diff --git a/llvm/include/llvm/Analysis/TargetTransformInfo.h b/llvm/include/llvm/Analysis/TargetTransformInfo.h
--- a/llvm/include/llvm/Analysis/TargetTransformInfo.h
+++ b/llvm/include/llvm/Analysis/TargetTransformInfo.h
@@ -687,6 +687,9 @@
   /// cost should return false, otherwise return true.
   bool isNumRegsMajorCostOfLSR() const;
 
+  /// Return true if LSR attempts to replace primary IV with other IV.
+  bool isAllowTerminatingConditionFoldingAfterLSR() const;
+
   /// \returns true if LSR should not optimize a chain that includes \p I.
   bool isProfitableLSRChainElement(Instruction *I) const;
 
@@ -1728,6 +1731,7 @@
   virtual bool isLSRCostLess(const TargetTransformInfo::LSRCost &C1,
                              const TargetTransformInfo::LSRCost &C2) = 0;
   virtual bool isNumRegsMajorCostOfLSR() = 0;
+  virtual bool isAllowTerminatingConditionFoldingAfterLSR() const = 0;
   virtual bool isProfitableLSRChainElement(Instruction *I) = 0;
   virtual bool canMacroFuseCmp() = 0;
   virtual bool canSaveCmp(Loop *L, BranchInst **BI, ScalarEvolution *SE,
@@ -2158,6 +2162,9 @@
   bool isNumRegsMajorCostOfLSR() override {
     return Impl.isNumRegsMajorCostOfLSR();
   }
+  bool isAllowTerminatingConditionFoldingAfterLSR() const override {
+    return Impl.isAllowTerminatingConditionFoldingAfterLSR();
+  }
   bool isProfitableLSRChainElement(Instruction *I) override {
     return Impl.isProfitableLSRChainElement(I);
   }
diff --git a/llvm/include/llvm/Analysis/TargetTransformInfoImpl.h b/llvm/include/llvm/Analysis/TargetTransformInfoImpl.h
--- a/llvm/include/llvm/Analysis/TargetTransformInfoImpl.h
+++ b/llvm/include/llvm/Analysis/TargetTransformInfoImpl.h
@@ -220,6 +220,8 @@
 
   bool isNumRegsMajorCostOfLSR() const { return true; }
 
+  bool isAllowTerminatingConditionFoldingAfterLSR() const { return false; }
+
   bool isProfitableLSRChainElement(Instruction *I) const { return false; }
 
   bool canMacroFuseCmp() const { return false; }
diff --git a/llvm/include/llvm/CodeGen/BasicTTIImpl.h b/llvm/include/llvm/CodeGen/BasicTTIImpl.h
--- a/llvm/include/llvm/CodeGen/BasicTTIImpl.h
+++ b/llvm/include/llvm/CodeGen/BasicTTIImpl.h
@@ -378,6 +378,11 @@
     return TargetTransformInfoImplBase::isNumRegsMajorCostOfLSR();
   }
 
+  bool isAllowTerminatingConditionFoldingAfterLSR() const {
+    return TargetTransformInfoImplBase::
+        isAllowTerminatingConditionFoldingAfterLSR();
+  }
+
   bool isProfitableLSRChainElement(Instruction *I) {
     return TargetTransformInfoImplBase::isProfitableLSRChainElement(I);
   }
diff --git a/llvm/lib/Analysis/TargetTransformInfo.cpp b/llvm/lib/Analysis/TargetTransformInfo.cpp
--- a/llvm/lib/Analysis/TargetTransformInfo.cpp
+++ b/llvm/lib/Analysis/TargetTransformInfo.cpp
@@ -393,6 +393,10 @@
   return TTIImpl->isNumRegsMajorCostOfLSR();
 }
 
+bool TargetTransformInfo::isAllowTerminatingConditionFoldingAfterLSR() const {
+  return TTIImpl->isAllowTerminatingConditionFoldingAfterLSR();
+}
+
 bool TargetTransformInfo::isProfitableLSRChainElement(Instruction *I) const {
   return TTIImpl->isProfitableLSRChainElement(I);
 }
diff --git a/llvm/lib/Target/RISCV/RISCVTargetTransformInfo.h b/llvm/lib/Target/RISCV/RISCVTargetTransformInfo.h
--- a/llvm/lib/Target/RISCV/RISCVTargetTransformInfo.h
+++ b/llvm/lib/Target/RISCV/RISCVTargetTransformInfo.h
@@ -357,6 +357,8 @@
 
   bool isLSRCostLess(const TargetTransformInfo::LSRCost &C1,
                      const TargetTransformInfo::LSRCost &C2);
+
+  bool isAllowTerminatingConditionFoldingAfterLSR() const;
 };
 
 } // end namespace llvm
diff --git a/llvm/lib/Target/RISCV/RISCVTargetTransformInfo.cpp b/llvm/lib/Target/RISCV/RISCVTargetTransformInfo.cpp
--- a/llvm/lib/Target/RISCV/RISCVTargetTransformInfo.cpp
+++ b/llvm/lib/Target/RISCV/RISCVTargetTransformInfo.cpp
@@ -1756,3 +1756,7 @@
                   C2.NumIVMuls, C2.NumBaseAdds,
                   C2.ScaleCost, C2.ImmCost, C2.SetupCost);
 }
+
+bool RISCVTTIImpl::isAllowTerminatingConditionFoldingAfterLSR() const {
+  return true;
+}
diff --git a/llvm/lib/Transforms/Scalar/LoopStrengthReduce.cpp b/llvm/lib/Transforms/Scalar/LoopStrengthReduce.cpp
--- a/llvm/lib/Transforms/Scalar/LoopStrengthReduce.cpp
+++ b/llvm/lib/Transforms/Scalar/LoopStrengthReduce.cpp
@@ -6908,7 +6908,8 @@
     }
   }
 
-  if (AllowTerminatingConditionFoldingAfterLSR) {
+  if (AllowTerminatingConditionFoldingAfterLSR ||
+      TTI.isAllowTerminatingConditionFoldingAfterLSR()) {
     if (auto Opt = canFoldTermCondOfLoop(L, SE, DT, LI)) {
       auto [ToFold, ToHelpFold, TermValueS, MustDrop] = *Opt;
 
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vector-strided-load-store-asm.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vector-strided-load-store-asm.ll
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vector-strided-load-store-asm.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vector-strided-load-store-asm.ll
@@ -13,7 +13,7 @@
 define void @gather(ptr noalias nocapture %A, ptr noalias nocapture readonly %B) {
 ; CHECK-LABEL: gather:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    addi a2, a0, 1024
 ; CHECK-NEXT:    li a3, 32
 ; CHECK-NEXT:    li a4, 5
 ; CHECK-NEXT:  .LBB0_1: # %vector.body
@@ -23,10 +23,9 @@
 ; CHECK-NEXT:    vle8.v v9, (a0)
 ; CHECK-NEXT:    vadd.vv v8, v9, v8
 ; CHECK-NEXT:    vse8.v v8, (a0)
-; CHECK-NEXT:    addi a2, a2, -32
 ; CHECK-NEXT:    addi a0, a0, 32
 ; CHECK-NEXT:    addi a1, a1, 160
-; CHECK-NEXT:    bnez a2, .LBB0_1
+; CHECK-NEXT:    bne a0, a2, .LBB0_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -54,7 +53,7 @@
 define void @gather_masked(ptr noalias nocapture %A, ptr noalias nocapture readonly %B, <32 x i8> %maskedoff) {
 ; V-LABEL: gather_masked:
 ; V:       # %bb.0: # %entry
-; V-NEXT:    li a2, 1024
+; V-NEXT:    addi a2, a0, 1024
 ; V-NEXT:    lui a3, 983765
 ; V-NEXT:    addiw a3, a3, 873
 ; V-NEXT:    vsetivli zero, 1, e32, mf2, ta, ma
@@ -69,16 +68,15 @@
 ; V-NEXT:    vle8.v v10, (a0)
 ; V-NEXT:    vadd.vv v9, v10, v9
 ; V-NEXT:    vse8.v v9, (a0)
-; V-NEXT:    addi a2, a2, -32
 ; V-NEXT:    addi a0, a0, 32
 ; V-NEXT:    addi a1, a1, 160
-; V-NEXT:    bnez a2, .LBB1_1
+; V-NEXT:    bne a0, a2, .LBB1_1
 ; V-NEXT:  # %bb.2: # %for.cond.cleanup
 ; V-NEXT:    ret
 ;
 ; ZVE32F-LABEL: gather_masked:
 ; ZVE32F:       # %bb.0: # %entry
-; ZVE32F-NEXT:    li a2, 1024
+; ZVE32F-NEXT:    addi a2, a0, 1024
 ; ZVE32F-NEXT:    lui a3, 983765
 ; ZVE32F-NEXT:    addiw a3, a3, 873
 ; ZVE32F-NEXT:    vsetivli zero, 1, e32, m1, ta, ma
@@ -93,10 +91,9 @@
 ; ZVE32F-NEXT:    vle8.v v10, (a0)
 ; ZVE32F-NEXT:    vadd.vv v9, v10, v9
 ; ZVE32F-NEXT:    vse8.v v9, (a0)
-; ZVE32F-NEXT:    addi a2, a2, -32
 ; ZVE32F-NEXT:    addi a0, a0, 32
 ; ZVE32F-NEXT:    addi a1, a1, 160
-; ZVE32F-NEXT:    bnez a2, .LBB1_1
+; ZVE32F-NEXT:    bne a0, a2, .LBB1_1
 ; ZVE32F-NEXT:  # %bb.2: # %for.cond.cleanup
 ; ZVE32F-NEXT:    ret
 entry:
@@ -125,7 +122,7 @@
 ; CHECK-LABEL: gather_negative_stride:
 ; CHECK:       # %bb.0: # %entry
 ; CHECK-NEXT:    addi a1, a1, 155
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    addi a2, a0, 1024
 ; CHECK-NEXT:    li a3, 32
 ; CHECK-NEXT:    li a4, -5
 ; CHECK-NEXT:  .LBB2_1: # %vector.body
@@ -135,10 +132,9 @@
 ; CHECK-NEXT:    vle8.v v9, (a0)
 ; CHECK-NEXT:    vadd.vv v8, v9, v8
 ; CHECK-NEXT:    vse8.v v8, (a0)
-; CHECK-NEXT:    addi a2, a2, -32
 ; CHECK-NEXT:    addi a0, a0, 32
 ; CHECK-NEXT:    addi a1, a1, 160
-; CHECK-NEXT:    bnez a2, .LBB2_1
+; CHECK-NEXT:    bne a0, a2, .LBB2_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -166,7 +162,7 @@
 define void @gather_zero_stride(ptr noalias nocapture %A, ptr noalias nocapture readonly %B) {
 ; CHECK-LABEL: gather_zero_stride:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    addi a2, a0, 1024
 ; CHECK-NEXT:    li a3, 32
 ; CHECK-NEXT:  .LBB3_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
@@ -175,10 +171,9 @@
 ; CHECK-NEXT:    vle8.v v8, (a0)
 ; CHECK-NEXT:    vadd.vx v8, v8, a4
 ; CHECK-NEXT:    vse8.v v8, (a0)
-; CHECK-NEXT:    addi a2, a2, -32
 ; CHECK-NEXT:    addi a0, a0, 32
 ; CHECK-NEXT:    addi a1, a1, 160
-; CHECK-NEXT:    bnez a2, .LBB3_1
+; CHECK-NEXT:    bne a0, a2, .LBB3_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -206,7 +201,7 @@
 define void @gather_zero_stride_unfold(ptr noalias nocapture %A, ptr noalias nocapture readonly %B) {
 ; V-LABEL: gather_zero_stride_unfold:
 ; V:       # %bb.0: # %entry
-; V-NEXT:    li a2, 1024
+; V-NEXT:    addi a2, a0, 1024
 ; V-NEXT:    li a3, 32
 ; V-NEXT:  .LBB4_1: # %vector.body
 ; V-NEXT:    # =>This Inner Loop Header: Depth=1
@@ -215,16 +210,15 @@
 ; V-NEXT:    vle8.v v9, (a0)
 ; V-NEXT:    vdivu.vv v8, v8, v9
 ; V-NEXT:    vse8.v v8, (a0)
-; V-NEXT:    addi a2, a2, -32
 ; V-NEXT:    addi a0, a0, 32
 ; V-NEXT:    addi a1, a1, 160
-; V-NEXT:    bnez a2, .LBB4_1
+; V-NEXT:    bne a0, a2, .LBB4_1
 ; V-NEXT:  # %bb.2: # %for.cond.cleanup
 ; V-NEXT:    ret
 ;
 ; ZVE32F-LABEL: gather_zero_stride_unfold:
 ; ZVE32F:       # %bb.0: # %entry
-; ZVE32F-NEXT:    li a2, 1024
+; ZVE32F-NEXT:    addi a2, a0, 1024
 ; ZVE32F-NEXT:    li a3, 32
 ; ZVE32F-NEXT:  .LBB4_1: # %vector.body
 ; ZVE32F-NEXT:    # =>This Inner Loop Header: Depth=1
@@ -233,16 +227,15 @@
 ; ZVE32F-NEXT:    vle8.v v9, (a0)
 ; ZVE32F-NEXT:    vdivu.vv v8, v8, v9
 ; ZVE32F-NEXT:    vse8.v v8, (a0)
-; ZVE32F-NEXT:    addi a2, a2, -32
 ; ZVE32F-NEXT:    addi a0, a0, 32
 ; ZVE32F-NEXT:    addi a1, a1, 160
-; ZVE32F-NEXT:    bnez a2, .LBB4_1
+; ZVE32F-NEXT:    bne a0, a2, .LBB4_1
 ; ZVE32F-NEXT:  # %bb.2: # %for.cond.cleanup
 ; ZVE32F-NEXT:    ret
 ;
 ; NOT-OPTIMIZED-LABEL: gather_zero_stride_unfold:
 ; NOT-OPTIMIZED:       # %bb.0: # %entry
-; NOT-OPTIMIZED-NEXT:    li a2, 1024
+; NOT-OPTIMIZED-NEXT:    addi a2, a0, 1024
 ; NOT-OPTIMIZED-NEXT:    li a3, 32
 ; NOT-OPTIMIZED-NEXT:  .LBB4_1: # %vector.body
 ; NOT-OPTIMIZED-NEXT:    # =>This Inner Loop Header: Depth=1
@@ -252,10 +245,9 @@
 ; NOT-OPTIMIZED-NEXT:    vmv.v.x v9, a4
 ; NOT-OPTIMIZED-NEXT:    vdivu.vv v8, v9, v8
 ; NOT-OPTIMIZED-NEXT:    vse8.v v8, (a0)
-; NOT-OPTIMIZED-NEXT:    addi a2, a2, -32
 ; NOT-OPTIMIZED-NEXT:    addi a0, a0, 32
 ; NOT-OPTIMIZED-NEXT:    addi a1, a1, 160
-; NOT-OPTIMIZED-NEXT:    bnez a2, .LBB4_1
+; NOT-OPTIMIZED-NEXT:    bne a0, a2, .LBB4_1
 ; NOT-OPTIMIZED-NEXT:  # %bb.2: # %for.cond.cleanup
 ; NOT-OPTIMIZED-NEXT:    ret
 entry:
@@ -287,7 +279,7 @@
 define void @scatter(ptr noalias nocapture %A, ptr noalias nocapture readonly %B) {
 ; CHECK-LABEL: scatter:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    addi a2, a1, 1024
 ; CHECK-NEXT:    li a3, 32
 ; CHECK-NEXT:    li a4, 5
 ; CHECK-NEXT:  .LBB5_1: # %vector.body
@@ -297,10 +289,9 @@
 ; CHECK-NEXT:    vlse8.v v9, (a0), a4
