[AMDGPU] Make SGPR spills exec mask agnostic

[critson]

Summary

Explicitly set the exec mask for SGPR spills and reloads.
This fixes a bug where SGPR spills to memory could be incorrect
if the exec mask was 0 (or differed between spill and reload).

Additionally pack scalar subregisters (upto 16/32 per VGPR),
so that the majority of scalar types can be spilt or reloaded
with a simple memory access. This should amortize some of the
additional overhead of manipulating the exec mask.

[arsenm]

Missing new tests / checks. This could probably use a MIR test

llvm/lib/Target/AMDGPU/SIRegisterInfo.cpp

838

Needs a comment explaining the point of this function

858–859

I don't understand these lane numbers; it picks the one past the midpoint lane?

867

I'm not sure I understand only sometimes saving exec_hi

870

We're trying to eliminate use of NoRegister in favor of using the default Register constructor, so you don't need the initializer here (or explicit checks for it later in the function)

880

getUndefRegState

896

Grammar, a odd size->an odd size

906–912

What other places do is have ExecMovOpc and ExecReg set from isWave32 and unify the BuildMI calls

907–908

Could this just turn on all lanes for the memory spill? mov -1 is always free for code size, but some other bitmask may not be

916

Should assert this isn't an SGPRSpill stack ID. You can't reuse the SGPR spill frame index for the real stack spill index

920–921

I think the correct thing to do is use use the base alignment here and add the offset to PtrInfo.getWithOffset

924–930

I know this is what this was doing before, but do we really need to use the intermediate spill pseudo here? As a follow on change could we directly emit the stack operation here?

[critson]

llvm/lib/Target/AMDGPU/SIRegisterInfo.cpp

858–859

Correct, because lanes before the midpoint may be used to store the SGPRs.
I have added a comment.

867

Made the comment cleared.
The idea is that we try to turn off all unused lanes for the load/store on the assumption that this /may/ have some benefit (performance or power). However if we do not have space for saving a copy of EXEC_HI then it is safe not to adjust the lanes.

870

I can get rid of the initializer, but I definitely need the explicit tests as getMatchingSuperReg returns NoRegister which must be handled correctly.

906–912

I have applied a simplification to the code based on this.

907–908

As mentioned above, the idea is that we try to only enable lanes that are used, in case there is some potential benefit.
The only lane combinations which required additional encoding space are S256, S512 and S1024.
Note that S1024 will not occur on Wave32, as we avoid filling all but half the lanes in the VGPR before calling this function.
I would argue that the encoding space is probably not an issue as most common spills are S32, S64, S128.
Of course I am working without any tangible assessment of the benefit of disabling lanes, it might be that for >=S256 we should just set the mask to -1 anyway.

916

Sorry I am not totally clear on what you are saying? This will only get here if it is not an SGPR to VGPR spill slot (tested in spillSGPR / restoreSGPR).

920–921

Again, I am not clear on this. Can you explain further?

924–930

Agreed, I will work on a follow up change to directly emit VMEM.

[critson]

• Address comments.
• Add MIR test.

[critson]

• Tidy test whitespace
• Directly output buffer_store/load from SGPR spill.
• Add assertion for type of spill slot.

[arsenm]

llvm/lib/Target/AMDGPU/SIRegisterInfo.cpp

870

These are the same thing. getMatchingSuperReg should return MCRegister. All of the unsigneds refering to registers should use Register

907–908

But is there an actual benefit? I don't think the hardware saves anything by having fewer lanes. I would rather not add complexity to handle a hypothetical case that doesn't provide a real benefit

916

It's hard to follow how all of the SGPR->VGPR spills work, and it doesn't hurt to be sure this is definitely not an SGPR frame index

920–921

I mean the usage of the MachineMemOperand isn't correct with respect to alignment. It's supposed to be expressed as a base MMO alignment with an offset applied. You're trying to figure out the total alignment. There are helpers for getting the memory operand with an offset

[critson]

llvm/lib/Target/AMDGPU/SIRegisterInfo.cpp

907–908

I have simplified the code.

916

Assertion added.

920–921

As this now directly calls buildSpillLoadStore it only needs to generate an MMO for the base pointer (with base alignment), and the offset is passed directly to buildSpillLoadStore which handles any further alignment requirements (in the way suggested).

[critson]

• Address outstanding comments.
• Add stack offset checks to MIR test.

[critson]

Update spill-wide-sgpr.ll for new tests in master.

[arsenm]

LGTM with some comment fixes

llvm/lib/Target/AMDGPU/SIRegisterInfo.cpp

872

The "In which case use VGPR." reads weirdly to me

885

Capitalize

958

Capitalize

llvm/lib/Target/AMDGPU/SIRegisterInfo.h

void resolveFrameIndex(MachineInstr &MI, Register BaseReg,
                          int64_t Offset) const override;
 
   bool isFrameOffsetLegal(const MachineInstr *MI, Register BaseReg,
                           int64_t Offset) const override;
 
   const TargetRegisterClass *getPointerRegClass(
     const MachineFunction &MF, unsigned Kind = 0) const override;
 
+  void buildSGPRSpillLoadStore(MachineBasicBlock::iterator MI, int Index,
+                               int Offset, unsigned EltSize, Register VGPR,
+                               int64_t VGPRLanes, bool IsLoad) const;
+
   /// If \p OnlyToVGPR is true, this will only succeed if this
   bool spillSGPR(MachineBasicBlock::iterator MI,
                  int FI, RegScavenger *RS,
                  bool OnlyToVGPR = false) const;
 
   bool restoreSGPR(MachineBasicBlock::iterator MI,
                    int FI, RegScavenger *RS,
                    bool OnlyToVGPR = false) const;

llvm/lib/Target/AMDGPU/SIRegisterInfo.cpp

void SIRegisterInfo::buildSpillLoadStore(MachineBasicBlock::iterator MI,
   if (ScratchOffsetRegDelta != 0) {
     // Subtract the offset we added to the ScratchOffset register.
     BuildMI(*MBB, MI, DL, TII->get(AMDGPU::S_SUB_U32), SOffset)
         .addReg(SOffset)
         .addImm(ScratchOffsetRegDelta);
   }
 }
 
+// Generate a VMEM access which loads or stores the VGPR containing an SGPR
+// spill such that all the lanes set in VGPRLanes are loaded or stored.
+// This generates exec mask manipulation and will use SGPRs available in MI
+// or VGPR lanes in the VGPR to save and restore the exec mask.
+void SIRegisterInfo::buildSGPRSpillLoadStore(MachineBasicBlock::iterator MI,
+                                             int Index, int Offset,
+                                             unsigned EltSize, Register VGPR,
+                                             int64_t VGPRLanes,
+                                             bool IsLoad) const {
+  MachineBasicBlock *MBB = MI->getParent();
+  MachineFunction *MF = MBB->getParent();
+  SIMachineFunctionInfo *MFI = MF->getInfo<SIMachineFunctionInfo>();
+  const SIInstrInfo *TII = ST.getInstrInfo();
+
+  Register SuperReg = MI->getOperand(0).getReg();
+  const TargetRegisterClass *RC = getPhysRegClass(SuperReg);
+  ArrayRef<int16_t> SplitParts = getRegSplitParts(RC, EltSize);
+  unsigned NumSubRegs = SplitParts.empty() ? 1 : SplitParts.size();
+  unsigned FirstPart = isWave32 ? Offset * 16 : Offset * 32;
+
+  bool IsKill = MI->getOperand(0).isKill();
+  const DebugLoc &DL = MI->getDebugLoc();
+
+  const bool SuperRegIsExec =
+      SuperReg == AMDGPU::EXEC || SuperReg == AMDGPU::EXEC_LO;
+
+  // If exec mask is stored in the VGPR, make sure it is stored after
+  // any lanes used by the spill (16 lanes on Wave32, 32 lanes on Wave64).
+  const unsigned ExecLoLane = SuperRegIsExec ? 0 : (isWave32 ? 16 : 32);
+  const unsigned ExecHiLane = SuperRegIsExec ? 1 : (isWave32 ? 17 : 33);
+
+  // Always try to use the src/dst SGPRs to hold a copy of the exec mask.
+  // This is not possible when the src value must be valid after spill
+  // or src is smaller than exec mask.  In which case use VGPR.
+  bool StoreExecInVGPR = !IsLoad && (SuperRegIsExec || !IsKill);
+
+  // On Wave32 only handle EXEC_LO.
+  // On Wave64 only update EXEC_HI if there is sufficent space for a copy.
+  bool OnlyExecLo = isWave32 || NumSubRegs == 1;
+
+  unsigned ExecMovOpc = OnlyExecLo ? AMDGPU::S_MOV_B32 : AMDGPU::S_MOV_B64;
+  unsigned ExecReg = OnlyExecLo ? AMDGPU::EXEC_LO : AMDGPU::EXEC;
+
+  Register SavedExecReg;
+
+  // Backup EXEC
+  if (SuperRegIsExec) {
+    // do nothing; exec is already stored in VGPR or will be overwritten
+  } else if (StoreExecInVGPR) {
+    BuildMI(*MBB, MI, DL, TII->getMCOpcodeFromPseudo(AMDGPU::V_WRITELANE_B32),
+            VGPR)
+        .addReg(AMDGPU::EXEC_LO)
+        .addImm(ExecLoLane)
+        .addReg(VGPR, getUndefRegState(IsLoad));
+
+    if (!isWave32) {
+      BuildMI(*MBB, MI, DL, TII->getMCOpcodeFromPseudo(AMDGPU::V_WRITELANE_B32),
+              VGPR)
+          .addReg(AMDGPU::EXEC_HI)
+          .addImm(ExecHiLane)
+          .addReg(VGPR);
+    }
+  } else {
+    if (OnlyExecLo) {
+      SavedExecReg = NumSubRegs == 1
+                         ? SuperReg
+                         : getSubReg(SuperReg, SplitParts[FirstPart]);
+    } else {
+      SavedExecReg =
+          getMatchingSuperReg(getSubReg(SuperReg, SplitParts[FirstPart]),
+                              AMDGPU::sub0, &AMDGPU::SGPR_64RegClass);
+      // If src/dst is an odd size it is possible subreg0 is not aligned.
+      if ((!SavedExecReg || SavedExecReg == AMDGPU::NoRegister) &&
+          NumSubRegs > 2)
+        SavedExecReg =
+            getMatchingSuperReg(getSubReg(SuperReg, SplitParts[FirstPart + 1]),
+                                AMDGPU::sub0, &AMDGPU::SGPR_64RegClass);
+    }
+
+    assert(SavedExecReg && SavedExecReg != AMDGPU::NoRegister);
+    BuildMI(*MBB, MI, DL, TII->get(ExecMovOpc), SavedExecReg).addReg(ExecReg);
+  }
+
+  // Setup EXEC
+  BuildMI(*MBB, MI, DL, TII->get(ExecMovOpc), ExecReg).addImm(VGPRLanes);
+
+  // Load/store VGPR
+  MachineFrameInfo &FrameInfo = MF->getFrameInfo();
+  Align Alignment = FrameInfo.getObjectAlign(Index);
+  MachinePointerInfo PtrInfo =
+      MachinePointerInfo::getFixedStack(*MF, Index, EltSize * Offset);
+  MachineMemOperand *MMO = MF->getMachineMemOperand(
+      PtrInfo, IsLoad ? MachineMemOperand::MOLoad : MachineMemOperand::MOStore,
+      EltSize, commonAlignment(Alignment, EltSize * Offset));
+
+  if (IsLoad) {
+    BuildMI(*MBB, MI, DL, TII->get(AMDGPU::SI_SPILL_V32_RESTORE), VGPR)
+        .addFrameIndex(Index)                // vaddr
+        .addReg(MFI->getScratchRSrcReg())    // srsrc
+        .addReg(MFI->getStackPtrOffsetReg()) // soffset
+        .addImm(Offset * EltSize)            // offset
+        .addMemOperand(MMO)
+        .addReg(VGPR, RegState::Implicit);
+  } else {
+    BuildMI(*MBB, MI, DL, TII->get(AMDGPU::SI_SPILL_V32_SAVE))
+        .addReg(VGPR, getKillRegState(!StoreExecInVGPR)) // src
+        .addFrameIndex(Index)                            // vaddr
+        .addReg(MFI->getScratchRSrcReg())                // srsrc
+        .addReg(MFI->getStackPtrOffsetReg())             // soffset
+        .addImm(Offset * EltSize)                        // offset
+        .addMemOperand(MMO);
+  }
+
+  // Restore EXEC
+  if (SuperRegIsExec && IsLoad) {
+    // do nothing; exec will be overwritten
+  } else if (StoreExecInVGPR) {
+    BuildMI(*MBB, MI, DL, TII->getMCOpcodeFromPseudo(AMDGPU::V_READLANE_B32),
+            AMDGPU::EXEC_LO)
+        .addReg(VGPR, getKillRegState(!IsLoad && isWave32))
+        .addImm(ExecLoLane);
+    if (!isWave32) {
+      BuildMI(*MBB, MI, DL, TII->getMCOpcodeFromPseudo(AMDGPU::V_READLANE_B32),
+              AMDGPU::EXEC_HI)
+          .addReg(VGPR, getKillRegState(!IsLoad))
+          .addImm(ExecHiLane);
+    }
+  } else {
+    assert(SavedExecReg && SavedExecReg != AMDGPU::NoRegister);
+    BuildMI(*MBB, MI, DL, TII->get(ExecMovOpc), ExecReg)
+        .addReg(SavedExecReg, RegState::Kill);
+  }
+}
+
 bool SIRegisterInfo::spillSGPR(MachineBasicBlock::iterator MI,
                                int Index,
                                RegScavenger *RS,
                                bool OnlyToVGPR) const {
   MachineBasicBlock *MBB = MI->getParent();
   MachineFunction *MF = MBB->getParent();
   SIMachineFunctionInfo *MFI = MF->getInfo<SIMachineFunctionInfo>();
   DenseSet<unsigned> SGPRSpillVGPRDefinedSet;
 
