Re: [Xen-devel] [PATCH] xen:rtds:fix bug in accounting budget

[Dario Faggioli <dario.faggioli@xxxxxxxxxx>]

On Wed, 2016-10-19 at 11:13 -0400, Meng Xu wrote:
> The bug is introduced in Xen 4.7 when we converted RTDS scheduler
> from quantum-driven model to event-driven model.
> We assumed rt_schedule() is always called for a VCPU
> before the VCPUs budget replenished handler.
>
No, we didn't.

Or at least, I've never done so, and tried as hard as I could to tell
you guys to not make any assumptions on who run first.

So, yes, I agree that, if there is code that only works under such
assumption, it's a bug.

> This assumption does not hold, when system is overloaded, or
> when the VCPU budget is almost equal its period.
> 
> Buggy behavior:
> 1) A VCPU may get less budget that assigned in a period.
> 2) A full capacity VCPU, i.e., a VCPU whose period is equal to
> budget,
>    may not get any budget in some period.
> 
So, there are two bugs. And things are very subtle, as far as I can
judge from both the bugs description and the code.

It would be, therefore, a lot more clear if you could send _one_ patch
per bug.

> Bug analysis:
> 1) A VCPU deadline can be fast-forwarded by more than one period.
>    However, the VCPU last_start time was not updated immediately.
>    If rt_schedule() is called after rt_update_deadline(), which
> happens
>    when VCPU budget is equal to period or when VCPU has deadline
> miss,
>    burn_budget() will burn the budget that was just replenished,
>    although the replenished budget should be used in the most recent
> period only.
> 
-EPARSE.

I've looked at the code and try to match current behavior, your
proposed change, and this description, but failed.

Can you be a little more precise and specific about what happens when?

I'll keep looking and thinking, but any help in making all this a bit
more clear would be very welcome.

> 2) When a full capacity VCPU depletes its budget and is context
> switching out,
>    but has not updated the cores current running VCPU,
>
"has not updated the cores current running VCPU,"

I've not idea what this sentence means. What is it that has not yet
been updated?

>    the budget replenish timer may be triggerred.
>    The replenish handler failed to re-schedule the full capacity VCPU
>    because it thought the VCPU is running.
> 
>    When a VCPU budget is replenished, we try to tickle a CPU.
>    When we find a core for a VCPU to tickle and the VCPU is context
> switching out,
>    we will always tickle the core where the VCPU was running,
>    if the VCPU cannot find another core to tickle
> 
Can't understand much again... I guess this is the description of the
solution to the bug?

> This bug was reported by Dagaen Golomb
> 
You can give credit by using the following tag:

Reported by: Dagaen Golomb <xxx@xxxxxx>

Thanks and Regards,
Dario
-- 
<<This happens because I choose it to happen!>> (Raistlin Majere)
-----------------------------------------------------------------
Dario Faggioli, Ph.D, http://about.me/dario.faggioli
Senior Software Engineer, Citrix Systems R&D Ltd., Cambridge (UK)

Attachment: signature.asc
Description: This is a digitally signed message part

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
https://lists.xen.org/xen-devel