[REGRESSION] external monitor+Dell dock in 6.8

[REGRESSION] external monitor+Dell dock in 6.8

[Andrei Gaponenko]

Hello,

I noticed a regression with the mailine kernel pre-compiled by EPEL.
I have just tried linux-6.9-rc1.tar.gz from kernel.org, and it still
misbehaves.

The default setup: a laptop is connected to a dock, Dell WD22TB4, via
a USB-C cable.  The dock is connected to an external monitor via a
Display Port cable.  With a "good" kernel everything works.  With a
"broken" kernel, the external monitor is still correctly identified by
the system, and is shown as enabled in plasma systemsettings. The
system also behaves like the monitor is working, for example, one can
move the mouse pointer off the laptop screen.  However the external
monitor screen stays black, and it eventually goes to sleep.

Everything worked with EPEL mainline kernels up to and including
kernel-ml-6.7.9-1.el9.elrepo.x86_64

The breakage is observed in

kernel-ml-6.8.1-1.el9.elrepo.x86_64
kernel-ml-6.8.2-1.el9.elrepo.x86_64
linux-6.9-rc1.tar.gz from kernel.org (with olddefconfig)

Other tests: using an HDMI cable instead of the Display Port cable
between the monitor and the dock does not change things, black screen
with the newer kernels.

Using a small HDMI-to-USB-C adapter instead of the dock results in a
working system, even with the newer kernels.  So the breakage appears
to be specific to the Dell WD22TB4 dock.

Operating System: AlmaLinux 9.3 (Shamrock Pampas Cat)

uname -mi: x86_64 x86_64

Laptop: Dell Precision 5470/02RK6V

lsusb |grep dock
Bus 003 Device 007: ID 413c:b06e Dell Computer Corp. Dell dock
Bus 003 Device 008: ID 413c:b06f Dell Computer Corp. Dell dock
Bus 003 Device 006: ID 0bda:5413 Realtek Semiconductor Corp. Dell dock
Bus 003 Device 005: ID 0bda:5487 Realtek Semiconductor Corp. Dell dock
Bus 002 Device 004: ID 0bda:0413 Realtek Semiconductor Corp. Dell dock
Bus 002 Device 003: ID 0bda:0487 Realtek Semiconductor Corp. Dell dock

dmesg and kernel config are attached to 
https://bugzilla.kernel.org/show_bug.cgi?id=218663

#regzbot introduced: v6.7.9..v6.8.1

Andrei

Re: [REGRESSION] external monitor+Dell dock in 6.8

[Linux regression tracking (Thorsten Leemhuis)]

Hi! Thx for reporting your problem!

On 31.03.24 07:59, Andrei Gaponenko wrote:
> 
> I noticed a regression with the mailine kernel pre-compiled by EPEL.
>
> I have just tried linux-6.9-rc1.tar.gz from kernel.org, and it still
> misbehaves.
> 
> The default setup: a laptop is connected to a dock, Dell WD22TB4, via
> a USB-C cable.  The dock is connected to an external monitor via a
> Display Port cable.  With a "good" kernel everything works.  With a
> "broken" kernel, the external monitor is still correctly identified by
> the system, and is shown as enabled in plasma systemsettings. The
> system also behaves like the monitor is working, for example, one can
> move the mouse pointer off the laptop screen.  However the external
> monitor screen stays black, and it eventually goes to sleep.
> 
> Everything worked with EPEL mainline kernels up to and including
> kernel-ml-6.7.9-1.el9.elrepo.x86_64
> 
> The breakage is observed in
> 
> kernel-ml-6.8.1-1.el9.elrepo.x86_64
> kernel-ml-6.8.2-1.el9.elrepo.x86_64
> linux-6.9-rc1.tar.gz from kernel.org (with olddefconfig)
> 
> Other tests: using an HDMI cable instead of the Display Port cable
> between the monitor and the dock does not change things, black screen
> with the newer kernels.
> 
> Using a small HDMI-to-USB-C adapter instead of the dock results in a
> working system, even with the newer kernels.  So the breakage appears
> to be specific to the Dell WD22TB4 dock.

