[TLX] Fix an issue in the blackwell gemm tutorial (#712)

Summary: Pull Request resolved: #712

Reviewed By: dshi7

Differential Revision: D88216358

Pulled By: htyu

fbshipit-source-id: bdd18b1eca23ace237fe844e43e89de6e09689d1

Author: htyu

third_party/tlx/tutorials/blackwell-gemm-ws_test.py

@@ -295,8 +295,7 @@ def benchmark(M, N, K, provider):
         ms, min_ms, max_ms = triton.testing.do_bench(lambda: torch.matmul(a, b), quantiles=quantiles, warmup=2000,
                                                      rep=2000)
     if provider == "triton_persistent":
-        ms, min_ms, max_ms = triton.testing.do_bench(lambda: matmul(a, b, False), quantiles=quantiles, warmup=2000,
-                                                     rep=2000)
+        ms, min_ms, max_ms = triton.testing.do_bench(lambda: matmul(a, b), quantiles=quantiles, warmup=2000, rep=2000)
 
     perf = lambda ms: 2 * M * N * K * 1e-12 / (ms * 1e-3)
     return perf(ms), perf(max_ms), perf(min_ms)