 ; CHECK-NEXT:    vadd.vv v8, v9, v8
 ; CHECK-NEXT:    vsse8.v v8, (a0), a4
-; CHECK-NEXT:    addi a2, a2, -32
 ; CHECK-NEXT:    addi a1, a1, 32
 ; CHECK-NEXT:    addi a0, a0, 160
-; CHECK-NEXT:    bnez a2, .LBB5_1
+; CHECK-NEXT:    bne a1, a2, .LBB5_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -328,7 +319,7 @@
 define void @scatter_masked(ptr noalias nocapture %A, ptr noalias nocapture readonly %B, <32 x i8> %maskedoff) {
 ; V-LABEL: scatter_masked:
 ; V:       # %bb.0: # %entry
-; V-NEXT:    li a2, 1024
+; V-NEXT:    addi a2, a1, 1024
 ; V-NEXT:    li a3, 32
 ; V-NEXT:    lui a4, 983765
 ; V-NEXT:    addiw a4, a4, 873
@@ -343,16 +334,15 @@
 ; V-NEXT:    vlse8.v v10, (a0), a4, v0.t
 ; V-NEXT:    vadd.vv v9, v10, v9
 ; V-NEXT:    vsse8.v v9, (a0), a4, v0.t
-; V-NEXT:    addi a2, a2, -32
 ; V-NEXT:    addi a1, a1, 32
 ; V-NEXT:    addi a0, a0, 160
-; V-NEXT:    bnez a2, .LBB6_1
+; V-NEXT:    bne a1, a2, .LBB6_1
 ; V-NEXT:  # %bb.2: # %for.cond.cleanup
 ; V-NEXT:    ret
 ;
 ; ZVE32F-LABEL: scatter_masked:
 ; ZVE32F:       # %bb.0: # %entry
-; ZVE32F-NEXT:    li a2, 1024
+; ZVE32F-NEXT:    addi a2, a1, 1024
 ; ZVE32F-NEXT:    li a3, 32
 ; ZVE32F-NEXT:    lui a4, 983765
 ; ZVE32F-NEXT:    addiw a4, a4, 873
@@ -367,10 +357,9 @@
 ; ZVE32F-NEXT:    vlse8.v v10, (a0), a4, v0.t
 ; ZVE32F-NEXT:    vadd.vv v9, v10, v9
 ; ZVE32F-NEXT:    vsse8.v v9, (a0), a4, v0.t
-; ZVE32F-NEXT:    addi a2, a2, -32
 ; ZVE32F-NEXT:    addi a1, a1, 32
 ; ZVE32F-NEXT:    addi a0, a0, 160
-; ZVE32F-NEXT:    bnez a2, .LBB6_1
+; ZVE32F-NEXT:    bne a1, a2, .LBB6_1
 ; ZVE32F-NEXT:  # %bb.2: # %for.cond.cleanup
 ; ZVE32F-NEXT:    ret
 entry:
@@ -402,7 +391,8 @@
 define void @gather_pow2(ptr noalias nocapture %A, ptr noalias nocapture readonly %B) {
 ; CHECK-LABEL: gather_pow2:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    lui a2, 1
+; CHECK-NEXT:    add a2, a0, a2
 ; CHECK-NEXT:    li a3, 16
 ; CHECK-NEXT:    li a4, 32
 ; CHECK-NEXT:  .LBB7_1: # %vector.body
@@ -415,10 +405,9 @@
 ; CHECK-NEXT:    vadd.vv v8, v9, v8
 ; CHECK-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-NEXT:    vse8.v v8, (a0)
-; CHECK-NEXT:    addi a2, a2, -8
 ; CHECK-NEXT:    addi a0, a0, 32
 ; CHECK-NEXT:    addi a1, a1, 128
-; CHECK-NEXT:    bnez a2, .LBB7_1
+; CHECK-NEXT:    bne a0, a2, .LBB7_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -450,7 +439,8 @@
 define void @scatter_pow2(ptr noalias nocapture %A, ptr noalias nocapture readonly %B) {
 ; CHECK-LABEL: scatter_pow2:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    lui a2, 1
+; CHECK-NEXT:    add a2, a1, a2
 ; CHECK-NEXT:    li a3, 32
 ; CHECK-NEXT:    li a4, 16
 ; CHECK-NEXT:  .LBB8_1: # %vector.body
@@ -461,10 +451,9 @@
 ; CHECK-NEXT:    vlse32.v v9, (a0), a4
 ; CHECK-NEXT:    vadd.vv v8, v9, v8
 ; CHECK-NEXT:    vsse32.v v8, (a0), a4
-; CHECK-NEXT:    addi a2, a2, -8
 ; CHECK-NEXT:    addi a1, a1, 32
 ; CHECK-NEXT:    addi a0, a0, 128
-; CHECK-NEXT:    bnez a2, .LBB8_1
+; CHECK-NEXT:    bne a1, a2, .LBB8_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -501,7 +490,8 @@
 ; CHECK-LABEL: struct_gather:
 ; CHECK:       # %bb.0: # %entry
 ; CHECK-NEXT:    addi a1, a1, 132
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    lui a2, 1
+; CHECK-NEXT:    add a2, a0, a2
 ; CHECK-NEXT:    li a3, 16
 ; CHECK-NEXT:    vsetivli zero, 8, e32, m1, ta, ma
 ; CHECK-NEXT:  .LBB9_1: # %vector.body
@@ -516,10 +506,9 @@
 ; CHECK-NEXT:    vadd.vv v9, v11, v9
 ; CHECK-NEXT:    vse32.v v8, (a0)
 ; CHECK-NEXT:    vse32.v v9, (a4)
-; CHECK-NEXT:    addi a2, a2, -16
 ; CHECK-NEXT:    addi a0, a0, 64
 ; CHECK-NEXT:    addi a1, a1, 256
-; CHECK-NEXT:    bnez a2, .LBB9_1
+; CHECK-NEXT:    bne a0, a2, .LBB9_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -650,7 +639,8 @@
 define void @gather_of_pointers(ptr noalias nocapture %arg, ptr noalias nocapture readonly %arg1) {
 ; V-LABEL: gather_of_pointers:
 ; V:       # %bb.0: # %bb
-; V-NEXT:    li a2, 1024
+; V-NEXT:    lui a2, 2
+; V-NEXT:    add a2, a0, a2
 ; V-NEXT:    li a3, 40
 ; V-NEXT:    vsetivli zero, 2, e64, m1, ta, ma
 ; V-NEXT:  .LBB11_1: # %bb2
@@ -661,22 +651,22 @@
 ; V-NEXT:    addi a4, a0, 16
 ; V-NEXT:    vse64.v v8, (a0)
 ; V-NEXT:    vse64.v v9, (a4)
-; V-NEXT:    addi a2, a2, -4
 ; V-NEXT:    addi a0, a0, 32
 ; V-NEXT:    addi a1, a1, 160
-; V-NEXT:    bnez a2, .LBB11_1
+; V-NEXT:    bne a0, a2, .LBB11_1
 ; V-NEXT:  # %bb.2: # %bb18
 ; V-NEXT:    ret
 ;
 ; ZVE32F-LABEL: gather_of_pointers:
 ; ZVE32F:       # %bb.0: # %bb
 ; ZVE32F-NEXT:    li a2, 0
-; ZVE32F-NEXT:    li a3, 1
-; ZVE32F-NEXT:    li a4, 1024
+; ZVE32F-NEXT:    lui a3, 2
+; ZVE32F-NEXT:    add a3, a0, a3
+; ZVE32F-NEXT:    li a4, 1
 ; ZVE32F-NEXT:    li a5, 40
 ; ZVE32F-NEXT:  .LBB11_1: # %bb2
 ; ZVE32F-NEXT:    # =>This Inner Loop Header: Depth=1
-; ZVE32F-NEXT:    mul a6, a3, a5
+; ZVE32F-NEXT:    mul a6, a4, a5
 ; ZVE32F-NEXT:    add a6, a1, a6
 ; ZVE32F-NEXT:    mul a7, a2, a5
 ; ZVE32F-NEXT:    add a7, a1, a7
@@ -689,10 +679,9 @@
 ; ZVE32F-NEXT:    sd a6, 24(a0)
 ; ZVE32F-NEXT:    sd a7, 16(a0)
 ; ZVE32F-NEXT:    addi a2, a2, 4
-; ZVE32F-NEXT:    addi a3, a3, 4
-; ZVE32F-NEXT:    addi a4, a4, -4
 ; ZVE32F-NEXT:    addi a0, a0, 32
-; ZVE32F-NEXT:    bnez a4, .LBB11_1
+; ZVE32F-NEXT:    addi a4, a4, 4
+; ZVE32F-NEXT:    bne a0, a3, .LBB11_1
 ; ZVE32F-NEXT:  # %bb.2: # %bb18
 ; ZVE32F-NEXT:    ret
 bb:
@@ -727,7 +716,8 @@
 define void @scatter_of_pointers(ptr noalias nocapture %arg, ptr noalias nocapture readonly %arg1) {
 ; V-LABEL: scatter_of_pointers:
 ; V:       # %bb.0: # %bb
-; V-NEXT:    li a2, 1024
+; V-NEXT:    lui a2, 2
+; V-NEXT:    add a2, a1, a2
 ; V-NEXT:    li a3, 40
 ; V-NEXT:    vsetivli zero, 2, e64, m1, ta, ma
 ; V-NEXT:  .LBB12_1: # %bb2
@@ -738,18 +728,18 @@
 ; V-NEXT:    addi a4, a0, 80
 ; V-NEXT:    vsse64.v v8, (a0), a3
 ; V-NEXT:    vsse64.v v9, (a4), a3
-; V-NEXT:    addi a2, a2, -4
 ; V-NEXT:    addi a1, a1, 32
 ; V-NEXT:    addi a0, a0, 160
-; V-NEXT:    bnez a2, .LBB12_1
+; V-NEXT:    bne a1, a2, .LBB12_1
 ; V-NEXT:  # %bb.2: # %bb18
 ; V-NEXT:    ret
 ;
 ; ZVE32F-LABEL: scatter_of_pointers:
 ; ZVE32F:       # %bb.0: # %bb
 ; ZVE32F-NEXT:    li a2, 0
-; ZVE32F-NEXT:    li a3, 1
-; ZVE32F-NEXT:    li a4, 1024
+; ZVE32F-NEXT:    lui a3, 2
+; ZVE32F-NEXT:    add a3, a1, a3
+; ZVE32F-NEXT:    li a4, 1
 ; ZVE32F-NEXT:    li a5, 40
 ; ZVE32F-NEXT:  .LBB12_1: # %bb2
 ; ZVE32F-NEXT:    # =>This Inner Loop Header: Depth=1
@@ -757,7 +747,7 @@
 ; ZVE32F-NEXT:    ld a7, 0(a1)
 ; ZVE32F-NEXT:    ld t0, 24(a1)
 ; ZVE32F-NEXT:    ld t1, 16(a1)
-; ZVE32F-NEXT:    mul t2, a3, a5
+; ZVE32F-NEXT:    mul t2, a4, a5
 ; ZVE32F-NEXT:    add t2, a0, t2
 ; ZVE32F-NEXT:    mul t3, a2, a5
 ; ZVE32F-NEXT:    add t3, a0, t3
@@ -766,10 +756,9 @@
 ; ZVE32F-NEXT:    sd t1, 80(t3)
 ; ZVE32F-NEXT:    sd t0, 80(t2)
 ; ZVE32F-NEXT:    addi a2, a2, 4
-; ZVE32F-NEXT:    addi a3, a3, 4
-; ZVE32F-NEXT:    addi a4, a4, -4
 ; ZVE32F-NEXT:    addi a1, a1, 32
-; ZVE32F-NEXT:    bnez a4, .LBB12_1
+; ZVE32F-NEXT:    addi a4, a4, 4
+; ZVE32F-NEXT:    bne a1, a3, .LBB12_1
 ; ZVE32F-NEXT:  # %bb.2: # %bb18
 ; ZVE32F-NEXT:    ret
 bb:
@@ -806,53 +795,56 @@
 ; CHECK-NEXT:    li a3, 1024
 ; CHECK-NEXT:    beq a2, a3, .LBB13_7
 ; CHECK-NEXT:  # %bb.1: # %bb3
-; CHECK-NEXT:    li a4, 1023
-; CHECK-NEXT:    subw a4, a4, a2
-; CHECK-NEXT:    li a5, 31
-; CHECK-NEXT:    mv a3, a2
-; CHECK-NEXT:    bltu a4, a5, .LBB13_5
+; CHECK-NEXT:    li a3, 1023
+; CHECK-NEXT:    subw a5, a3, a2
+; CHECK-NEXT:    li a6, 31
+; CHECK-NEXT:    mv a4, a2
+; CHECK-NEXT:    bltu a5, a6, .LBB13_5
 ; CHECK-NEXT:  # %bb.2: # %bb9
-; CHECK-NEXT:    slli a4, a4, 32
-; CHECK-NEXT:    srli a4, a4, 32
-; CHECK-NEXT:    addi a4, a4, 1
-; CHECK-NEXT:    andi a5, a4, -32
-; CHECK-NEXT:    add a3, a5, a2
-; CHECK-NEXT:    slli a7, a2, 2
-; CHECK-NEXT:    add a6, a0, a2
+; CHECK-NEXT:    slli a5, a5, 32
+; CHECK-NEXT:    srli a5, a5, 32
+; CHECK-NEXT:    addi a5, a5, 1
+; CHECK-NEXT:    andi a6, a5, -32
+; CHECK-NEXT:    add a4, a6, a2
+; CHECK-NEXT:    slli t0, a2, 2
+; CHECK-NEXT:    add a7, a0, a2
 ; CHECK-NEXT:    add a2, a1, a2
-; CHECK-NEXT:    add a2, a2, a7
-; CHECK-NEXT:    li a7, 32
-; CHECK-NEXT:    li t0, 5
-; CHECK-NEXT:    mv t1, a5
+; CHECK-NEXT:    add a2, a2, t0
+; CHECK-NEXT:    add t0, a4, a0
+; CHECK-NEXT:    li t1, 32
+; CHECK-NEXT:    li t2, 5
 ; CHECK-NEXT:  .LBB13_3: # %bb15
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
-; CHECK-NEXT:    vsetvli zero, a7, e8, m1, ta, ma
-; CHECK-NEXT:    vlse8.v v8, (a2), t0
-; CHECK-NEXT:    vle8.v v9, (a6)
+; CHECK-NEXT:    vsetvli zero, t1, e8, m1, ta, ma
+; CHECK-NEXT:    vlse8.v v8, (a2), t2
+; CHECK-NEXT:    vle8.v v9, (a7)
 ; CHECK-NEXT:    vadd.vv v8, v9, v8
-; CHECK-NEXT:    vse8.v v8, (a6)
-; CHECK-NEXT:    addi t1, t1, -32
-; CHECK-NEXT:    addi a6, a6, 32
+; CHECK-NEXT:    vse8.v v8, (a7)
+; CHECK-NEXT:    addi a7, a7, 32
 ; CHECK-NEXT:    addi a2, a2, 160
-; CHECK-NEXT:    bnez t1, .LBB13_3
+; CHECK-NEXT:    bne a7, t0, .LBB13_3
 ; CHECK-NEXT:  # %bb.4: # %bb30
-; CHECK-NEXT:    beq a4, a5, .LBB13_7
+; CHECK-NEXT:    beq a5, a6, .LBB13_7
 ; CHECK-NEXT:  .LBB13_5: # %bb32
-; CHECK-NEXT:    addiw a2, a3, -1024
-; CHECK-NEXT:    add a0, a0, a3
-; CHECK-NEXT:    slli a4, a3, 2
-; CHECK-NEXT:    add a1, a1, a3
+; CHECK-NEXT:    add a2, a0, a4
+; CHECK-NEXT:    slli a5, a4, 2
 ; CHECK-NEXT:    add a1, a1, a4
+; CHECK-NEXT:    add a1, a1, a5
+; CHECK-NEXT:    subw a3, a3, a4
+; CHECK-NEXT:    slli a3, a3, 32
+; CHECK-NEXT:    srli a3, a3, 32
+; CHECK-NEXT:    add a0, a4, a0
+; CHECK-NEXT:    add a0, a0, a3
+; CHECK-NEXT:    addi a0, a0, 1
 ; CHECK-NEXT:  .LBB13_6: # %bb35
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
 ; CHECK-NEXT:    lbu a3, 0(a1)
-; CHECK-NEXT:    lbu a4, 0(a0)
+; CHECK-NEXT:    lbu a4, 0(a2)
 ; CHECK-NEXT:    add a3, a4, a3
-; CHECK-NEXT:    sb a3, 0(a0)
-; CHECK-NEXT:    addiw a2, a2, 1
-; CHECK-NEXT:    addi a0, a0, 1
+; CHECK-NEXT:    sb a3, 0(a2)
+; CHECK-NEXT:    addi a2, a2, 1
 ; CHECK-NEXT:    addi a1, a1, 5
-; CHECK-NEXT:    bnez a2, .LBB13_6
+; CHECK-NEXT:    bne a2, a0, .LBB13_6
 ; CHECK-NEXT:  .LBB13_7: # %bb34
 ; CHECK-NEXT:    ret
 bb:
@@ -926,6 +918,10 @@
 ; CHECK-NEXT:    slli a2, a2, 4
 ; CHECK-NEXT:    beqz a2, .LBB14_3
 ; CHECK-NEXT:  # %bb.1: # %bb2
+; CHECK-NEXT:    addi a2, a2, -16
+; CHECK-NEXT:    andi a2, a2, -16
+; CHECK-NEXT:    add a2, a2, a0
+; CHECK-NEXT:    addi a2, a2, 16
 ; CHECK-NEXT:    li a3, 5
 ; CHECK-NEXT:    vsetivli zero, 16, e8, mf2, ta, ma
 ; CHECK-NEXT:  .LBB14_2: # %bb4
@@ -934,10 +930,9 @@
 ; CHECK-NEXT:    vle8.v v9, (a0)
 ; CHECK-NEXT:    vadd.vv v8, v9, v8
 ; CHECK-NEXT:    vse8.v v8, (a0)
-; CHECK-NEXT:    addi a2, a2, -16
 ; CHECK-NEXT:    addi a0, a0, 16
 ; CHECK-NEXT:    addi a1, a1, 80
