R0CKSTAR 272aad8b98 musa : define __CUDA_ARCH__ for device passes (#29508) ...

The MUSA vendor header never defined __CUDA_ARCH__, so every architecture
test in the shared ggml-cuda sources evaluated to 0.  Kernel bodies gated on
the architecture therefore compiled to nothing, for example the q8_0 -> f16
dequantization kernel in convert.cu, whose NO_DEVICE_CODE fallback expands to
an empty body in host code.

Report the newest architecture like the HIP backend does and exclude the
NVIDIA-only features explicitly, as they are not usable on MUSA.  Define it
for device passes only: CUB uses defined(__CUDA_ARCH__) to detect device
compilation, which is also how nvcc behaves.

Drop the now-redundant defined(__CUDA_ARCH__) checks in the architecture
comparisons: __CUDA_ARCH__ is undefined in host passes for CUDA and MUSA, and
HIP defines it for every pass, so both forms select the same branch.

2026-09-30 09:14:37 +02:00

..

cmake

Enables Windows ARM64 build with MSVC cl.exe (#28362)

2026-09-28 10:07:27 +02:00

include

ci: add zdnn backend build but not test (#29541)

2026-09-29 20:45:14 +03:00

src

musa : define __CUDA_ARCH__ for device passes (#29508)

2026-09-30 09:14:37 +02:00

.gitignore

vulkan : cmake integration (#8119)

2024-07-13 18:12:39 +02:00

CMakeLists.txt

sync : ggml (#29396)

2026-09-24 22:44:05 +03:00