From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jeremy Fitzhardinge Subject: Re: [RFC v2 4/7] change kernel accounting to include steal time Date: Mon, 30 Aug 2010 10:30:04 -0700 Message-ID: <4C7BEA9C.1060605@goop.org> References: <1283184391-7785-1-git-send-email-glommer@redhat.com> <1283184391-7785-2-git-send-email-glommer@redhat.com> <1283184391-7785-3-git-send-email-glommer@redhat.com> <1283184391-7785-4-git-send-email-glommer@redhat.com> <1283184391-7785-5-git-send-email-glommer@redhat.com> <1283184391-7785-6-git-send-email-glommer@redhat.com> <1283184391-7785-7-git-send-email-glommer@redhat.com> <1283184391-7785-8-git-send-email-glommer@redhat.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-15 Content-Transfer-Encoding: 7bit Cc: kvm@vger.kernel.org, avi@redhat.com, zamsden@redhat.com, mtosatti@redhat.com, riel@redhat.com, peterz@infradead.org, mingo@elte.hu To: Glauber Costa Return-path: Received: from claw.goop.org ([74.207.240.146]:34894 "EHLO claw.goop.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1756042Ab0H3RaJ (ORCPT ); Mon, 30 Aug 2010 13:30:09 -0400 In-Reply-To: <1283184391-7785-8-git-send-email-glommer@redhat.com> Sender: kvm-owner@vger.kernel.org List-ID: On 08/30/2010 09:06 AM, Glauber Costa wrote: > This patch proposes a common steal time implementation. When no > steal time is accounted, we just add a branch to the current > accounting code, that shouldn't add much overhead. How is stolen time logically any different from a CPU running slowly due to HT or power management? Is it worth trying to handle them in the same way? (I'm mostly picking on the "_from_hypervisor" part, since that seems over-specific.) Why not have a get_unstolen_time() function which just returns sched_clock() in the normal case, but can return less? > When we do want to register steal time, we proceed as following: > - if we would account user or system time in this tick, and there is > out-of-cpu time registered, we skip it altogether, and account steal > time only. > - if we would account user or system time in this tick, and we got the > cpu for the whole slice, we proceed normaly. > - if we are idle in this tick, we flush out-of-cpu time to give it the > chance to update whatever last-measure internal variable it may have. > > This approach is simple, but proved to work well for my test scenarios. > in a UP guest on UP host, with a cpu-hog in both guest and host shows > ~ 50 % steal time. steal time is also accounted proportionally, if > nice values are given to the host cpu-hog. > > A cpu-hog in the host with no load in the guest, produces 0 % steal time, > with 100 % idle, as one would expect. > > Signed-off-by: Glauber Costa > --- > include/linux/sched.h | 1 + > kernel/sched.c | 29 +++++++++++++++++++++++++++++ > 2 files changed, 30 insertions(+), 0 deletions(-) > > diff --git a/include/linux/sched.h b/include/linux/sched.h > index 0478888..e571ddd 100644 > --- a/include/linux/sched.h > +++ b/include/linux/sched.h > @@ -312,6 +312,7 @@ long io_schedule_timeout(long timeout); > extern void cpu_init (void); > extern void trap_init(void); > extern void update_process_times(int user); > +extern cputime_t (*hypervisor_steal_time)(void); > extern void scheduler_tick(void); > > extern void sched_show_task(struct task_struct *p); > diff --git a/kernel/sched.c b/kernel/sched.c > index f52a880..9695c92 100644 > --- a/kernel/sched.c > +++ b/kernel/sched.c > @@ -3157,6 +3157,16 @@ unsigned long long thread_group_sched_runtime(struct task_struct *p) > return ns; > } > > +cputime_t (*hypervisor_steal_time)(void) = NULL; > + > +static inline cputime_t get_steal_time_from_hypervisor(void) > +{ > + if (!hypervisor_steal_time) > + return 0; > + return hypervisor_steal_time(); > +} > + > + > /* > * Account user cpu time to a process. > * @p: the process that the cpu time gets accounted to > @@ -3169,6 +3179,12 @@ void account_user_time(struct task_struct *p, cputime_t cputime, > struct cpu_usage_stat *cpustat = &kstat_this_cpu.cpustat; > cputime64_t tmp; > > + tmp = get_steal_time_from_hypervisor(); > + if (tmp) { > + account_steal_time(tmp); > + return; > + } Is that all? Does the scheduler use account_steal_time() to adjust its scheduling decisions, or is it just something that gets shown to users? I thought just the latter. But if all you're doing is calling account_steal_time(), why bother with all this get_steal_time_from_hypervisor() stuff? The hypervisor-specific code can just call account_steal_time() directly. > + > /* Add user time to process. */ > p->utime = cputime_add(p->utime, cputime); > p->utimescaled = cputime_add(p->utimescaled, cputime_scaled); > @@ -3234,6 +3250,12 @@ void account_system_time(struct task_struct *p, int hardirq_offset, > return; > } > > + tmp = get_steal_time_from_hypervisor(); > + if (tmp) { > + account_steal_time(tmp); > + return; > + } > + > /* Add system time to process. */ > p->stime = cputime_add(p->stime, cputime); > p->stimescaled = cputime_add(p->stimescaled, cputime_scaled); > @@ -3276,6 +3298,13 @@ void account_idle_time(cputime_t cputime) > cputime64_t cputime64 = cputime_to_cputime64(cputime); > struct rq *rq = this_rq(); > > + /* > + * if we're idle, we don't account it as steal time, since we did > + * not want to run anyway. We do call the steal function, however, to > + * give the guest the chance to flush its internal buffers > + */ > + get_steal_time_from_hypervisor(); Eh? This doesn't make much sense. What side-effects is get_steal_time_from_hypervisor() expected to have? If there's some hypervisor-specific implementation detail, why not wrap that up in a specific function rather than overloading this one? J