NetBSD Problem Report #58026

From www@netbsd.org  Sun Mar 10 08:49:18 2024
Return-Path: <www@netbsd.org>
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 67C861A9241
	for <gnats-bugs@gnats.NetBSD.org>; Sun, 10 Mar 2024 08:49:18 +0000 (UTC)
Message-Id: <20240310084917.0AB401A9242@mollari.NetBSD.org>
Date: Sun, 10 Mar 2024 08:49:17 +0000 (UTC)
From: mrg@eterna23.net
Reply-To: mrg@eterna23.net
To: gnats-bugs@NetBSD.org
Subject: amdgpu suspend/resume almost works
X-Send-Pr-Version: www-1.0

>Number:         58026
>Category:       kern
>Synopsis:       amdgpu suspend/resume almost works
>Confidential:   no
>Severity:       serious
>Priority:       medium
>Responsible:    kern-bug-people
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Sun Mar 10 08:50:02 +0000 2024
>Last-Modified:  Mon Mar 11 11:05:01 +0000 2024
>Originator:     matthew green
>Release:        -current
>Organization:
>Environment:
amd64, thinkpad A475
>Description:
after suspend/resume cycle where X11 was the active console, the display
shows a blank screen with a cursor in the top-left, and requires switching
the vt away from X and then back to it, where it works fine.  it seems like the display is left in "text" mode at this point..

on half-generation earlier amd HP system, which has radeon for drm not
amdgpu (but are both 4-gen bulldozer APU systems), this works fine and it
resumes into X no problem, so something about amdgpu vs radeon may be the
problem here...
>How-To-Repeat:
run X on thinkpad A475 and suspend/resume.
>Fix:

>Audit-Trail:
From: Taylor R Campbell <riastradh@NetBSD.org>
To: mrg@NetBSD.org
Cc: gnats-bugs@NetBSD.org, netbsd-bugs@NetBSD.org, macallan@NetBSD.org
Subject: Re: kern/58026: amdgpu suspend/resume almost works
Date: Mon, 11 Mar 2024 11:01:01 +0000

 > after suspend/resume cycle where X11 was the active console, the
 > display shows a blank screen with a cursor in the top-left, and
 > requires switching the vt away from X and then back to it, where it
 > works fine.  it seems like the display is left in "text" mode at
 > this point..

 I see this on my Thinkpad T480 with Intel graphics too.  Annoying but
 I never got around to tracking it down.  Probably involves spelunking
 through the same logic as the usl_detachtimeout state machine horror.
 But we should fix it even if that sounds like an unappetizing
 adventure.

Home	PR Database Search