LeiWang1999 commented on code in PR #15106:
URL: https://github.com/apache/tvm/pull/15106#discussion_r1232990653
##########
python/tvm/testing/tir.py:
##########
@@ -128,3 +128,114 @@ def tile_wmma_fragment(block_read, height, width):
sch.tensorize(sch.get_loops(C_warp)[-2], mma_store_intrin)
return sch
+
+
+def mfma_schedule(
+ workload,
+ k_inner,
+ in_dtype,
+ b_transposed,
+ i_factors,
+ j_factors,
+ k_factors,
+ index_map_A,
+ index_map_B,
+ index_map_C,
+ ldmatrix_a_intrin,
+ ldmatrix_b_intrin,
+ mfma_intrin,
+ mfma_fill_intrin,
+ mfma_store_intrin,
+ shared_scope="shared",
+):
+ import tvm
+ ir_module = tvm.IRModule({"main": workload})
+ sch = tvm.tir.Schedule(ir_module)
+
+ wmma_m = 16
+ wmma_n = 16
+ wmma_k = k_inner
+ warp_size = 64
+ block = sch.get_block("C")
+ i, j, k = sch.get_loops(block)
+ i, i_tc = sch.split(i, factors=[None, wmma_m])
+ j, j_tc = sch.split(j, factors=[None, wmma_n])
+ k, k_tc = sch.split(k, factors=[None, k_inner])
+
+ sch.reorder(i, j, k, i_tc, j_tc, k_tc)
+
+ block_inner = sch.blockize(i_tc)
+ block_outer, block_inner = block_inner, block
+
+ num_ty = i_factors[2] * j_factors[2]
+
+ i0, i1, i2, i3, i4 = sch.split(i, factors=i_factors)
+ j0, j1, j2, j3, j4 = sch.split(j, factors=j_factors)
+ k0, k1, k2 = sch.split(k, k_factors)
+
+ sch.reorder(i0, j0, i1, j1, j2, i2, k0, k1, i3, j3, k2, i4, j4)
+
+ block_idx = sch.fuse(i0, j0)
+ block_idy = sch.fuse(i1, j1)
+ thread_idy = sch.fuse(j2, i2)
+ sch.bind(block_idx, "blockIdx.x")
+ sch.bind(block_idy, "blockIdx.y")
+ sch.bind(thread_idy, "threadIdx.y")
+
+ def fetch_to_shared(block, idx, ndim):
+ block_read = sch.cache_read(block, idx, shared_scope)
+ sch.compute_at(block_read, k0)
+ vector_size = 16 if in_dtype == "int8" else 8
+ fused = sch.fuse(*sch.get_loops(block_read)[-ndim:])
+ _, f_1, f_2, f_3 = sch.split(
+ fused, factors=[None, num_ty, warp_size, vector_size])
+ sch.bind(f_2, "threadIdx.x")
+ sch.bind(f_1, "threadIdx.y")
+ sch.vectorize(f_3)
+ return block_read
+
+ fetch_to_shared(block_outer, 0, 2)
+ fetch_to_shared(block_outer, 1, 2)
+
+ A_warp = sch.cache_read(block_outer, 0, "local")
+ B_warp = sch.cache_read(block_outer, 1, "local")
+
+ sch.compute_at(A_warp, k1)
+ sch.compute_at(B_warp, k1)
+ C_warp = sch.cache_write(block_outer, 0, "local")
+ sch.reverse_compute_at(C_warp, thread_idy)
+
+ ii, jj = sch.get_loops(C_warp)[-2:]
+ io, ii = sch.split(ii, factors=[None, 16])
+ jo, ji = sch.split(jj, factors=[None, 16])
+ sch.reorder(io, jo, ii, ji)
+
+ sch.decompose_reduction(block_outer, sch.get_loops(block_outer)[3])
+ block_init_c = sch.get_block("C_init")
+
+ def tile_wmma_fragment(block_read, height, width):
+ i, j = sch.get_loops(block_read)[-2:]
+ i0, i1 = sch.split(i, factors=[None, height])
+ j0, j1 = sch.split(j, factors=[None, width])
+ sch.reorder(i0, j0, i1, j1)
+ return i1
+
+ loop_a = tile_wmma_fragment(A_warp, 16, k_inner)
+
+ if b_transposed:
+ loop_b = tile_wmma_fragment(B_warp, 16, k_inner)
+ else:
+ loop_b = tile_wmma_fragment(B_warp, k_inner, 16)
+
+ sch.transform_layout(A_warp, ("write", 0), index_map_A)
+ sch.transform_layout(B_warp, ("write", 0), index_map_B)
+ sch.transform_layout(C_warp, ("read", 0), index_map_C)
+ print(sch.mod["main"].script())
Review Comment:
hi sy, did you mean that this part of the code overlaps with mma_schedule,
so we can make changes on mma_shcedule part instead of create a new mfma one?
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
