Re: hackbench vs select_idle_sibling; was: [tip:sched/core] sched/fair, cpumask: Export for_each_cpu_wrap()

2017-06-09 Thread Chris Mason
On 06/06/2017 05:21 AM, Peter Zijlstra wrote: On Mon, Jun 05, 2017 at 02:00:21PM +0100, Matt Fleming wrote: On Fri, 19 May, at 04:00:35PM, Matt Fleming wrote: On Wed, 17 May, at 12:53:50PM, Peter Zijlstra wrote: Please test.. Results are still coming in but things do look better with your

Re: hackbench vs select_idle_sibling; was: [tip:sched/core] sched/fair, cpumask: Export for_each_cpu_wrap()

2017-06-09 Thread Chris Mason
On 06/06/2017 05:21 AM, Peter Zijlstra wrote: On Mon, Jun 05, 2017 at 02:00:21PM +0100, Matt Fleming wrote: On Fri, 19 May, at 04:00:35PM, Matt Fleming wrote: On Wed, 17 May, at 12:53:50PM, Peter Zijlstra wrote: Please test.. Results are still coming in but things do look better with your

Re: hackbench vs select_idle_sibling; was: [tip:sched/core] sched/fair, cpumask: Export for_each_cpu_wrap()

2017-06-06 Thread Peter Zijlstra
On Mon, Jun 05, 2017 at 02:00:21PM +0100, Matt Fleming wrote: > On Fri, 19 May, at 04:00:35PM, Matt Fleming wrote: > > On Wed, 17 May, at 12:53:50PM, Peter Zijlstra wrote: > > > > > > Please test.. > > > > Results are still coming in but things do look better with your patch > > applied. > > >

Re: hackbench vs select_idle_sibling; was: [tip:sched/core] sched/fair, cpumask: Export for_each_cpu_wrap()

2017-06-06 Thread Peter Zijlstra
On Mon, Jun 05, 2017 at 02:00:21PM +0100, Matt Fleming wrote: > On Fri, 19 May, at 04:00:35PM, Matt Fleming wrote: > > On Wed, 17 May, at 12:53:50PM, Peter Zijlstra wrote: > > > > > > Please test.. > > > > Results are still coming in but things do look better with your patch > > applied. > > >

Re: hackbench vs select_idle_sibling; was: [tip:sched/core] sched/fair, cpumask: Export for_each_cpu_wrap()

2017-06-05 Thread Matt Fleming
On Fri, 19 May, at 04:00:35PM, Matt Fleming wrote: > On Wed, 17 May, at 12:53:50PM, Peter Zijlstra wrote: > > > > Please test.. > > Results are still coming in but things do look better with your patch > applied. > > It does look like there's a regression when running hackbench in > process

Re: hackbench vs select_idle_sibling; was: [tip:sched/core] sched/fair, cpumask: Export for_each_cpu_wrap()

2017-06-05 Thread Matt Fleming
On Fri, 19 May, at 04:00:35PM, Matt Fleming wrote: > On Wed, 17 May, at 12:53:50PM, Peter Zijlstra wrote: > > > > Please test.. > > Results are still coming in but things do look better with your patch > applied. > > It does look like there's a regression when running hackbench in > process

Re: hackbench vs select_idle_sibling; was: [tip:sched/core] sched/fair, cpumask: Export for_each_cpu_wrap()

2017-05-19 Thread Matt Fleming
On Wed, 17 May, at 12:53:50PM, Peter Zijlstra wrote: > > Please test.. Results are still coming in but things do look better with your patch applied. It does look like there's a regression when running hackbench in process mode and when the CPUs are not fully utilised, e.g. check this out:

Re: hackbench vs select_idle_sibling; was: [tip:sched/core] sched/fair, cpumask: Export for_each_cpu_wrap()

2017-05-19 Thread Matt Fleming
On Wed, 17 May, at 12:53:50PM, Peter Zijlstra wrote: > > Please test.. Results are still coming in but things do look better with your patch applied. It does look like there's a regression when running hackbench in process mode and when the CPUs are not fully utilised, e.g. check this out:

Re: hackbench vs select_idle_sibling; was: [tip:sched/core] sched/fair, cpumask: Export for_each_cpu_wrap()

2017-05-17 Thread Chris Mason
On 05/17/2017 06:53 AM, Peter Zijlstra wrote: On Mon, May 15, 2017 at 02:03:11AM -0700, tip-bot for Peter Zijlstra wrote: sched/fair, cpumask: Export for_each_cpu_wrap() -static int cpumask_next_wrap(int n, const struct cpumask *mask, int start, int *wrapped) -{ - next =

Re: hackbench vs select_idle_sibling; was: [tip:sched/core] sched/fair, cpumask: Export for_each_cpu_wrap()

2017-05-17 Thread Chris Mason
On 05/17/2017 06:53 AM, Peter Zijlstra wrote: On Mon, May 15, 2017 at 02:03:11AM -0700, tip-bot for Peter Zijlstra wrote: sched/fair, cpumask: Export for_each_cpu_wrap() -static int cpumask_next_wrap(int n, const struct cpumask *mask, int start, int *wrapped) -{ - next =

Re: hackbench vs select_idle_sibling; was: [tip:sched/core] sched/fair, cpumask: Export for_each_cpu_wrap()

2017-05-17 Thread Matt Fleming
On Wed, 17 May, at 12:53:50PM, Peter Zijlstra wrote: > On Mon, May 15, 2017 at 02:03:11AM -0700, tip-bot for Peter Zijlstra wrote: > > sched/fair, cpumask: Export for_each_cpu_wrap() > > > -static int cpumask_next_wrap(int n, const struct cpumask *mask, int start, > > int *wrapped) > > -{ > > >

Re: hackbench vs select_idle_sibling; was: [tip:sched/core] sched/fair, cpumask: Export for_each_cpu_wrap()

2017-05-17 Thread Matt Fleming
On Wed, 17 May, at 12:53:50PM, Peter Zijlstra wrote: > On Mon, May 15, 2017 at 02:03:11AM -0700, tip-bot for Peter Zijlstra wrote: > > sched/fair, cpumask: Export for_each_cpu_wrap() > > > -static int cpumask_next_wrap(int n, const struct cpumask *mask, int start, > > int *wrapped) > > -{ > > >

hackbench vs select_idle_sibling; was: [tip:sched/core] sched/fair, cpumask: Export for_each_cpu_wrap()

2017-05-17 Thread Peter Zijlstra
On Mon, May 15, 2017 at 02:03:11AM -0700, tip-bot for Peter Zijlstra wrote: > sched/fair, cpumask: Export for_each_cpu_wrap() > -static int cpumask_next_wrap(int n, const struct cpumask *mask, int start, > int *wrapped) > -{ > - next = find_next_bit(cpumask_bits(mask), nr_cpumask_bits,

hackbench vs select_idle_sibling; was: [tip:sched/core] sched/fair, cpumask: Export for_each_cpu_wrap()

2017-05-17 Thread Peter Zijlstra
On Mon, May 15, 2017 at 02:03:11AM -0700, tip-bot for Peter Zijlstra wrote: > sched/fair, cpumask: Export for_each_cpu_wrap() > -static int cpumask_next_wrap(int n, const struct cpumask *mask, int start, > int *wrapped) > -{ > - next = find_next_bit(cpumask_bits(mask), nr_cpumask_bits,