Index: llvm/trunk/lib/Target/AMDGPU/AMDGPUCallLowering.cpp
===================================================================
--- llvm/trunk/lib/Target/AMDGPU/AMDGPUCallLowering.cpp
+++ llvm/trunk/lib/Target/AMDGPU/AMDGPUCallLowering.cpp
@@ -37,7 +37,7 @@
   if (Val)
     return false;
 
-  MIRBuilder.buildInstr(AMDGPU::S_ENDPGM);
+  MIRBuilder.buildInstr(AMDGPU::S_ENDPGM).addImm(0);
   return true;
 }
 
Index: llvm/trunk/lib/Target/AMDGPU/AsmParser/AMDGPUAsmParser.cpp
===================================================================
--- llvm/trunk/lib/Target/AMDGPU/AsmParser/AMDGPUAsmParser.cpp
+++ llvm/trunk/lib/Target/AMDGPU/AsmParser/AMDGPUAsmParser.cpp
@@ -174,6 +174,7 @@
     ImmTyNegHi,
     ImmTySwizzle,
     ImmTyGprIdxMode,
+    ImmTyEndpgm,
     ImmTyHigh
   };
 
@@ -517,6 +518,7 @@
   bool isGPRIdxMode() const;
   bool isS16Imm() const;
   bool isU16Imm() const;
+  bool isEndpgm() const;
 
   StringRef getExpressionAsToken() const {
     assert(isExpr());
@@ -706,6 +708,9 @@
     case ImmTySwizzle: OS << "Swizzle"; break;
     case ImmTyGprIdxMode: OS << "GprIdxMode"; break;
     case ImmTyHigh: OS << "High"; break;
+    case ImmTyEndpgm:
+      OS << "Endpgm";
+      break;
     }
   }
 
@@ -1187,6 +1192,9 @@
   void cvtSdwaVOPC(MCInst &Inst, const OperandVector &Operands);
   void cvtSDWA(MCInst &Inst, const OperandVector &Operands,
                 uint64_t BasicInstType, bool skipVcc = false);
+
+  OperandMatchResultTy parseEndpgmOp(OperandVector &Operands);
+  AMDGPUOperand::Ptr defaultEndpgmImmOperands() const;
 };
 
 struct OptionalOperand {
@@ -5523,6 +5531,10 @@
   return AMDGPUOperand::CreateImm(this, 0xf, SMLoc(), AMDGPUOperand::ImmTyDppRowMask);
 }
 
+AMDGPUOperand::Ptr AMDGPUAsmParser::defaultEndpgmImmOperands() const {
+  return AMDGPUOperand::CreateImm(this, 0, SMLoc(), AMDGPUOperand::ImmTyEndpgm);
+}
+
 AMDGPUOperand::Ptr AMDGPUAsmParser::defaultBankMask() const {
   return AMDGPUOperand::CreateImm(this, 0xf, SMLoc(), AMDGPUOperand::ImmTyDppBankMask);
 }
@@ -5799,3 +5811,28 @@
     return Match_InvalidOperand;
   }
 }
+
+//===----------------------------------------------------------------------===//
+// endpgm
+//===----------------------------------------------------------------------===//
+
+OperandMatchResultTy AMDGPUAsmParser::parseEndpgmOp(OperandVector &Operands) {
+  SMLoc S = Parser.getTok().getLoc();
+  int64_t Imm = 0;
+
+  if (!parseExpr(Imm)) {
+    // The operand is optional, if not present default to 0
+    Imm = 0;
+  }
+
+  if (!isUInt<16>(Imm)) {
+    Error(S, "expected a 16-bit value");
+    return MatchOperand_ParseFail;
+  }
+
+  Operands.push_back(
+      AMDGPUOperand::CreateImm(this, Imm, S, AMDGPUOperand::ImmTyEndpgm));
+  return MatchOperand_Success;
+}
+
+bool AMDGPUOperand::isEndpgm() const { return isImmTy(ImmTyEndpgm); }
Index: llvm/trunk/lib/Target/AMDGPU/InstPrinter/AMDGPUInstPrinter.h
===================================================================
--- llvm/trunk/lib/Target/AMDGPU/InstPrinter/AMDGPUInstPrinter.h
+++ llvm/trunk/lib/Target/AMDGPU/InstPrinter/AMDGPUInstPrinter.h
@@ -213,6 +213,8 @@
                      const MCSubtargetInfo &STI, raw_ostream &O);
   void printHwreg(const MCInst *MI, unsigned OpNo, const MCSubtargetInfo &STI,
                   raw_ostream &O);
+  void printEndpgm(const MCInst *MI, unsigned OpNo, const MCSubtargetInfo &STI,
+                   raw_ostream &O);
 };
 
 class R600InstPrinter : public MCInstPrinter {
Index: llvm/trunk/lib/Target/AMDGPU/InstPrinter/AMDGPUInstPrinter.cpp
===================================================================
--- llvm/trunk/lib/Target/AMDGPU/InstPrinter/AMDGPUInstPrinter.cpp
+++ llvm/trunk/lib/Target/AMDGPU/InstPrinter/AMDGPUInstPrinter.cpp
@@ -1209,6 +1209,17 @@
   O << ')';
 }
 
+void AMDGPUInstPrinter::printEndpgm(const MCInst *MI, unsigned OpNo,
+                                    const MCSubtargetInfo &STI,
+                                    raw_ostream &O) {
+  uint16_t Imm = MI->getOperand(OpNo).getImm();
+  if (Imm == 0) {
+    return;
+  }
+
+  O << formatDec(Imm);
+}
+
 #include "AMDGPUGenAsmWriter.inc"
 
 void R600InstPrinter::printInst(const MCInst *MI, raw_ostream &O,
Index: llvm/trunk/lib/Target/AMDGPU/SIInsertSkips.cpp
===================================================================
--- llvm/trunk/lib/Target/AMDGPU/SIInsertSkips.cpp
+++ llvm/trunk/lib/Target/AMDGPU/SIInsertSkips.cpp
@@ -176,7 +176,7 @@
     .addImm(0); // en
 
   // ... and terminate wavefront.
-  BuildMI(*SkipBB, Insert, DL, TII->get(AMDGPU::S_ENDPGM));
+  BuildMI(*SkipBB, Insert, DL, TII->get(AMDGPU::S_ENDPGM)).addImm(0);
 
   return true;
 }
Index: llvm/trunk/lib/Target/AMDGPU/SIInstrInfo.cpp
===================================================================
--- llvm/trunk/lib/Target/AMDGPU/SIInstrInfo.cpp
+++ llvm/trunk/lib/Target/AMDGPU/SIInstrInfo.cpp
@@ -1137,9 +1137,13 @@
 
   if (MBB.succ_empty()) {
     bool HasNoTerminator = MBB.getFirstTerminator() == MBB.end();
-    if (HasNoTerminator)
-      BuildMI(MBB, MBB.end(), DebugLoc(),
-              get(Info->returnsVoid() ? AMDGPU::S_ENDPGM : AMDGPU::SI_RETURN_TO_EPILOG));
+    if (HasNoTerminator) {
+      if (Info->returnsVoid()) {
+        BuildMI(MBB, MBB.end(), DebugLoc(), get(AMDGPU::S_ENDPGM)).addImm(0);
+      } else {
+        BuildMI(MBB, MBB.end(), DebugLoc(), get(AMDGPU::SI_RETURN_TO_EPILOG));
+      }
+    }
   }
 }
 
