[AMDGPU] GFX11 CodeGen support for MIMG instructions

[foad]

Summary

This includes:

• New llvm.amdgcn.image.msaa.load.* intrinsics
• NSA changes, because MIMG-NSA is now limited to 3 dwords
• Split CD forms of IMAGE_SAMPLE instructions out into separate test files since they are no longer supported in GFX11

[rampitec]

Any tests for intersect_ray?

[foad]

Added intersect_ray tests.

[foad]

In D127837#3586539, @rampitec wrote:

Any tests for intersect_ray?

I have added them now. (I didn't bother before because there are no codegen changes in the patch specifically for GFX11 intersect_ray.)

[rampitec]

LGTM

llvm/include/llvm/IR/IntrinsicsAMDGPU.td

defset list<AMDGPUImageDimIntrinsic> AMDGPUImageDimIntrinsics = {
   //////////////////////////////////////////////////////////////////////////
   foreach dim = AMDGPUDims.Msaa in {
     def int_amdgcn_image_msaa_load_x # _ # dim.Name:
         AMDGPUImageDimIntrinsic<
             AMDGPUDimNoSampleProfile<"MSAA_LOAD_X", dim, [llvm_any_ty], []>,
             [IntrReadMem], [SDNPMemOperand]>;
   }
 
+  foreach dim = AMDGPUDims.Msaa in {
+    def int_amdgcn_image_msaa_load # _ # dim.Name:
+        AMDGPUImageDimIntrinsic<
+            AMDGPUDimNoSampleProfile<"MSAA_LOAD", dim, [llvm_any_ty], []>,
+            [IntrReadMem], [SDNPMemOperand]>;
+  }
+
   //////////////////////////////////////////////////////////////////////////
   // sample and getlod intrinsics
   //////////////////////////////////////////////////////////////////////////
   multiclass AMDGPUImageDimSampleDims<string opmod,
                                       AMDGPUSampleVariant sample,
                                       bit NoMem = false> {
     foreach dim = AMDGPUDims.NoMsaa in {
       def !strconcat(NAME, "_", dim.Name) : AMDGPUImageDimIntrinsic<

llvm/lib/Target/AMDGPU/AMDGPUInstructionSelector.cpp

bool AMDGPUInstructionSelector::selectImageIntrinsic(
   const DebugLoc &DL = MI.getDebugLoc();
 
   const AMDGPU::MIMGBaseOpcodeInfo *BaseOpcode =
     AMDGPU::getMIMGBaseOpcodeInfo(Intr->BaseOpcode);
 
   const AMDGPU::MIMGDimInfo *DimInfo = AMDGPU::getMIMGDimInfo(Intr->Dim);
   unsigned IntrOpcode = Intr->BaseOpcode;
   const bool IsGFX10Plus = AMDGPU::isGFX10Plus(STI);
+  const bool IsGFX11Plus = AMDGPU::isGFX11Plus(STI);
 
   const unsigned ArgOffset = MI.getNumExplicitDefs() + 1;
 
   Register VDataIn, VDataOut;
   LLT VDataTy;
   int NumVDataDwords = -1;
   bool IsD16 = MI.getOpcode() == AMDGPU::G_AMDGPU_INTRIN_IMAGE_LOAD_D16 ||
                MI.getOpcode() == AMDGPU::G_AMDGPU_INTRIN_IMAGE_STORE_D16;
if (UseNSA && !STI.hasFeature(AMDGPU::FeatureNSAEncoding)) {
     LLVM_DEBUG(dbgs() << "Trying to use NSA on non-NSA target\n");
     return false;
   }
 
   if (IsTexFail)
     ++NumVDataDwords;
 
   int Opcode = -1;
-  if (IsGFX10Plus) {
+  if (IsGFX11Plus) {
+    Opcode = AMDGPU::getMIMGOpcode(IntrOpcode,
+                                   UseNSA ? AMDGPU::MIMGEncGfx11NSA
+                                          : AMDGPU::MIMGEncGfx11Default,
+                                   NumVDataDwords, NumVAddrDwords);
+  } else if (IsGFX10Plus) {
     Opcode = AMDGPU::getMIMGOpcode(IntrOpcode,
                                    UseNSA ? AMDGPU::MIMGEncGfx10NSA
                                           : AMDGPU::MIMGEncGfx10Default,
                                    NumVDataDwords, NumVAddrDwords);
   } else {
     if (Subtarget->hasGFX90AInsts()) {
       Opcode = AMDGPU::getMIMGOpcode(IntrOpcode, AMDGPU::MIMGEncGfx90a,
                                      NumVDataDwords, NumVAddrDwords);

llvm/lib/Target/AMDGPU/AMDGPULegalizerInfo.cpp

if (IsA16 || IsG16) {
     // wash in terms of code size or even better.
     //
     // However, we currently have no way of hinting to the register allocator
     // that MIMG addresses should be placed contiguously when it is possible to
     // do so, so force non-NSA for the common 2-address case as a heuristic.
     //
     // SIShrinkInstructions will convert NSA encodings to non-NSA after register
     // allocation when possible.
+    //
+    // TODO: we can actually allow partial NSA where the final register is a
+    // contiguous set of the remaining addresses.
+    // This could help where there are more addresses than supported.
     const bool UseNSA = ST.hasNSAEncoding() && CorrectedNumVAddrs >= 3 &&
                         CorrectedNumVAddrs <= ST.getNSAMaxSize();
 
     if (!UseNSA && Intr->NumVAddrs > 1)
       convertImageAddrToPacked(B, MI, ArgOffset + Intr->VAddrStart,
                                Intr->NumVAddrs);
   }
 
bool AMDGPULegalizerInfo::legalizeDebugTrapIntrinsic(
   return true;
 }
 
 bool AMDGPULegalizerInfo::legalizeBVHIntrinsic(MachineInstr &MI,
                                                MachineIRBuilder &B) const {
   MachineRegisterInfo &MRI = *B.getMRI();
   const LLT S16 = LLT::scalar(16);
   const LLT S32 = LLT::scalar(32);
+  const LLT V2S16 = LLT::fixed_vector(2, 16);
+  const LLT V3S32 = LLT::fixed_vector(3, 32);
 
   Register DstReg = MI.getOperand(0).getReg();
   Register NodePtr = MI.getOperand(2).getReg();
   Register RayExtent = MI.getOperand(3).getReg();
   Register RayOrigin = MI.getOperand(4).getReg();
   Register RayDir = MI.getOperand(5).getReg();
   Register RayInvDir = MI.getOperand(6).getReg();
   Register TDescr = MI.getOperand(7).getReg();
 
   if (!ST.hasGFX10_AEncoding()) {
     DiagnosticInfoUnsupported BadIntrin(B.getMF().getFunction(),
                                         "intrinsic not supported on subtarget",
                                         MI.getDebugLoc());
     B.getMF().getFunction().getContext().diagnose(BadIntrin);
     return false;
   }
 
+  const bool IsGFX11Plus = AMDGPU::isGFX11Plus(ST);
   const bool IsA16 = MRI.getType(RayDir).getElementType().getSizeInBits() == 16;
   const bool Is64 = MRI.getType(NodePtr).getSizeInBits() == 64;
   const unsigned NumVDataDwords = 4;
   const unsigned NumVAddrDwords = IsA16 ? (Is64 ? 9 : 8) : (Is64 ? 12 : 11);
-  const bool UseNSA =
-      ST.hasNSAEncoding() && NumVAddrDwords <= ST.getNSAMaxSize();
+  const unsigned NumVAddrs = IsGFX11Plus ? (IsA16 ? 4 : 5) : NumVAddrDwords;
+  const bool UseNSA = ST.hasNSAEncoding() && NumVAddrs <= ST.getNSAMaxSize();
   const unsigned BaseOpcodes[2][2] = {
       {AMDGPU::IMAGE_BVH_INTERSECT_RAY, AMDGPU::IMAGE_BVH_INTERSECT_RAY_a16},
       {AMDGPU::IMAGE_BVH64_INTERSECT_RAY,
        AMDGPU::IMAGE_BVH64_INTERSECT_RAY_a16}};
   int Opcode;
   if (UseNSA) {
-    Opcode =
-        AMDGPU::getMIMGOpcode(BaseOpcodes[Is64][IsA16], AMDGPU::MIMGEncGfx10NSA,
+    Opcode = AMDGPU::getMIMGOpcode(BaseOpcodes[Is64][IsA16],
+                                   IsGFX11Plus ? AMDGPU::MIMGEncGfx11NSA
+                                               : AMDGPU::MIMGEncGfx10NSA,
                                    NumVDataDwords, NumVAddrDwords);
   } else {
-    Opcode = AMDGPU::getMIMGOpcode(BaseOpcodes[Is64][IsA16],
-                                   AMDGPU::MIMGEncGfx10Default, NumVDataDwords,
-                                   PowerOf2Ceil(NumVAddrDwords));
+    Opcode = AMDGPU::getMIMGOpcode(
+        BaseOpcodes[Is64][IsA16],
+        IsGFX11Plus ? AMDGPU::MIMGEncGfx11Default : AMDGPU::MIMGEncGfx10Default,
+        NumVDataDwords, PowerOf2Ceil(NumVAddrDwords));
   }
   assert(Opcode != -1);
 
   SmallVector<Register, 12> Ops;
+  if (UseNSA && IsGFX11Plus) {
+    auto packLanes = [&Ops, &S32, &V3S32, &B](Register Src) {
+      auto Unmerge = B.buildUnmerge({S32, S32, S32}, Src);
+      auto Merged = B.buildMerge(
+          V3S32, {Unmerge.getReg(0), Unmerge.getReg(1), Unmerge.getReg(2)});
+      Ops.push_back(Merged.getReg(0));
+    };
+
+    Ops.push_back(NodePtr);
+    Ops.push_back(RayExtent);
+    packLanes(RayOrigin);
+
+    if (IsA16) {
+      auto UnmergeRayDir = B.buildUnmerge({S16, S16, S16}, RayDir);
+      auto UnmergeRayInvDir = B.buildUnmerge({S16, S16, S16}, RayInvDir);
+      auto MergedDir = B.buildMerge(
+          V3S32,
+          {B.buildBitcast(S32, B.buildMerge(V2S16, {UnmergeRayInvDir.getReg(0),
+                                                    UnmergeRayDir.getReg(0)}))
+               .getReg(0),
+           B.buildBitcast(S32, B.buildMerge(V2S16, {UnmergeRayInvDir.getReg(1),
+                                                    UnmergeRayDir.getReg(1)}))
+               .getReg(0),
+           B.buildBitcast(S32, B.buildMerge(V2S16, {UnmergeRayInvDir.getReg(2),
+                                                    UnmergeRayDir.getReg(2)}))
+               .getReg(0)});
+      Ops.push_back(MergedDir.getReg(0));
+    } else {
+      packLanes(RayDir);
+      packLanes(RayInvDir);
+    }
+  } else {
     if (Is64) {
       auto Unmerge = B.buildUnmerge({S32, S32}, NodePtr);
       Ops.push_back(Unmerge.getReg(0));
       Ops.push_back(Unmerge.getReg(1));
     } else {
       Ops.push_back(NodePtr);
     }
     Ops.push_back(RayExtent);
 
     auto packLanes = [&Ops, &S32, &B](Register Src) {
       auto Unmerge = B.buildUnmerge({S32, S32, S32}, Src);
       Ops.push_back(Unmerge.getReg(0));
       Ops.push_back(Unmerge.getReg(1));
       Ops.push_back(Unmerge.getReg(2));
     };
 
     packLanes(RayOrigin);
     if (IsA16) {
       auto UnmergeRayDir = B.buildUnmerge({S16, S16, S16}, RayDir);
       auto UnmergeRayInvDir = B.buildUnmerge({S16, S16, S16}, RayInvDir);
       Register R1 = MRI.createGenericVirtualRegister(S32);
       Register R2 = MRI.createGenericVirtualRegister(S32);
       Register R3 = MRI.createGenericVirtualRegister(S32);
       B.buildMerge(R1, {UnmergeRayDir.getReg(0), UnmergeRayDir.getReg(1)});
       B.buildMerge(R2, {UnmergeRayDir.getReg(2), UnmergeRayInvDir.getReg(0)});
-    B.buildMerge(R3, {UnmergeRayInvDir.getReg(1), UnmergeRayInvDir.getReg(2)});
+      B.buildMerge(R3,
+                   {UnmergeRayInvDir.getReg(1), UnmergeRayInvDir.getReg(2)});
       Ops.push_back(R1);
       Ops.push_back(R2);
       Ops.push_back(R3);
     } else {
       packLanes(RayDir);
       packLanes(RayInvDir);
     }
+  }
 
   if (!UseNSA) {
     // Build a single vector containing all the operands so far prepared.
     LLT OpTy = LLT::fixed_vector(Ops.size(), 32);
     Register MergedOps = B.buildMerge(OpTy, Ops).getReg(0);
     Ops.clear();
     Ops.push_back(MergedOps);
   }

llvm/lib/Target/AMDGPU/AMDGPURegisterBankInfo.cpp

case AMDGPU::G_AMDGPU_INTRIN_BVH_INTERSECT_RAY: {
     if (N == 3) {
       // Sequential form: all operands combined into VGPR256/VGPR512
       unsigned Size = MRI.getType(MI.getOperand(2).getReg()).getSizeInBits();
       if (Size > 256)
         Size = 512;
       OpdsMapping[2] = AMDGPU::getValueMapping(AMDGPU::VGPRRegBankID, Size);
     } else {
       // NSA form
-      for (unsigned I = 2; I < N; ++I)
-        OpdsMapping[I] = AMDGPU::getValueMapping(AMDGPU::VGPRRegBankID, 32);
+      for (unsigned I = 2; I < N; ++I) {
+        unsigned Size = MRI.getType(MI.getOperand(I).getReg()).getSizeInBits();
+        OpdsMapping[I] = AMDGPU::getValueMapping(AMDGPU::VGPRRegBankID, Size);
+      }
     }
     break;
   }
   case AMDGPU::G_INTRINSIC_W_SIDE_EFFECTS: {
     auto IntrID = MI.getIntrinsicID();
     switch (IntrID) {
     case Intrinsic::amdgcn_s_getreg:
     case Intrinsic::amdgcn_s_memtime:

llvm/lib/Target/AMDGPU/GCNNSAReassign.cpp

 /// in NSA image instructions. Later SIShrinkInstructions pass will replace NSA
 /// with sequential versions where possible.
 ///
 //===----------------------------------------------------------------------===//
 
 #include "AMDGPU.h"
 #include "GCNSubtarget.h"
 #include "SIMachineFunctionInfo.h"
+#include "SIRegisterInfo.h"
 #include "llvm/ADT/Statistic.h"
 #include "llvm/CodeGen/LiveIntervals.h"
 #include "llvm/CodeGen/LiveRegMatrix.h"
 #include "llvm/CodeGen/MachineFunctionPass.h"
 #include "llvm/CodeGen/VirtRegMap.h"
 #include "llvm/InitializePasses.h"
 
 using namespace llvm;
GCNNSAReassign::scavengeRegs(SmallVectorImpl<LiveInterval *> &Intervals) const {
   }
 
   return false;
 }
 
 GCNNSAReassign::NSA_Status
 GCNNSAReassign::CheckNSA(const MachineInstr &MI, bool Fast) const {
   const AMDGPU::MIMGInfo *Info = AMDGPU::getMIMGInfo(MI.getOpcode());
-  if (!Info || Info->MIMGEncoding != AMDGPU::MIMGEncGfx10NSA)
+  if (!Info)
     return NSA_Status::NOT_NSA;
 
+  switch (Info->MIMGEncoding) {
+  case AMDGPU::MIMGEncGfx10NSA:
+  case AMDGPU::MIMGEncGfx11NSA:
+    break;
+  default:
+    return NSA_Status::NOT_NSA;
+  }
+
   int VAddr0Idx =
     AMDGPU::getNamedOperandIdx(MI.getOpcode(), AMDGPU::OpName::vaddr0);
 
   unsigned VgprBase = 0;
   bool NSA = false;
-  for (unsigned I = 0; I < Info->VAddrDwords; ++I) {
+  for (unsigned I = 0; I < Info->VAddrOperands; ++I) {
     const MachineOperand &Op = MI.getOperand(VAddr0Idx + I);
     Register Reg = Op.getReg();
     if (Reg.isPhysical() || !VRM->isAssignedReg(Reg))
       return NSA_Status::FIXED;
 
     Register PhysReg = VRM->getPhys(Reg);
 
     if (!Fast) {
       if (!PhysReg)
         return NSA_Status::FIXED;
 
+      // TODO: address the below limitation to handle GFX11 BVH instructions
       // Bail if address is not a VGPR32. That should be possible to extend the
       // optimization to work with subregs of a wider register tuples, but the
       // logic to find free registers will be much more complicated with much
       // less chances for success. That seems reasonable to assume that in most
       // cases a tuple is used because a vector variable contains different
       // parts of an address and it is either already consecutive or cannot
       // be reassigned if not. If needed it is better to rely on register
       // coalescer to process such address tuples.
-      if (MRI->getRegClass(Reg) != &AMDGPU::VGPR_32RegClass || Op.getSubReg())
+      if (TRI->getRegSizeInBits(*MRI->getRegClass(Reg)) != 32 || Op.getSubReg())
         return NSA_Status::FIXED;
 
       // InlineSpiller does not call LRM::assign() after an LI split leaving
       // it in an inconsistent state, so we cannot call LRM::unassign().
       // See llvm bug #48911.
       // Skip reassign if a register has originated from such split.
       // FIXME: Remove the workaround when bug #48911 is fixed.
       if (VRM->getPreSplitReg(Reg))
for (auto &C : Candidates) {
 
     const AMDGPU::MIMGInfo *Info = AMDGPU::getMIMGInfo(MI->getOpcode());
     int VAddr0Idx =
       AMDGPU::getNamedOperandIdx(MI->getOpcode(), AMDGPU::OpName::vaddr0);
 
     SmallVector<LiveInterval *, 16> Intervals;
     SmallVector<MCRegister, 16> OrigRegs;
     SlotIndex MinInd, MaxInd;
-    for (unsigned I = 0; I < Info->VAddrDwords; ++I) {
+    for (unsigned I = 0; I < Info->VAddrOperands; ++I) {
       const MachineOperand &Op = MI->getOperand(VAddr0Idx + I);
       Register Reg = Op.getReg();
       LiveInterval *LI = &LIS->getInterval(Reg);
       if (llvm::is_contained(Intervals, LI)) {
         // Same register used, unable to make sequential
         Intervals.clear();
         break;
       }
if (!Success) {
         if (I->second && CheckNSA(*I->first, true) < NSA_Status::CONTIGUOUS) {
           Success = false;
           LLVM_DEBUG(dbgs() << "\tNSA conversion conflict with " << *I->first);
         }
       }
     }
 
     if (!Success) {
-      for (unsigned I = 0; I < Info->VAddrDwords; ++I)
+      for (unsigned I = 0; I < Info->VAddrOperands; ++I)
         if (VRM->hasPhys(Intervals[I]->reg()))
           LRM->unassign(*Intervals[I]);
 
-      for (unsigned I = 0; I < Info->VAddrDwords; ++I)
+      for (unsigned I = 0; I < Info->VAddrOperands; ++I)
         LRM->assign(*Intervals[I], OrigRegs[I]);
 
       continue;
     }
 
     C.second = true;
     ++NumNSAConverted;
     LLVM_DEBUG(

llvm/lib/Target/AMDGPU/SIISelLowering.cpp

SDValue SITargetLowering::lowerImage(SDValue Op,
   SDLoc DL(Op);
   MachineFunction &MF = DAG.getMachineFunction();
   const GCNSubtarget* ST = &MF.getSubtarget<GCNSubtarget>();
   const AMDGPU::MIMGBaseOpcodeInfo *BaseOpcode =
       AMDGPU::getMIMGBaseOpcodeInfo(Intr->BaseOpcode);
   const AMDGPU::MIMGDimInfo *DimInfo = AMDGPU::getMIMGDimInfo(Intr->Dim);
   unsigned IntrOpcode = Intr->BaseOpcode;
   bool IsGFX10Plus = AMDGPU::isGFX10Plus(*Subtarget);
+  bool IsGFX11Plus = AMDGPU::isGFX11Plus(*Subtarget);
 
   SmallVector<EVT, 3> ResultTypes(Op->values());
   SmallVector<EVT, 3> OrigResultTypes(Op->values());
   bool IsD16 = false;
   bool IsG16 = false;
   bool IsA16 = false;
   SDValue VData;
   int NumVDataDwords;
SDValue SITargetLowering::lowerImage(SDValue Op,
   // wash in terms of code size or even better.
   //
   // However, we currently have no way of hinting to the register allocator that
   // MIMG addresses should be placed contiguously when it is possible to do so,
   // so force non-NSA for the common 2-address case as a heuristic.
   //
   // SIShrinkInstructions will convert NSA encodings to non-NSA after register
   // allocation when possible.
+  //
+  // TODO: we can actually allow partial NSA where the final register is a
+  // contiguous set of the remaining addresses.
+  // This could help where there are more addresses than supported.
   bool UseNSA = ST->hasFeature(AMDGPU::FeatureNSAEncoding) &&
                 VAddrs.size() >= 3 &&
                 VAddrs.size() <= (unsigned)ST->getNSAMaxSize();
   SDValue VAddr;
   if (!UseNSA)
     VAddr = getBuildDwordsVector(DAG, DL, VAddrs);
 
   SDValue True = DAG.getTargetConstant(1, DL, MVT::i1);
if (BaseOpcode->HasD16)
     Ops.push_back(IsD16 ? True : False);
   if (isa<MemSDNode>(Op))
     Ops.push_back(Op.getOperand(0)); // chain
 
   int NumVAddrDwords =
       UseNSA ? VAddrs.size() : VAddr.getValueType().getSizeInBits() / 32;
   int Opcode = -1;
 
-  if (IsGFX10Plus) {
+  if (IsGFX11Plus) {
+    Opcode = AMDGPU::getMIMGOpcode(IntrOpcode,
+                                   UseNSA ? AMDGPU::MIMGEncGfx11NSA
+                                          : AMDGPU::MIMGEncGfx11Default,
+                                   NumVDataDwords, NumVAddrDwords);
+  } else if (IsGFX10Plus) {
     Opcode = AMDGPU::getMIMGOpcode(IntrOpcode,
                                    UseNSA ? AMDGPU::MIMGEncGfx10NSA
                                           : AMDGPU::MIMGEncGfx10Default,
                                    NumVDataDwords, NumVAddrDwords);
   } else {
     if (Subtarget->hasGFX90AInsts()) {
       Opcode = AMDGPU::getMIMGOpcode(IntrOpcode, AMDGPU::MIMGEncGfx90a,
                                      NumVDataDwords, NumVAddrDwords);
case Intrinsic::amdgcn_image_bvh_intersect_ray: {
     assert(RayDir.getValueType() == MVT::v3f16 ||
            RayDir.getValueType() == MVT::v3f32);
 
     if (!Subtarget->hasGFX10_AEncoding()) {
       emitRemovedIntrinsicError(DAG, DL, Op.getValueType());
       return SDValue();
     }
 
+    const bool IsGFX11Plus = AMDGPU::isGFX11Plus(*Subtarget);
     const bool IsA16 = RayDir.getValueType().getVectorElementType() == MVT::f16;
     const bool Is64 = NodePtr.getValueType() == MVT::i64;
     const unsigned NumVDataDwords = 4;
     const unsigned NumVAddrDwords = IsA16 ? (Is64 ? 9 : 8) : (Is64 ? 12 : 11);
-    const bool UseNSA = Subtarget->hasNSAEncoding() &&
-                        NumVAddrDwords <= Subtarget->getNSAMaxSize();
+    const unsigned NumVAddrs = IsGFX11Plus ? (IsA16 ? 4 : 5) : NumVAddrDwords;
+    const bool UseNSA =
+        Subtarget->hasNSAEncoding() && NumVAddrs <= Subtarget->getNSAMaxSize();
     const unsigned BaseOpcodes[2][2] = {
         {AMDGPU::IMAGE_BVH_INTERSECT_RAY, AMDGPU::IMAGE_BVH_INTERSECT_RAY_a16},
         {AMDGPU::IMAGE_BVH64_INTERSECT_RAY,
          AMDGPU::IMAGE_BVH64_INTERSECT_RAY_a16}};
     int Opcode;
     if (UseNSA) {
       Opcode = AMDGPU::getMIMGOpcode(BaseOpcodes[Is64][IsA16],
-                                     AMDGPU::MIMGEncGfx10NSA, NumVDataDwords,
-                                     NumVAddrDwords);
+                                     IsGFX11Plus ? AMDGPU::MIMGEncGfx11NSA
+                                                 : AMDGPU::MIMGEncGfx10NSA,
+                                     NumVDataDwords, NumVAddrDwords);
     } else {
-      Opcode = AMDGPU::getMIMGOpcode(
-          BaseOpcodes[Is64][IsA16], AMDGPU::MIMGEncGfx10Default, NumVDataDwords,
-          PowerOf2Ceil(NumVAddrDwords));
+      Opcode =
+          AMDGPU::getMIMGOpcode(BaseOpcodes[Is64][IsA16],
+                                IsGFX11Plus ? AMDGPU::MIMGEncGfx11Default
+                                            : AMDGPU::MIMGEncGfx10Default,
+                                NumVDataDwords, PowerOf2Ceil(NumVAddrDwords));
     }
     assert(Opcode != -1);
 
     SmallVector<SDValue, 16> Ops;
 
     auto packLanes = [&DAG, &Ops, &DL] (SDValue Op, bool IsAligned) {
       SmallVector<SDValue, 3> Lanes;
       DAG.ExtractVectorElements(Op, Lanes, 0, 3);
auto packLanes = [&DAG, &Ops, &DL] (SDValue Op, bool IsAligned) {
           Ops.push_back(
             DAG.getBitcast(MVT::i32,
                            DAG.getBuildVector(MVT::v2f16, DL,
                                               { Lanes[1], Lanes[2] })));
         }
       }
     };
 
+    if (UseNSA && IsGFX11Plus) {
+      Ops.push_back(NodePtr);
+      Ops.push_back(DAG.getBitcast(MVT::i32, RayExtent));
+      Ops.push_back(RayOrigin);
+      if (IsA16) {
+        SmallVector<SDValue, 3> DirLanes, InvDirLanes, MergedLanes;
+        DAG.ExtractVectorElements(RayDir, DirLanes, 0, 3);
+        DAG.ExtractVectorElements(RayInvDir, InvDirLanes, 0, 3);
+        for (unsigned I = 0; I < 3; ++I) {
+          MergedLanes.push_back(DAG.getBitcast(
+              MVT::i32, DAG.getBuildVector(MVT::v2f16, DL,
+                                           {DirLanes[I], InvDirLanes[I]})));
+        }
+        Ops.push_back(DAG.getBuildVector(MVT::v3i32, DL, MergedLanes));
+      } else {
+        Ops.push_back(RayDir);
+        Ops.push_back(RayInvDir);
+      }
+    } else {
       if (Is64)
-      DAG.ExtractVectorElements(DAG.getBitcast(MVT::v2i32, NodePtr), Ops, 0, 2);
+        DAG.ExtractVectorElements(DAG.getBitcast(MVT::v2i32, NodePtr), Ops, 0,
+                                  2);
       else
         Ops.push_back(NodePtr);
 
       Ops.push_back(DAG.getBitcast(MVT::i32, RayExtent));
       packLanes(RayOrigin, true);
       packLanes(RayDir, true);
       packLanes(RayInvDir, false);
+    }
 
     if (!UseNSA) {
       // Build a single vector containing all the operands so far prepared.
       if (NumVAddrDwords > 8) {
         SDValue Undef = DAG.getUNDEF(MVT::i32);
         Ops.append(16 - Ops.size(), Undef);
       }
       assert(Ops.size() == 8 || Ops.size() == 16);

llvm/lib/Target/AMDGPU/SIShrinkInstructions.cpp

if ((TII->sopkIsZext(SOPKOpc) && isKUImmOperand(Src1)) ||
       (!TII->sopkIsZext(SOPKOpc) && isKImmOperand(Src1))) {
     MI.setDesc(NewDesc);
   }
 }
 
 // Shrink NSA encoded instructions with contiguous VGPRs to non-NSA encoding.
 void SIShrinkInstructions::shrinkMIMG(MachineInstr &MI) const {
   const AMDGPU::MIMGInfo *Info = AMDGPU::getMIMGInfo(MI.getOpcode());
-  if (!Info || Info->MIMGEncoding != AMDGPU::MIMGEncGfx10NSA)
+  if (!Info)
     return;
 
+  uint8_t NewEncoding;
+  switch (Info->MIMGEncoding) {
+  case AMDGPU::MIMGEncGfx10NSA:
+    NewEncoding = AMDGPU::MIMGEncGfx10Default;
+    break;
+  case AMDGPU::MIMGEncGfx11NSA:
+    NewEncoding = AMDGPU::MIMGEncGfx11Default;
+    break;
+  default:
+    return;
+  }
+
   int VAddr0Idx =
       AMDGPU::getNamedOperandIdx(MI.getOpcode(), AMDGPU::OpName::vaddr0);
   unsigned NewAddrDwords = Info->VAddrDwords;
   const TargetRegisterClass *RC;
 
   if (Info->VAddrDwords == 2) {
     RC = &AMDGPU::VReg_64RegClass;
   } else if (Info->VAddrDwords == 3) {
void SIShrinkInstructions::shrinkMIMG(MachineInstr &MI) const {
   } else if (Info->VAddrDwords == 8) {
     RC = &AMDGPU::VReg_256RegClass;
   } else {
     RC = &AMDGPU::VReg_512RegClass;
     NewAddrDwords = 16;
   }
 
   unsigned VgprBase = 0;
+  unsigned NextVgpr = 0;
   bool IsUndef = true;
   bool IsKill = NewAddrDwords == Info->VAddrDwords;
-  for (unsigned i = 0; i < Info->VAddrDwords; ++i) {
-    const MachineOperand &Op = MI.getOperand(VAddr0Idx + i);
+  for (unsigned Idx = 0; Idx < Info->VAddrOperands; ++Idx) {
+    const MachineOperand &Op = MI.getOperand(VAddr0Idx + Idx);
     unsigned Vgpr = TRI->getHWRegIndex(Op.getReg());
+    unsigned Dwords = TRI->getRegSizeInBits(Op.getReg(), *MRI) / 32;
+    assert(Dwords > 0 && "Un-implemented for less than 32 bit regs");
 
-    if (i == 0) {
+    if (Idx == 0) {
       VgprBase = Vgpr;
-    } else if (VgprBase + i != Vgpr)
+      NextVgpr = Vgpr + Dwords;
+    } else if (Vgpr == NextVgpr) {
+      NextVgpr = Vgpr + Dwords;
+    } else {
       return;
+    }
 
     if (!Op.isUndef())
       IsUndef = false;
     if (!Op.isKill())
       IsKill = false;
   }
 
   if (VgprBase + NewAddrDwords > 256)
for (unsigned i = LWEIdx + 1, e = MI.getNumOperands(); i != e; ++i) {
             ToUntie == -1 &&
             "found more than one tied implicit operand when expecting only 1");
         ToUntie = i;
         MI.untieRegOperand(ToUntie);
       }
     }
   }
 
-  unsigned NewOpcode =
-      AMDGPU::getMIMGOpcode(Info->BaseOpcode, AMDGPU::MIMGEncGfx10Default,
+  unsigned NewOpcode = AMDGPU::getMIMGOpcode(Info->BaseOpcode, NewEncoding,
                                              Info->VDataDwords, NewAddrDwords);
   MI.setDesc(TII->get(NewOpcode));
   MI.getOperand(VAddr0Idx).setReg(RC->getRegister(VgprBase));
   MI.getOperand(VAddr0Idx).setIsUndef(IsUndef);
   MI.getOperand(VAddr0Idx).setIsKill(IsKill);
 
-  for (unsigned i = 1; i < Info->VAddrDwords; ++i)
+  for (int i = 1; i < Info->VAddrOperands; ++i)
     MI.removeOperand(VAddr0Idx + 1);
 
   if (ToUntie >= 0) {
     MI.tieOperands(
         AMDGPU::getNamedOperandIdx(MI.getOpcode(), AMDGPU::OpName::vdata),
-        ToUntie - (Info->VAddrDwords - 1));
+        ToUntie - (Info->VAddrOperands - 1));
   }
 }
 
 // Shrink MAD to MADAK/MADMK and FMA to FMAAK/FMAMK.
 void SIShrinkInstructions::shrinkMadFma(MachineInstr &MI) const {
   if (!ST->hasVOP3Literal())
     return;
 

llvm/test/CodeGen/AMDGPU/GlobalISel/legalize-llvm.amdgcn.image.sample.a16.ll

 ; NOTE: Assertions have been autogenerated by utils/update_mir_test_checks.py
 ; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx900 -stop-after=legalizer -o - %s | FileCheck -check-prefix=GFX9 %s
 ; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx1010 -stop-after=legalizer -o - %s | FileCheck -check-prefix=GFX10 %s
+; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx1100 -stop-after=legalizer -o - %s | FileCheck -check-prefix=GFX10 %s
 
 define amdgpu_ps <4 x float> @sample_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, half %s) {
   ; GFX9-LABEL: name: sample_1d
   ; GFX9: bb.1.main_body:
   ; GFX9-NEXT:   liveins: $sgpr2, $sgpr3, $sgpr4, $sgpr5, $sgpr6, $sgpr7, $sgpr8, $sgpr9, $sgpr10, $sgpr11, $sgpr12, $sgpr13, $vgpr0
   ; GFX9-NEXT: {{  $}}
   ; GFX9-NEXT:   [[COPY:%[0-9]+]]:_(s32) = COPY $sgpr2
   ; GFX9-NEXT:   [[COPY1:%[0-9]+]]:_(s32) = COPY $sgpr3

llvm/test/CodeGen/AMDGPU/GlobalISel/llvm.amdgcn.image.atomic.dim.a16.ll

 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
 ; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx900 -o - %s | FileCheck -check-prefix=GFX9 %s
 ; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx1010 -o - %s | FileCheck -check-prefix=GFX10 %s
+; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx1100 -o - %s | FileCheck -check-prefix=GFX10 %s
 
 define amdgpu_ps float @atomic_swap_i32_1d(<8 x i32> inreg %rsrc, i32 %data, i16 %s) {
 ; GFX9-LABEL: atomic_swap_i32_1d:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    s_mov_b32 s0, s2
 ; GFX9-NEXT:    s_mov_b32 s1, s3
 ; GFX9-NEXT:    s_mov_b32 s2, s4
 ; GFX9-NEXT:    s_mov_b32 s3, s5

llvm/test/CodeGen/AMDGPU/GlobalISel/llvm.amdgcn.image.atomic.dim.ll

 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
 ; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=tahiti -o - %s | FileCheck -check-prefix=GFX6 %s
 ; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=fiji -o - %s | FileCheck -check-prefix=GFX8 %s
 ; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx900 -o - %s | FileCheck -check-prefix=GFX900 %s
 ; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx90a -o - %s | FileCheck -check-prefix=GFX90A %s
 ; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx1010 -o - %s | FileCheck -check-prefix=GFX10 %s
+; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx1100 -o - %s | FileCheck -check-prefix=GFX10 %s
 
 define amdgpu_ps float @atomic_swap_i32_1d(<8 x i32> inreg %rsrc, i32 %data, i32 %s) {
 ; GFX6-LABEL: atomic_swap_i32_1d:
 ; GFX6:       ; %bb.0: ; %main_body
 ; GFX6-NEXT:    s_mov_b32 s0, s2
 ; GFX6-NEXT:    s_mov_b32 s1, s3
 ; GFX6-NEXT:    s_mov_b32 s2, s4
 ; GFX6-NEXT:    s_mov_b32 s3, s5

llvm/test/CodeGen/AMDGPU/GlobalISel/llvm.amdgcn.image.gather4.a16.dim.ll

 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
 ; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx900 -o - %s | FileCheck -check-prefix=GFX9 %s
 ; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx1010 -o - %s | FileCheck -check-prefix=GFX10NSA %s
+; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx1100 -o - %s | FileCheck -check-prefix=GFX10NSA %s
 
 define amdgpu_ps <4 x float> @gather4_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, half %s, half %t) {
 ; GFX9-LABEL: gather4_2d:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    s_mov_b64 s[14:15], exec
 ; GFX9-NEXT:    s_mov_b32 s0, s2
 ; GFX9-NEXT:    s_wqm_b64 exec, exec
 ; GFX9-NEXT:    v_mov_b32_e32 v2, 0xffff

llvm/test/CodeGen/AMDGPU/GlobalISel/llvm.amdgcn.image.gather4.dim.ll

 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
 ; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=tahiti -o - %s | FileCheck -check-prefix=GFX6 %s
 ; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx1010 -o - %s | FileCheck -check-prefix=GFX10NSA %s
+; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx1100 -o - %s | FileCheck -check-prefix=GFX10NSA %s
 
 define amdgpu_ps <4 x float> @gather4_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %s, float %t) {
 ; GFX6-LABEL: gather4_2d:
 ; GFX6:       ; %bb.0: ; %main_body
 ; GFX6-NEXT:    s_mov_b64 s[14:15], exec
 ; GFX6-NEXT:    s_mov_b32 s0, s2
 ; GFX6-NEXT:    s_mov_b32 s1, s3
 ; GFX6-NEXT:    s_mov_b32 s2, s4

llvm/test/CodeGen/AMDGPU/GlobalISel/llvm.amdgcn.image.getresinfo.a16.ll

 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
 ; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx900 -verify-machineinstrs < %s | FileCheck -check-prefix=GFX9 %s
 ; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx1010 -verify-machineinstrs < %s | FileCheck -check-prefix=GFX10 %s
+; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx1100 -verify-machineinstrs < %s | FileCheck -check-prefix=GFX10 %s
 
 define amdgpu_ps <4 x float> @getresinfo_1d(<8 x i32> inreg %rsrc, i16 %mip) {
 ; GFX9-LABEL: getresinfo_1d:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    s_mov_b32 s0, s2
 ; GFX9-NEXT:    s_mov_b32 s1, s3
 ; GFX9-NEXT:    s_mov_b32 s2, s4
 ; GFX9-NEXT:    s_mov_b32 s3, s5

llvm/test/CodeGen/AMDGPU/GlobalISel/llvm.amdgcn.image.getresinfo.ll

 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
 ; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=tahiti -verify-machineinstrs < %s | FileCheck -check-prefix=GFX6 %s
 ; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=fiji -verify-machineinstrs < %s | FileCheck -check-prefix=GFX8 %s
 ; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx1010 -verify-machineinstrs < %s | FileCheck -check-prefix=GFX10 %s
+; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx1100 -verify-machineinstrs < %s | FileCheck -check-prefix=GFX10 %s
 
 define amdgpu_ps <4 x float> @getresinfo_1d(<8 x i32> inreg %rsrc, i32 %mip) {
 ; GFX6-LABEL: getresinfo_1d:
 ; GFX6:       ; %bb.0: ; %main_body
 ; GFX6-NEXT:    s_mov_b32 s0, s2
 ; GFX6-NEXT:    s_mov_b32 s1, s3
 ; GFX6-NEXT:    s_mov_b32 s2, s4
 ; GFX6-NEXT:    s_mov_b32 s3, s5

llvm/test/CodeGen/AMDGPU/GlobalISel/llvm.amdgcn.image.load.1d.d16.ll

 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
 ; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=tonga -verify-machineinstrs < %s | FileCheck -check-prefix=GFX8-UNPACKED %s
 ; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx810 -verify-machineinstrs < %s | FileCheck -check-prefix=GFX8-PACKED %s
 ; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx900 -verify-machineinstrs < %s | FileCheck -check-prefix=GFX9 %s
-; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx1010 -verify-machineinstrs < %s | FileCheck -check-prefix=GFX10 %s
+; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx1010 -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX10PLUS,GFX10 %s
+; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx1100 -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX10PLUS,GFX11 %s
 
 define amdgpu_ps half @load_1d_f16_x(<8 x i32> inreg %rsrc, i32 %s) {
 ; GFX8-UNPACKED-LABEL: load_1d_f16_x:
 ; GFX8-UNPACKED:       ; %bb.0:
 ; GFX8-UNPACKED-NEXT:    s_mov_b32 s0, s2
 ; GFX8-UNPACKED-NEXT:    s_mov_b32 s1, s3
 ; GFX8-UNPACKED-NEXT:    s_mov_b32 s2, s4
 ; GFX8-UNPACKED-NEXT:    s_mov_b32 s3, s5

 ; GFX9-NEXT:    s_mov_b32 s4, s6
 ; GFX9-NEXT:    s_mov_b32 s5, s7
 ; GFX9-NEXT:    s_mov_b32 s6, s8
 ; GFX9-NEXT:    s_mov_b32 s7, s9
 ; GFX9-NEXT:    image_load v0, v0, s[0:7] dmask:0x1 unorm d16
 ; GFX9-NEXT:    s_waitcnt vmcnt(0)
 ; GFX9-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: load_1d_f16_x:
-; GFX10:       ; %bb.0:
-; GFX10-NEXT:    s_mov_b32 s0, s2
-; GFX10-NEXT:    s_mov_b32 s1, s3
-; GFX10-NEXT:    s_mov_b32 s2, s4
-; GFX10-NEXT:    s_mov_b32 s3, s5
-; GFX10-NEXT:    s_mov_b32 s4, s6
-; GFX10-NEXT:    s_mov_b32 s5, s7
-; GFX10-NEXT:    s_mov_b32 s6, s8
-; GFX10-NEXT:    s_mov_b32 s7, s9
-; GFX10-NEXT:    image_load v0, v0, s[0:7] dmask:0x1 dim:SQ_RSRC_IMG_1D unorm d16
-; GFX10-NEXT:    s_waitcnt vmcnt(0)
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: load_1d_f16_x:
+; GFX10PLUS:       ; %bb.0:
+; GFX10PLUS-NEXT:    s_mov_b32 s0, s2
+; GFX10PLUS-NEXT:    s_mov_b32 s1, s3
+; GFX10PLUS-NEXT:    s_mov_b32 s2, s4
+; GFX10PLUS-NEXT:    s_mov_b32 s3, s5
+; GFX10PLUS-NEXT:    s_mov_b32 s4, s6
+; GFX10PLUS-NEXT:    s_mov_b32 s5, s7
+; GFX10PLUS-NEXT:    s_mov_b32 s6, s8
+; GFX10PLUS-NEXT:    s_mov_b32 s7, s9
+; GFX10PLUS-NEXT:    image_load v0, v0, s[0:7] dmask:0x1 dim:SQ_RSRC_IMG_1D unorm d16
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
   %v = call half @llvm.amdgcn.image.load.1d.half.i32(i32 1, i32 %s, <8 x i32> %rsrc, i32 0, i32 0)
   ret half %v
 }
 
 define amdgpu_ps half @load_1d_f16_y(<8 x i32> inreg %rsrc, i32 %s) {
 ; GFX8-UNPACKED-LABEL: load_1d_f16_y:
 ; GFX8-UNPACKED:       ; %bb.0:
 ; GFX8-UNPACKED-NEXT:    s_mov_b32 s0, s2

 ; GFX9-NEXT:    s_mov_b32 s4, s6
 ; GFX9-NEXT:    s_mov_b32 s5, s7
 ; GFX9-NEXT:    s_mov_b32 s6, s8
 ; GFX9-NEXT:    s_mov_b32 s7, s9
 ; GFX9-NEXT:    image_load v0, v0, s[0:7] dmask:0x2 unorm d16
 ; GFX9-NEXT:    s_waitcnt vmcnt(0)
 ; GFX9-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: load_1d_f16_y:
-; GFX10:       ; %bb.0:
-; GFX10-NEXT:    s_mov_b32 s0, s2
-; GFX10-NEXT:    s_mov_b32 s1, s3
-; GFX10-NEXT:    s_mov_b32 s2, s4
-; GFX10-NEXT:    s_mov_b32 s3, s5
-; GFX10-NEXT:    s_mov_b32 s4, s6
-; GFX10-NEXT:    s_mov_b32 s5, s7
-; GFX10-NEXT:    s_mov_b32 s6, s8
-; GFX10-NEXT:    s_mov_b32 s7, s9
-; GFX10-NEXT:    image_load v0, v0, s[0:7] dmask:0x2 dim:SQ_RSRC_IMG_1D unorm d16
-; GFX10-NEXT:    s_waitcnt vmcnt(0)
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: load_1d_f16_y:
+; GFX10PLUS:       ; %bb.0:
+; GFX10PLUS-NEXT:    s_mov_b32 s0, s2
+; GFX10PLUS-NEXT:    s_mov_b32 s1, s3
+; GFX10PLUS-NEXT:    s_mov_b32 s2, s4
+; GFX10PLUS-NEXT:    s_mov_b32 s3, s5
+; GFX10PLUS-NEXT:    s_mov_b32 s4, s6
+; GFX10PLUS-NEXT:    s_mov_b32 s5, s7
+; GFX10PLUS-NEXT:    s_mov_b32 s6, s8
+; GFX10PLUS-NEXT:    s_mov_b32 s7, s9
+; GFX10PLUS-NEXT:    image_load v0, v0, s[0:7] dmask:0x2 dim:SQ_RSRC_IMG_1D unorm d16
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
   %v = call half @llvm.amdgcn.image.load.1d.half.i32(i32 2, i32 %s, <8 x i32> %rsrc, i32 0, i32 0)
   ret half %v
 }
 
 define amdgpu_ps half @load_1d_f16_z(<8 x i32> inreg %rsrc, i32 %s) {
 ; GFX8-UNPACKED-LABEL: load_1d_f16_z:
 ; GFX8-UNPACKED:       ; %bb.0:
 ; GFX8-UNPACKED-NEXT:    s_mov_b32 s0, s2

 ; GFX9-NEXT:    s_mov_b32 s4, s6
 ; GFX9-NEXT:    s_mov_b32 s5, s7
 ; GFX9-NEXT:    s_mov_b32 s6, s8
 ; GFX9-NEXT:    s_mov_b32 s7, s9
 ; GFX9-NEXT:    image_load v0, v0, s[0:7] dmask:0x4 unorm d16
 ; GFX9-NEXT:    s_waitcnt vmcnt(0)
 ; GFX9-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: load_1d_f16_z:
-; GFX10:       ; %bb.0:
-; GFX10-NEXT:    s_mov_b32 s0, s2
-; GFX10-NEXT:    s_mov_b32 s1, s3
-; GFX10-NEXT:    s_mov_b32 s2, s4
-; GFX10-NEXT:    s_mov_b32 s3, s5
-; GFX10-NEXT:    s_mov_b32 s4, s6
-; GFX10-NEXT:    s_mov_b32 s5, s7
-; GFX10-NEXT:    s_mov_b32 s6, s8
-; GFX10-NEXT:    s_mov_b32 s7, s9
-; GFX10-NEXT:    image_load v0, v0, s[0:7] dmask:0x4 dim:SQ_RSRC_IMG_1D unorm d16
-; GFX10-NEXT:    s_waitcnt vmcnt(0)
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: load_1d_f16_z:
+; GFX10PLUS:       ; %bb.0:
+; GFX10PLUS-NEXT:    s_mov_b32 s0, s2
+; GFX10PLUS-NEXT:    s_mov_b32 s1, s3
+; GFX10PLUS-NEXT:    s_mov_b32 s2, s4
+; GFX10PLUS-NEXT:    s_mov_b32 s3, s5
+; GFX10PLUS-NEXT:    s_mov_b32 s4, s6
+; GFX10PLUS-NEXT:    s_mov_b32 s5, s7
+; GFX10PLUS-NEXT:    s_mov_b32 s6, s8
+; GFX10PLUS-NEXT:    s_mov_b32 s7, s9
+; GFX10PLUS-NEXT:    image_load v0, v0, s[0:7] dmask:0x4 dim:SQ_RSRC_IMG_1D unorm d16
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
   %v = call half @llvm.amdgcn.image.load.1d.half.i32(i32 4, i32 %s, <8 x i32> %rsrc, i32 0, i32 0)
   ret half %v
 }
 
 define amdgpu_ps half @load_1d_f16_w(<8 x i32> inreg %rsrc, i32 %s) {
 ; GFX8-UNPACKED-LABEL: load_1d_f16_w:
 ; GFX8-UNPACKED:       ; %bb.0:
 ; GFX8-UNPACKED-NEXT:    s_mov_b32 s0, s2

 ; GFX9-NEXT:    s_mov_b32 s4, s6
 ; GFX9-NEXT:    s_mov_b32 s5, s7
 ; GFX9-NEXT:    s_mov_b32 s6, s8
 ; GFX9-NEXT:    s_mov_b32 s7, s9
 ; GFX9-NEXT:    image_load v0, v0, s[0:7] dmask:0x8 unorm d16
 ; GFX9-NEXT:    s_waitcnt vmcnt(0)
 ; GFX9-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: load_1d_f16_w:
-; GFX10:       ; %bb.0:
-; GFX10-NEXT:    s_mov_b32 s0, s2
-; GFX10-NEXT:    s_mov_b32 s1, s3
-; GFX10-NEXT:    s_mov_b32 s2, s4
-; GFX10-NEXT:    s_mov_b32 s3, s5
-; GFX10-NEXT:    s_mov_b32 s4, s6
-; GFX10-NEXT:    s_mov_b32 s5, s7
-; GFX10-NEXT:    s_mov_b32 s6, s8
-; GFX10-NEXT:    s_mov_b32 s7, s9
-; GFX10-NEXT:    image_load v0, v0, s[0:7] dmask:0x8 dim:SQ_RSRC_IMG_1D unorm d16
-; GFX10-NEXT:    s_waitcnt vmcnt(0)
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: load_1d_f16_w:
+; GFX10PLUS:       ; %bb.0:
+; GFX10PLUS-NEXT:    s_mov_b32 s0, s2
+; GFX10PLUS-NEXT:    s_mov_b32 s1, s3
+; GFX10PLUS-NEXT:    s_mov_b32 s2, s4
+; GFX10PLUS-NEXT:    s_mov_b32 s3, s5
+; GFX10PLUS-NEXT:    s_mov_b32 s4, s6
+; GFX10PLUS-NEXT:    s_mov_b32 s5, s7
+; GFX10PLUS-NEXT:    s_mov_b32 s6, s8
+; GFX10PLUS-NEXT:    s_mov_b32 s7, s9
+; GFX10PLUS-NEXT:    image_load v0, v0, s[0:7] dmask:0x8 dim:SQ_RSRC_IMG_1D unorm d16
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
   %v = call half @llvm.amdgcn.image.load.1d.half.i32(i32 8, i32 %s, <8 x i32> %rsrc, i32 0, i32 0)
   ret half %v
 }
 
 define amdgpu_ps <2 x half> @load_1d_v2f16_xy(<8 x i32> inreg %rsrc, i32 %s) {
 ; GFX8-UNPACKED-LABEL: load_1d_v2f16_xy:
 ; GFX8-UNPACKED:       ; %bb.0:
 ; GFX8-UNPACKED-NEXT:    s_mov_b32 s0, s2

 ; GFX9-NEXT:    s_mov_b32 s4, s6
 ; GFX9-NEXT:    s_mov_b32 s5, s7
 ; GFX9-NEXT:    s_mov_b32 s6, s8
 ; GFX9-NEXT:    s_mov_b32 s7, s9
 ; GFX9-NEXT:    image_load v0, v0, s[0:7] dmask:0x3 unorm d16
 ; GFX9-NEXT:    s_waitcnt vmcnt(0)
 ; GFX9-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: load_1d_v2f16_xy:
-; GFX10:       ; %bb.0:
-; GFX10-NEXT:    s_mov_b32 s0, s2
-; GFX10-NEXT:    s_mov_b32 s1, s3
-; GFX10-NEXT:    s_mov_b32 s2, s4
-; GFX10-NEXT:    s_mov_b32 s3, s5
-; GFX10-NEXT:    s_mov_b32 s4, s6
-; GFX10-NEXT:    s_mov_b32 s5, s7
-; GFX10-NEXT:    s_mov_b32 s6, s8
-; GFX10-NEXT:    s_mov_b32 s7, s9
-; GFX10-NEXT:    image_load v0, v0, s[0:7] dmask:0x3 dim:SQ_RSRC_IMG_1D unorm d16
-; GFX10-NEXT:    s_waitcnt vmcnt(0)
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: load_1d_v2f16_xy:
+; GFX10PLUS:       ; %bb.0:
+; GFX10PLUS-NEXT:    s_mov_b32 s0, s2
+; GFX10PLUS-NEXT:    s_mov_b32 s1, s3
+; GFX10PLUS-NEXT:    s_mov_b32 s2, s4
+; GFX10PLUS-NEXT:    s_mov_b32 s3, s5
+; GFX10PLUS-NEXT:    s_mov_b32 s4, s6
+; GFX10PLUS-NEXT:    s_mov_b32 s5, s7
+; GFX10PLUS-NEXT:    s_mov_b32 s6, s8
+; GFX10PLUS-NEXT:    s_mov_b32 s7, s9
+; GFX10PLUS-NEXT:    image_load v0, v0, s[0:7] dmask:0x3 dim:SQ_RSRC_IMG_1D unorm d16
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
   %v = call <2 x half> @llvm.amdgcn.image.load.1d.v2f16.i32(i32 3, i32 %s, <8 x i32> %rsrc, i32 0, i32 0)
   ret <2 x half> %v
 }
 
 define amdgpu_ps <2 x half> @load_1d_v2f16_xz(<8 x i32> inreg %rsrc, i32 %s) {
 ; GFX8-UNPACKED-LABEL: load_1d_v2f16_xz:
 ; GFX8-UNPACKED:       ; %bb.0:
 ; GFX8-UNPACKED-NEXT:    s_mov_b32 s0, s2

 ; GFX9-NEXT:    s_mov_b32 s4, s6
 ; GFX9-NEXT:    s_mov_b32 s5, s7
 ; GFX9-NEXT:    s_mov_b32 s6, s8
 ; GFX9-NEXT:    s_mov_b32 s7, s9
 ; GFX9-NEXT:    image_load v0, v0, s[0:7] dmask:0x5 unorm d16
 ; GFX9-NEXT:    s_waitcnt vmcnt(0)
 ; GFX9-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: load_1d_v2f16_xz:
-; GFX10:       ; %bb.0:
-; GFX10-NEXT:    s_mov_b32 s0, s2
-; GFX10-NEXT:    s_mov_b32 s1, s3
-; GFX10-NEXT:    s_mov_b32 s2, s4
-; GFX10-NEXT:    s_mov_b32 s3, s5
-; GFX10-NEXT:    s_mov_b32 s4, s6
-; GFX10-NEXT:    s_mov_b32 s5, s7
-; GFX10-NEXT:    s_mov_b32 s6, s8
-; GFX10-NEXT:    s_mov_b32 s7, s9
-; GFX10-NEXT:    image_load v0, v0, s[0:7] dmask:0x5 dim:SQ_RSRC_IMG_1D unorm d16
-; GFX10-NEXT:    s_waitcnt vmcnt(0)
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: load_1d_v2f16_xz:
+; GFX10PLUS:       ; %bb.0:
+; GFX10PLUS-NEXT:    s_mov_b32 s0, s2
+; GFX10PLUS-NEXT:    s_mov_b32 s1, s3
+; GFX10PLUS-NEXT:    s_mov_b32 s2, s4
+; GFX10PLUS-NEXT:    s_mov_b32 s3, s5
+; GFX10PLUS-NEXT:    s_mov_b32 s4, s6
+; GFX10PLUS-NEXT:    s_mov_b32 s5, s7
+; GFX10PLUS-NEXT:    s_mov_b32 s6, s8
+; GFX10PLUS-NEXT:    s_mov_b32 s7, s9
+; GFX10PLUS-NEXT:    image_load v0, v0, s[0:7] dmask:0x5 dim:SQ_RSRC_IMG_1D unorm d16
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
   %v = call <2 x half> @llvm.amdgcn.image.load.1d.v2f16.i32(i32 5, i32 %s, <8 x i32> %rsrc, i32 0, i32 0)
   ret <2 x half> %v
 }
 
 define amdgpu_ps <2 x half> @load_1d_v2f16_xw(<8 x i32> inreg %rsrc, i32 %s) {
 ; GFX8-UNPACKED-LABEL: load_1d_v2f16_xw:
 ; GFX8-UNPACKED:       ; %bb.0:
 ; GFX8-UNPACKED-NEXT:    s_mov_b32 s0, s2

 ; GFX9-NEXT:    s_mov_b32 s4, s6
 ; GFX9-NEXT:    s_mov_b32 s5, s7
 ; GFX9-NEXT:    s_mov_b32 s6, s8
 ; GFX9-NEXT:    s_mov_b32 s7, s9
 ; GFX9-NEXT:    image_load v0, v0, s[0:7] dmask:0x9 unorm d16
 ; GFX9-NEXT:    s_waitcnt vmcnt(0)
 ; GFX9-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: load_1d_v2f16_xw:
-; GFX10:       ; %bb.0:
-; GFX10-NEXT:    s_mov_b32 s0, s2
-; GFX10-NEXT:    s_mov_b32 s1, s3
-; GFX10-NEXT:    s_mov_b32 s2, s4
-; GFX10-NEXT:    s_mov_b32 s3, s5
-; GFX10-NEXT:    s_mov_b32 s4, s6
-; GFX10-NEXT:    s_mov_b32 s5, s7
-; GFX10-NEXT:    s_mov_b32 s6, s8
-; GFX10-NEXT:    s_mov_b32 s7, s9
-; GFX10-NEXT:    image_load v0, v0, s[0:7] dmask:0x9 dim:SQ_RSRC_IMG_1D unorm d16
-; GFX10-NEXT:    s_waitcnt vmcnt(0)
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: load_1d_v2f16_xw:
+; GFX10PLUS:       ; %bb.0:
+; GFX10PLUS-NEXT:    s_mov_b32 s0, s2
+; GFX10PLUS-NEXT:    s_mov_b32 s1, s3
+; GFX10PLUS-NEXT:    s_mov_b32 s2, s4
+; GFX10PLUS-NEXT:    s_mov_b32 s3, s5
+; GFX10PLUS-NEXT:    s_mov_b32 s4, s6
+; GFX10PLUS-NEXT:    s_mov_b32 s5, s7
+; GFX10PLUS-NEXT:    s_mov_b32 s6, s8
+; GFX10PLUS-NEXT:    s_mov_b32 s7, s9
+; GFX10PLUS-NEXT:    image_load v0, v0, s[0:7] dmask:0x9 dim:SQ_RSRC_IMG_1D unorm d16
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
   %v = call <2 x half> @llvm.amdgcn.image.load.1d.v2f16.i32(i32 9, i32 %s, <8 x i32> %rsrc, i32 0, i32 0)
   ret <2 x half> %v
 }
 
 define amdgpu_ps <2 x half> @load_1d_v2f16_yz(<8 x i32> inreg %rsrc, i32 %s) {
 ; GFX8-UNPACKED-LABEL: load_1d_v2f16_yz:
 ; GFX8-UNPACKED:       ; %bb.0:
 ; GFX8-UNPACKED-NEXT:    s_mov_b32 s0, s2

 ; GFX9-NEXT:    s_mov_b32 s4, s6
 ; GFX9-NEXT:    s_mov_b32 s5, s7
 ; GFX9-NEXT:    s_mov_b32 s6, s8
 ; GFX9-NEXT:    s_mov_b32 s7, s9
 ; GFX9-NEXT:    image_load v0, v0, s[0:7] dmask:0x6 unorm d16
 ; GFX9-NEXT:    s_waitcnt vmcnt(0)
 ; GFX9-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: load_1d_v2f16_yz:
-; GFX10:       ; %bb.0:
-; GFX10-NEXT:    s_mov_b32 s0, s2
-; GFX10-NEXT:    s_mov_b32 s1, s3
-; GFX10-NEXT:    s_mov_b32 s2, s4
-; GFX10-NEXT:    s_mov_b32 s3, s5
-; GFX10-NEXT:    s_mov_b32 s4, s6
-; GFX10-NEXT:    s_mov_b32 s5, s7
-; GFX10-NEXT:    s_mov_b32 s6, s8
-; GFX10-NEXT:    s_mov_b32 s7, s9
-; GFX10-NEXT:    image_load v0, v0, s[0:7] dmask:0x6 dim:SQ_RSRC_IMG_1D unorm d16
-; GFX10-NEXT:    s_waitcnt vmcnt(0)
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: load_1d_v2f16_yz:
+; GFX10PLUS:       ; %bb.0:
+; GFX10PLUS-NEXT:    s_mov_b32 s0, s2
+; GFX10PLUS-NEXT:    s_mov_b32 s1, s3
+; GFX10PLUS-NEXT:    s_mov_b32 s2, s4
+; GFX10PLUS-NEXT:    s_mov_b32 s3, s5
+; GFX10PLUS-NEXT:    s_mov_b32 s4, s6
+; GFX10PLUS-NEXT:    s_mov_b32 s5, s7
+; GFX10PLUS-NEXT:    s_mov_b32 s6, s8
+; GFX10PLUS-NEXT:    s_mov_b32 s7, s9
+; GFX10PLUS-NEXT:    image_load v0, v0, s[0:7] dmask:0x6 dim:SQ_RSRC_IMG_1D unorm d16
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
   %v = call <2 x half> @llvm.amdgcn.image.load.1d.v2f16.i32(i32 6, i32 %s, <8 x i32> %rsrc, i32 0, i32 0)
   ret <2 x half> %v
 }
 
 define amdgpu_ps <3 x half> @load_1d_v3f16_xyz(<8 x i32> inreg %rsrc, i32 %s) {
 ; GFX8-UNPACKED-LABEL: load_1d_v3f16_xyz:
 ; GFX8-UNPACKED:       ; %bb.0:
 ; GFX8-UNPACKED-NEXT:    s_mov_b32 s0, s2

 ; GFX10-NEXT:    s_waitcnt_depctr 0xffe3
 ; GFX10-NEXT:    s_lshl_b32 s0, s0, 16
 ; GFX10-NEXT:    s_waitcnt vmcnt(0)
 ; GFX10-NEXT:    v_lshrrev_b32_e32 v2, 16, v0
 ; GFX10-NEXT:    v_and_or_b32 v1, 0xffff, v1, s0
 ; GFX10-NEXT:    v_lshlrev_b32_e32 v2, 16, v2
 ; GFX10-NEXT:    v_and_or_b32 v0, 0xffff, v0, v2
 ; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: load_1d_v3f16_xyz:
+; GFX11:       ; %bb.0:
+; GFX11-NEXT:    s_mov_b32 s0, s2
+; GFX11-NEXT:    s_mov_b32 s1, s3
+; GFX11-NEXT:    s_mov_b32 s2, s4
+; GFX11-NEXT:    s_mov_b32 s3, s5
+; GFX11-NEXT:    s_mov_b32 s4, s6
+; GFX11-NEXT:    s_mov_b32 s5, s7
+; GFX11-NEXT:    s_mov_b32 s6, s8
+; GFX11-NEXT:    s_mov_b32 s7, s9
+; GFX11-NEXT:    image_load v[0:1], v0, s[0:7] dmask:0x7 dim:SQ_RSRC_IMG_1D unorm d16
+; GFX11-NEXT:    s_lshl_b32 s0, s0, 16
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    v_lshrrev_b32_e32 v2, 16, v0
+; GFX11-NEXT:    v_and_or_b32 v1, 0xffff, v1, s0
+; GFX11-NEXT:    v_lshlrev_b32_e32 v2, 16, v2
+; GFX11-NEXT:    v_and_or_b32 v0, 0xffff, v0, v2
+; GFX11-NEXT:    ; return to shader part epilog
   %v = call <3 x half> @llvm.amdgcn.image.load.1d.v3f16.i32(i32 7, i32 %s, <8 x i32> %rsrc, i32 0, i32 0)
   ret <3 x half> %v
 }
 
 define amdgpu_ps <4 x half> @load_1d_v4f16_xyzw(<8 x i32> inreg %rsrc, i32 %s) {
 ; GFX8-UNPACKED-LABEL: load_1d_v4f16_xyzw:
 ; GFX8-UNPACKED:       ; %bb.0:
 ; GFX8-UNPACKED-NEXT:    s_mov_b32 s0, s2

 ; GFX9-NEXT:    s_mov_b32 s4, s6
 ; GFX9-NEXT:    s_mov_b32 s5, s7
 ; GFX9-NEXT:    s_mov_b32 s6, s8
 ; GFX9-NEXT:    s_mov_b32 s7, s9
 ; GFX9-NEXT:    image_load v[0:1], v0, s[0:7] dmask:0xf unorm d16
 ; GFX9-NEXT:    s_waitcnt vmcnt(0)
 ; GFX9-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: load_1d_v4f16_xyzw:
-; GFX10:       ; %bb.0:
-; GFX10-NEXT:    s_mov_b32 s0, s2
-; GFX10-NEXT:    s_mov_b32 s1, s3
-; GFX10-NEXT:    s_mov_b32 s2, s4
-; GFX10-NEXT:    s_mov_b32 s3, s5
-; GFX10-NEXT:    s_mov_b32 s4, s6
-; GFX10-NEXT:    s_mov_b32 s5, s7
-; GFX10-NEXT:    s_mov_b32 s6, s8
-; GFX10-NEXT:    s_mov_b32 s7, s9
-; GFX10-NEXT:    image_load v[0:1], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm d16
-; GFX10-NEXT:    s_waitcnt vmcnt(0)
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: load_1d_v4f16_xyzw:
+; GFX10PLUS:       ; %bb.0:
+; GFX10PLUS-NEXT:    s_mov_b32 s0, s2
+; GFX10PLUS-NEXT:    s_mov_b32 s1, s3
+; GFX10PLUS-NEXT:    s_mov_b32 s2, s4
+; GFX10PLUS-NEXT:    s_mov_b32 s3, s5
+; GFX10PLUS-NEXT:    s_mov_b32 s4, s6
+; GFX10PLUS-NEXT:    s_mov_b32 s5, s7
+; GFX10PLUS-NEXT:    s_mov_b32 s6, s8
+; GFX10PLUS-NEXT:    s_mov_b32 s7, s9
+; GFX10PLUS-NEXT:    image_load v[0:1], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm d16
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
   %v = call <4 x half> @llvm.amdgcn.image.load.1d.v4f16.i32(i32 15, i32 %s, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x half> %v
 }
 
 define amdgpu_ps float @load_1d_f16_tfe_dmask_x(<8 x i32> inreg %rsrc, i32 %s) {
 ; GFX8-UNPACKED-LABEL: load_1d_f16_tfe_dmask_x:
 ; GFX8-UNPACKED:       ; %bb.0:
 ; GFX8-UNPACKED-NEXT:    v_mov_b32_e32 v1, 0

 ; GFX9-NEXT:    s_mov_b32 s6, s8
 ; GFX9-NEXT:    s_mov_b32 s7, s9
 ; GFX9-NEXT:    v_mov_b32_e32 v2, v1
 ; GFX9-NEXT:    image_load v[1:2], v0, s[0:7] dmask:0x1 unorm tfe d16
 ; GFX9-NEXT:    s_waitcnt vmcnt(0)
 ; GFX9-NEXT:    v_mov_b32_e32 v0, v2
 ; GFX9-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: load_1d_f16_tfe_dmask_x:
-; GFX10:       ; %bb.0:
-; GFX10-NEXT:    v_mov_b32_e32 v1, 0
-; GFX10-NEXT:    s_mov_b32 s0, s2
-; GFX10-NEXT:    s_mov_b32 s1, s3
-; GFX10-NEXT:    s_mov_b32 s2, s4
-; GFX10-NEXT:    s_mov_b32 s3, s5
-; GFX10-NEXT:    s_mov_b32 s4, s6
-; GFX10-NEXT:    s_mov_b32 s5, s7
-; GFX10-NEXT:    s_mov_b32 s6, s8
-; GFX10-NEXT:    s_mov_b32 s7, s9
-; GFX10-NEXT:    v_mov_b32_e32 v2, v1
-; GFX10-NEXT:    image_load v[1:2], v0, s[0:7] dmask:0x1 dim:SQ_RSRC_IMG_1D unorm tfe d16
-; GFX10-NEXT:    s_waitcnt vmcnt(0)
-; GFX10-NEXT:    v_mov_b32_e32 v0, v2
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: load_1d_f16_tfe_dmask_x:
+; GFX10PLUS:       ; %bb.0:
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v1, 0
+; GFX10PLUS-NEXT:    s_mov_b32 s0, s2
+; GFX10PLUS-NEXT:    s_mov_b32 s1, s3
+; GFX10PLUS-NEXT:    s_mov_b32 s2, s4
+; GFX10PLUS-NEXT:    s_mov_b32 s3, s5
+; GFX10PLUS-NEXT:    s_mov_b32 s4, s6
+; GFX10PLUS-NEXT:    s_mov_b32 s5, s7
+; GFX10PLUS-NEXT:    s_mov_b32 s6, s8
+; GFX10PLUS-NEXT:    s_mov_b32 s7, s9
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v2, v1
+; GFX10PLUS-NEXT:    image_load v[1:2], v0, s[0:7] dmask:0x1 dim:SQ_RSRC_IMG_1D unorm tfe d16
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v0, v2
+; GFX10PLUS-NEXT:    ; return to shader part epilog
   %v = call { half, i32 } @llvm.amdgcn.image.load.1d.sl_f16i32s.i32(i32 1, i32 %s, <8 x i32> %rsrc, i32 1, i32 0)
   %v.err = extractvalue { half, i32 } %v, 1
   %vv = bitcast i32 %v.err to float
   ret float %vv
 }
 
 define amdgpu_ps float @load_1d_v2f16_tfe_dmask_xy(<8 x i32> inreg %rsrc, i32 %s) {
 ; GFX8-UNPACKED-LABEL: load_1d_v2f16_tfe_dmask_xy:

 ; GFX9-NEXT:    s_mov_b32 s6, s8
 ; GFX9-NEXT:    s_mov_b32 s7, s9
 ; GFX9-NEXT:    v_mov_b32_e32 v2, v1
 ; GFX9-NEXT:    image_load v[1:2], v0, s[0:7] dmask:0x3 unorm tfe d16
 ; GFX9-NEXT:    s_waitcnt vmcnt(0)
 ; GFX9-NEXT:    v_mov_b32_e32 v0, v2
 ; GFX9-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: load_1d_v2f16_tfe_dmask_xy:
-; GFX10:       ; %bb.0:
-; GFX10-NEXT:    v_mov_b32_e32 v1, 0
-; GFX10-NEXT:    s_mov_b32 s0, s2
-; GFX10-NEXT:    s_mov_b32 s1, s3
-; GFX10-NEXT:    s_mov_b32 s2, s4
-; GFX10-NEXT:    s_mov_b32 s3, s5
-; GFX10-NEXT:    s_mov_b32 s4, s6
-; GFX10-NEXT:    s_mov_b32 s5, s7
-; GFX10-NEXT:    s_mov_b32 s6, s8
-; GFX10-NEXT:    s_mov_b32 s7, s9
-; GFX10-NEXT:    v_mov_b32_e32 v2, v1
-; GFX10-NEXT:    image_load v[1:2], v0, s[0:7] dmask:0x3 dim:SQ_RSRC_IMG_1D unorm tfe d16
-; GFX10-NEXT:    s_waitcnt vmcnt(0)
-; GFX10-NEXT:    v_mov_b32_e32 v0, v2
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: load_1d_v2f16_tfe_dmask_xy:
+; GFX10PLUS:       ; %bb.0:
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v1, 0
+; GFX10PLUS-NEXT:    s_mov_b32 s0, s2
+; GFX10PLUS-NEXT:    s_mov_b32 s1, s3
+; GFX10PLUS-NEXT:    s_mov_b32 s2, s4
+; GFX10PLUS-NEXT:    s_mov_b32 s3, s5
+; GFX10PLUS-NEXT:    s_mov_b32 s4, s6
+; GFX10PLUS-NEXT:    s_mov_b32 s5, s7
+; GFX10PLUS-NEXT:    s_mov_b32 s6, s8
+; GFX10PLUS-NEXT:    s_mov_b32 s7, s9
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v2, v1
+; GFX10PLUS-NEXT:    image_load v[1:2], v0, s[0:7] dmask:0x3 dim:SQ_RSRC_IMG_1D unorm tfe d16
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v0, v2
+; GFX10PLUS-NEXT:    ; return to shader part epilog
   %v = call { <2 x half>, i32 } @llvm.amdgcn.image.load.1d.sl_v2f16i32s.i32(i32 3, i32 %s, <8 x i32> %rsrc, i32 1, i32 0)
   %v.err = extractvalue { <2 x half>, i32 } %v, 1
   %vv = bitcast i32 %v.err to float
   ret float %vv
 }
 
 define amdgpu_ps float @load_1d_v3f16_tfe_dmask_xyz(<8 x i32> inreg %rsrc, i32 %s) {
 ; GFX8-UNPACKED-LABEL: load_1d_v3f16_tfe_dmask_xyz:

 ; GFX9-NEXT:    s_mov_b32 s7, s9
 ; GFX9-NEXT:    v_mov_b32_e32 v2, v1
 ; GFX9-NEXT:    v_mov_b32_e32 v3, v1
 ; GFX9-NEXT:    image_load v[1:3], v0, s[0:7] dmask:0x7 unorm tfe d16
 ; GFX9-NEXT:    s_waitcnt vmcnt(0)
 ; GFX9-NEXT:    v_mov_b32_e32 v0, v3
 ; GFX9-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: load_1d_v3f16_tfe_dmask_xyz:
-; GFX10:       ; %bb.0:
-; GFX10-NEXT:    v_mov_b32_e32 v1, 0
-; GFX10-NEXT:    s_mov_b32 s0, s2
-; GFX10-NEXT:    s_mov_b32 s1, s3
-; GFX10-NEXT:    s_mov_b32 s2, s4
-; GFX10-NEXT:    s_mov_b32 s3, s5
-; GFX10-NEXT:    s_mov_b32 s4, s6
-; GFX10-NEXT:    s_mov_b32 s5, s7
-; GFX10-NEXT:    s_mov_b32 s6, s8
-; GFX10-NEXT:    s_mov_b32 s7, s9
-; GFX10-NEXT:    v_mov_b32_e32 v2, v1
-; GFX10-NEXT:    v_mov_b32_e32 v3, v1
-; GFX10-NEXT:    image_load v[1:3], v0, s[0:7] dmask:0x7 dim:SQ_RSRC_IMG_1D unorm tfe d16
-; GFX10-NEXT:    s_waitcnt vmcnt(0)
-; GFX10-NEXT:    v_mov_b32_e32 v0, v3
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: load_1d_v3f16_tfe_dmask_xyz:
+; GFX10PLUS:       ; %bb.0:
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v1, 0
+; GFX10PLUS-NEXT:    s_mov_b32 s0, s2
+; GFX10PLUS-NEXT:    s_mov_b32 s1, s3
+; GFX10PLUS-NEXT:    s_mov_b32 s2, s4
+; GFX10PLUS-NEXT:    s_mov_b32 s3, s5
+; GFX10PLUS-NEXT:    s_mov_b32 s4, s6
+; GFX10PLUS-NEXT:    s_mov_b32 s5, s7
+; GFX10PLUS-NEXT:    s_mov_b32 s6, s8
+; GFX10PLUS-NEXT:    s_mov_b32 s7, s9
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v2, v1
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v3, v1
+; GFX10PLUS-NEXT:    image_load v[1:3], v0, s[0:7] dmask:0x7 dim:SQ_RSRC_IMG_1D unorm tfe d16
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v0, v3
+; GFX10PLUS-NEXT:    ; return to shader part epilog
   %v = call { <3 x half>, i32 } @llvm.amdgcn.image.load.1d.sl_v3f16i32s.i32(i32 7, i32 %s, <8 x i32> %rsrc, i32 1, i32 0)
   %v.err = extractvalue { <3 x half>, i32 } %v, 1
   %vv = bitcast i32 %v.err to float
   ret float %vv
 }
 
 define amdgpu_ps float @load_1d_v4f16_tfe_dmask_xyzw(<8 x i32> inreg %rsrc, i32 %s) {
 ; GFX8-UNPACKED-LABEL: load_1d_v4f16_tfe_dmask_xyzw:

 ; GFX9-NEXT:    s_mov_b32 s6, s8
 ; GFX9-NEXT:    s_mov_b32 s7, s9
 ; GFX9-NEXT:    v_mov_b32_e32 v2, v1
 ; GFX9-NEXT:    image_load v[1:2], v0, s[0:7] dmask:0x10 unorm tfe d16
 ; GFX9-NEXT:    s_waitcnt vmcnt(0)
 ; GFX9-NEXT:    v_mov_b32_e32 v0, v2
 ; GFX9-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: load_1d_v4f16_tfe_dmask_xyzw:
-; GFX10:       ; %bb.0:
-; GFX10-NEXT:    v_mov_b32_e32 v1, 0
-; GFX10-NEXT:    s_mov_b32 s0, s2
-; GFX10-NEXT:    s_mov_b32 s1, s3
-; GFX10-NEXT:    s_mov_b32 s2, s4
-; GFX10-NEXT:    s_mov_b32 s3, s5
-; GFX10-NEXT:    s_mov_b32 s4, s6
-; GFX10-NEXT:    s_mov_b32 s5, s7
-; GFX10-NEXT:    s_mov_b32 s6, s8
-; GFX10-NEXT:    s_mov_b32 s7, s9
-; GFX10-NEXT:    v_mov_b32_e32 v2, v1
-; GFX10-NEXT:    image_load v[1:2], v0, s[0:7] dmask:0x10 dim:SQ_RSRC_IMG_1D unorm tfe d16
-; GFX10-NEXT:    s_waitcnt vmcnt(0)
-; GFX10-NEXT:    v_mov_b32_e32 v0, v2
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: load_1d_v4f16_tfe_dmask_xyzw:
+; GFX10PLUS:       ; %bb.0:
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v1, 0
+; GFX10PLUS-NEXT:    s_mov_b32 s0, s2
+; GFX10PLUS-NEXT:    s_mov_b32 s1, s3
+; GFX10PLUS-NEXT:    s_mov_b32 s2, s4
+; GFX10PLUS-NEXT:    s_mov_b32 s3, s5
+; GFX10PLUS-NEXT:    s_mov_b32 s4, s6
+; GFX10PLUS-NEXT:    s_mov_b32 s5, s7
+; GFX10PLUS-NEXT:    s_mov_b32 s6, s8
+; GFX10PLUS-NEXT:    s_mov_b32 s7, s9
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v2, v1
+; GFX10PLUS-NEXT:    image_load v[1:2], v0, s[0:7] dmask:0x10 dim:SQ_RSRC_IMG_1D unorm tfe d16
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v0, v2
+; GFX10PLUS-NEXT:    ; return to shader part epilog
   %v = call { <4 x half>, i32 } @llvm.amdgcn.image.load.1d.sl_v4f16i32s.i32(i32 16, i32 %s, <8 x i32> %rsrc, i32 1, i32 0)
   %v.err = extractvalue { <4 x half>, i32 } %v, 1
   %vv = bitcast i32 %v.err to float
   ret float %vv
 }
 
 declare half @llvm.amdgcn.image.load.1d.half.i32(i32 immarg, i32, <8 x i32>, i32 immarg, i32 immarg) #0
 declare <2 x half> @llvm.amdgcn.image.load.1d.v2f16.i32(i32 immarg, i32, <8 x i32>, i32 immarg, i32 immarg) #0

llvm/test/CodeGen/AMDGPU/GlobalISel/llvm.amdgcn.image.load.1d.ll

 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
 ; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=tahiti -verify-machineinstrs < %s | FileCheck -check-prefix=GFX68 %s
 ; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=fiji -verify-machineinstrs < %s | FileCheck -check-prefix=GFX68 %s
 ; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx1010 -verify-machineinstrs < %s | FileCheck -check-prefix=GFX10 %s
 ; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx1010 -mattr=-enable-prt-strict-null -verify-machineinstrs < %s | FileCheck -check-prefix=NOPRT %s
+; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx1100 -verify-machineinstrs < %s | FileCheck -check-prefix=GFX10 %s
 
 define amdgpu_ps float @load_1d_f32_x(<8 x i32> inreg %rsrc, i32 %s) {
 ; GFX68-LABEL: load_1d_f32_x:
 ; GFX68:       ; %bb.0:
 ; GFX68-NEXT:    s_mov_b32 s0, s2
 ; GFX68-NEXT:    s_mov_b32 s1, s3
 ; GFX68-NEXT:    s_mov_b32 s2, s4
 ; GFX68-NEXT:    s_mov_b32 s3, s5

llvm/test/CodeGen/AMDGPU/GlobalISel/llvm.amdgcn.image.load.2d.ll

 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
 ; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=tahiti -verify-machineinstrs < %s | FileCheck -check-prefix=GFX6 %s
-; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx1010 -verify-machineinstrs < %s | FileCheck -check-prefix=GFX10 %s
+; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx1010 -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX10PLUS,GFX10 %s
+; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx1100 -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX10PLUS,GFX11 %s
 
 define amdgpu_ps <4 x float> @load_2d_v4f32_xyzw(<8 x i32> inreg %rsrc, i32 %s, i32 %t) {
 ; GFX6-LABEL: load_2d_v4f32_xyzw:
 ; GFX6:       ; %bb.0:
 ; GFX6-NEXT:    s_mov_b32 s0, s2
 ; GFX6-NEXT:    s_mov_b32 s1, s3
 ; GFX6-NEXT:    s_mov_b32 s2, s4
 ; GFX6-NEXT:    s_mov_b32 s3, s5
 ; GFX6-NEXT:    s_mov_b32 s4, s6
 ; GFX6-NEXT:    s_mov_b32 s5, s7
 ; GFX6-NEXT:    s_mov_b32 s6, s8
 ; GFX6-NEXT:    s_mov_b32 s7, s9
 ; GFX6-NEXT:    image_load v[0:3], v[0:1], s[0:7] dmask:0xf unorm
 ; GFX6-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: load_2d_v4f32_xyzw:
-; GFX10:       ; %bb.0:
-; GFX10-NEXT:    s_mov_b32 s0, s2
-; GFX10-NEXT:    s_mov_b32 s1, s3
-; GFX10-NEXT:    s_mov_b32 s2, s4
-; GFX10-NEXT:    s_mov_b32 s3, s5
-; GFX10-NEXT:    s_mov_b32 s4, s6
-; GFX10-NEXT:    s_mov_b32 s5, s7
-; GFX10-NEXT:    s_mov_b32 s6, s8
-; GFX10-NEXT:    s_mov_b32 s7, s9
-; GFX10-NEXT:    image_load v[0:3], v[0:1], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D unorm
-; GFX10-NEXT:    s_waitcnt vmcnt(0)
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: load_2d_v4f32_xyzw:
+; GFX10PLUS:       ; %bb.0:
+; GFX10PLUS-NEXT:    s_mov_b32 s0, s2
+; GFX10PLUS-NEXT:    s_mov_b32 s1, s3
+; GFX10PLUS-NEXT:    s_mov_b32 s2, s4
+; GFX10PLUS-NEXT:    s_mov_b32 s3, s5
+; GFX10PLUS-NEXT:    s_mov_b32 s4, s6
+; GFX10PLUS-NEXT:    s_mov_b32 s5, s7
+; GFX10PLUS-NEXT:    s_mov_b32 s6, s8
+; GFX10PLUS-NEXT:    s_mov_b32 s7, s9
+; GFX10PLUS-NEXT:    image_load v[0:3], v[0:1], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D unorm
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
   %v = call <4 x float> @llvm.amdgcn.image.load.2d.v4f32.i32(i32 15, i32 %s, i32 %t, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @load_2d_v4f32_xyzw_tfe(<8 x i32> inreg %rsrc, i32 addrspace(1)* inreg %out, i32 %s, i32 %t) {
 ; GFX6-LABEL: load_2d_v4f32_xyzw_tfe:
 ; GFX6:       ; %bb.0:
 ; GFX6-NEXT:    v_mov_b32_e32 v5, v0

 ; GFX10-NEXT:    v_mov_b32_e32 v2, v9
 ; GFX10-NEXT:    v_mov_b32_e32 v3, v10
 ; GFX10-NEXT:    v_mov_b32_e32 v4, v11
 ; GFX10-NEXT:    image_load v[0:4], v[5:6], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D unorm tfe
 ; GFX10-NEXT:    s_waitcnt vmcnt(0)
 ; GFX10-NEXT:    global_store_dword v7, v4, s[10:11]
 ; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
 ; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: load_2d_v4f32_xyzw_tfe:
+; GFX11:       ; %bb.0:
+; GFX11-NEXT:    v_mov_b32_e32 v7, 0
+; GFX11-NEXT:    v_mov_b32_e32 v5, v0
+; GFX11-NEXT:    v_mov_b32_e32 v6, v1
+; GFX11-NEXT:    s_mov_b32 s0, s2
+; GFX11-NEXT:    s_mov_b32 s1, s3
+; GFX11-NEXT:    v_mov_b32_e32 v8, v7
+; GFX11-NEXT:    v_mov_b32_e32 v9, v7
+; GFX11-NEXT:    v_mov_b32_e32 v10, v7
+; GFX11-NEXT:    v_mov_b32_e32 v11, v7
+; GFX11-NEXT:    s_mov_b32 s2, s4
+; GFX11-NEXT:    s_mov_b32 s3, s5
+; GFX11-NEXT:    s_mov_b32 s4, s6
+; GFX11-NEXT:    s_mov_b32 s5, s7
+; GFX11-NEXT:    s_mov_b32 s6, s8
+; GFX11-NEXT:    s_mov_b32 s7, s9
+; GFX11-NEXT:    v_mov_b32_e32 v0, v7
+; GFX11-NEXT:    v_mov_b32_e32 v1, v8
+; GFX11-NEXT:    v_mov_b32_e32 v2, v9
+; GFX11-NEXT:    v_mov_b32_e32 v3, v10
+; GFX11-NEXT:    v_mov_b32_e32 v4, v11
+; GFX11-NEXT:    image_load v[0:4], v[5:6], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D unorm tfe
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    global_store_b32 v7, v4, s[10:11]
+; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX11-NEXT:    ; return to shader part epilog
   %v = call { <4 x float>, i32 } @llvm.amdgcn.image.load.2d.sl_v4f32i32s.i32(i32 15, i32 %s, i32 %t, <8 x i32> %rsrc, i32 1, i32 0)
   %v.vec = extractvalue { <4 x float>, i32 } %v, 0
   %v.err = extractvalue { <4 x float>, i32 } %v, 1
   store i32 %v.err, i32 addrspace(1)* %out, align 4
   ret <4 x float> %v.vec
 }
 
 define amdgpu_ps <4 x float> @load_2d_v4f32_xyzw_tfe_lwe(<8 x i32> inreg %rsrc, i32 addrspace(1)* inreg %out, i32 %s, i32 %t) {

 ; GFX10-NEXT:    v_mov_b32_e32 v2, v9
 ; GFX10-NEXT:    v_mov_b32_e32 v3, v10
 ; GFX10-NEXT:    v_mov_b32_e32 v4, v11
 ; GFX10-NEXT:    image_load v[0:4], v[5:6], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D unorm tfe lwe
 ; GFX10-NEXT:    s_waitcnt vmcnt(0)
 ; GFX10-NEXT:    global_store_dword v7, v4, s[10:11]
 ; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
 ; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: load_2d_v4f32_xyzw_tfe_lwe:
+; GFX11:       ; %bb.0:
+; GFX11-NEXT:    v_mov_b32_e32 v7, 0
+; GFX11-NEXT:    v_mov_b32_e32 v5, v0
+; GFX11-NEXT:    v_mov_b32_e32 v6, v1
+; GFX11-NEXT:    s_mov_b32 s0, s2
+; GFX11-NEXT:    s_mov_b32 s1, s3
+; GFX11-NEXT:    v_mov_b32_e32 v8, v7
+; GFX11-NEXT:    v_mov_b32_e32 v9, v7
+; GFX11-NEXT:    v_mov_b32_e32 v10, v7
+; GFX11-NEXT:    v_mov_b32_e32 v11, v7
+; GFX11-NEXT:    s_mov_b32 s2, s4
+; GFX11-NEXT:    s_mov_b32 s3, s5
+; GFX11-NEXT:    s_mov_b32 s4, s6
+; GFX11-NEXT:    s_mov_b32 s5, s7
+; GFX11-NEXT:    s_mov_b32 s6, s8
+; GFX11-NEXT:    s_mov_b32 s7, s9
+; GFX11-NEXT:    v_mov_b32_e32 v0, v7
+; GFX11-NEXT:    v_mov_b32_e32 v1, v8
+; GFX11-NEXT:    v_mov_b32_e32 v2, v9
+; GFX11-NEXT:    v_mov_b32_e32 v3, v10
+; GFX11-NEXT:    v_mov_b32_e32 v4, v11
+; GFX11-NEXT:    image_load v[0:4], v[5:6], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D unorm tfe lwe
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    global_store_b32 v7, v4, s[10:11]
+; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX11-NEXT:    ; return to shader part epilog
   %v = call { <4 x float>, i32 } @llvm.amdgcn.image.load.2d.sl_v4f32i32s.i32(i32 15, i32 %s, i32 %t, <8 x i32> %rsrc, i32 3, i32 0)
   %v.vec = extractvalue { <4 x float>, i32 } %v, 0
   %v.err = extractvalue { <4 x float>, i32 } %v, 1
   store i32 %v.err, i32 addrspace(1)* %out, align 4
   ret <4 x float> %v.vec
 }
 
 declare <4 x float> @llvm.amdgcn.image.load.2d.v4f32.i32(i32 immarg, i32, i32, <8 x i32>, i32 immarg, i32 immarg) #0
 declare { <4 x float>, i32 } @llvm.amdgcn.image.load.2d.sl_v4f32i32s.i32(i32 immarg, i32, i32, <8 x i32>, i32 immarg, i32 immarg) #0
 
 attributes #0 = { nounwind readonly }

llvm/test/CodeGen/AMDGPU/GlobalISel/llvm.amdgcn.image.load.2darraymsaa.a16.ll

 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
 ; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx900 -verify-machineinstrs < %s | FileCheck -check-prefix=GFX9 %s
-; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx1010 -verify-machineinstrs < %s | FileCheck -check-prefix=GFX10 %s
+; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx1010 -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX10PLUS,GFX10 %s
+; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx1100 -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX10PLUS,GFX11 %s
 
 define amdgpu_ps <4 x float> @load_2darraymsaa_v4f32_xyzw(<8 x i32> inreg %rsrc, i16 %s, i16 %t, i16 %slice, i16 %fragid) {
 ; GFX9-LABEL: load_2darraymsaa_v4f32_xyzw:
 ; GFX9:       ; %bb.0:
 ; GFX9-NEXT:    v_mov_b32_e32 v4, 0xffff
 ; GFX9-NEXT:    v_lshlrev_b32_e32 v1, 16, v1
 ; GFX9-NEXT:    v_and_or_b32 v0, v0, v4, v1
 ; GFX9-NEXT:    v_lshlrev_b32_e32 v1, 16, v3
 ; GFX9-NEXT:    s_mov_b32 s0, s2
 ; GFX9-NEXT:    s_mov_b32 s1, s3
 ; GFX9-NEXT:    s_mov_b32 s2, s4
 ; GFX9-NEXT:    s_mov_b32 s3, s5
 ; GFX9-NEXT:    s_mov_b32 s4, s6
 ; GFX9-NEXT:    s_mov_b32 s5, s7
 ; GFX9-NEXT:    s_mov_b32 s6, s8
 ; GFX9-NEXT:    s_mov_b32 s7, s9
 ; GFX9-NEXT:    v_and_or_b32 v1, v2, v4, v1
 ; GFX9-NEXT:    image_load v[0:3], v[0:1], s[0:7] dmask:0xf unorm a16 da
 ; GFX9-NEXT:    s_waitcnt vmcnt(0)
 ; GFX9-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: load_2darraymsaa_v4f32_xyzw:
-; GFX10:       ; %bb.0:
-; GFX10-NEXT:    v_lshlrev_b32_e32 v1, 16, v1
-; GFX10-NEXT:    v_lshlrev_b32_e32 v3, 16, v3
-; GFX10-NEXT:    s_mov_b32 s0, s2
-; GFX10-NEXT:    s_mov_b32 s1, s3
-; GFX10-NEXT:    s_mov_b32 s2, s4
-; GFX10-NEXT:    v_and_or_b32 v0, 0xffff, v0, v1
-; GFX10-NEXT:    v_and_or_b32 v1, 0xffff, v2, v3
-; GFX10-NEXT:    s_mov_b32 s3, s5
-; GFX10-NEXT:    s_mov_b32 s4, s6
-; GFX10-NEXT:    s_mov_b32 s5, s7
-; GFX10-NEXT:    s_mov_b32 s6, s8
-; GFX10-NEXT:    s_mov_b32 s7, s9
-; GFX10-NEXT:    image_load v[0:3], v[0:1], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_MSAA_ARRAY unorm a16
-; GFX10-NEXT:    s_waitcnt vmcnt(0)
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: load_2darraymsaa_v4f32_xyzw:
+; GFX10PLUS:       ; %bb.0:
+; GFX10PLUS-NEXT:    v_lshlrev_b32_e32 v1, 16, v1
+; GFX10PLUS-NEXT:    v_lshlrev_b32_e32 v3, 16, v3
+; GFX10PLUS-NEXT:    s_mov_b32 s0, s2
+; GFX10PLUS-NEXT:    s_mov_b32 s1, s3
+; GFX10PLUS-NEXT:    s_mov_b32 s2, s4
+; GFX10PLUS-NEXT:    v_and_or_b32 v0, 0xffff, v0, v1
+; GFX10PLUS-NEXT:    v_and_or_b32 v1, 0xffff, v2, v3
+; GFX10PLUS-NEXT:    s_mov_b32 s3, s5
+; GFX10PLUS-NEXT:    s_mov_b32 s4, s6
+; GFX10PLUS-NEXT:    s_mov_b32 s5, s7
+; GFX10PLUS-NEXT:    s_mov_b32 s6, s8
+; GFX10PLUS-NEXT:    s_mov_b32 s7, s9
+; GFX10PLUS-NEXT:    image_load v[0:3], v[0:1], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_MSAA_ARRAY unorm a16
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
   %v = call <4 x float> @llvm.amdgcn.image.load.2darraymsaa.v4f32.i16(i32 15, i16 %s, i16 %t, i16 %slice, i16 %fragid, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @load_2darraymsaa_v4f32_xyzw_tfe(<8 x i32> inreg %rsrc, i32 addrspace(1)* inreg %out, i16 %s, i16 %t, i16 %slice, i16 %fragid) {
 ; GFX9-LABEL: load_2darraymsaa_v4f32_xyzw_tfe:
 ; GFX9:       ; %bb.0:
 ; GFX9-NEXT:    v_mov_b32_e32 v4, 0xffff

