debug: zero-fill sP to check deadlock

dsv4/kernels/attention/fmha.py

@@ -360,20 +360,12 @@ class FmhaKernel:
                     cute.copy(tiled_tmem_store, rP_words, tTMEM_STOREtP)
                     cute.arch.fence_view_async_tmem_store()
                 else:
-                    # SMEM-P: write P to sP using coordinate-indexed store.
-                    # Uses tTMEM_LOADcS identity tensor to get (m, k) coordinates.
-                    # DEBUG: Write a known pattern to sP to verify the coordinate mapping.
-                    # Pattern: sP[m, k] = (m + k) % 256 as BF16 (unique per position)
+                    # SMEM-P: TEMPORARILY zero-fill sP (debugging deadlock).
+                    # The coordinate-indexed write causes a deadlock at hd=256.
+                    # TODO: Fix the SMEM-P write path.
                     for j0 in range(32):
                         for j1 in range(4):
-                            coord = tTMEM_LOADcS[(j0, 0), j1, 0, 0]
-                            m_coord = coord[0]
-                            k_coord = coord[1]
-                            k0 = k_coord % 16
-                            k1 = (k_coord // 16) % 4
-                            k2 = k_coord // 64
-                            # Debug: write (m + k) mod 256 instead of actual P value
-                            _sP_nostage[(m_coord, k0), 0, (k1, k2)] = rP_bf16[(j0, 0), j1, 0, 0]
+                            _sP_nostage[(j0, j1), 0, (0, 0)] = BFloat16(0.0)
                     cute.arch.fence_proxy("async.shared", space="cta")
                 if kt > 0:
                     tTMrO = cute.make_rmem_tensor(