Gengyuan Bai 1d155f4b80 [Cleanup] Replace root reproducers with CPU regression coverage (#2878) ...

* [Cleanup] Replace root vec_type reproducer with CPU regression test (#2875)

Remove the root-level reproducer scripts and add a proper pytest under
testing/python/cpu covering vectorized vec_type arithmetic (+ - * /) on
the CPU C target: the generated source must reach a float4 expression,
compile as C++, and execute with numerically correct results.

* [Test] Pin CPU vec_type test tensors to float32/cpu

CodeRabbit review finding: torch.randn/rand/empty respected PyTorch's
process-wide default dtype/device, so another test changing
set_default_dtype/set_default_device could create float64 or non-CPU
tensors and break the float32 CPU kernel contract. Pin all six tensors
explicitly without changing process-wide defaults.

2026-08-05 14:57:38 +08:00

..

test_tilelang_cpu_gemm.py

[Transform] Materialize kernel launch per backend (#2387)

2026-06-12 17:14:02 +08:00

test_tilelang_cpu_tgemm.py

[Transform] Materialize kernel launch per backend (#2387)

2026-06-12 17:14:02 +08:00

test_tilelang_cpu_transpose.py

[BugFix] Make T.transpose swap only the final two axes (#2757)

2026-07-24 13:29:24 +08:00

test_tilelang_cpu_vec_type.py

[Cleanup] Replace root reproducers with CPU regression coverage (#2878)

2026-08-05 14:57:38 +08:00

test_tilelang_cpu_while_repro.py

[Transform] Materialize kernel launch per backend (#2387)

2026-06-12 17:14:02 +08:00