http://bugzilla.opensuse.org/show_bug.cgi?id=914184 Ulrich Windl changed: What |Removed |Added ---------------------------------------------------------------------------- Flags|needinfo?(Ulrich.Windl@rz.u | |ni-regensburg.de) | --- Comment #2 from Ulrich Windl --- I booted the kernel as requested, but DRM info is very little. Also I did not see the message, and I had an idea: I did something else until the screen blanked (but screen's powersave mode was not active (visible in the color of its LED)). When I logged in again, the EDID message appeared once. So this should give a good starting point, I guess. Here are all kernel messages that match "drm": [ 0.000000] Command line: BOOT_IMAGE=/boot/vmlinuz-3.16.7-7-desktop root=UUID= resume=/dev/system/swap splash=silent quiet showopts drm_debug=0x0e [ 0.000000] Kernel command line: BOOT_IMAGE=/boot/vmlinuz-3.16.7-7-desktop root=UUID= resume=/dev/system/swap splash=silent quiet showopts drm_debug=0x0e [ 1.624994] [drm] Initialized drm 1.1.0 20060810 [ 1.630752] [drm] Memory usable by graphics device = 2048M [ 1.630754] [drm] Replacing VGA console driver [ 1.630757] fb: switching to inteldrmfb from VESA VGA [ 1.642200] [drm] Supports vblank timestamp caching Rev 2 (21.10.2013). [ 1.642202] [drm] Driver supports precise vblank timestamp query. [ 1.739960] fbcon: inteldrmfb (fb0) is primary device [ 1.746246] i915 0000:00:02.0: fb0: inteldrmfb frame buffer device [ 1.774330] [drm] Initialized i915 1.6.0 20080730 for 0000:00:02.0 on minor 0 [ 2.726631] [drm] Enabling RC6 states: RC6 on, RC6p off, RC6pp off [ 1631.797585] [drm:drm_edid_block_valid] *ERROR* EDID checksum is invalid, remainder is 189 -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=914184 --- Comment #4 from Ulrich Windl --- Created attachment 622679 --> http://bugzilla.opensuse.org/attachment.cgi?id=622679&action=edit hwinfo without debug messages -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=914184 http://bugzilla.opensuse.org/show_bug.cgi?id=914184#c7 --- Comment #7 from Ulrich Windl --- Sorry for not having replied in time; I was just very busy with other things. Today I had a look again. I saved the correct EDID from /sys/class/drm/card0-HDMI-A-1/edid, comparing it hexdump with the bad hexdums shown in the logs: It seems the transfer gets erros at some point, triggering a checksum error. For example: Good block: 00 ff ff ff ff ff ff 00 15 c3 78 20 01 01 01 01 0a 14 01 03 80 34 21 78 ea 1e c5 ae 4f 34 b1 26 0e 50 54 a1 08 00 a9 40 81 80 81 40 b3 00 01 01 01 01 01 01 01 01 28 3c 80 a0 70 b0 23 40 30 20 36 00 07 44 21 00 00 1a 00 00 00 ff 00 30 32 37 33 34 30 32 30 0a 20 20 20 20 00 00 00 fd 00 3b 3d 1f 4c 11 00 0a 20 20 20 20 20 20 00 00 00 fc 00 53 58 32 34 36 32 57 0a 20 20 20 20 20 00 2e Bad block: 00 ff ff ff ff ff ff 00 15 c3 78 20 01 01 01 01 0a 14 01 03 80 34 21 78 ea 1e c5 ae 4f 34 b1 26 0e 50 54 a1 08 00 a9 40 81 80 81 40 b3 00 01 01 01 01 01 01 01 01 28 3c 80 a0 70 b0 23 40 30 20 36 00 07 44 21 00 00 1a 00 00 00 ff 01 61 65 6f 67 69 61 65 61 15 41 41 41 41 01 01 01 fb 01 77 7b 3f 99 23 01 15 41 41 41 41 41 41 01 01 01 f9 01 a7 b1 65 69 6d 65 af 15 41 41 41 41 41 01 5d About frequency of the error: # journalctl |grep EDID.*vendor |wc -l 118 # journalctl |grep EDID.*checksum.*invalid |wc -l 24 So the error occurs frequently enough to believe there is a problem. -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=914184 http://bugzilla.opensuse.org/show_bug.cgi?id=914184#c9 Ulrich Windl changed: What |Removed |Added ---------------------------------------------------------------------------- Attachment #665326|0 |1 is obsolete| | --- Comment #9 from Ulrich Windl --- Created attachment 665327 --> http://bugzilla.opensuse.org/attachment.cgi?id=665327&action=edit Output of "journalctl -k -b", with suspend/resume (compressed) I was able to trigger an EDID checksum error by suspending, then waiting until the monitor goes into power save mode (yellow light), and then resuming. -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=914184 http://bugzilla.opensuse.org/show_bug.cgi?id=914184#c11 Egbert Eich changed: What |Removed |Added ---------------------------------------------------------------------------- Status|REOPENED |RESOLVED Resolution|--- |NORESPONSE --- Comment #11 from Egbert Eich --- (In reply to Ulrich Windl from comment #8)

Created attachment 665326 [details] Output of "journalctl -k -b |grep drm" (compressed)

I was not able to reproduce an EDID checksum error with kernel 3.16.7-29-desktop and xorg-x11-server-7.6_1.16.1-31.1.x86_64 within a short time; However I saw the error two days ago. Maybe the system has to go to a deeper sleep to make the problem appear...

This information doesn't help us much as we don't know the version numbers of the Xserver / kernel you did observe this problem. I'm also not sure if this isn't a problem on the HW side as we have not seen EDID issues with Intel for over 2 years. I'd suggest you use a kernel/Xserver/driver & software combination which seems to have a great chance to get this problem and instead exchange all HW components after the graphics card to see if the issue persists. -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=914184 http://bugzilla.opensuse.org/show_bug.cgi?id=914184#c13 --- Comment #13 from Ulrich Windl --- (In reply to Egbert Eich from comment #11)

This information doesn't help us much as we don't know the version numbers of the Xserver / kernel you did observe this problem.

I wasn't aware you knew that I updated the kernel and X server since my last report. Sorry.

I'm also not sure if this isn't a problem on the HW side as we have not seen EDID issues with Intel for over 2 years. I'd suggest you use a kernel/Xserver/driver & software combination which seems to have a great chance to get this problem and instead exchange all HW components after the graphics card to see if the issue persists.

So what software do you recommend, and where do I get the hardware to replace? ;-) Maybe re-read what I've written and what you have written. -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=914184 Egbert Eich changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |xorg-maintainer-bugs@forge. | |provo.novell.com -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=914184 http://bugzilla.opensuse.org/show_bug.cgi?id=914184#c15 --- Comment #15 from Ulrich Windl --- Created attachment 720321 --> http://bugzilla.opensuse.org/attachment.cgi?id=720321&action=edit EDID errors from syslog (compressed) After upgrading to openSUSE Leap 42.2, I still see the problem: My guess is like this: If you leave the computer, the screen will lock and the monitor will enter power save mode (LED changes from green to orange). When I hit a key and log in, the monitor is awakened again (LED changes to green), and maybe the driver tries to receiving the EDID when the monitor (EIZO SX2462W, vendor "ENC", prod id 8312) is not fully ready. That's what these "FF"s suggest to me. At boot time the EDID is always read without an error. Maybe if you look at the binary representation of the patterns you get an idea. Especially the "00 ff ff ff ff ff ff 00" pattern at the beginning of a bad EDID seems very common. -- You are receiving this mail because: You are on the CC list for the bug.