Re: [PATCH 13/29] target/riscv: Use GDBFeature for dynamic XML

[Alex Bennée]

Alex Bennée <alex.bennee@linaro.org> writes:

> Alex Bennée <alex.bennee@linaro.org> writes:
>
>> From: Akihiko Odaki <akihiko.odaki@daynix.com>
>>
>> In preparation for a change to use GDBFeature as a parameter of
>> gdb_register_coprocessor(), convert the internal representation of
>> dynamic feature from plain XML to GDBFeature.
>>
>> Signed-off-by: Akihiko Odaki <akihiko.odaki@daynix.com>
>> Message-Id: <20231025093128.33116-7-akihiko.odaki@daynix.com>
>> Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
>
> I bisected the following failure:
>
>  ./qemu-riscv64 -g 1234 ./tests/tcg/riscv64-linux-user/sha512
>
> and:
>
>  gdb-multiarch ./tests/tcg/riscv64-linux-user/sha512 -ex "target remote 
> localhost:1234" -x ../../tests/tcg/multiarch/gdbstub/registers.py
>
> gives:
>
> warning: Architecture rejected target-supplied description
> Ignoring packet error, continuing...
> Ignoring packet error, continuing...
> Ignoring packet error, continuing...
> Ignoring packet error, continuing...

I don't know if the discontinuous register numbering could be the
problem:

  <feature name="org.gnu.gdb.riscv.csr">
    <reg name="fflags" bitsize="64" type="int" regnum="67"/>
    <reg name="frm" bitsize="64" type="int" regnum="68"/>
    <reg name="fcsr" bitsize="64" type="int" regnum="69"/>
    <reg name="cycle" bitsize="64" type="int" regnum="3138"/>
    <reg name="time" bitsize="64" type="int" regnum="3139"/>
    <reg name="instret" bitsize="64" type="int" regnum="3140"/>
  </feature>

?

-- 
Alex Bennée
Virtualisation Tech Lead @ Linaro