Re: [Xen-devel] Bug in x86 instruction emulator?

["Jan Beulich" <jbeulich@xxxxxxxx>]

>>> <wogiz@xxxxxxxxxxxxxxx> 04/07/16 3:28 AM >>>
>On 2016-04-06 10:55, Andrew Cooper wrote:
>> Can you grab the full register state at the point of Xorgs crash?  
>> `info
>> regs` in gdb?
>> 
>> The instruction in use, `movaps` is specified to fault if the memory
>> operand isn't aligned on a 16byte boundary.  Therefore, if %rax in this
>> case isn't a multiple of 16, this is a code generation bug, rather than
>> an emulation bug.
>
>Attached is the full register state.

So it is even page aligned. Which raises the question whether we're
mishandling something here when the page needs bringing in from
disk by the guest.

>I'm very interested in getting to the bottom of this, so please let me 
>know if I can do anything to help.

We'd need to know which exact exception (including error code and,
in the case of #PF, CR2 value) gets raised to the guest by what
specific piece of code in the hypervisor. That'll likely mean some
instrumentation of the hypervisor code.

Jan


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