Leave IRQs enabled for per-cpu page allocations
Status from this message: [merged mm-stable] mm-page_alloc-leave-irqs-enabled-for-per-cpu-page-allocations.patch removed from -mm tree
Get series with b4
b4 am [email protected]Versions newest first · linked by author and title
- v3 Latest2 patches 6 repliesApplied 2/2
- v21 patch 7 repliesSuperseded
- v11 patch 7 repliesSuperseded
Changes in v3 from the cover letter
o Fix list corruption issue (hughd)
o Tighen up locking protocol (hughd)
o Micro-optimisations (hughd)
Patches in v3 All 2 applied
| # | Subject and files | Size | Replies | Applied |
|---|---|---|---|---|
| 1/2 | mm/page_alloc: Always remove pages from temporary listmm/ | +2 | 1 reply | Applied |
| 2/2 | mm/page_alloc: Leave IRQs enabled for per-cpu page allocationsmm/ | +53 | 5 replies | Applied |
Reviews trailers per patch
All 2 patches have a Reviewed-by or Acked-by.
| Patch | Vlastimil Babka |
|---|---|
| 1/2 | R-b |
| 2/2 | R-b A-b R-b |
Trailers given in replies are filled; those carried in the patches from earlier versions are outlined with dashes.
Cover letter [PATCH v3 0/2] Leave IRQs enabled for per-cpu page allocations
Changelog since V2
o Fix list corruption issue (hughd)
o Tighen up locking protocol (hughd)
o Micro-optimisations (hughd)
Changelog since V1
o Use trylock in free_unref_page_list due to IO completion from
softirq context (yuzhao)
This was a standalone patch but now split in two. The main changes since v2 are fixing the issues exposed by Hugh's shmfs stress test.
Patch 1 is a long-standing issue that is technically a bug but happened to work because of how it was used.
Patch 2 leaves IRQs enabled for most PCP allocations with more details in the changelog itself.
Signature · 1 line
2.35.3