diff --git a/impl/kfc/kernel_kfc.h b/impl/kfc/kernel_kfc.h
index 147a9bf13e15e59fc3b8ee65736e4162bce5e70a..b11e3bbe6bfcf0db731358bb010ab6df22e2869d 100644
--- a/impl/kfc/kernel_kfc.h
+++ b/impl/kfc/kernel_kfc.h
@@ -38,7 +38,7 @@ __aicore__ inline void clearWorkspace(__gm__ uint8_t *workspace)
         AscendC::ClearWorkspaceImpl(workspace);
         AscendC::NotifyEvent<PIPE_MTE3>(WORKSPACE_SYNC_ID);
     } else {
-        AscendC::AscendCUtils::SetMask<uint64_t>((uint64_t)-1, (uint64_t)-1);
+        SetVectorMask<uint64_t, MaskMode::NORMAL>((uint64_t)-1, (uint64_t)-1);
         AscendC::SetMaskNorm();
     }
 #endif
diff --git a/impl/normalization/deepnorm/deepnorm_common_impl.h b/impl/normalization/deepnorm/deepnorm_common_impl.h
index 33d3f8d8de08f38d8bc99ff5698f27f9b7e86657..fbb10ca223bf594ed117343196801a5a066ace35 100644
--- a/impl/normalization/deepnorm/deepnorm_common_impl.h
+++ b/impl/normalization/deepnorm/deepnorm_common_impl.h
@@ -38,8 +38,8 @@ __aicore__ inline bool IsDeepNormParamValid(DeepNormTiling& tiling)
     const bool bsDivBy8 = ((tiling.bLength * tiling.sLength) % BASIC_BLOCK_BSLENGTH == 0);
     if constexpr (isBasicBlock) {
         ASCENDC_ASSERT(hDivBy64 && bsDivBy8,
-            {KERNEL_LOG(KERNEL_ERROR, "In DeepNorm, when isBasicBlock is true, input must have hLength % 64 = 0, " \
-                "originalHLegnth % 64 = 0 and (bLength * sLength) % 8 = 0 !");});
+            {KERNEL_LOG(KERNEL_ERROR, "In DeepNorm, when isBasicBlock is true, input must have hLength %% 64 = 0, " \
+                "originalHLegnth %% 64 = 0 and (bLength * sLength) % 8 = 0 !");});
     }
 
     return true;
diff --git a/impl/normalization/layernorm/layernorm_common_impl.h b/impl/normalization/layernorm/layernorm_common_impl.h
index ae1aab74595bbc65981eb38c024fd4bbea8666ab..2ec5cd5fe0d1cb5948a35a3f4a662499bc3b8c79 100644
--- a/impl/normalization/layernorm/layernorm_common_impl.h
+++ b/impl/normalization/layernorm/layernorm_common_impl.h
@@ -466,7 +466,7 @@ __aicore__ inline void WelfordUpdateImpl(const LocalTensor<U>& outputMean, const
     });
     ASCENDC_ASSERT((para.abComputeLength > 0), {
         KERNEL_LOG(KERNEL_ERROR,
-        "Failed to check para.abComputeLength, para.abComputeLength should be greater than 0.",
+        "Failed to check para.abComputeLength, para.abComputeLength (%u) should be greater than 0.",
             para.abComputeLength, para.abLength);
     });
     ASCENDC_ASSERT((para.rnLength == 1), {
diff --git a/impl/normalization/welfordfinalize/welfordfinalize_common_impl.h b/impl/normalization/welfordfinalize/welfordfinalize_common_impl.h
index 08d405a33ce453623f516cbeedb235eec727f063..c23609b1dc06b31fffd3571f7b0eea8c56485bf5 100644
--- a/impl/normalization/welfordfinalize/welfordfinalize_common_impl.h
+++ b/impl/normalization/welfordfinalize/welfordfinalize_common_impl.h
@@ -421,7 +421,7 @@ __aicore__ inline void GetWelfordFinalizeTensorInfo(const LocalTensor<float> &st
     // 0x2 indicates reserving two buffers for the storage of mean and variance
     const uint32_t minTmpOutSize = B32_LEN * 0x2;
     ASCENDC_ASSERT((stackBuffer.GetSize() >= (minTmpSize + minTmpOutSize)),
-                { KERNEL_LOG(KERNEL_ERROR, "sharedTmpBuffer size must >= %d Bytes!", (minTmpSize + minTmpOutSize) * sizeof(float)); });
+                { KERNEL_LOG(KERNEL_ERROR, "sharedTmpBuffer size must >= %ld Bytes!", (minTmpSize + minTmpOutSize) * sizeof(float)); });
 
     const uint32_t expFactor = (stackBuffer.GetSize() - minTmpOutSize) / minTmpSize;
     tiling.computeLength = expFactor * B32_LEN;
diff --git a/impl/sort/topk/topk_common_impl.h b/impl/sort/topk/topk_common_impl.h
index b29b0870bcbaab897cf9cdd0f1bc63edff1f1653..45fda042b607d7bb266538a2ad7490b5a97a6fe4 100644
--- a/impl/sort/topk/topk_common_impl.h
+++ b/impl/sort/topk/topk_common_impl.h
@@ -123,7 +123,7 @@ __aicore__ inline void TopKNSmall(const LocalTensor<T> &dstValueLocal, const Loc
     LocalTensor<uint8_t> stackTensor;
     PopStackBuffer<uint8_t, TPosition::LCM>(stackTensor);
     ASCENDC_ASSERT((stackTensor.GetSize() / sizeof(T) >= tilling.tmpLocalSize), {KERNEL_LOG(KERNEL_ERROR, "The stack "
-        "buffer is insufficient, TopK api need %d, but only %d exists.", tilling.tmpLocalSize, 
+        "buffer is insufficient, TopK api need %d, but only %ld exists.", tilling.tmpLocalSize, 
         stackTensor.GetSize() / sizeof(T));});
     stackTensor.SetSize(tilling.tmpLocalSize * sizeof(T));
 
