Re: [PATCH] Fortran: degree trigonometric functions and procedure pointers [PR121366]

Hi Mikael,

On 7/6/26 22:05, Mikael Morin wrote:
> Le 05/07/2026 à 22:39, Harald Anlauf a écrit :
>> [Argh, pressed send too early...]
>>
>> Dear all,
>>
>> this patch fills in a gap for the degree trigonometric functions: there
>> is a library version with C calling conventions (by value), but the
>> Fortran wrappers were missing.  For their use as procedure pointer
>> target one needed to use the proper isym in resolution and generate
>> the proper library name in line with our conventions.
>>
>> The new wrappers were coded manually. 
> The radian functions were made one way, and now the degree ones some 
> other way.  That's a bit unfortunate, even if the maintenance burden 
> should be rather low.  What didn't work with regenerate.sh and 
> Makefile.am?  Was it the autoconf/automake version?  I can see that you 
> managed to regenerate Makefile.in.

I did install the versions of autoconf and automake required and could
use them here, otherwise regenerate.sh would have failed for me.

However, I did not see a way to reuse regenerate.sh for the functions
in question.  The degree trigonometric functions would need prototypes
to be added so that the autotests would find them, produce macros like
HAVE_ATAN2DL etc., possibly a platform nightmare.  So I thought it is
not worth it.

Just pushed it as gcc-17-2179-g216d08b2cc1 .

> The rest looks good.  I understand little of it though.
> Mikael

Thanks for the review!

Harald

>> I was not entirely sure if
>> the kind=16/17 handling is fine, so feedback is appreciated.
>>
>> The testcase checks both the F2023 variants as well as the existing
>> GNU intrinsic extensions.  (Non-existing GNU extensions were removed
>> from the documentation.)
>>
>> I checked that the testcase passes locally also when specifying
>> -freal-4-real-8, -freal-4-real-10, -freal-4-real-16,
>> -freal-8-real-10, and -freal-8-real-16,
>> but I chose to not test this explicitly in the testsuite.
>> If it is felt important, it can be added.
>>
>> Regtested cleanly on x86_64-pc-linux-gnu.  OK for mainline?
>>
>> Do we want a backport to 16-branch?  If this is the case,
>> I would need to adjust gfortran.map accordingly.
>>
>> Thanks,
>> Harald
>>
> 
>