Per cpu relocation to ZERO and x86_32 percpu ops on x86_64
Get series with b4
b4 am [email protected]Versions newest first
- v1 Latest3 patches 19 repliesIn review
Patches in v1 No applies recorded yet
| # | Subject and files | Size | Replies | Applied |
|---|---|---|---|---|
| 1/3 | Percpu infrastructure to rebase the per cpu area to 0ULinclude/ include/ include/ +2 more | +42 | 2 replies | |
| 2/3 | X86_64: Declare pda as per cpu data thereby moving it into the cpu areaarch/ arch/ arch/ +3 more | +21 | 0 replies | |
| 3/3 | x86_64: Make the x86_32 percpu operations usable on x86_64arch/ arch/ arch/ +1 more | +19 | 1 reply |
Reviews trailers per patch
No Reviewed-by, Acked-by or Tested-by trailers for v1 yet.
Cover letter [patch 0/3] Per cpu relocation to ZERO and x86_32 percpu ops on x86_64
This patchset allows the use of x86_32 percpu ops on x86_64 while maintaining %gs pointing to the pda. It does that by moving the x86_64 pda into the percpu area (thereby pointing %gs at the per cpu area) and then relocating the x86_64 per cpu variables to start at 0.
Patch applies on top of the per cpu cleanup patches V2. See http://marc.info/?l=linux-kernel&m=119628478316525&w=2
Ultimately I think we can make the per cpu accessors arch independent (see the RFC at http://marc.info/?l=linux-kernel&m=119552126330405&w=2).
There is a performance benefit from using these in core code.