[BUG] Hyprland 0.57.0 rendering regression on Intel HD Graphics 4000 (Ivy Bridge / i915)

[MaTziap]

Hyprland version

0.57.0

Is this a regression?

Yes

What's the bug?

System Information

• GPU: Intel HD Graphics 4000
• GPU Architecture: Ivy Bridge
• Kernel Driver: i915
• Laptop: Dell Inspiron 3521
• Display: LVDS-1, 1366x768 @ 60.04Hz
• Kernel: Linux 7.2.11-arch1-1
• Hyprland version: 0.57.0
• Distribution: Arch-based Linux (Ryoku)
• Mesa version: [Please confirm]

Description

There appears to be a rendering regression in Hyprland 0.57.0 affecting older Intel integrated GPUs, particularly Ivy Bridge (Intel HD Graphics 4000).

The issue is related to how Hyprland handles rendering and DMA-BUF formats on these GPUs.

After upgrading from Hyprland 0.56.2 to 0.57.0, the desktop may become partially or completely black. Wallpaper, desktop elements, and certain application windows may fail to render correctly.

This is particularly concerning for older Intel hardware that relies on the i915 driver.

Expected Behavior

Hyprland should render the desktop, wallpaper, panels, and application windows correctly on Intel HD Graphics 4000, just as previous versions did.

Actual Behavior

• Desktop background and wallpaper may not render.
• Some desktop components may appear black or remain invisible.
• Certain GUI applications may fail to display correctly.
• The issue appears to be related to rendering and DMA-BUF format handling rather than configuration errors.

Related Reports

This issue appears to be related to other reports involving Hyprland 0.57.0 and older Intel GPUs:

• Hyprland Discussion Hyprland 0.57.0 regression: EGL applications fail with "No EGLConfigs returned" on Intel HD 4600 + NVIDIA GTX 960M #16702: EGL applications failing on older Intel graphics.
• Reports involving Ivy Bridge / Intel HD Graphics 4000 and missing alpha-capable DMA-BUF formats.

A user with the same Intel HD Graphics 4000 architecture reported that changing the following option resolves the EGL rendering issue:

quirks {
skip_non_kms_dmabuf_formats = false
}

This suggests that the default behavior introduced in 0.57.0 may cause compatibility problems with GPUs whose scanout formats do not support alpha.

Questions

1. Is the change to "skip_non_kms_dmabuf_formats" in Hyprland 0.57.0 responsible for this regression?
2. Could the default behavior be adjusted to preserve compatibility with older Intel GPUs?
3. Is there a recommended workaround for Ivy Bridge users until a permanent fix is available?
4. Could this regression be addressed in a future patch release?

Additional Information

Intel HD Graphics 4000 is still used in many older laptops and relies on the i915 driver.

Maintaining compatibility with these GPUs would be appreciated.

Thank you for your work on Hyprland.

Reproduction steps

1. Run Hyprland 0.56.2 on an Intel HD Graphics 4000 (Ivy Bridge) GPU using the i915 driver.
2. Upgrade Hyprland to version 0.57.0.
3. Restart the system and log into the Hyprland session.
4. Check whether the wallpaper, desktop components, and application windows render correctly.
5. Compare the rendering behavior with version 0.56.2.

The issue appears to be related to DMA-BUF format handling on older Intel GPUs.

Attachments

No response

I acknowledge that:

• I have searched all (open and closed) discussions and issues and my problem is not there.

• I have read the Code of Conduct and agree to follow it.

• I wrote this request with my own hands, my own keyboard, and in my own words.

[vaxerski]

slop slop slop you even linked your own dupe which is FIXED and STAGED

[MaTziap]

slop slop slop you even linked your own dupe which is FIXED and STAGED

If you have nothing useful to say, shut the fuck up. If it's a duplicate, explain why like an adult instead of acting like an arrogant, obnoxious asshole. Being rude doesn't make you smarter.

[wabarTM]

Being rude doesn't make you smarter
@
looks inside
@
shut the fuck up
@
like an arrogant, obnoxious asshole
@
:wires:

[erstarr]

/strike-author

AI interaction with dev facing channels. Copy pasting shitslop doesn't make it not-shitslop

[MaTziap]

😂😂

…

On Sun, Nov 1, 2026, 22:10 Wabar ***@***.***> wrote: Being rude doesn't make you smarter @ looks inside @ shut the fuck up @ like an arrogant, obnoxious asshole @ :wires: — Reply to this email directly, view it on GitHub <#16787?email_source=notifications&email_token=BAMFFRORBAOCUXRPTHNE45T5W57RFA5CNFSNUABIM5UWIORPF5TWS5BNNB2WEL2ENFZWG5LTONUW63SDN5WW2ZLOOQXTCOJTGM3DKOBTUZZGKYLTN5XKMYLVORUG64VFMV3GK3TUVRTG633UMVZF6Y3MNFRWW#discussioncomment-19336583>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/BAMFFRK75MLVLHCNTEE7UJT5W57RFAVCNFSNUABIKJSXA33TNF2G64TZHM2DOMBXGMYDMNBYHNCGS43DOVZXG2LPNY5TCMJSGAYTINBVUF3AE> . You are receiving this because you authored the thread.Message ID: ***@***.***>

[github-actions[bot]]

⚠️ Strike 1/3 issued to @MaTziap by @erstarr.