diff --git a/llvm/lib/Target/X86/X86InstrInfo.cpp b/llvm/lib/Target/X86/X86InstrInfo.cpp
index 09b951befac9f..205fd24e6d402 100644
--- a/llvm/lib/Target/X86/X86InstrInfo.cpp
+++ b/llvm/lib/Target/X86/X86InstrInfo.cpp
@@ -6104,7 +6104,8 @@ MachineInstr *X86InstrInfo::foldMemoryOperandCustom(
       const TargetRegisterInfo &TRI = *MF.getSubtarget().getRegisterInfo();
       const TargetRegisterClass *RC = getRegClass(MI.getDesc(), OpNum, &RI, MF);
       unsigned RCSize = TRI.getRegSizeInBits(*RC) / 8;
-      if ((Size == 0 || Size >= 16) && RCSize >= 16 && Alignment >= Align(4)) {
+      if ((Size == 0 || Size >= 16) && RCSize >= 16 &&
+          (MI.getOpcode() != X86::INSERTPSrr || Alignment >= Align(4))) {
         int PtrOffset = SrcIdx * 4;
         unsigned NewImm = (DstIdx << 4) | ZMask;
         unsigned NewOpCode =
diff --git a/llvm/test/CodeGen/X86/avx.ll b/llvm/test/CodeGen/X86/avx.ll
index 6a10d855ea95c..4ce092c099b08 100644
--- a/llvm/test/CodeGen/X86/avx.ll
+++ b/llvm/test/CodeGen/X86/avx.ll
@@ -184,14 +184,12 @@ define <4 x float> @nofold_insertps(ptr %a, <4 x float> %b) {
 ; X86-LABEL: nofold_insertps:
 ; X86:       ## %bb.0:
 ; X86-NEXT:    movl {{[0-9]+}}(%esp), %eax
-; X86-NEXT:    vmovups (%eax), %xmm1
-; X86-NEXT:    vinsertps {{.*#+}} xmm0 = xmm0[0,1,2],xmm1[2]
+; X86-NEXT:    vinsertps $48, 8(%eax), %xmm0, %xmm0 ## xmm0 = xmm0[0,1,2],mem[0]
 ; X86-NEXT:    retl
 ;
 ; X64-LABEL: nofold_insertps:
 ; X64:       ## %bb.0:
-; X64-NEXT:    vmovups (%rdi), %xmm1
-; X64-NEXT:    vinsertps {{.*#+}} xmm0 = xmm0[0,1,2],xmm1[2]
+; X64-NEXT:    vinsertps $48, 8(%rdi), %xmm0, %xmm0 ## xmm0 = xmm0[0,1,2],mem[0]
 ; X64-NEXT:    retq
   %1 = load <4 x float>, ptr %a, align 1
   %2 = shufflevector <4 x float> %b, <4 x float> %1, <4 x i32> <i32 0, i32 1, i32 2, i32 6>