Index: llvm/trunk/lib/Target/AMDGPU/SIInstrInfo.td
===================================================================
--- llvm/trunk/lib/Target/AMDGPU/SIInstrInfo.td
+++ llvm/trunk/lib/Target/AMDGPU/SIInstrInfo.td
@@ -668,6 +668,14 @@
   let IsOptional = 1;
 }
 
+def EndpgmMatchClass : AsmOperandClass {
+  let Name = "EndpgmImm";
+  let PredicateMethod = "isEndpgm";
+  let ParserMethod = "parseEndpgmOp";
+  let RenderMethod = "addImmOperands";
+  let IsOptional = 1;
+}
+
 def ExpTgtMatchClass : AsmOperandClass {
   let Name = "ExpTgt";
   let PredicateMethod = "isExpTgt";
@@ -685,6 +693,11 @@
   let ParserMatchClass = SwizzleMatchClass;
 }
 
+def EndpgmImm : Operand<i16> {
+  let PrintMethod = "printEndpgm";
+  let ParserMatchClass = EndpgmMatchClass;
+}
+
 def SWaitMatchClass : AsmOperandClass {
   let Name = "SWaitCnt";
   let RenderMethod = "addImmOperands";
Index: llvm/trunk/lib/Target/AMDGPU/SIOptimizeExecMaskingPreRA.cpp
===================================================================
--- llvm/trunk/lib/Target/AMDGPU/SIOptimizeExecMaskingPreRA.cpp
+++ llvm/trunk/lib/Target/AMDGPU/SIOptimizeExecMaskingPreRA.cpp
@@ -247,9 +247,10 @@
 
       // Skip this if the endpgm has any implicit uses, otherwise we would need
       // to be careful to update / remove them.
+      // S_ENDPGM always has a single imm operand that is not used other than to
+      // end up in the encoding
       MachineInstr &Term = MBB.back();
-      if (Term.getOpcode() != AMDGPU::S_ENDPGM ||
-          Term.getNumOperands() != 0)
+      if (Term.getOpcode() != AMDGPU::S_ENDPGM || Term.getNumOperands() != 1)
         continue;
 
       SmallVector<MachineBasicBlock*, 4> Blocks({&MBB});
Index: llvm/trunk/lib/Target/AMDGPU/SOPInstructions.td
===================================================================
--- llvm/trunk/lib/Target/AMDGPU/SOPInstructions.td
+++ llvm/trunk/lib/Target/AMDGPU/SOPInstructions.td
@@ -866,9 +866,7 @@
 
 let isTerminator = 1 in {
 
-def S_ENDPGM : SOPP <0x00000001, (ins), "s_endpgm",
-  [(AMDGPUendpgm)]> {
-  let simm16 = 0;
+def S_ENDPGM : SOPP <0x00000001, (ins EndpgmImm:$simm16), "s_endpgm $simm16"> {
   let isBarrier = 1;
   let isReturn = 1;
 }
@@ -1045,6 +1043,11 @@
 //===----------------------------------------------------------------------===//
 
 def : GCNPat <
+  (AMDGPUendpgm),
+    (S_ENDPGM (i16 0))
+>;
+
+def : GCNPat <
   (i64 (ctpop i64:$src)),
     (i64 (REG_SEQUENCE SReg_64,
      (i32 (COPY_TO_REGCLASS (S_BCNT1_I32_B64 $src), SReg_32)), sub0,
Index: llvm/trunk/test/CodeGen/AMDGPU/GlobalISel/legalize-block-addr.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/GlobalISel/legalize-block-addr.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/GlobalISel/legalize-block-addr.mir
@@ -21,8 +21,8 @@
   bb.1 (%ir-block.0):
     ; CHECK-LABEL: name: test_blockaddress
     ; CHECK: [[BLOCK_ADDR:%[0-9]+]]:_(p0) = G_BLOCK_ADDR blockaddress(@test_blockaddress, %ir-block.block)
-    ; CHECK: S_ENDPGM implicit [[BLOCK_ADDR]](p0)
+    ; CHECK: S_ENDPGM 0, implicit [[BLOCK_ADDR]](p0)
     %0:_(p0) = G_BLOCK_ADDR blockaddress(@test_blockaddress, %ir-block.block)
-    S_ENDPGM implicit %0
+    S_ENDPGM 0, implicit %0
 
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/GlobalISel/legalize-constant.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/GlobalISel/legalize-constant.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/GlobalISel/legalize-constant.mir
@@ -32,9 +32,9 @@
 
     ; CHECK-LABEL: name: test_constant_s1
     ; CHECK: [[C:%[0-9]+]]:_(s1) = G_CONSTANT i1 false
-    ; CHECK: S_ENDPGM implicit [[C]](s1)
+    ; CHECK: S_ENDPGM 0, implicit [[C]](s1)
     %1:_(s1) = G_CONSTANT i1 0
-    S_ENDPGM implicit %1
+    S_ENDPGM 0, implicit %1
 ...
 
 ---
Index: llvm/trunk/test/CodeGen/AMDGPU/GlobalISel/regbankselect-block-addr.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/GlobalISel/regbankselect-block-addr.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/GlobalISel/regbankselect-block-addr.mir
@@ -22,8 +22,8 @@
   bb.1 (%ir-block.0):
     ; CHECK-LABEL: name: test_blockaddress
     ; CHECK: [[BLOCK_ADDR:%[0-9]+]]:sgpr(p0) = G_BLOCK_ADDR blockaddress(@test_blockaddress, %ir-block.block)
-    ; CHECK: S_ENDPGM implicit [[BLOCK_ADDR]](p0)
+    ; CHECK: S_ENDPGM 0, implicit [[BLOCK_ADDR]](p0)
     %0:_(p0) = G_BLOCK_ADDR blockaddress(@test_blockaddress, %ir-block.block)
-    S_ENDPGM implicit %0
+    S_ENDPGM 0, implicit %0
 
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/GlobalISel/regbankselect-merge-values.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/GlobalISel/regbankselect-merge-values.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/GlobalISel/regbankselect-merge-values.mir
@@ -14,12 +14,12 @@
     ; CHECK: [[EXTRACT:%[0-9]+]]:sgpr(s32) = G_EXTRACT [[COPY]](s64), 0
     ; CHECK: [[EXTRACT1:%[0-9]+]]:sgpr(s32) = G_EXTRACT [[COPY]](s64), 32
     ; CHECK: [[MV:%[0-9]+]]:sgpr(s64) = G_MERGE_VALUES [[EXTRACT]](s32), [[EXTRACT1]](s32)
-    ; CHECK: S_ENDPGM implicit [[MV]](s64)
+    ; CHECK: S_ENDPGM 0, implicit [[MV]](s64)
     %0:_(s64) = COPY $sgpr0_sgpr1
     %1:_(s32) = G_EXTRACT %0, 0
     %2:_(s32) = G_EXTRACT %0, 32
     %3:_(s64) = G_MERGE_VALUES %1, %2
-    S_ENDPGM implicit %3
+    S_ENDPGM 0, implicit %3
 ...
 
