Kernel

memcg reclaim demotion wrt. isolation (Michal Hocko) https://lkml.kernel.org/r/Y5idFucjKVbjatqc@dhcp22.suse.cz Currently pages allocated for demotion targets includes __GFP_KSWAPD_RECALIM, and therefore memcg triggered reclaim could result in reclaims of pages in different hierarchy. Michal suggests simple modification of it. [GIT PULL] MM updates for 6.2-rc1 (Andrew Morton) (1+ msgs) (Andrew Morton) https://lkml.kernel.org/r/20221213142614.525a9240506119abc3cef6b6@linux-foundation.org Andrew Morton sent the MM pull request for v6.2-rc1 to Linus Torvalds. It includes 28 DAMON patches for cleanup and DAMOS tried regions sysfs....

[PATCH v3] [mm-unstable] mm: Fix memcg reclaim on memory tiered systems (Mina Almasry) https://lkml.kernel.org/r/20221206023406.3182800-1-almasrymina@google.com There was a patch for enabling demotion via memory.reclaim memcg file. It made try_to_free_mem_cgroup_pages() behavior changed. As the changed behavior is somewhat the users of the caller didn’t expect, this patch fixes it. [PATCHv8 00/14] mm, x86/cc: Implement support for unaccepted memory (Kirill A. Shutemov) https://lkml.kernel.org/r/20221207014933.8435-1-kirill.shutemov@linux.intel.com Some confidential VM approaches like Intel TDX or AMD SEV-SNP requires the guests to accept memory that they will use before use....

[PATCH 1/2] lockdep: allow instrumenting lockdep.c with KMSAN (Alexander Potapenko) https://lkml.kernel.org/r/20221128094541.2645890-1-glider@google.com Lockdep and KMSAN are known to work together nowadays, though it could make kernel run slower. This patch fixes KMSAN’s false positives on wq_head->lock->dep_map. [PATCH] mm/huge_memory: add TRANSPARENT_HUGEPAGE_NEVER for THP (ye xingchen) https://lkml.kernel.org/r/202211301651462590168@zte.com.cn This patch adds a new kernel config called TRANSPARENT_HUGEPAGE_NEVER, which will set the value of /sys/kernel/mm/transparent_hugepage/enabled [never]. [RFC PATCH v2] mm: Add nodes= arg to memory.reclaim (Mina Almasry) https://lkml....

[PATCH 00/12] Introduce CONFIG_SLUB_TINY and deprecate SLOB (Vlastimil Babka) https://lkml.kernel.org/r/20221121171202.22080-1-vbabka@suse.cz Vlastimil has posted a mail[1] for a discussion about SLOB deprecation. As no one has objected at the idea so far, Vlastimil has sent a patchset for that. [1] https://lore.kernel.org/all/b35c3f82-f67b-2103-7d82-7a7ba7521439@suse.cz/ [RFC PATCH v1 0/4] Introduce merge identical pages mechanism (Alexey Romanov) https://lkml.kernel.org/r/20221121190020.66548-1-avromanov@sberdevices.ru This RFC patchset introduces a mechanism for merging identical compressed pages, which could be common in zram like use cases....

[PATCH] mm: Make drop_caches keep reclaiming on all nodes (Jan Kara) https://lkml.kernel.org/r/20221115123255.12559-1-jack@suse.cz drop_caches works in somewhat loop-until-no-progress manner. It could result in not so aggressive reclamation. This patch makes it to continue reclamation on other node until every node makes no progress. [RFC][PATCH 0/4] cachestat: a new syscall for page cache state of files (Nha Pham) https://lkml.kernel.org/r/20221115182901.2755368-1-nphamcs@gmail.com Users can know the cache state of files using mincore() system call, but it scales bad for large files....