Public bug reported:
All the information is directly from my PC although I am not very
experience in this kind bug reporting and I want to be as helpful as
possible. The information that follows was only organized by chatGPT for
me so maybe whoever reads this can understand it better. If I can help
with anymore info or testing please let me know.
AMDGPU Regression Report — Linux Mint (Ubuntu 24.04 Base)
Summary:
System hard-locks shortly after boot or during normal operation when running
kernel 6.14.0-33-generic.
Issue is fully resolved when downgrading to 6.14.0-32-generic, indicating a
regression in the newer kernel’s AMDGPU module or firmware handling.
System Information:
Distro: Linux Mint 21.3 (Ubuntu 24.04 base)
Working Kernel: 6.14.0-32-generic
Failing Kernel: 6.14.0-33-generic
GPU: AMD Radeon RX 9060 XT (gfx1200, RDNA3)
Driver: amdgpu (in-kernel)
linux-firmware: 20240318.git3b128b60-0ubuntu2.17
Problem Description:
• On kernel 6.14.0-33, the system freezes completely (no mouse, keyboard, or
SSH access) within minutes of login.
• One freeze occurred during boot, several others randomly after boot.
• Rebooting into kernel 6.14.0-32 restores full stability.
• Reinstalling linux-firmware did not resolve the issue.
Relevant Log Excerpts (from journalctl -b -1 on failing kernel):
amdgpu 0000:0e:00.0: amdgpu: Failed to load firmware "amdgpu/gfx1200_mec2.bin"
amdgpu 0000:0e:00.0: amdgpu: [gfxhub] timeout 0x00000010
amdgpu 0000:0e:00.0: amdgpu: Fatal error during GPU init
WARNING: CPU: 8 PID: 202 at amdgpu_irq_put+0x9c/0xb0 [amdgpu]
UBSAN: array-index-out-of-bounds in dml2_core_dcn4_calcs.c
I/O error, dev sda, sector 5992552
watchdog: task blocked for more than 122 seconds
Steps to Reproduce:
1. Boot kernel 6.14.0-33-generic with AMD RX 9060 XT.
2. Log into desktop and wait a few minutes or open GPU-accelerated
applications.
3. System locks up completely (no recovery except power cycle).
4. Boot into 6.14.0-32-generic → no freezes.
Expected vs Actual:
Expected: GPU initializes normally, stable desktop operation.
Actual: GPU firmware load timeout (-110), kernel bug in amdgpu display code,
and full system hang.
Workaround:
Booting kernel 6.14.0-32-generic avoids the problem entirely.
Notes:
The issue persists even after reinstalling the linux-firmware package and
regenerating initramfs.
Appears to be a regression in amdgpu initialization code introduced between
-32 and -33.
** Affects: linux (Ubuntu)
Importance: Undecided
Status: New
** Attachment added: "responces of sudo dmesg > dmesg_6.14.0-33.txt sudo
journalctl -b -1 > journal_6.14.0-33.txt uname -a > uname.txt"
https://bugs.launchpad.net/bugs/2126854/+attachment/5915190/+files/logs.zip
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux in Ubuntu.
https://bugs.launchpad.net/bugs/2126854
Title:
AMDGPU firmware load timeout and system freeze on kernel 6.14.0-33 (RX
9060 XT, RDNA3)
Status in linux package in Ubuntu:
New
Bug description:
All the information is directly from my PC although I am not very
experience in this kind bug reporting and I want to be as helpful as
possible. The information that follows was only organized by chatGPT
for me so maybe whoever reads this can understand it better. If I can
help with anymore info or testing please let me know.
AMDGPU Regression Report — Linux Mint (Ubuntu 24.04 Base)
Summary:
System hard-locks shortly after boot or during normal operation when running
kernel 6.14.0-33-generic.
Issue is fully resolved when downgrading to 6.14.0-32-generic, indicating a
regression in the newer kernel’s AMDGPU module or firmware handling.
System Information:
Distro: Linux Mint 21.3 (Ubuntu 24.04 base)
Working Kernel: 6.14.0-32-generic
Failing Kernel: 6.14.0-33-generic
GPU: AMD Radeon RX 9060 XT (gfx1200, RDNA3)
Driver: amdgpu (in-kernel)
linux-firmware: 20240318.git3b128b60-0ubuntu2.17
Problem Description:
• On kernel 6.14.0-33, the system freezes completely (no mouse, keyboard,
or SSH access) within minutes of login.
• One freeze occurred during boot, several others randomly after boot.
• Rebooting into kernel 6.14.0-32 restores full stability.
• Reinstalling linux-firmware did not resolve the issue.
Relevant Log Excerpts (from journalctl -b -1 on failing kernel):
amdgpu 0000:0e:00.0: amdgpu: Failed to load firmware
"amdgpu/gfx1200_mec2.bin"
amdgpu 0000:0e:00.0: amdgpu: [gfxhub] timeout 0x00000010
amdgpu 0000:0e:00.0: amdgpu: Fatal error during GPU init
WARNING: CPU: 8 PID: 202 at amdgpu_irq_put+0x9c/0xb0 [amdgpu]
UBSAN: array-index-out-of-bounds in dml2_core_dcn4_calcs.c
I/O error, dev sda, sector 5992552
watchdog: task blocked for more than 122 seconds
Steps to Reproduce:
1. Boot kernel 6.14.0-33-generic with AMD RX 9060 XT.
2. Log into desktop and wait a few minutes or open GPU-accelerated
applications.
3. System locks up completely (no recovery except power cycle).
4. Boot into 6.14.0-32-generic → no freezes.
Expected vs Actual:
Expected: GPU initializes normally, stable desktop operation.
Actual: GPU firmware load timeout (-110), kernel bug in amdgpu display
code, and full system hang.
Workaround:
Booting kernel 6.14.0-32-generic avoids the problem entirely.
Notes:
The issue persists even after reinstalling the linux-firmware package and
regenerating initramfs.
Appears to be a regression in amdgpu initialization code introduced between
-32 and -33.
To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/2126854/+subscriptions
--
Mailing list: https://launchpad.net/~kernel-packages
Post to : [email protected]
Unsubscribe : https://launchpad.net/~kernel-packages
More help : https://help.launchpad.net/ListHelp
