Peter Maydell <peter.mayd...@linaro.org> writes:
> On Thu, 26 Nov 2020 at 12:12, Daniel P. Berrangé <berra...@redhat.com> wrote:
>> On Thu, Nov 26, 2020 at 11:57:14AM +0000, Peter Maydell wrote:
>> > This is an interesting case, because it's a header intended
>> > for external use (people compiling plugins), not part of
>> > QEMU proper. I don't know whether we want to impose the
>> > same clang/gcc requirements on plugin builders, though it's
>> > probably not a bad idea to do so. Alex ?
>>
>> IIUC, we expect that the plugins are built specifically to go with
>> the QEMU you have built.
>
> No, the point of the plugin interface is that it actually is
> a (constrained) stable ABI with version-querying and checking
> so you don't have to build a plugin against the exact matching
> QEMU version. (But "don't use really ancient gcc" doesn't seem
> like a very major thing to ask of its users.)
Quite. But I agree using a GCC from this decade seems reasonable
especially as the feature has only been in QEMU for a few cycles.
Acked-by: Alex Bennée <alex.ben...@linaro.org>
--
Alex Bennée
