Re: part_blk on fiasco.oc

[Sebastian S. <Seb...@ge...>]

Hi,

On 04/27/2013 10:51 AM, yuqing wang wrote:
> Hi,
> I used the foc_x86_32 platform to try part_blk run script ,there was a
> error.
> 
> [init -> atapi_drv] 5 and below: yes enabled: yes
> [init -> atapi_drv] PRD base at 002f9000 (physical) at 00002000 (virtual)
> [init -> atapi_drv] Device initialized! Enabling interrupts ...
> Error: Test execution timed out
> make: *** [run/part_blk] Error 254
> 
> I try to find where is wrong and I get it 

You don't get an interrupt. Glancing at the run script, I see that the
ACPI driver is missing. On hardware and even Qemu it is usually required
to find out the correct GSI (global system interrupt) values. It then
overwrites the ones in the PCI config space, that are legacy PIC
interrupt numbers. Also the driver parses the MADT in order to retrieve
any interrupt overrides. Since you also seem to use the AHCI run script,
you can configure the ACPI server and the ATAPI driver in the same way
as it's done there.

Regards,

Sebastian