[Xen-devel] progress and a problem

[ron minnich <rminnich@xxxxxxxx>]

I've got xen calling into my hypervisor_callback but i'm seeing a very 
weird problem. On the first instruction, which is a PUSHL %eax, I take a 
page fault. The values of %ss and %esp on stack make no sense, and in 
fact change from run to run of the guest OS. All other values stay the 
same. PC does look right. 

If i just put an IRETL as the first instruction of the hypervisor_callback
all is well -- no problem. Does Xen pass in a read-only stack segment to
the callback? I don't see similar problems on the Linux side.

If I get a stack dump, everything EXCEPT ss/esp look fine -- cs, ds, etc. 
are all right. 

I have done a HYPERVISOR_stack_switch call, and all other interrupts work 
save this one.

what function in Xen calls the hypervisor_callback in the guest?

ron



-------------------------------------------------------
This SF.Net email sponsored by Black Hat Briefings & Training.
Attend Black Hat Briefings & Training, Las Vegas July 24-29 - 
digital self defense, top technical experts, no vendor pitches, 
unmatched networking opportunities. Visit www.blackhat.com
_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxxxx
https://lists.sourceforge.net/lists/listinfo/xen-devel