The regression is exactly like Josh Boyer described it (https://lkml.org/lkml/2013/5/3/467), but on my system (dell xps 13 ivy bridge 1080p version) it only occurs after the reverted revert (657445fe8660100ad174600ebfa61536392b7624) [1] that fixed the problem on the macbook. I had to revert the reverted revert to regain an usable display. :) [1] http://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/commit/?id=657445fe8660100ad174600ebfa61536392b7624
Can you please boot with drm.debug=0xe added to your kernel cmdline and then attach the complete dmesg for - working 3.9.x kernel - broken 3.10-rc kernel - working 3.10-rc kernel with the revert reverted Let's see which kernel comes up with which exact eDP link config.
Created attachment 104881 [details] dmesg output 3.9.0 - good
Created attachment 104891 [details] dmesg output 3.10-rc6 - bad
Created attachment 104901 [details] dmesg output 3.10-rc6 reverted commit - good
the 3.10-rc6 dmesg looks okay. The right modes get selected and it looks pretty similar to 3.10-rc-6-reverted, but without the revert, the screen is definetly messed up. It looks like if you selected a too high refreshrate on an old crt.
@Daniel: do you need any other logs or info?
*** This bug has been marked as a duplicate of bug 59841 ***