   ArrayRef<SIMachineFunctionInfo::SpilledReg> VGPRSpills
     = MFI->getSGPRToVGPRSpills(Index);
   bool SpillToVGPR = !VGPRSpills.empty();
   if (OnlyToVGPR && !SpillToVGPR)
     return false;
 
   const SIInstrInfo *TII = ST.getInstrInfo();
 
   Register SuperReg = MI->getOperand(0).getReg();
   bool IsKill = MI->getOperand(0).isKill();
   const DebugLoc &DL = MI->getDebugLoc();
 
-  MachineFrameInfo &FrameInfo = MF->getFrameInfo();
-
   assert(SpillToVGPR || (SuperReg != MFI->getStackPtrOffsetReg() &&
                          SuperReg != MFI->getFrameOffsetReg()));
 
   assert(SuperReg != AMDGPU::M0 && "m0 should never spill");
 
   unsigned EltSize = 4;
   const TargetRegisterClass *RC = getPhysRegClass(SuperReg);
 
   ArrayRef<int16_t> SplitParts = getRegSplitParts(RC, EltSize);
   unsigned NumSubRegs = SplitParts.empty() ? 1 : SplitParts.size();
 
-  // Scavenged temporary VGPR to use. It must be scavenged once for any number
-  // of spilled subregs.
-  Register TmpVGPR;
-
-  // SubReg carries the "Kill" flag when SubReg == SuperReg.
-  unsigned SubKillState = getKillRegState((NumSubRegs == 1) && IsKill);
+  if (SpillToVGPR) {
     for (unsigned i = 0, e = NumSubRegs; i < e; ++i) {
       Register SubReg =
           NumSubRegs == 1 ? SuperReg : getSubReg(SuperReg, SplitParts[i]);
-
-    if (SpillToVGPR) {
       SIMachineFunctionInfo::SpilledReg Spill = VGPRSpills[i];
 
       // During SGPR spilling to VGPR, determine if the VGPR is defined. The
       // only circumstance in which we say it is undefined is when it is the
       // first spill to this VGPR in the first basic block.
       bool VGPRDefined = true;
       if (MBB == &MF->front())
         VGPRDefined = !SGPRSpillVGPRDefinedSet.insert(Spill.VGPR).second;
 
       // Mark the "old value of vgpr" input undef only if this is the first sgpr
       // spill to this specific vgpr in the first basic block.
       BuildMI(*MBB, MI, DL,
               TII->getMCOpcodeFromPseudo(AMDGPU::V_WRITELANE_B32),
               Spill.VGPR)
         .addReg(SubReg, getKillRegState(IsKill))
         .addImm(Spill.Lane)
         .addReg(Spill.VGPR, VGPRDefined ? 0 : RegState::Undef);
 
       // FIXME: Since this spills to another register instead of an actual
       // frame index, we should delete the frame index when all references to
       // it are fixed.
-    } else {
-      // XXX - Can to VGPR spill fail for some subregisters but not others?
-      if (OnlyToVGPR)
-        return false;
+    }
+    // Scavenged temporary VGPR to use. It must be scavenged once for any number
+    // of spilled subregs.
+    Register TmpVGPR = RS->scavengeRegister(&AMDGPU::VGPR_32RegClass, MI, 0);
+
+    // SubReg carries the "Kill" flag when SubReg == SuperReg.
+    unsigned SubKillState = getKillRegState((NumSubRegs == 1) && IsKill);
+
+    unsigned PerVGPR = isWave32 ? 16 : 32;
+    unsigned NumVGPRs = (NumSubRegs + (PerVGPR - 1)) / PerVGPR;
+    int64_t VGPRLanes = (1LL << std::min(PerVGPR, NumSubRegs)) - 1LL;
+
+    for (unsigned Offset = 0; Offset < NumVGPRs; ++Offset) {
+      unsigned TmpVGPRFlags = RegState::Undef;
+
+      // Write sub registers into the VGPR
+      for (unsigned i = Offset * PerVGPR,
+                    e = std::min((Offset + 1) * PerVGPR, NumSubRegs);
+           i < e; ++i) {
+        Register SubReg =
+            NumSubRegs == 1 ? SuperReg : getSubReg(SuperReg, SplitParts[i]);
 
-      // Spill SGPR to a frame index.
-      if (!TmpVGPR.isValid())
-        TmpVGPR = RS->scavengeRegister(&AMDGPU::VGPR_32RegClass, MI, 0);
-
-      MachineInstrBuilder Mov
-        = BuildMI(*MBB, MI, DL, TII->get(AMDGPU::V_MOV_B32_e32), TmpVGPR)
-        .addReg(SubReg, SubKillState);
+        MachineInstrBuilder WriteLane =
+            BuildMI(*MBB, MI, DL,
+                    TII->getMCOpcodeFromPseudo(AMDGPU::V_WRITELANE_B32),
+                    TmpVGPR)
+                .addReg(SubReg, SubKillState)
+                .addImm(i % PerVGPR)
+                .addReg(TmpVGPR, TmpVGPRFlags);
+        TmpVGPRFlags = 0;
 
         // There could be undef components of a spilled super register.
         // TODO: Can we detect this and skip the spill?
         if (NumSubRegs > 1) {
           // The last implicit use of the SuperReg carries the "Kill" flag.
           unsigned SuperKillState = 0;
-        if (i + 1 == e)
+          if (i + 1 == NumSubRegs)
             SuperKillState |= getKillRegState(IsKill);
-        Mov.addReg(SuperReg, RegState::Implicit | SuperKillState);
+          WriteLane.addReg(SuperReg, RegState::Implicit | SuperKillState);
+        }
       }
 
-      Align Alignment = FrameInfo.getObjectAlign(Index);
-      MachinePointerInfo PtrInfo
-        = MachinePointerInfo::getFixedStack(*MF, Index, EltSize * i);
-      MachineMemOperand *MMO =
-          MF->getMachineMemOperand(PtrInfo, MachineMemOperand::MOStore, EltSize,
-                                   commonAlignment(Alignment, EltSize * i));
-      BuildMI(*MBB, MI, DL, TII->get(AMDGPU::SI_SPILL_V32_SAVE))
-        .addReg(TmpVGPR, RegState::Kill)      // src
-        .addFrameIndex(Index)                 // vaddr
-        .addReg(MFI->getScratchRSrcReg())     // srrsrc
-        .addReg(MFI->getStackPtrOffsetReg())  // soffset
-        .addImm(i * 4)                        // offset
-        .addMemOperand(MMO);
+      // Write out VGPR
+      buildSGPRSpillLoadStore(MI, Index, Offset, EltSize, TmpVGPR, VGPRLanes,
+                              false);
     }
   }
 
   MI->eraseFromParent();
   MFI->addToSpilledSGPRs(NumSubRegs);
   return true;
 }
 
 bool SIRegisterInfo::restoreSGPR(MachineBasicBlock::iterator MI,
                                  int Index,
                                  RegScavenger *RS,
                                  bool OnlyToVGPR) const {
   MachineFunction *MF = MI->getParent()->getParent();
   MachineBasicBlock *MBB = MI->getParent();
   SIMachineFunctionInfo *MFI = MF->getInfo<SIMachineFunctionInfo>();
 
   ArrayRef<SIMachineFunctionInfo::SpilledReg> VGPRSpills
     = MFI->getSGPRToVGPRSpills(Index);
   bool SpillToVGPR = !VGPRSpills.empty();
   if (OnlyToVGPR && !SpillToVGPR)
     return false;
 
-  MachineFrameInfo &FrameInfo = MF->getFrameInfo();
   const SIInstrInfo *TII = ST.getInstrInfo();
   const DebugLoc &DL = MI->getDebugLoc();
 
   Register SuperReg = MI->getOperand(0).getReg();
 
   assert(SuperReg != AMDGPU::M0 && "m0 should never spill");
 
   unsigned EltSize = 4;
 
   const TargetRegisterClass *RC = getPhysRegClass(SuperReg);
 
   ArrayRef<int16_t> SplitParts = getRegSplitParts(RC, EltSize);
   unsigned NumSubRegs = SplitParts.empty() ? 1 : SplitParts.size();
 
-  Register TmpVGPR;
-
+  if (SpillToVGPR) {
     for (unsigned i = 0, e = NumSubRegs; i < e; ++i) {
       Register SubReg =
           NumSubRegs == 1 ? SuperReg : getSubReg(SuperReg, SplitParts[i]);
 
