Re: [Xen-devel] Reproducible system crash

[Keir Fraser <Keir.Fraser@xxxxxxxxxxxx>]

On 16 Mar 2005, at 18:39, Peter Joanes wrote:

> On Wednesday 16 March 2005 16:22, Keir Fraser wrote:
> > The stack backtrace also indicates that the code at address 0xc010990a
> > in XenLinux may also be interesting (that is the code point that 
> > caused
> > the crash).
> > 'objdump -d vmlinux' and then grep the output for that address....
> Here's what I think is the relevant section:
Not so useful unfortunately. Here's a slightly harder thing to try:

1. Edit linux/include/asm-xen/asm-i386/pgtable-2level.h, and change 
set_pte_batched() to use xen_l1_pgentry_update rather than 
queue_l1_pgentry_update. Also, kill current definition of set_pte() and 
redefine it to the same as the new set_pte_batched().
2. Edit xen/arch/x86/memory.c, at the test that fails for you ('Bad L1 
type'). In the error path, before returning 0, add:
    extern void show_guest_stack(void);
    show_guest_stack();

This should cause us to get a backtrace of the guest kernel state when 
the bad PTE is written. If you send a link to that along with your 
vmlinux file then we can work out where the _PAGE_GLOBAL bit is coming 
from.
 -- Keir



-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users.
Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click
_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxxxx
https://lists.sourceforge.net/lists/listinfo/xen-devel