[PATCH V4 2/2] riscv: board: Support OpenPiton SoC

[Bin Meng]

On Fri, May 7, 2021 at 9:17 PM Sean Anderson <seanga2 at gmail.com> wrote:
>
> On 5/7/21 9:11 AM, Bin Meng wrote:
> > On Fri, May 7, 2021 at 9:03 PM Sean Anderson <seanga2 at gmail.com> wrote:
> >>
> >> On 5/6/21 11:48 PM, Tianrui Wei wrote:
> >>>
> >>> On 5/7/2021 11:41 AM, Sean Anderson wrote:
> >>>> On 5/6/21 11:28 PM, Tianrui Wei wrote:
> >>>>>
> >>>>> On 5/7/2021 11:15 AM, Sean Anderson wrote:
> >>>>>> On 5/6/21 11:06 PM, Tianrui Wei wrote:
> >>>>>>> Hi Sean,
> >>>>>>>
> >>>>>>>
> >>>>>>> Many thanks again for reviewing our code! We really appreciate it. Will fix the things you're suggesting ;p Though I have a few questions in line in the comment. Also, checkpatch didn't catch any of the identation issues. I was wondering if there are some specific flags to enable some of the checks? I'm running ./utils/checkpatch.pl this.patch right now.
> >>>>>>>
> >
> > [snip]
> >
> >>>>>>>
> >>>>>>>
> >>>>>>> No we don't support fit yet for some hardware reasons.
> >>>>>>
> >>>>>> Can you elaborate on that? I wasn't aware of any restrictions in this
> >>>>>> area.
> >>>>>
> >>>>>
> >>>>> So OpenPiton is actually some kind of SoC generator that generates
> >>>>> different SoC on FPGA. The device tree was generated at bitstream
> >>>>> creation time to facilitate different configurations, so each board
> >>>>> can have different device trees. We're aware of any way to do this in
> >>>>> FIT.
> >>>>
> >>>> Oh, so you're saying that the devicetree is placed at a specific address
> >>>> in hardware? Does it need any fixups?
> >>>
> >>>
> >>> No for now we just embed it in SPL, will be converted into the bootrom for the chip with a small ZSBL.
> >>
> >> Well, atm you are using the DTS embedded in U-Boot.
> >>
> >> I don't know what the correct way to do this is...
> >>
> >> +CC Bin, Heinrich: Do you have a comment on this?
> >
> > I am not sure I followed this correctly. Ideally we should use OF_SEPARATE.
>
> As I understand it, Tianrui would like to use the same FDT that U-Boot
> uses for Linux as well. Is doing something like
>
> >> +    "fdt addr ${fdtcontroladdr}; " \
> >> +    "fdt move ${fdtcontroladdr} ${fdt_addr_r}; " \
> >> +    "booti ${loadaddr} - ${fdt_addr_r}; "

Does "booti loadaddr - fdtcontroladdr" work? Or we have to do "fdt move"?

>
> the right way to go about this? Can any other platforms be used as a
> reference here (perhaps qemu)?

I think this way is okay, as long as U-Boot's device tree is the
superset of the Linux one.

Regards,
Bin