How do I disable edid check on kernel command line?

[Flash]

Current Kernel is 3.14.9-200 x86_64. This is a multihead setup; i915 and 
NV96 video cards (displays HMDI2 = i915 and NV96 DVI-I-1 and VGA-2 [both 
X233H Acer]). kernel bootmenu is presented from DVI-I-1, after kernel 
selection, LUKS prompt appears on HMDI2 and DVI-I-1 is disconnected. In 
multi-user env HDMI2 and VGA-2 are active and dmesg shows edid complaint 
about DVI-I-1. /usr/bin/edid-decode complains about edid block for both 
NV96 displays DVI and VGA but for whatever reason DVI is only one 
disconnected.

So far searches in various forum threads and bugzilla suggest buggy DVI 
edid problems been around for awhile. The DVI edid block error follows 
DVI connector, meaning, moving DVI cable to monitor previously with 
label VGA-2 will produce same edid message. I haven't found complete 
instructions to build CustomEdid yet, that seems the way to go since 
commandline settings like drm_edid_strict, video=DVI-I-1:e are ignored.

Pointers to documentation to build customedid or kernel commandline to 
disable DVI edid checking will be greatly appreciated.

Thanks in advance.

[poma]

On 01.07.2014 18:50, flashl(a)dailybrood.com wrote:
>  Current Kernel is 3.14.9-200 x86_64. This is a multihead setup; i915
> and
>  NV96 video cards (displays HMDI2 = i915 and NV96 DVI-I-1 and VGA-2 [both
>  X233H Acer]). kernel bootmenu is presented from DVI-I-1, after kernel
>  selection, LUKS prompt appears on HMDI2 and DVI-I-1 is disconnected. In
>  multi-user env HDMI2 and VGA-2 are active and dmesg shows edid complaint
>  about DVI-I-1. /usr/bin/edid-decode complains about edid block for both
>  NV96 displays DVI and VGA but for whatever reason DVI is only one
>  disconnected.
>
>  So far searches in various forum threads and bugzilla suggest buggy DVI
>  edid problems been around for awhile. The DVI edid block error follows
>  DVI connector, meaning, moving DVI cable to monitor previously with
>  label VGA-2 will produce same edid message. I haven't found complete
>  instructions to build CustomEdid yet, that seems the way to go since
>  commandline settings like drm_edid_strict, video=DVI-I-1:e are ignored.
>
>  Pointers to documentation to build customedid or kernel commandline to
>  disable DVI edid checking will be greatly appreciated.
>
>  Thanks in advance.
>  
https://www.kernel.org/doc/Documentation/kernel-parameters.txt
drm_kms_helper.edid_firmware=[<connector>:]<file>
		Broken monitors, graphic adapters and KVMs may
		send no or incorrect EDID data sets. This parameter
		allows to specify an EDID data set in the
		/lib/firmware directory that is used instead.
		Generic built-in EDID data sets are used, if one of
		edid/1024x768.bin, edid/1280x1024.bin,
		edid/1680x1050.bin, or edid/1920x1080.bin is given
		and no file with the same name exists. Details and
		instructions how to build your own EDID data are
		available in Documentation/EDID/HOWTO.txt. An EDID
		data set will only be used for a particular connector,
		if its name and a colon are prepended to the EDID
		name.

https://www.kernel.org/doc/Documentation/EDID/HOWTO.txt


poma

[Flash]

Thanks for reply. Finally realized why I couldn't locate EDID folder, 
kernel docs weren't installed.
Built default edids, copied 1920x1080.bin to /lib/firmware and 
/lib/firmware/edid, adds 
drm_kms_helper.edid_firmware=DVI-I-1:edid/1920x1080.bin drm_debug=14 
log_buf_len=16M to commandline. No change in behavior, commandline 
firmware helper ignored. Boot DVI-I-1 signal drops during and/or after 
LUKS prompt. Next, invalid EDID message appears although 
/usr/bin/edid-decode validates /lib/firmware/1920x1080.bin.

My guesstimation so far is (1) once DVI-I-1 signal drops during boot 
nothing I've done so far 're-connects' DVI-I-1; (2) although 
1920x1080.bin is placed in /lib/firmware and /lib/firmware/edid neither 
is found yet there is no notification/complaint; (3) for whatever 
reason, video:DVI-I-1:e is also ignored. Is there another commandline 
option to keep signal on for DVI-I-1?

