[AMDGPU][MC][GFX8][GFX9] Corrected names of integer v_{add/addc/sub/subrev/subb/subbrev}

[dp]

Summary

Summary of changes in gfx8:

• gfx7's v_add_i32, v_sub_i32, v_subrev_i32 are renamed to v_*_u32.

Summary of changes in gfx9:

• gfx7's v_add_i32, v_sub_i32, v_subrev_i32, v_addc_u32, v_subb_u32, v_subbrev_u32 are renamed to v_*co_u32.
• added gfx9-specific v_add_u32, v_sub_u32, v_subrev_u32, v_add_i32, v_sub_i32.

See bug 34765: https://bugs.llvm.org//show_bug.cgi?id=34765

Note that changes in script-generated tests (gfx*_asm_all.s, etc) are very large. They are not included in the diff because of Differential limitations.

[arsenm]

lib/Target/AMDGPU/SIInstrFormats.td

100 ↗

(On Diff #123036)

Why are you getting rid of this? We need to do something to understand the register def changes (although I don't think I ever committed the patch that uses this). We probably need to introduce new opcodes for the gfx9 variants that preserve the high bits that have this set (which may or may not need this bit)

test/CodeGen/AMDGPU/pack.v2i16.ll

84 ↗

(On Diff #123036)

Since this is a GFX9 specific check, this shouldn't need the regex on the add name.

I have a patch that starts selecting the carry-outless versions that I'll probably post today. The test changes are a subset of these, so it might be slightly easier to commit my patch first.

[dp]

lib/Target/AMDGPU/SIInstrFormats.td

100 ↗

(On Diff #123036)

Actually, F16_ZFILL was introduced to allow mapping of pseudo to MC via a separate table. Currently this bit has no other uses.

I renamed it to indicate its purpose clearly because other GFX9 opcodes also requires special mapping.

Should I preserve F16_ZFILL and add a separate bit renamedInGFX9?

test/CodeGen/AMDGPU/pack.v2i16.ll

84 ↗

(On Diff #123036)

Sure, I'll correct this. Please let me know when you submit your changes.

[rampitec]

lib/Target/AMDGPU/VOP2Instructions.td

717 ↗

(On Diff #123036)

This is not needed anymore.

741 ↗

(On Diff #123036)

This is not needed anymore.

[dp]

Updated and corrected as suggested by Matt and Stas

[rampitec]

LGTM

[artem.tamazov]

Good job.

llvm/trunk/lib/Target/AMDGPU/AMDGPUInstrInfo.cpp

case AMDGPUSubtarget::NORTHERN_ISLANDS:
     return SIEncodingFamily::SI;
   }
 
   llvm_unreachable("Unknown subtarget generation!");
 }
 
 int AMDGPUInstrInfo::pseudoToMCOpcode(int Opcode) const {
   SIEncodingFamily Gen = subtargetEncodingFamily(ST);
-  if (get(Opcode).TSFlags & SIInstrFlags::SDWA)
-    Gen = ST.getGeneration() == AMDGPUSubtarget::GFX9 ? SIEncodingFamily::SDWA9
-                                                      : SIEncodingFamily::SDWA;
 
-  if ((get(Opcode).TSFlags & SIInstrFlags::F16_ZFILL) != 0 &&
-      ST.getGeneration() >= AMDGPUSubtarget::GFX9)
+  if ((get(Opcode).TSFlags & SIInstrFlags::renamedInGFX9) != 0 &&
     Gen = SIEncodingFamily::GFX9;
 
+  if (get(Opcode).TSFlags & SIInstrFlags::SDWA)
+    Gen = ST.getGeneration() == AMDGPUSubtarget::GFX9 ? SIEncodingFamily::SDWA9
+                                                      : SIEncodingFamily::SDWA;
+
   int MCOp = AMDGPU::getMCOpcode(Opcode, Gen);
 
   // -1 means that Opcode is already a native instruction.
   if (MCOp == -1)
     return Opcode;
 
   // (uint16_t)-1 means that Opcode is a pseudo instruction that has
   // no encoding in the given subtarget generation.
   if (MCOp == (uint16_t)-1)
     return -1;
 
   return MCOp;
 }

llvm/trunk/lib/Target/AMDGPU/Disassembler/AMDGPUDisassembler.cpp

do {
     if (Bytes.size() < 4) break;
     const uint32_t DW = eatBytes<uint32_t>(Bytes);
     Res = tryDecodeInst(DecoderTableVI32, MI, DW, Address);
     if (Res) break;
 
     Res = tryDecodeInst(DecoderTableAMDGPU32, MI, DW, Address);
     if (Res) break;
 
+    Res = tryDecodeInst(DecoderTableGFX932, MI, DW, Address);
+    if (Res) break;
+
     if (Bytes.size() < 4) break;
     const uint64_t QW = ((uint64_t)eatBytes<uint32_t>(Bytes) << 32) | DW;
     Res = tryDecodeInst(DecoderTableVI64, MI, QW, Address);
     if (Res) break;
 
     Res = tryDecodeInst(DecoderTableAMDGPU64, MI, QW, Address);
     if (Res) break;
 

llvm/trunk/lib/Target/AMDGPU/SIDefines.h

// TODO: Should this be spilt into VOP3 a and b?
   DisableWQM = UINT64_C(1) << 36,
   Gather4 = UINT64_C(1) << 37,
   SOPK_ZEXT = UINT64_C(1) << 38,
   SCALAR_STORE = UINT64_C(1) << 39,
   FIXED_SIZE = UINT64_C(1) << 40,
   VOPAsmPrefer32Bit = UINT64_C(1) << 41,
   VOP3_OPSEL = UINT64_C(1) << 42,
   maybeAtomic = UINT64_C(1) << 43,
-  F16_ZFILL = UINT64_C(1) << 44,
+  renamedInGFX9 = UINT64_C(1) << 44,
 
   // Is a clamp on FP type.
   FPClamp = UINT64_C(1) << 45,
 
   // Is an integer clamp
   IntClamp = UINT64_C(1) << 46,
 
   // Clamps lo component of register.

llvm/trunk/lib/Target/AMDGPU/SIInstrFormats.td

class InstSI <dag outs, dag ins, string asm = "",
 
   // This bit indicates that this is a VOP3 opcode which supports op_sel
   // modifier (gfx9 only).
   field bit VOP3_OPSEL = 0;
 
   // Is it possible for this instruction to be atomic?
   field bit maybeAtomic = 0;
 
-  // This bit indicates that this is a 16-bit instruction which zero-fills
-  // unused bits in dst. Note that new GFX9 opcodes preserve unused bits.
-  field bit F16_ZFILL = 0;
+  // This bit indicates that this is a VI instruction which is renamed
+  // in GFX9. Required for correct mapping from pseudo to MC.
+  field bit renamedInGFX9 = 0;
 
   // This bit indicates that this has a floating point result type, so
   // the clamp modifier has floating point semantics.
   field bit FPClamp = 0;
 
   // This bit indicates that instruction may support integer clamping
   // which depends on GPU features.
   field bit IntClamp = 0;
class InstSI <dag outs, dag ins, string asm = "",
 
   let TSFlags{38} = SOPKZext;
   let TSFlags{39} = ScalarStore;
   let TSFlags{40} = FixedSize;
   let TSFlags{41} = VOPAsmPrefer32Bit;
   let TSFlags{42} = VOP3_OPSEL;
 
   let TSFlags{43} = maybeAtomic;
-  let TSFlags{44} = F16_ZFILL;
+  let TSFlags{44} = renamedInGFX9;
 
   let TSFlags{45} = FPClamp;
   let TSFlags{46} = IntClamp;
   let TSFlags{47} = ClampLo;
   let TSFlags{48} = ClampHi;
 
   let TSFlags{49} = IsPacked;
 

llvm/trunk/lib/Target/AMDGPU/SIInstructions.td

 
 def : FPMed3Pat<f32, V_MED3_F32>;
 
 let OtherPredicates = [isGFX9] in {
 def : FP16Med3Pat<f16, V_MED3_F16>;
 def : Int16Med3Pat<V_MED3_I16, smax, smax_oneuse, smin_oneuse, i16>;
 def : Int16Med3Pat<V_MED3_U16, umax, umax_oneuse, umin_oneuse, i16>;
 } // End Predicates = [isGFX9]
-
-//============================================================================//
-// Assembler aliases
-//============================================================================//
-
-multiclass NoCarryAlias<string Inst,
-  Instruction Inst32NC, Instruction Inst64NC,
-  Instruction Inst32CO, Instruction Inst64CO> {
-  def : InstAlias<Inst#" $vdst, $src0, $src1",
-    (Inst32NC VGPR_32:$vdst, VSrc_b32:$src0, VGPR_32:$src1), 1000>,
-    Requires<[HasAddNoCarryInsts]>;
-
-  def : InstAlias<Inst#" $vdst, $src0, $src1",
-    (Inst64NC VGPR_32:$vdst, VCSrc_b32:$src0, VCSrc_b32:$src1), -10>,
-    Requires<[HasAddNoCarryInsts]>;
-
-  def : InstAlias<Inst#" $vdst, vcc, $src0, $src1",
-    (Inst32CO VGPR_32:$vdst, VSrc_b32:$src0, VGPR_32:$src1), 1000>,
-    Requires<[HasAddNoCarryInsts]>;
-
-  def : InstAlias<Inst#" $vdst, $sdst, $src0, $src1",
-    (Inst64CO VGPR_32:$vdst, SReg_64:$sdst, VSrc_b32:$src0, VGPR_32:$src1), -10>,
-    Requires<[HasAddNoCarryInsts]>;
-}
-
-// gfx9 made a mess of add instruction names. The existing add
-// instructions add _co added to the names, and their old names were
-// repurposed to a version without carry out.
-// TODO: Do we need SubtargetPredicates for MnemonicAliases?
-let Predicates = [HasAddNoCarryInsts] in {
-defm : NoCarryAlias<"v_add_u32", V_ADD_U32_e32_vi, V_ADD_U32_e64_vi,
-                                 V_ADD_I32_e32_vi, V_ADD_I32_e64_vi>;
-defm : NoCarryAlias<"v_sub_u32", V_SUB_U32_e32_vi, V_SUB_U32_e64_vi,
-                                 V_SUB_I32_e32_vi, V_SUB_I32_e64_vi>;
-defm : NoCarryAlias<"v_subrev_u32",
-  V_SUBREV_U32_e32_vi, V_SUBREV_U32_e64_vi,
-  V_SUBREV_I32_e32_vi, V_SUBREV_I32_e64_vi>;
-}
-
-let Predicates = [NotHasAddNoCarryInsts] in {
-def : MnemonicAlias<"v_add_u32", "v_add_i32">;
-def : MnemonicAlias<"v_sub_u32", "v_sub_i32">;
-def : MnemonicAlias<"v_subrev_u32", "v_subrev_i32">;
-}

llvm/trunk/lib/Target/AMDGPU/VOP2Instructions.td

multiclass VOP2Inst <string opName,
 
   def _sdwa  : VOP2_SDWA_Pseudo <opName, P>;
 }
 
 multiclass VOP2bInst <string opName,
                       VOPProfile P,
                       SDPatternOperator node = null_frag,
                       string revOp = opName,
+                      bit GFX9Renamed = 0,
                       bit useSGPRInput = !eq(P.NumSrcArgs, 3)> {
-
+  let renamedInGFX9 = GFX9Renamed in {
     let SchedRW = [Write32Bit, WriteSALU] in {
       let Uses = !if(useSGPRInput, [VCC, EXEC], [EXEC]), Defs = [VCC] in {
         def _e32 : VOP2_Pseudo <opName, P>,
                    Commutable_REV<revOp#"_e32", !eq(revOp, opName)>;
 
         def _sdwa  : VOP2_SDWA_Pseudo <opName, P> {
           let AsmMatchConverter = "cvtSdwaVOP2b";
         }
       }
 
       def _e64 : VOP3_Pseudo <opName, P, getVOP2Pat64<node, P>.ret>,
                  Commutable_REV<revOp#"_e64", !eq(revOp, opName)>;
     }
   }
+}
 
 multiclass VOP2eInst <string opName,
                       VOPProfile P,
                       SDPatternOperator node = null_frag,
                       string revOp = opName,
                       bit useSGPRInput = !eq(P.NumSrcArgs, 3)> {
 
   let SchedRW = [Write32Bit] in {
def VOP2b_I32_I1_I32_I32_I1 : VOPProfile<[i32, i32, i32, i1]> {
   let Outs64 = (outs DstRC:$vdst, SReg_64:$sdst);
 
   // Suppress src2 implied by type since the 32-bit encoding uses an
   // implicit VCC use.
   let Ins32 = (ins Src0RC32:$src0, Src1RC32:$src1);
 
   let InsSDWA = (ins Src0ModSDWA:$src0_modifiers, Src0SDWA:$src0,
                      Src1ModSDWA:$src1_modifiers, Src1SDWA:$src1,
-                     clampmod:$clamp, omod:$omod,
+                     clampmod:$clamp,
                      dst_sel:$dst_sel, dst_unused:$dst_unused,
                      src0_sel:$src0_sel, src1_sel:$src1_sel);
 
   let InsDPP = (ins DstRCDPP:$old,
-                    Src0Mod:$src0_modifiers, Src0DPP:$src0,
-                    Src1Mod:$src1_modifiers, Src1DPP:$src1,
+                    Src0DPP:$src0,
+                    Src1DPP:$src1,
                     dpp_ctrl:$dpp_ctrl, row_mask:$row_mask,
                     bank_mask:$bank_mask, bound_ctrl:$bound_ctrl);
   let HasExt = 1;
   let HasSDWA9 = 1;
 }
 
 // Read in from vcc or arbitrary SGPR
 def VOP2e_I32_I32_I32_I1 : VOPProfile<[i32, i32, i32, i1]> {

 
 def V_MADAK_F32 : VOP2_Pseudo <"v_madak_f32", VOP_MADAK_F32, [], "">;
 
 // No patterns so that the scalar instructions are always selected.
 // The scalar versions will be replaced with vector when needed later.
 
 // V_ADD_I32, V_SUB_I32, and V_SUBREV_I32 where renamed to *_U32 in VI,
 // but the VI instructions behave the same as the SI versions.
-defm V_ADD_I32 : VOP2bInst <"v_add_i32", VOP2b_I32_I1_I32_I32>;
-defm V_SUB_I32 : VOP2bInst <"v_sub_i32", VOP2b_I32_I1_I32_I32>;
-defm V_SUBREV_I32 : VOP2bInst <"v_subrev_i32", VOP2b_I32_I1_I32_I32, null_frag, "v_sub_i32">;
-defm V_ADDC_U32 : VOP2bInst <"v_addc_u32", VOP2b_I32_I1_I32_I32_I1>;
-defm V_SUBB_U32 : VOP2bInst <"v_subb_u32", VOP2b_I32_I1_I32_I32_I1>;
-defm V_SUBBREV_U32 : VOP2bInst <"v_subbrev_u32", VOP2b_I32_I1_I32_I32_I1, null_frag, "v_subb_u32">;
+defm V_ADD_I32 : VOP2bInst <"v_add_i32", VOP2b_I32_I1_I32_I32, null_frag, "v_add_i32", 1>;
+defm V_SUB_I32 : VOP2bInst <"v_sub_i32", VOP2b_I32_I1_I32_I32, null_frag, "v_sub_i32", 1>;
+defm V_SUBREV_I32 : VOP2bInst <"v_subrev_i32", VOP2b_I32_I1_I32_I32, null_frag, "v_sub_i32", 1>;
+defm V_ADDC_U32 : VOP2bInst <"v_addc_u32", VOP2b_I32_I1_I32_I32_I1, null_frag, "v_addc_u32", 1>;
+defm V_SUBB_U32 : VOP2bInst <"v_subb_u32", VOP2b_I32_I1_I32_I32_I1, null_frag, "v_subb_u32", 1>;
+defm V_SUBBREV_U32 : VOP2bInst <"v_subbrev_u32", VOP2b_I32_I1_I32_I32_I1, null_frag, "v_subb_u32", 1>;
 
 
 let SubtargetPredicate = HasAddNoCarryInsts in {
 defm V_ADD_U32 : VOP2Inst <"v_add_u32", VOP_I32_I32_I32>;
 defm V_SUB_U32 : VOP2Inst <"v_sub_u32", VOP_I32_I32_I32>;
 defm V_SUBREV_U32 : VOP2Inst <"v_subrev_u32", VOP_I32_I32_I32, null_frag, "v_sub_u32">;
 }
 

 defm V_CVT_PK_U16_U32     : VOP2_Real_e32e64_si <0x30>;
 defm V_CVT_PK_I16_I32     : VOP2_Real_e32e64_si <0x31>;
 
 
 //===----------------------------------------------------------------------===//
 // VI
 //===----------------------------------------------------------------------===//
 
-class VOP2_DPP <bits<6> op, VOP2_Pseudo ps, VOPProfile P = ps.Pfl> :
-  VOP_DPP <ps.OpName, P> {
+class VOP2_DPP <bits<6> op, VOP2_Pseudo ps, string OpName = ps.OpName, VOPProfile P = ps.Pfl> :
+  VOP_DPP <OpName, P> {
   let Defs = ps.Defs;
   let Uses = ps.Uses;
   let SchedRW = ps.SchedRW;
   let hasSideEffects = ps.hasSideEffects;
 
   bits<8> vdst;
   bits<8> src1;
   let Inst{8-0}   = 0xfa; //dpp
def _e64_vi :
     VOP3e_vi <op, !cast<VOP3_Pseudo>(NAME#"_e64").Pfl> {
       // Hack to stop printing _e64
       VOP3_Pseudo ps = !cast<VOP3_Pseudo>(NAME#"_e64");
       let OutOperandList = (outs VGPR_32:$vdst);
       let AsmString = ps.Mnemonic # " " # ps.AsmOperands;
     }
 }
 
-multiclass Base_VOP2be_Real_e32e64_vi <bits<6> op> : VOP2_Real_e32_vi<op> {
-  def _e64_vi :
-    VOP3_Real<!cast<VOP3_Pseudo>(NAME#"_e64"), SIEncodingFamily.VI>,
-    VOP3be_vi <{0, 1, 0, 0, op{5-0}}, !cast<VOP3_Pseudo>(NAME#"_e64").Pfl>;
-}
-
 multiclass Base_VOP2_Real_e32e64_vi <bits<6> op> :
   VOP2_Real_e32_vi<op>,
   VOP2_Real_e64_vi<{0, 1, 0, 0, op{5-0}}>;
 
 } // End AssemblerPredicates = [isVI], DecoderNamespace = "VI"
 
 multiclass VOP2_SDWA_Real <bits<6> op> {
   def _sdwa_vi :
     VOP_SDWA_Real <!cast<VOP2_SDWA_Pseudo>(NAME#"_sdwa")>,
     VOP2_SDWAe <op{5-0}, !cast<VOP2_SDWA_Pseudo>(NAME#"_sdwa").Pfl>;
 }
 
 multiclass VOP2_SDWA9_Real <bits<6> op> {
   def _sdwa_gfx9 :
     VOP_SDWA9_Real <!cast<VOP2_SDWA_Pseudo>(NAME#"_sdwa")>,
     VOP2_SDWA9Ae <op{5-0}, !cast<VOP2_SDWA_Pseudo>(NAME#"_sdwa").Pfl>;
 }
 
-multiclass VOP2be_Real_e32e64_vi <bits<6> op> :
-  Base_VOP2be_Real_e32e64_vi<op>, VOP2_SDWA_Real<op>, VOP2_SDWA9_Real<op> {
-  // For now left dpp only for asm/dasm
-  // TODO: add corresponding pseudo
-  def _dpp : VOP2_DPP<op, !cast<VOP2_Pseudo>(NAME#"_e32")>;
+let AssemblerPredicates = [isVIOnly] in {
+
+multiclass VOP2be_Real_e32e64_vi_only <bits<6> op, string OpName, string AsmName> {
+  def _e32_vi :
+    VOP2_Real<!cast<VOP2_Pseudo>(OpName#"_e32"), SIEncodingFamily.VI>,
+    VOP2e<op{5-0}, !cast<VOP2_Pseudo>(OpName#"_e32").Pfl> {
+      VOP2_Pseudo ps = !cast<VOP2_Pseudo>(OpName#"_e32");
+      let AsmString = AsmName # ps.AsmOperands;
+      let DecoderNamespace = "VI";
+    }
+  def _e64_vi :
+    VOP3_Real<!cast<VOP3_Pseudo>(OpName#"_e64"), SIEncodingFamily.VI>,
+    VOP3be_vi <{0, 1, 0, 0, op{5-0}}, !cast<VOP3_Pseudo>(OpName#"_e64").Pfl> {
+      VOP3_Pseudo ps = !cast<VOP3_Pseudo>(OpName#"_e64");
+      let AsmString = AsmName # ps.AsmOperands;
+      let DecoderNamespace = "VI";
+    }
+  def _sdwa_vi :
+    VOP_SDWA_Real <!cast<VOP2_SDWA_Pseudo>(OpName#"_sdwa")>,
+    VOP2_SDWAe <op{5-0}, !cast<VOP2_SDWA_Pseudo>(OpName#"_sdwa").Pfl> {
+      VOP2_SDWA_Pseudo ps = !cast<VOP2_SDWA_Pseudo>(OpName#"_sdwa");
+      let AsmString = AsmName # ps.AsmOperands;
+    }
+  def _dpp :
+    VOP2_DPP<op, !cast<VOP2_Pseudo>(OpName#"_e32"), AsmName>;
+}
 }
 
+let AssemblerPredicates = [isGFX9] in {
+
+multiclass VOP2be_Real_e32e64_gfx9 <bits<6> op, string OpName, string AsmName> {
+  def _e32_gfx9 :
+    VOP2_Real<!cast<VOP2_Pseudo>(OpName#"_e32"), SIEncodingFamily.GFX9>,
+    VOP2e<op{5-0}, !cast<VOP2_Pseudo>(OpName#"_e32").Pfl> {
+      VOP2_Pseudo ps = !cast<VOP2_Pseudo>(OpName#"_e32");
+      let AsmString = AsmName # ps.AsmOperands;
+      let DecoderNamespace = "GFX9";
+    }
+  def _e64_gfx9 :
+    VOP3_Real<!cast<VOP3_Pseudo>(OpName#"_e64"), SIEncodingFamily.GFX9>,
+    VOP3be_vi <{0, 1, 0, 0, op{5-0}}, !cast<VOP3_Pseudo>(OpName#"_e64").Pfl> {
+      VOP3_Pseudo ps = !cast<VOP3_Pseudo>(OpName#"_e64");
+      let AsmString = AsmName # ps.AsmOperands;
+      let DecoderNamespace = "GFX9";
+    }
+  def _sdwa_gfx9 :
+    VOP_SDWA9_Real <!cast<VOP2_SDWA_Pseudo>(OpName#"_sdwa")>,
+    VOP2_SDWA9Ae <op{5-0}, !cast<VOP2_SDWA_Pseudo>(OpName#"_sdwa").Pfl> {
+      VOP2_SDWA_Pseudo ps = !cast<VOP2_SDWA_Pseudo>(OpName#"_sdwa");
+      let AsmString = AsmName # ps.AsmOperands;
+    }
+  def _dpp_gfx9 :
+    VOP2_DPP<op, !cast<VOP2_Pseudo>(OpName#"_e32"), AsmName> {
+      let DecoderNamespace = "SDWA9";
+    }
+}
+
+multiclass VOP2_Real_e32e64_gfx9 <bits<6> op> {
+  def _e32_gfx9 :
+    VOP2_Real<!cast<VOP2_Pseudo>(NAME#"_e32"), SIEncodingFamily.GFX9>,
+    VOP2e<op{5-0}, !cast<VOP2_Pseudo>(NAME#"_e32").Pfl>{
+      let DecoderNamespace = "GFX9";
+    }
+  def _e64_gfx9 :
+    VOP3_Real<!cast<VOP3_Pseudo>(NAME#"_e64"), SIEncodingFamily.GFX9>,
+    VOP3e_vi <{0, 1, 0, 0, op{5-0}}, !cast<VOP3_Pseudo>(NAME#"_e64").Pfl> {
+      let DecoderNamespace = "GFX9";
+    }
+  def _sdwa_gfx9 :
+    VOP_SDWA9_Real <!cast<VOP2_SDWA_Pseudo>(NAME#"_sdwa")>,
+    VOP2_SDWA9Ae <op{5-0}, !cast<VOP2_SDWA_Pseudo>(NAME#"_sdwa").Pfl> {
+    }
+  def _dpp_gfx9 :
+    VOP2_DPP<op, !cast<VOP2_Pseudo>(NAME#"_e32")> {
+      let DecoderNamespace = "SDWA9";
+    }
+}
+
+} // AssemblerPredicates = [isGFX9]
+
 multiclass VOP2_Real_e32e64_vi <bits<6> op> :
   Base_VOP2_Real_e32e64_vi<op>, VOP2_SDWA_Real<op>, VOP2_SDWA9_Real<op> {
   // For now left dpp only for asm/dasm
   // TODO: add corresponding pseudo
   def _dpp : VOP2_DPP<op, !cast<VOP2_Pseudo>(NAME#"_e32")>;
 }
 
 defm V_CNDMASK_B32        : Base_VOP2_Real_e32e64_vi <0x0>;

 defm V_ASHRREV_I32        : VOP2_Real_e32e64_vi <0x11>;
 defm V_LSHLREV_B32        : VOP2_Real_e32e64_vi <0x12>;
 defm V_AND_B32            : VOP2_Real_e32e64_vi <0x13>;
 defm V_OR_B32             : VOP2_Real_e32e64_vi <0x14>;
 defm V_XOR_B32            : VOP2_Real_e32e64_vi <0x15>;
 defm V_MAC_F32            : VOP2_Real_e32e64_vi <0x16>;
 defm V_MADMK_F32          : VOP2_Real_MADK_vi <0x17>;
 defm V_MADAK_F32          : VOP2_Real_MADK_vi <0x18>;
-defm V_ADD_I32            : VOP2be_Real_e32e64_vi <0x19>;
-defm V_SUB_I32            : VOP2be_Real_e32e64_vi <0x1a>;
-defm V_SUBREV_I32         : VOP2be_Real_e32e64_vi <0x1b>;
-defm V_ADDC_U32           : VOP2be_Real_e32e64_vi <0x1c>;
-defm V_SUBB_U32           : VOP2be_Real_e32e64_vi <0x1d>;
-defm V_SUBBREV_U32        : VOP2be_Real_e32e64_vi <0x1e>;
+
+defm V_ADD_U32            : VOP2be_Real_e32e64_vi_only <0x19, "V_ADD_I32",     "v_add_u32">;
+defm V_SUB_U32            : VOP2be_Real_e32e64_vi_only <0x1a, "V_SUB_I32",     "v_sub_u32">;
+defm V_SUBREV_U32         : VOP2be_Real_e32e64_vi_only <0x1b, "V_SUBREV_I32",  "v_subrev_u32">;
+defm V_ADDC_U32           : VOP2be_Real_e32e64_vi_only <0x1c, "V_ADDC_U32",    "v_addc_u32">;
+defm V_SUBB_U32           : VOP2be_Real_e32e64_vi_only <0x1d, "V_SUBB_U32",    "v_subb_u32">;
+defm V_SUBBREV_U32        : VOP2be_Real_e32e64_vi_only <0x1e, "V_SUBBREV_U32", "v_subbrev_u32">;
+
+defm V_ADD_CO_U32         : VOP2be_Real_e32e64_gfx9 <0x19, "V_ADD_I32",     "v_add_co_u32">;
+defm V_SUB_CO_U32         : VOP2be_Real_e32e64_gfx9 <0x1a, "V_SUB_I32",     "v_sub_co_u32">;
+defm V_SUBREV_CO_U32      : VOP2be_Real_e32e64_gfx9 <0x1b, "V_SUBREV_I32",  "v_subrev_co_u32">;
+defm V_ADDC_CO_U32        : VOP2be_Real_e32e64_gfx9 <0x1c, "V_ADDC_U32",    "v_addc_co_u32">;
+defm V_SUBB_CO_U32        : VOP2be_Real_e32e64_gfx9 <0x1d, "V_SUBB_U32",    "v_subb_co_u32">;
+defm V_SUBBREV_CO_U32     : VOP2be_Real_e32e64_gfx9 <0x1e, "V_SUBBREV_U32", "v_subbrev_co_u32">;
+
+defm V_ADD_U32            : VOP2_Real_e32e64_gfx9 <0x34>;
+defm V_SUB_U32            : VOP2_Real_e32e64_gfx9 <0x35>;
+defm V_SUBREV_U32         : VOP2_Real_e32e64_gfx9 <0x36>;
 
 defm V_READLANE_B32       : VOP32_Real_vi <0x289>;
 defm V_WRITELANE_B32      : VOP32_Real_vi <0x28a>;
 
 defm V_BFM_B32            : VOP2_Real_e64only_vi <0x293>;
 defm V_BCNT_U32_B32       : VOP2_Real_e64only_vi <0x28b>;
 defm V_MBCNT_LO_U32_B32   : VOP2_Real_e64only_vi <0x28c>;
 defm V_MBCNT_HI_U32_B32   : VOP2_Real_e64only_vi <0x28d>;

 
 def : SI2_VI3Alias <"v_ldexp_f32", V_LDEXP_F32_e64_vi>;
 def : SI2_VI3Alias <"v_cvt_pkaccum_u8_f32", V_CVT_PKACCUM_U8_F32_e64_vi>;
 def : SI2_VI3Alias <"v_cvt_pknorm_i16_f32", V_CVT_PKNORM_I16_F32_e64_vi>;
 def : SI2_VI3Alias <"v_cvt_pknorm_u16_f32", V_CVT_PKNORM_U16_F32_e64_vi>;
 def : SI2_VI3Alias <"v_cvt_pkrtz_f16_f32", V_CVT_PKRTZ_F16_F32_e64_vi>;
 
 } // End SubtargetPredicate = isVI
-
-let SubtargetPredicate = HasAddNoCarryInsts in {
-defm V_ADD_U32 : VOP2_Real_e32e64_vi <0x34>;
-defm V_SUB_U32 : VOP2_Real_e32e64_vi <0x35>;
-defm V_SUBREV_U32 : VOP2_Real_e32e64_vi <0x36>;
-}

llvm/trunk/lib/Target/AMDGPU/VOP3Instructions.td

 } // End SchedRW = [WriteDouble, WriteSALU]
 } // End isCommutable = 1
 
 } // End SubtargetPredicate = isCIVI
 
 
 let SubtargetPredicate = Has16BitInsts in {
 
-let F16_ZFILL = 1 in {
+let renamedInGFX9 = 1 in {
 def V_DIV_FIXUP_F16 : VOP3Inst <"v_div_fixup_f16", VOP3_Profile<VOP_F16_F16_F16_F16>, AMDGPUdiv_fixup>;
 }
 let SubtargetPredicate = isGFX9 in {
 def V_DIV_FIXUP_F16_gfx9 : VOP3Inst <"v_div_fixup_f16_gfx9", VOP3_Profile<VOP_F16_F16_F16_F16, VOP3_OPSEL>>;
 }
 
 let isCommutable = 1 in {
 
-let F16_ZFILL = 1 in {
+let renamedInGFX9 = 1 in {
 def V_MAD_F16 : VOP3Inst <"v_mad_f16", VOP3_Profile<VOP_F16_F16_F16_F16>, fmad>;
 def V_MAD_U16 : VOP3Inst <"v_mad_u16", VOP3_Profile<VOP_I16_I16_I16_I16, VOP3_CLAMP>>;
 def V_MAD_I16 : VOP3Inst <"v_mad_i16", VOP3_Profile<VOP_I16_I16_I16_I16, VOP3_CLAMP>>;
 def V_FMA_F16 : VOP3Inst <"v_fma_f16", VOP3_Profile<VOP_F16_F16_F16_F16>, fma>;
 }
 
 let SubtargetPredicate = isGFX9 in {
 def V_MAD_F16_gfx9   : VOP3Inst <"v_mad_f16_gfx9", VOP3_Profile<VOP_F16_F16_F16_F16, VOP3_OPSEL>>;

 def V_ADD_I16 : VOP3Inst <"v_add_i16", VOP3_Profile<VOP_I16_I16_I16, VOP3_OPSEL>>;
 def V_SUB_I16 : VOP3Inst <"v_sub_i16", VOP3_Profile<VOP_I16_I16_I16, VOP3_OPSEL>>;
 
 def V_MAD_U32_U16 : VOP3Inst <"v_mad_u32_u16", VOP3_Profile<VOP_I32_I16_I16_I32, VOP3_OPSEL>>;
 def V_MAD_I32_I16 : VOP3Inst <"v_mad_i32_i16", VOP3_Profile<VOP_I32_I16_I16_I32, VOP3_OPSEL>>;
 
 def V_CVT_PKNORM_I16_F16 : VOP3Inst <"v_cvt_pknorm_i16_f16", VOP3_Profile<VOP_B32_F16_F16, VOP3_OPSEL>>;
 def V_CVT_PKNORM_U16_F16 : VOP3Inst <"v_cvt_pknorm_u16_f16", VOP3_Profile<VOP_B32_F16_F16, VOP3_OPSEL>>;
+
+def V_ADD_I32_gfx9 : VOP3Inst <"v_add_i32_gfx9", VOP3_Profile<VOP_I32_I32_I32>>;
+def V_SUB_I32_gfx9 : VOP3Inst <"v_sub_i32_gfx9", VOP3_Profile<VOP_I32_I32_I32>>;
 } // End SubtargetPredicate = isGFX9
 
