* Activate DVI-I behind KVM on FX 5200
@ 2016-01-05 3:08 Thomas Richter
[not found] ` <568B339D.2090102-Jqd5/X81recL63KmMnjC+CEWGD4kr0XT@public.gmane.org>
0 siblings, 1 reply; 7+ messages in thread
From: Thomas Richter @ 2016-01-05 3:08 UTC (permalink / raw)
To: nouveau-PD4FTy7X32lNgt0PjOBp9y5qC8QIuHrW
Hi folks,
I don't seem to be able to enable the DVI-I output of an old FX 5200
behind a KVM switch. Autodetection works fine if the FX 5200 DVI output
is switched to the monitor, but when it is not, I have not found a way
to force-enable it.
Here is what I tried:
*) used the video=DVI-I-1:1280x1024-24@60e kernel parameter
(framebuffer still sits at 1024x768)
*) used "options drm_kms_helper edid-firmware=edid/edid.bin"
in /etc/modprobe.d/nouveau.conf. "no initrd here, drm-kms-helper.ko is
loaded during bootstrap)
*) downloaded the edid from the monitor and validated its correctness.
I placed the edid data into /lib/firmware/edid/edid.bin, and added the
module option "options drm_kms_helper edid-firmware=edid/edid.bin"
as one can see from
"/sys/module/drm_kms_helper/parameters/edid_firmware", the parameter is
accepted, but does not resolve the problem. Framebuffer stays at
1024x768, and "xinit" does not give any display.
Any hints appreciated. It would be very helpful to have a working display.
Greetings,
Thomas
_______________________________________________
Nouveau mailing list
Nouveau@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/nouveau
^ permalink raw reply [flat|nested] 7+ messages in thread[parent not found: <568B339D.2090102-Jqd5/X81recL63KmMnjC+CEWGD4kr0XT@public.gmane.org>]
* Re: Activate DVI-I behind KVM on FX 5200 [not found] ` <568B339D.2090102-Jqd5/X81recL63KmMnjC+CEWGD4kr0XT@public.gmane.org> @ 2016-01-05 10:41 ` poma [not found] ` <568B9DCE.8080407-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org> 2016-01-06 14:33 ` poma 1 sibling, 1 reply; 7+ messages in thread From: poma @ 2016-01-05 10:41 UTC (permalink / raw) To: richter-Jqd5/X81recL63KmMnjC+CEWGD4kr0XT, nouveau-PD4FTy7X32lNgt0PjOBp9y5qC8QIuHrW On 05.01.2016 04:08, Thomas Richter wrote: > Hi folks, > > I don't seem to be able to enable the DVI-I output of an old FX 5200 > behind a KVM switch. Autodetection works fine if the FX 5200 DVI output > is switched to the monitor, but when it is not, I have not found a way > to force-enable it. > > Here is what I tried: > > *) used the video=DVI-I-1:1280x1024-24@60e kernel parameter > (framebuffer still sits at 1024x768) > > *) used "options drm_kms_helper edid-firmware=edid/edid.bin" > in /etc/modprobe.d/nouveau.conf. "no initrd here, drm-kms-helper.ko is > loaded during bootstrap) > > *) downloaded the edid from the monitor and validated its correctness. > I placed the edid data into /lib/firmware/edid/edid.bin, and added the > module option "options drm_kms_helper edid-firmware=edid/edid.bin" > append to kernel cmdline: drm_kms_helper.edid_firmware=DVI-I-1:edid/edid.bin $ cat /proc/cmdline ... drm_kms_helper.edid_firmware=DVI-I-1:edid/edid.bin ... > as one can see from > "/sys/module/drm_kms_helper/parameters/edid_firmware", the parameter is > accepted, but does not resolve the problem. Framebuffer stays at > 1024x768, and "xinit" does not give any display. > > Any hints appreciated. It would be very helpful to have a working display. > > Greetings, > Thomas > _______________________________________________ > Nouveau mailing list > Nouveau@lists.freedesktop.org > http://lists.freedesktop.org/mailman/listinfo/nouveau > _______________________________________________ Nouveau mailing list Nouveau@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/nouveau ^ permalink raw reply [flat|nested] 7+ messages in thread