-    if (SpillToVGPR) {
       SIMachineFunctionInfo::SpilledReg Spill = VGPRSpills[i];
       auto MIB =
         BuildMI(*MBB, MI, DL, TII->getMCOpcodeFromPseudo(AMDGPU::V_READLANE_B32),
                 SubReg)
         .addReg(Spill.VGPR)
         .addImm(Spill.Lane);
 
       if (NumSubRegs > 1 && i == 0)
         MIB.addReg(SuperReg, RegState::ImplicitDefine);
-    } else {
-      if (OnlyToVGPR)
-        return false;
-
-      // Restore SGPR from a stack slot.
-      // FIXME: We should use S_LOAD_DWORD here for VI.
-      if (!TmpVGPR.isValid())
-        TmpVGPR = RS->scavengeRegister(&AMDGPU::VGPR_32RegClass, MI, 0);
-      Align Alignment = FrameInfo.getObjectAlign(Index);
-
-      MachinePointerInfo PtrInfo
-        = MachinePointerInfo::getFixedStack(*MF, Index, EltSize * i);
-
-      MachineMemOperand *MMO =
-          MF->getMachineMemOperand(PtrInfo, MachineMemOperand::MOLoad, EltSize,
-                                   commonAlignment(Alignment, EltSize * i));
+    }
+    Register TmpVGPR = RS->scavengeRegister(&AMDGPU::VGPR_32RegClass, MI, 0);
 
-      BuildMI(*MBB, MI, DL, TII->get(AMDGPU::SI_SPILL_V32_RESTORE), TmpVGPR)
-        .addFrameIndex(Index)                 // vaddr
-        .addReg(MFI->getScratchRSrcReg())     // srsrc
-        .addReg(MFI->getStackPtrOffsetReg())  // soffset
-        .addImm(i * 4)                        // offset
-        .addMemOperand(MMO);
+    unsigned PerVGPR = isWave32 ? 16 : 32;
+    unsigned NumVGPRs = (NumSubRegs + (PerVGPR - 1)) / PerVGPR;
+    int64_t VGPRLanes = (1LL << std::min(PerVGPR, NumSubRegs)) - 1LL;
+
+    for (unsigned Offset = 0; Offset < NumVGPRs; ++Offset) {
+      // Load in VGPR data
+      buildSGPRSpillLoadStore(MI, Index, Offset, EltSize, TmpVGPR, VGPRLanes,
+                              true);
+
+      // Unpack lanes
+      for (unsigned i = Offset * PerVGPR,
+                    e = std::min((Offset + 1) * PerVGPR, NumSubRegs);
+           i < e; ++i) {
+        Register SubReg =
+            NumSubRegs == 1 ? SuperReg : getSubReg(SuperReg, SplitParts[i]);
 
+        bool LastSubReg = (i + 1 == e);
         auto MIB =
-        BuildMI(*MBB, MI, DL, TII->get(AMDGPU::V_READFIRSTLANE_B32), SubReg)
-        .addReg(TmpVGPR, RegState::Kill);
+            BuildMI(*MBB, MI, DL,
+                    TII->getMCOpcodeFromPseudo(AMDGPU::V_READLANE_B32), SubReg)
+                .addReg(TmpVGPR, getKillRegState(LastSubReg))
+                .addImm(i);
 
-      if (NumSubRegs > 1)
-        MIB.addReg(MI->getOperand(0).getReg(), RegState::ImplicitDefine);
+        if (NumSubRegs > 1 && i == 0)
+          MIB.addReg(SuperReg, RegState::ImplicitDefine);
+      }
     }
   }
 
   MI->eraseFromParent();
   return true;
 }
 
 /// Special case of eliminateFrameIndex. Returns true if the SGPR was spilled to

llvm/test/CodeGen/AMDGPU/control-flow-fastregalloc.ll

 
 ; Spill load
 ; GCN: buffer_store_dword [[LOAD0]], off, s[0:3], 0 offset:[[LOAD0_OFFSET:[0-9]+]] ; 4-byte Folded Spill
 
 ; Spill saved exec
 ; VGPR: v_writelane_b32 [[SPILL_VGPR:v[0-9]+]], s[[SAVEEXEC_LO]], [[SAVEEXEC_LO_LANE:[0-9]+]]
 ; VGPR: v_writelane_b32 [[SPILL_VGPR]], s[[SAVEEXEC_HI]], [[SAVEEXEC_HI_LANE:[0-9]+]]
 
-; VMEM: v_mov_b32_e32 v[[V_SAVEEXEC_LO:[0-9]+]], s[[SAVEEXEC_LO]]
-; VMEM: buffer_store_dword v[[V_SAVEEXEC_LO]], off, s[0:3], 0 offset:20 ; 4-byte Folded Spill
-; VMEM: v_mov_b32_e32 v[[V_SAVEEXEC_HI:[0-9]+]], s[[SAVEEXEC_HI]]
-; VMEM: buffer_store_dword v[[V_SAVEEXEC_HI]], off, s[0:3], 0 offset:24 ; 4-byte Folded Spill
+; VMEM: v_writelane_b32 v[[V_SAVEEXEC:[0-9]+]], s[[SAVEEXEC_LO]], 0
+; VMEM: v_writelane_b32 v[[V_SAVEEXEC]], s[[SAVEEXEC_HI]], 1
+; VMEM: buffer_store_dword v[[V_SAVEEXEC]], off, s[0:3], 0 offset:20 ; 4-byte Folded Spill
 
 ; GCN: s_mov_b64 exec, s{{\[}}[[ANDEXEC_LO]]:[[ANDEXEC_HI]]{{\]}}
 
 ; GCN: s_cbranch_execz [[ENDIF:BB[0-9]+_[0-9]+]]
 
 ; GCN: ; %bb.{{[0-9]+}}: ; %if
 ; GCN: s_mov_b32 m0, -1
 ; GCN: ds_read_b32 [[LOAD1:v[0-9]+]]
 ; GCN: buffer_load_dword [[RELOAD_LOAD0:v[0-9]+]], off, s[0:3], 0 offset:[[LOAD0_OFFSET]] ; 4-byte Folded Reload
 ; GCN: s_waitcnt vmcnt(0) lgkmcnt(0)
 
 
 ; Spill val register
 ; GCN: v_add_i32_e32 [[VAL:v[0-9]+]], vcc, [[LOAD1]], [[RELOAD_LOAD0]]
 ; GCN: buffer_store_dword [[VAL]], off, s[0:3], 0 offset:[[VAL_OFFSET:[0-9]+]] ; 4-byte Folded Spill
 
 ; VMEM: [[ENDIF]]:
 
 ; Reload and restore exec mask
 ; VGPR: v_readlane_b32 s[[S_RELOAD_SAVEEXEC_LO:[0-9]+]], [[SPILL_VGPR]], [[SAVEEXEC_LO_LANE]]
 ; VGPR: v_readlane_b32 s[[S_RELOAD_SAVEEXEC_HI:[0-9]+]], [[SPILL_VGPR]], [[SAVEEXEC_HI_LANE]]
 
 
 
-; VMEM: buffer_load_dword v[[V_RELOAD_SAVEEXEC_LO:[0-9]+]], off, s[0:3], 0 offset:20 ; 4-byte Folded Reload
+; VMEM: buffer_load_dword v[[V_RELOAD_SAVEEXEC:[0-9]+]], off, s[0:3], 0 offset:20 ; 4-byte Folded Reload
 ; VMEM: s_waitcnt vmcnt(0)
-; VMEM: v_readfirstlane_b32 s[[S_RELOAD_SAVEEXEC_LO:[0-9]+]], v[[V_RELOAD_SAVEEXEC_LO]]
-
-; VMEM: buffer_load_dword v[[V_RELOAD_SAVEEXEC_HI:[0-9]+]], off, s[0:3], 0 offset:24 ; 4-byte Folded Reload
-; VMEM: s_waitcnt vmcnt(0)
-; VMEM: v_readfirstlane_b32 s[[S_RELOAD_SAVEEXEC_HI:[0-9]+]], v[[V_RELOAD_SAVEEXEC_HI]]
+; VMEM: v_readlane_b32 s[[S_RELOAD_SAVEEXEC_LO:[0-9]+]], v[[V_RELOAD_SAVEEXEC]], 0
+; VMEM: v_readlane_b32 s[[S_RELOAD_SAVEEXEC_HI:[0-9]+]], v[[V_RELOAD_SAVEEXEC]], 1
 
 ; GCN: s_or_b64 exec, exec, s{{\[}}[[S_RELOAD_SAVEEXEC_LO]]:[[S_RELOAD_SAVEEXEC_HI]]{{\]}}
 
 ; Restore val
 ; GCN: buffer_load_dword [[RELOAD_VAL:v[0-9]+]], off, s[0:3], 0 offset:[[VAL_OFFSET]] ; 4-byte Folded Reload
 
 ; GCN: flat_store_dword v{{\[[0-9]+:[0-9]+\]}}, [[RELOAD_VAL]]
 define amdgpu_kernel void @divergent_if_endif(i32 addrspace(1)* %out) #0 {

 ; Spill load
 ; GCN: buffer_store_dword [[LOAD0]], off, s[0:3], 0 offset:[[LOAD0_OFFSET:[0-9]+]] ; 4-byte Folded Spill
 
 ; Spill saved exec
 ; VGPR: v_writelane_b32 [[SPILL_VGPR:v[0-9]+]], s[[SAVEEXEC_LO]], [[SAVEEXEC_LO_LANE:[0-9]+]]
 ; VGPR: v_writelane_b32 [[SPILL_VGPR]], s[[SAVEEXEC_HI]], [[SAVEEXEC_HI_LANE:[0-9]+]]
 
 
-; VMEM: v_mov_b32_e32 v[[V_SAVEEXEC_LO:[0-9]+]], s[[SAVEEXEC_LO]]
-; VMEM: buffer_store_dword v[[V_SAVEEXEC_LO]], off, s[0:3], 0 offset:24 ; 4-byte Folded Spill
-; VMEM: v_mov_b32_e32 v[[V_SAVEEXEC_HI:[0-9]+]], s[[SAVEEXEC_HI]]
-; VMEM: buffer_store_dword v[[V_SAVEEXEC_HI]], off, s[0:3], 0 offset:28 ; 4-byte Folded Spill
+; VMEM: v_writelane_b32 v[[V_SAVEEXEC:[0-9]+]], s[[SAVEEXEC_LO]], 0
+; VMEM: v_writelane_b32 v[[V_SAVEEXEC]], s[[SAVEEXEC_HI]], 1
+; VMEM: buffer_store_dword v[[V_SAVEEXEC]], off, s[0:3], 0 offset:24 ; 4-byte Folded Spill
 
 ; GCN: s_mov_b64 exec, s{{\[}}[[ANDEXEC_LO]]:[[ANDEXEC_HI]]{{\]}}
 
 ; GCN-NEXT: s_cbranch_execz [[END:BB[0-9]+_[0-9]+]]
 
