Debug

[h-vetinari]

Trying to debug flame/blis#514

[conda-forge-linter]

Hi! This is the friendly automated conda-forge-linting service.

I just wanted to let you know that I linted all conda-recipes in your PR (recipe) and found it was in an excellent condition.

[conda-forge-linter]

Hi! This is the friendly automated conda-forge-linting service.

I wanted to let you know that I linted all conda-recipes in your PR (recipe) and found some lint.

Here's what I've got...

For recipe:

• Failed to even lint the recipe, probably because of a conda-smithy bug 😢. This likely indicates a problem in your meta.yaml, though. To get a traceback to help figure out what's going on, install conda-smithy and run conda smithy recipe-lint . from the recipe directory.

[conda-forge-linter]

Hi! This is the friendly automated conda-forge-linting service.

I just wanted to let you know that I linted all conda-recipes in your PR (recipe) and found it was in an excellent condition.

[h-vetinari]

@conda-forge/help-c-cpp, with the help of some numpy folks, I've managed to figure out some of the numpy failures with blis, and opened an issue upstream. The developers there asked me to build and run a (potential) C-reproducer, but I cannot seem to get the compilation to work here.

Pointers would be much appreciated. 🙃

[isuruf]

What's the error you are getting? Wouldn't it be better to test in numpy-feedstock where the turnaround time is less due to blis being already compiled?

[h-vetinari]

What's the error you are getting?

Error is:

+ /home/conda/feedstock_root/build_artifacts/blis_1625166386839/_test_env_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_p/bin/x86_64-conda-linux-gnu-cc -I/home/conda/feedstock_root/build_artifacts/blis_1625166386839/_test_env_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_p/include debug_blis.c
debug_blis.c:1:10: fatal error: blis.h: No such file or directory
    1 | #include <blis.h>

Which surprises me, because I'm explicitly including $PREFIX/include, and blis.h is tested to be in there.

Wouldn't it be better to test in numpy-feedstock where the turnaround time is less due to blis being already compiled?

I've thought about that, but the turn-around isn't massively better (in fact worse for all but windows), but more importantly IMO, if there are patches for blis to try out, this is the feedstock to try them out with - hence why I chose this feedstock.

[h-vetinari]

Thanks a lot for the help here, would have been a slow process...

I was wondering if the windows incantation needs something for the runtime linkage as well - e.g. /MT or so?

[h-vetinari]

This PR has served its purpose, closing.