 //===----------------------------------------------------------------------===//
 // Integer Clamp Patterns
 //===----------------------------------------------------------------------===//
 
 class getClampPat<VOPProfile P, SDPatternOperator node> {
   dag ret3 = (P.DstVT (node P.Src0VT:$src0, P.Src1VT:$src1, P.Src2VT:$src2));

 multiclass VOP3OpSel_F16_Real_gfx9<bits<10> op, string AsmName> {
   def _vi : VOP3_Real<!cast<VOP3_Pseudo>(NAME), SIEncodingFamily.GFX9>,
             VOP3OpSel_gfx9 <op, !cast<VOP3_Pseudo>(NAME).Pfl> {
               VOP3_Pseudo ps = !cast<VOP3_Pseudo>(NAME);
               let AsmString = AsmName # ps.AsmOperands;
             }
 }
 
+multiclass VOP3_Real_gfx9<bits<10> op, string AsmName> {
+  def _vi : VOP3_Real<!cast<VOP3_Pseudo>(NAME), SIEncodingFamily.GFX9>,
+              VOP3e_vi <op, !cast<VOP3_Pseudo>(NAME).Pfl> {
+              VOP3_Pseudo ps = !cast<VOP3_Pseudo>(NAME);
+              let AsmString = AsmName # ps.AsmOperands;
+            }
+}
+
 } // End AssemblerPredicates = [isGFX9], DecoderNamespace = "GFX9"
 
 defm V_MAD_U64_U32      : VOP3be_Real_vi <0x1E8>;
 defm V_MAD_I64_I32      : VOP3be_Real_vi <0x1E9>;
 
 defm V_MAD_LEGACY_F32   : VOP3_Real_vi <0x1c0>;
 defm V_MAD_F32          : VOP3_Real_vi <0x1c1>;
 defm V_MAD_I32_I24      : VOP3_Real_vi <0x1c2>;

 defm V_DIV_FIXUP_LEGACY_F16 : VOP3_F16_Real_gfx9 <0x1ef, "V_DIV_FIXUP_F16", "v_div_fixup_legacy_f16">;
 
 defm V_MAD_F16_gfx9         : VOP3OpSel_F16_Real_gfx9 <0x203, "v_mad_f16">;
 defm V_MAD_U16_gfx9         : VOP3OpSel_F16_Real_gfx9 <0x204, "v_mad_u16">;
 defm V_MAD_I16_gfx9         : VOP3OpSel_F16_Real_gfx9 <0x205, "v_mad_i16">;
 defm V_FMA_F16_gfx9         : VOP3OpSel_F16_Real_gfx9 <0x206, "v_fma_f16">;
 defm V_DIV_FIXUP_F16_gfx9   : VOP3OpSel_F16_Real_gfx9 <0x207, "v_div_fixup_f16">;
 
+defm V_ADD_I32_gfx9         : VOP3_Real_gfx9 <0x29c, "v_add_i32">;
+defm V_SUB_I32_gfx9         : VOP3_Real_gfx9 <0x29d, "v_sub_i32">;
+
 defm V_INTERP_P1_F32_e64  : VOP3Interp_Real_vi <0x270>;
 defm V_INTERP_P2_F32_e64  : VOP3Interp_Real_vi <0x271>;
 defm V_INTERP_MOV_F32_e64 : VOP3Interp_Real_vi <0x272>;
 
 defm V_INTERP_P1LL_F16  : VOP3Interp_Real_vi <0x274>;
 defm V_INTERP_P1LV_F16  : VOP3Interp_Real_vi <0x275>;
 defm V_INTERP_P2_F16    : VOP3Interp_Real_vi <0x276>;
 defm V_ADD_F64          : VOP3_Real_vi <0x280>;

llvm/trunk/test/CodeGen/AMDGPU/add.v2i16.ll

define amdgpu_kernel void @s_test_add_self_v2i16(<2 x i16> addrspace(1)* %out, <2 x i16> addrspace(2)* %in0) #1 {
   store <2 x i16> %add, <2 x i16> addrspace(1)* %out
   ret void
 }
 
 ; FIXME: VI should not scalarize arg access.
 ; GCN-LABEL: {{^}}s_test_add_v2i16_kernarg:
 ; GFX9: v_pk_add_u16 v{{[0-9]+}}, s{{[0-9]+}}, v{{[0-9]+}}
 
-; VI: v_add_i32
-; VI: v_add_i32_sdwa
+; VI: v_add_u32
+; VI: v_add_u32_sdwa
 define amdgpu_kernel void @s_test_add_v2i16_kernarg(<2 x i16> addrspace(1)* %out, <2 x i16> %a, <2 x i16> %b) #1 {
   %add = add <2 x i16> %a, %b
   store <2 x i16> %add, <2 x i16> addrspace(1)* %out
   ret void
 }
 
 ; GCN-LABEL: {{^}}v_test_add_v2i16_constant:
 ; GFX9: s_mov_b32 [[CONST:s[0-9]+]], 0x1c8007b{{$}}

llvm/trunk/test/CodeGen/AMDGPU/amdgcn.private-memory.ll

 ; RUN: llc -mattr=+promote-alloca -verify-machineinstrs -march=amdgcn < %s | FileCheck -check-prefix=GCN -check-prefix=GCN-PROMOTE %s
 ; RUN: llc -mattr=+promote-alloca,-flat-for-global -verify-machineinstrs -mtriple=amdgcn--amdhsa -mcpu=kaveri < %s | FileCheck -check-prefix=GCN -check-prefix=GCN-PROMOTE -check-prefix=HSA %s
 ; RUN: llc -mattr=-promote-alloca -verify-machineinstrs -march=amdgcn < %s | FileCheck -check-prefix=GCN -check-prefix=GCN-ALLOCA %s
 ; RUN: llc -mattr=-promote-alloca,-flat-for-global -verify-machineinstrs -mtriple=amdgcn-amdhsa -mcpu=kaveri < %s | FileCheck  -check-prefix=GCN -check-prefix=GCN-ALLOCA -check-prefix=HSA %s
 ; RUN: llc -mattr=+promote-alloca -verify-machineinstrs -march=amdgcn -mcpu=tonga -mattr=-flat-for-global < %s | FileCheck -check-prefix=GCN -check-prefix=GCN-PROMOTE %s
 ; RUN: llc -mattr=-promote-alloca -verify-machineinstrs -march=amdgcn -mcpu=tonga -mattr=-flat-for-global < %s | FileCheck -check-prefix=GCN  -check-prefix=GCN-ALLOCA %s
 
 
 declare i32 @llvm.amdgcn.workitem.id.x() nounwind readnone
 
 
 ; Make sure we don't overwrite workitem information with private memory
 
 ; GCN-LABEL: {{^}}work_item_info:
 ; GCN-NOT: v0
-; GCN: v_add_i32_e32 [[RESULT:v[0-9]+]], vcc, v0, v{{[0-9]+}}
+; GCN: v_add_{{[iu]}}32_e32 [[RESULT:v[0-9]+]], vcc, v0, v{{[0-9]+}}
 ; GCN: buffer_store_dword [[RESULT]]
 define amdgpu_kernel void @work_item_info(i32 addrspace(1)* %out, i32 %in) {
 entry:
   %0 = alloca [2 x i32]
   %1 = getelementptr [2 x i32], [2 x i32]* %0, i32 0, i32 0
   %2 = getelementptr [2 x i32], [2 x i32]* %0, i32 0, i32 1
   store i32 0, i32* %1
   store i32 1, i32* %2
   %3 = getelementptr [2 x i32], [2 x i32]* %0, i32 0, i32 %in
   %4 = load i32, i32* %3
   %5 = call i32 @llvm.amdgcn.workitem.id.x()
   %6 = add i32 %4, %5
   store i32 %6, i32 addrspace(1)* %out
   ret void
 }

llvm/trunk/test/CodeGen/AMDGPU/amdgpu-shader-calling-convention.ll

 ; RUN: llc -march=amdgcn -mcpu=tahiti -verify-machineinstrs < %s | FileCheck -check-prefix=GCN %s
 ; RUN: llc -march=amdgcn -mcpu=tonga -verify-machineinstrs < %s | FileCheck -check-prefix=GCN %s
 
 
 ; GCN-LABEL: {{^}}shader_cc:
-; GCN: v_add_i32_e32 v0, vcc, s8, v0
+; GCN: v_add_{{[iu]}}32_e32 v0, vcc, s8, v0
 define amdgpu_cs float @shader_cc(<4 x i32> inreg, <4 x i32> inreg, i32 inreg %w, float %v) {
   %vi = bitcast float %v to i32
   %x = add i32 %vi, %w
   %xf = bitcast i32 %x to float
   ret float %xf
 }
 
 ; GCN-LABEL: {{^}}kernel_cc:
 ; GCN: s_endpgm
 define amdgpu_kernel void @kernel_cc(<4 x i32> inreg, <4 x i32> inreg, i32 inreg %w, float %v) {
   %vi = bitcast float %v to i32
   %x = add i32 %vi, %w
   %xf = bitcast i32 %x to float
   store float %xf, float addrspace(1)* undef
   ret void
 }

llvm/trunk/test/CodeGen/AMDGPU/amdgpu.private-memory.ll

 }
 
 ; AMDGPUPromoteAlloca does not know how to handle ptrtoint.  When it
 ; finds one, it should stop trying to promote.
 
 ; FUNC-LABEL: ptrtoint:
 ; SI-NOT: ds_write
 ; SI: buffer_store_dword v{{[0-9]+}}, v{{[0-9]+}}, s[{{[0-9]+:[0-9]+}}], s{{[0-9]+}} offen
-; SI: v_add_i32_e32 [[ADD_OFFSET:v[0-9]+]], vcc, 5,
+; SI: v_add_{{[iu]}}32_e32 [[ADD_OFFSET:v[0-9]+]], vcc, 5,
 ; SI: buffer_load_dword v{{[0-9]+}}, [[ADD_OFFSET:v[0-9]+]], s[{{[0-9]+:[0-9]+}}], s{{[0-9]+}} offen ;
 define amdgpu_kernel void @ptrtoint(i32 addrspace(1)* %out, i32 %a, i32 %b) #0 {
   %alloca = alloca [16 x i32]
   %tmp0 = getelementptr [16 x i32], [16 x i32]* %alloca, i32 0, i32 %a
   store i32 5, i32* %tmp0
   %tmp1 = ptrtoint [16 x i32]* %alloca to i32
   %tmp2 = add i32 %tmp1, 5
   %tmp3 = inttoptr i32 %tmp2 to i32*

llvm/trunk/test/CodeGen/AMDGPU/bfe-combine.ll

 ; RUN: llc -march=amdgcn -mcpu=fiji -amdgpu-sdwa-peephole=0 < %s | FileCheck --check-prefix=GCN --check-prefix=VI %s
 ; RUN: llc -march=amdgcn -mcpu=fiji < %s | FileCheck --check-prefix=GCN --check-prefix=VI-SDWA %s
 ; RUN: llc -march=amdgcn -mcpu=bonaire < %s | FileCheck --check-prefix=GCN --check-prefix=CI %s
 
 ; GCN-LABEL: {{^}}bfe_combine8:
 ; VI: v_bfe_u32 v[[BFE:[0-9]+]], v{{[0-9]+}}, 8, 8
 ; VI: v_lshlrev_b32_e32 v[[ADDRBASE:[0-9]+]], 2, v[[BFE]]
 ; VI-SDWA: v_mov_b32_e32 v[[SHIFT:[0-9]+]], 2
 ; VI-SDWA: v_lshlrev_b32_sdwa v[[ADDRBASE:[0-9]+]], v[[SHIFT]], v{{[0-9]+}} dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:DWORD src1_sel:BYTE_1
 ; CI: v_lshrrev_b32_e32 v[[SHR:[0-9]+]], 6, v{{[0-9]+}}
 ; CI: v_and_b32_e32 v[[ADDRLO:[0-9]+]], 0x3fc, v[[SHR]]
-; VI: v_add_i32_e32 v[[ADDRLO:[0-9]+]], vcc, s{{[0-9]+}}, v[[ADDRBASE]]
-; VI-SDWA: v_add_i32_e32 v[[ADDRLO:[0-9]+]], vcc, s{{[0-9]+}}, v[[ADDRBASE]]
+; VI: v_add_u32_e32 v[[ADDRLO:[0-9]+]], vcc, s{{[0-9]+}}, v[[ADDRBASE]]
+; VI-SDWA: v_add_u32_e32 v[[ADDRLO:[0-9]+]], vcc, s{{[0-9]+}}, v[[ADDRBASE]]
 ; GCN: load_dword v{{[0-9]+}}, v{{\[}}[[ADDRLO]]:
 define amdgpu_kernel void @bfe_combine8(i32 addrspace(1)* nocapture %arg, i32 %x) {
   %id = tail call i32 @llvm.amdgcn.workitem.id.x() #2
   %idx = add i32 %x, %id
   %srl = lshr i32 %idx, 8
   %and = and i32 %srl, 255
   %ptr = getelementptr inbounds i32, i32 addrspace(1)* %arg, i32 %and
   %val = load i32, i32 addrspace(1)* %ptr, align 4
   store i32 %val, i32 addrspace(1)* %arg, align 4
   ret void
 }
 
 ; GCN-LABEL: {{^}}bfe_combine16:
 ; VI: v_bfe_u32 v[[BFE:[0-9]+]], v{{[0-9]+}}, 16, 16
 ; VI: v_lshlrev_b32_e32 v[[ADDRBASE:[0-9]+]], {{[^,]+}}, v[[BFE]]
 ; VI-SDWA: v_mov_b32_e32 v[[SHIFT:[0-9]+]], 15
 ; VI-SDWA: v_lshlrev_b32_sdwa v[[ADDRBASE1:[0-9]+]], v[[SHIFT]], v{{[0-9]+}} dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:DWORD src1_sel:WORD_1
 ; VI-SDWA: v_lshlrev_b64 v{{\[}}[[ADDRBASE:[0-9]+]]:{{[^\]+}}], 2, v{{\[}}[[ADDRBASE1]]:{{[^\]+}}]
-; VI-SDWA: v_add_i32_e32 v[[ADDRLO:[0-9]+]], vcc, s{{[0-9]+}}, v[[ADDRBASE]]
+; VI-SDWA: v_add_u32_e32 v[[ADDRLO:[0-9]+]], vcc, s{{[0-9]+}}, v[[ADDRBASE]]
 ; CI: v_lshrrev_b32_e32 v[[SHR:[0-9]+]], 1, v{{[0-9]+}}
 ; CI: v_and_b32_e32 v[[AND:[0-9]+]], 0x7fff8000, v[[SHR]]
 ; CI: v_lshl_b64 v{{\[}}[[ADDRLO:[0-9]+]]:{{[^\]+}}], v{{\[}}[[AND]]:{{[^\]+}}], 2
-; VI: v_add_i32_e32 v[[ADDRLO:[0-9]+]], vcc, s{{[0-9]+}}, v[[ADDRBASE]]
+; VI: v_add_u32_e32 v[[ADDRLO:[0-9]+]], vcc, s{{[0-9]+}}, v[[ADDRBASE]]
 ; GCN: load_dword v{{[0-9]+}}, v{{\[}}[[ADDRLO]]:
 define amdgpu_kernel void @bfe_combine16(i32 addrspace(1)* nocapture %arg, i32 %x) {
   %id = tail call i32 @llvm.amdgcn.workitem.id.x() #2
   %idx = add i32 %x, %id
   %srl = lshr i32 %idx, 1
   %and = and i32 %srl, 2147450880
   %ptr = getelementptr inbounds i32, i32 addrspace(1)* %arg, i32 %and
   %val = load i32, i32 addrspace(1)* %ptr, align 4
   store i32 %val, i32 addrspace(1)* %arg, align 4
   ret void
 }
 
 declare i32 @llvm.amdgcn.workitem.id.x() #1

llvm/trunk/test/CodeGen/AMDGPU/bfe-patterns.ll

define amdgpu_kernel void @v_ubfe_sub_i32(i32 addrspace(1)* %out, i32 addrspace(1)* %in0, i32 addrspace(1)* %in1) #1 {
   %bfe = lshr i32 %shl, %sub
   store i32 %bfe, i32 addrspace(1)* %out.gep
   ret void
 }
 
 ; GCN-LABEL: {{^}}v_ubfe_sub_multi_use_shl_i32:
 ; GCN: {{buffer|flat}}_load_dword [[SRC:v[0-9]+]]
 ; GCN: {{buffer|flat}}_load_dword [[WIDTH:v[0-9]+]]
-; GCN: v_sub_i32_e32 [[SUB:v[0-9]+]], vcc, 32, [[WIDTH]]
+; GCN: v_sub_{{[iu]}}32_e32 [[SUB:v[0-9]+]], vcc, 32, [[WIDTH]]
 
 ; SI-NEXT: v_lshl_b32_e32 [[SHL:v[0-9]+]], [[SRC]], [[SUB]]
 ; SI-NEXT: v_lshr_b32_e32 [[BFE:v[0-9]+]], [[SHL]], [[SUB]]
 
 ; VI-NEXT: v_lshlrev_b32_e32 [[SHL:v[0-9]+]], [[SUB]], [[SRC]]
 ; VI-NEXT: v_lshrrev_b32_e32 [[BFE:v[0-9]+]], [[SUB]], [[SHL]]
 
 ; GCN: [[BFE]]
define amdgpu_kernel void @v_sbfe_sub_i32(i32 addrspace(1)* %out, i32 addrspace(1)* %in0, i32 addrspace(1)* %in1) #1 {
   %bfe = ashr i32 %shl, %sub
   store i32 %bfe, i32 addrspace(1)* %out.gep
   ret void
 }
 
 ; GCN-LABEL: {{^}}v_sbfe_sub_multi_use_shl_i32:
 ; GCN: {{buffer|flat}}_load_dword [[SRC:v[0-9]+]]
 ; GCN: {{buffer|flat}}_load_dword [[WIDTH:v[0-9]+]]
-; GCN: v_sub_i32_e32 [[SUB:v[0-9]+]], vcc, 32, [[WIDTH]]
+; GCN: v_sub_{{[iu]}}32_e32 [[SUB:v[0-9]+]], vcc, 32, [[WIDTH]]
 
 ; SI-NEXT: v_lshl_b32_e32 [[SHL:v[0-9]+]], [[SRC]], [[SUB]]
 ; SI-NEXT: v_ashr_i32_e32 [[BFE:v[0-9]+]], [[SHL]], [[SUB]]
 
 ; VI-NEXT: v_lshlrev_b32_e32 [[SHL:v[0-9]+]], [[SUB]], [[SRC]]
 ; VI-NEXT: v_ashrrev_i32_e32 [[BFE:v[0-9]+]], [[SUB]], [[SHL]]
 
 ; GCN: [[BFE]]

llvm/trunk/test/CodeGen/AMDGPU/byval-frame-setup.ll

 ; GCN-DAG: buffer_store_dword v32
 ; GCN-DAG: buffer_store_dword v33
 ; GCN-NOT: v_writelane_b32 v{{[0-9]+}}, s32
 ; GCN: v_writelane_b32
 
 ; GCN-DAG: s_add_u32 s32, s32, 0xb00{{$}}
 
 ; GCN-DAG: buffer_load_dword [[LOAD0:v[0-9]+]], off, s[0:3], s5 offset:4{{$}}
-; GCN: v_add_i32_e32 [[ADD0:v[0-9]+]], vcc, 1, [[LOAD0]]
+; GCN: v_add_{{[iu]}}32_e32 [[ADD0:v[0-9]+]], vcc, 1, [[LOAD0]]
 ; GCN: buffer_store_dword [[ADD0]], off, s[0:3], s5 offset:4{{$}}
 
 ; GCN: buffer_load_dword [[LOAD1:v[0-9]+]], off, s[0:3], s5 offset:20{{$}}
-; GCN: v_add_i32_e32 [[ADD1:v[0-9]+]], vcc, 2, [[LOAD1]]
+; GCN: v_add_{{[iu]}}32_e32 [[ADD1:v[0-9]+]], vcc, 2, [[LOAD1]]
 
 ; GCN: s_swappc_b64
 
 ; GCN: buffer_store_dword [[ADD1]], off, s[0:3], s5 offset:20{{$}}
 
 ; GCN: v_readlane_b32
 ; GCN-NOT: v_readlane_b32 s32
 ; GCN: buffer_load_dword v32,

llvm/trunk/test/CodeGen/AMDGPU/ctlz.ll

 ; SI-DAG: v_ffbh_u32_e32 [[FFBH:v[0-9]+]], [[VAL]]
 ; VI-DAG: v_ffbh_u32_sdwa [[FFBH:v[0-9]+]], [[VAL]] dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_0
 ; SI: v_cmp_ne_u32_e32 vcc, 0, [[VAL]]
 ; VI: v_cmp_ne_u16_e32 vcc, 0, [[VAL]]
 
 ; GCN: v_cndmask_b32_e32 [[SELECT:v[0-9]+]], 32, [[FFBH]], vcc
 
 ; SI: v_subrev_i32_e32 [[RESULT:v[0-9]+]], vcc, 24, [[SELECT]]
-; VI: v_add_i32_e32 [[RESULT:v[0-9]+]], vcc, -16, [[SELECT]]
+; VI: v_add_u32_e32 [[RESULT:v[0-9]+]], vcc, -16, [[SELECT]]
 ; GCN: buffer_store_byte [[RESULT]],
 ; GCN: s_endpgm
 define amdgpu_kernel void @v_ctlz_i8(i8 addrspace(1)* noalias %out, i8 addrspace(1)* noalias %valptr) nounwind {
   %val = load i8, i8 addrspace(1)* %valptr
   %ctlz = call i8 @llvm.ctlz.i8(i8 %val, i1 false) nounwind readnone
   store i8 %ctlz, i8 addrspace(1)* %out
   ret void
 }
define amdgpu_kernel void @s_ctlz_i64_trunc(i32 addrspace(1)* noalias %out, i64 %val) nounwind {
   store i32 %trunc, i32 addrspace(1)* %out
   ret void
 }
 
 ; FUNC-LABEL: {{^}}v_ctlz_i64:
 ; GCN-DAG: {{buffer|flat}}_load_dwordx2 v{{\[}}[[LO:[0-9]+]]:[[HI:[0-9]+]]{{\]}}
 ; GCN-DAG: v_cmp_eq_u32_e32 vcc, 0, v[[HI]]
 ; GCN-DAG: v_ffbh_u32_e32 [[FFBH_LO:v[0-9]+]], v[[LO]]
-; GCN-DAG: v_add_i32_e32 [[ADD:v[0-9]+]], vcc, 32, [[FFBH_LO]]
+; GCN-DAG: v_add_{{[iu]}}32_e32 [[ADD:v[0-9]+]], vcc, 32, [[FFBH_LO]]
 ; GCN-DAG: v_ffbh_u32_e32 [[FFBH_HI:v[0-9]+]], v[[HI]]
 ; GCN-DAG: v_cndmask_b32_e32 v[[CTLZ:[0-9]+]], [[FFBH_HI]], [[ADD]], vcc
 ; GCN-DAG: v_or_b32_e32 [[OR:v[0-9]+]], v[[LO]], v[[HI]]
 ; GCN-DAG: v_cmp_ne_u32_e32 vcc, 0, [[OR]]
 ; GCN-DAG: v_cndmask_b32_e32 v[[CLTZ_LO:[0-9]+]], 64, v[[CTLZ:[0-9]+]], vcc
 ; GCN: {{buffer|flat}}_store_dwordx2 {{.*}}v{{\[}}[[CLTZ_LO]]:[[CTLZ_HI:[0-9]+]]{{\]}}
 define amdgpu_kernel void @v_ctlz_i64(i64 addrspace(1)* noalias %out, i64 addrspace(1)* noalias %in) nounwind {
   %tid = call i32 @llvm.r600.read.tidig.x()

llvm/trunk/test/CodeGen/AMDGPU/ctlz_zero_undef.ll

define amdgpu_kernel void @s_ctlz_zero_undef_i64_trunc(i32 addrspace(1)* noalias %out, i64 %val) nounwind {
   store i32 %trunc, i32 addrspace(1)* %out
   ret void
 }
 
 ; FUNC-LABEL: {{^}}v_ctlz_zero_undef_i64:
 ; GCN-DAG: {{buffer|flat}}_load_dwordx2 v{{\[}}[[LO:[0-9]+]]:[[HI:[0-9]+]]{{\]}}
 ; GCN-DAG: v_cmp_eq_u32_e32 vcc, 0, v[[HI]]
 ; GCN-DAG: v_ffbh_u32_e32 [[FFBH_LO:v[0-9]+]], v[[LO]]
-; GCN-DAG: v_add_i32_e32 [[ADD:v[0-9]+]], vcc, 32, [[FFBH_LO]]
+; GCN-DAG: v_add_{{[iu]}}32_e32 [[ADD:v[0-9]+]], vcc, 32, [[FFBH_LO]]
 ; GCN-DAG: v_ffbh_u32_e32 [[FFBH_HI:v[0-9]+]], v[[HI]]
 ; GCN-DAG: v_cndmask_b32_e32 v[[CTLZ:[0-9]+]], [[FFBH_HI]], [[FFBH_LO]]
 ; GCN: {{buffer|flat}}_store_dwordx2 {{.*}}v{{\[}}[[CTLZ]]:[[CTLZ_HI:[0-9]+]]{{\]}}
 define amdgpu_kernel void @v_ctlz_zero_undef_i64(i64 addrspace(1)* noalias %out, i64 addrspace(1)* noalias %in) nounwind {
   %tid = call i32 @llvm.r600.read.tidig.x()
   %in.gep = getelementptr i64, i64 addrspace(1)* %in, i32 %tid
   %out.gep = getelementptr i64, i64 addrspace(1)* %out, i32 %tid
   %val = load i64, i64 addrspace(1)* %in.gep

llvm/trunk/test/CodeGen/AMDGPU/ctpop64.ll

 ; VI: flat_load_dwordx4   v{{\[}}[[VAL0:[0-9]+]]:[[VAL3:[0-9]+]]{{\]}}, v{{\[[0-9]+:[0-9]+\]}}
 
 ; GCN-DAG: v_bcnt_u32_b32{{(_e64)*}} [[MIDRESULT0:v[0-9]+]], v{{[0-9]+}}, 0
 ; GCN-DAG: v_bcnt_u32_b32{{(_e32)*(_e64)*}} [[MIDRESULT1:v[0-9]+]], v[[VAL3]], [[MIDRESULT0]]
 
 ; GCN-DAG: v_bcnt_u32_b32{{(_e64)*}} [[MIDRESULT2:v[0-9]+]], v[[VAL0]], 0
 ; GCN-DAG: v_bcnt_u32_b32{{(_e32)*(_e64)*}} [[MIDRESULT3:v[0-9]+]], v{{[0-9]+}}, [[MIDRESULT2]]
 
-; GCN: v_add_i32_e32 [[RESULT:v[0-9]+]], vcc, [[MIDRESULT1]], [[MIDRESULT2]]
+; GCN: v_add_{{[iu]}}32_e32 [[RESULT:v[0-9]+]], vcc, [[MIDRESULT1]], [[MIDRESULT2]]
 
 ; GCN: buffer_store_dword [[RESULT]],
 ; GCN: s_endpgm
 define amdgpu_kernel void @v_ctpop_i128(i32 addrspace(1)* noalias %out, i128 addrspace(1)* noalias %in) nounwind {
   %tid = call i32 @llvm.r600.read.tidig.x()
   %in.gep = getelementptr i128, i128 addrspace(1)* %in, i32 %tid
   %val = load i128, i128 addrspace(1)* %in.gep, align 8
   %ctpop = call i128 @llvm.ctpop.i128(i128 %val) nounwind readnone
   %truncctpop = trunc i128 %ctpop to i32
   store i32 %truncctpop, i32 addrspace(1)* %out, align 4
   ret void
 }

llvm/trunk/test/CodeGen/AMDGPU/cvt_f32_ubyte.ll

define amdgpu_kernel void @load_v8i8_to_v8f32(<8 x float> addrspace(1)* noalias %out, <8 x i8> addrspace(1)* noalias %in) nounwind {
   %load = load <8 x i8>, <8 x i8> addrspace(1)* %gep, align 8
   %cvt = uitofp <8 x i8> %load to <8 x float>
   store <8 x float> %cvt, <8 x float> addrspace(1)* %out, align 16
   ret void
 }
 
 ; GCN-LABEL: {{^}}i8_zext_inreg_i32_to_f32:
 ; GCN: {{buffer|flat}}_load_dword [[LOADREG:v[0-9]+]],
-; GCN: v_add_i32_e32 [[ADD:v[0-9]+]], vcc, 2, [[LOADREG]]
+; GCN: v_add_{{[iu]}}32_e32 [[ADD:v[0-9]+]], vcc, 2, [[LOADREG]]
 ; GCN-NEXT: v_cvt_f32_ubyte0_e32 [[CONV:v[0-9]+]], [[ADD]]
 ; GCN: buffer_store_dword [[CONV]],
 define amdgpu_kernel void @i8_zext_inreg_i32_to_f32(float addrspace(1)* noalias %out, i32 addrspace(1)* noalias %in) nounwind {
   %tid = call i32 @llvm.amdgcn.workitem.id.x()
   %gep = getelementptr i32, i32 addrspace(1)* %in, i32 %tid
   %load = load i32, i32 addrspace(1)* %gep, align 4
   %add = add i32 %load, 2
   %inreg = and i32 %add, 255

llvm/trunk/test/CodeGen/AMDGPU/ds-combine-large-stride.ll

 ; RUN: llc -mtriple=amdgcn--amdhsa -mcpu=tonga -verify-machineinstrs < %s | FileCheck -check-prefix=GCN %s
 ; RUN: llc -mtriple=amdgcn--amdhsa -mcpu=gfx900 -verify-machineinstrs < %s | FileCheck -check-prefix=GFX9 -check-prefix=GCN %s
 
 ; GCN-LABEL: ds_read32_combine_stride_400:
 ; GCN:     s_load_dword [[ARG:s[0-9]+]], s[4:5], 0x0
 ; GCN:     v_mov_b32_e32 [[BASE:v[0-9]+]], [[ARG]]
-; GCN-DAG: v_add_i32_e32 [[B1:v[0-9]+]], vcc, 0x320, [[BASE]]
-; GCN-DAG: v_add_i32_e32 [[B2:v[0-9]+]], vcc, 0x640, [[BASE]]
-; GCN-DAG: v_add_i32_e32 [[B3:v[0-9]+]], vcc, 0x960, [[BASE]]
-; GFX9-DAG: v_add_i32_e32 [[B1:v[0-9]+]], vcc, 0x320, [[BASE]]
-; GFX9-DAG: v_add_i32_e32 [[B2:v[0-9]+]], vcc, 0x640, [[BASE]]
-; GFX9-DAG: v_add_i32_e32 [[B3:v[0-9]+]], vcc, 0x960, [[BASE]]
+; GCN-DAG: v_add_{{[_co]*}}u32_e32 [[B1:v[0-9]+]], vcc, 0x320, [[BASE]]
+; GCN-DAG: v_add_{{[_co]*}}u32_e32 [[B2:v[0-9]+]], vcc, 0x640, [[BASE]]
+; GCN-DAG: v_add_{{[_co]*}}u32_e32 [[B3:v[0-9]+]], vcc, 0x960, [[BASE]]
+; GFX9-DAG: v_add_{{[_co]*}}u32_e32 [[B1:v[0-9]+]], vcc, 0x320, [[BASE]]
+; GFX9-DAG: v_add_{{[_co]*}}u32_e32 [[B2:v[0-9]+]], vcc, 0x640, [[BASE]]
+; GFX9-DAG: v_add_{{[_co]*}}u32_e32 [[B3:v[0-9]+]], vcc, 0x960, [[BASE]]
 ; GCN-DAG: ds_read2_b32  v[{{[0-9]+:[0-9]+}}], [[BASE]] offset1:100
 ; GCN-DAG: ds_read2_b32  v[{{[0-9]+:[0-9]+}}], [[B1]] offset1:100
 ; GCN-DAG: ds_read2_b32  v[{{[0-9]+:[0-9]+}}], [[B2]] offset1:100
 ; GCN-DAG: ds_read2_b32  v[{{[0-9]+:[0-9]+}}], [[B3]] offset1:100
 define amdgpu_kernel void @ds_read32_combine_stride_400(float addrspace(3)* nocapture readonly %arg, float *nocapture %arg1) {
 bb:
   %tmp = load float, float addrspace(3)* %arg, align 4
   %tmp2 = fadd float %tmp, 0.000000e+00
bb:
   %tmp23 = fadd float %tmp20, %tmp22
   store float %tmp23, float *%arg1, align 4
   ret void
 }
 
 ; GCN-LABEL: ds_read32_combine_stride_400_back:
 ; GCN:     s_load_dword [[ARG:s[0-9]+]], s[4:5], 0x0
 ; GCN:     v_mov_b32_e32 [[BASE:v[0-9]+]], [[ARG]]
-; GCN-DAG: v_add_i32_e32 [[B1:v[0-9]+]], vcc, 0x320, [[BASE]]
-; GCN-DAG: v_add_i32_e32 [[B2:v[0-9]+]], vcc, 0x640, [[BASE]]
-; GCN-DAG: v_add_i32_e32 [[B3:v[0-9]+]], vcc, 0x960, [[BASE]]
-; GFX9-DAG: v_add_i32_e32 [[B1:v[0-9]+]], vcc, 0x320, [[BASE]]
-; GFX9-DAG: v_add_i32_e32 [[B2:v[0-9]+]], vcc, 0x640, [[BASE]]
-; GFX9-DAG: v_add_i32_e32 [[B3:v[0-9]+]], vcc, 0x960, [[BASE]]
+; GCN-DAG: v_add_{{[_co]*}}u32_e32 [[B1:v[0-9]+]], vcc, 0x320, [[BASE]]
+; GCN-DAG: v_add_{{[_co]*}}u32_e32 [[B2:v[0-9]+]], vcc, 0x640, [[BASE]]
+; GCN-DAG: v_add_{{[_co]*}}u32_e32 [[B3:v[0-9]+]], vcc, 0x960, [[BASE]]
+; GFX9-DAG: v_add_{{[_co]*}}u32_e32 [[B1:v[0-9]+]], vcc, 0x320, [[BASE]]
+; GFX9-DAG: v_add_{{[_co]*}}u32_e32 [[B2:v[0-9]+]], vcc, 0x640, [[BASE]]
+; GFX9-DAG: v_add_{{[_co]*}}u32_e32 [[B3:v[0-9]+]], vcc, 0x960, [[BASE]]
 ; GCN-DAG: ds_read2_b32  v[{{[0-9]+:[0-9]+}}], [[BASE]] offset1:100
 ; GCN-DAG: ds_read2_b32  v[{{[0-9]+:[0-9]+}}], [[B1]] offset1:100
 ; GCN-DAG: ds_read2_b32  v[{{[0-9]+:[0-9]+}}], [[B2]] offset1:100
 ; GCN-DAG: ds_read2_b32  v[{{[0-9]+:[0-9]+}}], [[B3]] offset1:100
 define amdgpu_kernel void @ds_read32_combine_stride_400_back(float addrspace(3)* nocapture readonly %arg, float *nocapture %arg1) {
 bb:
   %tmp = getelementptr inbounds float, float addrspace(3)* %arg, i32 700
   %tmp2 = load float, float addrspace(3)* %tmp, align 4
bb:
   %tmp23 = fadd float %tmp20, %tmp22
   store float %tmp23, float *%arg1, align 4
   ret void
 }
 
 ; GCN-LABEL: ds_read32_combine_stride_8192_shifted:
 ; GCN:     s_load_dword [[ARG:s[0-9]+]], s[4:5], 0x0
 ; GCN:     v_mov_b32_e32 [[BASE:v[0-9]+]], [[ARG]]
-; GCN-DAG: v_add_i32_e32 [[B1:v[0-9]+]], vcc, 8, [[BASE]]
-; GCN-DAG: v_add_i32_e32 [[B2:v[0-9]+]], vcc, 0x4008, [[BASE]]
-; GCN-DAG: v_add_i32_e32 [[B3:v[0-9]+]], vcc, 0x8008, [[BASE]]
-; GFX9-DAG: v_add_i32_e32 [[B1:v[0-9]+]], vcc, 8, [[BASE]]
-; GFX9-DAG: v_add_i32_e32 [[B2:v[0-9]+]], vcc, 0x4008, [[BASE]]
-; GFX9-DAG: v_add_i32_e32 [[B3:v[0-9]+]], vcc, 0x8008, [[BASE]]
+; GCN-DAG: v_add_{{[_co]*}}u32_e32 [[B1:v[0-9]+]], vcc, 8, [[BASE]]
+; GCN-DAG: v_add_{{[_co]*}}u32_e32 [[B2:v[0-9]+]], vcc, 0x4008, [[BASE]]
+; GCN-DAG: v_add_{{[_co]*}}u32_e32 [[B3:v[0-9]+]], vcc, 0x8008, [[BASE]]
+; GFX9-DAG: v_add_{{[_co]*}}u32_e32 [[B1:v[0-9]+]], vcc, 8, [[BASE]]
+; GFX9-DAG: v_add_{{[_co]*}}u32_e32 [[B2:v[0-9]+]], vcc, 0x4008, [[BASE]]
+; GFX9-DAG: v_add_{{[_co]*}}u32_e32 [[B3:v[0-9]+]], vcc, 0x8008, [[BASE]]
 ; GCN-DAG: ds_read2st64_b32 v[{{[0-9]+:[0-9]+}}], [[B1]] offset1:32
 ; GCN-DAG: ds_read2st64_b32 v[{{[0-9]+:[0-9]+}}], [[B2]] offset1:32
 ; GCN-DAG: ds_read2st64_b32 v[{{[0-9]+:[0-9]+}}], [[B3]] offset1:32
 define amdgpu_kernel void @ds_read32_combine_stride_8192_shifted(float addrspace(3)* nocapture readonly %arg, float *nocapture %arg1) {
 bb:
   %tmp = getelementptr inbounds float, float addrspace(3)* %arg, i32 2
   %tmp2 = load float, float addrspace(3)* %tmp, align 4
   %tmp3 = fadd float %tmp2, 0.000000e+00
bb:
   %tmp18 = fadd float %tmp15, %tmp17
   store float %tmp18, float *%arg1, align 4
   ret void
 }
 
 ; GCN-LABEL: ds_read64_combine_stride_400:
 ; GCN:     s_load_dword [[ARG:s[0-9]+]], s[4:5], 0x0
 ; GCN:     v_mov_b32_e32 [[BASE:v[0-9]+]], [[ARG]]
-; GCN-DAG: v_add_i32_e32 [[B1:v[0-9]+]], vcc, 0x960, [[BASE]]
-; GFX9-DAG: v_add_i32_e32 [[B1:v[0-9]+]], vcc, 0x960, [[BASE]]
+; GCN-DAG: v_add_{{[_co]*}}u32_e32 [[B1:v[0-9]+]], vcc, 0x960, [[BASE]]
+; GFX9-DAG: v_add_{{[_co]*}}u32_e32 [[B1:v[0-9]+]], vcc, 0x960, [[BASE]]
 ; GCN-DAG: ds_read2_b64  v[{{[0-9]+:[0-9]+}}], [[BASE]] offset1:50
 ; GCN-DAG: ds_read2_b64  v[{{[0-9]+:[0-9]+}}], [[BASE]] offset0:100 offset1:150
 ; GCN-DAG: ds_read2_b64  v[{{[0-9]+:[0-9]+}}], [[BASE]] offset0:200 offset1:250
 ; GCN-DAG: ds_read2_b64  v[{{[0-9]+:[0-9]+}}], [[B1]] offset1:50
 define amdgpu_kernel void @ds_read64_combine_stride_400(double addrspace(3)* nocapture readonly %arg, double *nocapture %arg1) {
 bb:
   %tmp = load double, double addrspace(3)* %arg, align 8
   %tmp2 = fadd double %tmp, 0.000000e+00
bb:
   %tmp23 = fadd double %tmp20, %tmp22
   store double %tmp23, double *%arg1, align 8
   ret void
 }
 
 ; GCN-LABEL: ds_read64_combine_stride_8192_shifted:
 ; GCN:     s_load_dword [[ARG:s[0-9]+]], s[4:5], 0x0
 ; GCN:     v_mov_b32_e32 [[BASE:v[0-9]+]], [[ARG]]
-; GCN-DAG: v_add_i32_e32 [[B1:v[0-9]+]], vcc, 8, [[BASE]]
-; GCN-DAG: v_add_i32_e32 [[B2:v[0-9]+]], vcc, 0x4008, [[BASE]]
-; GCN-DAG: v_add_i32_e32 [[B3:v[0-9]+]], vcc, 0x8008, [[BASE]]
-; GFX9-DAG: v_add_i32_e32 [[B1:v[0-9]+]], vcc, 8, [[BASE]]
-; GFX9-DAG: v_add_i32_e32 [[B2:v[0-9]+]], vcc, 0x4008, [[BASE]]
-; GFX9-DAG: v_add_i32_e32 [[B3:v[0-9]+]], vcc, 0x8008, [[BASE]]
+; GCN-DAG: v_add_{{[_co]*}}u32_e32 [[B1:v[0-9]+]], vcc, 8, [[BASE]]
+; GCN-DAG: v_add_{{[_co]*}}u32_e32 [[B2:v[0-9]+]], vcc, 0x4008, [[BASE]]
+; GCN-DAG: v_add_{{[_co]*}}u32_e32 [[B3:v[0-9]+]], vcc, 0x8008, [[BASE]]
+; GFX9-DAG: v_add_{{[_co]*}}u32_e32 [[B1:v[0-9]+]], vcc, 8, [[BASE]]
+; GFX9-DAG: v_add_{{[_co]*}}u32_e32 [[B2:v[0-9]+]], vcc, 0x4008, [[BASE]]
+; GFX9-DAG: v_add_{{[_co]*}}u32_e32 [[B3:v[0-9]+]], vcc, 0x8008, [[BASE]]
 ; GCN-DAG: ds_read2st64_b64 v[{{[0-9]+:[0-9]+}}], [[B1]] offset1:16
 ; GCN-DAG: ds_read2st64_b64 v[{{[0-9]+:[0-9]+}}], [[B2]] offset1:16
 ; GCN-DAG: ds_read2st64_b64 v[{{[0-9]+:[0-9]+}}], [[B3]] offset1:16
 define amdgpu_kernel void @ds_read64_combine_stride_8192_shifted(double addrspace(3)* nocapture readonly %arg, double *nocapture %arg1) {
 bb:
   %tmp = getelementptr inbounds double, double addrspace(3)* %arg, i32 1
   %tmp2 = load double, double addrspace(3)* %tmp, align 8
   %tmp3 = fadd double %tmp2, 0.000000e+00
bb:
   %tmp18 = fadd double %tmp15, %tmp17
   store double %tmp18, double *%arg1, align 8
   ret void
 }
 
 ; GCN-LABEL: ds_write32_combine_stride_400:
 ; GCN:     s_load_dword [[ARG:s[0-9]+]], s[4:5], 0x0
 ; GCN:     v_mov_b32_e32 [[BASE:v[0-9]+]], [[ARG]]
-; GCN-DAG: v_add_i32_e32 [[B1:v[0-9]+]], vcc, 0x320, [[BASE]]
-; GCN-DAG: v_add_i32_e32 [[B2:v[0-9]+]], vcc, 0x640, [[BASE]]
-; GCN-DAG: v_add_i32_e32 [[B3:v[0-9]+]], vcc, 0x960, [[BASE]]
-; GFX9-DAG: v_add_i32_e32 [[B1:v[0-9]+]], vcc, 0x320, [[BASE]]
-; GFX9-DAG: v_add_i32_e32 [[B2:v[0-9]+]], vcc, 0x640, [[BASE]]
-; GFX9-DAG: v_add_i32_e32 [[B3:v[0-9]+]], vcc, 0x960, [[BASE]]
+; GCN-DAG: v_add_{{[_co]*}}u32_e32 [[B1:v[0-9]+]], vcc, 0x320, [[BASE]]
+; GCN-DAG: v_add_{{[_co]*}}u32_e32 [[B2:v[0-9]+]], vcc, 0x640, [[BASE]]
+; GCN-DAG: v_add_{{[_co]*}}u32_e32 [[B3:v[0-9]+]], vcc, 0x960, [[BASE]]
+; GFX9-DAG: v_add_{{[_co]*}}u32_e32 [[B1:v[0-9]+]], vcc, 0x320, [[BASE]]
+; GFX9-DAG: v_add_{{[_co]*}}u32_e32 [[B2:v[0-9]+]], vcc, 0x640, [[BASE]]
+; GFX9-DAG: v_add_{{[_co]*}}u32_e32 [[B3:v[0-9]+]], vcc, 0x960, [[BASE]]
 ; GCN-DAG: ds_write2_b32 [[BASE]], v{{[0-9]+}}, v{{[0-9]+}} offset1:100
 ; GCN-DAG: ds_write2_b32 [[B1]], v{{[0-9]+}}, v{{[0-9]+}} offset1:100
 ; GCN-DAG: ds_write2_b32 [[B2]], v{{[0-9]+}}, v{{[0-9]+}} offset1:100
 ; GCN-DAG: ds_write2_b32 [[B3]], v{{[0-9]+}}, v{{[0-9]+}} offset1:100
 define amdgpu_kernel void @ds_write32_combine_stride_400(float addrspace(3)* nocapture %arg) {
 bb:
   store float 1.000000e+00, float addrspace(3)* %arg, align 4
   %tmp = getelementptr inbounds float, float addrspace(3)* %arg, i32 100
bb:
   %tmp6 = getelementptr inbounds float, float addrspace(3)* %arg, i32 700
   store float 1.000000e+00, float addrspace(3)* %tmp6, align 4
   ret void
 }
 
 ; GCN-LABEL: ds_write32_combine_stride_400_back:
 ; GCN:     s_load_dword [[ARG:s[0-9]+]], s[4:5], 0x0
 ; GCN:     v_mov_b32_e32 [[BASE:v[0-9]+]], [[ARG]]
-; GCN-DAG: v_add_i32_e32 [[B1:v[0-9]+]], vcc, 0x320, [[BASE]]
-; GCN-DAG: v_add_i32_e32 [[B2:v[0-9]+]], vcc, 0x640, [[BASE]]
-; GCN-DAG: v_add_i32_e32 [[B3:v[0-9]+]], vcc, 0x960, [[BASE]]
-; GFX9-DAG: v_add_i32_e32 [[B1:v[0-9]+]], vcc, 0x320, [[BASE]]
-; GFX9-DAG: v_add_i32_e32 [[B2:v[0-9]+]], vcc, 0x640, [[BASE]]
-; GFX9-DAG: v_add_i32_e32 [[B3:v[0-9]+]], vcc, 0x960, [[BASE]]
+; GCN-DAG: v_add_{{[_co]*}}u32_e32 [[B1:v[0-9]+]], vcc, 0x320, [[BASE]]
+; GCN-DAG: v_add_{{[_co]*}}u32_e32 [[B2:v[0-9]+]], vcc, 0x640, [[BASE]]
+; GCN-DAG: v_add_{{[_co]*}}u32_e32 [[B3:v[0-9]+]], vcc, 0x960, [[BASE]]
+; GFX9-DAG: v_add_{{[_co]*}}u32_e32 [[B1:v[0-9]+]], vcc, 0x320, [[BASE]]
+; GFX9-DAG: v_add_{{[_co]*}}u32_e32 [[B2:v[0-9]+]], vcc, 0x640, [[BASE]]
+; GFX9-DAG: v_add_{{[_co]*}}u32_e32 [[B3:v[0-9]+]], vcc, 0x960, [[BASE]]
 ; GCN-DAG: ds_write2_b32 [[BASE]], v{{[0-9]+}}, v{{[0-9]+}} offset1:100
 ; GCN-DAG: ds_write2_b32 [[B1]], v{{[0-9]+}}, v{{[0-9]+}} offset1:100
 ; GCN-DAG: ds_write2_b32 [[B2]], v{{[0-9]+}}, v{{[0-9]+}} offset1:100
 ; GCN-DAG: ds_write2_b32 [[B3]], v{{[0-9]+}}, v{{[0-9]+}} offset1:100
 define amdgpu_kernel void @ds_write32_combine_stride_400_back(float addrspace(3)* nocapture %arg) {
 bb:
   %tmp = getelementptr inbounds float, float addrspace(3)* %arg, i32 700
   store float 1.000000e+00, float addrspace(3)* %tmp, align 4
bb:
   %tmp6 = getelementptr inbounds float, float addrspace(3)* %arg, i32 14336
   store float 1.000000e+00, float addrspace(3)* %tmp6, align 4
   ret void
 }
 
 ; GCN-LABEL: ds_write32_combine_stride_8192_shifted:
 ; GCN:     s_load_dword [[ARG:s[0-9]+]], s[4:5], 0x0
 ; GCN:     v_mov_b32_e32 [[BASE:v[0-9]+]], [[ARG]]
-; GCN-DAG: v_add_i32_e32 [[B1:v[0-9]+]], vcc, 4, [[BASE]]
-; GCN-DAG: v_add_i32_e32 [[B2:v[0-9]+]], vcc, 0x4004, [[BASE]]
-; GCN-DAG: v_add_i32_e32 [[B3:v[0-9]+]], vcc, 0x8004, [[BASE]]
-; GFX9-DAG: v_add_i32_e32 [[B1:v[0-9]+]], vcc, 4, [[BASE]]
-; GFX9-DAG: v_add_i32_e32 [[B2:v[0-9]+]], vcc, 0x4004, [[BASE]]
-; GFX9-DAG: v_add_i32_e32 [[B3:v[0-9]+]], vcc, 0x8004, [[BASE]]
+; GCN-DAG: v_add_{{[_co]*}}u32_e32 [[B1:v[0-9]+]], vcc, 4, [[BASE]]
+; GCN-DAG: v_add_{{[_co]*}}u32_e32 [[B2:v[0-9]+]], vcc, 0x4004, [[BASE]]
+; GCN-DAG: v_add_{{[_co]*}}u32_e32 [[B3:v[0-9]+]], vcc, 0x8004, [[BASE]]
+; GFX9-DAG: v_add_{{[_co]*}}u32_e32 [[B1:v[0-9]+]], vcc, 4, [[BASE]]
+; GFX9-DAG: v_add_{{[_co]*}}u32_e32 [[B2:v[0-9]+]], vcc, 0x4004, [[BASE]]
+; GFX9-DAG: v_add_{{[_co]*}}u32_e32 [[B3:v[0-9]+]], vcc, 0x8004, [[BASE]]
 ; GCN-DAG: ds_write2st64_b32 [[B1]], v{{[0-9]+}}, v{{[0-9]+}} offset1:32
 ; GCN-DAG: ds_write2st64_b32 [[B2]], v{{[0-9]+}}, v{{[0-9]+}} offset1:32
 ; GCN-DAG: ds_write2st64_b32 [[B3]], v{{[0-9]+}}, v{{[0-9]+}} offset1:32
 define amdgpu_kernel void @ds_write32_combine_stride_8192_shifted(float addrspace(3)* nocapture %arg) {
 bb:
   %tmp = getelementptr inbounds float, float addrspace(3)* %arg, i32 1
   store float 1.000000e+00, float addrspace(3)* %tmp, align 4
   %tmp1 = getelementptr inbounds float, float addrspace(3)* %arg, i32 2049
   store float 1.000000e+00, float addrspace(3)* %tmp1, align 4
   %tmp2 = getelementptr inbounds float, float addrspace(3)* %arg, i32 4097
   store float 1.000000e+00, float addrspace(3)* %tmp2, align 4
   %tmp3 = getelementptr inbounds float, float addrspace(3)* %arg, i32 6145
   store float 1.000000e+00, float addrspace(3)* %tmp3, align 4
   %tmp4 = getelementptr inbounds float, float addrspace(3)* %arg, i32 8193
   store float 1.000000e+00, float addrspace(3)* %tmp4, align 4
   %tmp5 = getelementptr inbounds float, float addrspace(3)* %arg, i32 10241
   store float 1.000000e+00, float addrspace(3)* %tmp5, align 4
   ret void
 }
 
 ; GCN-LABEL: ds_write64_combine_stride_400:
 ; GCN:     s_load_dword [[ARG:s[0-9]+]], s[4:5], 0x0
 ; GCN:     v_mov_b32_e32 [[BASE:v[0-9]+]], [[ARG]]
-; GCN-DAG: v_add_i32_e32 [[B1:v[0-9]+]], vcc, 0x960, [[BASE]]
-; GFX9-DAG: v_add_i32_e32 [[B1:v[0-9]+]], vcc, 0x960, [[BASE]]
+; GCN-DAG: v_add_{{[_co]*}}u32_e32 [[B1:v[0-9]+]], vcc, 0x960, [[BASE]]
+; GFX9-DAG: v_add_{{[_co]*}}u32_e32 [[B1:v[0-9]+]], vcc, 0x960, [[BASE]]
 ; GCN-DAG: ds_write2_b64 [[BASE]], v[{{[0-9]+:[0-9]+}}], v[{{[0-9]+:[0-9]+}}] offset1:50
 ; GCN-DAG: ds_write2_b64 [[BASE]], v[{{[0-9]+:[0-9]+}}], v[{{[0-9]+:[0-9]+}}] offset0:100 offset1:150
 ; GCN-DAG: ds_write2_b64 [[BASE]], v[{{[0-9]+:[0-9]+}}], v[{{[0-9]+:[0-9]+}}] offset0:200 offset1:250
 ; GCN-DAG: ds_write2_b64 [[B1]],   v[{{[0-9]+:[0-9]+}}], v[{{[0-9]+:[0-9]+}}] offset1:50
 define amdgpu_kernel void @ds_write64_combine_stride_400(double addrspace(3)* nocapture %arg) {
 bb:
   store double 1.000000e+00, double addrspace(3)* %arg, align 8
   %tmp = getelementptr inbounds double, double addrspace(3)* %arg, i32 50
bb:
   %tmp6 = getelementptr inbounds double, double addrspace(3)* %arg, i32 350
   store double 1.000000e+00, double addrspace(3)* %tmp6, align 8
   ret void
 }
 
 ; GCN-LABEL: ds_write64_combine_stride_8192_shifted:
 ; GCN:     s_load_dword [[ARG:s[0-9]+]], s[4:5], 0x0
 ; GCN:     v_mov_b32_e32 [[BASE:v[0-9]+]], [[ARG]]
-; GCN-DAG: v_add_i32_e32 [[B1:v[0-9]+]], vcc, 8, [[BASE]]
-; GCN-DAG: v_add_i32_e32 [[B2:v[0-9]+]], vcc, 0x4008, [[BASE]]
-; GCN-DAG: v_add_i32_e32 [[B3:v[0-9]+]], vcc, 0x8008, [[BASE]]
-; GFX9-DAG: v_add_i32_e32 [[B1:v[0-9]+]], vcc, 8, [[BASE]]
-; GFX9-DAG: v_add_i32_e32 [[B2:v[0-9]+]], vcc, 0x4008, [[BASE]]
-; GFX9-DAG: v_add_i32_e32 [[B3:v[0-9]+]], vcc, 0x8008, [[BASE]]
+; GCN-DAG: v_add_{{[_co]*}}u32_e32 [[B1:v[0-9]+]], vcc, 8, [[BASE]]
+; GCN-DAG: v_add_{{[_co]*}}u32_e32 [[B2:v[0-9]+]], vcc, 0x4008, [[BASE]]
+; GCN-DAG: v_add_{{[_co]*}}u32_e32 [[B3:v[0-9]+]], vcc, 0x8008, [[BASE]]
+; GFX9-DAG: v_add_{{[_co]*}}u32_e32 [[B1:v[0-9]+]], vcc, 8, [[BASE]]
+; GFX9-DAG: v_add_{{[_co]*}}u32_e32 [[B2:v[0-9]+]], vcc, 0x4008, [[BASE]]
+; GFX9-DAG: v_add_{{[_co]*}}u32_e32 [[B3:v[0-9]+]], vcc, 0x8008, [[BASE]]
 ; GCN-DAG: ds_write2st64_b64 [[B1]], v[{{[0-9]+:[0-9]+}}], v[{{[0-9]+:[0-9]+}}] offset1:16
 ; GCN-DAG: ds_write2st64_b64 [[B2]], v[{{[0-9]+:[0-9]+}}], v[{{[0-9]+:[0-9]+}}] offset1:16
 ; GCN-DAG: ds_write2st64_b64 [[B3]], v[{{[0-9]+:[0-9]+}}], v[{{[0-9]+:[0-9]+}}] offset1:16
 define amdgpu_kernel void @ds_write64_combine_stride_8192_shifted(double addrspace(3)* nocapture %arg) {
 bb:
   %tmp = getelementptr inbounds double, double addrspace(3)* %arg, i32 1
   store double 1.000000e+00, double addrspace(3)* %tmp, align 8
   %tmp1 = getelementptr inbounds double, double addrspace(3)* %arg, i32 1025

llvm/trunk/test/CodeGen/AMDGPU/fence-barrier.ll

; <label>:7:                                      ; preds = %6, %1
   %22 = load i64, i64 addrspace(2)* %21, align 8
   %23 = add i64 %22, %20
   %24 = getelementptr inbounds i32, i32 addrspace(1)* %9, i64 %23
   store i32 %8, i32 addrspace(1)* %24, align 4
   ret void
 }
 
 ; GCN-LABEL: {{^}}test_global
-; GCN: v_add_i32_e32 v{{[0-9]+}}, vcc, 0x888, v{{[0-9]+}}
+; GCN: v_add_u32_e32 v{{[0-9]+}}, vcc, 0x888, v{{[0-9]+}}
 ; GCN: flat_store_dword
 ; GCN: s_waitcnt vmcnt(0) lgkmcnt(0){{$}}
 ; GCN-NEXT: s_barrier
 define amdgpu_kernel void @test_global(i32 addrspace(1)*) {
   %2 = alloca i32 addrspace(1)*, align 4
   %3 = alloca i32, align 4
   store i32 addrspace(1)* %0, i32 addrspace(1)** %2, align 4
   store i32 0, i32* %3, align 4

llvm/trunk/test/CodeGen/AMDGPU/function-args.ll

define void @void_func_i1_zeroext(i1 zeroext %arg0) #0 {
   %ext = zext i1 %arg0 to i32
   %add = add i32 %ext, 12
   store i32 %add, i32 addrspace(1)* undef
   ret void
 }
 
 ; GCN-LABEL: {{^}}void_func_i1_signext:
 ; GCN: s_waitcnt
-; GCN-NEXT: v_add_i32_e32 v0, vcc, 12, v0
+; GCN-NEXT: v_add_{{[_coiu]*}}32_e32 v0, vcc, 12, v0
 ; GCN-NOT: v0
 ; GCN: buffer_store_dword v0, off
 define void @void_func_i1_signext(i1 signext %arg0) #0 {
   %ext = sext i1 %arg0 to i32
   %add = add i32 %ext, 12
   store i32 %add, i32 addrspace(1)* undef
   ret void
 }

 ; GCN: buffer_store_byte v0, off
 define void @void_func_i8(i8 %arg0) #0 {
   store i8 %arg0, i8 addrspace(1)* undef
   ret void
 }
 
 ; GCN-LABEL: {{^}}void_func_i8_zeroext:
 ; GCN-NOT: and_b32
-; GCN: v_add_i32_e32 v0, vcc, 12, v0
+; GCN: v_add_{{[_coiu]*}}32_e32 v0, vcc, 12, v0
 define void @void_func_i8_zeroext(i8 zeroext %arg0) #0 {
   %ext = zext i8 %arg0 to i32
   %add = add i32 %ext, 12
   store i32 %add, i32 addrspace(1)* undef
   ret void
 }
 
 ; GCN-LABEL: {{^}}void_func_i8_signext:
 ; GCN-NOT: v_bfe_i32
-; GCN: v_add_i32_e32 v0, vcc, 12, v0
+; GCN: v_add_{{[_coiu]*}}32_e32 v0, vcc, 12, v0
 define void @void_func_i8_signext(i8 signext %arg0) #0 {
   %ext = sext i8 %arg0 to i32
   %add = add i32 %ext, 12
   store i32 %add, i32 addrspace(1)* undef
   ret void
 }
 
 ; GCN-LABEL: {{^}}void_func_i16:
 ; GCN: buffer_store_short v0, off
 define void @void_func_i16(i16 %arg0) #0 {
   store i16 %arg0, i16 addrspace(1)* undef
   ret void
 }
 
 ; GCN-LABEL: {{^}}void_func_i16_zeroext:
 ; GCN-NOT: v0
-; GCN: v_add_i32_e32 v0, vcc, 12, v0
+; GCN: v_add_{{[_coiu]*}}32_e32 v0, vcc, 12, v0
 define void @void_func_i16_zeroext(i16 zeroext %arg0) #0 {
   %ext = zext i16 %arg0 to i32
   %add = add i32 %ext, 12
   store i32 %add, i32 addrspace(1)* undef
   ret void
 }
 
 ; GCN-LABEL: {{^}}void_func_i16_signext:
 ; GCN-NOT: v0
-; GCN: v_add_i32_e32 v0, vcc, 12, v0
+; GCN: v_add_{{[_coiu]*}}32_e32 v0, vcc, 12, v0
 define void @void_func_i16_signext(i16 signext %arg0) #0 {
   %ext = sext i16 %arg0 to i32
   %add = add i32 %ext, 12
   store i32 %add, i32 addrspace(1)* undef
   ret void
 }
 
 ; GCN-LABEL: {{^}}void_func_i32:

llvm/trunk/test/CodeGen/AMDGPU/function-returns.ll

 
 ; GCN-LABEL: {{^}}v33i32_func_void:
 ; GCN: buffer_store_dword v32, off, s[0:3], s5 offset:12 ; 4-byte Folded Spill
 ; GCN: buffer_store_dword v33, off, s[0:3], s5 offset:8 ; 4-byte Folded Spill
 ; GCN: buffer_store_dword v34, off, s[0:3], s5 offset:4 ; 4-byte Folded Spill
 
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, v0, s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_4:v[0-9]+]], vcc, 4, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_4:v[0-9]+]], vcc, 4, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_4]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_8:v[0-9]+]], vcc, 8, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_8:v[0-9]+]], vcc, 8, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_8]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_12:v[0-9]+]], vcc, 12, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_12:v[0-9]+]], vcc, 12, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_12]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_16:v[0-9]+]], vcc, 16, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_16:v[0-9]+]], vcc, 16, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_16]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_20:v[0-9]+]], vcc, 20, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_20:v[0-9]+]], vcc, 20, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_20]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_24:v[0-9]+]], vcc, 24, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_24:v[0-9]+]], vcc, 24, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_24]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_28:v[0-9]+]], vcc, 28, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_28:v[0-9]+]], vcc, 28, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_28]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_32:v[0-9]+]], vcc, 32, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_32:v[0-9]+]], vcc, 32, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_32]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_36:v[0-9]+]], vcc, 36, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_36:v[0-9]+]], vcc, 36, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_36]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_40:v[0-9]+]], vcc, 40, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_40:v[0-9]+]], vcc, 40, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_40]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_44:v[0-9]+]], vcc, 44, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_44:v[0-9]+]], vcc, 44, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_44]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_48:v[0-9]+]], vcc, 48, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_48:v[0-9]+]], vcc, 48, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_48]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_52:v[0-9]+]], vcc, 52, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_52:v[0-9]+]], vcc, 52, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_52]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_56:v[0-9]+]], vcc, 56, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_56:v[0-9]+]], vcc, 56, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_56]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_60:v[0-9]+]], vcc, 60, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_60:v[0-9]+]], vcc, 60, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_60]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_64:v[0-9]+]], vcc, 64, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_64:v[0-9]+]], vcc, 64, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_64]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_68:v[0-9]+]], vcc, 0x44, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_68:v[0-9]+]], vcc, 0x44, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_68]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_72:v[0-9]+]], vcc, 0x48, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_72:v[0-9]+]], vcc, 0x48, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_72]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_76:v[0-9]+]], vcc, 0x4c, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_76:v[0-9]+]], vcc, 0x4c, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_76]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_80:v[0-9]+]], vcc, 0x50, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_80:v[0-9]+]], vcc, 0x50, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_80]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_84:v[0-9]+]], vcc, 0x54, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_84:v[0-9]+]], vcc, 0x54, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_84]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_88:v[0-9]+]], vcc, 0x58, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_88:v[0-9]+]], vcc, 0x58, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_88]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_92:v[0-9]+]], vcc, 0x5c, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_92:v[0-9]+]], vcc, 0x5c, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_92]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_96:v[0-9]+]], vcc, 0x60, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_96:v[0-9]+]], vcc, 0x60, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_96]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_100:v[0-9]+]], vcc, 0x64, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_100:v[0-9]+]], vcc, 0x64, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_100]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_104:v[0-9]+]], vcc, 0x68, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_104:v[0-9]+]], vcc, 0x68, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_104]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_108:v[0-9]+]], vcc, 0x6c, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_108:v[0-9]+]], vcc, 0x6c, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_108]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_112:v[0-9]+]], vcc, 0x70, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_112:v[0-9]+]], vcc, 0x70, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_112]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_116:v[0-9]+]], vcc, 0x74, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_116:v[0-9]+]], vcc, 0x74, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_116]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_120:v[0-9]+]], vcc, 0x78, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_120:v[0-9]+]], vcc, 0x78, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_120]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_124:v[0-9]+]], vcc, 0x7c, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_124:v[0-9]+]], vcc, 0x7c, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_124]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_128:v[0-9]+]], vcc, 0x80, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_128:v[0-9]+]], vcc, 0x80, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_128]], s[0:3], s4 offen{{$}}
 
 ; GCN: buffer_load_dword v34
 ; GCN: buffer_load_dword v33
 ; GCN: buffer_load_dword v32
 ; GCN: s_waitcnt vmcnt(0)
 ; GCN-NEXT: s_setpc_b64
 define <33 x i32> @v33i32_func_void() #0 {
   %ptr = load volatile <33 x i32> addrspace(1)*, <33 x i32> addrspace(1)* addrspace(2)* undef
   %val = load <33 x i32>, <33 x i32> addrspace(1)* %ptr
   ret <33 x i32> %val
 }
 
 ; GCN-LABEL: {{^}}struct_v32i32_i32_func_void:
 ; GCN: buffer_store_dword v32, off, s[0:3], s5 offset:12 ; 4-byte Folded Spill
 ; GCN: buffer_store_dword v33, off, s[0:3], s5 offset:8 ; 4-byte Folded Spill
 ; GCN: buffer_store_dword v34, off, s[0:3], s5 offset:4 ; 4-byte Folded Spill
 