 ; GFX10-NEXT:    v_mov_b32_e32 v2, v7
 ; GFX10-NEXT:    v_mov_b32_e32 v3, v8
 ; GFX10-NEXT:    v_mov_b32_e32 v4, v9
 ; GFX10-NEXT:    image_load v[0:4], v[10:11], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_MSAA_ARRAY unorm a16 tfe
 ; GFX10-NEXT:    s_waitcnt vmcnt(0)
 ; GFX10-NEXT:    global_store_dword v5, v4, s[10:11]
 ; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
 ; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: load_2darraymsaa_v4f32_xyzw_tfe:
+; GFX11:       ; %bb.0:
+; GFX11-NEXT:    v_mov_b32_e32 v5, 0
+; GFX11-NEXT:    v_lshlrev_b32_e32 v1, 16, v1
+; GFX11-NEXT:    v_lshlrev_b32_e32 v3, 16, v3
+; GFX11-NEXT:    s_mov_b32 s0, s2
+; GFX11-NEXT:    s_mov_b32 s1, s3
+; GFX11-NEXT:    v_mov_b32_e32 v6, v5
+; GFX11-NEXT:    v_mov_b32_e32 v7, v5
+; GFX11-NEXT:    v_mov_b32_e32 v8, v5
+; GFX11-NEXT:    v_mov_b32_e32 v9, v5
+; GFX11-NEXT:    v_and_or_b32 v10, 0xffff, v0, v1
+; GFX11-NEXT:    v_and_or_b32 v11, 0xffff, v2, v3
+; GFX11-NEXT:    s_mov_b32 s2, s4
+; GFX11-NEXT:    s_mov_b32 s3, s5
+; GFX11-NEXT:    s_mov_b32 s4, s6
+; GFX11-NEXT:    s_mov_b32 s5, s7
+; GFX11-NEXT:    s_mov_b32 s6, s8
+; GFX11-NEXT:    s_mov_b32 s7, s9
+; GFX11-NEXT:    v_mov_b32_e32 v0, v5
+; GFX11-NEXT:    v_mov_b32_e32 v1, v6
+; GFX11-NEXT:    v_mov_b32_e32 v2, v7
+; GFX11-NEXT:    v_mov_b32_e32 v3, v8
+; GFX11-NEXT:    v_mov_b32_e32 v4, v9
+; GFX11-NEXT:    image_load v[0:4], v[10:11], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_MSAA_ARRAY unorm a16 tfe
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    global_store_b32 v5, v4, s[10:11]
+; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX11-NEXT:    ; return to shader part epilog
   %v = call { <4 x float>, i32 } @llvm.amdgcn.image.load.2darraymsaa.sl_v4f32i32s.i16(i32 15, i16 %s, i16 %t, i16 %slice, i16 %fragid, <8 x i32> %rsrc, i32 1, i32 0)
   %v.vec = extractvalue { <4 x float>, i32 } %v, 0
   %v.err = extractvalue { <4 x float>, i32 } %v, 1
   store i32 %v.err, i32 addrspace(1)* %out, align 4
   ret <4 x float> %v.vec
 }
 
 define amdgpu_ps <4 x float> @load_2darraymsaa_v4f32_xyzw_tfe_lwe(<8 x i32> inreg %rsrc, i32 addrspace(1)* inreg %out, i16 %s, i16 %t, i16 %slice, i16 %fragid) {

 ; GFX10-NEXT:    v_mov_b32_e32 v2, v7
 ; GFX10-NEXT:    v_mov_b32_e32 v3, v8
 ; GFX10-NEXT:    v_mov_b32_e32 v4, v9
 ; GFX10-NEXT:    image_load v[0:4], v[10:11], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_MSAA_ARRAY unorm a16 tfe lwe
 ; GFX10-NEXT:    s_waitcnt vmcnt(0)
 ; GFX10-NEXT:    global_store_dword v5, v4, s[10:11]
 ; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
 ; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: load_2darraymsaa_v4f32_xyzw_tfe_lwe:
+; GFX11:       ; %bb.0:
+; GFX11-NEXT:    v_mov_b32_e32 v5, 0
+; GFX11-NEXT:    v_lshlrev_b32_e32 v1, 16, v1
+; GFX11-NEXT:    v_lshlrev_b32_e32 v3, 16, v3
+; GFX11-NEXT:    s_mov_b32 s0, s2
+; GFX11-NEXT:    s_mov_b32 s1, s3
+; GFX11-NEXT:    v_mov_b32_e32 v6, v5
+; GFX11-NEXT:    v_mov_b32_e32 v7, v5
+; GFX11-NEXT:    v_mov_b32_e32 v8, v5
+; GFX11-NEXT:    v_mov_b32_e32 v9, v5
+; GFX11-NEXT:    v_and_or_b32 v10, 0xffff, v0, v1
+; GFX11-NEXT:    v_and_or_b32 v11, 0xffff, v2, v3
+; GFX11-NEXT:    s_mov_b32 s2, s4
+; GFX11-NEXT:    s_mov_b32 s3, s5
+; GFX11-NEXT:    s_mov_b32 s4, s6
+; GFX11-NEXT:    s_mov_b32 s5, s7
+; GFX11-NEXT:    s_mov_b32 s6, s8
+; GFX11-NEXT:    s_mov_b32 s7, s9
+; GFX11-NEXT:    v_mov_b32_e32 v0, v5
+; GFX11-NEXT:    v_mov_b32_e32 v1, v6
+; GFX11-NEXT:    v_mov_b32_e32 v2, v7
+; GFX11-NEXT:    v_mov_b32_e32 v3, v8
+; GFX11-NEXT:    v_mov_b32_e32 v4, v9
+; GFX11-NEXT:    image_load v[0:4], v[10:11], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_MSAA_ARRAY unorm a16 tfe lwe
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    global_store_b32 v5, v4, s[10:11]
+; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX11-NEXT:    ; return to shader part epilog
   %v = call { <4 x float>, i32 } @llvm.amdgcn.image.load.2darraymsaa.sl_v4f32i32s.i16(i32 15, i16 %s, i16 %t, i16 %slice, i16 %fragid, <8 x i32> %rsrc, i32 3, i32 0)
   %v.vec = extractvalue { <4 x float>, i32 } %v, 0
   %v.err = extractvalue { <4 x float>, i32 } %v, 1
   store i32 %v.err, i32 addrspace(1)* %out, align 4
   ret <4 x float> %v.vec
 }
 
 declare <4 x float> @llvm.amdgcn.image.load.2darraymsaa.v4f32.i16(i32 immarg, i16, i16, i16, i16, <8 x i32>, i32 immarg, i32 immarg) #0
 declare { <4 x float>, i32 } @llvm.amdgcn.image.load.2darraymsaa.sl_v4f32i32s.i16(i32 immarg, i16, i16, i16, i16, <8 x i32>, i32 immarg, i32 immarg) #0
 
 attributes #0 = { nounwind readonly }

llvm/test/CodeGen/AMDGPU/GlobalISel/llvm.amdgcn.image.load.2darraymsaa.ll

 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
 ; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=tahiti -verify-machineinstrs < %s | FileCheck -check-prefix=GFX6 %s
-; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx1010 -verify-machineinstrs < %s | FileCheck -check-prefix=GFX10 %s
+; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx1010 -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX10PLUS,GFX10 %s
+; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx1100 -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX10PLUS,GFX11 %s
 
 define amdgpu_ps <4 x float> @load_2darraymsaa_v4f32_xyzw(<8 x i32> inreg %rsrc, i32 %s, i32 %t, i32 %slice, i32 %fragid) {
 ; GFX6-LABEL: load_2darraymsaa_v4f32_xyzw:
 ; GFX6:       ; %bb.0:
 ; GFX6-NEXT:    s_mov_b32 s0, s2
 ; GFX6-NEXT:    s_mov_b32 s1, s3
 ; GFX6-NEXT:    s_mov_b32 s2, s4
 ; GFX6-NEXT:    s_mov_b32 s3, s5
 ; GFX6-NEXT:    s_mov_b32 s4, s6
 ; GFX6-NEXT:    s_mov_b32 s5, s7
 ; GFX6-NEXT:    s_mov_b32 s6, s8
 ; GFX6-NEXT:    s_mov_b32 s7, s9
 ; GFX6-NEXT:    image_load v[0:3], v[0:3], s[0:7] dmask:0xf unorm da
 ; GFX6-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: load_2darraymsaa_v4f32_xyzw:
-; GFX10:       ; %bb.0:
-; GFX10-NEXT:    s_mov_b32 s0, s2
-; GFX10-NEXT:    s_mov_b32 s1, s3
-; GFX10-NEXT:    s_mov_b32 s2, s4
-; GFX10-NEXT:    s_mov_b32 s3, s5
-; GFX10-NEXT:    s_mov_b32 s4, s6
-; GFX10-NEXT:    s_mov_b32 s5, s7
-; GFX10-NEXT:    s_mov_b32 s6, s8
-; GFX10-NEXT:    s_mov_b32 s7, s9
-; GFX10-NEXT:    image_load v[0:3], v[0:3], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_MSAA_ARRAY unorm
-; GFX10-NEXT:    s_waitcnt vmcnt(0)
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: load_2darraymsaa_v4f32_xyzw:
+; GFX10PLUS:       ; %bb.0:
+; GFX10PLUS-NEXT:    s_mov_b32 s0, s2
+; GFX10PLUS-NEXT:    s_mov_b32 s1, s3
+; GFX10PLUS-NEXT:    s_mov_b32 s2, s4
+; GFX10PLUS-NEXT:    s_mov_b32 s3, s5
+; GFX10PLUS-NEXT:    s_mov_b32 s4, s6
+; GFX10PLUS-NEXT:    s_mov_b32 s5, s7
+; GFX10PLUS-NEXT:    s_mov_b32 s6, s8
+; GFX10PLUS-NEXT:    s_mov_b32 s7, s9
+; GFX10PLUS-NEXT:    image_load v[0:3], v[0:3], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_MSAA_ARRAY unorm
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
   %v = call <4 x float> @llvm.amdgcn.image.load.2darraymsaa.v4f32.i32(i32 15, i32 %s, i32 %t, i32 %slice, i32 %fragid, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @load_2darraymsaa_v4f32_xyzw_tfe(<8 x i32> inreg %rsrc, i32 addrspace(1)* inreg %out, i32 %s, i32 %t, i32 %slice, i32 %fragid) {
 ; GFX6-LABEL: load_2darraymsaa_v4f32_xyzw_tfe:
 ; GFX6:       ; %bb.0:
 ; GFX6-NEXT:    v_mov_b32_e32 v5, v0

 ; GFX10-NEXT:    v_mov_b32_e32 v2, v11
 ; GFX10-NEXT:    v_mov_b32_e32 v3, v12
 ; GFX10-NEXT:    v_mov_b32_e32 v4, v13
 ; GFX10-NEXT:    image_load v[0:4], v[5:8], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_MSAA_ARRAY unorm tfe
 ; GFX10-NEXT:    s_waitcnt vmcnt(0)
 ; GFX10-NEXT:    global_store_dword v9, v4, s[10:11]
 ; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
 ; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: load_2darraymsaa_v4f32_xyzw_tfe:
+; GFX11:       ; %bb.0:
+; GFX11-NEXT:    v_mov_b32_e32 v9, 0
+; GFX11-NEXT:    v_mov_b32_e32 v5, v0
+; GFX11-NEXT:    v_mov_b32_e32 v6, v1
+; GFX11-NEXT:    v_mov_b32_e32 v7, v2
+; GFX11-NEXT:    v_mov_b32_e32 v8, v3
+; GFX11-NEXT:    v_mov_b32_e32 v10, v9
+; GFX11-NEXT:    v_mov_b32_e32 v11, v9
+; GFX11-NEXT:    v_mov_b32_e32 v12, v9
+; GFX11-NEXT:    v_mov_b32_e32 v13, v9
+; GFX11-NEXT:    s_mov_b32 s0, s2
+; GFX11-NEXT:    s_mov_b32 s1, s3
+; GFX11-NEXT:    s_mov_b32 s2, s4
+; GFX11-NEXT:    s_mov_b32 s3, s5
+; GFX11-NEXT:    s_mov_b32 s4, s6
+; GFX11-NEXT:    s_mov_b32 s5, s7
+; GFX11-NEXT:    s_mov_b32 s6, s8
+; GFX11-NEXT:    s_mov_b32 s7, s9
+; GFX11-NEXT:    v_mov_b32_e32 v0, v9
+; GFX11-NEXT:    v_mov_b32_e32 v1, v10
+; GFX11-NEXT:    v_mov_b32_e32 v2, v11
+; GFX11-NEXT:    v_mov_b32_e32 v3, v12
+; GFX11-NEXT:    v_mov_b32_e32 v4, v13
+; GFX11-NEXT:    image_load v[0:4], v[5:8], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_MSAA_ARRAY unorm tfe
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    global_store_b32 v9, v4, s[10:11]
+; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX11-NEXT:    ; return to shader part epilog
   %v = call { <4 x float>, i32 } @llvm.amdgcn.image.load.2darraymsaa.sl_v4f32i32s.i32(i32 15, i32 %s, i32 %t, i32 %slice, i32 %fragid, <8 x i32> %rsrc, i32 1, i32 0)
   %v.vec = extractvalue { <4 x float>, i32 } %v, 0
   %v.err = extractvalue { <4 x float>, i32 } %v, 1
   store i32 %v.err, i32 addrspace(1)* %out, align 4
   ret <4 x float> %v.vec
 }
 
 define amdgpu_ps <4 x float> @load_2darraymsaa_v4f32_xyzw_tfe_lwe(<8 x i32> inreg %rsrc, i32 addrspace(1)* inreg %out, i32 %s, i32 %t, i32 %slice, i32 %fragid) {

 ; GFX10-NEXT:    v_mov_b32_e32 v2, v11
 ; GFX10-NEXT:    v_mov_b32_e32 v3, v12
 ; GFX10-NEXT:    v_mov_b32_e32 v4, v13
 ; GFX10-NEXT:    image_load v[0:4], v[5:8], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_MSAA_ARRAY unorm tfe lwe
 ; GFX10-NEXT:    s_waitcnt vmcnt(0)
 ; GFX10-NEXT:    global_store_dword v9, v4, s[10:11]
 ; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
 ; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: load_2darraymsaa_v4f32_xyzw_tfe_lwe:
+; GFX11:       ; %bb.0:
+; GFX11-NEXT:    v_mov_b32_e32 v9, 0
+; GFX11-NEXT:    v_mov_b32_e32 v5, v0
+; GFX11-NEXT:    v_mov_b32_e32 v6, v1
+; GFX11-NEXT:    v_mov_b32_e32 v7, v2
+; GFX11-NEXT:    v_mov_b32_e32 v8, v3
+; GFX11-NEXT:    v_mov_b32_e32 v10, v9
+; GFX11-NEXT:    v_mov_b32_e32 v11, v9
+; GFX11-NEXT:    v_mov_b32_e32 v12, v9
+; GFX11-NEXT:    v_mov_b32_e32 v13, v9
+; GFX11-NEXT:    s_mov_b32 s0, s2
+; GFX11-NEXT:    s_mov_b32 s1, s3
+; GFX11-NEXT:    s_mov_b32 s2, s4
+; GFX11-NEXT:    s_mov_b32 s3, s5
+; GFX11-NEXT:    s_mov_b32 s4, s6
+; GFX11-NEXT:    s_mov_b32 s5, s7
+; GFX11-NEXT:    s_mov_b32 s6, s8
+; GFX11-NEXT:    s_mov_b32 s7, s9
+; GFX11-NEXT:    v_mov_b32_e32 v0, v9
+; GFX11-NEXT:    v_mov_b32_e32 v1, v10
+; GFX11-NEXT:    v_mov_b32_e32 v2, v11
+; GFX11-NEXT:    v_mov_b32_e32 v3, v12
+; GFX11-NEXT:    v_mov_b32_e32 v4, v13
+; GFX11-NEXT:    image_load v[0:4], v[5:8], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_MSAA_ARRAY unorm tfe lwe
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    global_store_b32 v9, v4, s[10:11]
+; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX11-NEXT:    ; return to shader part epilog
   %v = call { <4 x float>, i32 } @llvm.amdgcn.image.load.2darraymsaa.sl_v4f32i32s.i32(i32 15, i32 %s, i32 %t, i32 %slice, i32 %fragid, <8 x i32> %rsrc, i32 3, i32 0)
   %v.vec = extractvalue { <4 x float>, i32 } %v, 0
   %v.err = extractvalue { <4 x float>, i32 } %v, 1
   store i32 %v.err, i32 addrspace(1)* %out, align 4
   ret <4 x float> %v.vec
 }
 
 declare <4 x float> @llvm.amdgcn.image.load.2darraymsaa.v4f32.i32(i32 immarg, i32, i32, i32, i32, <8 x i32>, i32 immarg, i32 immarg) #0
 declare { <4 x float>, i32 } @llvm.amdgcn.image.load.2darraymsaa.sl_v4f32i32s.i32(i32 immarg, i32, i32, i32, i32, <8 x i32>, i32 immarg, i32 immarg) #0
 
 attributes #0 = { nounwind readonly }

llvm/test/CodeGen/AMDGPU/GlobalISel/llvm.amdgcn.image.load.3d.a16.ll

 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
 ; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx900 -verify-machineinstrs < %s | FileCheck -check-prefix=GFX9 %s
-; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx1010 -verify-machineinstrs < %s | FileCheck -check-prefix=GFX10 %s
+; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx1010 -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX10PLUS,GFX10 %s
+; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx1100 -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX10PLUS,GFX11 %s
 
 define amdgpu_ps <4 x float> @load_3d_v4f32_xyzw(<8 x i32> inreg %rsrc, i16 %s, i16 %t, i16 %r) {
 ; GFX9-LABEL: load_3d_v4f32_xyzw:
 ; GFX9:       ; %bb.0:
 ; GFX9-NEXT:    s_mov_b32 s0, s2
 ; GFX9-NEXT:    s_mov_b32 s2, s4
 ; GFX9-NEXT:    s_mov_b32 s4, s6
 ; GFX9-NEXT:    s_mov_b32 s6, s8
 ; GFX9-NEXT:    v_mov_b32_e32 v3, 0xffff
 ; GFX9-NEXT:    v_lshlrev_b32_e32 v1, 16, v1
 ; GFX9-NEXT:    s_lshl_b32 s8, s0, 16
 ; GFX9-NEXT:    s_mov_b32 s1, s3
 ; GFX9-NEXT:    s_mov_b32 s3, s5
 ; GFX9-NEXT:    s_mov_b32 s5, s7
 ; GFX9-NEXT:    s_mov_b32 s7, s9
 ; GFX9-NEXT:    v_and_or_b32 v0, v0, v3, v1
 ; GFX9-NEXT:    v_and_or_b32 v1, v2, v3, s8
 ; GFX9-NEXT:    image_load v[0:3], v[0:1], s[0:7] dmask:0xf unorm a16
 ; GFX9-NEXT:    s_waitcnt vmcnt(0)
 ; GFX9-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: load_3d_v4f32_xyzw:
-; GFX10:       ; %bb.0:
-; GFX10-NEXT:    v_lshlrev_b32_e32 v3, 16, v1
-; GFX10-NEXT:    s_mov_b32 s0, s2
-; GFX10-NEXT:    s_mov_b32 s2, s4
-; GFX10-NEXT:    s_mov_b32 s4, s6
-; GFX10-NEXT:    s_mov_b32 s6, s8
-; GFX10-NEXT:    s_lshl_b32 s8, s0, 16
-; GFX10-NEXT:    v_and_or_b32 v0, 0xffff, v0, v3
-; GFX10-NEXT:    v_and_or_b32 v1, 0xffff, v2, s8
-; GFX10-NEXT:    s_mov_b32 s1, s3
-; GFX10-NEXT:    s_mov_b32 s3, s5
-; GFX10-NEXT:    s_mov_b32 s5, s7
-; GFX10-NEXT:    s_mov_b32 s7, s9
-; GFX10-NEXT:    image_load v[0:3], v[0:1], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_3D unorm a16
-; GFX10-NEXT:    s_waitcnt vmcnt(0)
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: load_3d_v4f32_xyzw:
+; GFX10PLUS:       ; %bb.0:
+; GFX10PLUS-NEXT:    v_lshlrev_b32_e32 v3, 16, v1
+; GFX10PLUS-NEXT:    s_mov_b32 s0, s2
+; GFX10PLUS-NEXT:    s_mov_b32 s2, s4
+; GFX10PLUS-NEXT:    s_mov_b32 s4, s6
+; GFX10PLUS-NEXT:    s_mov_b32 s6, s8
+; GFX10PLUS-NEXT:    s_lshl_b32 s8, s0, 16
+; GFX10PLUS-NEXT:    v_and_or_b32 v0, 0xffff, v0, v3
+; GFX10PLUS-NEXT:    v_and_or_b32 v1, 0xffff, v2, s8
+; GFX10PLUS-NEXT:    s_mov_b32 s1, s3
+; GFX10PLUS-NEXT:    s_mov_b32 s3, s5
+; GFX10PLUS-NEXT:    s_mov_b32 s5, s7
+; GFX10PLUS-NEXT:    s_mov_b32 s7, s9
+; GFX10PLUS-NEXT:    image_load v[0:3], v[0:1], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_3D unorm a16
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
   %v = call <4 x float> @llvm.amdgcn.image.load.3d.v4f32.i16(i32 15, i16 %s, i16 %t, i16 %r, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @load_3d_v4f32_xyzw_tfe(<8 x i32> inreg %rsrc, i32 addrspace(1)* inreg %out, i16 %s, i16 %t, i16 %r) {
 ; GFX9-LABEL: load_3d_v4f32_xyzw_tfe:
 ; GFX9:       ; %bb.0:
 ; GFX9-NEXT:    s_mov_b32 s0, s2

 ; GFX10-NEXT:    v_mov_b32_e32 v2, v7
 ; GFX10-NEXT:    v_mov_b32_e32 v3, v8
 ; GFX10-NEXT:    v_mov_b32_e32 v4, v9
 ; GFX10-NEXT:    image_load v[0:4], v[10:11], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_3D unorm a16 tfe
 ; GFX10-NEXT:    s_waitcnt vmcnt(0)
 ; GFX10-NEXT:    global_store_dword v5, v4, s[10:11]
 ; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
 ; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: load_3d_v4f32_xyzw_tfe:
+; GFX11:       ; %bb.0:
+; GFX11-NEXT:    v_mov_b32_e32 v5, 0
+; GFX11-NEXT:    v_lshlrev_b32_e32 v1, 16, v1
+; GFX11-NEXT:    s_mov_b32 s0, s2
+; GFX11-NEXT:    s_mov_b32 s2, s4
+; GFX11-NEXT:    s_mov_b32 s4, s6
+; GFX11-NEXT:    s_mov_b32 s6, s8
+; GFX11-NEXT:    s_lshl_b32 s8, s0, 16
+; GFX11-NEXT:    v_mov_b32_e32 v6, v5
+; GFX11-NEXT:    v_mov_b32_e32 v7, v5
+; GFX11-NEXT:    v_mov_b32_e32 v8, v5
+; GFX11-NEXT:    v_mov_b32_e32 v9, v5
+; GFX11-NEXT:    v_and_or_b32 v10, 0xffff, v0, v1
+; GFX11-NEXT:    v_and_or_b32 v11, 0xffff, v2, s8
+; GFX11-NEXT:    s_mov_b32 s1, s3
+; GFX11-NEXT:    s_mov_b32 s3, s5
+; GFX11-NEXT:    s_mov_b32 s5, s7
+; GFX11-NEXT:    s_mov_b32 s7, s9
+; GFX11-NEXT:    v_mov_b32_e32 v0, v5
+; GFX11-NEXT:    v_mov_b32_e32 v1, v6
+; GFX11-NEXT:    v_mov_b32_e32 v2, v7
+; GFX11-NEXT:    v_mov_b32_e32 v3, v8
+; GFX11-NEXT:    v_mov_b32_e32 v4, v9
+; GFX11-NEXT:    image_load v[0:4], v[10:11], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_3D unorm a16 tfe
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    global_store_b32 v5, v4, s[10:11]
+; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX11-NEXT:    ; return to shader part epilog
   %v = call { <4 x float>, i32 } @llvm.amdgcn.image.load.3d.sl_v4f32i32s.i16(i32 15, i16 %s, i16 %t, i16 %r, <8 x i32> %rsrc, i32 1, i32 0)
   %v.vec = extractvalue { <4 x float>, i32 } %v, 0
   %v.err = extractvalue { <4 x float>, i32 } %v, 1
   store i32 %v.err, i32 addrspace(1)* %out, align 4
   ret <4 x float> %v.vec
 }
 
 define amdgpu_ps <4 x float> @load_3d_v4f32_xyzw_tfe_lwe(<8 x i32> inreg %rsrc, i32 addrspace(1)* inreg %out, i16 %s, i16 %t, i16 %r) {

 ; GFX10-NEXT:    v_mov_b32_e32 v2, v7
 ; GFX10-NEXT:    v_mov_b32_e32 v3, v8
 ; GFX10-NEXT:    v_mov_b32_e32 v4, v9
 ; GFX10-NEXT:    image_load v[0:4], v[10:11], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_3D unorm a16 tfe lwe
 ; GFX10-NEXT:    s_waitcnt vmcnt(0)
 ; GFX10-NEXT:    global_store_dword v5, v4, s[10:11]
 ; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
 ; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: load_3d_v4f32_xyzw_tfe_lwe:
+; GFX11:       ; %bb.0:
+; GFX11-NEXT:    v_mov_b32_e32 v5, 0
+; GFX11-NEXT:    v_lshlrev_b32_e32 v1, 16, v1
+; GFX11-NEXT:    s_mov_b32 s0, s2
+; GFX11-NEXT:    s_mov_b32 s2, s4
+; GFX11-NEXT:    s_mov_b32 s4, s6
+; GFX11-NEXT:    s_mov_b32 s6, s8
+; GFX11-NEXT:    s_lshl_b32 s8, s0, 16
+; GFX11-NEXT:    v_mov_b32_e32 v6, v5
+; GFX11-NEXT:    v_mov_b32_e32 v7, v5
+; GFX11-NEXT:    v_mov_b32_e32 v8, v5
+; GFX11-NEXT:    v_mov_b32_e32 v9, v5
+; GFX11-NEXT:    v_and_or_b32 v10, 0xffff, v0, v1
+; GFX11-NEXT:    v_and_or_b32 v11, 0xffff, v2, s8
+; GFX11-NEXT:    s_mov_b32 s1, s3
+; GFX11-NEXT:    s_mov_b32 s3, s5
+; GFX11-NEXT:    s_mov_b32 s5, s7
+; GFX11-NEXT:    s_mov_b32 s7, s9
+; GFX11-NEXT:    v_mov_b32_e32 v0, v5
+; GFX11-NEXT:    v_mov_b32_e32 v1, v6
+; GFX11-NEXT:    v_mov_b32_e32 v2, v7
+; GFX11-NEXT:    v_mov_b32_e32 v3, v8
+; GFX11-NEXT:    v_mov_b32_e32 v4, v9
+; GFX11-NEXT:    image_load v[0:4], v[10:11], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_3D unorm a16 tfe lwe
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    global_store_b32 v5, v4, s[10:11]
+; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX11-NEXT:    ; return to shader part epilog
   %v = call { <4 x float>, i32 } @llvm.amdgcn.image.load.3d.sl_v4f32i32s.i16(i32 15, i16 %s, i16 %t, i16 %r, <8 x i32> %rsrc, i32 3, i32 0)
   %v.vec = extractvalue { <4 x float>, i32 } %v, 0
   %v.err = extractvalue { <4 x float>, i32 } %v, 1
   store i32 %v.err, i32 addrspace(1)* %out, align 4
   ret <4 x float> %v.vec
 }
 
 declare <4 x float> @llvm.amdgcn.image.load.3d.v4f32.i16(i32 immarg, i16, i16, i16, <8 x i32>, i32 immarg, i32 immarg) #0
 declare { <4 x float>, i32 } @llvm.amdgcn.image.load.3d.sl_v4f32i32s.i16(i32 immarg, i16, i16, i16, <8 x i32>, i32 immarg, i32 immarg) #0
 
 attributes #0 = { nounwind readonly }

llvm/test/CodeGen/AMDGPU/GlobalISel/llvm.amdgcn.image.load.3d.ll

 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
 ; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=tahiti -verify-machineinstrs < %s | FileCheck -check-prefix=GFX6 %s
-; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx1010 -verify-machineinstrs < %s | FileCheck -check-prefix=GFX10 %s
+; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx1010 -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX10PLUS,GFX10 %s
+; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx1100 -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX10PLUS,GFX11 %s
 
 define amdgpu_ps <4 x float> @load_3d_v4f32_xyzw(<8 x i32> inreg %rsrc, i32 %s, i32 %t, i32 %r) {
 ; GFX6-LABEL: load_3d_v4f32_xyzw:
 ; GFX6:       ; %bb.0:
 ; GFX6-NEXT:    s_mov_b32 s0, s2
 ; GFX6-NEXT:    s_mov_b32 s1, s3
 ; GFX6-NEXT:    s_mov_b32 s2, s4
 ; GFX6-NEXT:    s_mov_b32 s3, s5
 ; GFX6-NEXT:    s_mov_b32 s4, s6
 ; GFX6-NEXT:    s_mov_b32 s5, s7
 ; GFX6-NEXT:    s_mov_b32 s6, s8
 ; GFX6-NEXT:    s_mov_b32 s7, s9
 ; GFX6-NEXT:    image_load v[0:3], v[0:2], s[0:7] dmask:0xf unorm
 ; GFX6-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: load_3d_v4f32_xyzw:
-; GFX10:       ; %bb.0:
-; GFX10-NEXT:    s_mov_b32 s0, s2
-; GFX10-NEXT:    s_mov_b32 s1, s3
-; GFX10-NEXT:    s_mov_b32 s2, s4
-; GFX10-NEXT:    s_mov_b32 s3, s5
-; GFX10-NEXT:    s_mov_b32 s4, s6
-; GFX10-NEXT:    s_mov_b32 s5, s7
-; GFX10-NEXT:    s_mov_b32 s6, s8
-; GFX10-NEXT:    s_mov_b32 s7, s9
-; GFX10-NEXT:    image_load v[0:3], v[0:2], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_3D unorm
-; GFX10-NEXT:    s_waitcnt vmcnt(0)
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: load_3d_v4f32_xyzw:
+; GFX10PLUS:       ; %bb.0:
+; GFX10PLUS-NEXT:    s_mov_b32 s0, s2
+; GFX10PLUS-NEXT:    s_mov_b32 s1, s3
+; GFX10PLUS-NEXT:    s_mov_b32 s2, s4
+; GFX10PLUS-NEXT:    s_mov_b32 s3, s5
+; GFX10PLUS-NEXT:    s_mov_b32 s4, s6
+; GFX10PLUS-NEXT:    s_mov_b32 s5, s7
+; GFX10PLUS-NEXT:    s_mov_b32 s6, s8
+; GFX10PLUS-NEXT:    s_mov_b32 s7, s9
+; GFX10PLUS-NEXT:    image_load v[0:3], v[0:2], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_3D unorm
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
   %v = call <4 x float> @llvm.amdgcn.image.load.3d.v4f32.i32(i32 15, i32 %s, i32 %t, i32 %r, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @load_3d_v4f32_xyzw_tfe(<8 x i32> inreg %rsrc, i32 addrspace(1)* inreg %out, i32 %s, i32 %t, i32 %r) {
 ; GFX6-LABEL: load_3d_v4f32_xyzw_tfe:
 ; GFX6:       ; %bb.0:
 ; GFX6-NEXT:    v_mov_b32_e32 v5, v0

 ; GFX10-NEXT:    v_mov_b32_e32 v2, v10
 ; GFX10-NEXT:    v_mov_b32_e32 v3, v11
 ; GFX10-NEXT:    v_mov_b32_e32 v4, v12
 ; GFX10-NEXT:    image_load v[0:4], v[5:7], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_3D unorm tfe
 ; GFX10-NEXT:    s_waitcnt vmcnt(0)
 ; GFX10-NEXT:    global_store_dword v8, v4, s[10:11]
 ; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
 ; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: load_3d_v4f32_xyzw_tfe:
+; GFX11:       ; %bb.0:
+; GFX11-NEXT:    v_mov_b32_e32 v8, 0
+; GFX11-NEXT:    v_mov_b32_e32 v5, v0
+; GFX11-NEXT:    v_mov_b32_e32 v6, v1
+; GFX11-NEXT:    v_mov_b32_e32 v7, v2
+; GFX11-NEXT:    s_mov_b32 s0, s2
+; GFX11-NEXT:    v_mov_b32_e32 v9, v8
+; GFX11-NEXT:    v_mov_b32_e32 v10, v8
+; GFX11-NEXT:    v_mov_b32_e32 v11, v8
+; GFX11-NEXT:    v_mov_b32_e32 v12, v8
+; GFX11-NEXT:    s_mov_b32 s1, s3
+; GFX11-NEXT:    s_mov_b32 s2, s4
+; GFX11-NEXT:    s_mov_b32 s3, s5
+; GFX11-NEXT:    s_mov_b32 s4, s6
+; GFX11-NEXT:    s_mov_b32 s5, s7
+; GFX11-NEXT:    s_mov_b32 s6, s8
+; GFX11-NEXT:    s_mov_b32 s7, s9
+; GFX11-NEXT:    v_mov_b32_e32 v0, v8
+; GFX11-NEXT:    v_mov_b32_e32 v1, v9
+; GFX11-NEXT:    v_mov_b32_e32 v2, v10
+; GFX11-NEXT:    v_mov_b32_e32 v3, v11
+; GFX11-NEXT:    v_mov_b32_e32 v4, v12
+; GFX11-NEXT:    image_load v[0:4], v[5:7], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_3D unorm tfe
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    global_store_b32 v8, v4, s[10:11]
+; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX11-NEXT:    ; return to shader part epilog
   %v = call { <4 x float>, i32 } @llvm.amdgcn.image.load.3d.sl_v4f32i32s.i32(i32 15, i32 %s, i32 %t, i32 %r, <8 x i32> %rsrc, i32 1, i32 0)
   %v.vec = extractvalue { <4 x float>, i32 } %v, 0
   %v.err = extractvalue { <4 x float>, i32 } %v, 1
   store i32 %v.err, i32 addrspace(1)* %out, align 4
   ret <4 x float> %v.vec
 }
 
 define amdgpu_ps <4 x float> @load_3d_v4f32_xyzw_tfe_lwe(<8 x i32> inreg %rsrc, i32 addrspace(1)* inreg %out, i32 %s, i32 %t, i32 %r) {

 ; GFX10-NEXT:    v_mov_b32_e32 v2, v10
 ; GFX10-NEXT:    v_mov_b32_e32 v3, v11
 ; GFX10-NEXT:    v_mov_b32_e32 v4, v12
 ; GFX10-NEXT:    image_load v[0:4], v[5:7], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_3D unorm tfe lwe
 ; GFX10-NEXT:    s_waitcnt vmcnt(0)
 ; GFX10-NEXT:    global_store_dword v8, v4, s[10:11]
 ; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
 ; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: load_3d_v4f32_xyzw_tfe_lwe:
+; GFX11:       ; %bb.0:
+; GFX11-NEXT:    v_mov_b32_e32 v8, 0
+; GFX11-NEXT:    v_mov_b32_e32 v5, v0
+; GFX11-NEXT:    v_mov_b32_e32 v6, v1
+; GFX11-NEXT:    v_mov_b32_e32 v7, v2
+; GFX11-NEXT:    s_mov_b32 s0, s2
+; GFX11-NEXT:    v_mov_b32_e32 v9, v8
+; GFX11-NEXT:    v_mov_b32_e32 v10, v8
+; GFX11-NEXT:    v_mov_b32_e32 v11, v8
+; GFX11-NEXT:    v_mov_b32_e32 v12, v8
+; GFX11-NEXT:    s_mov_b32 s1, s3
+; GFX11-NEXT:    s_mov_b32 s2, s4
+; GFX11-NEXT:    s_mov_b32 s3, s5
+; GFX11-NEXT:    s_mov_b32 s4, s6
+; GFX11-NEXT:    s_mov_b32 s5, s7
+; GFX11-NEXT:    s_mov_b32 s6, s8
+; GFX11-NEXT:    s_mov_b32 s7, s9
+; GFX11-NEXT:    v_mov_b32_e32 v0, v8
+; GFX11-NEXT:    v_mov_b32_e32 v1, v9
+; GFX11-NEXT:    v_mov_b32_e32 v2, v10
+; GFX11-NEXT:    v_mov_b32_e32 v3, v11
+; GFX11-NEXT:    v_mov_b32_e32 v4, v12
+; GFX11-NEXT:    image_load v[0:4], v[5:7], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_3D unorm tfe lwe
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    global_store_b32 v8, v4, s[10:11]
+; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX11-NEXT:    ; return to shader part epilog
   %v = call { <4 x float>, i32 } @llvm.amdgcn.image.load.3d.sl_v4f32i32s.i32(i32 15, i32 %s, i32 %t, i32 %r, <8 x i32> %rsrc, i32 3, i32 0)
   %v.vec = extractvalue { <4 x float>, i32 } %v, 0
   %v.err = extractvalue { <4 x float>, i32 } %v, 1
   store i32 %v.err, i32 addrspace(1)* %out, align 4
   ret <4 x float> %v.vec
 }
 
 declare <4 x float> @llvm.amdgcn.image.load.3d.v4f32.i32(i32 immarg, i32, i32, i32, <8 x i32>, i32 immarg, i32 immarg) #0
 declare { <4 x float>, i32 } @llvm.amdgcn.image.load.3d.sl_v4f32i32s.i32(i32 immarg, i32, i32, i32, <8 x i32>, i32 immarg, i32 immarg) #0
 
 attributes #0 = { nounwind readonly }

llvm/test/CodeGen/AMDGPU/GlobalISel/llvm.amdgcn.image.sample.cd.g16.ll

+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
+; RUN: llc -global-isel -march=amdgcn -mcpu=gfx1010 -verify-machineinstrs < %s | FileCheck -check-prefix=GFX10 %s
+
+define amdgpu_ps <4 x float> @sample_cd_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, half %dsdh, half %dsdv, float %s) {
+; GFX10-LABEL: sample_cd_1d:
+; GFX10:       ; %bb.0: ; %main_body
+; GFX10-NEXT:    s_lshl_b32 s12, s0, 16
+; GFX10-NEXT:    v_and_or_b32 v0, 0xffff, v0, s12
+; GFX10-NEXT:    v_and_or_b32 v1, 0xffff, v1, s12
+; GFX10-NEXT:    image_sample_cd_g16 v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    ; return to shader part epilog
+main_body:
+  %v = call <4 x float> @llvm.amdgcn.image.sample.cd.1d.v4f32.f16.f32(i32 15, half %dsdh, half %dsdv, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
+  ret <4 x float> %v
+}
+
+define amdgpu_ps <4 x float> @sample_cd_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t) {
+; GFX10-LABEL: sample_cd_2d:
+; GFX10:       ; %bb.0: ; %main_body
+; GFX10-NEXT:    v_lshlrev_b32_e32 v1, 16, v1
+; GFX10-NEXT:    v_lshlrev_b32_e32 v3, 16, v3
+; GFX10-NEXT:    v_and_or_b32 v0, 0xffff, v0, v1
+; GFX10-NEXT:    v_and_or_b32 v1, 0xffff, v2, v3
+; GFX10-NEXT:    image_sample_cd_g16 v[0:3], [v0, v1, v4, v5], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    ; return to shader part epilog
+main_body:
+  %v = call <4 x float> @llvm.amdgcn.image.sample.cd.2d.v4f32.f16.f32(i32 15, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
+  ret <4 x float> %v
+}
+
+define amdgpu_ps <4 x float> @sample_c_cd_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, half %dsdh, half %dsdv, float %s) {
+; GFX10-LABEL: sample_c_cd_1d:
+; GFX10:       ; %bb.0: ; %main_body
+; GFX10-NEXT:    s_lshl_b32 s12, s0, 16
+; GFX10-NEXT:    v_and_or_b32 v1, 0xffff, v1, s12
+; GFX10-NEXT:    v_and_or_b32 v2, 0xffff, v2, s12
+; GFX10-NEXT:    image_sample_c_cd_g16 v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    ; return to shader part epilog
+main_body:
+  %v = call <4 x float> @llvm.amdgcn.image.sample.c.cd.1d.v4f32.f16.f32(i32 15, float %zcompare, half %dsdh, half %dsdv, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
+  ret <4 x float> %v
+}
+
+define amdgpu_ps <4 x float> @sample_c_cd_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t) {
+; GFX10-LABEL: sample_c_cd_2d:
+; GFX10:       ; %bb.0: ; %main_body
+; GFX10-NEXT:    v_lshlrev_b32_e32 v2, 16, v2
+; GFX10-NEXT:    v_lshlrev_b32_e32 v4, 16, v4
+; GFX10-NEXT:    v_and_or_b32 v1, 0xffff, v1, v2
+; GFX10-NEXT:    v_and_or_b32 v2, 0xffff, v3, v4
+; GFX10-NEXT:    image_sample_c_cd_g16 v[0:3], [v0, v1, v2, v5, v6], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    ; return to shader part epilog
+main_body:
+  %v = call <4 x float> @llvm.amdgcn.image.sample.c.cd.2d.v4f32.f16.f32(i32 15, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
+  ret <4 x float> %v
+}
+
+define amdgpu_ps <4 x float> @sample_cd_cl_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, half %dsdh, half %dsdv, float %s, float %clamp) {
+; GFX10-LABEL: sample_cd_cl_1d:
+; GFX10:       ; %bb.0: ; %main_body
+; GFX10-NEXT:    s_lshl_b32 s12, s0, 16
+; GFX10-NEXT:    v_and_or_b32 v0, 0xffff, v0, s12
+; GFX10-NEXT:    v_and_or_b32 v1, 0xffff, v1, s12
+; GFX10-NEXT:    image_sample_cd_cl_g16 v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    ; return to shader part epilog
+main_body:
+  %v = call <4 x float> @llvm.amdgcn.image.sample.cd.cl.1d.v4f32.f16.f32(i32 15, half %dsdh, half %dsdv, float %s, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
+  ret <4 x float> %v
+}
+
+define amdgpu_ps <4 x float> @sample_cd_cl_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, float %clamp) {
+; GFX10-LABEL: sample_cd_cl_2d:
+; GFX10:       ; %bb.0: ; %main_body
+; GFX10-NEXT:    v_lshlrev_b32_e32 v1, 16, v1
+; GFX10-NEXT:    v_lshlrev_b32_e32 v3, 16, v3
+; GFX10-NEXT:    v_and_or_b32 v0, 0xffff, v0, v1
+; GFX10-NEXT:    v_and_or_b32 v1, 0xffff, v2, v3
+; GFX10-NEXT:    image_sample_cd_cl_g16 v[0:3], [v0, v1, v4, v5, v6], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    ; return to shader part epilog
+main_body:
+  %v = call <4 x float> @llvm.amdgcn.image.sample.cd.cl.2d.v4f32.f16.f32(i32 15, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
+  ret <4 x float> %v
+}
+
+define amdgpu_ps <4 x float> @sample_c_cd_cl_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, half %dsdh, half %dsdv, float %s, float %clamp) {
+; GFX10-LABEL: sample_c_cd_cl_1d:
+; GFX10:       ; %bb.0: ; %main_body
+; GFX10-NEXT:    s_lshl_b32 s12, s0, 16
+; GFX10-NEXT:    v_and_or_b32 v1, 0xffff, v1, s12
+; GFX10-NEXT:    v_and_or_b32 v2, 0xffff, v2, s12
+; GFX10-NEXT:    image_sample_c_cd_cl_g16 v[0:3], v[0:4], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    ; return to shader part epilog
+main_body:
+  %v = call <4 x float> @llvm.amdgcn.image.sample.c.cd.cl.1d.v4f32.f16.f32(i32 15, float %zcompare, half %dsdh, half %dsdv, float %s, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
+  ret <4 x float> %v
+}
+
+define amdgpu_ps <4 x float> @sample_c_cd_cl_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, float %clamp) {
+; GFX10-LABEL: sample_c_cd_cl_2d:
+; GFX10:       ; %bb.0: ; %main_body
+; GFX10-NEXT:    v_mov_b32_e32 v8, v2
+; GFX10-NEXT:    v_mov_b32_e32 v9, v3
+; GFX10-NEXT:    v_mov_b32_e32 v2, v0
+; GFX10-NEXT:    v_lshlrev_b32_e32 v4, 16, v4
+; GFX10-NEXT:    v_lshlrev_b32_e32 v0, 16, v8
+; GFX10-NEXT:    v_and_or_b32 v4, 0xffff, v9, v4
+; GFX10-NEXT:    v_and_or_b32 v3, 0xffff, v1, v0
+; GFX10-NEXT:    image_sample_c_cd_cl_g16 v[0:3], v[2:7], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    ; return to shader part epilog
+main_body:
+  %v = call <4 x float> @llvm.amdgcn.image.sample.c.cd.cl.2d.v4f32.f16.f32(i32 15, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
+  ret <4 x float> %v
+}
+
+declare <4 x float> @llvm.amdgcn.image.sample.cd.1d.v4f32.f16.f32(i32, half, half, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
+declare <4 x float> @llvm.amdgcn.image.sample.cd.2d.v4f32.f16.f32(i32, half, half, half, half, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
+declare <4 x float> @llvm.amdgcn.image.sample.c.cd.1d.v4f32.f16.f32(i32, float, half, half, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
+declare <4 x float> @llvm.amdgcn.image.sample.c.cd.2d.v4f32.f16.f32(i32, float, half, half, half, half, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
+declare <4 x float> @llvm.amdgcn.image.sample.cd.cl.1d.v4f32.f16.f32(i32, half, half, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
+declare <4 x float> @llvm.amdgcn.image.sample.cd.cl.2d.v4f32.f16.f32(i32, half, half, half, half, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
+declare <4 x float> @llvm.amdgcn.image.sample.c.cd.cl.1d.v4f32.f16.f32(i32, float, half, half, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
+declare <4 x float> @llvm.amdgcn.image.sample.c.cd.cl.2d.v4f32.f16.f32(i32, float, half, half, half, half, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
+
+attributes #0 = { nounwind }
+attributes #1 = { nounwind readonly }
+attributes #2 = { nounwind readnone }

llvm/test/CodeGen/AMDGPU/GlobalISel/llvm.amdgcn.image.sample.g16.ll

 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
 ; RUN: llc -global-isel -march=amdgcn -mcpu=gfx1010 -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX10 %s
+; RUN: llc -global-isel -march=amdgcn -mcpu=gfx1100 -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX10 %s
 
 define amdgpu_ps <4 x float> @sample_d_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, half %dsdh, half %dsdv, float %s) {
 ; GFX10-LABEL: sample_d_1d:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    s_lshl_b32 s12, s0, 16
 ; GFX10-NEXT:    v_and_or_b32 v0, 0xffff, v0, s12
 ; GFX10-NEXT:    v_and_or_b32 v1, 0xffff, v1, s12
 ; GFX10-NEXT:    image_sample_d_g16 v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D

 ; GFX10-NEXT:    image_sample_c_d_cl_g16 v[0:3], v[2:7], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D
 ; GFX10-NEXT:    s_waitcnt vmcnt(0)
 ; GFX10-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.c.d.cl.2d.v4f32.f16.f32(i32 15, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v
 }
 
-define amdgpu_ps <4 x float> @sample_cd_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, half %dsdh, half %dsdv, float %s) {
-; GFX10-LABEL: sample_cd_1d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    s_lshl_b32 s12, s0, 16
-; GFX10-NEXT:    v_and_or_b32 v0, 0xffff, v0, s12
-; GFX10-NEXT:    v_and_or_b32 v1, 0xffff, v1, s12
-; GFX10-NEXT:    image_sample_cd_g16 v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D
-; GFX10-NEXT:    s_waitcnt vmcnt(0)
-; GFX10-NEXT:    ; return to shader part epilog
-main_body:
-  %v = call <4 x float> @llvm.amdgcn.image.sample.cd.1d.v4f32.f16.f32(i32 15, half %dsdh, half %dsdv, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
-  ret <4 x float> %v
-}
-
-define amdgpu_ps <4 x float> @sample_cd_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t) {
-; GFX10-LABEL: sample_cd_2d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    v_lshlrev_b32_e32 v1, 16, v1
-; GFX10-NEXT:    v_lshlrev_b32_e32 v3, 16, v3
-; GFX10-NEXT:    v_and_or_b32 v0, 0xffff, v0, v1
-; GFX10-NEXT:    v_and_or_b32 v1, 0xffff, v2, v3
-; GFX10-NEXT:    image_sample_cd_g16 v[0:3], [v0, v1, v4, v5], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D
-; GFX10-NEXT:    s_waitcnt vmcnt(0)
-; GFX10-NEXT:    ; return to shader part epilog
-main_body:
-  %v = call <4 x float> @llvm.amdgcn.image.sample.cd.2d.v4f32.f16.f32(i32 15, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
-  ret <4 x float> %v
-}
-
-define amdgpu_ps <4 x float> @sample_c_cd_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, half %dsdh, half %dsdv, float %s) {
-; GFX10-LABEL: sample_c_cd_1d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    s_lshl_b32 s12, s0, 16
-; GFX10-NEXT:    v_and_or_b32 v1, 0xffff, v1, s12
-; GFX10-NEXT:    v_and_or_b32 v2, 0xffff, v2, s12
-; GFX10-NEXT:    image_sample_c_cd_g16 v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D
-; GFX10-NEXT:    s_waitcnt vmcnt(0)
-; GFX10-NEXT:    ; return to shader part epilog
-main_body:
-  %v = call <4 x float> @llvm.amdgcn.image.sample.c.cd.1d.v4f32.f16.f32(i32 15, float %zcompare, half %dsdh, half %dsdv, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
-  ret <4 x float> %v
-}
-
-define amdgpu_ps <4 x float> @sample_c_cd_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t) {
-; GFX10-LABEL: sample_c_cd_2d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    v_lshlrev_b32_e32 v2, 16, v2
-; GFX10-NEXT:    v_lshlrev_b32_e32 v4, 16, v4
-; GFX10-NEXT:    v_and_or_b32 v1, 0xffff, v1, v2
-; GFX10-NEXT:    v_and_or_b32 v2, 0xffff, v3, v4
-; GFX10-NEXT:    image_sample_c_cd_g16 v[0:3], [v0, v1, v2, v5, v6], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D
-; GFX10-NEXT:    s_waitcnt vmcnt(0)
-; GFX10-NEXT:    ; return to shader part epilog
-main_body:
-  %v = call <4 x float> @llvm.amdgcn.image.sample.c.cd.2d.v4f32.f16.f32(i32 15, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
-  ret <4 x float> %v
-}
-
-define amdgpu_ps <4 x float> @sample_cd_cl_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, half %dsdh, half %dsdv, float %s, float %clamp) {
-; GFX10-LABEL: sample_cd_cl_1d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    s_lshl_b32 s12, s0, 16
-; GFX10-NEXT:    v_and_or_b32 v0, 0xffff, v0, s12
-; GFX10-NEXT:    v_and_or_b32 v1, 0xffff, v1, s12
-; GFX10-NEXT:    image_sample_cd_cl_g16 v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D
-; GFX10-NEXT:    s_waitcnt vmcnt(0)
-; GFX10-NEXT:    ; return to shader part epilog
-main_body:
-  %v = call <4 x float> @llvm.amdgcn.image.sample.cd.cl.1d.v4f32.f16.f32(i32 15, half %dsdh, half %dsdv, float %s, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
-  ret <4 x float> %v
-}
-
-define amdgpu_ps <4 x float> @sample_cd_cl_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, float %clamp) {
-; GFX10-LABEL: sample_cd_cl_2d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    v_lshlrev_b32_e32 v1, 16, v1
-; GFX10-NEXT:    v_lshlrev_b32_e32 v3, 16, v3
-; GFX10-NEXT:    v_and_or_b32 v0, 0xffff, v0, v1
-; GFX10-NEXT:    v_and_or_b32 v1, 0xffff, v2, v3
-; GFX10-NEXT:    image_sample_cd_cl_g16 v[0:3], [v0, v1, v4, v5, v6], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D
-; GFX10-NEXT:    s_waitcnt vmcnt(0)
-; GFX10-NEXT:    ; return to shader part epilog
-main_body:
-  %v = call <4 x float> @llvm.amdgcn.image.sample.cd.cl.2d.v4f32.f16.f32(i32 15, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
-  ret <4 x float> %v
-}
-
-define amdgpu_ps <4 x float> @sample_c_cd_cl_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, half %dsdh, half %dsdv, float %s, float %clamp) {
-; GFX10-LABEL: sample_c_cd_cl_1d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    s_lshl_b32 s12, s0, 16
-; GFX10-NEXT:    v_and_or_b32 v1, 0xffff, v1, s12
-; GFX10-NEXT:    v_and_or_b32 v2, 0xffff, v2, s12
-; GFX10-NEXT:    image_sample_c_cd_cl_g16 v[0:3], v[0:4], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D
-; GFX10-NEXT:    s_waitcnt vmcnt(0)
-; GFX10-NEXT:    ; return to shader part epilog
-main_body:
-  %v = call <4 x float> @llvm.amdgcn.image.sample.c.cd.cl.1d.v4f32.f16.f32(i32 15, float %zcompare, half %dsdh, half %dsdv, float %s, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
-  ret <4 x float> %v
-}
-
-define amdgpu_ps <4 x float> @sample_c_cd_cl_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, float %clamp) {
-; GFX10-LABEL: sample_c_cd_cl_2d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    v_mov_b32_e32 v8, v2
-; GFX10-NEXT:    v_mov_b32_e32 v9, v3
-; GFX10-NEXT:    v_mov_b32_e32 v2, v0
-; GFX10-NEXT:    v_lshlrev_b32_e32 v4, 16, v4
-; GFX10-NEXT:    v_lshlrev_b32_e32 v0, 16, v8
-; GFX10-NEXT:    v_and_or_b32 v4, 0xffff, v9, v4
-; GFX10-NEXT:    v_and_or_b32 v3, 0xffff, v1, v0
-; GFX10-NEXT:    image_sample_c_cd_cl_g16 v[0:3], v[2:7], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D
-; GFX10-NEXT:    s_waitcnt vmcnt(0)
-; GFX10-NEXT:    ; return to shader part epilog
-main_body:
-  %v = call <4 x float> @llvm.amdgcn.image.sample.c.cd.cl.2d.v4f32.f16.f32(i32 15, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
-  ret <4 x float> %v
-}
-
 define amdgpu_ps float @sample_c_d_o_2darray_V1(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, i32 %offset, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, float %slice) {
 ; GFX10-LABEL: sample_c_d_o_2darray_V1:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    v_mov_b32_e32 v9, v3
 ; GFX10-NEXT:    v_mov_b32_e32 v10, v2
 ; GFX10-NEXT:    v_mov_b32_e32 v11, v4
 ; GFX10-NEXT:    v_mov_b32_e32 v2, v0
 ; GFX10-NEXT:    v_mov_b32_e32 v3, v1

 declare <4 x float> @llvm.amdgcn.image.sample.d.3d.v4f32.f16.f32(i32, half, half, half, half, half, half, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 declare <4 x float> @llvm.amdgcn.image.sample.c.d.1d.v4f32.f16.f32(i32, float, half, half, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 declare <4 x float> @llvm.amdgcn.image.sample.c.d.2d.v4f32.f16.f32(i32, float, half, half, half, half, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 declare <4 x float> @llvm.amdgcn.image.sample.d.cl.1d.v4f32.f16.f32(i32, half, half, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 declare <4 x float> @llvm.amdgcn.image.sample.d.cl.2d.v4f32.f16.f32(i32, half, half, half, half, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 declare <4 x float> @llvm.amdgcn.image.sample.c.d.cl.1d.v4f32.f16.f32(i32, float, half, half, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 declare <4 x float> @llvm.amdgcn.image.sample.c.d.cl.2d.v4f32.f16.f32(i32, float, half, half, half, half, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 
-declare <4 x float> @llvm.amdgcn.image.sample.cd.1d.v4f32.f16.f32(i32, half, half, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
-declare <4 x float> @llvm.amdgcn.image.sample.cd.2d.v4f32.f16.f32(i32, half, half, half, half, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
-declare <4 x float> @llvm.amdgcn.image.sample.c.cd.1d.v4f32.f16.f32(i32, float, half, half, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
-declare <4 x float> @llvm.amdgcn.image.sample.c.cd.2d.v4f32.f16.f32(i32, float, half, half, half, half, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
-declare <4 x float> @llvm.amdgcn.image.sample.cd.cl.1d.v4f32.f16.f32(i32, half, half, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
-declare <4 x float> @llvm.amdgcn.image.sample.cd.cl.2d.v4f32.f16.f32(i32, half, half, half, half, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
-declare <4 x float> @llvm.amdgcn.image.sample.c.cd.cl.1d.v4f32.f16.f32(i32, float, half, half, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
-declare <4 x float> @llvm.amdgcn.image.sample.c.cd.cl.2d.v4f32.f16.f32(i32, float, half, half, half, half, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
-
 declare float @llvm.amdgcn.image.sample.c.d.o.2darray.f16.f32.f32(i32, i32, float, half, half, half, half, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 declare <2 x float> @llvm.amdgcn.image.sample.c.d.o.2darray.v2f32.f16.f32(i32, i32, float, half, half, half, half, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 
 attributes #0 = { nounwind }
 attributes #1 = { nounwind readonly }
 attributes #2 = { nounwind readnone }

llvm/test/CodeGen/AMDGPU/GlobalISel/llvm.amdgcn.image.store.2d.ll

 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
 ; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=tahiti -o - %s | FileCheck -check-prefix=GFX6 %s
 ; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=fiji -o - %s | FileCheck -check-prefix=GFX8 %s
 ; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx1010 -o - %s | FileCheck -check-prefix=GFX10 %s
+; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx1100 -o - %s | FileCheck -check-prefix=GFX10 %s
 
 define amdgpu_ps void @image_store_f32(<8 x i32> inreg %rsrc, i32 %s, i32 %t, float %data) {
 ; GFX6-LABEL: image_store_f32:
 ; GFX6:       ; %bb.0:
 ; GFX6-NEXT:    s_mov_b32 s0, s2
 ; GFX6-NEXT:    s_mov_b32 s1, s3
 ; GFX6-NEXT:    s_mov_b32 s2, s4
 ; GFX6-NEXT:    s_mov_b32 s3, s5

llvm/test/CodeGen/AMDGPU/GlobalISel/llvm.amdgcn.intersect_ray.ll

 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
-; RUN: llc -global-isel -march=amdgcn -mcpu=gfx1030 -verify-machineinstrs < %s | FileCheck -check-prefixes=GCN,GFX1030 %s
-; RUN: llc -global-isel -march=amdgcn -mcpu=gfx1013 -verify-machineinstrs < %s | FileCheck -check-prefixes=GCN,GFX1013 %s
+; RUN: llc -global-isel -march=amdgcn -mcpu=gfx1030 -verify-machineinstrs < %s | FileCheck -check-prefixes=GCN,GFX10,GFX1030 %s
+; RUN: llc -global-isel -march=amdgcn -mcpu=gfx1013 -verify-machineinstrs < %s | FileCheck -check-prefixes=GCN,GFX10,GFX1013 %s
+; RUN: llc -global-isel -march=amdgcn -mcpu=gfx1100 -verify-machineinstrs < %s | FileCheck -check-prefixes=GCN,GFX11 %s
 ; RUN: not --crash llc -global-isel -march=amdgcn -mcpu=gfx1012 -verify-machineinstrs < %s -o /dev/null 2>&1 | FileCheck -check-prefix=ERR %s
 
 ; uint4 llvm.amdgcn.image.bvh.intersect.ray.i32.v4f32(uint node_ptr, float ray_extent, float3 ray_origin, float3 ray_dir, float3 ray_inv_dir, uint4 texture_descr)
 ; uint4 llvm.amdgcn.image.bvh.intersect.ray.i32.v4f16(uint node_ptr, float ray_extent, float3 ray_origin, half3 ray_dir, half3 ray_inv_dir, uint4 texture_descr)
 ; uint4 llvm.amdgcn.image.bvh.intersect.ray.i64.v4f32(ulong node_ptr, float ray_extent, float3 ray_origin, float3 ray_dir, float3 ray_inv_dir, uint4 texture_descr)
 ; uint4 llvm.amdgcn.image.bvh.intersect.ray.i64.v4f16(ulong node_ptr, float ray_extent, float3 ray_origin, half3 ray_dir, half3 ray_inv_dir, uint4 texture_descr)
 
 declare <4 x i32> @llvm.amdgcn.image.bvh.intersect.ray.i32.v4f32(i32, float, <3 x float>, <3 x float>, <3 x float>, <4 x i32>)
; GCN-NEXT:    ; return to shader part epilog
   %ray_inv_dir1 = insertelement <3 x float> %ray_inv_dir0, float %ray_inv_dir_y, i32 1
   %ray_inv_dir = insertelement <3 x float> %ray_inv_dir1, float %ray_inv_dir_z, i32 2
   %v = call <4 x i32> @llvm.amdgcn.image.bvh.intersect.ray.i32.v4f32(i32 %node_ptr, float %ray_extent, <3 x float> %ray_origin, <3 x float> %ray_dir, <3 x float> %ray_inv_dir, <4 x i32> %tdescr)
  %r = bitcast <4 x i32> %v to <4 x float>
  ret <4 x float> %r
 }
 
 define amdgpu_ps <4 x float> @image_bvh_intersect_ray_a16(i32 %node_ptr, float %ray_extent, <3 x float> %ray_origin, <3 x half> %ray_dir, <3 x half> %ray_inv_dir, <4 x i32> inreg %tdescr) {
-; GCN-LABEL: image_bvh_intersect_ray_a16:
-; GCN:       ; %bb.0:
-; GCN-NEXT:    v_lshrrev_b32_e32 v9, 16, v5
-; GCN-NEXT:    v_and_b32_e32 v10, 0xffff, v7
-; GCN-NEXT:    v_and_b32_e32 v8, 0xffff, v8
-; GCN-NEXT:    v_lshlrev_b32_e32 v9, 16, v9
-; GCN-NEXT:    v_lshlrev_b32_e32 v10, 16, v10
-; GCN-NEXT:    v_alignbit_b32 v7, v8, v7, 16
-; GCN-NEXT:    v_and_or_b32 v5, v5, 0xffff, v9
-; GCN-NEXT:    v_and_or_b32 v6, v6, 0xffff, v10
-; GCN-NEXT:    image_bvh_intersect_ray v[0:3], v[0:7], s[0:3] a16
-; GCN-NEXT:    s_waitcnt vmcnt(0)
-; GCN-NEXT:    ; return to shader part epilog
+; GFX10-LABEL: image_bvh_intersect_ray_a16:
+; GFX10:       ; %bb.0:
+; GFX10-NEXT:    v_lshrrev_b32_e32 v9, 16, v5
+; GFX10-NEXT:    v_and_b32_e32 v10, 0xffff, v7
+; GFX10-NEXT:    v_and_b32_e32 v8, 0xffff, v8
+; GFX10-NEXT:    v_lshlrev_b32_e32 v9, 16, v9
+; GFX10-NEXT:    v_lshlrev_b32_e32 v10, 16, v10
+; GFX10-NEXT:    v_alignbit_b32 v7, v8, v7, 16
+; GFX10-NEXT:    v_and_or_b32 v5, v5, 0xffff, v9
+; GFX10-NEXT:    v_and_or_b32 v6, v6, 0xffff, v10
+; GFX10-NEXT:    image_bvh_intersect_ray v[0:3], v[0:7], s[0:3] a16
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: image_bvh_intersect_ray_a16:
+; GFX11:       ; %bb.0:
+; GFX11-NEXT:    v_lshrrev_b32_e32 v9, 16, v5
+; GFX11-NEXT:    v_lshrrev_b32_e32 v10, 16, v7
+; GFX11-NEXT:    v_lshlrev_b32_e32 v5, 16, v5
+; GFX11-NEXT:    v_lshlrev_b32_e32 v11, 16, v6
+; GFX11-NEXT:    v_lshlrev_b32_e32 v9, 16, v9
+; GFX11-NEXT:    v_and_or_b32 v5, 0xffff, v7, v5
+; GFX11-NEXT:    v_and_or_b32 v7, 0xffff, v8, v11
+; GFX11-NEXT:    v_and_or_b32 v6, 0xffff, v10, v9
+; GFX11-NEXT:    image_bvh_intersect_ray v[0:3], v[0:7], s[0:3] a16
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    ; return to shader part epilog
   %v = call <4 x i32> @llvm.amdgcn.image.bvh.intersect.ray.i32.v4f16(i32 %node_ptr, float %ray_extent, <3 x float> %ray_origin, <3 x half> %ray_dir, <3 x half> %ray_inv_dir, <4 x i32> %tdescr)
   %r = bitcast <4 x i32> %v to <4 x float>
   ret <4 x float> %r
 }
 
 define amdgpu_ps <4 x float> @image_bvh64_intersect_ray(i64 %node_ptr, float %ray_extent, <3 x float> %ray_origin, <3 x float> %ray_dir, <3 x float> %ray_inv_dir, <4 x i32> inreg %tdescr) {
 ; GCN-LABEL: image_bvh64_intersect_ray:
 ; GCN:       ; %bb.0:
; GCN-NEXT:    ; return to shader part epilog
   %ray_inv_dir1 = insertelement <3 x float> %ray_inv_dir0, float %ray_inv_dir_y, i32 1
   %ray_inv_dir = insertelement <3 x float> %ray_inv_dir1, float %ray_inv_dir_z, i32 2
   %v = call <4 x i32> @llvm.amdgcn.image.bvh.intersect.ray.i64.v4f32(i64 %node_ptr, float %ray_extent, <3 x float> %ray_origin, <3 x float> %ray_dir, <3 x float> %ray_inv_dir, <4 x i32> %tdescr)
  %r = bitcast <4 x i32> %v to <4 x float>
  ret <4 x float> %r
 }
 
 define amdgpu_ps <4 x float> @image_bvh64_intersect_ray_a16(i64 %node_ptr, float %ray_extent, <3 x float> %ray_origin, <3 x half> %ray_dir, <3 x half> %ray_inv_dir, <4 x i32> inreg %tdescr) {
-; GCN-LABEL: image_bvh64_intersect_ray_a16:
-; GCN:       ; %bb.0:
-; GCN-NEXT:    v_lshrrev_b32_e32 v10, 16, v6
-; GCN-NEXT:    v_and_b32_e32 v11, 0xffff, v8
-; GCN-NEXT:    v_and_b32_e32 v9, 0xffff, v9
-; GCN-NEXT:    v_lshlrev_b32_e32 v10, 16, v10
-; GCN-NEXT:    v_lshlrev_b32_e32 v11, 16, v11
-; GCN-NEXT:    v_alignbit_b32 v8, v9, v8, 16
-; GCN-NEXT:    v_and_or_b32 v6, v6, 0xffff, v10
-; GCN-NEXT:    v_and_or_b32 v7, v7, 0xffff, v11
-; GCN-NEXT:    image_bvh64_intersect_ray v[0:3], v[0:15], s[0:3] a16
-; GCN-NEXT:    s_waitcnt vmcnt(0)
-; GCN-NEXT:    ; return to shader part epilog
+; GFX10-LABEL: image_bvh64_intersect_ray_a16:
+; GFX10:       ; %bb.0:
+; GFX10-NEXT:    v_lshrrev_b32_e32 v10, 16, v6
+; GFX10-NEXT:    v_and_b32_e32 v11, 0xffff, v8
+; GFX10-NEXT:    v_and_b32_e32 v9, 0xffff, v9
+; GFX10-NEXT:    v_lshlrev_b32_e32 v10, 16, v10
+; GFX10-NEXT:    v_lshlrev_b32_e32 v11, 16, v11
+; GFX10-NEXT:    v_alignbit_b32 v8, v9, v8, 16
+; GFX10-NEXT:    v_and_or_b32 v6, v6, 0xffff, v10
+; GFX10-NEXT:    v_and_or_b32 v7, v7, 0xffff, v11
+; GFX10-NEXT:    image_bvh64_intersect_ray v[0:3], v[0:15], s[0:3] a16
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: image_bvh64_intersect_ray_a16:
+; GFX11:       ; %bb.0:
+; GFX11-NEXT:    v_lshrrev_b32_e32 v10, 16, v6
+; GFX11-NEXT:    v_lshrrev_b32_e32 v11, 16, v8
+; GFX11-NEXT:    v_lshlrev_b32_e32 v6, 16, v6
+; GFX11-NEXT:    v_lshlrev_b32_e32 v12, 16, v7
+; GFX11-NEXT:    v_lshlrev_b32_e32 v10, 16, v10
+; GFX11-NEXT:    v_and_or_b32 v6, 0xffff, v8, v6
+; GFX11-NEXT:    v_and_or_b32 v8, 0xffff, v9, v12
+; GFX11-NEXT:    v_and_or_b32 v7, 0xffff, v11, v10
+; GFX11-NEXT:    image_bvh64_intersect_ray v[0:3], v[0:15], s[0:3] a16
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    ; return to shader part epilog
   %v = call <4 x i32> @llvm.amdgcn.image.bvh.intersect.ray.i64.v4f16(i64 %node_ptr, float %ray_extent, <3 x float> %ray_origin, <3 x half> %ray_dir, <3 x half> %ray_inv_dir, <4 x i32> %tdescr)
   %r = bitcast <4 x i32> %v to <4 x float>
   ret <4 x float> %r
 }
 
 define amdgpu_ps <4 x float> @image_bvh_intersect_ray_vgpr_descr(i32 %node_ptr, float %ray_extent, <3 x float> %ray_origin, <3 x float> %ray_dir, <3 x float> %ray_inv_dir, <4 x i32> %tdescr) {
 ; GFX1030-LABEL: image_bvh_intersect_ray_vgpr_descr:
 ; GFX1030:       ; %bb.0:

 ; GFX1013-NEXT:  ; %bb.2:
 ; GFX1013-NEXT:    s_mov_b32 exec_lo, s1
 ; GFX1013-NEXT:    s_waitcnt vmcnt(0)
 ; GFX1013-NEXT:    v_mov_b32_e32 v0, v20
 ; GFX1013-NEXT:    v_mov_b32_e32 v1, v21
 ; GFX1013-NEXT:    v_mov_b32_e32 v2, v22
 ; GFX1013-NEXT:    v_mov_b32_e32 v3, v23
 ; GFX1013-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: image_bvh_intersect_ray_vgpr_descr:
+; GFX11:       ; %bb.0:
+; GFX11-NEXT:    v_mov_b32_e32 v18, v0
+; GFX11-NEXT:    v_mov_b32_e32 v19, v1
+; GFX11-NEXT:    v_mov_b32_e32 v15, v2
+; GFX11-NEXT:    v_mov_b32_e32 v16, v3
+; GFX11-NEXT:    v_mov_b32_e32 v17, v4
+; GFX11-NEXT:    s_mov_b32 s1, exec_lo
+; GFX11-NEXT:  .LBB6_1: ; =>This Inner Loop Header: Depth=1
+; GFX11-NEXT:    v_readfirstlane_b32 s4, v11
+; GFX11-NEXT:    v_readfirstlane_b32 s5, v12
+; GFX11-NEXT:    v_readfirstlane_b32 s6, v13
+; GFX11-NEXT:    v_readfirstlane_b32 s7, v14
+; GFX11-NEXT:    v_cmp_eq_u64_e32 vcc_lo, s[4:5], v[11:12]
+; GFX11-NEXT:    v_cmp_eq_u64_e64 s0, s[6:7], v[13:14]
+; GFX11-NEXT:    s_and_b32 s0, vcc_lo, s0
+; GFX11-NEXT:    s_and_saveexec_b32 s0, s0
+; GFX11-NEXT:    image_bvh_intersect_ray v[0:3], [v18, v19, v[15:17], v[5:7], v[8:10]], s[4:7]
+; GFX11-NEXT:    ; implicit-def: $vgpr11
+; GFX11-NEXT:    ; implicit-def: $vgpr18
+; GFX11-NEXT:    ; implicit-def: $vgpr19
+; GFX11-NEXT:    ; implicit-def: $vgpr15_vgpr16_vgpr17
+; GFX11-NEXT:    ; implicit-def: $vgpr5_vgpr6_vgpr7
+; GFX11-NEXT:    ; implicit-def: $vgpr8_vgpr9_vgpr10
+; GFX11-NEXT:    ; implicit-def: $vgpr11_vgpr12_vgpr13_vgpr14
+; GFX11-NEXT:    s_xor_b32 exec_lo, exec_lo, s0
+; GFX11-NEXT:    s_cbranch_execnz .LBB6_1
+; GFX11-NEXT:  ; %bb.2:
+; GFX11-NEXT:    s_mov_b32 exec_lo, s1
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    ; return to shader part epilog
   %v = call <4 x i32> @llvm.amdgcn.image.bvh.intersect.ray.i32.v4f32(i32 %node_ptr, float %ray_extent, <3 x float> %ray_origin, <3 x float> %ray_dir, <3 x float> %ray_inv_dir, <4 x i32> %tdescr)
   %r = bitcast <4 x i32> %v to <4 x float>
   ret <4 x float> %r
 }
 
 define amdgpu_ps <4 x float> @image_bvh_intersect_ray_a16_vgpr_descr(i32 %node_ptr, float %ray_extent, <3 x float> %ray_origin, <3 x half> %ray_dir, <3 x half> %ray_inv_dir, <4 x i32> %tdescr) {
 ; GFX1030-LABEL: image_bvh_intersect_ray_a16_vgpr_descr:
 ; GFX1030:       ; %bb.0:

 ; GFX1013-NEXT:  ; %bb.2:
 ; GFX1013-NEXT:    s_mov_b32 exec_lo, s1
 ; GFX1013-NEXT:    s_waitcnt vmcnt(0)
 ; GFX1013-NEXT:    v_mov_b32_e32 v0, v13
 ; GFX1013-NEXT:    v_mov_b32_e32 v1, v14
 ; GFX1013-NEXT:    v_mov_b32_e32 v2, v15
 ; GFX1013-NEXT:    v_mov_b32_e32 v3, v16
 ; GFX1013-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: image_bvh_intersect_ray_a16_vgpr_descr:
+; GFX11:       ; %bb.0:
+; GFX11-NEXT:    v_mov_b32_e32 v13, v0
+; GFX11-NEXT:    v_lshrrev_b32_e32 v0, 16, v5
+; GFX11-NEXT:    v_mov_b32_e32 v14, v1
+; GFX11-NEXT:    v_mov_b32_e32 v15, v2
+; GFX11-NEXT:    v_mov_b32_e32 v16, v3
+; GFX11-NEXT:    v_lshrrev_b32_e32 v1, 16, v7
+; GFX11-NEXT:    v_lshlrev_b32_e32 v2, 16, v5
+; GFX11-NEXT:    v_lshlrev_b32_e32 v0, 16, v0
+; GFX11-NEXT:    v_lshlrev_b32_e32 v3, 16, v6
+; GFX11-NEXT:    v_mov_b32_e32 v17, v4
+; GFX11-NEXT:    s_mov_b32 s1, exec_lo
+; GFX11-NEXT:    v_and_or_b32 v4, 0xffff, v7, v2
+; GFX11-NEXT:    v_and_or_b32 v5, 0xffff, v1, v0
+; GFX11-NEXT:    v_and_or_b32 v6, 0xffff, v8, v3
+; GFX11-NEXT:  .LBB7_1: ; =>This Inner Loop Header: Depth=1
+; GFX11-NEXT:    v_readfirstlane_b32 s4, v9
+; GFX11-NEXT:    v_readfirstlane_b32 s5, v10
+; GFX11-NEXT:    v_readfirstlane_b32 s6, v11
+; GFX11-NEXT:    v_readfirstlane_b32 s7, v12
+; GFX11-NEXT:    v_cmp_eq_u64_e32 vcc_lo, s[4:5], v[9:10]
+; GFX11-NEXT:    v_cmp_eq_u64_e64 s0, s[6:7], v[11:12]
+; GFX11-NEXT:    s_and_b32 s0, vcc_lo, s0
+; GFX11-NEXT:    s_and_saveexec_b32 s0, s0
+; GFX11-NEXT:    image_bvh_intersect_ray v[0:3], [v13, v14, v[15:17], v[4:6]], s[4:7] a16
+; GFX11-NEXT:    ; implicit-def: $vgpr9
+; GFX11-NEXT:    ; implicit-def: $vgpr13
+; GFX11-NEXT:    ; implicit-def: $vgpr14
+; GFX11-NEXT:    ; implicit-def: $vgpr15_vgpr16_vgpr17
+; GFX11-NEXT:    ; implicit-def: $vgpr4_vgpr5_vgpr6
+; GFX11-NEXT:    ; implicit-def: $vgpr9_vgpr10_vgpr11_vgpr12
+; GFX11-NEXT:    s_xor_b32 exec_lo, exec_lo, s0
+; GFX11-NEXT:    s_cbranch_execnz .LBB7_1
+; GFX11-NEXT:  ; %bb.2:
+; GFX11-NEXT:    s_mov_b32 exec_lo, s1
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    ; return to shader part epilog
   %v = call <4 x i32> @llvm.amdgcn.image.bvh.intersect.ray.i32.v4f16(i32 %node_ptr, float %ray_extent, <3 x float> %ray_origin, <3 x half> %ray_dir, <3 x half> %ray_inv_dir, <4 x i32> %tdescr)
   %r = bitcast <4 x i32> %v to <4 x float>
   ret <4 x float> %r
 }
 
 define amdgpu_ps <4 x float> @image_bvh64_intersect_ray_vgpr_descr(i64 %node_ptr, float %ray_extent, <3 x float> %ray_origin, <3 x float> %ray_dir, <3 x float> %ray_inv_dir, <4 x i32> %tdescr) {
 ; GFX1030-LABEL: image_bvh64_intersect_ray_vgpr_descr:
 ; GFX1030:       ; %bb.0:

 ; GFX1013-NEXT:  ; %bb.2:
 ; GFX1013-NEXT:    s_mov_b32 exec_lo, s1
 ; GFX1013-NEXT:    s_waitcnt vmcnt(0)
 ; GFX1013-NEXT:    v_mov_b32_e32 v0, v20
 ; GFX1013-NEXT:    v_mov_b32_e32 v1, v21
 ; GFX1013-NEXT:    v_mov_b32_e32 v2, v22
 ; GFX1013-NEXT:    v_mov_b32_e32 v3, v23
 ; GFX1013-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: image_bvh64_intersect_ray_vgpr_descr:
+; GFX11:       ; %bb.0:
+; GFX11-NEXT:    v_mov_b32_e32 v19, v0
+; GFX11-NEXT:    v_mov_b32_e32 v20, v1
+; GFX11-NEXT:    v_mov_b32_e32 v21, v2
+; GFX11-NEXT:    v_mov_b32_e32 v16, v3
+; GFX11-NEXT:    v_mov_b32_e32 v17, v4
+; GFX11-NEXT:    v_mov_b32_e32 v18, v5
+; GFX11-NEXT:    s_mov_b32 s1, exec_lo
+; GFX11-NEXT:  .LBB8_1: ; =>This Inner Loop Header: Depth=1
+; GFX11-NEXT:    v_readfirstlane_b32 s4, v12
+; GFX11-NEXT:    v_readfirstlane_b32 s5, v13
+; GFX11-NEXT:    v_readfirstlane_b32 s6, v14
+; GFX11-NEXT:    v_readfirstlane_b32 s7, v15
+; GFX11-NEXT:    v_cmp_eq_u64_e32 vcc_lo, s[4:5], v[12:13]
+; GFX11-NEXT:    v_cmp_eq_u64_e64 s0, s[6:7], v[14:15]
+; GFX11-NEXT:    s_and_b32 s0, vcc_lo, s0
+; GFX11-NEXT:    s_and_saveexec_b32 s0, s0
+; GFX11-NEXT:    image_bvh64_intersect_ray v[0:3], [v[19:20], v21, v[16:18], v[6:8], v[9:11]], s[4:7]
+; GFX11-NEXT:    ; implicit-def: $vgpr12
+; GFX11-NEXT:    ; implicit-def: $vgpr19_vgpr20
+; GFX11-NEXT:    ; implicit-def: $vgpr21
+; GFX11-NEXT:    ; implicit-def: $vgpr16_vgpr17_vgpr18
+; GFX11-NEXT:    ; implicit-def: $vgpr6_vgpr7_vgpr8
+; GFX11-NEXT:    ; implicit-def: $vgpr9_vgpr10_vgpr11
+; GFX11-NEXT:    ; implicit-def: $vgpr12_vgpr13_vgpr14_vgpr15
+; GFX11-NEXT:    s_xor_b32 exec_lo, exec_lo, s0
+; GFX11-NEXT:    s_cbranch_execnz .LBB8_1
+; GFX11-NEXT:  ; %bb.2:
+; GFX11-NEXT:    s_mov_b32 exec_lo, s1
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    ; return to shader part epilog
   %v = call <4 x i32> @llvm.amdgcn.image.bvh.intersect.ray.i64.v4f32(i64 %node_ptr, float %ray_extent, <3 x float> %ray_origin, <3 x float> %ray_dir, <3 x float> %ray_inv_dir, <4 x i32> %tdescr)
   %r = bitcast <4 x i32> %v to <4 x float>
   ret <4 x float> %r
 }
 
 define amdgpu_ps <4 x float> @image_bvh64_intersect_ray_a16_vgpr_descr(i64 %node_ptr, float %ray_extent, <3 x float> %ray_origin, <3 x half> %ray_dir, <3 x half> %ray_inv_dir, <4 x i32> %tdescr) {
 ; GFX1030-LABEL: image_bvh64_intersect_ray_a16_vgpr_descr:
 ; GFX1030:       ; %bb.0:

 ; GFX1013-NEXT:  ; %bb.2:
 ; GFX1013-NEXT:    s_mov_b32 exec_lo, s1
 ; GFX1013-NEXT:    s_waitcnt vmcnt(0)
 ; GFX1013-NEXT:    v_mov_b32_e32 v0, v20
 ; GFX1013-NEXT:    v_mov_b32_e32 v1, v21
 ; GFX1013-NEXT:    v_mov_b32_e32 v2, v22
 ; GFX1013-NEXT:    v_mov_b32_e32 v3, v23
 ; GFX1013-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: image_bvh64_intersect_ray_a16_vgpr_descr:
+; GFX11:       ; %bb.0:
+; GFX11-NEXT:    v_mov_b32_e32 v14, v0
+; GFX11-NEXT:    v_lshrrev_b32_e32 v0, 16, v6
+; GFX11-NEXT:    v_mov_b32_e32 v15, v1
+; GFX11-NEXT:    v_mov_b32_e32 v16, v2
+; GFX11-NEXT:    v_mov_b32_e32 v17, v3
+; GFX11-NEXT:    v_lshrrev_b32_e32 v1, 16, v8
+; GFX11-NEXT:    v_lshlrev_b32_e32 v2, 16, v6
+; GFX11-NEXT:    v_lshlrev_b32_e32 v0, 16, v0
+; GFX11-NEXT:    v_lshlrev_b32_e32 v3, 16, v7
+; GFX11-NEXT:    v_mov_b32_e32 v18, v4
+; GFX11-NEXT:    v_mov_b32_e32 v19, v5
+; GFX11-NEXT:    v_and_or_b32 v4, 0xffff, v8, v2
+; GFX11-NEXT:    v_and_or_b32 v5, 0xffff, v1, v0
+; GFX11-NEXT:    v_and_or_b32 v6, 0xffff, v9, v3
+; GFX11-NEXT:    s_mov_b32 s1, exec_lo
+; GFX11-NEXT:  .LBB9_1: ; =>This Inner Loop Header: Depth=1
+; GFX11-NEXT:    v_readfirstlane_b32 s4, v10
+; GFX11-NEXT:    v_readfirstlane_b32 s5, v11
+; GFX11-NEXT:    v_readfirstlane_b32 s6, v12
+; GFX11-NEXT:    v_readfirstlane_b32 s7, v13
+; GFX11-NEXT:    v_cmp_eq_u64_e32 vcc_lo, s[4:5], v[10:11]
+; GFX11-NEXT:    v_cmp_eq_u64_e64 s0, s[6:7], v[12:13]
+; GFX11-NEXT:    s_and_b32 s0, vcc_lo, s0
+; GFX11-NEXT:    s_and_saveexec_b32 s0, s0
+; GFX11-NEXT:    image_bvh64_intersect_ray v[0:3], [v[14:15], v16, v[17:19], v[4:6]], s[4:7] a16
+; GFX11-NEXT:    ; implicit-def: $vgpr10
+; GFX11-NEXT:    ; implicit-def: $vgpr14_vgpr15
+; GFX11-NEXT:    ; implicit-def: $vgpr16
+; GFX11-NEXT:    ; implicit-def: $vgpr17_vgpr18_vgpr19
+; GFX11-NEXT:    ; implicit-def: $vgpr4_vgpr5_vgpr6
+; GFX11-NEXT:    ; implicit-def: $vgpr10_vgpr11_vgpr12_vgpr13
+; GFX11-NEXT:    s_xor_b32 exec_lo, exec_lo, s0
+; GFX11-NEXT:    s_cbranch_execnz .LBB9_1
+; GFX11-NEXT:  ; %bb.2:
+; GFX11-NEXT:    s_mov_b32 exec_lo, s1
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    ; return to shader part epilog
   %v = call <4 x i32> @llvm.amdgcn.image.bvh.intersect.ray.i64.v4f16(i64 %node_ptr, float %ray_extent, <3 x float> %ray_origin, <3 x half> %ray_dir, <3 x half> %ray_inv_dir, <4 x i32> %tdescr)
   %r = bitcast <4 x i32> %v to <4 x float>
   ret <4 x float> %r
 }
 
 define amdgpu_kernel void @image_bvh_intersect_ray_nsa_reassign(i32* %p_node_ptr, float* %p_ray, <4 x i32> inreg %tdescr) {
 ; GFX1030-LABEL: image_bvh_intersect_ray_nsa_reassign:
 ; GFX1030:       ; %bb.0:

 ; GFX1013-NEXT:    v_mov_b32_e32 v3, 1.0
 ; GFX1013-NEXT:    v_mov_b32_e32 v4, 2.0
 ; GFX1013-NEXT:    v_mov_b32_e32 v5, 0x40400000
 ; GFX1013-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
 ; GFX1013-NEXT:    image_bvh_intersect_ray v[0:3], v[0:15], s[8:11]
 ; GFX1013-NEXT:    s_waitcnt vmcnt(0)
 ; GFX1013-NEXT:    flat_store_dwordx4 v[0:1], v[0:3]
 ; GFX1013-NEXT:    s_endpgm
+;
+; GFX11-LABEL: image_bvh_intersect_ray_nsa_reassign:
+; GFX11:       ; %bb.0:
+; GFX11-NEXT:    s_clause 0x1
+; GFX11-NEXT:    s_load_b128 s[4:7], s[0:1], 0x24
+; GFX11-NEXT:    s_load_b128 s[0:3], s[0:1], 0x34
+; GFX11-NEXT:    v_lshlrev_b32_e32 v4, 2, v0
+; GFX11-NEXT:    s_mov_b32 s9, 0x40400000
+; GFX11-NEXT:    s_mov_b32 s12, 0x40c00000
+; GFX11-NEXT:    s_mov_b32 s8, 2.0
+; GFX11-NEXT:    s_mov_b32 s11, 0x40a00000
+; GFX11-NEXT:    s_mov_b32 s10, 4.0
+; GFX11-NEXT:    s_mov_b32 s14, 0x41000000
+; GFX11-NEXT:    s_mov_b32 s13, 0x40e00000
+; GFX11-NEXT:    v_mov_b32_e32 v6, s12
+; GFX11-NEXT:    v_mov_b32_e32 v7, s13
+; GFX11-NEXT:    v_mov_b32_e32 v8, s14
+; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
+; GFX11-NEXT:    v_mov_b32_e32 v0, s4
+; GFX11-NEXT:    v_mov_b32_e32 v1, s5
+; GFX11-NEXT:    v_mov_b32_e32 v2, s6
+; GFX11-NEXT:    v_mov_b32_e32 v3, s7
+; GFX11-NEXT:    s_mov_b32 s6, 0
+; GFX11-NEXT:    v_add_co_u32 v0, vcc_lo, v0, v4
+; GFX11-NEXT:    v_add_co_ci_u32_e32 v1, vcc_lo, 0, v1, vcc_lo
+; GFX11-NEXT:    v_add_co_u32 v2, vcc_lo, v2, v4
+; GFX11-NEXT:    v_add_co_ci_u32_e32 v3, vcc_lo, 0, v3, vcc_lo
+; GFX11-NEXT:    s_mov_b32 s7, 1.0
+; GFX11-NEXT:    flat_load_b32 v9, v[0:1]
+; GFX11-NEXT:    flat_load_b32 v10, v[2:3]
+; GFX11-NEXT:    v_mov_b32_e32 v0, s6
+; GFX11-NEXT:    v_mov_b32_e32 v3, s9
+; GFX11-NEXT:    v_mov_b32_e32 v1, s7
+; GFX11-NEXT:    v_mov_b32_e32 v2, s8
+; GFX11-NEXT:    v_mov_b32_e32 v4, s10
+; GFX11-NEXT:    v_mov_b32_e32 v5, s11
+; GFX11-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
+; GFX11-NEXT:    image_bvh_intersect_ray v[0:3], [v9, v10, v[0:2], v[3:5], v[6:8]], s[0:3]
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    flat_store_b128 v[0:1], v[0:3]
+; GFX11-NEXT:    s_endpgm
   %lid = tail call i32 @llvm.amdgcn.workitem.id.x()
   %gep_node_ptr = getelementptr inbounds i32, i32* %p_node_ptr, i32 %lid
   %node_ptr = load i32, i32* %gep_node_ptr, align 4
   %gep_ray = getelementptr inbounds float, float* %p_ray, i32 %lid
   %ray_extent = load float, float* %gep_ray, align 4
   %ray_origin0 = insertelement <3 x float> undef, float 0.0, i32 0
   %ray_origin1 = insertelement <3 x float> %ray_origin0, float 1.0, i32 1
   %ray_origin = insertelement <3 x float> %ray_origin1, float 2.0, i32 2

 ; GFX1013-NEXT:    v_mov_b32_e32 v5, s1
 ; GFX1013-NEXT:    v_mov_b32_e32 v6, s0
 ; GFX1013-NEXT:    v_mov_b32_e32 v7, s2
 ; GFX1013-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
 ; GFX1013-NEXT:    image_bvh_intersect_ray v[0:3], v[0:7], s[8:11] a16
 ; GFX1013-NEXT:    s_waitcnt vmcnt(0)
 ; GFX1013-NEXT:    flat_store_dwordx4 v[0:1], v[0:3]
 ; GFX1013-NEXT:    s_endpgm
+;
+; GFX11-LABEL: image_bvh_intersect_ray_a16_nsa_reassign:
+; GFX11:       ; %bb.0:
+; GFX11-NEXT:    s_clause 0x1
+; GFX11-NEXT:    s_load_b128 s[4:7], s[0:1], 0x24
+; GFX11-NEXT:    s_load_b128 s[0:3], s[0:1], 0x34
+; GFX11-NEXT:    v_lshlrev_b32_e32 v4, 2, v0
+; GFX11-NEXT:    s_mov_b32 s9, 0x42004600
+; GFX11-NEXT:    s_mov_b32 s8, 2.0
+; GFX11-NEXT:    s_mov_b32 s10, 0x44004700
+; GFX11-NEXT:    s_mov_b32 s11, 0x45004800
+; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
+; GFX11-NEXT:    v_mov_b32_e32 v0, s4
+; GFX11-NEXT:    v_mov_b32_e32 v1, s5
+; GFX11-NEXT:    v_mov_b32_e32 v2, s6
+; GFX11-NEXT:    v_mov_b32_e32 v3, s7
+; GFX11-NEXT:    s_mov_b32 s6, 0
+; GFX11-NEXT:    v_add_co_u32 v0, vcc_lo, v0, v4
+; GFX11-NEXT:    v_add_co_ci_u32_e32 v1, vcc_lo, 0, v1, vcc_lo
+; GFX11-NEXT:    v_add_co_u32 v2, vcc_lo, v2, v4
+; GFX11-NEXT:    v_add_co_ci_u32_e32 v3, vcc_lo, 0, v3, vcc_lo
+; GFX11-NEXT:    s_mov_b32 s7, 1.0
+; GFX11-NEXT:    flat_load_b32 v6, v[0:1]
+; GFX11-NEXT:    flat_load_b32 v7, v[2:3]
+; GFX11-NEXT:    v_mov_b32_e32 v0, s6
+; GFX11-NEXT:    v_mov_b32_e32 v3, s9
+; GFX11-NEXT:    v_mov_b32_e32 v1, s7
+; GFX11-NEXT:    v_mov_b32_e32 v2, s8
+; GFX11-NEXT:    v_mov_b32_e32 v4, s10
+; GFX11-NEXT:    v_mov_b32_e32 v5, s11
+; GFX11-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
+; GFX11-NEXT:    image_bvh_intersect_ray v[0:3], [v6, v7, v[0:2], v[3:5]], s[0:3] a16
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    flat_store_b128 v[0:1], v[0:3]
+; GFX11-NEXT:    s_endpgm
   %lid = tail call i32 @llvm.amdgcn.workitem.id.x()
   %gep_node_ptr = getelementptr inbounds i32, i32* %p_node_ptr, i32 %lid
   %node_ptr = load i32, i32* %gep_node_ptr, align 4
   %gep_ray = getelementptr inbounds float, float* %p_ray, i32 %lid
   %ray_extent = load float, float* %gep_ray, align 4
   %ray_origin0 = insertelement <3 x float> undef, float 0.0, i32 0
   %ray_origin1 = insertelement <3 x float> %ray_origin0, float 1.0, i32 1
   %ray_origin = insertelement <3 x float> %ray_origin1, float 2.0, i32 2

 ; GFX1013-NEXT:    flat_load_dword v2, v[0:1]
 ; GFX1013-NEXT:    v_mov_b32_e32 v0, 0xb36211c7
 ; GFX1013-NEXT:    v_mov_b32_e32 v1, 0x102
 ; GFX1013-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
 ; GFX1013-NEXT:    image_bvh64_intersect_ray v[0:3], v[0:15], s[4:7]
 ; GFX1013-NEXT:    s_waitcnt vmcnt(0)
 ; GFX1013-NEXT:    flat_store_dwordx4 v[0:1], v[0:3]
 ; GFX1013-NEXT:    s_endpgm
+;
+; GFX11-LABEL: image_bvh64_intersect_ray_nsa_reassign:
+; GFX11:       ; %bb.0:
+; GFX11-NEXT:    s_clause 0x1
+; GFX11-NEXT:    s_load_b64 s[4:5], s[0:1], 0x24
+; GFX11-NEXT:    s_load_b128 s[0:3], s[0:1], 0x34
+; GFX11-NEXT:    v_lshlrev_b32_e32 v2, 2, v0
+; GFX11-NEXT:    s_mov_b32 s6, 0
+; GFX11-NEXT:    s_mov_b32 s9, 0x40400000
+; GFX11-NEXT:    s_mov_b32 s12, 0x40c00000
+; GFX11-NEXT:    s_mov_b32 s8, 2.0
+; GFX11-NEXT:    s_mov_b32 s7, 1.0
+; GFX11-NEXT:    s_mov_b32 s11, 0x40a00000
+; GFX11-NEXT:    s_mov_b32 s10, 4.0
+; GFX11-NEXT:    s_mov_b32 s14, 0x41000000
+; GFX11-NEXT:    s_mov_b32 s13, 0x40e00000
+; GFX11-NEXT:    v_mov_b32_e32 v3, s9
+; GFX11-NEXT:    v_mov_b32_e32 v6, s12
+; GFX11-NEXT:    v_mov_b32_e32 v4, s10
+; GFX11-NEXT:    v_mov_b32_e32 v5, s11
+; GFX11-NEXT:    v_mov_b32_e32 v7, s13
+; GFX11-NEXT:    v_mov_b32_e32 v8, s14
+; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
+; GFX11-NEXT:    v_mov_b32_e32 v0, s4
+; GFX11-NEXT:    v_mov_b32_e32 v1, s5
+; GFX11-NEXT:    s_mov_b32 s4, 0xb36211c7
+; GFX11-NEXT:    s_movk_i32 s5, 0x102
+; GFX11-NEXT:    v_mov_b32_e32 v10, s5
+; GFX11-NEXT:    v_add_co_u32 v0, vcc_lo, v0, v2
+; GFX11-NEXT:    v_add_co_ci_u32_e32 v1, vcc_lo, 0, v1, vcc_lo
+; GFX11-NEXT:    v_mov_b32_e32 v9, s4
+; GFX11-NEXT:    flat_load_b32 v11, v[0:1]
+; GFX11-NEXT:    v_mov_b32_e32 v0, s6
+; GFX11-NEXT:    v_mov_b32_e32 v1, s7
+; GFX11-NEXT:    v_mov_b32_e32 v2, s8
+; GFX11-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
+; GFX11-NEXT:    image_bvh64_intersect_ray v[0:3], [v[9:10], v11, v[0:2], v[3:5], v[6:8]], s[0:3]
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    flat_store_b128 v[0:1], v[0:3]
+; GFX11-NEXT:    s_endpgm
   %lid = tail call i32 @llvm.amdgcn.workitem.id.x()
   %gep_ray = getelementptr inbounds float, float* %p_ray, i32 %lid
   %ray_extent = load float, float* %gep_ray, align 4
   %ray_origin0 = insertelement <3 x float> undef, float 0.0, i32 0
   %ray_origin1 = insertelement <3 x float> %ray_origin0, float 1.0, i32 1
   %ray_origin = insertelement <3 x float> %ray_origin1, float 2.0, i32 2
   %ray_dir0 = insertelement <3 x float> undef, float 3.0, i32 0
   %ray_dir1 = insertelement <3 x float> %ray_dir0, float 4.0, i32 1

