Re: [PATCH] Fix: disable sys_membarrier when nohz_full is enabled

On Thu, Nov 17, 2016 at 11:46:34AM +, Mathieu Desnoyers wrote: > - On Nov 17, 2016, at 1:51 AM, Lai Jiangshan jiangshan...@gmail.com wrote: > > > On Fri, Nov 4, 2016 at 12:29 AM, Mathieu Desnoyers > > wrote: > >> Userspace applications should be allowed to

Re: [PATCH] Fix: disable sys_membarrier when nohz_full is enabled

On Thu, Nov 17, 2016 at 11:46:34AM +, Mathieu Desnoyers wrote: > - On Nov 17, 2016, at 1:51 AM, Lai Jiangshan jiangshan...@gmail.com wrote: > > > On Fri, Nov 4, 2016 at 12:29 AM, Mathieu Desnoyers > > wrote: > >> Userspace applications should be allowed to expect the membarrier system >

Re: [PATCH] Fix: disable sys_membarrier when nohz_full is enabled

On Thu, Nov 17, 2016 at 01:54:27PM +, Mathieu Desnoyers wrote: > - On Nov 17, 2016, at 8:40 AM, Paul E. McKenney > paul...@linux.vnet.ibm.com wrote: > > > On Thu, Nov 17, 2016 at 11:46:34AM +, Mathieu Desnoyers wrote: > >> - On Nov 17, 2016, at 1:51 AM, Lai Jiangshan

Re: [PATCH] Fix: disable sys_membarrier when nohz_full is enabled

On Thu, Nov 17, 2016 at 01:54:27PM +, Mathieu Desnoyers wrote: > - On Nov 17, 2016, at 8:40 AM, Paul E. McKenney > paul...@linux.vnet.ibm.com wrote: > > > On Thu, Nov 17, 2016 at 11:46:34AM +, Mathieu Desnoyers wrote: > >> - On Nov 17, 2016, at 1:51 AM, Lai Jiangshan

Re: [PATCH] Fix: disable sys_membarrier when nohz_full is enabled

On Thu, Nov 17, 2016 at 10:17:25AM -0500, Steven Rostedt wrote: > On Thu, 17 Nov 2016 15:02:18 + (UTC) > Mathieu Desnoyers wrote: > > > > That's an interesting approach. I would be tempted to give it a > > per-thread (rather than per-process) scope. > >

Re: [PATCH] Fix: disable sys_membarrier when nohz_full is enabled

On Thu, Nov 17, 2016 at 10:17:25AM -0500, Steven Rostedt wrote: > On Thu, 17 Nov 2016 15:02:18 + (UTC) > Mathieu Desnoyers wrote: > > > > That's an interesting approach. I would be tempted to give it a > > per-thread (rather than per-process) scope. > > Sure, per thread, but have it

Re: [PATCH] Fix: disable sys_membarrier when nohz_full is enabled

On Thu, 17 Nov 2016 15:02:18 + (UTC) Mathieu Desnoyers wrote: > That's an interesting approach. I would be tempted to give it a > per-thread (rather than per-process) scope. Sure, per thread, but have it inherit to child processes. > > E.g., a thread could

Re: [PATCH] Fix: disable sys_membarrier when nohz_full is enabled

On Thu, 17 Nov 2016 15:02:18 + (UTC) Mathieu Desnoyers wrote: > That's an interesting approach. I would be tempted to give it a > per-thread (rather than per-process) scope. Sure, per thread, but have it inherit to child processes. > > E.g., a thread could do the following to ask to be >

Re: [PATCH] Fix: disable sys_membarrier when nohz_full is enabled

- On Nov 17, 2016, at 9:50 AM, rostedt rost...@goodmis.org wrote: > On Thu, 17 Nov 2016 13:54:27 + (UTC) > Mathieu Desnoyers wrote: > > >> >> > >> >> > Acked-by: Lai Jiangshan >> >> > >> >> > But I'm afraid, in the future,

