diff --git a/src/hotspot/share/opto/vectornode.cpp b/src/hotspot/share/opto/vectornode.cpp
index 7592139f2d3..1a7954f6616 100644
--- a/src/hotspot/share/opto/vectornode.cpp
+++ b/src/hotspot/share/opto/vectornode.cpp
@@ -1051,7 +1051,7 @@ Node* VectorNode::Ideal(PhaseGVN* phase, bool can_reshape) {
 //
 // Due to the unique nature of vector masks, for specific IR patterns,
 // VectorMaskCast does not affect the output results. For example:
-//   (VectorStoreMask (VectorMaskCast ... (VectorLoadMask x))) => (x)
+//   (VectorStoreMask (VectorMaskCast* (VectorLoadMask x))) => (x)
 //   x remains to be a bool vector with no changes.
 // This function can be used to eliminate the VectorMaskCast in such patterns.
 Node* VectorNode::uncast_mask(Node* n) {
@@ -1486,7 +1486,7 @@ Node* VectorLoadMaskNode::Identity(PhaseGVN* phase) {
 
 Node* VectorStoreMaskNode::Identity(PhaseGVN* phase) {
   // Identity transformation on boolean vectors.
-  //   VectorStoreMask (VectorMaskCast ... VectorLoadMask bv) elem_size ==> bv
+  //   VectorStoreMask (VectorMaskCast* VectorLoadMask bv) elem_size ==> bv
   //   vector[n]{bool} => vector[n]{t} => vector[n]{bool}
   Node* in1 = VectorNode::uncast_mask(in(1));
   if (in1->Opcode() == Op_VectorLoadMask && length() == in1->as_Vector()->length()) {
@@ -1920,7 +1920,7 @@ Node* VectorMaskOpNode::Ideal(PhaseGVN* phase, bool can_reshape) {
 }
 
 Node* VectorMaskCastNode::Identity(PhaseGVN* phase) {
-  // (VectorMaskCast (VectorMaskCast ... x)) => (x)
+  // (VectorMaskCast+ x) => (x)
   // If the types of the input and output nodes in a VectorMaskCast chain are
   // exactly the same, the intermediate VectorMaskCast nodes can be eliminated.
   Node* n = VectorNode::uncast_mask(this);
diff --git a/test/hotspot/jtreg/compiler/vectorapi/VectorMaskCastIdentityTest.java b/test/hotspot/jtreg/compiler/vectorapi/VectorMaskCastIdentityTest.java
index 4bcd0bf1c1a..c6ff2f31d9f 100644
--- a/test/hotspot/jtreg/compiler/vectorapi/VectorMaskCastIdentityTest.java
+++ b/test/hotspot/jtreg/compiler/vectorapi/VectorMaskCastIdentityTest.java
@@ -60,7 +60,7 @@ public class VectorMaskCastIdentityTest {
         VectorMask<Integer> mInt128 = VectorMask.fromArray(IntVector.SPECIES_128, mr, 0);
         mInt128 = mInt128.cast(IntVector.SPECIES_128);
         // Insert a not() to prevent the casts being optimized by the optimization:
-        // (VectorStoreMask (VectorMaskCast ... (VectorLoadMask x))) => x
+        // (VectorStoreMask (VectorMaskCast* (VectorLoadMask x))) => x
         return mInt128.not().trueCount();
     }
 
diff --git a/test/hotspot/jtreg/compiler/vectorapi/VectorMaskToLongTest.java b/test/hotspot/jtreg/compiler/vectorapi/VectorMaskToLongTest.java
index 2860d0971bb..b638bdaab91 100644
--- a/test/hotspot/jtreg/compiler/vectorapi/VectorMaskToLongTest.java
+++ b/test/hotspot/jtreg/compiler/vectorapi/VectorMaskToLongTest.java
@@ -239,7 +239,7 @@ public class VectorMaskToLongTest {
     // Test VectorMask.fromLong().toLong() with Float species.
     // For floating-point types, VectorMaskCast is inserted between fromLong and toLong to convert
     // between float and integer types. There are two relevant optimizations:
-    //   1. (VectorStoreMask (VectorMaskCast ... (VectorLoadMask x))) => (x)
+    //   1. (VectorStoreMask (VectorMaskCast* (VectorLoadMask x))) => (x)
     //   2. (VectorMaskToLong (VectorLongToMask x)) => (x)
     // The optimization behavior varies by architecture:
     // - SVE with bitperm: IR chain is (VectorMaskToLong (VectorStoreMask (VectorMaskCast