Does you laptop offer a HDMI or DP connector? Have you tried if that
that works any better? If it does not, then the DRM developers might be
willing to look into this.

Anyway: could you try to bisect the problem as described in
https://docs.kernel.org/admin-guide/verify-bugs-and-bisect-regressions.html

I strongly suspect that this will be needed to get this resolved: this
could be caused by various subsystems, so no developer might be willing
to look into this without a bisection result. :-/

Ciao, Thorsten

> Operating System: AlmaLinux 9.3 (Shamrock Pampas Cat)
> 
> uname -mi: x86_64 x86_64
> 
> Laptop: Dell Precision 5470/02RK6V
> 
> lsusb |grep dock
> Bus 003 Device 007: ID 413c:b06e Dell Computer Corp. Dell dock
> Bus 003 Device 008: ID 413c:b06f Dell Computer Corp. Dell dock
> Bus 003 Device 006: ID 0bda:5413 Realtek Semiconductor Corp. Dell dock
> Bus 003 Device 005: ID 0bda:5487 Realtek Semiconductor Corp. Dell dock
> Bus 002 Device 004: ID 0bda:0413 Realtek Semiconductor Corp. Dell dock
> Bus 002 Device 003: ID 0bda:0487 Realtek Semiconductor Corp. Dell dock
> 
> dmesg and kernel config are attached to 
> https://bugzilla.kernel.org/show_bug.cgi?id=218663
> 
> #regzbot introduced: v6.7.9..v6.8.1
> 
> Andrei
> 
> 

Re: [REGRESSION] external monitor+Dell dock in 6.8

[Andrei Gaponenko]

Hi Thorsten,

On Sun, 31 Mar 2024, Linux regression tracking (Thorsten Leemhuis) wrote:

> Does you laptop offer a HDMI or DP connector? Have you tried if that
> that works any better? If it does not, then the DRM developers might be
> willing to look into this.

There are only USB-C port on this machine.  However the external
monitor still works if I use a small (passive?) HDMI-to-USB-C adapter
instead of the dock.

> Anyway: could you try to bisect the problem as described in
> https://docs.kernel.org/admin-guide/verify-bugs-and-bisect-regressions.html

Will do.

Thank you for your reply!
Andrei

Re: [REGRESSION] external monitor+Dell dock in 6.8

[Linux regression tracking (Thorsten Leemhuis)]

On 31.03.24 09:15, Andrei Gaponenko wrote:
> 
> On Sun, 31 Mar 2024, Linux regression tracking (Thorsten Leemhuis) wrote:
> 
>> Does you laptop offer a HDMI or DP connector? Have you tried if that
>> that works any better? If it does not, then the DRM developers might be
>> willing to look into this.
> 
> There are only USB-C port on this machine.  However the external
> monitor still works if I use a small (passive?) HDMI-to-USB-C adapter
> instead of the dock.

Yeah, you had mentioned that.

>> Anyway: could you try to bisect the problem as described in
>> https://docs.kernel.org/admin-guide/verify-bugs-and-bisect-regressions.html
> Will do.
> 
> Thank you for your reply!

np, but stupid me forgot something (sorry): I had opened your dmesg
file, but not looked into it. :-/

I see that it contains a warning from nouveau that might be hinting at
the problem. Then I'd say: go and file a ticket here:
https://gitlab.freedesktop.org/drm/nouveau/-/issues

Please afterwards drop a link to the ticket here please.

But a bisection might still be needed; if you do not get a reply from a
developer withing two or three days, you definitely want to perform one.

Ciao, Thorsten (wearing his 'the Linux kernel's regression tracker' hat)
--
Everything you wanna know about Linux kernel regression tracking:
https://linux-regtracking.leemhuis.info/about/#tldr
If I did something stupid, please tell me, as explained on that page.

