[J-core] qemu updates / roadmap

[Christopher Friedt]

+Peter Maydell (sorry for cross-posting)

Entire thread can be found here

http://lists.j-core.org/pipermail/j-core/2016-May/000095.html

On Fri, May 20, 2016 at 10:21 AM, Christopher Friedt
<chrisfriedt at gmail.com> wrote:
> On Fri, May 20, 2016 at 7:36 AM, Christopher Friedt
> <chrisfriedt at gmail.com> wrote:
>>
>> On May 19, 2016 5:16 PM, "Christopher Friedt" <chrisfriedt at gmail.com> wrote:
>>> https://github.com/cfriedt/qemu/tree/feature/j2-sh2-dirty-hacks
>>
>> That branch is now passing CAS tests.
>>
>> Hitting 'mova' corner cases now.
>
> Ugh... the MOVA corner cases (_testmov2) seem to be a bit trickier.
> I'm dealing with that in
>
> https://github.com/cfriedt/qemu/tree/for-j2-sh2-dirty-hacks
>
> right now, although it isn't yet clear to me if this is a delay slot
> issue or a separate quirk.
>
> According to the SH12 programming manual:
> "If this instruction is placed immediately after a delayed branch
> instruction, the PC must point to an address specified by (the
> starting address of the branch destination) + 2."
>
> Running into so many failing corner cases (in the original SH Qemu
> port) was not something I anticipated. Sorry for any delays that may
> result.
>
> I think I'll ping-pong between "for-j2-sh2-dirty-hacks " and
> "feature/j2-sh2-dirty-hacks" for a while, adding support to
> "feature/j2-sh2-dirty-hacks" as I go along. Any pointers are welcome.

Incidentally, RISU sounds quite useful as a testbench for J2 / SH* on Qemu.

http://events.linuxfoundation.org/sites/events/files/slides/Validating%20and%20defending%20QEMU%20TCG%20targets-FINAL_0.pdf

Considering that I would like to rewrite a fair bit of _decode_opc()
for SH, it might be a worthwhile thing to investigate.

I'm sure the Qemu folks would appreciate that level of testing too,
before accepting patches.