llvm-project.git/libc/src/math/amdgpu, branch users/nico/python-2

[libc][NFC] Remove unused CMake for gpu math

2025-06-07T14:25:27+00:00

[libc] Cleanup unimplemented math functions (#143173)

2025-06-06T17:27:13+00:00

Summary:
This patch cleans up the leftoever files that were either implemented or
are still unimplemented stubs.

[libc] Add `lgamma` and `lgamma_r` stubs for the GPU (#102019)

2024-08-05T19:53:05+00:00

Summary:
These functions are used by the  implementation in libc++ and
cause a lot of tests to fail. For now we provide these through the
vendor abstraction until we have a real version. The NVPTX version
doesn't even update the output correctly so these are just temporary.

[libc][math] Implement fast pass for double precision pow function with up to 1ULP error. (#101926)

2024-08-05T17:08:59+00:00

[libc] Use generic 'atan2' function for the GPU

2024-08-05T12:14:41+00:00

Summary:
I forget to enable this one in the previous patch.

[libc] Migrate to using LIBC_NAMESPACE_DECL for namespace declaration (#98597)

2024-07-12T16:28:41+00:00

This is a part of #97655.

Revert "[libc] Migrate to using LIBC_NAMESPACE_DECL for namespace declaration" (#98593)

2024-07-12T07:12:13+00:00

Reverts llvm/llvm-project#98075

bots are broken

[libc] Migrate to using LIBC_NAMESPACE_DECL for namespace declaration (#98075)

2024-07-11T19:35:22+00:00

This is a part of #97655.

[libc] Add support for `powi` as an LLVM libc extension on the GPU (#98236)

2024-07-10T01:51:36+00:00

Summary:
This function is used by the CUDA / HIP / OpenMP headers and exists as
an NVIDIA extension basically. This function is implemented in the C23
standard as `pown`, but for now we need to provide `powi` for backwards
compatibility. In the future this entrypoint will just be a redirect to
`pown` once that is implemented.

[libc] Fix integer `rint` variants on the GPU (#98095)

2024-07-09T01:07:45+00:00

Summary:
Currently these are implemented as a static cast on `__builtin_rint()`
to a long. Howver, this is not strictly correct. The standard states
that the output is unspecified, but most implementations, and the LLVM
libc implementation, do some kind of guarantee on this beahvior. This is
not guaranteed by just doing a cast. This patch just uses the generic
versions until we implement `__builitin_lrint` correctly.