On 26/07/26 15:25, Bruno Haible wrote:
> Collin Funk wrote:
>> Does the attached patch work? ...
>>
>> Note that I also haven't pushed it yet, since I suspect we will want to
>> update the outdated commentary to reflect the fact that this is now
>> standardized. And also consider not compiling these files
>> unconditionally.
>
> While your patch is correct, and what you say is correct as well, I see it
> as a glibc bug in the first place. Reported at
> https://sourceware.org/bugzilla/show_bug.cgi?id=34437
This approach saves us to export an additional symbol and I did not foresee that
such configure checks would fail, since they bypass the C headers and
deliberately
redeclares the function with a bogus prototype.
This will add the additional constraint on exporting all extensions that are
ratified by POSIX as extra symbols, not ideal but doable.
Is this really a glibc bug? If so, I think it would be good that we properly
document this extra implicit ABI constraints, where only providing new symbols
thought asm redirection is not suffice for build systems.
