Bug 2536023

Summary:	amd-gpu-firmware 20260910: yellow_carp DMUB 0x0400004A breaks resume on Rembrandt (Radeon 680M) – "Error getting DMUB FW meta info"
Product:	[Fedora] Fedora	Reporter:	jeremie.willemin
Component:	linux-firmware	Assignee:	David Woodhouse <dwmw2>
Status:	CLOSED DUPLICATE	QA Contact:	Fedora Extras Quality Assurance <extras-qa>
Severity:	high	Docs Contact:
Priority:	unspecified
Version:	44	CC:	dwmw2, jforbes, jwboyer, kernel-maint, pbrobinson
Target Milestone:	---	Keywords:	Bugfix, Upgrades
Target Release:	---
Hardware:	x86_64
OS:	Linux
Whiteboard:
Fixed In Version:		Doc Type:	---
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	2026-09-17 13:53:34 UTC	Type:	---
Regression:	---	Mount Type:	---
Documentation:	---	CRM:
Verified Versions:		Category:	---
oVirt Team:	---	RHEL 7.3 requirements from Atomic Host:
Cloudforms Team:	---	Target Upstream Version:
Embargoed:

Description jeremie.willemin 2026-09-17 13:51:14 UTC

Hi,

Honestly I used AI to solve my issue since I am not an expert at all. Seems like it did solve it and found the rootcause. Here are the findings.

Description of problem:
  After updating amd-gpu-firmware 20260810 -> 20260910, the screen stays black
  after resume from suspend (S3) on an AMD Rembrandt APU (Radeon 680M).
  The system itself wakes up, but the display never comes back; a hard
  power-off is needed.

  Hardware: GMKtec NucBox M6 (BIOS 103), Radeon 680M (amdgpu), KDE Plasma
  Wayland
  Kernels affected: 7.2.5-200.fc44, 7.2.5-100.fc43

Reproducible: Always

Steps to Reproduce:
1.Install amd-gpu-firmware-20260910, regenerate initramfs
2.Boot, suspend (deep/S3)
Actual Results:
Black screen. On every boot with this firmware:
    [drm] *ERROR* Error getting DMUB FW meta info: 4
    [drm] *ERROR* dc_dmub_srv_log_diagnostic_data: DMCUB error (40-800x per
  boot)
  After resume:
    amdgpu: [CRTC:376:crtc-0] flip_done timed out
    kwin_wayland: Pageflip timed out! This is a bug in the amdgpu kernel driver

Expected Results:
Display resumes normally.

Additional Information:
- Worked for 6 months (300+ S3 resumes, kernels 6.18 -> 7.2.4) with DMUB
  0x45..0x4C.
  - 7.2.4 initramfs built with the 20260810 firmware (DMUB 0x4C): no errors.
  - Downgrading to amd-gpu-firmware-20260309 (DMUB 0x47) + dracut -f
  --regenerate-all
    fixes it: "DMUB hardware initialized: version=0x04000047", no errors,
  resume works.
  - The new DMUB 0x4A has a lower version number than the 0x4C it replaced,
    so the wrong blob may have been shipped.

Comment 1 Peter Robinson 2026-09-17 13:53:34 UTC

Please do a basic check for duplicates.

*** This bug has been marked as a duplicate of bug 2532947 ***