 
 ; GCN: [[LOOP:BB[0-9]+_[0-9]+]]:
 ; GCN: buffer_load_dword v[[VAL_LOOP_RELOAD:[0-9]+]], off, s[0:3], 0 offset:[[LOAD0_OFFSET]] ; 4-byte Folded Reload
 ; GCN: v_subrev_i32_e32 [[VAL_LOOP:v[0-9]+]], vcc, v{{[0-9]+}}, v[[VAL_LOOP_RELOAD]]
 ; GCN: s_cmp_lg_u32
 ; GCN: buffer_store_dword [[VAL_LOOP]], off, s[0:3], 0 offset:{{[0-9]+}} ; 4-byte Folded Spill
 ; GCN-NEXT: s_cbranch_scc1 [[LOOP]]
 
 ; GCN: buffer_store_dword [[VAL_LOOP]], off, s[0:3], 0 offset:[[VAL_SUB_OFFSET:[0-9]+]] ; 4-byte Folded Spill
 
 ; GCN: [[END]]:
 ; VGPR: v_readlane_b32 s[[S_RELOAD_SAVEEXEC_LO:[0-9]+]], [[SPILL_VGPR]], [[SAVEEXEC_LO_LANE]]
 ; VGPR: v_readlane_b32 s[[S_RELOAD_SAVEEXEC_HI:[0-9]+]], [[SPILL_VGPR]], [[SAVEEXEC_HI_LANE]]
 
-; VMEM: buffer_load_dword v[[V_RELOAD_SAVEEXEC_LO:[0-9]+]], off, s[0:3], 0 offset:24 ; 4-byte Folded Reload
-; VMEM: s_waitcnt vmcnt(0)
-; VMEM: v_readfirstlane_b32 s[[S_RELOAD_SAVEEXEC_LO:[0-9]+]], v[[V_RELOAD_SAVEEXEC_LO]]
-
-; VMEM: buffer_load_dword v[[V_RELOAD_SAVEEXEC_HI:[0-9]+]], off, s[0:3], 0 offset:28 ; 4-byte Folded Reload
+; VMEM: buffer_load_dword v[[V_RELOAD_SAVEEXEC:[0-9]+]], off, s[0:3], 0 offset:24 ; 4-byte Folded Reload
 ; VMEM: s_waitcnt vmcnt(0)
-; VMEM: v_readfirstlane_b32 s[[S_RELOAD_SAVEEXEC_HI:[0-9]+]], v[[V_RELOAD_SAVEEXEC_HI]]
+; VMEM: v_readlane_b32 s[[S_RELOAD_SAVEEXEC_LO:[0-9]+]], v[[V_RELOAD_SAVEEXEC]], 0
+; VMEM: v_readlane_b32 s[[S_RELOAD_SAVEEXEC_HI:[0-9]+]], v[[V_RELOAD_SAVEEXEC]], 1
 
 ; GCN: s_or_b64 exec, exec, s{{\[}}[[S_RELOAD_SAVEEXEC_LO]]:[[S_RELOAD_SAVEEXEC_HI]]{{\]}}
 ; GCN: buffer_load_dword v[[VAL_END:[0-9]+]], off, s[0:3], 0 offset:[[VAL_SUB_OFFSET]] ; 4-byte Folded Reload
 
 ; GCN: flat_store_dword v{{\[[0-9]+:[0-9]+\]}}, v[[VAL_END]]
 define amdgpu_kernel void @divergent_loop(i32 addrspace(1)* %out) #0 {
 entry:
   %tid = call i32 @llvm.amdgcn.workitem.id.x()

 
 ; Spill load
 ; GCN: buffer_store_dword [[LOAD0]], off, s[0:3], 0 offset:[[LOAD0_OFFSET:[0-9]+]] ; 4-byte Folded Spill
 
 ; Spill saved exec
 ; VGPR: v_writelane_b32 [[SPILL_VGPR:v[0-9]+]], s[[SAVEEXEC_LO]], [[SAVEEXEC_LO_LANE:[0-9]+]]
 ; VGPR: v_writelane_b32 [[SPILL_VGPR]], s[[SAVEEXEC_HI]], [[SAVEEXEC_HI_LANE:[0-9]+]]
 
-; VMEM: v_mov_b32_e32 v[[V_SAVEEXEC_LO:[0-9]+]], s[[SAVEEXEC_LO]]
-; VMEM: buffer_store_dword v[[V_SAVEEXEC_LO]], off, s[0:3], 0 offset:[[SAVEEXEC_LO_OFFSET:[0-9]+]] ; 4-byte Folded Spill
-; VMEM: v_mov_b32_e32 v[[V_SAVEEXEC_HI:[0-9]+]], s[[SAVEEXEC_HI]]
-; VMEM: buffer_store_dword v[[V_SAVEEXEC_HI]], off, s[0:3], 0 offset:[[SAVEEXEC_HI_OFFSET:[0-9]+]] ; 4-byte Folded Spill
+; VMEM: v_writelane_b32 v[[V_SAVEEXEC:[0-9]+]], s[[SAVEEXEC_LO]], 0
+; VMEM: v_writelane_b32 v[[V_SAVEEXEC]], s[[SAVEEXEC_HI]], 1
+; VMEM: buffer_store_dword v[[V_SAVEEXEC]], off, s[0:3], 0 offset:[[SAVEEXEC_OFFSET:[0-9]+]] ; 4-byte Folded Spill
 
 ; GCN: s_mov_b64 exec, [[CMP0]]
 
 ; FIXME: It makes no sense to put this skip here
 ; GCN: s_cbranch_execz [[FLOW:BB[0-9]+_[0-9]+]]
 ; GCN-NEXT: s_branch [[ELSE:BB[0-9]+_[0-9]+]]
 
 ; GCN: [[FLOW]]: ; %Flow
 ; VGPR: v_readlane_b32 s[[FLOW_S_RELOAD_SAVEEXEC_LO:[0-9]+]], [[SPILL_VGPR]], [[SAVEEXEC_LO_LANE]]
 ; VGPR: v_readlane_b32 s[[FLOW_S_RELOAD_SAVEEXEC_HI:[0-9]+]], [[SPILL_VGPR]], [[SAVEEXEC_HI_LANE]]
 
 
-; VMEM: buffer_load_dword v[[FLOW_V_RELOAD_SAVEEXEC_LO:[0-9]+]], off, s[0:3], 0 offset:[[SAVEEXEC_LO_OFFSET]]
+; VMEM: buffer_load_dword v[[FLOW_V_RELOAD_SAVEEXEC:[0-9]+]], off, s[0:3], 0 offset:[[SAVEEXEC_OFFSET]]
 ; VMEM: s_waitcnt vmcnt(0)
-; VMEM: v_readfirstlane_b32 s[[FLOW_S_RELOAD_SAVEEXEC_LO:[0-9]+]], v[[FLOW_V_RELOAD_SAVEEXEC_LO]]
-
-; VMEM: buffer_load_dword v[[FLOW_V_RELOAD_SAVEEXEC_HI:[0-9]+]], off, s[0:3], 0 offset:[[SAVEEXEC_HI_OFFSET]] ; 4-byte Folded Reload
-; VMEM: s_waitcnt vmcnt(0)
-; VMEM: v_readfirstlane_b32 s[[FLOW_S_RELOAD_SAVEEXEC_HI:[0-9]+]], v[[FLOW_V_RELOAD_SAVEEXEC_HI]]
+; VMEM: v_readlane_b32 s[[FLOW_S_RELOAD_SAVEEXEC_LO:[0-9]+]], v[[FLOW_V_RELOAD_SAVEEXEC]], 0
+; VMEM: v_readlane_b32 s[[FLOW_S_RELOAD_SAVEEXEC_HI:[0-9]+]], v[[FLOW_V_RELOAD_SAVEEXEC]], 1
 
 ; GCN: s_or_saveexec_b64 s{{\[}}[[FLOW_S_RELOAD_SAVEEXEC_LO]]:[[FLOW_S_RELOAD_SAVEEXEC_HI]]{{\]}}, s{{\[}}[[FLOW_S_RELOAD_SAVEEXEC_LO]]:[[FLOW_S_RELOAD_SAVEEXEC_HI]]{{\]}}
 
 ; Regular spill value restored after exec modification
 ; GCN: buffer_load_dword [[FLOW_VAL:v[0-9]+]], off, s[0:3], 0 offset:[[FLOW_VAL_OFFSET:[0-9]+]] ; 4-byte Folded Reload
 
 
 ; Spill saved exec
 ; VGPR: v_writelane_b32 [[SPILL_VGPR]], s[[FLOW_S_RELOAD_SAVEEXEC_LO]], [[FLOW_SAVEEXEC_LO_LANE:[0-9]+]]
 ; VGPR: v_writelane_b32 [[SPILL_VGPR]], s[[FLOW_S_RELOAD_SAVEEXEC_HI]], [[FLOW_SAVEEXEC_HI_LANE:[0-9]+]]
 
 
-; VMEM: v_mov_b32_e32 v[[FLOW_V_SAVEEXEC_LO:[0-9]+]], s[[FLOW_S_RELOAD_SAVEEXEC_LO]]
-; VMEM: buffer_store_dword v[[FLOW_V_SAVEEXEC_LO]], off, s[0:3], 0 offset:[[FLOW_SAVEEXEC_LO_OFFSET:[0-9]+]] ; 4-byte Folded Spill
-; VMEM: v_mov_b32_e32 v[[FLOW_V_SAVEEXEC_HI:[0-9]+]], s[[FLOW_S_RELOAD_SAVEEXEC_HI]]
-; VMEM: buffer_store_dword v[[FLOW_V_SAVEEXEC_HI]], off, s[0:3], 0 offset:[[FLOW_SAVEEXEC_HI_OFFSET:[0-9]+]] ; 4-byte Folded Spill
+; VMEM: v_writelane_b32 v[[FLOW_V_SAVEEXEC:[0-9]+]], s[[FLOW_S_RELOAD_SAVEEXEC_LO]], 0
+; VMEM: v_writelane_b32 v[[FLOW_V_SAVEEXEC]], s[[FLOW_S_RELOAD_SAVEEXEC_HI]], 1
+; VMEM: buffer_store_dword v[[FLOW_V_SAVEEXEC]], off, s[0:3], 0 offset:[[FLOW_SAVEEXEC_OFFSET:[0-9]+]] ; 4-byte Folded Spill
 
 ; GCN: buffer_store_dword [[FLOW_VAL]], off, s[0:3], 0 offset:[[RESULT_OFFSET:[0-9]+]] ; 4-byte Folded Spill
 ; GCN: s_xor_b64 exec, exec, s{{\[}}[[FLOW_S_RELOAD_SAVEEXEC_LO]]:[[FLOW_S_RELOAD_SAVEEXEC_HI]]{{\]}}
 ; GCN-NEXT: s_cbranch_execz [[ENDIF:BB[0-9]+_[0-9]+]]
 
 
 ; GCN: ; %bb.{{[0-9]+}}: ; %if
 ; GCN: ds_read_b32
 ; GCN: buffer_load_dword v[[LOAD0_RELOAD:[0-9]+]], off, s[0:3], 0 offset:[[LOAD0_OFFSET]] ; 4-byte Folded Reload
 ; GCN: v_add_i32_e32 [[ADD:v[0-9]+]], vcc, v{{[0-9]+}}, v[[LOAD0_RELOAD]]
 ; GCN: buffer_store_dword [[ADD]], off, s[0:3], 0 offset:[[RESULT_OFFSET]] ; 4-byte Folded Spill
 ; GCN-NEXT: s_branch [[ENDIF:BB[0-9]+_[0-9]+]]
 