-; CHECK-NEXT:    bnez a2, .LBB14_2
+; CHECK-NEXT:    bne a0, a2, .LBB14_2
 ; CHECK-NEXT:  .LBB14_3: # %bb16
 ; CHECK-NEXT:    ret
 bb:
diff --git a/llvm/test/CodeGen/RISCV/rvv/sink-splat-operands.ll b/llvm/test/CodeGen/RISCV/rvv/sink-splat-operands.ll
--- a/llvm/test/CodeGen/RISCV/rvv/sink-splat-operands.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/sink-splat-operands.ll
@@ -5,16 +5,16 @@
 define void @sink_splat_mul(ptr nocapture %a, i32 signext %x) {
 ; CHECK-LABEL: sink_splat_mul:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    lui a2, 1
+; CHECK-NEXT:    add a2, a0, a2
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:  .LBB0_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
 ; CHECK-NEXT:    vle32.v v8, (a0)
 ; CHECK-NEXT:    vmul.vx v8, v8, a1
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a2, a2, -4
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a2, .LBB0_1
+; CHECK-NEXT:    bne a0, a2, .LBB0_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -39,16 +39,16 @@
 define void @sink_splat_add(ptr nocapture %a, i32 signext %x) {
 ; CHECK-LABEL: sink_splat_add:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    lui a2, 1
+; CHECK-NEXT:    add a2, a0, a2
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:  .LBB1_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
 ; CHECK-NEXT:    vle32.v v8, (a0)
 ; CHECK-NEXT:    vadd.vx v8, v8, a1
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a2, a2, -4
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a2, .LBB1_1
+; CHECK-NEXT:    bne a0, a2, .LBB1_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -73,16 +73,16 @@
 define void @sink_splat_sub(ptr nocapture %a, i32 signext %x) {
 ; CHECK-LABEL: sink_splat_sub:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    lui a2, 1
+; CHECK-NEXT:    add a2, a0, a2
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:  .LBB2_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
 ; CHECK-NEXT:    vle32.v v8, (a0)
 ; CHECK-NEXT:    vsub.vx v8, v8, a1
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a2, a2, -4
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a2, .LBB2_1
+; CHECK-NEXT:    bne a0, a2, .LBB2_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -107,16 +107,16 @@
 define void @sink_splat_rsub(ptr nocapture %a, i32 signext %x) {
 ; CHECK-LABEL: sink_splat_rsub:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    lui a2, 1
+; CHECK-NEXT:    add a2, a0, a2
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:  .LBB3_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
 ; CHECK-NEXT:    vle32.v v8, (a0)
 ; CHECK-NEXT:    vrsub.vx v8, v8, a1
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a2, a2, -4
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a2, .LBB3_1
+; CHECK-NEXT:    bne a0, a2, .LBB3_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -141,16 +141,16 @@
 define void @sink_splat_and(ptr nocapture %a, i32 signext %x) {
 ; CHECK-LABEL: sink_splat_and:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    lui a2, 1
+; CHECK-NEXT:    add a2, a0, a2
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:  .LBB4_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
 ; CHECK-NEXT:    vle32.v v8, (a0)
 ; CHECK-NEXT:    vand.vx v8, v8, a1
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a2, a2, -4
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a2, .LBB4_1
+; CHECK-NEXT:    bne a0, a2, .LBB4_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -175,16 +175,16 @@
 define void @sink_splat_or(ptr nocapture %a, i32 signext %x) {
 ; CHECK-LABEL: sink_splat_or:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    lui a2, 1
+; CHECK-NEXT:    add a2, a0, a2
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:  .LBB5_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
 ; CHECK-NEXT:    vle32.v v8, (a0)
 ; CHECK-NEXT:    vor.vx v8, v8, a1
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a2, a2, -4
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a2, .LBB5_1
+; CHECK-NEXT:    bne a0, a2, .LBB5_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -209,16 +209,16 @@
 define void @sink_splat_xor(ptr nocapture %a, i32 signext %x) {
 ; CHECK-LABEL: sink_splat_xor:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    lui a2, 1
+; CHECK-NEXT:    add a2, a0, a2
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:  .LBB6_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
 ; CHECK-NEXT:    vle32.v v8, (a0)
 ; CHECK-NEXT:    vxor.vx v8, v8, a1
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a2, a2, -4
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a2, .LBB6_1
+; CHECK-NEXT:    bne a0, a2, .LBB6_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -269,17 +269,17 @@
 ; CHECK-NEXT:  # %bb.4: # %middle.block
 ; CHECK-NEXT:    beqz a4, .LBB7_7
 ; CHECK-NEXT:  .LBB7_5: # %for.body.preheader
-; CHECK-NEXT:    addi a3, a2, -1024
 ; CHECK-NEXT:    slli a2, a2, 2
-; CHECK-NEXT:    add a0, a0, a2
+; CHECK-NEXT:    add a2, a0, a2
+; CHECK-NEXT:    lui a3, 1
+; CHECK-NEXT:    add a0, a0, a3
 ; CHECK-NEXT:  .LBB7_6: # %for.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
-; CHECK-NEXT:    lw a2, 0(a0)
-; CHECK-NEXT:    mul a2, a2, a1
-; CHECK-NEXT:    sw a2, 0(a0)
-; CHECK-NEXT:    addi a3, a3, 1
-; CHECK-NEXT:    addi a0, a0, 4
-; CHECK-NEXT:    bnez a3, .LBB7_6
+; CHECK-NEXT:    lw a3, 0(a2)
+; CHECK-NEXT:    mul a3, a3, a1
+; CHECK-NEXT:    sw a3, 0(a2)
+; CHECK-NEXT:    addi a2, a2, 4
+; CHECK-NEXT:    bne a2, a0, .LBB7_6
 ; CHECK-NEXT:  .LBB7_7: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -360,17 +360,17 @@
 ; CHECK-NEXT:  # %bb.4: # %middle.block
 ; CHECK-NEXT:    beqz a4, .LBB8_7
 ; CHECK-NEXT:  .LBB8_5: # %for.body.preheader
-; CHECK-NEXT:    addi a3, a2, -1024
 ; CHECK-NEXT:    slli a2, a2, 2
-; CHECK-NEXT:    add a0, a0, a2
+; CHECK-NEXT:    add a2, a0, a2
+; CHECK-NEXT:    lui a3, 1
+; CHECK-NEXT:    add a0, a0, a3
 ; CHECK-NEXT:  .LBB8_6: # %for.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
-; CHECK-NEXT:    lw a2, 0(a0)
-; CHECK-NEXT:    add a2, a2, a1
-; CHECK-NEXT:    sw a2, 0(a0)
-; CHECK-NEXT:    addi a3, a3, 1
-; CHECK-NEXT:    addi a0, a0, 4
-; CHECK-NEXT:    bnez a3, .LBB8_6
+; CHECK-NEXT:    lw a3, 0(a2)
+; CHECK-NEXT:    add a3, a3, a1
+; CHECK-NEXT:    sw a3, 0(a2)
+; CHECK-NEXT:    addi a2, a2, 4
+; CHECK-NEXT:    bne a2, a0, .LBB8_6
 ; CHECK-NEXT:  .LBB8_7: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -451,17 +451,17 @@
 ; CHECK-NEXT:  # %bb.4: # %middle.block
 ; CHECK-NEXT:    beqz a4, .LBB9_7
 ; CHECK-NEXT:  .LBB9_5: # %for.body.preheader
-; CHECK-NEXT:    addi a3, a2, -1024
 ; CHECK-NEXT:    slli a2, a2, 2
-; CHECK-NEXT:    add a0, a0, a2
+; CHECK-NEXT:    add a2, a0, a2
+; CHECK-NEXT:    lui a3, 1
+; CHECK-NEXT:    add a0, a0, a3
 ; CHECK-NEXT:  .LBB9_6: # %for.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
-; CHECK-NEXT:    lw a2, 0(a0)
-; CHECK-NEXT:    add a2, a2, a1
-; CHECK-NEXT:    sw a2, 0(a0)
-; CHECK-NEXT:    addi a3, a3, 1
-; CHECK-NEXT:    addi a0, a0, 4
-; CHECK-NEXT:    bnez a3, .LBB9_6
+; CHECK-NEXT:    lw a3, 0(a2)
+; CHECK-NEXT:    add a3, a3, a1
+; CHECK-NEXT:    sw a3, 0(a2)
+; CHECK-NEXT:    addi a2, a2, 4
+; CHECK-NEXT:    bne a2, a0, .LBB9_6
 ; CHECK-NEXT:  .LBB9_7: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -542,17 +542,17 @@
 ; CHECK-NEXT:  # %bb.4: # %middle.block
 ; CHECK-NEXT:    beqz a4, .LBB10_7
 ; CHECK-NEXT:  .LBB10_5: # %for.body.preheader
-; CHECK-NEXT:    addi a3, a2, -1024
 ; CHECK-NEXT:    slli a2, a2, 2
-; CHECK-NEXT:    add a0, a0, a2
+; CHECK-NEXT:    add a2, a0, a2
+; CHECK-NEXT:    lui a3, 1
+; CHECK-NEXT:    add a0, a0, a3
 ; CHECK-NEXT:  .LBB10_6: # %for.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
-; CHECK-NEXT:    lw a2, 0(a0)
-; CHECK-NEXT:    subw a2, a1, a2
-; CHECK-NEXT:    sw a2, 0(a0)
-; CHECK-NEXT:    addi a3, a3, 1
-; CHECK-NEXT:    addi a0, a0, 4
-; CHECK-NEXT:    bnez a3, .LBB10_6
+; CHECK-NEXT:    lw a3, 0(a2)
+; CHECK-NEXT:    subw a3, a1, a3
+; CHECK-NEXT:    sw a3, 0(a2)
+; CHECK-NEXT:    addi a2, a2, 4
+; CHECK-NEXT:    bne a2, a0, .LBB10_6
 ; CHECK-NEXT:  .LBB10_7: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -633,17 +633,17 @@
 ; CHECK-NEXT:  # %bb.4: # %middle.block
 ; CHECK-NEXT:    beqz a4, .LBB11_7
 ; CHECK-NEXT:  .LBB11_5: # %for.body.preheader
-; CHECK-NEXT:    addi a3, a2, -1024
 ; CHECK-NEXT:    slli a2, a2, 2
-; CHECK-NEXT:    add a0, a0, a2
+; CHECK-NEXT:    add a2, a0, a2
+; CHECK-NEXT:    lui a3, 1
+; CHECK-NEXT:    add a0, a0, a3
 ; CHECK-NEXT:  .LBB11_6: # %for.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
-; CHECK-NEXT:    lw a2, 0(a0)
-; CHECK-NEXT:    and a2, a2, a1
-; CHECK-NEXT:    sw a2, 0(a0)
-; CHECK-NEXT:    addi a3, a3, 1
-; CHECK-NEXT:    addi a0, a0, 4
-; CHECK-NEXT:    bnez a3, .LBB11_6
+; CHECK-NEXT:    lw a3, 0(a2)
+; CHECK-NEXT:    and a3, a3, a1
+; CHECK-NEXT:    sw a3, 0(a2)
+; CHECK-NEXT:    addi a2, a2, 4
+; CHECK-NEXT:    bne a2, a0, .LBB11_6
 ; CHECK-NEXT:  .LBB11_7: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -724,17 +724,17 @@
 ; CHECK-NEXT:  # %bb.4: # %middle.block
 ; CHECK-NEXT:    beqz a4, .LBB12_7
 ; CHECK-NEXT:  .LBB12_5: # %for.body.preheader
-; CHECK-NEXT:    addi a3, a2, -1024
 ; CHECK-NEXT:    slli a2, a2, 2
-; CHECK-NEXT:    add a0, a0, a2
+; CHECK-NEXT:    add a2, a0, a2
+; CHECK-NEXT:    lui a3, 1
+; CHECK-NEXT:    add a0, a0, a3
 ; CHECK-NEXT:  .LBB12_6: # %for.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
-; CHECK-NEXT:    lw a2, 0(a0)
-; CHECK-NEXT:    or a2, a2, a1
-; CHECK-NEXT:    sw a2, 0(a0)
-; CHECK-NEXT:    addi a3, a3, 1
-; CHECK-NEXT:    addi a0, a0, 4
-; CHECK-NEXT:    bnez a3, .LBB12_6
+; CHECK-NEXT:    lw a3, 0(a2)
+; CHECK-NEXT:    or a3, a3, a1
+; CHECK-NEXT:    sw a3, 0(a2)
+; CHECK-NEXT:    addi a2, a2, 4
+; CHECK-NEXT:    bne a2, a0, .LBB12_6
 ; CHECK-NEXT:  .LBB12_7: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -815,17 +815,17 @@
 ; CHECK-NEXT:  # %bb.4: # %middle.block
 ; CHECK-NEXT:    beqz a4, .LBB13_7
 ; CHECK-NEXT:  .LBB13_5: # %for.body.preheader
-; CHECK-NEXT:    addi a3, a2, -1024
 ; CHECK-NEXT:    slli a2, a2, 2
-; CHECK-NEXT:    add a0, a0, a2
+; CHECK-NEXT:    add a2, a0, a2
+; CHECK-NEXT:    lui a3, 1
+; CHECK-NEXT:    add a0, a0, a3
 ; CHECK-NEXT:  .LBB13_6: # %for.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
-; CHECK-NEXT:    lw a2, 0(a0)
-; CHECK-NEXT:    xor a2, a2, a1
-; CHECK-NEXT:    sw a2, 0(a0)
-; CHECK-NEXT:    addi a3, a3, 1
-; CHECK-NEXT:    addi a0, a0, 4
-; CHECK-NEXT:    bnez a3, .LBB13_6
+; CHECK-NEXT:    lw a3, 0(a2)
+; CHECK-NEXT:    xor a3, a3, a1
+; CHECK-NEXT:    sw a3, 0(a2)
+; CHECK-NEXT:    addi a2, a2, 4
+; CHECK-NEXT:    bne a2, a0, .LBB13_6