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, v0, s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_4:v[0-9]+]], vcc, 4, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_4:v[0-9]+]], vcc, 4, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_4]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_8:v[0-9]+]], vcc, 8, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_8:v[0-9]+]], vcc, 8, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_8]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_12:v[0-9]+]], vcc, 12, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_12:v[0-9]+]], vcc, 12, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_12]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_16:v[0-9]+]], vcc, 16, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_16:v[0-9]+]], vcc, 16, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_16]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_20:v[0-9]+]], vcc, 20, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_20:v[0-9]+]], vcc, 20, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_20]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_24:v[0-9]+]], vcc, 24, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_24:v[0-9]+]], vcc, 24, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_24]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_28:v[0-9]+]], vcc, 28, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_28:v[0-9]+]], vcc, 28, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_28]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_32:v[0-9]+]], vcc, 32, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_32:v[0-9]+]], vcc, 32, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_32]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_36:v[0-9]+]], vcc, 36, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_36:v[0-9]+]], vcc, 36, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_36]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_40:v[0-9]+]], vcc, 40, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_40:v[0-9]+]], vcc, 40, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_40]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_44:v[0-9]+]], vcc, 44, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_44:v[0-9]+]], vcc, 44, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_44]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_48:v[0-9]+]], vcc, 48, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_48:v[0-9]+]], vcc, 48, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_48]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_52:v[0-9]+]], vcc, 52, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_52:v[0-9]+]], vcc, 52, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_52]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_56:v[0-9]+]], vcc, 56, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_56:v[0-9]+]], vcc, 56, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_56]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_60:v[0-9]+]], vcc, 60, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_60:v[0-9]+]], vcc, 60, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_60]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_64:v[0-9]+]], vcc, 64, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_64:v[0-9]+]], vcc, 64, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_64]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_68:v[0-9]+]], vcc, 0x44, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_68:v[0-9]+]], vcc, 0x44, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_68]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_72:v[0-9]+]], vcc, 0x48, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_72:v[0-9]+]], vcc, 0x48, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_72]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_76:v[0-9]+]], vcc, 0x4c, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_76:v[0-9]+]], vcc, 0x4c, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_76]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_80:v[0-9]+]], vcc, 0x50, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_80:v[0-9]+]], vcc, 0x50, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_80]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_84:v[0-9]+]], vcc, 0x54, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_84:v[0-9]+]], vcc, 0x54, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_84]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_88:v[0-9]+]], vcc, 0x58, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_88:v[0-9]+]], vcc, 0x58, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_88]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_92:v[0-9]+]], vcc, 0x5c, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_92:v[0-9]+]], vcc, 0x5c, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_92]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_96:v[0-9]+]], vcc, 0x60, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_96:v[0-9]+]], vcc, 0x60, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_96]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_100:v[0-9]+]], vcc, 0x64, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_100:v[0-9]+]], vcc, 0x64, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_100]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_104:v[0-9]+]], vcc, 0x68, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_104:v[0-9]+]], vcc, 0x68, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_104]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_108:v[0-9]+]], vcc, 0x6c, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_108:v[0-9]+]], vcc, 0x6c, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_108]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_112:v[0-9]+]], vcc, 0x70, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_112:v[0-9]+]], vcc, 0x70, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_112]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_116:v[0-9]+]], vcc, 0x74, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_116:v[0-9]+]], vcc, 0x74, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_116]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_120:v[0-9]+]], vcc, 0x78, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_120:v[0-9]+]], vcc, 0x78, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_120]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_124:v[0-9]+]], vcc, 0x7c, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_124:v[0-9]+]], vcc, 0x7c, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_124]], s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_128:v[0-9]+]], vcc, 0x80, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_128:v[0-9]+]], vcc, 0x80, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_128]], s[0:3], s4 offen{{$}}
 
 ; GCN: buffer_load_dword v34
 ; GCN: buffer_load_dword v33
 ; GCN: buffer_load_dword v32
 ; GCN: s_waitcnt vmcnt(0)
 ; GCN-NEXT: s_setpc_b64
 define { <32 x i32>, i32 } @struct_v32i32_i32_func_void() #0 {
   %ptr = load volatile { <32 x i32>, i32 } addrspace(1)*, { <32 x i32>, i32 } addrspace(1)* addrspace(2)* undef
   %val = load { <32 x i32>, i32 }, { <32 x i32>, i32 } addrspace(1)* %ptr
   ret { <32 x i32>, i32 }%val
 }
 
 ; GCN-LABEL: {{^}}struct_i32_v32i32_func_void:
 ; GCN: buffer_store_dword v32, off, s[0:3], s5 offset:8 ; 4-byte Folded Spill
 ; GCN: buffer_store_dword v33, off, s[0:3], s5 offset:4 ; 4-byte Folded Spill
 
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, v0, s[0:3], s4 offen{{$}}
 
-; GCN-DAG: v_add_i32_e32 [[ADD_128:v[0-9]+]], vcc, 0x80, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_128:v[0-9]+]], vcc, 0x80, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_128]], s[0:3], s4 offen{{$}}
 
 
-; GCN-DAG: v_add_i32_e32 [[ADD_256:v[0-9]+]], vcc, 0xfc, v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 [[ADD_256:v[0-9]+]], vcc, 0xfc, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[ADD_256]], s[0:3], s4 offen{{$}}
 
 ; GCN: buffer_load_dword v33
 ; GCN: buffer_load_dword v32
 ; GCN: s_waitcnt vmcnt(0)
 ; GCN-NEXT: s_setpc_b64
 define { i32, <32 x i32> } @struct_i32_v32i32_func_void() #0 {
   %ptr = load volatile { i32, <32 x i32> } addrspace(1)*, { i32, <32 x i32> } addrspace(1)* addrspace(2)* undef
   %val = load { i32, <32 x i32> }, { i32, <32 x i32> } addrspace(1)* %ptr
   ret { i32, <32 x i32> }%val
 }
 
 attributes #0 = { nounwind }

llvm/trunk/test/CodeGen/AMDGPU/llvm.amdgcn.buffer.load.ll

 ;CHECK: s_waitcnt
 define amdgpu_ps <2 x float> @buffer_load_x2(<4 x i32> inreg %rsrc, i32 %idx, i32 %ofs) {
 main_body:
   %data = call <2 x float> @llvm.amdgcn.buffer.load.v2f32(<4 x i32> %rsrc, i32 %idx, i32 %ofs, i1 0, i1 0)
   ret <2 x float> %data
 }
 
 ;CHECK-LABEL: {{^}}buffer_load_negative_offset:
-;CHECK: v_add_i32_e32 [[VOFS:v[0-9]+]], vcc, -16, v0
+;CHECK: v_add_{{[iu]}}32_e32 [[VOFS:v[0-9]+]], vcc, -16, v0
 ;CHECK: buffer_load_dwordx4 v[0:3], [[VOFS]], s[0:3], 0 offen
 define amdgpu_ps <4 x float> @buffer_load_negative_offset(<4 x i32> inreg, i32 %ofs) {
 main_body:
   %ofs.1 = add i32 %ofs, -16
   %data = call <4 x float> @llvm.amdgcn.buffer.load.v4f32(<4 x i32> %0, i32 0, i32 %ofs.1, i1 0, i1 0)
   ret <4 x float> %data
 }
 

llvm/trunk/test/CodeGen/AMDGPU/llvm.amdgcn.init.exec.ll

main_body:
   ret float %s
 }
 
 ; GCN-LABEL: {{^}}reuse_input:
 ; GCN: s_bfe_u32 s1, s0, 0x70013
 ; GCN: s_bfm_b64 exec, s1, 0
 ; GCN: s_cmp_eq_u32 s1, 64
 ; GCN: s_cmov_b64 exec, -1
-; GCN: v_add_i32_e32 v0, vcc, s0, v0
+; GCN: v_add_co_u32_e32 v0, vcc, s0, v0
 define amdgpu_ps float @reuse_input(i32 inreg %count, i32 %a) {
 main_body:
   call void @llvm.amdgcn.init.exec.from.input(i32 %count, i32 19)
   %s = add i32 %a, %count
   %f = sitofp i32 %s to float
   ret float %f
 }
 
 ; GCN-LABEL: {{^}}reuse_input2:
 ; GCN: s_bfe_u32 s1, s0, 0x70013
 ; GCN: s_bfm_b64 exec, s1, 0
 ; GCN: s_cmp_eq_u32 s1, 64
 ; GCN: s_cmov_b64 exec, -1
-; GCN: v_add_i32_e32 v0, vcc, s0, v0
+; GCN: v_add_co_u32_e32 v0, vcc, s0, v0
 define amdgpu_ps float @reuse_input2(i32 inreg %count, i32 %a) {
 main_body:
   %s = add i32 %a, %count
   %f = sitofp i32 %s to float
   call void @llvm.amdgcn.init.exec.from.input(i32 %count, i32 19)
   ret float %f
 }
 
 declare void @llvm.amdgcn.init.exec(i64) #1
 declare void @llvm.amdgcn.init.exec.from.input(i32, i32) #1
 
 attributes #1 = { convergent }

llvm/trunk/test/CodeGen/AMDGPU/llvm.amdgcn.sbfe.ll

define amdgpu_kernel void @bfe_sext_in_reg_i24(i32 addrspace(1)* %out, i32 addrspace(1)* %in) #0 {
   store i32 %ashr, i32 addrspace(1)* %out, align 4
   ret void
 }
 
 ; GCN-LABEL: @simplify_demanded_bfe_sdiv
 ; GCN: buffer_load_dword [[LOAD:v[0-9]+]]
 ; GCN: v_bfe_i32 [[BFE:v[0-9]+]], [[LOAD]], 1, 16
 ; GCN: v_lshrrev_b32_e32 [[TMP0:v[0-9]+]], 31, [[BFE]]
-; GCN: v_add_i32_e32 [[TMP1:v[0-9]+]], vcc, [[TMP0]], [[BFE]]
+; GCN: v_add_{{[iu]}}32_e32 [[TMP1:v[0-9]+]], vcc, [[TMP0]], [[BFE]]
 ; GCN: v_ashrrev_i32_e32 [[TMP2:v[0-9]+]], 1, [[TMP1]]
 ; GCN: buffer_store_dword [[TMP2]]
 define amdgpu_kernel void @simplify_demanded_bfe_sdiv(i32 addrspace(1)* %out, i32 addrspace(1)* %in) #0 {
   %src = load i32, i32 addrspace(1)* %in, align 4
   %bfe = call i32 @llvm.amdgcn.sbfe.i32(i32 %src, i32 1, i32 16)
   %div = sdiv i32 %bfe, 2
   store i32 %div, i32 addrspace(1)* %out, align 4
   ret void

llvm/trunk/test/CodeGen/AMDGPU/llvm.amdgcn.ubfe.ll

define amdgpu_kernel void @bfe_u32_zextload_i8(i32 addrspace(1)* %out, i8 addrspace(1)* %in) #0 {
   %ext = zext i8 %load to i32
   %bfe = call i32 @llvm.amdgcn.ubfe.i32(i32 %ext, i32 0, i32 8)
   store i32 %bfe, i32 addrspace(1)* %out, align 4
   ret void
 }
 
 ; GCN-LABEL: {{^}}bfe_u32_zext_in_reg_i8:
 ; GCN: buffer_load_dword
-; GCN: v_add_i32
+; GCN: v_add_{{[iu]}}32
 ; GCN-NEXT: v_and_b32_e32
 ; FIXME: Should be using s_add_i32
 ; GCN-NOT: {{[^@]}}bfe
 ; GCN: s_endpgm
 define amdgpu_kernel void @bfe_u32_zext_in_reg_i8(i32 addrspace(1)* %out, i32 addrspace(1)* %in) #0 {
   %load = load i32, i32 addrspace(1)* %in, align 4
   %add = add i32 %load, 1
   %ext = and i32 %add, 255
   %bfe = call i32 @llvm.amdgcn.ubfe.i32(i32 %ext, i32 0, i32 8)
   store i32 %bfe, i32 addrspace(1)* %out, align 4
   ret void
 }
 
 ; GCN-LABEL: {{^}}bfe_u32_zext_in_reg_i16:
 ; GCN: buffer_load_dword
-; GCN: v_add_i32
+; GCN: v_add_{{[iu]}}32
 ; GCN-NEXT: v_and_b32_e32
 ; GCN-NOT: {{[^@]}}bfe
 ; GCN: s_endpgm
 define amdgpu_kernel void @bfe_u32_zext_in_reg_i16(i32 addrspace(1)* %out, i32 addrspace(1)* %in) #0 {
   %load = load i32, i32 addrspace(1)* %in, align 4
   %add = add i32 %load, 1
   %ext = and i32 %add, 65535
   %bfe = call i32 @llvm.amdgcn.ubfe.i32(i32 %ext, i32 0, i32 16)
   store i32 %bfe, i32 addrspace(1)* %out, align 4
   ret void
 }
 
 ; GCN-LABEL: {{^}}bfe_u32_zext_in_reg_i8_offset_1:
 ; GCN: buffer_load_dword
-; GCN: v_add_i32
+; GCN: v_add_{{[iu]}}32
 ; GCN: bfe
 ; GCN: s_endpgm
 define amdgpu_kernel void @bfe_u32_zext_in_reg_i8_offset_1(i32 addrspace(1)* %out, i32 addrspace(1)* %in) #0 {
   %load = load i32, i32 addrspace(1)* %in, align 4
   %add = add i32 %load, 1
   %ext = and i32 %add, 255
   %bfe = call i32 @llvm.amdgcn.ubfe.i32(i32 %ext, i32 1, i32 8)
   store i32 %bfe, i32 addrspace(1)* %out, align 4
   ret void
 }
 
 ; GCN-LABEL: {{^}}bfe_u32_zext_in_reg_i8_offset_3:
 ; GCN: buffer_load_dword
-; GCN: v_add_i32
+; GCN: v_add_{{[iu]}}32
 ; GCN-NEXT: v_and_b32_e32 {{v[0-9]+}}, 0xf8
 ; GCN-NEXT: bfe
 ; GCN: s_endpgm
 define amdgpu_kernel void @bfe_u32_zext_in_reg_i8_offset_3(i32 addrspace(1)* %out, i32 addrspace(1)* %in) #0 {
   %load = load i32, i32 addrspace(1)* %in, align 4
   %add = add i32 %load, 1
   %ext = and i32 %add, 255
   %bfe = call i32 @llvm.amdgcn.ubfe.i32(i32 %ext, i32 3, i32 8)
   store i32 %bfe, i32 addrspace(1)* %out, align 4
   ret void
 }
 
 ; GCN-LABEL: {{^}}bfe_u32_zext_in_reg_i8_offset_7:
 ; GCN: buffer_load_dword
-; GCN: v_add_i32
+; GCN: v_add_{{[iu]}}32
 ; GCN-NEXT: v_and_b32_e32 {{v[0-9]+}}, 0x80
 ; GCN-NEXT: bfe
 ; GCN: s_endpgm
 define amdgpu_kernel void @bfe_u32_zext_in_reg_i8_offset_7(i32 addrspace(1)* %out, i32 addrspace(1)* %in) #0 {
   %load = load i32, i32 addrspace(1)* %in, align 4
   %add = add i32 %load, 1
   %ext = and i32 %add, 255
   %bfe = call i32 @llvm.amdgcn.ubfe.i32(i32 %ext, i32 7, i32 8)
   store i32 %bfe, i32 addrspace(1)* %out, align 4
   ret void
 }
 
 ; GCN-LABEL: {{^}}bfe_u32_zext_in_reg_i16_offset_8:
 ; GCN: buffer_load_dword
-; GCN: v_add_i32
+; GCN: v_add_{{[iu]}}32
 ; GCN-NEXT: bfe
 ; GCN: s_endpgm
 define amdgpu_kernel void @bfe_u32_zext_in_reg_i16_offset_8(i32 addrspace(1)* %out, i32 addrspace(1)* %in) #0 {
   %load = load i32, i32 addrspace(1)* %in, align 4
   %add = add i32 %load, 1
   %ext = and i32 %add, 65535
   %bfe = call i32 @llvm.amdgcn.ubfe.i32(i32 %ext, i32 8, i32 8)
   store i32 %bfe, i32 addrspace(1)* %out, align 4

llvm/trunk/test/CodeGen/AMDGPU/pack.v2f16.ll

 
 ; GCN-LABEL: {{^}}v_pack_v2f16_user:
 ; GFX9: global_load_dword [[VAL0:v[0-9]+]]
 ; GFX9: global_load_dword [[VAL1:v[0-9]+]]
 
 ; GFX9: v_and_b32_e32 [[ELT0:v[0-9]+]], 0xffff, [[VAL0]]
 ; GFX9: v_lshl_or_b32 [[PACKED:v[0-9]+]], [[VAL1]], 16, [[ELT0]]
 
-; GFX9: v_add_i32_e32 v{{[0-9]+}}, vcc, 9, [[PACKED]]
+; GFX9: v_add_{{[_coiu]*}}32_e32 v{{[0-9]+}}, vcc, 9, [[PACKED]]
 define amdgpu_kernel void @v_pack_v2f16_user(i32 addrspace(1)* %in0, i32 addrspace(1)* %in1) #0 {
   %tid = call i32 @llvm.amdgcn.workitem.id.x()
   %tid.ext = sext i32 %tid to i64
   %in0.gep = getelementptr inbounds i32, i32 addrspace(1)* %in0, i64 %tid.ext
   %in1.gep = getelementptr inbounds i32, i32 addrspace(1)* %in1, i64 %tid.ext
   %val0 = load volatile i32, i32 addrspace(1)* %in0.gep
   %val1 = load volatile i32, i32 addrspace(1)* %in1.gep
   %lo.i = trunc i32 %val0 to i16

llvm/trunk/test/CodeGen/AMDGPU/pack.v2i16.ll

 
 ; GCN-LABEL: {{^}}v_pack_v2i16_user:
 ; GFX9: global_load_dword [[VAL0:v[0-9]+]]
 ; GFX9: global_load_dword [[VAL1:v[0-9]+]]
 
 ; GFX9: v_and_b32_e32 [[MASKED:v[0-9]+]], 0xffff, [[VAL0]]
 ; GFX9: v_lshl_or_b32 [[PACKED:v[0-9]+]], [[VAL1]], 16, [[MASKED]]
 
-; GFX9: v_add_i32_e32 v{{[0-9]+}}, vcc, 9, [[PACKED]]
+; GFX9: v_add_co_u32_e32 v{{[0-9]+}}, vcc, 9, [[PACKED]]
 define amdgpu_kernel void @v_pack_v2i16_user(i32 addrspace(1)* %in0, i32 addrspace(1)* %in1) #0 {
   %tid = call i32 @llvm.amdgcn.workitem.id.x()
   %tid.ext = sext i32 %tid to i64
   %in0.gep = getelementptr inbounds i32, i32 addrspace(1)* %in0, i64 %tid.ext
   %in1.gep = getelementptr inbounds i32, i32 addrspace(1)* %in1, i64 %tid.ext
   %val0 = load volatile i32, i32 addrspace(1)* %in0.gep
   %val1 = load volatile i32, i32 addrspace(1)* %in1.gep
   %lo = trunc i32 %val0 to i16

llvm/trunk/test/CodeGen/AMDGPU/rotl.i64.ll

entry:
   %3 = or i64 %0, %2
   store i64 %3, i64 addrspace(1)* %in
   ret void
 }
 
 ; BOTH-LABEL: {{^}}v_rotl_i64:
 ; SI-DAG: v_lshl_b64
 ; VI-DAG: v_lshlrev_b64
-; BOTH-DAG: v_sub_i32
+; BOTH-DAG: v_sub_{{[iu]}}32
 ; SI: v_lshr_b64
 ; VI: v_lshrrev_b64
 ; BOTH: v_or_b32
 ; BOTH: v_or_b32
 ; BOTH: s_endpgm
 define amdgpu_kernel void @v_rotl_i64(i64 addrspace(1)* %in, i64 addrspace(1)* %xptr, i64 addrspace(1)* %yptr) {
 entry:
   %x = load i64, i64 addrspace(1)* %xptr, align 8
   %y = load i64, i64 addrspace(1)* %yptr, align 8
   %tmp0 = shl i64 %x, %y
   %tmp1 = sub i64 64, %y
   %tmp2 = lshr i64 %x, %tmp1
   %tmp3 = or i64 %tmp0, %tmp2
   store i64 %tmp3, i64 addrspace(1)* %in, align 8
   ret void
 }

llvm/trunk/test/CodeGen/AMDGPU/rotr.i64.ll

entry:
   %tmp1 = shl i64 %x, %tmp0
   %tmp2 = lshr i64 %x, %y
   %tmp3 = or i64 %tmp1, %tmp2
   store i64 %tmp3, i64 addrspace(1)* %in
   ret void
 }
 
 ; BOTH-LABEL: {{^}}v_rotr_i64:
-; BOTH-DAG: v_sub_i32
+; BOTH-DAG: v_sub_{{[iu]}}32
 ; SI-DAG: v_lshr_b64
 ; SI-DAG: v_lshl_b64
 ; VI-DAG: v_lshrrev_b64
 ; VI-DAG: v_lshlrev_b64
 ; BOTH: v_or_b32
 ; BOTH: v_or_b32
 define amdgpu_kernel void @v_rotr_i64(i64 addrspace(1)* %in, i64 addrspace(1)* %xptr, i64 addrspace(1)* %yptr) {
 entry:

llvm/trunk/test/CodeGen/AMDGPU/saddo.ll

define amdgpu_kernel void @s_saddo_i64(i64 addrspace(1)* %out, i1 addrspace(1)* %carryout, i64 %a, i64 %b) nounwind {
   %val = extractvalue { i64, i1 } %sadd, 0
   %carry = extractvalue { i64, i1 } %sadd, 1
   store i64 %val, i64 addrspace(1)* %out, align 8
   store i1 %carry, i1 addrspace(1)* %carryout
   ret void
 }
 
 ; FUNC-LABEL: {{^}}v_saddo_i64:
-; SI: v_add_i32
+; SI: v_add_{{[iu]}}32
 ; SI: v_addc_u32
 define amdgpu_kernel void @v_saddo_i64(i64 addrspace(1)* %out, i1 addrspace(1)* %carryout, i64 addrspace(1)* %aptr, i64 addrspace(1)* %bptr) nounwind {
   %a = load i64, i64 addrspace(1)* %aptr, align 4
   %b = load i64, i64 addrspace(1)* %bptr, align 4
   %sadd = call { i64, i1 } @llvm.sadd.with.overflow.i64(i64 %a, i64 %b) nounwind
   %val = extractvalue { i64, i1 } %sadd, 0
   %carry = extractvalue { i64, i1 } %sadd, 1
   store i64 %val, i64 addrspace(1)* %out, align 8
   store i1 %carry, i1 addrspace(1)* %carryout
   ret void
 }

llvm/trunk/test/CodeGen/AMDGPU/scratch-buffer.ll

done:
   ret void
 
 }
 
 ; GCN-LABEL: {{^}}legal_offset_fi_offset:
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, v{{[0-9]+}}, s[{{[0-9]+}}:{{[0-9]+}}], s{{[0-9]+}} offen{{$}}
 ; This constant isn't folded, because it has multiple uses.
 ; GCN-DAG: v_mov_b32_e32 [[K8000:v[0-9]+]], 0x8004
-; GCN-DAG: v_add_i32_e32 [[OFFSET:v[0-9]+]], vcc, [[K8000]]
+; GCN-DAG: v_add_{{[iu]}}32_e32 [[OFFSET:v[0-9]+]], vcc, [[K8000]]
 ; GCN: buffer_store_dword v{{[0-9]+}}, [[OFFSET]], s[{{[0-9]+}}:{{[0-9]+}}], s{{[0-9]+}} offen{{$}}
 
 define amdgpu_kernel void @legal_offset_fi_offset(i32 addrspace(1)* %out, i32 %cond, i32 addrspace(1)* %offsets, i32 %if_offset, i32 %else_offset) {
 entry:
   %scratch0 = alloca [8192 x i32]
   %scratch1 = alloca [8192 x i32]
 
   %offset0 = load i32, i32 addrspace(1)* %offsets

 
 done:
   %value = phi i32 [%if_value, %if], [%else_value, %else]
   store i32 %value, i32 addrspace(1)* %out
   ret void
 }
 
 ; GCN-LABEL: {{^}}neg_vaddr_offset_inbounds:
-; GCN: v_add_i32_e32 [[ADD:v[0-9]+]], vcc, 16, v{{[0-9]+}}
+; GCN: v_add_{{[iu]}}32_e32 [[ADD:v[0-9]+]], vcc, 16, v{{[0-9]+}}
 ; GCN: buffer_store_dword v{{[0-9]+}}, [[ADD]], s[{{[0-9]+:[0-9]+}}], s{{[0-9]+}} offen{{$}}
 define amdgpu_kernel void @neg_vaddr_offset_inbounds(i32 %offset) {
 entry:
   %array = alloca [8192 x i32]
   %ptr_offset = add i32 %offset, 4
   %ptr = getelementptr inbounds [8192 x i32], [8192 x i32]* %array, i32 0, i32 %ptr_offset
   store i32 0, i32* %ptr
   ret void
 }
 
 ; GCN-LABEL: {{^}}neg_vaddr_offset:
-; GCN: v_add_i32_e32 [[ADD:v[0-9]+]], vcc, 16, v{{[0-9]+}}
+; GCN: v_add_{{[iu]}}32_e32 [[ADD:v[0-9]+]], vcc, 16, v{{[0-9]+}}
 ; GCN: buffer_store_dword v{{[0-9]+}}, [[ADD]], s[{{[0-9]+:[0-9]+}}], s{{[0-9]+}} offen{{$}}
 define amdgpu_kernel void @neg_vaddr_offset(i32 %offset) {
 entry:
   %array = alloca [8192 x i32]
   %ptr_offset = add i32 %offset, 4
   %ptr = getelementptr [8192 x i32], [8192 x i32]* %array, i32 0, i32 %ptr_offset
   store i32 0, i32* %ptr
   ret void

llvm/trunk/test/CodeGen/AMDGPU/sdiv.ll

 
 ; Multiply by a weird constant to make sure setIntDivIsCheap is
 ; working.
 
 ; FUNC-LABEL: {{^}}slow_sdiv_i32_3435:
 ; SI-DAG: buffer_load_dword [[VAL:v[0-9]+]],
 ; SI-DAG: v_mov_b32_e32 [[MAGIC:v[0-9]+]], 0x98a1930b
 ; SI: v_mul_hi_i32 [[TMP:v[0-9]+]], [[VAL]], [[MAGIC]]
-; SI: v_add_i32
+; SI: v_add_{{[iu]}}32
 ; SI: v_lshrrev_b32
 ; SI: v_ashrrev_i32
-; SI: v_add_i32
+; SI: v_add_{{[iu]}}32
 ; SI: buffer_store_dword
 ; SI: s_endpgm
 define amdgpu_kernel void @slow_sdiv_i32_3435(i32 addrspace(1)* %out, i32 addrspace(1)* %in) {
   %num = load i32, i32 addrspace(1) * %in
   %result = sdiv i32 %num, 3435
   store i32 %result, i32 addrspace(1)* %out
   ret void
 }

llvm/trunk/test/CodeGen/AMDGPU/sdwa-peephole.ll

 ; RUN:  llc -amdgpu-scalarize-global-loads=false  -march=amdgcn -mcpu=fiji -amdgpu-sdwa-peephole=0 -mattr=-fp64-fp16-denormals -verify-machineinstrs < %s | FileCheck -check-prefix=NOSDWA -check-prefix=GCN %s
 ; RUN:  llc -amdgpu-scalarize-global-loads=false  -march=amdgcn -mcpu=fiji -amdgpu-sdwa-peephole -mattr=-fp64-fp16-denormals -verify-machineinstrs < %s | FileCheck -check-prefix=VI -check-prefix=SDWA -check-prefix=GCN %s
 ; RUN:  llc -amdgpu-scalarize-global-loads=false  -march=amdgcn -mcpu=gfx900 -amdgpu-sdwa-peephole -mattr=-fp64-fp16-denormals -verify-machineinstrs < %s | FileCheck -check-prefix=GFX9 -check-prefix=SDWA -check-prefix=GCN %s
 
 ; GCN-LABEL: {{^}}add_shr_i32:
 ; NOSDWA: v_lshrrev_b32_e32 v[[DST:[0-9]+]], 16, v{{[0-9]+}}
-; NOSDWA: v_add_i32_e32 v{{[0-9]+}}, vcc, v{{[0-9]+}}, v[[DST]]
-; NOSDWA-NOT: v_add_i32_sdwa
+; NOSDWA: v_add_u32_e32 v{{[0-9]+}}, vcc, v{{[0-9]+}}, v[[DST]]
+; NOSDWA-NOT: v_add_{{[_cou]*}}32_sdwa
 
-; SDWA: v_add_i32_sdwa v{{[0-9]+}}, vcc, v{{[0-9]+}}, v{{[0-9]+}} dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:DWORD src1_sel:WORD_1
+; SDWA: v_add_{{[_cou]*}}32_sdwa v{{[0-9]+}}, vcc, v{{[0-9]+}}, v{{[0-9]+}} dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:DWORD src1_sel:WORD_1
 
 define amdgpu_kernel void @add_shr_i32(i32 addrspace(1)* %out, i32 addrspace(1)* %in) {
   %a = load i32, i32 addrspace(1)* %in, align 4
   %shr = lshr i32 %a, 16
   %add = add i32 %a, %shr
   store i32 %add, i32 addrspace(1)* %out, align 4
   ret void
 }
 
 ; GCN-LABEL: {{^}}sub_shr_i32:
 ; NOSDWA: v_lshrrev_b32_e32 v[[DST:[0-9]+]], 16, v{{[0-9]+}}
-; NOSDWA: v_subrev_i32_e32 v{{[0-9]+}}, vcc, v{{[0-9]+}}, v[[DST]]
-; NOSDWA-NOT: v_subrev_i32_sdwa
+; NOSDWA: v_subrev_u32_e32 v{{[0-9]+}}, vcc, v{{[0-9]+}}, v[[DST]]
+; NOSDWA-NOT: v_subrev_{{[_cou]*}}32_sdwa
 
-; SDWA: v_subrev_i32_sdwa v{{[0-9]+}}, vcc, v{{[0-9]+}}, v{{[0-9]+}} dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:DWORD src1_sel:WORD_1
+; SDWA: v_subrev_{{[_cou]*}}32_sdwa v{{[0-9]+}}, vcc, v{{[0-9]+}}, v{{[0-9]+}} dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:DWORD src1_sel:WORD_1
 
 define amdgpu_kernel void @sub_shr_i32(i32 addrspace(1)* %out, i32 addrspace(1)* %in) {
   %a = load i32, i32 addrspace(1)* %in, align 4
   %shr = lshr i32 %a, 16
   %sub = sub i32 %shr, %a
   store i32 %sub, i32 addrspace(1)* %out, align 4
   ret void
 }
entry:
   %b = load <2 x i16>, <2 x i16> addrspace(1)* %inb, align 4
   %mul = mul <2 x i16> %a, %b
   %mul2 = mul <2 x i16> %mul, %b
   store <2 x i16> %mul2, <2 x i16> addrspace(1)* %out, align 4
   ret void
 }
 
 ; GCN-LABEL: {{^}}add_bb_v2i16:
-; NOSDWA-NOT: v_add_i32_sdwa
+; NOSDWA-NOT: v_add_{{[_cou]*}}32_sdwa
 
-; VI: v_add_i32_sdwa v{{[0-9]+}}, vcc, v{{[0-9]+}}, v{{[0-9]+}} dst_sel:WORD_1 dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:WORD_1
+; VI: v_add_u32_sdwa v{{[0-9]+}}, vcc, v{{[0-9]+}}, v{{[0-9]+}} dst_sel:WORD_1 dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:WORD_1
 
 ; GFX9: v_pk_add_u16 v{{[0-9]+}}, v{{[0-9]+}}, v{{[0-9]+}}
 
 define amdgpu_kernel void @add_bb_v2i16(<2 x i16> addrspace(1)* %out, <2 x i16> addrspace(1)* %ina, <2 x i16> addrspace(1)* %inb) {
 entry:
   %a = load <2 x i16>, <2 x i16> addrspace(1)* %ina, align 4
   %b = load <2 x i16>, <2 x i16> addrspace(1)* %inb, align 4
   br label %add_label

llvm/trunk/test/CodeGen/AMDGPU/sgpr-copy-duplicate-operand.ll

 ; RUN:  llc -amdgpu-scalarize-global-loads=false  -march=amdgcn -verify-machineinstrs< %s | FileCheck -check-prefix=SI %s
 ; RUN:  llc -amdgpu-scalarize-global-loads=false  -march=amdgcn -mcpu=tonga -verify-machineinstrs< %s | FileCheck -check-prefix=SI %s
 
 ; Copy VGPR -> SGPR used twice as an instruction operand, which is then
 ; used in an REG_SEQUENCE that also needs to be handled.
 
 ; SI-LABEL: {{^}}test_dup_operands:
-; SI: v_add_i32_e32
+; SI: v_add_{{[iu]}}32_e32
 define amdgpu_kernel void @test_dup_operands(<2 x i32> addrspace(1)* noalias %out, <2 x i32> addrspace(1)* noalias %in) {
   %a = load <2 x i32>, <2 x i32> addrspace(1)* %in
   %lo = extractelement <2 x i32> %a, i32 0
   %hi = extractelement <2 x i32> %a, i32 1
   %add = add i32 %lo, %lo
   %vec0 = insertelement <2 x i32> undef, i32 %add, i32 0
   %vec1 = insertelement <2 x i32> %vec0, i32 %hi, i32 1
   store <2 x i32> %vec1, <2 x i32> addrspace(1)* %out, align 8
   ret void
 }
 

llvm/trunk/test/CodeGen/AMDGPU/sgpr-copy.ll

 
 ; This test checks that image_sample resource descriptors aren't loaded into
 ; vgprs.  The verifier will fail if this happens.
 ; CHECK-LABEL:{{^}}sample_rsrc
 
 ; CHECK: s_cmp_eq_u32
 ; CHECK: s_cbranch_scc0 [[END:BB[0-9]+_[0-9]+]]
 
-; CHECK: v_add_i32_e32 v[[ADD:[0-9]+]], vcc, 1, v{{[0-9]+}}
+; CHECK: v_add_{{[iu]}}32_e32 v[[ADD:[0-9]+]], vcc, 1, v{{[0-9]+}}
 
 ; [[END]]:
 ; CHECK: image_sample v{{\[[0-9]+:[0-9]+\]}}, v{{\[[0-9]+}}:[[ADD]]{{\]}}
 ; CHECK: s_endpgm
 define amdgpu_ps void @sample_rsrc([6 x <4 x i32>] addrspace(2)* byval %arg, [17 x <4 x i32>] addrspace(2)* byval %arg1, [16 x <4 x i32>] addrspace(2)* byval %arg2, [32 x <8 x i32>] addrspace(2)* byval %arg3, float inreg %arg4, i32 inreg %arg5, <2 x i32> %arg6, <2 x i32> %arg7, <2 x i32> %arg8, <3 x i32> %arg9, <2 x i32> %arg10, <2 x i32> %arg11, <2 x i32> %arg12, float %arg13, float %arg14, float %arg15, float %arg16, float %arg17, float %arg18, i32 %arg19, float %arg20, float %arg21) #0 {
 bb:
   %tmp = getelementptr [17 x <4 x i32>], [17 x <4 x i32>] addrspace(2)* %arg1, i32 0, i32 0
   %tmp22 = load <4 x i32>, <4 x i32> addrspace(2)* %tmp, !tbaa !3

llvm/trunk/test/CodeGen/AMDGPU/shl-add-to-add-shl.ll

 ; RUN: llc -march=amdgcn -mcpu=fiji < %s | FileCheck %s
 
 ; Check transformation shl (or|add x, c2), c1 => or|add (shl x, c1), (c2 << c1)
 ; Only one shift if expected, GEP shall not produce a separate shift
 
 ; CHECK-LABEL: {{^}}add_const_offset:
 ; CHECK: v_lshlrev_b32_e32 v[[SHL:[0-9]+]], 4, v0
-; CHECK: v_add_i32_e32 v[[ADD:[0-9]+]], vcc, 0xc80, v[[SHL]]
+; CHECK: v_add_u32_e32 v[[ADD:[0-9]+]], vcc, 0xc80, v[[SHL]]
 ; CHECK-NOT: v_lshl
-; CHECK: v_add_i32_e32 v[[ADDRLO:[0-9]+]], vcc, s{{[0-9]+}}, v[[ADD]]
+; CHECK: v_add_u32_e32 v[[ADDRLO:[0-9]+]], vcc, s{{[0-9]+}}, v[[ADD]]
 ; CHECK: load_dword v{{[0-9]+}}, v{{\[}}[[ADDRLO]]:
 define amdgpu_kernel void @add_const_offset(i32 addrspace(1)* nocapture %arg) {
 bb:
   %id = tail call i32 @llvm.amdgcn.workitem.id.x()
   %add = add i32 %id, 200
   %shl = shl i32 %add, 2
   %ptr = getelementptr inbounds i32, i32 addrspace(1)* %arg, i32 %shl
   %val = load i32, i32 addrspace(1)* %ptr, align 4
   store i32 %val, i32 addrspace(1)* %arg, align 4
   ret void
 }
 
 ; CHECK-LABEL: {{^}}or_const_offset:
 ; CHECK: v_lshlrev_b32_e32 v[[SHL:[0-9]+]], 4, v0
 ; CHECK: v_or_b32_e32 v[[OR:[0-9]+]], 0x1000, v[[SHL]]
 ; CHECK-NOT: v_lshl
-; CHECK: v_add_i32_e32 v[[ADDRLO:[0-9]+]], vcc, s{{[0-9]+}}, v[[OR]]
+; CHECK: v_add_u32_e32 v[[ADDRLO:[0-9]+]], vcc, s{{[0-9]+}}, v[[OR]]
 ; CHECK: load_dword v{{[0-9]+}}, v{{\[}}[[ADDRLO]]:
 define amdgpu_kernel void @or_const_offset(i32 addrspace(1)* nocapture %arg) {
 bb:
   %id = tail call i32 @llvm.amdgcn.workitem.id.x()
   %add = or i32 %id, 256
   %shl = shl i32 %add, 2
   %ptr = getelementptr inbounds i32, i32 addrspace(1)* %arg, i32 %shl
   %val = load i32, i32 addrspace(1)* %ptr, align 4
   store i32 %val, i32 addrspace(1)* %arg, align 4
   ret void
 }
 
 declare i32 @llvm.amdgcn.workitem.id.x()

llvm/trunk/test/CodeGen/AMDGPU/shl_add_ptr.ll

 }
 
 ; Make sure once the first use is folded into the addressing mode, the
 ; remaining add use goes through the normal shl + add constant fold.
 
 ; GCN-LABEL: {{^}}load_shl_base_lds_1:
 ; GCN: v_lshlrev_b32_e32 [[PTR:v[0-9]+]], 2, {{v[0-9]+}}
 ; GCN: ds_read_b32 [[RESULT:v[0-9]+]], [[PTR]] offset:8
-; GCN: v_add_i32_e32 [[ADDUSE:v[0-9]+]], vcc, 8, v{{[0-9]+}}
+; GCN: v_add_{{[iu]}}32_e32 [[ADDUSE:v[0-9]+]], vcc, 8, v{{[0-9]+}}
 ; GCN-DAG: buffer_store_dword [[RESULT]]
 ; GCN-DAG: buffer_store_dword [[ADDUSE]]
 ; GCN: s_endpgm
 define amdgpu_kernel void @load_shl_base_lds_1(float addrspace(1)* %out, i32 addrspace(1)* %add_use) #0 {
   %tid.x = tail call i32 @llvm.amdgcn.workitem.id.x() #1
   %idx.0 = add nsw i32 %tid.x, 2
   %arrayidx0 = getelementptr inbounds [512 x float], [512 x float] addrspace(3)* @lds0, i32 0, i32 %idx.0
   %val0 = load float, float addrspace(3)* %arrayidx0, align 4
define void @shl_add_ptr_combine_2use_lds(i32 %idx) #0 {
   store volatile i32 10, i32 addrspace(3)* %ptr1
   ret void
 }
 
 ; GCN-LABEL: {{^}}shl_add_ptr_combine_2use_max_lds_offset:
 ; GCN-DAG: v_lshlrev_b32_e32 [[SCALE0:v[0-9]+]], 3, v0
 ; GCN-DAG: v_lshlrev_b32_e32 [[SCALE1:v[0-9]+]], 4, v0
 ; GCN-DAG: ds_write_b32 [[SCALE0]], v{{[0-9]+}} offset:65528
-; GCN-DAG: v_add_i32_e32 [[ADD1:v[0-9]+]], vcc, 0x1fff0, [[SCALE1]]
+; GCN-DAG: v_add_{{[iu]}}32_e32 [[ADD1:v[0-9]+]], vcc, 0x1fff0, [[SCALE1]]
 ; GCN: ds_write_b32 [[ADD1]], v{{[0-9]+$}}
 define void @shl_add_ptr_combine_2use_max_lds_offset(i32 %idx) #0 {
   %idx.add = add nuw i32 %idx, 8191
   %shl0 = shl i32 %idx.add, 3
   %shl1 = shl i32 %idx.add, 4
   %ptr0 = inttoptr i32 %shl0 to i32 addrspace(3)*
   %ptr1 = inttoptr i32 %shl1 to i32 addrspace(3)*
   store volatile i32 9, i32 addrspace(3)* %ptr0
   store volatile i32 10, i32 addrspace(3)* %ptr1
   ret void
 }
 
 ; GCN-LABEL: {{^}}shl_add_ptr_combine_2use_both_max_lds_offset:
-; GCN: v_add_i32_e32 [[ADD:v[0-9]+]], vcc, 0x1000, v0
+; GCN: v_add_{{[iu]}}32_e32 [[ADD:v[0-9]+]], vcc, 0x1000, v0
 ; GCN-DAG: v_lshlrev_b32_e32 [[SCALE0:v[0-9]+]], 4, [[ADD]]
 ; GCN-DAG: v_lshlrev_b32_e32 [[SCALE1:v[0-9]+]], 5, [[ADD]]
 ; GCN-DAG: ds_write_b32 [[SCALE0]], v{{[0-9]+$}}
 ; GCN: ds_write_b32 [[SCALE1]], v{{[0-9]+$}}
 define void @shl_add_ptr_combine_2use_both_max_lds_offset(i32 %idx) #0 {
   %idx.add = add nuw i32 %idx, 4096
   %shl0 = shl i32 %idx.add, 4
   %shl1 = shl i32 %idx.add, 5
define void @shl_add_ptr_combine_2use_private(i16 zeroext %idx.arg) #0 {
   store volatile i32 10, i32* %ptr1
   ret void
 }
 
 ; GCN-LABEL: {{^}}shl_add_ptr_combine_2use_max_private_offset:
 ; GCN-DAG: v_lshlrev_b32_e32 [[SCALE0:v[0-9]+]], 3, v0
 ; GCN-DAG: v_lshlrev_b32_e32 [[SCALE1:v[0-9]+]], 4, v0
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[SCALE0]], s[0:3], s4 offen offset:4088
-; GCN-DAG: v_add_i32_e32 [[ADD:v[0-9]+]], vcc, 0x1ff0, [[SCALE1]]
+; GCN-DAG: v_add_{{[iu]}}32_e32 [[ADD:v[0-9]+]], vcc, 0x1ff0, [[SCALE1]]
 ; GCN: buffer_store_dword v{{[0-9]+}}, [[ADD]], s[0:3], s4 offen{{$}}
 define void @shl_add_ptr_combine_2use_max_private_offset(i16 zeroext %idx.arg) #0 {
   %idx = zext i16 %idx.arg to i32
   %idx.add = add nuw i32 %idx, 511
   %shl0 = shl i32 %idx.add, 3
   %shl1 = shl i32 %idx.add, 4
   %ptr0 = inttoptr i32 %shl0 to i32*
   %ptr1 = inttoptr i32 %shl1 to i32*
   store volatile i32 9, i32* %ptr0
   store volatile i32 10, i32* %ptr1
   ret void
 }
 ; GCN-LABEL: {{^}}shl_add_ptr_combine_2use_both_max_private_offset:
-; GCN: v_add_i32_e32 [[ADD:v[0-9]+]], vcc, 0x100, v0
+; GCN: v_add_{{[iu]}}32_e32 [[ADD:v[0-9]+]], vcc, 0x100, v0
 ; GCN-DAG: v_lshlrev_b32_e32 [[SCALE0:v[0-9]+]], 4, [[ADD]]
 ; GCN-DAG: v_lshlrev_b32_e32 [[SCALE1:v[0-9]+]], 5, [[ADD]]
 ; GCN-DAG: buffer_store_dword v{{[0-9]+}}, [[SCALE0]], s[0:3], s4 offen{{$}}
 ; GCN: buffer_store_dword v{{[0-9]+}}, [[SCALE1]], s[0:3], s4 offen{{$}}
 define void @shl_add_ptr_combine_2use_both_max_private_offset(i16 zeroext %idx.arg) #0 {
   %idx = zext i16 %idx.arg to i32
   %idx.add = add nuw i32 %idx, 256
   %shl0 = shl i32 %idx.add, 4

llvm/trunk/test/CodeGen/AMDGPU/shrink-add-sub-constant.ll

 ; RUN: llc -march=amdgcn -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=SI %s
 ; RUN: llc -march=amdgcn -mcpu=tonga -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=VI %s
 
 ; Test that add/sub with a constant is swapped to sub/add with negated
 ; constant to minimize code size.
 
 ; GCN-LABEL: {{^}}v_test_i32_x_sub_64:
 ; GCN: {{buffer|flat}}_load_dword [[X:v[0-9]+]]
-; GCN: v_subrev_i32_e32 v{{[0-9]+}}, vcc, 64, [[X]]
+; GCN: v_subrev_{{[iu]}}32_e32 v{{[0-9]+}}, vcc, 64, [[X]]
 define amdgpu_kernel void @v_test_i32_x_sub_64(i32 addrspace(1)* %out, i32 addrspace(1)* %in) #0 {
   %tid = call i32 @llvm.amdgcn.workitem.id.x()
   %tid.ext = sext i32 %tid to i64
   %gep = getelementptr inbounds i32, i32 addrspace(1)* %in, i64 %tid.ext
   %gep.out = getelementptr inbounds i32, i32 addrspace(1)* %out, i64 %tid.ext
   %x = load i32, i32 addrspace(1)* %gep
   %result = sub i32 %x, 64
   store i32 %result, i32 addrspace(1)* %gep.out
   ret void
 }
 
 ; GCN-LABEL: {{^}}v_test_i32_x_sub_64_multi_use:
 ; GCN: {{buffer|flat}}_load_dword [[X:v[0-9]+]]
 ; GCN: {{buffer|flat}}_load_dword [[Y:v[0-9]+]]
-; GCN-DAG: v_subrev_i32_e32 v{{[0-9]+}}, vcc, 64, [[X]]
-; GCN-DAG: v_subrev_i32_e32 v{{[0-9]+}}, vcc, 64, [[Y]]
+; GCN-DAG: v_subrev_{{[iu]}}32_e32 v{{[0-9]+}}, vcc, 64, [[X]]
+; GCN-DAG: v_subrev_{{[iu]}}32_e32 v{{[0-9]+}}, vcc, 64, [[Y]]
 define amdgpu_kernel void @v_test_i32_x_sub_64_multi_use(i32 addrspace(1)* %out, i32 addrspace(1)* %in) #0 {
   %tid = call i32 @llvm.amdgcn.workitem.id.x()
   %tid.ext = sext i32 %tid to i64
   %gep = getelementptr inbounds i32, i32 addrspace(1)* %in, i64 %tid.ext
   %gep.out = getelementptr inbounds i32, i32 addrspace(1)* %out, i64 %tid.ext
   %x = load volatile i32, i32 addrspace(1)* %gep
   %y = load volatile i32, i32 addrspace(1)* %gep
   %result0 = sub i32 %x, 64
   %result1 = sub i32 %y, 64
   store volatile i32 %result0, i32 addrspace(1)* %gep.out
   store volatile i32 %result1, i32 addrspace(1)* %gep.out
   ret void
 }
 
 ; GCN-LABEL: {{^}}v_test_i32_64_sub_x:
 ; GCN: {{buffer|flat}}_load_dword [[X:v[0-9]+]]
-; GCN: v_sub_i32_e32 v{{[0-9]+}}, vcc, 64, [[X]]
+; GCN: v_sub_{{[iu]}}32_e32 v{{[0-9]+}}, vcc, 64, [[X]]
 define amdgpu_kernel void @v_test_i32_64_sub_x(i32 addrspace(1)* %out, i32 addrspace(1)* %in) #0 {
   %tid = call i32 @llvm.amdgcn.workitem.id.x()
   %tid.ext = sext i32 %tid to i64
   %gep = getelementptr inbounds i32, i32 addrspace(1)* %in, i64 %tid.ext
   %gep.out = getelementptr inbounds i32, i32 addrspace(1)* %out, i64 %tid.ext
   %x = load i32, i32 addrspace(1)* %gep
   %result = sub i32 64, %x
   store i32 %result, i32 addrspace(1)* %gep.out
   ret void
 }
 
 ; GCN-LABEL: {{^}}v_test_i32_x_sub_65:
 ; GCN: {{buffer|flat}}_load_dword [[X:v[0-9]+]]
-; GCN: v_add_i32_e32 v{{[0-9]+}}, vcc, 0xffffffbf, [[X]]
+; GCN: v_add_{{[iu]}}32_e32 v{{[0-9]+}}, vcc, 0xffffffbf, [[X]]
 define amdgpu_kernel void @v_test_i32_x_sub_65(i32 addrspace(1)* %out, i32 addrspace(1)* %in) #0 {
   %tid = call i32 @llvm.amdgcn.workitem.id.x()
   %tid.ext = sext i32 %tid to i64
   %gep = getelementptr inbounds i32, i32 addrspace(1)* %in, i64 %tid.ext
   %gep.out = getelementptr inbounds i32, i32 addrspace(1)* %out, i64 %tid.ext
   %x = load i32, i32 addrspace(1)* %gep
   %result = sub i32 %x, 65
   store i32 %result, i32 addrspace(1)* %gep.out
   ret void
 }
 
 ; GCN-LABEL: {{^}}v_test_i32_65_sub_x:
 ; GCN: {{buffer|flat}}_load_dword [[X:v[0-9]+]]
-; GCN: v_sub_i32_e32 v{{[0-9]+}}, vcc, 0x41, [[X]]
+; GCN: v_sub_{{[iu]}}32_e32 v{{[0-9]+}}, vcc, 0x41, [[X]]
 define amdgpu_kernel void @v_test_i32_65_sub_x(i32 addrspace(1)* %out, i32 addrspace(1)* %in) #0 {
   %tid = call i32 @llvm.amdgcn.workitem.id.x()
   %tid.ext = sext i32 %tid to i64
   %gep = getelementptr inbounds i32, i32 addrspace(1)* %in, i64 %tid.ext
   %gep.out = getelementptr inbounds i32, i32 addrspace(1)* %out, i64 %tid.ext
   %x = load i32, i32 addrspace(1)* %gep
   %result = sub i32 65, %x
   store i32 %result, i32 addrspace(1)* %gep.out
   ret void
 }
 
 ; GCN-LABEL: {{^}}v_test_i32_x_sub_neg16:
 ; GCN: {{buffer|flat}}_load_dword [[X:v[0-9]+]]
-; GCN: v_add_i32_e32 v{{[0-9]+}}, vcc, 16, [[X]]
+; GCN: v_add_{{[iu]}}32_e32 v{{[0-9]+}}, vcc, 16, [[X]]
 define amdgpu_kernel void @v_test_i32_x_sub_neg16(i32 addrspace(1)* %out, i32 addrspace(1)* %in) #0 {
   %tid = call i32 @llvm.amdgcn.workitem.id.x()
   %tid.ext = sext i32 %tid to i64
   %gep = getelementptr inbounds i32, i32 addrspace(1)* %in, i64 %tid.ext
   %gep.out = getelementptr inbounds i32, i32 addrspace(1)* %out, i64 %tid.ext
   %x = load i32, i32 addrspace(1)* %gep
   %result = sub i32 %x, -16
   store i32 %result, i32 addrspace(1)* %gep.out
   ret void
 }
 
 ; GCN-LABEL: {{^}}v_test_i32_neg16_sub_x:
 ; GCN: {{buffer|flat}}_load_dword [[X:v[0-9]+]]
-; GCN: v_sub_i32_e32 v{{[0-9]+}}, vcc, -16, [[X]]
+; GCN: v_sub_{{[iu]}}32_e32 v{{[0-9]+}}, vcc, -16, [[X]]
 define amdgpu_kernel void @v_test_i32_neg16_sub_x(i32 addrspace(1)* %out, i32 addrspace(1)* %in) #0 {
   %tid = call i32 @llvm.amdgcn.workitem.id.x()
   %tid.ext = sext i32 %tid to i64
   %gep = getelementptr inbounds i32, i32 addrspace(1)* %in, i64 %tid.ext
   %gep.out = getelementptr inbounds i32, i32 addrspace(1)* %out, i64 %tid.ext
   %x = load i32, i32 addrspace(1)* %gep
   %result = sub i32 -16, %x
   store i32 %result, i32 addrspace(1)* %gep.out
   ret void
 }
 
 ; GCN-LABEL: {{^}}v_test_i32_x_sub_neg17:
 ; GCN: {{buffer|flat}}_load_dword [[X:v[0-9]+]]
-; GCN: v_add_i32_e32 v{{[0-9]+}}, vcc, 17, [[X]]
+; GCN: v_add_{{[iu]}}32_e32 v{{[0-9]+}}, vcc, 17, [[X]]
 define amdgpu_kernel void @v_test_i32_x_sub_neg17(i32 addrspace(1)* %out, i32 addrspace(1)* %in) #0 {
   %tid = call i32 @llvm.amdgcn.workitem.id.x()
   %tid.ext = sext i32 %tid to i64
   %gep = getelementptr inbounds i32, i32 addrspace(1)* %in, i64 %tid.ext
   %gep.out = getelementptr inbounds i32, i32 addrspace(1)* %out, i64 %tid.ext
   %x = load i32, i32 addrspace(1)* %gep
   %result = sub i32 %x, -17
   store i32 %result, i32 addrspace(1)* %gep.out
   ret void
 }
 
 ; GCN-LABEL: {{^}}v_test_i32_neg17_sub_x:
 ; GCN: {{buffer|flat}}_load_dword [[X:v[0-9]+]]
-; GCN: v_sub_i32_e32 v{{[0-9]+}}, vcc, 0xffffffef, [[X]]
+; GCN: v_sub_{{[iu]}}32_e32 v{{[0-9]+}}, vcc, 0xffffffef, [[X]]
 define amdgpu_kernel void @v_test_i32_neg17_sub_x(i32 addrspace(1)* %out, i32 addrspace(1)* %in) #0 {
   %tid = call i32 @llvm.amdgcn.workitem.id.x()
   %tid.ext = sext i32 %tid to i64
   %gep = getelementptr inbounds i32, i32 addrspace(1)* %in, i64 %tid.ext
   %gep.out = getelementptr inbounds i32, i32 addrspace(1)* %out, i64 %tid.ext
   %x = load i32, i32 addrspace(1)* %gep
   %result = sub i32 -17, %x
   store i32 %result, i32 addrspace(1)* %gep.out

llvm/trunk/test/CodeGen/AMDGPU/sibling-call.ll

 ; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=fiji -mattr=-flat-for-global -amdgpu-sroa=0 -verify-machineinstrs < %s | FileCheck -enable-var-scope -check-prefixes=GCN,VI,MESA %s
 ; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=hawaii -amdgpu-sroa=0 -verify-machineinstrs < %s | FileCheck -enable-var-scope -check-prefixes=GCN,CI,MESA %s
 ; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx900 -mattr=-flat-for-global -amdgpu-sroa=0 -verify-machineinstrs < %s | FileCheck -enable-var-scope -check-prefixes=GCN,GFX9,VI,MESA %s
 
 ; GCN-LABEL: {{^}}i32_fastcc_i32_i32:
 ; GCN: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
-; GCN-NEXT: v_add_i32_e32 v0, vcc, v1, v0
+; GCN-NEXT: v_add_{{[_coiu]*}}32_e32 v0, vcc, v1, v0
 ; GCN-NEXT: s_setpc_b64
 define fastcc i32 @i32_fastcc_i32_i32(i32 %arg0, i32 %arg1) #1 {
   %add0 = add i32 %arg0, %arg1
   ret i32 %add0
 }
 
 ; GCN-LABEL: {{^}}i32_fastcc_i32_i32_stack_object:
 ; GCN: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
-; GCN: v_add_i32_e32 v0, vcc, v1, v
+; GCN: v_add_{{[_coiu]*}}32_e32 v0, vcc, v1, v
 ; GCN: s_mov_b32 s5, s32
 ; GCN: buffer_store_dword v{{[0-9]+}}, off, s[0:3], s5 offset:24
 ; GCN: s_waitcnt vmcnt(0)
 ; GCN: s_setpc_b64
 ; GCN: ; ScratchSize: 68
 define fastcc i32 @i32_fastcc_i32_i32_stack_object(i32 %arg0, i32 %arg1) #1 {
   %alloca = alloca [16 x i32], align 4
   %gep = getelementptr inbounds [16 x i32], [16 x i32]* %alloca, i32 0, i32 5
entry:
   ret void
 }
 
 ; GCN-LABEL: {{^}}i32_fastcc_i32_byval_i32:
 ; GCN: s_waitcnt
 ; GCN-NEXT: s_mov_b32 s5, s32
 ; GCN-NEXT: buffer_load_dword v1, off, s[0:3], s5 offset:4
 ; GCN-NEXT: s_waitcnt vmcnt(0)
-; GCN-NEXT: v_add_i32_e32 v0, vcc, v1, v0
+; GCN-NEXT: v_add_{{[_coiu]*}}32_e32 v0, vcc, v1, v0
 ; GCN-NEXT: s_setpc_b64 s[30:31]
 define fastcc i32 @i32_fastcc_i32_byval_i32(i32 %arg0, i32* byval align 4 %arg1) #1 {
   %arg1.load = load i32, i32* %arg1, align 4
   %add0 = add i32 %arg0, %arg1.load
   ret i32 %add0
 }
 
 ; Tail call disallowed with byval in parent.
entry:
   %ret = tail call fastcc i32 @i32_fastcc_i32_byval_i32(i32 %a, i32* inttoptr (i32 16 to i32*))
   ret i32 %ret
 }
 
 ; GCN-LABEL: {{^}}i32_fastcc_i32_i32_a32i32:
 ; GCN: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
 ; GCN-DAG: buffer_load_dword [[LOAD_0:v[0-9]+]], off, s[0:3], s5 offset:4
 ; GCN-DAG: buffer_load_dword [[LOAD_1:v[0-9]+]], off, s[0:3], s5 offset:8
-; GCN-DAG: v_add_i32_e32 v0, vcc, v1, v0
-; GCN: v_add_i32_e32 v0, vcc, [[LOAD_0]], v0
-; GCN: v_add_i32_e32 v0, vcc, [[LOAD_1]], v0
+; GCN-DAG: v_add_{{[_coiu]*}}32_e32 v0, vcc, v1, v0
+; GCN: v_add_{{[_coiu]*}}32_e32 v0, vcc, [[LOAD_0]], v0
+; GCN: v_add_{{[_coiu]*}}32_e32 v0, vcc, [[LOAD_1]], v0
 ; GCN-NEXT: s_setpc_b64
 define fastcc i32 @i32_fastcc_i32_i32_a32i32(i32 %arg0, i32 %arg1, [32 x i32] %large) #1 {
   %val_firststack = extractvalue [32 x i32] %large, 30
   %val_laststack = extractvalue [32 x i32] %large, 31
   %add0 = add i32 %arg0, %arg1
   %add1 = add i32 %add0, %val_firststack
   %add2 = add i32 %add1, %val_laststack
   ret i32 %add2

llvm/trunk/test/CodeGen/AMDGPU/sminmax.ll

define amdgpu_kernel void @s_abs_i32(i32 addrspace(1)* %out, i32 %val) nounwind {
   %cond = icmp sgt i32 %val, %neg
   %res = select i1 %cond, i32 %val, i32 %neg
   %res2 = add i32 %res, 2
   store i32 %res2, i32 addrspace(1)* %out, align 4
   ret void
 }
 
 ; FUNC-LABEL: {{^}}v_abs_i32:
-; GCN: v_sub_i32_e32 [[NEG:v[0-9]+]], vcc, 0, [[SRC:v[0-9]+]]
+; GCN: v_sub_{{[iu]}}32_e32 [[NEG:v[0-9]+]], vcc, 0, [[SRC:v[0-9]+]]
 ; GCN: v_max_i32_e32 {{v[0-9]+}}, [[SRC]], [[NEG]]
-; GCN: v_add_i32
+; GCN: v_add_{{[iu]}}32
 
 ; EG: MAX_INT
 define amdgpu_kernel void @v_abs_i32(i32 addrspace(1)* %out, i32 addrspace(1)* %src) nounwind {
   %val = load i32, i32 addrspace(1)* %src, align 4
   %neg = sub i32 0, %val
   %cond = icmp sgt i32 %val, %neg
   %res = select i1 %cond, i32 %val, i32 %neg
   %res2 = add i32 %res, 2
   store i32 %res2, i32 addrspace(1)* %out, align 4
   ret void
 }
 
 ; GCN-LABEL: {{^}}v_abs_i32_repeat_user:
-; GCN: v_sub_i32_e32 [[NEG:v[0-9]+]], vcc, 0, [[SRC:v[0-9]+]]
+; GCN: v_sub_{{[iu]}}32_e32 [[NEG:v[0-9]+]], vcc, 0, [[SRC:v[0-9]+]]
 ; GCN: v_max_i32_e32 [[MAX:v[0-9]+]], [[SRC]], [[NEG]]
 ; GCN: v_mul_lo_i32 v{{[0-9]+}}, [[MAX]], [[MAX]]
 define amdgpu_kernel void @v_abs_i32_repeat_user(i32 addrspace(1)* %out, i32 addrspace(1)* %src) nounwind {
   %val = load i32, i32 addrspace(1)* %src, align 4
   %neg = sub i32 0, %val
   %cond = icmp sgt i32 %val, %neg
   %res = select i1 %cond, i32 %val, i32 %neg
   %mul = mul i32 %res, %res
define amdgpu_kernel void @s_abs_v2i32(<2 x i32> addrspace(1)* %out, <2 x i32> %val) nounwind {
   %cond = icmp sgt <2 x i32> %val, %neg
   %res = select <2 x i1> %cond, <2 x i32> %val, <2 x i32> %neg
   %res2 = add <2 x i32> %res, %t1
   store <2 x i32> %res2, <2 x i32> addrspace(1)* %out, align 4
   ret void
 }
 
 ; FUNC-LABEL: {{^}}v_abs_v2i32:
-; GCN-DAG: v_sub_i32_e32 [[NEG0:v[0-9]+]], vcc, 0, [[SRC0:v[0-9]+]]
-; GCN-DAG: v_sub_i32_e32 [[NEG1:v[0-9]+]], vcc, 0, [[SRC1:v[0-9]+]]
+; GCN-DAG: v_sub_{{[iu]}}32_e32 [[NEG0:v[0-9]+]], vcc, 0, [[SRC0:v[0-9]+]]
+; GCN-DAG: v_sub_{{[iu]}}32_e32 [[NEG1:v[0-9]+]], vcc, 0, [[SRC1:v[0-9]+]]
 
 ; GCN-DAG: v_max_i32_e32 {{v[0-9]+}}, [[SRC0]], [[NEG0]]
 ; GCN-DAG: v_max_i32_e32 {{v[0-9]+}}, [[SRC1]], [[NEG1]]
 
-; GCN: v_add_i32
-; GCN: v_add_i32
+; GCN: v_add_{{[iu]}}32
+; GCN: v_add_{{[iu]}}32
 
 ; EG: MAX_INT
 ; EG: MAX_INT
 define amdgpu_kernel void @v_abs_v2i32(<2 x i32> addrspace(1)* %out, <2 x i32> addrspace(1)* %src) nounwind {
   %z0 = insertelement <2 x i32> undef, i32 0, i32 0
   %z1 = insertelement <2 x i32> %z0, i32 0, i32 1
   %t0 = insertelement <2 x i32> undef, i32 2, i32 0
   %t1 = insertelement <2 x i32> %t0, i32 2, i32 1
define amdgpu_kernel void @s_abs_v4i32(<4 x i32> addrspace(1)* %out, <4 x i32> %val) nounwind {
   %cond = icmp sgt <4 x i32> %val, %neg
   %res = select <4 x i1> %cond, <4 x i32> %val, <4 x i32> %neg
   %res2 = add <4 x i32> %res, %t3
   store <4 x i32> %res2, <4 x i32> addrspace(1)* %out, align 4
   ret void
 }
 
 ; FUNC-LABEL: {{^}}v_abs_v4i32:
-; GCN-DAG: v_sub_i32_e32 [[NEG0:v[0-9]+]], vcc, 0, [[SRC0:v[0-9]+]]
-; GCN-DAG: v_sub_i32_e32 [[NEG1:v[0-9]+]], vcc, 0, [[SRC1:v[0-9]+]]
-; GCN-DAG: v_sub_i32_e32 [[NEG2:v[0-9]+]], vcc, 0, [[SRC2:v[0-9]+]]
-; GCN-DAG: v_sub_i32_e32 [[NEG3:v[0-9]+]], vcc, 0, [[SRC3:v[0-9]+]]
+; GCN-DAG: v_sub_{{[iu]}}32_e32 [[NEG0:v[0-9]+]], vcc, 0, [[SRC0:v[0-9]+]]
+; GCN-DAG: v_sub_{{[iu]}}32_e32 [[NEG1:v[0-9]+]], vcc, 0, [[SRC1:v[0-9]+]]
+; GCN-DAG: v_sub_{{[iu]}}32_e32 [[NEG2:v[0-9]+]], vcc, 0, [[SRC2:v[0-9]+]]
+; GCN-DAG: v_sub_{{[iu]}}32_e32 [[NEG3:v[0-9]+]], vcc, 0, [[SRC3:v[0-9]+]]
 
 ; GCN-DAG: v_max_i32_e32 {{v[0-9]+}}, [[SRC0]], [[NEG0]]
 ; GCN-DAG: v_max_i32_e32 {{v[0-9]+}}, [[SRC1]], [[NEG1]]
 ; GCN-DAG: v_max_i32_e32 {{v[0-9]+}}, [[SRC2]], [[NEG2]]
 ; GCN-DAG: v_max_i32_e32 {{v[0-9]+}}, [[SRC3]], [[NEG3]]
 
-; GCN: v_add_i32
-; GCN: v_add_i32
-; GCN: v_add_i32
-; GCN: v_add_i32
+; GCN: v_add_{{[iu]}}32
+; GCN: v_add_{{[iu]}}32
+; GCN: v_add_{{[iu]}}32
+; GCN: v_add_{{[iu]}}32
 
 ; EG: MAX_INT
 ; EG: MAX_INT
 ; EG: MAX_INT
 ; EG: MAX_INT
 define amdgpu_kernel void @v_abs_v4i32(<4 x i32> addrspace(1)* %out, <4 x i32> addrspace(1)* %src) nounwind {
   %z0 = insertelement <4 x i32> undef, i32 0, i32 0
   %z1 = insertelement <4 x i32> %z0, i32 0, i32 1

llvm/trunk/test/CodeGen/AMDGPU/sminmax.v2i16.ll

 ; RUN:  llc -amdgpu-scalarize-global-loads=false  -march=amdgcn -mcpu=gfx901 -mattr=-flat-for-global -verify-machineinstrs -enable-packed-inlinable-literals < %s | FileCheck -check-prefix=GFX9 -check-prefix=GCN %s
 ; RUN:  llc -amdgpu-scalarize-global-loads=false  -march=amdgcn -mcpu=tonga -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=VI -check-prefix=CIVI -check-prefix=GCN %s
 ; RUN:  llc -amdgpu-scalarize-global-loads=false  -march=amdgcn -mcpu=bonaire -verify-machineinstrs < %s | FileCheck -check-prefix=CI -check-prefix=CIVI -check-prefix=GCN %s
 
 ; GCN-LABEL: {{^}}s_abs_v2i16:
 ; GFX9: s_load_dword [[VAL:s[0-9]+]]
 ; GFX9: v_pk_sub_i16 [[SUB:v[0-9]+]], 0, [[VAL]]
 ; GFX9: v_pk_max_i16 [[MAX:v[0-9]+]], [[VAL]], [[SUB]]
 ; GFX9: v_pk_add_u16 [[ADD:v[0-9]+]], [[MAX]], 2
 
-; VI: v_sub_i32_e32
-; VI-DAG: v_sub_i32_e32
+; VI: v_sub_u32_e32
+; VI-DAG: v_sub_u32_e32
 ; VI: v_max_i32_sdwa v{{[0-9]+}}, sext(v{{[0-9]+}}), sext(v{{[0-9]+}}) dst_sel:WORD_1 dst_unused:UNUSED_PAD src0_sel:WORD_0 src1_sel:WORD_0
 ; VI: v_max_i32_sdwa v{{[0-9]+}}, sext(v{{[0-9]+}}), sext(v{{[0-9]+}}) dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_0 src1_sel:WORD_0
-; VI: v_add_i32_e32
-; VI: v_add_i32_e32
+; VI: v_add_u32_e32
+; VI: v_add_u32_e32
 ; VI: v_or_b32_sdwa v{{[0-9]+}}, v{{[0-9]+}}, v{{[0-9]+}} dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_0 src1_sel:DWORD
 
 ; CI: v_sub_i32_e32
 ; CI-DAG: v_sub_i32_e32
 ; CI: v_bfe_i32
 ; CI-DAG: v_bfe_i32
 ; CI-DAG: v_add_i32_e32
 ; CI-DAG: v_lshlrev_b32_e32 v{{[0-9]+}}, 16

llvm/trunk/test/CodeGen/AMDGPU/smrd.ll

 main_body:
   %off = add i32 %offset, 4095
   %r = call float @llvm.SI.load.const.v4i32(<4 x i32> %desc, i32 %off)
   ret float %r
 }
 
 ; GCN-LABEL: {{^}}smrd_vgpr_offset_imm_too_large:
 ; GCN-NEXT: BB#
-; GCN-NEXT: v_add_i32_e32 v0, vcc, 0x1000, v0
+; GCN-NEXT: v_add_{{[_coiu]*}}32_e32 v0, vcc, 0x1000, v0
 ; GCN-NEXT: buffer_load_dword v{{[0-9]}}, v0, s[0:3], 0 offen ;
 define amdgpu_ps float @smrd_vgpr_offset_imm_too_large(<4 x i32> inreg %desc, i32 %offset) #0 {
 main_body:
   %off = add i32 %offset, 4096
   %r = call float @llvm.SI.load.const.v4i32(<4 x i32> %desc, i32 %off)
   ret float %r
 }
 

llvm/trunk/test/CodeGen/AMDGPU/srem.ll

define amdgpu_kernel void @srem_i32_4(i32 addrspace(1)* %out, i32 addrspace(1)* %in) {
   store i32 %result, i32 addrspace(1)* %out
   ret void
 }
 
 ; FUNC-LABEL: {{^}}srem_i32_7:
 ; SI: v_mov_b32_e32 [[MAGIC:v[0-9]+]], 0x92492493
 ; SI: v_mul_hi_i32 {{v[0-9]+}}, {{v[0-9]+}}, [[MAGIC]]
 ; SI: v_mul_lo_i32
-; SI: v_sub_i32
+; SI: v_sub_{{[iu]}}32
 ; SI: s_endpgm
 define amdgpu_kernel void @srem_i32_7(i32 addrspace(1)* %out, i32 addrspace(1)* %in) {
   %num = load i32, i32 addrspace(1) * %in
   %result = srem i32 %num, 7
   store i32 %result, i32 addrspace(1)* %out
   ret void
 }
 

llvm/trunk/test/CodeGen/AMDGPU/ssubo.ll

define amdgpu_kernel void @s_ssubo_i64(i64 addrspace(1)* %out, i1 addrspace(1)* %carryout, i64 %a, i64 %b) nounwind {
   %val = extractvalue { i64, i1 } %ssub, 0
   %carry = extractvalue { i64, i1 } %ssub, 1
   store i64 %val, i64 addrspace(1)* %out, align 8
   store i1 %carry, i1 addrspace(1)* %carryout
   ret void
 }
 
 ; FUNC-LABEL: {{^}}v_ssubo_i64:
-; SI: v_sub_i32_e32
+; SI: v_sub_{{[iu]}}32_e32
 ; SI: v_subb_u32_e32
 define amdgpu_kernel void @v_ssubo_i64(i64 addrspace(1)* %out, i1 addrspace(1)* %carryout, i64 addrspace(1)* %aptr, i64 addrspace(1)* %bptr) nounwind {
   %a = load i64, i64 addrspace(1)* %aptr, align 4
   %b = load i64, i64 addrspace(1)* %bptr, align 4
   %ssub = call { i64, i1 } @llvm.ssub.with.overflow.i64(i64 %a, i64 %b) nounwind
   %val = extractvalue { i64, i1 } %ssub, 0
   %carry = extractvalue { i64, i1 } %ssub, 1
   store i64 %val, i64 addrspace(1)* %out, align 8
   store i1 %carry, i1 addrspace(1)* %carryout
   ret void
 }

llvm/trunk/test/CodeGen/AMDGPU/store-hi16.ll

entry:
   store i8 %hi, i8 addrspace(1)* %out
   ret void
 }
 
 ; GCN-LABEL: {{^}}store_global_hi_v2i16_max_offset:
 ; GCN: s_waitcnt
 ; GFX9-NEXT: global_store_short_d16_hi v[0:1], v2, off offset:4094
 
-; VI-DAG: v_add_i32_e32
+; VI-DAG: v_add_u32_e32
 ; VI-DAG: v_addc_u32_e32
 ; VI-DAG: v_lshrrev_b32_e32 v2, 16, v2
 
 ; VI: flat_store_short v[0:1], v2{{$}}
 ; GCN-NEXT: s_waitcnt
 ; GCN-NEXT: s_setpc_b64
 define void @store_global_hi_v2i16_max_offset(i16 addrspace(1)* %out, i32 %arg) #0 {
 entry:
   ; FIXME: ABI for pre-gfx9
   %value = bitcast i32 %arg to <2 x i16>
   %hi = extractelement <2 x i16> %value, i32 1
   %gep = getelementptr inbounds i16, i16 addrspace(1)* %out, i64 2047
   store i16 %hi, i16 addrspace(1)* %gep
   ret void
 }
 
 ; GCN-LABEL: {{^}}store_global_hi_v2i16_min_offset:
 ; GCN: s_waitcnt
 ; GFX9-NEXT: global_store_short_d16_hi v[0:1], v2, off offset:-4096{{$}}
 
-; VI-DAG: v_add_i32_e32
+; VI-DAG: v_add_u32_e32
 ; VI-DAG: v_addc_u32_e32
 ; VI-DAG: v_lshrrev_b32_e32 v2, 16, v2
 
 ; VI: flat_store_short v[0:1], v{{[0-9]$}}
 ; GCN-NEXT: s_waitcnt
 ; GCN-NEXT: s_setpc_b64
 define void @store_global_hi_v2i16_min_offset(i16 addrspace(1)* %out, i32 %arg) #0 {
 entry:
   %value = bitcast i32 %arg to <2 x i16>
   %hi = extractelement <2 x i16> %value, i32 1
   %gep = getelementptr inbounds i16, i16 addrspace(1)* %out, i64 -2048
   store i16 %hi, i16 addrspace(1)* %gep
   ret void
 }
 
 ; GCN-LABEL: {{^}}store_global_hi_v2i16_i8_max_offset:
 ; GCN: s_waitcnt
 ; GFX9-NEXT: global_store_byte_d16_hi v[0:1], v2, off offset:4095
 
-; VI-DAG: v_add_i32_e32
+; VI-DAG: v_add_u32_e32
 ; VI-DAG: v_addc_u32_e32
 ; VI-DAG: v_lshrrev_b32_e32 v2, 16, v2
 ; VI: flat_store_byte v[0:1], v{{[0-9]$}}
 
 ; GCN-NEXT: s_waitcnt
 ; GCN-NEXT: s_setpc_b64
 define void @store_global_hi_v2i16_i8_max_offset(i8 addrspace(1)* %out, i32 %arg) #0 {
 entry:
   %value = bitcast i32 %arg to <2 x i16>
   %hi = extractelement <2 x i16> %value, i32 1
   %trunc = trunc i16 %hi to i8
   %gep = getelementptr inbounds i8, i8 addrspace(1)* %out, i64 4095
   store i8 %trunc, i8 addrspace(1)* %gep
   ret void
 }
 
 ; GCN-LABEL: {{^}}store_global_hi_v2i16_i8_min_offset:
 ; GCN: s_waitcnt
 ; GFX9-NEXT: global_store_byte_d16_hi v[0:1], v2, off offset:-4095
 
-; VI-DAG: v_add_i32_e32
+; VI-DAG: v_add_u32_e32
 ; VI-DAG: v_addc_u32_e32
 ; VI-DAG: v_lshrrev_b32_e32 v2, 16, v2
 
 ; VI: flat_store_byte v[0:1], v{{[0-9]$}}
 ; GCN-NEXT: s_waitcnt
 ; GCN-NEXT: s_setpc_b64
 define void @store_global_hi_v2i16_i8_min_offset(i8 addrspace(1)* %out, i32 %arg) #0 {
 entry:
entry:
   store i8 %hi, i8 addrspace(4)* %out
   ret void
 }
 
 ; GCN-LABEL: {{^}}store_flat_hi_v2i16_max_offset:
 ; GCN: s_waitcnt
 ; GFX9-NEXT: flat_store_short_d16_hi v[0:1], v2 offset:4094{{$}}
 
-; VI-DAG: v_add_i32_e32
+; VI-DAG: v_add_u32_e32
 ; VI-DAG: v_addc_u32_e32
 ; VI-DAG: v_lshrrev_b32_e32 v2, 16, v2
 ; VI: flat_store_short v[0:1], v2{{$}}
 ; GCN-NEXT: s_waitcnt
 ; GCN-NEXT: s_setpc_b64
 define void @store_flat_hi_v2i16_max_offset(i16 addrspace(4)* %out, i32 %arg) #0 {
 entry:
   %value = bitcast i32 %arg to <2 x i16>
   %hi = extractelement <2 x i16> %value, i32 1
   %gep = getelementptr inbounds i16, i16 addrspace(4)* %out, i64 2047
   store i16 %hi, i16 addrspace(4)* %gep
   ret void
 }
 
 ; GCN-LABEL: {{^}}store_flat_hi_v2i16_neg_offset:
 ; GCN: s_waitcnt
-; GCN: v_add_i32_e32
-; GCN: v_addc_u32_e32
+; GCN: v_add_{{[_cou]*}}32_e32
+; VI: v_addc_u32_e32
+; GFX9: v_addc_co_u32_e32
 
 ; GFX9-NEXT: flat_store_short_d16_hi v[0:1], v2{{$}}
 ; VI: flat_store_short v[0:1], v2{{$}}
 ; GCN-NEXT: s_waitcnt
 ; GCN-NEXT: s_setpc_b64
 define void @store_flat_hi_v2i16_neg_offset(i16 addrspace(4)* %out, i32 %arg) #0 {
 entry:
   %value = bitcast i32 %arg to <2 x i16>
   %hi = extractelement <2 x i16> %value, i32 1
   %gep = getelementptr inbounds i16, i16 addrspace(4)* %out, i64 -1023
   store i16 %hi, i16 addrspace(4)* %gep
   ret void
 }
 
 ; GCN-LABEL: {{^}}store_flat_hi_v2i16_i8_max_offset:
 ; GCN: s_waitcnt
 ; GFX9-NEXT: flat_store_byte_d16_hi v[0:1], v2 offset:4095{{$}}
 
 ; VI-DAG: v_lshrrev_b32_e32 v2, 16, v2
-; VI-DAG: v_add_i32_e32
+; VI-DAG: v_add_u32_e32
 ; VI-DAG: v_addc_u32_e32
 ; VI: flat_store_byte v[0:1], v2{{$}}
 ; GCN-NEXT: s_waitcnt
 ; GCN-NEXT: s_setpc_b64
 define void @store_flat_hi_v2i16_i8_max_offset(i8 addrspace(4)* %out, i32 %arg) #0 {
 entry:
   %value = bitcast i32 %arg to <2 x i16>
   %hi = extractelement <2 x i16> %value, i32 1
   %trunc = trunc i16 %hi to i8
   %gep = getelementptr inbounds i8, i8 addrspace(4)* %out, i64 4095
   store i8 %trunc, i8 addrspace(4)* %gep
   ret void
 }
 
 ; GCN-LABEL: {{^}}store_flat_hi_v2i16_i8_neg_offset:
 ; GCN: s_waitcnt
-; GCN-DAG: v_add_i32_e32
-; GCN-DAG: v_addc_u32_e32
+; GCN-DAG: v_add_{{[_cou]*}}32_e32
+; VI-DAG: v_addc_u32_e32
+; GFX9-DAG: v_addc_co_u32_e32
 
 ; GFX9-NEXT: flat_store_byte_d16_hi v[0:1], v2{{$}}
 ; VI-DAG: v_lshrrev_b32_e32 v2, 16, v2
 ; VI: flat_store_byte v[0:1], v2{{$}}
 ; GCN-NEXT: s_waitcnt
 ; GCN-NEXT: s_setpc_b64
 define void @store_flat_hi_v2i16_i8_neg_offset(i8 addrspace(4)* %out, i32 %arg) #0 {
 entry:

llvm/trunk/test/CodeGen/AMDGPU/sub.v2i16.ll

define amdgpu_kernel void @s_test_sub_self_v2i16(<2 x i16> addrspace(1)* %out, <2 x i16> addrspace(2)* %in0) #1 {
   store <2 x i16> %add, <2 x i16> addrspace(1)* %out
   ret void
 }
 
 ; FIXME: VI should not scalarize arg access.
 ; GCN-LABEL: {{^}}s_test_sub_v2i16_kernarg:
 ; GFX9: v_pk_sub_i16 v{{[0-9]+}}, s{{[0-9]+}}, v{{[0-9]+}}
 
-; VI: v_subrev_i32_e32
-; VI: v_subrev_i32_sdwa v{{[0-9]+}}, vcc, v{{[0-9]+}}, v{{[0-9]+}} dst_sel:WORD_1 dst_unused:UNUSED_PAD src0_sel:DWORD src1_sel:DWORD
+; VI: v_subrev_u32_e32
+; VI: v_subrev_u32_sdwa v{{[0-9]+}}, vcc, v{{[0-9]+}}, v{{[0-9]+}} dst_sel:WORD_1 dst_unused:UNUSED_PAD src0_sel:DWORD src1_sel:DWORD
 define amdgpu_kernel void @s_test_sub_v2i16_kernarg(<2 x i16> addrspace(1)* %out, <2 x i16> %a, <2 x i16> %b) #1 {
   %add = sub <2 x i16> %a, %b
   store <2 x i16> %add, <2 x i16> addrspace(1)* %out
   ret void
 }
 
 ; GCN-LABEL: {{^}}v_test_sub_v2i16_constant:
 ; GFX9: s_mov_b32 [[CONST:s[0-9]+]], 0x1c8007b{{$}}

llvm/trunk/test/CodeGen/AMDGPU/uaddo.ll

define amdgpu_kernel void @s_uaddo_i64_zext(i64 addrspace(1)* %out, i64 %a, i64 %b) #0 {
   %add2 = add i64 %val, %ext
   store i64 %add2, i64 addrspace(1)* %out, align 8
   ret void
 }
 
 ; FIXME: Could do scalar
 
 ; FUNC-LABEL: {{^}}s_uaddo_i32:
-; GCN: v_add_i32_e32 v{{[0-9]+}}, vcc, s{{[0-9]+}}, v{{[0-9]+}}
+; GCN: v_add_{{[iu]}}32_e32 v{{[0-9]+}}, vcc, s{{[0-9]+}}, v{{[0-9]+}}
 ; GCN: v_cndmask_b32_e64 v{{[0-9]+}}, 0, 1, vcc
 
 ; EG: ADDC_UINT
 ; EG: ADD_INT
 define amdgpu_kernel void @s_uaddo_i32(i32 addrspace(1)* %out, i1 addrspace(1)* %carryout, i32 %a, i32 %b) #0 {
   %uadd = call { i32, i1 } @llvm.uadd.with.overflow.i32(i32 %a, i32 %b)
   %val = extractvalue { i32, i1 } %uadd, 0
   %carry = extractvalue { i32, i1 } %uadd, 1
   store i32 %val, i32 addrspace(1)* %out, align 4
   store i1 %carry, i1 addrspace(1)* %carryout
   ret void
 }
 
 ; FUNC-LABEL: {{^}}v_uaddo_i32:
-; GCN: v_add_i32_e32 v{{[0-9]+}}, vcc, v{{[0-9]+}}, v{{[0-9]+}}
+; GCN: v_add_{{[iu]}}32_e32 v{{[0-9]+}}, vcc, v{{[0-9]+}}, v{{[0-9]+}}
 ; GCN: v_cndmask_b32_e64 v{{[0-9]+}}, 0, 1, vcc
 
 ; EG: ADDC_UINT
 ; EG: ADD_INT
 define amdgpu_kernel void @v_uaddo_i32(i32 addrspace(1)* %out, i1 addrspace(1)* %carryout, i32 addrspace(1)* %a.ptr, i32 addrspace(1)* %b.ptr) #0 {
   %tid = call i32 @llvm.amdgcn.workitem.id.x()
   %tid.ext = sext i32 %tid to i64
   %a.gep = getelementptr inbounds i32, i32 addrspace(1)* %a.ptr
   %b.gep = getelementptr inbounds i32, i32 addrspace(1)* %b.ptr
   %a = load i32, i32 addrspace(1)* %a.gep, align 4
   %b = load i32, i32 addrspace(1)* %b.gep, align 4
   %uadd = call { i32, i1 } @llvm.uadd.with.overflow.i32(i32 %a, i32 %b)
   %val = extractvalue { i32, i1 } %uadd, 0
   %carry = extractvalue { i32, i1 } %uadd, 1
   store i32 %val, i32 addrspace(1)* %out, align 4
   store i1 %carry, i1 addrspace(1)* %carryout
   ret void
 }
 
 ; FUNC-LABEL: {{^}}v_uaddo_i32_novcc:
-; GCN: v_add_i32_e32 v{{[0-9]+}}, vcc, v{{[0-9]+}}, v{{[0-9]+}}
+; GCN: v_add_{{[iu]}}32_e32 v{{[0-9]+}}, vcc, v{{[0-9]+}}, v{{[0-9]+}}
 ; GCN: v_cndmask_b32_e64 v{{[0-9]+}}, 0, 1, vcc
 
 ; EG: ADDC_UINT
 ; EG: ADD_INT
 define amdgpu_kernel void @v_uaddo_i32_novcc(i32 addrspace(1)* %out, i1 addrspace(1)* %carryout, i32 addrspace(1)* %a.ptr, i32 addrspace(1)* %b.ptr) #0 {
   %tid = call i32 @llvm.amdgcn.workitem.id.x()
   %tid.ext = sext i32 %tid to i64
   %a.gep = getelementptr inbounds i32, i32 addrspace(1)* %a.ptr
define amdgpu_kernel void @s_uaddo_i64(i64 addrspace(1)* %out, i1 addrspace(1)* %carryout, i64 %a, i64 %b) #0 {
   %val = extractvalue { i64, i1 } %uadd, 0
   %carry = extractvalue { i64, i1 } %uadd, 1
   store i64 %val, i64 addrspace(1)* %out, align 8
   store i1 %carry, i1 addrspace(1)* %carryout
   ret void
 }
 
 ; FUNC-LABEL: {{^}}v_uaddo_i64:
-; GCN: v_add_i32
+; GCN: v_add_{{[iu]}}32
 ; GCN: v_addc_u32
 
 ; EG: ADDC_UINT
 ; EG: ADD_INT
 define amdgpu_kernel void @v_uaddo_i64(i64 addrspace(1)* %out, i1 addrspace(1)* %carryout, i64 addrspace(1)* %a.ptr, i64 addrspace(1)* %b.ptr) #0 {
   %tid = call i32 @llvm.amdgcn.workitem.id.x()
   %tid.ext = sext i32 %tid to i64
   %a.gep = getelementptr inbounds i64, i64 addrspace(1)* %a.ptr

llvm/trunk/test/CodeGen/AMDGPU/udivrem.ll

 ; EG-DAG: ADD_INT
 ; EG-DAG: SUB_INT
 ; EG-DAG: CNDE_INT
 ; EG-DAG: CNDE_INT
 
 ; SI: v_rcp_iflag_f32_e32 [[RCP:v[0-9]+]]
 ; SI-DAG: v_mul_hi_u32 [[RCP_HI:v[0-9]+]], [[RCP]]
 ; SI-DAG: v_mul_lo_i32 [[RCP_LO:v[0-9]+]], [[RCP]]
-; SI-DAG: v_sub_i32_e32 [[NEG_RCP_LO:v[0-9]+]], vcc, 0, [[RCP_LO]]
+; SI-DAG: v_sub_{{[iu]}}32_e32 [[NEG_RCP_LO:v[0-9]+]], vcc, 0, [[RCP_LO]]
 ; SI: v_cndmask_b32_e64
 ; SI: v_mul_hi_u32 [[E:v[0-9]+]], {{v[0-9]+}}, [[RCP]]
-; SI-DAG: v_add_i32_e32 [[RCP_A_E:v[0-9]+]], vcc, [[E]], [[RCP]]
-; SI-DAG: v_subrev_i32_e32 [[RCP_S_E:v[0-9]+]], vcc, [[E]], [[RCP]]
+; SI-DAG: v_add_{{[iu]}}32_e32 [[RCP_A_E:v[0-9]+]], vcc, [[E]], [[RCP]]
+; SI-DAG: v_subrev_{{[iu]}}32_e32 [[RCP_S_E:v[0-9]+]], vcc, [[E]], [[RCP]]
 ; SI: v_cndmask_b32_e64
 ; SI: v_mul_hi_u32 [[Quotient:v[0-9]+]]
 ; SI: v_mul_lo_i32 [[Num_S_Remainder:v[0-9]+]]
-; SI-DAG: v_add_i32_e32 [[Quotient_A_One:v[0-9]+]], vcc, 1, [[Quotient]]
-; SI-DAG: v_sub_i32_e32 [[Remainder:v[0-9]+]], vcc, {{[vs][0-9]+}}, [[Num_S_Remainder]]
+; SI-DAG: v_add_{{[iu]}}32_e32 [[Quotient_A_One:v[0-9]+]], vcc, 1, [[Quotient]]
+; SI-DAG: v_sub_{{[iu]}}32_e32 [[Remainder:v[0-9]+]], vcc, {{[vs][0-9]+}}, [[Num_S_Remainder]]
 ; SI-DAG: v_cndmask_b32_e64
 ; SI-DAG: v_cndmask_b32_e64
-; SI-DAG: v_subrev_i32_e32 [[Quotient_S_One:v[0-9]+]],
-; SI-DAG: v_subrev_i32_e32 [[Remainder_S_Den:v[0-9]+]],
+; SI-DAG: v_subrev_{{[iu]}}32_e32 [[Quotient_S_One:v[0-9]+]],
+; SI-DAG: v_subrev_{{[iu]}}32_e32 [[Remainder_S_Den:v[0-9]+]],
 ; SI: v_and_b32_e32 [[Tmp1:v[0-9]+]]
 ; SI-DAG: v_cndmask_b32_e64
 ; SI-DAG: v_cndmask_b32_e64
-; SI-DAG: v_add_i32_e32 [[Remainder_A_Den:v[0-9]+]],
+; SI-DAG: v_add_{{[iu]}}32_e32 [[Remainder_A_Den:v[0-9]+]],
 ; SI-DAG: v_cndmask_b32_e64
 ; SI-DAG: v_cndmask_b32_e64
 ; SI: s_endpgm
 define amdgpu_kernel void @test_udivrem(i32 addrspace(1)* %out0, i32 addrspace(1)* %out1, i32 %x, i32 %y) {
   %result0 = udiv i32 %x, %y
   store i32 %result0, i32 addrspace(1)* %out0
   %result1 = urem i32 %x, %y
   store i32 %result1, i32 addrspace(1)* %out1

 
 ; For SI, we used to have checks for the input and output registers
 ; of the instructions, but these are way too fragile.  The division for
 ; the two vector elements can be intermixed which makes it impossible to
 ; accurately check all the operands.
 ; SI-DAG: v_rcp_iflag_f32_e32
 ; SI-DAG: v_mul_hi_u32
 ; SI-DAG: v_mul_lo_i32
-; SI-DAG: v_sub_i32_e32
+; SI-DAG: v_sub_{{[iu]}}32_e32
 ; SI-DAG: v_cndmask_b32_e64
 ; SI-DAG: v_mul_hi_u32
-; SI-DAG: v_add_i32_e32
-; SI-DAG: v_subrev_i32_e32
+; SI-DAG: v_add_{{[iu]}}32_e32
+; SI-DAG: v_subrev_{{[iu]}}32_e32
 ; SI-DAG: v_cndmask_b32_e64
 ; SI-DAG: v_mul_hi_u32
 ; SI-DAG: v_mul_lo_i32
-; SI-DAG: v_subrev_i32_e32
+; SI-DAG: v_subrev_{{[iu]}}32_e32
 ; SI-DAG: v_cndmask_b32_e64
 ; SI-DAG: v_cndmask_b32_e64
 ; SI-DAG: v_and_b32_e32
-; SI-DAG: v_add_i32_e32
-; SI-DAG: v_subrev_i32_e32
+; SI-DAG: v_add_{{[iu]}}32_e32
+; SI-DAG: v_subrev_{{[iu]}}32_e32
 ; SI-DAG: v_cndmask_b32_e64
 ; SI-DAG: v_cndmask_b32_e64
-; SI-DAG: v_add_i32_e32
-; SI-DAG: v_subrev_i32_e32
+; SI-DAG: v_add_{{[iu]}}32_e32
+; SI-DAG: v_subrev_{{[iu]}}32_e32
 ; SI-DAG: v_cndmask_b32_e64
 ; SI-DAG: v_cndmask_b32_e64
 ; SI-DAG: v_rcp_iflag_f32_e32
 ; SI-DAG: v_mul_hi_u32
 ; SI-DAG: v_mul_lo_i32
-; SI-DAG: v_sub_i32_e32
+; SI-DAG: v_sub_{{[iu]}}32_e32
 ; SI-DAG: v_cndmask_b32_e64
 ; SI-DAG: v_mul_hi_u32
-; SI-DAG: v_add_i32_e32
-; SI-DAG: v_subrev_i32_e32
+; SI-DAG: v_add_{{[iu]}}32_e32
+; SI-DAG: v_subrev_{{[iu]}}32_e32
 ; SI-DAG: v_cndmask_b32_e64
 ; SI-DAG: v_mul_hi_u32
 ; SI-DAG: v_mul_lo_i32
-; SI-DAG: v_subrev_i32_e32
+; SI-DAG: v_subrev_{{[iu]}}32_e32
 ; SI-DAG: v_cndmask_b32_e64
 ; SI-DAG: v_cndmask_b32_e64
 ; SI-DAG: v_and_b32_e32
-; SI-DAG: v_add_i32_e32
-; SI-DAG: v_subrev_i32_e32
+; SI-DAG: v_add_{{[iu]}}32_e32
+; SI-DAG: v_subrev_{{[iu]}}32_e32
 ; SI-DAG: v_cndmask_b32_e64
 ; SI-DAG: v_cndmask_b32_e64
-; SI-DAG: v_add_i32_e32
-; SI-DAG: v_subrev_i32_e32
+; SI-DAG: v_add_{{[iu]}}32_e32
+; SI-DAG: v_subrev_{{[iu]}}32_e32
 ; SI-DAG: v_cndmask_b32_e64
 ; SI-DAG: v_cndmask_b32_e64
 ; SI: s_endpgm
 define amdgpu_kernel void @test_udivrem_v2(<2 x i32> addrspace(1)* %out, <2 x i32> %x, <2 x i32> %y) {
   %result0 = udiv <2 x i32> %x, %y
   store <2 x i32> %result0, <2 x i32> addrspace(1)* %out
   %result1 = urem <2 x i32> %x, %y
   store <2 x i32> %result1, <2 x i32> addrspace(1)* %out

 ; EG-DAG: ADD_INT
 ; EG-DAG: SUB_INT
 ; EG-DAG: CNDE_INT
 ; EG-DAG: CNDE_INT
 
 ; SI-DAG: v_rcp_iflag_f32_e32
 ; SI-DAG: v_mul_hi_u32
 ; SI-DAG: v_mul_lo_i32
-; SI-DAG: v_sub_i32_e32
+; SI-DAG: v_sub_{{[iu]}}32_e32
 ; SI-DAG: v_cndmask_b32_e64
 ; SI-DAG: v_mul_hi_u32
-; SI-DAG: v_add_i32_e32
-; SI-DAG: v_subrev_i32_e32
+; SI-DAG: v_add_{{[iu]}}32_e32
+; SI-DAG: v_subrev_{{[iu]}}32_e32
 ; SI-DAG: v_cndmask_b32_e64
 ; SI-DAG: v_mul_hi_u32
 ; SI-DAG: v_mul_lo_i32
-; SI-DAG: v_subrev_i32_e32
+; SI-DAG: v_subrev_{{[iu]}}32_e32
 ; SI-DAG: v_cndmask_b32_e64
 ; SI-DAG: v_cndmask_b32_e64
 ; SI-DAG: v_and_b32_e32
-; SI-DAG: v_add_i32_e32
-; SI-DAG: v_subrev_i32_e32
+; SI-DAG: v_add_{{[iu]}}32_e32
+; SI-DAG: v_subrev_{{[iu]}}32_e32
 ; SI-DAG: v_cndmask_b32_e64
 ; SI-DAG: v_cndmask_b32_e64
-; SI-DAG: v_add_i32_e32
-; SI-DAG: v_subrev_i32_e32
+; SI-DAG: v_add_{{[iu]}}32_e32
+; SI-DAG: v_subrev_{{[iu]}}32_e32
 ; SI-DAG: v_cndmask_b32_e64
 ; SI-DAG: v_cndmask_b32_e64
 ; SI-DAG: v_rcp_iflag_f32_e32
 ; SI-DAG: v_mul_hi_u32
 ; SI-DAG: v_mul_lo_i32
-; SI-DAG: v_sub_i32_e32
+; SI-DAG: v_sub_{{[iu]}}32_e32
 ; SI-DAG: v_cndmask_b32_e64
 ; SI-DAG: v_mul_hi_u32
-; SI-DAG: v_add_i32_e32
-; SI-DAG: v_subrev_i32_e32
+; SI-DAG: v_add_{{[iu]}}32_e32
+; SI-DAG: v_subrev_{{[iu]}}32_e32
 ; SI-DAG: v_cndmask_b32_e64
 ; SI-DAG: v_mul_hi_u32
 ; SI-DAG: v_mul_lo_i32
-; SI-DAG: v_subrev_i32_e32
+; SI-DAG: v_subrev_{{[iu]}}32_e32
 ; SI-DAG: v_cndmask_b32_e64
 ; SI-DAG: v_cndmask_b32_e64
 ; SI-DAG: v_and_b32_e32
-; SI-DAG: v_add_i32_e32
-; SI-DAG: v_subrev_i32_e32
+; SI-DAG: v_add_{{[iu]}}32_e32
+; SI-DAG: v_subrev_{{[iu]}}32_e32
 ; SI-DAG: v_cndmask_b32_e64
 ; SI-DAG: v_cndmask_b32_e64
-; SI-DAG: v_add_i32_e32
-; SI-DAG: v_subrev_i32_e32
+; SI-DAG: v_add_{{[iu]}}32_e32
+; SI-DAG: v_subrev_{{[iu]}}32_e32
 ; SI-DAG: v_cndmask_b32_e64
 ; SI-DAG: v_cndmask_b32_e64
 ; SI-DAG: v_rcp_iflag_f32_e32
 ; SI-DAG: v_mul_hi_u32
 ; SI-DAG: v_mul_lo_i32
-; SI-DAG: v_sub_i32_e32
+; SI-DAG: v_sub_{{[iu]}}32_e32
 ; SI-DAG: v_cndmask_b32_e64
 ; SI-DAG: v_mul_hi_u32
-; SI-DAG: v_add_i32_e32
-; SI-DAG: v_subrev_i32_e32
+; SI-DAG: v_add_{{[iu]}}32_e32
+; SI-DAG: v_subrev_{{[iu]}}32_e32
 ; SI-DAG: v_cndmask_b32_e64
 ; SI-DAG: v_mul_hi_u32
 ; SI-DAG: v_mul_lo_i32
-; SI-DAG: v_subrev_i32_e32
+; SI-DAG: v_subrev_{{[iu]}}32_e32
 ; SI-DAG: v_cndmask_b32_e64
 ; SI-DAG: v_cndmask_b32_e64
 ; SI-DAG: v_and_b32_e32
-; SI-DAG: v_add_i32_e32
-; SI-DAG: v_subrev_i32_e32
+; SI-DAG: v_add_{{[iu]}}32_e32
+; SI-DAG: v_subrev_{{[iu]}}32_e32
 ; SI-DAG: v_cndmask_b32_e64
 ; SI-DAG: v_cndmask_b32_e64
-; SI-DAG: v_add_i32_e32
-; SI-DAG: v_subrev_i32_e32
+; SI-DAG: v_add_{{[iu]}}32_e32
+; SI-DAG: v_subrev_{{[iu]}}32_e32
 ; SI-DAG: v_cndmask_b32_e64
 ; SI-DAG: v_cndmask_b32_e64
 ; SI-DAG: v_rcp_iflag_f32_e32
 ; SI-DAG: v_mul_hi_u32
 ; SI-DAG: v_mul_lo_i32
-; SI-DAG: v_sub_i32_e32
+; SI-DAG: v_sub_{{[iu]}}32_e32
 ; SI-DAG: v_cndmask_b32_e64
 ; SI-DAG: v_mul_hi_u32
-; SI-DAG: v_add_i32_e32
-; SI-DAG: v_subrev_i32_e32
+; SI-DAG: v_add_{{[iu]}}32_e32
+; SI-DAG: v_subrev_{{[iu]}}32_e32
 ; SI-DAG: v_cndmask_b32_e64
 ; SI: s_endpgm
 define amdgpu_kernel void @test_udivrem_v4(<4 x i32> addrspace(1)* %out, <4 x i32> %x, <4 x i32> %y) {
   %result0 = udiv <4 x i32> %x, %y
   store <4 x i32> %result0, <4 x i32> addrspace(1)* %out
   %result1 = urem <4 x i32> %x, %y
   store <4 x i32> %result1, <4 x i32> addrspace(1)* %out
   ret void
 }

llvm/trunk/test/CodeGen/AMDGPU/uint_to_fp.i64.ll

 ; GCN: v_ffbh_u32
 ; GCN: v_ffbh_u32
 ; GCN: v_cndmask
 ; GCN: v_cndmask
 
 ; GCN-DAG: v_cmp_eq_u64
 ; GCN-DAG: v_cmp_gt_u64
 
-; GCN: v_add_i32_e32 [[VR:v[0-9]+]]
+; GCN: v_add_{{[iu]}}32_e32 [[VR:v[0-9]+]]
 ; GCN: v_cvt_f16_f32_e32 [[VR_F16:v[0-9]+]], [[VR]]
 ; GCN: {{buffer|flat}}_store_short {{.*}}[[VR_F16]]
 define amdgpu_kernel void @v_uint_to_fp_i64_to_f16(half addrspace(1)* %out, i64 addrspace(1)* %in) #0 {
   %tid = call i32 @llvm.r600.read.tidig.x()
   %in.gep = getelementptr i64, i64 addrspace(1)* %in, i32 %tid
   %out.gep = getelementptr half, half addrspace(1)* %out, i32 %tid
   %val = load i64, i64 addrspace(1)* %in.gep
   %result = uitofp i64 %val to half

 ; GCN: v_ffbh_u32
 ; GCN: v_ffbh_u32
 ; GCN: v_cndmask
 ; GCN: v_cndmask
 
 ; GCN-DAG: v_cmp_eq_u64
 ; GCN-DAG: v_cmp_gt_u64
 
-; GCN: v_add_i32_e32 [[VR:v[0-9]+]]
+; GCN: v_add_{{[iu]}}32_e32 [[VR:v[0-9]+]]
 ; GCN: {{buffer|flat}}_store_dword {{.*}}[[VR]]
 define amdgpu_kernel void @v_uint_to_fp_i64_to_f32(float addrspace(1)* %out, i64 addrspace(1)* %in) #0 {
   %tid = call i32 @llvm.r600.read.tidig.x()
   %in.gep = getelementptr i64, i64 addrspace(1)* %in, i32 %tid
   %out.gep = getelementptr float, float addrspace(1)* %out, i32 %tid
   %val = load i64, i64 addrspace(1)* %in.gep
   %result = uitofp i64 %val to float
   store float %result, float addrspace(1)* %out.gep

llvm/trunk/test/CodeGen/AMDGPU/uniform-cfg.ll

 
 done:
   %value = phi i32 [0, %if], [1, %else]
   store i32 %value, i32 addrspace(1)* %out
   ret void
 }
 
 ; GCN-LABEL: {{^}}move_to_valu_vgpr_operand_phi:
-; GCN: v_add_i32_e32
+; GCN: v_add_{{[iu]}}32_e32
 ; GCN: ds_write_b32
 define void @move_to_valu_vgpr_operand_phi(i32 addrspace(3)* %out) {
 bb0:
   br label %bb1
 
 bb1:                                              ; preds = %bb3, %bb0
   %tmp0 = phi i32 [ 8, %bb0 ], [ %tmp4, %bb3 ]
   %tmp1 = add nsw i32 %tmp0, -1

llvm/trunk/test/CodeGen/AMDGPU/urem.ll

define amdgpu_kernel void @test_urem_i32(i32 addrspace(1)* %out, i32 addrspace(1)* %in) {
   %result = urem i32 %a, %b
   store i32 %result, i32 addrspace(1)* %out
   ret void
 }
 
 ; FUNC-LABEL: {{^}}test_urem_i32_7:
 ; SI: v_mov_b32_e32 [[MAGIC:v[0-9]+]], 0x24924925
 ; SI: v_mul_hi_u32 [[MAGIC]], {{v[0-9]+}}
-; SI: v_subrev_i32
+; SI: v_subrev_{{[iu]}}32
 ; SI: v_mul_lo_i32
-; SI: v_sub_i32
+; SI: v_sub_{{[iu]}}32
 ; SI: buffer_store_dword
 ; SI: s_endpgm
 define amdgpu_kernel void @test_urem_i32_7(i32 addrspace(1)* %out, i32 addrspace(1)* %in) {
   %num = load i32, i32 addrspace(1) * %in
   %result = urem i32 %num, 7
   store i32 %result, i32 addrspace(1)* %out
   ret void
 }

llvm/trunk/test/CodeGen/AMDGPU/usubo.ll

define amdgpu_kernel void @s_usubo_i64_zext(i64 addrspace(1)* %out, i64 %a, i64 %b) #0 {
   %add2 = add i64 %val, %ext
   store i64 %add2, i64 addrspace(1)* %out, align 8
   ret void
 }
 
 ; FIXME: Could do scalar
 
 ; FUNC-LABEL: {{^}}s_usubo_i32:
-; GCN: v_sub_i32_e32 v{{[0-9]+}}, vcc, s{{[0-9]+}}, v{{[0-9]+}}
+; GCN: v_sub_{{[iu]}}32_e32 v{{[0-9]+}}, vcc, s{{[0-9]+}}, v{{[0-9]+}}
 ; GCN: v_cndmask_b32_e64 v{{[0-9]+}}, 0, 1, vcc
 
 ; EG-DAG: SUBB_UINT
 ; EG-DAG: SUB_INT
 define amdgpu_kernel void @s_usubo_i32(i32 addrspace(1)* %out, i1 addrspace(1)* %carryout, i32 %a, i32 %b) #0 {
   %usub = call { i32, i1 } @llvm.usub.with.overflow.i32(i32 %a, i32 %b)
   %val = extractvalue { i32, i1 } %usub, 0
   %carry = extractvalue { i32, i1 } %usub, 1
   store i32 %val, i32 addrspace(1)* %out, align 4
   store i1 %carry, i1 addrspace(1)* %carryout
   ret void
 }
 
 ; FUNC-LABEL: {{^}}v_usubo_i32:
-; GCN: v_sub_i32_e32 v{{[0-9]+}}, vcc, v{{[0-9]+}}, v{{[0-9]+}}
+; GCN: v_sub_{{[iu]}}32_e32 v{{[0-9]+}}, vcc, v{{[0-9]+}}, v{{[0-9]+}}
 ; GCN: v_cndmask_b32_e64 v{{[0-9]+}}, 0, 1, vcc
 
 ; EG-DAG: SUBB_UINT
 ; EG-DAG: SUB_INT
 define amdgpu_kernel void @v_usubo_i32(i32 addrspace(1)* %out, i1 addrspace(1)* %carryout, i32 addrspace(1)* %a.ptr, i32 addrspace(1)* %b.ptr) #0 {
   %tid = call i32 @llvm.amdgcn.workitem.id.x()
   %tid.ext = sext i32 %tid to i64
   %a.gep = getelementptr inbounds i32, i32 addrspace(1)* %a.ptr
   %b.gep = getelementptr inbounds i32, i32 addrspace(1)* %b.ptr
   %a = load i32, i32 addrspace(1)* %a.gep, align 4
   %b = load i32, i32 addrspace(1)* %b.gep, align 4
   %usub = call { i32, i1 } @llvm.usub.with.overflow.i32(i32 %a, i32 %b)
   %val = extractvalue { i32, i1 } %usub, 0
   %carry = extractvalue { i32, i1 } %usub, 1
   store i32 %val, i32 addrspace(1)* %out, align 4
   store i1 %carry, i1 addrspace(1)* %carryout
   ret void
 }
 
 ; FUNC-LABEL: {{^}}v_usubo_i32_novcc:
-; GCN: v_sub_i32_e32 v{{[0-9]+}}, vcc, v{{[0-9]+}}, v{{[0-9]+}}
+; GCN: v_sub_{{[iu]}}32_e32 v{{[0-9]+}}, vcc, v{{[0-9]+}}, v{{[0-9]+}}
 ; GCN: v_cndmask_b32_e64 v{{[0-9]+}}, 0, 1, vcc
 
 ; EG-DAG: SUBB_UINT
 ; EG-DAG: SUB_INT
 define amdgpu_kernel void @v_usubo_i32_novcc(i32 addrspace(1)* %out, i1 addrspace(1)* %carryout, i32 addrspace(1)* %a.ptr, i32 addrspace(1)* %b.ptr) #0 {
   %tid = call i32 @llvm.amdgcn.workitem.id.x()
   %tid.ext = sext i32 %tid to i64
   %a.gep = getelementptr inbounds i32, i32 addrspace(1)* %a.ptr
define amdgpu_kernel void @s_usubo_i64(i64 addrspace(1)* %out, i1 addrspace(1)* %carryout, i64 %a, i64 %b) #0 {
   %val = extractvalue { i64, i1 } %usub, 0
   %carry = extractvalue { i64, i1 } %usub, 1
   store i64 %val, i64 addrspace(1)* %out, align 8
   store i1 %carry, i1 addrspace(1)* %carryout
   ret void
 }
 
 ; FUNC-LABEL: {{^}}v_usubo_i64:
-; GCN: v_sub_i32
+; GCN: v_sub_{{[iu]}}32
 ; GCN: v_subb_u32
 
 ; EG-DAG: SUBB_UINT
 ; EG-DAG: SUB_INT
 ; EG-DAG: SUB_INT
 ; EG: SUB_INT
 define amdgpu_kernel void @v_usubo_i64(i64 addrspace(1)* %out, i1 addrspace(1)* %carryout, i64 addrspace(1)* %a.ptr, i64 addrspace(1)* %b.ptr) #0 {
   %tid = call i32 @llvm.amdgcn.workitem.id.x()

llvm/trunk/test/CodeGen/AMDGPU/vop-shrink.ll

 ; RUN: llc -march=amdgcn -verify-machineinstrs < %s | FileCheck -check-prefix=SI -check-prefix=FUNC %s
 ; RUN: llc -march=amdgcn -mcpu=tonga -verify-machineinstrs < %s | FileCheck -check-prefix=SI -check-prefix=FUNC %s
 
 ; Test that we correctly commute a sub instruction
 ; FUNC-LABEL: {{^}}sub_rev:
 ; SI-NOT: v_sub_i32_e32 v{{[0-9]+}}, vcc, s
-; SI: v_subrev_i32_e32 v{{[0-9]+}}, vcc, s
+; SI: v_subrev_{{[iu]}}32_e32 v{{[0-9]+}}, vcc, s
 
 ; ModuleID = 'vop-shrink.ll'
 
 define amdgpu_kernel void @sub_rev(i32 addrspace(1)* %out, <4 x i32> %sgpr, i32 %cond) {
 entry:
   %vgpr = call i32 @llvm.amdgcn.workitem.id.x() #1
   %tmp = icmp eq i32 %cond, 0
   br i1 %tmp, label %if, label %else

llvm/trunk/test/CodeGen/AMDGPU/wqm.ll

 }
 
 ; Same as above, but with an integer type.
 ;
 ;CHECK-LABEL: {{^}}test_wwm2:
 ;CHECK: s_or_saveexec_b64 s{{\[[0-9]+:[0-9]+\]}}, -1
 ;CHECK: buffer_load_dword
 ;CHECK: buffer_load_dword
-;CHECK: v_add_i32_e32
+;CHECK: v_add_{{[iu]}}32_e32
 define amdgpu_ps float @test_wwm2(i32 inreg %idx0, i32 inreg %idx1) {
 main_body:
   %src0 = call float @llvm.amdgcn.buffer.load.f32(<4 x i32> undef, i32 %idx0, i32 0, i1 0, i1 0)
   %src1 = call float @llvm.amdgcn.buffer.load.f32(<4 x i32> undef, i32 %idx1, i32 0, i1 0, i1 0)
   %src0.0 = bitcast float %src0 to i32
   %src1.0 = bitcast float %src1 to i32
   %out = add i32 %src0.0, %src1.0
   %out.0 = call i32 @llvm.amdgcn.wwm.i32(i32 %out)

 ; Check that @llvm.amdgcn.set.inactive disables WWM.
 ;
 ;CHECK-LABEL: {{^}}test_set_inactive1:
 ;CHECK: buffer_load_dword
 ;CHECK: s_not_b64 exec, exec
 ;CHECK: v_mov_b32_e32
 ;CHECK: s_not_b64 exec, exec
 ;CHECK: s_or_saveexec_b64 s{{\[[0-9]+:[0-9]+\]}}, -1
-;CHECK: v_add_i32_e32
+;CHECK: v_add_{{[iu]}}32_e32
 define amdgpu_ps void @test_set_inactive1(i32 inreg %idx) {
 main_body:
   %src = call float @llvm.amdgcn.buffer.load.f32(<4 x i32> undef, i32 %idx, i32 0, i1 0, i1 0)
   %src.0 = bitcast float %src to i32
   %src.1 = call i32 @llvm.amdgcn.set.inactive.i32(i32 %src.0, i32 0)
   %out = add i32 %src.1, %src.1
   %out.0 = call i32 @llvm.amdgcn.wwm.i32(i32 %out)
   %out.1 = bitcast i32 %out.0 to float

llvm/trunk/test/MC/AMDGPU/add-sub-no-carry.s

-// RUN: llvm-mc -arch=amdgcn -mcpu=gfx900 -show-encoding %s | FileCheck -check-prefixes=GCN,GFX9 %s
-// RUN: not llvm-mc -arch=amdgcn -mcpu=fiji -show-encoding %s | FileCheck -check-prefixes=GCN,VI %s
+// RUN: llvm-mc -arch=amdgcn -mcpu=gfx900 -show-encoding %s | FileCheck -check-prefixes=GFX9 %s
 
 // RUN: not llvm-mc -arch=amdgcn -mcpu=fiji %s 2>&1 | FileCheck -check-prefixes=ERR-SICIVI %s
 // RUN: not llvm-mc -arch=amdgcn -mcpu=bonaire %s 2>&1 | FileCheck -check-prefixes=ERR-SICIVI %s
 // FIXME: pre-gfx9 errors should be more useful
 
 
-// FIXME: These should parse to VOP2 encoding
 v_add_u32 v1, v2, v3
-// GFX9: v_add_u32_e64 v1, v2, v3        ; encoding: [0x01,0x00,0x34,0xd1,0x02,0x07,0x02,0x00]
-// ERR-SICIVI: :15: error: invalid operand for instruction
+// GFX9: v_add_u32_e32 v1, v2, v3       ; encoding: [0x02,0x07,0x02,0x68]
+// ERR-SICIVI: error: instruction not supported on this GPU
 
 v_add_u32 v1, v2, s1
 // GFX9: v_add_u32_e64 v1, v2, s1        ; encoding: [0x01,0x00,0x34,0xd1,0x02,0x03,0x00,0x00]
-// ERR-SICIVI: :15: error: invalid operand for instruction
+// ERR-SICIVI: error: instruction not supported on this GPU
 
 v_add_u32 v1, s1, v2
-// GFX9: v_add_u32_e64 v1, s1, v2        ; encoding: [0x01,0x00,0x34,0xd1,0x01,0x04,0x02,0x00]
-// ERR-SICIVI: :15: error: invalid operand for instruction
+// GFX9: v_add_u32_e32 v1, s1, v2        ; encoding: [0x01,0x04,0x02,0x68]
+// ERR-SICIVI: error: instruction not supported on this GPU
 
 v_add_u32 v1, 4.0, v2
-// GFX9: v_add_u32_e64 v1, 4.0, v2       ; encoding: [0x01,0x00,0x34,0xd1,0xf6,0x04,0x02,0x00]
-// ERR-SICIVI: :15: error: invalid operand for instruction
+// GFX9: v_add_u32_e32 v1, 4.0, v2       ; encoding: [0xf6,0x04,0x02,0x68]
+// ERR-SICIVI: error: instruction not supported on this GPU
 
 v_add_u32 v1, v2, 4.0
 // GFX9: v_add_u32_e64 v1, v2, 4.0       ; encoding: [0x01,0x00,0x34,0xd1,0x02,0xed,0x01,0x00]
-// ERR-SICIVI: :15: error: invalid operand for instruction
+// ERR-SICIVI: error: instruction not supported on this GPU
 
 v_add_u32_e32 v1, v2, v3
 // GFX9: v_add_u32_e32 v1, v2, v3        ; encoding: [0x02,0x07,0x02,0x68]
-// ERR-SICIVI: :19: error: invalid operand for instruction
+// ERR-SICIVI: error: instruction not supported on this GPU
 
 v_add_u32_e32 v1, s1, v3
 // GFX9: v_add_u32_e32 v1, s1, v3        ; encoding: [0x01,0x06,0x02,0x68]
-// ERR-SICIVI: :19: error: invalid operand for instruction
+// ERR-SICIVI: error: instruction not supported on this GPU
 
 
 
 v_sub_u32 v1, v2, v3
-// GFX9: v_sub_u32_e64 v1, v2, v3        ; encoding: [0x01,0x00,0x35,0xd1,0x02,0x07,0x02,0x00]
-// ERR-SICIVI: :15: error: invalid operand for instruction
+// GFX9: v_sub_u32_e32 v1, v2, v3        ; encoding: [0x02,0x07,0x02,0x6a]
+// ERR-SICIVI: error: instruction not supported on this GPU
 
 v_sub_u32 v1, v2, s1
 // GFX9: v_sub_u32_e64 v1, v2, s1        ; encoding: [0x01,0x00,0x35,0xd1,0x02,0x03,0x00,0x00]
-// ERR-SICIVI: :15: error: invalid operand for instruction
+// ERR-SICIVI: error: instruction not supported on this GPU
 
 v_sub_u32 v1, s1, v2
-// GFX9: v_sub_u32_e64 v1, s1, v2        ; encoding: [0x01,0x00,0x35,0xd1,0x01,0x04,0x02,0x00]
-// ERR-SICIVI: :15: error: invalid operand for instruction
+// GFX9: v_sub_u32_e32 v1, s1, v2        ; encoding: [0x01,0x04,0x02,0x6a]
+// ERR-SICIVI: error: instruction not supported on this GPU
 
 v_sub_u32 v1, 4.0, v2
-// GFX9: v_sub_u32_e64 v1, 4.0, v2       ; encoding: [0x01,0x00,0x35,0xd1,0xf6,0x04,0x02,0x00]
-// ERR-SICIVI: :15: error: invalid operand for instruction
+// GFX9: v_sub_u32_e32 v1, 4.0, v2       ; encoding: [0xf6,0x04,0x02,0x6a]
+// ERR-SICIVI: error: instruction not supported on this GPU
 
 v_sub_u32 v1, v2, 4.0
 // GFX9: v_sub_u32_e64 v1, v2, 4.0       ; encoding: [0x01,0x00,0x35,0xd1,0x02,0xed,0x01,0x00]
-// ERR-SICIVI: :15: error: invalid operand for instruction
+// ERR-SICIVI: error: instruction not supported on this GPU
 
 v_sub_u32_e32 v1, v2, v3
 // GFX9: v_sub_u32_e32 v1, v2, v3        ; encoding: [0x02,0x07,0x02,0x6a]
-// ERR-SICIVI: :19: error: invalid operand for instruction
+// ERR-SICIVI: error: instruction not supported on this GPU
 
 v_sub_u32_e32 v1, s1, v3
 // GFX9: v_sub_u32_e32 v1, s1, v3        ; encoding: [0x01,0x06,0x02,0x6a]
-// ERR-SICIVI: :19: error: invalid operand for instruction
+// ERR-SICIVI: error: instruction not supported on this GPU
 
 
 
 v_subrev_u32 v1, v2, v3
-// GFX9: v_subrev_u32_e64 v1, v2, v3     ; encoding: [0x01,0x00,0x36,0xd1,0x02,0x07,0x02,0x00]
-// ERR-SICIVI: :18: error: invalid operand for instruction
+// GFX9: v_subrev_u32_e32 v1, v2, v3     ; encoding: [0x02,0x07,0x02,0x6c]
+// ERR-SICIVI: error: instruction not supported on this GPU
 
 v_subrev_u32 v1, v2, s1
 // GFX9: v_subrev_u32_e64 v1, v2, s1     ; encoding: [0x01,0x00,0x36,0xd1,0x02,0x03,0x00,0x00]
-// ERR-SICIVI: :18: error: invalid operand for instruction
+// ERR-SICIVI: error: instruction not supported on this GPU
 
 v_subrev_u32 v1, s1, v2
-// GFX9: v_subrev_u32_e64 v1, s1, v2     ; encoding: [0x01,0x00,0x36,0xd1,0x01,0x04,0x02,0x00]
-// ERR-SICIVI: :18: error: invalid operand for instruction
+// GFX9: v_subrev_u32_e32 v1, s1, v2     ; encoding: [0x01,0x04,0x02,0x6c]
+// ERR-SICIVI: error: instruction not supported on this GPU
 
 v_subrev_u32 v1, 4.0, v2
-// GFX9: v_subrev_u32_e64 v1, 4.0, v2    ; encoding: [0x01,0x00,0x36,0xd1,0xf6,0x04,0x02,0x00]
-// ERR-SICIVI: :18: error: invalid operand for instruction
+// GFX9: v_subrev_u32_e32 v1, 4.0, v2    ; encoding: [0xf6,0x04,0x02,0x6c]
+// ERR-SICIVI: error: instruction not supported on this GPU
 
 v_subrev_u32 v1, v2, 4.0
 // GFX9: v_subrev_u32_e64 v1, v2, 4.0    ; encoding: [0x01,0x00,0x36,0xd1,0x02,0xed,0x01,0x00]
-// ERR-SICIVI: :18: error: invalid operand for instruction
+// ERR-SICIVI: error: instruction not supported on this GPU
 
 v_subrev_u32_e32 v1, v2, v3
 // GFX9: v_subrev_u32_e32 v1, v2, v3     ; encoding: [0x02,0x07,0x02,0x6c]
-// ERR-SICIVI: :22: error: invalid operand for instruction
+// ERR-SICIVI: error: instruction not supported on this GPU
 
 v_subrev_u32_e32 v1, s1, v3
 // GFX9: v_subrev_u32_e32 v1, s1, v3     ; encoding: [0x01,0x06,0x02,0x6c]
-// ERR-SICIVI: :22: error: invalid operand for instruction
-
-
-
-v_add_u32 v1, vcc, v2, v3
-// GCN: v_add_i32_e32 v1, vcc, v2, v3   ; encoding: [0x02,0x07,0x02,0x32]
-
-v_add_u32 v1, s[0:1], v2, v3
-// GCN: v_add_i32_e64 v1, s[0:1], v2, v3 ; encoding: [0x01,0x00,0x19,0xd1,0x02,0x07,0x02,0x00]
+// ERR-SICIVI: error: instruction not supported on this GPU

llvm/trunk/test/MC/AMDGPU/regression/bug28413.s

 // SICI: v_cmpx_eq_u32_e32 vcc, 0x40480000, v0 ; encoding: [0xff,0x00,0xa4,0x7d,0x00,0x00,0x48,0x40]
 // VI: v_cmpx_eq_u32_e32 vcc, 0x40480000, v0 ; encoding: [0xff,0x00,0xb4,0x7d,0x00,0x00,0x48,0x40]
 
 v_mov_b32_e32 v0, 0.5
 // GCN: v_mov_b32_e32 v0, 0.5 ; encoding: [0xf0,0x02,0x00,0x7e]
 
 v_mov_b32_e32 v0, 3.125
 // GCN: v_mov_b32_e32 v0, 0x40480000 ; encoding: [0xff,0x02,0x00,0x7e,0x00,0x00,0x48,0x40]
-
-v_add_i32 v0, vcc, 0.5, v0
-// SICI: v_add_i32_e32 v0, vcc, 0.5, v0 ; encoding: [0xf0,0x00,0x00,0x4a]
-// VI: v_add_i32_e32 v0, vcc, 0.5, v0 ; encoding: [0xf0,0x00,0x00,0x32]
-
-v_add_i32 v0, vcc, 3.125, v0
-// SICI: v_add_i32_e32 v0, vcc, 0x40480000, v0 ; encoding: [0xff,0x00,0x00,0x4a,0x00,0x00,0x48,0x40]
-// VI: v_add_i32_e32 v0, vcc, 0x40480000, v0 ; encoding: [0xff,0x00,0x00,0x32,0x00,0x00,0x48,0x40]
 

llvm/trunk/test/MC/AMDGPU/vop2.s

 // SICI: v_mul_i32_i24_e64 v1, s2, 3 ; encoding: [0x01,0x00,0x12,0xd2,0x02,0x06,0x01,0x00]
 v_mul_i32_i24_e64 v1, s2, 3
 
 // src0 inline src1 sgpr
 // SICI: v_mul_i32_i24_e64 v1, 3, s3 ; encoding: [0x01,0x00,0x12,0xd2,0x83,0x06,0x00,0x00]
 v_mul_i32_i24_e64 v1, 3, s3
 
 // SICI: v_add_i32_e32 v0, vcc, 0.5, v0 ; encoding: [0xf0,0x00,0x00,0x4a]
-// VI: v_add_i32_e32 v0, vcc, 0.5, v0 ; encoding: [0xf0,0x00,0x00,0x32]
+// NOVI: error: instruction not supported on this GPU
 v_add_i32_e32 v0, vcc, 0.5, v0
 
 // SICI: v_add_i32_e32 v0, vcc, 0x40480000, v0 ; encoding: [0xff,0x00,0x00,0x4a,0x00,0x00,0x48,0x40]
-// VI: v_add_i32_e32 v0, vcc, 0x40480000, v0 ; encoding: [0xff,0x00,0x00,0x32,0x00,0x00,0x48,0x40]
+// NOVI: error: instruction not supported on this GPU
 v_add_i32_e32 v0, vcc, 3.125, v0
 
 //===----------------------------------------------------------------------===//
 // Instructions
 //===----------------------------------------------------------------------===//
 
 // GCN: v_cndmask_b32_e32 v1, v2, v3, vcc ; encoding: [0x02,0x07,0x02,0x00]
 v_cndmask_b32 v1, v2, v3, vcc

 // VI:   v_mbcnt_lo_u32_b32 v1, v2, v3 ; encoding: [0x01,0x00,0x8c,0xd2,0x02,0x07,0x02,0x00]
 v_mbcnt_lo_u32_b32_e64 v1, v2, v3
 
 // SICI: v_mbcnt_hi_u32_b32_e64 v1, v2, v3 ; encoding: [0x01,0x00,0x48,0xd2,0x02,0x07,0x02,0x00]
 // VI:   v_mbcnt_hi_u32_b32 v1, v2, v3 ; encoding: [0x01,0x00,0x8d,0xd2,0x02,0x07,0x02,0x00]
 v_mbcnt_hi_u32_b32_e64 v1, v2, v3
 
 // SICI: v_add_i32_e32 v1, vcc, v2, v3 ; encoding: [0x02,0x07,0x02,0x4a]
-// VI:   v_add_i32_e32 v1, vcc, v2, v3 ; encoding: [0x02,0x07,0x02,0x32]
+// NOVI: error: instruction not supported on this GPU
 v_add_i32_e32 v1, vcc, v2, v3
 
 // SICI: v_add_i32_e64 v1, s[0:1], v2, v3 ; encoding: [0x01,0x00,0x4a,0xd2,0x02,0x07,0x02,0x00]
-// VI:   v_add_i32_e64 v1, s[0:1], v2, v3 ; encoding: [0x01,0x00,0x19,0xd1,0x02,0x07,0x02,0x00]
+// NOVI: error: instruction not supported on this GPU
 v_add_i32 v1, s[0:1], v2, v3
 
 // SICI: v_add_i32_e64 v1, s[0:1], v2, v3 ; encoding: [0x01,0x00,0x4a,0xd2,0x02,0x07,0x02,0x00]
-// VI:   v_add_i32_e64 v1, s[0:1], v2, v3 ; encoding: [0x01,0x00,0x19,0xd1,0x02,0x07,0x02,0x00]
+// NOVI: error: instruction not supported on this GPU
 v_add_i32_e64 v1, s[0:1], v2, v3
 
 // SICI: v_add_i32_e64 v1, vcc, v2, v3 ; encoding: [0x01,0x6a,0x4a,0xd2,0x02,0x07,0x02,0x00]
-// VI:   v_add_i32_e64 v1, vcc, v2, v3 ; encoding: [0x01,0x6a,0x19,0xd1,0x02,0x07,0x02,0x00]
+// NOVI: error: instruction not supported on this GPU
 v_add_i32_e64 v1, vcc, v2, v3
 
-// SICI: v_add_i32_e32 v1, vcc, v2, v3 ; encoding: [0x02,0x07,0x02,0x4a]
-// VI:   v_add_i32_e32 v1, vcc, v2, v3 ; encoding: [0x02,0x07,0x02,0x32]
+// NOSICI: error: instruction not supported on this GPU
+// VI: v_add_u32_e32 v1, vcc, v2, v3   ; encoding: [0x02,0x07,0x02,0x32]
 v_add_u32 v1, vcc, v2, v3
 
-// SICI: v_add_i32_e64 v1, s[0:1], v2, v3 ; encoding: [0x01,0x00,0x4a,0xd2,0x02,0x07,0x02,0x00]
-// VI:   v_add_i32_e64 v1, s[0:1], v2, v3 ; encoding: [0x01,0x00,0x19,0xd1,0x02,0x07,0x02,0x00]
+// NOSICI: error: instruction not supported on this GPU
+// VI: v_add_u32_e64 v1, s[0:1], v2, v3 ; encoding: [0x01,0x00,0x19,0xd1,0x02,0x07,0x02,0x00]
 v_add_u32 v1, s[0:1], v2, v3
 
 // SICI: v_sub_i32_e32 v1, vcc, v2, v3 ; encoding: [0x02,0x07,0x02,0x4c]
-// VI:   v_sub_i32_e32 v1, vcc, v2, v3 ; encoding: [0x02,0x07,0x02,0x34]
+// NOVI: error: instruction not supported on this GPU
 v_sub_i32 v1, vcc, v2, v3
 
 // SICI: v_sub_i32_e64 v1, s[0:1], v2, v3 ; encoding: [0x01,0x00,0x4c,0xd2,0x02,0x07,0x02,0x00]
-// VI:   v_sub_i32_e64 v1, s[0:1], v2, v3 ; encoding: [0x01,0x00,0x1a,0xd1,0x02,0x07,0x02,0x00]
+// NOVI: error: instruction not supported on this GPU
 v_sub_i32 v1, s[0:1], v2, v3
 
-// SICI: v_sub_i32_e32 v1, vcc, v2, v3 ; encoding: [0x02,0x07,0x02,0x4c]
-// VI:   v_sub_i32_e32 v1, vcc, v2, v3 ; encoding: [0x02,0x07,0x02,0x34]
+// NOSICI: error: instruction not supported on this GPU
+// VI:   v_sub_u32_e32 v1, vcc, v2, v3 ; encoding: [0x02,0x07,0x02,0x34]
 v_sub_u32 v1, vcc, v2, v3
 
-// SICI: v_sub_i32_e64 v1, s[0:1], v2, v3 ; encoding: [0x01,0x00,0x4c,0xd2,0x02,0x07,0x02,0x00]
-// VI:   v_sub_i32_e64 v1, s[0:1], v2, v3 ; encoding: [0x01,0x00,0x1a,0xd1,0x02,0x07,0x02,0x00]
+// NOSICI: error: instruction not supported on this GPU
+// VI:   v_sub_u32_e64 v1, s[0:1], v2, v3 ; encoding: [0x01,0x00,0x1a,0xd1,0x02,0x07,0x02,0x00]
 v_sub_u32 v1, s[0:1], v2, v3
 
 // SICI: v_subrev_i32_e32 v1, vcc, v2, v3 ; encoding: [0x02,0x07,0x02,0x4e]
-// VI:   v_subrev_i32_e32 v1, vcc, v2, v3 ; encoding: [0x02,0x07,0x02,0x36]
+// NOVI: error: instruction not supported on this GPU
 v_subrev_i32 v1, vcc, v2, v3
 
 // SICI: v_subrev_i32_e64 v1, s[0:1], v2, v3 ; encoding: [0x01,0x00,0x4e,0xd2,0x02,0x07,0x02,0x00]
-// VI:   v_subrev_i32_e64 v1, s[0:1], v2, v3 ; encoding: [0x01,0x00,0x1b,0xd1,0x02,0x07,0x02,0x00]
+// NOVI: error: instruction not supported on this GPU
 v_subrev_i32 v1, s[0:1], v2, v3
 
-// SICI: v_subrev_i32_e32 v1, vcc, v2, v3 ; encoding: [0x02,0x07,0x02,0x4e]
-// VI:   v_subrev_i32_e32 v1, vcc, v2, v3 ; encoding: [0x02,0x07,0x02,0x36]
+// NOSICI: error: instruction not supported on this GPU
+// VI:   v_subrev_u32_e32 v1, vcc, v2, v3 ; encoding: [0x02,0x07,0x02,0x36]
 v_subrev_u32 v1, vcc, v2, v3
 
-// SICI: v_subrev_i32_e64 v1, s[0:1], v2, v3 ; encoding: [0x01,0x00,0x4e,0xd2,0x02,0x07,0x02,0x00]
-// VI:   v_subrev_i32_e64 v1, s[0:1], v2, v3 ; encoding: [0x01,0x00,0x1b,0xd1,0x02,0x07,0x02,0x00]
+// NOSICI: error: instruction not supported on this GPU
+// VI:   v_subrev_u32_e64 v1, s[0:1], v2, v3 ; encoding: [0x01,0x00,0x1b,0xd1,0x02,0x07,0x02,0x00]
 v_subrev_u32 v1, s[0:1], v2, v3
 
 // SICI: v_addc_u32_e32 v1, vcc, v2, v3, vcc ; encoding: [0x02,0x07,0x02,0x50]
 // VI:   v_addc_u32_e32 v1, vcc, v2, v3, vcc ; encoding: [0x02,0x07,0x02,0x38]
 v_addc_u32 v1, vcc, v2, v3, vcc
 
 // SICI: v_addc_u32_e32 v1, vcc, v2, v3, vcc ; encoding: [0x02,0x07,0x02,0x50]
 // VI:   v_addc_u32_e32 v1, vcc, v2, v3, vcc ; encoding: [0x02,0x07,0x02,0x38]

llvm/trunk/test/MC/AMDGPU/vop_dpp.s

 // VI9: v_min_i16_dpp v1, v2, v3 row_shl:1 row_mask:0xa bank_mask:0x1 bound_ctrl:0 ; encoding: [0xfa,0x06,0x02,0x64,0x02,0x01,0x09,0xa1]
 v_min_i16 v1, v2, v3 row_shl:1 row_mask:0xa bank_mask:0x1 bound_ctrl:0
 
 // NOSICI: error:
 // VI9: v_ldexp_f16_dpp v1, v2, v3 row_shl:1 row_mask:0xa bank_mask:0x1 bound_ctrl:0 ; encoding: [0xfa,0x06,0x02,0x66,0x02,0x01,0x09,0xa1]
 v_ldexp_f16 v1, v2, v3 row_shl:1 row_mask:0xa bank_mask:0x1 bound_ctrl:0
 
 // NOSICI: error:
-// VI9: v_add_i32_dpp v1, vcc, v2, v3 row_shl:1 row_mask:0xa bank_mask:0x1 bound_ctrl:0 ; encoding: [0xfa,0x06,0x02,0x32,0x02,0x01,0x09,0xa1]
-v_add_i32 v1, vcc, v2, v3 row_shl:1 row_mask:0xa bank_mask:0x1 bound_ctrl:0
+// NOGFX9: error:
+// VI: v_add_u32_dpp v1, vcc, v2, v3 row_shl:1 row_mask:0xa bank_mask:0x1 bound_ctrl:0 ; encoding: [0xfa,0x06,0x02,0x32,0x02,0x01,0x09,0xa1]
+v_add_u32 v1, vcc, v2, v3 row_shl:1 row_mask:0xa bank_mask:0x1 bound_ctrl:0
 
 // NOSICI: error:
-// VI9: v_sub_i32_dpp v1, vcc, v2, v3 row_shl:1 row_mask:0xa bank_mask:0x1 bound_ctrl:0 ; encoding: [0xfa,0x06,0x02,0x34,0x02,0x01,0x09,0xa1]
-v_sub_i32 v1, vcc, v2, v3 row_shl:1 row_mask:0xa bank_mask:0x1 bound_ctrl:0
+// NOGFX9: error:
+// VI: v_sub_u32_dpp v1, vcc, v2, v3 row_shl:1 row_mask:0xa bank_mask:0x1 bound_ctrl:0 ; encoding: [0xfa,0x06,0x02,0x34,0x02,0x01,0x09,0xa1]
+v_sub_u32 v1, vcc, v2, v3 row_shl:1 row_mask:0xa bank_mask:0x1 bound_ctrl:0
 
 // NOSICI: error:
-// VI9: v_subrev_i32_dpp v1, vcc, v2, v3 row_shl:1 row_mask:0xa bank_mask:0x1 bound_ctrl:0 ; encoding: [0xfa,0x06,0x02,0x36,0x02,0x01,0x09,0xa1]
-v_subrev_i32 v1, vcc, v2, v3 row_shl:1 row_mask:0xa bank_mask:0x1 bound_ctrl:0
+// NOGFX9: error:
+// VI: v_subrev_u32_dpp v1, vcc, v2, v3 row_shl:1 row_mask:0xa bank_mask:0x1 bound_ctrl:0 ; encoding: [0xfa,0x06,0x02,0x36,0x02,0x01,0x09,0xa1]
+v_subrev_u32 v1, vcc, v2, v3 row_shl:1 row_mask:0xa bank_mask:0x1 bound_ctrl:0
 
 // NOSICI: error:
-// VI9: v_addc_u32_dpp v1, vcc, v2, v3, vcc row_shl:1 row_mask:0xa bank_mask:0x1 bound_ctrl:0 ; encoding: [0xfa,0x06,0x02,0x38,0x02,0x01,0x09,0xa1]
+// NOGFX9: error:
+// VI: v_addc_u32_dpp v1, vcc, v2, v3, vcc row_shl:1 row_mask:0xa bank_mask:0x1 bound_ctrl:0 ; encoding: [0xfa,0x06,0x02,0x38,0x02,0x01,0x09,0xa1]
 v_addc_u32 v1, vcc, v2, v3, vcc row_shl:1 row_mask:0xa bank_mask:0x1 bound_ctrl:0
 
 // NOSICI: error:
-// VI9: v_subb_u32_dpp v1, vcc, v2, v3, vcc row_shl:1 row_mask:0xa bank_mask:0x1 bound_ctrl:0 ; encoding: [0xfa,0x06,0x02,0x3a,0x02,0x01,0x09,0xa1]
+// NOGFX9: error:
+// VI: v_subb_u32_dpp v1, vcc, v2, v3, vcc row_shl:1 row_mask:0xa bank_mask:0x1 bound_ctrl:0 ; encoding: [0xfa,0x06,0x02,0x3a,0x02,0x01,0x09,0xa1]
 v_subb_u32 v1, vcc, v2, v3, vcc row_shl:1 row_mask:0xa bank_mask:0x1 bound_ctrl:0
 
 // NOSICI: error:
-// VI9: v_subbrev_u32_dpp v1, vcc, v2, v3, vcc row_shl:1 row_mask:0xa bank_mask:0x1 bound_ctrl:0 ; encoding: [0xfa,0x06,0x02,0x3c,0x02,0x01,0x09,0xa1]
+// NOGFX9: error:
+// VI: v_subbrev_u32_dpp v1, vcc, v2, v3, vcc row_shl:1 row_mask:0xa bank_mask:0x1 bound_ctrl:0 ; encoding: [0xfa,0x06,0x02,0x3c,0x02,0x01,0x09,0xa1]
 v_subbrev_u32 v1, vcc, v2, v3, vcc row_shl:1 row_mask:0xa bank_mask:0x1 bound_ctrl:0
 
+// NOSICI: error:
+// NOVI: error:
+// GFX9: v_add_co_u32_dpp v1, vcc, v2, v3 row_shl:1 row_mask:0xa bank_mask:0x1 bound_ctrl:0 ; encoding: [0xfa,0x06,0x02,0x32,0x02,0x01,0x09,0xa1]
+v_add_co_u32 v1, vcc, v2, v3 row_shl:1 row_mask:0xa bank_mask:0x1 bound_ctrl:0
+
+// NOSICI: error:
+// NOVI: error:
+// GFX9: v_sub_co_u32_dpp v1, vcc, v2, v3 row_shl:1 row_mask:0xa bank_mask:0x1 bound_ctrl:0 ; encoding: [0xfa,0x06,0x02,0x34,0x02,0x01,0x09,0xa1]
+v_sub_co_u32 v1, vcc, v2, v3 row_shl:1 row_mask:0xa bank_mask:0x1 bound_ctrl:0
+
+// NOSICI: error:
+// NOVI: error:
+// GFX9: v_subrev_co_u32_dpp v1, vcc, v2, v3 row_shl:1 row_mask:0xa bank_mask:0x1 bound_ctrl:0 ; encoding: [0xfa,0x06,0x02,0x36,0x02,0x01,0x09,0xa1]
+v_subrev_co_u32 v1, vcc, v2, v3 row_shl:1 row_mask:0xa bank_mask:0x1 bound_ctrl:0
+
+// NOSICI: error:
+// NOVI: error:
+// GFX9: v_addc_co_u32_dpp v1, vcc, v2, v3, vcc row_shl:1 row_mask:0xa bank_mask:0x1 bound_ctrl:0 ; encoding: [0xfa,0x06,0x02,0x38,0x02,0x01,0x09,0xa1]
+v_addc_co_u32 v1, vcc, v2, v3, vcc row_shl:1 row_mask:0xa bank_mask:0x1 bound_ctrl:0
+
+// NOSICI: error:
+// NOVI: error:
+// GFX9: v_subb_co_u32_dpp v1, vcc, v2, v3, vcc row_shl:1 row_mask:0xa bank_mask:0x1 bound_ctrl:0 ; encoding: [0xfa,0x06,0x02,0x3a,0x02,0x01,0x09,0xa1]
+v_subb_co_u32 v1, vcc, v2, v3, vcc row_shl:1 row_mask:0xa bank_mask:0x1 bound_ctrl:0
+
+// NOSICI: error:
+// NOVI: error:
+// GFX9: v_subbrev_co_u32_dpp v1, vcc, v2, v3, vcc row_shl:1 row_mask:0xa bank_mask:0x1 bound_ctrl:0 ; encoding: [0xfa,0x06,0x02,0x3c,0x02,0x01,0x09,0xa1]
+v_subbrev_co_u32 v1, vcc, v2, v3, vcc row_shl:1 row_mask:0xa bank_mask:0x1 bound_ctrl:0
+
 //===----------------------------------------------------------------------===//
 // Check that immideates and scalar regs are not supported
 //===----------------------------------------------------------------------===//
 
 // NOSICI: error:
 // NOVI: error:
 // NOGFX9: error:
 v_mov_b32 v0, 1 row_shl:1 row_mask:0xa bank_mask:0x1 bound_ctrl:0

llvm/trunk/test/MC/AMDGPU/vop_sdwa.s

 // GFX89: v_min_i16_sdwa v1, v2, v3 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:BYTE_2 ; encoding: [0xf9,0x06,0x02,0x64,0x02,0x06,0x05,0x02]
 v_min_i16 v1, v2, v3 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:BYTE_2
 
 // NOSICI: error:
 // GFX89: v_ldexp_f16_sdwa v1, v2, v3 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:BYTE_2 ; encoding: [0xf9,0x06,0x02,0x66,0x02,0x06,0x05,0x02]
 v_ldexp_f16 v1, v2, v3 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:BYTE_2
 
 // NOSICI: error:
-// GFX89: v_add_i32_sdwa v1, vcc, v2, v3 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:BYTE_2 ; encoding: [0xf9,0x06,0x02,0x32,0x02,0x06,0x05,0x02]
-v_add_i32_sdwa v1, vcc, v2, v3 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:BYTE_2
+// NOGFX9: error:
+// VI: v_add_u32_sdwa v1, vcc, v2, v3 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:BYTE_2 ; encoding: [0xf9,0x06,0x02,0x32,0x02,0x06,0x05,0x02]
+v_add_u32_sdwa v1, vcc, v2, v3 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:BYTE_2
 
 // NOSICI: error:
-// GFX89: v_sub_i32_sdwa v1, vcc, v2, v3 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:BYTE_2 ; encoding: [0xf9,0x06,0x02,0x34,0x02,0x06,0x05,0x02]
-v_sub_i32_sdwa v1, vcc, v2, v3 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:BYTE_2
+// NOGFX9: error:
+// VI: v_sub_u32_sdwa v1, vcc, v2, v3 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:BYTE_2 ; encoding: [0xf9,0x06,0x02,0x34,0x02,0x06,0x05,0x02]
+v_sub_u32_sdwa v1, vcc, v2, v3 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:BYTE_2
 
 // NOSICI: error:
-// GFX89: v_subrev_i32_sdwa v1, vcc, v2, v3 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:BYTE_2 ; encoding: [0xf9,0x06,0x02,0x36,0x02,0x06,0x05,0x02]
-v_subrev_i32_sdwa v1, vcc, v2, v3 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:BYTE_2
+// NOGFX9: error:
+// VI: v_subrev_u32_sdwa v1, vcc, v2, v3 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:BYTE_2 ; encoding: [0xf9,0x06,0x02,0x36,0x02,0x06,0x05,0x02]
+v_subrev_u32_sdwa v1, vcc, v2, v3 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:BYTE_2
 
 // NOSICI: error:
-// GFX89: v_addc_u32_sdwa v1, vcc, v2, v3, vcc dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:BYTE_2 ; encoding: [0xf9,0x06,0x02,0x38,0x02,0x06,0x05,0x02]
+// NOGFX9: error:
+// VI: v_addc_u32_sdwa v1, vcc, v2, v3, vcc dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:BYTE_2 ; encoding: [0xf9,0x06,0x02,0x38,0x02,0x06,0x05,0x02]
 v_addc_u32_sdwa v1, vcc, v2, v3, vcc dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:BYTE_2
 
 // NOSICI: error:
-// GFX89: v_subb_u32_sdwa v1, vcc, v2, v3, vcc dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:BYTE_2 ; encoding: [0xf9,0x06,0x02,0x3a,0x02,0x06,0x05,0x02]
+// NOGFX9: error:
+// VI: v_subb_u32_sdwa v1, vcc, v2, v3, vcc dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:BYTE_2 ; encoding: [0xf9,0x06,0x02,0x3a,0x02,0x06,0x05,0x02]
 v_subb_u32_sdwa v1, vcc, v2, v3, vcc dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:BYTE_2
 
 // NOSICI: error:
-// GFX89: v_subbrev_u32_sdwa v1, vcc, v2, v3, vcc dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:BYTE_2 ; encoding: [0xf9,0x06,0x02,0x3c,0x02,0x06,0x05,0x02]
+// NOGFX9: error:
+// VI: v_subbrev_u32_sdwa v1, vcc, v2, v3, vcc dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:BYTE_2 ; encoding: [0xf9,0x06,0x02,0x3c,0x02,0x06,0x05,0x02]
 v_subbrev_u32_sdwa v1, vcc, v2, v3, vcc dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:BYTE_2
 
+// NOSICI: error:
+// NOVI: error:
+// GFX9: v_add_co_u32_sdwa v1, vcc, v2, v3 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:BYTE_2 ; encoding: [0xf9,0x06,0x02,0x32,0x02,0x06,0x05,0x02]
+v_add_co_u32_sdwa v1, vcc, v2, v3 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:BYTE_2
+
+// NOSICI: error:
+// NOVI: error:
+// GFX9: v_sub_co_u32_sdwa v1, vcc, v2, v3 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:BYTE_2 ; encoding: [0xf9,0x06,0x02,0x34,0x02,0x06,0x05,0x02]
+v_sub_co_u32_sdwa v1, vcc, v2, v3 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:BYTE_2
+
+// NOSICI: error:
+// NOVI: error:
+// GFX9: v_subrev_co_u32_sdwa v1, vcc, v2, v3 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:BYTE_2 ; encoding: [0xf9,0x06,0x02,0x36,0x02,0x06,0x05,0x02]
+v_subrev_co_u32_sdwa v1, vcc, v2, v3 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:BYTE_2
+
+// NOSICI: error:
+// NOVI: error:
+// GFX9: v_addc_co_u32_sdwa v1, vcc, v2, v3, vcc dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:BYTE_2 ; encoding: [0xf9,0x06,0x02,0x38,0x02,0x06,0x05,0x02]
+v_addc_co_u32_sdwa v1, vcc, v2, v3, vcc dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:BYTE_2
+
+// NOSICI: error:
+// NOVI: error:
+// GFX9: v_subb_co_u32_sdwa v1, vcc, v2, v3, vcc dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:BYTE_2 ; encoding: [0xf9,0x06,0x02,0x3a,0x02,0x06,0x05,0x02]
+v_subb_co_u32_sdwa v1, vcc, v2, v3, vcc dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:BYTE_2
+
+// NOSICI: error:
+// NOVI: error:
+// GFX9: v_subbrev_co_u32_sdwa v1, vcc, v2, v3, vcc dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:BYTE_2 ; encoding: [0xf9,0x06,0x02,0x3c,0x02,0x06,0x05,0x02]
+v_subbrev_co_u32_sdwa v1, vcc, v2, v3, vcc dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:BYTE_2
+
 //===----------------------------------------------------------------------===//
 // Check VOPC opcodes
 //===----------------------------------------------------------------------===//
 
 // NOSICI: error:
 // VI: v_cmp_eq_f32 vcc, v1, v2 src0_sel:BYTE_2 src1_sel:WORD_0 ; encoding: [0xf9,0x04,0x84,0x7c,0x01,0x16,0x02,0x04]
 // GFX9: v_cmp_eq_f32_sdwa vcc, v1, v2 src0_sel:BYTE_2 src1_sel:WORD_0 ; encoding: [0xf9,0x04,0x84,0x7c,0x01,0x00,0x02,0x04]
 v_cmp_eq_f32_sdwa vcc, v1, v2 src0_sel:BYTE_2 src1_sel:WORD_0

 // NOVI: error:
 // GFX9: v_add_f32_sdwa v0, v0, v0 mul:2 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:BYTE_2 ; encoding: [0xf9,0x00,0x00,0x02,0x00,0x46,0x05,0x02]
 v_add_f32 v0, v0, v0 mul:2 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:BYTE_2
 
 // NOSICI: error:
 // NOVI: error:
 // GFX9: v_add_f32_sdwa v0, v0, v0 clamp div:2 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:BYTE_2 ; encoding: [0xf9,0x00,0x00,0x02,0x00,0xe6,0x05,0x02]
 v_add_f32 v0, v0, v0 clamp div:2 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:BYTE_2
 No newline at end of file

llvm/trunk/test/MC/Disassembler/AMDGPU/vop2_vi.txt

 0x01 0x00 0x8b 0xd2 0x02 0x07 0x02 0x00
 
 # VI:   v_mbcnt_lo_u32_b32 v1, v2, v3 ; encoding: [0x01,0x00,0x8c,0xd2,0x02,0x07,0x02,0x00]
 0x01 0x00 0x8c 0xd2 0x02 0x07 0x02 0x00
 
 # VI:   v_mbcnt_hi_u32_b32 v1, v2, v3 ; encoding: [0x01,0x00,0x8d,0xd2,0x02,0x07,0x02,0x00]
 0x01 0x00 0x8d 0xd2 0x02 0x07 0x02 0x00
 
-# VI:   v_add_i32_e32 v1, vcc, v2, v3 ; encoding: [0x02,0x07,0x02,0x32]
+# VI:   v_add_u32_e32 v1, vcc, v2, v3 ; encoding: [0x02,0x07,0x02,0x32]
 0x02 0x07 0x02 0x32
 
-# VI:   v_add_i32_e64 v1, s[0:1], v2, v3 ; encoding: [0x01,0x00,0x19,0xd1,0x02,0x07,0x02,0x00]
+# VI:   v_add_u32_e64 v1, s[0:1], v2, v3 ; encoding: [0x01,0x00,0x19,0xd1,0x02,0x07,0x02,0x00]
 0x01 0x00 0x19 0xd1 0x02 0x07 0x02 0x00
 
-# VI:   v_add_i32_e64 v1, s[0:1], v2, v3 ; encoding: [0x01,0x00,0x19,0xd1,0x02,0x07,0x02,0x00]
+# VI:   v_add_u32_e64 v1, s[0:1], v2, v3 ; encoding: [0x01,0x00,0x19,0xd1,0x02,0x07,0x02,0x00]
 0x01 0x00 0x19 0xd1 0x02 0x07 0x02 0x00
 
-# VI:   v_add_i32_e64 v1, vcc, v2, v3 ; encoding: [0x01,0x6a,0x19,0xd1,0x02,0x07,0x02,0x00]
+# VI:   v_add_u32_e64 v1, vcc, v2, v3 ; encoding: [0x01,0x6a,0x19,0xd1,0x02,0x07,0x02,0x00]
 0x01 0x6a 0x19 0xd1 0x02 0x07 0x02 0x00
 
-# VI:   v_add_i32_e32 v1, vcc, v2, v3 ; encoding: [0x02,0x07,0x02,0x32]
+# VI:   v_add_u32_e32 v1, vcc, v2, v3 ; encoding: [0x02,0x07,0x02,0x32]
 0x02 0x07 0x02 0x32
 
-# VI:   v_add_i32_e64 v1, s[0:1], v2, v3 ; encoding: [0x01,0x00,0x19,0xd1,0x02,0x07,0x02,0x00]
+# VI:   v_add_u32_e64 v1, s[0:1], v2, v3 ; encoding: [0x01,0x00,0x19,0xd1,0x02,0x07,0x02,0x00]
 0x01 0x00 0x19 0xd1 0x02 0x07 0x02 0x00
 
-# VI:   v_sub_i32_e32 v1, vcc, v2, v3 ; encoding: [0x02,0x07,0x02,0x34]
+# VI:   v_sub_u32_e32 v1, vcc, v2, v3 ; encoding: [0x02,0x07,0x02,0x34]
 0x02 0x07 0x02 0x34
 
-# VI:   v_sub_i32_e64 v1, s[0:1], v2, v3 ; encoding: [0x01,0x00,0x1a,0xd1,0x02,0x07,0x02,0x00]
+# VI:   v_sub_u32_e64 v1, s[0:1], v2, v3 ; encoding: [0x01,0x00,0x1a,0xd1,0x02,0x07,0x02,0x00]
 0x01 0x00 0x1a 0xd1 0x02 0x07 0x02 0x00
 
-# VI:   v_sub_i32_e32 v1, vcc, v2, v3 ; encoding: [0x02,0x07,0x02,0x34]
+# VI:   v_sub_u32_e32 v1, vcc, v2, v3 ; encoding: [0x02,0x07,0x02,0x34]
 0x02 0x07 0x02 0x34
 
-# VI:   v_sub_i32_e64 v1, s[0:1], v2, v3 ; encoding: [0x01,0x00,0x1a,0xd1,0x02,0x07,0x02,0x00]
+# VI:   v_sub_u32_e64 v1, s[0:1], v2, v3 ; encoding: [0x01,0x00,0x1a,0xd1,0x02,0x07,0x02,0x00]
 0x01 0x00 0x1a 0xd1 0x02 0x07 0x02 0x00
 
-# VI:   v_subrev_i32_e32 v1, vcc, v2, v3 ; encoding: [0x02,0x07,0x02,0x36]
+# VI:   v_subrev_u32_e32 v1, vcc, v2, v3 ; encoding: [0x02,0x07,0x02,0x36]
 0x02 0x07 0x02 0x36
 
-# VI:   v_subrev_i32_e64 v1, s[0:1], v2, v3 ; encoding: [0x01,0x00,0x1b,0xd1,0x02,0x07,0x02,0x00]
+# VI:   v_subrev_u32_e64 v1, s[0:1], v2, v3 ; encoding: [0x01,0x00,0x1b,0xd1,0x02,0x07,0x02,0x00]
 0x01 0x00 0x1b 0xd1 0x02 0x07 0x02 0x00
 
-# VI:   v_subrev_i32_e32 v1, vcc, v2, v3 ; encoding: [0x02,0x07,0x02,0x36]
+# VI:   v_subrev_u32_e32 v1, vcc, v2, v3 ; encoding: [0x02,0x07,0x02,0x36]
 0x02 0x07 0x02 0x36
 
-# VI:   v_subrev_i32_e64 v1, s[0:1], v2, v3 ; encoding: [0x01,0x00,0x1b,0xd1,0x02,0x07,0x02,0x00]
+# VI:   v_subrev_u32_e64 v1, s[0:1], v2, v3 ; encoding: [0x01,0x00,0x1b,0xd1,0x02,0x07,0x02,0x00]
 0x01 0x00 0x1b 0xd1 0x02 0x07 0x02 0x00
 
 # VI:   v_addc_u32_e32 v1, vcc, v2, v3, vcc ; encoding: [0x02,0x07,0x02,0x38]
 0x02 0x07 0x02 0x38
 
 # VI:   v_addc_u32_e32 v1, vcc, v2, v3, vcc ; encoding: [0x02,0x07,0x02,0x38]
 0x02 0x07 0x02 0x38
 

llvm/trunk/test/Object/AMDGPU/objdump.s

 // RUN: llvm-mc -arch=amdgcn -mcpu=tonga %s -filetype=obj | llvm-objdump -disassemble -arch-name=amdgcn -mcpu=tonga - | FileCheck %s
 
   .text
 
   .amdgpu_hsa_kernel hello_world
 hello_world:
   .amd_kernel_code_t
   .end_amd_kernel_code_t
 
   s_mov_b32 m0, 0x10000
   s_load_dwordx2 s[0:1], s[4:5], 0x8
   s_waitcnt lgkmcnt(0)
   s_add_u32 s0, s7, s0
 BB0:
-  v_add_i32_e32 v1, vcc, s0, v1
+  v_add_u32_e32 v1, vcc, s0, v1
 BB1:
   s_movk_i32 s0, 0x483
   v_cmp_ge_i32_e32 vcc, s0, v0
   s_and_saveexec_b64 s[0:1], vcc
   v_lshlrev_b32_e32 v4, 2, v0
 BB3:
   s_cbranch_execz 21
   s_mov_b64 s[2:3], exec
   s_mov_b64 s[10:11], exec
   v_mov_b32_e32 v3, v0
         s_endpgm
 
   .amdgpu_hsa_kernel hello_world2
 hello_world2:
   .amd_kernel_code_t
   .end_amd_kernel_code_t
 
   s_and_saveexec_b64 s[0:1], vcc
   s_cbranch_execz 85
   s_load_dwordx4 s[8:11], s[4:5], 0x40
 BB5:
   v_ashrrev_i32_e32 v77, 31, v76
   v_lshlrev_b64 v[10:11], 2, v[76:77]
   s_waitcnt lgkmcnt(0)
-  v_add_i32_e32 v10, vcc, s8, v10
+  v_add_u32_e32 v10, vcc, s8, v10
   v_mov_b32_e32 v6, s9
   v_addc_u32_e32 v11, vcc, v6, v11, vcc
   flat_load_dword v0, v[10:11]
   v_lshlrev_b32_e32 v6, 5, v8
   v_lshlrev_b32_e32 v7, 2, v7
         s_endpgm
 
 // CHECK:  file format ELF64-amdgpu
 // CHECK:  Disassembly of section .text:
 // CHECK:  hello_world:
 // CHECK:  s_mov_b32 m0, 0x10000                                      // 000000000100: BEFC00FF 00010000
 // CHECK:  s_load_dwordx2 s[0:1], s[4:5], 0x8                         // 000000000108: C0060002 00000008
 // CHECK:  s_waitcnt lgkmcnt(0)                                       // 000000000110: BF8C007F
 // CHECK:  s_add_u32 s0, s7, s0                                       // 000000000114: 80000007
 // CHECK:  BB0:
-// CHECK:  v_add_i32_e32 v1, vcc, s0, v1                              // 000000000118: 32020200
+// CHECK:  v_add_u32_e32 v1, vcc, s0, v1                              // 000000000118: 32020200
 // CHECK:  BB1:
 // CHECK:  s_movk_i32 s0, 0x483                                       // 00000000011C: B0000483
 // CHECK:  v_cmp_ge_i32_e32 vcc, s0, v0                               // 000000000120: 7D8C0000
 // CHECK:  s_and_saveexec_b64 s[0:1], vcc                             // 000000000124: BE80206A
 // CHECK:  v_lshlrev_b32_e32 v4, 2, v0                                // 000000000128: 24080082
 // CHECK:  BB3:
 // CHECK:  s_cbranch_execz 21                                         // 00000000012C: BF880015
 // CHECK:  s_mov_b64 s[2:3], exec                                     // 000000000130: BE82017E
 // CHECK:  s_mov_b64 s[10:11], exec                                   // 000000000134: BE8A017E
 // CHECK:  v_mov_b32_e32 v3, v0                                       // 000000000138: 7E060300
 // CHECK:  s_endpgm                                                   // 00000000013C: BF810000
 
 // CHECK:  hello_world2:
 // CHECK:  s_and_saveexec_b64 s[0:1], vcc                             // 000000000240: BE80206A
 // CHECK:  s_cbranch_execz 85                                         // 000000000244: BF880055
 // CHECK:  s_load_dwordx4 s[8:11], s[4:5], 0x40                       // 000000000248: C00A0202 00000040
 // CHECK:  BB5:
 // CHECK:  v_ashrrev_i32_e32 v77, 31, v76                             // 000000000250: 229A989F
 // CHECK:  v_lshlrev_b64 v[10:11], 2, v[76:77]                        // 000000000254: D28F000A 00029882
 // CHECK:  s_waitcnt lgkmcnt(0)                                       // 00000000025C: BF8C007F
-// CHECK:  v_add_i32_e32 v10, vcc, s8, v10                            // 000000000260: 32141408
+// CHECK:  v_add_u32_e32 v10, vcc, s8, v10                            // 000000000260: 32141408
 // CHECK:  v_mov_b32_e32 v6, s9                                       // 000000000264: 7E0C0209
 // CHECK:  v_addc_u32_e32 v11, vcc, v6, v11, vcc                      // 000000000268: 38161706
 // CHECK:  flat_load_dword v0, v[10:11]                               // 00000000026C: DC500000 0000000A
 // CHECK:  v_lshlrev_b32_e32 v6, 5, v8                                // 000000000274: 240C1085
 // CHECK:  v_lshlrev_b32_e32 v7, 2, v7                                // 000000000278: 240E0E82
 // CHECK:  s_endpgm                                                   // 00000000027C: BF810000

llvm/trunk/test/tools/llvm-objdump/AMDGPU/source-lines.ll

 ; RUN: sed -e "s,SRC_COMPDIR,%/p/Inputs,g" %s > %t.ll
 ; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx800 -filetype=obj -O0 -o %t.o %t.ll
 ; RUN: llvm-objdump -triple=amdgcn-amd-amdhsa -mcpu=gfx800 -disassemble -line-numbers %t.o | FileCheck --check-prefix=LINE %t.ll
 ; RUN: llvm-objdump -triple=amdgcn-amd-amdhsa -mcpu=gfx800 -disassemble -source %t.o | FileCheck --check-prefix=SOURCE %t.ll
 
 ; Prologue.
 ; LINE:      source_lines_test:
 ; LINE-NEXT: ; {{.*}}source-lines.cl:1
 ; Kernel.
 ; LINE: ; {{.*}}source-lines.cl:2
 ; LINE: v_mov_b32_e32 v{{[0-9]+}}, 0x777
 ; LINE: ; {{.*}}source-lines.cl:3
 ; LINE: v_mov_b32_e32 v{{[0-9]+}}, 0x888
 ; LINE: ; {{.*}}source-lines.cl:4
-; LINE: v_add_i32_e32
+; LINE: v_add_u32_e32
 ; LINE: ; {{.*}}source-lines.cl:5
 ; LINE: flat_store_dword
 ; Epilogue.
 ; LINE:      ; {{.*}}source-lines.cl:6
 ; LINE-NEXT: s_endpgm
 
 ; Prologue.
 ; SOURCE:      source_lines_test:
 ; SOURCE-NEXT: ; kernel void source_lines_test(global int *Out) {
 ; Kernel.
 ; SOURCE: ; int var0 = 0x777;
 ; SOURCE: v_mov_b32_e32 v{{[0-9]+}}, 0x777
 ; SOURCE: ; int var1 = 0x888;
 ; SOURCE: v_mov_b32_e32 v{{[0-9]+}}, 0x888
 ; SOURCE: ; int var2 = var0 + var1;
-; SOURCE: v_add_i32_e32
+; SOURCE: v_add_u32_e32
 ; SOURCE: ; *Out = var2;
 ; SOURCE: flat_store_dword
 ; Epilogue.
 ; SOURCE:      ; }
 ; SOURCE-NEXT: s_endpgm
 
 ; ModuleID = 'source-lines.cl'
 source_filename = "source-lines.cl"