Graham Markall comments

Results 776 comments of


                                            Graham Markall

Allow numba functions within zip files to be cached

From a quick skim it looks like the tests aren't platform-agnostic WRT the way paths are formed (the root, and separators) - I haven't looked deeper, but perhaps `pathlib.Path` could...

Allow numba functions within zip files to be cached

> This is ready for review! Oops, sorry, I forgot to check back when I was waiting for CI the other day.

Allow numba functions within zip files to be cached

> But, this gives new flake8 errors: I think CI runs `flake8 numba` so this wouldn't usually be an issue - all the errors are from files outside the `numba`...

Omitted default argument not working

I also tried to work around this, with: ```python import numba as nb from numba import types from numba.typed import Dict my_set = {"Ab", "ab", "abcde"} @nb.njit def create_numba_dict(elements: set)...

Removes `context.compile_internal` where easy

gpuci run tests

Removes `context.compile_internal` where easy

gpuci run tests

Removes `context.compile_internal` where easy

(Note that the gpuCI fail is just an issue inherited from `main`, not this PR)

Removes `context.compile_internal` where easy

gpuci run tests (just to make everything look nice and green now #8537 is merged)

Removes `context.compile_internal` where easy

This message in the traceback: `Only accept returning of array passed into the function as argument` I see popping up when NRT is not available (usually in the CUDA target,...

Add ability to link CUDA functions with in-memory PTX.

@ed-o-saurus Apologies for the confusion here - are you using Linux and CUDA 12.0 or later? If so, the workaround is to add [`pynvjitlink`](https://github.com/rapidsai/pynvjitlink) as a dependency and call ```python...