system boots extremely slow with AMD R9 GPU

Xen list at xenhideout.nl
Fri Mar 10 11:06:12 UTC 2017


Xen schreef op 10-03-2017 8:55:

> that my Linux system boots very slowly when I have inserted a AMD R9
> 380 graphics card.

Lol I don't know what changed in between.

I activated a bunch of BIOS settings for other reasons and now suddenly 
the system boots quickly.

This is the difference:

[    3.444927] Console: switching to colour frame buffer device 180x56
[    3.451101] amdgpu 0000:01:00.0: fb0: amdgpudrmfb frame buffer device

versus

[  139.328143] Console: switching to colour frame buffer device 180x56
[  139.333613] amdgpu 0000:01:00.0: fb0: amdgpudrmfb frame buffer device

I also installed some AMD drivers directly from their homepage but I 
don't think it installed any kernel module.

The reason I don't know is because I went to do something else while the 
system was booting, so not knowing how long it actually took.

So from second 3.4 to second 139 it was actually enmeshed in sending 
those error messages, or perhaps some ringbuffer filling up?

This immediately followed the recognition of an USB keyboard.

No other differences were observed other than:

[    0.000000] ACPI BIOS Warning (bug): 32/64X length mismatch in 
FADT/Pm1aEventBlock: 32/8 (20150930/tbfadt-623)
[    0.000000] ACPI BIOS Warning (bug): 32/64X length mismatch in 
FADT/Pm1aControlBlock: 16/8 (20150930/tbfadt-623)
[    0.000000] ACPI BIOS Warning (bug): 32/64X length mismatch in 
FADT/PmTimerBlock: 32/8 (20150930/tbfadt-623)
[    0.000000] ACPI BIOS Warning (bug): 32/64X length mismatch in 
FADT/Gpe0Block: 64/8 (20150930/tbfadt-623)
[    0.000000] ACPI BIOS Warning (bug): Invalid length for 
FADT/Pm1aEventBlock: 8, using default 32 (20150930/tbfadt-704)
[    0.000000] ACPI BIOS Warning (bug): Invalid length for 
FADT/Pm1aControlBlock: 8, using default 16 (20150930/tbfadt-704)
[    0.000000] ACPI BIOS Warning (bug): Invalid length for 
FADT/PmTimerBlock: 8, using default 32 (20150930/tbfadt-704)

That wasn't there before (possibly due to a bios option being 
activated), and...

[    0.007220] ACPI: 2 ACPI AML tables successfully acquired and loaded

versus

[    0.005935] ACPI: 1 ACPI AML tables successfully acquired and loaded

in the original, possibly also due to all kinds of BIOS options being 
activated by me.

To me it's still a mystery.

Quite possibly it was an USB error because the kernel can hang on 
malfunctioning USB devices.




More information about the ubuntu-users mailing list