Comment 50 for bug 712075

This Bug affects me. I'm using a Samsung SF510 ("shark design") notebook with i915 kernel module and am connecting an Acer AL2216W 1680x1050 flat-screen Monitor via HDMI, using a hdmi-to-dvi-d cable, randr output: "HDMI1".

Today this happened for the second or third time. I think it is triggered by the activation of the screensaver. The laptop lid was closed and randr output "LVDS1" manually turned off, and "HDMI1" in default mode whenever the problem occured. Today was especially annoying because it took me quiet a long time to make the external monitor useable again. even after powering the monitor and laptop off, the problem (edid checksum error in dmesg output, monitor "disconnected" in xrandr output) persisted. I then powered off both devices and disconnected them from the power supply for ~1 minute. This finally helped.

I now extracted both edid bit-strings, the one reported by the kernel to have a bad checksum, as well as the one reported by "xrandr --verbose" now, which is ok, and analyzed the differences. It turns out to be only one byte:

byte 119 (0x77) reads 0x90 instead of 0x34

i have no idea why !?

$ cat edid
00 ff ff ff ff ff ff 00 04 72 74 ad 9c 04 70 64
2f 10 01 03 e8 2f 1e 78 2e c5 85 a4 59 49 9a 24
12 50 54 bf ef 00 81 80 81 40 71 4f 95 00 95 0f
b3 00 81 c0 8b c0 21 39 90 30 62 1a 27 40 68 b0
36 00 d9 28 11 00 00 1c 00 00 00 fd 00 38 4c 1e
52 10 00 0a 20 20 20 20 20 20 00 00 00 fc 00 41
63 65 72 20 41 4c 32 32 31 36 57 0a 00 00 00 ff
00 4c 37 34 30 39 30 34 36 33 38 32 30 20 00 f7

$ diff edid bad-edid
8c8
< 00 4c 37 34 30 39 30 34 36 33 38 32 30 20 00 f7
---
> 00 4c 37 34 30 39 30 90 36 33 38 32 30 20 00 f7