Methodlower(
self, layout_map: dict, target: Target, thread_bounds: Range, thread_var: tir.Var, mbar_phase_expr:
tilelang/metal/op/gemm/gemm_metal.py:23
Functionmain(
A: T.Tensor(A_shape, in_dtype),
B: T.Tensor(B_shape, in_dtype),
C: T.Tensor((M, N),
maint/gemm/correctness_evaluation_tcgen05.py:29
Functionmain(
A: T.Tensor((M, K), in_dtype),
B: T.Tensor((K, N), in_dtype),
C: T.Tensor((M, N), ou
maint/gemm/correctness_evaluation_tcgen05_2cta.py:22
Functionmain(
A: T.Tensor(A_shape, in_dtype),
B: T.Tensor(B_shape, in_dtype),
C: T.Tensor((M, N),
maint/gemm/correctness_evaluation_sm70.py:29
Functionmain(
A: T.Tensor(A_shape, in_dtype),
B: T.Tensor(B_shape, in_dtype),
C: T.Tensor((M, N),
maint/gemm/correctness_evaluation.py:30
Functionmain(
A: T.Tensor((M, K), dtype),
B: T.Tensor((K, N), dtype),
C: T.Tensor((M, N), dtype),
maint/host_checks/common.py:8
Functionmain(
cb: T.Tensor((batch, nchunks, ngroups, chunk_size, chunk_size), dtype), # type: ignore
x: T
benchmark/mamba2/benchmark_mamba_chunk_scan.py:211
Functionmain(
Q: T.Tensor(shape, dtype),
K: T.Tensor(shape, dtype),
V: T.Tensor(shape,
benchmark/blocksparse_attention/benchmark_tilelang_block_sparse_fmha.py:121
Functionmain(
X: T.Tensor((N, D), in_dtype),
gamma: T.Tensor((D,), in_dtype),
beta: T.Tensor((D,),
examples/norm/layernorm.py:11
Functionmain(
A: T.Buffer(A_shape, in_dtype),
B: T.Buffer(B_shape, in_dtype),
C: T.Buffer((M, N),
examples/bitnet-1.58b/kernel_benchmark/tl_int8xint8.py:104
Functionmain(
A: T.Tensor(A_shape, in_dtype),
B: T.Tensor(B_shape, in_dtype),
C: T.Tensor((M, N),
examples/gemm_fp8/example_tilelang_gemm_amd_fp8_preshuffle.py:103
Functionmain(
a: T.Buffer((B, M, N), T.bfloat16),
a_out: T.Buffer((B, M, N), T.float32),
)
testing/python/layout/test_tilelang_layout_fused_replicate.py:16
Functionmain(A: T.Tensor((128, 4), T.float), B: T.Tensor((128, 4), T.float))
testing/python/layout/test_tilelang_layout_inference.py:10
Functionmain(
A: T.Tensor(A_shape, in_dtype),
B: T.Tensor(B_shape, in_dtype),
C: T.Tensor((M, N),
testing/python/components/test_tilelang_pass_config_disable_warp_specialized.py:27
Functionmain(
A: T.Tensor((N,), dtype),
B: T.Tensor((N,), dtype),
C: T.Tensor((N,), dtype),
)
testing/python/language/test_tilelang_language_chain_equal.py:12
Functionmain(
A: T.Tensor((128, 128), "float32"),
B: T.Tensor((128, 128), "float32"),
C: T.Tensor(
testing/python/language/test_tilelang_language_min_blocks_per_sm.py:11
Functionmain(
A: T.Tensor((M, K), in_dtype),
B: T.Tensor((N, K), in_dtype),
C: T.Tensor((M, N), ou
testing/python/language/test_tilelang_language_atom_mma.py:81
Functionmain(
A: T.Tensor((N,), dtype),
B: T.Tensor((N // M, M), dtype),
)
testing/python/language/test_tilelang_language_reshape.py:10
Functionmain(
A: T.Tensor((M, N), dtype),
B: T.Tensor((N, M), dtype),
)
testing/python/language/test_tilelang_language_transpose.py:21
Functionmain(
A: T.Tensor((M, N), dtype),
B: T.Tensor((M, N), dtype),
)
testing/python/language/test_tilelang_language_ternary.py:12
Functionmain(
A: T.Tensor[(M,), dtype_A], # noqa: F821
B: T.Tensor[(M,), dtype_B], # noqa: F821
)
testing/python/language/test_tilelang_language_vectorized_cast.py:12
Functionmain(
A: T.Tensor((64,), T.uint8),
B: T.Tensor((64,), T.uint8),
)
testing/python/language/test_tilelang_language_access_ptr_codegen.py:13
Functionmain(
A: T.Tensor[(128, 8), T.float32],
B: T.Tensor[(128, 8), T.float32],
)
testing/python/language/test_tilelang_language_select.py:9
Functionmain(
A: T.Tensor((M, K), dtype),
B: T.Tensor((K, N), dtype),
BlockMask: T.Tensor(block_ma
testing/python/language/test_tilelang_language_any_of.py:40
Functionmain(
A: T.Tensor((M, N), dtype),
B: T.Tensor((M * N), dtype),
)
testing/python/language/test_tilelang_language_composable_index.py:11
Functionmain(
A: T.Tensor((128, 128), T.bfloat16),
B: T.Tensor((128, 128), T.bfloat16),
D: T.Tenso
testing/python/language/test_tilelang_language_tcgen05_gemm.py:11
Functionmain(
A: T.Tensor(A_shape, in_dtype),
B: T.Tensor(B_shape, in_dtype),
C: T.Tensor((M, N),
testing/python/language/test_tilelang_language_tma_store.py:36
Functionmain(
A: T.Tensor((M, N), src_dtype),
B: T.Tensor((M, N), dst_dtype),
)
testing/python/language/test_tilelang_language_copy.py:13
Functionmain(
A: T.Tensor(A_shape, in_dtype),
B: T.Tensor(B_shape, in_dtype),
C: T.Tensor((M, N),
testing/python/language/test_tilelang_language_pipeline.py:28
Functionmain(
A: T.Tensor[(1,), T.int32],
B: T.Tensor[(1,), T.int32],
)
testing/python/language/test_tilelang_language_augassign.py:24
Functionmain(
A: T.Tensor((M, K), dtype),
B: T.Tensor((N, K), dtype),
C: T.Tensor((M, N), dtype),
testing/python/language/test_tilelang_language_clear.py:9
Functionmain(
A: T.Tensor((M, K), dtype),
B: T.Tensor((N, K), dtype),
C: T.Tensor((M, N), dtype),
testing/python/language/test_tilelang_language_alias.py:9
Functionmain(
a_ptr: T.ptr,
b_ptr: T.ptr,
c_ptr: T.ptr,
m: T.int32,
n: T.int32,
testing/python/language/test_tilelang_language_ptr.py:11
Functionmain(
A: T.Tensor((64, 16), T.float16),
B: T.Tensor((16, 64), T.float16),
D: T.Tensor((64,
testing/python/language/test_tilelang_language_wgmma_gemm.py:11
Functionmain(
x: T.Tensor((n,), T.float32),
y: T.Tensor((n,), T.int32),
)
testing/python/language/test_tilelang_language_isfinite_codegen.py:12
Functionmain(
A: T.Tensor[(M_val,), "float32"], # noqa: F821
B: T.Tensor[(M_val,), "float8_e4m3fn
testing/python/language/test_tilelang_cast_rounding.py:74
Functionmain(
A: T.Tensor((N,), dtype),
B: T.Tensor((N,), dtype),
)
testing/python/language/test_tilelang_language_alloc.py:12
Functionmain(
A: T.Tensor((M, K), dtype),
B: T.Tensor((K, N), dtype),
BlockMask: T.Tensor(block_ma
testing/python/language/test_tilelang_language_all_of.py:40
Functionmain(
A: T.StridedTensor[(N, M), (1, stride_A), T.float32], # noqa: F821
B: T.StridedTensor[(N, M
testing/python/language/test_tilelang_language_vectorize.py:12
Functionmain(
A: T.Tensor((length,), dtype),
B: T.Tensor((length,), dtype),
)
testing/python/language/test_tilelang_language_parallel.py:13
Functionmain(
A: T.Tensor((M, N), dtype),
B: T.Tensor((M, N), dtype),
)
testing/python/language/test_tilelang_language_mask_op.py:10
Functionmain(
A: T.Tensor((N,), T.float32),
B: T.Tensor((N,), T.float32),
)
testing/python/language/test_tilelang_language_source_kernel.py:28
Functionmain(
A: T.Tensor((M, N), dtype),
B: T.Tensor((M, N), dtype),
)
testing/python/language/test_tilelang_language_annotate_safe_value.py:11
Functionmain(A: T.Tensor((M, N), dtype), B: T.Tensor((M,), dtype))
testing/python/language/test_tilelang_language_reduce.py:103
Functionmain(
Src: T.Tensor((N, K), in_dtype),
Idx: T.Tensor((4,), "int32"),
