Re: [PATCH v3 tip/core/rcu 40/40] rcu: Make non-preemptive schedule be Tasks RCU quiescent state

On Fri, Sep 29, 2017 at 12:01:24PM +0200, Paolo Bonzini wrote: > On 29/09/2017 11:30, Boqun Feng wrote: > > On Thu, Sep 28, 2017 at 04:05:14PM +, Paul E. McKenney wrote: > > [...] > >>> __schedule+0x201/0x2240 kernel/sched/core.c:3292 > >>> schedule+0x113/0x460 kernel/sched/core.c:3421 > >>>

Re: [PATCH v3 tip/core/rcu 40/40] rcu: Make non-preemptive schedule be Tasks RCU quiescent state

On Fri, Sep 29, 2017 at 12:01:24PM +0200, Paolo Bonzini wrote: > On 29/09/2017 11:30, Boqun Feng wrote: > > On Thu, Sep 28, 2017 at 04:05:14PM +, Paul E. McKenney wrote: > > [...] > >>> __schedule+0x201/0x2240 kernel/sched/core.c:3292 > >>> schedule+0x113/0x460 kernel/sched/core.c:3421 > >>>

Re: [PATCH v3 tip/core/rcu 40/40] rcu: Make non-preemptive schedule be Tasks RCU quiescent state

On Fri, Sep 29, 2017 at 01:44:56PM +0200, Paolo Bonzini wrote: > On 29/09/2017 12:34, Peter Zijlstra wrote: > > On Fri, Sep 29, 2017 at 12:01:24PM +0200, Paolo Bonzini wrote: > >>> Does this mean whenever we get a page fault in a RCU read-side critical > >>> section, we may hit this? > >>> > >>>

Re: [PATCH v3 tip/core/rcu 40/40] rcu: Make non-preemptive schedule be Tasks RCU quiescent state

On Fri, Sep 29, 2017 at 01:44:56PM +0200, Paolo Bonzini wrote: > On 29/09/2017 12:34, Peter Zijlstra wrote: > > On Fri, Sep 29, 2017 at 12:01:24PM +0200, Paolo Bonzini wrote: > >>> Does this mean whenever we get a page fault in a RCU read-side critical > >>> section, we may hit this? > >>> > >>>

Re: [PATCH v3 tip/core/rcu 40/40] rcu: Make non-preemptive schedule be Tasks RCU quiescent state

On 29/09/2017 12:34, Peter Zijlstra wrote: > On Fri, Sep 29, 2017 at 12:01:24PM +0200, Paolo Bonzini wrote: >>> Does this mean whenever we get a page fault in a RCU read-side critical >>> section, we may hit this? >>> >>> Could we simply avoid to schedule() in kvm_async_pf_task_wait() if the >>>

Re: [PATCH v3 tip/core/rcu 40/40] rcu: Make non-preemptive schedule be Tasks RCU quiescent state

On 29/09/2017 12:34, Peter Zijlstra wrote: > On Fri, Sep 29, 2017 at 12:01:24PM +0200, Paolo Bonzini wrote: >>> Does this mean whenever we get a page fault in a RCU read-side critical >>> section, we may hit this? >>> >>> Could we simply avoid to schedule() in kvm_async_pf_task_wait() if the >>>

Re: [PATCH v3 tip/core/rcu 40/40] rcu: Make non-preemptive schedule be Tasks RCU quiescent state

On Fri, Sep 29, 2017 at 12:01:24PM +0200, Paolo Bonzini wrote: > > Does this mean whenever we get a page fault in a RCU read-side critical > > section, we may hit this? > > > > Could we simply avoid to schedule() in kvm_async_pf_task_wait() if the > > fault process is in a RCU read-side critical

Re: [PATCH v3 tip/core/rcu 40/40] rcu: Make non-preemptive schedule be Tasks RCU quiescent state

On Fri, Sep 29, 2017 at 12:01:24PM +0200, Paolo Bonzini wrote: > > Does this mean whenever we get a page fault in a RCU read-side critical > > section, we may hit this? > > > > Could we simply avoid to schedule() in kvm_async_pf_task_wait() if the > > fault process is in a RCU read-side critical

