Debug: skip PV step entirely

tests/unit/test_fmha_ts_full.cu

@@ -128,9 +128,8 @@ test_fmha_ts_full(const bf16_t* __restrict__ q, const bf16_t* __restrict__ k,
     }
     __syncthreads();
 
-    // ===== STEP 3: PV GEMM (TS) =====
-    // P(128,128) × V(128,16) → O(128,16)
-    // 8 K-tiles: A = P cols [16*kt..16*kt+15), B = V K-tile kt
+    // ===== STEP 3: PV GEMM (TS) — SKIPPED FOR DEBUG =====
+    /*
     {
         uint32_t idesc_pv = make_idesc(BLOCK_MN, HD);
 
@@ -147,6 +146,7 @@ test_fmha_ts_full(const bf16_t* __restrict__ q, const bf16_t* __restrict__ k,
     }
     asm volatile("tcgen05.fence::after_thread_sync;" ::: "memory");
     __syncthreads();
+    */
 
     // ===== STEP 4: Epilogue — read O from TMEM =====
     // MMA output is scaled by 0.5, so multiply by 2.0