fix performance regression on blockgemm v3 pipe

include/ck/tensor_operation/gpu/block/blockwise_gemm_pipeline_xdlops_v3.hpp

@@ -477,6 +477,9 @@ struct BlockwiseGemmXdlops_pipeline_v3<BlockGemmPipelineScheduler::Intrawave,
                                        b_thread_buf_tail);
                 });
             });
+
+            HotLoopScheduler();
+            __builtin_amdgcn_sched_barrier(0);
         }
     }
 
@@ -692,6 +695,9 @@ struct BlockwiseGemmXdlops_pipeline_v3<BlockGemmPipelineScheduler::Intrawave,
                 });
             });
 
+            HotLoopScheduler();
+            __builtin_amdgcn_sched_barrier(0);
+
             static_for<0, KRepeat, 1>{}([&](auto k0) {
                 static_for<0, MRepeat, 1>{}([&](auto m0) {
                     static_for<0, NRepeat, 1>{}([&](auto n0) {