Re: [PATCH 8/9] locktorture: Support rwsems

2014-09-12 Thread Paul E. McKenney
On Thu, Sep 11, 2014 at 09:41:30PM -0700, Davidlohr Bueso wrote: > We can easily do so with our new reader lock support. Just an arbitrary > design default: readers have higher (5x) critical region latencies than > writers: 50 ms and 10 ms, respectively. Except in the massive contention case,

Re: [PATCH 8/9] locktorture: Support rwsems

2014-09-12 Thread Davidlohr Bueso
On Fri, 2014-09-12 at 09:37 +0200, Peter Zijlstra wrote: > On Thu, Sep 11, 2014 at 09:41:30PM -0700, Davidlohr Bueso wrote: > > We can easily do so with our new reader lock support. Just an arbitrary > > design default: readers have higher (5x) critical region latencies than > > writers: 50 ms and

Re: [PATCH 8/9] locktorture: Support rwsems

2014-09-12 Thread Peter Zijlstra
On Thu, Sep 11, 2014 at 09:41:30PM -0700, Davidlohr Bueso wrote: > We can easily do so with our new reader lock support. Just an arbitrary > design default: readers have higher (5x) critical region latencies than > writers: 50 ms and 10 ms, respectively. > Nice, could you copy/paste this into a

Re: [PATCH 8/9] locktorture: Support rwsems

2014-09-12 Thread Peter Zijlstra
On Thu, Sep 11, 2014 at 09:41:30PM -0700, Davidlohr Bueso wrote: We can easily do so with our new reader lock support. Just an arbitrary design default: readers have higher (5x) critical region latencies than writers: 50 ms and 10 ms, respectively. Nice, could you copy/paste this into a

Re: [PATCH 8/9] locktorture: Support rwsems

2014-09-12 Thread Davidlohr Bueso
On Fri, 2014-09-12 at 09:37 +0200, Peter Zijlstra wrote: On Thu, Sep 11, 2014 at 09:41:30PM -0700, Davidlohr Bueso wrote: We can easily do so with our new reader lock support. Just an arbitrary design default: readers have higher (5x) critical region latencies than writers: 50 ms and 10 ms,

Re: [PATCH 8/9] locktorture: Support rwsems

2014-09-12 Thread Paul E. McKenney
On Thu, Sep 11, 2014 at 09:41:30PM -0700, Davidlohr Bueso wrote: We can easily do so with our new reader lock support. Just an arbitrary design default: readers have higher (5x) critical region latencies than writers: 50 ms and 10 ms, respectively. Except in the massive contention case, where

[PATCH 8/9] locktorture: Support rwsems

2014-09-11 Thread Davidlohr Bueso
We can easily do so with our new reader lock support. Just an arbitrary design default: readers have higher (5x) critical region latencies than writers: 50 ms and 10 ms, respectively. Signed-off-by: Davidlohr Bueso --- Documentation/locking/locktorture.txt | 2 ++ kernel/locking/locktorture.c

[PATCH 8/9] locktorture: Support rwsems

2014-09-11 Thread Davidlohr Bueso
We can easily do so with our new reader lock support. Just an arbitrary design default: readers have higher (5x) critical region latencies than writers: 50 ms and 10 ms, respectively. Signed-off-by: Davidlohr Bueso dbu...@suse.de --- Documentation/locking/locktorture.txt | 2 ++