 ---
@@ -34,11 +34,11 @@
     ; CHECK: [[EXTRACT:%[0-9]+]]:vgpr(s32) = G_EXTRACT [[COPY]](s64), 0
     ; CHECK: [[EXTRACT1:%[0-9]+]]:vgpr(s32) = G_EXTRACT [[COPY]](s64), 32
     ; CHECK: [[MV:%[0-9]+]]:vgpr(s64) = G_MERGE_VALUES [[EXTRACT]](s32), [[EXTRACT1]](s32)
-    ; CHECK: S_ENDPGM implicit [[MV]](s64)
+    ; CHECK: S_ENDPGM 0, implicit [[MV]](s64)
     %0:_(s64) = COPY $vgpr0_vgpr1
     %1:_(s32) = G_EXTRACT %0, 0
     %2:_(s32) = G_EXTRACT %0, 32
     %3:_(s64) = G_MERGE_VALUES %1, %2
-    S_ENDPGM implicit %3
+    S_ENDPGM 0, implicit %3
 ...
 
Index: llvm/trunk/test/CodeGen/AMDGPU/break-smem-soft-clauses.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/break-smem-soft-clauses.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/break-smem-soft-clauses.mir
@@ -9,9 +9,9 @@
   bb.0:
     ; GCN-LABEL: name: trivial_smem_clause_load_smrd4_x1
     ; GCN: $sgpr0 = S_LOAD_DWORD_IMM $sgpr10_sgpr11, 0, 0
-    ; GCN-NEXT: S_ENDPGM
+    ; GCN-NEXT: S_ENDPGM 0
     $sgpr0 = S_LOAD_DWORD_IMM $sgpr10_sgpr11, 0, 0
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # Trivial clause at beginning of program
@@ -22,10 +22,10 @@
     ; GCN-LABEL: name: trivial_smem_clause_load_smrd4_x2
     ; GCN: $sgpr0 = S_LOAD_DWORD_IMM $sgpr10_sgpr11, 0, 0
     ; GCN-NEXT: $sgpr1 = S_LOAD_DWORD_IMM $sgpr12_sgpr13, 0, 0
-    ; GCN-NEXT: S_ENDPGM
+    ; GCN-NEXT: S_ENDPGM 0
     $sgpr0 = S_LOAD_DWORD_IMM $sgpr10_sgpr11, 0, 0
     $sgpr1 = S_LOAD_DWORD_IMM $sgpr12_sgpr13, 0, 0
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # Trivial clause at beginning of program
@@ -37,11 +37,11 @@
     ; GCN: $sgpr0 = S_LOAD_DWORD_IMM $sgpr12_sgpr13, 0, 0
     ; GCN-NEXT: $sgpr1 = S_LOAD_DWORD_IMM $sgpr6_sgpr7, 0, 0
     ; GCN-NEXT: $sgpr2 = S_LOAD_DWORD_IMM $sgpr14_sgpr15, 0, 0
-    ; GCN-NEXT: S_ENDPGM
+    ; GCN-NEXT: S_ENDPGM 0
     $sgpr0 = S_LOAD_DWORD_IMM $sgpr12_sgpr13, 0, 0
     $sgpr1 = S_LOAD_DWORD_IMM $sgpr6_sgpr7, 0, 0
     $sgpr2 = S_LOAD_DWORD_IMM $sgpr14_sgpr15, 0, 0
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # Trivial clause at beginning of program
@@ -54,12 +54,12 @@
     ; GCN-NEXT: $sgpr1 = S_LOAD_DWORD_IMM $sgpr8_sgpr9, 0, 0
     ; GCN-NEXT: $sgpr2 = S_LOAD_DWORD_IMM $sgpr14_sgpr15, 0, 0
     ; GCN-NEXT: $sgpr3 = S_LOAD_DWORD_IMM $sgpr16_sgpr17, 0, 0
-    ; GCN-NEXT: S_ENDPGM
+    ; GCN-NEXT: S_ENDPGM 0
     $sgpr0 = S_LOAD_DWORD_IMM $sgpr12_sgpr13, 0, 0
     $sgpr1 = S_LOAD_DWORD_IMM $sgpr8_sgpr9, 0, 0
     $sgpr2 = S_LOAD_DWORD_IMM $sgpr14_sgpr15, 0, 0
     $sgpr3 = S_LOAD_DWORD_IMM $sgpr16_sgpr17, 0, 0
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # Reuse of same input pointer is OK
@@ -69,10 +69,10 @@
     ; GCN-LABEL: name: trivial_smem_clause_load_smrd4_x2_sameptr
     ; GCN: $sgpr12 = S_LOAD_DWORD_IMM $sgpr10_sgpr11, 0, 0
     ; GCN-NEXT: $sgpr13 = S_LOAD_DWORD_IMM $sgpr10_sgpr11, 0, 0
-    ; GCN-NEXT: S_ENDPGM
+    ; GCN-NEXT: S_ENDPGM 0
     $sgpr12 = S_LOAD_DWORD_IMM $sgpr10_sgpr11, 0, 0
     $sgpr13 = S_LOAD_DWORD_IMM $sgpr10_sgpr11, 0, 0
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # 32-bit load partially clobbers its own ptr reg
@@ -82,9 +82,9 @@
   bb.0:
     ; GCN-LABEL: name: smrd_load4_overwrite_ptr_lo
     ; GCN: $sgpr10 = S_LOAD_DWORD_IMM $sgpr10_sgpr11, 0, 0
-    ; GCN-NEXT: S_ENDPGM
+    ; GCN-NEXT: S_ENDPGM 0
     $sgpr10 = S_LOAD_DWORD_IMM $sgpr10_sgpr11, 0, 0
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # 32-bit load partially clobbers its own ptr reg
@@ -94,9 +94,9 @@
   bb.0:
     ; GCN-LABEL: name: smrd_load4_overwrite_ptr_hi
     ; GCN: $sgpr11 = S_LOAD_DWORD_IMM $sgpr10_sgpr11, 0, 0
-    ; GCN-NEXT: S_ENDPGM
+    ; GCN-NEXT: S_ENDPGM 0
     $sgpr11 = S_LOAD_DWORD_IMM $sgpr10_sgpr11, 0, 0
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # 64-bit load clobbers its own ptr reg
@@ -106,9 +106,9 @@
   bb.0:
     ; GCN-LABEL: name: smrd_load8_overwrite_ptr
     ; GCN: $sgpr10_sgpr11 = S_LOAD_DWORDX2_IMM $sgpr10_sgpr11, 0, 0
-    ; GCN-NEXT: S_ENDPGM
+    ; GCN-NEXT: S_ENDPGM 0
     $sgpr10_sgpr11 = S_LOAD_DWORDX2_IMM $sgpr10_sgpr11, 0, 0
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # vmcnt has 4 bits, so maximum 16 outstanding loads. The waitcnt
@@ -137,7 +137,7 @@
     ; GCN-NEXT: $sgpr28 = S_LOAD_DWORD_IMM $sgpr10_sgpr11, 0, 0
     ; GCN-NEXT: $sgpr0 = S_LOAD_DWORD_IMM $sgpr30_sgpr31, 0, 0
     ; GCN-NEXT: $sgpr0 = S_MOV_B32 $sgpr0, implicit $sgpr13, implicit $sgpr14, implicit $sgpr15, implicit $sgpr16, implicit $sgpr17, implicit $sgpr18, implicit $sgpr19, implicit $sgpr20, implicit $sgpr21, implicit $sgpr22, implicit $sgpr23, implicit $sgpr24, implicit $sgpr25, implicit $sgpr26, implicit $sgpr27, implicit $sgpr28
-    ; GCN-NEXT: S_ENDPGM
+    ; GCN-NEXT: S_ENDPGM 0
     $sgpr13 = S_LOAD_DWORD_IMM $sgpr10_sgpr11, 0, 0
     $sgpr14 = S_LOAD_DWORD_IMM $sgpr10_sgpr11, 0, 0
     $sgpr15 = S_LOAD_DWORD_IMM $sgpr10_sgpr11, 0, 0
@@ -160,7 +160,7 @@
 
