Re: [PATCH] rwsem: steal writing sem for better performance

2013-02-06 Thread Ingo Molnar
* Linus Torvalds wrote: > On Wed, Feb 6, 2013 at 10:28 PM, Ingo Molnar wrote: > > > > Linus, Andrew, what is your thinking about the patch and about > > the timing of the patch? > > Not for 3.8. Queue it for 3.9, with possibly a stable tag with > a big comment "apply after much testing". Ok,

Re: [PATCH] rwsem: steal writing sem for better performance

2013-02-06 Thread Linus Torvalds
On Wed, Feb 6, 2013 at 10:28 PM, Ingo Molnar wrote: > > Linus, Andrew, what is your thinking about the patch and about > the timing of the patch? Not for 3.8. Queue it for 3.9, with possibly a stable tag with a big comment "apply after much testing". Linus -- To unsubscribe from th

Re: [PATCH] rwsem: steal writing sem for better performance

2013-02-06 Thread Ingo Molnar
* Alex Shi wrote: > On 02/05/2013 10:58 PM, Ingo Molnar wrote: > > > > * Alex Shi wrote: > > > >> Commit 5a50508 change to rwsem from mutex, that cause aim7 > >> fork_test dropped 50%. Yuanhan liu does a good analysis, > >> find it caused by strict sequential writing. Ingo suggest > >> ste

Re: [PATCH] rwsem: steal writing sem for better performance

2013-02-05 Thread Alex Shi
On 02/05/2013 10:58 PM, Ingo Molnar wrote: > > * Alex Shi wrote: > >> Commit 5a50508 change to rwsem from mutex, that cause aim7 fork_test >> dropped 50%. Yuanhan liu does a good analysis, find it caused by >> strict sequential writing. Ingo suggest stealing sem writing from >> front task in wai

Re: [PATCH] rwsem: steal writing sem for better performance

2013-02-05 Thread Ingo Molnar
* Alex Shi wrote: > Commit 5a50508 change to rwsem from mutex, that cause aim7 fork_test > dropped 50%. Yuanhan liu does a good analysis, find it caused by > strict sequential writing. Ingo suggest stealing sem writing from > front task in waiting queue. https://lkml.org/lkml/2013/1/29/84 > So h

[PATCH] rwsem: steal writing sem for better performance

2013-02-05 Thread Alex Shi
Commit 5a50508 change to rwsem from mutex, that cause aim7 fork_test dropped 50%. Yuanhan liu does a good analysis, find it caused by strict sequential writing. Ingo suggest stealing sem writing from front task in waiting queue. https://lkml.org/lkml/2013/1/29/84 So has this patch. In this patch,