I can easily trigger an endless stream of the following three lines
[drm:radeon_cp_idle] *ERROR* radeon_cp_idle called without lock held,
held 0 owner eb0908c0 eb0908c0
[drm:radeon_cp_reset] *ERROR* radeon_cp_reset called without lock held,
held 0 owner eb0908c0 eb0908c0
[drm:radeon_cp_start] *ERROR* radeon_cp_start called without lock held,
held 0 owner eb0908c0 eb0908c0
and thereby also lock my system completely up.
I did not dig into this problem, except for applying the attached patch
in order to get some more info and BUG_ON after the first round of
errors, so that sysRq remained usable and I could produce a bunch of
extra data in the dmesg.
Attached is also my config and the output of `lspci -vv' and `dmesg'.
Any ideas?
Simon Holm Thøgersen
> I can easily trigger an endless stream of the following three lines
>
> [drm:radeon_cp_idle] *ERROR* radeon_cp_idle called without lock held,
> held 0 owner eb0908c0 eb0908c0
> [drm:radeon_cp_reset] *ERROR* radeon_cp_reset called without lock held,
> held 0 owner eb0908c0 eb0908c0
> [drm:radeon_cp_start] *ERROR* radeon_cp_start called without lock held,
Its a userspace problem not a kernel side one, granted we might throttle
the messages. Are you running KDE compositing on radeon or something
likethat?
Dave.
>
> Backtrace:
> 0: X(xf86SigHandler+0x80) [0x80d29f0]
> 1: [0xb7f54400]
> 2: /usr/lib/xorg/modules/input//evdev_drv.so [0xb7aab90e]
> 3: X [0x80d2b87]
> 4: X [0x80b6065]
> 5: [0xb7f54400]
> 6: X(Dispatch+0x81) [0x808d5d1]
> 7: X(main+0x485) [0x8074bb5]
> 8: /lib/libc.so.6(__libc_start_main+0xdc) [0xb7c813ec]
> 9: X(FontFileCompleteXLFD+0x229) [0x8073f21]
>
> Fatal server error:
> Caught signal 11. Server aborting
Looks like X is dying in its input driver, which is causing the video
driver to do something without the lock, so its an X.org configuration or
build issues.. best ask on [email protected]
Dave.
søn, 01 06 2008 kl. 23:11 +0100, skrev Dave Airlie:
> >
> > Backtrace:
> > 0: X(xf86SigHandler+0x80) [0x80d29f0]
> > 1: [0xb7f54400]
> > 2: /usr/lib/xorg/modules/input//evdev_drv.so [0xb7aab90e]
> > 3: X [0x80d2b87]
> > 4: X [0x80b6065]
> > 5: [0xb7f54400]
> > 6: X(Dispatch+0x81) [0x808d5d1]
> > 7: X(main+0x485) [0x8074bb5]
> > 8: /lib/libc.so.6(__libc_start_main+0xdc) [0xb7c813ec]
> > 9: X(FontFileCompleteXLFD+0x229) [0x8073f21]
> >
> > Fatal server error:
> > Caught signal 11. Server aborting
>
> Looks like X is dying in its input driver, which is causing the video
> driver to do something without the lock, so its an X.org configuration or
> build issues.. best ask on [email protected]
You are right, this was simply freedesktop.org bug #13991 [1] which has
been solved since January [2], but not made it into a release of
xf86-input-evdev yet.
So I still want to do something about the throttling issue of the kernel
messages, but that's gonna wait as I'm off to bed now.
Thanks Dave.
Simon Holm Thøgersen
[1] http://bugs.freedesktop.org/show_bug.cgi?id=13991
[2]
http://gitweb.freedesktop.org/?p=xorg/driver/xf86-input-evdev.git;a=commit;h=5adbc6b98ca1da9a426528ce0df64a992feff3bd