* Direct rendering causing lock ups?
@ 2002-10-16 12:27 Paul Furness
2002-10-16 18:45 ` Glynn Clements
0 siblings, 1 reply; 4+ messages in thread
From: Paul Furness @ 2002-10-16 12:27 UTC (permalink / raw)
To: linux-admin
Hi.
I just (for various reasons) re-built my workstation from scratch, using
Red Hat 7.3 and then applying updates.
After getting it all working nicely, I noticed that when I switch from X
to a text terminal (ctrl-alt-F1) and then back to X, the whole box
hangs. It won't allow telnet or ssh and I can't kill the X server
(ctrl-alt-BS) so the only fix is the 'Big Red Switch' solution.
After some messing around, and chatting to a friend who also found this
problem, we have come to the conclusion that it is related to DRI mode
Radeon configuration; if we turn on DRI, X is much faster but crashes
the box when switching VT. If we turn off DRI, no crash.
My questions:
1) Has anyone else seen this behavior? Is it restricted just to ATI
Radeon cards?
2) Does anyone know how to _stop_ this behavior? I'd really like to keep
DRI because of the speed, but I'd also really like to have a system that
lets me do what it's supposed to!
Any thoughts, anyone?
Paul.
--
Paul Furness
Systems Manager
2+2=5 for extremely large values of 2.
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: Direct rendering causing lock ups?
2002-10-16 12:27 Direct rendering causing lock ups? Paul Furness
@ 2002-10-16 18:45 ` Glynn Clements
2002-10-17 0:22 ` Benoit Joseph
0 siblings, 1 reply; 4+ messages in thread
From: Glynn Clements @ 2002-10-16 18:45 UTC (permalink / raw)
To: Paul Furness; +Cc: linux-admin
Paul Furness wrote:
> I just (for various reasons) re-built my workstation from scratch, using
> Red Hat 7.3 and then applying updates.
>
> After getting it all working nicely, I noticed that when I switch from X
> to a text terminal (ctrl-alt-F1) and then back to X, the whole box
> hangs. It won't allow telnet or ssh and I can't kill the X server
> (ctrl-alt-BS) so the only fix is the 'Big Red Switch' solution.
>
> After some messing around, and chatting to a friend who also found this
> problem, we have come to the conclusion that it is related to DRI mode
> Radeon configuration; if we turn on DRI, X is much faster but crashes
> the box when switching VT. If we turn off DRI, no crash.
>
> My questions:
>
> 1) Has anyone else seen this behavior? Is it restricted just to ATI
> Radeon cards?
I haven't seen that exact behaviour. However, my experience (with a
Matrox G400) is that DRI problems tend to be fatal. If an OpenGL
program crashes when using DRI, any recovery is normally limited to
Alt-SysRq-K to kill the X server (requires "Magic SysRq" support in
the kernel), followed by Ctrl-Alt-Delete for a clean shutdown. Trying
to do anything else normally results in a hard lockup that even
Alt-SysRq-* can't touch.
> 2) Does anyone know how to _stop_ this behavior? I'd really like to keep
> DRI because of the speed, but I'd also really like to have a system that
> lets me do what it's supposed to!
I'm pretty sure that you have to choose between DRI or reliability;
they seem to be mutually exclusive.
Personally, I set LIBGL_ALWAY_INDIRECT=1 in /etc/profile, so OpenGL
uses indirect (i.e. software) rendering by default, and explicitly
unset it if I need the perfomance (but I'll save files, close programs
and run "sync" beforehand).
--
Glynn Clements <glynn.clements@virgin.net>
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: Direct rendering causing lock ups?
2002-10-16 18:45 ` Glynn Clements
@ 2002-10-17 0:22 ` Benoit Joseph
0 siblings, 0 replies; 4+ messages in thread
From: Benoit Joseph @ 2002-10-17 0:22 UTC (permalink / raw)
To: linux-admin
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Hello,
I'm using an ATI radeon QD with DRI inabled and openGL without any
problem.
Benoit
Le Mercredi 16 Octobre 2002 20:45, Glynn Clements a écrit :
> Paul Furness wrote:
> > I just (for various reasons) re-built my workstation from scratch,
> > using Red Hat 7.3 and then applying updates.
> >
> > After getting it all working nicely, I noticed that when I switch
> > from X to a text terminal (ctrl-alt-F1) and then back to X, the
> > whole box hangs. It won't allow telnet or ssh and I can't kill the
> > X server (ctrl-alt-BS) so the only fix is the 'Big Red Switch'
> > solution.
