[lkml]   [2014]   [Mar]   [19]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
Messages in this thread
Subject[BUG] Paravirtual time accounting / IRQ time accounting
In consolidated environments, when there are multiple virtual machines (VMs)
running on one CPU core, timekeeping will be a problem to the guest OS.
Here, I report my findings about Linux process scheduler.

Linux CFS relies on rq->clock_task to charge each task, determine
vruntime, etc.

When CONFIG_IRQ_TIME_ACCOUNTING is enabled, the time spent on serving IRQ
will be excluded from updating rq->clock_task.
When CONFIG_PARAVIRT_TIME_ACCOUNTING is enabled, the time stolen by
the hypervisor
will also be excluded from updating rq->clock_task.

I put three KVM guests on one core and run hackbench in each guest. I
find that
in the guests, rq->clock_task stays *unchanged*. The malfunction
embarrasses CFS.

static void update_rq_clock_task(struct rq *rq, s64 delta)
... ...
irq_delta = irq_time_read(cpu_of(rq)) - rq->prev_irq_time;
... ...
rq->prev_irq_time += irq_delta;
delta -= irq_delta;

if (static_key_false((&paravirt_steal_rq_enabled))) {
steal = paravirt_steal_clock(cpu_of(rq));
steal -= rq->prev_steal_time_rq;
... ...
rq->prev_steal_time_rq += steal;
delta -= steal;

rq->clock_task += delta;
... ...
"delta" -> the intended increment to rq->clock_task
"irq_delta" -> the time spent on serving IRQ (hard + soft)
"steal" -> the time stolen by the underlying hypervisor
"irq_delta" is calculated based on sched_clock_cpu(), which is vulnerable
to VM scheduling delays. "irq_delta" can include part or whole of "steal".
I observe that [irq_delta + steal >> delta].
As a result, "delta" becomes zero. That is why rq->clock_task stops.

Please confirm this bug. Thanks.

Luwei Cheng
CS student
The University of Hong Kong

 \ /
  Last update: 2014-03-19 11:21    [W:0.111 / U:5.004 seconds]
©2003-2018 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site