Bug 27490 - [KMS] dvi monitor is faulty reported as disconnected due to invalid EDID (with Y-cable)
Summary: [KMS] dvi monitor is faulty reported as disconnected due to invalid EDID (wit...
Status: RESOLVED INVALID
Alias: None
Product: xorg
Classification: Unclassified
Component: Driver/intel (show other bugs)
Version: unspecified
Hardware: x86 (IA32) Linux (All)
: medium normal
Assignee: ykzhao
QA Contact: Xorg Project Team
URL:
Whiteboard:
Keywords: NEEDINFO
Depends on:
Blocks:
 
Reported: 2010-04-06 08:41 UTC by grenouille
Modified: 2010-06-19 10:11 UTC (History)
0 users

See Also:
i915 platform:
i915 features:


Attachments
dmesg output (52.62 KB, application/octet-stream)
2010-04-18 10:46 UTC, grenouille
no flags Details
vbios dump (64.00 KB, application/octet-stream)
2010-04-18 10:49 UTC, grenouille
no flags Details
Xorg.log with Option ModeDebug True (71.80 KB, application/octet-stream)
2010-05-08 01:37 UTC, grenouille
no flags Details
dmsg with different Monitor (50.12 KB, text/plain)
2010-05-15 13:51 UTC, grenouille
no flags Details
dmesg output after cat /sys/.../drm/card0/card0-DVI-D-1/status (123.60 KB, text/plain)
2010-05-15 13:55 UTC, grenouille
no flags Details

Description grenouille 2010-04-06 08:41:12 UTC
If I boot with kms enabled (kerneloption i915.modeset=1) the DVI connected monitor is shown as disconneced which is not true. (devices named from xrandr as VGA1 and DVI1).
Also it is not possible to enable the output via

xrandr --newmode "1280x1024_x"  135 1280 1296 1440 1688 1024 1025 1028 1066 +HSync +VSync
xrandr --addmode DVI1 1280x1024_x
xrandr --output DVI1 --mode 1280x1024_x

If I boot with kms disabled (kerneloption i915.modeset=0)i, both connected Displays are correct shown as connected (devices named from xrandr as VGA and TMDS-1)

-- chipset:  Intel Corporation Mobile GM965/GL960 Integrated Graphics Controller (rev 03)
-- system architecture: 32-bit
-- xf86-video-intel: xserver-xorg-video-intel 2:2.9.1-3
-- xserver: xserver-xorg 1:7.5+5
-- mesa: 7.7-4
-- libdrm: libdrm-intel1 2.4.18-3
-- kernel: 2.6.32 and 2.6.33
-- Linux distribution: Debian sid
-- Machine or mobo model:
-- Display connector: vga and dvi

I also found this:
http://www.mail-archive.com/debian-bugs-dist@lists.debian.org/msg763368.html

Thanks
Comment 1 Gordon Jin 2010-04-07 05:56:30 UTC
This looks like bug#24458. Can you try newer driver (xf86-video-intel) like 2.10 or 2.11?
Comment 2 ykzhao 2010-04-07 19:16:56 UTC
Will you please try the 2.6.33 kernel and attach the output of dmesg? Please add the boot option of "drm.debug=0x04".

Please also attach the output of vbios.dump on this box, which can be obtained by using the following command:
  >1. echo 1 > /sys/devices/pci0000:00/0000:00:02.0/rom
    2. cat /sys/devices/pci0000:00/0000:00:02.0/rom >vbios.dump
    3. echo 0 > /sys/devices/pci0000:00/0000:00:02.0/rom

Thanks.
Comment 3 grenouille 2010-04-08 01:08:14 UTC
Thanks both of you for reply.

> This looks like bug#24458. Can you try newer driver (xf86-video-intel) like
> 2.10 or 2.11?

Sorry to report, same result.

> Will you please try the 2.6.33 kernel and attach the output of dmesg? Please
> add the boot option of "drm.debug=0x04".
> Please also attach the output of vbios.dump

I will, but first: 

- I'm not able to get a bios or grub screen via my DVI-Display. Sorry I never 
  give them any attention while I have no problems with both displays running X
  without kms (use them over a year now). Also I should mention, both displays
  are connected via one y-cable to one single DVI-plug (AOpen i965GMt-LA). So I
  will try to exclude bios or hardware issues first. I will give feedback If
  I'm done. 

