NetBSD Problem Report #58053
From paul@whooppee.com Tue Mar 19 16:24:29 2024
Return-Path: <paul@whooppee.com>
Received: from mail.netbsd.org (mail.netbsd.org [199.233.217.200])
(using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits))
(Client CN "mail.NetBSD.org", Issuer "mail.NetBSD.org CA" (not verified))
by mollari.NetBSD.org (Postfix) with ESMTPS id 4A56F1A9239
for <gnats-bugs@gnats.NetBSD.org>; Tue, 19 Mar 2024 16:24:29 +0000 (UTC)
Message-Id: <20240319162404.248AB5E33C5@speedy.whooppee.com>
Date: Tue, 19 Mar 2024 09:24:04 -0700 (PDT)
From: paul@whooppee.com
Reply-To: paul@whooppee.com
To: gnats-bugs@NetBSD.org
Subject: Issues on -current with nouveau graphics
X-Send-Pr-Version: 3.95
>Number: 58053
>Category: kern
>Synopsis: Issues on -current with nouveau graphics
>Confidential: no
>Severity: serious
>Priority: high
>Responsible: kern-bug-people
>State: open
>Class: sw-bug
>Submitter-Id: net
>Arrival-Date: Tue Mar 19 16:25:00 +0000 2024
>Last-Modified: Sat May 18 17:45:01 +0000 2024
>Originator: Paul Goyette
>Release: NetBSD 10.99.10
>Organization:
+---------------------+--------------------------+----------------------+
| Paul Goyette (.sig) | PGP Key fingerprint: | E-mail addresses: |
| (Retired) | 1B11 1849 721C 56C8 F63A | paul@whooppee.com |
| Software Developer | 6E2E 05FD 15CE 9F2D 5102 | pgoyette@netbsd.org |
| & Network Engineer | | pgoyette99@gmail.com |
+---------------------+--------------------------+----------------------+
>Environment:
System: NetBSD speedy.whooppee.com 10.99.10 NetBSD 10.99.10 (SPEEDY 2024-03-18 02:46:01 UTC) #0: Mon Mar 18 05:00:00 UTC 2024 paul@speedy.whooppee.com:/build/netbsd-local/obj/amd64/sys/arch/amd64/compile/SPEEDY amd64
Architecture: x86_64
Machine: amd64
>Description:
>How-To-Repeat:
With -current sources updated within the last few days, on both a
GENERIC and local-custom kernel I am seeing some issues with the
nouveau video driver. For system details, the dmesg is posted
on nycbug. Of particular note is the graphics card, a GeForce
730 with 2GB GDDR5:
nouveau0 at pci1 dev 0 function 0: vendor 10de product
1287 (rev. 0xa1)
...
nouveau0: NVIDIA GK208B (b06070b1)
nouveau0: bios: version 80.28.b8.00.13
nouveau0: interrupting at msi14 vec 0 (nouveau0)
nouveau0: fb: 2048 MiB GDDR5
Zone kernel: Available graphics memory: 9007199250255680 KiB
Zone dma32: Available graphics memory: 2097152 KiB
nouveau0: DRM: VRAM: 2048 MiB
nouveau0: DRM: GART: 1048576 MiB
nouveau0: DRM: TMDS table version 2.0
nouveau0: DRM: DCB version 4.0
nouveau0: DRM: DCB outp 00: 01000f02 00020030
nouveau0: DRM: DCB outp 01: 02011f62 00020010
nouveau0: DRM: DCB outp 02: 02022f10 00000000
nouveau0: DRM: DCB conn 00: 00001031
nouveau0: DRM: DCB conn 01: 00002161
nouveau0: DRM: DCB conn 02: 00000200
nouveau0: DRM: MM: using COPY for buffer copies
[drm] Supports vblank timestamp caching Rev 2 (21.10.2013).
[drm] Driver supports precise vblank timestamp query.
nouveaufb0 at nouveau0
[drm] Initialized nouveau 1.3.1 20120801 for nouveau0 on
minor 0
nouveaufb0: framebuffer at 0xfce80c0000, size 1920x1080,
depth 32, stride 7680
nouveau0: autoconfiguration error: error: DRM: core
notifier timeout
nouveau0: autoconfiguration error: error: DRM: base-0:
timeout
no data for est. mode 640x480x67
The most common problem results in the following console message:
nouveau0: autoconfiguration error: error: DRM: core
notifier timeout
This occurs frequently whenever I switch back and forth from a
console session to an X session using ctrl-alt-f<n>. Often there
is a delay before which keystrokes are not recognized for a few
seconds. Most of the time the system recovers, but not always.
