| View previous topic :: View next topic |
| Author |
Message |
jpalko
n00b
Joined: 12 Mar 2003
Posts: 65
Location: Piispanristi, Finland
|
 Posted: Mon Apr 29, 2013 10:18 am Post subject: [SOLVED] RadeonHD7770 + KMS and 3 monitors crash on resume |
 |
|
Anyone using this sort of combination that I use:
1. a Radeon HD 7770 2GB card
| Code: | lspci | grep VGA
02:00.0 VGA compatible controller: Advanced Micro Devices [AMD] nee ATI Cape Verde XT [Radeon HD 7770 GHz Edition] |
2. 3 monitors
| Code: | [ 59.860] (II) RADEON(0): Manufacturer: SAM Model: 4d5 Serial#: 1263088180
[ 59.860] (II) RADEON(0): Manufacturer's mask: 0
[ 59.860] (II) RADEON(0): Monitor name: SyncMaster
[ 59.923] (II) RADEON(0): Manufacturer: SAM Model: 68b Serial#: 1110590276
[ 59.923] (II) RADEON(0): Manufacturer's mask: 0
[ 59.923] (II) RADEON(0): Monitor name: SMBX2440
[ 59.986] (II) RADEON(0): Manufacturer: SAM Model: 68b Serial#: 1110590276
[ 59.986] (II) RADEON(0): Manufacturer's mask: 0
[ 59.986] (II) RADEON(0): Monitor name: SMBX2440 |
that are configured with resolutions
| Code: | $ xrandr | grep connect
DisplayPort-0 connected 1920x1080+3840+0 (normal left inverted right x axis y axis) 160mm x 90mm
HDMI-0 connected 1920x1080+1920+0 (normal left inverted right x axis y axis) 531mm x 299mm
DVI-0 connected 1920x1080+0+0 (normal left inverted right x axis y axis) 531mm x 299mm
DVI-1 disconnected (normal left inverted right x axis y axis) |
The problem is that when I let the powersaving happen on these, DisplayPort and HDMI wake back up but not the DVI which is followed after a moment with a crash of the whole system. If it matters, the problem display is of the model Samsung SMBX2440.
Any ideas how to debug this problem? The system remains stable until I try to wake up my displays. If I do not touch the machine at all, it is perfectly remotely accessible but as soon as that one monitor should wake up, everything crashes. No log output, no nothing, just the whole system stuck in a few moments after being able to move the mouse. As a further detail, the way I configured my X is that I have no full xorg.conf placed in /etc/X11 but I command the displays on login to be located thus:
| Code: | | /usr/bin/xrandr --output DVI-0 --left-of HDMI-0 --output DisplayPort-0 --right-of HDMI-0 |
I have been trying to get this working with hints from radeon - Gentoo Wiki and radeon - Gentoo Linux Wiki and several other places but I keep bumping into the problem that if the DVI attached monitor goes to powersaving, my whole machine crashes. Any ideas would be greatly appreciated.
_________________
Jukka Palko
Piispanristi, Finland
Linux Counter ID #219161
Last edited by jpalko on Thu May 02, 2013 5:59 am; edited 2 times in total |
|
| Back to top |
|
 |
chithanh
Developer
Joined: 05 Aug 2006
Posts: 2158
Location: Berlin, Germany
|
| Posted: Tue Apr 30, 2013 9:31 am Post subject: |
|
|
How hard is the system crash? Does the computer still respond to ssh, ping, magic SysRQ? Is kernel 3.9 affected?
As you have a few moments beforel the system freezes, you may be able to see if something is printed to dmesg or Xorg.0.log. If necessary, increase dmesg verbosity by booting with drm.debug=14 kernel parameter. |
|
| Back to top |
|
|
jpalko
n00b
Joined: 12 Mar 2003
Posts: 65
Location: Piispanristi, Finland
|
| Posted: Wed May 01, 2013 5:53 pm Post subject: |
|
|
The crash kills the system quite completely. I've had an ssh session open with tail running on /var/log/Xorg.0.log /var/log/messages /var/log/lightdm/lightdm.log /var/log/lightdm/x-0* and nothing gets printed out to any of them when I wake the screens and that session goes completely dead. The system also stop responding to pings at the same time. I'm running only 3.7.10-gentoo-r1 at the moment. Your recommendation would be to go with 3.9 first and then if the problem still exists then activate debugging?
Magic SysRQ I haven't tried yet as I have to say that my experience is pretty low with it but I'll gladly learn more and it looks like kernel source Documentation/sysrq.txt is pretty decent document on it. I'll try if anything there works after the crash.
I'll try stuff tomorrow at work with the machine as we had a Finnish holiday today.
_________________
Jukka Palko
Piispanristi, Finland
Linux Counter ID #219161 |
|
| Back to top |
|
|
jpalko
n00b
Joined: 12 Mar 2003
Posts: 65
Location: Piispanristi, Finland
|
| Posted: Thu May 02, 2013 5:57 am Post subject: |
|
|
Well, I upgraded the kernel to 3.9.0-gentoo and no problem any more thus far as the third display wakes up perfectly also from blanking. Thanks for mentioning the 3.9 kernel. I guess I should have figured out that with this new hardware that I'd have to go a bit further then the kernel that is currently marked stable. 
_________________
Jukka Palko
Piispanristi, Finland
Linux Counter ID #219161 |
|
| Back to top |
|
|
jpalko
n00b
Joined: 12 Mar 2003
Posts: 65
Location: Piispanristi, Finland
|
| Posted: Fri May 03, 2013 6:42 am Post subject: |
|
|
One interesting thing still popped up with display power management. The displays I have are attached via a DVI cable, a HDMI-DVI cable and a DisplayPort-HDMI cable
The DisplayPort-HDMI connected monitor didn't wake up from xscreensaver locking this morning which had the power management options enabled but X didn't crash either. It had just dropped the output to utilise only the DVI and HDMI-DVI connected monitors. I had yesterday set my DPMS settings to:
| Code: | $ xset q | tail -4
DPMS (Energy Star):
Standby: 600 Suspend: 1200 Off: 2400
DPMS is Enabled
Monitor is On |
I tried this morning with the following setting and didn't get the issue recreated.
| Code: | $ xset q | tail -4
DPMS (Energy Star):
Standby: 60 Suspend: 120 Off: 0
DPMS is Enabled
Monitor is On |
_________________
Jukka Palko
Piispanristi, Finland
Linux Counter ID #219161 |
|
| Back to top |
|
|
|
Desktop Environments