Re: [PATCH v3 tip/core/rcu 40/40] rcu: Make non-preemptive schedule be Tasks RCU quiescent state

On Fri, Sep 29, 2017 at 10:01:24AM +, Paolo Bonzini wrote: > On 29/09/2017 11:30, Boqun Feng wrote: > > On Thu, Sep 28, 2017 at 04:05:14PM +, Paul E. McKenney wrote: > > [...] > >>> __schedule+0x201/0x2240 kernel/sched/core.c:3292 > >>> schedule+0x113/0x460 kernel/sched/core.c:3421 > >>>

Re: [PATCH v3 tip/core/rcu 40/40] rcu: Make non-preemptive schedule be Tasks RCU quiescent state

On Fri, Sep 29, 2017 at 10:01:24AM +, Paolo Bonzini wrote: > On 29/09/2017 11:30, Boqun Feng wrote: > > On Thu, Sep 28, 2017 at 04:05:14PM +, Paul E. McKenney wrote: > > [...] > >>> __schedule+0x201/0x2240 kernel/sched/core.c:3292 > >>> schedule+0x113/0x460 kernel/sched/core.c:3421 > >>>

Re: [PATCH v3 tip/core/rcu 40/40] rcu: Make non-preemptive schedule be Tasks RCU quiescent state

On 29/09/2017 11:30, Boqun Feng wrote: > On Thu, Sep 28, 2017 at 04:05:14PM +, Paul E. McKenney wrote: > [...] >>> __schedule+0x201/0x2240 kernel/sched/core.c:3292 >>> schedule+0x113/0x460 kernel/sched/core.c:3421 >>> kvm_async_pf_task_wait+0x43f/0x940 arch/x86/kernel/kvm.c:158 >> >> It is

Re: [PATCH v3 tip/core/rcu 40/40] rcu: Make non-preemptive schedule be Tasks RCU quiescent state

On 29/09/2017 11:30, Boqun Feng wrote: > On Thu, Sep 28, 2017 at 04:05:14PM +, Paul E. McKenney wrote: > [...] >>> __schedule+0x201/0x2240 kernel/sched/core.c:3292 >>> schedule+0x113/0x460 kernel/sched/core.c:3421 >>> kvm_async_pf_task_wait+0x43f/0x940 arch/x86/kernel/kvm.c:158 >> >> It is

Re: [PATCH v3 tip/core/rcu 40/40] rcu: Make non-preemptive schedule be Tasks RCU quiescent state

On Thu, Sep 28, 2017 at 04:05:14PM +, Paul E. McKenney wrote: [...] > > __schedule+0x201/0x2240 kernel/sched/core.c:3292 > > schedule+0x113/0x460 kernel/sched/core.c:3421 > > kvm_async_pf_task_wait+0x43f/0x940 arch/x86/kernel/kvm.c:158 > > It is kvm_async_pf_task_wait() that calls

Re: [PATCH v3 tip/core/rcu 40/40] rcu: Make non-preemptive schedule be Tasks RCU quiescent state

On Thu, Sep 28, 2017 at 04:05:14PM +, Paul E. McKenney wrote: [...] > > __schedule+0x201/0x2240 kernel/sched/core.c:3292 > > schedule+0x113/0x460 kernel/sched/core.c:3421 > > kvm_async_pf_task_wait+0x43f/0x940 arch/x86/kernel/kvm.c:158 > > It is kvm_async_pf_task_wait() that calls

Re: [PATCH v3 tip/core/rcu 40/40] rcu: Make non-preemptive schedule be Tasks RCU quiescent state

On Thu, Sep 28, 2017 at 06:18:50PM +0200, Peter Zijlstra wrote: > On Thu, Sep 28, 2017 at 09:05:14AM -0700, Paul E. McKenney wrote: > > > do_async_page_fault+0x72/0x90 arch/x86/kernel/kvm.c:271 > > > async_page_fault+0x22/0x30 arch/x86/entry/entry_64.S:1069 > > > RIP:

Re: [PATCH v3 tip/core/rcu 40/40] rcu: Make non-preemptive schedule be Tasks RCU quiescent state

