Hi,

I ported DWC PCIE part from fsl-imx7.c to my new cusotm virt machine based
on RISC-V "virt".

It looks ok so far. Now i'm wondering whether there is a "standard/ready to
use for most of cases" device tree description for the synopsis pcie host
device ?

Looking this way:
https://mjmwired.net/kernel/Documentation/devicetree/bindings/pci/designware-pcie.txt
but would be glad to hear some advices if any.

Best regards,
Arthur

On Fri, Jun 21, 2024 at 1:30 PM Thomas Huth <th...@redhat.com> wrote:

> On 21/06/2024 11.16, Peter Maydell wrote:
> > On Fri, 21 Jun 2024 at 08:07, Arthur Tumanyan <arthurtuman...@gmail.com>
> wrote:
> >>
> >> Hi,
> >>
> >> I just tried to run mcimx7d-sabre machine this way:
> >>
> >> ${HOME}/cosim/usr/local/bin/qemu-system-arm -M mcimx7d-sabre -m 2G \
> >> -kernel ${HOME}/cosim-arm/buildroot/output/images/uImage \
> >>          --initrd
> ${HOME}/cosim-arm/buildroot/output/images/rootfs.cpio.gz \
> >> -nographic \
> >> -net nic -net user
> >>
> >> and it just prints this and do nothing: qemu-system-arm: warning: nic
> imx.enet.1 has no peer
> >>
> >> Based on what I see in the mcimx7d-sabre.c , it configures just very
> basic things, no PCIe at all (may be I'm wrong ;) )
> >
> > The machine model in mcimx7d-sabre.c creates the SoC object
> > (TYPE_FSL_IMX7). It's the code for that in hw/arm/fsl-imx7.c
> > that creates all the SoC devices including the PCIe controller.
> > (This structure is similar to real hardware where you have a
> > board, which has one or two chips on it like RAM but most of
> > the complicated stuff is inside the one big SoC chip.)
> >
> >> Is there any idea what goes wrong here ? Maybe someone has experience
> with running this machine ?
> >
> > "No output" usually means "your guest kernel is not configured
> > correctly for this machine type", and/or possibly "you didn't
> > tell the kernel to output on the serial port".
>
> By the way, it seems like we don't even have an avocado test for that
> machine available. Peter, do you know whether there is a kernel for this
> machine available somewhere that we could use for testing?
>
>   Thomas
>
>
>

Reply via email to