|
Description
|
When detecting a monitor as DFP/LCD, radeon driver uses its own validate
process which eventally calls RADEONSortModes() to sort the EDID modes it
probed. But RADEONSortModes() has a minor problem such that new mode A is
sorted below an existing mode B only when both A's HDisplay and VDisplay
are smaller than B's. In other words, if A's VDisplay is smaller than B's,
but HDisplay for them remain the same, then A is sorted above B, which is
obviously wrong.
This happens when I used a DELL 24"1 monitor, where mode "1920x1200" in
detailed timing sector is first added to the mode list, then a mode
"1600x1200" from standard timing sector is added. It ended up with mode
"1600x1200" being on top of "1920x1200" and therefore selected.
|