[LIVE] Partial support for old ATI Cards not supported by FGLRX - Printable Version
+- XBMC Community Forum (http://forum.xbmc.org)
+-- Forum: Development (/forumdisplay.php?fid=32)
+--- Forum: Development (/forumdisplay.php?fid=93)
+--- Thread: [LIVE] Partial support for old ATI Cards not supported by FGLRX (/showthread.php?tid=81008)
[LIVE] Partial support for old ATI Cards not supported by FGLRX - anonym1te - 2010-09-13 10:52
Presently the upstart job for XBMC-live utilises the numeric output of lspci -nn to perform Vendor ID checking. This check is then utilised to configure one of the 2 presently supported proprietary drivers, NVidia or AMD (fglrx), if there is no xorg.conf extant. If the vendor ID does not match either 10de (NVidia) or 1002 (ATI/AMD), no configuration of the proprietary drivers are performed and X is left to load its in-built default config (providing a config does not already exist)
If the Vendor ID matches 1002, configuration for the fglrx proprietary driver is performed. However, a lot of older ATI cards have the same Vendor ID but are no longer supported by the fglrx driver. This currently leads to silent failure of the boot process with no explanation of the issue that caused it.
Many of the fglrx unsupported cards have hardware assisted direct rendering support in the ati driver, which is the default driver loaded by X when it detects an ATI card. By checking to see if the ATI/AMD card is supported by fglrx prior to attempting to configure it as an fglrx card, and if not allow X to attempt to configure, it would greatly increase the chances of the user seeing the live-cd boot in the instance that they owned one of these older graphics cards.
A simple way to check for a card not supported by the fglrx driver is to run aticonfig and check for a non zero return code.
I have made a slight modification to the /etc/init/xbmc-live.conf that performs this check. However, before submitting a patch I wished to bring this up for discussion with the devs here in case there is a reason for not providing this support.
Please let me have your thoughts and indicate whether I should submit a patch for this.
- spiff - 2010-09-13 11:45
i'm 97% certain there's nothing more to this than an oversight.
Patch submitted - anonym1te - 2010-09-19 05:47
Ticket 10236 has been raised with the patch attached for your review and inclusion.
As per your patch instructions I have created the patch against trunk head (33961), however it would be great if this can make it into Dharma.