Re: [PATCH] blk: clean up plug

2015-04-17 Thread Jeff Moyer
Shaohua Li writes: >> Also, Jens or Shaohua or anyone, please review my blk-mq plug fix (patch >> 1/2 of aforementioned thread). ;) > > You are not alone :), I posted 2 times too > http://marc.info/?l=linux-kernel=142627559617005=2 Oh, sorry! I think Jens had mentioned that you had a patch

Re: [PATCH] blk: clean up plug

2015-04-17 Thread Shaohua Li
On Fri, Apr 17, 2015 at 04:54:40PM -0400, Jeff Moyer wrote: > Shaohua Li writes: > > > Current code looks like inner plug gets flushed with a > > blk_finish_plug(). Actually it's a nop. All requests/callbacks are added > > to current->plug, while only outmost plug is assigned to current->plug. >

Re: [PATCH] blk: clean up plug

2015-04-17 Thread Jeff Moyer
Shaohua Li writes: > Current code looks like inner plug gets flushed with a > blk_finish_plug(). Actually it's a nop. All requests/callbacks are added > to current->plug, while only outmost plug is assigned to current->plug. > So inner plug always has empty request/callback list, which makes >

Re: [PATCH] blk: clean up plug

2015-04-17 Thread Jeff Moyer
Shaohua Li s...@fb.com writes: Current code looks like inner plug gets flushed with a blk_finish_plug(). Actually it's a nop. All requests/callbacks are added to current-plug, while only outmost plug is assigned to current-plug. So inner plug always has empty request/callback list, which

Re: [PATCH] blk: clean up plug

2015-04-17 Thread Shaohua Li
On Fri, Apr 17, 2015 at 04:54:40PM -0400, Jeff Moyer wrote: Shaohua Li s...@fb.com writes: Current code looks like inner plug gets flushed with a blk_finish_plug(). Actually it's a nop. All requests/callbacks are added to current-plug, while only outmost plug is assigned to current-plug.

Re: [PATCH] blk: clean up plug

2015-04-17 Thread Jeff Moyer
Shaohua Li s...@fb.com writes: Also, Jens or Shaohua or anyone, please review my blk-mq plug fix (patch 1/2 of aforementioned thread). ;) You are not alone :), I posted 2 times too http://marc.info/?l=linux-kernelm=142627559617005w=2 Oh, sorry! I think Jens had mentioned that you had a

[PATCH] blk: clean up plug

2015-04-16 Thread Shaohua Li
Current code looks like inner plug gets flushed with a blk_finish_plug(). Actually it's a nop. All requests/callbacks are added to current->plug, while only outmost plug is assigned to current->plug. So inner plug always has empty request/callback list, which makes blk_flush_plug_list() a nop.

[PATCH] blk: clean up plug

2015-04-16 Thread Shaohua Li
Current code looks like inner plug gets flushed with a blk_finish_plug(). Actually it's a nop. All requests/callbacks are added to current-plug, while only outmost plug is assigned to current-plug. So inner plug always has empty request/callback list, which makes blk_flush_plug_list() a nop. This