Description of problem: Running Fedora 12 + updates If a default xorg.conf (created by Xorg -configure) is used, with "Option ForceLowPowerMode 1" the video display hangs on startup system still responds via ssh,but cannot be used locally (including VT switching) This is occuring with the 2.6.32 kernel & with KMS disabled, in order to try to gain better battery efficiency until power management is added to drm in a future kernel version Version-Release number of selected component (if applicable): kernel-2.6.32.8-58.fc12.x86_64 xorg-x11-drv-ati-6.13.0-0.21.20100219gite68d3a389.fc12.x86_64 How reproducible: Occurs every time Steps to Reproduce: 1.create default xorg.conf with X -configure 2. copy new config to /etc/X11/xorg.conf 3. reboot/check display works normally 4. edit /etc/x11/xorg.conf & change "ForceLowPowerMode" to 1 5. restart 6. note the video hang Actual results: Video display becomes unresponsive - errors are logged via kernel drm mostly of the form [drm] wait idle failed status : 0xA0003030 0x00000003 Expected results: Video display operates normally power saving is achieved over standard settings Additional info: Power saving is important on a laptop. Whilst the W500 also has intel gfx, this won't drive DVI. Since I dock/undock from the laptop during the day/week, and dynamic display switching is not *YET* supported, I need to keep in ATI discrete mode. However since some of that time is on battery, achieving power savings is also important. Only current option for power saving appears to be - Disable KMS - Make use of existing driver clock speed reduction (just desktop compiz) Longer term better option is - Make use of new KMS power management (too early? in rawhide/f13?) - Make use of ability to dynamically switch screens between Intel and ATI (? f13/rawhide?)
Created attachment 396041 [details] kernel log messages
Created attachment 396042 [details] X11 config file
Created attachment 396043 [details] xorg log file
Created attachment 396044 [details] output from tailing /var/log/messages
*** This bug has been marked as a duplicate of bug 504427 ***
Not entirely clear that this defect matches the dup - seemingly different symptoms - originator indicates bug is fixed in recent rpms (mine isn't) - problem there appeared to relate to sideport memory, this is about power optimizations ?