[Bug 2144712] Re: On Dell system, the internal OLED display drops to a visibly low FPS after suspend/resume

** Tags removed: verification-needed-noble-linux-oem-6.17
** Tags added: verification-done-noble-linux-oem-6.17

--
You received this bug notification because you are subscribed to linux
in Ubuntu.
Matching subscriptions: Bgg, Bmail, Nb
https://bugs.launchpad.net/bugs/2144712

Title:
On Dell system, the internal OLED display drops to a visibly low FPS
after suspend/resume

Status in HWE Next:
New
Status in linux package in Ubuntu:
In Progress
Status in linux-oem-6.17 package in Ubuntu:
Invalid
Status in linux source package in Noble:
Won't Fix
Status in linux-oem-6.17 source package in Noble:
Fix Committed
Status in linux source package in Questing:
In Progress
Status in linux-oem-6.17 source package in Questing:
Invalid
Status in linux source package in Resolute:
In Progress
Status in linux-oem-6.17 source package in Resolute:
Invalid

Bug description:
[Impact]
On Dell systems (CID: 202511-38076), the internal OLED display drops to a
visibly lower refresh rate after suspend/resume. The state persists until a
cold reboot. Applications like glmark2 still report rendering at 120Hz but
the panel visually operates at a much lower frame rate.

Steps to reproduce:
1. Unplug all accessories
2. Boot into OS
3. Drag a window around the desktop and observe the FPS (smooth 120Hz)
4. Suspend then resume
5. Drag a window around the desktop — FPS is now visibly and permanently lower
Failure rate: 100% on affected hardware
Affected test case: com.canonical.certification::suspend/display_after_suspend

Root cause: The i915 driver initializes Panel Replay (an OLED power-saving
feature) via DPCD negotiation during resume. On this specific panel, Panel
Replay causes the display to be locked at a degraded frame rate after resume.
The panel does not correctly handle Panel Replay activation in this
configuration, resulting in permanently reduced FPS until a cold reboot.

[Fix]
Add a new quirk QUIRK_DISABLE_PANEL_REPLAY to the Intel i915 display driver
that prevents Panel Replay DPCD initialization for problematic panel setups.
Apply this quirk for Dell systems with device ID 0xb080, subsystem vendor
0x1028, subsystem device 0x0db9, and sink OUI 0x00:0x22:0xb9.

When the quirk is present, _panel_replay_init_dpcd() returns early before
reading Panel Replay capability from the panel's DPCD registers, effectively
disabling the Panel Replay feature for this hardware combination.

The patch is currently under review in the Intel DRM/i915 mailing list:
https://patchwork.freedesktop.org/series/163265/

Upstream issue tracker:
https://gitlab.freedesktop.org/drm/xe/kernel/-/issues/7521

[Test Plan]
On a Dell system (CID: 202511-38076) with Intel integrated graphics:

1. Suspend and resume the system:
$ sudo rtcwake -m mem -s 20
2. After resume, drag a window around the desktop and observe the frame rate.

Without the patch: The display FPS drops permanently after resume and stays
degraded until cold reboot.
With the patch: The display maintains normal 120Hz refresh rate after
suspend/resume.

[Where problems could occur]
It may break Panel Replay functionality on Intel i915 display configurations.

The quirk check is inserted early in _panel_replay_init_dpcd(), before any
DPCD reads occur. If the quirk match logic incorrectly identifies a panel
(e.g., due to incorrect OUI or PCI subsystem ID matching), Panel Replay will
be silently disabled on panels where it would otherwise work correctly. This
would cause affected systems to lose the power-saving benefits of Panel Replay
without any visible error — users would see slightly higher power consumption
on OLED panels.

Additionally, if the quirk infrastructure itself has a bug (e.g., the
intel_has_dpcd_quirk() function returning unexpected values), it could either
fail to disable Panel Replay on affected hardware (regression still present)
or disable it too broadly on unaffected hardware.

[Other Info]
The patch is under review in the Intel DRM mailing list (patchwork series
163265) and has not yet landed in the mainline kernel.

To manage notifications about this bug go to:
https://bugs.launchpad.net/hwe-next/+bug/2144712/+subscriptions