Mailinglist Archive: radeonhd (436 mails)

[Egbert Eich <eich@xxxxxxx>]

Hans Ulrich Niedermann writes:
> Sven Lankes wrote:
> > On Sat, Jul 19, 2008 at 01:31:13PM +0200, Sven Lankes wrote:
> >
> > > I have attached the Xorg.0.log 
> >
> > And here comes the promised attachment ... 
>
> -------<snip>---------------------------------------
> Backtrace:
> 0: /usr/bin/Xorg(xf86SigHandler+0x79) [0x80bc249]
> 1: [0x132400]
> 2: /usr/bin/Xorg [0x80e381c]
> 3: /usr/bin/Xorg(xf86InitialConfiguration+0x1131) [0x80e7181]
> 4: /usr/lib/xorg/modules/drivers//radeonhd_drv.so(RHDRandrPreInit+0x56a) 
> [0x5139fa]
> 5: /usr/lib/xorg/modules/drivers//radeonhd_drv.so [0x501326]
> 6: /usr/bin/Xorg(InitOutput+0x96f) [0x80a40bf]
> 7: /usr/bin/Xorg(main+0x279) [0x806b479]
> 8: /lib/libc.so.6(__libc_start_main+0xe5) [0x1c95d5]
> 9: /usr/bin/Xorg(FontFileCompleteXLFD+0x221) [0x806aa61]
>
> Fatal server error:
> Caught signal 11.  Server aborting
> -------<snip>---------------------------------------
>
> I guess either
>   - I set me up a rawhide system for testing and hope I can
>     reproduce the issue
>   - some radeonhd dev has an idea

Not really. A more verbose log may help here.
The crash doesn't seem to happen in the driver.
Appearantly a corner case condition is triggered 
because no output is detected to be connected.

A more verbose log may actually help to get the
information we need.

Sven: please log into the text console and do:

X -logverbose 7

After the Xserver has crashed, please email the new log file.

>   - you will need to start Xorg in gdb, and do the backtrace dance:
>     [ SIGSEGV thingie ]
>     (gdb) set pagination off
>     (gdb) bt
>     (gdb) bt full
>
> BTW: Can one give gdb a file to write the "bt"/"bt full" output to?

This would indeed help to spot the cause of the crash. But unless
debug symbols are available the output will be of limited value.

Cheers,
        Egbert.

-- 
To unsubscribe, e-mail: radeonhd+unsubscribe@xxxxxxxxxxxx
For additional commands, e-mail: radeonhd+help@xxxxxxxxxxxx