On Thu, Sep 28, 2017 at 06:18:50PM +0200, Peter Zijlstra wrote: > On Thu, Sep 28, 2017 at 09:05:14AM -0700, Paul E. McKenney wrote: > > > do_async_page_fault+0x72/0x90 arch/x86/kernel/kvm.c:271 > > > async_page_fault+0x22/0x30 arch/x86/entry/entry_64.S:1069 > > > RIP:

Re: [PATCH v3 tip/core/rcu 40/40] rcu: Make non-preemptive schedule be Tasks RCU quiescent state

On Thu, Sep 28, 2017 at 06:08:56PM +0200, Peter Zijlstra wrote: >On Thu, Sep 28, 2017 at 03:38:16PM +, Levin, Alexander (Sasha Levin) wrote: >> On Thu, Sep 28, 2017 at 05:30:55AM -0700, Paul E. McKenney wrote: > >> >Hmmm... kernel/rcu/tree_plugin.h:329 thinks that someone slept (as opposed >>

Re: [PATCH v3 tip/core/rcu 40/40] rcu: Make non-preemptive schedule be Tasks RCU quiescent state

On Thu, Sep 28, 2017 at 06:08:56PM +0200, Peter Zijlstra wrote: >On Thu, Sep 28, 2017 at 03:38:16PM +, Levin, Alexander (Sasha Levin) wrote: >> On Thu, Sep 28, 2017 at 05:30:55AM -0700, Paul E. McKenney wrote: > >> >Hmmm... kernel/rcu/tree_plugin.h:329 thinks that someone slept (as opposed >>

Re: [PATCH v3 tip/core/rcu 40/40] rcu: Make non-preemptive schedule be Tasks RCU quiescent state

On Thu, Sep 28, 2017 at 09:05:14AM -0700, Paul E. McKenney wrote: > > do_async_page_fault+0x72/0x90 arch/x86/kernel/kvm.c:271 > > async_page_fault+0x22/0x30 arch/x86/entry/entry_64.S:1069 > > RIP: 0010:format_decode+0x240/0x830 lib/vsprintf.c:1996 > > RSP: 0018:88003b2df520 EFLAGS: 00010283

Re: [PATCH v3 tip/core/rcu 40/40] rcu: Make non-preemptive schedule be Tasks RCU quiescent state

On Thu, Sep 28, 2017 at 09:05:14AM -0700, Paul E. McKenney wrote: > > do_async_page_fault+0x72/0x90 arch/x86/kernel/kvm.c:271 > > async_page_fault+0x22/0x30 arch/x86/entry/entry_64.S:1069 > > RIP: 0010:format_decode+0x240/0x830 lib/vsprintf.c:1996 > > RSP: 0018:88003b2df520 EFLAGS: 00010283

Re: [PATCH v3 tip/core/rcu 40/40] rcu: Make non-preemptive schedule be Tasks RCU quiescent state

On Thu, Sep 28, 2017 at 03:38:16PM +, Levin, Alexander (Sasha Levin) wrote: > On Thu, Sep 28, 2017 at 05:30:55AM -0700, Paul E. McKenney wrote: > >Hmmm... kernel/rcu/tree_plugin.h:329 thinks that someone slept (as opposed > >to was preempted) in an RCU read-side critical section. Sasha, was

Re: [PATCH v3 tip/core/rcu 40/40] rcu: Make non-preemptive schedule be Tasks RCU quiescent state

On Thu, Sep 28, 2017 at 03:38:16PM +, Levin, Alexander (Sasha Levin) wrote: > On Thu, Sep 28, 2017 at 05:30:55AM -0700, Paul E. McKenney wrote: > >Hmmm... kernel/rcu/tree_plugin.h:329 thinks that someone slept (as opposed > >to was preempted) in an RCU read-side critical section. Sasha, was

Re: [PATCH v3 tip/core/rcu 40/40] rcu: Make non-preemptive schedule be Tasks RCU quiescent state

On Thu, Sep 28, 2017 at 03:38:16PM +, Levin, Alexander (Sasha Levin) wrote: > On Thu, Sep 28, 2017 at 05:30:55AM -0700, Paul E. McKenney wrote: > >On Thu, Sep 28, 2017 at 02:37:20AM -0700, Sasha Levin wrote: > >> On Wed, Apr 19, 2017 at 9:58 AM, Paul E. McKenney > >>

