmem-rss:0kB allocate: page allocation failure: order:0, mode:0x14280ca(GFP_HIGHUSER_MOVABLE|__GFP_ZERO), nodemask=(null) allocate cpuset=/ mems_allowed=0 CPU: 1 PID: 492 Comm: allocate Not tainted 4.12.0-rc1-mm1+ #181 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS Ubuntu-1.8.2-1ubuntu1 04/01/2014 Call Trace: __alloc_pages_slowpath+0xd32/0xd40 __alloc_pages_nodemask+0x245/0x260 alloc_pages_vma+0xa2/0x270 __handle_mm_fault+0xca9/0x10c0 handle_mm_fault+0xf3/0x210 __do_page_fault+0x240/0x4e0 trace_do_page_fault+0x37/0xe0 do_async_page_fault+0x19/0x70 async_page_fault+0x28/0x30 ... oom_reaper: reaped process 492 (allocate), now anon-rss:0kB, file-rss:0kB, shmem-rss:0kB ... allocate invoked oom-killer: gfp_mask=0x0(), nodemask=(null), order=0, oom_score_adj=0 allocate cpuset=/ mems_allowed=0 CPU: 1 PID: 492 Comm: allocate Not tainted 4.12.0-rc1-mm1+ #181 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS Ubuntu-1.8.2-1ubuntu1 04/01/2014 Call Trace: oom_kill_process+0x219/0x3e0 out_of_memory+0x11d/0x480 pagefault_out_of_memory+0x68/0x80 mm_fault_error+0x8f/0x190 ? handle_mm_fault+0xf3/0x210 __do_page_fault+0x4b2/0x4e0 trace_do_page_fault+0x37/0xe0 do_async_page_fault+0x19/0x70 async_page_fault+0x28/0x30 ... Out of memory: Kill process 233 (firewalld) score 10 or sacrifice child Killed process 233 (firewalld) total-vm:246076kB, anon-rss:20956kB, file-rss:0kB, shmem-rss:0kB There is a race window that the OOM reaper completes reclaiming the first victim's memory while nothing but mutex_trylock() prevents the first victim from calling out_of_memory() from pagefault_out_of_memory() after memory allocation for page fault path failed due to being selected as an OOM victim. This is a side effect of commit 9a67f6488eca926f ("mm: consolidate GFP_NOFAIL checks in the allocator slowpath") because that commit silently changed the behavior from /* Avoid allocations with no watermarks from looping endlessly */ to /* * Give up allocations without trying memory reserves if selected * as an OOM victim */ in __alloc_pages_slowpath() by moving the location to check TIF_MEMDIE flag. I have noticed this change but I didn't post a patch because I thought it is an acceptable change other than noise by warn_alloc() because !__GFP_NOFAIL allocations are allowed to fail. But we overlooked that failing memory allocation from page fault path makes difference due to the race window explained above. While it might be possible to add a check to pagefault_out_of_memory() that prevents the first victim from calling out_of_memory() or remove out_of_memory() from pagefault_out_of_memory(), changing pagefault_out_of_memory() does not suppress noise by warn_alloc() when allocating thread was selected as an OOM victim. There is little point with printing similar backtraces and memory information from both out_of_memory() and warn_alloc(). Instead, if we guarantee that current thread can try allocations with no watermarks once when current thread looping inside __alloc_pages_slowpath() was selected as an OOM victim, we can follow "who can use memory reserves" rules and suppress noise by warn_alloc() and prevent memory allocations from page fault path from calling pagefault_out_of_memory(). If we take the comment literally, this patch would do - if (test_thread_flag(TIF_MEMDIE)) - goto nopage; + if (alloc_flags == ALLOC_NO_WATERMARKS || (gfp_mask & __GFP_NOMEMALLOC)) + goto nopage; because gfp_pfmemalloc_allowed() returns false if __GFP_NOMEMALLOC is given. But if I recall correctly (I couldn't find the message), the condition is meant to apply to only OOM victims despite the comment. Therefore, this patch preserves TIF_MEMDIE check. Fixes: 9a67f6488eca926f ("mm: consolidate GFP_NOFAIL checks in the allocator slowpath") Link: http://lkml.kernel.org/r/201705192112.IAF69238.OQOHSJLFOFFMtV@I-love.SAKURA.ne.jp Signed-off-by: Tetsuo Handa Reported-by: Roman Gushchin Tested-by: Roman Gushchin Acked-by: Michal Hocko Cc: Johannes Weiner Cc: Vladimir Davydov Cc: [4.11] Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/page_alloc.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'mm/page_alloc.c') diff --git a/mm/page_alloc.c b/mm/page_alloc.c index f9e450c6b6e4..b7a6f583a373 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -3870,7 +3870,9 @@ retry: goto got_pg; /* Avoid allocations with no watermarks from looping endlessly */ - if (test_thread_flag(TIF_MEMDIE)) + if (test_thread_flag(TIF_MEMDIE) && + (alloc_flags == ALLOC_NO_WATERMARKS || + (gfp_mask & __GFP_NOMEMALLOC))) goto nopage; /* Retry as long as the OOM killer is making progress */ -- cgit From 864b9a393dcb5aed09b8fd31b9bbda0fdda99374 Mon Sep 17 00:00:00 2001 From: Michal Hocko Date: Fri, 2 Jun 2017 14:46:49 -0700 Subject: mm: consider memblock reservations for deferred memory initialization sizing We have seen an early OOM killer invocation on ppc64 systems with crashkernel=4096M: kthreadd invoked oom-killer: gfp_mask=0x16040c0(GFP_KERNEL|__GFP_COMP|__GFP_NOTRACK), nodemask=7, order=0, oom_score_adj=0 kthreadd cpuset=/ mems_allowed=7 CPU: 0 PID: 2 Comm: kthreadd Not tainted 4.4.68-1.gd7fe927-default #1 Call Trace: dump_stack+0xb0/0xf0 (unreliable) dump_header+0xb0/0x258 out_of_memory+0x5f0/0x640 __alloc_pages_nodemask+0xa8c/0xc80 kmem_getpages+0x84/0x1a0 fallback_alloc+0x2a4/0x320 kmem_cache_alloc_node+0xc0/0x2e0 copy_process.isra.25+0x260/0x1b30 _do_fork+0x94/0x470 kernel_thread+0x48/0x60 kthreadd+0x264/0x330 ret_from_kernel_thread+0x5c/0xa4 Mem-Info: active_anon:0 inactive_anon:0 isolated_anon:0 active_file:0 inactive_file:0 isolated_file:0 unevictable:0 dirty:0 writeback:0 unstable:0 slab_reclaimable:5 slab_unreclaimable:73 mapped:0 shmem:0 pagetables:0 bounce:0 free:0 free_pcp:0 free_cma:0 Node 7 DMA free:0kB min:0kB low:0kB high:0kB active_anon:0kB inactive_anon:0kB active_file:0kB inactive_file:0kB unevictable:0kB isolated(anon):0kB isolated(file):0kB present:52428800kB managed:110016kB mlocked:0kB dirty:0kB writeback:0kB mapped:0kB shmem:0kB slab_reclaimable:320kB slab_unreclaimable:4672kB kernel_stack:1152kB pagetables:0kB unstable:0kB bounce:0kB free_pcp:0kB local_pcp:0kB free_cma:0kB writeback_tmp:0kB pages_scanned:0 all_unreclaimable? yes lowmem_reserve[]: 0 0 0 0 Node 7 DMA: 0*64kB 0*128kB 0*256kB 0*512kB 0*1024kB 0*2048kB 0*4096kB 0*8192kB 0*16384kB = 0kB 0 total pagecache pages 0 pages in swap cache Swap cache stats: add 0, delete 0, find 0/0 Free swap = 0kB Total swap = 0kB 819200 pages RAM 0 pages HighMem/MovableOnly 817481 pages reserved 0 pages cma reserved 0 pages hwpoisoned the reason is that the managed memory is too low (only 110MB) while the rest of the the 50GB is still waiting for the deferred intialization to be done. update_defer_init estimates the initial memoty to initialize to 2GB at least but it doesn't consider any memory allocated in that range. In this particular case we've had Reserving 4096MB of memory at 128MB for crashkernel (System RAM: 51200MB) so the low 2GB is mostly depleted. Fix this by considering memblock allocations in the initial static initialization estimation. Move the max_initialise to reset_deferred_meminit and implement a simple memblock_reserved_memory helper which iterates all reserved blocks and sums the size of all that start below the given address. The cumulative size is than added on top of the initial estimation. This is still not ideal because reset_deferred_meminit doesn't consider holes and so reservation might be above the initial estimation whihch we ignore but let's make the logic simpler until we really need to handle more complicated cases. Fixes: 3a80a7fa7989 ("mm: meminit: initialise a subset of struct pages if CONFIG_DEFERRED_STRUCT_PAGE_INIT is set") Link: http://lkml.kernel.org/r/20170531104010.GI27783@dhcp22.suse.cz Signed-off-by: Michal Hocko Acked-by: Mel Gorman Tested-by: Srikar Dronamraju Cc: [4.2+] Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/page_alloc.c | 33 ++++++++++++++++++++++----------- 1 file changed, 22 insertions(+), 11 deletions(-) (limited to 'mm/page_alloc.c') diff --git a/mm/page_alloc.c b/mm/page_alloc.c index b7a6f583a373..2302f250d6b1 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -292,6 +292,26 @@ int page_group_by_mobility_disabled __read_mostly; #ifdef CONFIG_DEFERRED_STRUCT_PAGE_INIT static inline void reset_deferred_meminit(pg_data_t *pgdat) { + unsigned long max_initialise; + unsigned long reserved_lowmem; + + /* + * Initialise at least 2G of a node but also take into account that + * two large system hashes that can take up 1GB for 0.25TB/node. + */ + max_initialise = max(2UL << (30 - PAGE_SHIFT), + (pgdat->node_spanned_pages >> 8)); + + /* + * Compensate the all the memblock reservations (e.g. crash kernel) + * from the initial estimation to make sure we will initialize enough + * memory to boot. + */ + reserved_lowmem = memblock_reserved_memory_within(pgdat->node_start_pfn, + pgdat->node_start_pfn + max_initialise); + max_initialise += reserved_lowmem; + + pgdat->static_init_size = min(max_initialise, pgdat->node_spanned_pages); pgdat->first_deferred_pfn = ULONG_MAX; } @@ -314,20 +334,11 @@ static inline bool update_defer_init(pg_data_t *pgdat, unsigned long pfn, unsigned long zone_end, unsigned long *nr_initialised) { - unsigned long max_initialise; - /* Always populate low zones for address-contrained allocations */ if (zone_end < pgdat_end_pfn(pgdat)) return true; - /* - * Initialise at least 2G of a node but also take into account that - * two large system hashes that can take up 1GB for 0.25TB/node. - */ - max_initialise = max(2UL << (30 - PAGE_SHIFT), - (pgdat->node_spanned_pages >> 8)); - (*nr_initialised)++; - if ((*nr_initialised > max_initialise) && + if ((*nr_initialised > pgdat->static_init_size) && (pfn & (PAGES_PER_SECTION - 1)) == 0) { pgdat->first_deferred_pfn = pfn; return false; @@ -6138,7 +6149,6 @@ void __paginginit free_area_init_node(int nid, unsigned long *zones_size, /* pg_data_t should be reset to zero when it's allocated */ WARN_ON(pgdat->nr_zones || pgdat->kswapd_classzone_idx); - reset_deferred_meminit(pgdat); pgdat->node_id = nid; pgdat->node_start_pfn = node_start_pfn; pgdat->per_cpu_nodestats = NULL; @@ -6160,6 +6170,7 @@ void __paginginit free_area_init_node(int nid, unsigned long *zones_size, (unsigned long)pgdat->node_mem_map); #endif + reset_deferred_meminit(pgdat); free_area_init_core(pgdat); } -- cgit