 ; GCN: [[ELSE]]: ; %else
 ; GCN: buffer_load_dword v[[LOAD0_RELOAD:[0-9]+]], off, s[0:3], 0 offset:[[LOAD0_OFFSET]] ; 4-byte Folded Reload
 ; GCN: v_subrev_i32_e32 [[SUB:v[0-9]+]], vcc, v{{[0-9]+}}, v[[LOAD0_RELOAD]]
 ; GCN: buffer_store_dword [[ADD]], off, s[0:3], 0 offset:[[FLOW_RESULT_OFFSET:[0-9]+]] ; 4-byte Folded Spill
 ; GCN-NEXT: s_branch [[FLOW]]
 
 ; GCN: [[ENDIF]]:
 ; VGPR: v_readlane_b32 s[[S_RELOAD_SAVEEXEC_LO:[0-9]+]], [[SPILL_VGPR]], [[FLOW_SAVEEXEC_LO_LANE]]
 ; VGPR: v_readlane_b32 s[[S_RELOAD_SAVEEXEC_HI:[0-9]+]], [[SPILL_VGPR]], [[FLOW_SAVEEXEC_HI_LANE]]
 
 
-; VMEM: buffer_load_dword v[[V_RELOAD_SAVEEXEC_LO:[0-9]+]], off, s[0:3], 0 offset:[[FLOW_SAVEEXEC_LO_OFFSET]] ; 4-byte Folded Reload
-; VMEM: s_waitcnt vmcnt(0)
-; VMEM: v_readfirstlane_b32 s[[S_RELOAD_SAVEEXEC_LO:[0-9]+]], v[[V_RELOAD_SAVEEXEC_LO]]
-
-; VMEM: buffer_load_dword v[[V_RELOAD_SAVEEXEC_HI:[0-9]+]], off, s[0:3], 0 offset:[[FLOW_SAVEEXEC_HI_OFFSET]] ; 4-byte Folded Reload
+; VMEM: buffer_load_dword v[[V_RELOAD_SAVEEXEC:[0-9]+]], off, s[0:3], 0 offset:[[FLOW_SAVEEXEC_OFFSET]] ; 4-byte Folded Reload
 ; VMEM: s_waitcnt vmcnt(0)
-; VMEM: v_readfirstlane_b32 s[[S_RELOAD_SAVEEXEC_HI:[0-9]+]], v[[V_RELOAD_SAVEEXEC_HI]]
+; VMEM: v_readlane_b32 s[[S_RELOAD_SAVEEXEC_LO:[0-9]+]], v[[V_RELOAD_SAVEEXEC]], 0
+; VMEM: v_readlane_b32 s[[S_RELOAD_SAVEEXEC_HI:[0-9]+]], v[[V_RELOAD_SAVEEXEC]], 1
 
 ; GCN: s_or_b64 exec, exec, s{{\[}}[[S_RELOAD_SAVEEXEC_LO]]:[[S_RELOAD_SAVEEXEC_HI]]{{\]}}
 
 ; GCN: buffer_load_dword v[[RESULT:[0-9]+]], off, s[0:3], 0 offset:[[RESULT_OFFSET]] ; 4-byte Folded Reload
 ; GCN: flat_store_dword v{{\[[0-9]+:[0-9]+\]}}, v[[RESULT]]
 define amdgpu_kernel void @divergent_if_else_endif(i32 addrspace(1)* %out) #0 {
 entry:
   %tid = call i32 @llvm.amdgcn.workitem.id.x()

llvm/test/CodeGen/AMDGPU/partial-sgpr-to-vgpr-spills.ll

 ; GCN-NEXT: v_readlane_b32 s{{[0-9]+}}, v31, 27
 ; GCN-NEXT: v_readlane_b32 s{{[0-9]+}}, v31, 28
 ; GCN-NEXT: v_readlane_b32 s{{[0-9]+}}, v31, 29
 ; GCN-NEXT: v_readlane_b32 s{{[0-9]+}}, v31, 30
 ; GCN-NEXT: v_readlane_b32 s[[USE_TMP_HI:[0-9]+]], v31, 31
 ; GCN: ; use s{{\[}}[[USE_TMP_LO]]:[[USE_TMP_HI]]{{\]}}
 
 ; GCN: buffer_load_dword v[[RESTORE_TMP:[0-9]+]], off, s{{\[[0-9]+:[0-9]+\]}}, 0
-; GCN: v_readfirstlane_b32 s[[USE_TMP_LO:[0-9]+]], v[[RESTORE_TMP]]
-; GCN: buffer_load_dword v{{[0-9]+}}, off, s{{\[[0-9]+:[0-9]+\]}}, 0
-; GCN: v_readfirstlane_b32 s[[USE_TMP_HI:[0-9]+]], v[[RESTORE_TMP]]
+; GCN: v_readlane_b32 s[[USE_TMP_LO:[0-9]+]], v[[RESTORE_TMP]], 0
+; GCN: v_readlane_b32 s[[USE_TMP_HI:[0-9]+]], v[[RESTORE_TMP]], 1
 ; GCN: ;;#ASMSTART
 ; GCN: ; use s{{\[}}[[USE_TMP_LO]]:[[USE_TMP_HI]]{{\]}}
 define amdgpu_kernel void @no_vgprs_last_sgpr_spill(i32 addrspace(1)* %out, i32 %in) #1 {
   call void asm sideeffect "", "~{v[0:7]}" () #0
   call void asm sideeffect "", "~{v[8:15]}" () #0
   call void asm sideeffect "", "~{v[16:23]}" () #0
   call void asm sideeffect "", "~{v[24:27]}"() #0
   call void asm sideeffect "", "~{v[28:29]}"() #0