- Is it possible to force the display recognized as connected via the sys file
  system if kms is enabled?
  /sys/devices/pci0000\:00/0000\:00\:02.0/drm/card0/card0-DVI-D-1/*

Best
Comment 4 grenouille 2010-04-18 10:46:40 UTC
Created attachment 35143 [details]
dmesg output

> Will you please try the 2.6.33 kernel and attach the output of dmesg? Please
> add the boot option of "drm.debug=0x04".

output attached
Comment 5 grenouille 2010-04-18 10:49:10 UTC
Created attachment 35144 [details]
vbios dump

> Please also attach the output of vbios.dump on this box

dump attached
Comment 6 grenouille 2010-04-18 10:55:47 UTC
FYI
I'm able to use both displays when I disable kms (without touching the connectors)

The error described at end of post, may also be helpful.

Thanks a lot for interest
Best Wulf

Also while I do

cat: /sys/devices/pci0000:00/0000:00:02.0/drm/card0/card0-DVI-D-1/status
I got the following erros:


[  424.211865] [drm:intel_sdvo_debug_write], SDVOB: W: 0B                         (SDVO_CMD_GET_ATTACHED_DISPLAYS)
[  424.226154] [drm:intel_sdvo_debug_response], SDVOB: R: 01 00                   (Success)
[  424.226161] [drm:intel_sdvo_detect], SDVO response 1 0
[  424.226165] [drm:intel_sdvo_debug_write], SDVOB: W: 7A 02                      (SDVO_CMD_SET_CONTROL_BUS_SWITCH)
[  424.336677] [drm:edid_is_valid] *ERROR* EDID checksum is invalid, remainder is 237
[  424.338306] [drm:edid_is_valid] *ERROR* Raw EDID:
[  424.339922] <3>00 ff ff ff ff ff ff 00 26 cd ac 46 48 10 00 00  ........&..FH...
[  424.339925] <3>01 0c 01 03 00 22 1b 78 2a 2d 90 a4 57 4a 9c 25  .....".x*-..WJ.%
[  424.339927] <3>11 50 54 bf ef 00 81 80 71 4f 01 01 01 01 01 01  .PT.....qO......
[  424.339929] <3>01 01 01 01 01 01 bc 34 00 98 51 00 2a 40 10 90  .......4..Q.*@..
[  424.339931] <3>13 00 54 0e 11 00 00 1e 00 00 00 ff 00 30 35 31  ..T..........051
[  424.339933] <3>39 36 47 30 30 30 30 30 30 30 00 00 00 fd 00 37  96G0000000.....7
[  424.339935] <3>4b 18 53 0e 00 0a 20 20 20 20 20 20 00 00 00 fc  K.S...      ....
[  424.339937] <3>00 50 4c 45 34 33 31 0a 20 20 20 20 20 20 00 b0  .PLE431.      ..
[  424.339939] 
[  424.339941] [drm:intel_sdvo_debug_write], SDVOB: W: 7A 02                      (SDVO_CMD_SET_CONTROL_BUS_SWITCH)
[  424.450521] [drm:edid_is_valid] *ERROR* EDID checksum is invalid, remainder is 237
[  424.452158] [drm:edid_is_valid] *ERROR* Raw EDID:
[  424.453804] <3>00 ff ff ff ff ff ff 00 26 cd ac 46 48 10 00 00  ........&..FH...
[  424.453807] <3>01 0c 01 03 00 22 1b 78 2a 2d 90 a4 57 4a 9c 25  .....".x*-..WJ.%
[  424.453809] <3>11 50 54 bf ef 00 81 80 71 4f 01 01 01 01 01 01  .PT.....qO......
[  424.453811] <3>01 01 01 01 01 01 bc 34 00 98 51 00 2a 40 10 90  .......4..Q.*@..
[  424.453813] <3>13 00 54 0e 11 00 00 1e 00 00 00 ff 00 30 35 31  ..T..........051
[  424.453815] <3>39 36 47 30 30 30 30 30 30 30 00 00 00 fd 00 37  96G0000000.....7
[  424.453817] <3>4b 18 53 0e 00 0a 20 20 20 20 20 20 00 00 00 fc  K.S...      ....
[  424.453819] <3>00 50 4c 45 34 33 31 0a 20 20 20 20 20 20 00 b0  .PLE431.      ..
[  424.453821] 
[  424.453830] [drm:intel_sdvo_debug_write], SDVOB: W: 7A 02                      (SDVO_CMD_SET_CONTROL_BUS_SWITCH)
[  424.564287] [drm:edid_is_valid] *ERROR* EDID checksum is invalid, remainder is 237
[  424.565918] [drm:edid_is_valid] *ERROR* Raw EDID:
[  424.567514] <3>00 ff ff ff ff ff ff 00 26 cd ac 46 48 10 00 00  ........&..FH...
[  424.567516] <3>01 0c 01 03 00 22 1b 78 2a 2d 90 a4 57 4a 9c 25  .....".x*-..WJ.%
[  424.567518] <3>11 50 54 bf ef 00 81 80 71 4f 01 01 01 01 01 01  .PT.....qO......
[  424.567520] <3>01 01 01 01 01 01 bc 34 00 98 51 00 2a 40 10 90  .......4..Q.*@..
[  424.567522] <3>13 00 54 0e 11 00 00 1e 00 00 00 ff 00 30 35 31  ..T..........051
[  424.567524] <3>39 36 47 30 30 30 30 30 30 30 00 00 00 fd 00 37  96G0000000.....7
[  424.567526] <3>4b 18 53 0e 00 0a 20 20 20 20 20 20 00 00 00 fc  K.S...      ....
[  424.567528] <3>00 50 4c 45 34 33 31 0a 20 20 20 20 20 20 00 b0  .PLE431.      ..
[  424.567530] 
[  424.567535] [drm:intel_sdvo_debug_write], SDVOB: W: 7A 02                      (SDVO_CMD_SET_CONTROL_BUS_SWITCH)
[  424.677996] [drm:edid_is_valid] *ERROR* EDID checksum is invalid, remainder is 237
[  424.679593] [drm:edid_is_valid] *ERROR* Raw EDID:
[  424.681178] <3>00 ff ff ff ff ff ff 00 26 cd ac 46 48 10 00 00  ........&..FH...
[  424.681181] <3>01 0c 01 03 00 22 1b 78 2a 2d 90 a4 57 4a 9c 25  .....".x*-..WJ.%
[  424.681184] <3>11 50 54 bf ef 00 81 80 71 4f 01 01 01 01 01 01  .PT.....qO......
[  424.681186] <3>01 01 01 01 01 01 bc 34 00 98 51 00 2a 40 10 90  .......4..Q.*@..
[  424.681189] <3>13 00 54 0e 11 00 00 1e 00 00 00 ff 00 30 35 31  ..T..........051
[  424.681192] <3>39 36 47 30 30 30 30 30 30 30 00 00 00 fd 00 37  96G0000000.....7
[  424.681194] <3>4b 18 53 0e 00 0a 20 20 20 20 20 20 00 00 00 fc  K.S...      ....
[  424.681197] <3>00 50 4c 45 34 33 31 0a 20 20 20 20 20 20 00 b0  .PLE431.      ..
[  424.681199] 
[  424.681202] i915 0000:00:02.0: DVI-D-1: EDID invalid.
Comment 7 Michael Fu 2010-04-22 20:15:50 UTC
just for confirm - is the Y-cable shipped with the motherboard? we've known that Aopen has its own DVI-I connector defintion , and to get EDID, we have to use its own Y-cable. An ordinary Y-cable won't be able to get EDID for both VGA and DVI on this connector.

When you disable KMS, would you pls submit an xorg.log with Option ModeDebug True set in the xorg.conf?  Using that as a reference would give us hints, too.
Comment 8 grenouille 2010-05-08 01:36:04 UTC
Sorry fore late answer.


> just for confirm - is the Y-cable shipped with the motherboard? we've known
> that Aopen has its own DVI-I connector defintion , and to get EDID, we have to
> use its own Y-cable. An ordinary Y-cable won't be able to get EDID for both VGA
> and DVI on this connector.

I was in contact with aopen support about the cable, it doesn't seem to
be extra. But anyhow I tried to by one to make sure this is not causing
problems. But unfortunately I did not succeed in germany.
http://global.aopen.com/products_detail.aspx?auno=2445
check 50.8EX07.0010 on that side.