     $sgpr0 = S_LOAD_DWORD_IMM $sgpr30_sgpr31, 0, 0
     $sgpr0 = S_MOV_B32 $sgpr0, implicit $sgpr13, implicit $sgpr14, implicit $sgpr15, implicit $sgpr16, implicit $sgpr17, implicit $sgpr18, implicit $sgpr19, implicit $sgpr20, implicit $sgpr21, implicit $sgpr22, implicit $sgpr23, implicit $sgpr24, implicit $sgpr25, implicit $sgpr26, implicit $sgpr27, implicit $sgpr28
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 
@@ -172,10 +172,10 @@
     ; GCN: $sgpr10 = S_LOAD_DWORD_IMM $sgpr10_sgpr11, 0, 0
     ; XNACK-NEXT: S_NOP 0
     ; GCN-NEXT: $sgpr12 = S_LOAD_DWORD_IMM $sgpr12_sgpr13, 0, 0
-    ; GCN-NEXT: S_ENDPGM
+    ; GCN-NEXT: S_ENDPGM 0
     $sgpr10 = S_LOAD_DWORD_IMM $sgpr10_sgpr11, 0, 0
     $sgpr12 = S_LOAD_DWORD_IMM $sgpr12_sgpr13, 0, 0
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 
@@ -186,10 +186,10 @@
     ; GCN-LABEL: name: break_smem_clause_simple_load_smrd4_hi_ptr
     ; GCN: $sgpr0 = S_LOAD_DWORD_IMM $sgpr10_sgpr11, 0, 0
     ; GCN-NEXT: $sgpr3 = S_LOAD_DWORD_IMM $sgpr12_sgpr13, 0, 0
-    ; GCN-NEXT: S_ENDPGM
+    ; GCN-NEXT: S_ENDPGM 0
     $sgpr0 = S_LOAD_DWORD_IMM $sgpr10_sgpr11, 0, 0
     $sgpr3 = S_LOAD_DWORD_IMM $sgpr12_sgpr13, 0, 0
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 
@@ -201,10 +201,10 @@
     ; GCN: $sgpr0_sgpr1 = S_LOAD_DWORDX2_IMM $sgpr10_sgpr11, 0, 0
     ; XNACK-NEXT: S_NOP 0
     ; GCN-NEXT: $sgpr10_sgpr11 = S_LOAD_DWORDX2_IMM $sgpr12_sgpr13, 0, 0
-    ; GCN-NEXT: S_ENDPGM
+    ; GCN-NEXT: S_ENDPGM 0
     $sgpr0_sgpr1 = S_LOAD_DWORDX2_IMM $sgpr10_sgpr11, 0, 0
     $sgpr10_sgpr11 = S_LOAD_DWORDX2_IMM $sgpr12_sgpr13, 0, 0
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 
@@ -215,10 +215,10 @@
     ; GCN-LABEL: name: break_smem_clause_simple_load_smrd16_ptr
     ; GCN: $sgpr0_sgpr1 = S_LOAD_DWORDX2_IMM $sgpr10_sgpr11, 0, 0
     ; GCN-NEXT: $sgpr12_sgpr13_sgpr14_sgpr15 = S_LOAD_DWORDX4_IMM $sgpr6_sgpr7, 0, 0
-    ; GCN-NEXT: S_ENDPGM
+    ; GCN-NEXT: S_ENDPGM 0
     $sgpr0_sgpr1 = S_LOAD_DWORDX2_IMM $sgpr10_sgpr11, 0, 0
     $sgpr12_sgpr13_sgpr14_sgpr15 = S_LOAD_DWORDX4_IMM $sgpr6_sgpr7, 0, 0
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 
@@ -232,13 +232,13 @@
   ; GCN: bb.1:
   ; XNACK-NEXT:   S_NOP 0
   ; GCN-NEXT:   $sgpr10_sgpr11 = S_LOAD_DWORDX2_IMM $sgpr12_sgpr13, 0, 0
-  ; GCN-NEXT:   S_ENDPGM
+  ; GCN-NEXT:   S_ENDPGM 0
   bb.0:
     $sgpr0_sgpr1 = S_LOAD_DWORDX2_IMM $sgpr10_sgpr11, 0, 0
 
   bb.1:
     $sgpr10_sgpr11 = S_LOAD_DWORDX2_IMM $sgpr12_sgpr13, 0, 0
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # The load clobbers the pointer of the store, so it needs to break.
@@ -250,10 +250,10 @@
     ; GCN-LABEL: name: break_smem_clause_store_load_into_ptr_smrd4
     ; GCN: S_STORE_DWORD_IMM $sgpr16, $sgpr10_sgpr11, 0, 0
     ; GCN-NEXT: $sgpr12 = S_LOAD_DWORD_IMM $sgpr14_sgpr15, 0, 0
-    ; GCN-NEXT: S_ENDPGM
+    ; GCN-NEXT: S_ENDPGM 0
     S_STORE_DWORD_IMM $sgpr16, $sgpr10_sgpr11, 0, 0
     $sgpr12 = S_LOAD_DWORD_IMM $sgpr14_sgpr15, 0, 0
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # The load clobbers the data of the store, so it needs to break.
@@ -266,10 +266,10 @@
     ; GCN-LABEL: name: break_smem_clause_store_load_into_data_smrd4
     ; GCN: S_STORE_DWORD_IMM $sgpr8, $sgpr10_sgpr11, 0, 0
     ; GCN-NEXT: $sgpr8 = S_LOAD_DWORD_IMM $sgpr12_sgpr13, 0, 0
-    ; GCN-NEXT: S_ENDPGM
+    ; GCN-NEXT: S_ENDPGM 0
     S_STORE_DWORD_IMM $sgpr8, $sgpr10_sgpr11, 0, 0
     $sgpr8 = S_LOAD_DWORD_IMM $sgpr12_sgpr13, 0, 0
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # Regular VALU instruction breaks clause, no nop needed
@@ -281,11 +281,11 @@
     ; GCN: $sgpr0 = S_LOAD_DWORD_IMM $sgpr10_sgpr11, 0, 0
     ; GCN-NEXT: $vgpr8 = V_MOV_B32_e32 0, implicit $exec
     ; GCN-NEXT: $sgpr2 = S_LOAD_DWORD_IMM $sgpr12_sgpr13, 0, 0
-    ; GCN-NEXT: S_ENDPGM
+    ; GCN-NEXT: S_ENDPGM 0
     $sgpr0 = S_LOAD_DWORD_IMM $sgpr10_sgpr11, 0, 0
     $vgpr8 = V_MOV_B32_e32 0, implicit $exec
     $sgpr2 = S_LOAD_DWORD_IMM $sgpr12_sgpr13, 0, 0
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # Regular SALU instruction breaks clause, no nop needed
@@ -297,11 +297,11 @@
     ; GCN: $sgpr0 = S_LOAD_DWORD_IMM $sgpr10_sgpr11, 0, 0
     ; GCN-NEXT: $sgpr8 = S_MOV_B32 0
     ; GCN-NEXT: $sgpr2 = S_LOAD_DWORD_IMM $sgpr12_sgpr13, 0, 0
-    ; GCN-NEXT: S_ENDPGM
+    ; GCN-NEXT: S_ENDPGM 0
     $sgpr0 = S_LOAD_DWORD_IMM $sgpr10_sgpr11, 0, 0
     $sgpr8 = S_MOV_B32 0
     $sgpr2 = S_LOAD_DWORD_IMM $sgpr12_sgpr13, 0, 0
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 name: ds_inst_breaks_smem_clause
@@ -312,11 +312,11 @@
     ; GCN: $sgpr0 = S_LOAD_DWORD_IMM $sgpr10_sgpr11, 0, 0
     ; GCN-NEXT: $vgpr8 = DS_READ_B32 $vgpr9, 0, 0, implicit $m0, implicit $exec
     ; GCN-NEXT: $sgpr2 = S_LOAD_DWORD_IMM $sgpr12_sgpr13, 0, 0
-    ; GCN-NEXT: S_ENDPGM
+    ; GCN-NEXT: S_ENDPGM 0
     $sgpr0 = S_LOAD_DWORD_IMM $sgpr10_sgpr11, 0, 0
     $vgpr8 = DS_READ_B32 $vgpr9, 0, 0, implicit $m0, implicit $exec
     $sgpr2 = S_LOAD_DWORD_IMM $sgpr12_sgpr13, 0, 0
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 
@@ -328,11 +328,11 @@
     ; GCN: $sgpr0 = S_LOAD_DWORD_IMM $sgpr10_sgpr11, 0, 0
     ; GCN-NEXT: $vgpr0 = FLAT_LOAD_DWORD $vgpr0_vgpr1, 0, 0, 0, implicit $exec, implicit $flat_scr
     ; GCN-NEXT: $sgpr2 = S_LOAD_DWORD_IMM $sgpr12_sgpr13, 0, 0
-    ; GCN-NEXT: S_ENDPGM
+    ; GCN-NEXT: S_ENDPGM 0
     $sgpr0 = S_LOAD_DWORD_IMM $sgpr10_sgpr11, 0, 0
     $vgpr0 = FLAT_LOAD_DWORD $vgpr0_vgpr1, 0, 0, 0, implicit $exec, implicit $flat_scr
     $sgpr2 = S_LOAD_DWORD_IMM $sgpr12_sgpr13, 0, 0
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # FIXME: Should this be handled?
@@ -344,8 +344,8 @@
     ; GCN: $sgpr0_sgpr1 = S_LOAD_DWORDX2_IMM $sgpr10_sgpr11, 0, 0, implicit $sgpr12_sgpr13
     ; XNACK-NEXT: S_NOP 0
     ; GCN-NEXT: $sgpr12_sgpr13 = S_LOAD_DWORDX2_IMM $sgpr6_sgpr7, 0, 0
-    ; GCN-NEXT: S_ENDPGM
+    ; GCN-NEXT: S_ENDPGM 0
     $sgpr0_sgpr1 = S_LOAD_DWORDX2_IMM $sgpr10_sgpr11, 0, 0, implicit $sgpr12_sgpr13
     $sgpr12_sgpr13 = S_LOAD_DWORDX2_IMM $sgpr6_sgpr7, 0, 0
-    S_ENDPGM
+    S_ENDPGM 0
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/break-vmem-soft-clauses.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/break-vmem-soft-clauses.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/break-vmem-soft-clauses.mir
@@ -8,10 +8,10 @@
   bb.0:
     ; GCN-LABEL: name: trivial_clause_load_flat4_x1
     ; GCN: $vgpr0 = FLAT_LOAD_DWORD $vgpr2_vgpr3, 0, 0, 0, implicit $exec, implicit $flat_scr
-    ; GCN-NEXT: S_ENDPGM
+    ; GCN-NEXT: S_ENDPGM 0
 
     $vgpr0 = FLAT_LOAD_DWORD $vgpr2_vgpr3, 0, 0, 0, implicit $exec, implicit $flat_scr
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # Trivial clause at beginning of program
@@ -22,11 +22,11 @@
     ; GCN-LABEL: name: trivial_clause_load_flat4_x2
     ; GCN: $vgpr0 = FLAT_LOAD_DWORD $vgpr2_vgpr3, 0, 0, 0, implicit $exec, implicit $flat_scr
     ; GCN-NEXT: $vgpr1 = FLAT_LOAD_DWORD $vgpr4_vgpr5, 0, 0, 0, implicit $exec, implicit $flat_scr
-    ; GCN-NEXT: S_ENDPGM
+    ; GCN-NEXT: S_ENDPGM 0
 
     $vgpr0 = FLAT_LOAD_DWORD $vgpr2_vgpr3, 0, 0, 0, implicit $exec, implicit $flat_scr
     $vgpr1 = FLAT_LOAD_DWORD $vgpr4_vgpr5, 0, 0, 0, implicit $exec, implicit $flat_scr
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # Trivial clause at beginning of program
@@ -38,12 +38,12 @@
     ; GCN: $vgpr0 = FLAT_LOAD_DWORD $vgpr3_vgpr4, 0, 0, 0, implicit $exec, implicit $flat_scr
     ; GCN-NEXT: $vgpr1 = FLAT_LOAD_DWORD $vgpr5_vgpr6, 0, 0, 0, implicit $exec, implicit $flat_scr
     ; GCN-NEXT: $vgpr2 = FLAT_LOAD_DWORD $vgpr7_vgpr8, 0, 0, 0, implicit $exec, implicit $flat_scr
-    ; GCN-NEXT: S_ENDPGM
+    ; GCN-NEXT: S_ENDPGM 0
 
     $vgpr0 = FLAT_LOAD_DWORD $vgpr3_vgpr4, 0, 0, 0, implicit $exec, implicit $flat_scr
     $vgpr1 = FLAT_LOAD_DWORD $vgpr5_vgpr6, 0, 0, 0, implicit $exec, implicit $flat_scr
     $vgpr2 = FLAT_LOAD_DWORD $vgpr7_vgpr8, 0, 0, 0, implicit $exec, implicit $flat_scr
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # Trivial clause at beginning of program
@@ -56,13 +56,13 @@
     ; GCN-NEXT: $vgpr1 = FLAT_LOAD_DWORD $vgpr6_vgpr7, 0, 0, 0, implicit $exec, implicit $flat_scr
     ; GCN-NEXT: $vgpr2 = FLAT_LOAD_DWORD $vgpr8_vgpr9, 0, 0, 0, implicit $exec, implicit $flat_scr
     ; GCN-NEXT: $vgpr3 = FLAT_LOAD_DWORD $vgpr10_vgpr11, 0, 0, 0, implicit $exec, implicit $flat_scr
-    ; GCN-NEXT: S_ENDPGM
+    ; GCN-NEXT: S_ENDPGM 0
 
     $vgpr0 = FLAT_LOAD_DWORD $vgpr4_vgpr5, 0, 0, 0, implicit $exec, implicit $flat_scr
     $vgpr1 = FLAT_LOAD_DWORD $vgpr6_vgpr7, 0, 0, 0, implicit $exec, implicit $flat_scr
     $vgpr2 = FLAT_LOAD_DWORD $vgpr8_vgpr9, 0, 0, 0, implicit $exec, implicit $flat_scr
     $vgpr3 = FLAT_LOAD_DWORD $vgpr10_vgpr11, 0, 0, 0, implicit $exec, implicit $flat_scr
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # Reuse of same input pointer is OK
@@ -73,11 +73,11 @@
     ; GCN-LABEL: name: trivial_clause_load_flat4_x2_sameptr
     ; GCN: $vgpr0 = FLAT_LOAD_DWORD $vgpr2_vgpr3, 0, 0, 0, implicit $exec, implicit $flat_scr
     ; GCN-NEXT: $vgpr1 = FLAT_LOAD_DWORD $vgpr2_vgpr3, 0, 0, 0, implicit $exec, implicit $flat_scr
-    ; GCN-NEXT: S_ENDPGM
+    ; GCN-NEXT: S_ENDPGM 0
 
     $vgpr0 = FLAT_LOAD_DWORD $vgpr2_vgpr3, 0, 0, 0, implicit $exec, implicit $flat_scr
     $vgpr1 = FLAT_LOAD_DWORD $vgpr2_vgpr3, 0, 0, 0, implicit $exec, implicit $flat_scr
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # 32-bit load partially clobbers its own ptr reg
@@ -87,10 +87,10 @@
   bb.0:
     ; GCN-LABEL: name: flat_load4_overwrite_ptr_lo
     ; GCN: $vgpr0 = FLAT_LOAD_DWORD $vgpr0_vgpr1, 0, 0, 0, implicit $exec, implicit $flat_scr
-    ; GCN-NEXT: S_ENDPGM
+    ; GCN-NEXT: S_ENDPGM 0
 
     $vgpr0 = FLAT_LOAD_DWORD $vgpr0_vgpr1, 0, 0, 0, implicit $exec, implicit $flat_scr
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # 32-bit load partially clobbers its own ptr reg
@@ -100,10 +100,10 @@
   bb.0:
     ; GCN-LABEL: name: flat_load4_overwrite_ptr_hi
     ; GCN: $vgpr1 = FLAT_LOAD_DWORD $vgpr0_vgpr1, 0, 0, 0, implicit $exec, implicit $flat_scr
-    ; GCN-NEXT: S_ENDPGM
+    ; GCN-NEXT: S_ENDPGM 0
 
     $vgpr1 = FLAT_LOAD_DWORD $vgpr0_vgpr1, 0, 0, 0, implicit $exec, implicit $flat_scr
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # 64-bit load clobbers its own ptr reg
@@ -113,10 +113,10 @@
   bb.0:
     ; GCN-LABEL: name: flat_load8_overwrite_ptr
     ; GCN: $vgpr2_vgpr3 = FLAT_LOAD_DWORDX2 $vgpr2_vgpr3, 0, 0, 0, implicit $exec, implicit $flat_scr
-    ; GCN-NEXT: S_ENDPGM
+    ; GCN-NEXT: S_ENDPGM 0
 
     $vgpr2_vgpr3 = FLAT_LOAD_DWORDX2 $vgpr2_vgpr3, 0, 0, 0, implicit $exec, implicit $flat_scr
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # vmcnt has 4 bits, so maximum 16 outstanding loads. The waitcnt
@@ -147,7 +147,7 @@
     ; XNACK-NEXT: S_NOP 0
     ; GCN-NEXT: $vgpr0 = FLAT_LOAD_DWORD $vgpr2_vgpr3, 0, 0, 0, implicit $exec, implicit $flat_scr
     ; GCN-NEXT: $sgpr0 = S_MOV_B32 $sgpr0, implicit $vgpr2, implicit $vgpr3, implicit $vgpr4, implicit $vgpr5, implicit $vgpr6, implicit $vgpr7, implicit $vgpr8, implicit $vgpr9, implicit $vgpr10, implicit $vgpr11, implicit $vgpr12, implicit $vgpr13, implicit $vgpr14, implicit $vgpr15, implicit $vgpr16, implicit $vgpr17, implicit $vgpr18
-    ; GCN-NEXT: S_ENDPGM
+    ; GCN-NEXT: S_ENDPGM 0
 
     $vgpr2 = FLAT_LOAD_DWORD $vgpr0_vgpr1, 0, 0, 0, implicit $exec, implicit $flat_scr
     $vgpr3 = FLAT_LOAD_DWORD $vgpr0_vgpr1, 0, 0, 0, implicit $exec, implicit $flat_scr
@@ -171,7 +171,7 @@
 
     $vgpr0 = FLAT_LOAD_DWORD $vgpr2_vgpr3, 0, 0, 0, implicit $exec, implicit $flat_scr
     $sgpr0 = S_MOV_B32 $sgpr0, implicit $vgpr2, implicit $vgpr3, implicit $vgpr4, implicit $vgpr5, implicit $vgpr6, implicit $vgpr7, implicit $vgpr8, implicit $vgpr9, implicit $vgpr10, implicit $vgpr11, implicit $vgpr12, implicit $vgpr13, implicit $vgpr14, implicit $vgpr15, implicit $vgpr16, implicit $vgpr17, implicit $vgpr18
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 
@@ -183,11 +183,11 @@
     ; GCN: $vgpr0 = FLAT_LOAD_DWORD $vgpr2_vgpr3, 0, 0, 0, implicit $exec, implicit $flat_scr
     ; XNACK-NEXT: S_NOP 0
     ; GCN-NEXT: $vgpr2 = FLAT_LOAD_DWORD $vgpr4_vgpr5, 0, 0, 0, implicit $exec, implicit $flat_scr
-    ; GCN-NEXT: S_ENDPGM
+    ; GCN-NEXT: S_ENDPGM 0
 
     $vgpr0 = FLAT_LOAD_DWORD $vgpr2_vgpr3, 0, 0, 0, implicit $exec, implicit $flat_scr
     $vgpr2 = FLAT_LOAD_DWORD $vgpr4_vgpr5, 0, 0, 0, implicit $exec, implicit $flat_scr
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 
@@ -199,11 +199,11 @@
     ; GCN: $vgpr0 = FLAT_LOAD_DWORD $vgpr2_vgpr3, 0, 0, 0, implicit $exec, implicit $flat_scr
     ; XNACK-NEXT: S_NOP 0
     ; GCN-NEXT: $vgpr3 = FLAT_LOAD_DWORD $vgpr4_vgpr5, 0, 0, 0, implicit $exec, implicit $flat_scr
-    ; GCN-NEXT: S_ENDPGM
+    ; GCN-NEXT: S_ENDPGM 0
 
