The attached patch will correct the ordering of trace output between request queue insertions (I) and unplug calls (U). Right now the insert precedes the unplug, which just isn't right: 65,128 0 1 67.699868965 7882 Q R 0 + 1 [aiod] 65,128 0 2 67.699876462 7882 G R 0 + 1 [aiod] 65,128 0 3 67.699878286 7882 P W [aiod] 65,128 0 4 67.699880491 7882 I R 0 + 1 [aiod] 65,128 0 5 67.699887589 7882 U R [aiod] 1 65,128 0 6 67.699898317 54 D R 0 + 1 [kblockd/0] 65,128 2 153 67.700126590 0 C R 0 + 1 [0] With the patch provided the unplug comes first: 65,128 3 1 0.000000000 7045 Q R 0 + 1 [aiod] 65,128 3 2 0.000002295 7045 G R 0 + 1 [aiod] 65,128 3 3 0.000002617 7045 P W [aiod] 65,128 3 4 0.000003685 7045 U R [aiod] 1 65,128 3 5 0.000004107 7045 I R 0 + 1 [aiod] 65,128 3 6 0.000009491 57 D R 0 + 1 [kblockd/3] 65,128 2 1 0.000232447 0 C R 0 + 1 [0] Jens: If you agree, the patch can be applied to your plug branch for git://git.kernel.dk/data/git/linux-2.6-block.git Thanks, Alan
| Greg Kroah-Hartman | [PATCH 008/196] Chinese: add translation of volatile-considered-harmful.txt |
| Amit K. Arora | [RFC] Heads up on sys_fallocate() |
| Bart Van Assche | Integration of SCST in the mainstream Linux kernel |
| Linus Torvalds | Re: Slow DOWN, please!!! |
git: | |
| Gerrit Renker | [PATCH 0/37] dccp: Feature negotiation - last call for comments |
| David Miller | Re: [PATCH] pkt_sched: Destroy gen estimators under rtnl_lock(). |
| David Miller | [GIT]: Networking |
| Natalie Protasevich | [BUG] New Kernel Bugs |
