Red Hat Bugzilla – Bug 1002497
[Quadro K1000M] three monitor configuration do not work for GK107
Last modified: 2017-08-01 01:57:53 EDT
Created attachment 791689 [details]
Description of problem:
Hi as Ben confirmed me GK107 should be able to handle 4 monitor setup.
I with Dell M4700 have in dock with two NEC E231W connected (via DP).
g-d-p actually shows the monitor but the third in row stays dark with no picture on it. The system actually act like it is on, since the virtual display is correctly enlarged over three displays and one can place windows on the third display actually.
Mirror do not work as well, or to be precise act the same.
Machine booted with discrete graphics only.
Version-Release number of selected component (if applicable):
Steps to Reproduce:
third monitor in row stays dark
GK107 should be able to handle panel+2 monitors connected
01:00.0 VGA compatible controller : NVIDIA Corporation GK107GLM [Quadro K1000M] [10de:0ffc] (rev a1) (prog-if 00 [VGA controller])
Subsystem: Dell Device [1028:053e]
Flags: bus master, fast devsel, latency 0, IRQ 16
Memory at f5000000 (32-bit, non-prefetchable) [size=16M]
Memory at e0000000 (64-bit, prefetchable) [size=256M]
Memory at f0000000 (64-bit, prefetchable) [size=32M]
I/O ports at e000 [size=128]
Expansion ROM at f6000000 [disabled] [size=512K]
Capabilities:  Power Management version 3
Capabilities:  MSI: Enable- Count=1/1 Maskable- 64bit+
Capabilities:  Express Endpoint, MSI 00
Capabilities: [b4] Vendor Specific Information: Len=14 <?>
Capabilities:  Virtual Channel
Capabilities:  Power Budgeting <?>
Capabilities:  Vendor Specific Information: ID=0001 Rev=1 Len=024 <?>
Capabilities:  #19
Kernel driver in use: nouveau
Kernel modules: nouveau, nvidiafb
Manufacturer: Dell Inc.
Product Name: Precision M4700
Serial Number: GB069W1
Wake-up Type: Power Switch
SKU Number: Not Specified
Family: Not Specified
I have this machine on my table I can access it for debugging if needed.
Created attachment 791691 [details]
Created attachment 791694 [details]
Are you able to test this with F19 + latest kernel? There's a very significant number of changes in nouveau since 6.4 (which are in 6.5), so it'd be useful to know the status there.
Created attachment 792179 [details]
Created attachment 792180 [details]
Created attachment 792181 [details]
(In reply to Tomas Pelka from comment #4)
> Created attachment 792179 [details]
Linux dhcp130-204.brq.redhat.com 3.10.9-200.fc19.x86_64 #1 SMP Wed Aug 21 19:27:58 UTC 2013 x86_64 x86_64 x86_64 GNU/Linux
Situation is exactly the same.
Does the non-working port work in a two monitor configuration?
(In reply to Ben Skeggs from comment #8)
> Does the non-working port work in a two monitor configuration?
Nope, g-d-p (xrandr) will see it but still no screen on it.
(In reply to Tomas Pelka from comment #9)
> (In reply to Ben Skeggs from comment #8)
> > Does the non-working port work in a two monitor configuration?
> Nope, g-d-p (xrandr) will see it but still no screen on it.
Ok, in tht case, it's not a triple-head problem (good, because that should work!).. But, bad.. Because, I now don't have a clear idea of what's wrong here.
Can you attach the vbios image (/sys/kernel/debug/dri/0/vbios.rom, might need to mount debugfs on /sys/kernel/debug first), and give me the xrandr output name of the non-working port?
(In reply to Ben Skeggs from comment #10)
> (In reply to Tomas Pelka from comment #9)
> Ok, in tht case, it's not a triple-head problem (good, because that should
> work!).. But, bad.. Because, I now don't have a clear idea of what's wrong
> Can you attach the vbios image (/sys/kernel/debug/dri/0/vbios.rom, might
> need to mount debugfs on /sys/kernel/debug first), and give me the xrandr
> output name of the non-working port?
The nonworking output is DP-1
My laptop have no /sys/kernel/debug/dri/0/vbios.rom, is that expected?
$ mount | grep debug
none on /sys/kernel/debug type debugfs (rw)
$ ls /sys/kernel/debug/dri/*
bufs clients gem_names name vm vma
bufs clients gem_names name vm vma
$ lsmod | grep nouveau
nouveau 876885 2
ttm 80957 1 nouveau
drm_kms_helper 40087 1 nouveau
drm 265866 4 nouveau,ttm,drm_kms_helper
i2c_algo_bit 5935 1 nouveau
i2c_core 31084 6 i2c_i801,videodev,nouveau,drm_kms_helper,drm,i2c_algo_bit
mxm_wmi 1967 1 nouveau
video 20674 1 nouveau
wmi 6287 2 nouveau,mxm_wmi
I suspect you're probably on one of the kernel versions where it was missing temporarily. Kernel 3.8 and up should have it back.
Created attachment 795868 [details]
vbios.rom from dri/0/
Created attachment 795870 [details]
vbios.rom from dri/64/
I have a feeling this is going to be one of those not-so-fun cases where I'm going to need the hardware to play with directly.
But, let's keep looking for a bit. Can you copy /sys/firmware/acpi/tables somewhere (yes, copy first, tar directly doesn't work well on sysfs), and tar it up and attach for me please?
Created attachment 796662 [details]
This request was not resolved in time for the current release.
Red Hat invites you to ask your support representative to
propose this request, if still desired, for consideration in
the next release of Red Hat Enterprise Linux.