     $vgpr0 = FLAT_LOAD_DWORD $vgpr2_vgpr3, 0, 0, 0, implicit $exec, implicit $flat_scr
     $vgpr3 = FLAT_LOAD_DWORD $vgpr4_vgpr5, 0, 0, 0, implicit $exec, implicit $flat_scr
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 
@@ -215,11 +215,11 @@
     ; GCN: $vgpr0_vgpr1 = FLAT_LOAD_DWORDX2 $vgpr2_vgpr3, 0, 0, 0, implicit $exec, implicit $flat_scr
     ; XNACK-NEXT: S_NOP 0
     ; GCN-NEXT: $vgpr2_vgpr3 = FLAT_LOAD_DWORDX2 $vgpr4_vgpr5, 0, 0, 0, implicit $exec, implicit $flat_scr
-    ; GCN-NEXT: S_ENDPGM
+    ; GCN-NEXT: S_ENDPGM 0
 
     $vgpr0_vgpr1 = FLAT_LOAD_DWORDX2 $vgpr2_vgpr3, 0, 0, 0, implicit $exec, implicit $flat_scr
     $vgpr2_vgpr3 = FLAT_LOAD_DWORDX2 $vgpr4_vgpr5, 0, 0, 0, implicit $exec, implicit $flat_scr
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 
@@ -232,10 +232,10 @@
     ; GCN: $vgpr0_vgpr1 = FLAT_LOAD_DWORDX2 $vgpr2_vgpr3, 0, 0, 0, implicit $exec, implicit $flat_scr
     ; XNACK-NEXT: S_NOP 0
     ; GCN-NEXT: $vgpr2_vgpr3_vgpr4_vgpr5 = FLAT_LOAD_DWORDX4 $vgpr6_vgpr7, 0, 0, 0, implicit $exec, implicit $flat_scr
-    ; GCN-NEXT: S_ENDPGM
+    ; GCN-NEXT: S_ENDPGM 0
     $vgpr0_vgpr1 = FLAT_LOAD_DWORDX2 $vgpr2_vgpr3, 0, 0, 0, implicit $exec, implicit $flat_scr
     $vgpr2_vgpr3_vgpr4_vgpr5 = FLAT_LOAD_DWORDX4 $vgpr6_vgpr7, 0, 0, 0, implicit $exec, implicit $flat_scr
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 
@@ -253,14 +253,14 @@
   ; GCN: bb.1:
   ; XNACK-NEXT:  S_NOP 0
   ; GCN-NEXT:   $vgpr2_vgpr3 = FLAT_LOAD_DWORDX2 $vgpr4_vgpr5, 0, 0, 0, implicit $exec, implicit $flat_scr
-  ; GCN-NEXT:   S_ENDPGM
+  ; GCN-NEXT:   S_ENDPGM 0
 
   bb.0:
     $vgpr0_vgpr1 = FLAT_LOAD_DWORDX2 $vgpr2_vgpr3, 0, 0, 0, implicit $exec, implicit $flat_scr
 
   bb.1:
     $vgpr2_vgpr3 = FLAT_LOAD_DWORDX2 $vgpr4_vgpr5, 0, 0, 0, implicit $exec, implicit $flat_scr
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # The load clobbers the pointer of the store, so it needs to break.
@@ -272,11 +272,11 @@
     ; GCN-LABEL: name: break_clause_store_load_into_ptr_flat4
     ; GCN: FLAT_STORE_DWORD $vgpr2_vgpr3, $vgpr0, 0, 0, 0, implicit $exec, implicit $flat_scr
     ; GCN-NEXT: $vgpr2 = FLAT_LOAD_DWORD $vgpr4_vgpr5, 0, 0, 0, implicit $exec, implicit $flat_scr
-    ; GCN-NEXT: S_ENDPGM
+    ; GCN-NEXT: S_ENDPGM 0
 
     FLAT_STORE_DWORD $vgpr2_vgpr3, $vgpr0, 0, 0, 0, implicit $exec, implicit $flat_scr
     $vgpr2 = FLAT_LOAD_DWORD $vgpr4_vgpr5, 0, 0, 0, implicit $exec, implicit $flat_scr
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # The load clobbers the data of the store, so it needs to break.
@@ -289,11 +289,11 @@
     ; GCN-LABEL: name: break_clause_store_load_into_data_flat4
     ; GCN: FLAT_STORE_DWORD $vgpr2_vgpr3, $vgpr0, 0, 0, 0, implicit $exec, implicit $flat_scr
     ; GCN-NEXT: $vgpr0 = FLAT_LOAD_DWORD $vgpr4_vgpr5, 0, 0, 0, implicit $exec, implicit $flat_scr
-    ; GCN-NEXT: S_ENDPGM
+    ; GCN-NEXT: S_ENDPGM 0
 
     FLAT_STORE_DWORD $vgpr2_vgpr3, $vgpr0, 0, 0, 0, implicit $exec, implicit $flat_scr
     $vgpr0 = FLAT_LOAD_DWORD $vgpr4_vgpr5, 0, 0, 0, implicit $exec, implicit $flat_scr
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # Regular VALU instruction breaks clause, no nop needed
@@ -307,12 +307,12 @@
     ; GCN-NEXT: $vgpr8 = V_MOV_B32_e32 0, implicit $exec
     ; XNACK-NEXT: S_NOP 0
     ; GCN-NEXT: $vgpr2 = FLAT_LOAD_DWORD $vgpr4_vgpr5, 0, 0, 0, implicit $exec, implicit $flat_scr
-    ; GCN-NEXT: S_ENDPGM
+    ; GCN-NEXT: S_ENDPGM 0
 
     $vgpr0 = FLAT_LOAD_DWORD $vgpr2_vgpr3, 0, 0, 0, implicit $exec, implicit $flat_scr
     $vgpr8 = V_MOV_B32_e32 0, implicit $exec
     $vgpr2 = FLAT_LOAD_DWORD $vgpr4_vgpr5, 0, 0, 0, implicit $exec, implicit $flat_scr
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # Regular SALU instruction breaks clause, no nop needed
@@ -326,12 +326,12 @@
     ; GCN-NEXT: $sgpr8 = S_MOV_B32 0
     ; XNACK-NEXT: S_NOP 0
     ; GCN-NEXT: $vgpr2 = FLAT_LOAD_DWORD $vgpr4_vgpr5, 0, 0, 0, implicit $exec, implicit $flat_scr
-    ; GCN-NEXT: S_ENDPGM
+    ; GCN-NEXT: S_ENDPGM 0
 
     $vgpr0 = FLAT_LOAD_DWORD $vgpr2_vgpr3, 0, 0, 0, implicit $exec, implicit $flat_scr
     $sgpr8 = S_MOV_B32 0
     $vgpr2 = FLAT_LOAD_DWORD $vgpr4_vgpr5, 0, 0, 0, implicit $exec, implicit $flat_scr
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 
@@ -344,12 +344,12 @@
     ; GCN-NEXT: $vgpr8 = DS_READ_B32 $vgpr9, 0, 0, implicit $m0, implicit $exec
     ; XNACK-NEXT: S_NOP 0
     ; GCN-NEXT: $vgpr2 = FLAT_LOAD_DWORD $vgpr4_vgpr5, 0, 0, 0, implicit $exec, implicit $flat_scr
-    ; GCN-NEXT: S_ENDPGM
+    ; GCN-NEXT: S_ENDPGM 0
 