Dst: T.Tensor((N, K),
testing/python/language/test_tilelang_language_tma_gather_scatter.py:28
Functionmain(
x: T.Tensor((N,), T.float32),
y: T.Tensor((N,), T.float32),
)
testing/python/language/test_tilelang_language_intrinsics_codegen.py:11
Functionmain(
A: T.Tensor((M, N), dtype),
B: T.Tensor((M, N), dtype),
BlockMask: T.Tensor(block_ma
testing/python/language/test_tilelang_language_let_layout.py:31
Functionmain(
A: T.Tensor((1,), "int32"),
B: T.Tensor((1,), "int32"),
)
testing/python/language/test_tilelang_language_warp_sync.py:11
Functionmain(
A: T.Tensor((N,), dtype),
B: T.Tensor(new_shape, new_dtype if new_dtype else dtype),
)
testing/python/language/test_tilelang_language_view.py:22
Functionmain(
A: T.Tensor(A_shape, in_dtype),
B: T.Tensor(B_shape, in_dtype),
C: T.Tensor((M, N),
testing/python/language/test_tilelang_language_tma_copy.py:40
Functionmain(
A: T.Tensor((N,), dtype),
B: T.Tensor((N,), dtype),
)
testing/python/language/test_tilelang_language_clamp.py:15
Functionmain(
A: T.Tensor((M, N), dtype),
B: T.Tensor((M, N), dtype),
)
testing/python/language/test_tilelang_language_if_range.py:12
Functionmain(flag: T.Buffer((1,), "int32"), out: T.Buffer((1,), "int32"))
testing/python/cpu/test_tilelang_cpu_while_repro.py:22
Functionmain(
A: T.Tensor(A_shape, dtype),
B: T.Tensor(B_shape, dtype),
C: T.Tensor((M, N), dtype)
testing/python/cpu/test_tilelang_cpu_tgemm.py:22
Functionmain(
A: T.Tensor(A_shape, in_dtype),
B: T.Tensor(B_shape, in_dtype),
C: T.Tensor((M, N),
testing/python/amd/test_tilelang_gemm_mfma_preshuffle.py:160
Functionmain(
src: T.Tensor((QN,), T.uint8),
dst: T.Tensor((QN,), T.uint8),
)
testing/python/amd/test_tilelang_mxfp4_gfx950.py:72
Functionmain(
A: T.Tensor(A_shape, in_dtype),
B: T.Tensor(B_shape, in_dtype),
C: T.Tensor((M, N),
testing/python/amd/test_tilelang_gemm_mfma_intrinsic.py:83
Functionmain(A: T.Tensor(A_shape, in_dtype), B: T.Tensor(B_shape, in_dtype), C: T.Tensor((M, N), out_dtype))
testing/python/amd/test_tilelang_test_amd.py:31
Functionmain(
A: T.Tensor(A_shape, in_dtype),
B: T.Tensor(B_shape, in_dtype),
C: T.Tensor((M, N),
testing/python/amd/test_tilelang_gfx950_copy_async.py:45
Functionmain(A: T.Tensor((16, 14), dtype=dtype), B: T.Tensor((16, 448), dtype=dtype))
testing/python/issue/test_tilelang_issue_1374.py:13
Functionmain(
A: T.Tensor[(2, 512), T.float32],
B: T.Tensor[(2, 512), T.float32],
C: T
testing/python/issue/test_tilelang_issue_1734.py:12
Functionmain(
A: T.Tensor((M, K), dtype),
B: T.Tensor((N, K), dtype),
C: T.Tensor((M, N), dtype),
testing/python/issue/test_tilelang_issue_96.py:9
Functionmain(
A_sparse: T.Tensor((M, K // 2), T.float16),
E: T.Tensor((M, K // e_factor), T.int8),
testing/python/issue/test_tilelang_issue_ws_simt_copy_full_producer_extent.py:31
Functionmain(
A: T.Tensor((M, K), T.float16),
B: T.Tensor((K, N), T.float16),
C: T.Tensor((M, N),
testing/python/issue/test_tilelang_issue_1697.py:9
Functionmain(
A: T.Tensor((length,), dtype),
B: T.Tensor((length,), dtype),
)
testing/python/analysis/test_tilelang_nested_loop_checker.py:34
Functionmain(
A: T.Tensor((M, N), dtype),
B: T.Tensor((M, N), dtype),
)
testing/python/fastmath/test_mathops_fastmath.py:69
Functionmain(
A: T.Tensor(A_shape, in_dtype),
B: T.Tensor(B_shape, in_dtype),
C: T.Tensor((M, N),
testing/python/jit/test_tilelang_jit_cutedsl.py:43