Xen devel - Apr 2007 - FATAL PAGE FAULT - xen-3.0.5-rc2 and xen-3.0.5-rc3

Hi,

I have the following errors:

Panic on CPU0:
FATAL PAGE FAULT
[error_code=0002]
Faulting linear address: fbfb1100

More details in snapshots as below as I don''t have serial console
cables thus unable to get full details of the errors which I will try
to get one and send the full error if needed:

http://choon.net/xenerrors/Image000.jpg
http://choon.net/xenerrors/Image001.jpg
http://choon.net/xenerrors/Image002.jpg

I am able to use xen-3.0-unstable from hg to boot with the same
configurations from date 11 Feb 2007 build but unable to get current
xen-3.0-unstable from hg to boot now with the same error as 3.0.5-rc2
and 3.0.5-rc3.

Thanks.

Kindest regards,
Choon

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

On 26/4/07 10:18, "Teck Choon Giam" <giamteckchoon@gmail.com>
wrote:
> I have the following errors:
> 
> Panic on CPU0:
> FATAL PAGE FAULT
> [error_code=0002]
> Faulting linear address: fbfb1100
> 
> More details in snapshots as below as I don''t have serial console
> cables thus unable to get full details of the errors which I will try
> to get one and send the full error if needed:
It''s bizarre that only you have seen this crash. It should be easy to
track
down though. If you''re happy to modify C code, can you try adding:
 printk("***TICKER_INIT %d %d\n", cpu, spc->ticker.cpu);
To csched_start_tickers() in common/sched_credit.c, immediately after the
line ''spc = ...''.

Alternatively I can send you a patch. The extra message should come out
twice immediately before the crash message and it would be interesting to
know what the messages say.

Also, the xen-syms file that matches the crash-message JPEGs you linked to
would be interesting. Perhaps you could zip that up and send it to me
privately if you still have it.

 -- Keir


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

On 4/26/07, Keir Fraser <keir@xensource.com>
wrote:
> On 26/4/07 10:18, "Teck Choon Giam"
<giamteckchoon@gmail.com> wrote:
>
> > I have the following errors:
> >
> > Panic on CPU0:
> > FATAL PAGE FAULT
> > [error_code=0002]
> > Faulting linear address: fbfb1100
> >
> > More details in snapshots as below as I don''t have serial
console
> > cables thus unable to get full details of the errors which I will try
> > to get one and send the full error if needed:
>
> It''s bizarre that only you have seen this crash. It should be easy
to track
> down though. If you''re happy to modify C code, can you try adding:
>  printk("***TICKER_INIT %d %d\n", cpu, spc->ticker.cpu);
> To csched_start_tickers() in common/sched_credit.c, immediately after the
> line ''spc = ...''.
>From your suggestion make me think it is related to sched=sedf option.
 So with that option I has problem to boot xen but without sched=sedf
option it boot nicely... ...
>
> Alternatively I can send you a patch. The extra message should come out
> twice immediately before the crash message and it would be interesting to
> know what the messages say.
You don''t need to send me the patch as I have followed your
instruction to add that line :)

Please see attached txt file for boot messages with and without sched=sedf.
>
> Also, the xen-syms file that matches the crash-message JPEGs you linked to
> would be interesting. Perhaps you could zip that up and send it to me
> privately if you still have it.
Don''t need the JPEGs I guess since the crash message can be found in
xen-3.0.5-rc3-with-sched_eq_sedf-crash.txt file attached with the
difference of your suggested below line of course :P

printk("***TICKER_INIT %d %d\n", cpu, spc->ticker.cpu);
>
>  -- Keir
Thanks Keir =)

Kindest regards,
Choon
P.S. Sorry, if anyone won''t like to see those files in attachment then
I will take note and won''t send in any attachments. 
Apology.
>


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

On 27/4/07 21:00, "Teck Choon Giam" <giamteckchoon@gmail.com>
wrote:
>> It''s bizarre that only you have seen this crash. It should be
easy to track
>> down though. If you''re happy to modify C code, can you try
adding:
>>  printk("***TICKER_INIT %d %d\n", cpu, spc->ticker.cpu);
>> To csched_start_tickers() in common/sched_credit.c, immediately after
the
>> line ''spc = ...''.
> 
>> From your suggestion make me think it is related to sched=sedf option.
>  So with that option I has problem to boot xen but without sched=sedf
> option it boot nicely... ...
Oh, I didn''t realise you were using sched=sedf. I can test this for
myself
and fix before 3.0.5, so sched=sedf continues to work.

 -- Keir



_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

On 4/28/07, Keir Fraser <Keir.Fraser@cl.cam.ac.uk> wrote:
> Oh, I didn''t realise you were using sched=sedf. I can test this
for myself
> and fix before 3.0.5, so sched=sedf continues to work.
TIA =)

Kindest regards,
Choon

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel