Created attachment 38883 [details] Xorg log from DisplayPort usage I'm on Ubuntu Hardy unsing the latest ati git build. XServer: 1.4.1~git Mesa: 7.0.3 DRI: libdrm 2.3.0 / DRI1 Kernel: 2.6.32 (Ubuntu Lucid backport without KMS) The VGA-0-Port is working. Attaching the monitor via a DVI <-> DisplayPort adapter seems to work fine too, except that the monitor doesn't get a signal.
Created attachment 38884 [details] Xorg log using the VGA port
Created attachment 38896 [details] Xorg.conf
$ DISPLAY=:0 xrandr Screen 0: minimum 320 x 200, current 1280 x 1024, maximum 1280 x 1280 DisplayPort-0 connected 1280x1024+0+0 (normal left inverted right x axis y axis) 338mm x 270mm 1280x1024 60.0*+ 75.0 59.9 1280x960 59.9 1152x864 75.0 74.8 1024x768 75.1 70.1 60.0 832x624 74.6 800x600 72.2 75.0 60.3 56.2 640x480 75.0 72.8 66.7 60.0 720x400 70.1 VGA-0 disconnected (normal left inverted right x axis y axis)
I just patched the source to build the current git version without DRM. Still the same result: the driver detects the monitor correctly, the vlc connection is working, but there is not monitor signal.
Created attachment 39641 [details] [review] Config DisplayPort adapters I finally found and (kind of) fixed the bug. atombios_maybe_hdmi_mode defaults to ATOM_ENCODER_MODE_HDMI, if EDID_COMPLETE_RAWDATA is not defined (as it's not in my old Xorg servers "edid.h"). The patch introduces the new option "DPFallbackAdapter" and changes the default to DVI adapter. EDID_COMPLETE_RAWDATA was introduced with Xorg server 1.6. I have a few more patches to compile git against Xorg server 1.4 (no exaGetPixmapDriverPrivate, disable-dri build, disabled EXA build), so I couldn't compile test this single patch on git HEAD alone.
Better to just use DVI if we don't know and avoid all the additional mess. Although DVI vs. HDMI shouldn't really matter as far as the monitor is concerned as they are both TMDS. Does this patch work? --- a/src/atombios_output.c +++ b/src/atombios_output.c @@ -410,7 +410,7 @@ atombios_maybe_hdmi_mode(xf86OutputPtr output) { #ifndef EDID_COMPLETE_RAWDATA /* there's no getting this right unless we have complete EDID */ - return ATOM_ENCODER_MODE_HDMI; + return ATOM_ENCODER_MODE_DVI; #else if (output && xf86MonitorIsHDMI(output->MonInfo)) return ATOM_ENCODER_MODE_HDMI;
It works for me, but I don't know, if it'll work for HDMI adapters (and I can't check). That's why I introduced the new option.
Should be fine. Thanks for tracking this down. Pushed: 1e3ede62c39307d0210cb2044badd619a4f44fec
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.