Re: [PATCH v4 2/5] sched_clock: Optimize cache line usage

2015-02-09 Thread Stephen Boyd
On 02/09/15 01:47, Daniel Thompson wrote: > On 09/02/15 09:28, Will Deacon wrote: >> On Sun, Feb 08, 2015 at 12:02:37PM +, Daniel Thompson wrote: >>> Currently sched_clock(), a very hot code path, is not optimized to >>> minimise its cache profile. In particular: >>> >>> 1. cd is not cach

Re: [PATCH v4 2/5] sched_clock: Optimize cache line usage

2015-02-09 Thread Daniel Thompson
On 09/02/15 09:28, Will Deacon wrote: > On Sun, Feb 08, 2015 at 12:02:37PM +, Daniel Thompson wrote: >> Currently sched_clock(), a very hot code path, is not optimized to >> minimise its cache profile. In particular: >> >> 1. cd is not cacheline_aligned, >> >> 2. struct clock_data does

Re: [PATCH v4 2/5] sched_clock: Optimize cache line usage

2015-02-08 Thread Will Deacon
On Sun, Feb 08, 2015 at 12:02:37PM +, Daniel Thompson wrote: > Currently sched_clock(), a very hot code path, is not optimized to > minimise its cache profile. In particular: > > 1. cd is not cacheline_aligned, > > 2. struct clock_data does not distinguish between hotpath and > c

[PATCH v4 2/5] sched_clock: Optimize cache line usage

2015-02-08 Thread Daniel Thompson
Currently sched_clock(), a very hot code path, is not optimized to minimise its cache profile. In particular: 1. cd is not cacheline_aligned, 2. struct clock_data does not distinguish between hotpath and coldpath data, reducing locality of reference in the hotpath, 3. Some hotpath