llvm/test/CodeGen/AMDGPU/sgpr-spill.mir

+# RUN: llc -mtriple=amdgcn -mcpu=gfx900 -verify-machineinstrs -run-pass=prologepilog %s -o - | FileCheck -check-prefix=CHECK -check-prefix=GCN64 %s
+# RUN: llc -mtriple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32,-wavefrontsize64 -verify-machineinstrs -run-pass=prologepilog %s -o - | FileCheck -check-prefix=CHECK -check-prefix=GCN32 %s
+
+
+# CHECK-LABEL: name: check_spill
+
+# S32 with kill
+# CHECK: V_WRITELANE
+# CHECK: $sgpr12 = S_MOV_B32 $exec_lo
+# CHECK: $exec_lo = S_MOV_B32 1
+# CHECK: BUFFER_STORE_DWORD_OFFSET
+# CHECK: $exec_lo = S_MOV_B32 killed $sgpr12
+
+# S32 without kill
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: $exec_lo = S_MOV_B32 1
+# CHECK: BUFFER_STORE_DWORD_OFFSET
+# CHECK: $exec_lo = V_READLANE
+
+# S64 with kill
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# GCN32: $sgpr12 = S_MOV_B32 $exec_lo
+# GCN64: $sgpr12_sgpr13 = S_MOV_B64 $exec
+# GCN32: $exec_lo = S_MOV_B32 3
+# GCN64: $exec = S_MOV_B64 3
+# CHECK: BUFFER_STORE_DWORD_OFFSET
+# GCN32: $exec_lo = S_MOV_B32 killed $sgpr12
+# GCN64: $exec = S_MOV_B64 killed $sgpr12_sgpr13
+
+# S64 without kill
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# GCN64: V_WRITELANE
+# GCN32: $exec_lo = S_MOV_B32 3
+# GCN64: $exec = S_MOV_B64 3
+# CHECK: BUFFER_STORE_DWORD_OFFSET
+# CHECK: $exec_lo = V_READLANE
+# GCN64: $exec_hi = V_READLANE
+
+# S96
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# GCN32: $sgpr12 = S_MOV_B32 $exec_lo
+# GCN64: $sgpr12_sgpr13 = S_MOV_B64 $exec
+# GCN32: $exec_lo = S_MOV_B32 7
+# GCN64: $exec = S_MOV_B64 7
+# CHECK: BUFFER_STORE_DWORD_OFFSET
+# GCN32: $exec_lo = S_MOV_B32 killed $sgpr12
+# GCN64: $exec = S_MOV_B64 killed $sgpr12_sgpr13
+
+# S128
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# GCN32: $sgpr12 = S_MOV_B32 $exec_lo
+# GCN64: $sgpr12_sgpr13 = S_MOV_B64 $exec
+# GCN32: $exec_lo = S_MOV_B32 15
+# GCN64: $exec = S_MOV_B64 15
+# CHECK: BUFFER_STORE_DWORD_OFFSET
+# GCN32: $exec_lo = S_MOV_B32 killed $sgpr12
+# GCN64: $exec = S_MOV_B64 killed $sgpr12_sgpr13
+
+# S160
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# GCN32: $sgpr12 = S_MOV_B32 $exec_lo
+# GCN64: $sgpr12_sgpr13 = S_MOV_B64 $exec
+# GCN32: $exec_lo = S_MOV_B32 31
+# GCN64: $exec = S_MOV_B64 31
+# CHECK: BUFFER_STORE_DWORD_OFFSET
+# GCN32: $exec_lo = S_MOV_B32 killed $sgpr12
+# GCN64: $exec = S_MOV_B64 killed $sgpr12_sgpr13
+
+# S256
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# GCN32: $sgpr12 = S_MOV_B32 $exec_lo
+# GCN64: $sgpr12_sgpr13 = S_MOV_B64 $exec
+# GCN32: $exec_lo = S_MOV_B32 255
+# GCN64: $exec = S_MOV_B64 255
+# CHECK: BUFFER_STORE_DWORD_OFFSET
+# GCN32: $exec_lo = S_MOV_B32 killed $sgpr12
+# GCN64: $exec = S_MOV_B64 killed $sgpr12_sgpr13
+
+# S512
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# GCN32: $sgpr12 = S_MOV_B32 $exec_lo
+# GCN64: $sgpr12_sgpr13 = S_MOV_B64 $exec
+# GCN32: $exec_lo = S_MOV_B32 65535
+# GCN64: $exec = S_MOV_B64 65535
+# CHECK: BUFFER_STORE_DWORD_OFFSET
+# GCN32: $exec_lo = S_MOV_B32 killed $sgpr12
+# GCN64: $exec = S_MOV_B64 killed $sgpr12_sgpr13
+
+# S1024
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# GCN32: $sgpr64 = S_MOV_B32 $exec_lo
+# GCN32: $exec_lo = S_MOV_B32 65535
+# GCN32: BUFFER_STORE_DWORD_OFFSET
+# GCN32: $exec_lo = S_MOV_B32 killed $sgpr64
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# CHECK: V_WRITELANE
+# GCN32: $sgpr80 = S_MOV_B32 $exec_lo
+# GCN64: $sgpr64_sgpr65 = S_MOV_B64 $exec
+# GCN32: $exec_lo = S_MOV_B32 65535
+# GCN64: $exec = S_MOV_B64 4294967295
+# CHECK: BUFFER_STORE_DWORD_OFFSET
+# GCN32: $exec_lo = S_MOV_B32 killed $sgpr80
+# GCN64: $exec = S_MOV_B64 killed $sgpr64_sgpr65
+
+--- |
+
+  define amdgpu_kernel void @check_spill() #0 {
+    ret void
+  }
+  
+  define amdgpu_kernel void @check_reload() #0 {
+    ret void
+  }
+
+  attributes #0 = {  "frame-pointer"="all" }
+...
+---
+name:            check_spill
+tracksRegLiveness: true
+liveins:         
+  - { reg: '$sgpr4_sgpr5' }
+  - { reg: '$sgpr6_sgpr7' }
+  - { reg: '$sgpr8' }
+frameInfo:       
+  maxAlignment:    4
+stack:           
+  - { id: 0, type: spill-slot, size: 4, alignment: 4 }
+  - { id: 1, type: spill-slot, size: 8, alignment: 4 }
+  - { id: 2, type: spill-slot, size: 12, alignment: 4 }
+  - { id: 3, type: spill-slot, size: 16, alignment: 4 }
+  - { id: 4, type: spill-slot, size: 20, alignment: 4 }
+  - { id: 5, type: spill-slot, size: 32, alignment: 4 }
+  - { id: 6, type: spill-slot, size: 64, alignment: 4 }
+  - { id: 7, type: spill-slot, size: 128, alignment: 4 }
+machineFunctionInfo: 
+  explicitKernArgSize: 660
+  maxKernArgAlign: 4
+  isEntryFunction: true
+  waveLimiter:     true
+  scratchRSrcReg:  '$sgpr96_sgpr97_sgpr98_sgpr99'
+  stackPtrOffsetReg: '$sgpr32'
+  frameOffsetReg: '$sgpr33'
+  argumentInfo:
+    privateSegmentBuffer: { reg: '$sgpr0_sgpr1_sgpr2_sgpr3' }
+    dispatchPtr:     { reg: '$sgpr4_sgpr5' }
+    kernargSegmentPtr: { reg: '$sgpr6_sgpr7' }
+    workGroupIDX:    { reg: '$sgpr8' }
+    privateSegmentWaveByteOffset: { reg: '$sgpr9' }
+body:             |
+  bb.0:
+    liveins: $sgpr8, $sgpr4_sgpr5, $sgpr6_sgpr7
+  
+    renamable $sgpr12 = IMPLICIT_DEF
+    SI_SPILL_S32_SAVE killed $sgpr12, %stack.0, implicit $exec, implicit $sgpr96_sgpr97_sgpr98_sgpr99, implicit $sgpr32
+    
+    renamable $sgpr12 = IMPLICIT_DEF
+    SI_SPILL_S32_SAVE $sgpr12, %stack.0, implicit $exec, implicit $sgpr96_sgpr97_sgpr98_sgpr99, implicit $sgpr32
+    
+    renamable $sgpr12_sgpr13 = IMPLICIT_DEF
+    SI_SPILL_S64_SAVE killed $sgpr12_sgpr13, %stack.1, implicit $exec, implicit $sgpr96_sgpr97_sgpr98_sgpr99, implicit $sgpr32
+
+    renamable $sgpr12_sgpr13 = IMPLICIT_DEF
+    SI_SPILL_S64_SAVE $sgpr12_sgpr13, %stack.1, implicit $exec, implicit $sgpr96_sgpr97_sgpr98_sgpr99, implicit $sgpr32
+    
+    renamable $sgpr12_sgpr13_sgpr14 = IMPLICIT_DEF
+    SI_SPILL_S96_SAVE killed $sgpr12_sgpr13_sgpr14, %stack.2, implicit $exec, implicit $sgpr96_sgpr97_sgpr98_sgpr99, implicit $sgpr32
+    
+    renamable $sgpr12_sgpr13_sgpr14_sgpr15 = IMPLICIT_DEF
+    SI_SPILL_S128_SAVE killed $sgpr12_sgpr13_sgpr14_sgpr15, %stack.3, implicit $exec, implicit $sgpr96_sgpr97_sgpr98_sgpr99, implicit $sgpr32
+    
+    renamable $sgpr12_sgpr13_sgpr14_sgpr15_sgpr16 = IMPLICIT_DEF
+    SI_SPILL_S160_SAVE killed $sgpr12_sgpr13_sgpr14_sgpr15_sgpr16, %stack.4, implicit $exec, implicit $sgpr96_sgpr97_sgpr98_sgpr99, implicit $sgpr32
+    
+    renamable $sgpr12_sgpr13_sgpr14_sgpr15_sgpr16_sgpr17_sgpr18_sgpr19 = IMPLICIT_DEF
+    SI_SPILL_S256_SAVE killed $sgpr12_sgpr13_sgpr14_sgpr15_sgpr16_sgpr17_sgpr18_sgpr19, %stack.5, implicit $exec, implicit $sgpr96_sgpr97_sgpr98_sgpr99, implicit $sgpr32
+    
+    renamable $sgpr12_sgpr13_sgpr14_sgpr15_sgpr16_sgpr17_sgpr18_sgpr19_sgpr20_sgpr21_sgpr22_sgpr23_sgpr24_sgpr25_sgpr26_sgpr27 = IMPLICIT_DEF
+    SI_SPILL_S512_SAVE killed $sgpr12_sgpr13_sgpr14_sgpr15_sgpr16_sgpr17_sgpr18_sgpr19_sgpr20_sgpr21_sgpr22_sgpr23_sgpr24_sgpr25_sgpr26_sgpr27, %stack.6, implicit $exec, implicit $sgpr96_sgpr97_sgpr98_sgpr99, implicit $sgpr32
+    
+    renamable $sgpr64_sgpr65_sgpr66_sgpr67_sgpr68_sgpr69_sgpr70_sgpr71_sgpr72_sgpr73_sgpr74_sgpr75_sgpr76_sgpr77_sgpr78_sgpr79_sgpr80_sgpr81_sgpr82_sgpr83_sgpr84_sgpr85_sgpr86_sgpr87_sgpr88_sgpr89_sgpr90_sgpr91_sgpr92_sgpr93_sgpr94_sgpr95 = IMPLICIT_DEF
+    SI_SPILL_S1024_SAVE killed $sgpr64_sgpr65_sgpr66_sgpr67_sgpr68_sgpr69_sgpr70_sgpr71_sgpr72_sgpr73_sgpr74_sgpr75_sgpr76_sgpr77_sgpr78_sgpr79_sgpr80_sgpr81_sgpr82_sgpr83_sgpr84_sgpr85_sgpr86_sgpr87_sgpr88_sgpr89_sgpr90_sgpr91_sgpr92_sgpr93_sgpr94_sgpr95, %stack.7, implicit $exec, implicit $sgpr96_sgpr97_sgpr98_sgpr99, implicit $sgpr32
+
+
+# CHECK-LABEL: name: check_reload
+
+# S32
+# CHECK: $sgpr12 = S_MOV_B32 $exec_lo
+# CHECK: $exec_lo = S_MOV_B32 1
+# CHECK: BUFFER_LOAD_DWORD_OFFSET
+# CHECK: $exec_lo = S_MOV_B32 killed $sgpr12
+# CHECK: $sgpr12 = V_READLANE
+
+# S64
+# GCN32: $sgpr12 = S_MOV_B32 $exec_lo
+# GCN64: $sgpr12_sgpr13 = S_MOV_B64 $exec
+# GCN32: $exec_lo = S_MOV_B32 3
+# GCN64: $exec = S_MOV_B64 3
+# CHECK: BUFFER_LOAD_DWORD_OFFSET
+# GCN32: $exec_lo = S_MOV_B32 killed $sgpr12
+# GCN64: $exec = S_MOV_B64 killed $sgpr12_sgpr13
+# CHECK: $sgpr12 = V_READLANE
+# CHECK: $sgpr13 = V_READLANE
+
+# S96
+# GCN32: $sgpr12 = S_MOV_B32 $exec_lo
+# GCN64: $sgpr12_sgpr13 = S_MOV_B64 $exec
+# GCN32: $exec_lo = S_MOV_B32 7
+# GCN64: $exec = S_MOV_B64 7
+# CHECK: BUFFER_LOAD_DWORD_OFFSET
+# GCN32: $exec_lo = S_MOV_B32 killed $sgpr12
+# GCN64: $exec = S_MOV_B64 killed $sgpr12_sgpr13
+# CHECK: $sgpr12 = V_READLANE
+# CHECK: $sgpr13 = V_READLANE
+# CHECK: $sgpr14 = V_READLANE
+
+# S128
+# GCN32: $sgpr12 = S_MOV_B32 $exec_lo
+# GCN64: $sgpr12_sgpr13 = S_MOV_B64 $exec
+# GCN32: $exec_lo = S_MOV_B32 15
+# GCN64: $exec = S_MOV_B64 15
+# CHECK: BUFFER_LOAD_DWORD_OFFSET
+# GCN32: $exec_lo = S_MOV_B32 killed $sgpr12
+# GCN64: $exec = S_MOV_B64 killed $sgpr12_sgpr13
+# CHECK: $sgpr12 = V_READLANE
+# CHECK: $sgpr13 = V_READLANE
+# CHECK: $sgpr14 = V_READLANE
+# CHECK: $sgpr15 = V_READLANE
+
+# S160
+# GCN32: $sgpr12 = S_MOV_B32 $exec_lo
+# GCN64: $sgpr12_sgpr13 = S_MOV_B64 $exec
+# GCN32: $exec_lo = S_MOV_B32 31
+# GCN64: $exec = S_MOV_B64 31
+# CHECK: BUFFER_LOAD_DWORD_OFFSET
+# GCN32: $exec_lo = S_MOV_B32 killed $sgpr12
+# GCN64: $exec = S_MOV_B64 killed $sgpr12_sgpr13
+# CHECK: $sgpr12 = V_READLANE
+# CHECK: $sgpr13 = V_READLANE
+# CHECK: $sgpr14 = V_READLANE
+# CHECK: $sgpr15 = V_READLANE
+# CHECK: $sgpr16 = V_READLANE
+
+# S256
+# GCN32: $sgpr12 = S_MOV_B32 $exec_lo
+# GCN64: $sgpr12_sgpr13 = S_MOV_B64 $exec
+# GCN32: $exec_lo = S_MOV_B32 255
+# GCN64: $exec = S_MOV_B64 255
+# CHECK: BUFFER_LOAD_DWORD_OFFSET
+# GCN32: $exec_lo = S_MOV_B32 killed $sgpr12
+# GCN64: $exec = S_MOV_B64 killed $sgpr12_sgpr13
+# CHECK: $sgpr12 = V_READLANE
+# CHECK: $sgpr13 = V_READLANE
+# CHECK: $sgpr14 = V_READLANE
+# CHECK: $sgpr15 = V_READLANE
+# CHECK: $sgpr16 = V_READLANE
+# CHECK: $sgpr17 = V_READLANE
+# CHECK: $sgpr18 = V_READLANE
+# CHECK: $sgpr19 = V_READLANE
+
+# S512
+# GCN32: $sgpr12 = S_MOV_B32 $exec_lo
+# GCN64: $sgpr12_sgpr13 = S_MOV_B64 $exec
+# GCN32: $exec_lo = S_MOV_B32 65535
+# GCN64: $exec = S_MOV_B64 65535
+# CHECK: BUFFER_LOAD_DWORD_OFFSET
+# GCN32: $exec_lo = S_MOV_B32 killed $sgpr12
+# GCN64: $exec = S_MOV_B64 killed $sgpr12_sgpr13
+# CHECK: $sgpr12 = V_READLANE
+# CHECK: $sgpr13 = V_READLANE
+# CHECK: $sgpr14 = V_READLANE
+# CHECK: $sgpr15 = V_READLANE
+# CHECK: $sgpr16 = V_READLANE
+# CHECK: $sgpr17 = V_READLANE
+# CHECK: $sgpr18 = V_READLANE
+# CHECK: $sgpr19 = V_READLANE
+# CHECK: $sgpr20 = V_READLANE
+# CHECK: $sgpr21 = V_READLANE
+# CHECK: $sgpr22 = V_READLANE
+# CHECK: $sgpr23 = V_READLANE
+# CHECK: $sgpr24 = V_READLANE
+# CHECK: $sgpr25 = V_READLANE
+# CHECK: $sgpr26 = V_READLANE
+# CHECK: $sgpr27 = V_READLANE
+
+# S1024
+# GCN32: $sgpr64 = S_MOV_B32 $exec_lo
+# GCN64: $sgpr64_sgpr65 = S_MOV_B64 $exec
+# GCN32: $exec_lo = S_MOV_B32 65535
+# GCN64: $exec = S_MOV_B64 4294967295
+# CHECK: BUFFER_LOAD_DWORD_OFFSET
+# GCN32: $exec_lo = S_MOV_B32 killed $sgpr64
+# GCN64: $exec = S_MOV_B64 killed $sgpr64_sgpr65
+# CHECK: $sgpr64 = V_READLANE
+# CHECK: $sgpr65 = V_READLANE
+# CHECK: $sgpr66 = V_READLANE
+# CHECK: $sgpr67 = V_READLANE
+# CHECK: $sgpr68 = V_READLANE
+# CHECK: $sgpr69 = V_READLANE
+# CHECK: $sgpr70 = V_READLANE
+# CHECK: $sgpr71 = V_READLANE
+# CHECK: $sgpr72 = V_READLANE
+# CHECK: $sgpr73 = V_READLANE
+# CHECK: $sgpr74 = V_READLANE
+# CHECK: $sgpr75 = V_READLANE
+# CHECK: $sgpr76 = V_READLANE
+# CHECK: $sgpr77 = V_READLANE
+# CHECK: $sgpr78 = V_READLANE
+# CHECK: $sgpr79 = V_READLANE
+# GCN32: $sgpr80 = S_MOV_B32 $exec_lo
+# GCN32: $exec_lo = S_MOV_B32 65535
+# GCN32: BUFFER_LOAD_DWORD_OFFSET
+# GCN32: $exec_lo = S_MOV_B32 killed $sgpr80
+# CHECK: $sgpr80 = V_READLANE
+# CHECK: $sgpr81 = V_READLANE
+# CHECK: $sgpr82 = V_READLANE
+# CHECK: $sgpr83 = V_READLANE
+# CHECK: $sgpr84 = V_READLANE
+# CHECK: $sgpr85 = V_READLANE
+# CHECK: $sgpr86 = V_READLANE
+# CHECK: $sgpr87 = V_READLANE
+# CHECK: $sgpr88 = V_READLANE
+# CHECK: $sgpr89 = V_READLANE
+# CHECK: $sgpr90 = V_READLANE
+# CHECK: $sgpr91 = V_READLANE
+# CHECK: $sgpr92 = V_READLANE
+# CHECK: $sgpr93 = V_READLANE
+# CHECK: $sgpr94 = V_READLANE
+# CHECK: $sgpr95 = V_READLANE
+
+---
+name:            check_reload
+tracksRegLiveness: true
+liveins:         
+  - { reg: '$sgpr4_sgpr5' }
+  - { reg: '$sgpr6_sgpr7' }
+  - { reg: '$sgpr8' }
+frameInfo:       
+  maxAlignment:    4
+stack:           
+  - { id: 0, type: spill-slot, size: 4, alignment: 4 }
+  - { id: 1, type: spill-slot, size: 8, alignment: 4 }
+  - { id: 2, type: spill-slot, size: 12, alignment: 4 }
+  - { id: 3, type: spill-slot, size: 16, alignment: 4 }
+  - { id: 4, type: spill-slot, size: 20, alignment: 4 }
+  - { id: 5, type: spill-slot, size: 32, alignment: 4 }
+  - { id: 6, type: spill-slot, size: 64, alignment: 4 }
+  - { id: 7, type: spill-slot, size: 128, alignment: 4 }
+machineFunctionInfo: 
+  explicitKernArgSize: 660
+  maxKernArgAlign: 4
+  isEntryFunction: true
+  waveLimiter:     true
+  scratchRSrcReg:  '$sgpr96_sgpr97_sgpr98_sgpr99'
+  stackPtrOffsetReg: '$sgpr32'
+  frameOffsetReg: '$sgpr33'
+  argumentInfo:
+    privateSegmentBuffer: { reg: '$sgpr0_sgpr1_sgpr2_sgpr3' }
+    dispatchPtr:     { reg: '$sgpr4_sgpr5' }
+    kernargSegmentPtr: { reg: '$sgpr6_sgpr7' }
+    workGroupIDX:    { reg: '$sgpr8' }
+    privateSegmentWaveByteOffset: { reg: '$sgpr9' }
+body:             |
+  bb.0:
+    liveins: $sgpr8, $sgpr4_sgpr5, $sgpr6_sgpr7
+  
+    renamable $sgpr12 = SI_SPILL_S32_RESTORE %stack.0, implicit $exec, implicit $sgpr96_sgpr97_sgpr98_sgpr99, implicit $sgpr32
+    
+    renamable $sgpr12_sgpr13 = SI_SPILL_S64_RESTORE %stack.1, implicit $exec, implicit $sgpr96_sgpr97_sgpr98_sgpr99, implicit $sgpr32
+
+    renamable $sgpr12_sgpr13_sgpr14 = SI_SPILL_S96_RESTORE %stack.2, implicit $exec, implicit $sgpr96_sgpr97_sgpr98_sgpr99, implicit $sgpr32
+    
+    renamable $sgpr12_sgpr13_sgpr14_sgpr15 = SI_SPILL_S128_RESTORE %stack.3, implicit $exec, implicit $sgpr96_sgpr97_sgpr98_sgpr99, implicit $sgpr32
+    
+    renamable $sgpr12_sgpr13_sgpr14_sgpr15_sgpr16 = SI_SPILL_S160_RESTORE %stack.4, implicit $exec, implicit $sgpr96_sgpr97_sgpr98_sgpr99, implicit $sgpr32
+    
+    renamable $sgpr12_sgpr13_sgpr14_sgpr15_sgpr16_sgpr17_sgpr18_sgpr19 = SI_SPILL_S256_RESTORE %stack.5, implicit $exec, implicit $sgpr96_sgpr97_sgpr98_sgpr99, implicit $sgpr32
+    
+    renamable $sgpr12_sgpr13_sgpr14_sgpr15_sgpr16_sgpr17_sgpr18_sgpr19_sgpr20_sgpr21_sgpr22_sgpr23_sgpr24_sgpr25_sgpr26_sgpr27 = SI_SPILL_S512_RESTORE %stack.6, implicit $exec, implicit $sgpr96_sgpr97_sgpr98_sgpr99, implicit $sgpr32
+    
+    renamable $sgpr64_sgpr65_sgpr66_sgpr67_sgpr68_sgpr69_sgpr70_sgpr71_sgpr72_sgpr73_sgpr74_sgpr75_sgpr76_sgpr77_sgpr78_sgpr79_sgpr80_sgpr81_sgpr82_sgpr83_sgpr84_sgpr85_sgpr86_sgpr87_sgpr88_sgpr89_sgpr90_sgpr91_sgpr92_sgpr93_sgpr94_sgpr95 = SI_SPILL_S1024_RESTORE %stack.7, implicit $exec, implicit $sgpr96_sgpr97_sgpr98_sgpr99, implicit $sgpr32

llvm/test/CodeGen/AMDGPU/si-spill-sgpr-stack.ll

 ; RUN: llc -march=amdgcn -mcpu=fiji -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=ALL -check-prefix=SGPR %s
 
 ; Make sure this doesn't crash.
 ; ALL-LABEL: {{^}}test:
 ; ALL: s_mov_b32 s[[LO:[0-9]+]], SCRATCH_RSRC_DWORD0
 ; ALL: s_mov_b32 s[[HI:[0-9]+]], 0xe80000
 
 ; Make sure we are handling hazards correctly.
-; SGPR: buffer_load_dword [[VHI:v[0-9]+]], off, s[{{[0-9]+:[0-9]+}}], 0 offset:16
+; SGPR: buffer_load_dword [[VHI:v[0-9]+]], off, s[{{[0-9]+:[0-9]+}}], 0 offset:4
+; SGPR-NEXT: s_mov_b64 exec, s[0:1]
 ; SGPR-NEXT: s_waitcnt vmcnt(0)
-; SGPR-NEXT: v_readfirstlane_b32 s[[HI:[0-9]+]], [[VHI]]
+; SGPR-NEXT: v_readlane_b32 s{{[0-9]+}}, [[VHI]], 0
+; SGPR-NEXT: v_readlane_b32 s{{[0-9]+}}, [[VHI]], 1
+; SGPR-NEXT: v_readlane_b32 s{{[0-9]+}}, [[VHI]], 2
+; SGPR-NEXT: v_readlane_b32 s[[HI:[0-9]+]], [[VHI]], 3
 ; SGPR-NEXT: s_nop 4
 ; SGPR-NEXT: buffer_store_dword v0, off, s[0:[[HI]]{{\]}}, 0
 
 ; ALL: s_endpgm
 define amdgpu_kernel void @test(i32 addrspace(1)* %out, i32 %in) {
   call void asm sideeffect "", "~{s[0:7]}" ()
   call void asm sideeffect "", "~{s[8:15]}" ()
   call void asm sideeffect "", "~{s[16:23]}" ()

llvm/test/CodeGen/AMDGPU/spill-m0.ll

 ; RUN: llc -O0 -amdgpu-spill-sgpr-to-vgpr=1 -march=amdgcn -verify-machineinstrs < %s | FileCheck -check-prefix=TOVGPR -check-prefix=GCN %s
 ; RUN: llc -O0 -amdgpu-spill-sgpr-to-vgpr=1 -march=amdgcn -mcpu=tonga  -verify-machineinstrs < %s | FileCheck -check-prefix=TOVGPR -check-prefix=GCN %s
 ; RUN: llc -O0 -amdgpu-spill-sgpr-to-vgpr=0 -march=amdgcn -verify-machineinstrs < %s | FileCheck -check-prefix=TOVMEM -check-prefix=GCN %s
 ; RUN: llc -O0 -amdgpu-spill-sgpr-to-vgpr=0 -march=amdgcn -mcpu=tonga -verify-machineinstrs < %s | FileCheck -check-prefix=TOVMEM -check-prefix=GCN %s
 
 ; XXX - Why does it like to use vcc?
 
 ; GCN-LABEL: {{^}}spill_m0:
 
 ; GCN-DAG: s_cmp_lg_u32
 
 ; TOVGPR-DAG: s_mov_b32 [[M0_COPY:s[0-9]+]], m0
 ; TOVGPR: v_writelane_b32 [[SPILL_VREG:v[0-9]+]], [[M0_COPY]], 2
 
 ; TOVMEM-DAG: s_mov_b32 [[M0_COPY:s[0-9]+]], m0
-; TOVMEM-DAG: v_mov_b32_e32 [[SPILL_VREG:v[0-9]+]], [[M0_COPY]]
+; TOVMEM-DAG: v_writelane_b32 [[SPILL_VREG:v[0-9]+]], [[M0_COPY]], 0
 ; TOVMEM: buffer_store_dword [[SPILL_VREG]], off, s{{\[[0-9]+:[0-9]+\]}}, 0 offset:12 ; 4-byte Folded Spill
 
 ; GCN: s_cbranch_scc1 [[ENDIF:BB[0-9]+_[0-9]+]]
 
 ; GCN: [[ENDIF]]:
 ; TOVGPR: v_readlane_b32 [[M0_RESTORE:s[0-9]+]], [[SPILL_VREG]], 2
 ; TOVGPR: s_mov_b32 m0, [[M0_RESTORE]]
 
 ; TOVMEM: buffer_load_dword [[RELOAD_VREG:v[0-9]+]], off, s{{\[[0-9]+:[0-9]+\]}}, 0 offset:12 ; 4-byte Folded Reload
 ; TOVMEM: s_waitcnt vmcnt(0)
-; TOVMEM: v_readfirstlane_b32 [[M0_RESTORE:s[0-9]+]], [[RELOAD_VREG]]
+; TOVMEM: v_readlane_b32 [[M0_RESTORE:s[0-9]+]], [[RELOAD_VREG]], 0
 ; TOVMEM: s_mov_b32 m0, [[M0_RESTORE]]
 
 ; GCN: s_add_i32 s{{[0-9]+}}, m0, 1
 define amdgpu_kernel void @spill_m0(i32 %cond, i32 addrspace(1)* %out) #0 {
 entry:
   %m0 = call i32 asm sideeffect "s_mov_b32 m0, 0", "={m0}"() #0
   %cmp0 = icmp eq i32 %cond, 0
   br i1 %cmp0, label %if, label %endif

llvm/test/CodeGen/AMDGPU/spill-scavenge-offset.ll

; mark most VGPR registers as used to increase register pressure
 
   %outptr = getelementptr <1280 x i32>, <1280 x i32> addrspace(1)* %out, i32 %tid
   store <1280 x i32> %a, <1280 x i32> addrspace(1)* %outptr
 
   ret void
 }
 
 ; CHECK-LABEL: test_limited_sgpr
-; GFX6: s_add_u32 s32, s32, 0x84100
+; GFX6: s_add_u32 s32, s32, 0x[[OFFSET:[0-9]+]]
 ; GFX6-NEXT: buffer_load_dword v{{[0-9]+}}, off, s[{{[0-9:]+}}], s32
-; GFX6-NEXT: s_sub_u32 s32, s32, 0x84100
+; GFX6-NEXT: s_sub_u32 s32, s32, 0x[[OFFSET:[0-9]+]]
 ; GFX6: NumSgprs: 48
 ; GFX6: ScratchSize: 8624
 define amdgpu_kernel void @test_limited_sgpr(<64 x i32> addrspace(1)* %out, <64 x i32> addrspace(1)* %in) #0 {
 entry:
   %lo = call i32 @llvm.amdgcn.mbcnt.lo(i32 -1, i32 0)
   %tid = call i32 @llvm.amdgcn.mbcnt.hi(i32 -1, i32 %lo)
 
 ; allocate enough scratch to go beyond 2^12 addressing

llvm/test/CodeGen/AMDGPU/spill-wide-sgpr.ll

 ; RUN: llc -O0 -march=amdgcn -mcpu=fiji -verify-machineinstrs < %s | FileCheck -check-prefix=ALL -check-prefix=VGPR %s
 ; RUN: llc -O0 -march=amdgcn -mcpu=fiji -amdgpu-spill-sgpr-to-vgpr=0 -verify-machineinstrs < %s | FileCheck -check-prefix=ALL -check-prefix=VMEM %s
 
 ; ALL-LABEL: {{^}}spill_sgpr_x2:
 
 ; VGPR: v_writelane_b32 v{{[0-9]+}}, s{{[0-9]+}}, 0
 ; VGPR: v_writelane_b32 v{{[0-9]+}}, s{{[0-9]+}}, 1
 ; VGPR: s_cbranch_scc1
 
 ; VGPR: v_readlane_b32 s{{[0-9]+}}, v{{[0-9]+}}, 0
 ; VGPR: v_readlane_b32 s{{[0-9]+}}, v{{[0-9]+}}, 1
 
 
 ; VMEM: buffer_store_dword
-; VMEM: buffer_store_dword
 ; VMEM: s_cbranch_scc1
 
 ; VMEM: buffer_load_dword
-; VMEM: buffer_load_dword
 define amdgpu_kernel void @spill_sgpr_x2(i32 addrspace(1)* %out, i32 %in) #0 {
   %wide.sgpr = call <2 x i32>  asm sideeffect "; def $0", "=s" () #0
   %cmp = icmp eq i32 %in, 0
   br i1 %cmp, label %bb0, label %ret
 
 bb0:
   call void asm sideeffect "; use $0", "s"(<2 x i32> %wide.sgpr) #0
   br label %ret

 ; VGPR: s_cbranch_scc1
 
 ; VGPR: v_readlane_b32 s{{[0-9]+}}, v{{[0-9]+}}, 0
 ; VGPR: v_readlane_b32 s{{[0-9]+}}, v{{[0-9]+}}, 1
 ; VGPR: v_readlane_b32 s{{[0-9]+}}, v{{[0-9]+}}, 2
 
 
 ; VMEM: buffer_store_dword
-; VMEM: buffer_store_dword
-; VMEM: buffer_store_dword
 ; VMEM: s_cbranch_scc1
 
 ; VMEM: buffer_load_dword
-; VMEM: buffer_load_dword
-; VMEM: buffer_load_dword
 define amdgpu_kernel void @spill_sgpr_x3(i32 addrspace(1)* %out, i32 %in) #0 {
   %wide.sgpr = call <3 x i32>  asm sideeffect "; def $0", "=s" () #0
   %cmp = icmp eq i32 %in, 0
   br i1 %cmp, label %bb0, label %ret
 
 bb0:
   call void asm sideeffect "; use $0", "s"(<3 x i32> %wide.sgpr) #0
   br label %ret

 
 ; VGPR: v_readlane_b32 s{{[0-9]+}}, v{{[0-9]+}}, 0
 ; VGPR: v_readlane_b32 s{{[0-9]+}}, v{{[0-9]+}}, 1
 ; VGPR: v_readlane_b32 s{{[0-9]+}}, v{{[0-9]+}}, 2
 ; VGPR: v_readlane_b32 s{{[0-9]+}}, v{{[0-9]+}}, 3
 
 
 ; VMEM: buffer_store_dword
-; VMEM: buffer_store_dword
-; VMEM: buffer_store_dword
-; VMEM: buffer_store_dword
 ; VMEM: s_cbranch_scc1
 
 ; VMEM: buffer_load_dword
-; VMEM: buffer_load_dword
-; VMEM: buffer_load_dword
-; VMEM: buffer_load_dword
 define amdgpu_kernel void @spill_sgpr_x4(i32 addrspace(1)* %out, i32 %in) #0 {
   %wide.sgpr = call <4 x i32>  asm sideeffect "; def $0", "=s" () #0
   %cmp = icmp eq i32 %in, 0
   br i1 %cmp, label %bb0, label %ret
 
 bb0:
   call void asm sideeffect "; use $0", "s"(<4 x i32> %wide.sgpr) #0
   br label %ret

 ; VGPR: v_readlane_b32 s{{[0-9]+}}, v{{[0-9]+}}, 0
 ; VGPR: v_readlane_b32 s{{[0-9]+}}, v{{[0-9]+}}, 1
 ; VGPR: v_readlane_b32 s{{[0-9]+}}, v{{[0-9]+}}, 2
 ; VGPR: v_readlane_b32 s{{[0-9]+}}, v{{[0-9]+}}, 3
 ; VGPR: v_readlane_b32 s{{[0-9]+}}, v{{[0-9]+}}, 4
 
 
 ; VMEM: buffer_store_dword
-; VMEM: buffer_store_dword
-; VMEM: buffer_store_dword
-; VMEM: buffer_store_dword
-; VMEM: buffer_store_dword
 ; VMEM: s_cbranch_scc1
 
 ; VMEM: buffer_load_dword
-; VMEM: buffer_load_dword
-; VMEM: buffer_load_dword
-; VMEM: buffer_load_dword
-; VMEM: buffer_load_dword
 define amdgpu_kernel void @spill_sgpr_x5(i32 addrspace(1)* %out, i32 %in) #0 {
   %wide.sgpr = call <5 x i32>  asm sideeffect "; def $0", "=s" () #0
   %cmp = icmp eq i32 %in, 0
   br i1 %cmp, label %bb0, label %ret
 
 bb0:
   call void asm sideeffect "; use $0", "s"(<5 x i32> %wide.sgpr) #0
   br label %ret

 ; VGPR: v_readlane_b32 s{{[0-9]+}}, v{{[0-9]+}}, 2
 ; VGPR: v_readlane_b32 s{{[0-9]+}}, v{{[0-9]+}}, 3
 ; VGPR: v_readlane_b32 s{{[0-9]+}}, v{{[0-9]+}}, 4
 ; VGPR: v_readlane_b32 s{{[0-9]+}}, v{{[0-9]+}}, 5
 ; VGPR: v_readlane_b32 s{{[0-9]+}}, v{{[0-9]+}}, 6
 ; VGPR: v_readlane_b32 s{{[0-9]+}}, v{{[0-9]+}}, 7
 
 ; VMEM: buffer_store_dword
-; VMEM: buffer_store_dword
-; VMEM: buffer_store_dword
-; VMEM: buffer_store_dword
-; VMEM: buffer_store_dword
-; VMEM: buffer_store_dword
-; VMEM: buffer_store_dword
-; VMEM: buffer_store_dword
 ; VMEM: s_cbranch_scc1
 
 ; VMEM: buffer_load_dword
-; VMEM: buffer_load_dword
-; VMEM: buffer_load_dword
-; VMEM: buffer_load_dword
-; VMEM: buffer_load_dword
-; VMEM: buffer_load_dword
-; VMEM: buffer_load_dword
-; VMEM: buffer_load_dword
 define amdgpu_kernel void @spill_sgpr_x8(i32 addrspace(1)* %out, i32 %in) #0 {
   %wide.sgpr = call <8 x i32>  asm sideeffect "; def $0", "=s" () #0
   %cmp = icmp eq i32 %in, 0
   br i1 %cmp, label %bb0, label %ret
 
 bb0:
   call void asm sideeffect "; use $0", "s"(<8 x i32> %wide.sgpr) #0
   br label %ret