In the X session, cursor movement can be "jerky". The cursor
will follow mouse movement for a bit, and then stop/pause before
continuing. The stop/pause usually occurs when the cursor is
ready to enter a region which needs to change the curson icon
(Ie, entering a button needsd cursor to become hand/finger). As
above, the stall only lasts a couple of seconds, after which (in
most cases) normal cursor tracking resumes.
I'm also getting the following errors, although I can't correlate
them to specific human interactions:
nouveau0: autoconfiguration error: error: fifo: fault
00 [READ] at 00000000000be000 engine 04 [BAR1] client
07 [HUB/HOST_CPU] reason 02 [PTE] on channel -1
[007fce7000 unknown]
(The specific values printed vary.)
The final issue occurs when xdm is exiting, either during system
shutdown or whwen I terminate an X session. The following message
gets logged, for channel values of 4 down to 1:
nouveau0: autoconfiguration error: error: user: failed
to idle channel <n> [user]
There is a roughly 11-second timeout associated with these errors,
but the system does recover; the timeout process takes 2+ minutes.
>Fix:
please
>Audit-Trail:
From: Paul Goyette <paul@whooppee.com>
To: gnats-bugs@netbsd.org
Cc:
Subject: Re: kern/58053: Issues on -current with nouveau graphics
Date: Sat, 18 May 2024 10:41:49 -0700 (PDT)
> The most common problem results in the following console message:
>
> nouveau0: autoconfiguration error: error: DRM: core
> notifier timeout
>
> This occurs frequently whenever I switch back and forth from a
> console session to an X session using ctrl-alt-f<n>. Often there
> is a delay before which keystrokes are not recognized for a few
> seconds. Most of the time the system recovers, but not always.
>
> In the X session, cursor movement can be "jerky". The cursor
> will follow mouse movement for a bit, and then stop/pause before
> continuing. The stop/pause usually occurs when the cursor is
> ready to enter a region which needs to change the curson icon
> (Ie, entering a button needsd cursor to become hand/finger). As
> above, the stall only lasts a couple of seconds, after which (in
> most cases) normal cursor tracking resumes.
This seems to happen _every_ time I switch between Ctrl-Alt-F1 and
Ctrl-Alt-F5.
The jerky/stalling cursor movement occurs only in certain apps.
It is consistent when running either gnucash or libreoffice (both
from pkgsrc). Interestingly, firefox seems to be unaffected.
The cursor stalling also generates the timeout messages, although
occassionally a different message is generated:
[ 318189.124414] nouveau0: autoconfiguration error: error:
DRM: base-0: timeout
[ 31564.485889] kqueue_register: cmTC_e3127[8440]: event
type 6 not supported for file type -1/? (error 22)
> nouveau0: autoconfiguration error: error: fifo: fault
> 00 [READ] at 00000000000be000 engine 04 [BAR1] client
> 07 [HUB/HOST_CPU] reason 02 [PTE] on channel -1
> [007fce7000 unknown]
>
> (The specific values printed vary.)
>
> The final issue occurs when xdm is exiting, either during system
> shutdown or whwen I terminate an X session. The following message
> gets logged, for channel values of 4 down to 1:
>
> nouveau0: autoconfiguration error: error: user: failed
> to idle channel <n> [user]
>
> There is a roughly 11-second timeout associated with these errors,
> but the system does recover; the timeout process takes 2+ minutes.
The actual timeout is usually closer to 15 seconds, and most of the
channels fail more than once. (Usually channel 4 fails twice, while
channel 2 can fail five or moore times.)
These errors make using the workstation cumbersome/awkward at best.
+---------------------+--------------------------+----------------------+
| Paul Goyette (.sig) | PGP Key fingerprint: | E-mail addresses: |
| (Retired) | 1B11 1849 721C 56C8 F63A | paul@whooppee.com |
| Software Developer | 6E2E 05FD 15CE 9F2D 5102 | pgoyette@netbsd.org |
| & Network Engineer | | pgoyette99@gmail.com |
+---------------------+--------------------------+----------------------+
>Unformatted:
(Contact us)
$NetBSD: query-full-pr,v 1.47 2022/09/11 19:34:41 kim Exp $
$NetBSD: gnats_config.sh,v 1.9 2014/08/02 14:16:04 spz Exp $
Copyright © 1994-2024
The NetBSD Foundation, Inc. ALL RIGHTS RESERVED.