Eric Blake writes:
> On Fri, Apr 04, 2025 at 02:16:03PM +0200, Simon Josefsson wrote:
>> Eric Blake writes:
>>
>> >> # define G_GNUC_IDN2_ATTRIBUTE_PURE __attribute__ ((pure))
>> ...
>> >> extern _IDN2_API const char *idn2_check_version (const char
>> >> *req_version)
>> >> G_GNUC_IDN2_
On Fri, Apr 04, 2025 at 02:16:03PM +0200, Simon Josefsson wrote:
> Eric Blake writes:
>
> >> # define G_GNUC_IDN2_ATTRIBUTE_PURE __attribute__ ((pure))
> ...
> >> extern _IDN2_API const char *idn2_check_version (const char *req_version)
> >> G_GNUC_IDN2_ATTRIBUTE_PURE;
> >
> > Figuring out
Eric Blake writes:
>> # define G_GNUC_IDN2_ATTRIBUTE_PURE __attribute__ ((pure))
...
>> extern _IDN2_API const char *idn2_check_version (const char *req_version)
>> G_GNUC_IDN2_ATTRIBUTE_PURE;
>
> Figuring out how to write the __attribute__-hiding macros conditional
> on compiler version w
On Fri, Apr 04, 2025 at 12:23:55PM +0200, Simon Josefsson wrote:
> Eric Blake writes:
>
> > I'm trying to run indent on the GNU M4 source code base before a
> > release (it looks like gnulib added the ability to run make indent
> > since the last time I made an m4 release). But one change that i
Eric Blake writes:
> I'm trying to run indent on the GNU M4 source code base before a
> release (it looks like gnulib added the ability to run make indent
> since the last time I made an m4 release). But one change that indent
> is insisting on is wrong:
>
> -extern void m4_error (int, int, cons