Re: [PATCH v3 tip/core/rcu 40/40] rcu: Make non-preemptive schedule be Tasks RCU quiescent state

On Thu, Sep 28, 2017 at 03:38:16PM +, Levin, Alexander (Sasha Levin) wrote: > On Thu, Sep 28, 2017 at 05:30:55AM -0700, Paul E. McKenney wrote: > >On Thu, Sep 28, 2017 at 02:37:20AM -0700, Sasha Levin wrote: > >> On Wed, Apr 19, 2017 at 9:58 AM, Paul E. McKenney > >> wrote: > >> > Currently,

Re: [PATCH v3 tip/core/rcu 40/40] rcu: Make non-preemptive schedule be Tasks RCU quiescent state

On Thu, Sep 28, 2017 at 05:30:55AM -0700, Paul E. McKenney wrote: >On Thu, Sep 28, 2017 at 02:37:20AM -0700, Sasha Levin wrote: >> On Wed, Apr 19, 2017 at 9:58 AM, Paul E. McKenney >> wrote: >> > Currently, a call to schedule() acts as a Tasks RCU quiescent state >> >

Re: [PATCH v3 tip/core/rcu 40/40] rcu: Make non-preemptive schedule be Tasks RCU quiescent state

On Thu, Sep 28, 2017 at 05:30:55AM -0700, Paul E. McKenney wrote: >On Thu, Sep 28, 2017 at 02:37:20AM -0700, Sasha Levin wrote: >> On Wed, Apr 19, 2017 at 9:58 AM, Paul E. McKenney >> wrote: >> > Currently, a call to schedule() acts as a Tasks RCU quiescent state >> > only if a context switch

Re: [PATCH v3 tip/core/rcu 40/40] rcu: Make non-preemptive schedule be Tasks RCU quiescent state

On Thu, Sep 28, 2017 at 02:37:20AM -0700, Sasha Levin wrote: > On Wed, Apr 19, 2017 at 9:58 AM, Paul E. McKenney > wrote: > > Currently, a call to schedule() acts as a Tasks RCU quiescent state > > only if a context switch actually takes place. However, just the > >

Re: [PATCH v3 tip/core/rcu 40/40] rcu: Make non-preemptive schedule be Tasks RCU quiescent state

On Thu, Sep 28, 2017 at 02:37:20AM -0700, Sasha Levin wrote: > On Wed, Apr 19, 2017 at 9:58 AM, Paul E. McKenney > wrote: > > Currently, a call to schedule() acts as a Tasks RCU quiescent state > > only if a context switch actually takes place. However, just the > > call to schedule() guarantees

Re: [PATCH v3 tip/core/rcu 40/40] rcu: Make non-preemptive schedule be Tasks RCU quiescent state

On Wed, Apr 19, 2017 at 9:58 AM, Paul E. McKenney wrote: > Currently, a call to schedule() acts as a Tasks RCU quiescent state > only if a context switch actually takes place. However, just the > call to schedule() guarantees that the calling task has moved off of >

Re: [PATCH v3 tip/core/rcu 40/40] rcu: Make non-preemptive schedule be Tasks RCU quiescent state

On Wed, Apr 19, 2017 at 9:58 AM, Paul E. McKenney wrote: > Currently, a call to schedule() acts as a Tasks RCU quiescent state > only if a context switch actually takes place. However, just the > call to schedule() guarantees that the calling task has moved off of > whatever tracing trampoline

[PATCH v3 tip/core/rcu 40/40] rcu: Make non-preemptive schedule be Tasks RCU quiescent state

Currently, a call to schedule() acts as a Tasks RCU quiescent state only if a context switch actually takes place. However, just the call to schedule() guarantees that the calling task has moved off of whatever tracing trampoline that it might have been one previously. This commit therefore

[PATCH v3 tip/core/rcu 40/40] rcu: Make non-preemptive schedule be Tasks RCU quiescent state

Currently, a call to schedule() acts as a Tasks RCU quiescent state only if a context switch actually takes place. However, just the call to schedule() guarantees that the calling task has moved off of whatever tracing trampoline that it might have been one previously. This commit therefore