cba41d500c
D1.3: Fix critical bug - add TMEM column offset for P0 in PV GEMM
...
The softmax warps store P at tmem_p0_offset=32. PV MMA must read from
the same offset. tOrP0 was missing the offset, causing PV to read from
TMEM column 0 (where S is) instead of column 32 (where P is).
This was the root cause of NaN/zeros in D1 tests.
2026-05-23 21:00:29 +00:00
ed71f25903
D1.3: Revert to d1.3-pre-sm100-helpers baseline for testing
2026-05-23 20:58:06 +00:00
79a17303c4
D1.3: DIAGNOSTIC - test epilogue_tma_store raw PV without any round-trips
2026-05-23 20:57:13 +00:00
b926a0e806
D1.3: Remove NO-OP round-trip, keep normalize + epilogue_tma_store
2026-05-23 20:56:13 +00:00
820d6921d9
D1.3: Full correction_epilog with TMA store, normalize in reg before SMEM write
...
One-way trip: TMEM->reg (normalize) ->SMEM->GMEM
Replicates epilogue_tma_store logic with normalize step added
Uses CUTLASS helpers for correct layout handling
2026-05-23 20:54:23 +00:00
1c74d02adb
D1.3: Apply transform_partitioned_tensor_layout before epilogue helpers
2026-05-23 20:52:42 +00:00
1cf7140ea3
D1.3: Replace NO-op TMEM round-trip with correction_epilog using epilogue_tmem_copy_and_partition + epilogue_smem_copy_and_partition
...
- Remove hand-constructed TMEM round-trips (3% layout mismatch error)
- Use CUTLASS get_tmem_load_op + get_smem_store_op paired atoms
- One-way trip: TMEM -> reg (normalize) -> SMEM -> GMEM
- SMEM-P path: zero-fill stub (proper copy TBD)
- Keep per-tile O rescale atoms for n>128 support
2026-05-23 20:50:23 +00:00
1e55e36919
SMEM-P: test permutation 4 (swap m↔n2)
2026-05-23 20:20:07 +00:00
c0e2fe43f1
SMEM-P: add iterator offset debug print
2026-05-23 20:19:22 +00:00
42880fa397
SMEM-P: add tCrP debug print, reset permute to 0
2026-05-23 20:14:32 +00:00
1f46a62f65
SMEM-P: test permutation 1 (swap m↔n0)
2026-05-23 20:14:21 +00:00
57d5e4c9e9
auto: pre-test commit
2026-05-23 20:13:47 +00:00
474363ce59
SMEM-P: add debug_permute flag for coordinate permutation testing
2026-05-23 20:13:44 +00:00
cf9050c333
SMEM-P: disable debug flags, revert to original mapping
2026-05-23 20:12:26 +00:00
c7f613644f
SMEM-P: fix scoping error, disable debug_p_one, enable debug_swap_mn
2026-05-23 20:11:50 +00:00
14524295c7
SMEM-P: add debug_swap_mn flag to test swapped coordinate mapping
2026-05-23 20:10:39 +00:00
aafbe16a81
SMEM-P: add debug_p_one flag to write constant P=1.0
2026-05-23 20:09:47 +00:00
75b3d0f4b4
auto: pre-test commit
2026-05-23 20:08:31 +00:00
b551287123
SMEM-P: implement two-phase softmax with normalization before SMEM write
2026-05-23 20:08:29 +00:00
c88ec626e0
Revert TMEM layout change (caused error)
2026-05-23 20:07:02 +00:00
bb7f53702f
SMEM-P: try using PV A-operand layout directly for TMEM-P
2026-05-23 20:06:23 +00:00
cbb85cf091
SMEM-P: debug compute offset for known coordinate
2026-05-23 20:05:02 +00:00
64aa69c421
SMEM-P: add crd2idx debug attempt
2026-05-23 20:04:28 +00:00
1854fa19d5
SMEM-P: test pattern based on fragment indices (k,j)
2026-05-23 20:03:22 +00:00
ae7b550c3b
auto: pre-test commit
2026-05-23 20:02:35 +00:00
f21167f029
SMEM-P: add shape debug prints
2026-05-23 20:02:32 +00:00
d2d0eec33a
SMEM-P: add debug prints for coordinates
2026-05-23 20:00:33 +00:00
8edf2d434c
SMEM-P: implement CUTLASS LLM fixes - dynamic frg_tile, local coordinate conversion
2026-05-23 19:59:52 +00:00
44fc59b8fa
auto: pre-test commit
2026-05-23 19:53:45 +00:00
2b51837465
SMEM-P: try transposed mapping (swap m/n)
2026-05-23 19:53:42 +00:00
01fe51b175
SMEM-P: current state - working but mapping wrong (cos 0.02)
2026-05-23 19:53:25 +00:00
3d044b4747
SMEM-P: debug with linear index pattern m*128+n
2026-05-23 19:52:46 +00:00
81630037bd
SMEM-P: debug with test pattern (k+j)*0.01
2026-05-23 19:52:02 +00:00
1eb0c1b47a
SMEM-P: fix coordinate access - qk_coord is (m,n) not ((m,n),0,0)
2026-05-23 19:38:11 +00:00
dd42245478
SMEM-P: fix scoping - define tTMEM_LOADcS_frg unconditionally
2026-05-23 19:37:34 +00:00
25bb52feef
auto: pre-test commit
2026-05-23 19:36:58 +00:00
e09c8057be
SMEM-P: implement full 128-value write in softmax loop using coordinate mapping
2026-05-23 19:36:56 +00:00
e118ad967d
SMEM-P: fix BF16 value creation (use constant)
2026-05-23 19:33:29 +00:00
58639aa634
SMEM-P: implement CUTLASS LLM coordinate mapping pattern (minimal test)
2026-05-23 19:32:11 +00:00
dfe1d3803b
SMEM-P: fix thread_idx tuple access
2026-05-23 19:30:09 +00:00
c01d8b493f
auto: pre-test commit
2026-05-23 19:29:29 +00:00
ec84d35cf8
SMEM-P: add debug to understand thread partitioning
2026-05-23 19:29:27 +00:00
ee969d4c46
Update STAGE_D.md: manual SMEM addressing blocked on layout mapping
2026-05-23 19:22:28 +00:00
e0407793af
SMEM-P: implement simple test pattern instead of coord lookup
2026-05-23 19:21:31 +00:00
841db091f7
auto: pre-test commit
2026-05-23 19:20:42 +00:00
9b72411ca7
Start implementing manual SMEM-P addressing (helpers are a trap)
2026-05-23 19:20:40 +00:00
e0a2d272f4
Implement manual SMEM-P copy instead of cute.copy (helpers are a trap)
2026-05-23 19:14:44 +00:00
4d6acaeef0
auto: pre-test commit
2026-05-23 19:14:02 +00:00
341527977c
Try flattening sP and rP_bf16_qk with group_modes to fix rank mismatch
2026-05-23 19:13:59 +00:00
c9448aca03
Add debug prints for SMEM-P partition layouts
2026-05-23 19:13:07 +00:00
