mm/madvise: introduce MADV_COLLAPSE sync hugepage collapse
This idea was introduced by David Rientjes[1]. Introduce a new madvise mode, MADV_COLLAPSE, that allows users to request a synchronous collapse of memory at their own expense. The benefits of this approach are: * CPU is charged to the process that wants to spend the cycles for the THP * Avoid unpredictable timing of khugepaged collapse Semantics This call is independent of the system-wide THP sysfs settings, but will fail for memory marked VM_NOHUGEPAGE. If the ranges provided span multiple VMAs, the semantics of the collapse over each VMA is independent from the others. This implies a hugepage cannot cross a VMA boundary. If collapse of a given hugepage-aligned/sized region fails, the operation may continue to attempt collapsing the remainder of memory specified. The memory ranges provided must be page-aligned, but are not required to be hugepage-aligned. If the memory ranges are not hugepage-aligned, the start/end of the range will be clamped to the first/last hugepage-aligned address covered by said range. The memory ranges must span at least one hugepage-sized region. All non-resident pages covered by the range will first be swapped/faulted-in, before being internally copied onto a freshly allocated hugepage. Unmapped pages will have their data directly initialized to 0 in the new hugepage. However, for every eligible hugepage aligned/sized region to-be collapsed, at least one page must currently be backed by memory (a PMD covering the address range must already exist). Allocation for the new hugepage may enter direct reclaim and/or compaction, regardless of VMA flags. When the system has multiple NUMA nodes, the hugepage will be allocated from the node providing the most native pages. This operation operates on the current state of the specified process and makes no persistent changes or guarantees on how pages will be mapped, constructed, or faulted in the future Return Value If all hugepage-sized/aligned regions covered by the provided range were either successfully collapsed, or were already PMD-mapped THPs, this operation will be deemed successful. On success, process_madvise(2) returns the number of bytes advised, and madvise(2) returns 0. Else, -1 is returned and errno is set to indicate the error for the most-recently attempted hugepage collapse. Note that many failures might have occurred, since the operation may continue to collapse in the event a single hugepage-sized/aligned region fails. ENOMEM Memory allocation failed or VMA not found EBUSY Memcg charging failed EAGAIN Required resource temporarily unavailable. Try again might succeed. EINVAL Other error: No PMD found, subpage doesn't have Present bit set, "Special" page no backed by struct page, VMA incorrectly sized, address not page-aligned, ... Most notable here is ENOMEM and EBUSY (new to madvise) which are intended to provide the caller with actionable feedback so they may take an appropriate fallback measure. Use Cases An immediate user of this new functionality are malloc() implementations that manage memory in hugepage-sized chunks, but sometimes subrelease memory back to the system in native-sized chunks via MADV_DONTNEED; zapping the pmd. Later, when the memory is hot, the implementation could madvise(MADV_COLLAPSE) to re-back the memory by THPs to regain hugepage coverage and dTLB performance. TCMalloc is such an implementation that could benefit from this[2]. Only privately-mapped anon memory is supported for now, but additional support for file, shmem, and HugeTLB high-granularity mappings[2] is expected. File and tmpfs/shmem support would permit: * Backing executable text by THPs. Current support provided by CONFIG_READ_ONLY_THP_FOR_FS may take a long time on a large system which might impair services from serving at their full rated load after (re)starting. Tricks like mremap(2)'ing text onto anonymous memory to immediately realize iTLB performance prevents page sharing and demand paging, both of which increase steady state memory footprint. With MADV_COLLAPSE, we get the best of both worlds: Peak upfront performance and lower RAM footprints. * Backing guest memory by hugapages after the memory contents have been migrated in native-page-sized chunks to a new host, in a userfaultfd-based live-migration stack. [1] https://lore.kernel.org/linux-mm/d098c392-273a-36a4-1a29-59731cdf5d3d@google.com/ [2] https://github.com/google/tcmalloc/tree/master/tcmalloc [jrdr.linux@gmail.com: avoid possible memory leak in failure path] Link: https://lkml.kernel.org/r/20220713024109.62810-1-jrdr.linux@gmail.com [zokeefe@google.com add missing kfree() to madvise_collapse()] Link: https://lore.kernel.org/linux-mm/20220713024109.62810-1-jrdr.linux@gmail.com/ Link: https://lkml.kernel.org/r/20220713161851.1879439-1-zokeefe@google.com [zokeefe@google.com: delay computation of hpage boundaries until use]] Link: https://lkml.kernel.org/r/20220720140603.1958773-4-zokeefe@google.com Link: https://lkml.kernel.org/r/20220706235936.2197195-10-zokeefe@google.com Signed-off-by:Zach O'Keefe <zokeefe@google.com> Signed-off-by:
Please register or sign in to comment