Re: [PATCH] Fix: disable sys_membarrier when nohz_full is enabled

On Thu, Nov 17, 2016 at 02:51:26PM +0800, Lai Jiangshan wrote: > On Fri, Nov 4, 2016 at 12:29 AM, Mathieu Desnoyers > wrote: > > Userspace applications should be allowed to expect the membarrier system > > call with MEMBARRIER_CMD_SHARED command to issue memory

Re: [PATCH] Fix: disable sys_membarrier when nohz_full is enabled

- On Nov 17, 2016, at 9:50 AM, rostedt rost...@goodmis.org wrote: > On Thu, 17 Nov 2016 13:54:27 + (UTC) > Mathieu Desnoyers wrote: > > >> >> > >> >> > Acked-by: Lai Jiangshan >> >> > >> >> > But I'm afraid, in the future, tick_nohz_full will become a default y >> >> > feature. thus

Re: [PATCH] Fix: disable sys_membarrier when nohz_full is enabled

On Thu, Nov 17, 2016 at 02:51:26PM +0800, Lai Jiangshan wrote: > On Fri, Nov 4, 2016 at 12:29 AM, Mathieu Desnoyers > wrote: > > Userspace applications should be allowed to expect the membarrier system > > call with MEMBARRIER_CMD_SHARED command to issue memory barriers on > > nohz_full CPUs, but

Re: [PATCH] Fix: disable sys_membarrier when nohz_full is enabled

On Thu, 17 Nov 2016 13:54:27 + (UTC) Mathieu Desnoyers wrote: > >> > > >> > Acked-by: Lai Jiangshan > >> > > >> > But I'm afraid, in the future, tick_nohz_full will become a default y > >> > feature. thus it makes sys_membarrier()

Re: [PATCH] Fix: disable sys_membarrier when nohz_full is enabled

On Thu, 17 Nov 2016 13:54:27 + (UTC) Mathieu Desnoyers wrote: > >> > > >> > Acked-by: Lai Jiangshan > >> > > >> > But I'm afraid, in the future, tick_nohz_full will become a default y > >> > feature. thus it makes sys_membarrier() always disabled. we might > >> > need a new

Re: [PATCH] Fix: disable sys_membarrier when nohz_full is enabled

- On Nov 17, 2016, at 8:40 AM, Paul E. McKenney paul...@linux.vnet.ibm.com wrote: > On Thu, Nov 17, 2016 at 11:46:34AM +, Mathieu Desnoyers wrote: >> - On Nov 17, 2016, at 1:51 AM, Lai Jiangshan jiangshan...@gmail.com >> wrote: >> >> > On Fri, Nov 4, 2016 at 12:29 AM, Mathieu

Re: [PATCH] Fix: disable sys_membarrier when nohz_full is enabled

- On Nov 17, 2016, at 8:40 AM, Paul E. McKenney paul...@linux.vnet.ibm.com wrote: > On Thu, Nov 17, 2016 at 11:46:34AM +, Mathieu Desnoyers wrote: >> - On Nov 17, 2016, at 1:51 AM, Lai Jiangshan jiangshan...@gmail.com >> wrote: >> >> > On Fri, Nov 4, 2016 at 12:29 AM, Mathieu

Re: [PATCH] Fix: disable sys_membarrier when nohz_full is enabled

- On Nov 17, 2016, at 1:51 AM, Lai Jiangshan jiangshan...@gmail.com wrote: > On Fri, Nov 4, 2016 at 12:29 AM, Mathieu Desnoyers > wrote: >> Userspace applications should be allowed to expect the membarrier system >> call with MEMBARRIER_CMD_SHARED command to

Re: [PATCH] Fix: disable sys_membarrier when nohz_full is enabled

- On Nov 17, 2016, at 1:51 AM, Lai Jiangshan jiangshan...@gmail.com wrote: > On Fri, Nov 4, 2016 at 12:29 AM, Mathieu Desnoyers > wrote: >> Userspace applications should be allowed to expect the membarrier system >> call with MEMBARRIER_CMD_SHARED command to issue memory barriers on >>

Re: [PATCH] Fix: disable sys_membarrier when nohz_full is enabled

On Fri, Nov 4, 2016 at 12:29 AM, Mathieu Desnoyers wrote: > Userspace applications should be allowed to expect the membarrier system > call with MEMBARRIER_CMD_SHARED command to issue memory barriers on > nohz_full CPUs, but synchronize_sched() does not take those

Re: [PATCH] Fix: disable sys_membarrier when nohz_full is enabled

On Fri, Nov 4, 2016 at 12:29 AM, Mathieu Desnoyers wrote: > Userspace applications should be allowed to expect the membarrier system > call with MEMBARRIER_CMD_SHARED command to issue memory barriers on > nohz_full CPUs, but synchronize_sched() does not take those into > account. > > Given that

Re: [PATCH] Fix: disable sys_membarrier when nohz_full is enabled

On Tue, Nov 08, 2016 at 11:15:51AM +, Mathieu Desnoyers wrote: > - On Nov 7, 2016, at 3:03 PM, Paul E. McKenney paul...@linux.vnet.ibm.com > wrote: > > > On Mon, Nov 07, 2016 at 06:10:14PM +, Mathieu Desnoyers wrote: > >> > >> - On Nov 7, 2016, at 1:06 PM, Paul E. McKenney > >>

Re: [PATCH] Fix: disable sys_membarrier when nohz_full is enabled

On Tue, Nov 08, 2016 at 11:15:51AM +, Mathieu Desnoyers wrote: > - On Nov 7, 2016, at 3:03 PM, Paul E. McKenney paul...@linux.vnet.ibm.com > wrote: > > > On Mon, Nov 07, 2016 at 06:10:14PM +, Mathieu Desnoyers wrote: > >> > >> - On Nov 7, 2016, at 1:06 PM, Paul E. McKenney > >>

Re: [PATCH] Fix: disable sys_membarrier when nohz_full is enabled

- On Nov 7, 2016, at 3:03 PM, Paul E. McKenney paul...@linux.vnet.ibm.com wrote: > On Mon, Nov 07, 2016 at 06:10:14PM +, Mathieu Desnoyers wrote: >> >> - On Nov 7, 2016, at 1:06 PM, Paul E. McKenney paul...@linux.vnet.ibm.com >> wrote: >> >> > On Mon, Nov 07, 2016 at 05:08:59PM

Re: [PATCH] Fix: disable sys_membarrier when nohz_full is enabled

- On Nov 7, 2016, at 3:03 PM, Paul E. McKenney paul...@linux.vnet.ibm.com wrote: > On Mon, Nov 07, 2016 at 06:10:14PM +, Mathieu Desnoyers wrote: >> >> - On Nov 7, 2016, at 1:06 PM, Paul E. McKenney paul...@linux.vnet.ibm.com >> wrote: >> >> > On Mon, Nov 07, 2016 at 05:08:59PM

Re: [PATCH] Fix: disable sys_membarrier when nohz_full is enabled

On Mon, Nov 07, 2016 at 06:10:14PM +, Mathieu Desnoyers wrote: > > - On Nov 7, 2016, at 1:06 PM, Paul E. McKenney paul...@linux.vnet.ibm.com > wrote: > > > On Mon, Nov 07, 2016 at 05:08:59PM +, Mathieu Desnoyers wrote: > >> - On Nov 3, 2016, at 1:49 PM, Paul E. McKenney > >>

Re: [PATCH] Fix: disable sys_membarrier when nohz_full is enabled

On Mon, Nov 07, 2016 at 06:10:14PM +, Mathieu Desnoyers wrote: > > - On Nov 7, 2016, at 1:06 PM, Paul E. McKenney paul...@linux.vnet.ibm.com > wrote: > > > On Mon, Nov 07, 2016 at 05:08:59PM +, Mathieu Desnoyers wrote: > >> - On Nov 3, 2016, at 1:49 PM, Paul E. McKenney > >>

Re: [PATCH] Fix: disable sys_membarrier when nohz_full is enabled

On Mon, Nov 07, 2016 at 05:08:59PM +, Mathieu Desnoyers wrote: > - On Nov 3, 2016, at 1:49 PM, Paul E. McKenney paul...@linux.vnet.ibm.com > wrote: > > > On Thu, Nov 03, 2016 at 10:29:28AM -0600, Mathieu Desnoyers wrote: > >> Userspace applications should be allowed to expect the