Thanks again

On 2014-07-01 15:41, poma wrote:
>  On 01.07.2014 18:50, flashl(a)dailybrood.com wrote:
> > Current Kernel is 3.14.9-200 x86_64. This is a multihead setup; i915 
> > and
> > NV96 video cards (displays HMDI2 = i915 and NV96 DVI-I-1 and VGA-2 
> > [both
> > X233H Acer]). kernel bootmenu is presented from DVI-I-1, after kernel
> > selection, LUKS prompt appears on HMDI2 and DVI-I-1 is disconnected. 
> > In
> > multi-user env HDMI2 and VGA-2 are active and dmesg shows edid 
> > complaint
> > about DVI-I-1. /usr/bin/edid-decode complains about edid block for 
> > both
> > NV96 displays DVI and VGA but for whatever reason DVI is only one
> > disconnected.
> > 
> > So far searches in various forum threads and bugzilla suggest buggy 
> > DVI
> > edid problems been around for awhile. The DVI edid block error follows
> > DVI connector, meaning, moving DVI cable to monitor previously with
> > label VGA-2 will produce same edid message. I haven't found complete
> > instructions to build CustomEdid yet, that seems the way to go since
> > commandline settings like drm_edid_strict, video=DVI-I-1:e are 
> > ignored.
> > 
> > Pointers to documentation to build customedid or kernel commandline to
> > disable DVI edid checking will be greatly appreciated.
> > 
> > Thanks in advance.
> > 
>  
>  https://www.kernel.org/doc/Documentation/kernel-parameters.txt
>  drm_kms_helper.edid_firmware=[<connector>:]<file>
>  	Broken monitors, graphic adapters and KVMs may
>  	send no or incorrect EDID data sets. This parameter
>  	allows to specify an EDID data set in the
>  	/lib/firmware directory that is used instead.
>  	Generic built-in EDID data sets are used, if one of
>  	edid/1024x768.bin, edid/1280x1024.bin,
>  	edid/1680x1050.bin, or edid/1920x1080.bin is given
>  	and no file with the same name exists. Details and
>  	instructions how to build your own EDID data are
>  	available in Documentation/EDID/HOWTO.txt. An EDID
>  	data set will only be used for a particular connector,
>  	if its name and a colon are prepended to the EDID
>  	name.
>  
>  https://www.kernel.org/doc/Documentation/EDID/HOWTO.txt
>  
>  
>  poma

[poma]

On 02.07.2014 15:12, Flash wrote:
>  Thanks for reply. Finally realized why I couldn't locate EDID
> folder,
>  kernel docs weren't installed.
>  Built default edids, copied 1920x1080.bin to /lib/firmware and
>  /lib/firmware/edid, adds
>  drm_kms_helper.edid_firmware=DVI-I-1:edid/1920x1080.bin drm_debug=14
>  log_buf_len=16M to commandline. No change in behavior, commandline
>  firmware helper ignored. Boot DVI-I-1 signal drops during and/or after
>  LUKS prompt. Next, invalid EDID message appears although
>  /usr/bin/edid-decode validates /lib/firmware/1920x1080.bin.
>
>  My guesstimation so far is (1) once DVI-I-1 signal drops during boot
>  nothing I've done so far 're-connects' DVI-I-1; (2) although
>  1920x1080.bin is placed in /lib/firmware and /lib/firmware/edid neither
>  is found yet there is no notification/complaint; (3) for whatever
>  reason, video:DVI-I-1:e is also ignored. Is there another commandline
>  option to keep signal on for DVI-I-1?
>
>  Thanks again
>  
Déjà vu?
Multi-head intel and nvidia configuration fails 100%
https://lists.fedoraproject.org/pipermail/test/2013-December/119277.html

What happens when you run Rawhide?


poma

[Flash]

Not Deja vu, per se. Found solution by adjusting bios parameters. IIRC, 
today's behaviors appeared after changes to systemd-udev. If I could 
backtrack to the point when oddities first started, I would downgrade 
system. I'll burn rawhide DVD and give it a try.

Thanks

On 2014-07-02 09:55, poma wrote:
>  On 02.07.2014 15:12, Flash wrote:
> > Thanks for reply. Finally realized why I couldn't locate EDID folder,
> > kernel docs weren't installed.
> > Built default edids, copied 1920x1080.bin to /lib/firmware and
> > /lib/firmware/edid, adds
> > drm_kms_helper.edid_firmware=DVI-I-1:edid/1920x1080.bin drm_debug=14
> > log_buf_len=16M to commandline. No change in behavior, commandline
> > firmware helper ignored. Boot DVI-I-1 signal drops during and/or after
> > LUKS prompt. Next, invalid EDID message appears although
> > /usr/bin/edid-decode validates /lib/firmware/1920x1080.bin.
> > 
> > My guesstimation so far is (1) once DVI-I-1 signal drops during boot
> > nothing I've done so far 're-connects' DVI-I-1; (2) although
> > 1920x1080.bin is placed in /lib/firmware and /lib/firmware/edid 
> > neither
> > is found yet there is no notification/complaint; (3) for whatever
> > reason, video:DVI-I-1:e is also ignored. Is there another commandline
> > option to keep signal on for DVI-I-1?
> > 
> > Thanks again
> > 
>  
>  Déjà vu?
>  Multi-head intel and nvidia configuration fails 100%
>  https://lists.fedoraproject.org/pipermail/test/2013-December/119277.html
>  
>  What happens when you run Rawhide?
>  
>  
>  poma