 ; GFX1013-NEXT:    v_mov_b32_e32 v6, s1
 ; GFX1013-NEXT:    v_mov_b32_e32 v7, s0
 ; GFX1013-NEXT:    v_mov_b32_e32 v8, s2
 ; GFX1013-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
 ; GFX1013-NEXT:    image_bvh64_intersect_ray v[0:3], v[0:15], s[4:7] a16
 ; GFX1013-NEXT:    s_waitcnt vmcnt(0)
 ; GFX1013-NEXT:    flat_store_dwordx4 v[0:1], v[0:3]
 ; GFX1013-NEXT:    s_endpgm
+;
+; GFX11-LABEL: image_bvh64_intersect_ray_a16_nsa_reassign:
+; GFX11:       ; %bb.0:
+; GFX11-NEXT:    s_clause 0x1
+; GFX11-NEXT:    s_load_b64 s[4:5], s[0:1], 0x24
+; GFX11-NEXT:    s_load_b128 s[0:3], s[0:1], 0x34
+; GFX11-NEXT:    v_lshlrev_b32_e32 v2, 2, v0
+; GFX11-NEXT:    s_mov_b32 s6, 0
+; GFX11-NEXT:    s_mov_b32 s9, 0x42004600
+; GFX11-NEXT:    s_mov_b32 s8, 2.0
+; GFX11-NEXT:    s_mov_b32 s7, 1.0
+; GFX11-NEXT:    s_mov_b32 s10, 0x44004700
+; GFX11-NEXT:    s_mov_b32 s11, 0x45004800
+; GFX11-NEXT:    v_mov_b32_e32 v3, s9
+; GFX11-NEXT:    v_mov_b32_e32 v4, s10
+; GFX11-NEXT:    v_mov_b32_e32 v5, s11
+; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
+; GFX11-NEXT:    v_mov_b32_e32 v0, s4
+; GFX11-NEXT:    v_mov_b32_e32 v1, s5
+; GFX11-NEXT:    s_mov_b32 s4, 0xb36211c6
+; GFX11-NEXT:    s_movk_i32 s5, 0x102
+; GFX11-NEXT:    v_mov_b32_e32 v7, s5
+; GFX11-NEXT:    v_add_co_u32 v0, vcc_lo, v0, v2
+; GFX11-NEXT:    v_add_co_ci_u32_e32 v1, vcc_lo, 0, v1, vcc_lo
+; GFX11-NEXT:    v_mov_b32_e32 v6, s4
+; GFX11-NEXT:    flat_load_b32 v8, v[0:1]
+; GFX11-NEXT:    v_mov_b32_e32 v0, s6
+; GFX11-NEXT:    v_mov_b32_e32 v1, s7
+; GFX11-NEXT:    v_mov_b32_e32 v2, s8
+; GFX11-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
+; GFX11-NEXT:    image_bvh64_intersect_ray v[0:3], [v[6:7], v8, v[0:2], v[3:5]], s[0:3] a16
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    flat_store_b128 v[0:1], v[0:3]
+; GFX11-NEXT:    s_endpgm
   %lid = tail call i32 @llvm.amdgcn.workitem.id.x()
   %gep_ray = getelementptr inbounds float, float* %p_ray, i32 %lid
   %ray_extent = load float, float* %gep_ray, align 4
   %ray_origin0 = insertelement <3 x float> undef, float 0.0, i32 0
   %ray_origin1 = insertelement <3 x float> %ray_origin0, float 1.0, i32 1
   %ray_origin = insertelement <3 x float> %ray_origin1, float 2.0, i32 2
   %ray_dir0 = insertelement <3 x half> undef, half 3.0, i32 0
   %ray_dir1 = insertelement <3 x half> %ray_dir0, half 4.0, i32 1
   %ray_dir = insertelement <3 x half> %ray_dir1, half 5.0, i32 2
   %ray_inv_dir0 = insertelement <3 x half> undef, half 6.0, i32 0
   %ray_inv_dir1 = insertelement <3 x half> %ray_inv_dir0, half 7.0, i32 1
   %ray_inv_dir = insertelement <3 x half> %ray_inv_dir1, half 8.0, i32 2
   %v = call <4 x i32> @llvm.amdgcn.image.bvh.intersect.ray.i64.v4f16(i64 1111111111110, float %ray_extent, <3 x float> %ray_origin, <3 x half> %ray_dir, <3 x half> %ray_inv_dir, <4 x i32> %tdescr)
   store <4 x i32> %v, <4 x i32>* undef
   ret void
 }

llvm/test/CodeGen/AMDGPU/image-load-d16-tfe.ll

 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
 ; RUN: llc -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx900 -verify-machineinstrs < %s | FileCheck -check-prefix=GFX9 %s
 ; RUN: llc -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx1010 -verify-machineinstrs < %s | FileCheck -check-prefix=GFX10 %s
+; RUN: llc -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx1100 -verify-machineinstrs < %s | FileCheck -check-prefix=GFX11 %s
 ; RUN: llc -mtriple=amdgcn-mesa-mesa3d -mcpu=tonga -verify-machineinstrs < %s | FileCheck -check-prefix=GFX8-UNPACKED %s
 
 define amdgpu_ps void @load_1d_f16_tfe_dmask0(<8 x i32> inreg %rsrc, i32 %s) {
 ; GFX9-LABEL: load_1d_f16_tfe_dmask0:
 ; GFX9:       ; %bb.0:
 ; GFX9-NEXT:    v_mov_b32_e32 v1, 0
 ; GFX9-NEXT:    s_mov_b32 s11, s9
 ; GFX9-NEXT:    s_mov_b32 s10, s8

 ; GFX10-NEXT:    image_load v[1:2], v0, s[4:11] dmask:0x1 dim:SQ_RSRC_IMG_1D unorm tfe d16
 ; GFX10-NEXT:    s_waitcnt vmcnt(0)
 ; GFX10-NEXT:    global_store_short v[0:1], v1, off
 ; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
 ; GFX10-NEXT:    global_store_dword v[0:1], v2, off
 ; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
 ; GFX10-NEXT:    s_endpgm
 ;
+; GFX11-LABEL: load_1d_f16_tfe_dmask0:
+; GFX11:       ; %bb.0:
+; GFX11-NEXT:    v_mov_b32_e32 v1, 0
+; GFX11-NEXT:    s_mov_b32 s11, s9
+; GFX11-NEXT:    s_mov_b32 s10, s8
+; GFX11-NEXT:    s_mov_b32 s9, s7
+; GFX11-NEXT:    s_mov_b32 s8, s6
+; GFX11-NEXT:    s_mov_b32 s7, s5
+; GFX11-NEXT:    s_mov_b32 s6, s4
+; GFX11-NEXT:    s_mov_b32 s5, s3
+; GFX11-NEXT:    s_mov_b32 s4, s2
+; GFX11-NEXT:    v_mov_b32_e32 v2, v1
+; GFX11-NEXT:    image_load v[1:2], v0, s[4:11] dmask:0x1 dim:SQ_RSRC_IMG_1D unorm tfe d16
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    global_store_b16 v[0:1], v1, off dlc
+; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX11-NEXT:    global_store_b32 v[0:1], v2, off dlc
+; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX11-NEXT:    s_endpgm
+;
 ; GFX8-UNPACKED-LABEL: load_1d_f16_tfe_dmask0:
 ; GFX8-UNPACKED:       ; %bb.0:
 ; GFX8-UNPACKED-NEXT:    v_mov_b32_e32 v1, 0
 ; GFX8-UNPACKED-NEXT:    s_mov_b32 s11, s9
 ; GFX8-UNPACKED-NEXT:    s_mov_b32 s10, s8
 ; GFX8-UNPACKED-NEXT:    s_mov_b32 s9, s7
 ; GFX8-UNPACKED-NEXT:    s_mov_b32 s8, s6
 ; GFX8-UNPACKED-NEXT:    s_mov_b32 s7, s5

 ; GFX10-NEXT:    image_load v[1:2], v0, s[4:11] dmask:0x1 dim:SQ_RSRC_IMG_1D unorm tfe d16
 ; GFX10-NEXT:    s_waitcnt vmcnt(0)
 ; GFX10-NEXT:    global_store_short v[0:1], v1, off
 ; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
 ; GFX10-NEXT:    global_store_dword v[0:1], v2, off
 ; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
 ; GFX10-NEXT:    s_endpgm
 ;
+; GFX11-LABEL: load_1d_f16_tfe_dmask1:
+; GFX11:       ; %bb.0:
+; GFX11-NEXT:    v_mov_b32_e32 v1, 0
+; GFX11-NEXT:    s_mov_b32 s11, s9
+; GFX11-NEXT:    s_mov_b32 s10, s8
+; GFX11-NEXT:    s_mov_b32 s9, s7
+; GFX11-NEXT:    s_mov_b32 s8, s6
+; GFX11-NEXT:    s_mov_b32 s7, s5
+; GFX11-NEXT:    s_mov_b32 s6, s4
+; GFX11-NEXT:    s_mov_b32 s5, s3
+; GFX11-NEXT:    s_mov_b32 s4, s2
+; GFX11-NEXT:    v_mov_b32_e32 v2, v1
+; GFX11-NEXT:    image_load v[1:2], v0, s[4:11] dmask:0x1 dim:SQ_RSRC_IMG_1D unorm tfe d16
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    global_store_b16 v[0:1], v1, off dlc
+; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX11-NEXT:    global_store_b32 v[0:1], v2, off dlc
+; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX11-NEXT:    s_endpgm
+;
 ; GFX8-UNPACKED-LABEL: load_1d_f16_tfe_dmask1:
 ; GFX8-UNPACKED:       ; %bb.0:
 ; GFX8-UNPACKED-NEXT:    v_mov_b32_e32 v1, 0
 ; GFX8-UNPACKED-NEXT:    s_mov_b32 s11, s9
 ; GFX8-UNPACKED-NEXT:    s_mov_b32 s10, s8
 ; GFX8-UNPACKED-NEXT:    s_mov_b32 s9, s7
 ; GFX8-UNPACKED-NEXT:    s_mov_b32 s8, s6
 ; GFX8-UNPACKED-NEXT:    s_mov_b32 s7, s5

 ; GFX10-NEXT:    image_load v[1:2], v0, s[4:11] dmask:0x1 dim:SQ_RSRC_IMG_1D unorm tfe d16
 ; GFX10-NEXT:    s_waitcnt vmcnt(0)
 ; GFX10-NEXT:    global_store_dword v[0:1], v1, off
 ; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
 ; GFX10-NEXT:    global_store_dword v[0:1], v2, off
 ; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
 ; GFX10-NEXT:    s_endpgm
 ;
+; GFX11-LABEL: load_1d_v2f16_tfe_dmask0:
+; GFX11:       ; %bb.0:
+; GFX11-NEXT:    v_mov_b32_e32 v1, 0
+; GFX11-NEXT:    s_mov_b32 s11, s9
+; GFX11-NEXT:    s_mov_b32 s10, s8
+; GFX11-NEXT:    s_mov_b32 s9, s7
+; GFX11-NEXT:    s_mov_b32 s8, s6
+; GFX11-NEXT:    s_mov_b32 s7, s5
+; GFX11-NEXT:    s_mov_b32 s6, s4
+; GFX11-NEXT:    s_mov_b32 s5, s3
+; GFX11-NEXT:    s_mov_b32 s4, s2
+; GFX11-NEXT:    v_mov_b32_e32 v2, v1
+; GFX11-NEXT:    image_load v[1:2], v0, s[4:11] dmask:0x1 dim:SQ_RSRC_IMG_1D unorm tfe d16
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    global_store_b32 v[0:1], v1, off dlc
+; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX11-NEXT:    global_store_b32 v[0:1], v2, off dlc
+; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX11-NEXT:    s_endpgm
+;
 ; GFX8-UNPACKED-LABEL: load_1d_v2f16_tfe_dmask0:
 ; GFX8-UNPACKED:       ; %bb.0:
 ; GFX8-UNPACKED-NEXT:    v_mov_b32_e32 v1, 0
 ; GFX8-UNPACKED-NEXT:    s_mov_b32 s11, s9
 ; GFX8-UNPACKED-NEXT:    s_mov_b32 s10, s8
 ; GFX8-UNPACKED-NEXT:    s_mov_b32 s9, s7
 ; GFX8-UNPACKED-NEXT:    s_mov_b32 s8, s6
 ; GFX8-UNPACKED-NEXT:    s_mov_b32 s7, s5

 ; GFX10-NEXT:    image_load v[1:2], v0, s[4:11] dmask:0x1 dim:SQ_RSRC_IMG_1D unorm tfe d16
 ; GFX10-NEXT:    s_waitcnt vmcnt(0)
 ; GFX10-NEXT:    global_store_dword v[0:1], v1, off
 ; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
 ; GFX10-NEXT:    global_store_dword v[0:1], v2, off
 ; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
 ; GFX10-NEXT:    s_endpgm
 ;
+; GFX11-LABEL: load_1d_v2f16_tfe_dmask1:
+; GFX11:       ; %bb.0:
+; GFX11-NEXT:    v_mov_b32_e32 v1, 0
+; GFX11-NEXT:    s_mov_b32 s11, s9
+; GFX11-NEXT:    s_mov_b32 s10, s8
+; GFX11-NEXT:    s_mov_b32 s9, s7
+; GFX11-NEXT:    s_mov_b32 s8, s6
+; GFX11-NEXT:    s_mov_b32 s7, s5
+; GFX11-NEXT:    s_mov_b32 s6, s4
+; GFX11-NEXT:    s_mov_b32 s5, s3
+; GFX11-NEXT:    s_mov_b32 s4, s2
+; GFX11-NEXT:    v_mov_b32_e32 v2, v1
+; GFX11-NEXT:    image_load v[1:2], v0, s[4:11] dmask:0x1 dim:SQ_RSRC_IMG_1D unorm tfe d16
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    global_store_b32 v[0:1], v1, off dlc
+; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX11-NEXT:    global_store_b32 v[0:1], v2, off dlc
+; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX11-NEXT:    s_endpgm
+;
 ; GFX8-UNPACKED-LABEL: load_1d_v2f16_tfe_dmask1:
 ; GFX8-UNPACKED:       ; %bb.0:
 ; GFX8-UNPACKED-NEXT:    v_mov_b32_e32 v1, 0
 ; GFX8-UNPACKED-NEXT:    s_mov_b32 s11, s9
 ; GFX8-UNPACKED-NEXT:    s_mov_b32 s10, s8
 ; GFX8-UNPACKED-NEXT:    s_mov_b32 s9, s7
 ; GFX8-UNPACKED-NEXT:    s_mov_b32 s8, s6
 ; GFX8-UNPACKED-NEXT:    s_mov_b32 s7, s5

 ; GFX10-NEXT:    image_load v[1:2], v0, s[4:11] dmask:0x3 dim:SQ_RSRC_IMG_1D unorm tfe d16
 ; GFX10-NEXT:    s_waitcnt vmcnt(0)
 ; GFX10-NEXT:    global_store_dword v[0:1], v1, off
 ; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
 ; GFX10-NEXT:    global_store_dword v[0:1], v2, off
 ; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
 ; GFX10-NEXT:    s_endpgm
 ;
+; GFX11-LABEL: load_1d_v2f16_tfe_dmask3:
+; GFX11:       ; %bb.0:
+; GFX11-NEXT:    v_mov_b32_e32 v1, 0
+; GFX11-NEXT:    s_mov_b32 s11, s9
+; GFX11-NEXT:    s_mov_b32 s10, s8
+; GFX11-NEXT:    s_mov_b32 s9, s7
+; GFX11-NEXT:    s_mov_b32 s8, s6
+; GFX11-NEXT:    s_mov_b32 s7, s5
+; GFX11-NEXT:    s_mov_b32 s6, s4
+; GFX11-NEXT:    s_mov_b32 s5, s3
+; GFX11-NEXT:    s_mov_b32 s4, s2
+; GFX11-NEXT:    v_mov_b32_e32 v2, v1
+; GFX11-NEXT:    image_load v[1:2], v0, s[4:11] dmask:0x3 dim:SQ_RSRC_IMG_1D unorm tfe d16
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    global_store_b32 v[0:1], v1, off dlc
+; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX11-NEXT:    global_store_b32 v[0:1], v2, off dlc
+; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX11-NEXT:    s_endpgm
+;
 ; GFX8-UNPACKED-LABEL: load_1d_v2f16_tfe_dmask3:
 ; GFX8-UNPACKED:       ; %bb.0:
 ; GFX8-UNPACKED-NEXT:    v_mov_b32_e32 v1, 0
 ; GFX8-UNPACKED-NEXT:    s_mov_b32 s11, s9
 ; GFX8-UNPACKED-NEXT:    s_mov_b32 s10, s8
 ; GFX8-UNPACKED-NEXT:    s_mov_b32 s9, s7
 ; GFX8-UNPACKED-NEXT:    s_mov_b32 s8, s6
 ; GFX8-UNPACKED-NEXT:    s_mov_b32 s7, s5

 ; GFX10-NEXT:    global_store_short v[0:1], v2, off
 ; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
 ; GFX10-NEXT:    global_store_dword v[0:1], v1, off
 ; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
 ; GFX10-NEXT:    global_store_dword v[0:1], v3, off
 ; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
 ; GFX10-NEXT:    s_endpgm
 ;
+; GFX11-LABEL: load_1d_v3f16_tfe_dmask7:
+; GFX11:       ; %bb.0:
+; GFX11-NEXT:    v_mov_b32_e32 v1, 0
+; GFX11-NEXT:    s_mov_b32 s11, s9
+; GFX11-NEXT:    s_mov_b32 s10, s8
+; GFX11-NEXT:    s_mov_b32 s9, s7
+; GFX11-NEXT:    s_mov_b32 s8, s6
+; GFX11-NEXT:    s_mov_b32 s7, s5
+; GFX11-NEXT:    s_mov_b32 s6, s4
+; GFX11-NEXT:    s_mov_b32 s5, s3
+; GFX11-NEXT:    s_mov_b32 s4, s2
+; GFX11-NEXT:    v_mov_b32_e32 v2, v1
+; GFX11-NEXT:    v_mov_b32_e32 v3, v1
+; GFX11-NEXT:    image_load v[1:3], v0, s[4:11] dmask:0x7 dim:SQ_RSRC_IMG_1D unorm tfe d16
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    global_store_b16 v[0:1], v2, off dlc
+; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX11-NEXT:    global_store_b32 v[0:1], v1, off dlc
+; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX11-NEXT:    global_store_b32 v[0:1], v3, off dlc
+; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX11-NEXT:    s_endpgm
+;
 ; GFX8-UNPACKED-LABEL: load_1d_v3f16_tfe_dmask7:
 ; GFX8-UNPACKED:       ; %bb.0:
 ; GFX8-UNPACKED-NEXT:    v_mov_b32_e32 v1, 0
 ; GFX8-UNPACKED-NEXT:    s_mov_b32 s11, s9
 ; GFX8-UNPACKED-NEXT:    s_mov_b32 s10, s8
 ; GFX8-UNPACKED-NEXT:    s_mov_b32 s9, s7
 ; GFX8-UNPACKED-NEXT:    s_mov_b32 s8, s6
 ; GFX8-UNPACKED-NEXT:    s_mov_b32 s7, s5

 ; GFX10-NEXT:    image_load v[1:3], v0, s[4:11] dmask:0xf dim:SQ_RSRC_IMG_1D unorm tfe d16
 ; GFX10-NEXT:    s_waitcnt vmcnt(0)
 ; GFX10-NEXT:    global_store_dwordx2 v[0:1], v[1:2], off
 ; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
 ; GFX10-NEXT:    global_store_dword v[0:1], v3, off
 ; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
 ; GFX10-NEXT:    s_endpgm
 ;
+; GFX11-LABEL: load_1d_v4f16_tfe_dmask15:
+; GFX11:       ; %bb.0:
+; GFX11-NEXT:    v_mov_b32_e32 v1, 0
+; GFX11-NEXT:    s_mov_b32 s11, s9
+; GFX11-NEXT:    s_mov_b32 s10, s8
+; GFX11-NEXT:    s_mov_b32 s9, s7
+; GFX11-NEXT:    s_mov_b32 s8, s6
+; GFX11-NEXT:    s_mov_b32 s7, s5
+; GFX11-NEXT:    s_mov_b32 s6, s4
+; GFX11-NEXT:    s_mov_b32 s5, s3
+; GFX11-NEXT:    s_mov_b32 s4, s2
+; GFX11-NEXT:    v_mov_b32_e32 v2, v1
+; GFX11-NEXT:    v_mov_b32_e32 v3, v1
+; GFX11-NEXT:    image_load v[1:3], v0, s[4:11] dmask:0xf dim:SQ_RSRC_IMG_1D unorm tfe d16
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    global_store_b64 v[0:1], v[1:2], off dlc
+; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX11-NEXT:    global_store_b32 v[0:1], v3, off dlc
+; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX11-NEXT:    s_endpgm
+;
 ; GFX8-UNPACKED-LABEL: load_1d_v4f16_tfe_dmask15:
 ; GFX8-UNPACKED:       ; %bb.0:
 ; GFX8-UNPACKED-NEXT:    v_mov_b32_e32 v1, 0
 ; GFX8-UNPACKED-NEXT:    s_mov_b32 s11, s9
 ; GFX8-UNPACKED-NEXT:    s_mov_b32 s10, s8
 ; GFX8-UNPACKED-NEXT:    s_mov_b32 s9, s7
 ; GFX8-UNPACKED-NEXT:    s_mov_b32 s8, s6
 ; GFX8-UNPACKED-NEXT:    s_mov_b32 s7, s5

llvm/test/CodeGen/AMDGPU/llvm.amdgcn.image.a16.dim.ll

 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
 ; RUN: llc -march=amdgcn -mcpu=gfx900 -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX9 %s
 ; RUN: llc -march=amdgcn -mcpu=gfx1010 -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX10 %s
+; RUN: llc -march=amdgcn -mcpu=gfx1100 -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX10 %s
 
 define amdgpu_ps <4 x float> @load_1d(<8 x i32> inreg %rsrc, <2 x i16> %coords) {
 ; GFX9-LABEL: load_1d:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    image_load v[0:3], v0, s[0:7] dmask:0xf unorm a16
 ; GFX9-NEXT:    s_waitcnt vmcnt(0)
 ; GFX9-NEXT:    ; return to shader part epilog
 ;

llvm/test/CodeGen/AMDGPU/llvm.amdgcn.image.a16.encode.ll

 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
 ; RUN: llc -march=amdgcn -mcpu=gfx900 -show-mc-encoding < %s | FileCheck -check-prefixes=GFX9 %s
 ; RUN: llc -march=amdgcn -mcpu=gfx1010 -show-mc-encoding < %s | FileCheck -check-prefixes=GFX10 %s
+; RUN: llc -march=amdgcn -mcpu=gfx1100 -show-mc-encoding < %s | FileCheck -check-prefixes=GFX11 %s
 
 define amdgpu_ps <4 x float> @load_1d(<8 x i32> inreg %rsrc, <2 x i16> %coords) {
 ; GFX9-LABEL: load_1d:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    image_load v[0:3], v0, s[0:7] dmask:0xf unorm a16 ; encoding: [0x00,0x9f,0x00,0xf0,0x00,0x00,0x00,0x00]
 ; GFX9-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x0f,0x8c,0xbf]
 ; GFX9-NEXT:    ; return to shader part epilog
 ;
 ; GFX10-LABEL: load_1d:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    image_load v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm a16 ; encoding: [0x00,0x1f,0x00,0xf0,0x00,0x00,0x00,0x40]
 ; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
 ; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: load_1d:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    image_load v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm a16 ; encoding: [0x80,0x0f,0x01,0xf0,0x00,0x00,0x00,0x00]
+; GFX11-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0xf7,0x03,0x89,0xbf]
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %s = extractelement <2 x i16> %coords, i32 0
   %v = call <4 x float> @llvm.amdgcn.image.load.1d.v4f32.i16(i32 15, i16 %s, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @load_2d(<8 x i32> inreg %rsrc, <2 x i16> %coords) {
 ; GFX9-LABEL: load_2d:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    image_load v[0:3], v0, s[0:7] dmask:0xf unorm a16 ; encoding: [0x00,0x9f,0x00,0xf0,0x00,0x00,0x00,0x00]
 ; GFX9-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x0f,0x8c,0xbf]
 ; GFX9-NEXT:    ; return to shader part epilog
 ;
 ; GFX10-LABEL: load_2d:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    image_load v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D unorm a16 ; encoding: [0x08,0x1f,0x00,0xf0,0x00,0x00,0x00,0x40]
 ; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
 ; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: load_2d:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    image_load v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D unorm a16 ; encoding: [0x84,0x0f,0x01,0xf0,0x00,0x00,0x00,0x00]
+; GFX11-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0xf7,0x03,0x89,0xbf]
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %s = extractelement <2 x i16> %coords, i32 0
   %t = extractelement <2 x i16> %coords, i32 1
   %v = call <4 x float> @llvm.amdgcn.image.load.2d.v4f32.i16(i32 15, i16 %s, i16 %t, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @load_3d(<8 x i32> inreg %rsrc, <2 x i16> %coords_lo, <2 x i16> %coords_hi) {
 ; GFX9-LABEL: load_3d:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    image_load v[0:3], v[0:1], s[0:7] dmask:0xf unorm a16 ; encoding: [0x00,0x9f,0x00,0xf0,0x00,0x00,0x00,0x00]
 ; GFX9-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x0f,0x8c,0xbf]
 ; GFX9-NEXT:    ; return to shader part epilog
 ;
 ; GFX10-LABEL: load_3d:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    image_load v[0:3], v[0:1], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_3D unorm a16 ; encoding: [0x10,0x1f,0x00,0xf0,0x00,0x00,0x00,0x40]
 ; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
 ; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: load_3d:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    image_load v[0:3], v[0:1], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_3D unorm a16 ; encoding: [0x88,0x0f,0x01,0xf0,0x00,0x00,0x00,0x00]
+; GFX11-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0xf7,0x03,0x89,0xbf]
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %s = extractelement <2 x i16> %coords_lo, i32 0
   %t = extractelement <2 x i16> %coords_lo, i32 1
   %r = extractelement <2 x i16> %coords_hi, i32 0
   %v = call <4 x float> @llvm.amdgcn.image.load.3d.v4f32.i16(i32 15, i16 %s, i16 %t, i16 %r, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @load_cube(<8 x i32> inreg %rsrc, <2 x i16> %coords_lo, <2 x i16> %coords_hi) {
 ; GFX9-LABEL: load_cube:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    image_load v[0:3], v[0:1], s[0:7] dmask:0xf unorm a16 da ; encoding: [0x00,0xdf,0x00,0xf0,0x00,0x00,0x00,0x00]
 ; GFX9-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x0f,0x8c,0xbf]
 ; GFX9-NEXT:    ; return to shader part epilog
 ;
 ; GFX10-LABEL: load_cube:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    image_load v[0:3], v[0:1], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_CUBE unorm a16 ; encoding: [0x18,0x1f,0x00,0xf0,0x00,0x00,0x00,0x40]
 ; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
 ; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: load_cube:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    image_load v[0:3], v[0:1], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_CUBE unorm a16 ; encoding: [0x8c,0x0f,0x01,0xf0,0x00,0x00,0x00,0x00]
+; GFX11-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0xf7,0x03,0x89,0xbf]
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %s = extractelement <2 x i16> %coords_lo, i32 0
   %t = extractelement <2 x i16> %coords_lo, i32 1
   %slice = extractelement <2 x i16> %coords_hi, i32 0
   %v = call <4 x float> @llvm.amdgcn.image.load.cube.v4f32.i16(i32 15, i16 %s, i16 %t, i16 %slice, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @load_1darray(<8 x i32> inreg %rsrc, <2 x i16> %coords) {
 ; GFX9-LABEL: load_1darray:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    image_load v[0:3], v0, s[0:7] dmask:0xf unorm a16 da ; encoding: [0x00,0xdf,0x00,0xf0,0x00,0x00,0x00,0x00]
 ; GFX9-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x0f,0x8c,0xbf]
 ; GFX9-NEXT:    ; return to shader part epilog
 ;
 ; GFX10-LABEL: load_1darray:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    image_load v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D_ARRAY unorm a16 ; encoding: [0x20,0x1f,0x00,0xf0,0x00,0x00,0x00,0x40]
 ; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
 ; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: load_1darray:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    image_load v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D_ARRAY unorm a16 ; encoding: [0x90,0x0f,0x01,0xf0,0x00,0x00,0x00,0x00]
+; GFX11-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0xf7,0x03,0x89,0xbf]
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %s = extractelement <2 x i16> %coords, i32 0
   %slice = extractelement <2 x i16> %coords, i32 1
   %v = call <4 x float> @llvm.amdgcn.image.load.1darray.v4f32.i16(i32 15, i16 %s, i16 %slice, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @load_2darray(<8 x i32> inreg %rsrc, <2 x i16> %coords_lo, <2 x i16> %coords_hi) {
 ; GFX9-LABEL: load_2darray:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    image_load v[0:3], v[0:1], s[0:7] dmask:0xf unorm a16 da ; encoding: [0x00,0xdf,0x00,0xf0,0x00,0x00,0x00,0x00]
 ; GFX9-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x0f,0x8c,0xbf]
 ; GFX9-NEXT:    ; return to shader part epilog
 ;
 ; GFX10-LABEL: load_2darray:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    image_load v[0:3], v[0:1], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_ARRAY unorm a16 ; encoding: [0x28,0x1f,0x00,0xf0,0x00,0x00,0x00,0x40]
 ; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
 ; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: load_2darray:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    image_load v[0:3], v[0:1], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_ARRAY unorm a16 ; encoding: [0x94,0x0f,0x01,0xf0,0x00,0x00,0x00,0x00]
+; GFX11-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0xf7,0x03,0x89,0xbf]
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %s = extractelement <2 x i16> %coords_lo, i32 0
   %t = extractelement <2 x i16> %coords_lo, i32 1
   %slice = extractelement <2 x i16> %coords_hi, i32 0
   %v = call <4 x float> @llvm.amdgcn.image.load.2darray.v4f32.i16(i32 15, i16 %s, i16 %t, i16 %slice, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @load_2dmsaa(<8 x i32> inreg %rsrc, <2 x i16> %coords_lo, <2 x i16> %coords_hi) {
 ; GFX9-LABEL: load_2dmsaa:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    image_load v[0:3], v[0:1], s[0:7] dmask:0xf unorm a16 ; encoding: [0x00,0x9f,0x00,0xf0,0x00,0x00,0x00,0x00]
 ; GFX9-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x0f,0x8c,0xbf]
 ; GFX9-NEXT:    ; return to shader part epilog
 ;
 ; GFX10-LABEL: load_2dmsaa:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    image_load v[0:3], v[0:1], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_MSAA unorm a16 ; encoding: [0x30,0x1f,0x00,0xf0,0x00,0x00,0x00,0x40]
 ; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
 ; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: load_2dmsaa:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    image_load v[0:3], v[0:1], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_MSAA unorm a16 ; encoding: [0x98,0x0f,0x01,0xf0,0x00,0x00,0x00,0x00]
+; GFX11-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0xf7,0x03,0x89,0xbf]
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %s = extractelement <2 x i16> %coords_lo, i32 0
   %t = extractelement <2 x i16> %coords_lo, i32 1
   %fragid = extractelement <2 x i16> %coords_hi, i32 0
   %v = call <4 x float> @llvm.amdgcn.image.load.2dmsaa.v4f32.i16(i32 15, i16 %s, i16 %t, i16 %fragid, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @load_2darraymsaa(<8 x i32> inreg %rsrc, <2 x i16> %coords_lo, <2 x i16> %coords_hi) {
 ; GFX9-LABEL: load_2darraymsaa:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    image_load v[0:3], v[0:1], s[0:7] dmask:0xf unorm a16 da ; encoding: [0x00,0xdf,0x00,0xf0,0x00,0x00,0x00,0x00]
 ; GFX9-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x0f,0x8c,0xbf]
 ; GFX9-NEXT:    ; return to shader part epilog
 ;
 ; GFX10-LABEL: load_2darraymsaa:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    image_load v[0:3], v[0:1], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_MSAA_ARRAY unorm a16 ; encoding: [0x38,0x1f,0x00,0xf0,0x00,0x00,0x00,0x40]
 ; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
 ; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: load_2darraymsaa:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    image_load v[0:3], v[0:1], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_MSAA_ARRAY unorm a16 ; encoding: [0x9c,0x0f,0x01,0xf0,0x00,0x00,0x00,0x00]
+; GFX11-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0xf7,0x03,0x89,0xbf]
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %s = extractelement <2 x i16> %coords_lo, i32 0
   %t = extractelement <2 x i16> %coords_lo, i32 1
   %slice = extractelement <2 x i16> %coords_hi, i32 0
   %fragid = extractelement <2 x i16> %coords_hi, i32 1
   %v = call <4 x float> @llvm.amdgcn.image.load.2darraymsaa.v4f32.i16(i32 15, i16 %s, i16 %t, i16 %slice, i16 %fragid, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @load_mip_1d(<8 x i32> inreg %rsrc, <2 x i16> %coords) {
 ; GFX9-LABEL: load_mip_1d:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    image_load_mip v[0:3], v0, s[0:7] dmask:0xf unorm a16 ; encoding: [0x00,0x9f,0x04,0xf0,0x00,0x00,0x00,0x00]
 ; GFX9-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x0f,0x8c,0xbf]
 ; GFX9-NEXT:    ; return to shader part epilog
 ;
 ; GFX10-LABEL: load_mip_1d:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    image_load_mip v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm a16 ; encoding: [0x00,0x1f,0x04,0xf0,0x00,0x00,0x00,0x40]
 ; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
 ; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: load_mip_1d:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    image_load_mip v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm a16 ; encoding: [0x80,0x0f,0x05,0xf0,0x00,0x00,0x00,0x00]
+; GFX11-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0xf7,0x03,0x89,0xbf]
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %s = extractelement <2 x i16> %coords, i32 0
   %mip = extractelement <2 x i16> %coords, i32 1
   %v = call <4 x float> @llvm.amdgcn.image.load.mip.1d.v4f32.i16(i32 15, i16 %s, i16 %mip, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @load_mip_2d(<8 x i32> inreg %rsrc, <2 x i16> %coords_lo, <2 x i16> %coords_hi) {
 ; GFX9-LABEL: load_mip_2d:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    image_load_mip v[0:3], v[0:1], s[0:7] dmask:0xf unorm a16 ; encoding: [0x00,0x9f,0x04,0xf0,0x00,0x00,0x00,0x00]
 ; GFX9-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x0f,0x8c,0xbf]
 ; GFX9-NEXT:    ; return to shader part epilog
 ;
 ; GFX10-LABEL: load_mip_2d:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    image_load_mip v[0:3], v[0:1], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D unorm a16 ; encoding: [0x08,0x1f,0x04,0xf0,0x00,0x00,0x00,0x40]
 ; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
 ; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: load_mip_2d:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    image_load_mip v[0:3], v[0:1], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D unorm a16 ; encoding: [0x84,0x0f,0x05,0xf0,0x00,0x00,0x00,0x00]
+; GFX11-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0xf7,0x03,0x89,0xbf]
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %s = extractelement <2 x i16> %coords_lo, i32 0
   %t = extractelement <2 x i16> %coords_lo, i32 1
   %mip = extractelement <2 x i16> %coords_hi, i32 0
   %v = call <4 x float> @llvm.amdgcn.image.load.mip.2d.v4f32.i16(i32 15, i16 %s, i16 %t, i16 %mip, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @load_mip_3d(<8 x i32> inreg %rsrc, <2 x i16> %coords_lo, <2 x i16> %coords_hi) {
 ; GFX9-LABEL: load_mip_3d:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    image_load_mip v[0:3], v[0:1], s[0:7] dmask:0xf unorm a16 ; encoding: [0x00,0x9f,0x04,0xf0,0x00,0x00,0x00,0x00]
 ; GFX9-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x0f,0x8c,0xbf]
 ; GFX9-NEXT:    ; return to shader part epilog
 ;
 ; GFX10-LABEL: load_mip_3d:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    image_load_mip v[0:3], v[0:1], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_3D unorm a16 ; encoding: [0x10,0x1f,0x04,0xf0,0x00,0x00,0x00,0x40]
 ; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
 ; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: load_mip_3d:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    image_load_mip v[0:3], v[0:1], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_3D unorm a16 ; encoding: [0x88,0x0f,0x05,0xf0,0x00,0x00,0x00,0x00]
+; GFX11-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0xf7,0x03,0x89,0xbf]
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %s = extractelement <2 x i16> %coords_lo, i32 0
   %t = extractelement <2 x i16> %coords_lo, i32 1
   %r = extractelement <2 x i16> %coords_hi, i32 0
   %mip = extractelement <2 x i16> %coords_hi, i32 1
   %v = call <4 x float> @llvm.amdgcn.image.load.mip.3d.v4f32.i16(i32 15, i16 %s, i16 %t, i16 %r, i16 %mip, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @load_mip_cube(<8 x i32> inreg %rsrc, <2 x i16> %coords_lo, <2 x i16> %coords_hi) {
 ; GFX9-LABEL: load_mip_cube:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    image_load_mip v[0:3], v[0:1], s[0:7] dmask:0xf unorm a16 da ; encoding: [0x00,0xdf,0x04,0xf0,0x00,0x00,0x00,0x00]
 ; GFX9-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x0f,0x8c,0xbf]
 ; GFX9-NEXT:    ; return to shader part epilog
 ;
 ; GFX10-LABEL: load_mip_cube:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    image_load_mip v[0:3], v[0:1], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_CUBE unorm a16 ; encoding: [0x18,0x1f,0x04,0xf0,0x00,0x00,0x00,0x40]
 ; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
 ; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: load_mip_cube:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    image_load_mip v[0:3], v[0:1], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_CUBE unorm a16 ; encoding: [0x8c,0x0f,0x05,0xf0,0x00,0x00,0x00,0x00]
+; GFX11-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0xf7,0x03,0x89,0xbf]
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %s = extractelement <2 x i16> %coords_lo, i32 0
   %t = extractelement <2 x i16> %coords_lo, i32 1
   %slice = extractelement <2 x i16> %coords_hi, i32 0
   %mip = extractelement <2 x i16> %coords_hi, i32 1
   %v = call <4 x float> @llvm.amdgcn.image.load.mip.cube.v4f32.i16(i32 15, i16 %s, i16 %t, i16 %slice, i16 %mip, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @load_mip_1darray(<8 x i32> inreg %rsrc, <2 x i16> %coords_lo, <2 x i16> %coords_hi) {
 ; GFX9-LABEL: load_mip_1darray:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    image_load_mip v[0:3], v[0:1], s[0:7] dmask:0xf unorm a16 da ; encoding: [0x00,0xdf,0x04,0xf0,0x00,0x00,0x00,0x00]
 ; GFX9-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x0f,0x8c,0xbf]
 ; GFX9-NEXT:    ; return to shader part epilog
 ;
 ; GFX10-LABEL: load_mip_1darray:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    image_load_mip v[0:3], v[0:1], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D_ARRAY unorm a16 ; encoding: [0x20,0x1f,0x04,0xf0,0x00,0x00,0x00,0x40]
 ; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
 ; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: load_mip_1darray:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    image_load_mip v[0:3], v[0:1], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D_ARRAY unorm a16 ; encoding: [0x90,0x0f,0x05,0xf0,0x00,0x00,0x00,0x00]
+; GFX11-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0xf7,0x03,0x89,0xbf]
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %s = extractelement <2 x i16> %coords_lo, i32 0
   %slice = extractelement <2 x i16> %coords_lo, i32 1
   %mip = extractelement <2 x i16> %coords_hi, i32 0
   %v = call <4 x float> @llvm.amdgcn.image.load.mip.1darray.v4f32.i16(i32 15, i16 %s, i16 %slice, i16 %mip, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @load_mip_2darray(<8 x i32> inreg %rsrc, <2 x i16> %coords_lo, <2 x i16> %coords_hi) {
 ; GFX9-LABEL: load_mip_2darray:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    image_load_mip v[0:3], v[0:1], s[0:7] dmask:0xf unorm a16 da ; encoding: [0x00,0xdf,0x04,0xf0,0x00,0x00,0x00,0x00]
 ; GFX9-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x0f,0x8c,0xbf]
 ; GFX9-NEXT:    ; return to shader part epilog
 ;
 ; GFX10-LABEL: load_mip_2darray:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    image_load_mip v[0:3], v[0:1], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_ARRAY unorm a16 ; encoding: [0x28,0x1f,0x04,0xf0,0x00,0x00,0x00,0x40]
 ; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
 ; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: load_mip_2darray:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    image_load_mip v[0:3], v[0:1], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_ARRAY unorm a16 ; encoding: [0x94,0x0f,0x05,0xf0,0x00,0x00,0x00,0x00]
+; GFX11-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0xf7,0x03,0x89,0xbf]
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %s = extractelement <2 x i16> %coords_lo, i32 0
   %t = extractelement <2 x i16> %coords_lo, i32 1
   %slice = extractelement <2 x i16> %coords_hi, i32 0
   %mip = extractelement <2 x i16> %coords_hi, i32 1
   %v = call <4 x float> @llvm.amdgcn.image.load.mip.2darray.v4f32.i16(i32 15, i16 %s, i16 %t, i16 %slice, i16 %mip, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps void @store_1d(<8 x i32> inreg %rsrc, <4 x float> %vdata, <2 x i16> %coords) {
 ; GFX9-LABEL: store_1d:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    image_store v[0:3], v4, s[0:7] dmask:0xf unorm a16 ; encoding: [0x00,0x9f,0x20,0xf0,0x04,0x00,0x00,0x00]
 ; GFX9-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
 ;
 ; GFX10-LABEL: store_1d:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    image_store v[0:3], v4, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm a16 ; encoding: [0x00,0x1f,0x20,0xf0,0x04,0x00,0x00,0x40]
 ; GFX10-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
+;
+; GFX11-LABEL: store_1d:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    image_store v[0:3], v4, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm a16 ; encoding: [0x80,0x0f,0x19,0xf0,0x04,0x00,0x00,0x00]
+; GFX11-NEXT:    s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
 main_body:
   %s = extractelement <2 x i16> %coords, i32 0
   call void @llvm.amdgcn.image.store.1d.v4f32.i16(<4 x float> %vdata, i32 15, i16 %s, <8 x i32> %rsrc, i32 0, i32 0)
   ret void
 }
 
 define amdgpu_ps void @store_2d(<8 x i32> inreg %rsrc, <4 x float> %vdata, <2 x i16> %coords) {
 ; GFX9-LABEL: store_2d:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    image_store v[0:3], v4, s[0:7] dmask:0xf unorm a16 ; encoding: [0x00,0x9f,0x20,0xf0,0x04,0x00,0x00,0x00]
 ; GFX9-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
 ;
 ; GFX10-LABEL: store_2d:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    image_store v[0:3], v4, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D unorm a16 ; encoding: [0x08,0x1f,0x20,0xf0,0x04,0x00,0x00,0x40]
 ; GFX10-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
+;
+; GFX11-LABEL: store_2d:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    image_store v[0:3], v4, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D unorm a16 ; encoding: [0x84,0x0f,0x19,0xf0,0x04,0x00,0x00,0x00]
+; GFX11-NEXT:    s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
 main_body:
   %s = extractelement <2 x i16> %coords, i32 0
   %t = extractelement <2 x i16> %coords, i32 1
   call void @llvm.amdgcn.image.store.2d.v4f32.i16(<4 x float> %vdata, i32 15, i16 %s, i16 %t, <8 x i32> %rsrc, i32 0, i32 0)
   ret void
 }
 
 define amdgpu_ps void @store_3d(<8 x i32> inreg %rsrc, <4 x float> %vdata, <2 x i16> %coords_lo, <2 x i16> %coords_hi) {
 ; GFX9-LABEL: store_3d:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    image_store v[0:3], v[4:5], s[0:7] dmask:0xf unorm a16 ; encoding: [0x00,0x9f,0x20,0xf0,0x04,0x00,0x00,0x00]
 ; GFX9-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
 ;
 ; GFX10-LABEL: store_3d:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    image_store v[0:3], v[4:5], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_3D unorm a16 ; encoding: [0x10,0x1f,0x20,0xf0,0x04,0x00,0x00,0x40]
 ; GFX10-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
+;
+; GFX11-LABEL: store_3d:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    image_store v[0:3], v[4:5], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_3D unorm a16 ; encoding: [0x88,0x0f,0x19,0xf0,0x04,0x00,0x00,0x00]
+; GFX11-NEXT:    s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
 main_body:
   %s = extractelement <2 x i16> %coords_lo, i32 0
   %t = extractelement <2 x i16> %coords_lo, i32 1
   %r = extractelement <2 x i16> %coords_hi, i32 0
   call void @llvm.amdgcn.image.store.3d.v4f32.i16(<4 x float> %vdata, i32 15, i16 %s, i16 %t, i16 %r, <8 x i32> %rsrc, i32 0, i32 0)
   ret void
 }
 
 define amdgpu_ps void @store_cube(<8 x i32> inreg %rsrc, <4 x float> %vdata, <2 x i16> %coords_lo, <2 x i16> %coords_hi) {
 ; GFX9-LABEL: store_cube:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    image_store v[0:3], v[4:5], s[0:7] dmask:0xf unorm a16 da ; encoding: [0x00,0xdf,0x20,0xf0,0x04,0x00,0x00,0x00]
 ; GFX9-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
 ;
 ; GFX10-LABEL: store_cube:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    image_store v[0:3], v[4:5], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_CUBE unorm a16 ; encoding: [0x18,0x1f,0x20,0xf0,0x04,0x00,0x00,0x40]
 ; GFX10-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
+;
+; GFX11-LABEL: store_cube:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    image_store v[0:3], v[4:5], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_CUBE unorm a16 ; encoding: [0x8c,0x0f,0x19,0xf0,0x04,0x00,0x00,0x00]
+; GFX11-NEXT:    s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
 main_body:
   %s = extractelement <2 x i16> %coords_lo, i32 0
   %t = extractelement <2 x i16> %coords_lo, i32 1
   %slice = extractelement <2 x i16> %coords_hi, i32 0
   call void @llvm.amdgcn.image.store.cube.v4f32.i16(<4 x float> %vdata, i32 15, i16 %s, i16 %t, i16 %slice, <8 x i32> %rsrc, i32 0, i32 0)
   ret void
 }
 
 define amdgpu_ps void @store_1darray(<8 x i32> inreg %rsrc, <4 x float> %vdata, <2 x i16> %coords) {
 ; GFX9-LABEL: store_1darray:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    image_store v[0:3], v4, s[0:7] dmask:0xf unorm a16 da ; encoding: [0x00,0xdf,0x20,0xf0,0x04,0x00,0x00,0x00]
 ; GFX9-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
 ;
 ; GFX10-LABEL: store_1darray:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    image_store v[0:3], v4, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D_ARRAY unorm a16 ; encoding: [0x20,0x1f,0x20,0xf0,0x04,0x00,0x00,0x40]
 ; GFX10-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
+;
+; GFX11-LABEL: store_1darray:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    image_store v[0:3], v4, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D_ARRAY unorm a16 ; encoding: [0x90,0x0f,0x19,0xf0,0x04,0x00,0x00,0x00]
+; GFX11-NEXT:    s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
 main_body:
   %s = extractelement <2 x i16> %coords, i32 0
   %slice = extractelement <2 x i16> %coords, i32 1
   call void @llvm.amdgcn.image.store.1darray.v4f32.i16(<4 x float> %vdata, i32 15, i16 %s, i16 %slice, <8 x i32> %rsrc, i32 0, i32 0)
   ret void
 }
 
 define amdgpu_ps void @store_2darray(<8 x i32> inreg %rsrc, <4 x float> %vdata, <2 x i16> %coords_lo, <2 x i16> %coords_hi) {
 ; GFX9-LABEL: store_2darray:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    image_store v[0:3], v[4:5], s[0:7] dmask:0xf unorm a16 da ; encoding: [0x00,0xdf,0x20,0xf0,0x04,0x00,0x00,0x00]
 ; GFX9-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
 ;
 ; GFX10-LABEL: store_2darray:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    image_store v[0:3], v[4:5], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_ARRAY unorm a16 ; encoding: [0x28,0x1f,0x20,0xf0,0x04,0x00,0x00,0x40]
 ; GFX10-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
+;
+; GFX11-LABEL: store_2darray:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    image_store v[0:3], v[4:5], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_ARRAY unorm a16 ; encoding: [0x94,0x0f,0x19,0xf0,0x04,0x00,0x00,0x00]
+; GFX11-NEXT:    s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
 main_body:
   %s = extractelement <2 x i16> %coords_lo, i32 0
   %t = extractelement <2 x i16> %coords_lo, i32 1
   %slice = extractelement <2 x i16> %coords_hi, i32 0
   call void @llvm.amdgcn.image.store.2darray.v4f32.i16(<4 x float> %vdata, i32 15, i16 %s, i16 %t, i16 %slice, <8 x i32> %rsrc, i32 0, i32 0)
   ret void
 }
 
 define amdgpu_ps void @store_2dmsaa(<8 x i32> inreg %rsrc, <4 x float> %vdata, <2 x i16> %coords_lo, <2 x i16> %coords_hi) {
 ; GFX9-LABEL: store_2dmsaa:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    image_store v[0:3], v[4:5], s[0:7] dmask:0xf unorm a16 ; encoding: [0x00,0x9f,0x20,0xf0,0x04,0x00,0x00,0x00]
 ; GFX9-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
 ;
 ; GFX10-LABEL: store_2dmsaa:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    image_store v[0:3], v[4:5], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_MSAA unorm a16 ; encoding: [0x30,0x1f,0x20,0xf0,0x04,0x00,0x00,0x40]
 ; GFX10-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
+;
+; GFX11-LABEL: store_2dmsaa:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    image_store v[0:3], v[4:5], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_MSAA unorm a16 ; encoding: [0x98,0x0f,0x19,0xf0,0x04,0x00,0x00,0x00]
+; GFX11-NEXT:    s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
 main_body:
   %s = extractelement <2 x i16> %coords_lo, i32 0
   %t = extractelement <2 x i16> %coords_lo, i32 1
   %fragid = extractelement <2 x i16> %coords_hi, i32 0
   call void @llvm.amdgcn.image.store.2dmsaa.v4f32.i16(<4 x float> %vdata, i32 15, i16 %s, i16 %t, i16 %fragid, <8 x i32> %rsrc, i32 0, i32 0)
   ret void
 }
 
 define amdgpu_ps void @store_2darraymsaa(<8 x i32> inreg %rsrc, <4 x float> %vdata, <2 x i16> %coords_lo, <2 x i16> %coords_hi) {
 ; GFX9-LABEL: store_2darraymsaa:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    image_store v[0:3], v[4:5], s[0:7] dmask:0xf unorm a16 da ; encoding: [0x00,0xdf,0x20,0xf0,0x04,0x00,0x00,0x00]
 ; GFX9-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
 ;
 ; GFX10-LABEL: store_2darraymsaa:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    image_store v[0:3], v[4:5], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_MSAA_ARRAY unorm a16 ; encoding: [0x38,0x1f,0x20,0xf0,0x04,0x00,0x00,0x40]
 ; GFX10-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
+;
+; GFX11-LABEL: store_2darraymsaa:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    image_store v[0:3], v[4:5], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_MSAA_ARRAY unorm a16 ; encoding: [0x9c,0x0f,0x19,0xf0,0x04,0x00,0x00,0x00]
+; GFX11-NEXT:    s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
 main_body:
   %s = extractelement <2 x i16> %coords_lo, i32 0
   %t = extractelement <2 x i16> %coords_lo, i32 1
   %slice = extractelement <2 x i16> %coords_hi, i32 0
   %fragid = extractelement <2 x i16> %coords_hi, i32 1
   call void @llvm.amdgcn.image.store.2darraymsaa.v4f32.i16(<4 x float> %vdata, i32 15, i16 %s, i16 %t, i16 %slice, i16 %fragid, <8 x i32> %rsrc, i32 0, i32 0)
   ret void
 }
 
 define amdgpu_ps void @store_mip_1d(<8 x i32> inreg %rsrc, <4 x float> %vdata, <2 x i16> %coords) {
 ; GFX9-LABEL: store_mip_1d:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    image_store_mip v[0:3], v4, s[0:7] dmask:0xf unorm a16 ; encoding: [0x00,0x9f,0x24,0xf0,0x04,0x00,0x00,0x00]
 ; GFX9-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
 ;
 ; GFX10-LABEL: store_mip_1d:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    image_store_mip v[0:3], v4, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm a16 ; encoding: [0x00,0x1f,0x24,0xf0,0x04,0x00,0x00,0x40]
 ; GFX10-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
+;
+; GFX11-LABEL: store_mip_1d:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    image_store_mip v[0:3], v4, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm a16 ; encoding: [0x80,0x0f,0x1d,0xf0,0x04,0x00,0x00,0x00]
+; GFX11-NEXT:    s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
 main_body:
   %s = extractelement <2 x i16> %coords, i32 0
   %mip = extractelement <2 x i16> %coords, i32 1
   call void @llvm.amdgcn.image.store.mip.1d.v4f32.i16(<4 x float> %vdata, i32 15, i16 %s, i16 %mip, <8 x i32> %rsrc, i32 0, i32 0)
   ret void
 }
 
 define amdgpu_ps void @store_mip_2d(<8 x i32> inreg %rsrc, <4 x float> %vdata, <2 x i16> %coords_lo, <2 x i16> %coords_hi) {
 ; GFX9-LABEL: store_mip_2d:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    image_store_mip v[0:3], v[4:5], s[0:7] dmask:0xf unorm a16 ; encoding: [0x00,0x9f,0x24,0xf0,0x04,0x00,0x00,0x00]
 ; GFX9-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
 ;
 ; GFX10-LABEL: store_mip_2d:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    image_store_mip v[0:3], v[4:5], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D unorm a16 ; encoding: [0x08,0x1f,0x24,0xf0,0x04,0x00,0x00,0x40]
 ; GFX10-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
+;
+; GFX11-LABEL: store_mip_2d:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    image_store_mip v[0:3], v[4:5], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D unorm a16 ; encoding: [0x84,0x0f,0x1d,0xf0,0x04,0x00,0x00,0x00]
+; GFX11-NEXT:    s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
 main_body:
   %s = extractelement <2 x i16> %coords_lo, i32 0
   %t = extractelement <2 x i16> %coords_lo, i32 1
   %mip = extractelement <2 x i16> %coords_hi, i32 0
   call void @llvm.amdgcn.image.store.mip.2d.v4f32.i16(<4 x float> %vdata, i32 15, i16 %s, i16 %t, i16 %mip, <8 x i32> %rsrc, i32 0, i32 0)
   ret void
 }
 
 define amdgpu_ps void @store_mip_3d(<8 x i32> inreg %rsrc, <4 x float> %vdata, <2 x i16> %coords_lo, <2 x i16> %coords_hi) {
 ; GFX9-LABEL: store_mip_3d:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    image_store_mip v[0:3], v[4:5], s[0:7] dmask:0xf unorm a16 ; encoding: [0x00,0x9f,0x24,0xf0,0x04,0x00,0x00,0x00]
 ; GFX9-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
 ;
 ; GFX10-LABEL: store_mip_3d:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    image_store_mip v[0:3], v[4:5], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_3D unorm a16 ; encoding: [0x10,0x1f,0x24,0xf0,0x04,0x00,0x00,0x40]
 ; GFX10-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
+;
+; GFX11-LABEL: store_mip_3d:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    image_store_mip v[0:3], v[4:5], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_3D unorm a16 ; encoding: [0x88,0x0f,0x1d,0xf0,0x04,0x00,0x00,0x00]
+; GFX11-NEXT:    s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
 main_body:
   %s = extractelement <2 x i16> %coords_lo, i32 0
   %t = extractelement <2 x i16> %coords_lo, i32 1
   %r = extractelement <2 x i16> %coords_hi, i32 0
   %mip = extractelement <2 x i16> %coords_hi, i32 1
   call void @llvm.amdgcn.image.store.mip.3d.v4f32.i16(<4 x float> %vdata, i32 15, i16 %s, i16 %t, i16 %r, i16 %mip, <8 x i32> %rsrc, i32 0, i32 0)
   ret void
 }
 
 define amdgpu_ps void @store_mip_cube(<8 x i32> inreg %rsrc, <4 x float> %vdata, <2 x i16> %coords_lo, <2 x i16> %coords_hi) {
 ; GFX9-LABEL: store_mip_cube:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    image_store_mip v[0:3], v[4:5], s[0:7] dmask:0xf unorm a16 da ; encoding: [0x00,0xdf,0x24,0xf0,0x04,0x00,0x00,0x00]
 ; GFX9-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
 ;
 ; GFX10-LABEL: store_mip_cube:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    image_store_mip v[0:3], v[4:5], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_CUBE unorm a16 ; encoding: [0x18,0x1f,0x24,0xf0,0x04,0x00,0x00,0x40]
 ; GFX10-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
+;
+; GFX11-LABEL: store_mip_cube:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    image_store_mip v[0:3], v[4:5], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_CUBE unorm a16 ; encoding: [0x8c,0x0f,0x1d,0xf0,0x04,0x00,0x00,0x00]
+; GFX11-NEXT:    s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
 main_body:
   %s = extractelement <2 x i16> %coords_lo, i32 0
   %t = extractelement <2 x i16> %coords_lo, i32 1
   %slice = extractelement <2 x i16> %coords_hi, i32 0
   %mip = extractelement <2 x i16> %coords_hi, i32 1
   call void @llvm.amdgcn.image.store.mip.cube.v4f32.i16(<4 x float> %vdata, i32 15, i16 %s, i16 %t, i16 %slice, i16 %mip, <8 x i32> %rsrc, i32 0, i32 0)
   ret void
 }
 
 define amdgpu_ps void @store_mip_1darray(<8 x i32> inreg %rsrc, <4 x float> %vdata, <2 x i16> %coords_lo, <2 x i16> %coords_hi) {
 ; GFX9-LABEL: store_mip_1darray:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    image_store_mip v[0:3], v[4:5], s[0:7] dmask:0xf unorm a16 da ; encoding: [0x00,0xdf,0x24,0xf0,0x04,0x00,0x00,0x00]
 ; GFX9-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
 ;
 ; GFX10-LABEL: store_mip_1darray:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    image_store_mip v[0:3], v[4:5], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D_ARRAY unorm a16 ; encoding: [0x20,0x1f,0x24,0xf0,0x04,0x00,0x00,0x40]
 ; GFX10-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
+;
+; GFX11-LABEL: store_mip_1darray:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    image_store_mip v[0:3], v[4:5], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D_ARRAY unorm a16 ; encoding: [0x90,0x0f,0x1d,0xf0,0x04,0x00,0x00,0x00]
+; GFX11-NEXT:    s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
 main_body:
   %s = extractelement <2 x i16> %coords_lo, i32 0
   %slice = extractelement <2 x i16> %coords_lo, i32 1
   %mip = extractelement <2 x i16> %coords_hi, i32 0
   call void @llvm.amdgcn.image.store.mip.1darray.v4f32.i16(<4 x float> %vdata, i32 15, i16 %s, i16 %slice, i16 %mip, <8 x i32> %rsrc, i32 0, i32 0)
   ret void
 }
 
 define amdgpu_ps void @store_mip_2darray(<8 x i32> inreg %rsrc, <4 x float> %vdata, <2 x i16> %coords_lo, <2 x i16> %coords_hi) {
 ; GFX9-LABEL: store_mip_2darray:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    image_store_mip v[0:3], v[4:5], s[0:7] dmask:0xf unorm a16 da ; encoding: [0x00,0xdf,0x24,0xf0,0x04,0x00,0x00,0x00]
 ; GFX9-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
 ;
 ; GFX10-LABEL: store_mip_2darray:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    image_store_mip v[0:3], v[4:5], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_ARRAY unorm a16 ; encoding: [0x28,0x1f,0x24,0xf0,0x04,0x00,0x00,0x40]
 ; GFX10-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
+;
+; GFX11-LABEL: store_mip_2darray:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    image_store_mip v[0:3], v[4:5], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_ARRAY unorm a16 ; encoding: [0x94,0x0f,0x1d,0xf0,0x04,0x00,0x00,0x00]
+; GFX11-NEXT:    s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
 main_body:
   %s = extractelement <2 x i16> %coords_lo, i32 0
   %t = extractelement <2 x i16> %coords_lo, i32 1
   %slice = extractelement <2 x i16> %coords_hi, i32 0
   %mip = extractelement <2 x i16> %coords_hi, i32 1
   call void @llvm.amdgcn.image.store.mip.2darray.v4f32.i16(<4 x float> %vdata, i32 15, i16 %s, i16 %t, i16 %slice, i16 %mip, <8 x i32> %rsrc, i32 0, i32 0)
   ret void
 }
 
 define amdgpu_ps <4 x float> @getresinfo_1d(<8 x i32> inreg %rsrc, <2 x i16> %coords) {
 ; GFX9-LABEL: getresinfo_1d:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    image_get_resinfo v[0:3], v0, s[0:7] dmask:0xf unorm a16 ; encoding: [0x00,0x9f,0x38,0xf0,0x00,0x00,0x00,0x00]
 ; GFX9-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x0f,0x8c,0xbf]
 ; GFX9-NEXT:    ; return to shader part epilog
 ;
 ; GFX10-LABEL: getresinfo_1d:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    image_get_resinfo v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm a16 ; encoding: [0x00,0x1f,0x38,0xf0,0x00,0x00,0x00,0x40]
 ; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
 ; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: getresinfo_1d:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    image_get_resinfo v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm a16 ; encoding: [0x80,0x0f,0x5d,0xf0,0x00,0x00,0x00,0x00]
+; GFX11-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0xf7,0x03,0x89,0xbf]
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %mip = extractelement <2 x i16> %coords, i32 0
   %v = call <4 x float> @llvm.amdgcn.image.getresinfo.1d.v4f32.i16(i32 15, i16 %mip, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @getresinfo_2d(<8 x i32> inreg %rsrc, <2 x i16> %coords) {
 ; GFX9-LABEL: getresinfo_2d:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    image_get_resinfo v[0:3], v0, s[0:7] dmask:0xf unorm a16 ; encoding: [0x00,0x9f,0x38,0xf0,0x00,0x00,0x00,0x00]
 ; GFX9-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x0f,0x8c,0xbf]
 ; GFX9-NEXT:    ; return to shader part epilog
 ;
 ; GFX10-LABEL: getresinfo_2d:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    image_get_resinfo v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D unorm a16 ; encoding: [0x08,0x1f,0x38,0xf0,0x00,0x00,0x00,0x40]
 ; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
 ; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: getresinfo_2d:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    image_get_resinfo v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D unorm a16 ; encoding: [0x84,0x0f,0x5d,0xf0,0x00,0x00,0x00,0x00]
+; GFX11-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0xf7,0x03,0x89,0xbf]
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %mip = extractelement <2 x i16> %coords, i32 0
   %v = call <4 x float> @llvm.amdgcn.image.getresinfo.2d.v4f32.i16(i32 15, i16 %mip, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @getresinfo_3d(<8 x i32> inreg %rsrc, <2 x i16> %coords) {
 ; GFX9-LABEL: getresinfo_3d:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    image_get_resinfo v[0:3], v0, s[0:7] dmask:0xf unorm a16 ; encoding: [0x00,0x9f,0x38,0xf0,0x00,0x00,0x00,0x00]
 ; GFX9-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x0f,0x8c,0xbf]
 ; GFX9-NEXT:    ; return to shader part epilog
 ;
 ; GFX10-LABEL: getresinfo_3d:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    image_get_resinfo v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_3D unorm a16 ; encoding: [0x10,0x1f,0x38,0xf0,0x00,0x00,0x00,0x40]
 ; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
 ; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: getresinfo_3d:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    image_get_resinfo v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_3D unorm a16 ; encoding: [0x88,0x0f,0x5d,0xf0,0x00,0x00,0x00,0x00]
+; GFX11-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0xf7,0x03,0x89,0xbf]
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %mip = extractelement <2 x i16> %coords, i32 0
   %v = call <4 x float> @llvm.amdgcn.image.getresinfo.3d.v4f32.i16(i32 15, i16 %mip, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @getresinfo_cube(<8 x i32> inreg %rsrc, <2 x i16> %coords) {
 ; GFX9-LABEL: getresinfo_cube:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    image_get_resinfo v[0:3], v0, s[0:7] dmask:0xf unorm a16 da ; encoding: [0x00,0xdf,0x38,0xf0,0x00,0x00,0x00,0x00]
 ; GFX9-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x0f,0x8c,0xbf]
 ; GFX9-NEXT:    ; return to shader part epilog
 ;
 ; GFX10-LABEL: getresinfo_cube:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    image_get_resinfo v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_CUBE unorm a16 ; encoding: [0x18,0x1f,0x38,0xf0,0x00,0x00,0x00,0x40]
 ; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
 ; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: getresinfo_cube:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    image_get_resinfo v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_CUBE unorm a16 ; encoding: [0x8c,0x0f,0x5d,0xf0,0x00,0x00,0x00,0x00]
+; GFX11-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0xf7,0x03,0x89,0xbf]
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %mip = extractelement <2 x i16> %coords, i32 0
   %v = call <4 x float> @llvm.amdgcn.image.getresinfo.cube.v4f32.i16(i32 15, i16 %mip, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @getresinfo_1darray(<8 x i32> inreg %rsrc, <2 x i16> %coords) {
 ; GFX9-LABEL: getresinfo_1darray:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    image_get_resinfo v[0:3], v0, s[0:7] dmask:0xf unorm a16 da ; encoding: [0x00,0xdf,0x38,0xf0,0x00,0x00,0x00,0x00]
 ; GFX9-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x0f,0x8c,0xbf]
 ; GFX9-NEXT:    ; return to shader part epilog
 ;
 ; GFX10-LABEL: getresinfo_1darray:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    image_get_resinfo v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D_ARRAY unorm a16 ; encoding: [0x20,0x1f,0x38,0xf0,0x00,0x00,0x00,0x40]
 ; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
 ; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: getresinfo_1darray:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    image_get_resinfo v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D_ARRAY unorm a16 ; encoding: [0x90,0x0f,0x5d,0xf0,0x00,0x00,0x00,0x00]
+; GFX11-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0xf7,0x03,0x89,0xbf]
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %mip = extractelement <2 x i16> %coords, i32 0
   %v = call <4 x float> @llvm.amdgcn.image.getresinfo.1darray.v4f32.i16(i32 15, i16 %mip, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @getresinfo_2darray(<8 x i32> inreg %rsrc, <2 x i16> %coords) {
 ; GFX9-LABEL: getresinfo_2darray:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    image_get_resinfo v[0:3], v0, s[0:7] dmask:0xf unorm a16 da ; encoding: [0x00,0xdf,0x38,0xf0,0x00,0x00,0x00,0x00]
 ; GFX9-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x0f,0x8c,0xbf]
 ; GFX9-NEXT:    ; return to shader part epilog
 ;
 ; GFX10-LABEL: getresinfo_2darray:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    image_get_resinfo v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_ARRAY unorm a16 ; encoding: [0x28,0x1f,0x38,0xf0,0x00,0x00,0x00,0x40]
 ; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
 ; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: getresinfo_2darray:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    image_get_resinfo v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_ARRAY unorm a16 ; encoding: [0x94,0x0f,0x5d,0xf0,0x00,0x00,0x00,0x00]
+; GFX11-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0xf7,0x03,0x89,0xbf]
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %mip = extractelement <2 x i16> %coords, i32 0
   %v = call <4 x float> @llvm.amdgcn.image.getresinfo.2darray.v4f32.i16(i32 15, i16 %mip, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @getresinfo_2dmsaa(<8 x i32> inreg %rsrc, <2 x i16> %coords) {
 ; GFX9-LABEL: getresinfo_2dmsaa:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    image_get_resinfo v[0:3], v0, s[0:7] dmask:0xf unorm a16 ; encoding: [0x00,0x9f,0x38,0xf0,0x00,0x00,0x00,0x00]
 ; GFX9-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x0f,0x8c,0xbf]
 ; GFX9-NEXT:    ; return to shader part epilog
 ;
 ; GFX10-LABEL: getresinfo_2dmsaa:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    image_get_resinfo v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_MSAA unorm a16 ; encoding: [0x30,0x1f,0x38,0xf0,0x00,0x00,0x00,0x40]
 ; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
 ; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: getresinfo_2dmsaa:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    image_get_resinfo v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_MSAA unorm a16 ; encoding: [0x98,0x0f,0x5d,0xf0,0x00,0x00,0x00,0x00]
+; GFX11-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0xf7,0x03,0x89,0xbf]
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %mip = extractelement <2 x i16> %coords, i32 0
   %v = call <4 x float> @llvm.amdgcn.image.getresinfo.2dmsaa.v4f32.i16(i32 15, i16 %mip, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @getresinfo_2darraymsaa(<8 x i32> inreg %rsrc, <2 x i16> %coords) {
 ; GFX9-LABEL: getresinfo_2darraymsaa:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    image_get_resinfo v[0:3], v0, s[0:7] dmask:0xf unorm a16 da ; encoding: [0x00,0xdf,0x38,0xf0,0x00,0x00,0x00,0x00]
 ; GFX9-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x0f,0x8c,0xbf]
 ; GFX9-NEXT:    ; return to shader part epilog
 ;
 ; GFX10-LABEL: getresinfo_2darraymsaa:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    image_get_resinfo v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_MSAA_ARRAY unorm a16 ; encoding: [0x38,0x1f,0x38,0xf0,0x00,0x00,0x00,0x40]
 ; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
 ; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: getresinfo_2darraymsaa:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    image_get_resinfo v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_MSAA_ARRAY unorm a16 ; encoding: [0x9c,0x0f,0x5d,0xf0,0x00,0x00,0x00,0x00]
+; GFX11-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0xf7,0x03,0x89,0xbf]
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %mip = extractelement <2 x i16> %coords, i32 0
   %v = call <4 x float> @llvm.amdgcn.image.getresinfo.2darraymsaa.v4f32.i16(i32 15, i16 %mip, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps float @load_1d_V1(<8 x i32> inreg %rsrc, <2 x i16> %coords) {
 ; GFX9-LABEL: load_1d_V1:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    image_load v0, v0, s[0:7] dmask:0x8 unorm a16 ; encoding: [0x00,0x98,0x00,0xf0,0x00,0x00,0x00,0x00]
 ; GFX9-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x0f,0x8c,0xbf]
 ; GFX9-NEXT:    ; return to shader part epilog
 ;
 ; GFX10-LABEL: load_1d_V1:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    image_load v0, v0, s[0:7] dmask:0x8 dim:SQ_RSRC_IMG_1D unorm a16 ; encoding: [0x00,0x18,0x00,0xf0,0x00,0x00,0x00,0x40]
 ; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
 ; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: load_1d_V1:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    image_load v0, v0, s[0:7] dmask:0x8 dim:SQ_RSRC_IMG_1D unorm a16 ; encoding: [0x80,0x08,0x01,0xf0,0x00,0x00,0x00,0x00]
+; GFX11-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0xf7,0x03,0x89,0xbf]
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %s = extractelement <2 x i16> %coords, i32 0
   %v = call float @llvm.amdgcn.image.load.1d.f32.i16(i32 8, i16 %s, <8 x i32> %rsrc, i32 0, i32 0)
   ret float %v
 }
 
 define amdgpu_ps <2 x float> @load_1d_V2(<8 x i32> inreg %rsrc, <2 x i16> %coords) {
 ; GFX9-LABEL: load_1d_V2:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    image_load v[0:1], v0, s[0:7] dmask:0x9 unorm a16 ; encoding: [0x00,0x99,0x00,0xf0,0x00,0x00,0x00,0x00]
 ; GFX9-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x0f,0x8c,0xbf]
 ; GFX9-NEXT:    ; return to shader part epilog
 ;
 ; GFX10-LABEL: load_1d_V2:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    image_load v[0:1], v0, s[0:7] dmask:0x9 dim:SQ_RSRC_IMG_1D unorm a16 ; encoding: [0x00,0x19,0x00,0xf0,0x00,0x00,0x00,0x40]
 ; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
 ; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: load_1d_V2:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    image_load v[0:1], v0, s[0:7] dmask:0x9 dim:SQ_RSRC_IMG_1D unorm a16 ; encoding: [0x80,0x09,0x01,0xf0,0x00,0x00,0x00,0x00]
+; GFX11-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0xf7,0x03,0x89,0xbf]
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %s = extractelement <2 x i16> %coords, i32 0
   %v = call <2 x float> @llvm.amdgcn.image.load.1d.v2f32.i16(i32 9, i16 %s, <8 x i32> %rsrc, i32 0, i32 0)
   ret <2 x float> %v
 }
 
 define amdgpu_ps void @store_1d_V1(<8 x i32> inreg %rsrc, float %vdata, <2 x i16> %coords) {
 ; GFX9-LABEL: store_1d_V1:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    image_store v0, v1, s[0:7] dmask:0x2 unorm a16 ; encoding: [0x00,0x92,0x20,0xf0,0x01,0x00,0x00,0x00]
 ; GFX9-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
 ;
 ; GFX10-LABEL: store_1d_V1:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    image_store v0, v1, s[0:7] dmask:0x2 dim:SQ_RSRC_IMG_1D unorm a16 ; encoding: [0x00,0x12,0x20,0xf0,0x01,0x00,0x00,0x40]
 ; GFX10-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
+;
+; GFX11-LABEL: store_1d_V1:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    image_store v0, v1, s[0:7] dmask:0x2 dim:SQ_RSRC_IMG_1D unorm a16 ; encoding: [0x80,0x02,0x19,0xf0,0x01,0x00,0x00,0x00]
+; GFX11-NEXT:    s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
 main_body:
   %s = extractelement <2 x i16> %coords, i32 0
   call void @llvm.amdgcn.image.store.1d.f32.i16(float %vdata, i32 2, i16 %s, <8 x i32> %rsrc, i32 0, i32 0)
   ret void
 }
 
 define amdgpu_ps void @store_1d_V2(<8 x i32> inreg %rsrc, <2 x float> %vdata, <2 x i16> %coords) {
 ; GFX9-LABEL: store_1d_V2:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    image_store v[0:1], v2, s[0:7] dmask:0xc unorm a16 ; encoding: [0x00,0x9c,0x20,0xf0,0x02,0x00,0x00,0x00]
 ; GFX9-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
 ;
 ; GFX10-LABEL: store_1d_V2:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    image_store v[0:1], v2, s[0:7] dmask:0xc dim:SQ_RSRC_IMG_1D unorm a16 ; encoding: [0x00,0x1c,0x20,0xf0,0x02,0x00,0x00,0x40]
 ; GFX10-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
+;
+; GFX11-LABEL: store_1d_V2:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    image_store v[0:1], v2, s[0:7] dmask:0xc dim:SQ_RSRC_IMG_1D unorm a16 ; encoding: [0x80,0x0c,0x19,0xf0,0x02,0x00,0x00,0x00]
+; GFX11-NEXT:    s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
 main_body:
   %s = extractelement <2 x i16> %coords, i32 0
   call void @llvm.amdgcn.image.store.1d.v2f32.i16(<2 x float> %vdata, i32 12, i16 %s, <8 x i32> %rsrc, i32 0, i32 0)
   ret void
 }
 
 define amdgpu_ps <4 x float> @load_1d_glc(<8 x i32> inreg %rsrc, <2 x i16> %coords) {
 ; GFX9-LABEL: load_1d_glc:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    image_load v[0:3], v0, s[0:7] dmask:0xf unorm glc a16 ; encoding: [0x00,0xbf,0x00,0xf0,0x00,0x00,0x00,0x00]
 ; GFX9-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x0f,0x8c,0xbf]
 ; GFX9-NEXT:    ; return to shader part epilog
 ;
 ; GFX10-LABEL: load_1d_glc:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    image_load v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm glc a16 ; encoding: [0x00,0x3f,0x00,0xf0,0x00,0x00,0x00,0x40]
 ; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
 ; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: load_1d_glc:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    image_load v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm glc a16 ; encoding: [0x80,0x4f,0x01,0xf0,0x00,0x00,0x00,0x00]
+; GFX11-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0xf7,0x03,0x89,0xbf]
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %s = extractelement <2 x i16> %coords, i32 0
   %v = call <4 x float> @llvm.amdgcn.image.load.1d.v4f32.i16(i32 15, i16 %s, <8 x i32> %rsrc, i32 0, i32 1)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @load_1d_slc(<8 x i32> inreg %rsrc, <2 x i16> %coords) {
 ; GFX9-LABEL: load_1d_slc:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    image_load v[0:3], v0, s[0:7] dmask:0xf unorm slc a16 ; encoding: [0x00,0x9f,0x00,0xf2,0x00,0x00,0x00,0x00]
 ; GFX9-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x0f,0x8c,0xbf]
 ; GFX9-NEXT:    ; return to shader part epilog
 ;
 ; GFX10-LABEL: load_1d_slc:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    image_load v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm slc a16 ; encoding: [0x00,0x1f,0x00,0xf2,0x00,0x00,0x00,0x40]
 ; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
 ; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: load_1d_slc:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    image_load v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm slc a16 ; encoding: [0x80,0x1f,0x01,0xf0,0x00,0x00,0x00,0x00]
+; GFX11-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0xf7,0x03,0x89,0xbf]
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %s = extractelement <2 x i16> %coords, i32 0
   %v = call <4 x float> @llvm.amdgcn.image.load.1d.v4f32.i16(i32 15, i16 %s, <8 x i32> %rsrc, i32 0, i32 2)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @load_1d_glc_slc(<8 x i32> inreg %rsrc, <2 x i16> %coords) {
 ; GFX9-LABEL: load_1d_glc_slc:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    image_load v[0:3], v0, s[0:7] dmask:0xf unorm glc slc a16 ; encoding: [0x00,0xbf,0x00,0xf2,0x00,0x00,0x00,0x00]
 ; GFX9-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x0f,0x8c,0xbf]
 ; GFX9-NEXT:    ; return to shader part epilog
 ;
 ; GFX10-LABEL: load_1d_glc_slc:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    image_load v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm glc slc a16 ; encoding: [0x00,0x3f,0x00,0xf2,0x00,0x00,0x00,0x40]
 ; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
 ; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: load_1d_glc_slc:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    image_load v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm glc slc a16 ; encoding: [0x80,0x5f,0x01,0xf0,0x00,0x00,0x00,0x00]
+; GFX11-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0xf7,0x03,0x89,0xbf]
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %s = extractelement <2 x i16> %coords, i32 0
   %v = call <4 x float> @llvm.amdgcn.image.load.1d.v4f32.i16(i32 15, i16 %s, <8 x i32> %rsrc, i32 0, i32 3)
   ret <4 x float> %v
 }
 
 define amdgpu_ps void @store_1d_glc(<8 x i32> inreg %rsrc, <4 x float> %vdata, <2 x i16> %coords) {
 ; GFX9-LABEL: store_1d_glc:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    image_store v[0:3], v4, s[0:7] dmask:0xf unorm glc a16 ; encoding: [0x00,0xbf,0x20,0xf0,0x04,0x00,0x00,0x00]
 ; GFX9-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
 ;
 ; GFX10-LABEL: store_1d_glc:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    image_store v[0:3], v4, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm glc a16 ; encoding: [0x00,0x3f,0x20,0xf0,0x04,0x00,0x00,0x40]
 ; GFX10-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
+;
+; GFX11-LABEL: store_1d_glc:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    image_store v[0:3], v4, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm glc a16 ; encoding: [0x80,0x4f,0x19,0xf0,0x04,0x00,0x00,0x00]
+; GFX11-NEXT:    s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
 main_body:
   %s = extractelement <2 x i16> %coords, i32 0
   call void @llvm.amdgcn.image.store.1d.v4f32.i16(<4 x float> %vdata, i32 15, i16 %s, <8 x i32> %rsrc, i32 0, i32 1)
   ret void
 }
 
 define amdgpu_ps void @store_1d_slc(<8 x i32> inreg %rsrc, <4 x float> %vdata, <2 x i16> %coords) {
 ; GFX9-LABEL: store_1d_slc:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    image_store v[0:3], v4, s[0:7] dmask:0xf unorm slc a16 ; encoding: [0x00,0x9f,0x20,0xf2,0x04,0x00,0x00,0x00]
 ; GFX9-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
 ;
 ; GFX10-LABEL: store_1d_slc:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    image_store v[0:3], v4, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm slc a16 ; encoding: [0x00,0x1f,0x20,0xf2,0x04,0x00,0x00,0x40]
 ; GFX10-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
+;
+; GFX11-LABEL: store_1d_slc:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    image_store v[0:3], v4, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm slc a16 ; encoding: [0x80,0x1f,0x19,0xf0,0x04,0x00,0x00,0x00]
+; GFX11-NEXT:    s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
 main_body:
   %s = extractelement <2 x i16> %coords, i32 0
   call void @llvm.amdgcn.image.store.1d.v4f32.i16(<4 x float> %vdata, i32 15, i16 %s, <8 x i32> %rsrc, i32 0, i32 2)
   ret void
 }
 
 define amdgpu_ps void @store_1d_glc_slc(<8 x i32> inreg %rsrc, <4 x float> %vdata, <2 x i16> %coords) {
 ; GFX9-LABEL: store_1d_glc_slc:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    image_store v[0:3], v4, s[0:7] dmask:0xf unorm glc slc a16 ; encoding: [0x00,0xbf,0x20,0xf2,0x04,0x00,0x00,0x00]
 ; GFX9-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
 ;
 ; GFX10-LABEL: store_1d_glc_slc:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    image_store v[0:3], v4, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm glc slc a16 ; encoding: [0x00,0x3f,0x20,0xf2,0x04,0x00,0x00,0x40]
 ; GFX10-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
+;
+; GFX11-LABEL: store_1d_glc_slc:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    image_store v[0:3], v4, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm glc slc a16 ; encoding: [0x80,0x5f,0x19,0xf0,0x04,0x00,0x00,0x00]
+; GFX11-NEXT:    s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
 main_body:
   %s = extractelement <2 x i16> %coords, i32 0
   call void @llvm.amdgcn.image.store.1d.v4f32.i16(<4 x float> %vdata, i32 15, i16 %s, <8 x i32> %rsrc, i32 0, i32 3)
   ret void
 }
 
 define amdgpu_ps <4 x float> @getresinfo_dmask0(<8 x i32> inreg %rsrc, <4 x float> %vdata, <2 x i16> %coords) #0 {
 ; GFX9-LABEL: getresinfo_dmask0:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    ; return to shader part epilog
 ;
 ; GFX10-LABEL: getresinfo_dmask0:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: getresinfo_dmask0:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %mip = extractelement <2 x i16> %coords, i32 0
   %r = call <4 x float> @llvm.amdgcn.image.getresinfo.1d.v4f32.i16(i32 0, i16 %mip, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x float> %r
 }
 
 declare <4 x float> @llvm.amdgcn.image.load.1d.v4f32.i16(i32, i16, <8 x i32>, i32, i32) #1
 declare <4 x float> @llvm.amdgcn.image.load.2d.v4f32.i16(i32, i16, i16, <8 x i32>, i32, i32) #1

llvm/test/CodeGen/AMDGPU/llvm.amdgcn.image.atomic.dim.ll

 ; RUN: llc -march=amdgcn -mcpu=verde -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=GFX6789 %s
 ; RUN: llc -march=amdgcn -mcpu=gfx900 -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=GFX6789 %s
 ; RUN: llc -march=amdgcn -mcpu=gfx90a -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=GFX90A %s
 ; RUN: llc -march=amdgcn -mcpu=gfx1010 -verify-machineinstrs -show-mc-encoding < %s | FileCheck -check-prefix=GCN -check-prefix=GFX10 %s
+; RUN: llc -march=amdgcn -mcpu=gfx1100 -verify-machineinstrs -show-mc-encoding < %s | FileCheck -check-prefix=GCN -check-prefix=GFX10 %s
 
 ; GCN-LABEL: {{^}}atomic_swap_1d:
 ; GFX6789: image_atomic_swap v0, v1, s[0:7] dmask:0x1 unorm glc{{$}}
 ; GFX90A: image_atomic_swap v0, v{{[02468]}}, s[0:7] dmask:0x1 unorm glc{{$}}
 ; GFX10: image_atomic_swap v0, v1, s[0:7] dmask:0x1 dim:SQ_RSRC_IMG_1D unorm glc ;
 define amdgpu_ps float @atomic_swap_1d(<8 x i32> inreg %rsrc, i32 %data, i32 %s) {
 main_body:
   %v = call i32 @llvm.amdgcn.image.atomic.swap.1d.i32.i32(i32 %data, i32 %s, <8 x i32> %rsrc, i32 0, i32 0)

llvm/test/CodeGen/AMDGPU/llvm.amdgcn.image.d16.dim.ll

 ; RUN: llc < %s -march=amdgcn -mcpu=tonga -verify-machineinstrs | FileCheck -check-prefixes=GCN,UNPACKED,GFX89 %s
 ; RUN: llc < %s -march=amdgcn -mcpu=gfx810 -verify-machineinstrs | FileCheck -check-prefixes=GCN,GFX81,GFX89 %s
 ; RUN: llc < %s -march=amdgcn -mcpu=gfx900 -verify-machineinstrs | FileCheck -check-prefixes=GCN,PACKED,GFX89 %s
 ; RUN: llc < %s -march=amdgcn -mcpu=gfx1010 -verify-machineinstrs | FileCheck -check-prefixes=GCN,GFX10 %s
+; RUN: llc < %s -march=amdgcn -mcpu=gfx1100 -verify-machineinstrs | FileCheck -check-prefixes=GCN,GFX10 %s
 
 ; GCN-LABEL: {{^}}image_load_f16:
 ; GFX89: image_load v0, v[0:1], s[0:7] dmask:0x1 unorm d16{{$}}
 ; GFX10: image_load v0, v[0:1], s[0:7] dmask:0x1 dim:SQ_RSRC_IMG_2D unorm d16{{$}}
 define amdgpu_ps half @image_load_f16(<8 x i32> inreg %rsrc, i32 %s, i32 %t) {
 main_body:
   %tex = call half @llvm.amdgcn.image.load.2d.f16.i32(i32 1, i32 %s, i32 %t, <8 x i32> %rsrc, i32 0, i32 0)
   ret half %tex

llvm/test/CodeGen/AMDGPU/llvm.amdgcn.image.dim.ll

 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
 ; RUN: llc -march=amdgcn -mcpu=verde -verify-machineinstrs < %s | FileCheck -check-prefixes=VERDE %s
 ; RUN: llc -march=amdgcn -mcpu=fiji -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefixes=FIJI %s
 ; RUN: llc -march=amdgcn -mcpu=gfx900 -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX6789 %s
 ; RUN: llc -march=amdgcn -mcpu=gfx900 -mattr=-enable-prt-strict-null -verify-machineinstrs < %s | FileCheck -check-prefixes=NOPRT %s
-; RUN: llc -march=amdgcn -mcpu=gfx1010 -verify-machineinstrs -show-mc-encoding < %s | FileCheck -check-prefixes=GFX10 %s
+; RUN: llc -march=amdgcn -mcpu=gfx1010 -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX10PLUS,GFX10 %s
+; RUN: llc -march=amdgcn -mcpu=gfx1100 -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX10PLUS,GFX11 %s
 
 define amdgpu_ps <4 x float> @load_1d(<8 x i32> inreg %rsrc, i32 %s) {
 ; VERDE-LABEL: load_1d:
 ; VERDE:       ; %bb.0: ; %main_body
 ; VERDE-NEXT:    image_load v[0:3], v0, s[0:7] dmask:0xf unorm
 ; VERDE-NEXT:    s_waitcnt vmcnt(0)
 ; VERDE-NEXT:    ; return to shader part epilog
 ;

 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
 ; NOPRT-LABEL: load_1d:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    image_load v[0:3], v0, s[0:7] dmask:0xf unorm
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: load_1d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_load v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm ; encoding: [0x00,0x1f,0x00,0xf0,0x00,0x00,0x00,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: load_1d:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_load v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.load.1d.v4f32.i32(i32 15, i32 %s, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @load_1d_tfe(<8 x i32> inreg %rsrc, i32 addrspace(1)* inreg %out, i32 %s) {
 ; VERDE-LABEL: load_1d_tfe:
 ; VERDE:       ; %bb.0: ; %main_body

 ; NOPRT-NEXT:    v_mov_b32_e32 v5, 0
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    global_store_dword v5, v4, s[8:9]
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
 ; GFX10-LABEL: load_1d_tfe:
 ; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    v_mov_b32_e32 v6, 0 ; encoding: [0x80,0x02,0x0c,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v5, v0 ; encoding: [0x00,0x03,0x0a,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v7, v6 ; encoding: [0x06,0x03,0x0e,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v8, v6 ; encoding: [0x06,0x03,0x10,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v9, v6 ; encoding: [0x06,0x03,0x12,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v10, v6 ; encoding: [0x06,0x03,0x14,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v0, v6 ; encoding: [0x06,0x03,0x00,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v1, v7 ; encoding: [0x07,0x03,0x02,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v2, v8 ; encoding: [0x08,0x03,0x04,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v3, v9 ; encoding: [0x09,0x03,0x06,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v4, v10 ; encoding: [0x0a,0x03,0x08,0x7e]
-; GFX10-NEXT:    image_load v[0:4], v5, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm tfe ; encoding: [0x00,0x1f,0x01,0xf0,0x05,0x00,0x00,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    global_store_dword v6, v4, s[8:9] ; encoding: [0x00,0x80,0x70,0xdc,0x06,0x04,0x08,0x00]
-; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0 ; encoding: [0x00,0x00,0xfd,0xbb]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10-NEXT:    v_mov_b32_e32 v6, 0
+; GFX10-NEXT:    v_mov_b32_e32 v5, v0
+; GFX10-NEXT:    v_mov_b32_e32 v7, v6
+; GFX10-NEXT:    v_mov_b32_e32 v8, v6
+; GFX10-NEXT:    v_mov_b32_e32 v9, v6
+; GFX10-NEXT:    v_mov_b32_e32 v10, v6
+; GFX10-NEXT:    v_mov_b32_e32 v0, v6
+; GFX10-NEXT:    v_mov_b32_e32 v1, v7
+; GFX10-NEXT:    v_mov_b32_e32 v2, v8
+; GFX10-NEXT:    v_mov_b32_e32 v3, v9
+; GFX10-NEXT:    v_mov_b32_e32 v4, v10
+; GFX10-NEXT:    image_load v[0:4], v5, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm tfe
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    global_store_dword v6, v4, s[8:9]
+; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: load_1d_tfe:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    v_mov_b32_e32 v6, 0
+; GFX11-NEXT:    v_mov_b32_e32 v5, v0
+; GFX11-NEXT:    v_mov_b32_e32 v7, v6
+; GFX11-NEXT:    v_mov_b32_e32 v8, v6
+; GFX11-NEXT:    v_mov_b32_e32 v9, v6
+; GFX11-NEXT:    v_mov_b32_e32 v10, v6
+; GFX11-NEXT:    v_mov_b32_e32 v0, v6
+; GFX11-NEXT:    v_mov_b32_e32 v1, v7
+; GFX11-NEXT:    v_mov_b32_e32 v2, v8
+; GFX11-NEXT:    v_mov_b32_e32 v3, v9
+; GFX11-NEXT:    v_mov_b32_e32 v4, v10
+; GFX11-NEXT:    image_load v[0:4], v5, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm tfe
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    global_store_b32 v6, v4, s[8:9]
+; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %v = call {<4 x float>,i32} @llvm.amdgcn.image.load.1d.v4f32i32.i32(i32 15, i32 %s, <8 x i32> %rsrc, i32 1, i32 0)
   %v.vec = extractvalue {<4 x float>, i32} %v, 0
   %v.err = extractvalue {<4 x float>, i32} %v, 1
   store i32 %v.err, i32 addrspace(1)* %out, align 4
   ret <4 x float> %v.vec
 }
 

 ; NOPRT-NEXT:    v_mov_b32_e32 v5, 0
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    global_store_dword v5, v4, s[8:9]
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
 ; GFX10-LABEL: load_1d_lwe:
 ; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    v_mov_b32_e32 v6, 0 ; encoding: [0x80,0x02,0x0c,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v5, v0 ; encoding: [0x00,0x03,0x0a,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v7, v6 ; encoding: [0x06,0x03,0x0e,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v8, v6 ; encoding: [0x06,0x03,0x10,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v9, v6 ; encoding: [0x06,0x03,0x12,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v10, v6 ; encoding: [0x06,0x03,0x14,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v0, v6 ; encoding: [0x06,0x03,0x00,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v1, v7 ; encoding: [0x07,0x03,0x02,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v2, v8 ; encoding: [0x08,0x03,0x04,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v3, v9 ; encoding: [0x09,0x03,0x06,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v4, v10 ; encoding: [0x0a,0x03,0x08,0x7e]
-; GFX10-NEXT:    image_load v[0:4], v5, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm lwe ; encoding: [0x00,0x1f,0x02,0xf0,0x05,0x00,0x00,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    global_store_dword v6, v4, s[8:9] ; encoding: [0x00,0x80,0x70,0xdc,0x06,0x04,0x08,0x00]
-; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0 ; encoding: [0x00,0x00,0xfd,0xbb]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10-NEXT:    v_mov_b32_e32 v6, 0
+; GFX10-NEXT:    v_mov_b32_e32 v5, v0
+; GFX10-NEXT:    v_mov_b32_e32 v7, v6
+; GFX10-NEXT:    v_mov_b32_e32 v8, v6
+; GFX10-NEXT:    v_mov_b32_e32 v9, v6
+; GFX10-NEXT:    v_mov_b32_e32 v10, v6
+; GFX10-NEXT:    v_mov_b32_e32 v0, v6
+; GFX10-NEXT:    v_mov_b32_e32 v1, v7
+; GFX10-NEXT:    v_mov_b32_e32 v2, v8
+; GFX10-NEXT:    v_mov_b32_e32 v3, v9
+; GFX10-NEXT:    v_mov_b32_e32 v4, v10
+; GFX10-NEXT:    image_load v[0:4], v5, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm lwe
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    global_store_dword v6, v4, s[8:9]
+; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: load_1d_lwe:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    v_mov_b32_e32 v6, 0
+; GFX11-NEXT:    v_mov_b32_e32 v5, v0
+; GFX11-NEXT:    v_mov_b32_e32 v7, v6
+; GFX11-NEXT:    v_mov_b32_e32 v8, v6
+; GFX11-NEXT:    v_mov_b32_e32 v9, v6
+; GFX11-NEXT:    v_mov_b32_e32 v10, v6
+; GFX11-NEXT:    v_mov_b32_e32 v0, v6
+; GFX11-NEXT:    v_mov_b32_e32 v1, v7
+; GFX11-NEXT:    v_mov_b32_e32 v2, v8
+; GFX11-NEXT:    v_mov_b32_e32 v3, v9
+; GFX11-NEXT:    v_mov_b32_e32 v4, v10
+; GFX11-NEXT:    image_load v[0:4], v5, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm lwe
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    global_store_b32 v6, v4, s[8:9]
+; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %v = call {<4 x float>, i32} @llvm.amdgcn.image.load.1d.v4f32i32.i32(i32 15, i32 %s, <8 x i32> %rsrc, i32 2, i32 0)
   %v.vec = extractvalue {<4 x float>, i32} %v, 0
   %v.err = extractvalue {<4 x float>, i32} %v, 1
   store i32 %v.err, i32 addrspace(1)* %out, align 4
   ret <4 x float> %v.vec
 }
 

