Hello. Linux 2.6.17 (Vanilla Sources), xorg-x11-7.0-r1, Radeon Mobility M9000 (radeon driver, not fglrx) on a Dell Inspiron 600m laptop (bios revision A17). I have a dual-monitor configuration that I sometimes enable. That is, laptop LCD and an external LCD. I also have the machine go into S3 suspend sometimes. When the external monitor is not connected, the machine suspends and resumes fine. When the external monitor is connected and the system resumes, the VTs do not work- they are just blank screens, or sometimes a single white flashing line across the screen. VT 7 (graphical, X) works in all cases. I have noticed that what determines the behavior is the presence of the second monitor at the time of resume. It does not matter what xrandr was set to, just if the VGA cable is plugged into the laptop. If I resume with monitor plugged in, VTs go bad, and without monitor connected, they are fine. If I resume with monitor attached, then unplug the monitor and do another suspend/resume cycle before reattaching it, the VTs are fine again. But it is annoying to reach back there and unplug it every time. I am aware of a bios quirk with dells (mine included) where the bios does not POST the video card during S3 resume, so I believe X is what does this task. (Without X running, the machine resumes into a working state but without anything on the screen). Does anyone have a patch or simple fix? I believe the problem is in the open-source radeon driver. This isn't very critical but it would be nice if it were fixed at some point.
Please confirm this with X.Org 7.1 -- add xorg-server, your drivers, mesa and mesa-progs to /etc/portage/package.keywords.
Please re-open the bug once you've tried 7.1.
Installed xorg-x11-7.1. Same exact behavior. VTs blank when resuming with external monitor connected, come back to life if unplug external monitor and run another suspend/resume cycle.
Alright, please find or create a bug upstream at https://bugs.freedesktop.org and post the link here. Thanks!
Upstream Bug: https://bugs.freedesktop.org/show_bug.cgi?id=8425
Much appreciated.