[poma]

On 02.07.2014 16:19, Flash wrote:
>  Not Deja vu, per se. Found solution by adjusting bios parameters.
> IIRC,
>  today's behaviors appeared after changes to systemd-udev. If I could
>  backtrack to the point when oddities first started, I would downgrade
>  system. I'll burn rawhide DVD and give it a try.
>
>  Thanks
>  
Care to clarify, these BIOS settings are exactly what?


poma

[Flash]

I discovered after rebooting what seems like a hundred times that 
adjusting MB MSI Z77A-G43 bios integrated graphic interface option 
changes  whether graphic card was discovered, discovery  sequence, 
system preference during boot, which graphic card recognized after boot 
sequence, naming conventions and other behaviors at multi-user target. 
i.e. dvi name in multi-user state could be dvi-i-1 or maybe dvi-i-1-1 or 
dvi-i-1-1-1 or Xorg may pick the device as primary; and, if I didn't 
include the i915 graphic card in the integrated graphic boot sequence, 
it would remain in a disconnected state.

I finally succeeded building a stable xorg.conf by referencing Xorg log 
messages and adjusting the file accordingly. Manpages, documentation, 
web searches, or other references to expected behavior didn't help.

On 2014-07-02 17:58, poma wrote:
>  On 02.07.2014 16:19, Flash wrote:
> > Not Deja vu, per se. Found solution by adjusting bios parameters. 
> > IIRC,
> > today's behaviors appeared after changes to systemd-udev. If I could
> > backtrack to the point when oddities first started, I would downgrade
> > system. I'll burn rawhide DVD and give it a try.
> > 
> > Thanks
> > 
>  
>  Care to clarify, these BIOS settings are exactly what?
>  
>  
>  poma

[poma]

On 03.07.2014 13:35, Flash wrote:
>  I discovered after rebooting what seems like a hundred times that
>  adjusting MB MSI Z77A-G43 bios integrated graphic interface option
>  changes  whether graphic card was discovered, discovery  sequence,
>  system preference during boot, which graphic card recognized after boot
>  sequence, naming conventions and other behaviors at multi-user target.
>  i.e. dvi name in multi-user state could be dvi-i-1 or maybe dvi-i-1-1 or
>  dvi-i-1-1-1 or Xorg may pick the device as primary; and, if I didn't
>  include the i915 graphic card in the integrated graphic boot sequence,
>  it would remain in a disconnected state.
>
>  I finally succeeded building a stable xorg.conf by referencing Xorg log
>  messages and adjusting the file accordingly. Manpages, documentation,
>  web searches, or other references to expected behavior didn't help. 
You probably feel like Columbus amid the discovery of new lands. :)
It would be nice to record somewhere publicly entire travel route, captain.


poma

[Flash]

On 2014-07-03 08:06, poma wrote:
>  On 03.07.2014 13:35, Flash wrote:
> > I discovered after rebooting what seems like a hundred times that
> > adjusting MB MSI Z77A-G43 bios integrated graphic interface option
> > changes  whether graphic card was discovered, discovery  sequence,
> > system preference during boot, which graphic card recognized after 
> > boot
> > sequence, naming conventions and other behaviors at multi-user target.
> > i.e. dvi name in multi-user state could be dvi-i-1 or maybe dvi-i-1-1 
> > or
> > dvi-i-1-1-1 or Xorg may pick the device as primary; and, if I didn't
> > include the i915 graphic card in the integrated graphic boot sequence,
> > it would remain in a disconnected state.
> > 
> > I finally succeeded building a stable xorg.conf by referencing Xorg 
> > log
> > messages and adjusting the file accordingly. Manpages, documentation,
> > web searches, or other references to expected behavior didn't help.
>  
>  You probably feel like Columbus amid the discovery of new lands. :)
>  It would be nice to record somewhere publicly entire travel route, 
>  captain.
>  
>  
>  poma 
Truly, I don't mind the odyssey. It just become a bit irritating for me 
sometimes without error messages guiding me. :D

thanks for taking time to respond
flash