> When you disable KMS, would you pls submit an xorg.log with Option ModeDebug
> True set in the xorg.conf?  Using that as a reference would give us hints, too.
Output attached
Thanks for help
Comment 9 grenouille 2010-05-08 01:37:24 UTC
Created attachment 35505 [details]
Xorg.log with Option ModeDebug True
Comment 10 ykzhao 2010-05-08 08:18:48 UTC
Sorry for the late response.
   From the log info in comment #4 it seems that the incorrect EDID is obtained. 
   >[drm:edid_is_valid] *ERROR* Raw EDID

   As the EDID is incorrect, it will be regarded as disconnected. Not sure whether this is related with the monitor.

   Do you have an opportunity to try another monitor and see whether the issue still exists?
Comment 11 grenouille 2010-05-09 01:49:27 UTC
>    >[drm:edid_is_valid] *ERROR* Raw EDID
> 
>    As the EDID is incorrect, it will be regarded as disconnected. Not sure
> whether this is related with the monitor.
> 
>    Do you have an opportunity to try another monitor and see whether the issue
> still exists?

Yes, but this will be after 12.05. because I'm not at home now.

Is there a possibility to hard set the EDID.

P.S. Also with kms disabled I have to correct the monitor settings via xrandr
Comment 12 ykzhao 2010-05-10 01:23:25 UTC
(In reply to comment #11)
> >    >[drm:edid_is_valid] *ERROR* Raw EDID
> > 
> >    As the EDID is incorrect, it will be regarded as disconnected. Not sure
> > whether this is related with the monitor.
> > 
> >    Do you have an opportunity to try another monitor and see whether the issue
> > still exists?
> 
> Yes, but this will be after 12.05. because I'm not at home now.

Ok. I will wait for your test result.

> 
> Is there a possibility to hard set the EDID.

Maybe what you wanted is to override the EDID. Right? Sorry that there is no interface that can support to override the EDID.

> 
> P.S. Also with kms disabled I have to correct the monitor settings via xrandr
Comment 13 grenouille 2010-05-15 13:51:50 UTC
Created attachment 35683 [details]
dmsg with different Monitor

>    Do you have an opportunity to try another monitor and see whether the issue
> still exists?

unlike my other monitor there is no comment about EDID in dmsg after I boot
with kms enbaled, but the DVI output is still reportet as disconnected.

If I run
  cat /sys/devices/pci0000\:00/0000\:00\:02.0/drm/card0/card0-DVI-D-1/status
I get the 
  [drm:edid_is_valid] *ERROR* Raw EDID:
error, see next attachment for details
Comment 14 grenouille 2010-05-15 13:55:12 UTC
Created attachment 35684 [details]
dmesg output after cat /sys/.../drm/card0/card0-DVI-D-1/status

dmsg output after
  cat /sys/devices/pci0000\:00/0000\:00\:02.0/drm/card0/card0-DVI-D-1/status
Comment 15 grenouille 2010-05-15 14:03:56 UTC
V(In reply to comment #12)
> (In reply to comment #11)
> > >    >[drm:edid_is_valid] *ERROR* Raw EDID
> > > 
> > >    As the EDID is incorrect, it will be regarded as disconnected. Not sure
> > > whether this is related with the monitor.
> > > 
> > >    Do you have an opportunity to try another monitor and see whether the issue
> > > still exists?
> > 
> > Yes, but this will be after 12.05. because I'm not at home now.
> > Is there a possibility to hard set the EDID.
> Maybe what you wanted is to override the EDID. Right? Sorry that there is no
> interface that can support to override the EDID.

Yes, that was my idea. Anyhow if there is any possibility to mark the output as
connected it would be fine. I don't care about wrong settings, if I can change
them via xrand to the right modeline.
Comment 16 grenouille 2010-06-19 10:11:32 UTC
I changed the y-cabel to AOpen Best. Nr: 50.8EX07.0010 (wich is not available
in Germany so I have to import it from us - regarding to AOpen-support there is nothing special with this y-cable, but I'm not sure about that) 
AND
I replaced the motherbord (AOpen i965GMt-LA) and now I have no problems with
kms, also the modelines are auto detected correct.

The new cable with the old motherboard didn't work, while the new motherboard
also doesn't work in combination with the old cable. Looks like lice in
combination with fleas ...

Thanks a lot for your support.


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.