Re: rapidfuzz failing 9.0 i386 2023Q2

[Greg Troxel <gdt%lexort.com@localhost>]

Thomas Klausner <wiz%gatalith.at@localhost> writes:

> On Sun, Jul 02, 2023 at 07:23:14AM -0400, Greg Troxel wrote:
>> http://victory.netbsd.org/pkgsrc/packages/reports/2023Q2/i386-10.0/20230630.1710/py310-rapidfuzz-3.1.1/install.log
>> (I don't see any pkgbuild bulk reports yet)
>> 
>> Looks like confusion with native extension vs not at first glance, which
>> I remember being addressed, but this is i386.
>
> -- Performing Test HAVE_CXX_ATOMICS64_WITH_LIB - Failed
> CMake Error at src/rapidfuzz/CMakeLists.txt:124 (message):
>   No native support for std::atomic, or libatomic not found!
>
> Perhaps adding mk/atomic64.mk will be enough?

It helped I think, but now:

  => Checking file-check results for py310-rapidfuzz-3.1.1
  ERROR: ************************************************************
  ERROR: The following files are in the PLIST but not in /usr/ANONCVS/pkgsrc/textproc/py-rapidfuzz/work/.destdir/usr/pkg:
  ERROR:         /usr/ANONCVS/pkgsrc/textproc/py-rapidfuzz/work/.destdir/usr/pkg/lib/python3.10/site-packages/rapidfuzz/_feature_detector_cpp.cpython-310.so
  ERROR:         /usr/ANONCVS/pkgsrc/textproc/py-rapidfuzz/work/.destdir/usr/pkg/lib/python3.10/site-packages/rapidfuzz/distance/metrics_cpp_avx2.cpython-310.so
  ERROR:         /usr/ANONCVS/pkgsrc/textproc/py-rapidfuzz/work/.destdir/usr/pkg/lib/python3.10/site-packages/rapidfuzz/fuzz_cpp_avx2.cpython-310.so
  *** Error code 1

And I can't say that declining to build avx2 is a bug, since pkgsrc
targets eiher i486 or i586.

I'll send you the full log, and happy to send it to anybody else.

But, I am thinking that we should just drop i386 from the avx2
conditional for PLIST.  I think it's only there because when fixing this
for "not x86" there was an assumption these files were generated on
"x86", rather than "architectures for which avx2 is guaranteed, meaning
x86_64".

So I'll try that and commit/pullup if it works.