blk-throttle: Couple of more fixes
Get series with b4
b4 am [email protected]Versions newest first
- v1 Latest2 patches 4 repliesIn review
Patches in v1 No applies recorded yet
| # | Subject and files | Size | Replies | Applied |
|---|---|---|---|---|
| 1/2 | blk-throttle: process limit change only through one functionblock/ | +1 | 0 replies | |
| 2/2 | blk-throttle: Some cleanups and race fixes in limit update codeblock/ | +40 | 9 replies |
Reviews trailers per patch
1 of 2 patches has a Reviewed-by or Acked-by.
| Patch | Oleg Nesterov | Paul E. McKenney |
|---|---|---|
| 1/2 | ||
| 2/2 | R-b | R-b |
Trailers given in replies are filled; those carried in the patches from earlier versions are outlined with dashes.
Cover letter [PATCH 0/2] blk-throttle: Couple of more fixes
Hi Jens,
Please find attached couple of more fixes for blk-throttle code. These are based on top of "for-linus" branch of your block tree.
Oleg had pointed out couple of race conditions in cgroup weight update code. I think these race conditions are hard to hit and not disastrous so I would not be too concerned about pushing these patches in 2.6.37 and can queue up for 2.6.38.
Paul,
Based on discussion in other mail thread, I have used xchg() based implementation for updating and processing limtis. Can you please have a look if it is correct implementation and do I need any ACCESS_ONCE() or barriers somewhere. If this implementation is not correct then I can go back to atomic variable based implementation as suggested by you in other mail thread. Appreciate the help.
Thanks
Vivek
Vivek Goyal (2):
blk-throttle: process limit change only through one function
blk-throttle: Some cleanups and race fixes in limit update code
block/blk-throttle.c | 104 ++++++++++++++++++++------------------------------ 1 files changed, 41 insertions(+), 63 deletions(-)