Re: hypervisor crash with latest xen-unstable

[Juergen Gross <juergen.gross@ts.fujitsu.com>]

On 11/25/10 18:14, Keir Fraser wrote:
> On 25/11/2010 10:27, "Juergen Gross"<juergen.gross@ts.fujitsu.com>  wrote:
>
>> (XEN) Xen call trace:
>> (XEN)    [<ffff82c4801212f9>] do_softirq+0x51/0x7a
>> (XEN)
>>
>> I suspect the waitqueue stuff broke the continue_hypercall_on_cpu() stuff.
>
> Then please bisect to confirm that.

Removing the ASSERT(!in_atomic()) in do_softirq() and schedule() makes it
work again. And these assertions were added with cs22395.

The question now is, whether these assertions are wrong or if it is just
working by chance without them.
It seems to be necessary to be able to call schedule() in this case, as adding

if (in_atomic()) return;

to do_softirq() made the hypercall hang.


Juergen

-- 
Juergen Gross                 Principal Developer Operating Systems
TSP ES&S SWE OS6                       Telephone: +49 (0) 89 3222 2967
Fujitsu Technology Solutions              e-mail: juergen.gross@ts.fujitsu.com
Domagkstr. 28                           Internet: ts.fujitsu.com
D-80807 Muenchen                 Company details: ts.fujitsu.com/imprint.html