#regzbot duplicate: https://bugzilla.kernel.org/show_bug.cgi?id=218663
#regzbot title: drm: nouveau: external monitor on Dell dock broke

Re: [REGRESSION] external monitor+Dell dock in 6.8

[Andrei Gaponenko]

Hi Thorsten,

On Sun, 31 Mar 2024, Linux regression tracking (Thorsten Leemhuis) wrote:

> I see that it contains a warning from nouveau that might be hinting at
> the problem. Then I'd say: go and file a ticket here:
> https://gitlab.freedesktop.org/drm/nouveau/-/issues

Working kernel versions issue a similar warning; I am not sure it it related to the regression.
Perhaps I should file a report for it, but I want to deal with one issue at a time.

git bisect pointed to a range of drm/i915 commits, so I filed an updated regression report at
https://gitlab.freedesktop.org/drm/intel/-/issues/10637

Best regards,
Andrei

Re: [REGRESSION] external monitor+Dell dock in 6.8

[Linux regression tracking (Thorsten Leemhuis)]

[Adding a few folks and list while dropping the stable list, as this is
unrelated to it]

On 31.03.24 07:59, Andrei Gaponenko wrote:
> 
> I noticed a regression with the mailine kernel pre-compiled by EPEL.
> I have just tried linux-6.9-rc1.tar.gz from kernel.org, and it still
> misbehaves.
> 
> The default setup: a laptop is connected to a dock, Dell WD22TB4, via
> a USB-C cable.  The dock is connected to an external monitor via a
> Display Port cable.  With a "good" kernel everything works.  With a
> "broken" kernel, the external monitor is still correctly identified by
> the system, and is shown as enabled in plasma systemsettings. The
> system also behaves like the monitor is working, for example, one can
> move the mouse pointer off the laptop screen.  However the external
> monitor screen stays black, and it eventually goes to sleep.

Just a quick heads up to ensure people are aware of it:

Imre Deak, turns out this is caused by a patch of yours: 55eaef16417448
("drm/i915/dp_mst: Handle the Synaptics HBlank expansion quirk"). Andrei
Gaponenko meanwhile filed a ticket about it here:

https://gitlab.freedesktop.org/drm/intel/-/issues/10637

Ciao, Thorsten

> Everything worked with EPEL mainline kernels up to and including
> kernel-ml-6.7.9-1.el9.elrepo.x86_64
> 
> The breakage is observed in
> 
> kernel-ml-6.8.1-1.el9.elrepo.x86_64
> kernel-ml-6.8.2-1.el9.elrepo.x86_64
> linux-6.9-rc1.tar.gz from kernel.org (with olddefconfig)
> 
> Other tests: using an HDMI cable instead of the Display Port cable
> between the monitor and the dock does not change things, black screen
> with the newer kernels.
> 
> Using a small HDMI-to-USB-C adapter instead of the dock results in a
> working system, even with the newer kernels.  So the breakage appears
> to be specific to the Dell WD22TB4 dock.
> 
> Operating System: AlmaLinux 9.3 (Shamrock Pampas Cat)
> 
> uname -mi: x86_64 x86_64
> 
> Laptop: Dell Precision 5470/02RK6V
> 
> lsusb |grep dock
> Bus 003 Device 007: ID 413c:b06e Dell Computer Corp. Dell dock
> Bus 003 Device 008: ID 413c:b06f Dell Computer Corp. Dell dock
> Bus 003 Device 006: ID 0bda:5413 Realtek Semiconductor Corp. Dell dock
> Bus 003 Device 005: ID 0bda:5487 Realtek Semiconductor Corp. Dell dock
> Bus 002 Device 004: ID 0bda:0413 Realtek Semiconductor Corp. Dell dock
> Bus 002 Device 003: ID 0bda:0487 Realtek Semiconductor Corp. Dell dock
> 
> dmesg and kernel config are attached to 
> https://bugzilla.kernel.org/show_bug.cgi?id=218663
> 
> #regzbot introduced: v6.7.9..v6.8.1

P.S.:

#regzbot duplicate: https://bugzilla.kernel.org/show_bug.cgi?id=218663
#regzbot duplicate: https://gitlab.freedesktop.org/drm/intel/-/issues/10637
#regzbot title: drm/i915/dp_mst: external monitor on Dell dock broke