On 9/2/2026 9:24 AM, Pierrick Bouvier wrote:
> On 9/2/2026 8:55 AM, Philippe Mathieu-Daudé wrote:
>> On 1/9/26 22:20, Pierrick Bouvier wrote:
>>> Now that we remove machine_typename, we don't need to keep duplicated
>>> file per target anymore. So we can just rely on target-info-stub.c.
>>>
>>> Signed-off-by: Pierrick Bouvier <[email protected]>
>>> ---
>>>   configs/targets/aarch64-softmmu.c | 26 --------------------------
>>>   configs/targets/arm-softmmu.c     | 26 --------------------------
>>>   configs/targets/meson.build       |  6 ------
>>>   configs/targets/riscv32-softmmu.c | 25 -------------------------
>>>   configs/targets/riscv64-softmmu.c | 25 -------------------------
>>>   meson.build                       |  7 +------
>>>   6 files changed, 1 insertion(+), 114 deletions(-)
>>>   delete mode 100644 configs/targets/aarch64-softmmu.c
>>>   delete mode 100644 configs/targets/arm-softmmu.c
>>>   delete mode 100644 configs/targets/meson.build
>>>   delete mode 100644 configs/targets/riscv32-softmmu.c
>>>   delete mode 100644 configs/targets/riscv64-softmmu.c
>>
>> We also need to update MAINTAINERS. Anyway I rather keep that for
>> now (including the following patch). I'am squashing removal of
>> "*machines-qom.h" include lines where appropriate.
> 
> I don't mind this too much.
> 
> Out of curiosity, what's the reason to keep it?
> 
> The only reason it was duplicated was to specify different typenames.
> It could be argued that writing them without defines helps to see
> definitions for each target, but not sure it's worth the copy paste IMHO.
>

Answer from Philippe:
idea is to keep them is to remove more cpu defines, and eradicate
cpu-param.h headers. Sounds good to me.

> Regards,
> Pierrick


Reply via email to