Thanks for Kevin merging this large patch I built. Since there are lots of changes involved, some issues will surface with more testing. In fact, I found a few bugs (like in DDC detection and DFP handling) in this patch after the patch was submitted. I haven't got time to make another patch yet. I'm thinking, after putting together the 2D support for all the new cards, I'll submit another patch. For now, I'd like to hear more feedback. Below I'll try to explain some issues raised.
> On Thu, 2002-07-11 at 22:27, Kevin E Martin wrote: > > I've just checked in a large patch from ATI. Here's the CHANGELOG > > entry: > > > > 199. ATI patch to: > > - Fix Dell OEM VE card support > > - Add better clone mode support > > - Fix large panel (>= 1600x1200) detection and initialization problems > > - Remove "PanelSize" and "CrtScreen" options since they are no longer > > needed with new CloneMode and improved flat panel support > > At least Option "PanelSize" would still be useful for systems where we > can't determine the panel parameters yet. Current Radeon driver supports two kinds of panel: TMDS panel (usually used for desktop, named as DFP in the driver for some legacy reason) and LVDS panel (usually used for laptop, named as LCD in the driver). The TMDS panels should be DDC capable, if not, the BIOS won't even be able to bring them up to text mode. For this kind of panel, the radeon driver will try to do DDC detection first. If it fails for some reason, the driver will used EDID data detected by BIOS. With the DDC info, the driver can figure out the panel size and bring it up automatically. The laptop (LVDS) panels usually are not DDC capable (some new ones are). The panel size and timing information are hard coded into video BIOS and radeon driver will use it to bring the panel up. The only potential problem here is that some custom version of BIOS doesn't follow the rule and put the panel info to a different place. I haven't come across such a laptop yet and would like to know if someone has one. Only knowing panel size (together with VESA timings) usually is not good enough to bring a LVDS panel up properly. Even with the correct modeline, it's better to know other things (like power delay) to safely turn the panel on/off. Let user to specify panel size can be error prone. Anyway, if there are people out there really requiring this option, I can put it back. As for non-native modes (resolution lower than panel size), by default, radeon driver will use the internal ratio matrix expansion (RMX) unit to scale the display down. It works not only for all standard VESA modes, it'll work also for any mode between 320x200 and panel size, for example 600x500. If the DDCMode option is enabled, the driver will use the modes listed in the EDID data, which will let the panel to scale down with its own scaler. > > - Add "DDCMode" option to detect and use DDC modes > > - Add "PanelOff" option to disable panel on laptops > > - Fix corrupted console problem > > - Other misc fixes > > (#A.1043, Hui Yu@ATI). > > > > I'd like get some feedback on how it works for others with both desktop > > and laptop Radeons. > > Doesn't quite work on a Titanium PowerBook III (without Option > "UseFBDev", with a 1280x854 modeline obtained by radeonfb via EDID which > works fine with that option). There are flickering lines on the right > side. Is the DDC detection OK? If so, can you try to use DDCMode option? I know something broken with DDC detection in this patch. If so, can you send me your log file? > Also, I think it would be nice if the radeon driver was better > documented, in a manpage and/or programs/Xserver/hw/xfree86/Options. I > don't know what the Clone options are for even after a quick look > through the code, how is an end user supposed to know? Clone mode is for 2nd head mirroring by using CRTC2. Quite a few people asked for this feature (a recent example http://www.xfree86.org/pipermail/xpert/2002-July/019082.html). When I wrote dual-head/panel support for X4.2, I didn't have time to add the 2nd head handling when only one screen is specified. It turns out some people plug in both of their monitors and only have one screen section in their config files. This results in various problems caused by the same CRTC driving two monitors with different capabilities. I agree we need to improve the documentation. Lots of questions appeared on this forum are because of lack of document. A couple of months ago, Achim Bohnet ([EMAIL PROTECTED]) offered to do this work. He helped me a lot in testing this patch. Together we have tested on 10+ different flat panels. And we found a few issues even after the patch was submitted. Not sure if he is still available now. Hui > > > I'll start working on the other patches that have been sent in shortly. > > Great, thanks. > > > -- > Earthling Michel D�nzer (MrCooper)/ Debian GNU/Linux (powerpc) developer > XFree86 and DRI project member / CS student, Free Software enthusiast > > _______________________________________________ > Xpert mailing list > [EMAIL PROTECTED] > http://XFree86.Org/mailman/listinfo/xpert > _______________________________________________ Xpert mailing list [EMAIL PROTECTED] http://XFree86.Org/mailman/listinfo/xpert
