FS#58666 - [linux, nvidia, xorg-server] Blank screen after boot

Attached to Project: Arch Linux
Opened by PF (greyeye) - Saturday, 19 May 2018, 20:35 GMT
Last edited by Doug Newgard (Scimmia) - Monday, 21 May 2018, 05:23 GMT
Task Type Bug Report
Category Packages: Extra
Status Closed
Assigned To No-one
Architecture All
Severity High
Priority Normal
Reported Version
Due in Version Undecided
Due Date Undecided
Percent Complete 100%
Votes 0
Private No

Details

Description:
Upgrading linux, nvidia and/or xorg-server causes blank screen after boot. This happens only on my desktop PC, on my laptop (without an Nvidia card) I have no issues.

Additional info:
Package versions
-linux-4.16.8-1 and 4.16.9-1
-nvidia-396.24
-xorg-server-1.20.0

Graphics card
-GeForce GTX 570

Steps to reproduce:
-Upgrade any or all of the previously mentioned packages
-Reboot and see the blank screen

Reverting back to linux-4.16.7-1, nvidia-390.48-13, xorg-server-1.19.6+13+gd0d1a694f-2 fixes the issue.
This task depends upon

Closed by  Doug Newgard (Scimmia)
Monday, 21 May 2018, 05:23 GMT
Reason for closing:  Duplicate
Additional comments about closing:   FS#58558 
Comment by loqs (loqs) - Saturday, 19 May 2018, 21:12 GMT
When you have a blank screen on boot is that because the system starts X on boot? If so what is the Xorg.#.log
In any case please include the dmesg for a blank screen boot.
Comment by Danny Helmke (Dittsche) - Sunday, 20 May 2018, 10:57 GMT
I do have the same issue. I ran the latest updates available yesterday, but still the same problem, black screen whilst the boot. On my machine it started last week, after having installed updates. First it started with graphic errors and then only black screens.
I have not yet checked the log files, but I will do. As a work around I deinstalled the Nvidia driver.
Comment by PF (greyeye) - Sunday, 20 May 2018, 11:40 GMT
Ok, solved it. Apparently Nvidia had dropped the support for GeForce GTX 570 in their main driver branch ('nvidia' package in repos). GeForce GTX 570 is from now on supported through legacy driver 390.xx ('nvidia-390xx' package in repos). (List of supported cards: http://www.nvidia.com/object/IO_32667.html) I installed nvidia-390xx and upgraded linux + xorg-server and everything works fine now. I wonder why I haven't seen any information about the dropped support... Don't know if it's possible but maybe there could be a message during nvidia upgrade if the card is not supported.
Comment by Johannes Wegener (hpfmn) - Sunday, 20 May 2018, 13:57 GMT
I don't know if that is the exact same issue. I'm starting Xorg manually and since upgrading to 396.24 it won't start. Downgrading to the nvidia-390xx packages resolved the issue.
The interesssting parts from the Xorg.0.log seem to be these.

[ 30.941] (EE) NVIDIA(GPU-0): Failed to initialize the NVIDIA GPU at PCI:1:0:0. Please
[ 30.941] (EE) NVIDIA(GPU-0): check your system's kernel log for additional error
[ 30.941] (EE) NVIDIA(GPU-0): messages and refer to Chapter 8: Common Problems in the
[ 30.941] (EE) NVIDIA(GPU-0): README for additional information.
[ 30.941] (EE) NVIDIA(GPU-0): Failed to initialize the NVIDIA graphics device!
[ 30.941] (EE) NVIDIA(0): Failing initialization of X screen 0

I'm using a Nvidia GTX 1060 on an Asus Z87 Pro board with an intel i5 4670k.

If the need arises I could also upload the whole log file.
Comment by Johannes Wegener (hpfmn) - Sunday, 20 May 2018, 14:02 GMT
Ah I guess the coressponding kernel messages are these:

May 20 15:47:46 hp810_Arch kernel: NVRM: rm_init_adapter failed for device bearing minor number 0
May 20 15:47:46 hp810_Arch kernel: NVRM: RmInitAdapter failed! (0x53:0xffff:1957)
May 20 15:47:26 hp810_Arch kernel: NVRM: Xid (PCI:0000:01:00): 62, 0ac0(2f10) 00000000 00000000
May 20 15:47:26 hp810_Arch kernel: NVRM: GPU Board Serial Number:
May 20 15:47:26 hp810_Arch kernel: NVRM: GPU at PCI:0000:01:00: GPU-13321be1-f3c0-0eec-c139-c3d4eab90da5
May 20 15:47:26 hp810_Arch kernel: caller _nv001112rm+0xe3/0x1d0 [nvidia] mapping multiple BARs
May 20 15:47:26 hp810_Arch kernel: resource sanity check: requesting [mem 0x000c0000-0x000fffff], which spans more than PCI Bus 0000:00 [mem 0x000d0000-0x000d>
Comment by Johannes Wegener (hpfmn) - Sunday, 20 May 2018, 14:06 GMT

Loading...