mirror of
https://github.com/tile-ai/tilelang.git
synced 2026-10-25 07:14:16 +08:00
48a8f4a6ce
* [Feature] Add new hierarchical reduce with less workspace size and new ptx * Lint * Add intrinsic functions and more types * Fix sm100a redux instruction * Fix data type * [Test] Add tests for reduce operations * [Refactor] Unify AllReduce barrier mechanism via policy template Replace the duplicated run()/run_hopper() methods in AllReduce with a single run() that accepts a Barrier policy template parameter (SyncThreadsBarrier or NamedBarrier<N>). Extract the shared inter-warp reduction logic into a warp_inter_reduce helper, and split the dispatch into private hierarchical_reduce/butterfly_reduce methods. Update codegen in reduce.cc and finalize_reducer.cc to emit NamedBarrier<all_threads> for SM >= 90 targets instead of the old all_threads + run_hopper pattern. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * Add handling for non-float types in warp_reduce function This update introduces a new case in the warp_reduce function to cast non-float types to float using static_cast. This enhancement improves type flexibility and ensures compatibility with a broader range of input types during reduction operations. * [Refactor] Remove hierarchical reduce because of performance regression * [BugFix] Buffer for reduction --------- Co-authored-by: LeiWang1999 <leiwang1999@outlook.com> Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
302 lines
8.5 KiB
Python
302 lines
8.5 KiB
Python
from tilelang import tvm as tvm
|
|
import tilelang.testing
|
|
import tilelang as tl
|
|
import tilelang.language as T
|
|
|
|
tilelang.testing.set_random_seed()
|
|
tilelang.disable_cache()
|
|
|
|
|
|
def _make_shared_reduce(M, N, dtype, reduce_cb):
|
|
@T.prim_func
|
|
def main(
|
|
A: T.Tensor((M, N), dtype),
|
|
B: T.Tensor((M,), dtype),
|
|
):
|
|
with T.Kernel(1) as _:
|
|
A_shared = T.alloc_shared((M, N), dtype)
|
|
B_shared = T.alloc_shared((M,), dtype)
|
|
|
|
T.copy(A, A_shared)
|
|
reduce_cb(T, A_shared, B_shared)
|
|
T.copy(B_shared, B)
|
|
|
|
return main
|
|
|
|
|
|
def _run_program(program, ref_program, atol=1e-2, rtol=1e-2):
|
|
jit_kernel = tl.compile(program, out_idx=-1)
|
|
profiler = jit_kernel.get_profiler()
|
|
profiler.assert_allclose(ref_program, atol=atol, rtol=rtol)
|
|
|
|
|
|
def reduce_test(M, N, dtype=T.float16, op="sum", threads=32):
|
|
import tilelang.language as T
|
|
|
|
@T.prim_func
|
|
def main(
|
|
A: T.Tensor((M, N), dtype),
|
|
B: T.Tensor((M,), dtype),
|
|
):
|
|
with T.Kernel(1, threads=threads) as _:
|
|
A_local = T.alloc_fragment((M, N), dtype)
|
|
B_local = T.alloc_fragment((M,), dtype)
|
|
|
|
T.copy(A, A_local)
|
|
if op == "sum":
|
|
T.reduce_sum(A_local, B_local, dim=1)
|
|
elif op == "max":
|
|
T.reduce_max(A_local, B_local, dim=1)
|
|
elif op == "min":
|
|
T.reduce_min(A_local, B_local, dim=1)
|
|
elif op == "abssum":
|
|
T.reduce_abssum(A_local, B_local, dim=1)
|
|
elif op == "absmax":
|
|
T.reduce_absmax(A_local, B_local, dim=1)
|
|
elif op == "bitand":
|
|
T.reduce_bitand(A_local, B_local, dim=1)
|
|
elif op == "bitor":
|
|
T.reduce_bitor(A_local, B_local, dim=1)
|
|
elif op == "bitxor":
|
|
T.reduce_bitxor(A_local, B_local, dim=1)
|
|
T.copy(B_local, B)
|
|
|
|
return main
|
|
|
|
|
|
def reduce_sum_ss(M, N, dtype=T.float32):
|
|
return _make_shared_reduce(M, N, dtype, lambda T, src, dst: T.reduce_sum(src, dst, dim=1))
|
|
|
|
|
|
def reduce_max_ss(M, N, dtype=T.float32):
|
|
return _make_shared_reduce(M, N, dtype, lambda T, src, dst: T.reduce_max(src, dst, dim=1))
|
|
|
|
|
|
def reduce_min_ss(M, N, dtype=T.float32):
|
|
return _make_shared_reduce(M, N, dtype, lambda T, src, dst: T.reduce_min(src, dst, dim=1))
|
|
|
|
|
|
def reduce_abssum_ss(M, N, dtype=T.float32):
|
|
return _make_shared_reduce(M, N, dtype, lambda T, src, dst: T.reduce_abssum(src, dst, dim=1))
|
|
|
|
|
|
def reduce_absmax_ss(M, N, dtype=T.float32):
|
|
return _make_shared_reduce(M, N, dtype, lambda T, src, dst: T.reduce_absmax(src, dst, dim=1))
|
|
|
|
|
|
def run_reduce(M, N, dtype=T.float32, op="sum", mode="rr", threads=32):
|
|
if mode == "rr":
|
|
program = reduce_test(M, N, dtype, op, threads)
|
|
elif mode == "ss":
|
|
assert op == "sum", f"shared reduce only supports sum, got {op}"
|
|
program = reduce_sum_ss(M, N, dtype)
|
|
else:
|
|
raise NotImplementedError(f"run_reduce only supports rr and ss, got {mode}")
|
|
|
|
import torch
|
|
|
|
def ref_fn(A):
|
|
if op == "sum":
|
|
res = A.sum(dim=1)
|
|
elif op == "max":
|
|
res = A.max(dim=1).values
|
|
elif op == "min":
|
|
res = A.min(dim=1).values
|
|
elif op == "abssum":
|
|
res = A.abs().sum(dim=1)
|
|
elif op == "absmax":
|
|
res = A.abs().max(dim=1).values
|
|
if A.dtype in [torch.uint32, torch.int32, torch.int64]:
|
|
return res.to(A.dtype)
|
|
return res
|
|
|
|
_run_program(program, ref_fn)
|
|
|
|
|
|
def run_shared_reduce(program_builder, ref_program, M, N, dtype=T.float32):
|
|
program = program_builder(M, N, dtype)
|
|
_run_program(program, ref_program)
|
|
|
|
|
|
def run_reduce_max(M, N, dtype=T.float16):
|
|
program = reduce_test(M, N, dtype, "max")
|
|
_run_program(program, lambda A: A.max(dim=1).values, atol=1e-2, rtol=1e-2)
|
|
|
|
|
|
def test_reduce_sum():
|
|
MN_zip = [(256, 256), (512, 128), (128, 512)]
|
|
for dtype in [T.float32, T.int32, T.int64]:
|
|
for M, N in MN_zip:
|
|
run_reduce(M, N, dtype, "sum")
|
|
|
|
|
|
def test_reduce_other_op():
|
|
MN_zip = [(256, 256), (512, 128)]
|
|
for op in ["max", "min", "abssum", "absmax"]:
|
|
for dtype in [T.float32, T.int32, T.int64]:
|
|
for M, N in MN_zip:
|
|
run_reduce(M, N, dtype, op)
|
|
|
|
|
|
def test_reduce_sum_threads():
|
|
run_reduce(32, 32, T.float32, "sum", mode="rr", threads=16)
|
|
run_reduce(16, 16, T.float32, "sum", mode="rr", threads=8)
|
|
|
|
|
|
def test_reduce_sum_shared():
|
|
run_reduce(64, 64, op="sum", mode="ss")
|
|
|
|
|
|
def test_reduce_max():
|
|
run_reduce_max(256, 256, T.float16)
|
|
run_reduce_max(512, 128, T.float16)
|
|
run_reduce_max(256, 256, T.float32)
|
|
|
|
|
|
def test_reduce_max_shared():
|
|
run_shared_reduce(reduce_max_ss, lambda A: A.max(dim=1).values, 64, 64, T.float32)
|
|
|
|
|
|
def test_reduce_min_shared():
|
|
run_shared_reduce(reduce_min_ss, lambda A: A.min(dim=1).values, 64, 64, T.float32)
|
|
|
|
|
|
def test_reduce_abssum_shared():
|
|
run_shared_reduce(reduce_abssum_ss, lambda A: A.abs().sum(dim=1), 64, 64, T.float32)
|
|
|
|
|
|
def test_reduce_absmax_shared():
|
|
run_shared_reduce(reduce_absmax_ss, lambda A: A.abs().max(dim=1).values, 64, 64, T.float32)
|
|
|
|
|
|
def reduce_sum_test_clear(M, N, dtype=T.float32):
|
|
import tilelang.language as T
|
|
|
|
@T.prim_func
|
|
def main(
|
|
A: T.Tensor((M, N), dtype),
|
|
B: T.Tensor((M,), dtype),
|
|
):
|
|
with T.Kernel(1, threads=32) as _:
|
|
A_local = T.alloc_fragment((M, N), dtype)
|
|
B_local = T.alloc_fragment((M,), dtype)
|
|
|
|
T.copy(A, A_local)
|
|
T.fill(B_local, 1)
|
|
T.reduce_sum(A_local, B_local, dim=1, clear=False)
|
|
T.copy(B_local, B)
|
|
|
|
return main
|
|
|
|
|
|
def run_reduce_sum_clear(M, N, dtype=T.float32, tl_func=reduce_sum_test_clear):
|
|
program = tl_func(M, N, dtype)
|
|
jit_kernel = tl.compile(program, out_idx=-1)
|
|
|
|
def ref_program(A):
|
|
return A.sum(dim=1) + 1
|
|
|
|
import torch
|
|
|
|
dummy_A = torch.randn((M, N), dtype=getattr(torch, dtype)).cuda()
|
|
ref_out = ref_program(dummy_A)
|
|
tl_out = jit_kernel(dummy_A)
|
|
torch.testing.assert_close(tl_out, ref_out, atol=1e-2, rtol=1e-2)
|
|
|
|
|
|
def test_reduce_sum_clear():
|
|
run_reduce_sum_clear(256, 256, T.float32)
|
|
run_reduce_sum_clear(512, 128, T.float32)
|
|
run_reduce_sum_clear(128, 512, T.float32)
|
|
|
|
|
|
def reduce_max_test_clear(M, N, dtype=T.float16):
|
|
import tilelang.language as T
|
|
|
|
@T.prim_func
|
|
def main(
|
|
A: T.Tensor((M, N), dtype),
|
|
B: T.Tensor((M,), dtype),
|
|
):
|
|
with T.Kernel(1, threads=32) as _:
|
|
A_local = T.alloc_fragment((M, N), dtype)
|
|
B_local = T.alloc_fragment((M,), dtype)
|
|
|
|
T.copy(A, A_local)
|
|
T.fill(B_local, -T.infinity(dtype))
|
|
T.reduce_max(A_local, B_local, dim=1, clear=False)
|
|
T.copy(B_local, B)
|
|
|
|
return main
|
|
|
|
|
|
def run_reduce_max_clear(M, N, dtype=T.float16):
|
|
program = reduce_max_test_clear(M, N, dtype)
|
|
jit_kernel = tl.compile(program, out_idx=-1)
|
|
|
|
def ref_program(A):
|
|
return A.max(dim=1).values
|
|
|
|
import torch
|
|
|
|
dummy_A = torch.randn((M, N), dtype=getattr(torch, dtype)).cuda()
|
|
ref_out = ref_program(dummy_A)
|
|
tl_out = jit_kernel(dummy_A)
|
|
torch.testing.assert_close(tl_out, ref_out, atol=1e-2, rtol=1e-2)
|
|
|
|
|
|
def test_reduce_max_clear():
|
|
run_reduce_max_clear(256, 256, T.float16)
|
|
|
|
|
|
def reduce_sum_test_clear_B_shared(M, N, dtype=T.float32):
|
|
import tilelang.language as T
|
|
|
|
@T.prim_func
|
|
def main(
|
|
A: T.Tensor((M, N), dtype),
|
|
B: T.Tensor((M,), dtype),
|
|
):
|
|
with T.Kernel(1, threads=32) as _:
|
|
A_local = T.alloc_fragment((M, N), dtype)
|
|
B_shared = T.alloc_shared((M,), dtype)
|
|
|
|
T.copy(A, A_local)
|
|
T.fill(B_shared, 1)
|
|
T.reduce_sum(A_local, B_shared, dim=1, clear=False)
|
|
T.copy(B_shared, B)
|
|
|
|
return main
|
|
|
|
|
|
def test_reduce_sum_clear_B_shared():
|
|
run_reduce_sum_clear(256, 256, T.float32, reduce_sum_test_clear_B_shared)
|
|
|
|
|
|
def reduce_sum_test_clear_AB_shared(M, N, dtype=T.float32):
|
|
import tilelang.language as T
|
|
|
|
@T.prim_func
|
|
def main(
|
|
A: T.Tensor((M, N), dtype),
|
|
B: T.Tensor((M,), dtype),
|
|
):
|
|
with T.Kernel(1, threads=32) as _:
|
|
A_shared = T.alloc_shared((M, N), dtype)
|
|
B_shared = T.alloc_shared((M,), dtype)
|
|
|
|
T.copy(A, A_shared, disable_tma=True)
|
|
T.fill(B_shared, 1)
|
|
T.reduce_sum(A_shared, B_shared, dim=1, clear=False)
|
|
T.copy(B_shared, B)
|
|
|
|
return main
|
|
|
|
|
|
def test_reduce_sum_clear_AB_shared():
|
|
run_reduce_sum_clear(64, 64, T.float32, reduce_sum_test_clear_AB_shared)
|
|
|
|
|
|
if __name__ == "__main__":
|
|
tilelang.testing.main()
|