2b3435f97c
D1: Remove debug prints, clean up
2026-05-24 04:06:26 +00:00
38c6486fc7
D1: const_expr for sP layout selection (CuTeDSL)
2026-05-24 04:05:17 +00:00
a945edea79
D1: Python if for sP layout (trace-time, not MLIR)
2026-05-24 04:04:27 +00:00
955b023164
D1: Tiny 4-mode sP placeholder for TMEM-P path
2026-05-24 04:03:28 +00:00
762f45a8a2
D1: Conditional sP allocation (saves 64KB SMEM for TMEM-P at hd=256)
2026-05-24 04:02:02 +00:00
2184c79732
D1: Fix sP dummy allocation
2026-05-24 04:00:19 +00:00
44d96f2cf8
D1: Skip sP allocation when use_smem_p=False (saves 64KB at hd=256)
2026-05-24 03:59:27 +00:00
8f9b6518f1
D1: Fix syntax (separate kv_stage line)
2026-05-24 03:58:12 +00:00
6597a1cd16
D1: Reduce kv_stage to 1 at hd>128 to avoid SMEM overflow
2026-05-24 03:55:44 +00:00
39367265e5
D1: FIX qk_mma_tiler K-dim = head_dim (was hardcoded to 64, broke hd>64)
2026-05-24 03:53:19 +00:00
13675a12f8
D1: Print qk_ik in _setup
2026-05-24 03:51:40 +00:00
58ed4f5e8b
D1: Add more debug prints (QK/PV mode2 sizes)
2026-05-24 03:49:55 +00:00
ebde1d67fd
D1: Add sP shape debug print
2026-05-24 03:46:27 +00:00
0f52c3453c
D1: Fix SMEM-P (coordinate store), LSE (FP32), add TMEM-P-only test
2026-05-24 03:27:14 +00:00
52570d94cb
D1: Fix SMEM-P - coordinate-indexed store (replaces make_tiled_copy_C)
2026-05-24 03:24:44 +00:00
70c9d93d28
feat: SMEM-P make_tiled_copy_C + zero-fill dest tensor
2026-05-24 03:23:53 +00:00
7cf9704584
feat: SMEM-P using make_tiled_copy_C(qk_mma) approach
2026-05-24 03:22:57 +00:00
fe47a5f882
fix: LSE type mismatch Float32→BFloat16
2026-05-24 03:20:26 +00:00
6313974fba
D1.5: Fix SMEM-P - use coordinate-indexed store (same proven pattern)
2026-05-24 03:19:32 +00:00
153db24be2
D1.5: Always output un-normalized O + LSE (epilogue_tma_store only, no TMEM round-trip normalize)
2026-05-24 03:18:38 +00:00
d68ab348bb
feat: SMEM-P using make_tiled_copy_A from PV MMA
2026-05-24 03:16:34 +00:00
b4a985631b
fix: fence_proxy not fence
2026-05-24 02:45:01 +00:00
228ec3c638
D1.5: Replace broken make_cotiled_copy SMEM-P with coordinate-indexed store
2026-05-24 02:43:42 +00:00
eda7d40df2
Merge branch 'master' of ssh://sweetapi.com:2222/biondizzle/nvfp4-megamoe-kernel
2026-05-24 02:41:39 +00:00
952c25e227
D1.5: Use tCtO_fake layout for epilogue_tma_store (needs STAGE dim)
2026-05-24 02:41:32 +00:00
0a980de7ad
feat: SMEM-P using make_cotiled_copy (one-row-per-thread)
2026-05-24 02:41:19 +00:00
85eb2bc4bb
D1.5: Remove duplicate tTMrO definition (keep unconditional one)
2026-05-24 02:40:40 +00:00
cd223e1b98
fix: reorder tTMrO definition after tTMEM_LOADcO
2026-05-24 02:36:08 +00:00
54e94d44ef
fix: tTMrO scoping + restore SMEM-P coordinate write
2026-05-24 02:34:53 +00:00
6ead708c7d
D1.5: Move tTMrO def before softmax loop (CuTeDSL scoping)
2026-05-24 02:32:51 +00:00
5a34865062
debug: zero-fill sP to check deadlock
2026-05-24 02:31:12 +00:00
81652629e3
D1.5: Use proven Stage C approach - normalize via TMEM round-trip + epilogue_tma_store
2026-05-24 02:30:25 +00:00
e50ba7212c
test: SMEM-P coordinate verification test
2026-05-24 01:58:32 +00:00
8166898cc3
D1.5: Fix bSG_gC slicing - group trailing modes (CUTLASS pattern)
2026-05-24 01:41:52 +00:00
c7a692b75e
D1.5: Dynamic slicing for tTR_gC (variable rest dims)
2026-05-24 01:40:44 +00:00
cb6eae4c4f
D1.5: Fix flat_divide slice coordinates (4 modes, no STAGE dim)
2026-05-24 01:39:21 +00:00
d6a607d12e
D1.5: Rewrite correction epilogue using CUTLASS pattern (transform_partitioned, flat_divide, paired atoms)
2026-05-24 01:37:53 +00:00
ef9fa86848
D1.5: Fix TMA store - group_modes on bSG_gC, use flat indexing
2026-05-24 01:36:01 +00:00
5a5cd19b81
D1.5: Fix TMA store - use flat_divide on tCgC instead of local_tile on mC
2026-05-24 01:35:10 +00:00
52a5aa61bc
D1.5: Fix TMA store - use 3D tile for local_tile on 3D mC
2026-05-24 01:20:33 +00:00
865832f669
D1.5: Use group_modes on sC for 2D TMA view (preserves swizzle)
2026-05-24 00:52:57 +00:00
a8939ced1f
D1.5: Use 2D sC_epi layout from c_smem_s for TMA partition
2026-05-24 00:51:18 +00:00
9705b07969
D1.5: Fix TMA store - use group_modes on sC and tCgC
2026-05-24 00:48:18 +00:00
d834836bb7
D1.5: Simplify TMA store - use 2D sC_epi and gC_epi views
2026-05-24 00:46:52 +00:00
c05a171e15
D1.5: Fix TMA store - use existing gC partition
2026-05-24 00:43:35 +00:00
74e1c0420a
D1.5: Implement correction epilog with paired atoms (get_tmem_load_op + get_smem_store_op)
...
One-way: TMEM → registers (normalize) → SMEM → GMEM
Based on CUTLASS FMHA reference's correction_epilog pattern.
Eliminates TMEM round-trip error for O normalization.
O rescale (kt>0) still uses old atoms (separate fix).
2026-05-24 00:41:27 +00:00
d96786ec44
D1.5: Add TODO for correction epilog - keeping working TMEM round-trip for now
2026-05-24 00:37:36 +00:00
ae7a1f5e0a
D1.5: Revert to pre-epilog backup - correction epilog refactor is complex, will do incrementally
...
The correction epilog (TMEM→reg→SMEM→GMEM one-way trip) is the right approach
but the TMA store from SMEM requires proper partitioning that needs more work.
Reverting to the known-working state (with 3% TMEM round-trip error) to focus
on the SMEM-P write first.
2026-05-24 00:35:00 +00:00
a59d57e4d5
D1.5: Fix TMA store - use local_tile with pv_mma_tiler
2026-05-24 00:32:35 +00:00
a6bf31a22e
D1.5: Fix TMA store rank mismatch - use 2D sC_epi view
2026-05-24 00:31:45 +00:00
