[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

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

On 23 February 2015 at 05:51, Dario Faggioli <dario.faggioli@xxxxxxxxxx> wrote:
> I see. And you've done the right thing in providing all these info! :-)
> I don't think these things about Dom0 have any influence in the repoted
> issue. However, since you're building Dom0, let me ask this: are you
> also building Xen from sources, or are you running it out of some distro
> package? If you're not building it, could you (e.g., to test debug
> patches, etc)?

I did build Xen from sources. I will rebuild with debug enabled.

> That's it: "debug=n". But you have to rebuild Xen to change this,

Ah my mistake. I just saw that in the output and assumed it was a boot

> Weird. I've never seen this before, TBH. You said you're on 4.5. Have
> you used earlier version of Xen? If yes, was this bug there?

I have only used Xen 4.5.

> I'll take a look. In the meanwhile, if you can setup a debug build and,
> when the bug manifests, send the stack trace again, that would
> (hopefully) help.

I have some potentially good news to report: I realized that the 3 of
my domU that were based on OpenWrt, whilst I had managed to get them
on the same kernel revision as the Dom0, were running as 32bit. I
fiddled around with the OpenWrt build system some more and got it to
build 3.18.7/x86_64 Xen images. I booted those, and now there aren't
any 32 bit guests at all. So far I have 2 days and 10 hours of uptime,
which is by far the longest the system has ever ran. I will prepare
the debug version of Xen, and then wait to install for another few
days. I want to see how long it stays stable like this. I know it
could just be a coincidence...


Xen-users mailing list



Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.