 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
 ; NOPRT-LABEL: load_2d:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    image_load v[0:3], v[0:1], s[0:7] dmask:0xf unorm
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: load_2d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_load v[0:3], v[0:1], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D unorm ; encoding: [0x08,0x1f,0x00,0xf0,0x00,0x00,0x00,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: load_2d:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_load v[0:3], v[0:1], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D unorm
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.load.2d.v4f32.i32(i32 15, i32 %s, i32 %t, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @load_2d_tfe(<8 x i32> inreg %rsrc, i32 addrspace(1)* inreg %out, i32 %s, i32 %t) {
 ; VERDE-LABEL: load_2d_tfe:
 ; VERDE:       ; %bb.0: ; %main_body

 ; NOPRT-NEXT:    v_mov_b32_e32 v5, 0
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    global_store_dword v5, v4, s[8:9]
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
 ; GFX10-LABEL: load_2d_tfe:
 ; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    v_mov_b32_e32 v7, 0 ; encoding: [0x80,0x02,0x0e,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v6, v1 ; encoding: [0x01,0x03,0x0c,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v5, v0 ; encoding: [0x00,0x03,0x0a,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v8, v7 ; encoding: [0x07,0x03,0x10,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v9, v7 ; encoding: [0x07,0x03,0x12,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v10, v7 ; encoding: [0x07,0x03,0x14,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v11, v7 ; encoding: [0x07,0x03,0x16,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v0, v7 ; encoding: [0x07,0x03,0x00,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v1, v8 ; encoding: [0x08,0x03,0x02,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v2, v9 ; encoding: [0x09,0x03,0x04,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v3, v10 ; encoding: [0x0a,0x03,0x06,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v4, v11 ; encoding: [0x0b,0x03,0x08,0x7e]
-; GFX10-NEXT:    image_load v[0:4], v[5:6], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D unorm tfe ; encoding: [0x08,0x1f,0x01,0xf0,0x05,0x00,0x00,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    global_store_dword v7, v4, s[8:9] ; encoding: [0x00,0x80,0x70,0xdc,0x07,0x04,0x08,0x00]
-; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0 ; encoding: [0x00,0x00,0xfd,0xbb]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10-NEXT:    v_mov_b32_e32 v7, 0
+; GFX10-NEXT:    v_mov_b32_e32 v6, v1
+; GFX10-NEXT:    v_mov_b32_e32 v5, v0
+; GFX10-NEXT:    v_mov_b32_e32 v8, v7
+; GFX10-NEXT:    v_mov_b32_e32 v9, v7
+; GFX10-NEXT:    v_mov_b32_e32 v10, v7
+; GFX10-NEXT:    v_mov_b32_e32 v11, v7
+; GFX10-NEXT:    v_mov_b32_e32 v0, v7
+; GFX10-NEXT:    v_mov_b32_e32 v1, v8
+; GFX10-NEXT:    v_mov_b32_e32 v2, v9
+; GFX10-NEXT:    v_mov_b32_e32 v3, v10
+; GFX10-NEXT:    v_mov_b32_e32 v4, v11
+; GFX10-NEXT:    image_load v[0:4], v[5:6], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D unorm tfe
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    global_store_dword v7, v4, s[8:9]
+; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: load_2d_tfe:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    v_mov_b32_e32 v7, 0
+; GFX11-NEXT:    v_mov_b32_e32 v6, v1
+; GFX11-NEXT:    v_mov_b32_e32 v5, v0
+; GFX11-NEXT:    v_mov_b32_e32 v8, v7
+; GFX11-NEXT:    v_mov_b32_e32 v9, v7
+; GFX11-NEXT:    v_mov_b32_e32 v10, v7
+; GFX11-NEXT:    v_mov_b32_e32 v11, v7
+; GFX11-NEXT:    v_mov_b32_e32 v0, v7
+; GFX11-NEXT:    v_mov_b32_e32 v1, v8
+; GFX11-NEXT:    v_mov_b32_e32 v2, v9
+; GFX11-NEXT:    v_mov_b32_e32 v3, v10
+; GFX11-NEXT:    v_mov_b32_e32 v4, v11
+; GFX11-NEXT:    image_load v[0:4], v[5:6], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D unorm tfe
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    global_store_b32 v7, v4, s[8:9]
+; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %v = call {<4 x float>,i32} @llvm.amdgcn.image.load.2d.v4f32i32.i32(i32 15, i32 %s, i32 %t, <8 x i32> %rsrc, i32 1, i32 0)
   %v.vec = extractvalue {<4 x float>, i32} %v, 0
   %v.err = extractvalue {<4 x float>, i32} %v, 1
   store i32 %v.err, i32 addrspace(1)* %out, align 4
   ret <4 x float> %v.vec
 }
 

 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
 ; NOPRT-LABEL: load_3d:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    image_load v[0:3], v[0:2], s[0:7] dmask:0xf unorm
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: load_3d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_load v[0:3], v[0:2], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_3D unorm ; encoding: [0x10,0x1f,0x00,0xf0,0x00,0x00,0x00,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: load_3d:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_load v[0:3], v[0:2], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_3D unorm
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.load.3d.v4f32.i32(i32 15, i32 %s, i32 %t, i32 %r, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @load_3d_tfe_lwe(<8 x i32> inreg %rsrc, i32 addrspace(1)* inreg %out, i32 %s, i32 %t, i32 %r) {
 ; VERDE-LABEL: load_3d_tfe_lwe:
 ; VERDE:       ; %bb.0: ; %main_body

 ; NOPRT-NEXT:    v_mov_b32_e32 v5, 0
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    global_store_dword v5, v4, s[8:9]
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
 ; GFX10-LABEL: load_3d_tfe_lwe:
 ; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    v_mov_b32_e32 v8, 0 ; encoding: [0x80,0x02,0x10,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v7, v2 ; encoding: [0x02,0x03,0x0e,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v6, v1 ; encoding: [0x01,0x03,0x0c,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v5, v0 ; encoding: [0x00,0x03,0x0a,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v9, v8 ; encoding: [0x08,0x03,0x12,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v10, v8 ; encoding: [0x08,0x03,0x14,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v11, v8 ; encoding: [0x08,0x03,0x16,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v12, v8 ; encoding: [0x08,0x03,0x18,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v0, v8 ; encoding: [0x08,0x03,0x00,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v1, v9 ; encoding: [0x09,0x03,0x02,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v2, v10 ; encoding: [0x0a,0x03,0x04,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v3, v11 ; encoding: [0x0b,0x03,0x06,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v4, v12 ; encoding: [0x0c,0x03,0x08,0x7e]
-; GFX10-NEXT:    image_load v[0:4], v[5:7], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_3D unorm tfe lwe ; encoding: [0x10,0x1f,0x03,0xf0,0x05,0x00,0x00,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    global_store_dword v8, v4, s[8:9] ; encoding: [0x00,0x80,0x70,0xdc,0x08,0x04,0x08,0x00]
-; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0 ; encoding: [0x00,0x00,0xfd,0xbb]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10-NEXT:    v_mov_b32_e32 v8, 0
+; GFX10-NEXT:    v_mov_b32_e32 v7, v2
+; GFX10-NEXT:    v_mov_b32_e32 v6, v1
+; GFX10-NEXT:    v_mov_b32_e32 v5, v0
+; GFX10-NEXT:    v_mov_b32_e32 v9, v8
+; GFX10-NEXT:    v_mov_b32_e32 v10, v8
+; GFX10-NEXT:    v_mov_b32_e32 v11, v8
+; GFX10-NEXT:    v_mov_b32_e32 v12, v8
+; GFX10-NEXT:    v_mov_b32_e32 v0, v8
+; GFX10-NEXT:    v_mov_b32_e32 v1, v9
+; GFX10-NEXT:    v_mov_b32_e32 v2, v10
+; GFX10-NEXT:    v_mov_b32_e32 v3, v11
+; GFX10-NEXT:    v_mov_b32_e32 v4, v12
+; GFX10-NEXT:    image_load v[0:4], v[5:7], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_3D unorm tfe lwe
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    global_store_dword v8, v4, s[8:9]
+; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: load_3d_tfe_lwe:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    v_mov_b32_e32 v8, 0
+; GFX11-NEXT:    v_mov_b32_e32 v7, v2
+; GFX11-NEXT:    v_mov_b32_e32 v6, v1
+; GFX11-NEXT:    v_mov_b32_e32 v5, v0
+; GFX11-NEXT:    v_mov_b32_e32 v9, v8
+; GFX11-NEXT:    v_mov_b32_e32 v10, v8
+; GFX11-NEXT:    v_mov_b32_e32 v11, v8
+; GFX11-NEXT:    v_mov_b32_e32 v12, v8
+; GFX11-NEXT:    v_mov_b32_e32 v0, v8
+; GFX11-NEXT:    v_mov_b32_e32 v1, v9
+; GFX11-NEXT:    v_mov_b32_e32 v2, v10
+; GFX11-NEXT:    v_mov_b32_e32 v3, v11
+; GFX11-NEXT:    v_mov_b32_e32 v4, v12
+; GFX11-NEXT:    image_load v[0:4], v[5:7], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_3D unorm tfe lwe
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    global_store_b32 v8, v4, s[8:9]
+; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %v = call {<4 x float>,i32} @llvm.amdgcn.image.load.3d.v4f32i32.i32(i32 15, i32 %s, i32 %t, i32 %r, <8 x i32> %rsrc, i32 3, i32 0)
   %v.vec = extractvalue {<4 x float>, i32} %v, 0
   %v.err = extractvalue {<4 x float>, i32} %v, 1
   store i32 %v.err, i32 addrspace(1)* %out, align 4
   ret <4 x float> %v.vec
 }
 

 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
 ; NOPRT-LABEL: load_cube:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    image_load v[0:3], v[0:2], s[0:7] dmask:0xf unorm da
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: load_cube:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_load v[0:3], v[0:2], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_CUBE unorm ; encoding: [0x18,0x1f,0x00,0xf0,0x00,0x00,0x00,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: load_cube:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_load v[0:3], v[0:2], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_CUBE unorm
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.load.cube.v4f32.i32(i32 15, i32 %s, i32 %t, i32 %slice, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @load_cube_lwe(<8 x i32> inreg %rsrc, i32 addrspace(1)* inreg %out, i32 %s, i32 %t, i32 %slice) {
 ; VERDE-LABEL: load_cube_lwe:
 ; VERDE:       ; %bb.0: ; %main_body

 ; NOPRT-NEXT:    v_mov_b32_e32 v5, 0
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    global_store_dword v5, v4, s[8:9]
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
 ; GFX10-LABEL: load_cube_lwe:
 ; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    v_mov_b32_e32 v8, 0 ; encoding: [0x80,0x02,0x10,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v7, v2 ; encoding: [0x02,0x03,0x0e,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v6, v1 ; encoding: [0x01,0x03,0x0c,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v5, v0 ; encoding: [0x00,0x03,0x0a,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v9, v8 ; encoding: [0x08,0x03,0x12,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v10, v8 ; encoding: [0x08,0x03,0x14,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v11, v8 ; encoding: [0x08,0x03,0x16,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v12, v8 ; encoding: [0x08,0x03,0x18,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v0, v8 ; encoding: [0x08,0x03,0x00,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v1, v9 ; encoding: [0x09,0x03,0x02,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v2, v10 ; encoding: [0x0a,0x03,0x04,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v3, v11 ; encoding: [0x0b,0x03,0x06,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v4, v12 ; encoding: [0x0c,0x03,0x08,0x7e]
-; GFX10-NEXT:    image_load v[0:4], v[5:7], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_CUBE unorm lwe ; encoding: [0x18,0x1f,0x02,0xf0,0x05,0x00,0x00,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    global_store_dword v8, v4, s[8:9] ; encoding: [0x00,0x80,0x70,0xdc,0x08,0x04,0x08,0x00]
-; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0 ; encoding: [0x00,0x00,0xfd,0xbb]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10-NEXT:    v_mov_b32_e32 v8, 0
+; GFX10-NEXT:    v_mov_b32_e32 v7, v2
+; GFX10-NEXT:    v_mov_b32_e32 v6, v1
+; GFX10-NEXT:    v_mov_b32_e32 v5, v0
+; GFX10-NEXT:    v_mov_b32_e32 v9, v8
+; GFX10-NEXT:    v_mov_b32_e32 v10, v8
+; GFX10-NEXT:    v_mov_b32_e32 v11, v8
+; GFX10-NEXT:    v_mov_b32_e32 v12, v8
+; GFX10-NEXT:    v_mov_b32_e32 v0, v8
+; GFX10-NEXT:    v_mov_b32_e32 v1, v9
+; GFX10-NEXT:    v_mov_b32_e32 v2, v10
+; GFX10-NEXT:    v_mov_b32_e32 v3, v11
+; GFX10-NEXT:    v_mov_b32_e32 v4, v12
+; GFX10-NEXT:    image_load v[0:4], v[5:7], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_CUBE unorm lwe
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    global_store_dword v8, v4, s[8:9]
+; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: load_cube_lwe:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    v_mov_b32_e32 v8, 0
+; GFX11-NEXT:    v_mov_b32_e32 v7, v2
+; GFX11-NEXT:    v_mov_b32_e32 v6, v1
+; GFX11-NEXT:    v_mov_b32_e32 v5, v0
+; GFX11-NEXT:    v_mov_b32_e32 v9, v8
+; GFX11-NEXT:    v_mov_b32_e32 v10, v8
+; GFX11-NEXT:    v_mov_b32_e32 v11, v8
+; GFX11-NEXT:    v_mov_b32_e32 v12, v8
+; GFX11-NEXT:    v_mov_b32_e32 v0, v8
+; GFX11-NEXT:    v_mov_b32_e32 v1, v9
+; GFX11-NEXT:    v_mov_b32_e32 v2, v10
+; GFX11-NEXT:    v_mov_b32_e32 v3, v11
+; GFX11-NEXT:    v_mov_b32_e32 v4, v12
+; GFX11-NEXT:    image_load v[0:4], v[5:7], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_CUBE unorm lwe
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    global_store_b32 v8, v4, s[8:9]
+; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %v = call {<4 x float>,i32} @llvm.amdgcn.image.load.cube.v4f32i32.i32(i32 15, i32 %s, i32 %t, i32 %slice, <8 x i32> %rsrc, i32 2, i32 0)
   %v.vec = extractvalue {<4 x float>, i32} %v, 0
   %v.err = extractvalue {<4 x float>, i32} %v, 1
   store i32 %v.err, i32 addrspace(1)* %out, align 4
   ret <4 x float> %v.vec
 }
 

 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
 ; NOPRT-LABEL: load_1darray:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    image_load v[0:3], v[0:1], s[0:7] dmask:0xf unorm da
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: load_1darray:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_load v[0:3], v[0:1], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D_ARRAY unorm ; encoding: [0x20,0x1f,0x00,0xf0,0x00,0x00,0x00,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: load_1darray:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_load v[0:3], v[0:1], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D_ARRAY unorm
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.load.1darray.v4f32.i32(i32 15, i32 %s, i32 %slice, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @load_1darray_tfe(<8 x i32> inreg %rsrc, i32 addrspace(1)* inreg %out, i32 %s, i32 %slice) {
 ; VERDE-LABEL: load_1darray_tfe:
 ; VERDE:       ; %bb.0: ; %main_body

 ; NOPRT-NEXT:    v_mov_b32_e32 v5, 0
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    global_store_dword v5, v4, s[8:9]
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
 ; GFX10-LABEL: load_1darray_tfe:
 ; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    v_mov_b32_e32 v7, 0 ; encoding: [0x80,0x02,0x0e,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v6, v1 ; encoding: [0x01,0x03,0x0c,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v5, v0 ; encoding: [0x00,0x03,0x0a,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v8, v7 ; encoding: [0x07,0x03,0x10,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v9, v7 ; encoding: [0x07,0x03,0x12,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v10, v7 ; encoding: [0x07,0x03,0x14,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v11, v7 ; encoding: [0x07,0x03,0x16,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v0, v7 ; encoding: [0x07,0x03,0x00,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v1, v8 ; encoding: [0x08,0x03,0x02,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v2, v9 ; encoding: [0x09,0x03,0x04,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v3, v10 ; encoding: [0x0a,0x03,0x06,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v4, v11 ; encoding: [0x0b,0x03,0x08,0x7e]
-; GFX10-NEXT:    image_load v[0:4], v[5:6], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D_ARRAY unorm tfe ; encoding: [0x20,0x1f,0x01,0xf0,0x05,0x00,0x00,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    global_store_dword v7, v4, s[8:9] ; encoding: [0x00,0x80,0x70,0xdc,0x07,0x04,0x08,0x00]
-; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0 ; encoding: [0x00,0x00,0xfd,0xbb]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10-NEXT:    v_mov_b32_e32 v7, 0
+; GFX10-NEXT:    v_mov_b32_e32 v6, v1
+; GFX10-NEXT:    v_mov_b32_e32 v5, v0
+; GFX10-NEXT:    v_mov_b32_e32 v8, v7
+; GFX10-NEXT:    v_mov_b32_e32 v9, v7
+; GFX10-NEXT:    v_mov_b32_e32 v10, v7
+; GFX10-NEXT:    v_mov_b32_e32 v11, v7
+; GFX10-NEXT:    v_mov_b32_e32 v0, v7
+; GFX10-NEXT:    v_mov_b32_e32 v1, v8
+; GFX10-NEXT:    v_mov_b32_e32 v2, v9
+; GFX10-NEXT:    v_mov_b32_e32 v3, v10
+; GFX10-NEXT:    v_mov_b32_e32 v4, v11
+; GFX10-NEXT:    image_load v[0:4], v[5:6], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D_ARRAY unorm tfe
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    global_store_dword v7, v4, s[8:9]
+; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: load_1darray_tfe:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    v_mov_b32_e32 v7, 0
+; GFX11-NEXT:    v_mov_b32_e32 v6, v1
+; GFX11-NEXT:    v_mov_b32_e32 v5, v0
+; GFX11-NEXT:    v_mov_b32_e32 v8, v7
+; GFX11-NEXT:    v_mov_b32_e32 v9, v7
+; GFX11-NEXT:    v_mov_b32_e32 v10, v7
+; GFX11-NEXT:    v_mov_b32_e32 v11, v7
+; GFX11-NEXT:    v_mov_b32_e32 v0, v7
+; GFX11-NEXT:    v_mov_b32_e32 v1, v8
+; GFX11-NEXT:    v_mov_b32_e32 v2, v9
+; GFX11-NEXT:    v_mov_b32_e32 v3, v10
+; GFX11-NEXT:    v_mov_b32_e32 v4, v11
+; GFX11-NEXT:    image_load v[0:4], v[5:6], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D_ARRAY unorm tfe
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    global_store_b32 v7, v4, s[8:9]
+; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %v = call {<4 x float>,i32} @llvm.amdgcn.image.load.1darray.v4f32i32.i32(i32 15, i32 %s, i32 %slice, <8 x i32> %rsrc, i32 1, i32 0)
   %v.vec = extractvalue {<4 x float>, i32} %v, 0
   %v.err = extractvalue {<4 x float>, i32} %v, 1
   store i32 %v.err, i32 addrspace(1)* %out, align 4
   ret <4 x float> %v.vec
 }
 

 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
 ; NOPRT-LABEL: load_2darray:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    image_load v[0:3], v[0:2], s[0:7] dmask:0xf unorm da
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: load_2darray:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_load v[0:3], v[0:2], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_ARRAY unorm ; encoding: [0x28,0x1f,0x00,0xf0,0x00,0x00,0x00,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: load_2darray:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_load v[0:3], v[0:2], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_ARRAY unorm
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.load.2darray.v4f32.i32(i32 15, i32 %s, i32 %t, i32 %slice, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @load_2darray_lwe(<8 x i32> inreg %rsrc, i32 addrspace(1)* inreg %out, i32 %s, i32 %t, i32 %slice) {
 ; VERDE-LABEL: load_2darray_lwe:
 ; VERDE:       ; %bb.0: ; %main_body

 ; NOPRT-NEXT:    v_mov_b32_e32 v5, 0
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    global_store_dword v5, v4, s[8:9]
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
 ; GFX10-LABEL: load_2darray_lwe:
 ; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    v_mov_b32_e32 v8, 0 ; encoding: [0x80,0x02,0x10,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v7, v2 ; encoding: [0x02,0x03,0x0e,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v6, v1 ; encoding: [0x01,0x03,0x0c,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v5, v0 ; encoding: [0x00,0x03,0x0a,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v9, v8 ; encoding: [0x08,0x03,0x12,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v10, v8 ; encoding: [0x08,0x03,0x14,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v11, v8 ; encoding: [0x08,0x03,0x16,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v12, v8 ; encoding: [0x08,0x03,0x18,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v0, v8 ; encoding: [0x08,0x03,0x00,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v1, v9 ; encoding: [0x09,0x03,0x02,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v2, v10 ; encoding: [0x0a,0x03,0x04,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v3, v11 ; encoding: [0x0b,0x03,0x06,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v4, v12 ; encoding: [0x0c,0x03,0x08,0x7e]
-; GFX10-NEXT:    image_load v[0:4], v[5:7], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_ARRAY unorm lwe ; encoding: [0x28,0x1f,0x02,0xf0,0x05,0x00,0x00,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    global_store_dword v8, v4, s[8:9] ; encoding: [0x00,0x80,0x70,0xdc,0x08,0x04,0x08,0x00]
-; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0 ; encoding: [0x00,0x00,0xfd,0xbb]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10-NEXT:    v_mov_b32_e32 v8, 0
+; GFX10-NEXT:    v_mov_b32_e32 v7, v2
+; GFX10-NEXT:    v_mov_b32_e32 v6, v1
+; GFX10-NEXT:    v_mov_b32_e32 v5, v0
+; GFX10-NEXT:    v_mov_b32_e32 v9, v8
+; GFX10-NEXT:    v_mov_b32_e32 v10, v8
+; GFX10-NEXT:    v_mov_b32_e32 v11, v8
+; GFX10-NEXT:    v_mov_b32_e32 v12, v8
+; GFX10-NEXT:    v_mov_b32_e32 v0, v8
+; GFX10-NEXT:    v_mov_b32_e32 v1, v9
+; GFX10-NEXT:    v_mov_b32_e32 v2, v10
+; GFX10-NEXT:    v_mov_b32_e32 v3, v11
+; GFX10-NEXT:    v_mov_b32_e32 v4, v12
+; GFX10-NEXT:    image_load v[0:4], v[5:7], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_ARRAY unorm lwe
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    global_store_dword v8, v4, s[8:9]
+; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: load_2darray_lwe:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    v_mov_b32_e32 v8, 0
+; GFX11-NEXT:    v_mov_b32_e32 v7, v2
+; GFX11-NEXT:    v_mov_b32_e32 v6, v1
+; GFX11-NEXT:    v_mov_b32_e32 v5, v0
+; GFX11-NEXT:    v_mov_b32_e32 v9, v8
+; GFX11-NEXT:    v_mov_b32_e32 v10, v8
+; GFX11-NEXT:    v_mov_b32_e32 v11, v8
+; GFX11-NEXT:    v_mov_b32_e32 v12, v8
+; GFX11-NEXT:    v_mov_b32_e32 v0, v8
+; GFX11-NEXT:    v_mov_b32_e32 v1, v9
+; GFX11-NEXT:    v_mov_b32_e32 v2, v10
+; GFX11-NEXT:    v_mov_b32_e32 v3, v11
+; GFX11-NEXT:    v_mov_b32_e32 v4, v12
+; GFX11-NEXT:    image_load v[0:4], v[5:7], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_ARRAY unorm lwe
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    global_store_b32 v8, v4, s[8:9]
+; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %v = call {<4 x float>,i32} @llvm.amdgcn.image.load.2darray.v4f32i32.i32(i32 15, i32 %s, i32 %t, i32 %slice, <8 x i32> %rsrc, i32 2, i32 0)
   %v.vec = extractvalue {<4 x float>, i32} %v, 0
   %v.err = extractvalue {<4 x float>, i32} %v, 1
   store i32 %v.err, i32 addrspace(1)* %out, align 4
   ret <4 x float> %v.vec
 }
 

 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
 ; NOPRT-LABEL: load_2dmsaa:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    image_load v[0:3], v[0:2], s[0:7] dmask:0xf unorm
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: load_2dmsaa:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_load v[0:3], v[0:2], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_MSAA unorm ; encoding: [0x30,0x1f,0x00,0xf0,0x00,0x00,0x00,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: load_2dmsaa:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_load v[0:3], v[0:2], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_MSAA unorm
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.load.2dmsaa.v4f32.i32(i32 15, i32 %s, i32 %t, i32 %fragid, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @load_2dmsaa_both(<8 x i32> inreg %rsrc, i32 addrspace(1)* inreg %out, i32 %s, i32 %t, i32 %fragid) {
 ; VERDE-LABEL: load_2dmsaa_both:
 ; VERDE:       ; %bb.0: ; %main_body

 ; NOPRT-NEXT:    v_mov_b32_e32 v5, 0
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    global_store_dword v5, v4, s[8:9]
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
 ; GFX10-LABEL: load_2dmsaa_both:
 ; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    v_mov_b32_e32 v8, 0 ; encoding: [0x80,0x02,0x10,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v7, v2 ; encoding: [0x02,0x03,0x0e,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v6, v1 ; encoding: [0x01,0x03,0x0c,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v5, v0 ; encoding: [0x00,0x03,0x0a,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v9, v8 ; encoding: [0x08,0x03,0x12,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v10, v8 ; encoding: [0x08,0x03,0x14,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v11, v8 ; encoding: [0x08,0x03,0x16,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v12, v8 ; encoding: [0x08,0x03,0x18,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v0, v8 ; encoding: [0x08,0x03,0x00,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v1, v9 ; encoding: [0x09,0x03,0x02,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v2, v10 ; encoding: [0x0a,0x03,0x04,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v3, v11 ; encoding: [0x0b,0x03,0x06,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v4, v12 ; encoding: [0x0c,0x03,0x08,0x7e]
-; GFX10-NEXT:    image_load v[0:4], v[5:7], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_MSAA unorm tfe lwe ; encoding: [0x30,0x1f,0x03,0xf0,0x05,0x00,0x00,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    global_store_dword v8, v4, s[8:9] ; encoding: [0x00,0x80,0x70,0xdc,0x08,0x04,0x08,0x00]
-; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0 ; encoding: [0x00,0x00,0xfd,0xbb]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10-NEXT:    v_mov_b32_e32 v8, 0
+; GFX10-NEXT:    v_mov_b32_e32 v7, v2
+; GFX10-NEXT:    v_mov_b32_e32 v6, v1
+; GFX10-NEXT:    v_mov_b32_e32 v5, v0
+; GFX10-NEXT:    v_mov_b32_e32 v9, v8
+; GFX10-NEXT:    v_mov_b32_e32 v10, v8
+; GFX10-NEXT:    v_mov_b32_e32 v11, v8
+; GFX10-NEXT:    v_mov_b32_e32 v12, v8
+; GFX10-NEXT:    v_mov_b32_e32 v0, v8
+; GFX10-NEXT:    v_mov_b32_e32 v1, v9
+; GFX10-NEXT:    v_mov_b32_e32 v2, v10
+; GFX10-NEXT:    v_mov_b32_e32 v3, v11
+; GFX10-NEXT:    v_mov_b32_e32 v4, v12
+; GFX10-NEXT:    image_load v[0:4], v[5:7], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_MSAA unorm tfe lwe
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    global_store_dword v8, v4, s[8:9]
+; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: load_2dmsaa_both:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    v_mov_b32_e32 v8, 0
+; GFX11-NEXT:    v_mov_b32_e32 v7, v2
+; GFX11-NEXT:    v_mov_b32_e32 v6, v1
+; GFX11-NEXT:    v_mov_b32_e32 v5, v0
+; GFX11-NEXT:    v_mov_b32_e32 v9, v8
+; GFX11-NEXT:    v_mov_b32_e32 v10, v8
+; GFX11-NEXT:    v_mov_b32_e32 v11, v8
+; GFX11-NEXT:    v_mov_b32_e32 v12, v8
+; GFX11-NEXT:    v_mov_b32_e32 v0, v8
+; GFX11-NEXT:    v_mov_b32_e32 v1, v9
+; GFX11-NEXT:    v_mov_b32_e32 v2, v10
+; GFX11-NEXT:    v_mov_b32_e32 v3, v11
+; GFX11-NEXT:    v_mov_b32_e32 v4, v12
+; GFX11-NEXT:    image_load v[0:4], v[5:7], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_MSAA unorm tfe lwe
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    global_store_b32 v8, v4, s[8:9]
+; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %v = call {<4 x float>,i32} @llvm.amdgcn.image.load.2dmsaa.v4f32i32.i32(i32 15, i32 %s, i32 %t, i32 %fragid, <8 x i32> %rsrc, i32 3, i32 0)
   %v.vec = extractvalue {<4 x float>, i32} %v, 0
   %v.err = extractvalue {<4 x float>, i32} %v, 1
   store i32 %v.err, i32 addrspace(1)* %out, align 4
   ret <4 x float> %v.vec
 }
 

 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
 ; NOPRT-LABEL: load_2darraymsaa:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    image_load v[0:3], v[0:3], s[0:7] dmask:0xf unorm da
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: load_2darraymsaa:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_load v[0:3], v[0:3], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_MSAA_ARRAY unorm ; encoding: [0x38,0x1f,0x00,0xf0,0x00,0x00,0x00,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: load_2darraymsaa:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_load v[0:3], v[0:3], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_MSAA_ARRAY unorm
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.load.2darraymsaa.v4f32.i32(i32 15, i32 %s, i32 %t, i32 %slice, i32 %fragid, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @load_2darraymsaa_tfe(<8 x i32> inreg %rsrc, i32 addrspace(1)* inreg %out, i32 %s, i32 %t, i32 %slice, i32 %fragid) {
 ; VERDE-LABEL: load_2darraymsaa_tfe:
 ; VERDE:       ; %bb.0: ; %main_body

 ; NOPRT-NEXT:    v_mov_b32_e32 v5, 0
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    global_store_dword v5, v4, s[8:9]
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
 ; GFX10-LABEL: load_2darraymsaa_tfe:
 ; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    v_mov_b32_e32 v9, 0 ; encoding: [0x80,0x02,0x12,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v8, v3 ; encoding: [0x03,0x03,0x10,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v7, v2 ; encoding: [0x02,0x03,0x0e,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v6, v1 ; encoding: [0x01,0x03,0x0c,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v5, v0 ; encoding: [0x00,0x03,0x0a,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v10, v9 ; encoding: [0x09,0x03,0x14,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v11, v9 ; encoding: [0x09,0x03,0x16,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v12, v9 ; encoding: [0x09,0x03,0x18,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v13, v9 ; encoding: [0x09,0x03,0x1a,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v0, v9 ; encoding: [0x09,0x03,0x00,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v1, v10 ; encoding: [0x0a,0x03,0x02,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v2, v11 ; encoding: [0x0b,0x03,0x04,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v3, v12 ; encoding: [0x0c,0x03,0x06,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v4, v13 ; encoding: [0x0d,0x03,0x08,0x7e]
-; GFX10-NEXT:    image_load v[0:4], v[5:8], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_MSAA_ARRAY unorm tfe ; encoding: [0x38,0x1f,0x01,0xf0,0x05,0x00,0x00,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    global_store_dword v9, v4, s[8:9] ; encoding: [0x00,0x80,0x70,0xdc,0x09,0x04,0x08,0x00]
-; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0 ; encoding: [0x00,0x00,0xfd,0xbb]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10-NEXT:    v_mov_b32_e32 v9, 0
+; GFX10-NEXT:    v_mov_b32_e32 v8, v3
+; GFX10-NEXT:    v_mov_b32_e32 v7, v2
+; GFX10-NEXT:    v_mov_b32_e32 v6, v1
+; GFX10-NEXT:    v_mov_b32_e32 v5, v0
+; GFX10-NEXT:    v_mov_b32_e32 v10, v9
+; GFX10-NEXT:    v_mov_b32_e32 v11, v9
+; GFX10-NEXT:    v_mov_b32_e32 v12, v9
+; GFX10-NEXT:    v_mov_b32_e32 v13, v9
+; GFX10-NEXT:    v_mov_b32_e32 v0, v9
+; GFX10-NEXT:    v_mov_b32_e32 v1, v10
+; GFX10-NEXT:    v_mov_b32_e32 v2, v11
+; GFX10-NEXT:    v_mov_b32_e32 v3, v12
+; GFX10-NEXT:    v_mov_b32_e32 v4, v13
+; GFX10-NEXT:    image_load v[0:4], v[5:8], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_MSAA_ARRAY unorm tfe
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    global_store_dword v9, v4, s[8:9]
+; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: load_2darraymsaa_tfe:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    v_mov_b32_e32 v9, 0
+; GFX11-NEXT:    v_mov_b32_e32 v8, v3
+; GFX11-NEXT:    v_mov_b32_e32 v7, v2
+; GFX11-NEXT:    v_mov_b32_e32 v6, v1
+; GFX11-NEXT:    v_mov_b32_e32 v5, v0
+; GFX11-NEXT:    v_mov_b32_e32 v10, v9
+; GFX11-NEXT:    v_mov_b32_e32 v11, v9
+; GFX11-NEXT:    v_mov_b32_e32 v12, v9
+; GFX11-NEXT:    v_mov_b32_e32 v13, v9
+; GFX11-NEXT:    v_mov_b32_e32 v0, v9
+; GFX11-NEXT:    v_mov_b32_e32 v1, v10
+; GFX11-NEXT:    v_mov_b32_e32 v2, v11
+; GFX11-NEXT:    v_mov_b32_e32 v3, v12
+; GFX11-NEXT:    v_mov_b32_e32 v4, v13
+; GFX11-NEXT:    image_load v[0:4], v[5:8], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_MSAA_ARRAY unorm tfe
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    global_store_b32 v9, v4, s[8:9]
+; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %v = call {<4 x float>,i32} @llvm.amdgcn.image.load.2darraymsaa.v4f32i32.i32(i32 15, i32 %s, i32 %t, i32 %slice, i32 %fragid, <8 x i32> %rsrc, i32 1, i32 0)
   %v.vec = extractvalue {<4 x float>, i32} %v, 0
   %v.err = extractvalue {<4 x float>, i32} %v, 1
   store i32 %v.err, i32 addrspace(1)* %out, align 4
   ret <4 x float> %v.vec
 }
 

 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
 ; NOPRT-LABEL: load_mip_1d:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    image_load_mip v[0:3], v[0:1], s[0:7] dmask:0xf unorm
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: load_mip_1d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_load_mip v[0:3], v[0:1], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm ; encoding: [0x00,0x1f,0x04,0xf0,0x00,0x00,0x00,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: load_mip_1d:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_load_mip v[0:3], v[0:1], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.load.mip.1d.v4f32.i32(i32 15, i32 %s, i32 %mip, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @load_mip_1d_lwe(<8 x i32> inreg %rsrc, i32 addrspace(1)* inreg %out, i32 %s, i32 %mip) {
 ; VERDE-LABEL: load_mip_1d_lwe:
 ; VERDE:       ; %bb.0: ; %main_body

 ; NOPRT-NEXT:    v_mov_b32_e32 v5, 0
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    global_store_dword v5, v4, s[8:9]
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
 ; GFX10-LABEL: load_mip_1d_lwe:
 ; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    v_mov_b32_e32 v7, 0 ; encoding: [0x80,0x02,0x0e,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v6, v1 ; encoding: [0x01,0x03,0x0c,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v5, v0 ; encoding: [0x00,0x03,0x0a,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v8, v7 ; encoding: [0x07,0x03,0x10,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v9, v7 ; encoding: [0x07,0x03,0x12,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v10, v7 ; encoding: [0x07,0x03,0x14,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v11, v7 ; encoding: [0x07,0x03,0x16,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v0, v7 ; encoding: [0x07,0x03,0x00,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v1, v8 ; encoding: [0x08,0x03,0x02,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v2, v9 ; encoding: [0x09,0x03,0x04,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v3, v10 ; encoding: [0x0a,0x03,0x06,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v4, v11 ; encoding: [0x0b,0x03,0x08,0x7e]
-; GFX10-NEXT:    image_load_mip v[0:4], v[5:6], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm lwe ; encoding: [0x00,0x1f,0x06,0xf0,0x05,0x00,0x00,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    global_store_dword v7, v4, s[8:9] ; encoding: [0x00,0x80,0x70,0xdc,0x07,0x04,0x08,0x00]
-; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0 ; encoding: [0x00,0x00,0xfd,0xbb]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10-NEXT:    v_mov_b32_e32 v7, 0
+; GFX10-NEXT:    v_mov_b32_e32 v6, v1
+; GFX10-NEXT:    v_mov_b32_e32 v5, v0
+; GFX10-NEXT:    v_mov_b32_e32 v8, v7
+; GFX10-NEXT:    v_mov_b32_e32 v9, v7
+; GFX10-NEXT:    v_mov_b32_e32 v10, v7
+; GFX10-NEXT:    v_mov_b32_e32 v11, v7
+; GFX10-NEXT:    v_mov_b32_e32 v0, v7
+; GFX10-NEXT:    v_mov_b32_e32 v1, v8
+; GFX10-NEXT:    v_mov_b32_e32 v2, v9
+; GFX10-NEXT:    v_mov_b32_e32 v3, v10
+; GFX10-NEXT:    v_mov_b32_e32 v4, v11
+; GFX10-NEXT:    image_load_mip v[0:4], v[5:6], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm lwe
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    global_store_dword v7, v4, s[8:9]
+; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: load_mip_1d_lwe:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    v_mov_b32_e32 v7, 0
+; GFX11-NEXT:    v_mov_b32_e32 v6, v1
+; GFX11-NEXT:    v_mov_b32_e32 v5, v0
+; GFX11-NEXT:    v_mov_b32_e32 v8, v7
+; GFX11-NEXT:    v_mov_b32_e32 v9, v7
+; GFX11-NEXT:    v_mov_b32_e32 v10, v7
+; GFX11-NEXT:    v_mov_b32_e32 v11, v7
+; GFX11-NEXT:    v_mov_b32_e32 v0, v7
+; GFX11-NEXT:    v_mov_b32_e32 v1, v8
+; GFX11-NEXT:    v_mov_b32_e32 v2, v9
+; GFX11-NEXT:    v_mov_b32_e32 v3, v10
+; GFX11-NEXT:    v_mov_b32_e32 v4, v11
+; GFX11-NEXT:    image_load_mip v[0:4], v[5:6], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm lwe
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    global_store_b32 v7, v4, s[8:9]
+; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %v = call {<4 x float>,i32} @llvm.amdgcn.image.load.mip.1d.v4f32i32.i32(i32 15, i32 %s, i32 %mip, <8 x i32> %rsrc, i32 2, i32 0)
   %v.vec = extractvalue {<4 x float>, i32} %v, 0
   %v.err = extractvalue {<4 x float>, i32} %v, 1
   store i32 %v.err, i32 addrspace(1)* %out, align 4
   ret <4 x float> %v.vec
 }
 

 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
 ; NOPRT-LABEL: load_mip_2d:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    image_load_mip v[0:3], v[0:2], s[0:7] dmask:0xf unorm
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: load_mip_2d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_load_mip v[0:3], v[0:2], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D unorm ; encoding: [0x08,0x1f,0x04,0xf0,0x00,0x00,0x00,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: load_mip_2d:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_load_mip v[0:3], v[0:2], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D unorm
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.load.mip.2d.v4f32.i32(i32 15, i32 %s, i32 %t, i32 %mip, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @load_mip_2d_tfe(<8 x i32> inreg %rsrc, i32 addrspace(1)* inreg %out, i32 %s, i32 %t, i32 %mip) {
 ; VERDE-LABEL: load_mip_2d_tfe:
 ; VERDE:       ; %bb.0: ; %main_body

 ; NOPRT-NEXT:    v_mov_b32_e32 v5, 0
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    global_store_dword v5, v4, s[8:9]
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
 ; GFX10-LABEL: load_mip_2d_tfe:
 ; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    v_mov_b32_e32 v8, 0 ; encoding: [0x80,0x02,0x10,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v7, v2 ; encoding: [0x02,0x03,0x0e,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v6, v1 ; encoding: [0x01,0x03,0x0c,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v5, v0 ; encoding: [0x00,0x03,0x0a,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v9, v8 ; encoding: [0x08,0x03,0x12,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v10, v8 ; encoding: [0x08,0x03,0x14,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v11, v8 ; encoding: [0x08,0x03,0x16,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v12, v8 ; encoding: [0x08,0x03,0x18,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v0, v8 ; encoding: [0x08,0x03,0x00,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v1, v9 ; encoding: [0x09,0x03,0x02,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v2, v10 ; encoding: [0x0a,0x03,0x04,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v3, v11 ; encoding: [0x0b,0x03,0x06,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v4, v12 ; encoding: [0x0c,0x03,0x08,0x7e]
-; GFX10-NEXT:    image_load_mip v[0:4], v[5:7], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D unorm tfe ; encoding: [0x08,0x1f,0x05,0xf0,0x05,0x00,0x00,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    global_store_dword v8, v4, s[8:9] ; encoding: [0x00,0x80,0x70,0xdc,0x08,0x04,0x08,0x00]
-; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0 ; encoding: [0x00,0x00,0xfd,0xbb]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10-NEXT:    v_mov_b32_e32 v8, 0
+; GFX10-NEXT:    v_mov_b32_e32 v7, v2
+; GFX10-NEXT:    v_mov_b32_e32 v6, v1
+; GFX10-NEXT:    v_mov_b32_e32 v5, v0
+; GFX10-NEXT:    v_mov_b32_e32 v9, v8
+; GFX10-NEXT:    v_mov_b32_e32 v10, v8
+; GFX10-NEXT:    v_mov_b32_e32 v11, v8
+; GFX10-NEXT:    v_mov_b32_e32 v12, v8
+; GFX10-NEXT:    v_mov_b32_e32 v0, v8
+; GFX10-NEXT:    v_mov_b32_e32 v1, v9
+; GFX10-NEXT:    v_mov_b32_e32 v2, v10
+; GFX10-NEXT:    v_mov_b32_e32 v3, v11
+; GFX10-NEXT:    v_mov_b32_e32 v4, v12
+; GFX10-NEXT:    image_load_mip v[0:4], v[5:7], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D unorm tfe
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    global_store_dword v8, v4, s[8:9]
+; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: load_mip_2d_tfe:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    v_mov_b32_e32 v8, 0
+; GFX11-NEXT:    v_mov_b32_e32 v7, v2
+; GFX11-NEXT:    v_mov_b32_e32 v6, v1
+; GFX11-NEXT:    v_mov_b32_e32 v5, v0
+; GFX11-NEXT:    v_mov_b32_e32 v9, v8
+; GFX11-NEXT:    v_mov_b32_e32 v10, v8
+; GFX11-NEXT:    v_mov_b32_e32 v11, v8
+; GFX11-NEXT:    v_mov_b32_e32 v12, v8
+; GFX11-NEXT:    v_mov_b32_e32 v0, v8
+; GFX11-NEXT:    v_mov_b32_e32 v1, v9
+; GFX11-NEXT:    v_mov_b32_e32 v2, v10
+; GFX11-NEXT:    v_mov_b32_e32 v3, v11
+; GFX11-NEXT:    v_mov_b32_e32 v4, v12
+; GFX11-NEXT:    image_load_mip v[0:4], v[5:7], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D unorm tfe
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    global_store_b32 v8, v4, s[8:9]
+; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %v = call {<4 x float>,i32} @llvm.amdgcn.image.load.mip.2d.v4f32i32.i32(i32 15, i32 %s, i32 %t, i32 %mip, <8 x i32> %rsrc, i32 1, i32 0)
   %v.vec = extractvalue {<4 x float>, i32} %v, 0
   %v.err = extractvalue {<4 x float>, i32} %v, 1
   store i32 %v.err, i32 addrspace(1)* %out, align 4
   ret <4 x float> %v.vec
 }
 

 ; NOPRT-LABEL: load_1d_V2_tfe_dmask0:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    v_mov_b32_e32 v1, 0
 ; NOPRT-NEXT:    image_load v[0:1], v0, s[0:7] dmask:0x1 unorm tfe
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    v_mov_b32_e32 v0, v1
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: load_1d_V2_tfe_dmask0:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    v_mov_b32_e32 v1, 0 ; encoding: [0x80,0x02,0x02,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v2, v1 ; encoding: [0x01,0x03,0x04,0x7e]
-; GFX10-NEXT:    image_load v[1:2], v0, s[0:7] dmask:0x1 dim:SQ_RSRC_IMG_1D unorm tfe ; encoding: [0x00,0x11,0x01,0xf0,0x00,0x01,0x00,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    v_mov_b32_e32 v0, v2 ; encoding: [0x02,0x03,0x00,0x7e]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: load_1d_V2_tfe_dmask0:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v1, 0
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v2, v1
+; GFX10PLUS-NEXT:    image_load v[1:2], v0, s[0:7] dmask:0x1 dim:SQ_RSRC_IMG_1D unorm tfe
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v0, v2
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call {<2 x float>,i32} @llvm.amdgcn.image.load.1d.v2f32i32.i32(i32 0, i32 %s, <8 x i32> %rsrc, i32 1, i32 0)
   %v.err = extractvalue {<2 x float>, i32} %v, 1
   %vv = bitcast i32 %v.err to float
   ret float %vv
 }
 
 define amdgpu_ps float @load_1d_V1_tfe_dmask0(<8 x i32> inreg %rsrc, i32 %s) {

 ; NOPRT-LABEL: load_1d_V1_tfe_dmask0:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    v_mov_b32_e32 v1, 0
 ; NOPRT-NEXT:    image_load v[0:1], v0, s[0:7] dmask:0x1 unorm tfe
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    v_mov_b32_e32 v0, v1
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: load_1d_V1_tfe_dmask0:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    v_mov_b32_e32 v1, 0 ; encoding: [0x80,0x02,0x02,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v2, v1 ; encoding: [0x01,0x03,0x04,0x7e]
-; GFX10-NEXT:    image_load v[1:2], v0, s[0:7] dmask:0x1 dim:SQ_RSRC_IMG_1D unorm tfe ; encoding: [0x00,0x11,0x01,0xf0,0x00,0x01,0x00,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    v_mov_b32_e32 v0, v2 ; encoding: [0x02,0x03,0x00,0x7e]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: load_1d_V1_tfe_dmask0:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v1, 0
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v2, v1
+; GFX10PLUS-NEXT:    image_load v[1:2], v0, s[0:7] dmask:0x1 dim:SQ_RSRC_IMG_1D unorm tfe
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v0, v2
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call {float,i32} @llvm.amdgcn.image.load.1d.f32i32.i32(i32 0, i32 %s, <8 x i32> %rsrc, i32 1, i32 0)
   %v.err = extractvalue {float, i32} %v, 1
   %vv = bitcast i32 %v.err to float
   ret float %vv
 }
 
 define amdgpu_ps float @load_mip_2d_tfe_dmask0(<8 x i32> inreg %rsrc, i32 %s, i32 %t, i32 %mip) {

 ; NOPRT-LABEL: load_mip_2d_tfe_dmask0:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    v_mov_b32_e32 v3, 0
 ; NOPRT-NEXT:    image_load_mip v[2:3], v[0:2], s[0:7] dmask:0x1 unorm tfe
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    v_mov_b32_e32 v0, v3
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: load_mip_2d_tfe_dmask0:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    v_mov_b32_e32 v3, 0 ; encoding: [0x80,0x02,0x06,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v4, v3 ; encoding: [0x03,0x03,0x08,0x7e]
-; GFX10-NEXT:    image_load_mip v[3:4], v[0:2], s[0:7] dmask:0x1 dim:SQ_RSRC_IMG_2D unorm tfe ; encoding: [0x08,0x11,0x05,0xf0,0x00,0x03,0x00,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    v_mov_b32_e32 v0, v4 ; encoding: [0x04,0x03,0x00,0x7e]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: load_mip_2d_tfe_dmask0:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v3, 0
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v4, v3
+; GFX10PLUS-NEXT:    image_load_mip v[3:4], v[0:2], s[0:7] dmask:0x1 dim:SQ_RSRC_IMG_2D unorm tfe
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v0, v4
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call {<4 x float>,i32} @llvm.amdgcn.image.load.mip.2d.v4f32i32.i32(i32 0, i32 %s, i32 %t, i32 %mip, <8 x i32> %rsrc, i32 1, i32 0)
   %v.err = extractvalue {<4 x float>, i32} %v, 1
   %vv = bitcast i32 %v.err to float
   ret float %vv
 }
 
 define amdgpu_ps float @load_mip_2d_tfe_nouse(<8 x i32> inreg %rsrc, i32 %s, i32 %t, i32 %mip) {

 ; NOPRT-LABEL: load_mip_2d_tfe_nouse:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    v_mov_b32_e32 v3, 0
 ; NOPRT-NEXT:    image_load_mip v[2:3], v[0:2], s[0:7] dmask:0x1 unorm tfe
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    v_mov_b32_e32 v0, v3
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: load_mip_2d_tfe_nouse:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    v_mov_b32_e32 v3, 0 ; encoding: [0x80,0x02,0x06,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v4, v3 ; encoding: [0x03,0x03,0x08,0x7e]
-; GFX10-NEXT:    image_load_mip v[3:4], v[0:2], s[0:7] dmask:0x1 dim:SQ_RSRC_IMG_2D unorm tfe ; encoding: [0x08,0x11,0x05,0xf0,0x00,0x03,0x00,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    v_mov_b32_e32 v0, v4 ; encoding: [0x04,0x03,0x00,0x7e]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: load_mip_2d_tfe_nouse:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v3, 0
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v4, v3
+; GFX10PLUS-NEXT:    image_load_mip v[3:4], v[0:2], s[0:7] dmask:0x1 dim:SQ_RSRC_IMG_2D unorm tfe
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v0, v4
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call {<4 x float>,i32} @llvm.amdgcn.image.load.mip.2d.v4f32i32.i32(i32 15, i32 %s, i32 %t, i32 %mip, <8 x i32> %rsrc, i32 1, i32 0)
   %v.err = extractvalue {<4 x float>, i32} %v, 1
   %vv = bitcast i32 %v.err to float
   ret float %vv
 }
 
 define amdgpu_ps float @load_mip_2d_tfe_nouse_V2(<8 x i32> inreg %rsrc, i32 %s, i32 %t, i32 %mip) {

 ; NOPRT-LABEL: load_mip_2d_tfe_nouse_V2:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    v_mov_b32_e32 v3, 0
 ; NOPRT-NEXT:    image_load_mip v[2:3], v[0:2], s[0:7] dmask:0x1 unorm tfe
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    v_mov_b32_e32 v0, v3
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: load_mip_2d_tfe_nouse_V2:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    v_mov_b32_e32 v3, 0 ; encoding: [0x80,0x02,0x06,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v4, v3 ; encoding: [0x03,0x03,0x08,0x7e]
-; GFX10-NEXT:    image_load_mip v[3:4], v[0:2], s[0:7] dmask:0x1 dim:SQ_RSRC_IMG_2D unorm tfe ; encoding: [0x08,0x11,0x05,0xf0,0x00,0x03,0x00,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    v_mov_b32_e32 v0, v4 ; encoding: [0x04,0x03,0x00,0x7e]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: load_mip_2d_tfe_nouse_V2:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v3, 0
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v4, v3
+; GFX10PLUS-NEXT:    image_load_mip v[3:4], v[0:2], s[0:7] dmask:0x1 dim:SQ_RSRC_IMG_2D unorm tfe
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v0, v4
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call {<2 x float>,i32} @llvm.amdgcn.image.load.mip.2d.v2f32i32.i32(i32 6, i32 %s, i32 %t, i32 %mip, <8 x i32> %rsrc, i32 1, i32 0)
   %v.err = extractvalue {<2 x float>, i32} %v, 1
   %vv = bitcast i32 %v.err to float
   ret float %vv
 }
 
 define amdgpu_ps float @load_mip_2d_tfe_nouse_V1(<8 x i32> inreg %rsrc, i32 %s, i32 %t, i32 %mip) {

 ; NOPRT-LABEL: load_mip_2d_tfe_nouse_V1:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    v_mov_b32_e32 v3, 0
 ; NOPRT-NEXT:    image_load_mip v[2:3], v[0:2], s[0:7] dmask:0x2 unorm tfe
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    v_mov_b32_e32 v0, v3
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: load_mip_2d_tfe_nouse_V1:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    v_mov_b32_e32 v3, 0 ; encoding: [0x80,0x02,0x06,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v4, v3 ; encoding: [0x03,0x03,0x08,0x7e]
-; GFX10-NEXT:    image_load_mip v[3:4], v[0:2], s[0:7] dmask:0x2 dim:SQ_RSRC_IMG_2D unorm tfe ; encoding: [0x08,0x12,0x05,0xf0,0x00,0x03,0x00,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    v_mov_b32_e32 v0, v4 ; encoding: [0x04,0x03,0x00,0x7e]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: load_mip_2d_tfe_nouse_V1:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v3, 0
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v4, v3
+; GFX10PLUS-NEXT:    image_load_mip v[3:4], v[0:2], s[0:7] dmask:0x2 dim:SQ_RSRC_IMG_2D unorm tfe
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v0, v4
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call {float, i32} @llvm.amdgcn.image.load.mip.2d.f32i32.i32(i32 2, i32 %s, i32 %t, i32 %mip, <8 x i32> %rsrc, i32 1, i32 0)
   %v.err = extractvalue {float, i32} %v, 1
   %vv = bitcast i32 %v.err to float
   ret float %vv
 }
 
 define amdgpu_ps <4 x float> @load_1d_tfe_V4_dmask3(<8 x i32> inreg %rsrc, i32 addrspace(1)* inreg %out, i32 %s) {

 ; NOPRT-NEXT:    v_mov_b32_e32 v4, 0
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    global_store_dword v4, v3, s[8:9]
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
 ; GFX10-LABEL: load_1d_tfe_V4_dmask3:
 ; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    v_mov_b32_e32 v5, 0 ; encoding: [0x80,0x02,0x0a,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v4, v0 ; encoding: [0x00,0x03,0x08,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v6, v5 ; encoding: [0x05,0x03,0x0c,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v7, v5 ; encoding: [0x05,0x03,0x0e,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v8, v5 ; encoding: [0x05,0x03,0x10,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v0, v5 ; encoding: [0x05,0x03,0x00,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v1, v6 ; encoding: [0x06,0x03,0x02,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v2, v7 ; encoding: [0x07,0x03,0x04,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v3, v8 ; encoding: [0x08,0x03,0x06,0x7e]
-; GFX10-NEXT:    image_load v[0:3], v4, s[0:7] dmask:0x7 dim:SQ_RSRC_IMG_1D unorm tfe ; encoding: [0x00,0x17,0x01,0xf0,0x04,0x00,0x00,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    global_store_dword v5, v3, s[8:9] ; encoding: [0x00,0x80,0x70,0xdc,0x05,0x03,0x08,0x00]
-; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0 ; encoding: [0x00,0x00,0xfd,0xbb]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10-NEXT:    v_mov_b32_e32 v5, 0
+; GFX10-NEXT:    v_mov_b32_e32 v4, v0
+; GFX10-NEXT:    v_mov_b32_e32 v6, v5
+; GFX10-NEXT:    v_mov_b32_e32 v7, v5
+; GFX10-NEXT:    v_mov_b32_e32 v8, v5
+; GFX10-NEXT:    v_mov_b32_e32 v0, v5
+; GFX10-NEXT:    v_mov_b32_e32 v1, v6
+; GFX10-NEXT:    v_mov_b32_e32 v2, v7
+; GFX10-NEXT:    v_mov_b32_e32 v3, v8
+; GFX10-NEXT:    image_load v[0:3], v4, s[0:7] dmask:0x7 dim:SQ_RSRC_IMG_1D unorm tfe
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    global_store_dword v5, v3, s[8:9]
+; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: load_1d_tfe_V4_dmask3:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    v_mov_b32_e32 v5, 0
+; GFX11-NEXT:    v_mov_b32_e32 v4, v0
+; GFX11-NEXT:    v_mov_b32_e32 v6, v5
+; GFX11-NEXT:    v_mov_b32_e32 v7, v5
+; GFX11-NEXT:    v_mov_b32_e32 v8, v5
+; GFX11-NEXT:    v_mov_b32_e32 v0, v5
+; GFX11-NEXT:    v_mov_b32_e32 v1, v6
+; GFX11-NEXT:    v_mov_b32_e32 v2, v7
+; GFX11-NEXT:    v_mov_b32_e32 v3, v8
+; GFX11-NEXT:    image_load v[0:3], v4, s[0:7] dmask:0x7 dim:SQ_RSRC_IMG_1D unorm tfe
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    global_store_b32 v5, v3, s[8:9]
+; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %v = call {<4 x float>,i32} @llvm.amdgcn.image.load.1d.v4f32i32.i32(i32 7, i32 %s, <8 x i32> %rsrc, i32 1, i32 0)
   %v.vec = extractvalue {<4 x float>, i32} %v, 0
   %v.err = extractvalue {<4 x float>, i32} %v, 1
   store i32 %v.err, i32 addrspace(1)* %out, align 4
   ret <4 x float> %v.vec
 }
 

 ; NOPRT-NEXT:    v_mov_b32_e32 v3, 0
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    global_store_dword v3, v2, s[8:9]
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
 ; GFX10-LABEL: load_1d_tfe_V4_dmask2:
 ; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    v_mov_b32_e32 v4, 0 ; encoding: [0x80,0x02,0x08,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v3, v0 ; encoding: [0x00,0x03,0x06,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v5, v4 ; encoding: [0x04,0x03,0x0a,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v6, v4 ; encoding: [0x04,0x03,0x0c,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v0, v4 ; encoding: [0x04,0x03,0x00,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v1, v5 ; encoding: [0x05,0x03,0x02,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v2, v6 ; encoding: [0x06,0x03,0x04,0x7e]
-; GFX10-NEXT:    image_load v[0:2], v3, s[0:7] dmask:0x6 dim:SQ_RSRC_IMG_1D unorm tfe ; encoding: [0x00,0x16,0x01,0xf0,0x03,0x00,0x00,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    global_store_dword v4, v2, s[8:9] ; encoding: [0x00,0x80,0x70,0xdc,0x04,0x02,0x08,0x00]
-; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0 ; encoding: [0x00,0x00,0xfd,0xbb]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10-NEXT:    v_mov_b32_e32 v4, 0
+; GFX10-NEXT:    v_mov_b32_e32 v3, v0
+; GFX10-NEXT:    v_mov_b32_e32 v5, v4
+; GFX10-NEXT:    v_mov_b32_e32 v6, v4
+; GFX10-NEXT:    v_mov_b32_e32 v0, v4
+; GFX10-NEXT:    v_mov_b32_e32 v1, v5
+; GFX10-NEXT:    v_mov_b32_e32 v2, v6
+; GFX10-NEXT:    image_load v[0:2], v3, s[0:7] dmask:0x6 dim:SQ_RSRC_IMG_1D unorm tfe
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    global_store_dword v4, v2, s[8:9]
+; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: load_1d_tfe_V4_dmask2:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    v_mov_b32_e32 v4, 0
+; GFX11-NEXT:    v_mov_b32_e32 v3, v0
+; GFX11-NEXT:    v_mov_b32_e32 v5, v4
+; GFX11-NEXT:    v_mov_b32_e32 v6, v4
+; GFX11-NEXT:    v_mov_b32_e32 v0, v4
+; GFX11-NEXT:    v_mov_b32_e32 v1, v5
+; GFX11-NEXT:    v_mov_b32_e32 v2, v6
+; GFX11-NEXT:    image_load v[0:2], v3, s[0:7] dmask:0x6 dim:SQ_RSRC_IMG_1D unorm tfe
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    global_store_b32 v4, v2, s[8:9]
+; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %v = call {<4 x float>,i32} @llvm.amdgcn.image.load.1d.v4f32i32.i32(i32 6, i32 %s, <8 x i32> %rsrc, i32 1, i32 0)
   %v.vec = extractvalue {<4 x float>, i32} %v, 0
   %v.err = extractvalue {<4 x float>, i32} %v, 1
   store i32 %v.err, i32 addrspace(1)* %out, align 4
   ret <4 x float> %v.vec
 }
 

 ; NOPRT-NEXT:    v_mov_b32_e32 v2, 0
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    global_store_dword v2, v1, s[8:9]
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
 ; GFX10-LABEL: load_1d_tfe_V4_dmask1:
 ; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    v_mov_b32_e32 v3, 0 ; encoding: [0x80,0x02,0x06,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v2, v0 ; encoding: [0x00,0x03,0x04,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v4, v3 ; encoding: [0x03,0x03,0x08,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v0, v3 ; encoding: [0x03,0x03,0x00,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v1, v4 ; encoding: [0x04,0x03,0x02,0x7e]
-; GFX10-NEXT:    image_load v[0:1], v2, s[0:7] dmask:0x8 dim:SQ_RSRC_IMG_1D unorm tfe ; encoding: [0x00,0x18,0x01,0xf0,0x02,0x00,0x00,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    global_store_dword v3, v1, s[8:9] ; encoding: [0x00,0x80,0x70,0xdc,0x03,0x01,0x08,0x00]
-; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0 ; encoding: [0x00,0x00,0xfd,0xbb]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10-NEXT:    v_mov_b32_e32 v3, 0
+; GFX10-NEXT:    v_mov_b32_e32 v2, v0
+; GFX10-NEXT:    v_mov_b32_e32 v4, v3
+; GFX10-NEXT:    v_mov_b32_e32 v0, v3
+; GFX10-NEXT:    v_mov_b32_e32 v1, v4
+; GFX10-NEXT:    image_load v[0:1], v2, s[0:7] dmask:0x8 dim:SQ_RSRC_IMG_1D unorm tfe
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    global_store_dword v3, v1, s[8:9]
+; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: load_1d_tfe_V4_dmask1:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    v_mov_b32_e32 v3, 0
+; GFX11-NEXT:    v_mov_b32_e32 v2, v0
+; GFX11-NEXT:    v_mov_b32_e32 v4, v3
+; GFX11-NEXT:    v_mov_b32_e32 v0, v3
+; GFX11-NEXT:    v_mov_b32_e32 v1, v4
+; GFX11-NEXT:    image_load v[0:1], v2, s[0:7] dmask:0x8 dim:SQ_RSRC_IMG_1D unorm tfe
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    global_store_b32 v3, v1, s[8:9]
+; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %v = call {<4 x float>,i32} @llvm.amdgcn.image.load.1d.v4f32i32.i32(i32 8, i32 %s, <8 x i32> %rsrc, i32 1, i32 0)
   %v.vec = extractvalue {<4 x float>, i32} %v, 0
   %v.err = extractvalue {<4 x float>, i32} %v, 1
   store i32 %v.err, i32 addrspace(1)* %out, align 4
   ret <4 x float> %v.vec
 }
 

 ; NOPRT-NEXT:    v_mov_b32_e32 v2, 0
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    global_store_dword v2, v1, s[8:9]
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
 ; GFX10-LABEL: load_1d_tfe_V2_dmask1:
 ; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    v_mov_b32_e32 v3, 0 ; encoding: [0x80,0x02,0x06,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v2, v0 ; encoding: [0x00,0x03,0x04,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v4, v3 ; encoding: [0x03,0x03,0x08,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v0, v3 ; encoding: [0x03,0x03,0x00,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v1, v4 ; encoding: [0x04,0x03,0x02,0x7e]
-; GFX10-NEXT:    image_load v[0:1], v2, s[0:7] dmask:0x8 dim:SQ_RSRC_IMG_1D unorm tfe ; encoding: [0x00,0x18,0x01,0xf0,0x02,0x00,0x00,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    global_store_dword v3, v1, s[8:9] ; encoding: [0x00,0x80,0x70,0xdc,0x03,0x01,0x08,0x00]
-; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0 ; encoding: [0x00,0x00,0xfd,0xbb]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10-NEXT:    v_mov_b32_e32 v3, 0
+; GFX10-NEXT:    v_mov_b32_e32 v2, v0
+; GFX10-NEXT:    v_mov_b32_e32 v4, v3
+; GFX10-NEXT:    v_mov_b32_e32 v0, v3
+; GFX10-NEXT:    v_mov_b32_e32 v1, v4
+; GFX10-NEXT:    image_load v[0:1], v2, s[0:7] dmask:0x8 dim:SQ_RSRC_IMG_1D unorm tfe
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    global_store_dword v3, v1, s[8:9]
+; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: load_1d_tfe_V2_dmask1:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    v_mov_b32_e32 v3, 0
+; GFX11-NEXT:    v_mov_b32_e32 v2, v0
+; GFX11-NEXT:    v_mov_b32_e32 v4, v3
+; GFX11-NEXT:    v_mov_b32_e32 v0, v3
+; GFX11-NEXT:    v_mov_b32_e32 v1, v4
+; GFX11-NEXT:    image_load v[0:1], v2, s[0:7] dmask:0x8 dim:SQ_RSRC_IMG_1D unorm tfe
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    global_store_b32 v3, v1, s[8:9]
+; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %v = call {<2 x float>,i32} @llvm.amdgcn.image.load.1d.v2f32i32.i32(i32 8, i32 %s, <8 x i32> %rsrc, i32 1, i32 0)
   %v.vec = extractvalue {<2 x float>, i32} %v, 0
   %v.err = extractvalue {<2 x float>, i32} %v, 1
   store i32 %v.err, i32 addrspace(1)* %out, align 4
   ret <2 x float> %v.vec
 }
 

 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
 ; NOPRT-LABEL: load_mip_3d:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    image_load_mip v[0:3], v[0:3], s[0:7] dmask:0xf unorm
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: load_mip_3d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_load_mip v[0:3], v[0:3], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_3D unorm ; encoding: [0x10,0x1f,0x04,0xf0,0x00,0x00,0x00,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: load_mip_3d:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_load_mip v[0:3], v[0:3], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_3D unorm
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.load.mip.3d.v4f32.i32(i32 15, i32 %s, i32 %t, i32 %r, i32 %mip, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @load_mip_cube(<8 x i32> inreg %rsrc, i32 %s, i32 %t, i32 %slice, i32 %mip) {
 ; VERDE-LABEL: load_mip_cube:
 ; VERDE:       ; %bb.0: ; %main_body

 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
 ; NOPRT-LABEL: load_mip_cube:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    image_load_mip v[0:3], v[0:3], s[0:7] dmask:0xf unorm da
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: load_mip_cube:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_load_mip v[0:3], v[0:3], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_CUBE unorm ; encoding: [0x18,0x1f,0x04,0xf0,0x00,0x00,0x00,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: load_mip_cube:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_load_mip v[0:3], v[0:3], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_CUBE unorm
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.load.mip.cube.v4f32.i32(i32 15, i32 %s, i32 %t, i32 %slice, i32 %mip, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @load_mip_1darray(<8 x i32> inreg %rsrc, i32 %s, i32 %slice, i32 %mip) {
 ; VERDE-LABEL: load_mip_1darray:
 ; VERDE:       ; %bb.0: ; %main_body

 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
 ; NOPRT-LABEL: load_mip_1darray:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    image_load_mip v[0:3], v[0:2], s[0:7] dmask:0xf unorm da
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: load_mip_1darray:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_load_mip v[0:3], v[0:2], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D_ARRAY unorm ; encoding: [0x20,0x1f,0x04,0xf0,0x00,0x00,0x00,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: load_mip_1darray:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_load_mip v[0:3], v[0:2], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D_ARRAY unorm
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.load.mip.1darray.v4f32.i32(i32 15, i32 %s, i32 %slice, i32 %mip, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @load_mip_2darray(<8 x i32> inreg %rsrc, i32 %s, i32 %t, i32 %slice, i32 %mip) {
 ; VERDE-LABEL: load_mip_2darray:
 ; VERDE:       ; %bb.0: ; %main_body

 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
 ; NOPRT-LABEL: load_mip_2darray:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    image_load_mip v[0:3], v[0:3], s[0:7] dmask:0xf unorm da
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: load_mip_2darray:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_load_mip v[0:3], v[0:3], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_ARRAY unorm ; encoding: [0x28,0x1f,0x04,0xf0,0x00,0x00,0x00,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: load_mip_2darray:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_load_mip v[0:3], v[0:3], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_ARRAY unorm
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.load.mip.2darray.v4f32.i32(i32 15, i32 %s, i32 %t, i32 %slice, i32 %mip, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps void @store_1d(<8 x i32> inreg %rsrc, <4 x float> %vdata, i32 %s) {
 ; VERDE-LABEL: store_1d:
 ; VERDE:       ; %bb.0: ; %main_body

 ; GFX6789-NEXT:    image_store v[0:3], v4, s[0:7] dmask:0xf unorm
 ; GFX6789-NEXT:    s_endpgm
 ;
 ; NOPRT-LABEL: store_1d:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    image_store v[0:3], v4, s[0:7] dmask:0xf unorm
 ; NOPRT-NEXT:    s_endpgm
 ;
-; GFX10-LABEL: store_1d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_store v[0:3], v4, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm ; encoding: [0x00,0x1f,0x20,0xf0,0x04,0x00,0x00,0x00]
-; GFX10-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
+; GFX10PLUS-LABEL: store_1d:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_store v[0:3], v4, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm
+; GFX10PLUS-NEXT:    s_endpgm
 main_body:
   call void @llvm.amdgcn.image.store.1d.v4f32.i32(<4 x float> %vdata, i32 15, i32 %s, <8 x i32> %rsrc, i32 0, i32 0)
   ret void
 }
 
 define amdgpu_ps void @store_2d(<8 x i32> inreg %rsrc, <4 x float> %vdata, i32 %s, i32 %t) {
 ; VERDE-LABEL: store_2d:
 ; VERDE:       ; %bb.0: ; %main_body

 ; GFX6789-NEXT:    image_store v[0:3], v[4:5], s[0:7] dmask:0xf unorm
 ; GFX6789-NEXT:    s_endpgm
 ;
 ; NOPRT-LABEL: store_2d:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    image_store v[0:3], v[4:5], s[0:7] dmask:0xf unorm
 ; NOPRT-NEXT:    s_endpgm
 ;
-; GFX10-LABEL: store_2d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_store v[0:3], v[4:5], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D unorm ; encoding: [0x08,0x1f,0x20,0xf0,0x04,0x00,0x00,0x00]
-; GFX10-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
+; GFX10PLUS-LABEL: store_2d:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_store v[0:3], v[4:5], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D unorm
+; GFX10PLUS-NEXT:    s_endpgm
 main_body:
   call void @llvm.amdgcn.image.store.2d.v4f32.i32(<4 x float> %vdata, i32 15, i32 %s, i32 %t, <8 x i32> %rsrc, i32 0, i32 0)
   ret void
 }
 
 define amdgpu_ps void @store_3d(<8 x i32> inreg %rsrc, <4 x float> %vdata, i32 %s, i32 %t, i32 %r) {
 ; VERDE-LABEL: store_3d:
 ; VERDE:       ; %bb.0: ; %main_body

 ; GFX6789-NEXT:    image_store v[0:3], v[4:6], s[0:7] dmask:0xf unorm
 ; GFX6789-NEXT:    s_endpgm
 ;
 ; NOPRT-LABEL: store_3d:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    image_store v[0:3], v[4:6], s[0:7] dmask:0xf unorm
 ; NOPRT-NEXT:    s_endpgm
 ;
-; GFX10-LABEL: store_3d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_store v[0:3], v[4:6], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_3D unorm ; encoding: [0x10,0x1f,0x20,0xf0,0x04,0x00,0x00,0x00]
-; GFX10-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
+; GFX10PLUS-LABEL: store_3d:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_store v[0:3], v[4:6], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_3D unorm
+; GFX10PLUS-NEXT:    s_endpgm
 main_body:
   call void @llvm.amdgcn.image.store.3d.v4f32.i32(<4 x float> %vdata, i32 15, i32 %s, i32 %t, i32 %r, <8 x i32> %rsrc, i32 0, i32 0)
   ret void
 }
 
 define amdgpu_ps void @store_cube(<8 x i32> inreg %rsrc, <4 x float> %vdata, i32 %s, i32 %t, i32 %slice) {
 ; VERDE-LABEL: store_cube:
 ; VERDE:       ; %bb.0: ; %main_body

 ; GFX6789-NEXT:    image_store v[0:3], v[4:6], s[0:7] dmask:0xf unorm da
 ; GFX6789-NEXT:    s_endpgm
 ;
 ; NOPRT-LABEL: store_cube:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    image_store v[0:3], v[4:6], s[0:7] dmask:0xf unorm da
 ; NOPRT-NEXT:    s_endpgm
 ;
-; GFX10-LABEL: store_cube:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_store v[0:3], v[4:6], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_CUBE unorm ; encoding: [0x18,0x1f,0x20,0xf0,0x04,0x00,0x00,0x00]
-; GFX10-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
+; GFX10PLUS-LABEL: store_cube:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_store v[0:3], v[4:6], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_CUBE unorm
+; GFX10PLUS-NEXT:    s_endpgm
 main_body:
   call void @llvm.amdgcn.image.store.cube.v4f32.i32(<4 x float> %vdata, i32 15, i32 %s, i32 %t, i32 %slice, <8 x i32> %rsrc, i32 0, i32 0)
   ret void
 }
 
 define amdgpu_ps void @store_1darray(<8 x i32> inreg %rsrc, <4 x float> %vdata, i32 %s, i32 %slice) {
 ; VERDE-LABEL: store_1darray:
 ; VERDE:       ; %bb.0: ; %main_body

 ; GFX6789-NEXT:    image_store v[0:3], v[4:5], s[0:7] dmask:0xf unorm da
 ; GFX6789-NEXT:    s_endpgm
 ;
 ; NOPRT-LABEL: store_1darray:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    image_store v[0:3], v[4:5], s[0:7] dmask:0xf unorm da
 ; NOPRT-NEXT:    s_endpgm
 ;
-; GFX10-LABEL: store_1darray:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_store v[0:3], v[4:5], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D_ARRAY unorm ; encoding: [0x20,0x1f,0x20,0xf0,0x04,0x00,0x00,0x00]
-; GFX10-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
+; GFX10PLUS-LABEL: store_1darray:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_store v[0:3], v[4:5], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D_ARRAY unorm
+; GFX10PLUS-NEXT:    s_endpgm
 main_body:
   call void @llvm.amdgcn.image.store.1darray.v4f32.i32(<4 x float> %vdata, i32 15, i32 %s, i32 %slice, <8 x i32> %rsrc, i32 0, i32 0)
   ret void
 }
 
 define amdgpu_ps void @store_2darray(<8 x i32> inreg %rsrc, <4 x float> %vdata, i32 %s, i32 %t, i32 %slice) {
 ; VERDE-LABEL: store_2darray:
 ; VERDE:       ; %bb.0: ; %main_body

 ; GFX6789-NEXT:    image_store v[0:3], v[4:6], s[0:7] dmask:0xf unorm da
 ; GFX6789-NEXT:    s_endpgm
 ;
 ; NOPRT-LABEL: store_2darray:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    image_store v[0:3], v[4:6], s[0:7] dmask:0xf unorm da
 ; NOPRT-NEXT:    s_endpgm
 ;
-; GFX10-LABEL: store_2darray:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_store v[0:3], v[4:6], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_ARRAY unorm ; encoding: [0x28,0x1f,0x20,0xf0,0x04,0x00,0x00,0x00]
-; GFX10-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
+; GFX10PLUS-LABEL: store_2darray:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_store v[0:3], v[4:6], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_ARRAY unorm
+; GFX10PLUS-NEXT:    s_endpgm
 main_body:
   call void @llvm.amdgcn.image.store.2darray.v4f32.i32(<4 x float> %vdata, i32 15, i32 %s, i32 %t, i32 %slice, <8 x i32> %rsrc, i32 0, i32 0)
   ret void
 }
 
 define amdgpu_ps void @store_2dmsaa(<8 x i32> inreg %rsrc, <4 x float> %vdata, i32 %s, i32 %t, i32 %fragid) {
 ; VERDE-LABEL: store_2dmsaa:
 ; VERDE:       ; %bb.0: ; %main_body

 ; GFX6789-NEXT:    image_store v[0:3], v[4:6], s[0:7] dmask:0xf unorm
 ; GFX6789-NEXT:    s_endpgm
 ;
 ; NOPRT-LABEL: store_2dmsaa:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    image_store v[0:3], v[4:6], s[0:7] dmask:0xf unorm
 ; NOPRT-NEXT:    s_endpgm
 ;
-; GFX10-LABEL: store_2dmsaa:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_store v[0:3], v[4:6], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_MSAA unorm ; encoding: [0x30,0x1f,0x20,0xf0,0x04,0x00,0x00,0x00]
-; GFX10-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
+; GFX10PLUS-LABEL: store_2dmsaa:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_store v[0:3], v[4:6], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_MSAA unorm
+; GFX10PLUS-NEXT:    s_endpgm
 main_body:
   call void @llvm.amdgcn.image.store.2dmsaa.v4f32.i32(<4 x float> %vdata, i32 15, i32 %s, i32 %t, i32 %fragid, <8 x i32> %rsrc, i32 0, i32 0)
   ret void
 }
 
 define amdgpu_ps void @store_2darraymsaa(<8 x i32> inreg %rsrc, <4 x float> %vdata, i32 %s, i32 %t, i32 %slice, i32 %fragid) {
 ; VERDE-LABEL: store_2darraymsaa:
 ; VERDE:       ; %bb.0: ; %main_body

 ; GFX6789-NEXT:    image_store v[0:3], v[4:7], s[0:7] dmask:0xf unorm da
 ; GFX6789-NEXT:    s_endpgm
 ;
 ; NOPRT-LABEL: store_2darraymsaa:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    image_store v[0:3], v[4:7], s[0:7] dmask:0xf unorm da
 ; NOPRT-NEXT:    s_endpgm
 ;
-; GFX10-LABEL: store_2darraymsaa:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_store v[0:3], v[4:7], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_MSAA_ARRAY unorm ; encoding: [0x38,0x1f,0x20,0xf0,0x04,0x00,0x00,0x00]
-; GFX10-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
+; GFX10PLUS-LABEL: store_2darraymsaa:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_store v[0:3], v[4:7], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_MSAA_ARRAY unorm
+; GFX10PLUS-NEXT:    s_endpgm
 main_body:
   call void @llvm.amdgcn.image.store.2darraymsaa.v4f32.i32(<4 x float> %vdata, i32 15, i32 %s, i32 %t, i32 %slice, i32 %fragid, <8 x i32> %rsrc, i32 0, i32 0)
   ret void
 }
 
 define amdgpu_ps void @store_mip_1d(<8 x i32> inreg %rsrc, <4 x float> %vdata, i32 %s, i32 %mip) {
 ; VERDE-LABEL: store_mip_1d:
 ; VERDE:       ; %bb.0: ; %main_body

 ; GFX6789-NEXT:    image_store_mip v[0:3], v[4:5], s[0:7] dmask:0xf unorm
 ; GFX6789-NEXT:    s_endpgm
 ;
 ; NOPRT-LABEL: store_mip_1d:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    image_store_mip v[0:3], v[4:5], s[0:7] dmask:0xf unorm
 ; NOPRT-NEXT:    s_endpgm
 ;
-; GFX10-LABEL: store_mip_1d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_store_mip v[0:3], v[4:5], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm ; encoding: [0x00,0x1f,0x24,0xf0,0x04,0x00,0x00,0x00]
-; GFX10-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
+; GFX10PLUS-LABEL: store_mip_1d:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_store_mip v[0:3], v[4:5], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm
+; GFX10PLUS-NEXT:    s_endpgm
 main_body:
   call void @llvm.amdgcn.image.store.mip.1d.v4f32.i32(<4 x float> %vdata, i32 15, i32 %s, i32 %mip, <8 x i32> %rsrc, i32 0, i32 0)
   ret void
 }
 
 define amdgpu_ps void @store_mip_2d(<8 x i32> inreg %rsrc, <4 x float> %vdata, i32 %s, i32 %t, i32 %mip) {
 ; VERDE-LABEL: store_mip_2d:
 ; VERDE:       ; %bb.0: ; %main_body

 ; GFX6789-NEXT:    image_store_mip v[0:3], v[4:6], s[0:7] dmask:0xf unorm
 ; GFX6789-NEXT:    s_endpgm
 ;
 ; NOPRT-LABEL: store_mip_2d:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    image_store_mip v[0:3], v[4:6], s[0:7] dmask:0xf unorm
 ; NOPRT-NEXT:    s_endpgm
 ;
-; GFX10-LABEL: store_mip_2d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_store_mip v[0:3], v[4:6], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D unorm ; encoding: [0x08,0x1f,0x24,0xf0,0x04,0x00,0x00,0x00]
-; GFX10-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
+; GFX10PLUS-LABEL: store_mip_2d:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_store_mip v[0:3], v[4:6], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D unorm
+; GFX10PLUS-NEXT:    s_endpgm
 main_body:
   call void @llvm.amdgcn.image.store.mip.2d.v4f32.i32(<4 x float> %vdata, i32 15, i32 %s, i32 %t, i32 %mip, <8 x i32> %rsrc, i32 0, i32 0)
   ret void
 }
 
 define amdgpu_ps void @store_mip_3d(<8 x i32> inreg %rsrc, <4 x float> %vdata, i32 %s, i32 %t, i32 %r, i32 %mip) {
 ; VERDE-LABEL: store_mip_3d:
 ; VERDE:       ; %bb.0: ; %main_body

 ; GFX6789-NEXT:    image_store_mip v[0:3], v[4:7], s[0:7] dmask:0xf unorm
 ; GFX6789-NEXT:    s_endpgm
 ;
 ; NOPRT-LABEL: store_mip_3d:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    image_store_mip v[0:3], v[4:7], s[0:7] dmask:0xf unorm
 ; NOPRT-NEXT:    s_endpgm
 ;
-; GFX10-LABEL: store_mip_3d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_store_mip v[0:3], v[4:7], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_3D unorm ; encoding: [0x10,0x1f,0x24,0xf0,0x04,0x00,0x00,0x00]
-; GFX10-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
+; GFX10PLUS-LABEL: store_mip_3d:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_store_mip v[0:3], v[4:7], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_3D unorm
+; GFX10PLUS-NEXT:    s_endpgm
 main_body:
   call void @llvm.amdgcn.image.store.mip.3d.v4f32.i32(<4 x float> %vdata, i32 15, i32 %s, i32 %t, i32 %r, i32 %mip, <8 x i32> %rsrc, i32 0, i32 0)
   ret void
 }
 
 define amdgpu_ps void @store_mip_cube(<8 x i32> inreg %rsrc, <4 x float> %vdata, i32 %s, i32 %t, i32 %slice, i32 %mip) {
 ; VERDE-LABEL: store_mip_cube:
 ; VERDE:       ; %bb.0: ; %main_body

 ; GFX6789-NEXT:    image_store_mip v[0:3], v[4:7], s[0:7] dmask:0xf unorm da
 ; GFX6789-NEXT:    s_endpgm
 ;
 ; NOPRT-LABEL: store_mip_cube:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    image_store_mip v[0:3], v[4:7], s[0:7] dmask:0xf unorm da
 ; NOPRT-NEXT:    s_endpgm
 ;
-; GFX10-LABEL: store_mip_cube:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_store_mip v[0:3], v[4:7], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_CUBE unorm ; encoding: [0x18,0x1f,0x24,0xf0,0x04,0x00,0x00,0x00]
-; GFX10-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
+; GFX10PLUS-LABEL: store_mip_cube:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_store_mip v[0:3], v[4:7], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_CUBE unorm
+; GFX10PLUS-NEXT:    s_endpgm
 main_body:
   call void @llvm.amdgcn.image.store.mip.cube.v4f32.i32(<4 x float> %vdata, i32 15, i32 %s, i32 %t, i32 %slice, i32 %mip, <8 x i32> %rsrc, i32 0, i32 0)
   ret void
 }
 
 define amdgpu_ps void @store_mip_1darray(<8 x i32> inreg %rsrc, <4 x float> %vdata, i32 %s, i32 %slice, i32 %mip) {
 ; VERDE-LABEL: store_mip_1darray:
 ; VERDE:       ; %bb.0: ; %main_body

 ; GFX6789-NEXT:    image_store_mip v[0:3], v[4:6], s[0:7] dmask:0xf unorm da
 ; GFX6789-NEXT:    s_endpgm
 ;
 ; NOPRT-LABEL: store_mip_1darray:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    image_store_mip v[0:3], v[4:6], s[0:7] dmask:0xf unorm da
 ; NOPRT-NEXT:    s_endpgm
 ;
-; GFX10-LABEL: store_mip_1darray:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_store_mip v[0:3], v[4:6], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D_ARRAY unorm ; encoding: [0x20,0x1f,0x24,0xf0,0x04,0x00,0x00,0x00]
-; GFX10-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
+; GFX10PLUS-LABEL: store_mip_1darray:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_store_mip v[0:3], v[4:6], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D_ARRAY unorm
+; GFX10PLUS-NEXT:    s_endpgm
 main_body:
   call void @llvm.amdgcn.image.store.mip.1darray.v4f32.i32(<4 x float> %vdata, i32 15, i32 %s, i32 %slice, i32 %mip, <8 x i32> %rsrc, i32 0, i32 0)
   ret void
 }
 
 define amdgpu_ps void @store_mip_2darray(<8 x i32> inreg %rsrc, <4 x float> %vdata, i32 %s, i32 %t, i32 %slice, i32 %mip) {
 ; VERDE-LABEL: store_mip_2darray:
 ; VERDE:       ; %bb.0: ; %main_body

 ; GFX6789-NEXT:    image_store_mip v[0:3], v[4:7], s[0:7] dmask:0xf unorm da
 ; GFX6789-NEXT:    s_endpgm
 ;
 ; NOPRT-LABEL: store_mip_2darray:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    image_store_mip v[0:3], v[4:7], s[0:7] dmask:0xf unorm da
 ; NOPRT-NEXT:    s_endpgm
 ;
-; GFX10-LABEL: store_mip_2darray:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_store_mip v[0:3], v[4:7], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_ARRAY unorm ; encoding: [0x28,0x1f,0x24,0xf0,0x04,0x00,0x00,0x00]
-; GFX10-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
+; GFX10PLUS-LABEL: store_mip_2darray:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_store_mip v[0:3], v[4:7], s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_ARRAY unorm
+; GFX10PLUS-NEXT:    s_endpgm
 main_body:
   call void @llvm.amdgcn.image.store.mip.2darray.v4f32.i32(<4 x float> %vdata, i32 15, i32 %s, i32 %t, i32 %slice, i32 %mip, <8 x i32> %rsrc, i32 0, i32 0)
   ret void
 }
 
 define amdgpu_ps <4 x float> @getresinfo_1d(<8 x i32> inreg %rsrc, i32 %mip) {
 ; VERDE-LABEL: getresinfo_1d:
 ; VERDE:       ; %bb.0: ; %main_body

 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
 ; NOPRT-LABEL: getresinfo_1d:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    image_get_resinfo v[0:3], v0, s[0:7] dmask:0xf unorm
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: getresinfo_1d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_get_resinfo v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm ; encoding: [0x00,0x1f,0x38,0xf0,0x00,0x00,0x00,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: getresinfo_1d:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_get_resinfo v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.getresinfo.1d.v4f32.i32(i32 15, i32 %mip, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @getresinfo_2d(<8 x i32> inreg %rsrc, i32 %mip) {
 ; VERDE-LABEL: getresinfo_2d:
 ; VERDE:       ; %bb.0: ; %main_body

 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
 ; NOPRT-LABEL: getresinfo_2d:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    image_get_resinfo v[0:3], v0, s[0:7] dmask:0xf unorm
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: getresinfo_2d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_get_resinfo v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D unorm ; encoding: [0x08,0x1f,0x38,0xf0,0x00,0x00,0x00,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: getresinfo_2d:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_get_resinfo v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D unorm
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.getresinfo.2d.v4f32.i32(i32 15, i32 %mip, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @getresinfo_3d(<8 x i32> inreg %rsrc, i32 %mip) {
 ; VERDE-LABEL: getresinfo_3d:
 ; VERDE:       ; %bb.0: ; %main_body

 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
 ; NOPRT-LABEL: getresinfo_3d:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    image_get_resinfo v[0:3], v0, s[0:7] dmask:0xf unorm
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: getresinfo_3d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_get_resinfo v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_3D unorm ; encoding: [0x10,0x1f,0x38,0xf0,0x00,0x00,0x00,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: getresinfo_3d:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_get_resinfo v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_3D unorm
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.getresinfo.3d.v4f32.i32(i32 15, i32 %mip, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @getresinfo_cube(<8 x i32> inreg %rsrc, i32 %mip) {
 ; VERDE-LABEL: getresinfo_cube:
 ; VERDE:       ; %bb.0: ; %main_body

 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
 ; NOPRT-LABEL: getresinfo_cube:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    image_get_resinfo v[0:3], v0, s[0:7] dmask:0xf unorm da
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: getresinfo_cube:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_get_resinfo v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_CUBE unorm ; encoding: [0x18,0x1f,0x38,0xf0,0x00,0x00,0x00,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: getresinfo_cube:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_get_resinfo v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_CUBE unorm
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.getresinfo.cube.v4f32.i32(i32 15, i32 %mip, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @getresinfo_1darray(<8 x i32> inreg %rsrc, i32 %mip) {
 ; VERDE-LABEL: getresinfo_1darray:
 ; VERDE:       ; %bb.0: ; %main_body

 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
 ; NOPRT-LABEL: getresinfo_1darray:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    image_get_resinfo v[0:3], v0, s[0:7] dmask:0xf unorm da
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: getresinfo_1darray:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_get_resinfo v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D_ARRAY unorm ; encoding: [0x20,0x1f,0x38,0xf0,0x00,0x00,0x00,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: getresinfo_1darray:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_get_resinfo v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D_ARRAY unorm
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.getresinfo.1darray.v4f32.i32(i32 15, i32 %mip, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @getresinfo_2darray(<8 x i32> inreg %rsrc, i32 %mip) {
 ; VERDE-LABEL: getresinfo_2darray:
 ; VERDE:       ; %bb.0: ; %main_body

 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
 ; NOPRT-LABEL: getresinfo_2darray:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    image_get_resinfo v[0:3], v0, s[0:7] dmask:0xf unorm da
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: getresinfo_2darray:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_get_resinfo v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_ARRAY unorm ; encoding: [0x28,0x1f,0x38,0xf0,0x00,0x00,0x00,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: getresinfo_2darray:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_get_resinfo v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_ARRAY unorm
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.getresinfo.2darray.v4f32.i32(i32 15, i32 %mip, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @getresinfo_2dmsaa(<8 x i32> inreg %rsrc, i32 %mip) {
 ; VERDE-LABEL: getresinfo_2dmsaa:
 ; VERDE:       ; %bb.0: ; %main_body

 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
 ; NOPRT-LABEL: getresinfo_2dmsaa:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    image_get_resinfo v[0:3], v0, s[0:7] dmask:0xf unorm
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: getresinfo_2dmsaa:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_get_resinfo v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_MSAA unorm ; encoding: [0x30,0x1f,0x38,0xf0,0x00,0x00,0x00,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: getresinfo_2dmsaa:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_get_resinfo v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_MSAA unorm
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.getresinfo.2dmsaa.v4f32.i32(i32 15, i32 %mip, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @getresinfo_2darraymsaa(<8 x i32> inreg %rsrc, i32 %mip) {
 ; VERDE-LABEL: getresinfo_2darraymsaa:
 ; VERDE:       ; %bb.0: ; %main_body

 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
 ; NOPRT-LABEL: getresinfo_2darraymsaa:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    image_get_resinfo v[0:3], v0, s[0:7] dmask:0xf unorm da
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: getresinfo_2darraymsaa:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_get_resinfo v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_MSAA_ARRAY unorm ; encoding: [0x38,0x1f,0x38,0xf0,0x00,0x00,0x00,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: getresinfo_2darraymsaa:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_get_resinfo v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_2D_MSAA_ARRAY unorm
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.getresinfo.2darraymsaa.v4f32.i32(i32 15, i32 %mip, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps float @load_1d_V1(<8 x i32> inreg %rsrc, i32 %s) {
 ; VERDE-LABEL: load_1d_V1:
 ; VERDE:       ; %bb.0: ; %main_body

 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
 ; NOPRT-LABEL: load_1d_V1:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    image_load v0, v0, s[0:7] dmask:0x8 unorm
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: load_1d_V1:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_load v0, v0, s[0:7] dmask:0x8 dim:SQ_RSRC_IMG_1D unorm ; encoding: [0x00,0x18,0x00,0xf0,0x00,0x00,0x00,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: load_1d_V1:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_load v0, v0, s[0:7] dmask:0x8 dim:SQ_RSRC_IMG_1D unorm
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call float @llvm.amdgcn.image.load.1d.f32.i32(i32 8, i32 %s, <8 x i32> %rsrc, i32 0, i32 0)
   ret float %v
 }
 
 define amdgpu_ps <2 x float> @load_1d_V2(<8 x i32> inreg %rsrc, i32 %s) {
 ; VERDE-LABEL: load_1d_V2:
 ; VERDE:       ; %bb.0: ; %main_body

 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
 ; NOPRT-LABEL: load_1d_V2:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    image_load v[0:1], v0, s[0:7] dmask:0x9 unorm
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: load_1d_V2:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_load v[0:1], v0, s[0:7] dmask:0x9 dim:SQ_RSRC_IMG_1D unorm ; encoding: [0x00,0x19,0x00,0xf0,0x00,0x00,0x00,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: load_1d_V2:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_load v[0:1], v0, s[0:7] dmask:0x9 dim:SQ_RSRC_IMG_1D unorm
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <2 x float> @llvm.amdgcn.image.load.1d.v2f32.i32(i32 9, i32 %s, <8 x i32> %rsrc, i32 0, i32 0)
   ret <2 x float> %v
 }
 
 define amdgpu_ps void @store_1d_V1(<8 x i32> inreg %rsrc, float %vdata, i32 %s) {
 ; VERDE-LABEL: store_1d_V1:
 ; VERDE:       ; %bb.0: ; %main_body

 ; GFX6789-NEXT:    image_store v0, v1, s[0:7] dmask:0x2 unorm
 ; GFX6789-NEXT:    s_endpgm
 ;
 ; NOPRT-LABEL: store_1d_V1:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    image_store v0, v1, s[0:7] dmask:0x2 unorm
 ; NOPRT-NEXT:    s_endpgm
 ;
-; GFX10-LABEL: store_1d_V1:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_store v0, v1, s[0:7] dmask:0x2 dim:SQ_RSRC_IMG_1D unorm ; encoding: [0x00,0x12,0x20,0xf0,0x01,0x00,0x00,0x00]
-; GFX10-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
+; GFX10PLUS-LABEL: store_1d_V1:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_store v0, v1, s[0:7] dmask:0x2 dim:SQ_RSRC_IMG_1D unorm
+; GFX10PLUS-NEXT:    s_endpgm
 main_body:
   call void @llvm.amdgcn.image.store.1d.f32.i32(float %vdata, i32 2, i32 %s, <8 x i32> %rsrc, i32 0, i32 0)
   ret void
 }
 
 define amdgpu_ps void @store_1d_V2(<8 x i32> inreg %rsrc, <2 x float> %vdata, i32 %s) {
 ; VERDE-LABEL: store_1d_V2:
 ; VERDE:       ; %bb.0: ; %main_body

 ; GFX6789-NEXT:    image_store v[0:1], v2, s[0:7] dmask:0xc unorm
 ; GFX6789-NEXT:    s_endpgm
 ;
 ; NOPRT-LABEL: store_1d_V2:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    image_store v[0:1], v2, s[0:7] dmask:0xc unorm
 ; NOPRT-NEXT:    s_endpgm
 ;
-; GFX10-LABEL: store_1d_V2:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_store v[0:1], v2, s[0:7] dmask:0xc dim:SQ_RSRC_IMG_1D unorm ; encoding: [0x00,0x1c,0x20,0xf0,0x02,0x00,0x00,0x00]
-; GFX10-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
+; GFX10PLUS-LABEL: store_1d_V2:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_store v[0:1], v2, s[0:7] dmask:0xc dim:SQ_RSRC_IMG_1D unorm
+; GFX10PLUS-NEXT:    s_endpgm
 main_body:
   call void @llvm.amdgcn.image.store.1d.v2f32.i32(<2 x float> %vdata, i32 12, i32 %s, <8 x i32> %rsrc, i32 0, i32 0)
   ret void
 }
 
 define amdgpu_ps <4 x float> @load_1d_glc(<8 x i32> inreg %rsrc, i32 %s) {
 ; VERDE-LABEL: load_1d_glc:
 ; VERDE:       ; %bb.0: ; %main_body

 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
 ; NOPRT-LABEL: load_1d_glc:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    image_load v[0:3], v0, s[0:7] dmask:0xf unorm glc
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: load_1d_glc:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_load v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm glc ; encoding: [0x00,0x3f,0x00,0xf0,0x00,0x00,0x00,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: load_1d_glc:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_load v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm glc
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.load.1d.v4f32.i32(i32 15, i32 %s, <8 x i32> %rsrc, i32 0, i32 1)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @load_1d_slc(<8 x i32> inreg %rsrc, i32 %s) {
 ; VERDE-LABEL: load_1d_slc:
 ; VERDE:       ; %bb.0: ; %main_body

 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
 ; NOPRT-LABEL: load_1d_slc:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    image_load v[0:3], v0, s[0:7] dmask:0xf unorm slc
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: load_1d_slc:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_load v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm slc ; encoding: [0x00,0x1f,0x00,0xf2,0x00,0x00,0x00,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: load_1d_slc:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_load v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm slc
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.load.1d.v4f32.i32(i32 15, i32 %s, <8 x i32> %rsrc, i32 0, i32 2)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @load_1d_glc_slc(<8 x i32> inreg %rsrc, i32 %s) {
 ; VERDE-LABEL: load_1d_glc_slc:
 ; VERDE:       ; %bb.0: ; %main_body

 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
 ; NOPRT-LABEL: load_1d_glc_slc:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    image_load v[0:3], v0, s[0:7] dmask:0xf unorm glc slc
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: load_1d_glc_slc:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_load v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm glc slc ; encoding: [0x00,0x3f,0x00,0xf2,0x00,0x00,0x00,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: load_1d_glc_slc:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_load v[0:3], v0, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm glc slc
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.load.1d.v4f32.i32(i32 15, i32 %s, <8 x i32> %rsrc, i32 0, i32 3)
   ret <4 x float> %v
 }
 
 define amdgpu_ps void @store_1d_glc(<8 x i32> inreg %rsrc, <4 x float> %vdata, i32 %s) {
 ; VERDE-LABEL: store_1d_glc:
 ; VERDE:       ; %bb.0: ; %main_body

 ; GFX6789-NEXT:    image_store v[0:3], v4, s[0:7] dmask:0xf unorm glc
 ; GFX6789-NEXT:    s_endpgm
 ;
 ; NOPRT-LABEL: store_1d_glc:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    image_store v[0:3], v4, s[0:7] dmask:0xf unorm glc
 ; NOPRT-NEXT:    s_endpgm
 ;
-; GFX10-LABEL: store_1d_glc:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_store v[0:3], v4, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm glc ; encoding: [0x00,0x3f,0x20,0xf0,0x04,0x00,0x00,0x00]
-; GFX10-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
+; GFX10PLUS-LABEL: store_1d_glc:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_store v[0:3], v4, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm glc
+; GFX10PLUS-NEXT:    s_endpgm
 main_body:
   call void @llvm.amdgcn.image.store.1d.v4f32.i32(<4 x float> %vdata, i32 15, i32 %s, <8 x i32> %rsrc, i32 0, i32 1)
   ret void
 }
 
 define amdgpu_ps void @store_1d_slc(<8 x i32> inreg %rsrc, <4 x float> %vdata, i32 %s) {
 ; VERDE-LABEL: store_1d_slc:
 ; VERDE:       ; %bb.0: ; %main_body

 ; GFX6789-NEXT:    image_store v[0:3], v4, s[0:7] dmask:0xf unorm slc
 ; GFX6789-NEXT:    s_endpgm
 ;
 ; NOPRT-LABEL: store_1d_slc:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    image_store v[0:3], v4, s[0:7] dmask:0xf unorm slc
 ; NOPRT-NEXT:    s_endpgm
 ;
-; GFX10-LABEL: store_1d_slc:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_store v[0:3], v4, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm slc ; encoding: [0x00,0x1f,0x20,0xf2,0x04,0x00,0x00,0x00]
-; GFX10-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
+; GFX10PLUS-LABEL: store_1d_slc:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_store v[0:3], v4, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm slc
+; GFX10PLUS-NEXT:    s_endpgm
 main_body:
   call void @llvm.amdgcn.image.store.1d.v4f32.i32(<4 x float> %vdata, i32 15, i32 %s, <8 x i32> %rsrc, i32 0, i32 2)
   ret void
 }
 
 define amdgpu_ps void @store_1d_glc_slc(<8 x i32> inreg %rsrc, <4 x float> %vdata, i32 %s) {
 ; VERDE-LABEL: store_1d_glc_slc:
 ; VERDE:       ; %bb.0: ; %main_body

 ; GFX6789-NEXT:    image_store v[0:3], v4, s[0:7] dmask:0xf unorm glc slc
 ; GFX6789-NEXT:    s_endpgm
 ;
 ; NOPRT-LABEL: store_1d_glc_slc:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    image_store v[0:3], v4, s[0:7] dmask:0xf unorm glc slc
 ; NOPRT-NEXT:    s_endpgm
 ;
-; GFX10-LABEL: store_1d_glc_slc:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_store v[0:3], v4, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm glc slc ; encoding: [0x00,0x3f,0x20,0xf2,0x04,0x00,0x00,0x00]
-; GFX10-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
+; GFX10PLUS-LABEL: store_1d_glc_slc:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_store v[0:3], v4, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm glc slc
+; GFX10PLUS-NEXT:    s_endpgm
 main_body:
   call void @llvm.amdgcn.image.store.1d.v4f32.i32(<4 x float> %vdata, i32 15, i32 %s, <8 x i32> %rsrc, i32 0, i32 3)
   ret void
 }
 
 define amdgpu_ps <3 x float> @getresinfo_dmask7(<8 x i32> inreg %rsrc, <4 x float> %vdata, i32 %mip) {
 ; GFX6-LABEL: getresinfo_dmask7:
 ; GFX6:       ; %bb.0: ; %main_body

 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
 ; NOPRT-LABEL: getresinfo_dmask7:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    image_get_resinfo v[0:2], v0, s[0:7] dmask:0x7 unorm
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: getresinfo_dmask7:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_get_resinfo v[0:2], v0, s[0:7] dmask:0x7 dim:SQ_RSRC_IMG_1D unorm ; encoding: [0x00,0x17,0x38,0xf0,0x00,0x00,0x00,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: getresinfo_dmask7:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_get_resinfo v[0:2], v0, s[0:7] dmask:0x7 dim:SQ_RSRC_IMG_1D unorm
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %r = call <3 x float> @llvm.amdgcn.image.getresinfo.1d.v3f32.i32(i32 7, i32 %mip, <8 x i32> %rsrc, i32 0, i32 0)
   ret <3 x float> %r
 }
 
 define amdgpu_ps <2 x float> @getresinfo_dmask3(<8 x i32> inreg %rsrc, <4 x float> %vdata, i32 %mip) {
 ; GFX6-LABEL: getresinfo_dmask3:
 ; GFX6:       ; %bb.0: ; %main_body

 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
 ; NOPRT-LABEL: getresinfo_dmask3:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    image_get_resinfo v[0:1], v0, s[0:7] dmask:0x3 unorm
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: getresinfo_dmask3:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_get_resinfo v[0:1], v0, s[0:7] dmask:0x3 dim:SQ_RSRC_IMG_1D unorm ; encoding: [0x00,0x13,0x38,0xf0,0x00,0x00,0x00,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: getresinfo_dmask3:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_get_resinfo v[0:1], v0, s[0:7] dmask:0x3 dim:SQ_RSRC_IMG_1D unorm
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %r = call <2 x float> @llvm.amdgcn.image.getresinfo.1d.v2f32.i32(i32 3, i32 %mip, <8 x i32> %rsrc, i32 0, i32 0)
   ret <2 x float> %r
 }
 
 define amdgpu_ps float @getresinfo_dmask1(<8 x i32> inreg %rsrc, <4 x float> %vdata, i32 %mip) {
 ; GFX6-LABEL: getresinfo_dmask1:
 ; GFX6:       ; %bb.0: ; %main_body

 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
 ; NOPRT-LABEL: getresinfo_dmask1:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    image_get_resinfo v0, v0, s[0:7] dmask:0x1 unorm
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: getresinfo_dmask1:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_get_resinfo v0, v0, s[0:7] dmask:0x1 dim:SQ_RSRC_IMG_1D unorm ; encoding: [0x00,0x11,0x38,0xf0,0x00,0x00,0x00,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: getresinfo_dmask1:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_get_resinfo v0, v0, s[0:7] dmask:0x1 dim:SQ_RSRC_IMG_1D unorm
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %r = call float @llvm.amdgcn.image.getresinfo.1d.f32.i32(i32 1, i32 %mip, <8 x i32> %rsrc, i32 0, i32 0)
   ret float %r
 }
 
 define amdgpu_ps <4 x float> @getresinfo_dmask0(<8 x i32> inreg %rsrc, <4 x float> %vdata, i32 %mip) #0 {
 ; VERDE-LABEL: getresinfo_dmask0:
 ; VERDE:       ; %bb.0: ; %main_body
 ; VERDE-NEXT:    ; return to shader part epilog
 ;
 ; FIJI-LABEL: getresinfo_dmask0:
 ; FIJI:       ; %bb.0: ; %main_body
 ; FIJI-NEXT:    ; return to shader part epilog
 ;
 ; GFX6789-LABEL: getresinfo_dmask0:
 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
 ; NOPRT-LABEL: getresinfo_dmask0:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: getresinfo_dmask0:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: getresinfo_dmask0:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %r = call <4 x float> @llvm.amdgcn.image.getresinfo.1d.v4f32.i32(i32 0, i32 %mip, <8 x i32> %rsrc, i32 0, i32 0)
   ret <4 x float> %r
 }
 
 ;
 define amdgpu_ps void @image_store_wait(<8 x i32> inreg %arg, <8 x i32> inreg %arg1, <8 x i32> inreg %arg2, <4 x float> %arg3, i32 %arg4) #0 {
 ; VERDE-LABEL: image_store_wait:

 ; NOPRT-LABEL: image_store_wait:
 ; NOPRT:       ; %bb.0: ; %main_body
 ; NOPRT-NEXT:    image_store v[0:3], v4, s[0:7] dmask:0xf unorm
 ; NOPRT-NEXT:    image_load v[0:3], v4, s[8:15] dmask:0xf unorm
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    image_store v[0:3], v4, s[16:23] dmask:0xf unorm
 ; NOPRT-NEXT:    s_endpgm
 ;
-; GFX10-LABEL: image_store_wait:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_store v[0:3], v4, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm ; encoding: [0x00,0x1f,0x20,0xf0,0x04,0x00,0x00,0x00]
-; GFX10-NEXT:    image_load v[0:3], v4, s[8:15] dmask:0xf dim:SQ_RSRC_IMG_1D unorm ; encoding: [0x00,0x1f,0x00,0xf0,0x04,0x00,0x02,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    image_store v[0:3], v4, s[16:23] dmask:0xf dim:SQ_RSRC_IMG_1D unorm ; encoding: [0x00,0x1f,0x20,0xf0,0x04,0x00,0x04,0x00]
-; GFX10-NEXT:    s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
+; GFX10PLUS-LABEL: image_store_wait:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_store v[0:3], v4, s[0:7] dmask:0xf dim:SQ_RSRC_IMG_1D unorm
+; GFX10PLUS-NEXT:    image_load v[0:3], v4, s[8:15] dmask:0xf dim:SQ_RSRC_IMG_1D unorm
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    image_store v[0:3], v4, s[16:23] dmask:0xf dim:SQ_RSRC_IMG_1D unorm
+; GFX10PLUS-NEXT:    s_endpgm
 main_body:
   call void @llvm.amdgcn.image.store.1d.v4f32.i32(<4 x float> %arg3, i32 15, i32 %arg4, <8 x i32> %arg, i32 0, i32 0)
   %data = call <4 x float> @llvm.amdgcn.image.load.1d.v4f32.i32(i32 15, i32 %arg4, <8 x i32> %arg1, i32 0, i32 0)
   call void @llvm.amdgcn.image.store.1d.v4f32.i32(<4 x float> %data, i32 15, i32 %arg4, <8 x i32> %arg2, i32 0, i32 0)
   ret void
 }
 
 define amdgpu_ps float @image_load_mmo(<8 x i32> inreg %rsrc, float addrspace(3)* %lds, <2 x i32> %c) #0 {

 ; NOPRT-NEXT:    ds_write2_b32 v0, v2, v2 offset1:4
 ; NOPRT-NEXT:    s_waitcnt vmcnt(0)
 ; NOPRT-NEXT:    v_mov_b32_e32 v0, v1
 ; NOPRT-NEXT:    s_waitcnt lgkmcnt(0)
 ; NOPRT-NEXT:    ; return to shader part epilog
 ;
 ; GFX10-LABEL: image_load_mmo:
 ; GFX10:       ; %bb.0:
-; GFX10-NEXT:    image_load v1, v[1:2], s[0:7] dmask:0x1 dim:SQ_RSRC_IMG_2D unorm ; encoding: [0x08,0x11,0x00,0xf0,0x01,0x01,0x00,0x00]
-; GFX10-NEXT:    v_mov_b32_e32 v2, 0 ; encoding: [0x80,0x02,0x04,0x7e]
-; GFX10-NEXT:    ds_write2_b32 v0, v2, v2 offset1:4 ; encoding: [0x00,0x04,0x38,0xd8,0x00,0x02,0x02,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    v_mov_b32_e32 v0, v1 ; encoding: [0x01,0x03,0x00,0x7e]
-; GFX10-NEXT:    s_waitcnt lgkmcnt(0) ; encoding: [0x7f,0xc0,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10-NEXT:    image_load v1, v[1:2], s[0:7] dmask:0x1 dim:SQ_RSRC_IMG_2D unorm
+; GFX10-NEXT:    v_mov_b32_e32 v2, 0
+; GFX10-NEXT:    ds_write2_b32 v0, v2, v2 offset1:4
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    v_mov_b32_e32 v0, v1
+; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
+; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: image_load_mmo:
+; GFX11:       ; %bb.0:
+; GFX11-NEXT:    image_load v1, v[1:2], s[0:7] dmask:0x1 dim:SQ_RSRC_IMG_2D unorm
+; GFX11-NEXT:    v_mov_b32_e32 v2, 0
+; GFX11-NEXT:    ds_store_2addr_b32 v0, v2, v2 offset1:4
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    v_mov_b32_e32 v0, v1
+; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
+; GFX11-NEXT:    ; return to shader part epilog
   store float 0.000000e+00, float addrspace(3)* %lds
   %c0 = extractelement <2 x i32> %c, i32 0
   %c1 = extractelement <2 x i32> %c, i32 1
   %tex = call float @llvm.amdgcn.image.load.2d.f32.i32(i32 1, i32 %c0, i32 %c1, <8 x i32> %rsrc, i32 0, i32 0)
   %tmp2 = getelementptr float, float addrspace(3)* %lds, i32 4
   store float 0.000000e+00, float addrspace(3)* %tmp2
   ret float %tex
 }

llvm/test/CodeGen/AMDGPU/llvm.amdgcn.image.gather4.a16.dim.ll

 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
 ; RUN: llc -march=amdgcn -mcpu=gfx900 -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX9 %s
 ; RUN: llc -march=amdgcn -mcpu=gfx1010 -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX10 %s
+; RUN: llc -march=amdgcn -mcpu=gfx1100 -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX10 %s
 
 define amdgpu_ps <4 x float> @gather4_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, half %s, half %t) {
 ; GFX9-LABEL: gather4_2d:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    s_mov_b64 s[12:13], exec
 ; GFX9-NEXT:    s_wqm_b64 exec, exec
 ; GFX9-NEXT:    v_and_b32_e32 v0, 0xffff, v0
 ; GFX9-NEXT:    v_lshl_or_b32 v0, v1, 16, v0

llvm/test/CodeGen/AMDGPU/llvm.amdgcn.image.gather4.d16.dim.ll

 ; RUN: llc < %s -march=amdgcn -mcpu=tonga -verify-machineinstrs | FileCheck -check-prefixes=GCN,UNPACKED %s
 ; RUN: llc < %s -march=amdgcn -mcpu=gfx810 -verify-machineinstrs | FileCheck --check-prefix=GCN %s
 ; RUN: llc < %s -march=amdgcn -mcpu=gfx900 -verify-machineinstrs | FileCheck -check-prefixes=GCN,GFX9 %s
 ; RUN: llc < %s -march=amdgcn -mcpu=gfx1010 -verify-machineinstrs | FileCheck -check-prefixes=GCN,GFX10 %s
+; RUN: llc < %s -march=amdgcn -mcpu=gfx1100 -verify-machineinstrs | FileCheck -check-prefixes=GCN,GFX10 %s
 
 ; GCN-LABEL: {{^}}image_gather4_b_2d_v4f16:
 ; UNPACKED: image_gather4_b v[0:3], v[0:2], s[0:7], s[8:11] dmask:0x4 d16{{$}}
 ; PACKED: image_gather4_b v[0:1], v[0:2], s[0:7], s[8:11] dmask:0x4 d16{{$}}
 ; GFX810: image_gather4_b v[0:3], v[0:2], s[0:7], s[8:11] dmask:0x4 d16{{$}}
 ; GFX9: image_gather4_b v[0:3], v[0:2], s[0:7], s[8:11] dmask:0x4 d16{{$}}
 ; GFX10: image_gather4_b v[0:1], v[0:2], s[0:7], s[8:11] dmask:0x4 dim:SQ_RSRC_IMG_2D d16{{$}}
 define amdgpu_ps <2 x float> @image_gather4_b_2d_v4f16(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %bias, float %s, float %t) {

llvm/test/CodeGen/AMDGPU/llvm.amdgcn.image.gather4.dim.ll

 ; RUN: llc -march=amdgcn -mcpu=verde -verify-machineinstrs < %s | FileCheck -check-prefixes=GCN,GFX6789 %s
 ; RUN: llc -march=amdgcn -mcpu=gfx900 -verify-machineinstrs < %s | FileCheck -check-prefixes=GCN,GFX6789 %s
 ; RUN: llc -march=amdgcn -mcpu=gfx1010 -verify-machineinstrs -show-mc-encoding < %s | FileCheck -check-prefixes=GCN,GFX10 %s
+; RUN: llc -march=amdgcn -mcpu=gfx1100 -verify-machineinstrs -show-mc-encoding < %s | FileCheck -check-prefixes=GCN,GFX10 %s
 
 ; GCN-LABEL: {{^}}gather4_2d:
 ; GFX6789: image_gather4 v[0:3], v[0:1], s[0:7], s[8:11] dmask:0x1{{$}}
 ; GFX10: image_gather4 v[0:3], v[0:1], s[0:7], s[8:11] dmask:0x1 dim:SQ_RSRC_IMG_2D ;
 define amdgpu_ps <4 x float> @gather4_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %s, float %t) {
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.gather4.2d.v4f32.f32(i32 1, float %s, float %t, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v

llvm/test/CodeGen/AMDGPU/llvm.amdgcn.image.getlod.dim.ll

 ; RUN: llc < %s -march=amdgcn -mcpu=verde -verify-machineinstrs | FileCheck --check-prefixes=GCN,PRE-GFX10 %s
 ; RUN: llc < %s -march=amdgcn -mcpu=gfx900 -verify-machineinstrs | FileCheck --check-prefixes=GCN,PRE-GFX10 %s
 ; RUN: llc < %s -march=amdgcn -mcpu=gfx1010 -verify-machineinstrs | FileCheck --check-prefixes=GCN,GFX10 %s
+; RUN: llc < %s -march=amdgcn -mcpu=gfx1100 -verify-machineinstrs | FileCheck --check-prefixes=GCN,GFX10 %s
 
 ; GCN-LABEL: {{^}}getlod_1d:
 ; PRE-GFX10: image_get_lod v[0:3], v0, s[0:7], s[8:11] dmask:0xf{{$}}
 ; GFX10: image_get_lod v[0:3], v0, s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D
 ; GCN: s_waitcnt vmcnt(0)
 define amdgpu_ps <4 x float> @getlod_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %s) {
 main_body:
   %r = call <4 x float> @llvm.amdgcn.image.getlod.1d.v4f32.f32(i32 15, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)

llvm/test/CodeGen/AMDGPU/llvm.amdgcn.image.load.a16.d16.ll

 ; RUN: llc -march=amdgcn -mcpu=gfx900 -verify-machineinstrs < %s | FileCheck -check-prefixes=GCN,GFX9 %s
 ; RUN: llc -march=amdgcn -mcpu=gfx1010 -verify-machineinstrs < %s | FileCheck -check-prefixes=GCN,GFX10 %s
+; RUN: llc -march=amdgcn -mcpu=gfx1100 -verify-machineinstrs < %s | FileCheck -check-prefixes=GCN,GFX10 %s
 
 ; GCN-LABEL: {{^}}load.f16.1d:
 ; GFX9: image_load v0, v0, s[0:7] dmask:0x1 unorm a16 d16
 ; GFX10: image_load v0, v0, s[0:7] dmask:0x1 dim:SQ_RSRC_IMG_1D unorm a16 d16
 define amdgpu_ps <4 x half> @load.f16.1d(<8 x i32> inreg %rsrc, <2 x i16> %coords) {
 main_body:
   %x = extractelement <2 x i16> %coords, i32 0
   %v = call <4 x half> @llvm.amdgcn.image.load.1d.v4f16.i16(i32 1, i16 %x, <8 x i32> %rsrc, i32 0, i32 0)

llvm/test/CodeGen/AMDGPU/llvm.amdgcn.image.load.a16.ll

 ; RUN: llc -march=amdgcn -mcpu=gfx900 -verify-machineinstrs < %s | FileCheck -check-prefixes=GCN,GFX9 %s
 ; RUN: llc -march=amdgcn -mcpu=gfx1010 -verify-machineinstrs < %s | FileCheck -check-prefixes=GCN,GFX10 %s
+; RUN: llc -march=amdgcn -mcpu=gfx1100 -verify-machineinstrs < %s | FileCheck -check-prefixes=GCN,GFX10 %s
 
 ; GCN-LABEL: {{^}}load.f32.1d:
 ; GFX9: image_load v0, v0, s[0:7] dmask:0x1 unorm a16
 ; GFX10: image_load v0, v0, s[0:7] dmask:0x1 dim:SQ_RSRC_IMG_1D unorm a16
 define amdgpu_ps <4 x float> @load.f32.1d(<8 x i32> inreg %rsrc, <2 x i16> %coords) {
 main_body:
   %x = extractelement <2 x i16> %coords, i32 0
   %v = call <4 x float> @llvm.amdgcn.image.load.1d.v4f32.i16(i32 1, i16 %x, <8 x i32> %rsrc, i32 0, i32 0)

llvm/test/CodeGen/AMDGPU/llvm.amdgcn.image.msaa.load.ll

+; RUN: llc -march=amdgcn -mcpu=gfx1100 -verify-machineinstrs -show-mc-encoding < %s | FileCheck -check-prefixes=GCN,GFX11 %s
+
+; GCN-LABEL: {{^}}load_2dmsaa:
+; GFX11: image_msaa_load v[0:3], v[0:2], s[0:7] dmask:0x1 dim:SQ_RSRC_IMG_2D_MSAA unorm ;
+define amdgpu_ps <4 x float> @load_2dmsaa(<8 x i32> inreg %rsrc, i32 %s, i32 %t, i32 %fragid) {
+main_body:
+  %v = call <4 x float> @llvm.amdgcn.image.msaa.load.2dmsaa.v4f32.i32(i32 1, i32 %s, i32 %t, i32 %fragid, <8 x i32> %rsrc, i32 0, i32 0)
+  ret <4 x float> %v
+}
+
+; GCN-LABEL: {{^}}load_2dmsaa_both:
+; GFX11: image_msaa_load v[0:4], v[{{[0-9]+:[0-9]+}}], s[0:7] dmask:0x2 dim:SQ_RSRC_IMG_2D_MSAA unorm tfe lwe ;
+define amdgpu_ps <4 x float> @load_2dmsaa_both(<8 x i32> inreg %rsrc, i32 addrspace(1)* inreg %out, i32 %s, i32 %t, i32 %fragid) {
+main_body:
+  %v = call {<4 x float>,i32} @llvm.amdgcn.image.msaa.load.2dmsaa.v4f32i32.i32(i32 2, i32 %s, i32 %t, i32 %fragid, <8 x i32> %rsrc, i32 3, i32 0)
+  %v.vec = extractvalue {<4 x float>, i32} %v, 0
+  %v.err = extractvalue {<4 x float>, i32} %v, 1
+  store i32 %v.err, i32 addrspace(1)* %out, align 4
+  ret <4 x float> %v.vec
+}
+
+; GCN-LABEL: {{^}}load_2darraymsaa:
+; GFX11: image_msaa_load v[0:3], v[0:3], s[0:7] dmask:0x4 dim:SQ_RSRC_IMG_2D_MSAA_ARRAY unorm ;
+define amdgpu_ps <4 x float> @load_2darraymsaa(<8 x i32> inreg %rsrc, i32 %s, i32 %t, i32 %slice, i32 %fragid) {
+main_body:
+  %v = call <4 x float> @llvm.amdgcn.image.msaa.load.2darraymsaa.v4f32.i32(i32 4, i32 %s, i32 %t, i32 %slice, i32 %fragid, <8 x i32> %rsrc, i32 0, i32 0)
+  ret <4 x float> %v
+}
+
+; GCN-LABEL: {{^}}load_2darraymsaa_tfe:
+; GFX11: image_msaa_load v[0:4], v[{{[0-9]+:[0-9]+}}], s[0:7] dmask:0x8 dim:SQ_RSRC_IMG_2D_MSAA_ARRAY unorm tfe ;
+define amdgpu_ps <4 x float> @load_2darraymsaa_tfe(<8 x i32> inreg %rsrc, i32 addrspace(1)* inreg %out, i32 %s, i32 %t, i32 %slice, i32 %fragid) {
+main_body:
+  %v = call {<4 x float>,i32} @llvm.amdgcn.image.msaa.load.2darraymsaa.v4f32i32.i32(i32 8, i32 %s, i32 %t, i32 %slice, i32 %fragid, <8 x i32> %rsrc, i32 1, i32 0)
+  %v.vec = extractvalue {<4 x float>, i32} %v, 0
+  %v.err = extractvalue {<4 x float>, i32} %v, 1
+  store i32 %v.err, i32 addrspace(1)* %out, align 4
+  ret <4 x float> %v.vec
+}
+
+; GCN-LABEL: {{^}}load_2dmsaa_glc:
+; GFX11: image_msaa_load v[0:3], v[{{[0-9]+:[0-9]+}}], s[0:7] dmask:0x1 dim:SQ_RSRC_IMG_2D_MSAA unorm glc ;
+define amdgpu_ps <4 x float> @load_2dmsaa_glc(<8 x i32> inreg %rsrc, i32 %s, i32 %t, i32 %fragid) {
+main_body:
+  %v = call <4 x float> @llvm.amdgcn.image.msaa.load.2dmsaa.v4f32.i32(i32 1, i32 %s, i32 %t, i32 %fragid, <8 x i32> %rsrc, i32 0, i32 1)
+  ret <4 x float> %v
+}
+
+; GCN-LABEL: {{^}}load_2dmsaa_slc:
+; GFX11: image_msaa_load v[0:3], v[{{[0-9]+:[0-9]+}}], s[0:7] dmask:0x1 dim:SQ_RSRC_IMG_2D_MSAA unorm slc ;
+define amdgpu_ps <4 x float> @load_2dmsaa_slc(<8 x i32> inreg %rsrc, i32 %s, i32 %t, i32 %fragid) {
+main_body:
+  %v = call <4 x float> @llvm.amdgcn.image.msaa.load.2dmsaa.v4f32.i32(i32 1, i32 %s, i32 %t, i32 %fragid, <8 x i32> %rsrc, i32 0, i32 2)
+  ret <4 x float> %v
+}
+
+; GCN-LABEL: {{^}}load_2dmsaa_glc_slc:
+; GFX11: image_msaa_load v[0:3], v[{{[0-9]+:[0-9]+}}], s[0:7] dmask:0x1 dim:SQ_RSRC_IMG_2D_MSAA unorm glc slc ;
+define amdgpu_ps <4 x float> @load_2dmsaa_glc_slc(<8 x i32> inreg %rsrc, i32 %s, i32 %t, i32 %fragid) {
+main_body:
+  %v = call <4 x float> @llvm.amdgcn.image.msaa.load.2dmsaa.v4f32.i32(i32 1, i32 %s, i32 %t, i32 %fragid, <8 x i32> %rsrc, i32 0, i32 3)
+  ret <4 x float> %v
+}
+
+; GCN-LABEL: {{^}}load_2dmsaa_d16:
+; GFX11: image_msaa_load v[0:1], v[0:2], s[0:7] dmask:0x1 dim:SQ_RSRC_IMG_2D_MSAA unorm d16 ;
+define amdgpu_ps <4 x half> @load_2dmsaa_d16(<8 x i32> inreg %rsrc, i32 %s, i32 %t, i32 %fragid) {
+main_body:
+  %v = call <4 x half> @llvm.amdgcn.image.msaa.load.2dmsaa.v4f16.i32(i32 1, i32 %s, i32 %t, i32 %fragid, <8 x i32> %rsrc, i32 0, i32 0)
+  ret <4 x half> %v
+}
+
+; GCN-LABEL: {{^}}load_2dmsaa_tfe_d16:
+; GFX11: image_msaa_load v[0:2], v[{{[0-9]+:[0-9]+}}], s[0:7] dmask:0x1 dim:SQ_RSRC_IMG_2D_MSAA unorm tfe d16 ;
+define amdgpu_ps <4 x half> @load_2dmsaa_tfe_d16(<8 x i32> inreg %rsrc, i32 addrspace(1)* inreg %out, i32 %s, i32 %t, i32 %fragid) {
+main_body:
+  %v = call {<4 x half>,i32} @llvm.amdgcn.image.msaa.load.2dmsaa.v4f16i32.i32(i32 1, i32 %s, i32 %t, i32 %fragid, <8 x i32> %rsrc, i32 1, i32 0)
+  %v.vec = extractvalue {<4 x half>, i32} %v, 0
+  %v.err = extractvalue {<4 x half>, i32} %v, 1
+  store i32 %v.err, i32 addrspace(1)* %out, align 4
+  ret <4 x half> %v.vec
+}
+
+; GCN-LABEL: {{^}}load_2darraymsaa_d16:
+; GFX11: image_msaa_load v[0:1], v[0:3], s[0:7] dmask:0x1 dim:SQ_RSRC_IMG_2D_MSAA_ARRAY unorm d16 ;
+define amdgpu_ps <4 x half> @load_2darraymsaa_d16(<8 x i32> inreg %rsrc, i32 %s, i32 %t, i32 %slice, i32 %fragid) {
+main_body:
+  %v = call <4 x half> @llvm.amdgcn.image.msaa.load.2darraymsaa.v4f16.i32(i32 1, i32 %s, i32 %t, i32 %slice, i32 %fragid, <8 x i32> %rsrc, i32 0, i32 0)
+  ret <4 x half> %v
+}
+
+; GCN-LABEL: {{^}}load_2darraymsaa_tfe_d16:
+; GFX11: image_msaa_load v[0:2], v[0:3], s[0:7] dmask:0x1 dim:SQ_RSRC_IMG_2D_MSAA_ARRAY unorm tfe d16 ;
+define amdgpu_ps <4 x half> @load_2darraymsaa_tfe_d16(<8 x i32> inreg %rsrc, i32 addrspace(1)* inreg %out, i32 %s, i32 %t, i32 %slice, i32 %fragid) {
+main_body:
+  %v = call {<4 x half>,i32} @llvm.amdgcn.image.msaa.load.2darraymsaa.v4f16i32.i32(i32 1, i32 %s, i32 %t, i32 %slice, i32 %fragid, <8 x i32> %rsrc, i32 1, i32 0)
+  %v.vec = extractvalue {<4 x half>, i32} %v, 0
+  %v.err = extractvalue {<4 x half>, i32} %v, 1
+  store i32 %v.err, i32 addrspace(1)* %out, align 4
+  ret <4 x half> %v.vec
+}
+
+; GCN-LABEL: {{^}}load_2dmsaa_a16:
+; GFX11: image_msaa_load v[0:3], v[1:2], s[0:7] dmask:0x1 dim:SQ_RSRC_IMG_2D_MSAA unorm a16 ;
+define amdgpu_ps <4 x float> @load_2dmsaa_a16(<8 x i32> inreg %rsrc, i16 %s, i16 %t, i16 %fragid) {
+main_body:
+  %v = call <4 x float> @llvm.amdgcn.image.msaa.load.2dmsaa.v4f32.i16(i32 1, i16 %s, i16 %t, i16 %fragid, <8 x i32> %rsrc, i32 0, i32 0)
+  ret <4 x float> %v
+}
+
+; GCN-LABEL: {{^}}load_2darraymsaa_a16:
+; GFX11: image_msaa_load v[0:3], v[1:2], s[0:7] dmask:0x4 dim:SQ_RSRC_IMG_2D_MSAA_ARRAY unorm a16 ;
+define amdgpu_ps <4 x float> @load_2darraymsaa_a16(<8 x i32> inreg %rsrc, i16 %s, i16 %t, i16 %slice, i16 %fragid) {
+main_body:
+  %v = call <4 x float> @llvm.amdgcn.image.msaa.load.2darraymsaa.v4f32.i16(i32 4, i16 %s, i16 %t, i16 %slice, i16 %fragid, <8 x i32> %rsrc, i32 0, i32 0)
+  ret <4 x float> %v
+}
+
+declare <4 x float> @llvm.amdgcn.image.msaa.load.2dmsaa.v4f32.i32(i32, i32, i32, i32, <8 x i32>, i32, i32) #1
+declare {<4 x float>,i32} @llvm.amdgcn.image.msaa.load.2dmsaa.v4f32i32.i32(i32, i32, i32, i32, <8 x i32>, i32, i32) #1
+declare <4 x float> @llvm.amdgcn.image.msaa.load.2darraymsaa.v4f32.i32(i32, i32, i32, i32, i32, <8 x i32>, i32, i32) #1
+declare {<4 x float>,i32} @llvm.amdgcn.image.msaa.load.2darraymsaa.v4f32i32.i32(i32, i32, i32, i32, i32, <8 x i32>, i32, i32) #1
+
+declare <4 x half> @llvm.amdgcn.image.msaa.load.2dmsaa.v4f16.i32(i32, i32, i32, i32, <8 x i32>, i32, i32) #1
+declare {<4 x half>,i32} @llvm.amdgcn.image.msaa.load.2dmsaa.v4f16i32.i32(i32, i32, i32, i32, <8 x i32>, i32, i32) #1
+declare <4 x half> @llvm.amdgcn.image.msaa.load.2darraymsaa.v4f16.i32(i32, i32, i32, i32, i32, <8 x i32>, i32, i32) #1
+declare {<4 x half>,i32} @llvm.amdgcn.image.msaa.load.2darraymsaa.v4f16i32.i32(i32, i32, i32, i32, i32, <8 x i32>, i32, i32) #1
+
+declare <4 x float> @llvm.amdgcn.image.msaa.load.2dmsaa.v4f32.i16(i32, i16, i16, i16, <8 x i32>, i32, i32) #1
+declare <4 x float> @llvm.amdgcn.image.msaa.load.2darraymsaa.v4f32.i16(i32, i16, i16, i16, i16, <8 x i32>, i32, i32) #1
+
+attributes #0 = { nounwind }
+attributes #1 = { nounwind readonly }

llvm/test/CodeGen/AMDGPU/llvm.amdgcn.image.nsa.ll

-; RUN: llc -march=amdgcn -mcpu=gfx1010 -mattr=-nsa-encoding -verify-machineinstrs < %s | FileCheck -check-prefixes=GCN,NONSA %s
-; RUN: llc -march=amdgcn -mcpu=gfx1010 -verify-machineinstrs < %s | FileCheck -check-prefixes=GCN,GFX1010,NSA %s
-; RUN: llc -march=amdgcn -mcpu=gfx1030 -verify-machineinstrs < %s | FileCheck -check-prefixes=GCN,GFX1030,NSA %s
+; RUN: llc -march=amdgcn -mcpu=gfx1010 -mattr=-nsa-encoding -verify-machineinstrs < %s | FileCheck -check-prefixes=GCN,NONSA,GFX10-NONSA %s
+; RUN: llc -march=amdgcn -mcpu=gfx1010 -verify-machineinstrs < %s | FileCheck -check-prefixes=GCN,NSA,GFX1010-NSA %s
+; RUN: llc -march=amdgcn -mcpu=gfx1030 -verify-machineinstrs < %s | FileCheck -check-prefixes=GCN,NSA,GFX1030-NSA %s
+; RUN: llc -march=amdgcn -mcpu=gfx1100 -mattr=-nsa-encoding -verify-machineinstrs < %s | FileCheck -check-prefixes=GCN,NONSA,GFX11-NONSA %s
+; RUN: llc -march=amdgcn -mcpu=gfx1100 -verify-machineinstrs < %s | FileCheck -check-prefixes=GCN,NSA,GFX11-NSA %s
 
 ; GCN-LABEL: {{^}}sample_2d:
 ;
 ; TODO: use NSA here
 ; GCN: v_mov_b32_e32 v2, v0
 ;
 ; GCN: image_sample v[0:3], v[1:2],
 define amdgpu_ps <4 x float> @sample_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %t, float %s) {
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.2d.v4f32.f32(i32 15, float %s, float %t, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 ; GCN-LABEL: {{^}}sample_3d:
 ; NONSA: v_mov_b32_e32 v3, v0
 ; NONSA: image_sample v[0:3], v[1:3],
 ; NSA: image_sample v[0:3], [v1, v2, v0],
 define amdgpu_ps <4 x float> @sample_3d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %r, float %s, float %t) {
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.3d.v4f32.f32(i32 15, float %s, float %t, float %r, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 ; GCN-LABEL: {{^}}sample_d_3d:
-; GFX1010: image_sample_d v[0:3], v[7:22],
-; GFX1030: image_sample_d v[0:3], [v3, v8, v7, v5, v4, v6, v0, v2, v1],
+; GFX1010-NSA: image_sample_d v[0:3], v[7:22],
+; GFX1030-NSA: image_sample_d v[0:3], [v3, v8, v7, v5, v4, v6, v0, v2, v1],
+; GFX11-NSA: image_sample_d v[0:3], v[7:22],
 define amdgpu_ps <4 x float> @sample_d_3d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %s, float %r, float %t, float %dsdh, float %dtdv, float %dsdv, float %drdv, float %drdh, float %dtdh) {
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.d.3d.v4f32.f32(i32 15, float %dsdh, float %dtdh, float %drdh, float %dsdv, float %dtdv, float %drdv, float %s, float %t, float %r, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 ; GCN-LABEL: {{^}}sample_contig_nsa:
-; NONSA: image_sample_c_l v5, v[0:4],
-; NSA: image_sample_c_l v{{[0-9]+}}, v[0:4],
-; NSA: image_sample v{{[0-9]+}}, [v6, v7, v5],
+; GFX10-NONSA: image_sample_c_l v5, v[0:4],
+; GFX11-NONSA: image_sample_c_l v0, v[0:4],
+; GFX1010-NSA: image_sample_c_l v8, v[0:4],
+; GFX1010-NSA: image_sample v9, [v6, v7, v5],
+; GFX1030-NSA: image_sample_c_l v0, v[0:4],
+; GFX1030-NSA: image_sample v1, [v6, v7, v5],
+; GFX11-NSA: image_sample_c_l v0, v[0:4],
+; GFX11-NSA: image_sample v1, [v6, v7, v5],
 define amdgpu_ps <2 x float> @sample_contig_nsa(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, float %s1, float %t1, float %r1, float %lod, float %r2, float %s2, float %t2) {
 main_body:
   %v1 = call float @llvm.amdgcn.image.sample.c.l.3d.f32.f32(i32 1, float %zcompare, float %s1, float %t1, float %r1, float %lod, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   %v2 = call float @llvm.amdgcn.image.sample.3d.f32.f32(i32 1, float %s2, float %t2, float %r2, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   %r.0 = insertelement <2 x float> undef, float %v1, i32 0
   %r = insertelement <2 x float> %r.0, float %v2, i32 1
   ret <2 x float> %r
 }
 
 ; GCN-LABEL: {{^}}sample_nsa_nsa:
-; NSA: image_sample_c_l v{{[0-9]+}}, [v1, v2, v3, v4, v0],
-; NSA: image_sample v{{[0-9]+}}, [v6, v7, v5],
+; GFX1010-NSA: image_sample_c_l v8, [v1, v2, v3, v4, v0],
+; GFX1010-NSA: image_sample v9, [v6, v7, v5],
+; GFX1030-NSA: image_sample_c_l v0, [v1, v2, v3, v4, v0],
+; GFX1030-NSA: image_sample v1, [v6, v7, v5],
+; GFX11-NSA: image_sample_c_l v0, [v1, v2, v3, v4, v0],
+; GFX11-NSA: image_sample v1, [v6, v7, v5],
 define amdgpu_ps <2 x float> @sample_nsa_nsa(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %lod, float %zcompare, float %s1, float %t1, float %r1, float %r2, float %s2, float %t2) {
 main_body:
   %v1 = call float @llvm.amdgcn.image.sample.c.l.3d.f32.f32(i32 1, float %zcompare, float %s1, float %t1, float %r1, float %lod, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   %v2 = call float @llvm.amdgcn.image.sample.3d.f32.f32(i32 1, float %s2, float %t2, float %r2, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   %r.0 = insertelement <2 x float> undef, float %v1, i32 0
   %r = insertelement <2 x float> %r.0, float %v2, i32 1
   ret <2 x float> %r
 }
 
 ; GCN-LABEL: {{^}}sample_nsa_contig:
-; NSA: image_sample_c_l v{{[0-9]+}}, [v1, v2, v3, v4, v0],
-; NSA: image_sample v{{[0-9]+}}, v[5:7],
+; GFX1010-NSA: image_sample_c_l v8, [v1, v2, v3, v4, v0],
+; GFX1010-NSA: image_sample v9, v[5:7],
+; GFX1030-NSA: image_sample_c_l v0, [v1, v2, v3, v4, v0],
+; GFX1030-NSA: image_sample v1, v[5:7],
+; GFX11-NSA: image_sample_c_l v0, [v1, v2, v3, v4, v0],
+; GFX11-NSA: image_sample v1, v[5:7],
 define amdgpu_ps <2 x float> @sample_nsa_contig(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %lod, float %zcompare, float %s1, float %t1, float %r1, float %s2, float %t2, float %r2) {
 main_body:
   %v1 = call float @llvm.amdgcn.image.sample.c.l.3d.f32.f32(i32 1, float %zcompare, float %s1, float %t1, float %r1, float %lod, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   %v2 = call float @llvm.amdgcn.image.sample.3d.f32.f32(i32 1, float %s2, float %t2, float %r2, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   %r.0 = insertelement <2 x float> undef, float %v1, i32 0
   %r = insertelement <2 x float> %r.0, float %v2, i32 1
   ret <2 x float> %r
 }
 
 ; GCN-LABEL: {{^}}sample_contig_contig:
-; NSA: image_sample_c_l v{{[0-9]+}}, v[0:4],
-; NSA: image_sample v{{[0-9]+}}, v[5:7],
-; NONSA: image_sample_c_l v{{[0-9]+}}, v[0:4],
-; NONSA: image_sample v{{[0-9]+}}, v[5:7],
+; GFX1010-NSA: image_sample_c_l v8, v[0:4],
+; GFX1010-NSA: image_sample v9, v[5:7],
+; GFX1030-NSA: image_sample_c_l v0, v[0:4],
+; GFX1030-NSA: image_sample v1, v[5:7],
+; GFX11-NSA: image_sample_c_l v0, v[0:4],
+; GFX11-NSA: image_sample v1, v[5:7],
+; GFX10-NONSA: image_sample_c_l v8, v[0:4],
+; GFX10-NONSA: image_sample v9, v[5:7],
+; GFX11-NONSA: image_sample_c_l v0, v[0:4],
+; GFX11-NONSA: image_sample v1, v[5:7],
 define amdgpu_ps <2 x float> @sample_contig_contig(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, float %s1, float %t1, float %r1, float %lod, float %s2, float %t2, float %r2) {
 main_body:
   %v1 = call float @llvm.amdgcn.image.sample.c.l.3d.f32.f32(i32 1, float %zcompare, float %s1, float %t1, float %r1, float %lod, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   %v2 = call float @llvm.amdgcn.image.sample.3d.f32.f32(i32 1, float %s2, float %t2, float %r2, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   %r.0 = insertelement <2 x float> undef, float %v1, i32 0
   %r = insertelement <2 x float> %r.0, float %v2, i32 1
   ret <2 x float> %r
 }

llvm/test/CodeGen/AMDGPU/llvm.amdgcn.image.sample.a16.dim.ll

 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
 ; RUN: llc -march=amdgcn -mcpu=gfx900 -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX9 %s
 ; RUN: llc -march=amdgcn -mcpu=gfx1010 -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX10 %s
+; RUN: llc -march=amdgcn -mcpu=gfx1100 -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX10 %s
 
 define amdgpu_ps <4 x float> @sample_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, half %s) {
 ; GFX9-LABEL: sample_1d:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    s_mov_b64 s[12:13], exec
 ; GFX9-NEXT:    s_wqm_b64 exec, exec
 ; GFX9-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; GFX9-NEXT:    image_sample v[0:3], v0, s[0:7], s[8:11] dmask:0xf a16

 ; GFX10-NEXT:    image_sample_c_d_cl_g16 v[0:3], [v0, v1, v3, v5, v7], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D a16
 ; GFX10-NEXT:    s_waitcnt vmcnt(0)
 ; GFX10-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.c.d.cl.2d.v4f32.f32.f16(i32 15, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, half %s, half %t, half %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v
 }
 
-define amdgpu_ps <4 x float> @sample_cd_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, half %dsdh, half %dsdv, half %s) {
-; GFX9-LABEL: sample_cd_1d:
-; GFX9:       ; %bb.0: ; %main_body
-; GFX9-NEXT:    image_sample_cd v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf a16
-; GFX9-NEXT:    s_waitcnt vmcnt(0)
-; GFX9-NEXT:    ; return to shader part epilog
-;
-; GFX10-LABEL: sample_cd_1d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_sample_cd_g16 v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D a16
-; GFX10-NEXT:    s_waitcnt vmcnt(0)
-; GFX10-NEXT:    ; return to shader part epilog
-main_body:
-  %v = call <4 x float> @llvm.amdgcn.image.sample.cd.1d.v4f32.f16.f16(i32 15, half %dsdh, half %dsdv, half %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
-  ret <4 x float> %v
-}
-
-define amdgpu_ps <4 x float> @sample_cd_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, half %dsdh, half %dtdh, half %dsdv, half %dtdv, half %s, half %t) {
-; GFX9-LABEL: sample_cd_2d:
-; GFX9:       ; %bb.0: ; %main_body
-; GFX9-NEXT:    v_and_b32_e32 v4, 0xffff, v4
-; GFX9-NEXT:    v_and_b32_e32 v2, 0xffff, v2
-; GFX9-NEXT:    v_and_b32_e32 v0, 0xffff, v0
-; GFX9-NEXT:    v_lshl_or_b32 v4, v5, 16, v4
-; GFX9-NEXT:    v_lshl_or_b32 v3, v3, 16, v2
-; GFX9-NEXT:    v_lshl_or_b32 v2, v1, 16, v0
-; GFX9-NEXT:    image_sample_cd v[0:3], v[2:4], s[0:7], s[8:11] dmask:0xf a16
-; GFX9-NEXT:    s_waitcnt vmcnt(0)
-; GFX9-NEXT:    ; return to shader part epilog
-;
-; GFX10-LABEL: sample_cd_2d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    v_and_b32_e32 v4, 0xffff, v4
-; GFX10-NEXT:    v_and_b32_e32 v2, 0xffff, v2
-; GFX10-NEXT:    v_and_b32_e32 v0, 0xffff, v0
-; GFX10-NEXT:    v_lshl_or_b32 v4, v5, 16, v4
-; GFX10-NEXT:    v_lshl_or_b32 v3, v3, 16, v2
-; GFX10-NEXT:    v_lshl_or_b32 v2, v1, 16, v0
-; GFX10-NEXT:    image_sample_cd_g16 v[0:3], v[2:4], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D a16
-; GFX10-NEXT:    s_waitcnt vmcnt(0)
-; GFX10-NEXT:    ; return to shader part epilog
-main_body:
-  %v = call <4 x float> @llvm.amdgcn.image.sample.cd.2d.v4f32.f16.f16(i32 15, half %dsdh, half %dtdh, half %dsdv, half %dtdv, half %s, half %t, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
-  ret <4 x float> %v
-}
-
-define amdgpu_ps <4 x float> @sample_c_cd_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, half %dsdh, half %dsdv, half %s) {
-; GFX9-LABEL: sample_c_cd_1d:
-; GFX9:       ; %bb.0: ; %main_body
-; GFX9-NEXT:    image_sample_c_cd v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf a16
-; GFX9-NEXT:    s_waitcnt vmcnt(0)
-; GFX9-NEXT:    ; return to shader part epilog
-;
-; GFX10-LABEL: sample_c_cd_1d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_sample_c_cd_g16 v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D a16
-; GFX10-NEXT:    s_waitcnt vmcnt(0)
-; GFX10-NEXT:    ; return to shader part epilog
-main_body:
-  %v = call <4 x float> @llvm.amdgcn.image.sample.c.cd.1d.v4f32.f32.f16(i32 15, float %zcompare, half %dsdh, half %dsdv, half %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
-  ret <4 x float> %v
-}
-
-define amdgpu_ps <4 x float> @sample_c_cd_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, half %s, half %t) {
-; GFX9-LABEL: sample_c_cd_2d:
-; GFX9:       ; %bb.0: ; %main_body
-; GFX9-NEXT:    v_mov_b32_e32 v7, v3
-; GFX9-NEXT:    v_mov_b32_e32 v8, v2
-; GFX9-NEXT:    v_and_b32_e32 v2, 0xffff, v5
-; GFX9-NEXT:    v_lshl_or_b32 v3, v6, 16, v2
-; GFX9-NEXT:    v_and_b32_e32 v2, 0xffff, v7
-; GFX9-NEXT:    v_and_b32_e32 v1, 0xffff, v1
-; GFX9-NEXT:    v_lshl_or_b32 v2, v4, 16, v2
-; GFX9-NEXT:    v_lshl_or_b32 v1, v8, 16, v1
-; GFX9-NEXT:    image_sample_c_cd v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf a16
-; GFX9-NEXT:    s_waitcnt vmcnt(0)
-; GFX9-NEXT:    ; return to shader part epilog
-;
-; GFX10-LABEL: sample_c_cd_2d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    v_and_b32_e32 v5, 0xffff, v5
-; GFX10-NEXT:    v_and_b32_e32 v3, 0xffff, v3
-; GFX10-NEXT:    v_and_b32_e32 v1, 0xffff, v1
-; GFX10-NEXT:    v_lshl_or_b32 v5, v6, 16, v5
-; GFX10-NEXT:    v_lshl_or_b32 v3, v4, 16, v3
-; GFX10-NEXT:    v_lshl_or_b32 v1, v2, 16, v1
-; GFX10-NEXT:    image_sample_c_cd_g16 v[0:3], [v0, v1, v3, v5], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D a16
-; GFX10-NEXT:    s_waitcnt vmcnt(0)
-; GFX10-NEXT:    ; return to shader part epilog
-main_body:
-  %v = call <4 x float> @llvm.amdgcn.image.sample.c.cd.2d.v4f32.f32.f16(i32 15, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, half %s, half %t, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
-  ret <4 x float> %v
-}
-
-define amdgpu_ps <4 x float> @sample_cd_cl_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, half %dsdh, half %dsdv, half %s, half %clamp) {
-; GFX9-LABEL: sample_cd_cl_1d:
-; GFX9:       ; %bb.0: ; %main_body
-; GFX9-NEXT:    v_and_b32_e32 v2, 0xffff, v2
-; GFX9-NEXT:    v_lshl_or_b32 v2, v3, 16, v2
-; GFX9-NEXT:    image_sample_cd_cl v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf a16
-; GFX9-NEXT:    s_waitcnt vmcnt(0)
-; GFX9-NEXT:    ; return to shader part epilog
-;
-; GFX10-LABEL: sample_cd_cl_1d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    v_and_b32_e32 v2, 0xffff, v2
-; GFX10-NEXT:    v_lshl_or_b32 v2, v3, 16, v2
-; GFX10-NEXT:    image_sample_cd_cl_g16 v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D a16
-; GFX10-NEXT:    s_waitcnt vmcnt(0)
-; GFX10-NEXT:    ; return to shader part epilog
-main_body:
-  %v = call <4 x float> @llvm.amdgcn.image.sample.cd.cl.1d.v4f32.f16.f16(i32 15, half %dsdh, half %dsdv, half %s, half %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
-  ret <4 x float> %v
-}
-
-define amdgpu_ps <4 x float> @sample_cd_cl_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, half %dsdh, half %dtdh, half %dsdv, half %dtdv, half %s, half %t, half %clamp) {
-; GFX9-LABEL: sample_cd_cl_2d:
-; GFX9:       ; %bb.0: ; %main_body
-; GFX9-NEXT:    v_and_b32_e32 v4, 0xffff, v4
-; GFX9-NEXT:    v_and_b32_e32 v2, 0xffff, v2
-; GFX9-NEXT:    v_and_b32_e32 v0, 0xffff, v0
-; GFX9-NEXT:    v_lshl_or_b32 v5, v5, 16, v4
-; GFX9-NEXT:    v_lshl_or_b32 v4, v3, 16, v2
-; GFX9-NEXT:    v_lshl_or_b32 v3, v1, 16, v0
-; GFX9-NEXT:    image_sample_cd_cl v[0:3], v[3:6], s[0:7], s[8:11] dmask:0xf a16
-; GFX9-NEXT:    s_waitcnt vmcnt(0)
-; GFX9-NEXT:    ; return to shader part epilog
-;
-; GFX10-LABEL: sample_cd_cl_2d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    v_and_b32_e32 v4, 0xffff, v4
-; GFX10-NEXT:    v_and_b32_e32 v2, 0xffff, v2
-; GFX10-NEXT:    v_and_b32_e32 v0, 0xffff, v0
-; GFX10-NEXT:    v_lshl_or_b32 v4, v5, 16, v4
-; GFX10-NEXT:    v_lshl_or_b32 v2, v3, 16, v2
-; GFX10-NEXT:    v_lshl_or_b32 v0, v1, 16, v0
-; GFX10-NEXT:    image_sample_cd_cl_g16 v[0:3], [v0, v2, v4, v6], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D a16
-; GFX10-NEXT:    s_waitcnt vmcnt(0)
-; GFX10-NEXT:    ; return to shader part epilog
-main_body:
-  %v = call <4 x float> @llvm.amdgcn.image.sample.cd.cl.2d.v4f32.f16.f16(i32 15, half %dsdh, half %dtdh, half %dsdv, half %dtdv, half %s, half %t, half %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
-  ret <4 x float> %v
-}
-
-define amdgpu_ps <4 x float> @sample_c_cd_cl_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, half %dsdh, half %dsdv, half %s, half %clamp) {
-; GFX9-LABEL: sample_c_cd_cl_1d:
-; GFX9:       ; %bb.0: ; %main_body
-; GFX9-NEXT:    v_and_b32_e32 v3, 0xffff, v3
-; GFX9-NEXT:    v_lshl_or_b32 v3, v4, 16, v3
-; GFX9-NEXT:    image_sample_c_cd_cl v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf a16
-; GFX9-NEXT:    s_waitcnt vmcnt(0)
-; GFX9-NEXT:    ; return to shader part epilog
-;
-; GFX10-LABEL: sample_c_cd_cl_1d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    v_and_b32_e32 v3, 0xffff, v3
-; GFX10-NEXT:    v_lshl_or_b32 v3, v4, 16, v3
-; GFX10-NEXT:    image_sample_c_cd_cl_g16 v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D a16
-; GFX10-NEXT:    s_waitcnt vmcnt(0)
-; GFX10-NEXT:    ; return to shader part epilog
-main_body:
-  %v = call <4 x float> @llvm.amdgcn.image.sample.c.cd.cl.1d.v4f32.f32.f16(i32 15, float %zcompare, half %dsdh, half %dsdv, half %s, half %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
-  ret <4 x float> %v
-}
-
-define amdgpu_ps <4 x float> @sample_c_cd_cl_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, half %s, half %t, half %clamp) {
-; GFX9-LABEL: sample_c_cd_cl_2d:
-; GFX9:       ; %bb.0: ; %main_body
-; GFX9-NEXT:    v_mov_b32_e32 v11, v7
-; GFX9-NEXT:    v_mov_b32_e32 v7, v0
-; GFX9-NEXT:    v_and_b32_e32 v0, 0xffff, v5
-; GFX9-NEXT:    v_lshl_or_b32 v10, v6, 16, v0
-; GFX9-NEXT:    v_and_b32_e32 v0, 0xffff, v3
-; GFX9-NEXT:    v_lshl_or_b32 v9, v4, 16, v0
-; GFX9-NEXT:    v_and_b32_e32 v0, 0xffff, v1
-; GFX9-NEXT:    v_lshl_or_b32 v8, v2, 16, v0
-; GFX9-NEXT:    image_sample_c_cd_cl v[0:3], v[7:11], s[0:7], s[8:11] dmask:0xf a16
-; GFX9-NEXT:    s_waitcnt vmcnt(0)
-; GFX9-NEXT:    ; return to shader part epilog
-;
-; GFX10-LABEL: sample_c_cd_cl_2d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    v_and_b32_e32 v5, 0xffff, v5
-; GFX10-NEXT:    v_and_b32_e32 v3, 0xffff, v3
-; GFX10-NEXT:    v_and_b32_e32 v1, 0xffff, v1
-; GFX10-NEXT:    v_lshl_or_b32 v5, v6, 16, v5
-; GFX10-NEXT:    v_lshl_or_b32 v3, v4, 16, v3
-; GFX10-NEXT:    v_lshl_or_b32 v1, v2, 16, v1
-; GFX10-NEXT:    image_sample_c_cd_cl_g16 v[0:3], [v0, v1, v3, v5, v7], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D a16
-; GFX10-NEXT:    s_waitcnt vmcnt(0)
-; GFX10-NEXT:    ; return to shader part epilog
-main_body:
-  %v = call <4 x float> @llvm.amdgcn.image.sample.c.cd.cl.2d.v4f32.f32.f16(i32 15, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, half %s, half %t, half %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
-  ret <4 x float> %v
-}
-
 define amdgpu_ps <4 x float> @sample_l_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, half %s, half %lod) {
 ; GFX9-LABEL: sample_l_1d:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    v_and_b32_e32 v0, 0xffff, v0
 ; GFX9-NEXT:    v_lshl_or_b32 v0, v1, 16, v0
 ; GFX9-NEXT:    image_sample_l v[0:3], v0, s[0:7], s[8:11] dmask:0xf a16
 ; GFX9-NEXT:    s_waitcnt vmcnt(0)
 ; GFX9-NEXT:    ; return to shader part epilog

 declare <4 x float> @llvm.amdgcn.image.sample.d.3d.v4f32.f16.f16(i32, half, half, half, half, half, half, half, half, half, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 declare <4 x float> @llvm.amdgcn.image.sample.c.d.1d.v4f32.f32.f16(i32, float, half, half, half, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 declare <4 x float> @llvm.amdgcn.image.sample.c.d.2d.v4f32.f32.f16(i32, float, half, half, half, half, half, half, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 declare <4 x float> @llvm.amdgcn.image.sample.d.cl.1d.v4f32.f16.f16(i32, half, half, half, half, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 declare <4 x float> @llvm.amdgcn.image.sample.d.cl.2d.v4f32.f16.f16(i32, half, half, half, half, half, half, half, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 declare <4 x float> @llvm.amdgcn.image.sample.c.d.cl.1d.v4f32.f32.f16(i32, float, half, half, half, half, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 declare <4 x float> @llvm.amdgcn.image.sample.c.d.cl.2d.v4f32.f32.f16(i32, float, half, half, half, half, half, half, half, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 
-declare <4 x float> @llvm.amdgcn.image.sample.cd.1d.v4f32.f16.f16(i32, half, half, half, <8 x i32>, <4 x i32>, i1, i32, i32) #1
-declare <4 x float> @llvm.amdgcn.image.sample.cd.2d.v4f32.f16.f16(i32, half, half, half, half, half, half, <8 x i32>, <4 x i32>, i1, i32, i32) #1
-declare <4 x float> @llvm.amdgcn.image.sample.c.cd.1d.v4f32.f32.f16(i32, float, half, half, half, <8 x i32>, <4 x i32>, i1, i32, i32) #1
-declare <4 x float> @llvm.amdgcn.image.sample.c.cd.2d.v4f32.f32.f16(i32, float, half, half, half, half, half, half, <8 x i32>, <4 x i32>, i1, i32, i32) #1
-declare <4 x float> @llvm.amdgcn.image.sample.cd.cl.1d.v4f32.f16.f16(i32, half, half, half, half, <8 x i32>, <4 x i32>, i1, i32, i32) #1
-declare <4 x float> @llvm.amdgcn.image.sample.cd.cl.2d.v4f32.f16.f16(i32, half, half, half, half, half, half, half, <8 x i32>, <4 x i32>, i1, i32, i32) #1
-declare <4 x float> @llvm.amdgcn.image.sample.c.cd.cl.1d.v4f32.f32.f16(i32, float, half, half, half, half, <8 x i32>, <4 x i32>, i1, i32, i32) #1
-declare <4 x float> @llvm.amdgcn.image.sample.c.cd.cl.2d.v4f32.f32.f16(i32, float, half, half, half, half, half, half, half, <8 x i32>, <4 x i32>, i1, i32, i32) #1
-
 declare <4 x float> @llvm.amdgcn.image.sample.l.1d.v4f32.f16(i32, half, half, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 declare <4 x float> @llvm.amdgcn.image.sample.l.2d.v4f32.f16(i32, half, half, half, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 declare <4 x float> @llvm.amdgcn.image.sample.c.l.1d.v4f32.f16(i32, float, half, half, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 declare <4 x float> @llvm.amdgcn.image.sample.c.l.2d.v4f32.f16(i32, float, half, half, half, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 
 declare <4 x float> @llvm.amdgcn.image.sample.lz.1d.v4f32.f16(i32, half, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 declare <4 x float> @llvm.amdgcn.image.sample.lz.2d.v4f32.f16(i32, half, half, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 declare <4 x float> @llvm.amdgcn.image.sample.c.lz.1d.v4f32.f16(i32, float, half, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 declare <4 x float> @llvm.amdgcn.image.sample.c.lz.2d.v4f32.f16(i32, float, half, half, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 
 declare float @llvm.amdgcn.image.sample.c.d.o.2darray.f32.f16.f16(i32, i32, float, half, half, half, half, half, half, half, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 declare <2 x float> @llvm.amdgcn.image.sample.c.d.o.2darray.v2f32.f32.f16(i32, i32, float, half, half, half, half, half, half, half, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 
 attributes #0 = { nounwind }
 attributes #1 = { nounwind readonly }
 attributes #2 = { nounwind readnone }

llvm/test/CodeGen/AMDGPU/llvm.amdgcn.image.sample.cd.a16.dim.ll

+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
+; RUN: llc -march=amdgcn -mcpu=gfx900 -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX9 %s
+; RUN: llc -march=amdgcn -mcpu=gfx1010 -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX10 %s
+
+define amdgpu_ps <4 x float> @sample_cd_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, half %dsdh, half %dsdv, half %s) {
+; GFX9-LABEL: sample_cd_1d:
+; GFX9:       ; %bb.0: ; %main_body
+; GFX9-NEXT:    image_sample_cd v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf a16
+; GFX9-NEXT:    s_waitcnt vmcnt(0)
+; GFX9-NEXT:    ; return to shader part epilog
+;
+; GFX10-LABEL: sample_cd_1d:
+; GFX10:       ; %bb.0: ; %main_body
+; GFX10-NEXT:    image_sample_cd_g16 v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D a16
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    ; return to shader part epilog
+main_body:
+  %v = call <4 x float> @llvm.amdgcn.image.sample.cd.1d.v4f32.f16.f16(i32 15, half %dsdh, half %dsdv, half %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
+  ret <4 x float> %v
+}
+
+define amdgpu_ps <4 x float> @sample_cd_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, half %dsdh, half %dtdh, half %dsdv, half %dtdv, half %s, half %t) {
+; GFX9-LABEL: sample_cd_2d:
+; GFX9:       ; %bb.0: ; %main_body
+; GFX9-NEXT:    v_and_b32_e32 v4, 0xffff, v4
+; GFX9-NEXT:    v_and_b32_e32 v2, 0xffff, v2
+; GFX9-NEXT:    v_and_b32_e32 v0, 0xffff, v0
+; GFX9-NEXT:    v_lshl_or_b32 v4, v5, 16, v4
+; GFX9-NEXT:    v_lshl_or_b32 v3, v3, 16, v2
+; GFX9-NEXT:    v_lshl_or_b32 v2, v1, 16, v0
+; GFX9-NEXT:    image_sample_cd v[0:3], v[2:4], s[0:7], s[8:11] dmask:0xf a16
+; GFX9-NEXT:    s_waitcnt vmcnt(0)
+; GFX9-NEXT:    ; return to shader part epilog
+;
+; GFX10-LABEL: sample_cd_2d:
+; GFX10:       ; %bb.0: ; %main_body
+; GFX10-NEXT:    v_and_b32_e32 v4, 0xffff, v4
+; GFX10-NEXT:    v_and_b32_e32 v2, 0xffff, v2
+; GFX10-NEXT:    v_and_b32_e32 v0, 0xffff, v0
+; GFX10-NEXT:    v_lshl_or_b32 v4, v5, 16, v4
+; GFX10-NEXT:    v_lshl_or_b32 v3, v3, 16, v2
+; GFX10-NEXT:    v_lshl_or_b32 v2, v1, 16, v0
+; GFX10-NEXT:    image_sample_cd_g16 v[0:3], v[2:4], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D a16
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    ; return to shader part epilog
+main_body:
+  %v = call <4 x float> @llvm.amdgcn.image.sample.cd.2d.v4f32.f16.f16(i32 15, half %dsdh, half %dtdh, half %dsdv, half %dtdv, half %s, half %t, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
+  ret <4 x float> %v
+}
+
+define amdgpu_ps <4 x float> @sample_c_cd_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, half %dsdh, half %dsdv, half %s) {
+; GFX9-LABEL: sample_c_cd_1d:
+; GFX9:       ; %bb.0: ; %main_body
+; GFX9-NEXT:    image_sample_c_cd v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf a16
+; GFX9-NEXT:    s_waitcnt vmcnt(0)
+; GFX9-NEXT:    ; return to shader part epilog
+;
+; GFX10-LABEL: sample_c_cd_1d:
+; GFX10:       ; %bb.0: ; %main_body
+; GFX10-NEXT:    image_sample_c_cd_g16 v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D a16
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    ; return to shader part epilog
+main_body:
+  %v = call <4 x float> @llvm.amdgcn.image.sample.c.cd.1d.v4f32.f32.f16(i32 15, float %zcompare, half %dsdh, half %dsdv, half %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
+  ret <4 x float> %v
+}
+
+define amdgpu_ps <4 x float> @sample_c_cd_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, half %s, half %t) {
+; GFX9-LABEL: sample_c_cd_2d:
+; GFX9:       ; %bb.0: ; %main_body
+; GFX9-NEXT:    v_mov_b32_e32 v7, v3
+; GFX9-NEXT:    v_mov_b32_e32 v8, v2
+; GFX9-NEXT:    v_and_b32_e32 v2, 0xffff, v5
+; GFX9-NEXT:    v_lshl_or_b32 v3, v6, 16, v2
+; GFX9-NEXT:    v_and_b32_e32 v2, 0xffff, v7
+; GFX9-NEXT:    v_and_b32_e32 v1, 0xffff, v1
+; GFX9-NEXT:    v_lshl_or_b32 v2, v4, 16, v2
+; GFX9-NEXT:    v_lshl_or_b32 v1, v8, 16, v1
+; GFX9-NEXT:    image_sample_c_cd v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf a16
+; GFX9-NEXT:    s_waitcnt vmcnt(0)
+; GFX9-NEXT:    ; return to shader part epilog
+;
+; GFX10-LABEL: sample_c_cd_2d:
+; GFX10:       ; %bb.0: ; %main_body
+; GFX10-NEXT:    v_and_b32_e32 v5, 0xffff, v5
+; GFX10-NEXT:    v_and_b32_e32 v3, 0xffff, v3
+; GFX10-NEXT:    v_and_b32_e32 v1, 0xffff, v1
+; GFX10-NEXT:    v_lshl_or_b32 v5, v6, 16, v5
+; GFX10-NEXT:    v_lshl_or_b32 v3, v4, 16, v3
+; GFX10-NEXT:    v_lshl_or_b32 v1, v2, 16, v1
+; GFX10-NEXT:    image_sample_c_cd_g16 v[0:3], [v0, v1, v3, v5], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D a16
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    ; return to shader part epilog
+main_body:
+  %v = call <4 x float> @llvm.amdgcn.image.sample.c.cd.2d.v4f32.f32.f16(i32 15, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, half %s, half %t, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
+  ret <4 x float> %v
+}
+
+define amdgpu_ps <4 x float> @sample_cd_cl_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, half %dsdh, half %dsdv, half %s, half %clamp) {
+; GFX9-LABEL: sample_cd_cl_1d:
+; GFX9:       ; %bb.0: ; %main_body
+; GFX9-NEXT:    v_and_b32_e32 v2, 0xffff, v2
+; GFX9-NEXT:    v_lshl_or_b32 v2, v3, 16, v2
+; GFX9-NEXT:    image_sample_cd_cl v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf a16
+; GFX9-NEXT:    s_waitcnt vmcnt(0)
+; GFX9-NEXT:    ; return to shader part epilog
+;
+; GFX10-LABEL: sample_cd_cl_1d:
+; GFX10:       ; %bb.0: ; %main_body
+; GFX10-NEXT:    v_and_b32_e32 v2, 0xffff, v2
+; GFX10-NEXT:    v_lshl_or_b32 v2, v3, 16, v2
+; GFX10-NEXT:    image_sample_cd_cl_g16 v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D a16
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    ; return to shader part epilog
+main_body:
+  %v = call <4 x float> @llvm.amdgcn.image.sample.cd.cl.1d.v4f32.f16.f16(i32 15, half %dsdh, half %dsdv, half %s, half %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
+  ret <4 x float> %v
+}
+
+define amdgpu_ps <4 x float> @sample_cd_cl_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, half %dsdh, half %dtdh, half %dsdv, half %dtdv, half %s, half %t, half %clamp) {
+; GFX9-LABEL: sample_cd_cl_2d:
+; GFX9:       ; %bb.0: ; %main_body
+; GFX9-NEXT:    v_and_b32_e32 v4, 0xffff, v4
+; GFX9-NEXT:    v_and_b32_e32 v2, 0xffff, v2
+; GFX9-NEXT:    v_and_b32_e32 v0, 0xffff, v0
+; GFX9-NEXT:    v_lshl_or_b32 v5, v5, 16, v4
+; GFX9-NEXT:    v_lshl_or_b32 v4, v3, 16, v2
+; GFX9-NEXT:    v_lshl_or_b32 v3, v1, 16, v0
+; GFX9-NEXT:    image_sample_cd_cl v[0:3], v[3:6], s[0:7], s[8:11] dmask:0xf a16
+; GFX9-NEXT:    s_waitcnt vmcnt(0)
+; GFX9-NEXT:    ; return to shader part epilog
+;
+; GFX10-LABEL: sample_cd_cl_2d:
+; GFX10:       ; %bb.0: ; %main_body
+; GFX10-NEXT:    v_and_b32_e32 v4, 0xffff, v4
+; GFX10-NEXT:    v_and_b32_e32 v2, 0xffff, v2
+; GFX10-NEXT:    v_and_b32_e32 v0, 0xffff, v0
+; GFX10-NEXT:    v_lshl_or_b32 v4, v5, 16, v4
+; GFX10-NEXT:    v_lshl_or_b32 v2, v3, 16, v2
+; GFX10-NEXT:    v_lshl_or_b32 v0, v1, 16, v0
+; GFX10-NEXT:    image_sample_cd_cl_g16 v[0:3], [v0, v2, v4, v6], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D a16
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    ; return to shader part epilog
+main_body:
+  %v = call <4 x float> @llvm.amdgcn.image.sample.cd.cl.2d.v4f32.f16.f16(i32 15, half %dsdh, half %dtdh, half %dsdv, half %dtdv, half %s, half %t, half %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
+  ret <4 x float> %v
+}
+
+define amdgpu_ps <4 x float> @sample_c_cd_cl_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, half %dsdh, half %dsdv, half %s, half %clamp) {
+; GFX9-LABEL: sample_c_cd_cl_1d:
+; GFX9:       ; %bb.0: ; %main_body
+; GFX9-NEXT:    v_and_b32_e32 v3, 0xffff, v3
+; GFX9-NEXT:    v_lshl_or_b32 v3, v4, 16, v3
+; GFX9-NEXT:    image_sample_c_cd_cl v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf a16
+; GFX9-NEXT:    s_waitcnt vmcnt(0)
+; GFX9-NEXT:    ; return to shader part epilog
+;
+; GFX10-LABEL: sample_c_cd_cl_1d:
+; GFX10:       ; %bb.0: ; %main_body
+; GFX10-NEXT:    v_and_b32_e32 v3, 0xffff, v3
+; GFX10-NEXT:    v_lshl_or_b32 v3, v4, 16, v3
+; GFX10-NEXT:    image_sample_c_cd_cl_g16 v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D a16
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    ; return to shader part epilog
+main_body:
+  %v = call <4 x float> @llvm.amdgcn.image.sample.c.cd.cl.1d.v4f32.f32.f16(i32 15, float %zcompare, half %dsdh, half %dsdv, half %s, half %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
+  ret <4 x float> %v
+}
+
+define amdgpu_ps <4 x float> @sample_c_cd_cl_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, half %s, half %t, half %clamp) {
+; GFX9-LABEL: sample_c_cd_cl_2d:
+; GFX9:       ; %bb.0: ; %main_body
+; GFX9-NEXT:    v_mov_b32_e32 v11, v7
+; GFX9-NEXT:    v_mov_b32_e32 v7, v0
+; GFX9-NEXT:    v_and_b32_e32 v0, 0xffff, v5
+; GFX9-NEXT:    v_lshl_or_b32 v10, v6, 16, v0
+; GFX9-NEXT:    v_and_b32_e32 v0, 0xffff, v3
+; GFX9-NEXT:    v_lshl_or_b32 v9, v4, 16, v0
+; GFX9-NEXT:    v_and_b32_e32 v0, 0xffff, v1
+; GFX9-NEXT:    v_lshl_or_b32 v8, v2, 16, v0
+; GFX9-NEXT:    image_sample_c_cd_cl v[0:3], v[7:11], s[0:7], s[8:11] dmask:0xf a16
+; GFX9-NEXT:    s_waitcnt vmcnt(0)
+; GFX9-NEXT:    ; return to shader part epilog
+;
+; GFX10-LABEL: sample_c_cd_cl_2d:
+; GFX10:       ; %bb.0: ; %main_body
+; GFX10-NEXT:    v_and_b32_e32 v5, 0xffff, v5
+; GFX10-NEXT:    v_and_b32_e32 v3, 0xffff, v3
+; GFX10-NEXT:    v_and_b32_e32 v1, 0xffff, v1
+; GFX10-NEXT:    v_lshl_or_b32 v5, v6, 16, v5
+; GFX10-NEXT:    v_lshl_or_b32 v3, v4, 16, v3
+; GFX10-NEXT:    v_lshl_or_b32 v1, v2, 16, v1
+; GFX10-NEXT:    image_sample_c_cd_cl_g16 v[0:3], [v0, v1, v3, v5, v7], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D a16
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    ; return to shader part epilog
+main_body:
+  %v = call <4 x float> @llvm.amdgcn.image.sample.c.cd.cl.2d.v4f32.f32.f16(i32 15, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, half %s, half %t, half %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
+  ret <4 x float> %v
+}
+
+declare <4 x float> @llvm.amdgcn.image.sample.cd.1d.v4f32.f16.f16(i32, half, half, half, <8 x i32>, <4 x i32>, i1, i32, i32) #1
+declare <4 x float> @llvm.amdgcn.image.sample.cd.2d.v4f32.f16.f16(i32, half, half, half, half, half, half, <8 x i32>, <4 x i32>, i1, i32, i32) #1
+declare <4 x float> @llvm.amdgcn.image.sample.c.cd.1d.v4f32.f32.f16(i32, float, half, half, half, <8 x i32>, <4 x i32>, i1, i32, i32) #1
+declare <4 x float> @llvm.amdgcn.image.sample.c.cd.2d.v4f32.f32.f16(i32, float, half, half, half, half, half, half, <8 x i32>, <4 x i32>, i1, i32, i32) #1
+declare <4 x float> @llvm.amdgcn.image.sample.cd.cl.1d.v4f32.f16.f16(i32, half, half, half, half, <8 x i32>, <4 x i32>, i1, i32, i32) #1
+declare <4 x float> @llvm.amdgcn.image.sample.cd.cl.2d.v4f32.f16.f16(i32, half, half, half, half, half, half, half, <8 x i32>, <4 x i32>, i1, i32, i32) #1
+declare <4 x float> @llvm.amdgcn.image.sample.c.cd.cl.1d.v4f32.f32.f16(i32, float, half, half, half, half, <8 x i32>, <4 x i32>, i1, i32, i32) #1
+declare <4 x float> @llvm.amdgcn.image.sample.c.cd.cl.2d.v4f32.f32.f16(i32, float, half, half, half, half, half, half, half, <8 x i32>, <4 x i32>, i1, i32, i32) #1
+
+attributes #0 = { nounwind }
+attributes #1 = { nounwind readonly }
+attributes #2 = { nounwind readnone }

llvm/test/CodeGen/AMDGPU/llvm.amdgcn.image.sample.cd.dim.ll

+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
+; RUN: llc -march=amdgcn -mcpu=verde -verify-machineinstrs < %s | FileCheck -check-prefixes=VERDE %s
+; RUN: llc -march=amdgcn -mcpu=gfx900 -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX6789 %s
+; RUN: llc -march=amdgcn -mcpu=gfx1010 -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX10 %s
+
+define amdgpu_ps <4 x float> @sample_cd_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %dsdh, float %dsdv, float %s) {
+; VERDE-LABEL: sample_cd_1d:
+; VERDE:       ; %bb.0: ; %main_body
+; VERDE-NEXT:    image_sample_cd v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf
+; VERDE-NEXT:    s_waitcnt vmcnt(0)
+; VERDE-NEXT:    ; return to shader part epilog
+;
+; GFX6789-LABEL: sample_cd_1d:
+; GFX6789:       ; %bb.0: ; %main_body
+; GFX6789-NEXT:    image_sample_cd v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf
+; GFX6789-NEXT:    s_waitcnt vmcnt(0)
+; GFX6789-NEXT:    ; return to shader part epilog
+;
+; GFX10-LABEL: sample_cd_1d:
+; GFX10:       ; %bb.0: ; %main_body
+; GFX10-NEXT:    image_sample_cd v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    ; return to shader part epilog
+main_body:
+  %v = call <4 x float> @llvm.amdgcn.image.sample.cd.1d.v4f32.f32.f32(i32 15, float %dsdh, float %dsdv, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
+  ret <4 x float> %v
+}
+
+define amdgpu_ps <4 x float> @sample_cd_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %dsdh, float %dtdh, float %dsdv, float %dtdv, float %s, float %t) {
+; VERDE-LABEL: sample_cd_2d:
+; VERDE:       ; %bb.0: ; %main_body
+; VERDE-NEXT:    image_sample_cd v[0:3], v[0:5], s[0:7], s[8:11] dmask:0xf
+; VERDE-NEXT:    s_waitcnt vmcnt(0)
+; VERDE-NEXT:    ; return to shader part epilog
+;
+; GFX6789-LABEL: sample_cd_2d:
+; GFX6789:       ; %bb.0: ; %main_body
+; GFX6789-NEXT:    image_sample_cd v[0:3], v[0:5], s[0:7], s[8:11] dmask:0xf
+; GFX6789-NEXT:    s_waitcnt vmcnt(0)
+; GFX6789-NEXT:    ; return to shader part epilog
+;
+; GFX10-LABEL: sample_cd_2d:
+; GFX10:       ; %bb.0: ; %main_body
+; GFX10-NEXT:    image_sample_cd v[0:3], v[0:5], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    ; return to shader part epilog
+main_body:
+  %v = call <4 x float> @llvm.amdgcn.image.sample.cd.2d.v4f32.f32.f32(i32 15, float %dsdh, float %dtdh, float %dsdv, float %dtdv, float %s, float %t, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
+  ret <4 x float> %v
+}
+
+define amdgpu_ps <4 x float> @sample_c_cd_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, float %dsdh, float %dsdv, float %s) {
+; VERDE-LABEL: sample_c_cd_1d:
+; VERDE:       ; %bb.0: ; %main_body
+; VERDE-NEXT:    image_sample_c_cd v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf
+; VERDE-NEXT:    s_waitcnt vmcnt(0)
+; VERDE-NEXT:    ; return to shader part epilog
+;
+; GFX6789-LABEL: sample_c_cd_1d:
+; GFX6789:       ; %bb.0: ; %main_body
+; GFX6789-NEXT:    image_sample_c_cd v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf
+; GFX6789-NEXT:    s_waitcnt vmcnt(0)
+; GFX6789-NEXT:    ; return to shader part epilog
+;
+; GFX10-LABEL: sample_c_cd_1d:
+; GFX10:       ; %bb.0: ; %main_body
+; GFX10-NEXT:    image_sample_c_cd v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    ; return to shader part epilog
+main_body:
+  %v = call <4 x float> @llvm.amdgcn.image.sample.c.cd.1d.v4f32.f32.f32(i32 15, float %zcompare, float %dsdh, float %dsdv, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
+  ret <4 x float> %v
+}
+
+define amdgpu_ps <4 x float> @sample_c_cd_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, float %dsdh, float %dtdh, float %dsdv, float %dtdv, float %s, float %t) {
+; VERDE-LABEL: sample_c_cd_2d:
+; VERDE:       ; %bb.0: ; %main_body
+; VERDE-NEXT:    image_sample_c_cd v[0:3], v[0:6], s[0:7], s[8:11] dmask:0xf
+; VERDE-NEXT:    s_waitcnt vmcnt(0)
+; VERDE-NEXT:    ; return to shader part epilog
+;
+; GFX6789-LABEL: sample_c_cd_2d:
+; GFX6789:       ; %bb.0: ; %main_body
+; GFX6789-NEXT:    image_sample_c_cd v[0:3], v[0:6], s[0:7], s[8:11] dmask:0xf
+; GFX6789-NEXT:    s_waitcnt vmcnt(0)
+; GFX6789-NEXT:    ; return to shader part epilog
+;
+; GFX10-LABEL: sample_c_cd_2d:
+; GFX10:       ; %bb.0: ; %main_body
+; GFX10-NEXT:    image_sample_c_cd v[0:3], v[0:6], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    ; return to shader part epilog
+main_body:
+  %v = call <4 x float> @llvm.amdgcn.image.sample.c.cd.2d.v4f32.f32.f32(i32 15, float %zcompare, float %dsdh, float %dtdh, float %dsdv, float %dtdv, float %s, float %t, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
+  ret <4 x float> %v
+}
+
+define amdgpu_ps <4 x float> @sample_cd_cl_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %dsdh, float %dsdv, float %s, float %clamp) {
+; VERDE-LABEL: sample_cd_cl_1d:
+; VERDE:       ; %bb.0: ; %main_body
+; VERDE-NEXT:    image_sample_cd_cl v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf
+; VERDE-NEXT:    s_waitcnt vmcnt(0)
+; VERDE-NEXT:    ; return to shader part epilog
+;
+; GFX6789-LABEL: sample_cd_cl_1d:
+; GFX6789:       ; %bb.0: ; %main_body
+; GFX6789-NEXT:    image_sample_cd_cl v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf
+; GFX6789-NEXT:    s_waitcnt vmcnt(0)
+; GFX6789-NEXT:    ; return to shader part epilog
+;
+; GFX10-LABEL: sample_cd_cl_1d:
+; GFX10:       ; %bb.0: ; %main_body
+; GFX10-NEXT:    image_sample_cd_cl v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    ; return to shader part epilog
+main_body:
+  %v = call <4 x float> @llvm.amdgcn.image.sample.cd.cl.1d.v4f32.f32.f32(i32 15, float %dsdh, float %dsdv, float %s, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
+  ret <4 x float> %v
+}
+
+define amdgpu_ps <4 x float> @sample_cd_cl_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %dsdh, float %dtdh, float %dsdv, float %dtdv, float %s, float %t, float %clamp) {
+; VERDE-LABEL: sample_cd_cl_2d:
+; VERDE:       ; %bb.0: ; %main_body
+; VERDE-NEXT:    image_sample_cd_cl v[0:3], v[0:6], s[0:7], s[8:11] dmask:0xf
+; VERDE-NEXT:    s_waitcnt vmcnt(0)
+; VERDE-NEXT:    ; return to shader part epilog
+;
+; GFX6789-LABEL: sample_cd_cl_2d:
+; GFX6789:       ; %bb.0: ; %main_body
+; GFX6789-NEXT:    image_sample_cd_cl v[0:3], v[0:6], s[0:7], s[8:11] dmask:0xf
+; GFX6789-NEXT:    s_waitcnt vmcnt(0)
+; GFX6789-NEXT:    ; return to shader part epilog
+;
+; GFX10-LABEL: sample_cd_cl_2d:
+; GFX10:       ; %bb.0: ; %main_body
+; GFX10-NEXT:    image_sample_cd_cl v[0:3], v[0:6], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    ; return to shader part epilog
+main_body:
+  %v = call <4 x float> @llvm.amdgcn.image.sample.cd.cl.2d.v4f32.f32.f32(i32 15, float %dsdh, float %dtdh, float %dsdv, float %dtdv, float %s, float %t, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
+  ret <4 x float> %v
+}
+
+define amdgpu_ps <4 x float> @sample_c_cd_cl_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, float %dsdh, float %dsdv, float %s, float %clamp) {
+; VERDE-LABEL: sample_c_cd_cl_1d:
+; VERDE:       ; %bb.0: ; %main_body
+; VERDE-NEXT:    image_sample_c_cd_cl v[0:3], v[0:4], s[0:7], s[8:11] dmask:0xf
+; VERDE-NEXT:    s_waitcnt vmcnt(0)
+; VERDE-NEXT:    ; return to shader part epilog
+;
+; GFX6789-LABEL: sample_c_cd_cl_1d:
+; GFX6789:       ; %bb.0: ; %main_body
+; GFX6789-NEXT:    image_sample_c_cd_cl v[0:3], v[0:4], s[0:7], s[8:11] dmask:0xf
+; GFX6789-NEXT:    s_waitcnt vmcnt(0)
+; GFX6789-NEXT:    ; return to shader part epilog
+;
+; GFX10-LABEL: sample_c_cd_cl_1d:
+; GFX10:       ; %bb.0: ; %main_body
+; GFX10-NEXT:    image_sample_c_cd_cl v[0:3], v[0:4], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    ; return to shader part epilog
+main_body:
+  %v = call <4 x float> @llvm.amdgcn.image.sample.c.cd.cl.1d.v4f32.f32.f32(i32 15, float %zcompare, float %dsdh, float %dsdv, float %s, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
+  ret <4 x float> %v
+}
+
+declare <4 x float> @llvm.amdgcn.image.sample.cd.1d.v4f32.f32.f32(i32, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
+declare <4 x float> @llvm.amdgcn.image.sample.cd.2d.v4f32.f32.f32(i32, float, float, float, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
+declare <4 x float> @llvm.amdgcn.image.sample.c.cd.1d.v4f32.f32.f32(i32, float, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
+declare <4 x float> @llvm.amdgcn.image.sample.c.cd.2d.v4f32.f32.f32(i32, float, float, float, float, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
+declare <4 x float> @llvm.amdgcn.image.sample.cd.cl.1d.v4f32.f32.f32(i32, float, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
+declare <4 x float> @llvm.amdgcn.image.sample.cd.cl.2d.v4f32.f32.f32(i32, float, float, float, float, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
+declare <4 x float> @llvm.amdgcn.image.sample.c.cd.cl.1d.v4f32.f32.f32(i32, float, float, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
+declare <4 x float> @llvm.amdgcn.image.sample.c.cd.cl.2d.v4f32.f32.f32(i32, float, float, float, float, float, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
+
+attributes #0 = { nounwind }
+attributes #1 = { nounwind readonly }
+attributes #2 = { nounwind readnone }

llvm/test/CodeGen/AMDGPU/llvm.amdgcn.image.sample.cd.g16.encode.ll

+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
+; RUN: llc -march=amdgcn -mcpu=gfx1010 -show-mc-encoding < %s | FileCheck -check-prefix=GFX10 %s
+
+define amdgpu_ps <4 x float> @sample_cd_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, half %dsdh, half %dsdv, float %s) {
+; GFX10-LABEL: sample_cd_1d:
+; GFX10:       ; %bb.0: ; %main_body
+; GFX10-NEXT:    image_sample_cd_g16 v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D ; encoding: [0x01,0x0f,0xa0,0xf1,0x00,0x00,0x40,0x00]
+; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
+; GFX10-NEXT:    ; return to shader part epilog
+main_body:
+  %v = call <4 x float> @llvm.amdgcn.image.sample.cd.1d.v4f32.f16.f32(i32 15, half %dsdh, half %dsdv, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
+  ret <4 x float> %v
+}
+
+define amdgpu_ps <4 x float> @sample_cd_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t) {
+; GFX10-LABEL: sample_cd_2d:
+; GFX10:       ; %bb.0: ; %main_body
+; GFX10-NEXT:    v_and_b32_e32 v2, 0xffff, v2 ; encoding: [0xff,0x04,0x04,0x36,0xff,0xff,0x00,0x00]
+; GFX10-NEXT:    v_and_b32_e32 v0, 0xffff, v0 ; encoding: [0xff,0x00,0x00,0x36,0xff,0xff,0x00,0x00]
+; GFX10-NEXT:    v_lshl_or_b32 v2, v3, 16, v2 ; encoding: [0x02,0x00,0x6f,0xd7,0x03,0x21,0x09,0x04]
+; GFX10-NEXT:    v_lshl_or_b32 v0, v1, 16, v0 ; encoding: [0x00,0x00,0x6f,0xd7,0x01,0x21,0x01,0x04]
+; GFX10-NEXT:    image_sample_cd_g16 v[0:3], [v0, v2, v4, v5], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D ; encoding: [0x0b,0x0f,0xa0,0xf1,0x00,0x00,0x40,0x00,0x02,0x04,0x05,0x00]
+; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
+; GFX10-NEXT:    ; return to shader part epilog
+main_body:
+  %v = call <4 x float> @llvm.amdgcn.image.sample.cd.2d.v4f32.f16.f32(i32 15, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
+  ret <4 x float> %v
+}
+
+define amdgpu_ps <4 x float> @sample_c_cd_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, half %dsdh, half %dsdv, float %s) {
+; GFX10-LABEL: sample_c_cd_1d:
+; GFX10:       ; %bb.0: ; %main_body
+; GFX10-NEXT:    image_sample_c_cd_g16 v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D ; encoding: [0x01,0x0f,0xa8,0xf1,0x00,0x00,0x40,0x00]
+; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
+; GFX10-NEXT:    ; return to shader part epilog
+main_body:
+  %v = call <4 x float> @llvm.amdgcn.image.sample.c.cd.1d.v4f32.f16.f32(i32 15, float %zcompare, half %dsdh, half %dsdv, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
+  ret <4 x float> %v
+}
+
+define amdgpu_ps <4 x float> @sample_c_cd_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t) {
+; GFX10-LABEL: sample_c_cd_2d:
+; GFX10:       ; %bb.0: ; %main_body
+; GFX10-NEXT:    v_and_b32_e32 v3, 0xffff, v3 ; encoding: [0xff,0x06,0x06,0x36,0xff,0xff,0x00,0x00]
+; GFX10-NEXT:    v_and_b32_e32 v1, 0xffff, v1 ; encoding: [0xff,0x02,0x02,0x36,0xff,0xff,0x00,0x00]
+; GFX10-NEXT:    v_lshl_or_b32 v3, v4, 16, v3 ; encoding: [0x03,0x00,0x6f,0xd7,0x04,0x21,0x0d,0x04]
+; GFX10-NEXT:    v_lshl_or_b32 v1, v2, 16, v1 ; encoding: [0x01,0x00,0x6f,0xd7,0x02,0x21,0x05,0x04]
+; GFX10-NEXT:    image_sample_c_cd_g16 v[0:3], [v0, v1, v3, v5, v6], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D ; encoding: [0x0b,0x0f,0xa8,0xf1,0x00,0x00,0x40,0x00,0x01,0x03,0x05,0x06]
+; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
+; GFX10-NEXT:    ; return to shader part epilog
+main_body:
+  %v = call <4 x float> @llvm.amdgcn.image.sample.c.cd.2d.v4f32.f16.f32(i32 15, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
+  ret <4 x float> %v
+}
+
+define amdgpu_ps <4 x float> @sample_cd_cl_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, half %dsdh, half %dsdv, float %s, float %clamp) {
+; GFX10-LABEL: sample_cd_cl_1d:
+; GFX10:       ; %bb.0: ; %main_body
+; GFX10-NEXT:    image_sample_cd_cl_g16 v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D ; encoding: [0x01,0x0f,0xa4,0xf1,0x00,0x00,0x40,0x00]
+; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
+; GFX10-NEXT:    ; return to shader part epilog
+main_body:
+  %v = call <4 x float> @llvm.amdgcn.image.sample.cd.cl.1d.v4f32.f16.f32(i32 15, half %dsdh, half %dsdv, float %s, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
+  ret <4 x float> %v
+}
+
+define amdgpu_ps <4 x float> @sample_cd_cl_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, float %clamp) {
+; GFX10-LABEL: sample_cd_cl_2d:
+; GFX10:       ; %bb.0: ; %main_body
+; GFX10-NEXT:    v_and_b32_e32 v2, 0xffff, v2 ; encoding: [0xff,0x04,0x04,0x36,0xff,0xff,0x00,0x00]
+; GFX10-NEXT:    v_and_b32_e32 v0, 0xffff, v0 ; encoding: [0xff,0x00,0x00,0x36,0xff,0xff,0x00,0x00]
+; GFX10-NEXT:    v_lshl_or_b32 v2, v3, 16, v2 ; encoding: [0x02,0x00,0x6f,0xd7,0x03,0x21,0x09,0x04]
+; GFX10-NEXT:    v_lshl_or_b32 v0, v1, 16, v0 ; encoding: [0x00,0x00,0x6f,0xd7,0x01,0x21,0x01,0x04]
+; GFX10-NEXT:    image_sample_cd_cl_g16 v[0:3], [v0, v2, v4, v5, v6], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D ; encoding: [0x0b,0x0f,0xa4,0xf1,0x00,0x00,0x40,0x00,0x02,0x04,0x05,0x06]
+; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
+; GFX10-NEXT:    ; return to shader part epilog
+main_body:
+  %v = call <4 x float> @llvm.amdgcn.image.sample.cd.cl.2d.v4f32.f16.f32(i32 15, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
+  ret <4 x float> %v
+}
+
+define amdgpu_ps <4 x float> @sample_c_cd_cl_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, half %dsdh, half %dsdv, float %s, float %clamp) {
+; GFX10-LABEL: sample_c_cd_cl_1d:
+; GFX10:       ; %bb.0: ; %main_body
+; GFX10-NEXT:    image_sample_c_cd_cl_g16 v[0:3], v[0:4], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D ; encoding: [0x01,0x0f,0xac,0xf1,0x00,0x00,0x40,0x00]
+; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
+; GFX10-NEXT:    ; return to shader part epilog
+main_body:
+  %v = call <4 x float> @llvm.amdgcn.image.sample.c.cd.cl.1d.v4f32.f16.f32(i32 15, float %zcompare, half %dsdh, half %dsdv, float %s, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
+  ret <4 x float> %v
+}
+
+define amdgpu_ps <4 x float> @sample_c_cd_cl_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, float %clamp) {
+; GFX10-LABEL: sample_c_cd_cl_2d:
+; GFX10:       ; %bb.0: ; %main_body
+; GFX10-NEXT:    v_mov_b32_e32 v8, v2 ; encoding: [0x02,0x03,0x10,0x7e]
+; GFX10-NEXT:    v_mov_b32_e32 v2, v0 ; encoding: [0x00,0x03,0x04,0x7e]
+; GFX10-NEXT:    v_and_b32_e32 v0, 0xffff, v3 ; encoding: [0xff,0x06,0x00,0x36,0xff,0xff,0x00,0x00]
+; GFX10-NEXT:    v_and_b32_e32 v1, 0xffff, v1 ; encoding: [0xff,0x02,0x02,0x36,0xff,0xff,0x00,0x00]
+; GFX10-NEXT:    v_lshl_or_b32 v4, v4, 16, v0 ; encoding: [0x04,0x00,0x6f,0xd7,0x04,0x21,0x01,0x04]
+; GFX10-NEXT:    v_lshl_or_b32 v3, v8, 16, v1 ; encoding: [0x03,0x00,0x6f,0xd7,0x08,0x21,0x05,0x04]
+; GFX10-NEXT:    image_sample_c_cd_cl_g16 v[0:3], v[2:7], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D ; encoding: [0x09,0x0f,0xac,0xf1,0x02,0x00,0x40,0x00]
+; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
+; GFX10-NEXT:    ; return to shader part epilog
+main_body:
+  %v = call <4 x float> @llvm.amdgcn.image.sample.c.cd.cl.2d.v4f32.f16.f32(i32 15, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
+  ret <4 x float> %v
+}
+
+declare <4 x float> @llvm.amdgcn.image.sample.cd.1d.v4f32.f16.f32(i32, half, half, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
+declare <4 x float> @llvm.amdgcn.image.sample.cd.2d.v4f32.f16.f32(i32, half, half, half, half, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
+declare <4 x float> @llvm.amdgcn.image.sample.c.cd.1d.v4f32.f16.f32(i32, float, half, half, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
+declare <4 x float> @llvm.amdgcn.image.sample.c.cd.2d.v4f32.f16.f32(i32, float, half, half, half, half, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
+declare <4 x float> @llvm.amdgcn.image.sample.cd.cl.1d.v4f32.f16.f32(i32, half, half, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
+declare <4 x float> @llvm.amdgcn.image.sample.cd.cl.2d.v4f32.f16.f32(i32, half, half, half, half, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
+declare <4 x float> @llvm.amdgcn.image.sample.c.cd.cl.1d.v4f32.f16.f32(i32, float, half, half, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
+declare <4 x float> @llvm.amdgcn.image.sample.c.cd.cl.2d.v4f32.f16.f32(i32, float, half, half, half, half, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
+
+attributes #0 = { nounwind }
+attributes #1 = { nounwind readonly }
+attributes #2 = { nounwind readnone }

llvm/test/CodeGen/AMDGPU/llvm.amdgcn.image.sample.cd.g16.ll

+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
+; RUN: llc -march=amdgcn -mcpu=gfx1010 -verify-machineinstrs < %s | FileCheck -check-prefix=GFX10 %s
+
+define amdgpu_ps <4 x float> @sample_cd_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, half %dsdh, half %dsdv, float %s) {
+; GFX10-LABEL: sample_cd_1d:
+; GFX10:       ; %bb.0: ; %main_body
+; GFX10-NEXT:    image_sample_cd_g16 v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    ; return to shader part epilog
+main_body:
+  %v = call <4 x float> @llvm.amdgcn.image.sample.cd.1d.v4f32.f16.f32(i32 15, half %dsdh, half %dsdv, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
+  ret <4 x float> %v
+}
+
+define amdgpu_ps <4 x float> @sample_cd_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t) {
+; GFX10-LABEL: sample_cd_2d:
+; GFX10:       ; %bb.0: ; %main_body
+; GFX10-NEXT:    v_and_b32_e32 v2, 0xffff, v2
+; GFX10-NEXT:    v_and_b32_e32 v0, 0xffff, v0
+; GFX10-NEXT:    v_lshl_or_b32 v2, v3, 16, v2
+; GFX10-NEXT:    v_lshl_or_b32 v0, v1, 16, v0
+; GFX10-NEXT:    image_sample_cd_g16 v[0:3], [v0, v2, v4, v5], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    ; return to shader part epilog
+main_body:
+  %v = call <4 x float> @llvm.amdgcn.image.sample.cd.2d.v4f32.f16.f32(i32 15, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
+  ret <4 x float> %v
+}
+
+define amdgpu_ps <4 x float> @sample_c_cd_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, half %dsdh, half %dsdv, float %s) {
+; GFX10-LABEL: sample_c_cd_1d:
+; GFX10:       ; %bb.0: ; %main_body
+; GFX10-NEXT:    image_sample_c_cd_g16 v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    ; return to shader part epilog
+main_body:
+  %v = call <4 x float> @llvm.amdgcn.image.sample.c.cd.1d.v4f32.f16.f32(i32 15, float %zcompare, half %dsdh, half %dsdv, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
+  ret <4 x float> %v
+}
+
+define amdgpu_ps <4 x float> @sample_c_cd_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t) {
+; GFX10-LABEL: sample_c_cd_2d:
+; GFX10:       ; %bb.0: ; %main_body
+; GFX10-NEXT:    v_and_b32_e32 v3, 0xffff, v3
+; GFX10-NEXT:    v_and_b32_e32 v1, 0xffff, v1
+; GFX10-NEXT:    v_lshl_or_b32 v3, v4, 16, v3
+; GFX10-NEXT:    v_lshl_or_b32 v1, v2, 16, v1
+; GFX10-NEXT:    image_sample_c_cd_g16 v[0:3], [v0, v1, v3, v5, v6], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    ; return to shader part epilog
+main_body:
+  %v = call <4 x float> @llvm.amdgcn.image.sample.c.cd.2d.v4f32.f16.f32(i32 15, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
+  ret <4 x float> %v
+}
+
+define amdgpu_ps <4 x float> @sample_cd_cl_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, half %dsdh, half %dsdv, float %s, float %clamp) {
+; GFX10-LABEL: sample_cd_cl_1d:
+; GFX10:       ; %bb.0: ; %main_body
+; GFX10-NEXT:    image_sample_cd_cl_g16 v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    ; return to shader part epilog
+main_body:
+  %v = call <4 x float> @llvm.amdgcn.image.sample.cd.cl.1d.v4f32.f16.f32(i32 15, half %dsdh, half %dsdv, float %s, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
+  ret <4 x float> %v
+}
+
+define amdgpu_ps <4 x float> @sample_cd_cl_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, float %clamp) {
+; GFX10-LABEL: sample_cd_cl_2d:
+; GFX10:       ; %bb.0: ; %main_body
+; GFX10-NEXT:    v_and_b32_e32 v2, 0xffff, v2
+; GFX10-NEXT:    v_and_b32_e32 v0, 0xffff, v0
+; GFX10-NEXT:    v_lshl_or_b32 v2, v3, 16, v2
+; GFX10-NEXT:    v_lshl_or_b32 v0, v1, 16, v0
+; GFX10-NEXT:    image_sample_cd_cl_g16 v[0:3], [v0, v2, v4, v5, v6], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    ; return to shader part epilog
+main_body:
+  %v = call <4 x float> @llvm.amdgcn.image.sample.cd.cl.2d.v4f32.f16.f32(i32 15, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
+  ret <4 x float> %v
+}
+
+define amdgpu_ps <4 x float> @sample_c_cd_cl_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, half %dsdh, half %dsdv, float %s, float %clamp) {
+; GFX10-LABEL: sample_c_cd_cl_1d:
+; GFX10:       ; %bb.0: ; %main_body
+; GFX10-NEXT:    image_sample_c_cd_cl_g16 v[0:3], v[0:4], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    ; return to shader part epilog
+main_body:
+  %v = call <4 x float> @llvm.amdgcn.image.sample.c.cd.cl.1d.v4f32.f16.f32(i32 15, float %zcompare, half %dsdh, half %dsdv, float %s, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
+  ret <4 x float> %v
+}
+
+define amdgpu_ps <4 x float> @sample_c_cd_cl_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, float %clamp) {
+; GFX10-LABEL: sample_c_cd_cl_2d:
+; GFX10:       ; %bb.0: ; %main_body
+; GFX10-NEXT:    v_mov_b32_e32 v8, v2
+; GFX10-NEXT:    v_mov_b32_e32 v2, v0
+; GFX10-NEXT:    v_and_b32_e32 v0, 0xffff, v3
+; GFX10-NEXT:    v_and_b32_e32 v1, 0xffff, v1
+; GFX10-NEXT:    v_lshl_or_b32 v4, v4, 16, v0
+; GFX10-NEXT:    v_lshl_or_b32 v3, v8, 16, v1
+; GFX10-NEXT:    image_sample_c_cd_cl_g16 v[0:3], v[2:7], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    ; return to shader part epilog
+main_body:
+  %v = call <4 x float> @llvm.amdgcn.image.sample.c.cd.cl.2d.v4f32.f16.f32(i32 15, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
+  ret <4 x float> %v
+}
+
+declare <4 x float> @llvm.amdgcn.image.sample.cd.1d.v4f32.f16.f32(i32, half, half, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
+declare <4 x float> @llvm.amdgcn.image.sample.cd.2d.v4f32.f16.f32(i32, half, half, half, half, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
+declare <4 x float> @llvm.amdgcn.image.sample.c.cd.1d.v4f32.f16.f32(i32, float, half, half, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
+declare <4 x float> @llvm.amdgcn.image.sample.c.cd.2d.v4f32.f16.f32(i32, float, half, half, half, half, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
+declare <4 x float> @llvm.amdgcn.image.sample.cd.cl.1d.v4f32.f16.f32(i32, half, half, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
+declare <4 x float> @llvm.amdgcn.image.sample.cd.cl.2d.v4f32.f16.f32(i32, half, half, half, half, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
+declare <4 x float> @llvm.amdgcn.image.sample.c.cd.cl.1d.v4f32.f16.f32(i32, float, half, half, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
+declare <4 x float> @llvm.amdgcn.image.sample.c.cd.cl.2d.v4f32.f16.f32(i32, float, half, half, half, half, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
+
+attributes #0 = { nounwind }
+attributes #1 = { nounwind readonly }
+attributes #2 = { nounwind readnone }

llvm/test/CodeGen/AMDGPU/llvm.amdgcn.image.sample.d16.dim.ll

 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
 ; RUN: llc < %s -march=amdgcn -mcpu=tonga -verify-machineinstrs | FileCheck -check-prefixes=TONGA %s
 ; RUN: llc < %s -march=amdgcn -mcpu=gfx810 -verify-machineinstrs | FileCheck -check-prefixes=GFX81 %s
 ; RUN: llc < %s -march=amdgcn -mcpu=gfx900 -verify-machineinstrs | FileCheck -check-prefixes=GFX9 %s
-; RUN: llc < %s -march=amdgcn -mcpu=gfx1010 -verify-machineinstrs | FileCheck -check-prefixes=GFX10 %s
+; RUN: llc < %s -march=amdgcn -mcpu=gfx1010 -verify-machineinstrs | FileCheck -check-prefixes=GFX10PLUS,GFX10 %s
+; RUN: llc < %s -march=amdgcn -mcpu=gfx1100 -verify-machineinstrs | FileCheck -check-prefixes=GFX10PLUS,GFX11 %s
 
 define amdgpu_ps half @image_sample_2d_f16(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %s, float %t) {
 ; TONGA-LABEL: image_sample_2d_f16:
 ; TONGA:       ; %bb.0: ; %main_body
 ; TONGA-NEXT:    s_mov_b64 s[12:13], exec
 ; TONGA-NEXT:    s_wqm_b64 exec, exec
 ; TONGA-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; TONGA-NEXT:    image_sample v0, v[0:1], s[0:7], s[8:11] dmask:0x1 d16

