Re: [PATCH 1/2] perf bench futex: cache align the worer struct

2016-10-19 Thread Sebastian Andrzej Siewior
On 2016-10-17 18:09:49 [-0700], Davidlohr Bueso wrote: > On Sun, 16 Oct 2016, Sebastian Andrzej Siewior wrote: > > > It popped up in perf testing that the worker consumes some amount of > > CPU. It boils down to the increment of `ops` which causes cache line > > bouncing between the individual thr

Re: [PATCH 1/2] perf bench futex: cache align the worer struct

2016-10-17 Thread Davidlohr Bueso
On Sun, 16 Oct 2016, Sebastian Andrzej Siewior wrote: It popped up in perf testing that the worker consumes some amount of CPU. It boils down to the increment of `ops` which causes cache line bouncing between the individual threads. Are you referring to this? │ for (i = 0; i <

[PATCH 1/2] perf bench futex: cache align the worer struct

2016-10-16 Thread Sebastian Andrzej Siewior
It popped up in perf testing that the worker consumes some amount of CPU. It boils down to the increment of `ops` which causes cache line bouncing between the individual threads. The patch aligns the struct by 256 bytes to ensure that not a cache line is shared among CPUs. 128 byte is the x86 worst