Re: [Xen-devel] Latest development (master) Xen fails to boot on HP ProLiant DL20 GEN10

[Roman Shaposhnik <roman@xxxxxxxxxx>]

Sorry -- was traveling last week, but I'm still very curious to get to
the bottom of this:

On Tue, Oct 1, 2019 at 1:25 AM Jan Beulich <jbeulich@xxxxxxxx> wrote:
>
> On 01.10.2019 00:38, Roman Shaposhnik wrote:
> > Btw, forgot to attach the patch with maxcpus=2 -- interestingly enough
> > Xen seems to hang much further down than before (basically after
> > attempting to build out Dom0)
>
> All 3 logs contain
>
> (XEN) TSC_DEADLINE disabled due to Errata; please update microcode to version 
> 0x52 (or later)

Ok. This makes some sense. Btw, in a situation like this, do we expect
Xen to cope with a broken microcode or we always expect microcode to
be updated?

> Please load up-to-date microcode on the system and,

I couldn't find any kind of HP guide on how to do this on this box.
Any chance someone here may have a pointer for me?

> preferably with Andrew's suggestions also applied, re-post the logs. I notice 
> that
> even logs 1 and 2 have "Brought up 4 CPUs", other than you've
> indicated in your initial report. This suggests something's broken
> _after_ bringup of secondary CPUs, not while bringing them up. Log
> 3 effectively seems to confirm this.
>
> Seeing that "max_cstate=1" did help, as another next step could you
> try whether "mwait-idle=0" makes enough of a difference (it'll
> likely make a difference initially, as it makes the system
> effectively stay in a "max_cstate=1"-like mode until Dom0 has booted
> up far enough; the question this is going to be whether a hang still
> occurs one Dom0 has uploaded C-state data)?

Ok. Will do all these experiments tomorrow.

My plan is to use latest Xen master + build it with extra debug info
as was suggested earlier in this thread.

Thanks,
Roman.

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxxx
https://lists.xenproject.org/mailman/listinfo/xen-devel