8350956: Fix repetitions of the word "the" in compiler component comments

Reviewed-by: rcastanedalo

Author: Thomas Schatzl

src/hotspot/cpu/arm/frame_arm.inline.hpp

@@ -62,7 +62,7 @@ inline void frame::init(intptr_t* sp, intptr_t* unextended_sp, intptr_t* fp, add
   if (original_pc != nullptr) {
     _pc = original_pc;
     assert(_cb->as_nmethod()->insts_contains_inclusive(_pc),
-           "original PC must be in the main code section of the the compiled method (or must be immediately following it)");
+           "original PC must be in the main code section of the compiled method (or must be immediately following it)");
     _deopt_state = is_deoptimized;
   } else {
     _deopt_state = not_deoptimized;

src/hotspot/share/opto/escape.cpp

@@ -4472,7 +4472,7 @@ void ConnectionGraph::split_unique_types(GrowableArray<Node *>  &alloc_worklist,
       }
     } else if (n->is_AddP()) {
       if (has_reducible_merge_base(n->as_AddP(), reducible_merges)) {
-        // This AddP will go away when we reduce the the Phi
+        // This AddP will go away when we reduce the Phi
         continue;
       }
       Node* addp_base = get_addp_base(n);

src/hotspot/share/opto/vectorization.hpp

@@ -445,7 +445,7 @@ class VLoopBody : public StackObj {
 // types (byte, char, short). In the C2 IR, their operations are
 // done with full int type with 4 byte precision (e.g. AddI, MulI).
 // Example:  char a,b,c;  a = (char)(b + c);
-// However, if we can prove the the upper bits are only truncated,
+// However, if we can prove the upper bits are only truncated,
 // and the lower bits for the narrower type computed correctly, we
 // can compute the operations in the narrower type directly (e.g we
 // perform the AddI or MulI with 1 or 2 bytes). This allows us to