Kernel module ae4dma caused my system to not boot

Kernel module ae4dma caused my system to not boot

[Jesper Dangaard Brouer]

Hi Basavaraj,

When rebasing my devel kernel to v6.14 (6.14.0-rc3-traits-009-bench+),
my AMD testlab system didn't boot due to the kernel module: ae4dma.
I disabled the kernel module via CONFIG_AMD_AE4DMA so I can boot my
system again.

CPU: AMD EPYC 9684X 96-Core Processor

As a help for you to debug here are the kernel console messages that
seems relevant:

[   13.853689] ae4dma 0000:85:00.1: enabling device (0000 -> 0002)
[   13.861287] genirq: Flags mismatch irq 0. 00000000 (0000:85:00.1) vs. 
00215a00 (timer)
[   13.870123] CPU: 0 UID: 0 PID: 8 Comm: kworker/0:0 Not tainted 
6.14.0-rc3-traits-009-bench+ #32
[   13.870125] Hardware name: Lenovo HR355M-V3-G12/HR355M_V3_HPM, BIOS 
HR355M_V3.G.026 10/13/2023
[   13.870126] Workqueue: events work_for_cpu_fn
[   13.870132] Call Trace:
[   13.870134]  <TASK>
[   13.870136]  dump_stack_lvl+0x64/0x80
[   13.870141]  __setup_irq+0x522/0x6d0
[   13.870145]  request_threaded_irq+0x10d/0x180
[   13.870146]  devm_request_threaded_irq+0x71/0xd0
[   13.870149]  ? __pfx_ae4_core_irq_handler+0x10/0x10 [ae4dma]
[   13.870153]  ae4_core_init+0x8f/0x2f0 [ae4dma]
[   13.870155]  ae4_pci_probe+0x1f3/0x270 [ae4dma]
[   13.870156]  local_pci_probe+0x42/0xa0
[   13.870161]  work_for_cpu_fn+0x17/0x30
[   13.870162]  process_one_work+0x181/0x3a0
[   13.870165]  worker_thread+0x251/0x360
[   13.870166]  ? __pfx_worker_thread+0x10/0x10
[   13.870168]  kthread+0xee/0x240
[   13.870170]  ? finish_task_switch.isra.0+0x88/0x2d0
[   13.870173]  ? __pfx_kthread+0x10/0x10
[   13.870174]  ret_from_fork+0x31/0x50
[   13.870177]  ? __pfx_kthread+0x10/0x10
[   13.870178]  ret_from_fork_asm+0x1a/0x30
[   13.870182]  </TASK>
[   14.010326] ccp 0000:03:00.5: SEV firmware updated from 1.55.32 to 
1.55.37
[   16.565527] ae4dma 0000:85:00.1: probe with driver ae4dma failed with 
error -16
[   16.574958] ae4dma 0000:c1:00.1: enabling device (0000 -> 0002)

Let me know if you need other data,
--Jesper

Re: Kernel module ae4dma caused my system to not boot

[Basavaraj Natikar]

Hi Jesper,

Could you please use below patch series which fixes below issue.

https://lore.kernel.org/all/20250203162511.911946-1-Basavaraj.Natikar@amd.com/


Thanks,
--
Basavaraj


On 2/28/2025 8:05 PM, Jesper Dangaard Brouer wrote:
> Hi Basavaraj,
>
> When rebasing my devel kernel to v6.14 (6.14.0-rc3-traits-009-bench+),
> my AMD testlab system didn't boot due to the kernel module: ae4dma.
> I disabled the kernel module via CONFIG_AMD_AE4DMA so I can boot my
> system again.
>
> CPU: AMD EPYC 9684X 96-Core Processor
>
> As a help for you to debug here are the kernel console messages that
> seems relevant:
>
> [   13.853689] ae4dma 0000:85:00.1: enabling device (0000 -> 0002)
> [   13.861287] genirq: Flags mismatch irq 0. 00000000 (0000:85:00.1) 
> vs. 00215a00 (timer)
> [   13.870123] CPU: 0 UID: 0 PID: 8 Comm: kworker/0:0 Not tainted 
> 6.14.0-rc3-traits-009-bench+ #32
> [   13.870125] Hardware name: Lenovo HR355M-V3-G12/HR355M_V3_HPM, BIOS 
> HR355M_V3.G.026 10/13/2023
> [   13.870126] Workqueue: events work_for_cpu_fn
> [   13.870132] Call Trace:
> [   13.870134]  <TASK>
> [   13.870136]  dump_stack_lvl+0x64/0x80
> [   13.870141]  __setup_irq+0x522/0x6d0
> [   13.870145]  request_threaded_irq+0x10d/0x180
> [   13.870146]  devm_request_threaded_irq+0x71/0xd0
> [   13.870149]  ? __pfx_ae4_core_irq_handler+0x10/0x10 [ae4dma]
> [   13.870153]  ae4_core_init+0x8f/0x2f0 [ae4dma]
> [   13.870155]  ae4_pci_probe+0x1f3/0x270 [ae4dma]
> [   13.870156]  local_pci_probe+0x42/0xa0
> [   13.870161]  work_for_cpu_fn+0x17/0x30
> [   13.870162]  process_one_work+0x181/0x3a0
> [   13.870165]  worker_thread+0x251/0x360
> [   13.870166]  ? __pfx_worker_thread+0x10/0x10
> [   13.870168]  kthread+0xee/0x240
> [   13.870170]  ? finish_task_switch.isra.0+0x88/0x2d0
> [   13.870173]  ? __pfx_kthread+0x10/0x10
> [   13.870174]  ret_from_fork+0x31/0x50
> [   13.870177]  ? __pfx_kthread+0x10/0x10
> [   13.870178]  ret_from_fork_asm+0x1a/0x30
> [   13.870182]  </TASK>
> [   14.010326] ccp 0000:03:00.5: SEV firmware updated from 1.55.32 to 
> 1.55.37
> [   16.565527] ae4dma 0000:85:00.1: probe with driver ae4dma failed 
> with error -16
> [   16.574958] ae4dma 0000:c1:00.1: enabling device (0000 -> 0002)
>
> Let me know if you need other data,
> --Jesper
>