linux-kernel - Re: [v2,1/2] drm/msm/dpu1: don't choke on disabling the writeback connector

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <ZzyqhK-FUwoAcgx1@hovoldconsulting.com>
Date: Tue, 19 Nov 2024 16:11:00 +0100
From: Johan Hovold <johan@...nel.org>
To: Leonard Lausen <leonard@...sen.nl>
Cc: György Kurucz <me@...uczgy.com>,
	Dmitry Baryshkov <dmitry.baryshkov@...aro.org>,
	Rob Clark <robdclark@...il.com>,
	Abhinav Kumar <quic_abhinavk@...cinc.com>,
	Sean Paul <sean@...rly.run>,
	Marijn Suijten <marijn.suijten@...ainline.org>,
	David Airlie <airlied@...il.com>, Daniel Vetter <daniel@...ll.ch>,
	linux-arm-msm@...r.kernel.org, dri-devel@...ts.freedesktop.org,
	freedreno@...ts.freedesktop.org, linux-kernel@...r.kernel.org,
	Jeykumar Sankaran <jsanka@...eaurora.org>, stable@...r.kernel.org,
	Abel Vesa <abel.vesa@...aro.org>
Subject: Re: [v2,1/2] drm/msm/dpu1: don't choke on disabling the writeback
 connector

On Tue, Nov 19, 2024 at 09:33:26AM -0500, Leonard Lausen wrote:

> > I'm seeing the same issue as György on the x1e80100 CRD and Lenovo
> > ThinkPad T14s. Without this patch, the internal display fails to resume
> > properly (switching VT brings it back) and the following errors are
> > logged:
> > 
> > 	[dpu error]connector not connected 3
> > 	[drm:drm_mode_config_helper_resume [drm_kms_helper]] *ERROR* Failed to resume (-22)
> > 
> > I see the same symptoms with Xorg as well as sway.
> 
> The issue of "internal display fails to resume properly (switching VT brings it back)"
> also affects sc7180 platform during some resumes. Do you see the issue consistently
> during every resume?

Yes, it happens on every suspend cycle here.

I didn't notice the issue initially as fbdev does not seem to be
affected, and I've been running with this patch applied to suppress the
resume errors since it was posted.

> > Can we please get this fixed and backported as soon as possible?
> > 
> > Even if there are further issues with some "Night Light" functionality
> > on one machine, keeping this bug as workaround does not seem warranted
> > given that it breaks basic functionality for users.
> 
> I suspect this is not about "further issues with some 'Night Light' functionality
> on one machine", but rather a more fundamental issue or race condition in the qcom
> DRM devices stack, that is exposed when applying this patch. With this patch applied
> DRM device state is lost after resume and setting the state is no longer possible.
> Lots of kernel errors are printed if attempting to set DRM state such as the
> Color Transform Matrix, when running a kernel with this patch applied.
> Back in July 2024 I tested this patch on top of 6.9.8 and next-20240709,
> observing below snippet being logged tens of times:
> 
> [drm:_dpu_rm_check_lm_and_get_connected_blks] [dpu error]failed to get dspp on lm 0
> [drm:_dpu_rm_make_reservation] [dpu error]unable to find appropriate mixers
> [drm:dpu_rm_reserve] [dpu error]failed to reserve hw resources: -119
> 
> Full logs are attached at https://gitlab.freedesktop.org/drm/msm/-/issues/58.

I would not be surprised if there are further issues here, but we can't
just leave things completely broken as they currently are.

> > The x1e80100 is the only platform I have access to with a writeback
> > connector, but this regression potentially affects a whole host of older
> > platforms as well.
> 
> Have you attempted setting CTM or other DRM state when running with this patch?

Nope, I just want basic suspend to work.

Johan