 ; CHECK-NEXT:  .LBB13_7: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -880,16 +880,16 @@
 define void @sink_splat_shl(ptr nocapture %a, i32 signext %x) {
 ; CHECK-LABEL: sink_splat_shl:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    lui a2, 1
+; CHECK-NEXT:    add a2, a0, a2
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:  .LBB14_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
 ; CHECK-NEXT:    vle32.v v8, (a0)
 ; CHECK-NEXT:    vsll.vx v8, v8, a1
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a2, a2, -4
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a2, .LBB14_1
+; CHECK-NEXT:    bne a0, a2, .LBB14_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -914,16 +914,16 @@
 define void @sink_splat_lshr(ptr nocapture %a, i32 signext %x) {
 ; CHECK-LABEL: sink_splat_lshr:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    lui a2, 1
+; CHECK-NEXT:    add a2, a0, a2
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:  .LBB15_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
 ; CHECK-NEXT:    vle32.v v8, (a0)
 ; CHECK-NEXT:    vsrl.vx v8, v8, a1
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a2, a2, -4
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a2, .LBB15_1
+; CHECK-NEXT:    bne a0, a2, .LBB15_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -948,16 +948,16 @@
 define void @sink_splat_ashr(ptr nocapture %a, i32 signext %x) {
 ; CHECK-LABEL: sink_splat_ashr:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    lui a2, 1
+; CHECK-NEXT:    add a2, a0, a2
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:  .LBB16_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
 ; CHECK-NEXT:    vle32.v v8, (a0)
 ; CHECK-NEXT:    vsra.vx v8, v8, a1
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a2, a2, -4
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a2, .LBB16_1
+; CHECK-NEXT:    bne a0, a2, .LBB16_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -1008,17 +1008,17 @@
 ; CHECK-NEXT:  # %bb.4: # %middle.block
 ; CHECK-NEXT:    beqz a4, .LBB17_7
 ; CHECK-NEXT:  .LBB17_5: # %for.body.preheader
-; CHECK-NEXT:    addi a3, a2, -1024
 ; CHECK-NEXT:    slli a2, a2, 2
-; CHECK-NEXT:    add a0, a0, a2
+; CHECK-NEXT:    add a2, a0, a2
+; CHECK-NEXT:    lui a3, 1
+; CHECK-NEXT:    add a0, a0, a3
 ; CHECK-NEXT:  .LBB17_6: # %for.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
-; CHECK-NEXT:    lw a2, 0(a0)
-; CHECK-NEXT:    sllw a2, a2, a1
-; CHECK-NEXT:    sw a2, 0(a0)
-; CHECK-NEXT:    addi a3, a3, 1
-; CHECK-NEXT:    addi a0, a0, 4
-; CHECK-NEXT:    bnez a3, .LBB17_6
+; CHECK-NEXT:    lw a3, 0(a2)
+; CHECK-NEXT:    sllw a3, a3, a1
+; CHECK-NEXT:    sw a3, 0(a2)
+; CHECK-NEXT:    addi a2, a2, 4
+; CHECK-NEXT:    bne a2, a0, .LBB17_6
 ; CHECK-NEXT:  .LBB17_7: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -1099,17 +1099,17 @@
 ; CHECK-NEXT:  # %bb.4: # %middle.block
 ; CHECK-NEXT:    beqz a4, .LBB18_7
 ; CHECK-NEXT:  .LBB18_5: # %for.body.preheader
-; CHECK-NEXT:    addi a3, a2, -1024
 ; CHECK-NEXT:    slli a2, a2, 2
-; CHECK-NEXT:    add a0, a0, a2
+; CHECK-NEXT:    add a2, a0, a2
+; CHECK-NEXT:    lui a3, 1
+; CHECK-NEXT:    add a0, a0, a3
 ; CHECK-NEXT:  .LBB18_6: # %for.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
-; CHECK-NEXT:    lw a2, 0(a0)
-; CHECK-NEXT:    srlw a2, a2, a1
-; CHECK-NEXT:    sw a2, 0(a0)
-; CHECK-NEXT:    addi a3, a3, 1
-; CHECK-NEXT:    addi a0, a0, 4
-; CHECK-NEXT:    bnez a3, .LBB18_6
+; CHECK-NEXT:    lw a3, 0(a2)
+; CHECK-NEXT:    srlw a3, a3, a1
+; CHECK-NEXT:    sw a3, 0(a2)
+; CHECK-NEXT:    addi a2, a2, 4
+; CHECK-NEXT:    bne a2, a0, .LBB18_6
 ; CHECK-NEXT:  .LBB18_7: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -1190,17 +1190,17 @@
 ; CHECK-NEXT:  # %bb.4: # %middle.block
 ; CHECK-NEXT:    beqz a3, .LBB19_7
 ; CHECK-NEXT:  .LBB19_5: # %for.body.preheader
-; CHECK-NEXT:    addi a2, a1, -1024
 ; CHECK-NEXT:    slli a1, a1, 2
-; CHECK-NEXT:    add a0, a0, a1
+; CHECK-NEXT:    add a1, a0, a1
+; CHECK-NEXT:    lui a2, 1
+; CHECK-NEXT:    add a0, a0, a2
 ; CHECK-NEXT:  .LBB19_6: # %for.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
-; CHECK-NEXT:    lw a1, 0(a0)
-; CHECK-NEXT:    srli a1, a1, 2
-; CHECK-NEXT:    sw a1, 0(a0)
-; CHECK-NEXT:    addi a2, a2, 1
-; CHECK-NEXT:    addi a0, a0, 4
-; CHECK-NEXT:    bnez a2, .LBB19_6
+; CHECK-NEXT:    lw a2, 0(a1)
+; CHECK-NEXT:    srli a2, a2, 2
+; CHECK-NEXT:    sw a2, 0(a1)
+; CHECK-NEXT:    addi a1, a1, 4
+; CHECK-NEXT:    bne a1, a0, .LBB19_6
 ; CHECK-NEXT:  .LBB19_7: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -1255,16 +1255,16 @@
 define void @sink_splat_fmul(ptr nocapture %a, float %x) {
 ; CHECK-LABEL: sink_splat_fmul:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a1, 1024
+; CHECK-NEXT:    lui a1, 1
+; CHECK-NEXT:    add a1, a0, a1
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:  .LBB20_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
 ; CHECK-NEXT:    vle32.v v8, (a0)
 ; CHECK-NEXT:    vfmul.vf v8, v8, fa0
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a1, a1, -4
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a1, .LBB20_1
+; CHECK-NEXT:    bne a0, a1, .LBB20_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -1289,16 +1289,16 @@
 define void @sink_splat_fdiv(ptr nocapture %a, float %x) {
 ; CHECK-LABEL: sink_splat_fdiv:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a1, 1024
+; CHECK-NEXT:    lui a1, 1
+; CHECK-NEXT:    add a1, a0, a1
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:  .LBB21_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
 ; CHECK-NEXT:    vle32.v v8, (a0)
 ; CHECK-NEXT:    vfdiv.vf v8, v8, fa0
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a1, a1, -4
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a1, .LBB21_1
+; CHECK-NEXT:    bne a0, a1, .LBB21_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -1323,16 +1323,16 @@
 define void @sink_splat_frdiv(ptr nocapture %a, float %x) {
 ; CHECK-LABEL: sink_splat_frdiv:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a1, 1024
+; CHECK-NEXT:    lui a1, 1
+; CHECK-NEXT:    add a1, a0, a1
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:  .LBB22_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
 ; CHECK-NEXT:    vle32.v v8, (a0)
 ; CHECK-NEXT:    vfrdiv.vf v8, v8, fa0
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a1, a1, -4
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a1, .LBB22_1
+; CHECK-NEXT:    bne a0, a1, .LBB22_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -1357,16 +1357,16 @@
 define void @sink_splat_fadd(ptr nocapture %a, float %x) {
 ; CHECK-LABEL: sink_splat_fadd:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a1, 1024
+; CHECK-NEXT:    lui a1, 1
+; CHECK-NEXT:    add a1, a0, a1
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:  .LBB23_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
 ; CHECK-NEXT:    vle32.v v8, (a0)
 ; CHECK-NEXT:    vfadd.vf v8, v8, fa0
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a1, a1, -4
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a1, .LBB23_1
+; CHECK-NEXT:    bne a0, a1, .LBB23_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -1391,16 +1391,16 @@
 define void @sink_splat_fsub(ptr nocapture %a, float %x) {
 ; CHECK-LABEL: sink_splat_fsub:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a1, 1024
+; CHECK-NEXT:    lui a1, 1
+; CHECK-NEXT:    add a1, a0, a1
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:  .LBB24_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
 ; CHECK-NEXT:    vle32.v v8, (a0)
 ; CHECK-NEXT:    vfsub.vf v8, v8, fa0
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a1, a1, -4
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a1, .LBB24_1
+; CHECK-NEXT:    bne a0, a1, .LBB24_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -1425,16 +1425,16 @@
 define void @sink_splat_frsub(ptr nocapture %a, float %x) {
 ; CHECK-LABEL: sink_splat_frsub:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a1, 1024
+; CHECK-NEXT:    lui a1, 1
+; CHECK-NEXT:    add a1, a0, a1
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:  .LBB25_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
 ; CHECK-NEXT:    vle32.v v8, (a0)
 ; CHECK-NEXT:    vfrsub.vf v8, v8, fa0
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a1, a1, -4
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a1, .LBB25_1
+; CHECK-NEXT:    bne a0, a1, .LBB25_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -1484,17 +1484,17 @@
 ; CHECK-NEXT:  # %bb.4: # %middle.block
 ; CHECK-NEXT:    beqz a4, .LBB26_7
 ; CHECK-NEXT:  .LBB26_5: # %for.body.preheader
-; CHECK-NEXT:    addi a2, a1, -1024
 ; CHECK-NEXT:    slli a1, a1, 2
-; CHECK-NEXT:    add a0, a0, a1
+; CHECK-NEXT:    add a1, a0, a1
+; CHECK-NEXT:    lui a2, 1
+; CHECK-NEXT:    add a0, a0, a2
 ; CHECK-NEXT:  .LBB26_6: # %for.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
-; CHECK-NEXT:    flw fa5, 0(a0)
+; CHECK-NEXT:    flw fa5, 0(a1)
 ; CHECK-NEXT:    fmul.s fa5, fa5, fa0
-; CHECK-NEXT:    fsw fa5, 0(a0)
-; CHECK-NEXT:    addi a2, a2, 1
-; CHECK-NEXT:    addi a0, a0, 4
-; CHECK-NEXT:    bnez a2, .LBB26_6
+; CHECK-NEXT:    fsw fa5, 0(a1)
+; CHECK-NEXT:    addi a1, a1, 4
+; CHECK-NEXT:    bne a1, a0, .LBB26_6
 ; CHECK-NEXT:  .LBB26_7: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -1574,17 +1574,17 @@
 ; CHECK-NEXT:  # %bb.4: # %middle.block
 ; CHECK-NEXT:    beqz a4, .LBB27_7
 ; CHECK-NEXT:  .LBB27_5: # %for.body.preheader
-; CHECK-NEXT:    addi a2, a1, -1024
 ; CHECK-NEXT:    slli a1, a1, 2
-; CHECK-NEXT:    add a0, a0, a1
+; CHECK-NEXT:    add a1, a0, a1
+; CHECK-NEXT:    lui a2, 1
+; CHECK-NEXT:    add a0, a0, a2
 ; CHECK-NEXT:  .LBB27_6: # %for.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
-; CHECK-NEXT:    flw fa5, 0(a0)
+; CHECK-NEXT:    flw fa5, 0(a1)
 ; CHECK-NEXT:    fdiv.s fa5, fa5, fa0
-; CHECK-NEXT:    fsw fa5, 0(a0)
-; CHECK-NEXT:    addi a2, a2, 1
-; CHECK-NEXT:    addi a0, a0, 4
-; CHECK-NEXT:    bnez a2, .LBB27_6
+; CHECK-NEXT:    fsw fa5, 0(a1)
+; CHECK-NEXT:    addi a1, a1, 4
+; CHECK-NEXT:    bne a1, a0, .LBB27_6
 ; CHECK-NEXT:  .LBB27_7: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -1664,17 +1664,17 @@
 ; CHECK-NEXT:  # %bb.4: # %middle.block
 ; CHECK-NEXT:    beqz a4, .LBB28_7
 ; CHECK-NEXT:  .LBB28_5: # %for.body.preheader
-; CHECK-NEXT:    addi a2, a1, -1024
 ; CHECK-NEXT:    slli a1, a1, 2
-; CHECK-NEXT:    add a0, a0, a1
+; CHECK-NEXT:    add a1, a0, a1
+; CHECK-NEXT:    lui a2, 1
+; CHECK-NEXT:    add a0, a0, a2
 ; CHECK-NEXT:  .LBB28_6: # %for.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
-; CHECK-NEXT:    flw fa5, 0(a0)
+; CHECK-NEXT:    flw fa5, 0(a1)
 ; CHECK-NEXT:    fdiv.s fa5, fa0, fa5
-; CHECK-NEXT:    fsw fa5, 0(a0)
-; CHECK-NEXT:    addi a2, a2, 1
-; CHECK-NEXT:    addi a0, a0, 4
-; CHECK-NEXT:    bnez a2, .LBB28_6
+; CHECK-NEXT:    fsw fa5, 0(a1)
+; CHECK-NEXT:    addi a1, a1, 4
+; CHECK-NEXT:    bne a1, a0, .LBB28_6
 ; CHECK-NEXT:  .LBB28_7: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -1754,17 +1754,17 @@
 ; CHECK-NEXT:  # %bb.4: # %middle.block
 ; CHECK-NEXT:    beqz a4, .LBB29_7
 ; CHECK-NEXT:  .LBB29_5: # %for.body.preheader
-; CHECK-NEXT:    addi a2, a1, -1024
 ; CHECK-NEXT:    slli a1, a1, 2
-; CHECK-NEXT:    add a0, a0, a1
+; CHECK-NEXT:    add a1, a0, a1
+; CHECK-NEXT:    lui a2, 1
+; CHECK-NEXT:    add a0, a0, a2
 ; CHECK-NEXT:  .LBB29_6: # %for.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
-; CHECK-NEXT:    flw fa5, 0(a0)
+; CHECK-NEXT:    flw fa5, 0(a1)
 ; CHECK-NEXT:    fadd.s fa5, fa5, fa0
-; CHECK-NEXT:    fsw fa5, 0(a0)
-; CHECK-NEXT:    addi a2, a2, 1
-; CHECK-NEXT:    addi a0, a0, 4
-; CHECK-NEXT:    bnez a2, .LBB29_6
+; CHECK-NEXT:    fsw fa5, 0(a1)
+; CHECK-NEXT:    addi a1, a1, 4
+; CHECK-NEXT:    bne a1, a0, .LBB29_6
 ; CHECK-NEXT:  .LBB29_7: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -1844,17 +1844,17 @@
 ; CHECK-NEXT:  # %bb.4: # %middle.block
 ; CHECK-NEXT:    beqz a4, .LBB30_7
 ; CHECK-NEXT:  .LBB30_5: # %for.body.preheader
-; CHECK-NEXT:    addi a2, a1, -1024
 ; CHECK-NEXT:    slli a1, a1, 2
-; CHECK-NEXT:    add a0, a0, a1
+; CHECK-NEXT:    add a1, a0, a1
+; CHECK-NEXT:    lui a2, 1
+; CHECK-NEXT:    add a0, a0, a2
 ; CHECK-NEXT:  .LBB30_6: # %for.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
-; CHECK-NEXT:    flw fa5, 0(a0)
+; CHECK-NEXT:    flw fa5, 0(a1)
 ; CHECK-NEXT:    fsub.s fa5, fa5, fa0
-; CHECK-NEXT:    fsw fa5, 0(a0)
-; CHECK-NEXT:    addi a2, a2, 1
-; CHECK-NEXT:    addi a0, a0, 4
-; CHECK-NEXT:    bnez a2, .LBB30_6
+; CHECK-NEXT:    fsw fa5, 0(a1)
+; CHECK-NEXT:    addi a1, a1, 4
+; CHECK-NEXT:    bne a1, a0, .LBB30_6
 ; CHECK-NEXT:  .LBB30_7: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -1934,17 +1934,17 @@
 ; CHECK-NEXT:  # %bb.4: # %middle.block
 ; CHECK-NEXT:    beqz a4, .LBB31_7
 ; CHECK-NEXT:  .LBB31_5: # %for.body.preheader
-; CHECK-NEXT:    addi a2, a1, -1024
 ; CHECK-NEXT:    slli a1, a1, 2
-; CHECK-NEXT:    add a0, a0, a1
+; CHECK-NEXT:    add a1, a0, a1
+; CHECK-NEXT:    lui a2, 1
+; CHECK-NEXT:    add a0, a0, a2
 ; CHECK-NEXT:  .LBB31_6: # %for.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
-; CHECK-NEXT:    flw fa5, 0(a0)
+; CHECK-NEXT:    flw fa5, 0(a1)
 ; CHECK-NEXT:    fsub.s fa5, fa0, fa5
-; CHECK-NEXT:    fsw fa5, 0(a0)
-; CHECK-NEXT:    addi a2, a2, 1
-; CHECK-NEXT:    addi a0, a0, 4
-; CHECK-NEXT:    bnez a2, .LBB31_6
+; CHECK-NEXT:    fsw fa5, 0(a1)
+; CHECK-NEXT:    addi a1, a1, 4
+; CHECK-NEXT:    bne a1, a0, .LBB31_6
 ; CHECK-NEXT:  .LBB31_7: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -1999,7 +1999,8 @@
 define void @sink_splat_fma(ptr noalias nocapture %a, ptr nocapture readonly %b, float %x) {
 ; CHECK-LABEL: sink_splat_fma:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    lui a2, 1
+; CHECK-NEXT:    add a2, a1, a2
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:  .LBB32_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
@@ -2007,10 +2008,9 @@
 ; CHECK-NEXT:    vle32.v v9, (a1)
 ; CHECK-NEXT:    vfmacc.vf v9, fa0, v8
 ; CHECK-NEXT:    vse32.v v9, (a0)
-; CHECK-NEXT:    addi a2, a2, -4
 ; CHECK-NEXT:    addi a1, a1, 16
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a2, .LBB32_1
+; CHECK-NEXT:    bne a1, a2, .LBB32_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -2037,7 +2037,8 @@
 define void @sink_splat_fma_commute(ptr noalias nocapture %a, ptr nocapture readonly %b, float %x) {
 ; CHECK-LABEL: sink_splat_fma_commute:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    lui a2, 1
+; CHECK-NEXT:    add a2, a1, a2
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:  .LBB33_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
@@ -2045,10 +2046,9 @@
 ; CHECK-NEXT:    vle32.v v9, (a1)
 ; CHECK-NEXT:    vfmacc.vf v9, fa0, v8
 ; CHECK-NEXT:    vse32.v v9, (a0)
-; CHECK-NEXT:    addi a2, a2, -4
 ; CHECK-NEXT:    addi a1, a1, 16
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a2, .LBB33_1
+; CHECK-NEXT:    bne a1, a2, .LBB33_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -2103,20 +2103,20 @@
 ; CHECK-NEXT:  # %bb.4: # %middle.block
 ; CHECK-NEXT:    beqz a5, .LBB34_7
 ; CHECK-NEXT:  .LBB34_5: # %for.body.preheader
-; CHECK-NEXT:    addi a3, a2, -1024
-; CHECK-NEXT:    slli a2, a2, 2
-; CHECK-NEXT:    add a1, a1, a2
-; CHECK-NEXT:    add a0, a0, a2
+; CHECK-NEXT:    slli a3, a2, 2
+; CHECK-NEXT:    add a2, a1, a3
+; CHECK-NEXT:    add a0, a0, a3
+; CHECK-NEXT:    lui a3, 1
+; CHECK-NEXT:    add a1, a1, a3
 ; CHECK-NEXT:  .LBB34_6: # %for.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
 ; CHECK-NEXT:    flw fa5, 0(a0)
-; CHECK-NEXT:    flw fa4, 0(a1)
+; CHECK-NEXT:    flw fa4, 0(a2)
 ; CHECK-NEXT:    fmadd.s fa5, fa5, fa0, fa4
 ; CHECK-NEXT:    fsw fa5, 0(a0)
-; CHECK-NEXT:    addi a3, a3, 1
-; CHECK-NEXT:    addi a1, a1, 4
+; CHECK-NEXT:    addi a2, a2, 4
 ; CHECK-NEXT:    addi a0, a0, 4
-; CHECK-NEXT:    bnez a3, .LBB34_6
+; CHECK-NEXT:    bne a2, a1, .LBB34_6
 ; CHECK-NEXT:  .LBB34_7: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -2203,20 +2203,20 @@
 ; CHECK-NEXT:  # %bb.4: # %middle.block
 ; CHECK-NEXT:    beqz a5, .LBB35_7
 ; CHECK-NEXT:  .LBB35_5: # %for.body.preheader
-; CHECK-NEXT:    addi a3, a2, -1024
-; CHECK-NEXT:    slli a2, a2, 2
-; CHECK-NEXT:    add a1, a1, a2
-; CHECK-NEXT:    add a0, a0, a2
+; CHECK-NEXT:    slli a3, a2, 2
+; CHECK-NEXT:    add a2, a1, a3
+; CHECK-NEXT:    add a0, a0, a3
+; CHECK-NEXT:    lui a3, 1
+; CHECK-NEXT:    add a1, a1, a3
 ; CHECK-NEXT:  .LBB35_6: # %for.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
 ; CHECK-NEXT:    flw fa5, 0(a0)
-; CHECK-NEXT:    flw fa4, 0(a1)
+; CHECK-NEXT:    flw fa4, 0(a2)
 ; CHECK-NEXT:    fmadd.s fa5, fa0, fa5, fa4
 ; CHECK-NEXT:    fsw fa5, 0(a0)
-; CHECK-NEXT:    addi a3, a3, 1
-; CHECK-NEXT:    addi a1, a1, 4
+; CHECK-NEXT:    addi a2, a2, 4
 ; CHECK-NEXT:    addi a0, a0, 4
-; CHECK-NEXT:    bnez a3, .LBB35_6
+; CHECK-NEXT:    bne a2, a1, .LBB35_6
 ; CHECK-NEXT:  .LBB35_7: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -2280,7 +2280,8 @@
 define void @sink_splat_icmp(ptr nocapture %x, i32 signext %y) {
 ; CHECK-LABEL: sink_splat_icmp:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    lui a2, 1
+; CHECK-NEXT:    add a2, a0, a2
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:    vmv.v.i v8, 0
 ; CHECK-NEXT:  .LBB36_1: # %vector.body
@@ -2288,9 +2289,8 @@
 ; CHECK-NEXT:    vle32.v v9, (a0)
 ; CHECK-NEXT:    vmseq.vx v0, v9, a1
 ; CHECK-NEXT:    vse32.v v8, (a0), v0.t
-; CHECK-NEXT:    addi a2, a2, -4
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a2, .LBB36_1
+; CHECK-NEXT:    bne a0, a2, .LBB36_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -2316,7 +2316,8 @@
 define void @sink_splat_fcmp(ptr nocapture %x, float %y) {
 ; CHECK-LABEL: sink_splat_fcmp:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a1, 1024
+; CHECK-NEXT:    lui a1, 1
+; CHECK-NEXT:    add a1, a0, a1
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:    vmv.v.i v8, 0
 ; CHECK-NEXT:  .LBB37_1: # %vector.body
@@ -2324,9 +2325,8 @@
 ; CHECK-NEXT:    vle32.v v9, (a0)
 ; CHECK-NEXT:    vmfeq.vf v0, v9, fa0
 ; CHECK-NEXT:    vse32.v v8, (a0), v0.t
-; CHECK-NEXT:    addi a1, a1, -4
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a1, .LBB37_1
+; CHECK-NEXT:    bne a0, a1, .LBB37_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -2352,16 +2352,16 @@
 define void @sink_splat_udiv(ptr nocapture %a, i32 signext %x) {
 ; CHECK-LABEL: sink_splat_udiv:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    lui a2, 1
+; CHECK-NEXT:    add a2, a0, a2
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:  .LBB38_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
 ; CHECK-NEXT:    vle32.v v8, (a0)
 ; CHECK-NEXT:    vdivu.vx v8, v8, a1
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a2, a2, -4
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a2, .LBB38_1
+; CHECK-NEXT:    bne a0, a2, .LBB38_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -2386,16 +2386,16 @@
 define void @sink_splat_sdiv(ptr nocapture %a, i32 signext %x) {
 ; CHECK-LABEL: sink_splat_sdiv:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    lui a2, 1
+; CHECK-NEXT:    add a2, a0, a2
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:  .LBB39_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
 ; CHECK-NEXT:    vle32.v v8, (a0)
 ; CHECK-NEXT:    vdiv.vx v8, v8, a1
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a2, a2, -4
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a2, .LBB39_1
+; CHECK-NEXT:    bne a0, a2, .LBB39_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -2420,16 +2420,16 @@
 define void @sink_splat_urem(ptr nocapture %a, i32 signext %x) {
 ; CHECK-LABEL: sink_splat_urem:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    lui a2, 1
+; CHECK-NEXT:    add a2, a0, a2
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:  .LBB40_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
 ; CHECK-NEXT:    vle32.v v8, (a0)
 ; CHECK-NEXT:    vremu.vx v8, v8, a1
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a2, a2, -4
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a2, .LBB40_1
+; CHECK-NEXT:    bne a0, a2, .LBB40_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -2454,16 +2454,16 @@
 define void @sink_splat_srem(ptr nocapture %a, i32 signext %x) {
 ; CHECK-LABEL: sink_splat_srem:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    lui a2, 1
+; CHECK-NEXT:    add a2, a0, a2
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:  .LBB41_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
 ; CHECK-NEXT:    vle32.v v8, (a0)
 ; CHECK-NEXT:    vrem.vx v8, v8, a1
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a2, a2, -4
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a2, .LBB41_1
+; CHECK-NEXT:    bne a0, a2, .LBB41_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -2514,17 +2514,17 @@
 ; CHECK-NEXT:  # %bb.4: # %middle.block
 ; CHECK-NEXT:    beqz a4, .LBB42_7
 ; CHECK-NEXT:  .LBB42_5: # %for.body.preheader
-; CHECK-NEXT:    addi a3, a2, -1024
 ; CHECK-NEXT:    slli a2, a2, 2
-; CHECK-NEXT:    add a0, a0, a2
+; CHECK-NEXT:    add a2, a0, a2
+; CHECK-NEXT:    lui a3, 1
+; CHECK-NEXT:    add a0, a0, a3
 ; CHECK-NEXT:  .LBB42_6: # %for.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
-; CHECK-NEXT:    lw a2, 0(a0)
-; CHECK-NEXT:    divuw a2, a2, a1
-; CHECK-NEXT:    sw a2, 0(a0)
-; CHECK-NEXT:    addi a3, a3, 1
-; CHECK-NEXT:    addi a0, a0, 4
-; CHECK-NEXT:    bnez a3, .LBB42_6
+; CHECK-NEXT:    lw a3, 0(a2)
+; CHECK-NEXT:    divuw a3, a3, a1
+; CHECK-NEXT:    sw a3, 0(a2)
+; CHECK-NEXT:    addi a2, a2, 4
+; CHECK-NEXT:    bne a2, a0, .LBB42_6
 ; CHECK-NEXT:  .LBB42_7: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -2605,17 +2605,17 @@
 ; CHECK-NEXT:  # %bb.4: # %middle.block
 ; CHECK-NEXT:    beqz a4, .LBB43_7
 ; CHECK-NEXT:  .LBB43_5: # %for.body.preheader
-; CHECK-NEXT:    addi a3, a2, -1024
 ; CHECK-NEXT:    slli a2, a2, 2
-; CHECK-NEXT:    add a0, a0, a2
+; CHECK-NEXT:    add a2, a0, a2
+; CHECK-NEXT:    lui a3, 1
+; CHECK-NEXT:    add a0, a0, a3
 ; CHECK-NEXT:  .LBB43_6: # %for.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
-; CHECK-NEXT:    lw a2, 0(a0)
-; CHECK-NEXT:    divw a2, a2, a1
-; CHECK-NEXT:    sw a2, 0(a0)
-; CHECK-NEXT:    addi a3, a3, 1
-; CHECK-NEXT:    addi a0, a0, 4
-; CHECK-NEXT:    bnez a3, .LBB43_6
+; CHECK-NEXT:    lw a3, 0(a2)
+; CHECK-NEXT:    divw a3, a3, a1
+; CHECK-NEXT:    sw a3, 0(a2)
+; CHECK-NEXT:    addi a2, a2, 4
+; CHECK-NEXT:    bne a2, a0, .LBB43_6
 ; CHECK-NEXT:  .LBB43_7: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -2696,17 +2696,17 @@
 ; CHECK-NEXT:  # %bb.4: # %middle.block
 ; CHECK-NEXT:    beqz a4, .LBB44_7
 ; CHECK-NEXT:  .LBB44_5: # %for.body.preheader
-; CHECK-NEXT:    addi a3, a2, -1024
 ; CHECK-NEXT:    slli a2, a2, 2
-; CHECK-NEXT:    add a0, a0, a2
+; CHECK-NEXT:    add a2, a0, a2
+; CHECK-NEXT:    lui a3, 1
+; CHECK-NEXT:    add a0, a0, a3
 ; CHECK-NEXT:  .LBB44_6: # %for.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
-; CHECK-NEXT:    lw a2, 0(a0)
-; CHECK-NEXT:    remuw a2, a2, a1
-; CHECK-NEXT:    sw a2, 0(a0)
-; CHECK-NEXT:    addi a3, a3, 1
-; CHECK-NEXT:    addi a0, a0, 4
-; CHECK-NEXT:    bnez a3, .LBB44_6
+; CHECK-NEXT:    lw a3, 0(a2)
+; CHECK-NEXT:    remuw a3, a3, a1
+; CHECK-NEXT:    sw a3, 0(a2)
+; CHECK-NEXT:    addi a2, a2, 4
+; CHECK-NEXT:    bne a2, a0, .LBB44_6
 ; CHECK-NEXT:  .LBB44_7: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -2787,17 +2787,17 @@
 ; CHECK-NEXT:  # %bb.4: # %middle.block
 ; CHECK-NEXT:    beqz a4, .LBB45_7
 ; CHECK-NEXT:  .LBB45_5: # %for.body.preheader
-; CHECK-NEXT:    addi a3, a2, -1024
 ; CHECK-NEXT:    slli a2, a2, 2
-; CHECK-NEXT:    add a0, a0, a2
+; CHECK-NEXT:    add a2, a0, a2
+; CHECK-NEXT:    lui a3, 1
+; CHECK-NEXT:    add a0, a0, a3
 ; CHECK-NEXT:  .LBB45_6: # %for.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
-; CHECK-NEXT:    lw a2, 0(a0)
-; CHECK-NEXT:    remw a2, a2, a1
-; CHECK-NEXT:    sw a2, 0(a0)
-; CHECK-NEXT:    addi a3, a3, 1
-; CHECK-NEXT:    addi a0, a0, 4
-; CHECK-NEXT:    bnez a3, .LBB45_6
+; CHECK-NEXT:    lw a3, 0(a2)
+; CHECK-NEXT:    remw a3, a3, a1
+; CHECK-NEXT:    sw a3, 0(a2)
+; CHECK-NEXT:    addi a2, a2, 4
+; CHECK-NEXT:    bne a2, a0, .LBB45_6
 ; CHECK-NEXT:  .LBB45_7: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -2854,7 +2854,8 @@
 define void @sink_splat_vp_mul(ptr nocapture %a, i32 signext %x, <4 x i1> %m, i32 zeroext %vl) {
 ; CHECK-LABEL: sink_splat_vp_mul:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a3, 1024
+; CHECK-NEXT:    lui a3, 1
+; CHECK-NEXT:    add a3, a0, a3
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:  .LBB46_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
@@ -2863,9 +2864,8 @@
 ; CHECK-NEXT:    vmul.vx v8, v8, a1, v0.t
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a3, a3, -4
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a3, .LBB46_1
+; CHECK-NEXT:    bne a0, a3, .LBB46_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -2892,7 +2892,8 @@
 define void @sink_splat_vp_add(ptr nocapture %a, i32 signext %x, <4 x i1> %m, i32 zeroext %vl) {
 ; CHECK-LABEL: sink_splat_vp_add:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a3, 1024
+; CHECK-NEXT:    lui a3, 1
+; CHECK-NEXT:    add a3, a0, a3
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:  .LBB47_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
@@ -2901,9 +2902,8 @@
 ; CHECK-NEXT:    vadd.vx v8, v8, a1, v0.t
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a3, a3, -4
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a3, .LBB47_1
+; CHECK-NEXT:    bne a0, a3, .LBB47_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -2928,7 +2928,8 @@
 define void @sink_splat_vp_add_commute(ptr nocapture %a, i32 signext %x, <4 x i1> %m, i32 zeroext %vl) {
 ; CHECK-LABEL: sink_splat_vp_add_commute:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a3, 1024
+; CHECK-NEXT:    lui a3, 1
+; CHECK-NEXT:    add a3, a0, a3
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:  .LBB48_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
@@ -2937,9 +2938,8 @@
 ; CHECK-NEXT:    vadd.vx v8, v8, a1, v0.t
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a3, a3, -4
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a3, .LBB48_1
+; CHECK-NEXT:    bne a0, a3, .LBB48_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -2966,7 +2966,8 @@
 define void @sink_splat_vp_sub(ptr nocapture %a, i32 signext %x, <4 x i1> %m, i32 zeroext %vl) {
 ; CHECK-LABEL: sink_splat_vp_sub:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a3, 1024
+; CHECK-NEXT:    lui a3, 1
+; CHECK-NEXT:    add a3, a0, a3
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:  .LBB49_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
@@ -2975,9 +2976,8 @@
 ; CHECK-NEXT:    vsub.vx v8, v8, a1, v0.t
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a3, a3, -4
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a3, .LBB49_1
+; CHECK-NEXT:    bne a0, a3, .LBB49_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -3002,7 +3002,8 @@
 define void @sink_splat_vp_rsub(ptr nocapture %a, i32 signext %x, <4 x i1> %m, i32 zeroext %vl) {
 ; CHECK-LABEL: sink_splat_vp_rsub:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a3, 1024
+; CHECK-NEXT:    lui a3, 1
+; CHECK-NEXT:    add a3, a0, a3
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:  .LBB50_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
@@ -3011,9 +3012,8 @@
 ; CHECK-NEXT:    vrsub.vx v8, v8, a1, v0.t
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a3, a3, -4
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a3, .LBB50_1
+; CHECK-NEXT:    bne a0, a3, .LBB50_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -3040,7 +3040,8 @@
 define void @sink_splat_vp_shl(ptr nocapture %a, i32 signext %x, <4 x i1> %m, i32 zeroext %vl) {
 ; CHECK-LABEL: sink_splat_vp_shl:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a3, 1024
+; CHECK-NEXT:    lui a3, 1
+; CHECK-NEXT:    add a3, a0, a3
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:  .LBB51_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
@@ -3049,9 +3050,8 @@
 ; CHECK-NEXT:    vsll.vx v8, v8, a1, v0.t
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a3, a3, -4
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a3, .LBB51_1
+; CHECK-NEXT:    bne a0, a3, .LBB51_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -3078,7 +3078,8 @@
 define void @sink_splat_vp_lshr(ptr nocapture %a, i32 signext %x, <4 x i1> %m, i32 zeroext %vl) {
 ; CHECK-LABEL: sink_splat_vp_lshr:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a3, 1024
+; CHECK-NEXT:    lui a3, 1
+; CHECK-NEXT:    add a3, a0, a3
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:  .LBB52_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
@@ -3087,9 +3088,8 @@
 ; CHECK-NEXT:    vsrl.vx v8, v8, a1, v0.t
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a3, a3, -4
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a3, .LBB52_1
+; CHECK-NEXT:    bne a0, a3, .LBB52_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -3116,7 +3116,8 @@
 define void @sink_splat_vp_ashr(ptr nocapture %a, i32 signext %x, <4 x i1> %m, i32 zeroext %vl) {
 ; CHECK-LABEL: sink_splat_vp_ashr:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a3, 1024
+; CHECK-NEXT:    lui a3, 1
+; CHECK-NEXT:    add a3, a0, a3
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:  .LBB53_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
@@ -3125,9 +3126,8 @@
 ; CHECK-NEXT:    vsra.vx v8, v8, a1, v0.t
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a3, a3, -4
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a3, .LBB53_1
+; CHECK-NEXT:    bne a0, a3, .LBB53_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -3154,7 +3154,8 @@
 define void @sink_splat_vp_fmul(ptr nocapture %a, float %x, <4 x i1> %m, i32 zeroext %vl) {
 ; CHECK-LABEL: sink_splat_vp_fmul:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    lui a2, 1
+; CHECK-NEXT:    add a2, a0, a2
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:  .LBB54_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
@@ -3163,9 +3164,8 @@
 ; CHECK-NEXT:    vfmul.vf v8, v8, fa0, v0.t
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a2, a2, -4
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a2, .LBB54_1
+; CHECK-NEXT:    bne a0, a2, .LBB54_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -3192,7 +3192,8 @@
 define void @sink_splat_vp_fdiv(ptr nocapture %a, float %x, <4 x i1> %m, i32 zeroext %vl) {
 ; CHECK-LABEL: sink_splat_vp_fdiv:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    lui a2, 1
+; CHECK-NEXT:    add a2, a0, a2
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:  .LBB55_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
@@ -3201,9 +3202,8 @@
 ; CHECK-NEXT:    vfdiv.vf v8, v8, fa0, v0.t
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a2, a2, -4
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a2, .LBB55_1
+; CHECK-NEXT:    bne a0, a2, .LBB55_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -3228,7 +3228,8 @@
 define void @sink_splat_vp_frdiv(ptr nocapture %a, float %x, <4 x i1> %m, i32 zeroext %vl) {
 ; CHECK-LABEL: sink_splat_vp_frdiv:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    lui a2, 1
+; CHECK-NEXT:    add a2, a0, a2
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:  .LBB56_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
@@ -3237,9 +3238,8 @@
 ; CHECK-NEXT:    vfrdiv.vf v8, v8, fa0, v0.t
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a2, a2, -4
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a2, .LBB56_1
+; CHECK-NEXT:    bne a0, a2, .LBB56_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -3266,7 +3266,8 @@
 define void @sink_splat_vp_fadd(ptr nocapture %a, float %x, <4 x i1> %m, i32 zeroext %vl) {
 ; CHECK-LABEL: sink_splat_vp_fadd:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    lui a2, 1
+; CHECK-NEXT:    add a2, a0, a2
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:  .LBB57_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
@@ -3275,9 +3276,8 @@
 ; CHECK-NEXT:    vfadd.vf v8, v8, fa0, v0.t
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a2, a2, -4
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a2, .LBB57_1
+; CHECK-NEXT:    bne a0, a2, .LBB57_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -3304,7 +3304,8 @@
 define void @sink_splat_vp_fsub(ptr nocapture %a, float %x, <4 x i1> %m, i32 zeroext %vl) {
 ; CHECK-LABEL: sink_splat_vp_fsub:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    lui a2, 1
+; CHECK-NEXT:    add a2, a0, a2
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:  .LBB58_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
@@ -3313,9 +3314,8 @@
 ; CHECK-NEXT:    vfsub.vf v8, v8, fa0, v0.t
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a2, a2, -4
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a2, .LBB58_1
+; CHECK-NEXT:    bne a0, a2, .LBB58_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -3342,7 +3342,8 @@
 define void @sink_splat_vp_frsub(ptr nocapture %a, float %x, <4 x i1> %m, i32 zeroext %vl) {
 ; CHECK-LABEL: sink_splat_vp_frsub:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    lui a2, 1
+; CHECK-NEXT:    add a2, a0, a2
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:  .LBB59_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
@@ -3351,9 +3352,8 @@
 ; CHECK-NEXT:    vfrsub.vf v8, v8, fa0, v0.t
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a2, a2, -4
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a2, .LBB59_1
+; CHECK-NEXT:    bne a0, a2, .LBB59_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -3380,7 +3380,8 @@
 define void @sink_splat_vp_udiv(ptr nocapture %a, i32 signext %x, <4 x i1> %m, i32 zeroext %vl) {
 ; CHECK-LABEL: sink_splat_vp_udiv:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a3, 1024
+; CHECK-NEXT:    lui a3, 1
+; CHECK-NEXT:    add a3, a0, a3
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:  .LBB60_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
@@ -3389,9 +3390,8 @@
 ; CHECK-NEXT:    vdivu.vx v8, v8, a1, v0.t
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a3, a3, -4
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a3, .LBB60_1
+; CHECK-NEXT:    bne a0, a3, .LBB60_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -3418,7 +3418,8 @@
 define void @sink_splat_vp_sdiv(ptr nocapture %a, i32 signext %x, <4 x i1> %m, i32 zeroext %vl) {
 ; CHECK-LABEL: sink_splat_vp_sdiv:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a3, 1024
+; CHECK-NEXT:    lui a3, 1
+; CHECK-NEXT:    add a3, a0, a3
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:  .LBB61_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
@@ -3427,9 +3428,8 @@
 ; CHECK-NEXT:    vdiv.vx v8, v8, a1, v0.t
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a3, a3, -4
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a3, .LBB61_1
+; CHECK-NEXT:    bne a0, a3, .LBB61_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -3456,7 +3456,8 @@
 define void @sink_splat_vp_urem(ptr nocapture %a, i32 signext %x, <4 x i1> %m, i32 zeroext %vl) {
 ; CHECK-LABEL: sink_splat_vp_urem:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a3, 1024
+; CHECK-NEXT:    lui a3, 1
+; CHECK-NEXT:    add a3, a0, a3
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:  .LBB62_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
@@ -3465,9 +3466,8 @@
 ; CHECK-NEXT:    vremu.vx v8, v8, a1, v0.t
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a3, a3, -4
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a3, .LBB62_1
+; CHECK-NEXT:    bne a0, a3, .LBB62_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -3494,7 +3494,8 @@
 define void @sink_splat_vp_srem(ptr nocapture %a, i32 signext %x, <4 x i1> %m, i32 zeroext %vl) {
 ; CHECK-LABEL: sink_splat_vp_srem:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a3, 1024
+; CHECK-NEXT:    lui a3, 1
+; CHECK-NEXT:    add a3, a0, a3
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:  .LBB63_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
@@ -3503,9 +3504,8 @@
 ; CHECK-NEXT:    vrem.vx v8, v8, a1, v0.t
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a3, a3, -4
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a3, .LBB63_1
+; CHECK-NEXT:    bne a0, a3, .LBB63_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -3534,7 +3534,8 @@
 ; CHECK:       # %bb.0: # %entry
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:    vmv.v.x v8, a1
-; CHECK-NEXT:    li a1, 1024
+; CHECK-NEXT:    lui a1, 1
+; CHECK-NEXT:    add a1, a0, a1
 ; CHECK-NEXT:  .LBB64_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
 ; CHECK-NEXT:    vle32.v v9, (a0)
@@ -3542,9 +3543,8 @@
 ; CHECK-NEXT:    vrem.vv v9, v8, v9, v0.t
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:    vse32.v v9, (a0)
-; CHECK-NEXT:    addi a1, a1, -4
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a1, .LBB64_1
+; CHECK-NEXT:    bne a0, a1, .LBB64_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -3571,7 +3571,8 @@
 define void @sink_splat_vp_fma(ptr noalias nocapture %a, ptr nocapture readonly %b, float %x, <4 x i1> %m, i32 zeroext %vl) {
 ; CHECK-LABEL: sink_splat_vp_fma:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a3, 1024
+; CHECK-NEXT:    lui a3, 1
+; CHECK-NEXT:    add a3, a1, a3
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:  .LBB65_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
@@ -3581,10 +3582,9 @@
 ; CHECK-NEXT:    vfmadd.vf v8, fa0, v9, v0.t
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a3, a3, -4
 ; CHECK-NEXT:    addi a1, a1, 16
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a3, .LBB65_1
+; CHECK-NEXT:    bne a1, a3, .LBB65_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -3611,7 +3611,8 @@
 define void @sink_splat_vp_fma_commute(ptr noalias nocapture %a, ptr nocapture readonly %b, float %x, <4 x i1> %m, i32 zeroext %vl) {
 ; CHECK-LABEL: sink_splat_vp_fma_commute:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a3, 1024
+; CHECK-NEXT:    lui a3, 1
+; CHECK-NEXT:    add a3, a1, a3
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:  .LBB66_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
@@ -3621,10 +3622,9 @@
 ; CHECK-NEXT:    vfmadd.vf v8, fa0, v9, v0.t
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a3, a3, -4
 ; CHECK-NEXT:    addi a1, a1, 16
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a3, .LBB66_1
+; CHECK-NEXT:    bne a1, a3, .LBB66_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -3652,16 +3652,16 @@
 define void @sink_splat_mul_lmul2(ptr nocapture %a, i64 signext %x) {
 ; CHECK-LABEL: sink_splat_mul_lmul2:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    lui a2, 2
+; CHECK-NEXT:    add a2, a0, a2
 ; CHECK-NEXT:    vsetivli zero, 4, e64, m2, ta, ma
 ; CHECK-NEXT:  .LBB67_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
 ; CHECK-NEXT:    vle64.v v8, (a0)
 ; CHECK-NEXT:    vmul.vx v8, v8, a1
 ; CHECK-NEXT:    vse64.v v8, (a0)
-; CHECK-NEXT:    addi a2, a2, -4
 ; CHECK-NEXT:    addi a0, a0, 32
-; CHECK-NEXT:    bnez a2, .LBB67_1
+; CHECK-NEXT:    bne a0, a2, .LBB67_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -3686,16 +3686,16 @@
 define void @sink_splat_add_lmul2(ptr nocapture %a, i64 signext %x) {
 ; CHECK-LABEL: sink_splat_add_lmul2:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    lui a2, 2
+; CHECK-NEXT:    add a2, a0, a2
 ; CHECK-NEXT:    vsetivli zero, 4, e64, m2, ta, ma
 ; CHECK-NEXT:  .LBB68_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
 ; CHECK-NEXT:    vle64.v v8, (a0)
 ; CHECK-NEXT:    vadd.vx v8, v8, a1
 ; CHECK-NEXT:    vse64.v v8, (a0)
-; CHECK-NEXT:    addi a2, a2, -4
 ; CHECK-NEXT:    addi a0, a0, 32
-; CHECK-NEXT:    bnez a2, .LBB68_1
+; CHECK-NEXT:    bne a0, a2, .LBB68_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -3720,16 +3720,16 @@
 define void @sink_splat_sub_lmul2(ptr nocapture %a, i64 signext %x) {
 ; CHECK-LABEL: sink_splat_sub_lmul2:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    lui a2, 2
+; CHECK-NEXT:    add a2, a0, a2
 ; CHECK-NEXT:    vsetivli zero, 4, e64, m2, ta, ma
 ; CHECK-NEXT:  .LBB69_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
 ; CHECK-NEXT:    vle64.v v8, (a0)
 ; CHECK-NEXT:    vsub.vx v8, v8, a1
 ; CHECK-NEXT:    vse64.v v8, (a0)
-; CHECK-NEXT:    addi a2, a2, -4
 ; CHECK-NEXT:    addi a0, a0, 32
-; CHECK-NEXT:    bnez a2, .LBB69_1
+; CHECK-NEXT:    bne a0, a2, .LBB69_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -3754,16 +3754,16 @@
 define void @sink_splat_rsub_lmul2(ptr nocapture %a, i64 signext %x) {
 ; CHECK-LABEL: sink_splat_rsub_lmul2:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    lui a2, 2
+; CHECK-NEXT:    add a2, a0, a2
 ; CHECK-NEXT:    vsetivli zero, 4, e64, m2, ta, ma
 ; CHECK-NEXT:  .LBB70_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
 ; CHECK-NEXT:    vle64.v v8, (a0)
 ; CHECK-NEXT:    vrsub.vx v8, v8, a1
 ; CHECK-NEXT:    vse64.v v8, (a0)
-; CHECK-NEXT:    addi a2, a2, -4
 ; CHECK-NEXT:    addi a0, a0, 32
-; CHECK-NEXT:    bnez a2, .LBB70_1
+; CHECK-NEXT:    bne a0, a2, .LBB70_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -3788,16 +3788,16 @@
 define void @sink_splat_and_lmul2(ptr nocapture %a, i64 signext %x) {
 ; CHECK-LABEL: sink_splat_and_lmul2:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    lui a2, 2
+; CHECK-NEXT:    add a2, a0, a2
 ; CHECK-NEXT:    vsetivli zero, 4, e64, m2, ta, ma
 ; CHECK-NEXT:  .LBB71_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
 ; CHECK-NEXT:    vle64.v v8, (a0)
 ; CHECK-NEXT:    vand.vx v8, v8, a1
 ; CHECK-NEXT:    vse64.v v8, (a0)
-; CHECK-NEXT:    addi a2, a2, -4
 ; CHECK-NEXT:    addi a0, a0, 32
-; CHECK-NEXT:    bnez a2, .LBB71_1
+; CHECK-NEXT:    bne a0, a2, .LBB71_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -3822,16 +3822,16 @@
 define void @sink_splat_or_lmul2(ptr nocapture %a, i64 signext %x) {
 ; CHECK-LABEL: sink_splat_or_lmul2:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    lui a2, 2
+; CHECK-NEXT:    add a2, a0, a2
 ; CHECK-NEXT:    vsetivli zero, 4, e64, m2, ta, ma
 ; CHECK-NEXT:  .LBB72_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
 ; CHECK-NEXT:    vle64.v v8, (a0)
 ; CHECK-NEXT:    vor.vx v8, v8, a1
 ; CHECK-NEXT:    vse64.v v8, (a0)
-; CHECK-NEXT:    addi a2, a2, -4
 ; CHECK-NEXT:    addi a0, a0, 32
-; CHECK-NEXT:    bnez a2, .LBB72_1
+; CHECK-NEXT:    bne a0, a2, .LBB72_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -3856,16 +3856,16 @@
 define void @sink_splat_xor_lmul2(ptr nocapture %a, i64 signext %x) {
 ; CHECK-LABEL: sink_splat_xor_lmul2:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    lui a2, 2
+; CHECK-NEXT:    add a2, a0, a2
 ; CHECK-NEXT:    vsetivli zero, 4, e64, m2, ta, ma
 ; CHECK-NEXT:  .LBB73_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
 ; CHECK-NEXT:    vle64.v v8, (a0)
 ; CHECK-NEXT:    vxor.vx v8, v8, a1
 ; CHECK-NEXT:    vse64.v v8, (a0)
-; CHECK-NEXT:    addi a2, a2, -4
 ; CHECK-NEXT:    addi a0, a0, 32
-; CHECK-NEXT:    bnez a2, .LBB73_1
+; CHECK-NEXT:    bne a0, a2, .LBB73_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -3890,7 +3890,8 @@
 define void @sink_splat_mul_lmul8(ptr nocapture %a, i32 signext %x) {
 ; CHECK-LABEL: sink_splat_mul_lmul8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    lui a2, 1
+; CHECK-NEXT:    add a2, a0, a2
 ; CHECK-NEXT:    li a3, 32
 ; CHECK-NEXT:  .LBB74_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
@@ -3898,9 +3899,8 @@
 ; CHECK-NEXT:    vle32.v v8, (a0)
 ; CHECK-NEXT:    vmul.vx v8, v8, a1
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a2, a2, -4
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a2, .LBB74_1
+; CHECK-NEXT:    bne a0, a2, .LBB74_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -3925,7 +3925,8 @@
 define void @sink_splat_add_lmul8(ptr nocapture %a, i32 signext %x) {
 ; CHECK-LABEL: sink_splat_add_lmul8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    lui a2, 1
+; CHECK-NEXT:    add a2, a0, a2
 ; CHECK-NEXT:    li a3, 32
 ; CHECK-NEXT:  .LBB75_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
@@ -3933,9 +3934,8 @@
 ; CHECK-NEXT:    vle32.v v8, (a0)
 ; CHECK-NEXT:    vadd.vx v8, v8, a1
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a2, a2, -4
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a2, .LBB75_1
+; CHECK-NEXT:    bne a0, a2, .LBB75_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -3960,7 +3960,8 @@
 define void @sink_splat_sub_lmul8(ptr nocapture %a, i32 signext %x) {
 ; CHECK-LABEL: sink_splat_sub_lmul8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    lui a2, 1
+; CHECK-NEXT:    add a2, a0, a2
 ; CHECK-NEXT:    li a3, 32
 ; CHECK-NEXT:  .LBB76_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
@@ -3968,9 +3969,8 @@
 ; CHECK-NEXT:    vle32.v v8, (a0)
 ; CHECK-NEXT:    vsub.vx v8, v8, a1
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a2, a2, -4
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a2, .LBB76_1
+; CHECK-NEXT:    bne a0, a2, .LBB76_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -3995,7 +3995,8 @@
 define void @sink_splat_rsub_lmul8(ptr nocapture %a, i32 signext %x) {
 ; CHECK-LABEL: sink_splat_rsub_lmul8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    lui a2, 1
+; CHECK-NEXT:    add a2, a0, a2
 ; CHECK-NEXT:    li a3, 32
 ; CHECK-NEXT:  .LBB77_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
@@ -4003,9 +4004,8 @@
 ; CHECK-NEXT:    vle32.v v8, (a0)
 ; CHECK-NEXT:    vrsub.vx v8, v8, a1
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a2, a2, -4
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a2, .LBB77_1
+; CHECK-NEXT:    bne a0, a2, .LBB77_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -4030,7 +4030,8 @@
 define void @sink_splat_and_lmul8(ptr nocapture %a, i32 signext %x) {
 ; CHECK-LABEL: sink_splat_and_lmul8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    lui a2, 1
+; CHECK-NEXT:    add a2, a0, a2
 ; CHECK-NEXT:    li a3, 32
 ; CHECK-NEXT:  .LBB78_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
@@ -4038,9 +4039,8 @@
 ; CHECK-NEXT:    vle32.v v8, (a0)
 ; CHECK-NEXT:    vand.vx v8, v8, a1
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a2, a2, -4
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a2, .LBB78_1
+; CHECK-NEXT:    bne a0, a2, .LBB78_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -4065,7 +4065,8 @@
 define void @sink_splat_or_lmul8(ptr nocapture %a, i32 signext %x) {
 ; CHECK-LABEL: sink_splat_or_lmul8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    lui a2, 1
+; CHECK-NEXT:    add a2, a0, a2
 ; CHECK-NEXT:    li a3, 32
 ; CHECK-NEXT:  .LBB79_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
@@ -4073,9 +4074,8 @@
 ; CHECK-NEXT:    vle32.v v8, (a0)
 ; CHECK-NEXT:    vor.vx v8, v8, a1
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a2, a2, -4
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a2, .LBB79_1
+; CHECK-NEXT:    bne a0, a2, .LBB79_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -4100,7 +4100,8 @@
 define void @sink_splat_xor_lmul8(ptr nocapture %a, i32 signext %x) {
 ; CHECK-LABEL: sink_splat_xor_lmul8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    lui a2, 1
+; CHECK-NEXT:    add a2, a0, a2
 ; CHECK-NEXT:    li a3, 32
 ; CHECK-NEXT:  .LBB80_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
@@ -4108,9 +4109,8 @@
 ; CHECK-NEXT:    vle32.v v8, (a0)
 ; CHECK-NEXT:    vxor.vx v8, v8, a1
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a2, a2, -4
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a2, .LBB80_1
+; CHECK-NEXT:    bne a0, a2, .LBB80_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -4135,16 +4135,16 @@
 define void @sink_splat_mul_lmulmf2(ptr nocapture %a, i32 signext %x) {
 ; CHECK-LABEL: sink_splat_mul_lmulmf2:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    lui a2, 2
+; CHECK-NEXT:    add a2, a0, a2
 ; CHECK-NEXT:    vsetivli zero, 2, e32, mf2, ta, ma
 ; CHECK-NEXT:  .LBB81_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
 ; CHECK-NEXT:    vle32.v v8, (a0)
 ; CHECK-NEXT:    vmul.vx v8, v8, a1
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a2, a2, -4
 ; CHECK-NEXT:    addi a0, a0, 32
-; CHECK-NEXT:    bnez a2, .LBB81_1
+; CHECK-NEXT:    bne a0, a2, .LBB81_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -4169,16 +4169,16 @@
 define void @sink_splat_add_lmulmf2(ptr nocapture %a, i32 signext %x) {
 ; CHECK-LABEL: sink_splat_add_lmulmf2:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    lui a2, 2
+; CHECK-NEXT:    add a2, a0, a2
 ; CHECK-NEXT:    vsetivli zero, 2, e32, mf2, ta, ma
 ; CHECK-NEXT:  .LBB82_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
 ; CHECK-NEXT:    vle32.v v8, (a0)
 ; CHECK-NEXT:    vadd.vx v8, v8, a1
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a2, a2, -4
 ; CHECK-NEXT:    addi a0, a0, 32
-; CHECK-NEXT:    bnez a2, .LBB82_1
+; CHECK-NEXT:    bne a0, a2, .LBB82_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -4203,16 +4203,16 @@
 define void @sink_splat_sub_lmulmf2(ptr nocapture %a, i32 signext %x) {
 ; CHECK-LABEL: sink_splat_sub_lmulmf2:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    lui a2, 2
+; CHECK-NEXT:    add a2, a0, a2
 ; CHECK-NEXT:    vsetivli zero, 2, e32, mf2, ta, ma
 ; CHECK-NEXT:  .LBB83_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
 ; CHECK-NEXT:    vle32.v v8, (a0)
 ; CHECK-NEXT:    vsub.vx v8, v8, a1
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a2, a2, -4
 ; CHECK-NEXT:    addi a0, a0, 32
-; CHECK-NEXT:    bnez a2, .LBB83_1
+; CHECK-NEXT:    bne a0, a2, .LBB83_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -4237,16 +4237,16 @@
 define void @sink_splat_rsub_lmulmf2(ptr nocapture %a, i32 signext %x) {
 ; CHECK-LABEL: sink_splat_rsub_lmulmf2:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    lui a2, 2
+; CHECK-NEXT:    add a2, a0, a2
 ; CHECK-NEXT:    vsetivli zero, 2, e32, mf2, ta, ma
 ; CHECK-NEXT:  .LBB84_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
 ; CHECK-NEXT:    vle32.v v8, (a0)
 ; CHECK-NEXT:    vrsub.vx v8, v8, a1
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a2, a2, -4
 ; CHECK-NEXT:    addi a0, a0, 32
-; CHECK-NEXT:    bnez a2, .LBB84_1
+; CHECK-NEXT:    bne a0, a2, .LBB84_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -4271,16 +4271,16 @@
 define void @sink_splat_and_lmulmf2(ptr nocapture %a, i32 signext %x) {
 ; CHECK-LABEL: sink_splat_and_lmulmf2:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    lui a2, 2
+; CHECK-NEXT:    add a2, a0, a2
 ; CHECK-NEXT:    vsetivli zero, 2, e32, mf2, ta, ma
 ; CHECK-NEXT:  .LBB85_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
 ; CHECK-NEXT:    vle32.v v8, (a0)
 ; CHECK-NEXT:    vand.vx v8, v8, a1
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a2, a2, -4
 ; CHECK-NEXT:    addi a0, a0, 32
-; CHECK-NEXT:    bnez a2, .LBB85_1
+; CHECK-NEXT:    bne a0, a2, .LBB85_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -4305,16 +4305,16 @@
 define void @sink_splat_or_lmulmf2(ptr nocapture %a, i32 signext %x) {
 ; CHECK-LABEL: sink_splat_or_lmulmf2:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    lui a2, 2
+; CHECK-NEXT:    add a2, a0, a2
 ; CHECK-NEXT:    vsetivli zero, 2, e32, mf2, ta, ma
 ; CHECK-NEXT:  .LBB86_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
 ; CHECK-NEXT:    vle32.v v8, (a0)
 ; CHECK-NEXT:    vor.vx v8, v8, a1
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a2, a2, -4
 ; CHECK-NEXT:    addi a0, a0, 32
-; CHECK-NEXT:    bnez a2, .LBB86_1
+; CHECK-NEXT:    bne a0, a2, .LBB86_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -4339,16 +4339,16 @@
 define void @sink_splat_xor_lmulmf2(ptr nocapture %a, i32 signext %x) {
 ; CHECK-LABEL: sink_splat_xor_lmulmf2:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    lui a2, 2
+; CHECK-NEXT:    add a2, a0, a2
 ; CHECK-NEXT:    vsetivli zero, 2, e32, mf2, ta, ma
 ; CHECK-NEXT:  .LBB87_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
 ; CHECK-NEXT:    vle32.v v8, (a0)
 ; CHECK-NEXT:    vxor.vx v8, v8, a1
 ; CHECK-NEXT:    vse32.v v8, (a0)
-; CHECK-NEXT:    addi a2, a2, -4
 ; CHECK-NEXT:    addi a0, a0, 32
-; CHECK-NEXT:    bnez a2, .LBB87_1
+; CHECK-NEXT:    bne a0, a2, .LBB87_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -4376,7 +4376,8 @@
 ; CHECK-LABEL: sink_splat_vp_icmp:
 ; CHECK:       # %bb.0: # %entry
 ; CHECK-NEXT:    vmv1r.v v8, v0
-; CHECK-NEXT:    li a3, 1024
+; CHECK-NEXT:    lui a3, 1
+; CHECK-NEXT:    add a3, a0, a3
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:    vmv.v.i v9, 0
 ; CHECK-NEXT:  .LBB88_1: # %vector.body
@@ -4387,9 +4388,8 @@
 ; CHECK-NEXT:    vmseq.vx v0, v10, a1, v0.t
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:    vse32.v v9, (a0), v0.t
-; CHECK-NEXT:    addi a3, a3, -4
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a3, .LBB88_1
+; CHECK-NEXT:    bne a0, a3, .LBB88_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry:
@@ -4417,7 +4417,8 @@
 ; CHECK-LABEL: sink_splat_vp_fcmp:
 ; CHECK:       # %bb.0: # %entry
 ; CHECK-NEXT:    vmv1r.v v8, v0
-; CHECK-NEXT:    li a2, 1024
+; CHECK-NEXT:    lui a2, 1
+; CHECK-NEXT:    add a2, a0, a2
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:    vmv.v.i v9, 0
 ; CHECK-NEXT:  .LBB89_1: # %vector.body
@@ -4428,9 +4429,8 @@
 ; CHECK-NEXT:    vmfeq.vf v0, v10, fa0, v0.t
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:    vse32.v v9, (a0), v0.t
-; CHECK-NEXT:    addi a2, a2, -4
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a2, .LBB89_1
+; CHECK-NEXT:    bne a0, a2, .LBB89_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry: