Re: [PATCH v2] sched/cputime: add steal clock warp handling

2016-06-06 Thread Paolo Bonzini
On 06/06/2016 16:22, Rik van Riel wrote: >> Sorry for being late---again, I'd like to give a shot to a fix in KVM >> guest code. > > This is in the guest code. The key word is "KVM" guest code, :) i.e. not in kernel/sched/cputime.c but in arch/x86/kernel/kvm.c. Thanks, Paolo > Do you mean a

Re: [PATCH v2] sched/cputime: add steal clock warp handling

2016-06-06 Thread Paolo Bonzini
On 06/06/2016 16:22, Rik van Riel wrote: >> Sorry for being late---again, I'd like to give a shot to a fix in KVM >> guest code. > > This is in the guest code. The key word is "KVM" guest code, :) i.e. not in kernel/sched/cputime.c but in arch/x86/kernel/kvm.c. Thanks, Paolo > Do you mean a

Re: [PATCH v2] sched/cputime: add steal clock warp handling

2016-06-06 Thread Rik van Riel
On Mon, 2016-06-06 at 15:44 +0200, Paolo Bonzini wrote: > > On 03/06/2016 15:10, Rik van Riel wrote: > > > > On Fri, 2016-06-03 at 13:21 +0800, Wanpeng Li wrote: > > > > > > From: Wanpeng Li > > > > > > I observed that sometimes st is 100% instantaneous, then idle is >

Re: [PATCH v2] sched/cputime: add steal clock warp handling

2016-06-06 Thread Rik van Riel
On Mon, 2016-06-06 at 15:44 +0200, Paolo Bonzini wrote: > > On 03/06/2016 15:10, Rik van Riel wrote: > > > > On Fri, 2016-06-03 at 13:21 +0800, Wanpeng Li wrote: > > > > > > From: Wanpeng Li > > > > > > I observed that sometimes st is 100% instantaneous, then idle is > > > 100%  > > > even if

Re: [PATCH v2] sched/cputime: add steal clock warp handling

2016-06-06 Thread Paolo Bonzini
On 03/06/2016 15:10, Rik van Riel wrote: > On Fri, 2016-06-03 at 13:21 +0800, Wanpeng Li wrote: >> From: Wanpeng Li >> >> I observed that sometimes st is 100% instantaneous, then idle is >> 100% >> even if there is a cpu hog on the guest cpu after the cpu hotplug >>

Re: [PATCH v2] sched/cputime: add steal clock warp handling

2016-06-06 Thread Paolo Bonzini
On 03/06/2016 15:10, Rik van Riel wrote: > On Fri, 2016-06-03 at 13:21 +0800, Wanpeng Li wrote: >> From: Wanpeng Li >> >> I observed that sometimes st is 100% instantaneous, then idle is >> 100% >> even if there is a cpu hog on the guest cpu after the cpu hotplug >> comes >> back(N.B. this

Re: [PATCH v2] sched/cputime: add steal clock warp handling

2016-06-03 Thread Rik van Riel
On Fri, 2016-06-03 at 13:21 +0800, Wanpeng Li wrote: > From: Wanpeng Li > > I observed that sometimes st is 100% instantaneous, then idle is > 100%  > even if there is a cpu hog on the guest cpu after the cpu hotplug > comes  > back(N.B. this can not always be readily

Re: [PATCH v2] sched/cputime: add steal clock warp handling

2016-06-03 Thread Rik van Riel
On Fri, 2016-06-03 at 13:21 +0800, Wanpeng Li wrote: > From: Wanpeng Li > > I observed that sometimes st is 100% instantaneous, then idle is > 100%  > even if there is a cpu hog on the guest cpu after the cpu hotplug > comes  > back(N.B. this can not always be readily reproduced). I add trace to 

[PATCH v2] sched/cputime: add steal clock warp handling

2016-06-02 Thread Wanpeng Li
From: Wanpeng Li I observed that sometimes st is 100% instantaneous, then idle is 100% even if there is a cpu hog on the guest cpu after the cpu hotplug comes back(N.B. this can not always be readily reproduced). I add trace to capture it as below: cpuhp/1-12[001]

[PATCH v2] sched/cputime: add steal clock warp handling

2016-06-02 Thread Wanpeng Li
From: Wanpeng Li I observed that sometimes st is 100% instantaneous, then idle is 100% even if there is a cpu hog on the guest cpu after the cpu hotplug comes back(N.B. this can not always be readily reproduced). I add trace to capture it as below: cpuhp/1-12[001] d.h1 167.461657: