The quilt patch titled Subject: mm: zswap: fix data loss on SWP_SYNCHRONOUS_IO devices has been removed from the -mm tree. Its filename was mm-zswap-fix-data-loss-on-swp_synchronous_io-devices.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: Johannes Weiner hannes@cmpxchg.org Subject: mm: zswap: fix data loss on SWP_SYNCHRONOUS_IO devices Date: Sun, 24 Mar 2024 17:04:47 -0400
Zhongkun He reports data corruption when combining zswap with zram.
The issue is the exclusive loads we're doing in zswap. They assume that all reads are going into the swapcache, which can assume authoritative ownership of the data and so the zswap copy can go.
However, zram files are marked SWP_SYNCHRONOUS_IO, and faults will try to bypass the swapcache. This results in an optimistic read of the swap data into a page that will be dismissed if the fault fails due to races. In this case, zswap mustn't drop its authoritative copy.
Link: https://lore.kernel.org/all/CACSyD1N+dUvsu8=zV9P691B9bVq33erwOXNTmEaUbi9DrDe... Fixes: b9c91c43412f ("mm: zswap: support exclusive loads") Link: https://lkml.kernel.org/r/20240324210447.956973-1-hannes@cmpxchg.org Signed-off-by: Johannes Weiner hannes@cmpxchg.org Reported-by: Zhongkun He hezhongkun.hzk@bytedance.com Tested-by: Zhongkun He hezhongkun.hzk@bytedance.com Acked-by: Yosry Ahmed yosryahmed@google.com Acked-by: Barry Song baohua@kernel.org Reviewed-by: Chengming Zhou chengming.zhou@linux.dev Reviewed-by: Nhat Pham nphamcs@gmail.com Acked-by: Chris Li chrisl@kernel.org Cc: stable@vger.kernel.org [6.5+] Signed-off-by: Andrew Morton akpm@linux-foundation.org ---
mm/zswap.c | 23 +++++++++++++++++++---- 1 file changed, 19 insertions(+), 4 deletions(-)
--- a/mm/zswap.c~mm-zswap-fix-data-loss-on-swp_synchronous_io-devices +++ a/mm/zswap.c @@ -1636,6 +1636,7 @@ bool zswap_load(struct folio *folio) swp_entry_t swp = folio->swap; pgoff_t offset = swp_offset(swp); struct page *page = &folio->page; + bool swapcache = folio_test_swapcache(folio); struct zswap_tree *tree = swap_zswap_tree(swp); struct zswap_entry *entry; u8 *dst; @@ -1648,7 +1649,20 @@ bool zswap_load(struct folio *folio) spin_unlock(&tree->lock); return false; } - zswap_rb_erase(&tree->rbroot, entry); + /* + * When reading into the swapcache, invalidate our entry. The + * swapcache can be the authoritative owner of the page and + * its mappings, and the pressure that results from having two + * in-memory copies outweighs any benefits of caching the + * compression work. + * + * (Most swapins go through the swapcache. The notable + * exception is the singleton fault on SWP_SYNCHRONOUS_IO + * files, which reads into a private page and may free it if + * the fault fails. We remain the primary owner of the entry.) + */ + if (swapcache) + zswap_rb_erase(&tree->rbroot, entry); spin_unlock(&tree->lock);
if (entry->length) @@ -1663,9 +1677,10 @@ bool zswap_load(struct folio *folio) if (entry->objcg) count_objcg_event(entry->objcg, ZSWPIN);
- zswap_entry_free(entry); - - folio_mark_dirty(folio); + if (swapcache) { + zswap_entry_free(entry); + folio_mark_dirty(folio); + }
return true; } _
Patches currently in -mm which might be from hannes@cmpxchg.org are
mm-zswap-optimize-zswap-pool-size-tracking.patch mm-zpool-return-pool-size-in-pages.patch mm-page_alloc-remove-pcppage-migratetype-caching.patch mm-page_alloc-optimize-free_unref_folios.patch mm-page_alloc-fix-up-block-types-when-merging-compatible-blocks.patch mm-page_alloc-move-free-pages-when-converting-block-during-isolation.patch mm-page_alloc-fix-move_freepages_block-range-error.patch mm-page_alloc-fix-freelist-movement-during-block-conversion.patch mm-page_alloc-close-migratetype-race-between-freeing-and-stealing.patch mm-page_isolation-prepare-for-hygienic-freelists.patch mm-page_isolation-prepare-for-hygienic-freelists-fix.patch mm-page_alloc-consolidate-free-page-accounting.patch