kernel/git/stable/linux.git/mm/compaction.c, branch linux-6.7.y

mm, vmscan: prevent infinite loop for costly GFP_NOIO | __GFP_RETRY_MAYFAIL allocations

2024-04-03T13:11:40+00:00

commit 803de9000f334b771afacb6ff3e78622916668b0 upstream. Sven reports an infinite loop in __alloc_pages_slowpath() for costly order __GFP_RETRY_MAYFAIL allocations that are also GFP_NOIO. Such combination can happen in a suspend/resume context where a GFP_KERNEL allocation can have __GFP_IO masked out via gfp_allowed_mask. Quoting Sven: 1. try to do a "costly" allocation (order > PAGE_ALLOC_COSTLY_ORDER) with __GFP_RETRY_MAYFAIL set. 2. page alloc's __alloc_pages_slowpath tries to get a page from the freelist. This fails because there is nothing free of that costly order. 3. page alloc tries to reclaim by calling __alloc_pages_direct_reclaim, which bails out because a zone is ready to be compacted; it pretends to have made a single page of progress. 4. page alloc tries to compact, but this always bails out early because __GFP_IO is not set (it's not passed by the snd allocator, and even if it were, we are suspending so the __GFP_IO flag would be cleared anyway). 5. page alloc believes reclaim progress was made (because of the pretense in item 3) and so it checks whether it should retry compaction. The compaction retry logic thinks it should try again, because: a) reclaim is needed because of the early bail-out in item 4 b) a zonelist is suitable for compaction 6. goto 2. indefinite stall. (end quote) The immediate root cause is confusing the COMPACT_SKIPPED returned from __alloc_pages_direct_compact() (step 4) due to lack of __GFP_IO to be indicating a lack of order-0 pages, and in step 5 evaluating that in should_compact_retry() as a reason to retry, before incrementing and limiting the number of retries. There are however other places that wrongly assume that compaction can happen while we lack __GFP_IO. To fix this, introduce gfp_compaction_allowed() to abstract the __GFP_IO evaluation and switch the open-coded test in try_to_compact_pages() to use it. Also use the new helper in: - compaction_ready(), which will make reclaim not bail out in step 3, so there's at least one attempt to actually reclaim, even if chances are small for a costly order - in_reclaim_compaction() which will make should_continue_reclaim() return false and we don't over-reclaim unnecessarily - in __alloc_pages_slowpath() to set a local variable can_compact, which is then used to avoid retrying reclaim/compaction for costly allocations (step 5) if we can't compact and also to skip the early compaction attempt that we do in some cases Link: https://lkml.kernel.org/r/20240221114357.13655-2-vbabka@suse.cz Fixes: 3250845d0526 ("Revert "mm, oom: prevent premature OOM killer invocation for high order request"") Signed-off-by: Vlastimil Babka Reported-by: Sven van Ashbrook Closes: https://lore.kernel.org/all/CAG-rBihs_xMKb3wrMO1%2B-%2Bp4fowP9oy1pa_OTkfxBzPUVOZF%2Bg@mail.gmail.com/ Tested-by: Karthikeyan Ramasubramanian Cc: Brian Geffon Cc: Curtis Malainey Cc: Jaroslav Kysela Cc: Mel Gorman Cc: Michal Hocko Cc: Takashi Iwai Cc: Signed-off-by: Andrew Morton Signed-off-by: Greg Kroah-Hartman

mm/compaction: factor out code to test if we should run compaction for target order

2023-10-04T17:32:19+00:00

We always do zone_watermark_ok check and compaction_suitable check together to test if compaction for target order should be ran. Factor these code out to remove repeat code. Link: https://lkml.kernel.org/r/20230901155141.249860-7-shikemeng@huaweicloud.com Signed-off-by: Kemeng Shi Reviewed-by: Baolin Wang Cc: David Hildenbrand Cc: Matthew Wilcox (Oracle) Cc: Mel Gorman Signed-off-by: Andrew Morton

mm/compaction: improve comment of is_via_compact_memory

2023-10-04T17:32:19+00:00

We do proactive compaction with order == -1 via 1. /proc/sys/vm/compact_memory 2. /sys/devices/system/node/nodex/compact 3. /proc/sys/vm/compaction_proactiveness Add missed situation in which order == -1. Link: https://lkml.kernel.org/r/20230901155141.249860-6-shikemeng@huaweicloud.com Signed-off-by: Kemeng Shi Reviewed-by: Baolin Wang Acked-by: Mel Gorman Cc: David Hildenbrand Cc: Matthew Wilcox (Oracle) Signed-off-by: Andrew Morton

mm/compaction: remove repeat compact_blockskip_flush check in reset_isolation_suitable

