CachyOS / CachyOS/linux-cachyos
[BUG] HP Victus 15-fa1xxx: i915 eDP link training failure after s2idle resume
- Dominant language
- Shell
- Stars
- 4.5k
- Forks
- 160
- Avg merge
- 2d 12h
- Merged PRs (30d)
- 1
Description
### Pre-flight checklist
- [x] I have searched existing issues and this is not a duplicate.
- [x] I have read the [Contributing Guidelines](https://github.com/CachyOS/linux-cachyos/blob/master/CONTRIBUTING.md).
- [x] I have verified the issue is reproducible with the latest available CachyOS kernel.
- [x] I have tried to reproduce the issue on Arch Linux's `linux` kernel.
### Upstream / vanilla kernel check
I have not tested with a vanilla/upstream kernel
### Kernel variant
linux-cachyos (EEVDF, Clang)
### System information (cachyos-bugreport.sh)
```text
❯ sudo cachyos-bugreport.sh
[sudo] password for luimi:
Starting with bugreport
Redacting personal information...
Do you want to upload this log to https://paste.cachyos.org? [Y]es/[N]o: y
Uploading Log
% Total % Received % Xferd Average Speed Time Time Time Current
Dload Upload Total Spent Left Speed
100 292.0k 100 10 100 292.0k 4 134.8k 00:02 00:02 205.7k
https://paste.cachyos.org/p/86599be.log
```
### Manual system information (if cachyos-bugreport.sh is unavailable)
```text
```
### Bug description
I am experiencing an intermittent internal display failure after suspend/resume on an HP Victus 15-fa1xxx.
The system uses s2idle as its only suspend mode. Suspend itself completes, but after resume the internal eDP panel sometimes takes several seconds to return, briefly appears and goes black again, or remains black until I hard power off the laptop.
The problem is intermittent: some suspend/resume cycles work correctly, while others fail.
Normal KDE/Wayland display power-off without actual suspend generally resumes correctly. The failure appears specifically around real s2idle suspend/resume cycles.
The kernel consistently logs i915/eDP errors around the problematic resumes:
i915 0000:00:02.0: [drm] *ERROR*
[CONNECTOR:508:eDP-1][ENCODER:507:DDI A/PHY A][DPRX]
Failed to enable link training
I also see:
[transcoder A] PSR aux error
i915 raw-wakerefs=1 wakelocks=1 on cleanup
WARNING at:
drivers/gpu/drm/i915/intel_runtime_pm.c
intel_runtime_pm_driver_release
and occasionally during boot:
failed to retrieve link info, disabling eDP
Hardware:
- HP Victus 15-fa1xxx
- SKU: 8B3R1UA#ABA
- Intel Alder Lake-P Iris Xe (i915)
- NVIDIA GeForce RTX 4060 Laptop GPU
- nvidia-open 615.71.09
- KDE Plasma / Wayland
- linux-cachyos 7.2.4-3
- suspend mode: [s2idle]
- BIOS F.18
NVIDIA suspend handling appears normal for the current driver:
- PreserveVideoMemoryAllocations: 2
- UseKernelSuspendNotifiers: 1
- Runtime D3: Enabled (fine-grained)
There appears to be a recent upstream Intel i915 patch series addressing this exact type of eDP/DPCD resume failure:
"drm/i915/dp: On DPCD init/caps wake the DPRx"
Patchwork series, rev6:
https://patchwork.freedesktop.org/series/161818/
A v6 test was performed on another HP Victus 15 using Linux 7.2.2. The tester reported multiple successful suspend/resume and lid-close cycles with zero "Failed to enable link training" errors:
https://www.mail-archive.com/intel-gfx@lists.freedesktop.org/msg384512.html
Would it be possible to consider carrying/backporting this patch in linux-cachyos once the maintainers consider the revision sufficiently safe or once it is accepted upstream?
I am happy to test a CachyOS kernel containing the patch and provide additional logs if useful.
Note regarding the pre-flight checklist:
I have NOT actually tested this on Arch Linux's stock `linux` kernel.
The issue form currently requires the checkbox to be selected in order to submit the report.
I am running CachyOS only, and the "Upstream / vanilla kernel check" field above correctly reflects that I have not tested with a vanilla/upstream kernel.
### Steps to reproduce
1. Boot CachyOS using linux-cachyos 7.2.4-3.
2. Log into KDE Plasma Wayland.
3. Allow the laptop to enter suspend, or run:
systemctl suspend
4. Resume the laptop by opening the lid or using an input device.
5. Repeat several times because the failure is intermittent.
6. Inspect the kernel log with:
journalctl -b -k
7. On failed/problematic resumes, i915 reports:
[DPRX] Failed to enable link training
### Expected behavior
The internal eDP display should wake immediately and remain stable after every suspend/resume cycle, without requiring a hard shutdown.
### Actual behavior
Resume is intermittent.
Sometimes the internal display returns normally. Other times it takes several seconds to recover, briefly displays an image and goes black again, or remains black until the laptop is forcibly powered off.
The kernel reports i915 eDP link-training and AUX errors after the affected resumes.
### Logs / stack traces
```text
PM: suspend entry (s2idle)
...
PM: suspend exit
...
kwin_wayland: Applying output configuration failed!
...
i915 0000:00:02.0: [drm] *ERROR*
[CONNECTOR:508:eDP-1][ENCODER:507:DDI A/PHY A][DPRX]
Failed to enable link training
```
### Additional system information
The internal display is driven by Intel i915. The NVIDIA GPU uses nvidia-open.
DPMS/display power-off without a real suspend generally works correctly, which is why I suspect the issue is specifically in the i915/eDP resume path rather than KDE's normal display blanking.
A similar issue exists outside CachyOS as well:
https://github.com/pop-os/pop/issues/3874
Contributor guide
Assessment
This issue has not been assessed yet.