Graham Markall comments

Results 776 comments of


                                            Graham Markall

Update llvmlite to be compatible with llvm-17

> Can you elaborate more on the advantages of doing it this way? Anything that uses llvmlite needs to be atomically on one pass manager or the other so having...

Update llvmlite to be compatible with llvm-17

> As a Numba user, I do like the feature `opaque pointers`, since sometimes I need to write some lowering code in Numba with llvmlite builder/IR APIs. And many times,...

Update llvmlite to be compatible with llvm-17

Closing - superseded by #1092.

Jitclasses do not properly implement `is`

We do have a bit of discussion around this issue in https://github.com/numba/numba/issues/8712 too (the `numba_id()` vs `id()` ideas, etc.)

Support zero-length arrays in numba.cuda

Thanks for the report / request!

Support zero-length arrays in numba.cuda

Yes, I don't think they're duplicates, I was just trying to gather together a few shape-related issues in case someone is looking at fixing up things in these areas.

Support zero-length arrays in numba.cuda

@AruparnaMaity If you can see a way to fix it, you could open a PR with the fix, if you'd like - note that as per https://numba.discourse.group/t/rfc-moving-the-cuda-target-to-a-new-package-maintained-by-nvidia/2628 new PRs should...

#8783, Namedtuple fixes with merged `main`

I think this shows that #8783 is still good WRT main today.

Cannot assign to NamedTuple array when @njit(parallel=True)

Is this related? https://numba.discourse.group/t/are-the-results-of-this-verification-using-namedtuple-correct/3074