On Tue 12/Nov/2024 10:09:31 +0100 Kilian eldre8 Hart wrote:
> Hi!
>
> Le 2024/11/12 à 03:31, Ludovic Bellière via Dng écrit:
>>
>> On Mon, 11 Nov 2024, Alessandro Vesely via Dng wrote:
>>
>> > Do you think something went wrong much earlier than the hanging point?
>>
>> Maybe, you didn't post the whole log so all I can do is trust your investigation
>> skills.
>>
> The traceback is almost sufficient as you already located where the issue is,
> on the radeon part, r100.


Actually, no. That trace belonged to the final, successful boot. It is certainly a problem, but not fatal. I found a similar trace in older boots, with the previous kernel.

The failed boots hanged right after setting up the network, which is done by a custom script.


> One other log is the /var/log/Xorg.0.log that would contain interesting hints.
> And lsmod to see which lkm is loaded although we can deduce those.


"lsmod| grep lkm" prints nothing.


> Just before r100_irq_set, we see radeon_enable_vblank_kms and drm,
> I tried to grep on all Kconfig about vblank, or radeon or drm inside my 5.15.x
> sources but got no hints; you should try on your kernel sources.


r100_irq_set is only called by r100_startup, after calling radeon_irq_kms_init. It should set up interrupts for the video card. Indeed, the video card capabilities show no MSI interrupts:

# lspci -vvv -s 01:03.0
01:03.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] ES1000 (rev 02) (prog-if 00 [VGA controller])
    Subsystem: Hewlett-Packard Company ProLiant DL360 G5
    Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping+ SERR+ FastB2B- DisINTx-
    Status: Cap+ 66MHz- UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
    Latency: 64 (2000ns min), Cache Line Size: 64 bytes
    Interrupt: pin A routed to IRQ 23
    Region 0: Memory at f0000000 (32-bit, prefetchable) [size=128M]
    Region 1: I/O ports at 3000 [size=256]
    Region 2: Memory at f9ff0000 (32-bit, non-prefetchable) [size=64K]
    Expansion ROM at 000c0000 [virtual] [disabled] [size=128K]
    Capabilities: [50] Power Management version 2
        Flags: PMEClk- DSI- D1+ D2+ AuxCurrent=0mA PME(D0-,D1-,D2-,D3hot-,D3cold-)
        Status: D0 NoSoftRst- PME-Enable- DSel=0 DScale=0 PME-
    Kernel driver in use: radeon
    Kernel modules: radeonfb, radeon, amdgpu

The device, however, works well. Firefox prints this:

[GFX1-]: glxtest: X error, error_code=10, request_code=151, minor_code=5

But I only use console and text editor on this server.


Best
Ale
--