Re: [Xen-users] [BUG] Xen BUG at sched_credit.c:243

[Dario Faggioli <dario.faggioli@xxxxxxxxxx>]

On Fri, 2015-03-06 at 10:03 +0000, George Dunlap wrote:
> On 02/25/2015 04:09 PM, C. B. wrote:

> > Well since I had your attention I decided to install the debug build
> > of xen.efi right away, and put back the 32 bit guests I had been using
> > previously, returning the system to the state it was in when I was
> > having issues. After 30 hours, it did explode again but unfortunately
> > I did not capture anything from the serial port. I am not sure if it
> > just didn't output anything, or if my TTL to USB serial cable let me
> > down. Either way, I restarted and I'm watching it closely again. At
> > the moment it's up to 15 hours of uptime.
> 
> CB,
> 
> Has there been any further information on this?
> 
Yeah, I also wanted to ask about this...

> The original bug is actually very concerning; I wonder if it's related
> to c/s e8bb583 , which went into the development branch about a month ago.
> 
That would be weird. That commit is all about /arch/arm, while CB is on
x86, IIRC.

Also, it fiddles with the .context_saved hook, which is only relevant
for Credit2 and RTDS, while the bug seems to be manifesting while using
Credit1.

> Any chance you could try the "master" branch of Xen 4.6-unstable, and
> see if your problem is fixed at all?
>
This is probably worth anyway, but apart that change, there really
hasn't happened much in scheduling (yet :-)) since 4.5 release...

Regards,
Dario

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

_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxx
http://lists.xen.org/xen-users