[parent not found: <568B9DCE.8080407-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>]
* Re: Activate DVI-I behind KVM on FX 5200 [not found] ` <568B9DCE.8080407-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org> @ 2016-01-05 13:47 ` Thomas Richter [not found] ` <568BC959.3070109-Jqd5/X81recL63KmMnjC+CEWGD4kr0XT@public.gmane.org> 0 siblings, 1 reply; 7+ messages in thread From: Thomas Richter @ 2016-01-05 13:47 UTC (permalink / raw) To: poma, nouveau-PD4FTy7X32lNgt0PjOBp9y5qC8QIuHrW Am 05.01.2016 um 11:41 schrieb poma: > > append to kernel cmdline: > drm_kms_helper.edid_firmware=DVI-I-1:edid/edid.bin > > $ cat /proc/cmdline > ... drm_kms_helper.edid_firmware=DVI-I-1:edid/edid.bin ... > Well, no banana. Yes, the kernel loads the edid, but the screen keeps blank if I switch the monitor to the system after bootstrap. )-: [ 20.319271] [drm] Got external EDID base block and 0 extensions from "edid/edid.bin" for connector "DVI-I-1" [ 20.347274] [drm] Got external EDID base block and 0 extensions from "edid/edid.bin" for connector "DVI-I-1" I also replaced that with the "dummy" edid/1280x1024.bin parameter, same thing. If I boot with the monitor connected, everything is fine. If I boot with it disconnected, no chance. I also tried to force the DVI connector on: video=DVI-I-1:e The result is that the kernel still loads the edid, but the screen remains now blank all the time, even if I boot with the monitor connected. Yes, it is really connected to DVI-I-1, at least according to xrandr without the "video" parameter, and with the monitor connected. Greetings, Thomas _______________________________________________ Nouveau mailing list Nouveau@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/nouveau ^ permalink raw reply [flat|nested] 7+ messages in thread
[parent not found: <568BC959.3070109-Jqd5/X81recL63KmMnjC+CEWGD4kr0XT@public.gmane.org>]
* Re: Activate DVI-I behind KVM on FX 5200 [not found] ` <568BC959.3070109-Jqd5/X81recL63KmMnjC+CEWGD4kr0XT@public.gmane.org> @ 2016-01-05 18:43 ` poma [not found] ` <568C0EE4.50307-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org> 0 siblings, 1 reply; 7+ messages in thread From: poma @ 2016-01-05 18:43 UTC (permalink / raw) To: richter-Jqd5/X81recL63KmMnjC+CEWGD4kr0XT, nouveau-PD4FTy7X32lNgt0PjOBp9y5qC8QIuHrW On 05.01.2016 14:47, Thomas Richter wrote: > Am 05.01.2016 um 11:41 schrieb poma: > >> >> append to kernel cmdline: >> drm_kms_helper.edid_firmware=DVI-I-1:edid/edid.bin >> >> $ cat /proc/cmdline >> ... drm_kms_helper.edid_firmware=DVI-I-1:edid/edid.bin ... >> > > Well, no banana. Yes, the kernel loads the edid, but the screen keeps > blank if I switch the monitor to the system after bootstrap. )-: > > [ 20.319271] [drm] Got external EDID base block and 0 extensions from > "edid/edid.bin" for connector "DVI-I-1" > [ 20.347274] [drm] Got external EDID base block and 0 extensions from > "edid/edid.bin" for connector "DVI-I-1" > > I also replaced that with the "dummy" edid/1280x1024.bin parameter, same > thing. > > If I boot with the monitor connected, everything is fine. If I boot with > it disconnected, no chance. > > I also tried to force the DVI connector on: > > video=DVI-I-1:e > > The result is that the kernel still loads the edid, but the screen > remains now blank all the time, even if I boot with the monitor connected. > > Yes, it is really connected to DVI-I-1, at least according to xrandr > without the "video" parameter, and with the monitor connected. > > Greetings, > Thomas > Try your luck here http://lists.freedesktop.org/mailman/listinfo/dri-devel BTW what is actual KVM switch device, vendor/product? _______________________________________________ Nouveau mailing list Nouveau@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/nouveau ^ permalink raw reply [flat|nested] 7+ messages in thread
[parent not found: <568C0EE4.50307-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>]
* Re: Activate DVI-I behind KVM on FX 5200 [not found] ` <568C0EE4.50307-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org> @ 2016-01-05 19:31 ` Thomas Richter [not found] ` <568C1A26.7010004-otB+qtk3XKcL63KmMnjC+CEWGD4kr0XT@public.gmane.org> 0 siblings, 1 reply; 7+ messages in thread From: Thomas Richter @ 2016-01-05 19:31 UTC (permalink / raw) To: nouveau-PD4FTy7X32lNgt0PjOBp9y5qC8QIuHrW Hi, > > Try your luck here > http://lists.freedesktop.org/mailman/listinfo/dri-devel The major problem is that there is no way to tell the nouveau kernel module to enforce a specific output. There is a video=XXX kernel parameter, but this only applies to the VGA framebuffer but not to the nouveau framebuffer. > BTW what is actual KVM switch device, vendor/product? This is an ATEN CS22D. It is "supposed to remember the EDID data", but frankly, it does not. Greetings, Thomas _______________________________________________ Nouveau mailing list Nouveau@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/nouveau ^ permalink raw reply [flat|nested] 7+ messages in thread
[parent not found: <568C1A26.7010004-otB+qtk3XKcL63KmMnjC+CEWGD4kr0XT@public.gmane.org>]
* Re: Activate DVI-I behind KVM on FX 5200 [not found] ` <568C1A26.7010004-otB+qtk3XKcL63KmMnjC+CEWGD4kr0XT@public.gmane.org> @ 2016-01-06 4:37 ` poma 0 siblings, 0 replies; 7+ messages in thread From: poma @ 2016-01-06 4:37 UTC (permalink / raw) To: Thomas Richter, nouveau-PD4FTy7X32lNgt0PjOBp9y5qC8QIuHrW On 05.01.2016 20:31, Thomas Richter wrote: > Hi, > >> >> Try your luck here >> http://lists.freedesktop.org/mailman/listinfo/dri-devel > > The major problem is that there is no way to tell the nouveau kernel > module to enforce a specific output. There is a video=XXX kernel > parameter, but this only applies to the VGA framebuffer but not to the > nouveau framebuffer. > Tested, in contrast to enablement, video=DVI-I-1:d becomes effective at the exact moment: fb: switching to nouveaufb from VESA VGA However this is a direct connection, without KVM switch. >> BTW what is actual KVM switch device, vendor/product? > > This is an ATEN CS22D. It is "supposed to remember the EDID data", but > frankly, it does not. > http://eservice.aten.com/eServiceCx/Common/FAQ/view.do?id=4746 "No, The CS22D/CS22U do not support Video Dynasync feature, it bypass's EDID." _______________________________________________ Nouveau mailing list Nouveau@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/nouveau ^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: Activate DVI-I behind KVM on FX 5200 [not found] ` <568B339D.2090102-Jqd5/X81recL63KmMnjC+CEWGD4kr0XT@public.gmane.org> 2016-01-05 10:41 ` poma @ 2016-01-06 14:33 ` poma 1 sibling, 0 replies; 7+ messages in thread From: poma @ 2016-01-06 14:33 UTC (permalink / raw) To: richter-Jqd5/X81recL63KmMnjC+CEWGD4kr0XT, nouveau-PD4FTy7X32lNgt0PjOBp9y5qC8QIuHrW On 05.01.2016 04:08, Thomas Richter wrote: > Hi folks, > > I don't seem to be able to enable the DVI-I output of an old FX 5200 > behind a KVM switch. Autodetection works fine if the FX 5200 DVI output > is switched to the monitor, but when it is not, I have not found a way > to force-enable it. > > Here is what I tried: > > *) used the video=DVI-I-1:1280x1024-24@60e kernel parameter > (framebuffer still sits at 1024x768) > ... Tested here, in "video=DVI-I-1:1280x1024-24@60e" directive problematic are: color depth (bpp) 24 -and- 'e' as "forced display enablement" causing display -switching off- at the exact moment: fb: switching to nouveaufb from VESA VGA Contrairement à ce que, these 3 examples have been successfully tested: 1. video=DVI-I-1:1280x1024 2. video=DVI-I-1:1280x1024@60 3. video=DVI-I-1:1280x1024-16@60 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ # cat /proc/cmdline ... video=DVI-I-1:1280x1024 ... # fbset -i mode "1280x1024" geometry 1280 1024 1280 1024 32 timings 0 0 0 0 0 0 0 accel true rgba 8/16,8/8,8/0,0/0 endmode Frame buffer device information: Name : nouveaufb Address : 0xd0006000 Size : 5242880 Type : PACKED PIXELS Visual : TRUECOLOR XPanStep : 1 YPanStep : 1 YWrapStep : 0 LineLength : 5120 Accelerator : No # dmesg -t | egrep fb:\ 0x nouveau 0000:02:00.0: DRM: allocated 1280x1024 fb: 0x50000, bo ffff88007f5d2800 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ # cat /proc/cmdline ... video=DVI-I-1:1280x1024@60 ... # fbset -i mode "1280x1024" geometry 1280 1024 1280 1024 32 timings 0 0 0 0 0 0 0 accel true rgba 8/16,8/8,8/0,0/0 endmode Frame buffer device information: Name : nouveaufb Address : 0xd0006000 Size : 5242880 Type : PACKED PIXELS Visual : TRUECOLOR XPanStep : 1 YPanStep : 1 YWrapStep : 0 LineLength : 5120 Accelerator : No ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ # cat /proc/cmdline ... video=DVI-I-1:1280x1024-16@60 ... # fbset -i mode "1280x1024" geometry 1280 1024 1280 1024 16 timings 0 0 0 0 0 0 0 accel true rgba 5/11,6/5,5/0,0/0 endmode Frame buffer device information: Name : nouveaufb Address : 0xd0006000 Size : 2621440 Type : PACKED PIXELS Visual : TRUECOLOR XPanStep : 1 YPanStep : 1 YWrapStep : 0 LineLength : 2560 Accelerator : No Ref. https://www.kernel.org/doc/Documentation/fb/modedb.txt p.s. Can you paste here the output of these two oneliners: $ cd /sys/class/drm ; ls -1d card0-* | sed s/card0-// | sort $ xrandr | grep connected | awk '{print $1}' | sort _______________________________________________ Nouveau mailing list Nouveau@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/nouveau ^ permalink raw reply [flat|nested] 7+ messages in thread
end of thread, other threads:[~2016-01-06 14:33 UTC | newest]
Thread overview: 7+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2016-01-05 3:08 Activate DVI-I behind KVM on FX 5200 Thomas Richter
[not found] ` <568B339D.2090102-Jqd5/X81recL63KmMnjC+CEWGD4kr0XT@public.gmane.org>
2016-01-05 10:41 ` poma
[not found] ` <568B9DCE.8080407-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
2016-01-05 13:47 ` Thomas Richter
[not found] ` <568BC959.3070109-Jqd5/X81recL63KmMnjC+CEWGD4kr0XT@public.gmane.org>
2016-01-05 18:43 ` poma
[not found] ` <568C0EE4.50307-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
2016-01-05 19:31 ` Thomas Richter
[not found] ` <568C1A26.7010004-otB+qtk3XKcL63KmMnjC+CEWGD4kr0XT@public.gmane.org>
2016-01-06 4:37 ` poma
2016-01-06 14:33 ` poma
This is an external index of several public inboxes, see mirroring instructions on how to clone and mirror all data and code used by this external index.