Re: [PATCH] Fix: disable sys_membarrier when nohz_full is enabled

On Mon, Nov 07, 2016 at 05:08:59PM +, Mathieu Desnoyers wrote: > - On Nov 3, 2016, at 1:49 PM, Paul E. McKenney paul...@linux.vnet.ibm.com > wrote: > > > On Thu, Nov 03, 2016 at 10:29:28AM -0600, Mathieu Desnoyers wrote: > >> Userspace applications should be allowed to expect the

Re: [PATCH] Fix: disable sys_membarrier when nohz_full is enabled

- On Nov 7, 2016, at 1:06 PM, Paul E. McKenney paul...@linux.vnet.ibm.com wrote: > On Mon, Nov 07, 2016 at 05:08:59PM +, Mathieu Desnoyers wrote: >> - On Nov 3, 2016, at 1:49 PM, Paul E. McKenney paul...@linux.vnet.ibm.com >> wrote: >> >> > On Thu, Nov 03, 2016 at 10:29:28AM -0600,

Re: [PATCH] Fix: disable sys_membarrier when nohz_full is enabled

- On Nov 7, 2016, at 1:06 PM, Paul E. McKenney paul...@linux.vnet.ibm.com wrote: > On Mon, Nov 07, 2016 at 05:08:59PM +, Mathieu Desnoyers wrote: >> - On Nov 3, 2016, at 1:49 PM, Paul E. McKenney paul...@linux.vnet.ibm.com >> wrote: >> >> > On Thu, Nov 03, 2016 at 10:29:28AM -0600,

Re: [PATCH] Fix: disable sys_membarrier when nohz_full is enabled

- On Nov 3, 2016, at 1:49 PM, Paul E. McKenney paul...@linux.vnet.ibm.com wrote: > On Thu, Nov 03, 2016 at 10:29:28AM -0600, Mathieu Desnoyers wrote: >> Userspace applications should be allowed to expect the membarrier system >> call with MEMBARRIER_CMD_SHARED command to issue memory

Re: [PATCH] Fix: disable sys_membarrier when nohz_full is enabled

- On Nov 3, 2016, at 1:49 PM, Paul E. McKenney paul...@linux.vnet.ibm.com wrote: > On Thu, Nov 03, 2016 at 10:29:28AM -0600, Mathieu Desnoyers wrote: >> Userspace applications should be allowed to expect the membarrier system >> call with MEMBARRIER_CMD_SHARED command to issue memory

Re: [PATCH] Fix: disable sys_membarrier when nohz_full is enabled

On Thu, Nov 03, 2016 at 10:29:28AM -0600, Mathieu Desnoyers wrote: > Userspace applications should be allowed to expect the membarrier system > call with MEMBARRIER_CMD_SHARED command to issue memory barriers on > nohz_full CPUs, but synchronize_sched() does not take those into > account. > >

Re: [PATCH] Fix: disable sys_membarrier when nohz_full is enabled

On Thu, Nov 03, 2016 at 10:29:28AM -0600, Mathieu Desnoyers wrote: > Userspace applications should be allowed to expect the membarrier system > call with MEMBARRIER_CMD_SHARED command to issue memory barriers on > nohz_full CPUs, but synchronize_sched() does not take those into > account. > >

[PATCH] Fix: disable sys_membarrier when nohz_full is enabled

Userspace applications should be allowed to expect the membarrier system call with MEMBARRIER_CMD_SHARED command to issue memory barriers on nohz_full CPUs, but synchronize_sched() does not take those into account. Given that we do not want unrelated processes to be able to affect real-time

[PATCH] Fix: disable sys_membarrier when nohz_full is enabled

Userspace applications should be allowed to expect the membarrier system call with MEMBARRIER_CMD_SHARED command to issue memory barriers on nohz_full CPUs, but synchronize_sched() does not take those into account. Given that we do not want unrelated processes to be able to affect real-time