2023-10-04T17:32:19+00:00

We have compact_blockskip_flush check in __reset_isolation_suitable, just remove repeat check before __reset_isolation_suitable in compact_blockskip_flush. Link: https://lkml.kernel.org/r/20230901155141.249860-5-shikemeng@huaweicloud.com Signed-off-by: Kemeng Shi Reviewed-by: Baolin Wang Acked-by: Mel Gorman Cc: David Hildenbrand Cc: Matthew Wilcox (Oracle) Signed-off-by: Andrew Morton

mm/compaction: correctly return failure with bogus compound_order in strict mode

2023-10-04T17:32:19+00:00

In strict mode, we should return 0 if there is any hole in pageblock. If we successfully isolated pages at beginning at pageblock and then have a bogus compound_order outside pageblock in next page. We will abort search loop with blockpfn > end_pfn. Although we will limit blockpfn to end_pfn, we will treat it as a successful isolation in strict mode as blockpfn is not < end_pfn and return partial isolated pages. Then isolate_freepages_range may success unexpectly with hole in isolated range. Link: https://lkml.kernel.org/r/20230901155141.249860-4-shikemeng@huaweicloud.com Fixes: 9fcd6d2e052e ("mm, compaction: skip compound pages by order in free scanner") Signed-off-by: Kemeng Shi Reviewed-by: Baolin Wang Acked-by: Mel Gorman Cc: David Hildenbrand Cc: Matthew Wilcox (Oracle) Signed-off-by: Andrew Morton

mm/compaction: call list_is_{first}/{last} more intuitively in move_freelist_{head}/{tail}

2023-10-04T17:32:19+00:00

We use move_freelist_head after list_for_each_entry_reverse to skip recent pages. And there is no need to do actual move if all freepages are searched in list_for_each_entry_reverse, e.g. freepage point to first page in freelist. It's more intuitively to call list_is_first with list entry as the first argument and list head as the second argument to check if list entry is the first list entry instead of call list_is_last with list entry and list head passed in reverse. Similarly, call list_is_last in move_freelist_tail is more intuitively. Link: https://lkml.kernel.org/r/20230901155141.249860-3-shikemeng@huaweicloud.com Signed-off-by: Kemeng Shi Reviewed-by: Baolin Wang Acked-by: Mel Gorman Cc: David Hildenbrand Cc: Matthew Wilcox (Oracle) Signed-off-by: Andrew Morton

mm/compaction: use correct list in move_freelist_{head}/{tail}

2023-10-04T17:32:19+00:00

Patch series "Fixes and cleanups to compaction", v3. This is a series to do fix and clean up to compaction. Patch 1-2 fix and clean up freepage list operation. Patch 3-4 fix and clean up isolation of freepages Patch 7 factor code to check if compaction is needed for allocation order. More details can be found in respective patches. This patch (of 6): The freepage is chained with buddy_list in freelist head. Use buddy_list instead of lru to correct the list operation. Link: https://lkml.kernel.org/r/20230901155141.249860-1-shikemeng@huaweicloud.com Link: https://lkml.kernel.org/r/20230901155141.249860-2-shikemeng@huaweicloud.com Signed-off-by: Kemeng Shi Reviewed-by: Baolin Wang Acked-by: Mel Gorman Cc: David Hildenbrand Cc: Matthew Wilcox (Oracle) Signed-off-by: Andrew Morton

merge mm-hotfixes-stable into mm-stable to pick up depended-upon changes

2023-08-21T21:26:20+00:00

mm/compaction: remove unused parameter pgdata of fragmentation_score_wmark

2023-08-21T20:37:50+00:00

Parameter pgdat is not used in fragmentation_score_wmark. Just remove it. Link: https://lkml.kernel.org/r/20230809094910.3092446-1-shikemeng@huaweicloud.com Signed-off-by: Kemeng Shi Reviewed-by: David Hildenbrand Acked-by: Mel Gorman Reviewed-by: Baolin Wang Cc: Matthew Wilcox Signed-off-by: Andrew Morton

mm/compaction: only set skip flag if cc->no_set_skip_hint is false

2023-08-21T20:37:43+00:00

Keep the same logic as update_pageblock_skip, only set skip if no_set_skip_hint is false which is more reasonable. Link: https://lkml.kernel.org/r/20230804110454.2935878-9-shikemeng@huaweicloud.com Signed-off-by: Kemeng Shi Reviewed-by: David Hildenbrand Cc: Baolin Wang Cc: Mel Gorman Signed-off-by: Andrew Morton