Bug 214901
| Summary: | amdgpu freezes HP laptop at start up | ||
|---|---|---|---|
| Product: | Drivers | Reporter: | spasswolf |
| Component: | Video(DRI - non Intel) | Assignee: | drivers_video-dri |
| Status: | RESOLVED DUPLICATE | ||
| Severity: | normal | CC: | towo |
| Priority: | P1 | ||
| Hardware: | All | ||
| OS: | Linux | ||
| Kernel Version: | 5.15.0 | Subsystem: | |
| Regression: | Yes | Bisected commit-id: | |
|
Description
spasswolf
2021-11-01 00:02:39 UTC
G 00:01.0 VGA compatible controller [0300]: Advanced Micro Devices, Inc. [AMD/ATI] Wani [Radeon R5/R6/R7 Graphics] [1002:9874] (rev ca) (prog-if 00 [VGA controller]) DeviceName: ATI EG BROADWAY Subsystem: Hewlett-Packard Company Wani [Radeon R5/R6/R7 Graphics] [103c:8332] Flags: bus master, fast devsel, latency 0, IRQ 37 Memory at e0000000 (64-bit, prefetchable) [size=256M] Memory at f0800000 (64-bit, prefetchable) [size=8M] I/O ports at 4000 [size=256] Memory at f0400000 (32-bit, non-prefetchable) [size=256K] Expansion ROM at 000c0000 [disabled] [size=128K] Capabilities: <access denied> Kernel driver in use: amdgpu Kernel modules: amdgpu There is another error message just before the Oops: Nov 1 00:22:49 bart kernel: [ 2.137397] amdgpu 0000:00:01.0: amdgpu: amdgpu_device_ip_init failed Nov 1 00:22:49 bart kernel: [ 2.137402] amdgpu 0000:00:01.0: amdgpu: Fatal error during GPU init Nov 1 00:22:49 bart kernel: [ 2.137406] amdgpu 0000:00:01.0: amdgpu: amdgpu: finishing device. Nov 1 00:22:49 bart kernel: [ 2.139639] BUG: kernel NULL pointer dereference, address: 00000000000001db Actually the above message is not complete: Nov 1 00:22:49 bart kernel: [ 2.136382] kfd kfd: amdgpu: Allocated 3969056 bytes on gart Nov 1 00:22:49 bart kernel: [ 2.136462] kfd kfd: amdgpu: error getting iommu info. is the iommu enabled? Nov 1 00:22:49 bart kernel: [ 2.136470] kfd kfd: amdgpu: Error initializing iommuv2 Nov 1 00:22:49 bart kernel: [ 2.137386] kfd kfd: amdgpu: device 1002:9874 NOT added due to errors Nov 1 00:22:49 bart kernel: [ 2.137393] kfd kfd: amdgpu: Failed to resume IOMMU for device 1002:9874 Nov 1 00:22:49 bart kernel: [ 2.137397] amdgpu 0000:00:01.0: amdgpu: amdgpu_device_ip_init failed Nov 1 00:22:49 bart kernel: [ 2.137402] amdgpu 0000:00:01.0: amdgpu: Fatal error during GPU init Nov 1 00:22:49 bart kernel: [ 2.137406] amdgpu 0000:00:01.0: amdgpu: amdgpu: finishing device. Nov 1 00:22:49 bart kernel: [ 2.139639] BUG: kernel NULL pointer dereference, address: 00000000000001db The messages from kfd have been there with older kernels, too, but were not fatal. They are caused by the HP Laptop 15-bw0xx/8332, not having a iommu or its BIOS not properly initializing it. But linux-5.15 has added the following lines to the amdgpu_device_ip_init: r = amdgpu_amdkfd_resume_iommu(adev); if (r) goto init_failed; which make causes the amdgpu_device_ip_init function to fail when kfd init fails. As a workaround one could remove these. A BIOS update could perhaps also solve the problem but seems to require a Windows running on the Laptop (which was actually sold without Windows) Just confirmed that removing the 3 lines
r = amdgpu_amdkfd_resume_iommu(adev);
if (r)
goto init_failed;
can be used as a workaround. Removing only the if (r) check is not enough,
just calling amdgpu_amdkfd_resume_iommu(adev) leads to freezing.
This commit leads to a freeze when starting https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/drivers/gpu/drm/amd/amdgpu?id=714d9e4574d54596973ee3b0624ee4a16264d700 After reverting it the kernel 5.15 boots normally Looks like the same problem as i stated here: https://bugzilla.kernel.org/show_bug.cgi?id=214859 *** This bug has been marked as a duplicate of bug 214859 *** |