> >
> > After some messing around, and chatting to a friend who also found
> > this problem, we have come to the conclusion that it is related to
> > DRI mode Radeon configuration; if we turn on DRI, X is much faster
> > but crashes the box when switching VT. If we turn off DRI, no
> > crash.
> >
> > My questions:
> >
> > 1) Has anyone else seen this behavior? Is it restricted just to ATI
> > Radeon cards?
>
> I haven't seen that exact behaviour. However, my experience (with a
> Matrox G400) is that DRI problems tend to be fatal. If an OpenGL
> program crashes when using DRI, any recovery is normally limited to
> Alt-SysRq-K to kill the X server (requires "Magic SysRq" support in
> the kernel), followed by Ctrl-Alt-Delete for a clean shutdown. Trying
> to do anything else normally results in a hard lockup that even
> Alt-SysRq-* can't touch.
>
> > 2) Does anyone know how to _stop_ this behavior? I'd really like to
> > keep DRI because of the speed, but I'd also really like to have a
> > system that lets me do what it's supposed to!
>
> I'm pretty sure that you have to choose between DRI or reliability;
> they seem to be mutually exclusive.
>
> Personally, I set LIBGL_ALWAY_INDIRECT=1 in /etc/profile, so OpenGL
> uses indirect (i.e. software) rendering by default, and explicitly
> unset it if I need the perfomance (but I'll save files, close
> programs and run "sync" beforehand).
- --
JOSEPH Benoit
benoit.joseph@teledisnet.be
joker@baby-linux.net
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (GNU/Linux)
Comment: For info see http://www.gnupg.org
iD8DBQE9rgK7LPnuiaZn1q4RAnyyAKDRG1y1xU+3mGsm8RZIAUByq7i2SgCfY0Xt
EhRPDqYf0XS4PJlaVcoGlTc=
=pEMD
-----END PGP SIGNATURE-----
-
To unsubscribe from this list: send the line "unsubscribe linux-admin" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: Direct rendering causing lock ups?
@ 2002-10-17 16:50 gerardo juarez-mondragon
0 siblings, 0 replies; 4+ messages in thread
From: gerardo juarez-mondragon @ 2002-10-17 16:50 UTC (permalink / raw)
To: linux-admin
Tridents do about the same thing in a perfectly
reproducible way. I have had to play with the bus
clock divider in the BIOS to make it work, but
the documentation of the X server about this
Trident problem suggests using the card's manual
to adjust the jumpers, I guess to slow down in
some way (wait states, caching/no-caching) access
to the card. I'm afraid is a question of speed
vs. reliability. take your pick.
Gerardo Juarez
---- Begin Original Message ----
From: Paul Furness <paul.furness@vil.ite.mee.com>
Sent: 16 Oct 2002 13:27:46 +0100
To: linux-admin@vger.kernel.org
Subject: Direct rendering causing lock ups?
Hi.
I just (for various reasons) re-built my
workstation from scratch, using
Red Hat 7.3 and then applying updates.
After getting it all working nicely, I noticed
that when I switch from X
to a text terminal (ctrl-alt-F1) and then back to
X, the whole box
hangs. It won't allow telnet or ssh and I can't
kill the X server
(ctrl-alt-BS) so the only fix is the 'Big Red
Switch' solution.
After some messing around, and chatting to a
friend who also found this
problem, we have come to the conclusion that it
is related to DRI mode
Radeon configuration; if we turn on DRI, X is
much faster but crashes
the box when switching VT. If we turn off DRI, no
crash.
My questions:
1) Has anyone else seen this behavior? Is it
restricted just to ATI
Radeon cards?
2) Does anyone know how to _stop_ this behavior?
I'd really like to keep
DRI because of the speed, but I'd also really
like to have a system that
lets me do what it's supposed to!
Any thoughts, anyone?
Paul.
--
Paul Furness
Systems Manager
2+2=5 for extremely large values of 2.
-
To unsubscribe from this list: send the line
"unsubscribe linux-admin" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at
http://vger.kernel.org/majordomo-info.html
---- End Original Message ----
Searching for the best free email? Try MetaCrawler Mail, from the #1 metasearch service on the Web, http://www.metacrawler.com
^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2002-10-17 16:50 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2002-10-16 12:27 Direct rendering causing lock ups? Paul Furness
2002-10-16 18:45 ` Glynn Clements
2002-10-17 0:22 ` Benoit Joseph
-- strict thread matches above, loose matches on Subject: below --
2002-10-17 16:50 gerardo juarez-mondragon
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox