[Kernel] Implement different SEQ Q/KV examples with block sparse (#133)

* Change default log level from WARNING to INFO in TileLang initialization

* Refactor Flash Attention Variable-Length MHA Example with Cython Backend Support

- Update `example_mha_fwd_varlen.py` to use Cython backend for kernel compilation
- Remove unused imports and simplify function signature
- Modify `flashattn` function to handle max sequence length as a separate argument
- Update kernel call to include max sequence length parameter
- Improve code readability and remove commented-out code
- Add print statement to confirm successful assertion

* Refactor code formatting in TileLang lowering and example files

- Improve line breaks and code formatting in `lower.py`, `wrapper.py`, and `tensor.py`
- Simplify line breaks and reduce unnecessary whitespace
- Enhance code readability by adjusting indentation and line breaks
- Update example MHA forward pass script with cleaner tensor initialization

* Update TileLang kernel test with import path changes for MMA layout and macro generator

- Modify import statements in test_tilelang_kernel_dequantize_gemm.py
- Replace bitblas imports with tilelang.intrinsics imports for MMA-related utilities
- Update main function to use tilelang.testing.main()

* Add Block Sparse Attention Examples for TileLang and Triton

- Implement block sparse attention kernels for both TileLang and Triton
- Add utility functions for generating sparse attention masks using top-k and threshold methods
- Support causal and variable-length attention scenarios
- Include test cases for different sequence length configurations
- Demonstrate block-level sparse attention with configurable parameters

* Refactor Block Sparse Attention Examples with Code Style Improvements

- Improve code formatting in block_sparse_attn_tilelang.py and block_sparse_attn_triton.py
- Enhance readability by adjusting line breaks and indentation
- Simplify kernel and function calls with better formatting
- Add whitespace and line break improvements for better code clarity

examples/flash_attention/example_mha_fwd_bhsd.py

@@ -4,7 +4,6 @@
 import torch
 import torch.nn.functional as F
 import tilelang
-from tilelang import Profiler
 from tilelang.autotuner import *
 import tilelang.language as T
 import itertools
@@ -28,9 +27,10 @@ def get_configs():
     return configs
 
 
-def flashattn(batch, heads, seq_len, dim, is_causal, tune=False):
+def flashattn(batch, heads, seq_q, seq_kv, dim, is_causal, tune=False):
     scale = (1.0 / dim)**0.5 * 1.44269504  # log2(e)
-    shape = [batch, heads, seq_len, dim]
+    q_shape = [batch, heads, seq_q, dim]
+    kv_shape = [batch, heads, seq_kv, dim]
     dtype = "float16"
     accum_dtype = "float"
 
@@ -38,7 +38,7 @@ def flashattn(batch, heads, seq_len, dim, is_causal, tune=False):
 
         @T.macro
         def MMA0(
-            K: T.Buffer(shape, dtype),
+            K: T.Buffer(kv_shape, dtype),
             Q_shared: T.Buffer([block_M, dim], dtype),
             K_shared: T.Buffer([block_N, dim], dtype),
             acc_s: T.Buffer([block_M, block_N], accum_dtype),
@@ -47,18 +47,20 @@ def flashattn(batch, heads, seq_len, dim, is_causal, tune=False):
             by: T.int32,
             bz: T.int32,
         ):
+            past_len = seq_kv - seq_q
             T.copy(K[bz, by, k * block_N:(k + 1) * block_N, :], K_shared)
             if is_causal:
                 for i, j in T.Parallel(block_M, block_N):
-                    acc_s[i, j] = T.if_then_else(bx * block_M + i >= k * block_N + j, 0,
-                                                 -T.infinity(acc_s.dtype))
+                    q_idx = bx * block_M + i + past_len
+                    k_idx = k * block_N + j
+                    acc_s[i, j] = T.if_then_else(q_idx >= k_idx, 0, -T.infinity(acc_s.dtype))
             else:
                 T.clear(acc_s)
             T.gemm(Q_shared, K_shared, acc_s, transpose_B=True, policy=T.GemmWarpPolicy.FullRow)
 
         @T.macro
         def MMA1(
-                V: T.Buffer(shape, dtype),
+                V: T.Buffer(kv_shape, dtype),
                 V_shared: T.Buffer([block_M, dim], dtype),
                 acc_s_cast: T.Buffer([block_M, block_N], dtype),
                 acc_o: T.Buffer([block_M, dim], accum_dtype),
@@ -89,6 +91,7 @@ def flashattn(batch, heads, seq_len, dim, is_causal, tune=False):
             #     scores_max[i] = T.if_then_else(scores_max[i] == -T.infinity(accum_dtype), 0, scores_max[i])
             for i in T.Parallel(block_M):
                 scores_scale[i] = T.exp2(scores_max_prev[i] * scale - scores_max[i] * scale)
+
             for i, j in T.Parallel(block_M, block_N):
                 # Instead of computing exp(x - max), we compute exp2(x * log_2(e) -
                 # max * log_2(e)) This allows the compiler to use the ffma
@@ -109,13 +112,12 @@ def flashattn(batch, heads, seq_len, dim, is_causal, tune=False):
 
         @T.prim_func
         def main(
-                Q: T.Buffer(shape, dtype),
-                K: T.Buffer(shape, dtype),
-                V: T.Buffer(shape, dtype),
-                Output: T.Buffer(shape, dtype),
+                Q: T.Buffer(q_shape, dtype),
+                K: T.Buffer(kv_shape, dtype),
+                V: T.Buffer(kv_shape, dtype),
+                Output: T.Buffer(q_shape, dtype),
         ):
-            with T.Kernel(
-                    T.ceildiv(seq_len, block_M), heads, batch, threads=threads) as (bx, by, bz):
+            with T.Kernel(T.ceildiv(seq_q, block_M), heads, batch, threads=threads) as (bx, by, bz):
                 Q_shared = T.alloc_shared([block_M, dim], dtype)
                 K_shared = T.alloc_shared([block_N, dim], dtype)
                 V_shared = T.alloc_shared([block_N, dim], dtype)
@@ -135,8 +137,8 @@ def flashattn(batch, heads, seq_len, dim, is_causal, tune=False):
                 T.fill(scores_max, -T.infinity(accum_dtype))
 
                 loop_range = (
-                    T.min(T.ceildiv(seq_len, block_N), T.ceildiv(
-                        (bx + 1) * block_M, block_N)) if is_causal else T.ceildiv(seq_len, block_N))
+                    T.min(T.ceildiv(seq_kv, block_N), T.ceildiv(
+                        (bx + 1) * block_M, block_N)) if is_causal else T.ceildiv(seq_kv, block_N))
 
                 for k in T.Pipelined(loop_range, num_stages=num_stages):
                     MMA0(K, Q_shared, K_shared, acc_s, k, bx, by, bz)
@@ -180,8 +182,9 @@ def ref_program(Q, K, V, is_causal):
     scores = torch.einsum('bhqd,bhkd->bhqk', Q, K)
     scores = scores / torch.sqrt(torch.tensor(dim, dtype=scores.dtype))
     if is_causal:
-        seq_len = Q.size(1)
-        mask = torch.tril(torch.ones(seq_len, seq_len, device=scores.device))
+        seq_q = Q.size(2)
+        seq_kv = K.size(2)
+        mask = torch.tril(torch.ones(seq_q, seq_kv, device=scores.device))
         mask = mask.unsqueeze(0).unsqueeze(0)
         scores = scores.masked_fill(mask == 0, float('-inf'))
     attention_weights = F.softmax(scores, dim=-1)
@@ -191,37 +194,38 @@ def ref_program(Q, K, V, is_causal):
 
 if __name__ == "__main__":
     parser = argparse.ArgumentParser()
-    parser.add_argument('--batch', type=int, default=8, help='batch size')
-    parser.add_argument('--heads', type=int, default=32, help='heads')
-    parser.add_argument('--seq_len', type=int, default=4096, help='sequence length')
-    parser.add_argument('--dim', type=int, default=128, help='dim')
+    parser.add_argument('--batch', type=int, default=1, help='batch size')
+    parser.add_argument('--heads', type=int, default=1, help='heads')
+    parser.add_argument('--seq_q', type=int, default=256, help='query sequence length')
+    parser.add_argument('--seq_kv', type=int, default=256, help='key/value sequence length')
+    parser.add_argument('--dim', type=int, default=64, help='dim')
     parser.add_argument('--is_causal', action='store_true', help='causal')
     parser.add_argument('--tune', action='store_true', help='tune configs')
     args = parser.parse_args()
-    batch, heads, seq_len, dim, is_causal = args.batch, args.heads, args.seq_len, args.dim, args.is_causal
-    flops_per_matmul = 2.0 * batch * heads * seq_len * seq_len * dim
+    batch, heads, seq_q, seq_kv, dim, is_causal = args.batch, args.heads, args.seq_q, args.seq_kv, args.dim, args.is_causal
+    flops_per_matmul = 2.0 * batch * heads * seq_q * seq_kv * dim
     total_flops = 2 * flops_per_matmul
     if is_causal:
         total_flops *= 0.5
 
     if (not args.tune):
         program = flashattn(
-            batch, heads, seq_len, dim, is_causal, tune=args.tune)(
-                block_M=128, block_N=128, num_stages=1, threads=128)
+            batch, heads, seq_q, seq_kv, dim, is_causal, tune=args.tune)(
+                block_M=64, block_N=64, num_stages=0, threads=128)
         ref_program = partial(ref_program, is_causal=is_causal)
-        mod, params = tilelang.lower(program)
-        mod = Profiler(mod, params, [3], tilelang.TensorSupplyType.Normal)
-        mod.assert_allclose(ref_program, rtol=0.01, atol=0.01)
+        kernel = tilelang.compile(program, out_idx=[3])
+        profiler = kernel.get_profiler()
+        profiler.assert_allclose(ref_program, rtol=0.01, atol=0.01)
         print("All checks pass.")
-        latency = mod.do_bench(ref_program, warmup=500)
+        latency = profiler.do_bench(ref_program, warmup=500)
         print("Ref: {:.2f} ms".format(latency))
         print("Ref: {:.2f} TFlops".format(total_flops / latency * 1e-9))
-        latency = mod.do_bench(mod.func, warmup=500)
+        latency = profiler.do_bench(profiler.mod, warmup=500)
         print("Tile-lang: {:.2f} ms".format(latency))
         print("Tile-lang: {:.2f} TFlops".format(total_flops / latency * 1e-9))
     else:
         best_latency, best_config, _ = flashattn(
-            batch, heads, seq_len, dim, is_causal, tune=args.tune)
+            batch, heads, seq_q, seq_kv, dim, is_causal, tune=args.tune)
         print(f"Best latency: {best_latency}")
         print(f"Best TFlops: {total_flops / best_latency * 1e-9}")
         print(f"Best config: {best_config}")

examples/seer_attention/block_sparse_attn_tilelang.py

+# Copyright (c) Microsoft Corporation.
+# Licensed under the MIT License.
+import math
+import torch
+
+import tilelang
+import tilelang.language as T
+import torch.nn.functional as F
+
+
+def get_sparse_attn_mask_from_topk(x, topk, use_dense_for_last_block=False):
+    bsz, num_head, downsample_len, _ = x.shape
+    # N_CTX = downsample_len * BLOCK
+    sparse_index = torch.topk(x, topk, dim=-1).indices
+    dense_mask = torch.full([bsz, num_head, downsample_len, downsample_len],
+                            False,
+                            dtype=torch.bool,
+                            device=x.device)
+    dense_mask.scatter_(-1, sparse_index, True)
+    if use_dense_for_last_block:
+        dense_mask[:, :, -2:, :] = True
+    dense_mask.tril_()
+    return dense_mask
+
+
+def get_sparse_attn_mask_from_threshold(x, threshold, use_dense_for_last_block=False):
+    dense_mask = x > threshold
+    if use_dense_for_last_block:
+        dense_mask[:, :, -2:, :] = True
+    dense_mask.tril_()
+    return dense_mask
+
+
+def blocksparse_flashattn(batch, heads, seq_q, seq_kv, dim, downsample_len, is_causal):
+    block_M = 64
+    block_N = 64
+    num_stages = 0
+    threads = 128
+    scale = (1.0 / dim)**0.5 * 1.44269504  # log2(e)
+    q_shape = [batch, heads, seq_q, dim]
+    kv_shape = [batch, heads, seq_kv, dim]
+    block_mask_shape = [batch, heads, downsample_len, downsample_len]
+
+    dtype = "float16"
+    accum_dtype = "float"
+    block_mask_dtype = "int8"
+
+    def kernel_func(block_M, block_N, num_stages, threads):
+
+        @T.macro
+        def Softmax(
+                acc_s: T.Buffer([block_M, block_N], accum_dtype),
+                acc_s_cast: T.Buffer([block_M, block_N], dtype),
+                scores_max: T.Buffer([block_M], accum_dtype),
+                scores_max_prev: T.Buffer([block_M], accum_dtype),
+                scores_scale: T.Buffer([block_M], accum_dtype),
+                scores_sum: T.Buffer([block_M], accum_dtype),
+                logsum: T.Buffer([block_M], accum_dtype),
+        ):
+            T.copy(scores_max, scores_max_prev)
+            T.fill(scores_max, -T.infinity(accum_dtype))
+            T.reduce_max(acc_s, scores_max, dim=1, clear=False)
+            # To do causal softmax, we need to set the scores_max to 0 if it is -inf
+            # This process is called Check_inf in FlashAttention3 code, and it only need to be done
+            # in the first ceil_div(kBlockM, kBlockN) steps.
+            # for i in T.Parallel(block_M):
+            #     scores_max[i] = T.if_then_else(scores_max[i] == -T.infinity(accum_dtype), 0, scores_max[i])
+            for i in T.Parallel(block_M):
+                scores_scale[i] = T.exp2(scores_max_prev[i] * scale - scores_max[i] * scale)
+            for i, j in T.Parallel(block_M, block_N):
+                # Instead of computing exp(x - max), we compute exp2(x * log_2(e) -
+                # max * log_2(e)) This allows the compiler to use the ffma
+                # instruction instead of fadd and fmul separately.
+                acc_s[i, j] = T.exp2(acc_s[i, j] * scale - scores_max[i] * scale)
+            T.reduce_sum(acc_s, scores_sum, dim=1)
+            for i in T.Parallel(block_M):
+                logsum[i] = logsum[i] * scores_scale[i] + scores_sum[i]
+            T.copy(acc_s, acc_s_cast)
+
+        @T.macro
+        def Rescale(
+                acc_o: T.Buffer([block_M, dim], accum_dtype),
+                scores_scale: T.Buffer([block_M], accum_dtype),
+        ):
+            for i, j in T.Parallel(block_M, dim):
+                acc_o[i, j] *= scores_scale[i]
+
+        @T.prim_func
+        def main(
+                Q: T.Buffer(q_shape, dtype),
+                K: T.Buffer(kv_shape, dtype),
+                V: T.Buffer(kv_shape, dtype),
+                BlockSparseMask: T.Buffer(block_mask_shape, block_mask_dtype),
+                Output: T.Buffer(q_shape, dtype),
+        ):
+            with T.Kernel(T.ceildiv(seq_q, block_M), heads, batch, threads=threads) as (bx, by, bz):
+                Q_shared = T.alloc_shared([block_M, dim], dtype)
+                K_shared = T.alloc_shared([block_N, dim], dtype)
+                V_shared = T.alloc_shared([block_N, dim], dtype)
+                O_shared = T.alloc_shared([block_M, dim], dtype)
+                acc_s = T.alloc_fragment([block_M, block_N], accum_dtype)
+                acc_s_cast = T.alloc_fragment([block_M, block_N], dtype)
+                acc_o = T.alloc_fragment([block_M, dim], accum_dtype)
+                scores_max = T.alloc_fragment([block_M], accum_dtype)
+                scores_max_prev = T.alloc_fragment([block_M], accum_dtype)
+                scores_scale = T.alloc_fragment([block_M], accum_dtype)
+                scores_sum = T.alloc_fragment([block_M], accum_dtype)
+                logsum = T.alloc_fragment([block_M], accum_dtype)
+                block_mask = T.alloc_local([downsample_len], block_mask_dtype)
+
+                T.copy(Q[bz, by, bx * block_M:(bx + 1) * block_M, :], Q_shared)
+                T.fill(acc_o, 0)
+                T.fill(logsum, 0)
+                T.fill(scores_max, -T.infinity(accum_dtype))
+
+                for vj in T.serial(downsample_len):
+                    block_mask[vj] = BlockSparseMask[bz, by, bx, vj]
+
+                loop_range = T.ceildiv(seq_kv, block_N)
+
+                for k in T.Pipelined(loop_range, num_stages=num_stages):
+                    if block_mask[k] != 0:
+                        T.copy(K[bz, by, k * block_N:(k + 1) * block_N, :], K_shared)
+                        if is_causal:
+                            past_len = seq_kv - seq_q
+                            for i, j in T.Parallel(block_M, block_N):
+                                acc_s[i, j] = T.if_then_else(
+                                    bx * block_M + i + past_len >= k * block_N + j, 0,
+                                    -T.infinity(acc_s.dtype))
+                        else:
+                            T.clear(acc_s)
+                        T.gemm(
+                            Q_shared,
+                            K_shared,
+                            acc_s,
+                            transpose_B=True,
+                            policy=T.GemmWarpPolicy.FullRow)
+
+                        Softmax(acc_s, acc_s_cast, scores_max, scores_max_prev, scores_scale,
+                                scores_sum, logsum)
+                        Rescale(acc_o, scores_scale)
+                        T.copy(V[bz, by, k * block_N:(k + 1) * block_N, :], V_shared)
+                        T.gemm(acc_s_cast, V_shared, acc_o, policy=T.GemmWarpPolicy.FullRow)
+
+                for i, j in T.Parallel(block_M, dim):
+                    acc_o[i, j] /= logsum[i]
+
+                T.copy(acc_o, O_shared)
+                T.copy(O_shared, Output[bz, by, bx * block_M:(bx + 1) * block_M, :])
+
+        return main
+
+    return kernel_func(block_M, block_N, num_stages, threads)
+
+
+def test_topk_sparse_attention():
+    # Config
+    BATCH, N_HEADS, SEQ_LEN, D_HEAD = 4, 2, 256, 64
+    TOPK = 2  # Keep top 8 elements per row
+    BLOCK = 64
+    torch.manual_seed(0)
+
+    # Create inputs
+    q = torch.randn(BATCH, N_HEADS, SEQ_LEN, D_HEAD, device='cuda', dtype=torch.float16)
+    k = torch.randn(BATCH, N_HEADS, SEQ_LEN, D_HEAD, device='cuda', dtype=torch.float16)
+    v = torch.randn(BATCH, N_HEADS, SEQ_LEN, D_HEAD, device='cuda', dtype=torch.float16)
+
+    sm_scale = 1.0 / (D_HEAD**0.5)
+
+    # Create sparse mask (downsampled to block level)
+    downsample_factor = BLOCK
+    downsample_len = math.ceil(SEQ_LEN / downsample_factor)
+    x_ds = torch.randn([BATCH, N_HEADS, downsample_len, downsample_len],
+                       device='cuda',
+                       dtype=torch.float16)
+    x_ds[:, :, :, 0] = 100
+    block_mask = get_sparse_attn_mask_from_topk(x_ds, topk=TOPK)
+
+    # Run Triton kernel
+    program = blocksparse_flashattn(
+        BATCH, N_HEADS, SEQ_LEN, SEQ_LEN, D_HEAD, downsample_len, is_causal=True)
+    kernel = tilelang.compile(program, out_idx=[4])
+    print(kernel.get_kernel_source())
+    tilelang_output = kernel(q, k, v, block_mask)
+
+    # Compute reference
+    # Expand block mask to full attention matrix
+    full_mask = torch.kron(block_mask.float(), torch.ones(BLOCK, BLOCK, device='cuda'))
+    full_mask = full_mask[..., :SEQ_LEN, :SEQ_LEN].bool()
+    full_mask = full_mask & torch.tril(torch.ones_like(full_mask))  # Apply causal
+
+    # PyTorch reference implementation
+    attn = torch.einsum('bhsd,bhtd->bhst', q, k) * sm_scale
+    attn = attn.masked_fill(~full_mask, float('-inf'))
+    attn = F.softmax(attn, dim=-1)
+    ref_output = torch.einsum('bhst,bhtd->bhsd', attn, v)
+
+    print("ref_output", ref_output)
+    print("tilelang_output", tilelang_output)
+
+    # Verify accuracy
+    assert torch.allclose(tilelang_output, ref_output, atol=1e-2, rtol=1e-2), \
+        "TileLang output doesn't match reference"
+    print("Pass topk sparse attention test with qlen == klen")
+
+
+def test_topk_sparse_attention_qlen_lt_klen():
+    # Config
+    BATCH, N_HEADS = 1, 1
+    Q_LEN, K_LEN, D_HEAD = 128, 256, 64  # qlen < klen; here, past_len = 256 - 128 = 128.
+    TOPK = 1
+    BLOCK = 64  # block size used in downsampling
+    torch.manual_seed(0)
+
+    # Create inputs.
+    q = torch.randn(BATCH, N_HEADS, Q_LEN, D_HEAD, device='cuda', dtype=torch.float16)
+    k = torch.randn(BATCH, N_HEADS, K_LEN, D_HEAD, device='cuda', dtype=torch.float16)
+    v = torch.randn(BATCH, N_HEADS, K_LEN, D_HEAD, device='cuda', dtype=torch.float16)
+    sm_scale = 1.0 / (D_HEAD**0.5)
+
+    downsample_factor = BLOCK
+    downsample_len = math.ceil(K_LEN / downsample_factor)  # number of blocks along one dimension
+    x_ds = torch.randn(
+        BATCH, N_HEADS, downsample_len, downsample_len, device='cuda', dtype=torch.float16)
+    # Force the first column to be high so that the first block is always selected.
+    x_ds[:, :, :, 0] = 100
+    block_mask = get_sparse_attn_mask_from_topk(x_ds, topk=TOPK)
+
+    program = blocksparse_flashattn(
+        BATCH, N_HEADS, Q_LEN, K_LEN, D_HEAD, downsample_len, is_causal=True)
+    print(program)
+    kernel = tilelang.compile(program, out_idx=[4])
+    print(kernel.get_kernel_source())
+    tilelang_output = kernel(q, k, v, block_mask)
+
+    # import flash_attn
+
+    # ref_out = flash_attn.flash_attn_func(q, k, v, causal=True)
+    # torch.testing.assert_close(tilelang_output, ref_out, atol=1e-2, rtol=1e-2)
+    # exit()
+
+    past_len = K_LEN - Q_LEN
+
+    attn = torch.einsum('bhsd,bhtd->bhst', q, k) * sm_scale
+
+    full_mask_full = torch.kron(block_mask.float(), torch.ones(BLOCK, BLOCK, device='cuda')).bool()
+    full_mask_full = full_mask_full[..., :K_LEN, :K_LEN]
+
+    effective_mask = full_mask_full[..., past_len:K_LEN, :]  # shape: (B, H, Q_LEN, K_LEN)
+
+    i_global = torch.arange(past_len, K_LEN, device=k.device).unsqueeze(1)  # shape: (Q_LEN, 1)
+    j_global = torch.arange(K_LEN, device=k.device).unsqueeze(0)  # shape: (1, K_LEN)
+    causal_mask = (j_global <= i_global)  # shape: (Q_LEN, K_LEN)
+
+    final_mask = effective_mask & causal_mask  # shape: (B, H, Q_LEN, K_LEN)
+
+    attn = attn.masked_fill(~final_mask, float('-inf'))
+    attn = F.softmax(attn, dim=-1)
+    ref_output = torch.einsum('bhst,bhtd->bhsd', attn, v)
+
+    print("ref_output", ref_output)
+    print("tilelang_output", tilelang_output)
+
+    # Verify accuracy.
+    torch.testing.assert_close(tilelang_output, ref_output, atol=1e-2, rtol=1e-2)
+
+    print("Pass topk sparse attention test with qlen < klen")
+
+
+if __name__ == "__main__":
+    # test_topk_sparse_attention()
+    test_topk_sparse_attention_qlen_lt_klen()

examples/seer_attention/block_sparse_attn_triton.py

+# Copyright (c) Microsoft Corporation.
+# Licensed under the MIT License.
+# ruff: noqa: E712
+import math
+import torch
+
+import triton
+import triton.language as tl
+import torch.nn.functional as F
+
+
+def is_hip():
+    return triton.runtime.driver.active.get_current_target().backend == "hip"
+
+
+def get_sparse_attn_mask_from_topk(x, topk, use_dense_for_last_block=False):
+    bsz, num_head, downsample_len, _ = x.shape
+    # N_CTX = downsample_len * BLOCK
+    sparse_index = torch.topk(x, topk, dim=-1).indices
+    dense_mask = torch.full([bsz, num_head, downsample_len, downsample_len],
+                            False,
+                            dtype=torch.bool,
+                            device=x.device)
+    dense_mask.scatter_(-1, sparse_index, True)
+    if use_dense_for_last_block:
+        dense_mask[:, :, -2:, :] = True
+    dense_mask.tril_()
+    return dense_mask
+
+
+def get_sparse_attn_mask_from_threshold(x, threshold, use_dense_for_last_block=False):
+    dense_mask = x > threshold
+    if use_dense_for_last_block:
+        dense_mask[:, :, -2:, :] = True
+    dense_mask.tril_()
+    return dense_mask
+
+
+@triton.jit
+def _fwd_kernel_inner(
+    acc,
+    l_i,
+    m_i,
+    q,
+    k_block_col_idx,
+    block_mask_ptr,
+    k_ptrs,
+    v_ptrs,
+    offs_m,
+    offs_n,
+    stride_kt,
+    stride_vt,
+    stride_bmask_n,
+    sm_scale,
+    past_len,
+    BLOCK_M: tl.constexpr,
+    BLOCK_N: tl.constexpr,
+):
+
+    mask_val = tl.load(block_mask_ptr + k_block_col_idx * stride_bmask_n)
+
+    if mask_val == True:
+        start_n = k_block_col_idx * BLOCK_N
+        # -- compute qk ----
+
+        k = tl.load(k_ptrs + start_n * stride_kt)
+
+        qk = tl.zeros([BLOCK_M, BLOCK_N], dtype=tl.float32)
+        qk += tl.dot(q, k)
+
+        qk *= sm_scale
+
+        # the following is needed only when LAST_K_BLOCK or BLOCK_M < BLOCK_N
+        qk += tl.where(offs_m[:, None] + past_len >= (start_n + offs_n[None, :]), 0, float('-inf'))
+
+        m_ij = tl.maximum(m_i, tl.max(qk, 1))
+        qk -= m_ij[:, None]
+        p = tl.exp(qk)
+        l_ij = tl.sum(p, 1)
+        alpha = tl.exp(m_i - m_ij)
+        l_i = l_i * alpha + l_ij
+        acc = acc * alpha[:, None]
+
+        # update acc
+        v = tl.load(v_ptrs + start_n * stride_vt)
+
+        p = p.to(v.type.element_ty)
+
+        acc += tl.dot(p, v)
+        # update m_i and l_i
+        m_i = m_ij
+    return acc, l_i, m_i
+
+
+@triton.jit
+def _fwd_kernel(
+    Q,
+    K,
+    V,
+    sm_scale,
+    block_mask_ptr,
+    Out,
+    stride_qz,
+    stride_qh,
+    stride_qm,
+    stride_qd,
+    stride_kz,
+    stride_kh,
+    stride_kn,
+    stride_kd,
+    stride_vz,
+    stride_vh,
+    stride_vn,
+    stride_vd,
+    stride_bmz,
+    stride_bmh,
+    stride_bmm,
+    stride_bmn,
+    stride_oz,
+    stride_oh,
+    stride_om,
+    stride_od,
+    H,
+    N_CTX,
+    PAST_LEN,
+    BLOCK_M: tl.constexpr,
+    BLOCK_N: tl.constexpr,
+    BLOCK_DMODEL: tl.constexpr,
+):
+    Q_LEN = N_CTX - PAST_LEN
+    start_m = tl.program_id(0)
+    off_hz = tl.program_id(1)
+    off_h = off_hz % H
+    off_z = off_hz // H
+    Q += off_z * stride_qz + off_h * stride_qh
+    K += off_z * stride_kz + off_h * stride_kh
+    V += off_z * stride_vz + off_h * stride_vh
+    block_mask_ptr += off_z * stride_bmz + off_h * stride_bmh
+
+    # initialize offsets
+    offs_m = start_m * BLOCK_M + tl.arange(0, BLOCK_M)
+    offs_n = tl.arange(0, BLOCK_N)
+    offs_d = tl.arange(0, BLOCK_DMODEL)
+    off_q = offs_m[:, None] * stride_qm + offs_d[None, :] * stride_qd
+    # off_k = offs_n[:, None] * stride_kn + offs_d[None, :] * stride_kd
+    off_k = offs_n[None, :] * stride_kn + offs_d[:, None] * stride_kd
+    off_v = offs_n[:, None] * stride_vn + offs_d[None, :] * stride_vd
+    # Initialize pointers to Q, K, V
+    q_ptrs = Q + off_q
+    k_ptrs = K + off_k
+    v_ptrs = V + off_v
+    mask_ptrs = block_mask_ptr + start_m * stride_bmm
+
+    m_i = tl.zeros([BLOCK_M], dtype=tl.float32) - float('inf')
+    l_i = tl.zeros([BLOCK_M], dtype=tl.float32)
+    acc = tl.zeros([BLOCK_M, BLOCK_DMODEL], dtype=tl.float32)
+
+    q = tl.load(q_ptrs, mask=offs_m[:, None] < Q_LEN)
+
+    k_block_start = 0
+    k_block_end = tl.cdiv((start_m + 1) * BLOCK_M, BLOCK_N)
+
+    # loop over k, v and update accumulator
+    for col_idx in range(k_block_start, k_block_end):
+        acc, l_i, m_i = _fwd_kernel_inner(
+            acc,
+            l_i,
+            m_i,
+            q,
+            col_idx,
+            mask_ptrs,
+            k_ptrs,
+            v_ptrs,
+            offs_m,
+            offs_n,
+            stride_kn,
+            stride_vn,
+            stride_bmn,
+            sm_scale,
+            PAST_LEN,
+            BLOCK_M,
+            BLOCK_N,
+        )
+
+    m_i += tl.math.log(l_i)
+    l_recip = 1 / l_i[:, None]
+    acc = acc * l_recip
+    acc = acc.to(Out.dtype.element_ty)
+
+    off_o = off_z * stride_oz + off_h * stride_oh + offs_m[:, None] * stride_om + offs_d[
+        None, :] * stride_od
+    out_ptrs = Out + off_o
+    tl.store(out_ptrs, acc, mask=offs_m[:, None] < N_CTX)
+
+
+def _forward(ctx,
+             q,
+             k,
+             v,
+             block_sparse_mask,
+             sm_scale,
+             BLOCK_M=64,
+             BLOCK_N=64,
+             num_warps=None,
+             num_stages=1,
+             out=None):
+
+    assert q.shape[-1] == k.shape[-1] == v.shape[-1]
+    assert k.shape[2] == v.shape[2]
+    o = out if out is not None else torch.empty_like(q).contiguous()
+    grid = (triton.cdiv(q.shape[2], BLOCK_M), q.shape[0] * q.shape[1])
+
+    assert q.shape[-1] in [64, 128]
+    BLOCK_DMODEL = q.shape[-1]
+
+    if is_hip():
+        num_warps, num_stages = 8, 1
+    else:
+        num_warps, num_stages = 4, 2
+
+    N_CTX = k.shape[2]
+    PAST_LEN = N_CTX - q.shape[2]
+    print("PAST_LEN", PAST_LEN)
+    H = q.shape[1]
+
+    _fwd_kernel[grid](
+        q,
+        k,
+        v,
+        sm_scale,
+        block_sparse_mask,
+        o,
+        *q.stride(),
+        *k.stride(),
+        *v.stride(),
+        *block_sparse_mask.stride(),
+        *o.stride(),
+        H,
+        N_CTX,
+        PAST_LEN,
+        BLOCK_M,
+        BLOCK_N,
+        BLOCK_DMODEL,
+        num_warps=num_warps,
+        num_stages=num_stages,
+    )
+
+    return o
+
+
+class _sparse_attention(torch.autograd.Function):
+
+    @staticmethod
+    def forward(ctx, q, k, v, block_sparse_dense, sm_scale):
+        # shape constraints
+        return _forward(ctx, q, k, v, block_sparse_dense, sm_scale)
+
+    @staticmethod
+    def backward(ctx, do):
+        # No gradient propagation.
+        raise NotImplementedError("It does not support gradient propagation yet")
+        return None, None, None, None, None
+
+
+block_sparse_triton_fn = _sparse_attention.apply
+
+
+def test_topk_sparse_attention():
+    # Config
+    BATCH, N_HEADS, SEQ_LEN, D_HEAD = 1, 1, 256, 64
+    TOPK = 2  # Keep top 8 elements per row
+    BLOCK = 64
+    torch.manual_seed(0)
+
+    # Create inputs
+    q = torch.randn(BATCH, N_HEADS, SEQ_LEN, D_HEAD, device='cuda', dtype=torch.bfloat16)
+    k = torch.randn(BATCH, N_HEADS, SEQ_LEN, D_HEAD, device='cuda', dtype=torch.bfloat16)
+    v = torch.randn(BATCH, N_HEADS, SEQ_LEN, D_HEAD, device='cuda', dtype=torch.bfloat16)
+    sm_scale = 1.0 / (D_HEAD**0.5)
+
+    # Create sparse mask (downsampled to block level)
+    downsample_factor = BLOCK
+    downsample_len = math.ceil(SEQ_LEN / downsample_factor)
+    print("downsample_len", downsample_len)
+
+    x_ds = torch.randn([BATCH, N_HEADS, downsample_len, downsample_len],
+                       device='cuda',
+                       dtype=torch.bfloat16)
+    x_ds[:, :, :, 0] = 100
+    print("x_ds.shape", x_ds.shape)
+    block_mask = get_sparse_attn_mask_from_topk(x_ds, topk=TOPK)
+    # print("block_mask", block_mask)
+    print("block_mask.shape", block_mask.shape)
+
+    # Run Triton kernel
+    triton_output = block_sparse_triton_fn(q, k, v, block_mask, sm_scale)
+
+    # Compute reference
+    # Expand block mask to full attention matrix
+    full_mask = torch.kron(block_mask.float(), torch.ones(BLOCK, BLOCK, device='cuda'))
+    full_mask = full_mask[..., :SEQ_LEN, :SEQ_LEN].bool()
+    full_mask = full_mask & torch.tril(torch.ones_like(full_mask))  # Apply causal
+
+    # PyTorch reference implementation
+    attn = torch.einsum('bhsd,bhtd->bhst', q, k) * sm_scale
+    attn = attn.masked_fill(~full_mask, float('-inf'))
+    attn = F.softmax(attn, dim=-1)
+    ref_output = torch.einsum('bhst,bhtd->bhsd', attn, v)
+
+    # print("ref_output", ref_output)
+    # print("triton_output", triton_output)
+
+    # Verify accuracy
+    assert torch.allclose(triton_output, ref_output, atol=1e-2, rtol=1e-2), \
+        "Triton output doesn't match reference"
+    print("Pass topk sparse attention test with qlen == klen")
+
+
+def test_topk_sparse_attention_qlt_kl():
+    BATCH, N_HEADS = 1, 1
+    Q_LEN, K_LEN, D_HEAD = 64, 256, 64  # qlen < klen; here, past_len = 256 - 128 = 128.
+    TOPK = 1
+    BLOCK = 64  # block size used in downsampling
+    torch.manual_seed(0)
+
+    # Create inputs.
+    q = torch.randn(BATCH, N_HEADS, Q_LEN, D_HEAD, device='cuda', dtype=torch.bfloat16)
+    k = torch.randn(BATCH, N_HEADS, K_LEN, D_HEAD, device='cuda', dtype=torch.bfloat16)
+    v = torch.randn(BATCH, N_HEADS, K_LEN, D_HEAD, device='cuda', dtype=torch.bfloat16)
+    # softmax scale
+    sm_scale = 1.0 / (D_HEAD**0.5)
+
+    downsample_factor = BLOCK
+    downsample_len = math.ceil(K_LEN / downsample_factor)  # number of blocks along one dimension
+    x_ds = torch.randn(
+        BATCH, N_HEADS, downsample_len, downsample_len, device='cuda', dtype=torch.bfloat16)
+    # Force the first column to be high so that the first block is always selected.
+    x_ds[:, :, :, 0] = 100
+    block_mask = get_sparse_attn_mask_from_topk(x_ds, topk=TOPK)
+    # Run Triton kernel.
+    triton_output = block_sparse_triton_fn(q, k, v, block_mask, sm_scale)
+
+    past_len = K_LEN - Q_LEN
+
+    attn = torch.einsum('bhsd,bhtd->bhst', q, k) * sm_scale
+
+    full_mask_full = torch.kron(block_mask.float(), torch.ones(BLOCK, BLOCK, device='cuda')).bool()
+    full_mask_full = full_mask_full[..., :K_LEN, :K_LEN]
+
+    effective_mask = full_mask_full[..., past_len:K_LEN, :]  # shape: (B, H, Q_LEN, K_LEN)
+
+    i_global = torch.arange(past_len, K_LEN, device=k.device).unsqueeze(1)  # shape: (Q_LEN, 1)
+    j_global = torch.arange(K_LEN, device=k.device).unsqueeze(0)  # shape: (1, K_LEN)
+    causal_mask = (j_global <= i_global)  # shape: (Q_LEN, K_LEN)
+
+    final_mask = effective_mask & causal_mask  # shape: (B, H, Q_LEN, K_LEN)
+
+    attn = attn.masked_fill(~final_mask, float('-inf'))
+    attn = F.softmax(attn, dim=-1)
+    ref_output = torch.einsum('bhst,bhtd->bhsd', attn, v)
+
+    # Verify accuracy.
+    assert torch.allclose(triton_output, ref_output, atol=1e-2, rtol=1e-2), \
+        "Triton output doesn't match reference when qlen < klen"
+
+    print("Pass topk sparse attention test with qlen < klen")
+
+
+if __name__ == "__main__":
+    test_topk_sparse_attention()
+    test_topk_sparse_attention_qlt_kl()

testing/python/kernel/test_tilelang_kernel_dequantize_gemm.py

@@ -350,8 +350,8 @@ def tl_matmul_with_ladder_weight_only_transform_block_reduce_int4(
     accum_dtype,
     transform_b,
 ):
-    from bitblas.tl.utils import make_mma_swizzle_layout as make_swizzle_layout
-    from bitblas.tl.mma_macro_generator import (
+    from tilelang.intrinsics.mma_layout import make_mma_swizzle_layout as make_swizzle_layout
+    from tilelang.intrinsics.mma_macro_generator import (
         TensorCoreIntrinEmitterWithLadderTransform,)
 
     from bitblas.gpu.intrin.lop3 import decode_i4_to_f16
@@ -641,6 +641,4 @@ def test_assert_tl_matmul_with_ladder_weight_only_transform_block_reduce_int4():
 
 
 if __name__ == "__main__":
-    # tilelang.testing.main()
-    assert_simple_impl_float16xfp4_gemm(256, 256, 256, "float16", "float16", "float32", 64, 64, 64,
-                                        1, 128)
+    tilelang.testing.main()

tilelang/language/print.py

@@ -93,6 +93,30 @@ def print_fragment_buffer_with_condition(condition: tir.PrimExpr,
             tir.call_extern("handle", "debug_print_buffer_value", msg, buffer.name, i, smem[coords])
 
 
+@macro
+def print_local_buffer_with_condition(condition: tir.PrimExpr,
+                                      buffer: tir.Buffer,
+                                      elems: int,
+                                      msg: str = "") -> tir.PrimExpr:
+    """
+    Conditionally prints the values of a flattened TIR buffer if the condition is True.
+    
+    Parameters:
+        condition (tir.PrimExpr): A TIR expression representing the condition to check.
+        buffer (tir.Buffer): The buffer whose values need to be printed.
+        elems (int): The number of elements in the buffer to print.
+        
+    Returns:
+        tir.PrimExpr: The TIR expression for the debug print operation.
+    """
+    if condition:
+        # Iterate through the buffer elements and print each one.
+        for i in serial(elems):
+            coords = index_to_coordinates(i, buffer.shape)
+            tir.call_extern("handle", "debug_print_buffer_value", msg, buffer.name, i,
+                            buffer[coords])
+
+
 def print(obj: Any, msg: str = "") -> tir.PrimExpr:
     """
     A generic print function that handles both TIR buffers and primitive expressions.
@@ -117,7 +141,16 @@ def print(obj: Any, msg: str = "") -> tir.PrimExpr:
 
         # Flatten the buffer for consistent printing. This assumes a 1D flattened buffer.
         buffer = obj
-        if buffer.scope() == "local.fragment":
+        if buffer.scope() == "local":
+            # Get the number of elements in the buffer.
+            elems = 1
+            for dim in buffer.shape:
+                elems *= dim
+            condition = True
+            if not msg:
+                msg = f"buffer<{buffer.name}, {buffer.dtype}>"
+            return print_local_buffer_with_condition(condition, buffer, elems, msg)
+        elif buffer.scope() == "local.fragment":
             # Get the number of elements in the buffer.
             elems = 1
             for dim in buffer.shape: