LKML News v6.10-rc5
[GIT PULL] hotfixes for 6.10-rc5 (Andrew Morton)
https://lore.kernel.org/20240617114712.45d4743f8bacb832dea4b5a9@linux-foundation.org
Andrew posted hotfixes pull request to Linus Torvalds.
[PATCH] Add accumulated call counter for memory allocation profiling (David Wang)
https://lore.kernel.org/20240617153250.9079-1-00107082@163.com
David sent a patch for adding accumulated call counter for memory allocation profiling. He argues this could help performance behaviors analysis.
[RFC PATCH v1 0/5] Alternative mTHP swap allocator improvements (Ryan Roberts)
https://lore.kernel.org/20240618232648.4090299-1-ryan.roberts@arm.com
Ryan posted this proof-of-concept patchset for achieving 0% fallback at mTHP swap.
Re: [PATCHSET v6] sched: Implement BPF extensible scheduler class (Thomas Gleixner)
https://lore.kernel.org/878qz0pcir.ffs@tglx
Thomas asked Linus to postpone merge of sched-ext for about three months. Linus said “no”, and Thomas posted this not-short response with a proposal of maintainers summit topic.
[PATCH v2 0/7] static key support for error injection functions (Vlastimil Babka)
https://lore.kernel.org/20240620-fault-injection-statickeys-v2-0-e23947d3d84b@suse.cz
Vlastimil drops the RFC tag and posted the second version of patch series that adds static key support for error injection functions.
[PATCH] mm/page_alloc: add one PCP list for THP (yangge)
https://lore.kernel.org/1718801672-30152-1-git-send-email-yangge1116@126.com
When large CMA memory is configured, starting a virtual machine with device passthrough gets stuck, and it also make negative impacts on CMA itself. This patch fixes the problem by adding one PCP list for THP.
[PATCH v1] memory tier: consolidate the initialization of memory tiers (Ho-Ren (Jack) Chuang)
https://lore.kernel.org/20240621044833.3953055-1-horen.chuang@linux.dev
Ho-Ren sent a patch consolidating memory tier initialization, which is
currently distributed in memory_tier_init()
and memory_tier_late_init()
.
[PATCH v2 0/1] tools/mm: Introduce a tool to assess swap entry allocation for thp_swapout (Barry Song)
https://lore.kernel.org/20240622071231.576056-1-21cnbao@gmail.com
Barry introduces a tool for accessing swap entry allocation, so that it can be used to assess efficiency of the underlying kernel mechanism.
MM global locks as core counts quadruple (David Rientjes)
https://lore.kernel.org/07e7d078-0c9d-6a1f-1ab5-295f86974b72@google.com
David and Namhyung found some global locks that used for memory managment have some high contention, and worrying about it for upcoming era which will have more CPU cores. Further they are asking how people testing it, and if someone working on it.
[PATCH] Fix initializing a static union variable (Yabin Cui)
https://lore.kernel.org/20240620181736.1270455-1-yabinc@google.com
Yabin found static union variable initialzied with {}
can incompletely
initialized according to C11 standard and some clang implementations. Based on
it, he fixes one such wrong use case. Some people including Linus are saying
it is better to fix combilers instead of only this single point of bug.
Linux 6.10-rc5 (Linus Torvalds)
https://lore.kernel.org/CAHk-=wiMMT-2pfJZ8ckbnGTbSHy5mvvE=+-MyA_ARPUqiy_C1w@mail.gmail.com
Another release candidate for Linux v6.10 is released. Linus says it continues the trend of being fairly calm.