 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    s_mov_b64 s[12:13], exec
 ; GFX9-NEXT:    s_wqm_b64 exec, exec
 ; GFX9-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; GFX9-NEXT:    image_sample v0, v[0:1], s[0:7], s[8:11] dmask:0x1 d16
 ; GFX9-NEXT:    s_waitcnt vmcnt(0)
 ; GFX9-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: image_sample_2d_f16:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    s_mov_b32 s12, exec_lo
-; GFX10-NEXT:    s_wqm_b32 exec_lo, exec_lo
-; GFX10-NEXT:    s_and_b32 exec_lo, exec_lo, s12
-; GFX10-NEXT:    image_sample v0, v[0:1], s[0:7], s[8:11] dmask:0x1 dim:SQ_RSRC_IMG_2D d16
-; GFX10-NEXT:    s_waitcnt vmcnt(0)
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: image_sample_2d_f16:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    s_mov_b32 s12, exec_lo
+; GFX10PLUS-NEXT:    s_wqm_b32 exec_lo, exec_lo
+; GFX10PLUS-NEXT:    s_and_b32 exec_lo, exec_lo, s12
+; GFX10PLUS-NEXT:    image_sample v0, v[0:1], s[0:7], s[8:11] dmask:0x1 dim:SQ_RSRC_IMG_2D d16
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %tex = call half @llvm.amdgcn.image.sample.2d.f16.f32(i32 1, float %s, float %t, <8 x i32> %rsrc, <4 x i32> %samp, i1 false, i32 0, i32 0)
   ret half %tex
 }
 
 define amdgpu_ps half @image_sample_2d_f16_tfe(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %s, float %t, i32 addrspace(1)* inreg %out) {
 ; TONGA-LABEL: image_sample_2d_f16_tfe:
 ; TONGA:       ; %bb.0: ; %main_body

 ; GFX10-NEXT:    v_mov_b32_e32 v3, v5
 ; GFX10-NEXT:    s_and_b32 exec_lo, exec_lo, s14
 ; GFX10-NEXT:    image_sample v[2:3], v[0:1], s[0:7], s[8:11] dmask:0x1 dim:SQ_RSRC_IMG_2D tfe d16
 ; GFX10-NEXT:    s_waitcnt vmcnt(0)
 ; GFX10-NEXT:    v_mov_b32_e32 v0, v2
 ; GFX10-NEXT:    global_store_dword v4, v3, s[12:13]
 ; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
 ; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: image_sample_2d_f16_tfe:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    s_mov_b32 s14, exec_lo
+; GFX11-NEXT:    s_wqm_b32 exec_lo, exec_lo
+; GFX11-NEXT:    v_mov_b32_e32 v4, 0
+; GFX11-NEXT:    v_mov_b32_e32 v5, v4
+; GFX11-NEXT:    v_mov_b32_e32 v2, v4
+; GFX11-NEXT:    v_mov_b32_e32 v3, v5
+; GFX11-NEXT:    s_and_b32 exec_lo, exec_lo, s14
+; GFX11-NEXT:    image_sample v[2:3], v[0:1], s[0:7], s[8:11] dmask:0x1 dim:SQ_RSRC_IMG_2D tfe d16
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    v_mov_b32_e32 v0, v2
+; GFX11-NEXT:    global_store_b32 v4, v3, s[12:13]
+; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %tex = call {half,i32} @llvm.amdgcn.image.sample.2d.f16i32.f32(i32 1, float %s, float %t, <8 x i32> %rsrc, <4 x i32> %samp, i1 false, i32 1, i32 0)
   %tex.vec = extractvalue {half, i32} %tex, 0
   %tex.err = extractvalue {half, i32} %tex, 1
   store i32 %tex.err, i32 addrspace(1)* %out, align 4
   ret half %tex.vec
 }
 

 ; GFX81-NEXT:    ; return to shader part epilog
 ;
 ; GFX9-LABEL: image_sample_c_d_1d_v2f16:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    image_sample_c_d v0, v[0:3], s[0:7], s[8:11] dmask:0x3 d16
 ; GFX9-NEXT:    s_waitcnt vmcnt(0)
 ; GFX9-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: image_sample_c_d_1d_v2f16:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_sample_c_d v0, v[0:3], s[0:7], s[8:11] dmask:0x3 dim:SQ_RSRC_IMG_1D d16
-; GFX10-NEXT:    s_waitcnt vmcnt(0)
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: image_sample_c_d_1d_v2f16:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_sample_c_d v0, v[0:3], s[0:7], s[8:11] dmask:0x3 dim:SQ_RSRC_IMG_1D d16
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %tex = call <2 x half> @llvm.amdgcn.image.sample.c.d.1d.v2f16.f32.f32(i32 3, float %zcompare, float %dsdh, float %dsdv, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 false, i32 0, i32 0)
   %r = bitcast <2 x half> %tex to float
   ret float %r
 }
 
 define amdgpu_ps <2 x float> @image_sample_c_d_1d_v2f16_tfe(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, float %dsdh, float %dsdv, float %s) {
 ; TONGA-LABEL: image_sample_c_d_1d_v2f16_tfe:

 ; GFX9-NEXT:    v_mov_b32_e32 v4, 0
 ; GFX9-NEXT:    v_mov_b32_e32 v5, v4
 ; GFX9-NEXT:    image_sample_c_d v[4:5], v[0:3], s[0:7], s[8:11] dmask:0x3 tfe d16
 ; GFX9-NEXT:    s_waitcnt vmcnt(0)
 ; GFX9-NEXT:    v_mov_b32_e32 v0, v4
 ; GFX9-NEXT:    v_mov_b32_e32 v1, v5
 ; GFX9-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: image_sample_c_d_1d_v2f16_tfe:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    v_mov_b32_e32 v5, v0
-; GFX10-NEXT:    v_mov_b32_e32 v0, 0
-; GFX10-NEXT:    v_mov_b32_e32 v4, v1
-; GFX10-NEXT:    v_mov_b32_e32 v1, v0
-; GFX10-NEXT:    image_sample_c_d v[0:1], [v5, v4, v2, v3], s[0:7], s[8:11] dmask:0x3 dim:SQ_RSRC_IMG_1D tfe d16
-; GFX10-NEXT:    s_waitcnt vmcnt(0)
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: image_sample_c_d_1d_v2f16_tfe:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v5, v0
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v0, 0
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v4, v1
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v1, v0
+; GFX10PLUS-NEXT:    image_sample_c_d v[0:1], [v5, v4, v2, v3], s[0:7], s[8:11] dmask:0x3 dim:SQ_RSRC_IMG_1D tfe d16
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %tex = call {<2 x half>,i32} @llvm.amdgcn.image.sample.c.d.1d.v2f16i32.f32.f32(i32 3, float %zcompare, float %dsdh, float %dsdv, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 false, i32 1, i32 0)
   %tex.vec = extractvalue {<2 x half>, i32} %tex, 0
   %tex.err = extractvalue {<2 x half>, i32} %tex, 1
   %tex.vecf = bitcast <2 x half> %tex.vec to float
   %r.0 = insertelement <2 x float> undef, float %tex.vecf, i32 0
   %tex.errf = bitcast i32 %tex.err to float
   %r = insertelement <2 x float> %r.0, float %tex.errf, i32 1

 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    s_mov_b64 s[12:13], exec
 ; GFX9-NEXT:    s_wqm_b64 exec, exec
 ; GFX9-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; GFX9-NEXT:    image_sample_b v[0:1], v[0:2], s[0:7], s[8:11] dmask:0x7 d16
 ; GFX9-NEXT:    s_waitcnt vmcnt(0)
 ; GFX9-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: image_sample_b_2d_v3f16:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    s_mov_b32 s12, exec_lo
-; GFX10-NEXT:    s_wqm_b32 exec_lo, exec_lo
-; GFX10-NEXT:    s_and_b32 exec_lo, exec_lo, s12
-; GFX10-NEXT:    image_sample_b v[0:1], v[0:2], s[0:7], s[8:11] dmask:0x7 dim:SQ_RSRC_IMG_2D d16
-; GFX10-NEXT:    s_waitcnt vmcnt(0)
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: image_sample_b_2d_v3f16:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    s_mov_b32 s12, exec_lo
+; GFX10PLUS-NEXT:    s_wqm_b32 exec_lo, exec_lo
+; GFX10PLUS-NEXT:    s_and_b32 exec_lo, exec_lo, s12
+; GFX10PLUS-NEXT:    image_sample_b v[0:1], v[0:2], s[0:7], s[8:11] dmask:0x7 dim:SQ_RSRC_IMG_2D d16
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %tex = call <3 x half> @llvm.amdgcn.image.sample.b.2d.v3f16.f32.f32(i32 7, float %bias, float %s, float %t, <8 x i32> %rsrc, <4 x i32> %samp, i1 false, i32 0, i32 0)
   %tex_wide = shufflevector <3 x half> %tex, <3 x half> undef, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
   %r = bitcast <4 x half> %tex_wide to <2 x float>
   ret <2 x float> %r
 }
 
 define amdgpu_ps <4 x float> @image_sample_b_2d_v3f16_tfe(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %bias, float %s, float %t) {

 ; GFX9-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; GFX9-NEXT:    image_sample_b v[3:5], v[0:2], s[0:7], s[8:11] dmask:0x7 tfe d16
 ; GFX9-NEXT:    s_waitcnt vmcnt(0)
 ; GFX9-NEXT:    v_mov_b32_e32 v0, v3
 ; GFX9-NEXT:    v_mov_b32_e32 v1, v4
 ; GFX9-NEXT:    v_mov_b32_e32 v2, v5
 ; GFX9-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: image_sample_b_2d_v3f16_tfe:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    s_mov_b32 s12, exec_lo
-; GFX10-NEXT:    s_wqm_b32 exec_lo, exec_lo
-; GFX10-NEXT:    v_mov_b32_e32 v3, v0
-; GFX10-NEXT:    v_mov_b32_e32 v0, 0
-; GFX10-NEXT:    v_mov_b32_e32 v5, v2
-; GFX10-NEXT:    v_mov_b32_e32 v4, v1
-; GFX10-NEXT:    v_mov_b32_e32 v1, v0
-; GFX10-NEXT:    v_mov_b32_e32 v2, v0
-; GFX10-NEXT:    s_and_b32 exec_lo, exec_lo, s12
-; GFX10-NEXT:    image_sample_b v[0:2], v[3:5], s[0:7], s[8:11] dmask:0x7 dim:SQ_RSRC_IMG_2D tfe d16
-; GFX10-NEXT:    s_waitcnt vmcnt(0)
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: image_sample_b_2d_v3f16_tfe:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    s_mov_b32 s12, exec_lo
+; GFX10PLUS-NEXT:    s_wqm_b32 exec_lo, exec_lo
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v3, v0
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v0, 0
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v5, v2
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v4, v1
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v1, v0
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v2, v0
+; GFX10PLUS-NEXT:    s_and_b32 exec_lo, exec_lo, s12
+; GFX10PLUS-NEXT:    image_sample_b v[0:2], v[3:5], s[0:7], s[8:11] dmask:0x7 dim:SQ_RSRC_IMG_2D tfe d16
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %tex = call {<3 x half>,i32} @llvm.amdgcn.image.sample.b.2d.v3f16i32.f32.f32(i32 7, float %bias, float %s, float %t, <8 x i32> %rsrc, <4 x i32> %samp, i1 false, i32 1, i32 0)
   %tex.vec = extractvalue {<3 x half>, i32} %tex, 0
   %tex.vec_wide = shufflevector <3 x half> %tex.vec, <3 x half> undef, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
   %tex.err = extractvalue {<3 x half>, i32} %tex, 1
   %tex.vecf = bitcast <4 x half> %tex.vec_wide to <2 x float>
   %tex.vecf.0 = extractelement <2 x float> %tex.vecf, i32 0
   %tex.vecf.1 = extractelement <2 x float> %tex.vecf, i32 1

 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    s_mov_b64 s[12:13], exec
 ; GFX9-NEXT:    s_wqm_b64 exec, exec
 ; GFX9-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; GFX9-NEXT:    image_sample_b v[0:1], v[0:2], s[0:7], s[8:11] dmask:0xf d16
 ; GFX9-NEXT:    s_waitcnt vmcnt(0)
 ; GFX9-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: image_sample_b_2d_v4f16:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    s_mov_b32 s12, exec_lo
-; GFX10-NEXT:    s_wqm_b32 exec_lo, exec_lo
-; GFX10-NEXT:    s_and_b32 exec_lo, exec_lo, s12
-; GFX10-NEXT:    image_sample_b v[0:1], v[0:2], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D d16
-; GFX10-NEXT:    s_waitcnt vmcnt(0)
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: image_sample_b_2d_v4f16:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    s_mov_b32 s12, exec_lo
+; GFX10PLUS-NEXT:    s_wqm_b32 exec_lo, exec_lo
+; GFX10PLUS-NEXT:    s_and_b32 exec_lo, exec_lo, s12
+; GFX10PLUS-NEXT:    image_sample_b v[0:1], v[0:2], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D d16
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %tex = call <4 x half> @llvm.amdgcn.image.sample.b.2d.v4f16.f32.f32(i32 15, float %bias, float %s, float %t, <8 x i32> %rsrc, <4 x i32> %samp, i1 false, i32 0, i32 0)
   %r = bitcast <4 x half> %tex to <2 x float>
   ret <2 x float> %r
 }
 
 define amdgpu_ps <4 x float> @image_sample_b_2d_v4f16_tfe(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %bias, float %s, float %t) {
 ; TONGA-LABEL: image_sample_b_2d_v4f16_tfe:

 ; GFX9-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; GFX9-NEXT:    image_sample_b v[3:5], v[0:2], s[0:7], s[8:11] dmask:0xf tfe d16
 ; GFX9-NEXT:    s_waitcnt vmcnt(0)
 ; GFX9-NEXT:    v_mov_b32_e32 v0, v3
 ; GFX9-NEXT:    v_mov_b32_e32 v1, v4
 ; GFX9-NEXT:    v_mov_b32_e32 v2, v5
 ; GFX9-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: image_sample_b_2d_v4f16_tfe:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    s_mov_b32 s12, exec_lo
-; GFX10-NEXT:    s_wqm_b32 exec_lo, exec_lo
-; GFX10-NEXT:    v_mov_b32_e32 v3, v0
-; GFX10-NEXT:    v_mov_b32_e32 v0, 0
-; GFX10-NEXT:    v_mov_b32_e32 v5, v2
-; GFX10-NEXT:    v_mov_b32_e32 v4, v1
-; GFX10-NEXT:    v_mov_b32_e32 v1, v0
-; GFX10-NEXT:    v_mov_b32_e32 v2, v0
-; GFX10-NEXT:    s_and_b32 exec_lo, exec_lo, s12
-; GFX10-NEXT:    image_sample_b v[0:2], v[3:5], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D tfe d16
-; GFX10-NEXT:    s_waitcnt vmcnt(0)
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: image_sample_b_2d_v4f16_tfe:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    s_mov_b32 s12, exec_lo
+; GFX10PLUS-NEXT:    s_wqm_b32 exec_lo, exec_lo
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v3, v0
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v0, 0
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v5, v2
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v4, v1
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v1, v0
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v2, v0
+; GFX10PLUS-NEXT:    s_and_b32 exec_lo, exec_lo, s12
+; GFX10PLUS-NEXT:    image_sample_b v[0:2], v[3:5], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D tfe d16
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %tex = call {<4 x half>,i32} @llvm.amdgcn.image.sample.b.2d.v4f16i32.f32.f32(i32 15, float %bias, float %s, float %t, <8 x i32> %rsrc, <4 x i32> %samp, i1 false, i32 1, i32 0)
   %tex.vec = extractvalue {<4 x half>, i32} %tex, 0
   %tex.err = extractvalue {<4 x half>, i32} %tex, 1
   %tex.vecf = bitcast <4 x half> %tex.vec to <2 x float>
   %tex.vecf.0 = extractelement <2 x float> %tex.vecf, i32 0
   %tex.vecf.1 = extractelement <2 x float> %tex.vecf, i32 1
   %r.0 = insertelement <4 x float> undef, float %tex.vecf.0, i32 0

llvm/test/CodeGen/AMDGPU/llvm.amdgcn.image.sample.dim.ll

 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
 ; RUN: llc -march=amdgcn -mcpu=verde -verify-machineinstrs < %s | FileCheck -check-prefixes=VERDE %s
 ; RUN: llc -march=amdgcn -mcpu=gfx900 -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX6789 %s
-; RUN: llc -march=amdgcn -mcpu=gfx1010 -verify-machineinstrs -show-mc-encoding < %s | FileCheck -check-prefixes=GFX10 %s
+; RUN: llc -march=amdgcn -mcpu=gfx1010 -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX10PLUS,GFX10 %s
+; RUN: llc -march=amdgcn -mcpu=gfx1100 -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX10PLUS,GFX11 %s
 
 define amdgpu_ps <4 x float> @sample_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %s) {
 ; VERDE-LABEL: sample_1d:
 ; VERDE:       ; %bb.0: ; %main_body
 ; VERDE-NEXT:    s_mov_b64 s[12:13], exec
 ; VERDE-NEXT:    s_wqm_b64 exec, exec
 ; VERDE-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; VERDE-NEXT:    image_sample v[0:3], v0, s[0:7], s[8:11] dmask:0xf
 ; VERDE-NEXT:    s_waitcnt vmcnt(0)
 ; VERDE-NEXT:    ; return to shader part epilog
 ;
 ; GFX6789-LABEL: sample_1d:
 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    s_mov_b64 s[12:13], exec
 ; GFX6789-NEXT:    s_wqm_b64 exec, exec
 ; GFX6789-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; GFX6789-NEXT:    image_sample v[0:3], v0, s[0:7], s[8:11] dmask:0xf
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: sample_1d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    s_mov_b32 s12, exec_lo ; encoding: [0x7e,0x03,0x8c,0xbe]
-; GFX10-NEXT:    s_wqm_b32 exec_lo, exec_lo ; encoding: [0x7e,0x09,0xfe,0xbe]
-; GFX10-NEXT:    s_and_b32 exec_lo, exec_lo, s12 ; encoding: [0x7e,0x0c,0x7e,0x87]
-; GFX10-NEXT:    image_sample v[0:3], v0, s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D ; encoding: [0x00,0x0f,0x80,0xf0,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: sample_1d:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    s_mov_b32 s12, exec_lo
+; GFX10PLUS-NEXT:    s_wqm_b32 exec_lo, exec_lo
+; GFX10PLUS-NEXT:    s_and_b32 exec_lo, exec_lo, s12
+; GFX10PLUS-NEXT:    image_sample v[0:3], v0, s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.1d.v4f32.f32(i32 15, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @sample_1d_tfe(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, i32 addrspace(1)* inreg %out, float %s) {
 ; VERDE-LABEL: sample_1d_tfe:
 ; VERDE:       ; %bb.0: ; %main_body

 ; GFX6789-NEXT:    image_sample v[0:4], v5, s[0:7], s[8:11] dmask:0xf tfe
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    global_store_dword v6, v4, s[12:13]
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
 ; GFX10-LABEL: sample_1d_tfe:
 ; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    s_mov_b32 s14, exec_lo ; encoding: [0x7e,0x03,0x8e,0xbe]
-; GFX10-NEXT:    s_wqm_b32 exec_lo, exec_lo ; encoding: [0x7e,0x09,0xfe,0xbe]
-; GFX10-NEXT:    v_mov_b32_e32 v6, 0 ; encoding: [0x80,0x02,0x0c,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v5, v0 ; encoding: [0x00,0x03,0x0a,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v7, v6 ; encoding: [0x06,0x03,0x0e,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v8, v6 ; encoding: [0x06,0x03,0x10,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v9, v6 ; encoding: [0x06,0x03,0x12,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v10, v6 ; encoding: [0x06,0x03,0x14,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v0, v6 ; encoding: [0x06,0x03,0x00,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v1, v7 ; encoding: [0x07,0x03,0x02,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v2, v8 ; encoding: [0x08,0x03,0x04,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v3, v9 ; encoding: [0x09,0x03,0x06,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v4, v10 ; encoding: [0x0a,0x03,0x08,0x7e]
-; GFX10-NEXT:    s_and_b32 exec_lo, exec_lo, s14 ; encoding: [0x7e,0x0e,0x7e,0x87]
-; GFX10-NEXT:    image_sample v[0:4], v5, s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D tfe ; encoding: [0x00,0x0f,0x81,0xf0,0x05,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    global_store_dword v6, v4, s[12:13] ; encoding: [0x00,0x80,0x70,0xdc,0x06,0x04,0x0c,0x00]
-; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0 ; encoding: [0x00,0x00,0xfd,0xbb]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10-NEXT:    s_mov_b32 s14, exec_lo
+; GFX10-NEXT:    s_wqm_b32 exec_lo, exec_lo
+; GFX10-NEXT:    v_mov_b32_e32 v6, 0
+; GFX10-NEXT:    v_mov_b32_e32 v5, v0
+; GFX10-NEXT:    v_mov_b32_e32 v7, v6
+; GFX10-NEXT:    v_mov_b32_e32 v8, v6
+; GFX10-NEXT:    v_mov_b32_e32 v9, v6
+; GFX10-NEXT:    v_mov_b32_e32 v10, v6
+; GFX10-NEXT:    v_mov_b32_e32 v0, v6
+; GFX10-NEXT:    v_mov_b32_e32 v1, v7
+; GFX10-NEXT:    v_mov_b32_e32 v2, v8
+; GFX10-NEXT:    v_mov_b32_e32 v3, v9
+; GFX10-NEXT:    v_mov_b32_e32 v4, v10
+; GFX10-NEXT:    s_and_b32 exec_lo, exec_lo, s14
+; GFX10-NEXT:    image_sample v[0:4], v5, s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D tfe
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    global_store_dword v6, v4, s[12:13]
+; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: sample_1d_tfe:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    s_mov_b32 s14, exec_lo
+; GFX11-NEXT:    s_wqm_b32 exec_lo, exec_lo
+; GFX11-NEXT:    v_mov_b32_e32 v6, 0
+; GFX11-NEXT:    v_mov_b32_e32 v5, v0
+; GFX11-NEXT:    v_mov_b32_e32 v7, v6
+; GFX11-NEXT:    v_mov_b32_e32 v8, v6
+; GFX11-NEXT:    v_mov_b32_e32 v9, v6
+; GFX11-NEXT:    v_mov_b32_e32 v10, v6
+; GFX11-NEXT:    v_mov_b32_e32 v0, v6
+; GFX11-NEXT:    v_mov_b32_e32 v1, v7
+; GFX11-NEXT:    v_mov_b32_e32 v2, v8
+; GFX11-NEXT:    v_mov_b32_e32 v3, v9
+; GFX11-NEXT:    v_mov_b32_e32 v4, v10
+; GFX11-NEXT:    s_and_b32 exec_lo, exec_lo, s14
+; GFX11-NEXT:    image_sample v[0:4], v5, s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D tfe
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    global_store_b32 v6, v4, s[12:13]
+; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %v = call {<4 x float>,i32} @llvm.amdgcn.image.sample.1d.v4f32i32.f32(i32 15, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 1, i32 0)
   %v.vec = extractvalue {<4 x float>, i32} %v, 0
   %v.err = extractvalue {<4 x float>, i32} %v, 1
   store i32 %v.err, i32 addrspace(1)* %out, align 4
   ret <4 x float> %v.vec
 }
 

 ; GFX6789-NEXT:    v_mov_b32_e32 v2, v0
 ; GFX6789-NEXT:    v_mov_b32_e32 v0, 0
 ; GFX6789-NEXT:    v_mov_b32_e32 v1, v0
 ; GFX6789-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; GFX6789-NEXT:    image_sample v[0:1], v2, s[0:7], s[8:11] dmask:0x1 tfe
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: sample_1d_tfe_adjust_writemask_1:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    s_mov_b32 s12, exec_lo ; encoding: [0x7e,0x03,0x8c,0xbe]
-; GFX10-NEXT:    s_wqm_b32 exec_lo, exec_lo ; encoding: [0x7e,0x09,0xfe,0xbe]
-; GFX10-NEXT:    v_mov_b32_e32 v2, v0 ; encoding: [0x00,0x03,0x04,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v0, 0 ; encoding: [0x80,0x02,0x00,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v1, v0 ; encoding: [0x00,0x03,0x02,0x7e]
-; GFX10-NEXT:    s_and_b32 exec_lo, exec_lo, s12 ; encoding: [0x7e,0x0c,0x7e,0x87]
-; GFX10-NEXT:    image_sample v[0:1], v2, s[0:7], s[8:11] dmask:0x1 dim:SQ_RSRC_IMG_1D tfe ; encoding: [0x00,0x01,0x81,0xf0,0x02,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: sample_1d_tfe_adjust_writemask_1:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    s_mov_b32 s12, exec_lo
+; GFX10PLUS-NEXT:    s_wqm_b32 exec_lo, exec_lo
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v2, v0
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v0, 0
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v1, v0
+; GFX10PLUS-NEXT:    s_and_b32 exec_lo, exec_lo, s12
+; GFX10PLUS-NEXT:    image_sample v[0:1], v2, s[0:7], s[8:11] dmask:0x1 dim:SQ_RSRC_IMG_1D tfe
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call {<4 x float>,i32} @llvm.amdgcn.image.sample.1d.v4f32i32.f32(i32 15, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 1, i32 0)
   %res.vec = extractvalue {<4 x float>,i32} %v, 0
   %res.f = extractelement <4 x float> %res.vec, i32 0
   %res.err = extractvalue {<4 x float>,i32} %v, 1
   %res.errf = bitcast i32 %res.err to float
   %res.tmp = insertelement <2 x float> undef, float %res.f, i32 0
   %res = insertelement <2 x float> %res.tmp, float %res.errf, i32 1

 ; GFX6789-NEXT:    v_mov_b32_e32 v2, v0
 ; GFX6789-NEXT:    v_mov_b32_e32 v0, 0
 ; GFX6789-NEXT:    v_mov_b32_e32 v1, v0
 ; GFX6789-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; GFX6789-NEXT:    image_sample v[0:1], v2, s[0:7], s[8:11] dmask:0x2 tfe
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: sample_1d_tfe_adjust_writemask_2:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    s_mov_b32 s12, exec_lo ; encoding: [0x7e,0x03,0x8c,0xbe]
-; GFX10-NEXT:    s_wqm_b32 exec_lo, exec_lo ; encoding: [0x7e,0x09,0xfe,0xbe]
-; GFX10-NEXT:    v_mov_b32_e32 v2, v0 ; encoding: [0x00,0x03,0x04,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v0, 0 ; encoding: [0x80,0x02,0x00,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v1, v0 ; encoding: [0x00,0x03,0x02,0x7e]
-; GFX10-NEXT:    s_and_b32 exec_lo, exec_lo, s12 ; encoding: [0x7e,0x0c,0x7e,0x87]
-; GFX10-NEXT:    image_sample v[0:1], v2, s[0:7], s[8:11] dmask:0x2 dim:SQ_RSRC_IMG_1D tfe ; encoding: [0x00,0x02,0x81,0xf0,0x02,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: sample_1d_tfe_adjust_writemask_2:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    s_mov_b32 s12, exec_lo
+; GFX10PLUS-NEXT:    s_wqm_b32 exec_lo, exec_lo
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v2, v0
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v0, 0
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v1, v0
+; GFX10PLUS-NEXT:    s_and_b32 exec_lo, exec_lo, s12
+; GFX10PLUS-NEXT:    image_sample v[0:1], v2, s[0:7], s[8:11] dmask:0x2 dim:SQ_RSRC_IMG_1D tfe
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call {<4 x float>,i32} @llvm.amdgcn.image.sample.1d.v4f32i32.f32(i32 15, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 1, i32 0)
   %res.vec = extractvalue {<4 x float>,i32} %v, 0
   %res.f = extractelement <4 x float> %res.vec, i32 1
   %res.err = extractvalue {<4 x float>,i32} %v, 1
   %res.errf = bitcast i32 %res.err to float
   %res.tmp = insertelement <2 x float> undef, float %res.f, i32 0
   %res = insertelement <2 x float> %res.tmp, float %res.errf, i32 1

 ; GFX6789-NEXT:    v_mov_b32_e32 v2, v0
 ; GFX6789-NEXT:    v_mov_b32_e32 v0, 0
 ; GFX6789-NEXT:    v_mov_b32_e32 v1, v0
 ; GFX6789-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; GFX6789-NEXT:    image_sample v[0:1], v2, s[0:7], s[8:11] dmask:0x4 tfe
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: sample_1d_tfe_adjust_writemask_3:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    s_mov_b32 s12, exec_lo ; encoding: [0x7e,0x03,0x8c,0xbe]
-; GFX10-NEXT:    s_wqm_b32 exec_lo, exec_lo ; encoding: [0x7e,0x09,0xfe,0xbe]
-; GFX10-NEXT:    v_mov_b32_e32 v2, v0 ; encoding: [0x00,0x03,0x04,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v0, 0 ; encoding: [0x80,0x02,0x00,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v1, v0 ; encoding: [0x00,0x03,0x02,0x7e]
-; GFX10-NEXT:    s_and_b32 exec_lo, exec_lo, s12 ; encoding: [0x7e,0x0c,0x7e,0x87]
-; GFX10-NEXT:    image_sample v[0:1], v2, s[0:7], s[8:11] dmask:0x4 dim:SQ_RSRC_IMG_1D tfe ; encoding: [0x00,0x04,0x81,0xf0,0x02,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: sample_1d_tfe_adjust_writemask_3:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    s_mov_b32 s12, exec_lo
+; GFX10PLUS-NEXT:    s_wqm_b32 exec_lo, exec_lo
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v2, v0
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v0, 0
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v1, v0
+; GFX10PLUS-NEXT:    s_and_b32 exec_lo, exec_lo, s12
+; GFX10PLUS-NEXT:    image_sample v[0:1], v2, s[0:7], s[8:11] dmask:0x4 dim:SQ_RSRC_IMG_1D tfe
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call {<4 x float>,i32} @llvm.amdgcn.image.sample.1d.v4f32i32.f32(i32 15, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 1, i32 0)
   %res.vec = extractvalue {<4 x float>,i32} %v, 0
   %res.f = extractelement <4 x float> %res.vec, i32 2
   %res.err = extractvalue {<4 x float>,i32} %v, 1
   %res.errf = bitcast i32 %res.err to float
   %res.tmp = insertelement <2 x float> undef, float %res.f, i32 0
   %res = insertelement <2 x float> %res.tmp, float %res.errf, i32 1

 ; GFX6789-NEXT:    v_mov_b32_e32 v2, v0
 ; GFX6789-NEXT:    v_mov_b32_e32 v0, 0
 ; GFX6789-NEXT:    v_mov_b32_e32 v1, v0
 ; GFX6789-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; GFX6789-NEXT:    image_sample v[0:1], v2, s[0:7], s[8:11] dmask:0x8 tfe
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: sample_1d_tfe_adjust_writemask_4:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    s_mov_b32 s12, exec_lo ; encoding: [0x7e,0x03,0x8c,0xbe]
-; GFX10-NEXT:    s_wqm_b32 exec_lo, exec_lo ; encoding: [0x7e,0x09,0xfe,0xbe]
-; GFX10-NEXT:    v_mov_b32_e32 v2, v0 ; encoding: [0x00,0x03,0x04,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v0, 0 ; encoding: [0x80,0x02,0x00,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v1, v0 ; encoding: [0x00,0x03,0x02,0x7e]
-; GFX10-NEXT:    s_and_b32 exec_lo, exec_lo, s12 ; encoding: [0x7e,0x0c,0x7e,0x87]
-; GFX10-NEXT:    image_sample v[0:1], v2, s[0:7], s[8:11] dmask:0x8 dim:SQ_RSRC_IMG_1D tfe ; encoding: [0x00,0x08,0x81,0xf0,0x02,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: sample_1d_tfe_adjust_writemask_4:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    s_mov_b32 s12, exec_lo
+; GFX10PLUS-NEXT:    s_wqm_b32 exec_lo, exec_lo
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v2, v0
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v0, 0
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v1, v0
+; GFX10PLUS-NEXT:    s_and_b32 exec_lo, exec_lo, s12
+; GFX10PLUS-NEXT:    image_sample v[0:1], v2, s[0:7], s[8:11] dmask:0x8 dim:SQ_RSRC_IMG_1D tfe
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call {<4 x float>,i32} @llvm.amdgcn.image.sample.1d.v4f32i32.f32(i32 15, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 1, i32 0)
   %res.vec = extractvalue {<4 x float>,i32} %v, 0
   %res.f = extractelement <4 x float> %res.vec, i32 3
   %res.err = extractvalue {<4 x float>,i32} %v, 1
   %res.errf = bitcast i32 %res.err to float
   %res.tmp = insertelement <2 x float> undef, float %res.f, i32 0
   %res = insertelement <2 x float> %res.tmp, float %res.errf, i32 1

 ; GFX6789-NEXT:    v_mov_b32_e32 v0, 0
 ; GFX6789-NEXT:    v_mov_b32_e32 v1, v0
 ; GFX6789-NEXT:    v_mov_b32_e32 v2, v0
 ; GFX6789-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; GFX6789-NEXT:    image_sample v[0:2], v3, s[0:7], s[8:11] dmask:0x3 tfe
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: sample_1d_tfe_adjust_writemask_12:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    s_mov_b32 s12, exec_lo ; encoding: [0x7e,0x03,0x8c,0xbe]
-; GFX10-NEXT:    s_wqm_b32 exec_lo, exec_lo ; encoding: [0x7e,0x09,0xfe,0xbe]
-; GFX10-NEXT:    v_mov_b32_e32 v3, v0 ; encoding: [0x00,0x03,0x06,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v0, 0 ; encoding: [0x80,0x02,0x00,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v1, v0 ; encoding: [0x00,0x03,0x02,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v2, v0 ; encoding: [0x00,0x03,0x04,0x7e]
-; GFX10-NEXT:    s_and_b32 exec_lo, exec_lo, s12 ; encoding: [0x7e,0x0c,0x7e,0x87]
-; GFX10-NEXT:    image_sample v[0:2], v3, s[0:7], s[8:11] dmask:0x3 dim:SQ_RSRC_IMG_1D tfe ; encoding: [0x00,0x03,0x81,0xf0,0x03,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: sample_1d_tfe_adjust_writemask_12:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    s_mov_b32 s12, exec_lo
+; GFX10PLUS-NEXT:    s_wqm_b32 exec_lo, exec_lo
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v3, v0
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v0, 0
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v1, v0
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v2, v0
+; GFX10PLUS-NEXT:    s_and_b32 exec_lo, exec_lo, s12
+; GFX10PLUS-NEXT:    image_sample v[0:2], v3, s[0:7], s[8:11] dmask:0x3 dim:SQ_RSRC_IMG_1D tfe
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call {<4 x float>,i32} @llvm.amdgcn.image.sample.1d.v4f32i32.f32(i32 15, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 1, i32 0)
   %res.vec = extractvalue {<4 x float>,i32} %v, 0
   %res.f1 = extractelement <4 x float> %res.vec, i32 0
   %res.f2 = extractelement <4 x float> %res.vec, i32 1
   %res.err = extractvalue {<4 x float>,i32} %v, 1
   %res.errf = bitcast i32 %res.err to float
   %res.tmp1 = insertelement <4 x float> undef, float %res.f1, i32 0

 ; GFX6789-NEXT:    v_mov_b32_e32 v0, 0
 ; GFX6789-NEXT:    v_mov_b32_e32 v1, v0
 ; GFX6789-NEXT:    v_mov_b32_e32 v2, v0
 ; GFX6789-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; GFX6789-NEXT:    image_sample v[0:2], v3, s[0:7], s[8:11] dmask:0xa tfe
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: sample_1d_tfe_adjust_writemask_24:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    s_mov_b32 s12, exec_lo ; encoding: [0x7e,0x03,0x8c,0xbe]
-; GFX10-NEXT:    s_wqm_b32 exec_lo, exec_lo ; encoding: [0x7e,0x09,0xfe,0xbe]
-; GFX10-NEXT:    v_mov_b32_e32 v3, v0 ; encoding: [0x00,0x03,0x06,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v0, 0 ; encoding: [0x80,0x02,0x00,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v1, v0 ; encoding: [0x00,0x03,0x02,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v2, v0 ; encoding: [0x00,0x03,0x04,0x7e]
-; GFX10-NEXT:    s_and_b32 exec_lo, exec_lo, s12 ; encoding: [0x7e,0x0c,0x7e,0x87]
-; GFX10-NEXT:    image_sample v[0:2], v3, s[0:7], s[8:11] dmask:0xa dim:SQ_RSRC_IMG_1D tfe ; encoding: [0x00,0x0a,0x81,0xf0,0x03,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: sample_1d_tfe_adjust_writemask_24:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    s_mov_b32 s12, exec_lo
+; GFX10PLUS-NEXT:    s_wqm_b32 exec_lo, exec_lo
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v3, v0
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v0, 0
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v1, v0
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v2, v0
+; GFX10PLUS-NEXT:    s_and_b32 exec_lo, exec_lo, s12
+; GFX10PLUS-NEXT:    image_sample v[0:2], v3, s[0:7], s[8:11] dmask:0xa dim:SQ_RSRC_IMG_1D tfe
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call {<4 x float>,i32} @llvm.amdgcn.image.sample.1d.v4f32i32.f32(i32 15, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 1, i32 0)
   %res.vec = extractvalue {<4 x float>,i32} %v, 0
   %res.f1 = extractelement <4 x float> %res.vec, i32 1
   %res.f2 = extractelement <4 x float> %res.vec, i32 3
   %res.err = extractvalue {<4 x float>,i32} %v, 1
   %res.errf = bitcast i32 %res.err to float
   %res.tmp1 = insertelement <4 x float> undef, float %res.f1, i32 0

 ; GFX6789-NEXT:    v_mov_b32_e32 v1, v0
 ; GFX6789-NEXT:    v_mov_b32_e32 v2, v0
 ; GFX6789-NEXT:    v_mov_b32_e32 v3, v0
 ; GFX6789-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; GFX6789-NEXT:    image_sample v[0:3], v4, s[0:7], s[8:11] dmask:0xd tfe
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: sample_1d_tfe_adjust_writemask_134:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    s_mov_b32 s12, exec_lo ; encoding: [0x7e,0x03,0x8c,0xbe]
-; GFX10-NEXT:    s_wqm_b32 exec_lo, exec_lo ; encoding: [0x7e,0x09,0xfe,0xbe]
-; GFX10-NEXT:    v_mov_b32_e32 v4, v0 ; encoding: [0x00,0x03,0x08,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v0, 0 ; encoding: [0x80,0x02,0x00,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v1, v0 ; encoding: [0x00,0x03,0x02,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v2, v0 ; encoding: [0x00,0x03,0x04,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v3, v0 ; encoding: [0x00,0x03,0x06,0x7e]
-; GFX10-NEXT:    s_and_b32 exec_lo, exec_lo, s12 ; encoding: [0x7e,0x0c,0x7e,0x87]
-; GFX10-NEXT:    image_sample v[0:3], v4, s[0:7], s[8:11] dmask:0xd dim:SQ_RSRC_IMG_1D tfe ; encoding: [0x00,0x0d,0x81,0xf0,0x04,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: sample_1d_tfe_adjust_writemask_134:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    s_mov_b32 s12, exec_lo
+; GFX10PLUS-NEXT:    s_wqm_b32 exec_lo, exec_lo
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v4, v0
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v0, 0
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v1, v0
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v2, v0
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v3, v0
+; GFX10PLUS-NEXT:    s_and_b32 exec_lo, exec_lo, s12
+; GFX10PLUS-NEXT:    image_sample v[0:3], v4, s[0:7], s[8:11] dmask:0xd dim:SQ_RSRC_IMG_1D tfe
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call {<4 x float>,i32} @llvm.amdgcn.image.sample.1d.v4f32i32.f32(i32 15, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 1, i32 0)
   %res.vec = extractvalue {<4 x float>,i32} %v, 0
   %res.f1 = extractelement <4 x float> %res.vec, i32 0
   %res.f2 = extractelement <4 x float> %res.vec, i32 2
   %res.f3 = extractelement <4 x float> %res.vec, i32 3
   %res.err = extractvalue {<4 x float>,i32} %v, 1
   %res.errf = bitcast i32 %res.err to float

 ; GFX6789-NEXT:    image_sample v[0:4], v5, s[0:7], s[8:11] dmask:0xf lwe
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    global_store_dword v6, v4, s[12:13]
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
 ; GFX10-LABEL: sample_1d_lwe:
 ; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    s_mov_b32 s14, exec_lo ; encoding: [0x7e,0x03,0x8e,0xbe]
-; GFX10-NEXT:    s_wqm_b32 exec_lo, exec_lo ; encoding: [0x7e,0x09,0xfe,0xbe]
-; GFX10-NEXT:    v_mov_b32_e32 v6, 0 ; encoding: [0x80,0x02,0x0c,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v5, v0 ; encoding: [0x00,0x03,0x0a,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v7, v6 ; encoding: [0x06,0x03,0x0e,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v8, v6 ; encoding: [0x06,0x03,0x10,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v9, v6 ; encoding: [0x06,0x03,0x12,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v10, v6 ; encoding: [0x06,0x03,0x14,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v0, v6 ; encoding: [0x06,0x03,0x00,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v1, v7 ; encoding: [0x07,0x03,0x02,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v2, v8 ; encoding: [0x08,0x03,0x04,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v3, v9 ; encoding: [0x09,0x03,0x06,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v4, v10 ; encoding: [0x0a,0x03,0x08,0x7e]
-; GFX10-NEXT:    s_and_b32 exec_lo, exec_lo, s14 ; encoding: [0x7e,0x0e,0x7e,0x87]
-; GFX10-NEXT:    image_sample v[0:4], v5, s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D lwe ; encoding: [0x00,0x0f,0x82,0xf0,0x05,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    global_store_dword v6, v4, s[12:13] ; encoding: [0x00,0x80,0x70,0xdc,0x06,0x04,0x0c,0x00]
-; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0 ; encoding: [0x00,0x00,0xfd,0xbb]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10-NEXT:    s_mov_b32 s14, exec_lo
+; GFX10-NEXT:    s_wqm_b32 exec_lo, exec_lo
+; GFX10-NEXT:    v_mov_b32_e32 v6, 0
+; GFX10-NEXT:    v_mov_b32_e32 v5, v0
+; GFX10-NEXT:    v_mov_b32_e32 v7, v6
+; GFX10-NEXT:    v_mov_b32_e32 v8, v6
+; GFX10-NEXT:    v_mov_b32_e32 v9, v6
+; GFX10-NEXT:    v_mov_b32_e32 v10, v6
+; GFX10-NEXT:    v_mov_b32_e32 v0, v6
+; GFX10-NEXT:    v_mov_b32_e32 v1, v7
+; GFX10-NEXT:    v_mov_b32_e32 v2, v8
+; GFX10-NEXT:    v_mov_b32_e32 v3, v9
+; GFX10-NEXT:    v_mov_b32_e32 v4, v10
+; GFX10-NEXT:    s_and_b32 exec_lo, exec_lo, s14
+; GFX10-NEXT:    image_sample v[0:4], v5, s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D lwe
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    global_store_dword v6, v4, s[12:13]
+; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: sample_1d_lwe:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    s_mov_b32 s14, exec_lo
+; GFX11-NEXT:    s_wqm_b32 exec_lo, exec_lo
+; GFX11-NEXT:    v_mov_b32_e32 v6, 0
+; GFX11-NEXT:    v_mov_b32_e32 v5, v0
+; GFX11-NEXT:    v_mov_b32_e32 v7, v6
+; GFX11-NEXT:    v_mov_b32_e32 v8, v6
+; GFX11-NEXT:    v_mov_b32_e32 v9, v6
+; GFX11-NEXT:    v_mov_b32_e32 v10, v6
+; GFX11-NEXT:    v_mov_b32_e32 v0, v6
+; GFX11-NEXT:    v_mov_b32_e32 v1, v7
+; GFX11-NEXT:    v_mov_b32_e32 v2, v8
+; GFX11-NEXT:    v_mov_b32_e32 v3, v9
+; GFX11-NEXT:    v_mov_b32_e32 v4, v10
+; GFX11-NEXT:    s_and_b32 exec_lo, exec_lo, s14
+; GFX11-NEXT:    image_sample v[0:4], v5, s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D lwe
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    global_store_b32 v6, v4, s[12:13]
+; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %v = call {<4 x float>,i32} @llvm.amdgcn.image.sample.1d.v4f32i32.f32(i32 15, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 2, i32 0)
   %v.vec = extractvalue {<4 x float>, i32} %v, 0
   %v.err = extractvalue {<4 x float>, i32} %v, 1
   store i32 %v.err, i32 addrspace(1)* %out, align 4
   ret <4 x float> %v.vec
 }
 

 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    s_mov_b64 s[12:13], exec
 ; GFX6789-NEXT:    s_wqm_b64 exec, exec
 ; GFX6789-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; GFX6789-NEXT:    image_sample v[0:3], v[0:1], s[0:7], s[8:11] dmask:0xf
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: sample_2d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    s_mov_b32 s12, exec_lo ; encoding: [0x7e,0x03,0x8c,0xbe]
-; GFX10-NEXT:    s_wqm_b32 exec_lo, exec_lo ; encoding: [0x7e,0x09,0xfe,0xbe]
-; GFX10-NEXT:    s_and_b32 exec_lo, exec_lo, s12 ; encoding: [0x7e,0x0c,0x7e,0x87]
-; GFX10-NEXT:    image_sample v[0:3], v[0:1], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D ; encoding: [0x08,0x0f,0x80,0xf0,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: sample_2d:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    s_mov_b32 s12, exec_lo
+; GFX10PLUS-NEXT:    s_wqm_b32 exec_lo, exec_lo
+; GFX10PLUS-NEXT:    s_and_b32 exec_lo, exec_lo, s12
+; GFX10PLUS-NEXT:    image_sample v[0:3], v[0:1], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.2d.v4f32.f32(i32 15, float %s, float %t, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @sample_3d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %s, float %t, float %r) {
 ; VERDE-LABEL: sample_3d:
 ; VERDE:       ; %bb.0: ; %main_body
 ; VERDE-NEXT:    s_mov_b64 s[12:13], exec
 ; VERDE-NEXT:    s_wqm_b64 exec, exec
 ; VERDE-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; VERDE-NEXT:    image_sample v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf
 ; VERDE-NEXT:    s_waitcnt vmcnt(0)
 ; VERDE-NEXT:    ; return to shader part epilog
 ;
 ; GFX6789-LABEL: sample_3d:
 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    s_mov_b64 s[12:13], exec
 ; GFX6789-NEXT:    s_wqm_b64 exec, exec
 ; GFX6789-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; GFX6789-NEXT:    image_sample v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: sample_3d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    s_mov_b32 s12, exec_lo ; encoding: [0x7e,0x03,0x8c,0xbe]
-; GFX10-NEXT:    s_wqm_b32 exec_lo, exec_lo ; encoding: [0x7e,0x09,0xfe,0xbe]
-; GFX10-NEXT:    s_and_b32 exec_lo, exec_lo, s12 ; encoding: [0x7e,0x0c,0x7e,0x87]
-; GFX10-NEXT:    image_sample v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_3D ; encoding: [0x10,0x0f,0x80,0xf0,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: sample_3d:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    s_mov_b32 s12, exec_lo
+; GFX10PLUS-NEXT:    s_wqm_b32 exec_lo, exec_lo
+; GFX10PLUS-NEXT:    s_and_b32 exec_lo, exec_lo, s12
+; GFX10PLUS-NEXT:    image_sample v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_3D
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.3d.v4f32.f32(i32 15, float %s, float %t, float %r, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @sample_cube(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %s, float %t, float %face) {
 ; VERDE-LABEL: sample_cube:
 ; VERDE:       ; %bb.0: ; %main_body
 ; VERDE-NEXT:    s_mov_b64 s[12:13], exec
 ; VERDE-NEXT:    s_wqm_b64 exec, exec
 ; VERDE-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; VERDE-NEXT:    image_sample v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf da
 ; VERDE-NEXT:    s_waitcnt vmcnt(0)
 ; VERDE-NEXT:    ; return to shader part epilog
 ;
 ; GFX6789-LABEL: sample_cube:
 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    s_mov_b64 s[12:13], exec
 ; GFX6789-NEXT:    s_wqm_b64 exec, exec
 ; GFX6789-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; GFX6789-NEXT:    image_sample v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf da
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: sample_cube:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    s_mov_b32 s12, exec_lo ; encoding: [0x7e,0x03,0x8c,0xbe]
-; GFX10-NEXT:    s_wqm_b32 exec_lo, exec_lo ; encoding: [0x7e,0x09,0xfe,0xbe]
-; GFX10-NEXT:    s_and_b32 exec_lo, exec_lo, s12 ; encoding: [0x7e,0x0c,0x7e,0x87]
-; GFX10-NEXT:    image_sample v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_CUBE ; encoding: [0x18,0x0f,0x80,0xf0,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: sample_cube:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    s_mov_b32 s12, exec_lo
+; GFX10PLUS-NEXT:    s_wqm_b32 exec_lo, exec_lo
+; GFX10PLUS-NEXT:    s_and_b32 exec_lo, exec_lo, s12
+; GFX10PLUS-NEXT:    image_sample v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_CUBE
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.cube.v4f32.f32(i32 15, float %s, float %t, float %face, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @sample_1darray(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %s, float %slice) {
 ; VERDE-LABEL: sample_1darray:
 ; VERDE:       ; %bb.0: ; %main_body
 ; VERDE-NEXT:    s_mov_b64 s[12:13], exec
 ; VERDE-NEXT:    s_wqm_b64 exec, exec
 ; VERDE-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; VERDE-NEXT:    image_sample v[0:3], v[0:1], s[0:7], s[8:11] dmask:0xf da
 ; VERDE-NEXT:    s_waitcnt vmcnt(0)
 ; VERDE-NEXT:    ; return to shader part epilog
 ;
 ; GFX6789-LABEL: sample_1darray:
 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    s_mov_b64 s[12:13], exec
 ; GFX6789-NEXT:    s_wqm_b64 exec, exec
 ; GFX6789-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; GFX6789-NEXT:    image_sample v[0:3], v[0:1], s[0:7], s[8:11] dmask:0xf da
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: sample_1darray:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    s_mov_b32 s12, exec_lo ; encoding: [0x7e,0x03,0x8c,0xbe]
-; GFX10-NEXT:    s_wqm_b32 exec_lo, exec_lo ; encoding: [0x7e,0x09,0xfe,0xbe]
-; GFX10-NEXT:    s_and_b32 exec_lo, exec_lo, s12 ; encoding: [0x7e,0x0c,0x7e,0x87]
-; GFX10-NEXT:    image_sample v[0:3], v[0:1], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D_ARRAY ; encoding: [0x20,0x0f,0x80,0xf0,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: sample_1darray:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    s_mov_b32 s12, exec_lo
+; GFX10PLUS-NEXT:    s_wqm_b32 exec_lo, exec_lo
+; GFX10PLUS-NEXT:    s_and_b32 exec_lo, exec_lo, s12
+; GFX10PLUS-NEXT:    image_sample v[0:3], v[0:1], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D_ARRAY
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.1darray.v4f32.f32(i32 15, float %s, float %slice, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @sample_2darray(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %s, float %t, float %slice) {
 ; VERDE-LABEL: sample_2darray:
 ; VERDE:       ; %bb.0: ; %main_body
 ; VERDE-NEXT:    s_mov_b64 s[12:13], exec
 ; VERDE-NEXT:    s_wqm_b64 exec, exec
 ; VERDE-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; VERDE-NEXT:    image_sample v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf da
 ; VERDE-NEXT:    s_waitcnt vmcnt(0)
 ; VERDE-NEXT:    ; return to shader part epilog
 ;
 ; GFX6789-LABEL: sample_2darray:
 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    s_mov_b64 s[12:13], exec
 ; GFX6789-NEXT:    s_wqm_b64 exec, exec
 ; GFX6789-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; GFX6789-NEXT:    image_sample v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf da
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: sample_2darray:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    s_mov_b32 s12, exec_lo ; encoding: [0x7e,0x03,0x8c,0xbe]
-; GFX10-NEXT:    s_wqm_b32 exec_lo, exec_lo ; encoding: [0x7e,0x09,0xfe,0xbe]
-; GFX10-NEXT:    s_and_b32 exec_lo, exec_lo, s12 ; encoding: [0x7e,0x0c,0x7e,0x87]
-; GFX10-NEXT:    image_sample v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D_ARRAY ; encoding: [0x28,0x0f,0x80,0xf0,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: sample_2darray:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    s_mov_b32 s12, exec_lo
+; GFX10PLUS-NEXT:    s_wqm_b32 exec_lo, exec_lo
+; GFX10PLUS-NEXT:    s_and_b32 exec_lo, exec_lo, s12
+; GFX10PLUS-NEXT:    image_sample v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D_ARRAY
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.2darray.v4f32.f32(i32 15, float %s, float %t, float %slice, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @sample_c_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, float %s) {
 ; VERDE-LABEL: sample_c_1d:
 ; VERDE:       ; %bb.0: ; %main_body
 ; VERDE-NEXT:    s_mov_b64 s[12:13], exec
 ; VERDE-NEXT:    s_wqm_b64 exec, exec
 ; VERDE-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; VERDE-NEXT:    image_sample_c v[0:3], v[0:1], s[0:7], s[8:11] dmask:0xf
 ; VERDE-NEXT:    s_waitcnt vmcnt(0)
 ; VERDE-NEXT:    ; return to shader part epilog
 ;
 ; GFX6789-LABEL: sample_c_1d:
 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    s_mov_b64 s[12:13], exec
 ; GFX6789-NEXT:    s_wqm_b64 exec, exec
 ; GFX6789-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; GFX6789-NEXT:    image_sample_c v[0:3], v[0:1], s[0:7], s[8:11] dmask:0xf
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: sample_c_1d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    s_mov_b32 s12, exec_lo ; encoding: [0x7e,0x03,0x8c,0xbe]
-; GFX10-NEXT:    s_wqm_b32 exec_lo, exec_lo ; encoding: [0x7e,0x09,0xfe,0xbe]
-; GFX10-NEXT:    s_and_b32 exec_lo, exec_lo, s12 ; encoding: [0x7e,0x0c,0x7e,0x87]
-; GFX10-NEXT:    image_sample_c v[0:3], v[0:1], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D ; encoding: [0x00,0x0f,0xa0,0xf0,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: sample_c_1d:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    s_mov_b32 s12, exec_lo
+; GFX10PLUS-NEXT:    s_wqm_b32 exec_lo, exec_lo
+; GFX10PLUS-NEXT:    s_and_b32 exec_lo, exec_lo, s12
+; GFX10PLUS-NEXT:    image_sample_c v[0:3], v[0:1], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.c.1d.v4f32.f32(i32 15, float %zcompare, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @sample_c_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, float %s, float %t) {
 ; VERDE-LABEL: sample_c_2d:
 ; VERDE:       ; %bb.0: ; %main_body
 ; VERDE-NEXT:    s_mov_b64 s[12:13], exec
 ; VERDE-NEXT:    s_wqm_b64 exec, exec
 ; VERDE-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; VERDE-NEXT:    image_sample_c v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf
 ; VERDE-NEXT:    s_waitcnt vmcnt(0)
 ; VERDE-NEXT:    ; return to shader part epilog
 ;
 ; GFX6789-LABEL: sample_c_2d:
 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    s_mov_b64 s[12:13], exec
 ; GFX6789-NEXT:    s_wqm_b64 exec, exec
 ; GFX6789-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; GFX6789-NEXT:    image_sample_c v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: sample_c_2d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    s_mov_b32 s12, exec_lo ; encoding: [0x7e,0x03,0x8c,0xbe]
-; GFX10-NEXT:    s_wqm_b32 exec_lo, exec_lo ; encoding: [0x7e,0x09,0xfe,0xbe]
-; GFX10-NEXT:    s_and_b32 exec_lo, exec_lo, s12 ; encoding: [0x7e,0x0c,0x7e,0x87]
-; GFX10-NEXT:    image_sample_c v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D ; encoding: [0x08,0x0f,0xa0,0xf0,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: sample_c_2d:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    s_mov_b32 s12, exec_lo
+; GFX10PLUS-NEXT:    s_wqm_b32 exec_lo, exec_lo
+; GFX10PLUS-NEXT:    s_and_b32 exec_lo, exec_lo, s12
+; GFX10PLUS-NEXT:    image_sample_c v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.c.2d.v4f32.f32(i32 15, float %zcompare, float %s, float %t, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @sample_cl_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %s, float %clamp) {
 ; VERDE-LABEL: sample_cl_1d:
 ; VERDE:       ; %bb.0: ; %main_body
 ; VERDE-NEXT:    s_mov_b64 s[12:13], exec
 ; VERDE-NEXT:    s_wqm_b64 exec, exec
 ; VERDE-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; VERDE-NEXT:    image_sample_cl v[0:3], v[0:1], s[0:7], s[8:11] dmask:0xf
 ; VERDE-NEXT:    s_waitcnt vmcnt(0)
 ; VERDE-NEXT:    ; return to shader part epilog
 ;
 ; GFX6789-LABEL: sample_cl_1d:
 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    s_mov_b64 s[12:13], exec
 ; GFX6789-NEXT:    s_wqm_b64 exec, exec
 ; GFX6789-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; GFX6789-NEXT:    image_sample_cl v[0:3], v[0:1], s[0:7], s[8:11] dmask:0xf
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: sample_cl_1d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    s_mov_b32 s12, exec_lo ; encoding: [0x7e,0x03,0x8c,0xbe]
-; GFX10-NEXT:    s_wqm_b32 exec_lo, exec_lo ; encoding: [0x7e,0x09,0xfe,0xbe]
-; GFX10-NEXT:    s_and_b32 exec_lo, exec_lo, s12 ; encoding: [0x7e,0x0c,0x7e,0x87]
-; GFX10-NEXT:    image_sample_cl v[0:3], v[0:1], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D ; encoding: [0x00,0x0f,0x84,0xf0,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: sample_cl_1d:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    s_mov_b32 s12, exec_lo
+; GFX10PLUS-NEXT:    s_wqm_b32 exec_lo, exec_lo
+; GFX10PLUS-NEXT:    s_and_b32 exec_lo, exec_lo, s12
+; GFX10PLUS-NEXT:    image_sample_cl v[0:3], v[0:1], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.cl.1d.v4f32.f32(i32 15, float %s, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @sample_cl_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %s, float %t, float %clamp) {
 ; VERDE-LABEL: sample_cl_2d:
 ; VERDE:       ; %bb.0: ; %main_body
 ; VERDE-NEXT:    s_mov_b64 s[12:13], exec
 ; VERDE-NEXT:    s_wqm_b64 exec, exec
 ; VERDE-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; VERDE-NEXT:    image_sample_cl v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf
 ; VERDE-NEXT:    s_waitcnt vmcnt(0)
 ; VERDE-NEXT:    ; return to shader part epilog
 ;
 ; GFX6789-LABEL: sample_cl_2d:
 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    s_mov_b64 s[12:13], exec
 ; GFX6789-NEXT:    s_wqm_b64 exec, exec
 ; GFX6789-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; GFX6789-NEXT:    image_sample_cl v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: sample_cl_2d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    s_mov_b32 s12, exec_lo ; encoding: [0x7e,0x03,0x8c,0xbe]
-; GFX10-NEXT:    s_wqm_b32 exec_lo, exec_lo ; encoding: [0x7e,0x09,0xfe,0xbe]
-; GFX10-NEXT:    s_and_b32 exec_lo, exec_lo, s12 ; encoding: [0x7e,0x0c,0x7e,0x87]
-; GFX10-NEXT:    image_sample_cl v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D ; encoding: [0x08,0x0f,0x84,0xf0,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: sample_cl_2d:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    s_mov_b32 s12, exec_lo
+; GFX10PLUS-NEXT:    s_wqm_b32 exec_lo, exec_lo
+; GFX10PLUS-NEXT:    s_and_b32 exec_lo, exec_lo, s12
+; GFX10PLUS-NEXT:    image_sample_cl v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.cl.2d.v4f32.f32(i32 15, float %s, float %t, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @sample_c_cl_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, float %s, float %clamp) {
 ; VERDE-LABEL: sample_c_cl_1d:
 ; VERDE:       ; %bb.0: ; %main_body
 ; VERDE-NEXT:    s_mov_b64 s[12:13], exec
 ; VERDE-NEXT:    s_wqm_b64 exec, exec
 ; VERDE-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; VERDE-NEXT:    image_sample_c_cl v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf
 ; VERDE-NEXT:    s_waitcnt vmcnt(0)
 ; VERDE-NEXT:    ; return to shader part epilog
 ;
 ; GFX6789-LABEL: sample_c_cl_1d:
 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    s_mov_b64 s[12:13], exec
 ; GFX6789-NEXT:    s_wqm_b64 exec, exec
 ; GFX6789-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; GFX6789-NEXT:    image_sample_c_cl v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: sample_c_cl_1d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    s_mov_b32 s12, exec_lo ; encoding: [0x7e,0x03,0x8c,0xbe]
-; GFX10-NEXT:    s_wqm_b32 exec_lo, exec_lo ; encoding: [0x7e,0x09,0xfe,0xbe]
-; GFX10-NEXT:    s_and_b32 exec_lo, exec_lo, s12 ; encoding: [0x7e,0x0c,0x7e,0x87]
-; GFX10-NEXT:    image_sample_c_cl v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D ; encoding: [0x00,0x0f,0xa4,0xf0,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: sample_c_cl_1d:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    s_mov_b32 s12, exec_lo
+; GFX10PLUS-NEXT:    s_wqm_b32 exec_lo, exec_lo
+; GFX10PLUS-NEXT:    s_and_b32 exec_lo, exec_lo, s12
+; GFX10PLUS-NEXT:    image_sample_c_cl v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.c.cl.1d.v4f32.f32(i32 15, float %zcompare, float %s, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @sample_c_cl_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, float %s, float %t, float %clamp) {
 ; VERDE-LABEL: sample_c_cl_2d:
 ; VERDE:       ; %bb.0: ; %main_body
 ; VERDE-NEXT:    s_mov_b64 s[12:13], exec
 ; VERDE-NEXT:    s_wqm_b64 exec, exec
 ; VERDE-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; VERDE-NEXT:    image_sample_c_cl v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf
 ; VERDE-NEXT:    s_waitcnt vmcnt(0)
 ; VERDE-NEXT:    ; return to shader part epilog
 ;
 ; GFX6789-LABEL: sample_c_cl_2d:
 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    s_mov_b64 s[12:13], exec
 ; GFX6789-NEXT:    s_wqm_b64 exec, exec
 ; GFX6789-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; GFX6789-NEXT:    image_sample_c_cl v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: sample_c_cl_2d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    s_mov_b32 s12, exec_lo ; encoding: [0x7e,0x03,0x8c,0xbe]
-; GFX10-NEXT:    s_wqm_b32 exec_lo, exec_lo ; encoding: [0x7e,0x09,0xfe,0xbe]
-; GFX10-NEXT:    s_and_b32 exec_lo, exec_lo, s12 ; encoding: [0x7e,0x0c,0x7e,0x87]
-; GFX10-NEXT:    image_sample_c_cl v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D ; encoding: [0x08,0x0f,0xa4,0xf0,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: sample_c_cl_2d:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    s_mov_b32 s12, exec_lo
+; GFX10PLUS-NEXT:    s_wqm_b32 exec_lo, exec_lo
+; GFX10PLUS-NEXT:    s_and_b32 exec_lo, exec_lo, s12
+; GFX10PLUS-NEXT:    image_sample_c_cl v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.c.cl.2d.v4f32.f32(i32 15, float %zcompare, float %s, float %t, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @sample_b_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %bias, float %s) {
 ; VERDE-LABEL: sample_b_1d:
 ; VERDE:       ; %bb.0: ; %main_body
 ; VERDE-NEXT:    s_mov_b64 s[12:13], exec
 ; VERDE-NEXT:    s_wqm_b64 exec, exec
 ; VERDE-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; VERDE-NEXT:    image_sample_b v[0:3], v[0:1], s[0:7], s[8:11] dmask:0xf
 ; VERDE-NEXT:    s_waitcnt vmcnt(0)
 ; VERDE-NEXT:    ; return to shader part epilog
 ;
 ; GFX6789-LABEL: sample_b_1d:
 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    s_mov_b64 s[12:13], exec
 ; GFX6789-NEXT:    s_wqm_b64 exec, exec
 ; GFX6789-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; GFX6789-NEXT:    image_sample_b v[0:3], v[0:1], s[0:7], s[8:11] dmask:0xf
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: sample_b_1d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    s_mov_b32 s12, exec_lo ; encoding: [0x7e,0x03,0x8c,0xbe]
-; GFX10-NEXT:    s_wqm_b32 exec_lo, exec_lo ; encoding: [0x7e,0x09,0xfe,0xbe]
-; GFX10-NEXT:    s_and_b32 exec_lo, exec_lo, s12 ; encoding: [0x7e,0x0c,0x7e,0x87]
-; GFX10-NEXT:    image_sample_b v[0:3], v[0:1], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D ; encoding: [0x00,0x0f,0x94,0xf0,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: sample_b_1d:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    s_mov_b32 s12, exec_lo
+; GFX10PLUS-NEXT:    s_wqm_b32 exec_lo, exec_lo
+; GFX10PLUS-NEXT:    s_and_b32 exec_lo, exec_lo, s12
+; GFX10PLUS-NEXT:    image_sample_b v[0:3], v[0:1], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.b.1d.v4f32.f32.f32(i32 15, float %bias, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @sample_b_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %bias, float %s, float %t) {
 ; VERDE-LABEL: sample_b_2d:
 ; VERDE:       ; %bb.0: ; %main_body
 ; VERDE-NEXT:    s_mov_b64 s[12:13], exec
 ; VERDE-NEXT:    s_wqm_b64 exec, exec
 ; VERDE-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; VERDE-NEXT:    image_sample_b v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf
 ; VERDE-NEXT:    s_waitcnt vmcnt(0)
 ; VERDE-NEXT:    ; return to shader part epilog
 ;
 ; GFX6789-LABEL: sample_b_2d:
 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    s_mov_b64 s[12:13], exec
 ; GFX6789-NEXT:    s_wqm_b64 exec, exec
 ; GFX6789-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; GFX6789-NEXT:    image_sample_b v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: sample_b_2d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    s_mov_b32 s12, exec_lo ; encoding: [0x7e,0x03,0x8c,0xbe]
-; GFX10-NEXT:    s_wqm_b32 exec_lo, exec_lo ; encoding: [0x7e,0x09,0xfe,0xbe]
-; GFX10-NEXT:    s_and_b32 exec_lo, exec_lo, s12 ; encoding: [0x7e,0x0c,0x7e,0x87]
-; GFX10-NEXT:    image_sample_b v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D ; encoding: [0x08,0x0f,0x94,0xf0,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: sample_b_2d:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    s_mov_b32 s12, exec_lo
+; GFX10PLUS-NEXT:    s_wqm_b32 exec_lo, exec_lo
+; GFX10PLUS-NEXT:    s_and_b32 exec_lo, exec_lo, s12
+; GFX10PLUS-NEXT:    image_sample_b v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.b.2d.v4f32.f32.f32(i32 15, float %bias, float %s, float %t, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @sample_c_b_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %bias, float %zcompare, float %s) {
 ; VERDE-LABEL: sample_c_b_1d:
 ; VERDE:       ; %bb.0: ; %main_body
 ; VERDE-NEXT:    s_mov_b64 s[12:13], exec
 ; VERDE-NEXT:    s_wqm_b64 exec, exec
 ; VERDE-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; VERDE-NEXT:    image_sample_c_b v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf
 ; VERDE-NEXT:    s_waitcnt vmcnt(0)
 ; VERDE-NEXT:    ; return to shader part epilog
 ;
 ; GFX6789-LABEL: sample_c_b_1d:
 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    s_mov_b64 s[12:13], exec
 ; GFX6789-NEXT:    s_wqm_b64 exec, exec
 ; GFX6789-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; GFX6789-NEXT:    image_sample_c_b v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: sample_c_b_1d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    s_mov_b32 s12, exec_lo ; encoding: [0x7e,0x03,0x8c,0xbe]
-; GFX10-NEXT:    s_wqm_b32 exec_lo, exec_lo ; encoding: [0x7e,0x09,0xfe,0xbe]
-; GFX10-NEXT:    s_and_b32 exec_lo, exec_lo, s12 ; encoding: [0x7e,0x0c,0x7e,0x87]
-; GFX10-NEXT:    image_sample_c_b v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D ; encoding: [0x00,0x0f,0xb4,0xf0,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: sample_c_b_1d:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    s_mov_b32 s12, exec_lo
+; GFX10PLUS-NEXT:    s_wqm_b32 exec_lo, exec_lo
+; GFX10PLUS-NEXT:    s_and_b32 exec_lo, exec_lo, s12
+; GFX10PLUS-NEXT:    image_sample_c_b v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.c.b.1d.v4f32.f32.f32(i32 15, float %bias, float %zcompare, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @sample_c_b_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %bias, float %zcompare, float %s, float %t) {
 ; VERDE-LABEL: sample_c_b_2d:
 ; VERDE:       ; %bb.0: ; %main_body
 ; VERDE-NEXT:    s_mov_b64 s[12:13], exec
 ; VERDE-NEXT:    s_wqm_b64 exec, exec
 ; VERDE-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; VERDE-NEXT:    image_sample_c_b v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf
 ; VERDE-NEXT:    s_waitcnt vmcnt(0)
 ; VERDE-NEXT:    ; return to shader part epilog
 ;
 ; GFX6789-LABEL: sample_c_b_2d:
 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    s_mov_b64 s[12:13], exec
 ; GFX6789-NEXT:    s_wqm_b64 exec, exec
 ; GFX6789-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; GFX6789-NEXT:    image_sample_c_b v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: sample_c_b_2d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    s_mov_b32 s12, exec_lo ; encoding: [0x7e,0x03,0x8c,0xbe]
-; GFX10-NEXT:    s_wqm_b32 exec_lo, exec_lo ; encoding: [0x7e,0x09,0xfe,0xbe]
-; GFX10-NEXT:    s_and_b32 exec_lo, exec_lo, s12 ; encoding: [0x7e,0x0c,0x7e,0x87]
-; GFX10-NEXT:    image_sample_c_b v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D ; encoding: [0x08,0x0f,0xb4,0xf0,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: sample_c_b_2d:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    s_mov_b32 s12, exec_lo
+; GFX10PLUS-NEXT:    s_wqm_b32 exec_lo, exec_lo
+; GFX10PLUS-NEXT:    s_and_b32 exec_lo, exec_lo, s12
+; GFX10PLUS-NEXT:    image_sample_c_b v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.c.b.2d.v4f32.f32.f32(i32 15, float %bias, float %zcompare, float %s, float %t, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @sample_b_cl_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %bias, float %s, float %clamp) {
 ; VERDE-LABEL: sample_b_cl_1d:
 ; VERDE:       ; %bb.0: ; %main_body
 ; VERDE-NEXT:    s_mov_b64 s[12:13], exec
 ; VERDE-NEXT:    s_wqm_b64 exec, exec
 ; VERDE-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; VERDE-NEXT:    image_sample_b_cl v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf
 ; VERDE-NEXT:    s_waitcnt vmcnt(0)
 ; VERDE-NEXT:    ; return to shader part epilog
 ;
 ; GFX6789-LABEL: sample_b_cl_1d:
 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    s_mov_b64 s[12:13], exec
 ; GFX6789-NEXT:    s_wqm_b64 exec, exec
 ; GFX6789-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; GFX6789-NEXT:    image_sample_b_cl v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: sample_b_cl_1d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    s_mov_b32 s12, exec_lo ; encoding: [0x7e,0x03,0x8c,0xbe]
-; GFX10-NEXT:    s_wqm_b32 exec_lo, exec_lo ; encoding: [0x7e,0x09,0xfe,0xbe]
-; GFX10-NEXT:    s_and_b32 exec_lo, exec_lo, s12 ; encoding: [0x7e,0x0c,0x7e,0x87]
-; GFX10-NEXT:    image_sample_b_cl v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D ; encoding: [0x00,0x0f,0x98,0xf0,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: sample_b_cl_1d:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    s_mov_b32 s12, exec_lo
+; GFX10PLUS-NEXT:    s_wqm_b32 exec_lo, exec_lo
+; GFX10PLUS-NEXT:    s_and_b32 exec_lo, exec_lo, s12
+; GFX10PLUS-NEXT:    image_sample_b_cl v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.b.cl.1d.v4f32.f32.f32(i32 15, float %bias, float %s, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @sample_b_cl_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %bias, float %s, float %t, float %clamp) {
 ; VERDE-LABEL: sample_b_cl_2d:
 ; VERDE:       ; %bb.0: ; %main_body
 ; VERDE-NEXT:    s_mov_b64 s[12:13], exec
 ; VERDE-NEXT:    s_wqm_b64 exec, exec
 ; VERDE-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; VERDE-NEXT:    image_sample_b_cl v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf
 ; VERDE-NEXT:    s_waitcnt vmcnt(0)
 ; VERDE-NEXT:    ; return to shader part epilog
 ;
 ; GFX6789-LABEL: sample_b_cl_2d:
 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    s_mov_b64 s[12:13], exec
 ; GFX6789-NEXT:    s_wqm_b64 exec, exec
 ; GFX6789-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; GFX6789-NEXT:    image_sample_b_cl v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: sample_b_cl_2d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    s_mov_b32 s12, exec_lo ; encoding: [0x7e,0x03,0x8c,0xbe]
-; GFX10-NEXT:    s_wqm_b32 exec_lo, exec_lo ; encoding: [0x7e,0x09,0xfe,0xbe]
-; GFX10-NEXT:    s_and_b32 exec_lo, exec_lo, s12 ; encoding: [0x7e,0x0c,0x7e,0x87]
-; GFX10-NEXT:    image_sample_b_cl v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D ; encoding: [0x08,0x0f,0x98,0xf0,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: sample_b_cl_2d:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    s_mov_b32 s12, exec_lo
+; GFX10PLUS-NEXT:    s_wqm_b32 exec_lo, exec_lo
+; GFX10PLUS-NEXT:    s_and_b32 exec_lo, exec_lo, s12
+; GFX10PLUS-NEXT:    image_sample_b_cl v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.b.cl.2d.v4f32.f32.f32(i32 15, float %bias, float %s, float %t, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @sample_c_b_cl_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %bias, float %zcompare, float %s, float %clamp) {
 ; VERDE-LABEL: sample_c_b_cl_1d:
 ; VERDE:       ; %bb.0: ; %main_body
 ; VERDE-NEXT:    s_mov_b64 s[12:13], exec
 ; VERDE-NEXT:    s_wqm_b64 exec, exec
 ; VERDE-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; VERDE-NEXT:    image_sample_c_b_cl v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf
 ; VERDE-NEXT:    s_waitcnt vmcnt(0)
 ; VERDE-NEXT:    ; return to shader part epilog
 ;
 ; GFX6789-LABEL: sample_c_b_cl_1d:
 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    s_mov_b64 s[12:13], exec
 ; GFX6789-NEXT:    s_wqm_b64 exec, exec
 ; GFX6789-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; GFX6789-NEXT:    image_sample_c_b_cl v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: sample_c_b_cl_1d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    s_mov_b32 s12, exec_lo ; encoding: [0x7e,0x03,0x8c,0xbe]
-; GFX10-NEXT:    s_wqm_b32 exec_lo, exec_lo ; encoding: [0x7e,0x09,0xfe,0xbe]
-; GFX10-NEXT:    s_and_b32 exec_lo, exec_lo, s12 ; encoding: [0x7e,0x0c,0x7e,0x87]
-; GFX10-NEXT:    image_sample_c_b_cl v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D ; encoding: [0x00,0x0f,0xb8,0xf0,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: sample_c_b_cl_1d:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    s_mov_b32 s12, exec_lo
+; GFX10PLUS-NEXT:    s_wqm_b32 exec_lo, exec_lo
+; GFX10PLUS-NEXT:    s_and_b32 exec_lo, exec_lo, s12
+; GFX10PLUS-NEXT:    image_sample_c_b_cl v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.c.b.cl.1d.v4f32.f32.f32(i32 15, float %bias, float %zcompare, float %s, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @sample_c_b_cl_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %bias, float %zcompare, float %s, float %t, float %clamp) {
 ; VERDE-LABEL: sample_c_b_cl_2d:
 ; VERDE:       ; %bb.0: ; %main_body
 ; VERDE-NEXT:    s_mov_b64 s[12:13], exec
 ; VERDE-NEXT:    s_wqm_b64 exec, exec
 ; VERDE-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; VERDE-NEXT:    image_sample_c_b_cl v[0:3], v[0:4], s[0:7], s[8:11] dmask:0xf
 ; VERDE-NEXT:    s_waitcnt vmcnt(0)
 ; VERDE-NEXT:    ; return to shader part epilog
 ;
 ; GFX6789-LABEL: sample_c_b_cl_2d:
 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    s_mov_b64 s[12:13], exec
 ; GFX6789-NEXT:    s_wqm_b64 exec, exec
 ; GFX6789-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; GFX6789-NEXT:    image_sample_c_b_cl v[0:3], v[0:4], s[0:7], s[8:11] dmask:0xf
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: sample_c_b_cl_2d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    s_mov_b32 s12, exec_lo ; encoding: [0x7e,0x03,0x8c,0xbe]
-; GFX10-NEXT:    s_wqm_b32 exec_lo, exec_lo ; encoding: [0x7e,0x09,0xfe,0xbe]
-; GFX10-NEXT:    s_and_b32 exec_lo, exec_lo, s12 ; encoding: [0x7e,0x0c,0x7e,0x87]
-; GFX10-NEXT:    image_sample_c_b_cl v[0:3], v[0:4], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D ; encoding: [0x08,0x0f,0xb8,0xf0,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: sample_c_b_cl_2d:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    s_mov_b32 s12, exec_lo
+; GFX10PLUS-NEXT:    s_wqm_b32 exec_lo, exec_lo
+; GFX10PLUS-NEXT:    s_and_b32 exec_lo, exec_lo, s12
+; GFX10PLUS-NEXT:    image_sample_c_b_cl v[0:3], v[0:4], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.c.b.cl.2d.v4f32.f32.f32(i32 15, float %bias, float %zcompare, float %s, float %t, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @sample_d_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %dsdh, float %dsdv, float %s) {
 ; VERDE-LABEL: sample_d_1d:
 ; VERDE:       ; %bb.0: ; %main_body
 ; VERDE-NEXT:    image_sample_d v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf
 ; VERDE-NEXT:    s_waitcnt vmcnt(0)
 ; VERDE-NEXT:    ; return to shader part epilog
 ;
 ; GFX6789-LABEL: sample_d_1d:
 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    image_sample_d v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: sample_d_1d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_sample_d v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D ; encoding: [0x00,0x0f,0x88,0xf0,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: sample_d_1d:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_sample_d v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.d.1d.v4f32.f32.f32(i32 15, float %dsdh, float %dsdv, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @sample_d_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %dsdh, float %dtdh, float %dsdv, float %dtdv, float %s, float %t) {
 ; VERDE-LABEL: sample_d_2d:
 ; VERDE:       ; %bb.0: ; %main_body
 ; VERDE-NEXT:    image_sample_d v[0:3], v[0:5], s[0:7], s[8:11] dmask:0xf
 ; VERDE-NEXT:    s_waitcnt vmcnt(0)
 ; VERDE-NEXT:    ; return to shader part epilog
 ;
 ; GFX6789-LABEL: sample_d_2d:
 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    image_sample_d v[0:3], v[0:5], s[0:7], s[8:11] dmask:0xf
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: sample_d_2d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_sample_d v[0:3], v[0:5], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D ; encoding: [0x08,0x0f,0x88,0xf0,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: sample_d_2d:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_sample_d v[0:3], v[0:5], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.d.2d.v4f32.f32.f32(i32 15, float %dsdh, float %dtdh, float %dsdv, float %dtdv, float %s, float %t, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @sample_c_d_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, float %dsdh, float %dsdv, float %s) {
 ; VERDE-LABEL: sample_c_d_1d:
 ; VERDE:       ; %bb.0: ; %main_body
 ; VERDE-NEXT:    image_sample_c_d v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf
 ; VERDE-NEXT:    s_waitcnt vmcnt(0)
 ; VERDE-NEXT:    ; return to shader part epilog
 ;
 ; GFX6789-LABEL: sample_c_d_1d:
 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    image_sample_c_d v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: sample_c_d_1d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_sample_c_d v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D ; encoding: [0x00,0x0f,0xa8,0xf0,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: sample_c_d_1d:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_sample_c_d v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.c.d.1d.v4f32.f32.f32(i32 15, float %zcompare, float %dsdh, float %dsdv, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @sample_c_d_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, float %dsdh, float %dtdh, float %dsdv, float %dtdv, float %s, float %t) {
 ; VERDE-LABEL: sample_c_d_2d:
 ; VERDE:       ; %bb.0: ; %main_body
 ; VERDE-NEXT:    image_sample_c_d v[0:3], v[0:6], s[0:7], s[8:11] dmask:0xf
 ; VERDE-NEXT:    s_waitcnt vmcnt(0)
 ; VERDE-NEXT:    ; return to shader part epilog
 ;
 ; GFX6789-LABEL: sample_c_d_2d:
 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    image_sample_c_d v[0:3], v[0:6], s[0:7], s[8:11] dmask:0xf
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: sample_c_d_2d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_sample_c_d v[0:3], v[0:6], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D ; encoding: [0x08,0x0f,0xa8,0xf0,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: sample_c_d_2d:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_sample_c_d v[0:3], v[0:6], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.c.d.2d.v4f32.f32.f32(i32 15, float %zcompare, float %dsdh, float %dtdh, float %dsdv, float %dtdv, float %s, float %t, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @sample_d_cl_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %dsdh, float %dsdv, float %s, float %clamp) {
 ; VERDE-LABEL: sample_d_cl_1d:
 ; VERDE:       ; %bb.0: ; %main_body
 ; VERDE-NEXT:    image_sample_d_cl v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf
 ; VERDE-NEXT:    s_waitcnt vmcnt(0)
 ; VERDE-NEXT:    ; return to shader part epilog
 ;
 ; GFX6789-LABEL: sample_d_cl_1d:
 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    image_sample_d_cl v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: sample_d_cl_1d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_sample_d_cl v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D ; encoding: [0x00,0x0f,0x8c,0xf0,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: sample_d_cl_1d:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_sample_d_cl v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.d.cl.1d.v4f32.f32.f32(i32 15, float %dsdh, float %dsdv, float %s, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @sample_d_cl_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %dsdh, float %dtdh, float %dsdv, float %dtdv, float %s, float %t, float %clamp) {
 ; VERDE-LABEL: sample_d_cl_2d:
 ; VERDE:       ; %bb.0: ; %main_body
 ; VERDE-NEXT:    image_sample_d_cl v[0:3], v[0:6], s[0:7], s[8:11] dmask:0xf
 ; VERDE-NEXT:    s_waitcnt vmcnt(0)
 ; VERDE-NEXT:    ; return to shader part epilog
 ;
 ; GFX6789-LABEL: sample_d_cl_2d:
 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    image_sample_d_cl v[0:3], v[0:6], s[0:7], s[8:11] dmask:0xf
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: sample_d_cl_2d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_sample_d_cl v[0:3], v[0:6], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D ; encoding: [0x08,0x0f,0x8c,0xf0,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: sample_d_cl_2d:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_sample_d_cl v[0:3], v[0:6], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.d.cl.2d.v4f32.f32.f32(i32 15, float %dsdh, float %dtdh, float %dsdv, float %dtdv, float %s, float %t, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @sample_c_d_cl_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, float %dsdh, float %dsdv, float %s, float %clamp) {
 ; VERDE-LABEL: sample_c_d_cl_1d:
 ; VERDE:       ; %bb.0: ; %main_body
 ; VERDE-NEXT:    image_sample_c_d_cl v[0:3], v[0:4], s[0:7], s[8:11] dmask:0xf
 ; VERDE-NEXT:    s_waitcnt vmcnt(0)
 ; VERDE-NEXT:    ; return to shader part epilog
 ;
 ; GFX6789-LABEL: sample_c_d_cl_1d:
 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    image_sample_c_d_cl v[0:3], v[0:4], s[0:7], s[8:11] dmask:0xf
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: sample_c_d_cl_1d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_sample_c_d_cl v[0:3], v[0:4], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D ; encoding: [0x00,0x0f,0xac,0xf0,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: sample_c_d_cl_1d:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_sample_c_d_cl v[0:3], v[0:4], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.c.d.cl.1d.v4f32.f32.f32(i32 15, float %zcompare, float %dsdh, float %dsdv, float %s, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @sample_c_d_cl_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, float %dsdh, float %dtdh, float %dsdv, float %dtdv, float %s, float %t, float %clamp) {
 ; VERDE-LABEL: sample_c_d_cl_2d:
 ; VERDE:       ; %bb.0: ; %main_body
 ; VERDE-NEXT:    image_sample_c_d_cl v[0:3], v[0:7], s[0:7], s[8:11] dmask:0xf
 ; VERDE-NEXT:    s_waitcnt vmcnt(0)
 ; VERDE-NEXT:    ; return to shader part epilog
 ;
 ; GFX6789-LABEL: sample_c_d_cl_2d:
 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    image_sample_c_d_cl v[0:3], v[0:7], s[0:7], s[8:11] dmask:0xf
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: sample_c_d_cl_2d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_sample_c_d_cl v[0:3], v[0:7], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D ; encoding: [0x08,0x0f,0xac,0xf0,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: sample_c_d_cl_2d:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_sample_c_d_cl v[0:3], v[0:7], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.c.d.cl.2d.v4f32.f32.f32(i32 15, float %zcompare, float %dsdh, float %dtdh, float %dsdv, float %dtdv, float %s, float %t, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v
 }
 
