Hello again,
attached you'll find another crash dump from today. Don't know if
it gives you more information than the last one.
Just FYI, this is a system with an Intel Mainboard (H87 chipset)
and a Core i5-4670 CPU.
Best regards
Thimo
Am 09.08.2013 23:44, schrieb Andrew Cooper:
On 09/08/13 22:40, Andrew Cooper
wrote:
So according to my debugging, we really have just pushed the
same irq which we have subsequently seen again unexpectedly.
This bug has only ever been seen on Haswell hardware, and
appears linked to running HVM guests.
So either there is an erroneous ACK the LAPIC which is clearing
the ISR before the PEOI stack is expecting (which I
"can't"
Apologies for the confusion.
~Andrew
obviously
see, looking at the code), or something more funky is going on
with the hardware.
CC'ing in the Intel maintainers: Do you have any ideas? Could
this be related to APICv?
~Andrew
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel