Re: [Qemu-devel] [PATCH] mips_malta: move i8259 initialization after piix4 initialization

[Stefan Weil <weil@mail.berlios.de>]

Am 14.09.2011 18:19, schrieb Andreas Färber:
> Am 14.09.2011 14:00, schrieb Avi Kivity:
>> On 09/12/2011 04:07 PM, Avi Kivity wrote:
>>> i8259 is an ISA device (or at least, depends on the ISA
>>> infrastructure to
>>> register its ioport); and the ISA bus is supplied by piix4. Later
>>> patches
>>> make this dependency explicit.
>>>
>>> Move the i8259 initialization until after the ISA bus is created; and
>>> supply
>>> a new qemu_irq to PCI initialization, since the i8259 isn't ready
>>> yet. Later
>>> wire the new qemu_irq to the i8259.
>>>
>>
>> Can this please be reviewed? It's in the front of the memory queue
>> logjam.
>
> With or without this patch, the mips and mipsel test images on qemu.org
> produce no output at all, on x64 host.
>
> Andreas

I just tested the mipsel test image on i386 and x64 hosts with the
same result. Adding -d in_asm shows that the linux kernel is started,
but runs in a loop at some point (timer initialization?). Timer interrupts
are still served (attach cross gdb to running emulation).

Could you already bisect this problem? I did not run these tests for
years.

My standard mips tests use debian mips kernels and show no problems.

Stefan