diff --git a/tests/activation/sigmoid/test_operator_vec_sigmoid.cpp b/tests/activation/sigmoid/test_operator_vec_sigmoid.cpp
index 4d387301a5892f088bb289e2fc2999d9e7ed3abe..7a65331fe954887b94cf858834d481225a93a2dc 100644
--- a/tests/activation/sigmoid/test_operator_vec_sigmoid.cpp
+++ b/tests/activation/sigmoid/test_operator_vec_sigmoid.cpp
@@ -30,14 +30,14 @@ void SigmoidKernel(__gm__ uint8_t* __restrict__ srcGm, __gm__ uint8_t* __restric
     tpipe.InitBuffer(tbuf2, dataSize * sizeof(T));
     LocalTensor<T> outputLocal = tbuf2.Get<T>();
 
-    AscendCUtils::SetMask<uint8_t>(256);
+    SetVectorMask<uint8_t, MaskMode::NORMAL>(256);
 
     DataCopy(inputLocal, inputGlobal, dataSize);
     event_t eventIdMTE2ToV = static_cast<event_t>(GetTPipePtr()->FetchEventID(HardEvent::MTE2_V));
     SetFlag<HardEvent::MTE2_V>(eventIdMTE2ToV);
     WaitFlag<HardEvent::MTE2_V>(eventIdMTE2ToV);
 
-    AscendCUtils::SetMask<uint8_t>(128);
+    SetVectorMask<uint8_t, MaskMode::NORMAL>(128);
 
     Sigmoid(outputLocal, inputLocal, dataSize);
 
diff --git a/tests/math/axpy/test_operator_axpy.cpp b/tests/math/axpy/test_operator_axpy.cpp
index 8a7126a17e7061da698f9c90bd3004a9164eabb1..ae348c646946f4a78c6ee974b91b037a39b9a35e 100644
--- a/tests/math/axpy/test_operator_axpy.cpp
+++ b/tests/math/axpy/test_operator_axpy.cpp
@@ -35,11 +35,11 @@ void AxpyKernel(__gm__ uint8_t* __restrict__ srcGm, __gm__ uint8_t* __restrict__
     LocalTensor<T> outputLocal = vecOutQue.AllocTensor<T>();
     LocalTensor<uint8_t> tmpLocal = vecTmpQue.AllocTensor<uint8_t>();
 
-    AscendCUtils::SetMask<uint8_t>(256);
+    SetVectorMask<uint8_t, MaskMode::NORMAL>(256);
     DataCopy(inputLocal, inputGlobal, dataSize);
 
     SetFlag<HardEvent::MTE2_V>(EVENT_ID0);
-    AscendCUtils::SetMask<uint8_t>(128);
+    SetVectorMask<uint8_t, MaskMode::NORMAL>(128);
     WaitFlag<HardEvent::MTE2_V>(EVENT_ID0);
     U scalar = 4;
 
diff --git a/tests/math/tanh/test_operator_tanh.cpp b/tests/math/tanh/test_operator_tanh.cpp
index 5a3abfd45a7ca74995153c48f37609c7868ce355..cba449a5b19d612cb742458707c225c392ca3e57 100644
--- a/tests/math/tanh/test_operator_tanh.cpp
+++ b/tests/math/tanh/test_operator_tanh.cpp
@@ -30,12 +30,12 @@ void TanhKernel(__gm__ uint8_t* __restrict__ srcGm, __gm__ uint8_t* __restrict__
     tpipe.InitBuffer(tbuf2, dataSize * sizeof(T));
     LocalTensor<T> outputLocal = tbuf2.Get<T>();
 
-    AscendCUtils::SetMask<uint8_t>(256);
+    SetVectorMask<uint8_t, MaskMode::NORMAL>(256);
 
     DataCopy(inputLocal, inputGlobal, dataSize);
 
     SetFlag<HardEvent::MTE2_V>(EVENT_ID0);
-    AscendCUtils::SetMask<uint8_t>(128);
+    SetVectorMask<uint8_t, MaskMode::NORMAL>(128);
     WaitFlag<HardEvent::MTE2_V>(EVENT_ID0);
 
     Tanh(outputLocal, inputLocal, dataSize);
diff --git a/tests/math/xor/test_operator_xor.cpp b/tests/math/xor/test_operator_xor.cpp
index edba1b78e62a67edf93c5f616f191aa554d2b2b2..b3c3daa8ede04df62b2a5a2574b5fdca9a63bcb0 100644
--- a/tests/math/xor/test_operator_xor.cpp
+++ b/tests/math/xor/test_operator_xor.cpp
@@ -59,13 +59,13 @@ void main_vec_xor_level2_demo(__gm__ uint8_t* __restrict__ dstGm,
         input1_local.SetValue(i, 1);
     }
 
-    AscendCUtils::SetMask<uint8_t>(256);
+    SetVectorMask<uint8_t, MaskMode::NORMAL>(256);
 
     DataCopy(input0_local, input0Global, dataSize);
     DataCopy(input1_local, input1Global, dataSize);
 
     SetFlag<HardEvent::MTE2_V>(EVENT_ID0);
-    AscendCUtils::SetMask<uint8_t>(128);
+    SetVectorMask<uint8_t, MaskMode::NORMAL>(128);
     WaitFlag<HardEvent::MTE2_V>(EVENT_ID0);
 
     if (testMode == NORMAL_MODE) {