-define amdgpu_ps <4 x float> @sample_cd_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %dsdh, float %dsdv, float %s) {
-; VERDE-LABEL: sample_cd_1d:
-; VERDE:       ; %bb.0: ; %main_body
-; VERDE-NEXT:    image_sample_cd v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf
-; VERDE-NEXT:    s_waitcnt vmcnt(0)
-; VERDE-NEXT:    ; return to shader part epilog
-;
-; GFX6789-LABEL: sample_cd_1d:
-; GFX6789:       ; %bb.0: ; %main_body
-; GFX6789-NEXT:    image_sample_cd v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf
-; GFX6789-NEXT:    s_waitcnt vmcnt(0)
-; GFX6789-NEXT:    ; return to shader part epilog
-;
-; GFX10-LABEL: sample_cd_1d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_sample_cd v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D ; encoding: [0x00,0x0f,0xa0,0xf1,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
-main_body:
-  %v = call <4 x float> @llvm.amdgcn.image.sample.cd.1d.v4f32.f32.f32(i32 15, float %dsdh, float %dsdv, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
-  ret <4 x float> %v
-}
-
-define amdgpu_ps <4 x float> @sample_cd_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %dsdh, float %dtdh, float %dsdv, float %dtdv, float %s, float %t) {
-; VERDE-LABEL: sample_cd_2d:
-; VERDE:       ; %bb.0: ; %main_body
-; VERDE-NEXT:    image_sample_cd v[0:3], v[0:5], s[0:7], s[8:11] dmask:0xf
-; VERDE-NEXT:    s_waitcnt vmcnt(0)
-; VERDE-NEXT:    ; return to shader part epilog
-;
-; GFX6789-LABEL: sample_cd_2d:
-; GFX6789:       ; %bb.0: ; %main_body
-; GFX6789-NEXT:    image_sample_cd v[0:3], v[0:5], s[0:7], s[8:11] dmask:0xf
-; GFX6789-NEXT:    s_waitcnt vmcnt(0)
-; GFX6789-NEXT:    ; return to shader part epilog
-;
-; GFX10-LABEL: sample_cd_2d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_sample_cd v[0:3], v[0:5], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D ; encoding: [0x08,0x0f,0xa0,0xf1,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
-main_body:
-  %v = call <4 x float> @llvm.amdgcn.image.sample.cd.2d.v4f32.f32.f32(i32 15, float %dsdh, float %dtdh, float %dsdv, float %dtdv, float %s, float %t, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
-  ret <4 x float> %v
-}
-
-define amdgpu_ps <4 x float> @sample_c_cd_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, float %dsdh, float %dsdv, float %s) {
-; VERDE-LABEL: sample_c_cd_1d:
-; VERDE:       ; %bb.0: ; %main_body
-; VERDE-NEXT:    image_sample_c_cd v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf
-; VERDE-NEXT:    s_waitcnt vmcnt(0)
-; VERDE-NEXT:    ; return to shader part epilog
-;
-; GFX6789-LABEL: sample_c_cd_1d:
-; GFX6789:       ; %bb.0: ; %main_body
-; GFX6789-NEXT:    image_sample_c_cd v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf
-; GFX6789-NEXT:    s_waitcnt vmcnt(0)
-; GFX6789-NEXT:    ; return to shader part epilog
-;
-; GFX10-LABEL: sample_c_cd_1d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_sample_c_cd v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D ; encoding: [0x00,0x0f,0xa8,0xf1,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
-main_body:
-  %v = call <4 x float> @llvm.amdgcn.image.sample.c.cd.1d.v4f32.f32.f32(i32 15, float %zcompare, float %dsdh, float %dsdv, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
-  ret <4 x float> %v
-}
-
-define amdgpu_ps <4 x float> @sample_c_cd_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, float %dsdh, float %dtdh, float %dsdv, float %dtdv, float %s, float %t) {
-; VERDE-LABEL: sample_c_cd_2d:
-; VERDE:       ; %bb.0: ; %main_body
-; VERDE-NEXT:    image_sample_c_cd v[0:3], v[0:6], s[0:7], s[8:11] dmask:0xf
-; VERDE-NEXT:    s_waitcnt vmcnt(0)
-; VERDE-NEXT:    ; return to shader part epilog
-;
-; GFX6789-LABEL: sample_c_cd_2d:
-; GFX6789:       ; %bb.0: ; %main_body
-; GFX6789-NEXT:    image_sample_c_cd v[0:3], v[0:6], s[0:7], s[8:11] dmask:0xf
-; GFX6789-NEXT:    s_waitcnt vmcnt(0)
-; GFX6789-NEXT:    ; return to shader part epilog
-;
-; GFX10-LABEL: sample_c_cd_2d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_sample_c_cd v[0:3], v[0:6], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D ; encoding: [0x08,0x0f,0xa8,0xf1,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
-main_body:
-  %v = call <4 x float> @llvm.amdgcn.image.sample.c.cd.2d.v4f32.f32.f32(i32 15, float %zcompare, float %dsdh, float %dtdh, float %dsdv, float %dtdv, float %s, float %t, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
-  ret <4 x float> %v
-}
-
-define amdgpu_ps <4 x float> @sample_cd_cl_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %dsdh, float %dsdv, float %s, float %clamp) {
-; VERDE-LABEL: sample_cd_cl_1d:
-; VERDE:       ; %bb.0: ; %main_body
-; VERDE-NEXT:    image_sample_cd_cl v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf
-; VERDE-NEXT:    s_waitcnt vmcnt(0)
-; VERDE-NEXT:    ; return to shader part epilog
-;
-; GFX6789-LABEL: sample_cd_cl_1d:
-; GFX6789:       ; %bb.0: ; %main_body
-; GFX6789-NEXT:    image_sample_cd_cl v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf
-; GFX6789-NEXT:    s_waitcnt vmcnt(0)
-; GFX6789-NEXT:    ; return to shader part epilog
-;
-; GFX10-LABEL: sample_cd_cl_1d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_sample_cd_cl v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D ; encoding: [0x00,0x0f,0xa4,0xf1,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
-main_body:
-  %v = call <4 x float> @llvm.amdgcn.image.sample.cd.cl.1d.v4f32.f32.f32(i32 15, float %dsdh, float %dsdv, float %s, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
-  ret <4 x float> %v
-}
-
-define amdgpu_ps <4 x float> @sample_cd_cl_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %dsdh, float %dtdh, float %dsdv, float %dtdv, float %s, float %t, float %clamp) {
-; VERDE-LABEL: sample_cd_cl_2d:
-; VERDE:       ; %bb.0: ; %main_body
-; VERDE-NEXT:    image_sample_cd_cl v[0:3], v[0:6], s[0:7], s[8:11] dmask:0xf
-; VERDE-NEXT:    s_waitcnt vmcnt(0)
-; VERDE-NEXT:    ; return to shader part epilog
-;
-; GFX6789-LABEL: sample_cd_cl_2d:
-; GFX6789:       ; %bb.0: ; %main_body
-; GFX6789-NEXT:    image_sample_cd_cl v[0:3], v[0:6], s[0:7], s[8:11] dmask:0xf
-; GFX6789-NEXT:    s_waitcnt vmcnt(0)
-; GFX6789-NEXT:    ; return to shader part epilog
-;
-; GFX10-LABEL: sample_cd_cl_2d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_sample_cd_cl v[0:3], v[0:6], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D ; encoding: [0x08,0x0f,0xa4,0xf1,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
-main_body:
-  %v = call <4 x float> @llvm.amdgcn.image.sample.cd.cl.2d.v4f32.f32.f32(i32 15, float %dsdh, float %dtdh, float %dsdv, float %dtdv, float %s, float %t, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
-  ret <4 x float> %v
-}
-
-define amdgpu_ps <4 x float> @sample_c_cd_cl_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, float %dsdh, float %dsdv, float %s, float %clamp) {
-; VERDE-LABEL: sample_c_cd_cl_1d:
-; VERDE:       ; %bb.0: ; %main_body
-; VERDE-NEXT:    image_sample_c_cd_cl v[0:3], v[0:4], s[0:7], s[8:11] dmask:0xf
-; VERDE-NEXT:    s_waitcnt vmcnt(0)
-; VERDE-NEXT:    ; return to shader part epilog
-;
-; GFX6789-LABEL: sample_c_cd_cl_1d:
-; GFX6789:       ; %bb.0: ; %main_body
-; GFX6789-NEXT:    image_sample_c_cd_cl v[0:3], v[0:4], s[0:7], s[8:11] dmask:0xf
-; GFX6789-NEXT:    s_waitcnt vmcnt(0)
-; GFX6789-NEXT:    ; return to shader part epilog
-;
-; GFX10-LABEL: sample_c_cd_cl_1d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_sample_c_cd_cl v[0:3], v[0:4], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D ; encoding: [0x00,0x0f,0xac,0xf1,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
-main_body:
-  %v = call <4 x float> @llvm.amdgcn.image.sample.c.cd.cl.1d.v4f32.f32.f32(i32 15, float %zcompare, float %dsdh, float %dsdv, float %s, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
-  ret <4 x float> %v
-}
-
-define amdgpu_ps <4 x float> @sample_c_cd_cl_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, float %dsdh, float %dtdh, float %dsdv, float %dtdv, float %s, float %t, float %clamp) {
-; VERDE-LABEL: sample_c_cd_cl_2d:
-; VERDE:       ; %bb.0: ; %main_body
-; VERDE-NEXT:    image_sample_c_cd_cl v[0:3], v[0:7], s[0:7], s[8:11] dmask:0xf
-; VERDE-NEXT:    s_waitcnt vmcnt(0)
-; VERDE-NEXT:    ; return to shader part epilog
-;
-; GFX6789-LABEL: sample_c_cd_cl_2d:
-; GFX6789:       ; %bb.0: ; %main_body
-; GFX6789-NEXT:    image_sample_c_cd_cl v[0:3], v[0:7], s[0:7], s[8:11] dmask:0xf
-; GFX6789-NEXT:    s_waitcnt vmcnt(0)
-; GFX6789-NEXT:    ; return to shader part epilog
-;
-; GFX10-LABEL: sample_c_cd_cl_2d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_sample_c_cd_cl v[0:3], v[0:7], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D ; encoding: [0x08,0x0f,0xac,0xf1,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
-main_body:
-  %v = call <4 x float> @llvm.amdgcn.image.sample.c.cd.cl.2d.v4f32.f32.f32(i32 15, float %zcompare, float %dsdh, float %dtdh, float %dsdv, float %dtdv, float %s, float %t, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
-  ret <4 x float> %v
-}
-
 define amdgpu_ps <4 x float> @sample_l_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %s, float %lod) {
 ; VERDE-LABEL: sample_l_1d:
 ; VERDE:       ; %bb.0: ; %main_body
 ; VERDE-NEXT:    image_sample_l v[0:3], v[0:1], s[0:7], s[8:11] dmask:0xf
 ; VERDE-NEXT:    s_waitcnt vmcnt(0)
 ; VERDE-NEXT:    ; return to shader part epilog
 ;
 ; GFX6789-LABEL: sample_l_1d:
 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    image_sample_l v[0:3], v[0:1], s[0:7], s[8:11] dmask:0xf
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: sample_l_1d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_sample_l v[0:3], v[0:1], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D ; encoding: [0x00,0x0f,0x90,0xf0,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: sample_l_1d:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_sample_l v[0:3], v[0:1], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.l.1d.v4f32.f32(i32 15, float %s, float %lod, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @sample_l_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %s, float %t, float %lod) {
 ; VERDE-LABEL: sample_l_2d:
 ; VERDE:       ; %bb.0: ; %main_body
 ; VERDE-NEXT:    image_sample_l v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf
 ; VERDE-NEXT:    s_waitcnt vmcnt(0)
 ; VERDE-NEXT:    ; return to shader part epilog
 ;
 ; GFX6789-LABEL: sample_l_2d:
 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    image_sample_l v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: sample_l_2d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_sample_l v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D ; encoding: [0x08,0x0f,0x90,0xf0,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: sample_l_2d:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_sample_l v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.l.2d.v4f32.f32(i32 15, float %s, float %t, float %lod, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @sample_c_l_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, float %s, float %lod) {
 ; VERDE-LABEL: sample_c_l_1d:
 ; VERDE:       ; %bb.0: ; %main_body
 ; VERDE-NEXT:    image_sample_c_l v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf
 ; VERDE-NEXT:    s_waitcnt vmcnt(0)
 ; VERDE-NEXT:    ; return to shader part epilog
 ;
 ; GFX6789-LABEL: sample_c_l_1d:
 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    image_sample_c_l v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: sample_c_l_1d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_sample_c_l v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D ; encoding: [0x00,0x0f,0xb0,0xf0,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: sample_c_l_1d:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_sample_c_l v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.c.l.1d.v4f32.f32(i32 15, float %zcompare, float %s, float %lod, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @sample_c_l_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, float %s, float %t, float %lod) {
 ; VERDE-LABEL: sample_c_l_2d:
 ; VERDE:       ; %bb.0: ; %main_body
 ; VERDE-NEXT:    image_sample_c_l v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf
 ; VERDE-NEXT:    s_waitcnt vmcnt(0)
 ; VERDE-NEXT:    ; return to shader part epilog
 ;
 ; GFX6789-LABEL: sample_c_l_2d:
 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    image_sample_c_l v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: sample_c_l_2d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_sample_c_l v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D ; encoding: [0x08,0x0f,0xb0,0xf0,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: sample_c_l_2d:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_sample_c_l v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.c.l.2d.v4f32.f32(i32 15, float %zcompare, float %s, float %t, float %lod, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @sample_lz_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %s) {
 ; VERDE-LABEL: sample_lz_1d:
 ; VERDE:       ; %bb.0: ; %main_body
 ; VERDE-NEXT:    image_sample_lz v[0:3], v0, s[0:7], s[8:11] dmask:0xf
 ; VERDE-NEXT:    s_waitcnt vmcnt(0)
 ; VERDE-NEXT:    ; return to shader part epilog
 ;
 ; GFX6789-LABEL: sample_lz_1d:
 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    image_sample_lz v[0:3], v0, s[0:7], s[8:11] dmask:0xf
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: sample_lz_1d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_sample_lz v[0:3], v0, s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D ; encoding: [0x00,0x0f,0x9c,0xf0,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: sample_lz_1d:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_sample_lz v[0:3], v0, s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.lz.1d.v4f32.f32(i32 15, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @sample_lz_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %s, float %t) {
 ; VERDE-LABEL: sample_lz_2d:
 ; VERDE:       ; %bb.0: ; %main_body
 ; VERDE-NEXT:    image_sample_lz v[0:3], v[0:1], s[0:7], s[8:11] dmask:0xf
 ; VERDE-NEXT:    s_waitcnt vmcnt(0)
 ; VERDE-NEXT:    ; return to shader part epilog
 ;
 ; GFX6789-LABEL: sample_lz_2d:
 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    image_sample_lz v[0:3], v[0:1], s[0:7], s[8:11] dmask:0xf
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: sample_lz_2d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_sample_lz v[0:3], v[0:1], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D ; encoding: [0x08,0x0f,0x9c,0xf0,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: sample_lz_2d:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_sample_lz v[0:3], v[0:1], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.lz.2d.v4f32.f32(i32 15, float %s, float %t, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @sample_c_lz_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, float %s) {
 ; VERDE-LABEL: sample_c_lz_1d:
 ; VERDE:       ; %bb.0: ; %main_body
 ; VERDE-NEXT:    image_sample_c_lz v[0:3], v[0:1], s[0:7], s[8:11] dmask:0xf
 ; VERDE-NEXT:    s_waitcnt vmcnt(0)
 ; VERDE-NEXT:    ; return to shader part epilog
 ;
 ; GFX6789-LABEL: sample_c_lz_1d:
 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    image_sample_c_lz v[0:3], v[0:1], s[0:7], s[8:11] dmask:0xf
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: sample_c_lz_1d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_sample_c_lz v[0:3], v[0:1], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D ; encoding: [0x00,0x0f,0xbc,0xf0,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: sample_c_lz_1d:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_sample_c_lz v[0:3], v[0:1], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.c.lz.1d.v4f32.f32(i32 15, float %zcompare, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @sample_c_lz_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, float %s, float %t) {
 ; VERDE-LABEL: sample_c_lz_2d:
 ; VERDE:       ; %bb.0: ; %main_body
 ; VERDE-NEXT:    image_sample_c_lz v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf
 ; VERDE-NEXT:    s_waitcnt vmcnt(0)
 ; VERDE-NEXT:    ; return to shader part epilog
 ;
 ; GFX6789-LABEL: sample_c_lz_2d:
 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    image_sample_c_lz v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: sample_c_lz_2d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_sample_c_lz v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D ; encoding: [0x08,0x0f,0xbc,0xf0,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: sample_c_lz_2d:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_sample_c_lz v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.c.lz.2d.v4f32.f32(i32 15, float %zcompare, float %s, float %t, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps float @sample_c_d_o_2darray_V1(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, i32 %offset, float %zcompare, float %dsdh, float %dtdh, float %dsdv, float %dtdv, float %s, float %t, float %slice) {
 ; VERDE-LABEL: sample_c_d_o_2darray_V1:
 ; VERDE:       ; %bb.0: ; %main_body
 ; VERDE-NEXT:    image_sample_c_d_o v0, v[0:15], s[0:7], s[8:11] dmask:0x4 da
 ; VERDE-NEXT:    s_waitcnt vmcnt(0)
 ; VERDE-NEXT:    ; return to shader part epilog
 ;
 ; GFX6789-LABEL: sample_c_d_o_2darray_V1:
 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    image_sample_c_d_o v0, v[0:15], s[0:7], s[8:11] dmask:0x4 da
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: sample_c_d_o_2darray_V1:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_sample_c_d_o v0, v[0:15], s[0:7], s[8:11] dmask:0x4 dim:SQ_RSRC_IMG_2D_ARRAY ; encoding: [0x28,0x04,0xe8,0xf0,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: sample_c_d_o_2darray_V1:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_sample_c_d_o v0, v[0:15], s[0:7], s[8:11] dmask:0x4 dim:SQ_RSRC_IMG_2D_ARRAY
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call float @llvm.amdgcn.image.sample.c.d.o.2darray.f32.f32.f32(i32 4, i32 %offset, float %zcompare, float %dsdh, float %dtdh, float %dsdv, float %dtdv, float %s, float %t, float %slice, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret float %v
 }
 
 define amdgpu_ps float @sample_c_d_o_2darray_V1_tfe(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, i32 %offset, float %zcompare, float %dsdh, float %dtdh, float %dsdv, float %dtdv, float %s, float %t, float %slice, i32 addrspace(1)* inreg %out) {
 ; VERDE-LABEL: sample_c_d_o_2darray_V1_tfe:
 ; VERDE:       ; %bb.0: ; %main_body

 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    v_mov_b32_e32 v0, v9
 ; GFX6789-NEXT:    global_store_dword v11, v10, s[12:13]
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
 ; GFX10-LABEL: sample_c_d_o_2darray_V1_tfe:
 ; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    v_mov_b32_e32 v11, 0 ; encoding: [0x80,0x02,0x16,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v12, v11 ; encoding: [0x0b,0x03,0x18,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v9, v11 ; encoding: [0x0b,0x03,0x12,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v10, v12 ; encoding: [0x0c,0x03,0x14,0x7e]
-; GFX10-NEXT:    image_sample_c_d_o v[9:10], v[0:15], s[0:7], s[8:11] dmask:0x4 dim:SQ_RSRC_IMG_2D_ARRAY tfe ; encoding: [0x28,0x04,0xe9,0xf0,0x00,0x09,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    v_mov_b32_e32 v0, v9 ; encoding: [0x09,0x03,0x00,0x7e]
-; GFX10-NEXT:    global_store_dword v11, v10, s[12:13] ; encoding: [0x00,0x80,0x70,0xdc,0x0b,0x0a,0x0c,0x00]
-; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0 ; encoding: [0x00,0x00,0xfd,0xbb]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10-NEXT:    v_mov_b32_e32 v11, 0
+; GFX10-NEXT:    v_mov_b32_e32 v12, v11
+; GFX10-NEXT:    v_mov_b32_e32 v9, v11
+; GFX10-NEXT:    v_mov_b32_e32 v10, v12
+; GFX10-NEXT:    image_sample_c_d_o v[9:10], v[0:15], s[0:7], s[8:11] dmask:0x4 dim:SQ_RSRC_IMG_2D_ARRAY tfe
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    v_mov_b32_e32 v0, v9
+; GFX10-NEXT:    global_store_dword v11, v10, s[12:13]
+; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: sample_c_d_o_2darray_V1_tfe:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    v_mov_b32_e32 v11, 0
+; GFX11-NEXT:    v_mov_b32_e32 v12, v11
+; GFX11-NEXT:    v_mov_b32_e32 v9, v11
+; GFX11-NEXT:    v_mov_b32_e32 v10, v12
+; GFX11-NEXT:    image_sample_c_d_o v[9:10], v[0:15], s[0:7], s[8:11] dmask:0x4 dim:SQ_RSRC_IMG_2D_ARRAY tfe
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    v_mov_b32_e32 v0, v9
+; GFX11-NEXT:    global_store_b32 v11, v10, s[12:13]
+; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %v = call {float,i32} @llvm.amdgcn.image.sample.c.d.o.2darray.f32i32.f32.f32(i32 4, i32 %offset, float %zcompare, float %dsdh, float %dtdh, float %dsdv, float %dtdv, float %s, float %t, float %slice, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 1, i32 0)
   %v.vec = extractvalue {float, i32} %v, 0
   %v.err = extractvalue {float, i32} %v, 1
   store i32 %v.err, i32 addrspace(1)* %out, align 4
   ret float %v.vec
 }
 
 define amdgpu_ps <2 x float> @sample_c_d_o_2darray_V2(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, i32 %offset, float %zcompare, float %dsdh, float %dtdh, float %dsdv, float %dtdv, float %s, float %t, float %slice) {
 ; VERDE-LABEL: sample_c_d_o_2darray_V2:
 ; VERDE:       ; %bb.0: ; %main_body
 ; VERDE-NEXT:    image_sample_c_d_o v[0:1], v[0:15], s[0:7], s[8:11] dmask:0x6 da
 ; VERDE-NEXT:    s_waitcnt vmcnt(0)
 ; VERDE-NEXT:    ; return to shader part epilog
 ;
 ; GFX6789-LABEL: sample_c_d_o_2darray_V2:
 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    image_sample_c_d_o v[0:1], v[0:15], s[0:7], s[8:11] dmask:0x6 da
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: sample_c_d_o_2darray_V2:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_sample_c_d_o v[0:1], v[0:15], s[0:7], s[8:11] dmask:0x6 dim:SQ_RSRC_IMG_2D_ARRAY ; encoding: [0x28,0x06,0xe8,0xf0,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: sample_c_d_o_2darray_V2:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    image_sample_c_d_o v[0:1], v[0:15], s[0:7], s[8:11] dmask:0x6 dim:SQ_RSRC_IMG_2D_ARRAY
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <2 x float> @llvm.amdgcn.image.sample.c.d.o.2darray.v2f32.f32.f32(i32 6, i32 %offset, float %zcompare, float %dsdh, float %dtdh, float %dsdv, float %dtdv, float %s, float %t, float %slice, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <2 x float> %v
 }
 
 define amdgpu_ps <4 x float> @sample_c_d_o_2darray_V2_tfe(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, i32 %offset, float %zcompare, float %dsdh, float %dtdh, float %dsdv, float %dtdv, float %s, float %t, float %slice) {
 ; VERDE-LABEL: sample_c_d_o_2darray_V2_tfe:
 ; VERDE:       ; %bb.0: ; %main_body

 ; GFX6789-NEXT:    v_mov_b32_e32 v11, v9
 ; GFX6789-NEXT:    image_sample_c_d_o v[9:11], v[0:15], s[0:7], s[8:11] dmask:0x6 tfe da
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    v_mov_b32_e32 v0, v9
 ; GFX6789-NEXT:    v_mov_b32_e32 v1, v10
 ; GFX6789-NEXT:    v_mov_b32_e32 v2, v11
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: sample_c_d_o_2darray_V2_tfe:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    v_mov_b32_e32 v9, 0 ; encoding: [0x80,0x02,0x12,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v10, v9 ; encoding: [0x09,0x03,0x14,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v11, v9 ; encoding: [0x09,0x03,0x16,0x7e]
-; GFX10-NEXT:    image_sample_c_d_o v[9:11], v[0:15], s[0:7], s[8:11] dmask:0x6 dim:SQ_RSRC_IMG_2D_ARRAY tfe ; encoding: [0x28,0x06,0xe9,0xf0,0x00,0x09,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    v_mov_b32_e32 v0, v9 ; encoding: [0x09,0x03,0x00,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v1, v10 ; encoding: [0x0a,0x03,0x02,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v2, v11 ; encoding: [0x0b,0x03,0x04,0x7e]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: sample_c_d_o_2darray_V2_tfe:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v9, 0
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v10, v9
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v11, v9
+; GFX10PLUS-NEXT:    image_sample_c_d_o v[9:11], v[0:15], s[0:7], s[8:11] dmask:0x6 dim:SQ_RSRC_IMG_2D_ARRAY tfe
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v0, v9
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v1, v10
+; GFX10PLUS-NEXT:    v_mov_b32_e32 v2, v11
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call {<2 x float>, i32} @llvm.amdgcn.image.sample.c.d.o.2darray.v2f32i32.f32.f32(i32 6, i32 %offset, float %zcompare, float %dsdh, float %dtdh, float %dsdv, float %dtdv, float %s, float %t, float %slice, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 1, i32 0)
   %v.vec = extractvalue {<2 x float>, i32} %v, 0
   %v.f1 = extractelement <2 x float> %v.vec, i32 0
   %v.f2 = extractelement <2 x float> %v.vec, i32 1
   %v.err = extractvalue {<2 x float>, i32} %v, 1
   %v.errf = bitcast i32 %v.err to float
   %res.0 = insertelement <4 x float> undef, float %v.f1, i32 0

 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    s_mov_b64 s[12:13], exec
 ; GFX6789-NEXT:    s_wqm_b64 exec, exec
 ; GFX6789-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; GFX6789-NEXT:    image_sample v[0:3], v0, s[0:7], s[8:11] dmask:0xf unorm
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: sample_1d_unorm:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    s_mov_b32 s12, exec_lo ; encoding: [0x7e,0x03,0x8c,0xbe]
-; GFX10-NEXT:    s_wqm_b32 exec_lo, exec_lo ; encoding: [0x7e,0x09,0xfe,0xbe]
-; GFX10-NEXT:    s_and_b32 exec_lo, exec_lo, s12 ; encoding: [0x7e,0x0c,0x7e,0x87]
-; GFX10-NEXT:    image_sample v[0:3], v0, s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D unorm ; encoding: [0x00,0x1f,0x80,0xf0,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: sample_1d_unorm:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    s_mov_b32 s12, exec_lo
+; GFX10PLUS-NEXT:    s_wqm_b32 exec_lo, exec_lo
+; GFX10PLUS-NEXT:    s_and_b32 exec_lo, exec_lo, s12
+; GFX10PLUS-NEXT:    image_sample v[0:3], v0, s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D unorm
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.1d.v4f32.f32(i32 15, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 1, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @sample_1d_glc(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %s) {
 ; VERDE-LABEL: sample_1d_glc:
 ; VERDE:       ; %bb.0: ; %main_body
 ; VERDE-NEXT:    s_mov_b64 s[12:13], exec
 ; VERDE-NEXT:    s_wqm_b64 exec, exec
 ; VERDE-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; VERDE-NEXT:    image_sample v[0:3], v0, s[0:7], s[8:11] dmask:0xf glc
 ; VERDE-NEXT:    s_waitcnt vmcnt(0)
 ; VERDE-NEXT:    ; return to shader part epilog
 ;
 ; GFX6789-LABEL: sample_1d_glc:
 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    s_mov_b64 s[12:13], exec
 ; GFX6789-NEXT:    s_wqm_b64 exec, exec
 ; GFX6789-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; GFX6789-NEXT:    image_sample v[0:3], v0, s[0:7], s[8:11] dmask:0xf glc
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: sample_1d_glc:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    s_mov_b32 s12, exec_lo ; encoding: [0x7e,0x03,0x8c,0xbe]
-; GFX10-NEXT:    s_wqm_b32 exec_lo, exec_lo ; encoding: [0x7e,0x09,0xfe,0xbe]
-; GFX10-NEXT:    s_and_b32 exec_lo, exec_lo, s12 ; encoding: [0x7e,0x0c,0x7e,0x87]
-; GFX10-NEXT:    image_sample v[0:3], v0, s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D glc ; encoding: [0x00,0x2f,0x80,0xf0,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: sample_1d_glc:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    s_mov_b32 s12, exec_lo
+; GFX10PLUS-NEXT:    s_wqm_b32 exec_lo, exec_lo
+; GFX10PLUS-NEXT:    s_and_b32 exec_lo, exec_lo, s12
+; GFX10PLUS-NEXT:    image_sample v[0:3], v0, s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D glc
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.1d.v4f32.f32(i32 15, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 1)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @sample_1d_slc(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %s) {
 ; VERDE-LABEL: sample_1d_slc:
 ; VERDE:       ; %bb.0: ; %main_body
 ; VERDE-NEXT:    s_mov_b64 s[12:13], exec
 ; VERDE-NEXT:    s_wqm_b64 exec, exec
 ; VERDE-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; VERDE-NEXT:    image_sample v[0:3], v0, s[0:7], s[8:11] dmask:0xf slc
 ; VERDE-NEXT:    s_waitcnt vmcnt(0)
 ; VERDE-NEXT:    ; return to shader part epilog
 ;
 ; GFX6789-LABEL: sample_1d_slc:
 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    s_mov_b64 s[12:13], exec
 ; GFX6789-NEXT:    s_wqm_b64 exec, exec
 ; GFX6789-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; GFX6789-NEXT:    image_sample v[0:3], v0, s[0:7], s[8:11] dmask:0xf slc
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: sample_1d_slc:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    s_mov_b32 s12, exec_lo ; encoding: [0x7e,0x03,0x8c,0xbe]
-; GFX10-NEXT:    s_wqm_b32 exec_lo, exec_lo ; encoding: [0x7e,0x09,0xfe,0xbe]
-; GFX10-NEXT:    s_and_b32 exec_lo, exec_lo, s12 ; encoding: [0x7e,0x0c,0x7e,0x87]
-; GFX10-NEXT:    image_sample v[0:3], v0, s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D slc ; encoding: [0x00,0x0f,0x80,0xf2,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: sample_1d_slc:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    s_mov_b32 s12, exec_lo
+; GFX10PLUS-NEXT:    s_wqm_b32 exec_lo, exec_lo
+; GFX10PLUS-NEXT:    s_and_b32 exec_lo, exec_lo, s12
+; GFX10PLUS-NEXT:    image_sample v[0:3], v0, s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D slc
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.1d.v4f32.f32(i32 15, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 2)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @sample_1d_glc_slc(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %s) {
 ; VERDE-LABEL: sample_1d_glc_slc:
 ; VERDE:       ; %bb.0: ; %main_body
 ; VERDE-NEXT:    s_mov_b64 s[12:13], exec
 ; VERDE-NEXT:    s_wqm_b64 exec, exec
 ; VERDE-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; VERDE-NEXT:    image_sample v[0:3], v0, s[0:7], s[8:11] dmask:0xf glc slc
 ; VERDE-NEXT:    s_waitcnt vmcnt(0)
 ; VERDE-NEXT:    ; return to shader part epilog
 ;
 ; GFX6789-LABEL: sample_1d_glc_slc:
 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    s_mov_b64 s[12:13], exec
 ; GFX6789-NEXT:    s_wqm_b64 exec, exec
 ; GFX6789-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; GFX6789-NEXT:    image_sample v[0:3], v0, s[0:7], s[8:11] dmask:0xf glc slc
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: sample_1d_glc_slc:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    s_mov_b32 s12, exec_lo ; encoding: [0x7e,0x03,0x8c,0xbe]
-; GFX10-NEXT:    s_wqm_b32 exec_lo, exec_lo ; encoding: [0x7e,0x09,0xfe,0xbe]
-; GFX10-NEXT:    s_and_b32 exec_lo, exec_lo, s12 ; encoding: [0x7e,0x0c,0x7e,0x87]
-; GFX10-NEXT:    image_sample v[0:3], v0, s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D glc slc ; encoding: [0x00,0x2f,0x80,0xf2,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: sample_1d_glc_slc:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    s_mov_b32 s12, exec_lo
+; GFX10PLUS-NEXT:    s_wqm_b32 exec_lo, exec_lo
+; GFX10PLUS-NEXT:    s_and_b32 exec_lo, exec_lo, s12
+; GFX10PLUS-NEXT:    image_sample v[0:3], v0, s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D glc slc
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.1d.v4f32.f32(i32 15, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 3)
   ret <4 x float> %v
 }
 
 define amdgpu_ps float @adjust_writemask_sample_0(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %s) {
 ; VERDE-LABEL: adjust_writemask_sample_0:
 ; VERDE:       ; %bb.0: ; %main_body
 ; VERDE-NEXT:    s_mov_b64 s[12:13], exec
 ; VERDE-NEXT:    s_wqm_b64 exec, exec
 ; VERDE-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; VERDE-NEXT:    image_sample v0, v0, s[0:7], s[8:11] dmask:0x1
 ; VERDE-NEXT:    s_waitcnt vmcnt(0)
 ; VERDE-NEXT:    ; return to shader part epilog
 ;
 ; GFX6789-LABEL: adjust_writemask_sample_0:
 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    s_mov_b64 s[12:13], exec
 ; GFX6789-NEXT:    s_wqm_b64 exec, exec
 ; GFX6789-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; GFX6789-NEXT:    image_sample v0, v0, s[0:7], s[8:11] dmask:0x1
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: adjust_writemask_sample_0:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    s_mov_b32 s12, exec_lo ; encoding: [0x7e,0x03,0x8c,0xbe]
-; GFX10-NEXT:    s_wqm_b32 exec_lo, exec_lo ; encoding: [0x7e,0x09,0xfe,0xbe]
-; GFX10-NEXT:    s_and_b32 exec_lo, exec_lo, s12 ; encoding: [0x7e,0x0c,0x7e,0x87]
-; GFX10-NEXT:    image_sample v0, v0, s[0:7], s[8:11] dmask:0x1 dim:SQ_RSRC_IMG_1D ; encoding: [0x00,0x01,0x80,0xf0,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: adjust_writemask_sample_0:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    s_mov_b32 s12, exec_lo
+; GFX10PLUS-NEXT:    s_wqm_b32 exec_lo, exec_lo
+; GFX10PLUS-NEXT:    s_and_b32 exec_lo, exec_lo, s12
+; GFX10PLUS-NEXT:    image_sample v0, v0, s[0:7], s[8:11] dmask:0x1 dim:SQ_RSRC_IMG_1D
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %r = call <4 x float> @llvm.amdgcn.image.sample.1d.v4f32.f32(i32 15, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   %elt0 = extractelement <4 x float> %r, i32 0
   ret float %elt0
 }
 
 define amdgpu_ps <2 x float> @adjust_writemask_sample_01(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %s) {
 ; VERDE-LABEL: adjust_writemask_sample_01:

 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    s_mov_b64 s[12:13], exec
 ; GFX6789-NEXT:    s_wqm_b64 exec, exec
 ; GFX6789-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; GFX6789-NEXT:    image_sample v[0:1], v0, s[0:7], s[8:11] dmask:0x3
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: adjust_writemask_sample_01:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    s_mov_b32 s12, exec_lo ; encoding: [0x7e,0x03,0x8c,0xbe]
-; GFX10-NEXT:    s_wqm_b32 exec_lo, exec_lo ; encoding: [0x7e,0x09,0xfe,0xbe]
-; GFX10-NEXT:    s_and_b32 exec_lo, exec_lo, s12 ; encoding: [0x7e,0x0c,0x7e,0x87]
-; GFX10-NEXT:    image_sample v[0:1], v0, s[0:7], s[8:11] dmask:0x3 dim:SQ_RSRC_IMG_1D ; encoding: [0x00,0x03,0x80,0xf0,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: adjust_writemask_sample_01:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    s_mov_b32 s12, exec_lo
+; GFX10PLUS-NEXT:    s_wqm_b32 exec_lo, exec_lo
+; GFX10PLUS-NEXT:    s_and_b32 exec_lo, exec_lo, s12
+; GFX10PLUS-NEXT:    image_sample v[0:1], v0, s[0:7], s[8:11] dmask:0x3 dim:SQ_RSRC_IMG_1D
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %r = call <4 x float> @llvm.amdgcn.image.sample.1d.v4f32.f32(i32 15, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   %out = shufflevector <4 x float> %r, <4 x float> undef, <2 x i32> <i32 0, i32 1>
   ret <2 x float> %out
 }
 
 define amdgpu_ps <3 x float> @adjust_writemask_sample_012(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %s) {
 ; VERDE-LABEL: adjust_writemask_sample_012:

 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    s_mov_b64 s[12:13], exec
 ; GFX6789-NEXT:    s_wqm_b64 exec, exec
 ; GFX6789-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; GFX6789-NEXT:    image_sample v[0:2], v0, s[0:7], s[8:11] dmask:0x7
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: adjust_writemask_sample_012:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    s_mov_b32 s12, exec_lo ; encoding: [0x7e,0x03,0x8c,0xbe]
-; GFX10-NEXT:    s_wqm_b32 exec_lo, exec_lo ; encoding: [0x7e,0x09,0xfe,0xbe]
-; GFX10-NEXT:    s_and_b32 exec_lo, exec_lo, s12 ; encoding: [0x7e,0x0c,0x7e,0x87]
-; GFX10-NEXT:    image_sample v[0:2], v0, s[0:7], s[8:11] dmask:0x7 dim:SQ_RSRC_IMG_1D ; encoding: [0x00,0x07,0x80,0xf0,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: adjust_writemask_sample_012:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    s_mov_b32 s12, exec_lo
+; GFX10PLUS-NEXT:    s_wqm_b32 exec_lo, exec_lo
+; GFX10PLUS-NEXT:    s_and_b32 exec_lo, exec_lo, s12
+; GFX10PLUS-NEXT:    image_sample v[0:2], v0, s[0:7], s[8:11] dmask:0x7 dim:SQ_RSRC_IMG_1D
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %r = call <4 x float> @llvm.amdgcn.image.sample.1d.v4f32.f32(i32 15, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   %out = shufflevector <4 x float> %r, <4 x float> undef, <3 x i32> <i32 0, i32 1, i32 2>
   ret <3 x float> %out
 }
 
 define amdgpu_ps <2 x float> @adjust_writemask_sample_12(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %s) {
 ; VERDE-LABEL: adjust_writemask_sample_12:

 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    s_mov_b64 s[12:13], exec
 ; GFX6789-NEXT:    s_wqm_b64 exec, exec
 ; GFX6789-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; GFX6789-NEXT:    image_sample v[0:1], v0, s[0:7], s[8:11] dmask:0x6
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: adjust_writemask_sample_12:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    s_mov_b32 s12, exec_lo ; encoding: [0x7e,0x03,0x8c,0xbe]
-; GFX10-NEXT:    s_wqm_b32 exec_lo, exec_lo ; encoding: [0x7e,0x09,0xfe,0xbe]
-; GFX10-NEXT:    s_and_b32 exec_lo, exec_lo, s12 ; encoding: [0x7e,0x0c,0x7e,0x87]
-; GFX10-NEXT:    image_sample v[0:1], v0, s[0:7], s[8:11] dmask:0x6 dim:SQ_RSRC_IMG_1D ; encoding: [0x00,0x06,0x80,0xf0,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: adjust_writemask_sample_12:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    s_mov_b32 s12, exec_lo
+; GFX10PLUS-NEXT:    s_wqm_b32 exec_lo, exec_lo
+; GFX10PLUS-NEXT:    s_and_b32 exec_lo, exec_lo, s12
+; GFX10PLUS-NEXT:    image_sample v[0:1], v0, s[0:7], s[8:11] dmask:0x6 dim:SQ_RSRC_IMG_1D
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %r = call <4 x float> @llvm.amdgcn.image.sample.1d.v4f32.f32(i32 15, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   %out = shufflevector <4 x float> %r, <4 x float> undef, <2 x i32> <i32 1, i32 2>
   ret <2 x float> %out
 }
 
 define amdgpu_ps <2 x float> @adjust_writemask_sample_03(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %s) {
 ; VERDE-LABEL: adjust_writemask_sample_03:

 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    s_mov_b64 s[12:13], exec
 ; GFX6789-NEXT:    s_wqm_b64 exec, exec
 ; GFX6789-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; GFX6789-NEXT:    image_sample v[0:1], v0, s[0:7], s[8:11] dmask:0x9
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: adjust_writemask_sample_03:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    s_mov_b32 s12, exec_lo ; encoding: [0x7e,0x03,0x8c,0xbe]
-; GFX10-NEXT:    s_wqm_b32 exec_lo, exec_lo ; encoding: [0x7e,0x09,0xfe,0xbe]
-; GFX10-NEXT:    s_and_b32 exec_lo, exec_lo, s12 ; encoding: [0x7e,0x0c,0x7e,0x87]
-; GFX10-NEXT:    image_sample v[0:1], v0, s[0:7], s[8:11] dmask:0x9 dim:SQ_RSRC_IMG_1D ; encoding: [0x00,0x09,0x80,0xf0,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: adjust_writemask_sample_03:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    s_mov_b32 s12, exec_lo
+; GFX10PLUS-NEXT:    s_wqm_b32 exec_lo, exec_lo
+; GFX10PLUS-NEXT:    s_and_b32 exec_lo, exec_lo, s12
+; GFX10PLUS-NEXT:    image_sample v[0:1], v0, s[0:7], s[8:11] dmask:0x9 dim:SQ_RSRC_IMG_1D
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %r = call <4 x float> @llvm.amdgcn.image.sample.1d.v4f32.f32(i32 15, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   %out = shufflevector <4 x float> %r, <4 x float> undef, <2 x i32> <i32 0, i32 3>
   ret <2 x float> %out
 }
 
 define amdgpu_ps <2 x float> @adjust_writemask_sample_13(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %s) {
 ; VERDE-LABEL: adjust_writemask_sample_13:

 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    s_mov_b64 s[12:13], exec
 ; GFX6789-NEXT:    s_wqm_b64 exec, exec
 ; GFX6789-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; GFX6789-NEXT:    image_sample v[0:1], v0, s[0:7], s[8:11] dmask:0xa
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: adjust_writemask_sample_13:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    s_mov_b32 s12, exec_lo ; encoding: [0x7e,0x03,0x8c,0xbe]
-; GFX10-NEXT:    s_wqm_b32 exec_lo, exec_lo ; encoding: [0x7e,0x09,0xfe,0xbe]
-; GFX10-NEXT:    s_and_b32 exec_lo, exec_lo, s12 ; encoding: [0x7e,0x0c,0x7e,0x87]
-; GFX10-NEXT:    image_sample v[0:1], v0, s[0:7], s[8:11] dmask:0xa dim:SQ_RSRC_IMG_1D ; encoding: [0x00,0x0a,0x80,0xf0,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: adjust_writemask_sample_13:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    s_mov_b32 s12, exec_lo
+; GFX10PLUS-NEXT:    s_wqm_b32 exec_lo, exec_lo
+; GFX10PLUS-NEXT:    s_and_b32 exec_lo, exec_lo, s12
+; GFX10PLUS-NEXT:    image_sample v[0:1], v0, s[0:7], s[8:11] dmask:0xa dim:SQ_RSRC_IMG_1D
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %r = call <4 x float> @llvm.amdgcn.image.sample.1d.v4f32.f32(i32 15, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   %out = shufflevector <4 x float> %r, <4 x float> undef, <2 x i32> <i32 1, i32 3>
   ret <2 x float> %out
 }
 
 define amdgpu_ps <3 x float> @adjust_writemask_sample_123(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %s) {
 ; VERDE-LABEL: adjust_writemask_sample_123:

 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    s_mov_b64 s[12:13], exec
 ; GFX6789-NEXT:    s_wqm_b64 exec, exec
 ; GFX6789-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; GFX6789-NEXT:    image_sample v[0:2], v0, s[0:7], s[8:11] dmask:0xe
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: adjust_writemask_sample_123:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    s_mov_b32 s12, exec_lo ; encoding: [0x7e,0x03,0x8c,0xbe]
-; GFX10-NEXT:    s_wqm_b32 exec_lo, exec_lo ; encoding: [0x7e,0x09,0xfe,0xbe]
-; GFX10-NEXT:    s_and_b32 exec_lo, exec_lo, s12 ; encoding: [0x7e,0x0c,0x7e,0x87]
-; GFX10-NEXT:    image_sample v[0:2], v0, s[0:7], s[8:11] dmask:0xe dim:SQ_RSRC_IMG_1D ; encoding: [0x00,0x0e,0x80,0xf0,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: adjust_writemask_sample_123:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    s_mov_b32 s12, exec_lo
+; GFX10PLUS-NEXT:    s_wqm_b32 exec_lo, exec_lo
+; GFX10PLUS-NEXT:    s_and_b32 exec_lo, exec_lo, s12
+; GFX10PLUS-NEXT:    image_sample v[0:2], v0, s[0:7], s[8:11] dmask:0xe dim:SQ_RSRC_IMG_1D
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %r = call <4 x float> @llvm.amdgcn.image.sample.1d.v4f32.f32(i32 15, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   %out = shufflevector <4 x float> %r, <4 x float> undef, <3 x i32> <i32 1, i32 2, i32 3>
   ret <3 x float> %out
 }
 
 define amdgpu_ps <4 x float> @adjust_writemask_sample_none_enabled(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %s) {
 ; VERDE-LABEL: adjust_writemask_sample_none_enabled:
 ; VERDE:       ; %bb.0: ; %main_body
 ; VERDE-NEXT:    ; return to shader part epilog
 ;
 ; GFX6789-LABEL: adjust_writemask_sample_none_enabled:
 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: adjust_writemask_sample_none_enabled:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: adjust_writemask_sample_none_enabled:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %r = call <4 x float> @llvm.amdgcn.image.sample.1d.v4f32.f32(i32 0, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %r
 }
 
 define amdgpu_ps <2 x float> @adjust_writemask_sample_123_to_12(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %s) {
 ; VERDE-LABEL: adjust_writemask_sample_123_to_12:
 ; VERDE:       ; %bb.0: ; %main_body
 ; VERDE-NEXT:    s_mov_b64 s[12:13], exec
 ; VERDE-NEXT:    s_wqm_b64 exec, exec
 ; VERDE-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; VERDE-NEXT:    image_sample v[0:1], v0, s[0:7], s[8:11] dmask:0x6
 ; VERDE-NEXT:    s_waitcnt vmcnt(0)
 ; VERDE-NEXT:    ; return to shader part epilog
 ;
 ; GFX6789-LABEL: adjust_writemask_sample_123_to_12:
 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    s_mov_b64 s[12:13], exec
 ; GFX6789-NEXT:    s_wqm_b64 exec, exec
 ; GFX6789-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; GFX6789-NEXT:    image_sample v[0:1], v0, s[0:7], s[8:11] dmask:0x6
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: adjust_writemask_sample_123_to_12:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    s_mov_b32 s12, exec_lo ; encoding: [0x7e,0x03,0x8c,0xbe]
-; GFX10-NEXT:    s_wqm_b32 exec_lo, exec_lo ; encoding: [0x7e,0x09,0xfe,0xbe]
-; GFX10-NEXT:    s_and_b32 exec_lo, exec_lo, s12 ; encoding: [0x7e,0x0c,0x7e,0x87]
-; GFX10-NEXT:    image_sample v[0:1], v0, s[0:7], s[8:11] dmask:0x6 dim:SQ_RSRC_IMG_1D ; encoding: [0x00,0x06,0x80,0xf0,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: adjust_writemask_sample_123_to_12:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    s_mov_b32 s12, exec_lo
+; GFX10PLUS-NEXT:    s_wqm_b32 exec_lo, exec_lo
+; GFX10PLUS-NEXT:    s_and_b32 exec_lo, exec_lo, s12
+; GFX10PLUS-NEXT:    image_sample v[0:1], v0, s[0:7], s[8:11] dmask:0x6 dim:SQ_RSRC_IMG_1D
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %r = call <4 x float> @llvm.amdgcn.image.sample.1d.v4f32.f32(i32 14, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   %out = shufflevector <4 x float> %r, <4 x float> undef, <2 x i32> <i32 0, i32 1>
   ret <2 x float> %out
 }
 
 define amdgpu_ps <2 x float> @adjust_writemask_sample_013_to_13(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %s) {
 ; VERDE-LABEL: adjust_writemask_sample_013_to_13:

 ; GFX6789:       ; %bb.0: ; %main_body
 ; GFX6789-NEXT:    s_mov_b64 s[12:13], exec
 ; GFX6789-NEXT:    s_wqm_b64 exec, exec
 ; GFX6789-NEXT:    s_and_b64 exec, exec, s[12:13]
 ; GFX6789-NEXT:    image_sample v[0:1], v0, s[0:7], s[8:11] dmask:0xa
 ; GFX6789-NEXT:    s_waitcnt vmcnt(0)
 ; GFX6789-NEXT:    ; return to shader part epilog
 ;
-; GFX10-LABEL: adjust_writemask_sample_013_to_13:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    s_mov_b32 s12, exec_lo ; encoding: [0x7e,0x03,0x8c,0xbe]
-; GFX10-NEXT:    s_wqm_b32 exec_lo, exec_lo ; encoding: [0x7e,0x09,0xfe,0xbe]
-; GFX10-NEXT:    s_and_b32 exec_lo, exec_lo, s12 ; encoding: [0x7e,0x0c,0x7e,0x87]
-; GFX10-NEXT:    image_sample v[0:1], v0, s[0:7], s[8:11] dmask:0xa dim:SQ_RSRC_IMG_1D ; encoding: [0x00,0x0a,0x80,0xf0,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
+; GFX10PLUS-LABEL: adjust_writemask_sample_013_to_13:
+; GFX10PLUS:       ; %bb.0: ; %main_body
+; GFX10PLUS-NEXT:    s_mov_b32 s12, exec_lo
+; GFX10PLUS-NEXT:    s_wqm_b32 exec_lo, exec_lo
+; GFX10PLUS-NEXT:    s_and_b32 exec_lo, exec_lo, s12
+; GFX10PLUS-NEXT:    image_sample v[0:1], v0, s[0:7], s[8:11] dmask:0xa dim:SQ_RSRC_IMG_1D
+; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0)
+; GFX10PLUS-NEXT:    ; return to shader part epilog
 main_body:
   %r = call <4 x float> @llvm.amdgcn.image.sample.1d.v4f32.f32(i32 11, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   %out = shufflevector <4 x float> %r, <4 x float> undef, <2 x i32> <i32 1, i32 2>
   ret <2 x float> %out
 }
 
 declare <4 x float> @llvm.amdgcn.image.sample.1d.v4f32.f32(i32, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 declare {<4 x float>,i32} @llvm.amdgcn.image.sample.1d.v4f32i32.f32(i32, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1

 declare <4 x float> @llvm.amdgcn.image.sample.d.2d.v4f32.f32.f32(i32, float, float, float, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 declare <4 x float> @llvm.amdgcn.image.sample.c.d.1d.v4f32.f32.f32(i32, float, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 declare <4 x float> @llvm.amdgcn.image.sample.c.d.2d.v4f32.f32.f32(i32, float, float, float, float, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 declare <4 x float> @llvm.amdgcn.image.sample.d.cl.1d.v4f32.f32.f32(i32, float, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 declare <4 x float> @llvm.amdgcn.image.sample.d.cl.2d.v4f32.f32.f32(i32, float, float, float, float, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 declare <4 x float> @llvm.amdgcn.image.sample.c.d.cl.1d.v4f32.f32.f32(i32, float, float, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 declare <4 x float> @llvm.amdgcn.image.sample.c.d.cl.2d.v4f32.f32.f32(i32, float, float, float, float, float, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 
-declare <4 x float> @llvm.amdgcn.image.sample.cd.1d.v4f32.f32.f32(i32, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
-declare <4 x float> @llvm.amdgcn.image.sample.cd.2d.v4f32.f32.f32(i32, float, float, float, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
-declare <4 x float> @llvm.amdgcn.image.sample.c.cd.1d.v4f32.f32.f32(i32, float, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
-declare <4 x float> @llvm.amdgcn.image.sample.c.cd.2d.v4f32.f32.f32(i32, float, float, float, float, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
-declare <4 x float> @llvm.amdgcn.image.sample.cd.cl.1d.v4f32.f32.f32(i32, float, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
-declare <4 x float> @llvm.amdgcn.image.sample.cd.cl.2d.v4f32.f32.f32(i32, float, float, float, float, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
-declare <4 x float> @llvm.amdgcn.image.sample.c.cd.cl.1d.v4f32.f32.f32(i32, float, float, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
-declare <4 x float> @llvm.amdgcn.image.sample.c.cd.cl.2d.v4f32.f32.f32(i32, float, float, float, float, float, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
-
 declare <4 x float> @llvm.amdgcn.image.sample.l.1d.v4f32.f32(i32, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 declare <4 x float> @llvm.amdgcn.image.sample.l.2d.v4f32.f32(i32, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 declare <4 x float> @llvm.amdgcn.image.sample.c.l.1d.v4f32.f32(i32, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 declare <4 x float> @llvm.amdgcn.image.sample.c.l.2d.v4f32.f32(i32, float, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 
 declare <4 x float> @llvm.amdgcn.image.sample.lz.1d.v4f32.f32(i32, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 declare <4 x float> @llvm.amdgcn.image.sample.lz.2d.v4f32.f32(i32, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 declare <4 x float> @llvm.amdgcn.image.sample.c.lz.1d.v4f32.f32(i32, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1

llvm/test/CodeGen/AMDGPU/llvm.amdgcn.image.sample.g16.encode.ll

 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
 ; RUN: llc -march=amdgcn -mcpu=gfx1010 -show-mc-encoding < %s | FileCheck -check-prefixes=GFX10 %s
+; RUN: llc -march=amdgcn -mcpu=gfx1100 -show-mc-encoding < %s | FileCheck -check-prefixes=GFX11 %s
 
 define amdgpu_ps <4 x float> @sample_d_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, half %dsdh, half %dsdv, float %s) {
 ; GFX10-LABEL: sample_d_1d:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    image_sample_d_g16 v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D ; encoding: [0x01,0x0f,0x88,0xf0,0x00,0x00,0x40,0x00]
 ; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
 ; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: sample_d_1d:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    image_sample_d_g16 v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D ; encoding: [0x00,0x0f,0xe4,0xf0,0x00,0x00,0x00,0x08]
+; GFX11-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0xf7,0x03,0x89,0xbf]
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.d.1d.v4f32.f16.f32(i32 15, half %dsdh, half %dsdv, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @sample_d_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t) {
 ; GFX10-LABEL: sample_d_2d:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    v_and_b32_e32 v2, 0xffff, v2 ; encoding: [0xff,0x04,0x04,0x36,0xff,0xff,0x00,0x00]
 ; GFX10-NEXT:    v_and_b32_e32 v0, 0xffff, v0 ; encoding: [0xff,0x00,0x00,0x36,0xff,0xff,0x00,0x00]
 ; GFX10-NEXT:    v_lshl_or_b32 v2, v3, 16, v2 ; encoding: [0x02,0x00,0x6f,0xd7,0x03,0x21,0x09,0x04]
 ; GFX10-NEXT:    v_lshl_or_b32 v0, v1, 16, v0 ; encoding: [0x00,0x00,0x6f,0xd7,0x01,0x21,0x01,0x04]
 ; GFX10-NEXT:    image_sample_d_g16 v[0:3], [v0, v2, v4, v5], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D ; encoding: [0x0b,0x0f,0x88,0xf0,0x00,0x00,0x40,0x00,0x02,0x04,0x05,0x00]
 ; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
 ; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: sample_d_2d:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    v_and_b32_e32 v2, 0xffff, v2 ; encoding: [0xff,0x04,0x04,0x36,0xff,0xff,0x00,0x00]
+; GFX11-NEXT:    v_and_b32_e32 v0, 0xffff, v0 ; encoding: [0xff,0x00,0x00,0x36,0xff,0xff,0x00,0x00]
+; GFX11-NEXT:    v_lshl_or_b32 v2, v3, 16, v2 ; encoding: [0x02,0x00,0x56,0xd6,0x03,0x21,0x09,0x04]
+; GFX11-NEXT:    v_lshl_or_b32 v0, v1, 16, v0 ; encoding: [0x00,0x00,0x56,0xd6,0x01,0x21,0x01,0x04]
+; GFX11-NEXT:    image_sample_d_g16 v[0:3], [v0, v2, v4, v5], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D ; encoding: [0x05,0x0f,0xe4,0xf0,0x00,0x00,0x00,0x08,0x02,0x04,0x05,0x00]
+; GFX11-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0xf7,0x03,0x89,0xbf]
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.d.2d.v4f32.f16.f32(i32 15, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @sample_d_3d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, half %dsdh, half %dtdh, half %drdh, half %dsdv, half %dtdv, half %drdv, float %s, float %t, float %r) {
 ; GFX10-LABEL: sample_d_3d:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    v_mov_b32_e32 v9, v3 ; encoding: [0x03,0x03,0x12,0x7e]
 ; GFX10-NEXT:    v_mov_b32_e32 v3, v2 ; encoding: [0x02,0x03,0x06,0x7e]
 ; GFX10-NEXT:    v_and_b32_e32 v0, 0xffff, v0 ; encoding: [0xff,0x00,0x00,0x36,0xff,0xff,0x00,0x00]
 ; GFX10-NEXT:    v_and_b32_e32 v2, 0xffff, v9 ; encoding: [0xff,0x12,0x04,0x36,0xff,0xff,0x00,0x00]
 ; GFX10-NEXT:    v_lshl_or_b32 v4, v4, 16, v2 ; encoding: [0x04,0x00,0x6f,0xd7,0x04,0x21,0x09,0x04]
 ; GFX10-NEXT:    v_lshl_or_b32 v2, v1, 16, v0 ; encoding: [0x02,0x00,0x6f,0xd7,0x01,0x21,0x01,0x04]
 ; GFX10-NEXT:    image_sample_d_g16 v[0:3], v[2:8], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_3D ; encoding: [0x11,0x0f,0x88,0xf0,0x02,0x00,0x40,0x00]
 ; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
 ; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: sample_d_3d:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    v_mov_b32_e32 v9, v3 ; encoding: [0x03,0x03,0x12,0x7e]
+; GFX11-NEXT:    v_mov_b32_e32 v3, v2 ; encoding: [0x02,0x03,0x06,0x7e]
+; GFX11-NEXT:    v_and_b32_e32 v0, 0xffff, v0 ; encoding: [0xff,0x00,0x00,0x36,0xff,0xff,0x00,0x00]
+; GFX11-NEXT:    v_and_b32_e32 v2, 0xffff, v9 ; encoding: [0xff,0x12,0x04,0x36,0xff,0xff,0x00,0x00]
+; GFX11-NEXT:    v_lshl_or_b32 v4, v4, 16, v2 ; encoding: [0x04,0x00,0x56,0xd6,0x04,0x21,0x09,0x04]
+; GFX11-NEXT:    v_lshl_or_b32 v2, v1, 16, v0 ; encoding: [0x02,0x00,0x56,0xd6,0x01,0x21,0x01,0x04]
+; GFX11-NEXT:    image_sample_d_g16 v[0:3], v[2:8], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_3D ; encoding: [0x08,0x0f,0xe4,0xf0,0x02,0x00,0x00,0x08]
+; GFX11-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0xf7,0x03,0x89,0xbf]
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.d.3d.v4f32.f16.f32(i32 15, half %dsdh, half %dtdh, half %drdh, half %dsdv, half %dtdv, half %drdv, float %s, float %t, float %r, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @sample_c_d_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, half %dsdh, half %dsdv, float %s) {
 ; GFX10-LABEL: sample_c_d_1d:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    image_sample_c_d_g16 v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D ; encoding: [0x01,0x0f,0xa8,0xf0,0x00,0x00,0x40,0x00]
 ; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
 ; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: sample_c_d_1d:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    image_sample_c_d_g16 v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D ; encoding: [0x00,0x0f,0xe8,0xf0,0x00,0x00,0x00,0x08]
+; GFX11-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0xf7,0x03,0x89,0xbf]
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.c.d.1d.v4f32.f16.f32(i32 15, float %zcompare, half %dsdh, half %dsdv, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @sample_c_d_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t) {
 ; GFX10-LABEL: sample_c_d_2d:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    v_and_b32_e32 v3, 0xffff, v3 ; encoding: [0xff,0x06,0x06,0x36,0xff,0xff,0x00,0x00]
 ; GFX10-NEXT:    v_and_b32_e32 v1, 0xffff, v1 ; encoding: [0xff,0x02,0x02,0x36,0xff,0xff,0x00,0x00]
 ; GFX10-NEXT:    v_lshl_or_b32 v3, v4, 16, v3 ; encoding: [0x03,0x00,0x6f,0xd7,0x04,0x21,0x0d,0x04]
 ; GFX10-NEXT:    v_lshl_or_b32 v1, v2, 16, v1 ; encoding: [0x01,0x00,0x6f,0xd7,0x02,0x21,0x05,0x04]
 ; GFX10-NEXT:    image_sample_c_d_g16 v[0:3], [v0, v1, v3, v5, v6], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D ; encoding: [0x0b,0x0f,0xa8,0xf0,0x00,0x00,0x40,0x00,0x01,0x03,0x05,0x06]
 ; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
 ; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: sample_c_d_2d:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    v_and_b32_e32 v3, 0xffff, v3 ; encoding: [0xff,0x06,0x06,0x36,0xff,0xff,0x00,0x00]
+; GFX11-NEXT:    v_and_b32_e32 v1, 0xffff, v1 ; encoding: [0xff,0x02,0x02,0x36,0xff,0xff,0x00,0x00]
+; GFX11-NEXT:    v_lshl_or_b32 v3, v4, 16, v3 ; encoding: [0x03,0x00,0x56,0xd6,0x04,0x21,0x0d,0x04]
+; GFX11-NEXT:    v_lshl_or_b32 v1, v2, 16, v1 ; encoding: [0x01,0x00,0x56,0xd6,0x02,0x21,0x05,0x04]
+; GFX11-NEXT:    image_sample_c_d_g16 v[0:3], [v0, v1, v3, v5, v6], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D ; encoding: [0x05,0x0f,0xe8,0xf0,0x00,0x00,0x00,0x08,0x01,0x03,0x05,0x06]
+; GFX11-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0xf7,0x03,0x89,0xbf]
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.c.d.2d.v4f32.f16.f32(i32 15, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @sample_d_cl_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, half %dsdh, half %dsdv, float %s, float %clamp) {
 ; GFX10-LABEL: sample_d_cl_1d:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    image_sample_d_cl_g16 v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D ; encoding: [0x01,0x0f,0x8c,0xf0,0x00,0x00,0x40,0x00]
 ; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
 ; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: sample_d_cl_1d:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    image_sample_d_cl_g16 v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D ; encoding: [0x00,0x0f,0x7c,0xf1,0x00,0x00,0x00,0x08]
+; GFX11-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0xf7,0x03,0x89,0xbf]
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.d.cl.1d.v4f32.f16.f32(i32 15, half %dsdh, half %dsdv, float %s, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @sample_d_cl_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, float %clamp) {
 ; GFX10-LABEL: sample_d_cl_2d:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    v_and_b32_e32 v2, 0xffff, v2 ; encoding: [0xff,0x04,0x04,0x36,0xff,0xff,0x00,0x00]
 ; GFX10-NEXT:    v_and_b32_e32 v0, 0xffff, v0 ; encoding: [0xff,0x00,0x00,0x36,0xff,0xff,0x00,0x00]
 ; GFX10-NEXT:    v_lshl_or_b32 v2, v3, 16, v2 ; encoding: [0x02,0x00,0x6f,0xd7,0x03,0x21,0x09,0x04]
 ; GFX10-NEXT:    v_lshl_or_b32 v0, v1, 16, v0 ; encoding: [0x00,0x00,0x6f,0xd7,0x01,0x21,0x01,0x04]
 ; GFX10-NEXT:    image_sample_d_cl_g16 v[0:3], [v0, v2, v4, v5, v6], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D ; encoding: [0x0b,0x0f,0x8c,0xf0,0x00,0x00,0x40,0x00,0x02,0x04,0x05,0x06]
 ; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
 ; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: sample_d_cl_2d:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    v_and_b32_e32 v2, 0xffff, v2 ; encoding: [0xff,0x04,0x04,0x36,0xff,0xff,0x00,0x00]
+; GFX11-NEXT:    v_and_b32_e32 v0, 0xffff, v0 ; encoding: [0xff,0x00,0x00,0x36,0xff,0xff,0x00,0x00]
+; GFX11-NEXT:    v_lshl_or_b32 v2, v3, 16, v2 ; encoding: [0x02,0x00,0x56,0xd6,0x03,0x21,0x09,0x04]
+; GFX11-NEXT:    v_lshl_or_b32 v0, v1, 16, v0 ; encoding: [0x00,0x00,0x56,0xd6,0x01,0x21,0x01,0x04]
+; GFX11-NEXT:    image_sample_d_cl_g16 v[0:3], [v0, v2, v4, v5, v6], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D ; encoding: [0x05,0x0f,0x7c,0xf1,0x00,0x00,0x00,0x08,0x02,0x04,0x05,0x06]
+; GFX11-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0xf7,0x03,0x89,0xbf]
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.d.cl.2d.v4f32.f16.f32(i32 15, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @sample_c_d_cl_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, half %dsdh, half %dsdv, float %s, float %clamp) {
 ; GFX10-LABEL: sample_c_d_cl_1d:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    image_sample_c_d_cl_g16 v[0:3], v[0:4], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D ; encoding: [0x01,0x0f,0xac,0xf0,0x00,0x00,0x40,0x00]
 ; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
 ; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: sample_c_d_cl_1d:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    image_sample_c_d_cl_g16 v[0:3], v[0:4], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D ; encoding: [0x00,0x0f,0x50,0xf1,0x00,0x00,0x00,0x08]
+; GFX11-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0xf7,0x03,0x89,0xbf]
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.c.d.cl.1d.v4f32.f16.f32(i32 15, float %zcompare, half %dsdh, half %dsdv, float %s, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v
 }
 
 define amdgpu_ps <4 x float> @sample_c_d_cl_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, float %clamp) {
 ; GFX10-LABEL: sample_c_d_cl_2d:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    v_mov_b32_e32 v8, v2 ; encoding: [0x02,0x03,0x10,0x7e]
 ; GFX10-NEXT:    v_mov_b32_e32 v2, v0 ; encoding: [0x00,0x03,0x04,0x7e]
 ; GFX10-NEXT:    v_and_b32_e32 v0, 0xffff, v3 ; encoding: [0xff,0x06,0x00,0x36,0xff,0xff,0x00,0x00]
 ; GFX10-NEXT:    v_and_b32_e32 v1, 0xffff, v1 ; encoding: [0xff,0x02,0x02,0x36,0xff,0xff,0x00,0x00]
 ; GFX10-NEXT:    v_lshl_or_b32 v4, v4, 16, v0 ; encoding: [0x04,0x00,0x6f,0xd7,0x04,0x21,0x01,0x04]
 ; GFX10-NEXT:    v_lshl_or_b32 v3, v8, 16, v1 ; encoding: [0x03,0x00,0x6f,0xd7,0x08,0x21,0x05,0x04]
 ; GFX10-NEXT:    image_sample_c_d_cl_g16 v[0:3], v[2:7], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D ; encoding: [0x09,0x0f,0xac,0xf0,0x02,0x00,0x40,0x00]
 ; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
 ; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: sample_c_d_cl_2d:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    v_mov_b32_e32 v8, v2 ; encoding: [0x02,0x03,0x10,0x7e]
+; GFX11-NEXT:    v_mov_b32_e32 v2, v0 ; encoding: [0x00,0x03,0x04,0x7e]
+; GFX11-NEXT:    v_and_b32_e32 v0, 0xffff, v3 ; encoding: [0xff,0x06,0x00,0x36,0xff,0xff,0x00,0x00]
+; GFX11-NEXT:    v_and_b32_e32 v1, 0xffff, v1 ; encoding: [0xff,0x02,0x02,0x36,0xff,0xff,0x00,0x00]
+; GFX11-NEXT:    v_lshl_or_b32 v4, v4, 16, v0 ; encoding: [0x04,0x00,0x56,0xd6,0x04,0x21,0x01,0x04]
+; GFX11-NEXT:    v_lshl_or_b32 v3, v8, 16, v1 ; encoding: [0x03,0x00,0x56,0xd6,0x08,0x21,0x05,0x04]
+; GFX11-NEXT:    image_sample_c_d_cl_g16 v[0:3], v[2:7], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D ; encoding: [0x04,0x0f,0x50,0xf1,0x02,0x00,0x00,0x08]
+; GFX11-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0xf7,0x03,0x89,0xbf]
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.c.d.cl.2d.v4f32.f16.f32(i32 15, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v
 }
 
-define amdgpu_ps <4 x float> @sample_cd_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, half %dsdh, half %dsdv, float %s) {
-; GFX10-LABEL: sample_cd_1d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_sample_cd_g16 v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D ; encoding: [0x01,0x0f,0xa0,0xf1,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
-main_body:
-  %v = call <4 x float> @llvm.amdgcn.image.sample.cd.1d.v4f32.f16.f32(i32 15, half %dsdh, half %dsdv, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
-  ret <4 x float> %v
-}
-
-define amdgpu_ps <4 x float> @sample_cd_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t) {
-; GFX10-LABEL: sample_cd_2d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    v_and_b32_e32 v2, 0xffff, v2 ; encoding: [0xff,0x04,0x04,0x36,0xff,0xff,0x00,0x00]
-; GFX10-NEXT:    v_and_b32_e32 v0, 0xffff, v0 ; encoding: [0xff,0x00,0x00,0x36,0xff,0xff,0x00,0x00]
-; GFX10-NEXT:    v_lshl_or_b32 v2, v3, 16, v2 ; encoding: [0x02,0x00,0x6f,0xd7,0x03,0x21,0x09,0x04]
-; GFX10-NEXT:    v_lshl_or_b32 v0, v1, 16, v0 ; encoding: [0x00,0x00,0x6f,0xd7,0x01,0x21,0x01,0x04]
-; GFX10-NEXT:    image_sample_cd_g16 v[0:3], [v0, v2, v4, v5], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D ; encoding: [0x0b,0x0f,0xa0,0xf1,0x00,0x00,0x40,0x00,0x02,0x04,0x05,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
-main_body:
-  %v = call <4 x float> @llvm.amdgcn.image.sample.cd.2d.v4f32.f16.f32(i32 15, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
-  ret <4 x float> %v
-}
-
-define amdgpu_ps <4 x float> @sample_c_cd_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, half %dsdh, half %dsdv, float %s) {
-; GFX10-LABEL: sample_c_cd_1d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_sample_c_cd_g16 v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D ; encoding: [0x01,0x0f,0xa8,0xf1,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
-main_body:
-  %v = call <4 x float> @llvm.amdgcn.image.sample.c.cd.1d.v4f32.f16.f32(i32 15, float %zcompare, half %dsdh, half %dsdv, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
-  ret <4 x float> %v
-}
-
-define amdgpu_ps <4 x float> @sample_c_cd_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t) {
-; GFX10-LABEL: sample_c_cd_2d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    v_and_b32_e32 v3, 0xffff, v3 ; encoding: [0xff,0x06,0x06,0x36,0xff,0xff,0x00,0x00]
-; GFX10-NEXT:    v_and_b32_e32 v1, 0xffff, v1 ; encoding: [0xff,0x02,0x02,0x36,0xff,0xff,0x00,0x00]
-; GFX10-NEXT:    v_lshl_or_b32 v3, v4, 16, v3 ; encoding: [0x03,0x00,0x6f,0xd7,0x04,0x21,0x0d,0x04]
-; GFX10-NEXT:    v_lshl_or_b32 v1, v2, 16, v1 ; encoding: [0x01,0x00,0x6f,0xd7,0x02,0x21,0x05,0x04]
-; GFX10-NEXT:    image_sample_c_cd_g16 v[0:3], [v0, v1, v3, v5, v6], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D ; encoding: [0x0b,0x0f,0xa8,0xf1,0x00,0x00,0x40,0x00,0x01,0x03,0x05,0x06]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
-main_body:
-  %v = call <4 x float> @llvm.amdgcn.image.sample.c.cd.2d.v4f32.f16.f32(i32 15, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
-  ret <4 x float> %v
-}
-
-define amdgpu_ps <4 x float> @sample_cd_cl_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, half %dsdh, half %dsdv, float %s, float %clamp) {
-; GFX10-LABEL: sample_cd_cl_1d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_sample_cd_cl_g16 v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D ; encoding: [0x01,0x0f,0xa4,0xf1,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
-main_body:
-  %v = call <4 x float> @llvm.amdgcn.image.sample.cd.cl.1d.v4f32.f16.f32(i32 15, half %dsdh, half %dsdv, float %s, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
-  ret <4 x float> %v
-}
-
-define amdgpu_ps <4 x float> @sample_cd_cl_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, float %clamp) {
-; GFX10-LABEL: sample_cd_cl_2d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    v_and_b32_e32 v2, 0xffff, v2 ; encoding: [0xff,0x04,0x04,0x36,0xff,0xff,0x00,0x00]
-; GFX10-NEXT:    v_and_b32_e32 v0, 0xffff, v0 ; encoding: [0xff,0x00,0x00,0x36,0xff,0xff,0x00,0x00]
-; GFX10-NEXT:    v_lshl_or_b32 v2, v3, 16, v2 ; encoding: [0x02,0x00,0x6f,0xd7,0x03,0x21,0x09,0x04]
-; GFX10-NEXT:    v_lshl_or_b32 v0, v1, 16, v0 ; encoding: [0x00,0x00,0x6f,0xd7,0x01,0x21,0x01,0x04]
-; GFX10-NEXT:    image_sample_cd_cl_g16 v[0:3], [v0, v2, v4, v5, v6], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D ; encoding: [0x0b,0x0f,0xa4,0xf1,0x00,0x00,0x40,0x00,0x02,0x04,0x05,0x06]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
-main_body:
-  %v = call <4 x float> @llvm.amdgcn.image.sample.cd.cl.2d.v4f32.f16.f32(i32 15, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
-  ret <4 x float> %v
-}
-
-define amdgpu_ps <4 x float> @sample_c_cd_cl_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, half %dsdh, half %dsdv, float %s, float %clamp) {
-; GFX10-LABEL: sample_c_cd_cl_1d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_sample_c_cd_cl_g16 v[0:3], v[0:4], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D ; encoding: [0x01,0x0f,0xac,0xf1,0x00,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
-main_body:
-  %v = call <4 x float> @llvm.amdgcn.image.sample.c.cd.cl.1d.v4f32.f16.f32(i32 15, float %zcompare, half %dsdh, half %dsdv, float %s, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
-  ret <4 x float> %v
-}
-
-define amdgpu_ps <4 x float> @sample_c_cd_cl_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, float %clamp) {
-; GFX10-LABEL: sample_c_cd_cl_2d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    v_mov_b32_e32 v8, v2 ; encoding: [0x02,0x03,0x10,0x7e]
-; GFX10-NEXT:    v_mov_b32_e32 v2, v0 ; encoding: [0x00,0x03,0x04,0x7e]
-; GFX10-NEXT:    v_and_b32_e32 v0, 0xffff, v3 ; encoding: [0xff,0x06,0x00,0x36,0xff,0xff,0x00,0x00]
-; GFX10-NEXT:    v_and_b32_e32 v1, 0xffff, v1 ; encoding: [0xff,0x02,0x02,0x36,0xff,0xff,0x00,0x00]
-; GFX10-NEXT:    v_lshl_or_b32 v4, v4, 16, v0 ; encoding: [0x04,0x00,0x6f,0xd7,0x04,0x21,0x01,0x04]
-; GFX10-NEXT:    v_lshl_or_b32 v3, v8, 16, v1 ; encoding: [0x03,0x00,0x6f,0xd7,0x08,0x21,0x05,0x04]
-; GFX10-NEXT:    image_sample_c_cd_cl_g16 v[0:3], v[2:7], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D ; encoding: [0x09,0x0f,0xac,0xf1,0x02,0x00,0x40,0x00]
-; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
-; GFX10-NEXT:    ; return to shader part epilog
-main_body:
-  %v = call <4 x float> @llvm.amdgcn.image.sample.c.cd.cl.2d.v4f32.f16.f32(i32 15, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
-  ret <4 x float> %v
-}
-
 define amdgpu_ps float @sample_c_d_o_2darray_V1(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, i32 %offset, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, float %slice) {
 ; GFX10-LABEL: sample_c_d_o_2darray_V1:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    v_mov_b32_e32 v9, v2 ; encoding: [0x02,0x03,0x12,0x7e]
 ; GFX10-NEXT:    v_mov_b32_e32 v10, v3 ; encoding: [0x03,0x03,0x14,0x7e]
 ; GFX10-NEXT:    v_mov_b32_e32 v3, v1 ; encoding: [0x01,0x03,0x06,0x7e]
 ; GFX10-NEXT:    v_mov_b32_e32 v2, v0 ; encoding: [0x00,0x03,0x04,0x7e]
 ; GFX10-NEXT:    v_and_b32_e32 v0, 0xffff, v4 ; encoding: [0xff,0x08,0x00,0x36,0xff,0xff,0x00,0x00]
 ; GFX10-NEXT:    v_and_b32_e32 v1, 0xffff, v9 ; encoding: [0xff,0x12,0x02,0x36,0xff,0xff,0x00,0x00]
 ; GFX10-NEXT:    v_lshl_or_b32 v5, v5, 16, v0 ; encoding: [0x05,0x00,0x6f,0xd7,0x05,0x21,0x01,0x04]
 ; GFX10-NEXT:    v_lshl_or_b32 v4, v10, 16, v1 ; encoding: [0x04,0x00,0x6f,0xd7,0x0a,0x21,0x05,0x04]
 ; GFX10-NEXT:    image_sample_c_d_o_g16 v0, v[2:8], s[0:7], s[8:11] dmask:0x4 dim:SQ_RSRC_IMG_2D_ARRAY ; encoding: [0x29,0x04,0xe8,0xf0,0x02,0x00,0x40,0x00]
 ; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
 ; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: sample_c_d_o_2darray_V1:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    v_mov_b32_e32 v9, v2 ; encoding: [0x02,0x03,0x12,0x7e]
+; GFX11-NEXT:    v_mov_b32_e32 v10, v3 ; encoding: [0x03,0x03,0x14,0x7e]
+; GFX11-NEXT:    v_mov_b32_e32 v3, v1 ; encoding: [0x01,0x03,0x06,0x7e]
+; GFX11-NEXT:    v_mov_b32_e32 v2, v0 ; encoding: [0x00,0x03,0x04,0x7e]
+; GFX11-NEXT:    v_and_b32_e32 v0, 0xffff, v4 ; encoding: [0xff,0x08,0x00,0x36,0xff,0xff,0x00,0x00]
+; GFX11-NEXT:    v_and_b32_e32 v1, 0xffff, v9 ; encoding: [0xff,0x12,0x02,0x36,0xff,0xff,0x00,0x00]
+; GFX11-NEXT:    v_lshl_or_b32 v5, v5, 16, v0 ; encoding: [0x05,0x00,0x56,0xd6,0x05,0x21,0x01,0x04]
+; GFX11-NEXT:    v_lshl_or_b32 v4, v10, 16, v1 ; encoding: [0x04,0x00,0x56,0xd6,0x0a,0x21,0x05,0x04]
+; GFX11-NEXT:    image_sample_c_d_o_g16 v0, v[2:8], s[0:7], s[8:11] dmask:0x4 dim:SQ_RSRC_IMG_2D_ARRAY ; encoding: [0x14,0x04,0xf0,0xf0,0x02,0x00,0x00,0x08]
+; GFX11-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0xf7,0x03,0x89,0xbf]
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %v = call float @llvm.amdgcn.image.sample.c.d.o.2darray.f16.f32.f32(i32 4, i32 %offset, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, float %slice, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret float %v
 }
 
 define amdgpu_ps <2 x float> @sample_c_d_o_2darray_V2(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, i32 %offset, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, float %slice) {
 ; GFX10-LABEL: sample_c_d_o_2darray_V2:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    v_mov_b32_e32 v9, v2 ; encoding: [0x02,0x03,0x12,0x7e]
 ; GFX10-NEXT:    v_mov_b32_e32 v10, v3 ; encoding: [0x03,0x03,0x14,0x7e]
 ; GFX10-NEXT:    v_mov_b32_e32 v3, v1 ; encoding: [0x01,0x03,0x06,0x7e]
 ; GFX10-NEXT:    v_mov_b32_e32 v2, v0 ; encoding: [0x00,0x03,0x04,0x7e]
 ; GFX10-NEXT:    v_and_b32_e32 v0, 0xffff, v4 ; encoding: [0xff,0x08,0x00,0x36,0xff,0xff,0x00,0x00]
 ; GFX10-NEXT:    v_and_b32_e32 v1, 0xffff, v9 ; encoding: [0xff,0x12,0x02,0x36,0xff,0xff,0x00,0x00]
 ; GFX10-NEXT:    v_lshl_or_b32 v5, v5, 16, v0 ; encoding: [0x05,0x00,0x6f,0xd7,0x05,0x21,0x01,0x04]
 ; GFX10-NEXT:    v_lshl_or_b32 v4, v10, 16, v1 ; encoding: [0x04,0x00,0x6f,0xd7,0x0a,0x21,0x05,0x04]
 ; GFX10-NEXT:    image_sample_c_d_o_g16 v[0:1], v[2:8], s[0:7], s[8:11] dmask:0x6 dim:SQ_RSRC_IMG_2D_ARRAY ; encoding: [0x29,0x06,0xe8,0xf0,0x02,0x00,0x40,0x00]
 ; GFX10-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0x70,0x3f,0x8c,0xbf]
 ; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: sample_c_d_o_2darray_V2:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    v_mov_b32_e32 v9, v2 ; encoding: [0x02,0x03,0x12,0x7e]
+; GFX11-NEXT:    v_mov_b32_e32 v10, v3 ; encoding: [0x03,0x03,0x14,0x7e]
+; GFX11-NEXT:    v_mov_b32_e32 v3, v1 ; encoding: [0x01,0x03,0x06,0x7e]
+; GFX11-NEXT:    v_mov_b32_e32 v2, v0 ; encoding: [0x00,0x03,0x04,0x7e]
+; GFX11-NEXT:    v_and_b32_e32 v0, 0xffff, v4 ; encoding: [0xff,0x08,0x00,0x36,0xff,0xff,0x00,0x00]
+; GFX11-NEXT:    v_and_b32_e32 v1, 0xffff, v9 ; encoding: [0xff,0x12,0x02,0x36,0xff,0xff,0x00,0x00]
+; GFX11-NEXT:    v_lshl_or_b32 v5, v5, 16, v0 ; encoding: [0x05,0x00,0x56,0xd6,0x05,0x21,0x01,0x04]
+; GFX11-NEXT:    v_lshl_or_b32 v4, v10, 16, v1 ; encoding: [0x04,0x00,0x56,0xd6,0x0a,0x21,0x05,0x04]
+; GFX11-NEXT:    image_sample_c_d_o_g16 v[0:1], v[2:8], s[0:7], s[8:11] dmask:0x6 dim:SQ_RSRC_IMG_2D_ARRAY ; encoding: [0x14,0x06,0xf0,0xf0,0x02,0x00,0x00,0x08]
+; GFX11-NEXT:    s_waitcnt vmcnt(0) ; encoding: [0xf7,0x03,0x89,0xbf]
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <2 x float> @llvm.amdgcn.image.sample.c.d.o.2darray.v2f32.f16.f32(i32 6, i32 %offset, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, float %slice, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <2 x float> %v
 }
 
 declare <4 x float> @llvm.amdgcn.image.sample.d.1d.v4f32.f16.f32(i32, half, half, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 declare <4 x float> @llvm.amdgcn.image.sample.d.2d.v4f32.f16.f32(i32, half, half, half, half, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 declare <4 x float> @llvm.amdgcn.image.sample.d.3d.v4f32.f16.f32(i32, half, half, half, half, half, half, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 declare <4 x float> @llvm.amdgcn.image.sample.c.d.1d.v4f32.f16.f32(i32, float, half, half, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 declare <4 x float> @llvm.amdgcn.image.sample.c.d.2d.v4f32.f16.f32(i32, float, half, half, half, half, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 declare <4 x float> @llvm.amdgcn.image.sample.d.cl.1d.v4f32.f16.f32(i32, half, half, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 declare <4 x float> @llvm.amdgcn.image.sample.d.cl.2d.v4f32.f16.f32(i32, half, half, half, half, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 declare <4 x float> @llvm.amdgcn.image.sample.c.d.cl.1d.v4f32.f16.f32(i32, float, half, half, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 declare <4 x float> @llvm.amdgcn.image.sample.c.d.cl.2d.v4f32.f16.f32(i32, float, half, half, half, half, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 
-declare <4 x float> @llvm.amdgcn.image.sample.cd.1d.v4f32.f16.f32(i32, half, half, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
-declare <4 x float> @llvm.amdgcn.image.sample.cd.2d.v4f32.f16.f32(i32, half, half, half, half, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
-declare <4 x float> @llvm.amdgcn.image.sample.c.cd.1d.v4f32.f16.f32(i32, float, half, half, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
-declare <4 x float> @llvm.amdgcn.image.sample.c.cd.2d.v4f32.f16.f32(i32, float, half, half, half, half, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
-declare <4 x float> @llvm.amdgcn.image.sample.cd.cl.1d.v4f32.f16.f32(i32, half, half, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
-declare <4 x float> @llvm.amdgcn.image.sample.cd.cl.2d.v4f32.f16.f32(i32, half, half, half, half, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
-declare <4 x float> @llvm.amdgcn.image.sample.c.cd.cl.1d.v4f32.f16.f32(i32, float, half, half, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
-declare <4 x float> @llvm.amdgcn.image.sample.c.cd.cl.2d.v4f32.f16.f32(i32, float, half, half, half, half, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
-
 declare float @llvm.amdgcn.image.sample.c.d.o.2darray.f16.f32.f32(i32, i32, float, half, half, half, half, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 declare <2 x float> @llvm.amdgcn.image.sample.c.d.o.2darray.v2f32.f16.f32(i32, i32, float, half, half, half, half, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 
 attributes #0 = { nounwind }
 attributes #1 = { nounwind readonly }
 attributes #2 = { nounwind readnone }

llvm/test/CodeGen/AMDGPU/llvm.amdgcn.image.sample.g16.ll

 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
 ; RUN: llc -march=amdgcn -mcpu=gfx1010 -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX10 %s
+; RUN: llc -march=amdgcn -mcpu=gfx1100 -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX10 %s
 
 define amdgpu_ps <4 x float> @sample_d_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, half %dsdh, half %dsdv, float %s) {
 ; GFX10-LABEL: sample_d_1d:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    image_sample_d_g16 v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D
 ; GFX10-NEXT:    s_waitcnt vmcnt(0)
 ; GFX10-NEXT:    ; return to shader part epilog
 main_body:

 ; GFX10-NEXT:    image_sample_c_d_cl_g16 v[0:3], v[2:7], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D
 ; GFX10-NEXT:    s_waitcnt vmcnt(0)
 ; GFX10-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x float> @llvm.amdgcn.image.sample.c.d.cl.2d.v4f32.f16.f32(i32 15, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
   ret <4 x float> %v
 }
 
-define amdgpu_ps <4 x float> @sample_cd_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, half %dsdh, half %dsdv, float %s) {
-; GFX10-LABEL: sample_cd_1d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_sample_cd_g16 v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D
-; GFX10-NEXT:    s_waitcnt vmcnt(0)
-; GFX10-NEXT:    ; return to shader part epilog
-main_body:
-  %v = call <4 x float> @llvm.amdgcn.image.sample.cd.1d.v4f32.f16.f32(i32 15, half %dsdh, half %dsdv, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
-  ret <4 x float> %v
-}
-
-define amdgpu_ps <4 x float> @sample_cd_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t) {
-; GFX10-LABEL: sample_cd_2d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    v_and_b32_e32 v2, 0xffff, v2
-; GFX10-NEXT:    v_and_b32_e32 v0, 0xffff, v0
-; GFX10-NEXT:    v_lshl_or_b32 v2, v3, 16, v2
-; GFX10-NEXT:    v_lshl_or_b32 v0, v1, 16, v0
-; GFX10-NEXT:    image_sample_cd_g16 v[0:3], [v0, v2, v4, v5], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D
-; GFX10-NEXT:    s_waitcnt vmcnt(0)
-; GFX10-NEXT:    ; return to shader part epilog
-main_body:
-  %v = call <4 x float> @llvm.amdgcn.image.sample.cd.2d.v4f32.f16.f32(i32 15, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
-  ret <4 x float> %v
-}
-
-define amdgpu_ps <4 x float> @sample_c_cd_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, half %dsdh, half %dsdv, float %s) {
-; GFX10-LABEL: sample_c_cd_1d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_sample_c_cd_g16 v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D
-; GFX10-NEXT:    s_waitcnt vmcnt(0)
-; GFX10-NEXT:    ; return to shader part epilog
-main_body:
-  %v = call <4 x float> @llvm.amdgcn.image.sample.c.cd.1d.v4f32.f16.f32(i32 15, float %zcompare, half %dsdh, half %dsdv, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
-  ret <4 x float> %v
-}
-
-define amdgpu_ps <4 x float> @sample_c_cd_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t) {
-; GFX10-LABEL: sample_c_cd_2d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    v_and_b32_e32 v3, 0xffff, v3
-; GFX10-NEXT:    v_and_b32_e32 v1, 0xffff, v1
-; GFX10-NEXT:    v_lshl_or_b32 v3, v4, 16, v3
-; GFX10-NEXT:    v_lshl_or_b32 v1, v2, 16, v1
-; GFX10-NEXT:    image_sample_c_cd_g16 v[0:3], [v0, v1, v3, v5, v6], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D
-; GFX10-NEXT:    s_waitcnt vmcnt(0)
-; GFX10-NEXT:    ; return to shader part epilog
-main_body:
-  %v = call <4 x float> @llvm.amdgcn.image.sample.c.cd.2d.v4f32.f16.f32(i32 15, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
-  ret <4 x float> %v
-}
-
-define amdgpu_ps <4 x float> @sample_cd_cl_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, half %dsdh, half %dsdv, float %s, float %clamp) {
-; GFX10-LABEL: sample_cd_cl_1d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_sample_cd_cl_g16 v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D
-; GFX10-NEXT:    s_waitcnt vmcnt(0)
-; GFX10-NEXT:    ; return to shader part epilog
-main_body:
-  %v = call <4 x float> @llvm.amdgcn.image.sample.cd.cl.1d.v4f32.f16.f32(i32 15, half %dsdh, half %dsdv, float %s, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
-  ret <4 x float> %v
-}
-
-define amdgpu_ps <4 x float> @sample_cd_cl_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, float %clamp) {
-; GFX10-LABEL: sample_cd_cl_2d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    v_and_b32_e32 v2, 0xffff, v2
-; GFX10-NEXT:    v_and_b32_e32 v0, 0xffff, v0
-; GFX10-NEXT:    v_lshl_or_b32 v2, v3, 16, v2
-; GFX10-NEXT:    v_lshl_or_b32 v0, v1, 16, v0
-; GFX10-NEXT:    image_sample_cd_cl_g16 v[0:3], [v0, v2, v4, v5, v6], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D
-; GFX10-NEXT:    s_waitcnt vmcnt(0)
-; GFX10-NEXT:    ; return to shader part epilog
-main_body:
-  %v = call <4 x float> @llvm.amdgcn.image.sample.cd.cl.2d.v4f32.f16.f32(i32 15, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
-  ret <4 x float> %v
-}
-
-define amdgpu_ps <4 x float> @sample_c_cd_cl_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, half %dsdh, half %dsdv, float %s, float %clamp) {
-; GFX10-LABEL: sample_c_cd_cl_1d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    image_sample_c_cd_cl_g16 v[0:3], v[0:4], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D
-; GFX10-NEXT:    s_waitcnt vmcnt(0)
-; GFX10-NEXT:    ; return to shader part epilog
-main_body:
-  %v = call <4 x float> @llvm.amdgcn.image.sample.c.cd.cl.1d.v4f32.f16.f32(i32 15, float %zcompare, half %dsdh, half %dsdv, float %s, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
-  ret <4 x float> %v
-}
-
-define amdgpu_ps <4 x float> @sample_c_cd_cl_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, float %clamp) {
-; GFX10-LABEL: sample_c_cd_cl_2d:
-; GFX10:       ; %bb.0: ; %main_body
-; GFX10-NEXT:    v_mov_b32_e32 v8, v2
-; GFX10-NEXT:    v_mov_b32_e32 v2, v0
-; GFX10-NEXT:    v_and_b32_e32 v0, 0xffff, v3
-; GFX10-NEXT:    v_and_b32_e32 v1, 0xffff, v1
-; GFX10-NEXT:    v_lshl_or_b32 v4, v4, 16, v0
-; GFX10-NEXT:    v_lshl_or_b32 v3, v8, 16, v1
-; GFX10-NEXT:    image_sample_c_cd_cl_g16 v[0:3], v[2:7], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D
-; GFX10-NEXT:    s_waitcnt vmcnt(0)
-; GFX10-NEXT:    ; return to shader part epilog
-main_body:
-  %v = call <4 x float> @llvm.amdgcn.image.sample.c.cd.cl.2d.v4f32.f16.f32(i32 15, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
-  ret <4 x float> %v
-}
-
 define amdgpu_ps float @sample_c_d_o_2darray_V1(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, i32 %offset, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, float %slice) {
 ; GFX10-LABEL: sample_c_d_o_2darray_V1:
 ; GFX10:       ; %bb.0: ; %main_body
 ; GFX10-NEXT:    v_mov_b32_e32 v9, v2
 ; GFX10-NEXT:    v_mov_b32_e32 v10, v3
 ; GFX10-NEXT:    v_mov_b32_e32 v3, v1
 ; GFX10-NEXT:    v_mov_b32_e32 v2, v0
 ; GFX10-NEXT:    v_and_b32_e32 v0, 0xffff, v4

 declare <4 x float> @llvm.amdgcn.image.sample.d.3d.v4f32.f16.f32(i32, half, half, half, half, half, half, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 declare <4 x float> @llvm.amdgcn.image.sample.c.d.1d.v4f32.f16.f32(i32, float, half, half, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 declare <4 x float> @llvm.amdgcn.image.sample.c.d.2d.v4f32.f16.f32(i32, float, half, half, half, half, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 declare <4 x float> @llvm.amdgcn.image.sample.d.cl.1d.v4f32.f16.f32(i32, half, half, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 declare <4 x float> @llvm.amdgcn.image.sample.d.cl.2d.v4f32.f16.f32(i32, half, half, half, half, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 declare <4 x float> @llvm.amdgcn.image.sample.c.d.cl.1d.v4f32.f16.f32(i32, float, half, half, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 declare <4 x float> @llvm.amdgcn.image.sample.c.d.cl.2d.v4f32.f16.f32(i32, float, half, half, half, half, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 
-declare <4 x float> @llvm.amdgcn.image.sample.cd.1d.v4f32.f16.f32(i32, half, half, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
-declare <4 x float> @llvm.amdgcn.image.sample.cd.2d.v4f32.f16.f32(i32, half, half, half, half, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
-declare <4 x float> @llvm.amdgcn.image.sample.c.cd.1d.v4f32.f16.f32(i32, float, half, half, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
-declare <4 x float> @llvm.amdgcn.image.sample.c.cd.2d.v4f32.f16.f32(i32, float, half, half, half, half, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
-declare <4 x float> @llvm.amdgcn.image.sample.cd.cl.1d.v4f32.f16.f32(i32, half, half, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
-declare <4 x float> @llvm.amdgcn.image.sample.cd.cl.2d.v4f32.f16.f32(i32, half, half, half, half, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
-declare <4 x float> @llvm.amdgcn.image.sample.c.cd.cl.1d.v4f32.f16.f32(i32, float, half, half, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
-declare <4 x float> @llvm.amdgcn.image.sample.c.cd.cl.2d.v4f32.f16.f32(i32, float, half, half, half, half, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
-
 declare float @llvm.amdgcn.image.sample.c.d.o.2darray.f16.f32.f32(i32, i32, float, half, half, half, half, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 declare <2 x float> @llvm.amdgcn.image.sample.c.d.o.2darray.v2f32.f16.f32(i32, i32, float, half, half, half, half, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
 
 attributes #0 = { nounwind }
 attributes #1 = { nounwind readonly }
 attributes #2 = { nounwind readnone }

llvm/test/CodeGen/AMDGPU/llvm.amdgcn.image.store.a16.d16.ll

 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
 ; RUN: llc -march=amdgcn -mcpu=gfx900 -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX9 %s
 ; RUN: llc -march=amdgcn -mcpu=gfx1010 -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX10 %s
+; RUN: llc -march=amdgcn -mcpu=gfx1100 -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX10 %s
 
 define amdgpu_ps void @store_f16_1d(<8 x i32> inreg %rsrc, <2 x i16> %coords, <2 x i32> %val) {
 ; GFX9-LABEL: store_f16_1d:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    image_store v[1:2], v0, s[0:7] dmask:0x1 unorm a16 d16
 ; GFX9-NEXT:    s_endpgm
 ;
 ; GFX10-LABEL: store_f16_1d:

llvm/test/CodeGen/AMDGPU/llvm.amdgcn.image.store.a16.ll

 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
 ; RUN: llc -march=amdgcn -mcpu=gfx900 -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX9 %s
 ; RUN: llc -march=amdgcn -mcpu=gfx1010 -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX10 %s
+; RUN: llc -march=amdgcn -mcpu=gfx1100 -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX10 %s
 
 define amdgpu_ps void @store_f32_1d(<8 x i32> inreg %rsrc, <2 x i16> %coords, <4 x float> %val) {
 ; GFX9-LABEL: store_f32_1d:
 ; GFX9:       ; %bb.0: ; %main_body
 ; GFX9-NEXT:    image_store v[1:4], v0, s[0:7] dmask:0x1 unorm a16
 ; GFX9-NEXT:    s_endpgm
 ;
 ; GFX10-LABEL: store_f32_1d:

llvm/test/CodeGen/AMDGPU/llvm.amdgcn.intersect_ray.ll

 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
-; RUN: llc -march=amdgcn -mcpu=gfx1013 -verify-machineinstrs < %s | FileCheck -check-prefixes=GCN,GFX1013 %s
-; RUN: llc -march=amdgcn -mcpu=gfx1030 -verify-machineinstrs < %s | FileCheck -check-prefixes=GCN,GFX1030 %s
+; RUN: llc -march=amdgcn -mcpu=gfx1013 -verify-machineinstrs < %s | FileCheck -check-prefixes=GCN,GFX10,GFX1013 %s
+; RUN: llc -march=amdgcn -mcpu=gfx1030 -verify-machineinstrs < %s | FileCheck -check-prefixes=GCN,GFX10,GFX1030 %s
 ; RUN: not --crash llc -march=amdgcn -mcpu=gfx1012 -verify-machineinstrs < %s 2>&1 | FileCheck -check-prefix=ERR %s
+; RUN: llc -march=amdgcn -mcpu=gfx1100 -verify-machineinstrs < %s | FileCheck -check-prefixes=GCN,GFX11 %s
 
 ; uint4 llvm.amdgcn.image.bvh.intersect.ray.i32.v4f32(uint node_ptr, float ray_extent, float3 ray_origin, float3 ray_dir, float3 ray_inv_dir, uint4 texture_descr)
 ; uint4 llvm.amdgcn.image.bvh.intersect.ray.i32.v4f16(uint node_ptr, float ray_extent, float3 ray_origin, half3 ray_dir, half3 ray_inv_dir, uint4 texture_descr)
 ; uint4 llvm.amdgcn.image.bvh.intersect.ray.i64.v4f32(ulong node_ptr, float ray_extent, float3 ray_origin, float3 ray_dir, float3 ray_inv_dir, uint4 texture_descr)
 ; uint4 llvm.amdgcn.image.bvh.intersect.ray.i64.v4f16(ulong node_ptr, float ray_extent, float3 ray_origin, half3 ray_dir, half3 ray_inv_dir, uint4 texture_descr)
 
 declare <4 x i32> @llvm.amdgcn.image.bvh.intersect.ray.i32.v4f32(i32, float, <3 x float>, <3 x float>, <3 x float>, <4 x i32>)
 declare <4 x i32> @llvm.amdgcn.image.bvh.intersect.ray.i32.v4f16(i32, float, <3 x float>, <3 x half>, <3 x half>, <4 x i32>)
main_body:
   %ray_inv_dir1 = insertelement <3 x float> %ray_inv_dir0, float %ray_inv_dir_y, i32 1
   %ray_inv_dir = insertelement <3 x float> %ray_inv_dir1, float %ray_inv_dir_z, i32 2
   %v = call <4 x i32> @llvm.amdgcn.image.bvh.intersect.ray.i32.v4f32(i32 %node_ptr, float %ray_extent, <3 x float> %ray_origin, <3 x float> %ray_dir, <3 x float> %ray_inv_dir, <4 x i32> %tdescr)
  %r = bitcast <4 x i32> %v to <4 x float>
  ret <4 x float> %r
 }
 
 define amdgpu_ps <4 x float> @image_bvh_intersect_ray_a16(i32 inreg %node_ptr, float inreg %ray_extent, <3 x float> inreg %ray_origin, <3 x half> inreg %ray_dir, <3 x half> inreg %ray_inv_dir, <4 x i32> inreg %tdescr) {
-; GCN-LABEL: image_bvh_intersect_ray_a16:
-; GCN:       ; %bb.0: ; %main_body
-; GCN-NEXT:    s_mov_b32 s15, s12
-; GCN-NEXT:    s_mov_b32 s12, s9
-; GCN-NEXT:    s_lshr_b32 s9, s7, 16
-; GCN-NEXT:    s_pack_ll_b32_b16 s6, s6, s7
-; GCN-NEXT:    s_pack_ll_b32_b16 s7, s9, s8
-; GCN-NEXT:    v_mov_b32_e32 v0, s0
-; GCN-NEXT:    v_mov_b32_e32 v1, s1
-; GCN-NEXT:    v_mov_b32_e32 v2, s2
-; GCN-NEXT:    v_mov_b32_e32 v3, s3
-; GCN-NEXT:    v_mov_b32_e32 v4, s4
-; GCN-NEXT:    v_mov_b32_e32 v5, s5
-; GCN-NEXT:    v_mov_b32_e32 v6, s6
-; GCN-NEXT:    v_mov_b32_e32 v7, s7
-; GCN-NEXT:    s_mov_b32 s14, s11
-; GCN-NEXT:    s_mov_b32 s13, s10
-; GCN-NEXT:    image_bvh_intersect_ray v[0:3], v[0:7], s[12:15] a16
-; GCN-NEXT:    s_waitcnt vmcnt(0)
-; GCN-NEXT:    ; return to shader part epilog
+; GFX10-LABEL: image_bvh_intersect_ray_a16:
+; GFX10:       ; %bb.0: ; %main_body
+; GFX10-NEXT:    s_mov_b32 s15, s12
+; GFX10-NEXT:    s_mov_b32 s12, s9
+; GFX10-NEXT:    s_lshr_b32 s9, s7, 16
+; GFX10-NEXT:    s_pack_ll_b32_b16 s6, s6, s7
+; GFX10-NEXT:    s_pack_ll_b32_b16 s7, s9, s8
+; GFX10-NEXT:    v_mov_b32_e32 v0, s0
+; GFX10-NEXT:    v_mov_b32_e32 v1, s1
+; GFX10-NEXT:    v_mov_b32_e32 v2, s2
+; GFX10-NEXT:    v_mov_b32_e32 v3, s3
+; GFX10-NEXT:    v_mov_b32_e32 v4, s4
+; GFX10-NEXT:    v_mov_b32_e32 v5, s5
+; GFX10-NEXT:    v_mov_b32_e32 v6, s6
+; GFX10-NEXT:    v_mov_b32_e32 v7, s7
+; GFX10-NEXT:    s_mov_b32 s14, s11
+; GFX10-NEXT:    s_mov_b32 s13, s10
+; GFX10-NEXT:    image_bvh_intersect_ray v[0:3], v[0:7], s[12:15] a16
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: image_bvh_intersect_ray_a16:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    v_mov_b32_e32 v0, s2
+; GFX11-NEXT:    v_mov_b32_e32 v1, s3
+; GFX11-NEXT:    s_lshr_b32 s2, s7, 16
+; GFX11-NEXT:    s_lshr_b32 s3, s5, 16
+; GFX11-NEXT:    v_mov_b32_e32 v2, s4
+; GFX11-NEXT:    s_pack_ll_b32_b16 s2, s3, s2
+; GFX11-NEXT:    s_pack_ll_b32_b16 s3, s5, s7
+; GFX11-NEXT:    s_pack_ll_b32_b16 s4, s6, s8
+; GFX11-NEXT:    v_mov_b32_e32 v3, s3
+; GFX11-NEXT:    v_mov_b32_e32 v4, s2
+; GFX11-NEXT:    v_mov_b32_e32 v5, s4
+; GFX11-NEXT:    v_mov_b32_e32 v6, s0
+; GFX11-NEXT:    v_mov_b32_e32 v7, s1
+; GFX11-NEXT:    s_mov_b32 s15, s12
+; GFX11-NEXT:    s_mov_b32 s14, s11
+; GFX11-NEXT:    s_mov_b32 s13, s10
+; GFX11-NEXT:    s_mov_b32 s12, s9
+; GFX11-NEXT:    image_bvh_intersect_ray v[0:3], [v6, v7, v[0:2], v[3:5]], s[12:15] a16
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x i32> @llvm.amdgcn.image.bvh.intersect.ray.i32.v4f16(i32 %node_ptr, float %ray_extent, <3 x float> %ray_origin, <3 x half> %ray_dir, <3 x half> %ray_inv_dir, <4 x i32> %tdescr)
   %r = bitcast <4 x i32> %v to <4 x float>
   ret <4 x float> %r
 }
 
 ; Arguments are flattened to represent the actual VGPR_A layout, so we have no
 ; extra moves in the generated kernel.
main_body:
   %ray_inv_dir1 = insertelement <3 x float> %ray_inv_dir0, float %ray_inv_dir_y, i32 1
   %ray_inv_dir = insertelement <3 x float> %ray_inv_dir1, float %ray_inv_dir_z, i32 2
   %v = call <4 x i32> @llvm.amdgcn.image.bvh.intersect.ray.i64.v4f32(i64 %node_ptr, float %ray_extent, <3 x float> %ray_origin, <3 x float> %ray_dir, <3 x float> %ray_inv_dir, <4 x i32> %tdescr)
  %r = bitcast <4 x i32> %v to <4 x float>
  ret <4 x float> %r
 }
 
 define amdgpu_ps <4 x float> @image_bvh64_intersect_ray_a16(i64 inreg %node_ptr, float inreg %ray_extent, <3 x float> inreg %ray_origin, <3 x half> inreg %ray_dir, <3 x half> inreg %ray_inv_dir, <4 x i32> inreg %tdescr) {
-; GCN-LABEL: image_bvh64_intersect_ray_a16:
-; GCN:       ; %bb.0: ; %main_body
-; GCN-NEXT:    s_mov_b32 s14, s12
-; GCN-NEXT:    s_mov_b32 s12, s10
-; GCN-NEXT:    s_lshr_b32 s10, s8, 16
-; GCN-NEXT:    s_pack_ll_b32_b16 s7, s7, s8
-; GCN-NEXT:    s_pack_ll_b32_b16 s8, s10, s9
-; GCN-NEXT:    v_mov_b32_e32 v0, s0
-; GCN-NEXT:    v_mov_b32_e32 v1, s1
-; GCN-NEXT:    v_mov_b32_e32 v2, s2
-; GCN-NEXT:    v_mov_b32_e32 v3, s3
-; GCN-NEXT:    v_mov_b32_e32 v4, s4
-; GCN-NEXT:    v_mov_b32_e32 v5, s5
-; GCN-NEXT:    v_mov_b32_e32 v6, s6
-; GCN-NEXT:    v_mov_b32_e32 v7, s7
-; GCN-NEXT:    v_mov_b32_e32 v8, s8
-; GCN-NEXT:    s_mov_b32 s15, s13
-; GCN-NEXT:    s_mov_b32 s13, s11
-; GCN-NEXT:    image_bvh64_intersect_ray v[0:3], v[0:15], s[12:15] a16
-; GCN-NEXT:    s_waitcnt vmcnt(0)
-; GCN-NEXT:    ; return to shader part epilog
+; GFX10-LABEL: image_bvh64_intersect_ray_a16:
+; GFX10:       ; %bb.0: ; %main_body
+; GFX10-NEXT:    s_mov_b32 s14, s12
+; GFX10-NEXT:    s_mov_b32 s12, s10
+; GFX10-NEXT:    s_lshr_b32 s10, s8, 16
+; GFX10-NEXT:    s_pack_ll_b32_b16 s7, s7, s8
+; GFX10-NEXT:    s_pack_ll_b32_b16 s8, s10, s9
+; GFX10-NEXT:    v_mov_b32_e32 v0, s0
+; GFX10-NEXT:    v_mov_b32_e32 v1, s1
+; GFX10-NEXT:    v_mov_b32_e32 v2, s2
+; GFX10-NEXT:    v_mov_b32_e32 v3, s3
+; GFX10-NEXT:    v_mov_b32_e32 v4, s4
+; GFX10-NEXT:    v_mov_b32_e32 v5, s5
+; GFX10-NEXT:    v_mov_b32_e32 v6, s6
+; GFX10-NEXT:    v_mov_b32_e32 v7, s7
+; GFX10-NEXT:    v_mov_b32_e32 v8, s8
+; GFX10-NEXT:    s_mov_b32 s15, s13
+; GFX10-NEXT:    s_mov_b32 s13, s11
+; GFX10-NEXT:    image_bvh64_intersect_ray v[0:3], v[0:15], s[12:15] a16
+; GFX10-NEXT:    s_waitcnt vmcnt(0)
+; GFX10-NEXT:    ; return to shader part epilog
+;
+; GFX11-LABEL: image_bvh64_intersect_ray_a16:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    v_mov_b32_e32 v0, s3
+; GFX11-NEXT:    v_mov_b32_e32 v6, s0
+; GFX11-NEXT:    s_lshr_b32 s0, s8, 16
+; GFX11-NEXT:    s_lshr_b32 s3, s6, 16
+; GFX11-NEXT:    v_mov_b32_e32 v7, s1
+; GFX11-NEXT:    s_pack_ll_b32_b16 s0, s3, s0
+; GFX11-NEXT:    s_pack_ll_b32_b16 s1, s6, s8
+; GFX11-NEXT:    s_pack_ll_b32_b16 s3, s7, s9
+; GFX11-NEXT:    v_mov_b32_e32 v1, s4
+; GFX11-NEXT:    v_mov_b32_e32 v2, s5
+; GFX11-NEXT:    v_mov_b32_e32 v3, s1
+; GFX11-NEXT:    v_mov_b32_e32 v4, s0
+; GFX11-NEXT:    v_mov_b32_e32 v5, s3
+; GFX11-NEXT:    v_mov_b32_e32 v8, s2
+; GFX11-NEXT:    s_mov_b32 s15, s13
+; GFX11-NEXT:    s_mov_b32 s14, s12
+; GFX11-NEXT:    s_mov_b32 s13, s11
+; GFX11-NEXT:    s_mov_b32 s12, s10
+; GFX11-NEXT:    image_bvh64_intersect_ray v[0:3], [v[6:7], v8, v[0:2], v[3:5]], s[12:15] a16
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    ; return to shader part epilog
 main_body:
   %v = call <4 x i32> @llvm.amdgcn.image.bvh.intersect.ray.i64.v4f16(i64 %node_ptr, float %ray_extent, <3 x float> %ray_origin, <3 x half> %ray_dir, <3 x half> %ray_inv_dir, <4 x i32> %tdescr)
   %r = bitcast <4 x i32> %v to <4 x float>
   ret <4 x float> %r
 }
 
 ; TODO: NSA reassign is very limited and cannot work with VGPR tuples and subregs.
 

 ; GFX1030-NEXT:    flat_load_dword v1, v[2:3]
 ; GFX1030-NEXT:    v_mov_b32_e32 v2, 0
 ; GFX1030-NEXT:    v_mov_b32_e32 v3, 1.0
 ; GFX1030-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
 ; GFX1030-NEXT:    image_bvh_intersect_ray v[0:3], v[0:15], s[0:3]
 ; GFX1030-NEXT:    s_waitcnt vmcnt(0)
 ; GFX1030-NEXT:    flat_store_dwordx4 v[0:1], v[0:3]
 ; GFX1030-NEXT:    s_endpgm
+;
+; GFX11-LABEL: image_bvh_intersect_ray_nsa_reassign:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    s_load_b128 s[4:7], s[0:1], 0x24
+; GFX11-NEXT:    v_lshlrev_b32_e32 v2, 2, v0
+; GFX11-NEXT:    s_load_b128 s[0:3], s[0:1], 0x34
+; GFX11-NEXT:    v_mov_b32_e32 v4, 4.0
+; GFX11-NEXT:    v_mov_b32_e32 v5, 0x40a00000
+; GFX11-NEXT:    v_mov_b32_e32 v6, 0
+; GFX11-NEXT:    v_mov_b32_e32 v7, 1.0
+; GFX11-NEXT:    v_mov_b32_e32 v8, 2.0
+; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
+; GFX11-NEXT:    v_add_co_u32 v0, s4, s4, v2
+; GFX11-NEXT:    v_add_co_ci_u32_e64 v1, null, s5, 0, s4
+; GFX11-NEXT:    v_add_co_u32 v2, s4, s6, v2
+; GFX11-NEXT:    v_add_co_ci_u32_e64 v3, null, s7, 0, s4
+; GFX11-NEXT:    flat_load_b32 v9, v[0:1]
+; GFX11-NEXT:    flat_load_b32 v10, v[2:3]
+; GFX11-NEXT:    v_mov_b32_e32 v0, 0x40c00000
+; GFX11-NEXT:    v_mov_b32_e32 v1, 0x40e00000
+; GFX11-NEXT:    v_mov_b32_e32 v2, 0x41000000
+; GFX11-NEXT:    v_mov_b32_e32 v3, 0x40400000
+; GFX11-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
+; GFX11-NEXT:    image_bvh_intersect_ray v[0:3], [v9, v10, v[6:8], v[3:5], v[0:2]], s[0:3]
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    flat_store_b128 v[0:1], v[0:3]
+; GFX11-NEXT:    s_endpgm
 main_body:
   %lid = tail call i32 @llvm.amdgcn.workitem.id.x()
   %gep_node_ptr = getelementptr inbounds i32, i32* %p_node_ptr, i32 %lid
   %node_ptr = load i32, i32* %gep_node_ptr, align 4
   %gep_ray = getelementptr inbounds float, float* %p_ray, i32 %lid
   %ray_extent = load float, float* %gep_ray, align 4
   %ray_origin0 = insertelement <3 x float> undef, float 0.0, i32 0
   %ray_origin1 = insertelement <3 x float> %ray_origin0, float 1.0, i32 1

 ; GFX1030-NEXT:    flat_load_dword v1, v[2:3]
 ; GFX1030-NEXT:    v_mov_b32_e32 v2, 0
 ; GFX1030-NEXT:    v_mov_b32_e32 v3, 1.0
 ; GFX1030-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
 ; GFX1030-NEXT:    image_bvh_intersect_ray v[0:3], v[0:7], s[0:3] a16
 ; GFX1030-NEXT:    s_waitcnt vmcnt(0)
 ; GFX1030-NEXT:    flat_store_dwordx4 v[0:1], v[0:3]
 ; GFX1030-NEXT:    s_endpgm
+;
+; GFX11-LABEL: image_bvh_intersect_ray_a16_nsa_reassign:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    s_load_b128 s[4:7], s[0:1], 0x24
+; GFX11-NEXT:    v_lshlrev_b32_e32 v2, 2, v0
+; GFX11-NEXT:    s_load_b128 s[0:3], s[0:1], 0x34
+; GFX11-NEXT:    v_mov_b32_e32 v4, 1.0
+; GFX11-NEXT:    v_mov_b32_e32 v5, 2.0
+; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
+; GFX11-NEXT:    v_add_co_u32 v0, s4, s4, v2
+; GFX11-NEXT:    v_add_co_ci_u32_e64 v1, null, s5, 0, s4
+; GFX11-NEXT:    v_add_co_u32 v2, s4, s6, v2
+; GFX11-NEXT:    v_add_co_ci_u32_e64 v3, null, s7, 0, s4
+; GFX11-NEXT:    flat_load_b32 v6, v[0:1]
+; GFX11-NEXT:    flat_load_b32 v7, v[2:3]
+; GFX11-NEXT:    v_mov_b32_e32 v0, 0x46004200
+; GFX11-NEXT:    v_mov_b32_e32 v1, 0x47004400
+; GFX11-NEXT:    v_mov_b32_e32 v2, 0x48004500
+; GFX11-NEXT:    v_mov_b32_e32 v3, 0
+; GFX11-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
+; GFX11-NEXT:    image_bvh_intersect_ray v[0:3], [v6, v7, v[3:5], v[0:2]], s[0:3] a16
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    flat_store_b128 v[0:1], v[0:3]
+; GFX11-NEXT:    s_endpgm
 main_body:
   %lid = tail call i32 @llvm.amdgcn.workitem.id.x()
   %gep_node_ptr = getelementptr inbounds i32, i32* %p_node_ptr, i32 %lid
   %node_ptr = load i32, i32* %gep_node_ptr, align 4
   %gep_ray = getelementptr inbounds float, float* %p_ray, i32 %lid
   %ray_extent = load float, float* %gep_ray, align 4
   %ray_origin0 = insertelement <3 x float> undef, float 0.0, i32 0
   %ray_origin1 = insertelement <3 x float> %ray_origin0, float 1.0, i32 1

 ; GFX1030-NEXT:    flat_load_dword v2, v[0:1]
 ; GFX1030-NEXT:    v_mov_b32_e32 v1, 0x102
 ; GFX1030-NEXT:    v_mov_b32_e32 v0, 0xb36211c7
 ; GFX1030-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
 ; GFX1030-NEXT:    image_bvh64_intersect_ray v[0:3], v[0:15], s[0:3]
 ; GFX1030-NEXT:    s_waitcnt vmcnt(0)
 ; GFX1030-NEXT:    flat_store_dwordx4 v[0:1], v[0:3]
 ; GFX1030-NEXT:    s_endpgm
+;
+; GFX11-LABEL: image_bvh64_intersect_ray_nsa_reassign:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    s_load_b64 s[4:5], s[0:1], 0x24
+; GFX11-NEXT:    v_lshlrev_b32_e32 v0, 2, v0
+; GFX11-NEXT:    s_load_b128 s[0:3], s[0:1], 0x34
+; GFX11-NEXT:    v_mov_b32_e32 v2, 0x41000000
+; GFX11-NEXT:    v_mov_b32_e32 v3, 0x40400000
+; GFX11-NEXT:    v_mov_b32_e32 v4, 4.0
+; GFX11-NEXT:    v_mov_b32_e32 v5, 0x40a00000
+; GFX11-NEXT:    v_mov_b32_e32 v6, 0
+; GFX11-NEXT:    v_mov_b32_e32 v7, 1.0
+; GFX11-NEXT:    v_mov_b32_e32 v8, 2.0
+; GFX11-NEXT:    v_mov_b32_e32 v9, 0xb36211c7
+; GFX11-NEXT:    v_mov_b32_e32 v10, 0x102
+; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
+; GFX11-NEXT:    v_add_co_u32 v0, s4, s4, v0
+; GFX11-NEXT:    v_add_co_ci_u32_e64 v1, null, s5, 0, s4
+; GFX11-NEXT:    flat_load_b32 v11, v[0:1]
+; GFX11-NEXT:    v_mov_b32_e32 v0, 0x40c00000
+; GFX11-NEXT:    v_mov_b32_e32 v1, 0x40e00000
+; GFX11-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
+; GFX11-NEXT:    image_bvh64_intersect_ray v[0:3], [v[9:10], v11, v[6:8], v[3:5], v[0:2]], s[0:3]
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    flat_store_b128 v[0:1], v[0:3]
+; GFX11-NEXT:    s_endpgm
 main_body:
   %lid = tail call i32 @llvm.amdgcn.workitem.id.x()
   %gep_ray = getelementptr inbounds float, float* %p_ray, i32 %lid
   %ray_extent = load float, float* %gep_ray, align 4
   %ray_origin0 = insertelement <3 x float> undef, float 0.0, i32 0
   %ray_origin1 = insertelement <3 x float> %ray_origin0, float 1.0, i32 1
   %ray_origin = insertelement <3 x float> %ray_origin1, float 2.0, i32 2
   %ray_dir0 = insertelement <3 x float> undef, float 3.0, i32 0

 ; GFX1030-NEXT:    flat_load_dword v2, v[0:1]
 ; GFX1030-NEXT:    v_mov_b32_e32 v1, 0x102
 ; GFX1030-NEXT:    v_mov_b32_e32 v0, 0xb36211c6
 ; GFX1030-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
 ; GFX1030-NEXT:    image_bvh64_intersect_ray v[0:3], v[0:15], s[0:3] a16
 ; GFX1030-NEXT:    s_waitcnt vmcnt(0)
 ; GFX1030-NEXT:    flat_store_dwordx4 v[0:1], v[0:3]
 ; GFX1030-NEXT:    s_endpgm
+;
+; GFX11-LABEL: image_bvh64_intersect_ray_a16_nsa_reassign:
+; GFX11:       ; %bb.0: ; %main_body
+; GFX11-NEXT:    s_load_b64 s[4:5], s[0:1], 0x24
+; GFX11-NEXT:    v_lshlrev_b32_e32 v0, 2, v0
+; GFX11-NEXT:    s_load_b128 s[0:3], s[0:1], 0x34
+; GFX11-NEXT:    v_mov_b32_e32 v2, 0x48004500
+; GFX11-NEXT:    v_mov_b32_e32 v3, 0
+; GFX11-NEXT:    v_mov_b32_e32 v4, 1.0
+; GFX11-NEXT:    v_mov_b32_e32 v5, 2.0
+; GFX11-NEXT:    v_mov_b32_e32 v6, 0xb36211c6
+; GFX11-NEXT:    v_mov_b32_e32 v7, 0x102
+; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
+; GFX11-NEXT:    v_add_co_u32 v0, s4, s4, v0
+; GFX11-NEXT:    v_add_co_ci_u32_e64 v1, null, s5, 0, s4
+; GFX11-NEXT:    flat_load_b32 v8, v[0:1]
+; GFX11-NEXT:    v_mov_b32_e32 v0, 0x46004200
+; GFX11-NEXT:    v_mov_b32_e32 v1, 0x47004400
+; GFX11-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
+; GFX11-NEXT:    image_bvh64_intersect_ray v[0:3], [v[6:7], v8, v[3:5], v[0:2]], s[0:3] a16
+; GFX11-NEXT:    s_waitcnt vmcnt(0)
+; GFX11-NEXT:    flat_store_b128 v[0:1], v[0:3]
+; GFX11-NEXT:    s_endpgm
 main_body:
   %lid = tail call i32 @llvm.amdgcn.workitem.id.x()
   %gep_ray = getelementptr inbounds float, float* %p_ray, i32 %lid
   %ray_extent = load float, float* %gep_ray, align 4
   %ray_origin0 = insertelement <3 x float> undef, float 0.0, i32 0
   %ray_origin1 = insertelement <3 x float> %ray_origin0, float 1.0, i32 1
   %ray_origin = insertelement <3 x float> %ray_origin1, float 2.0, i32 2
   %ray_dir0 = insertelement <3 x half> undef, half 3.0, i32 0

llvm/test/CodeGen/AMDGPU/merge-image-load-gfx11.mir

+# RUN: llc -march=amdgcn -mcpu=gfx1100 -verify-machineinstrs -run-pass si-load-store-opt -o - %s | FileCheck -check-prefix=GFX11 %s
+
+# GFX11-LABEL: name: image_load_merged_v1v3
+# GFX11: %{{[0-9]+}}:vreg_128 = IMAGE_LOAD_V4_V2_gfx11 %5, %3, 15, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec, implicit $exec :: (dereferenceable load (s128), align 4, addrspace 4)
+# GFX11: %{{[0-9]+}}:vgpr_32 = COPY %8.sub0
+# GFX11: %{{[0-9]+}}:vreg_96 = COPY killed %8.sub1_sub2_sub3
+
+name: image_load_merged_v1v3
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vreg_64 = BUFFER_LOAD_DWORDX2_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = IMAGE_LOAD_V1_V2_gfx11 %5:vreg_64, %3:sgpr_256, 1, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_LOAD_V3_V2_gfx11 %5:vreg_64, %3:sgpr_256, 14, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+# GFX11-LABEL: name: image_load_merged_v1v3_reversed
+# GFX11: %{{[0-9]+}}:vreg_128 = IMAGE_LOAD_V4_V2_gfx11 %5, %3, 15, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec, implicit $exec :: (dereferenceable load (s128), addrspace 4)
+# GFX11: %{{[0-9]+}}:vgpr_32 = COPY %8.sub3
+# GFX11: %{{[0-9]+}}:vreg_96 = COPY killed %8.sub0_sub1_sub2
+
+name: image_load_merged_v1v3_reversed
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vreg_64 = BUFFER_LOAD_DWORDX2_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = IMAGE_LOAD_V1_V2_gfx11 %5:vreg_64, %3:sgpr_256, 8, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_LOAD_V3_V2_gfx11 %5:vreg_64, %3:sgpr_256, 7, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+# GFX11-LABEL: name: image_load_merged_v2v2
+# GFX11: %{{[0-9]+}}:vreg_128 = IMAGE_LOAD_V4_V2_gfx11 %5, %3, 15, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec, implicit $exec :: (dereferenceable load (s128), addrspace 4)
+# GFX11: %{{[0-9]+}}:vreg_64 = COPY %8.sub0_sub1
+# GFX11: %{{[0-9]+}}:vreg_64 = COPY killed %8.sub2_sub3
+
+name: image_load_merged_v2v2
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vreg_64 = BUFFER_LOAD_DWORDX2_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vreg_64 = IMAGE_LOAD_V2_V2_gfx11 %5:vreg_64, %3:sgpr_256, 3, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 8, align 16, addrspace 4)
+    %7:vreg_64 = IMAGE_LOAD_V2_V2_gfx11 %5:vreg_64, %3:sgpr_256, 12, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 8, align 16, addrspace 4)
+...
+---
+
+# GFX11-LABEL: name: image_load_merged_v2v2_reversed
+# GFX11: %{{[0-9]+}}:vreg_128 = IMAGE_LOAD_V4_V2_gfx11 %5, %3, 15, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec, implicit $exec :: (dereferenceable load (s128), addrspace 4)
+# GFX11: %{{[0-9]+}}:vreg_64 = COPY %8.sub2_sub3
+# GFX11: %{{[0-9]+}}:vreg_64 = COPY killed %8.sub0_sub1
+
+name: image_load_merged_v2v2_reversed
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vreg_64 = BUFFER_LOAD_DWORDX2_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vreg_64 = IMAGE_LOAD_V2_V2_gfx11 %5:vreg_64, %3:sgpr_256, 12, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 8, align 16, addrspace 4)
+    %7:vreg_64 = IMAGE_LOAD_V2_V2_gfx11 %5:vreg_64, %3:sgpr_256, 3, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 8, align 16, addrspace 4)
+...
+---
+
+# GFX11-LABEL: name: image_load_merged_v3v1
+# GFX11: %{{[0-9]+}}:vreg_128 = IMAGE_LOAD_V4_V2_gfx11 %5, %3, 15, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec, implicit $exec :: (dereferenceable load (s128), addrspace 4)
+# GFX11: %{{[0-9]+}}:vreg_96 = COPY %8.sub0_sub1_sub2
+# GFX11: %{{[0-9]+}}:vgpr_32 = COPY killed %8.sub3
+
+name: image_load_merged_v3v1
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vreg_64 = BUFFER_LOAD_DWORDX2_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vreg_96 = IMAGE_LOAD_V3_V2_gfx11 %5:vreg_64, %3:sgpr_256, 7, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+    %7:vgpr_32 = IMAGE_LOAD_V1_V2_gfx11 %5:vreg_64, %3:sgpr_256, 8, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+...
+---
+
+# GFX11-LABEL: name: image_load_merged_v3v1_reversed
+# GFX11: %{{[0-9]+}}:vreg_128 = IMAGE_LOAD_V4_V2_gfx11 %5, %3, 15, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec, implicit $exec :: (dereferenceable load (s128), align 4, addrspace 4)
+# GFX11: %{{[0-9]+}}:vreg_96 = COPY %8.sub1_sub2_sub3
+# GFX11: %{{[0-9]+}}:vgpr_32 = COPY killed %8.sub0
+
+name: image_load_merged_v3v1_reversed
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vreg_64 = BUFFER_LOAD_DWORDX2_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vreg_96 = IMAGE_LOAD_V3_V2_gfx11 %5:vreg_64, %3:sgpr_256, 14, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+    %7:vgpr_32 = IMAGE_LOAD_V1_V2_gfx11 %5:vreg_64, %3:sgpr_256, 1, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+...
+---
+
+# GFX11-LABEL: name: image_load_divided_merged
+# GFX11: %{{[0-9]+}}:vreg_128 = IMAGE_LOAD_V4_V2_gfx11 %5, %3, 15, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec, implicit $exec :: (dereferenceable load (s128), addrspace 4)
+
+name: image_load_divided_merged
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vreg_64 = BUFFER_LOAD_DWORDX2_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = IMAGE_LOAD_V1_V2_gfx11 %5:vreg_64, %3:sgpr_256, 8, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_64 = BUFFER_LOAD_DWORDX2_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %8:vreg_128 = BUFFER_LOAD_DWORDX4_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %9:vreg_96 = IMAGE_LOAD_V3_V2_gfx11 %7:vreg_64, %3:sgpr_256, 7, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+    %10:vreg_128 = BUFFER_LOAD_DWORDX4_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %11:vreg_96 = IMAGE_LOAD_V3_V2_gfx11 %5:vreg_64, %3:sgpr_256, 7, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+# GFX11-LABEL: name: image_load_divided_not_merged
+# GFX11: %{{[0-9]+}}:vgpr_32 = IMAGE_LOAD_V1_V2_gfx11 %5, %3, 8, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s32), addrspace 4)
+# GFX11: %{{[0-9]+}}:vreg_96 = IMAGE_LOAD_V3_V2_gfx11 %5, %3, 7, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s96), align 16, addrspace 4)
+
+name: image_load_divided_not_merged
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vreg_128 = COPY %2
+    %5:vreg_64 = BUFFER_LOAD_DWORDX2_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = IMAGE_LOAD_V1_V2_gfx11 %5:vreg_64, %3:sgpr_256, 8, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    IMAGE_STORE_V4_V2 %4:vreg_128, %5:vreg_64, %3:sgpr_256, 15, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (store 16)
+    %7:vreg_96 = IMAGE_LOAD_V3_V2_gfx11 %5:vreg_64, %3:sgpr_256, 7, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+# GFX11-LABEL: name: image_load_dmask_overlapped_not_merged
+# GFX11: %{{[0-9]+}}:vgpr_32 = IMAGE_LOAD_V1_V2_gfx11 %5, %3, 4, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s32), addrspace 4)
+# GFX11: %{{[0-9]+}}:vreg_96 = IMAGE_LOAD_V3_V2_gfx11 %5, %3, 7, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s96), align 16, addrspace 4)
+
+name: image_load_dmask_overlapped_not_merged
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vreg_64 = BUFFER_LOAD_DWORDX2_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = IMAGE_LOAD_V1_V2_gfx11 %5:vreg_64, %3:sgpr_256, 4, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_LOAD_V3_V2_gfx11 %5:vreg_64, %3:sgpr_256, 7, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+# GFX11-LABEL: name: image_load_dmask_not_disjoint_not_merged
+# GFX11: %{{[0-9]+}}:vgpr_32 = IMAGE_LOAD_V1_V2_gfx11 %5, %3, 4, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s32), addrspace 4)
+# GFX11: %{{[0-9]+}}:vreg_96 = IMAGE_LOAD_V3_V2_gfx11 %5, %3, 11, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s96), align 16, addrspace 4)
+
+name: image_load_dmask_not_disjoint_not_merged
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vreg_64 = BUFFER_LOAD_DWORDX2_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = IMAGE_LOAD_V1_V2_gfx11 %5:vreg_64, %3:sgpr_256, 4, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_LOAD_V3_V2_gfx11 %5:vreg_64, %3:sgpr_256, 11, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+# GFX11-LABEL: name: image_load_not_merged_0
+# GFX11: %{{[0-9]+}}:vgpr_32 = IMAGE_LOAD_V1_V2_gfx11 %5, %3, 8, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s32), addrspace 4)
+# GFX11: %{{[0-9]+}}:vreg_96 = IMAGE_LOAD_V3_V2_gfx11 %6, %3, 7, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s96), align 16, addrspace 4)
+
+name: image_load_not_merged_0
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vreg_64 = BUFFER_LOAD_DWORDX2_OFFSET %2, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vreg_64 = BUFFER_LOAD_DWORDX2_OFFSET %2, 1, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %7:vgpr_32 = IMAGE_LOAD_V1_V2_gfx11 %5, %3, 8, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %8:vreg_96 = IMAGE_LOAD_V3_V2_gfx11 %6, %3, 7, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+# GFX11-LABEL: name: image_load_not_merged_1
+# GFX11: %{{[0-9]+}}:vgpr_32 = IMAGE_LOAD_V1_V2_gfx11 %6, %3, 8, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s32), addrspace 4)
+# GFX11: %{{[0-9]+}}:vreg_96 = IMAGE_LOAD_V3_V2_gfx11 %6, %4, 7, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s96), align 16, addrspace 4)
+
+name: image_load_not_merged_1
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %5:vgpr_32 = COPY %2.sub3
+    %6:vreg_64 = BUFFER_LOAD_DWORDX2_OFFSET %2, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %7:vgpr_32 = IMAGE_LOAD_V1_V2_gfx11 %6, %3, 8, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %8:vreg_96 = IMAGE_LOAD_V3_V2_gfx11 %6, %4, 7, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+# GFX11-LABEL: name: image_load_not_merged_3
+# GFX11: %{{[0-9]+}}:vgpr_32 = IMAGE_LOAD_V1_V2_gfx11 %5, %3, 8, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s32), addrspace 4)
+# GFX11: %{{[0-9]+}}:vreg_96 = IMAGE_LOAD_V3_V2_gfx11 %5, %3, 7, 1, 0, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s96), align 16, addrspace 4)
+
+name: image_load_not_merged_3
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vreg_64 = BUFFER_LOAD_DWORDX2_OFFSET %2, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = IMAGE_LOAD_V1_V2_gfx11 %5, %3, 8, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_LOAD_V3_V2_gfx11 %5, %3, 7, 1, 0, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+# GFX11-LABEL: name: image_load_not_merged_4
+# GFX11: %{{[0-9]+}}:vgpr_32 = IMAGE_LOAD_V1_V2_gfx11 %5, %3, 8, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s32), addrspace 4)
+# GFX11: %{{[0-9]+}}:vreg_96 = IMAGE_LOAD_V3_V2_gfx11 %5, %3, 7, 1, -1, 1, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s96), align 16, addrspace 4)
+
+name: image_load_not_merged_4
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vreg_64 = BUFFER_LOAD_DWORDX2_OFFSET %2, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = IMAGE_LOAD_V1_V2_gfx11 %5, %3, 8, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_LOAD_V3_V2_gfx11 %5, %3, 7, 1, -1, 1, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+# GFX11-LABEL: name: image_load_not_merged_5
+# GFX11: %{{[0-9]+}}:vgpr_32 = IMAGE_LOAD_V1_V2_gfx11 %5, %3, 8, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s32), addrspace 4)
+# GFX11: %{{[0-9]+}}:vreg_96 = IMAGE_LOAD_V3_V2_gfx11 %5, %3, 7, 1, -1, 0, 1, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s96), align 16, addrspace 4)
+
+name: image_load_not_merged_5
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vreg_64 = BUFFER_LOAD_DWORDX2_OFFSET %2, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = IMAGE_LOAD_V1_V2_gfx11 %5, %3, 8, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_LOAD_V3_V2_gfx11 %5, %3, 7, 1, -1, 0, 1, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+# GFX11-LABEL: name: image_load_not_merged_6
+# GFX11: %{{[0-9]+}}:vgpr_32 = IMAGE_LOAD_V1_V2_gfx11 %5, %3, 8, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s32), addrspace 4)
+# GFX11: %{{[0-9]+}}:vreg_96 = IMAGE_LOAD_V3_V2_gfx11 %5, %3, 7, 1, -1, 1, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s96), align 16, addrspace 4)
+
+name: image_load_not_merged_6
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vreg_64 = BUFFER_LOAD_DWORDX2_OFFSET %2, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = IMAGE_LOAD_V1_V2_gfx11 %5, %3, 8, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_LOAD_V3_V2_gfx11 %5, %3, 7, 1, -1, 1, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+# GFX11-LABEL: name: image_load_not_merged_7
+# GFX11: %{{[0-9]+}}:vgpr_32 = IMAGE_LOAD_V1_V2_gfx11 %5, %3, 8, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s32), addrspace 4)
+# GFX11: %{{[0-9]+}}:vreg_96 = IMAGE_LOAD_V3_V2_gfx11 %5, %3, 7, 1, -1, 0, 1, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s96), align 16, addrspace 4)
+
+name: image_load_not_merged_7
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vreg_64 = BUFFER_LOAD_DWORDX2_OFFSET %2, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = IMAGE_LOAD_V1_V2_gfx11 %5, %3, 8, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_LOAD_V3_V2_gfx11 %5, %3, 7, 1, -1, 0, 1, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+# GFX11-LABEL: name: image_load_not_merged_8
+# GFX11: %{{[0-9]+}}:vgpr_32 = IMAGE_LOAD_V1_V1_gfx11 %6, %3, 8, 1, -1, 0, 0, 1, 0, 0, 0, implicit $exec :: (dereferenceable load (s64), align 16, addrspace 4)
+# GFX11: %{{[0-9]+}}:vreg_96 = IMAGE_LOAD_V3_V2_gfx11 %5, %3, 7, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s96), align 16, addrspace 4)
+
+name: image_load_not_merged_8
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vreg_64 = BUFFER_LOAD_DWORDX2_OFFSET %2, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = COPY %5.sub0
+    %7:vgpr_32 = IMAGE_LOAD_V1_V1_gfx11 %6, %3, 8, 1, -1, 0, 0, 1, 0, 0, 0, implicit $exec :: (dereferenceable load 8, align 16, addrspace 4)
+    %8:vreg_96 = IMAGE_LOAD_V3_V2_gfx11 %5, %3, 7, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+# GFX11-LABEL: name: image_load_not_merged_9
+# GFX11: %{{[0-9]+}}:vreg_64 = IMAGE_LOAD_V2_V2_gfx11 %5, %3, 8, 1, -1, 0, 0, 0, 1, 0, 0, implicit $exec :: (dereferenceable load (s32), addrspace 4)
+# GFX11: %{{[0-9]+}}:vreg_96 = IMAGE_LOAD_V3_V2_gfx11 %5, %3, 7, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s96), align 16, addrspace 4)
+
+name: image_load_not_merged_9
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vreg_64 = BUFFER_LOAD_DWORDX2_OFFSET %2, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vreg_64 = IMAGE_LOAD_V2_V2_gfx11 %5, %3, 8, 1, -1, 0, 0, 0, 1, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_LOAD_V3_V2_gfx11 %5, %3, 7, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+# GFX11-LABEL: name: image_load_not_merged_10
+# GFX11: %{{[0-9]+}}:vreg_64 = IMAGE_LOAD_V2_V2_gfx11 %5, %3, 8, 1, -1, 0, 0, 0, 0, 1, 0, implicit $exec :: (dereferenceable load (s32), addrspace 4)
+# GFX11: %{{[0-9]+}}:vreg_96 = IMAGE_LOAD_V3_V2_gfx11 %5, %3, 7, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s96), align 16, addrspace 4)
+
+name: image_load_not_merged_10
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vreg_64 = BUFFER_LOAD_DWORDX2_OFFSET %2, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vreg_64 = IMAGE_LOAD_V2_V2_gfx11 %5, %3, 8, 1, -1, 0, 0, 0, 0, 1, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_LOAD_V3_V2_gfx11 %5, %3, 7, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+# GFX11-LABEL: name: image_load_not_merged_11
+# GFX11: %{{[0-9]+}}:vgpr_32 = IMAGE_LOAD_V1_V2_gfx11 %5, %3, 8, 1, -1, 0, 0, 0, 0, 0, 1, implicit $exec :: (dereferenceable load (s32), addrspace 4)
+# GFX11: %{{[0-9]+}}:vreg_96 = IMAGE_LOAD_V3_V2_gfx11 %5, %3, 7, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s96), align 16, addrspace 4)
+
+name: image_load_not_merged_11
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vreg_64 = BUFFER_LOAD_DWORDX2_OFFSET %2, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = IMAGE_LOAD_V1_V2_gfx11 %5, %3, 8, 1, -1, 0, 0, 0, 0, 0, 1, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_LOAD_V3_V2_gfx11 %5, %3, 7, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+# GFX11-LABEL: name: image_load_mip_merged_v1v3
+# GFX11: %{{[0-9]+}}:vreg_128 = IMAGE_LOAD_MIP_V4_V3_gfx11 %5, %3, 15, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec, implicit $exec :: (dereferenceable load (s128), align 4, addrspace 4)
+# GFX11: %{{[0-9]+}}:vgpr_32 = COPY %8.sub0
+# GFX11: %{{[0-9]+}}:vreg_96 = COPY killed %8.sub1_sub2_sub3
+
+name: image_load_mip_merged_v1v3
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vreg_96 = BUFFER_LOAD_DWORDX3_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = IMAGE_LOAD_MIP_V1_V3_gfx11 %5:vreg_96, %3:sgpr_256, 1, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_LOAD_MIP_V3_V3_gfx11 %5:vreg_96, %3:sgpr_256, 14, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+
+
+# GFX11-LABEL: name: image_load_mip_pck_merged_v1v3
+# GFX11: %{{[0-9]+}}:vreg_128 = IMAGE_LOAD_MIP_PCK_V4_V3_gfx11 %5, %3, 15, 1, -1, 0, 0, 0, 0, 0, implicit $exec, implicit $exec :: (dereferenceable load (s128), align 4, addrspace 4)
+# GFX11: %{{[0-9]+}}:vgpr_32 = COPY %8.sub0
+# GFX11: %{{[0-9]+}}:vreg_96 = COPY killed %8.sub1_sub2_sub3
+
+name: image_load_mip_pck_merged_v1v3
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vreg_96 = BUFFER_LOAD_DWORDX3_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = IMAGE_LOAD_MIP_PCK_V1_V3_gfx11 %5:vreg_96, %3:sgpr_256, 1, 1, -1, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_LOAD_MIP_PCK_V3_V3_gfx11 %5:vreg_96, %3:sgpr_256, 14, 1, -1, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+
+
+# GFX11-LABEL: name: image_load_mip_pck_sgn_merged_v1v3
+# GFX11: %{{[0-9]+}}:vreg_128 = IMAGE_LOAD_MIP_PCK_SGN_V4_V3_gfx11 %5, %3, 15, 1, -1, 0, 0, 0, 0, 0, implicit $exec, implicit $exec :: (dereferenceable load (s128), align 4, addrspace 4)
+# GFX11: %{{[0-9]+}}:vgpr_32 = COPY %8.sub0
+# GFX11: %{{[0-9]+}}:vreg_96 = COPY killed %8.sub1_sub2_sub3
+
+name: image_load_mip_pck_sgn_merged_v1v3
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vreg_96 = BUFFER_LOAD_DWORDX3_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = IMAGE_LOAD_MIP_PCK_SGN_V1_V3_gfx11 %5:vreg_96, %3:sgpr_256, 1, 1, -1, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_LOAD_MIP_PCK_SGN_V3_V3_gfx11 %5:vreg_96, %3:sgpr_256, 14, 1, -1, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+# GFX11-LABEL: name: image_load_pck_merged_v1v3
+# GFX11: %{{[0-9]+}}:vreg_128 = IMAGE_LOAD_PCK_V4_V2_gfx11 %5, %3, 15, 1, -1, 0, 0, 0, 0, 0, implicit $exec, implicit $exec :: (dereferenceable load (s128), align 4, addrspace 4)
+# GFX11: %{{[0-9]+}}:vgpr_32 = COPY %8.sub0
+# GFX11: %{{[0-9]+}}:vreg_96 = COPY killed %8.sub1_sub2_sub3
+
+name: image_load_pck_merged_v1v3
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vreg_64 = BUFFER_LOAD_DWORDX2_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = IMAGE_LOAD_PCK_V1_V2_gfx11 %5:vreg_64, %3:sgpr_256, 1, 1, -1, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_LOAD_PCK_V3_V2_gfx11 %5:vreg_64, %3:sgpr_256, 14, 1, -1, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+# GFX11-LABEL: name: image_load_pck_sgn_merged_v1v3
+# GFX11: %{{[0-9]+}}:vreg_128 = IMAGE_LOAD_PCK_SGN_V4_V2_gfx11 %5, %3, 15, 1, -1, 0, 0, 0, 0, 0, implicit $exec, implicit $exec :: (dereferenceable load (s128), align 4, addrspace 4)
+# GFX11: %{{[0-9]+}}:vgpr_32 = COPY %8.sub0
+# GFX11: %{{[0-9]+}}:vreg_96 = COPY killed %8.sub1_sub2_sub3
+
+name: image_load_pck_sgn_merged_v1v3
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vreg_64 = BUFFER_LOAD_DWORDX2_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = IMAGE_LOAD_PCK_SGN_V1_V2_gfx11 %5:vreg_64, %3:sgpr_256, 1, 1, -1, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_LOAD_PCK_SGN_V3_V2_gfx11 %5:vreg_64, %3:sgpr_256, 14, 1, -1, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---

llvm/test/CodeGen/AMDGPU/merge-image-sample-gfx11.mir

+# RUN: llc -march=amdgcn -mcpu=gfx1100 -verify-machineinstrs -run-pass si-load-store-opt -o - %s | FileCheck -check-prefix=GFX11 %s
+
+# GFX11-LABEL: name: image_sample_l_merged_v1v3
+# GFX11: %{{[0-9]+}}:vreg_128 = IMAGE_SAMPLE_L_V4_V3_nsa_gfx11 %5, %5, %5, %3, %2, 15, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec, implicit $exec :: (dereferenceable load (s128), align 4, addrspace 4)
+# GFX11: %{{[0-9]+}}:vgpr_32 = COPY %8.sub0
+# GFX11: %{{[0-9]+}}:vreg_96 = COPY killed %8.sub1_sub2_sub3
+
+name: image_sample_l_merged_v1v3
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = IMAGE_SAMPLE_L_V1_V3_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 1, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_SAMPLE_L_V3_V3_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 14, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+# GFX11-LABEL: name: image_sample_l_merged_v1v3_reversed
+# GFX11: %{{[0-9]+}}:vreg_128 = IMAGE_SAMPLE_L_V4_V3_nsa_gfx11 %5, %5, %5, %3, %2, 15, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec, implicit $exec :: (dereferenceable load (s128), addrspace 4)
+# GFX11: %{{[0-9]+}}:vgpr_32 = COPY %8.sub3
+# GFX11: %{{[0-9]+}}:vreg_96 = COPY killed %8.sub0_sub1_sub2
+
+name: image_sample_l_merged_v1v3_reversed
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = IMAGE_SAMPLE_L_V1_V3_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 8, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_SAMPLE_L_V3_V3_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 7, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+# GFX11-LABEL: name: image_sample_l_merged_v2v2
+# GFX11: %{{[0-9]+}}:vreg_128 = IMAGE_SAMPLE_L_V4_V3_nsa_gfx11 %5, %5, %5, %3, %2, 15, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec, implicit $exec :: (dereferenceable load (s128), addrspace 4)
+# GFX11: %{{[0-9]+}}:vreg_64 = COPY %8.sub0_sub1
+# GFX11: %{{[0-9]+}}:vreg_64 = COPY killed %8.sub2_sub3
+
+name: image_sample_l_merged_v2v2
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vreg_64 = IMAGE_SAMPLE_L_V2_V3_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 3, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 8, align 16, addrspace 4)
+    %7:vreg_64 = IMAGE_SAMPLE_L_V2_V3_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 12, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 8, align 16, addrspace 4)
+...
+---
+
+# GFX11-LABEL: name: image_sample_l_merged_v2v2_reversed
+# GFX11: %{{[0-9]+}}:vreg_128 = IMAGE_SAMPLE_L_V4_V3_nsa_gfx11 %5, %5, %5, %3, %2, 15, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec, implicit $exec :: (dereferenceable load (s128), addrspace 4)
+# GFX11: %{{[0-9]+}}:vreg_64 = COPY %8.sub2_sub3
+# GFX11: %{{[0-9]+}}:vreg_64 = COPY killed %8.sub0_sub1
+
+name: image_sample_l_merged_v2v2_reversed
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vreg_64 = IMAGE_SAMPLE_L_V2_V3_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 12, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 8, align 16, addrspace 4)
+    %7:vreg_64 = IMAGE_SAMPLE_L_V2_V3_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 3, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 8, align 16, addrspace 4)
+...
+---
+
+# GFX11-LABEL: name: image_sample_l_merged_v3v1
+# GFX11: %{{[0-9]+}}:vreg_128 = IMAGE_SAMPLE_L_V4_V3_nsa_gfx11 %5, %5, %5, %3, %2, 15, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec, implicit $exec :: (dereferenceable load (s128), addrspace 4)
+# GFX11: %{{[0-9]+}}:vreg_96 = COPY %8.sub0_sub1_sub2
+# GFX11: %{{[0-9]+}}:vgpr_32 = COPY killed %8.sub3
+
+name: image_sample_l_merged_v3v1
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vreg_96 = IMAGE_SAMPLE_L_V3_V3_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 7, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+    %7:vgpr_32 = IMAGE_SAMPLE_L_V1_V3_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 8, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+...
+---
+
+# GFX11-LABEL: name: image_sample_l_merged_v3v1_reversed
+# GFX11: %{{[0-9]+}}:vreg_128 = IMAGE_SAMPLE_L_V4_V3_nsa_gfx11 %5, %5, %5, %3, %2, 15, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec, implicit $exec :: (dereferenceable load (s128), align 4, addrspace 4)
+# GFX11: %{{[0-9]+}}:vreg_96 = COPY %8.sub1_sub2_sub3
+# GFX11: %{{[0-9]+}}:vgpr_32 = COPY killed %8.sub0
+
+name: image_sample_l_merged_v3v1_reversed
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vreg_96 = IMAGE_SAMPLE_L_V3_V3_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 14, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+    %7:vgpr_32 = IMAGE_SAMPLE_L_V1_V3_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 1, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+...
+---
+
+# GFX11-LABEL: name: image_sample_l_divided_merged
+# GFX11: %{{[0-9]+}}:vreg_128 = IMAGE_SAMPLE_L_V4_V3_nsa_gfx11 %5, %5, %5, %3, %2, 15, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec, implicit $exec :: (dereferenceable load (s128), addrspace 4)
+
+name: image_sample_l_divided_merged
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = IMAGE_SAMPLE_L_V1_V3_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 8, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %8:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %9:vreg_96 = IMAGE_SAMPLE_L_V3_V3_nsa_gfx11 %7:vgpr_32, %7:vgpr_32, %7:vgpr_32, %3:sgpr_256, %2:sgpr_128, 7, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+    %10:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %11:vreg_96 = IMAGE_SAMPLE_L_V3_V3_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 7, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+# GFX11-LABEL: name: image_sample_l_divided_not_merged
+# GFX11: %{{[0-9]+}}:vgpr_32 = IMAGE_SAMPLE_L_V1_V3_nsa_gfx11 %5, %5, %5, %3, %2, 8, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s32), addrspace 4)
+# GFX11: %{{[0-9]+}}:vreg_96 = IMAGE_SAMPLE_L_V3_V3_nsa_gfx11 %5, %5, %5, %3, %2, 7, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s96), align 16, addrspace 4)
+
+name: image_sample_l_divided_not_merged
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vreg_128 = COPY %2
+    %5:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = IMAGE_SAMPLE_L_V1_V3_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 8, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    IMAGE_STORE_V4_V2_nsa_gfx11 %4:vreg_128, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, 15, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (store 16)
+    %7:vreg_96 = IMAGE_SAMPLE_L_V3_V3_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 7, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+# GFX11-LABEL: name: image_sample_l_dmask_overlapped_not_merged
+# GFX11: %{{[0-9]+}}:vgpr_32 = IMAGE_SAMPLE_L_V1_V3_nsa_gfx11 %5, %5, %5, %3, %2, 4, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s32), addrspace 4)
+# GFX11: %{{[0-9]+}}:vreg_96 = IMAGE_SAMPLE_L_V3_V3_nsa_gfx11 %5, %5, %5, %3, %2, 7, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s96), align 16, addrspace 4)
+
+name: image_sample_l_dmask_overlapped_not_merged
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = IMAGE_SAMPLE_L_V1_V3_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 4, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_SAMPLE_L_V3_V3_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 7, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+# GFX11-LABEL: name: image_sample_l_dmask_not_disjoint_not_merged
+# GFX11: %{{[0-9]+}}:vgpr_32 = IMAGE_SAMPLE_L_V1_V3_nsa_gfx11 %5, %5, %5, %3, %2, 4, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s32), addrspace 4)
+# GFX11: %{{[0-9]+}}:vreg_96 = IMAGE_SAMPLE_L_V3_V3_nsa_gfx11 %5, %5, %5, %3, %2, 11, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s96), align 16, addrspace 4)
+
+name: image_sample_l_dmask_not_disjoint_not_merged
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = IMAGE_SAMPLE_L_V1_V3_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 4, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_SAMPLE_L_V3_V3_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 11, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+# GFX11-LABEL: name: image_sample_l_not_merged_0
+# GFX11: %{{[0-9]+}}:vgpr_32 = IMAGE_SAMPLE_L_V1_V3_nsa_gfx11 %5, %5, %5, %3, %2, 8, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s32), addrspace 4)
+# GFX11: %{{[0-9]+}}:vreg_96 = IMAGE_SAMPLE_L_V3_V3_nsa_gfx11 %6, %6, %6, %3, %2, 7, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s96), align 16, addrspace 4)
+
+name: image_sample_l_not_merged_0
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET %2, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET %2, 1, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %7:vgpr_32 = IMAGE_SAMPLE_L_V1_V3_nsa_gfx11 %5, %5, %5, %3, %2, 8, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %8:vreg_96 = IMAGE_SAMPLE_L_V3_V3_nsa_gfx11 %6, %6, %6, %3, %2, 7, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+# GFX11-LABEL: name: image_sample_l_not_merged_1
+# GFX11: %{{[0-9]+}}:vgpr_32 = IMAGE_SAMPLE_L_V1_V3_nsa_gfx11 %6, %6, %6, %3, %2, 8, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s32), addrspace 4)
+# GFX11: %{{[0-9]+}}:vreg_96 = IMAGE_SAMPLE_L_V3_V3_nsa_gfx11 %6, %6, %6, %4, %2, 7, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s96), align 16, addrspace 4)
+
+name: image_sample_l_not_merged_1
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %5:vgpr_32 = COPY %2.sub3
+    %6:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET %2, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %7:vgpr_32 = IMAGE_SAMPLE_L_V1_V3_nsa_gfx11 %6, %6, %6, %3, %2, 8, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %8:vreg_96 = IMAGE_SAMPLE_L_V3_V3_nsa_gfx11 %6, %6, %6, %4, %2, 7, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+# GFX11-LABEL: name: image_sample_l_not_merged_2
+# GFX11: %{{[0-9]+}}:vgpr_32 = IMAGE_SAMPLE_L_V1_V3_nsa_gfx11 %6, %6, %6, %4, %2, 8, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s32), addrspace 4)
+# GFX11: %{{[0-9]+}}:vreg_96 = IMAGE_SAMPLE_L_V3_V3_nsa_gfx11 %6, %6, %6, %4, %3, 7, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s96), align 16, addrspace 4)
+
+name: image_sample_l_not_merged_2
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_128 = COPY $sgpr92_sgpr93_sgpr94_sgpr95
+    %4:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %5:vgpr_32 = COPY %2.sub3
+    %6:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET %2, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %7:vgpr_32 = IMAGE_SAMPLE_L_V1_V3_nsa_gfx11 %6, %6, %6, %4, %2, 8, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %8:vreg_96 = IMAGE_SAMPLE_L_V3_V3_nsa_gfx11 %6, %6, %6, %4, %3, 7, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+# GFX11-LABEL: name: image_sample_l_not_merged_3
+# GFX11: %{{[0-9]+}}:vgpr_32 = IMAGE_SAMPLE_L_V1_V3_nsa_gfx11 %5, %5, %5, %3, %2, 8, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s32), addrspace 4)
+# GFX11: %{{[0-9]+}}:vreg_96 = IMAGE_SAMPLE_L_V3_V3_nsa_gfx11 %5, %5, %5, %3, %2, 7, 1, -1, 1, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s96), align 16, addrspace 4)
+
+name: image_sample_l_not_merged_3
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET %2, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = IMAGE_SAMPLE_L_V1_V3_nsa_gfx11 %5, %5, %5, %3, %2, 8, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_SAMPLE_L_V3_V3_nsa_gfx11 %5, %5, %5, %3, %2, 7, 1, -1, 1, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+# GFX11-LABEL: name: image_sample_l_not_merged_4
+# GFX11: %{{[0-9]+}}:vgpr_32 = IMAGE_SAMPLE_L_V1_V3_nsa_gfx11 %5, %5, %5, %3, %2, 8, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s32), addrspace 4)
+# GFX11: %{{[0-9]+}}:vreg_96 = IMAGE_SAMPLE_L_V3_V3_nsa_gfx11 %5, %5, %5, %3, %2, 7, 1, -1, 0, 1, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s96), align 16, addrspace 4)
+
+name: image_sample_l_not_merged_4
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET %2, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = IMAGE_SAMPLE_L_V1_V3_nsa_gfx11 %5, %5, %5, %3, %2, 8, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_SAMPLE_L_V3_V3_nsa_gfx11 %5, %5, %5, %3, %2, 7, 1, -1, 0, 1, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+# GFX11-LABEL: name: image_sample_l_not_merged_5
+# GFX11: %{{[0-9]+}}:vgpr_32 = IMAGE_SAMPLE_L_V1_V3_nsa_gfx11 %5, %5, %5, %3, %2, 8, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s32), addrspace 4)
+# GFX11: %{{[0-9]+}}:vreg_96 = IMAGE_SAMPLE_L_V3_V3_nsa_gfx11 %5, %5, %5, %3, %2, 7, 1, -1, 1, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s96), align 16, addrspace 4)
+
+name: image_sample_l_not_merged_5
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET %2, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = IMAGE_SAMPLE_L_V1_V3_nsa_gfx11 %5, %5, %5, %3, %2, 8, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_SAMPLE_L_V3_V3_nsa_gfx11 %5, %5, %5, %3, %2, 7, 1, -1, 1, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+# GFX11-LABEL: name: image_sample_l_not_merged_6
+# GFX11: %{{[0-9]+}}:vgpr_32 = IMAGE_SAMPLE_L_V1_V3_nsa_gfx11 %5, %5, %5, %3, %2, 8, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s32), addrspace 4)
+# GFX11: %{{[0-9]+}}:vreg_96 = IMAGE_SAMPLE_L_V3_V3_nsa_gfx11 %5, %5, %5, %3, %2, 7, 1, -1, 0, 1, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s96), align 16, addrspace 4)
+
+name: image_sample_l_not_merged_6
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET %2, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = IMAGE_SAMPLE_L_V1_V3_nsa_gfx11 %5, %5, %5, %3, %2, 8, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_SAMPLE_L_V3_V3_nsa_gfx11 %5, %5, %5, %3, %2, 7, 1, -1, 0, 1, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+# GFX11-LABEL: name: image_sample_l_not_merged_7
+# GFX11: %{{[0-9]+}}:vgpr_32 = IMAGE_SAMPLE_L_V1_V2_nsa_gfx11 %5, %5, %3, %2, 8, 1, -1, 0, 0, 1, 0, 0, 0, implicit $exec :: (dereferenceable load (s32), addrspace 4)
+# GFX11: %{{[0-9]+}}:vreg_96 = IMAGE_SAMPLE_L_V3_V3_nsa_gfx11 %5, %5, %5, %3, %2, 7, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s96), align 16, addrspace 4)
+
+name: image_sample_l_not_merged_7
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET %2, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = IMAGE_SAMPLE_L_V1_V2_nsa_gfx11 %5, %5, %3, %2, 8, 1, -1, 0, 0, 1, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_SAMPLE_L_V3_V3_nsa_gfx11 %5, %5, %5, %3, %2, 7, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+# GFX11-LABEL: name: image_sample_l_not_merged_8
+# GFX11: %{{[0-9]+}}:vreg_64 = IMAGE_SAMPLE_L_V2_V3_nsa_gfx11 %5, %5, %5, %3, %2, 8, 1, -1, 0, 0, 0, 1, 0, 0, implicit $exec :: (dereferenceable load (s64), addrspace 4)
+# GFX11: %{{[0-9]+}}:vreg_96 = IMAGE_SAMPLE_L_V3_V3_nsa_gfx11 %5, %5, %5, %3, %2, 7, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s96), align 16, addrspace 4)
+
+name: image_sample_l_not_merged_8
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET %2, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vreg_64 = IMAGE_SAMPLE_L_V2_V3_nsa_gfx11 %5, %5, %5, %3, %2, 8, 1, -1, 0, 0, 0, 1, 0, 0, implicit $exec :: (dereferenceable load 8, addrspace 4)
+    %7:vreg_96 = IMAGE_SAMPLE_L_V3_V3_nsa_gfx11 %5, %5, %5, %3, %2, 7, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+# GFX11-LABEL: name: image_sample_l_not_merged_9
+# GFX11: %{{[0-9]+}}:vreg_64 = IMAGE_SAMPLE_L_V2_V3_nsa_gfx11 %5, %5, %5, %3, %2, 8, 1, -1, 0, 0, 0, 0, 1, 0, implicit $exec :: (dereferenceable load (s64), addrspace 4)
+# GFX11: %{{[0-9]+}}:vreg_96 = IMAGE_SAMPLE_L_V3_V3_nsa_gfx11 %5, %5, %5, %3, %2, 7, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s96), align 16, addrspace 4)
+
+name: image_sample_l_not_merged_9
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET %2, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vreg_64 = IMAGE_SAMPLE_L_V2_V3_nsa_gfx11 %5, %5, %5, %3, %2, 8, 1, -1, 0, 0, 0, 0, 1, 0, implicit $exec :: (dereferenceable load 8, addrspace 4)
+    %7:vreg_96 = IMAGE_SAMPLE_L_V3_V3_nsa_gfx11 %5, %5, %5, %3, %2, 7, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+# GFX11-LABEL: name: image_sample_l_not_merged_10
+# GFX11: %{{[0-9]+}}:vgpr_32 = IMAGE_SAMPLE_L_V1_V3_nsa_gfx11 %5, %5, %5, %3, %2, 8, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s32), addrspace 4)
+# GFX11: %{{[0-9]+}}:vreg_96 = IMAGE_SAMPLE_L_V3_V3_nsa_gfx11 %5, %5, %5, %3, %2, 7, 1, -1, 0, 0, 0, 0, 0, 1, implicit $exec :: (dereferenceable load (s96), align 16, addrspace 4)
+
+name: image_sample_l_not_merged_10
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET %2, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = IMAGE_SAMPLE_L_V1_V3_nsa_gfx11 %5, %5, %5, %3, %2, 8, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_SAMPLE_L_V3_V3_nsa_gfx11 %5, %5, %5, %3, %2, 7, 1, -1, 0, 0, 0, 0, 0, 1, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+
+
+
+# GFX11-LABEL: name: image_sample_merged_v1v3
+# GFX11: %{{[0-9]+}}:vreg_128 = IMAGE_SAMPLE_V4_V2_nsa_gfx11 %5, %5, %3, %2, 15, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec, implicit $exec :: (dereferenceable load (s128), align 4, addrspace 4)
+# GFX11: %{{[0-9]+}}:vgpr_32 = COPY %8.sub0
+# GFX11: %{{[0-9]+}}:vreg_96 = COPY killed %8.sub1_sub2_sub3
+
+name: image_sample_merged_v1v3
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = IMAGE_SAMPLE_V1_V2_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 1, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_SAMPLE_V3_V2_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 14, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+
+# GFX11-LABEL: name: image_sample_b_merged_v1v3
+# GFX11: %{{[0-9]+}}:vreg_128 = IMAGE_SAMPLE_B_V4_V3_nsa_gfx11 %5, %5, %5, %3, %2, 15, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec, implicit $exec :: (dereferenceable load (s128), align 4, addrspace 4)
+# GFX11: %{{[0-9]+}}:vgpr_32 = COPY %8.sub0
+# GFX11: %{{[0-9]+}}:vreg_96 = COPY killed %8.sub1_sub2_sub3
+
+name: image_sample_b_merged_v1v3
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = IMAGE_SAMPLE_B_V1_V3_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 1, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_SAMPLE_B_V3_V3_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 14, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+
+# GFX11-LABEL: name: image_sample_b_cl_merged_v1v3
+# GFX11: %{{[0-9]+}}:vreg_128 = IMAGE_SAMPLE_B_CL_V4_V4_nsa_gfx11 %5, %5, %5, %5, %3, %2, 15, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec, implicit $exec :: (dereferenceable load (s128), align 4, addrspace 4)
+# GFX11: %{{[0-9]+}}:vgpr_32 = COPY %8.sub0
+# GFX11: %{{[0-9]+}}:vreg_96 = COPY killed %8.sub1_sub2_sub3
+
+name: image_sample_b_cl_merged_v1v3
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = IMAGE_SAMPLE_B_CL_V1_V4_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 1, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_SAMPLE_B_CL_V3_V4_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 14, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+
+# GFX11-LABEL: name: image_sample_b_cl_o_merged_v1v3
+# GFX11: %{{[0-9]+}}:vreg_128 = IMAGE_SAMPLE_B_CL_O_V4_V5_nsa_gfx11 %5, %5, %5, %5, %5, %3, %2, 15, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec, implicit $exec :: (dereferenceable load (s128), align 4, addrspace 4)
+# GFX11: %{{[0-9]+}}:vgpr_32 = COPY %8.sub0
+# GFX11: %{{[0-9]+}}:vreg_96 = COPY killed %8.sub1_sub2_sub3
+
+name: image_sample_b_cl_o_merged_v1v3
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = IMAGE_SAMPLE_B_CL_O_V1_V5_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 1, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_SAMPLE_B_CL_O_V3_V5_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 14, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+
+# GFX11-LABEL: name: image_sample_b_o_merged_v1v3
+# GFX11: %{{[0-9]+}}:vreg_128 = IMAGE_SAMPLE_B_O_V4_V4_nsa_gfx11 %5, %5, %5, %5, %3, %2, 15, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec, implicit $exec :: (dereferenceable load (s128), align 4, addrspace 4)
+# GFX11: %{{[0-9]+}}:vgpr_32 = COPY %8.sub0
+# GFX11: %{{[0-9]+}}:vreg_96 = COPY killed %8.sub1_sub2_sub3
+
+name: image_sample_b_o_merged_v1v3
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = IMAGE_SAMPLE_B_O_V1_V4_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 1, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_SAMPLE_B_O_V3_V4_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 14, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+
+# GFX11-LABEL: name: image_sample_c_merged_v1v3
+# GFX11: %{{[0-9]+}}:vreg_128 = IMAGE_SAMPLE_C_V4_V3_nsa_gfx11 %5, %5, %5, %3, %2, 15, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec, implicit $exec :: (dereferenceable load (s128), align 4, addrspace 4)
+# GFX11: %{{[0-9]+}}:vgpr_32 = COPY %8.sub0
+# GFX11: %{{[0-9]+}}:vreg_96 = COPY killed %8.sub1_sub2_sub3
+
+name: image_sample_c_merged_v1v3
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = IMAGE_SAMPLE_C_V1_V3_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 1, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_SAMPLE_C_V3_V3_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 14, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+
+# GFX11-LABEL: name: image_sample_cl_merged_v1v3
+# GFX11: %{{[0-9]+}}:vreg_128 = IMAGE_SAMPLE_CL_V4_V3_nsa_gfx11 %5, %5, %5, %3, %2, 15, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec, implicit $exec :: (dereferenceable load (s128), align 4, addrspace 4)
+# GFX11: %{{[0-9]+}}:vgpr_32 = COPY %8.sub0
+# GFX11: %{{[0-9]+}}:vreg_96 = COPY killed %8.sub1_sub2_sub3
+
+name: image_sample_cl_merged_v1v3
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = IMAGE_SAMPLE_CL_V1_V3_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 1, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_SAMPLE_CL_V3_V3_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 14, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+
+# GFX11-LABEL: name: image_sample_cl_o_merged_v1v3
+# GFX11: %{{[0-9]+}}:vreg_128 = IMAGE_SAMPLE_CL_O_V4_V4_nsa_gfx11 %5, %5, %5, %5, %3, %2, 15, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec, implicit $exec :: (dereferenceable load (s128), align 4, addrspace 4)
+# GFX11: %{{[0-9]+}}:vgpr_32 = COPY %8.sub0
+# GFX11: %{{[0-9]+}}:vreg_96 = COPY killed %8.sub1_sub2_sub3
+
+name: image_sample_cl_o_merged_v1v3
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = IMAGE_SAMPLE_CL_O_V1_V4_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 1, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_SAMPLE_CL_O_V3_V4_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 14, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+
+# GFX11-LABEL: name: image_sample_c_b_merged_v1v3
+# GFX11: %{{[0-9]+}}:vreg_128 = IMAGE_SAMPLE_C_B_V4_V4_nsa_gfx11 %5, %5, %5, %5, %3, %2, 15, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec, implicit $exec :: (dereferenceable load (s128), align 4, addrspace 4)
+# GFX11: %{{[0-9]+}}:vgpr_32 = COPY %8.sub0
+# GFX11: %{{[0-9]+}}:vreg_96 = COPY killed %8.sub1_sub2_sub3
+
+name: image_sample_c_b_merged_v1v3
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = IMAGE_SAMPLE_C_B_V1_V4_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 1, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_SAMPLE_C_B_V3_V4_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 14, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+
+# GFX11-LABEL: name: image_sample_c_b_cl_merged_v1v3
+# GFX11: %{{[0-9]+}}:vreg_128 = IMAGE_SAMPLE_C_B_CL_V4_V5_nsa_gfx11 %5, %5, %5, %5, %5, %3, %2, 15, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec, implicit $exec :: (dereferenceable load (s128), align 4, addrspace 4)
+# GFX11: %{{[0-9]+}}:vgpr_32 = COPY %8.sub0
+# GFX11: %{{[0-9]+}}:vreg_96 = COPY killed %8.sub1_sub2_sub3
+
+name: image_sample_c_b_cl_merged_v1v3
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = IMAGE_SAMPLE_C_B_CL_V1_V5_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 1, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_SAMPLE_C_B_CL_V3_V5_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 14, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+
+# GFX11-LABEL: name: image_sample_c_b_cl_o_merged_v1v3
+# GFX11: %{{[0-9]+}}:vreg_128 = IMAGE_SAMPLE_C_B_CL_O_V4_V6_gfx11 %5, %3, %2, 15, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec, implicit $exec :: (dereferenceable load (s128), align 4, addrspace 4)
+# GFX11: %{{[0-9]+}}:vgpr_32 = COPY %8.sub0
+# GFX11: %{{[0-9]+}}:vreg_96 = COPY killed %8.sub1_sub2_sub3
+
+name: image_sample_c_b_cl_o_merged_v1v3
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vreg_192 = IMPLICIT_DEF
+    %6:vgpr_32 = IMAGE_SAMPLE_C_B_CL_O_V1_V6_gfx11 %5:vreg_192, %3:sgpr_256, %2:sgpr_128, 1, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_SAMPLE_C_B_CL_O_V3_V6_gfx11 %5:vreg_192, %3:sgpr_256, %2:sgpr_128, 14, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+
+# GFX11-LABEL: name: image_sample_c_b_o_merged_v1v3
+# GFX11: %{{[0-9]+}}:vreg_128 = IMAGE_SAMPLE_C_B_O_V4_V5_nsa_gfx11 %5, %5, %5, %5, %5, %3, %2, 15, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec, implicit $exec :: (dereferenceable load (s128), align 4, addrspace 4)
+# GFX11: %{{[0-9]+}}:vgpr_32 = COPY %8.sub0
+# GFX11: %{{[0-9]+}}:vreg_96 = COPY killed %8.sub1_sub2_sub3
+
+name: image_sample_c_b_o_merged_v1v3
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = IMAGE_SAMPLE_C_B_O_V1_V5_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 1, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_SAMPLE_C_B_O_V3_V5_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 14, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+
+# GFX11-LABEL: name: image_sample_c_cl_merged_v1v3
+# GFX11: %{{[0-9]+}}:vreg_128 = IMAGE_SAMPLE_C_CL_V4_V4_nsa_gfx11 %5, %5, %5, %5, %3, %2, 15, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec, implicit $exec :: (dereferenceable load (s128), align 4, addrspace 4)
+# GFX11: %{{[0-9]+}}:vgpr_32 = COPY %8.sub0
+# GFX11: %{{[0-9]+}}:vreg_96 = COPY killed %8.sub1_sub2_sub3
+
+name: image_sample_c_cl_merged_v1v3
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = IMAGE_SAMPLE_C_CL_V1_V4_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 1, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_SAMPLE_C_CL_V3_V4_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 14, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+
+# GFX11-LABEL: name: image_sample_c_cl_o_merged_v1v3
+# GFX11: %{{[0-9]+}}:vreg_128 = IMAGE_SAMPLE_C_CL_O_V4_V5_nsa_gfx11 %5, %5, %5, %5, %5, %3, %2, 15, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec, implicit $exec :: (dereferenceable load (s128), align 4, addrspace 4)
+# GFX11: %{{[0-9]+}}:vgpr_32 = COPY %8.sub0
+# GFX11: %{{[0-9]+}}:vreg_96 = COPY killed %8.sub1_sub2_sub3
+
+name: image_sample_c_cl_o_merged_v1v3
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = IMAGE_SAMPLE_C_CL_O_V1_V5_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 1, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_SAMPLE_C_CL_O_V3_V5_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 14, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+
+# GFX11-LABEL: name: image_sample_c_d_merged_v1v3
+# GFX11: %{{[0-9]+}}:vreg_128 = IMAGE_SAMPLE_C_D_V4_V7_gfx11 %5, %3, %2, 15, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec, implicit $exec :: (dereferenceable load (s128), align 4, addrspace 4)
+# GFX11: %{{[0-9]+}}:vgpr_32 = COPY %8.sub0
+# GFX11: %{{[0-9]+}}:vreg_96 = COPY killed %8.sub1_sub2_sub3
+
+name: image_sample_c_d_merged_v1v3
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vreg_224 = IMPLICIT_DEF
+    %6:vgpr_32 = IMAGE_SAMPLE_C_D_V1_V7_gfx11 %5:vreg_224, %3:sgpr_256, %2:sgpr_128, 1, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_SAMPLE_C_D_V3_V7_gfx11 %5:vreg_224, %3:sgpr_256, %2:sgpr_128, 14, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+
+# GFX11-LABEL: name: image_sample_c_d_cl_merged_v1v3
+# GFX11: %{{[0-9]+}}:vreg_128 = IMAGE_SAMPLE_C_D_CL_V4_V8_gfx11 %5, %3, %2, 15, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec, implicit $exec :: (dereferenceable load (s128), align 4, addrspace 4)
+# GFX11: %{{[0-9]+}}:vgpr_32 = COPY %8.sub0
+# GFX11: %{{[0-9]+}}:vreg_96 = COPY killed %8.sub1_sub2_sub3
+
+name: image_sample_c_d_cl_merged_v1v3
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vreg_256 = IMPLICIT_DEF
+    %6:vgpr_32 = IMAGE_SAMPLE_C_D_CL_V1_V8_gfx11 %5:vreg_256, %3:sgpr_256, %2:sgpr_128, 1, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_SAMPLE_C_D_CL_V3_V8_gfx11 %5:vreg_256, %3:sgpr_256, %2:sgpr_128, 14, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+
+# GFX11-LABEL: name: image_sample_c_d_cl_o_merged_v1v3
+# GFX11: %{{[0-9]+}}:vreg_128 = IMAGE_SAMPLE_C_D_CL_O_V4_V16_gfx11 %5, %3, %2, 15, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec, implicit $exec :: (dereferenceable load (s128), align 4, addrspace 4)
+# GFX11: %{{[0-9]+}}:vgpr_32 = COPY %8.sub0
+# GFX11: %{{[0-9]+}}:vreg_96 = COPY killed %8.sub1_sub2_sub3
+
+name: image_sample_c_d_cl_o_merged_v1v3
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vreg_512 = IMPLICIT_DEF
+    %6:vgpr_32 = IMAGE_SAMPLE_C_D_CL_O_V1_V16_gfx11 %5:vreg_512, %3:sgpr_256, %2:sgpr_128, 1, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_SAMPLE_C_D_CL_O_V3_V16_gfx11 %5:vreg_512, %3:sgpr_256, %2:sgpr_128, 14, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+
+# GFX11-LABEL: name: image_sample_c_d_o_merged_v1v3
+# GFX11: %{{[0-9]+}}:vreg_128 = IMAGE_SAMPLE_C_D_O_V4_V8_gfx11 %5, %3, %2, 15, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec, implicit $exec :: (dereferenceable load (s128), align 4, addrspace 4)
+# GFX11: %{{[0-9]+}}:vgpr_32 = COPY %8.sub0
+# GFX11: %{{[0-9]+}}:vreg_96 = COPY killed %8.sub1_sub2_sub3
+
+name: image_sample_c_d_o_merged_v1v3
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vreg_256 = IMPLICIT_DEF
+    %6:vgpr_32 = IMAGE_SAMPLE_C_D_O_V1_V8_gfx11 %5:vreg_256, %3:sgpr_256, %2:sgpr_128, 1, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_SAMPLE_C_D_O_V3_V8_gfx11 %5:vreg_256, %3:sgpr_256, %2:sgpr_128, 14, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+
+# GFX11-LABEL: name: image_sample_c_l_merged_v1v3
+# GFX11: %{{[0-9]+}}:vreg_128 = IMAGE_SAMPLE_C_L_V4_V4_nsa_gfx11 %5, %5, %5, %5, %3, %2, 15, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec, implicit $exec :: (dereferenceable load (s128), align 4, addrspace 4)
+# GFX11: %{{[0-9]+}}:vgpr_32 = COPY %8.sub0
+# GFX11: %{{[0-9]+}}:vreg_96 = COPY killed %8.sub1_sub2_sub3
+
+name: image_sample_c_l_merged_v1v3
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = IMAGE_SAMPLE_C_L_V1_V4_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 1, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_SAMPLE_C_L_V3_V4_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 14, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+
+# GFX11-LABEL: name: image_sample_c_lz_merged_v1v3
+# GFX11: %{{[0-9]+}}:vreg_128 = IMAGE_SAMPLE_C_LZ_V4_V3_nsa_gfx11 %5, %5, %5, %3, %2, 15, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec, implicit $exec :: (dereferenceable load (s128), align 4, addrspace 4)
+# GFX11: %{{[0-9]+}}:vgpr_32 = COPY %8.sub0
+# GFX11: %{{[0-9]+}}:vreg_96 = COPY killed %8.sub1_sub2_sub3
+
+name: image_sample_c_lz_merged_v1v3
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = IMAGE_SAMPLE_C_LZ_V1_V3_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 1, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_SAMPLE_C_LZ_V3_V3_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 14, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+
+# GFX11-LABEL: name: image_sample_c_lz_o_merged_v1v3
+# GFX11: %{{[0-9]+}}:vreg_128 = IMAGE_SAMPLE_C_LZ_O_V4_V4_nsa_gfx11 %5, %5, %5, %5, %3, %2, 15, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec, implicit $exec :: (dereferenceable load (s128), align 4, addrspace 4)
+# GFX11: %{{[0-9]+}}:vgpr_32 = COPY %8.sub0
+# GFX11: %{{[0-9]+}}:vreg_96 = COPY killed %8.sub1_sub2_sub3
+
+name: image_sample_c_lz_o_merged_v1v3
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = IMAGE_SAMPLE_C_LZ_O_V1_V4_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 1, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_SAMPLE_C_LZ_O_V3_V4_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 14, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+
+# GFX11-LABEL: name: image_sample_c_l_o_merged_v1v3
+# GFX11: %{{[0-9]+}}:vreg_128 = IMAGE_SAMPLE_C_L_O_V4_V5_nsa_gfx11 %5, %5, %5, %5, %5, %3, %2, 15, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec, implicit $exec :: (dereferenceable load (s128), align 4, addrspace 4)
+# GFX11: %{{[0-9]+}}:vgpr_32 = COPY %8.sub0
+# GFX11: %{{[0-9]+}}:vreg_96 = COPY killed %8.sub1_sub2_sub3
+
+name: image_sample_c_l_o_merged_v1v3
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = IMAGE_SAMPLE_C_L_O_V1_V5_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 1, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_SAMPLE_C_L_O_V3_V5_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 14, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+
+# GFX11-LABEL: name: image_sample_c_o_merged_v1v3
+# GFX11: %{{[0-9]+}}:vreg_128 = IMAGE_SAMPLE_C_O_V4_V4_nsa_gfx11 %5, %5, %5, %5, %3, %2, 15, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec, implicit $exec :: (dereferenceable load (s128), align 4, addrspace 4)
+# GFX11: %{{[0-9]+}}:vgpr_32 = COPY %8.sub0
+# GFX11: %{{[0-9]+}}:vreg_96 = COPY killed %8.sub1_sub2_sub3
+
+name: image_sample_c_o_merged_v1v3
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = IMAGE_SAMPLE_C_O_V1_V4_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 1, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_SAMPLE_C_O_V3_V4_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 14, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+
+# GFX11-LABEL: name: image_sample_d_merged_v1v3
+# GFX11: %{{[0-9]+}}:vreg_128 = IMAGE_SAMPLE_D_V4_V6_gfx11 %5, %3, %2, 15, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec, implicit $exec :: (dereferenceable load (s128), align 4, addrspace 4)
+# GFX11: %{{[0-9]+}}:vgpr_32 = COPY %8.sub0
+# GFX11: %{{[0-9]+}}:vreg_96 = COPY killed %8.sub1_sub2_sub3
+
+name: image_sample_d_merged_v1v3
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vreg_192 = IMPLICIT_DEF
+    %6:vgpr_32 = IMAGE_SAMPLE_D_V1_V6_gfx11 %5:vreg_192, %3:sgpr_256, %2:sgpr_128, 1, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_SAMPLE_D_V3_V6_gfx11 %5:vreg_192, %3:sgpr_256, %2:sgpr_128, 14, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+
+# GFX11-LABEL: name: image_sample_d_cl_merged_v1v3
+# GFX11: %{{[0-9]+}}:vreg_128 = IMAGE_SAMPLE_D_CL_V4_V7_gfx11 %5, %3, %2, 15, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec, implicit $exec :: (dereferenceable load (s128), align 4, addrspace 4)
+# GFX11: %{{[0-9]+}}:vgpr_32 = COPY %8.sub0
+# GFX11: %{{[0-9]+}}:vreg_96 = COPY killed %8.sub1_sub2_sub3
+
+name: image_sample_d_cl_merged_v1v3
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vreg_224 = IMPLICIT_DEF
+    %6:vgpr_32 = IMAGE_SAMPLE_D_CL_V1_V7_gfx11 %5:vreg_224, %3:sgpr_256, %2:sgpr_128, 1, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_SAMPLE_D_CL_V3_V7_gfx11 %5:vreg_224, %3:sgpr_256, %2:sgpr_128, 14, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+
+# GFX11-LABEL: name: image_sample_d_cl_o_merged_v1v3
+# GFX11: %{{[0-9]+}}:vreg_128 = IMAGE_SAMPLE_D_CL_O_V4_V8_gfx11 %5, %3, %2, 15, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec, implicit $exec :: (dereferenceable load (s128), align 4, addrspace 4)
+# GFX11: %{{[0-9]+}}:vgpr_32 = COPY %8.sub0
+# GFX11: %{{[0-9]+}}:vreg_96 = COPY killed %8.sub1_sub2_sub3
+
+name: image_sample_d_cl_o_merged_v1v3
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vreg_256 = IMPLICIT_DEF
+    %6:vgpr_32 = IMAGE_SAMPLE_D_CL_O_V1_V8_gfx11 %5:vreg_256, %3:sgpr_256, %2:sgpr_128, 1, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_SAMPLE_D_CL_O_V3_V8_gfx11 %5:vreg_256, %3:sgpr_256, %2:sgpr_128, 14, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+
+# GFX11-LABEL: name: image_sample_d_o_merged_v1v3
+# GFX11: %{{[0-9]+}}:vreg_128 = IMAGE_SAMPLE_D_O_V4_V7_gfx11 %5, %3, %2, 15, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec, implicit $exec :: (dereferenceable load (s128), align 4, addrspace 4)
+# GFX11: %{{[0-9]+}}:vgpr_32 = COPY %8.sub0
+# GFX11: %{{[0-9]+}}:vreg_96 = COPY killed %8.sub1_sub2_sub3
+
+name: image_sample_d_o_merged_v1v3
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vreg_224 = IMPLICIT_DEF
+    %6:vgpr_32 = IMAGE_SAMPLE_D_O_V1_V7_gfx11 %5:vreg_224, %3:sgpr_256, %2:sgpr_128, 1, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_SAMPLE_D_O_V3_V7_gfx11 %5:vreg_224, %3:sgpr_256, %2:sgpr_128, 14, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+
+# GFX11-LABEL: name: image_sample_lz_merged_v1v3
+# GFX11: %{{[0-9]+}}:vreg_128 = IMAGE_SAMPLE_LZ_V4_V2_nsa_gfx11 %5, %5, %3, %2, 15, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec, implicit $exec :: (dereferenceable load (s128), align 4, addrspace 4)
+# GFX11: %{{[0-9]+}}:vgpr_32 = COPY %8.sub0
+# GFX11: %{{[0-9]+}}:vreg_96 = COPY killed %8.sub1_sub2_sub3
+
+name: image_sample_lz_merged_v1v3
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = IMAGE_SAMPLE_LZ_V1_V2_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 1, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_SAMPLE_LZ_V3_V2_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 14, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+
+# GFX11-LABEL: name: image_sample_lz_o_merged_v1v3
+# GFX11: %{{[0-9]+}}:vreg_128 = IMAGE_SAMPLE_LZ_O_V4_V3_nsa_gfx11 %5, %5, %5, %3, %2, 15, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec, implicit $exec :: (dereferenceable load (s128), align 4, addrspace 4)
+# GFX11: %{{[0-9]+}}:vgpr_32 = COPY %8.sub0
+# GFX11: %{{[0-9]+}}:vreg_96 = COPY killed %8.sub1_sub2_sub3
+
+name: image_sample_lz_o_merged_v1v3
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = IMAGE_SAMPLE_LZ_O_V1_V3_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 1, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_SAMPLE_LZ_O_V3_V3_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 14, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+
+# GFX11-LABEL: name: image_sample_l_o_merged_v1v3
+# GFX11: %{{[0-9]+}}:vreg_128 = IMAGE_SAMPLE_L_O_V4_V4_nsa_gfx11 %5, %5, %5, %5, %3, %2, 15, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec, implicit $exec :: (dereferenceable load (s128), align 4, addrspace 4)
+# GFX11: %{{[0-9]+}}:vgpr_32 = COPY %8.sub0
+# GFX11: %{{[0-9]+}}:vreg_96 = COPY killed %8.sub1_sub2_sub3
+
+name: image_sample_l_o_merged_v1v3
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = IMAGE_SAMPLE_L_O_V1_V4_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 1, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_SAMPLE_L_O_V3_V4_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 14, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---
+
+
+# GFX11-LABEL: name: image_sample_o_merged_v1v3
+# GFX11: %{{[0-9]+}}:vreg_128 = IMAGE_SAMPLE_O_V4_V3_nsa_gfx11 %5, %5, %5, %3, %2, 15, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec, implicit $exec :: (dereferenceable load (s128), align 4, addrspace 4)
+# GFX11: %{{[0-9]+}}:vgpr_32 = COPY %8.sub0
+# GFX11: %{{[0-9]+}}:vreg_96 = COPY killed %8.sub1_sub2_sub3
+
+name: image_sample_o_merged_v1v3
+body:             |
+  bb.0.entry:
+    %0:sgpr_64 = COPY $sgpr0_sgpr1
+    %1:sreg_64_xexec = S_LOAD_DWORDX2_IMM %0, 36, 0
+    %2:sgpr_128 = COPY $sgpr96_sgpr97_sgpr98_sgpr99
+    %3:sgpr_256 = S_LOAD_DWORDX8_IMM %1, 208, 0
+    %4:vgpr_32 = COPY %2.sub3
+    %5:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET %2:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable invariant load 16)
+    %6:vgpr_32 = IMAGE_SAMPLE_O_V1_V3_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 1, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4, addrspace 4)
+    %7:vreg_96 = IMAGE_SAMPLE_O_V3_V3_nsa_gfx11 %5:vgpr_32, %5:vgpr_32, %5:vgpr_32, %3:sgpr_256, %2:sgpr_128, 14, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12, align 16, addrspace 4)
+...
+---