Created attachment 43526 [details] [review] dmesg Radeon Driver reports EDID errors every 10 seconds, which causes related log files increased over 100M in size every day. OS: Ubuntu Maverick 10.10 Desktop Kernel_version: Ubuntu 2.6.35-25.44-generic 2.6.35.10 Mother Board: ECS A740GM-M Chipset: AMD 740G, integrated ATI Radeon 2100
Created attachment 43527 [details] [review] Xorg.0.log Xorg.0.log provided.
Created attachment 43529 [details] [review] dump lspci
Created attachment 43530 [details] vbios
a little more description about the problem. 1. Existing for long The problem has been reported from April 2010 or before, on all Linux platform. Appears the Radeon driver has repeated EDID error and RAW dump reported into syslog and related log files, causes log file rapidly increasing in size of 100M per day. 2. Specific to my HW. The mother board ECS A740GM-M version 7.0 uses chipset AMD 740G, which integrated ATI Radeon 2100. The chip support VGA and DVI ports, but to this version of mother board, the manufacture only has VGA connector installed. There no DVI port at all on this mother board. The Radeon driver repeatedly scans the video connectors, reports a connection detected on DVI-1, but with invalid EDID checksum. 3. Severity of the problem Since the great amount of dump to log happens in every 10 seconds, the system suffering in obvious performance degrade. In some slower system, report says the system appears frozen about 5 seconds in every 10 seconds. In non-GUI configuration, the repeated EDID error message dump into console for every 10 seconds, makes the system hardly management. 4. Whos fault The board has no DVI connector at all, but the Radeon driver repeatedly find a connection on DVI-1 port. It possibly because the manufacture does not handle the unconnected DVI signals properly, which causes the driver keep claim finding a monitor connected into DVI-1 port. However, these repeated message were useless to ordinary end user, instead of annoying and lower system performance. I will like rather see such error message in log only once for every system reboot, or provide a debug switch in xorg.conf to allow the error to be repeatedly reported (in case a debug engineer want to on-site). Other thought is, the driver could better to provide option in xorg.conf to disable detect on specified video port, so the driver will NOT to scan and detect those ports over and over again.
Is this still an issue with a newer kernel?
The problem is known in Ubuntu: see https://bugs.launchpad.net/ubuntu/+source/linux/+bug/810926. Either upgrade your system to Ubuntu Oneiric or install the fixed kernel (with backported kernel patches) from Launchpad PPA for Ubuntu Maverick. Both solutions will fix the problem, as it did for the others having this special buggy chipset. btw the Ubuntu bug has been closed with status Fix released.
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.