[RFC PATCH] mm/oom: detect and kill task which has allocation forbidden by cpuset limit (Feng Tang) https://lkml.kernel.org/r/1630399085-70431-1-git-send-email-feng.tang@intel.com
Docker needs to allocate memory in non-movable zones for creation of an OS (e.g., GFP_HIGHUSER). However, if it is binded to a cpuset which has only movable zone, the allocation fails and OOM storm attacks other innocent processes. This patch detects the case and kill only the allocation failing process.
Folios: Can we resolve this please?
[GIT PULL] Memory folios for v5.15 (Matthew Wilcox) https://lkml.kernel.org/r/YSPwmNNuuQhXNToQ@casper.infradead.org
Matthew Wilcox sent a pull request for his page folio works to Linus Torvalds. Linus Torvalds replied he wants this to work in reverse. That is, calling head pages as normal pages but calling tail pages something other, as the term, page, is more common and de-facto standard in the industry.
mmotm 2021-08-23-16-42 uploaded (Andrew Morton) https://lkml.kernel.org/r/20210823234329.H9WD-du1K%akpm@linux-foundation.org
MMOTM is rebased on v5.
[PATCH v2 0/3] virtio-mem: disallow mapping virtio-mem memory via /dev/mem (David Hildenbrand) https://lkml.kernel.org/r/20210816142505.28359-1-david@redhat.com
Once the virtio-mem driver is plugging/unplugging memory of virtual machines, accessing the memory region via /dev/mem from the user space makes no sense. To avoid such insane use case, this patchset adds a basic infrastructure to exclude some physical memory regions completely from /dev/mem.
[PATCH v2 00/61] Introducing the Maple Tree (Liam Howlett) https://lkml.kernel.org/r/20210817154651.1570984-1-Liam.Howlett@oracle.com
This patchset introduces a new data structure called maple tree, which intended to replace the rbtree for vm_area_struct.
[PATCH v9 1/2] mm: introduce process_mrelease system call (Suren Baghdasaryan) https://lkml.kernel.org/r/20210809185259.405936-1-surenb@google.com
9th version of the patchset which introducing a new system call for releasing memory of a process in a speedy manner.
mmotm 2021-08-09-19-18 uploaded (Andrew Morton) https://lkml.kernel.org/r/20210810021934.XcpwGUEMn%akpm@linux-foundation.org
-mm tree updated.
[PATCH 0/5] x86: Impplement support for unaccepted memory (Kirill A. Shutemov) https://lkml.kernel.org/r/20210810062626.1012-1-kirill.shutemov@linux.intel.com
Some platforms like Intel TDX or AMD SEV-SNP require memory to be accepted before it can be used by the guest.
[PATCH 0/3] mm: optimize thp for reclaim and migration (Yu Zhao) https://lkml.kernel.org/r/20210731063938.1391602-1-yuzhao@google.com
This patchset drops subpages of THPs which contain only zero contents when splitting those, to reduce memory pressure overheads due to THP internal fragmentations.
[PATCH v4 1/2] mm: introduce process_mrelease system call (Suren Baghdasaryan) https://lkml.kernel.org/r/20210802221431.2251210-1-surenb@google.com
This patchset introduces a new system call, namely process_mrelease(). It can be used to quickly release memory of a process.
[PATCH v7 0/5] Introduce multi-preference mempolicy (Feng Tang) https://lkml.
mmotm 2021-07-28-18-24 uploaded (Andrew Morton) https://lkml.kernel.org/r/20210729012524.S8-WP%akpm@linux-foundation.org
Andrew Morton is updating the -mm tree.
incoming (Andrew Morton) https://lkml.kernel.org/r/20210729145259.24681c326dc3ed18194cf9e5@linux-foundation.org
Seven fixup patches from Andrew Morton.
[PATCH 0/3] mm: optimize thp for reclaim and migration (Yu Zhao) https://lkml.kernel.org/r/20210731063938.1391602-1-yuzhao@google.com
This patch optimizes THP split for reclaim or migration by dropping subpages that contain only zeroes.
Linux 5.14-rc4 (Linus Torvalds) https://lkml.kernel.org/r/CAHk-=whv0g5YqoRXc8oxqJX3r+GP1dN3ia_Qg_V6UTjhi8vg8w@mail.gmail.com
After one week, entirely normal rc4 is released. 5.14 is not so far now.
[PATCH 0/7] Free user PTE page table pages (Qi Zheng) https://lkml.kernel.org/r/20210718043034.76431-1-zhengqi.arch@bytedance.com
This patchset frees user PTE page table pages when all PTE entries are empty, to reduce the memory overhead due to page tables.
[PATCH v15 00/17] Folio support in block + iomap layers (Matthew Wilcox) https://lkml.kernel.org/r/20210719184001.1750630-1-willy@infradead.org
Seems memory folios patchset is now merged in -mm tree and -next tree. This patchset further updates iomap to use folios.
Runtime Memory Validation in Intel-TDX and AMD-SNP (Joerg Roedel) https://lkml.
[PATCH v6 0/6] Introduce multi-preference mempolicy (Feng Tang) https://lkml.kernel.org/r/1626077374-81682-1-git-send-email-feng.tang@intel.com
Six version of the patchset that allows applications to set multiple preferred memory policy.
[PATCH v1] madvise.2: Document MADV_POPULATE_READ and MADV_POPULATE_WRITE (David Hildenbrand) https://lkml.kernel.org/r/20210712083917.16361-1-david@redhat.com
This patch documents the new madvise() options, MADV_POPULATE_(READ|WRITE).
Request for folios (David Howells) https://lkml.kernel.org/r/3398985.1626104609@warthog.procyon.org.uk
David is working for improving local filesystem’s caching and network filesystem support library. He thinks the folio work would make his work easier, though it also makes some of his work challenging.
incoming (Andrew Morton) https://lkml.kernel.org/r/20210628193256.008961950a714730751c1423@linux-foundation.org
The first pull request from Andrew Morton for the v5.14 merge window.
[PATCH v3 00/18] Folio conversion of memcg (Matthew Wilcox) https://lkml.kernel.org/r/20210630040034.1155892-1-willy@infradead.org
This patchset was a part of the memory folio patchset, but has split out to individual patchset.
incoming (Andrew Morton) https://lkml.kernel.org/r/20210630184624.9ca1937310b0dd5ce66b30e7@linux-foundation.org
Rest of the -mm tree, which dependent on things in -next. Andrew will trickle that over next week, before the merge window is closed.
[PATCH v4] mm: introduce reference pages (Peter Collingbourne) https://lkml.kernel.org/r/20210619092002.1791322-1-pcc@google.com
This patch introduces a new syscall, refpage_create(). This system creates a file descriptor which can be mmap-ed using mmap(), which will be similar to anonymous mapping but the memory will be backed by a reference page instead of the zero page.
[PATCH 0/3] mm/page_reporting: Make page reporting work on arm64 with 64KB page size (Gavin Shan) https://lkml.kernel.org/r/20210621051152.305224-1-gshan@redhat.com
This patchset makes the page reporting to work on ARM with more realistic threshold, which is 2MB instead of 512MB, when the system’s page size is 64KB.