     $vgpr0 = FLAT_LOAD_DWORD $vgpr2_vgpr3, 0, 0, 0, implicit $exec, implicit $flat_scr
     $vgpr8 = DS_READ_B32 $vgpr9, 0, 0, implicit $m0, implicit $exec
     $vgpr2 = FLAT_LOAD_DWORD $vgpr4_vgpr5, 0, 0, 0, implicit $exec, implicit $flat_scr
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 
@@ -361,12 +361,12 @@
     ; GCN: $vgpr0 = FLAT_LOAD_DWORD $vgpr2_vgpr3, 0, 0, 0, implicit $exec, implicit $flat_scr
     ; GCN-NEXT: $sgpr8 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 0, 0
     ; GCN-NEXT: $vgpr2 = FLAT_LOAD_DWORD $vgpr4_vgpr5, 0, 0, 0, implicit $exec, implicit $flat_scr
-    ; GCN-NEXT: S_ENDPGM
+    ; GCN-NEXT: S_ENDPGM 0
 
     $vgpr0 = FLAT_LOAD_DWORD $vgpr2_vgpr3, 0, 0, 0, implicit $exec, implicit $flat_scr
     $sgpr8 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 0, 0
     $vgpr2 = FLAT_LOAD_DWORD $vgpr4_vgpr5, 0, 0, 0, implicit $exec, implicit $flat_scr
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # FIXME: Should this be handled?
@@ -378,11 +378,11 @@
     ; GCN: $vgpr0_vgpr1 = FLAT_LOAD_DWORDX2 $vgpr2_vgpr3, 0, 0, 0, implicit $exec, implicit $flat_scr, implicit $vgpr4_vgpr5
     ; XNACK-NEXT: S_NOP 0
     ; GCN-NEXT: $vgpr4_vgpr5 = FLAT_LOAD_DWORDX2 $vgpr6_vgpr7, 0, 0, 0, implicit $exec, implicit $flat_scr
-    ; GCN-NEXT: S_ENDPGM
+    ; GCN-NEXT: S_ENDPGM 0
 
     $vgpr0_vgpr1 = FLAT_LOAD_DWORDX2 $vgpr2_vgpr3, 0, 0, 0, implicit $exec, implicit $flat_scr, implicit $vgpr4_vgpr5
     $vgpr4_vgpr5 = FLAT_LOAD_DWORDX2 $vgpr6_vgpr7, 0, 0, 0, implicit $exec, implicit $flat_scr
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 name: trivial_clause_load_mubuf4_x2
@@ -392,11 +392,11 @@
     ; GCN-LABEL: name: trivial_clause_load_mubuf4_x2
     ; GCN: $vgpr1 = BUFFER_LOAD_DWORD_OFFEN $vgpr2, $sgpr0_sgpr1_sgpr2_sgpr3, $sgpr4, 0, 0, 0, 0, implicit $exec
     ; GCN-NEXT: $vgpr3 = BUFFER_LOAD_DWORD_OFFEN $vgpr4, $sgpr0_sgpr1_sgpr2_sgpr3, $sgpr4, 0, 0, 0, 0, implicit $exec
-    ; GCN-NEXT: S_ENDPGM
+    ; GCN-NEXT: S_ENDPGM 0
 
     $vgpr1 = BUFFER_LOAD_DWORD_OFFEN $vgpr2, $sgpr0_sgpr1_sgpr2_sgpr3, $sgpr4, 0, 0, 0, 0, implicit $exec
     $vgpr3 = BUFFER_LOAD_DWORD_OFFEN $vgpr4, $sgpr0_sgpr1_sgpr2_sgpr3, $sgpr4, 0, 0, 0, 0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 name: break_clause_simple_load_mubuf_offen_ptr
@@ -407,11 +407,11 @@
     ; GCN: $vgpr1 = BUFFER_LOAD_DWORD_OFFEN $vgpr2, $sgpr0_sgpr1_sgpr2_sgpr3, $sgpr4, 0, 0, 0, 0, implicit $exec
     ; XNACK-NEXT: S_NOP 0
     ; GCN-NEXT: $vgpr2 = BUFFER_LOAD_DWORD_OFFEN $vgpr3, $sgpr0_sgpr1_sgpr2_sgpr3, $sgpr4, 0, 0, 0, 0, implicit $exec
-    ; GCN-NEXT: S_ENDPGM
+    ; GCN-NEXT: S_ENDPGM 0
 
     $vgpr1 = BUFFER_LOAD_DWORD_OFFEN $vgpr2, $sgpr0_sgpr1_sgpr2_sgpr3, $sgpr4, 0, 0, 0, 0, implicit $exec
     $vgpr2 = BUFFER_LOAD_DWORD_OFFEN $vgpr3, $sgpr0_sgpr1_sgpr2_sgpr3, $sgpr4, 0, 0, 0, 0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # BUFFER instructions overwriting their own inputs is supposedly OK.
@@ -424,11 +424,11 @@
     ; GCN: $vgpr0 = BUFFER_LOAD_DWORD_OFFEN $vgpr0, $sgpr0_sgpr1_sgpr2_sgpr3, $sgpr4, 0, 0, 0, 0, implicit $exec
     ; GCN-NEXT: $vgpr1 = V_MOV_B32_e32 0, implicit $exec
     ; GCN-NEXT: $vgpr2 = V_MOV_B32_e32 $vgpr0, implicit $exec
-    ; GCN-NEXT: S_ENDPGM
+    ; GCN-NEXT: S_ENDPGM 0
     $vgpr0 = BUFFER_LOAD_DWORD_OFFEN $vgpr0, $sgpr0_sgpr1_sgpr2_sgpr3, $sgpr4, 0, 0, 0, 0, implicit $exec
     $vgpr1 = V_MOV_B32_e32 0, implicit $exec
     $vgpr2 = V_MOV_B32_e32 $vgpr0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # Break a clause from interference between mubuf and flat instructions
@@ -441,11 +441,11 @@
     ; GCN: $vgpr0 = FLAT_LOAD_DWORD $vgpr2_vgpr3, 0, 0, 0, implicit $exec, implicit $flat_scr
     ; XNACK-NEXT: S_NOP 0
     ; GCN-NEXT: $vgpr2 = BUFFER_LOAD_DWORD_OFFEN $vgpr1, $sgpr0_sgpr1_sgpr2_sgpr3, $sgpr4, 0, 0, 0, 0, implicit $exec
-    ; GCN-NEXT: S_ENDPGM
+    ; GCN-NEXT: S_ENDPGM 0
 
     $vgpr0 = FLAT_LOAD_DWORD $vgpr2_vgpr3, 0, 0, 0, implicit $exec, implicit $flat_scr
     $vgpr2 = BUFFER_LOAD_DWORD_OFFEN $vgpr1, $sgpr0_sgpr1_sgpr2_sgpr3, $sgpr4, 0, 0, 0, 0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 # Break a clause from interference between mubuf and flat instructions
 
@@ -454,7 +454,7 @@
 # GCN-NEXT: $vgpr0 = BUFFER_LOAD_DWORD_OFFEN $vgpr1, $sgpr0_sgpr1_sgpr2_sgpr3, $sgpr4
 # XNACK-NEXT: S_NOP 0
 # GCN-NEXT: $vgpr1 = FLAT_LOAD_DWORD $vgpr2_vgpr3
-# GCN-NEXT: S_ENDPGM
+# GCN-NEXT: S_ENDPGM 0
 name: break_clause_mubuf_load_flat_load
 
 body: |
@@ -462,7 +462,7 @@
     $vgpr0 = BUFFER_LOAD_DWORD_OFFEN $vgpr1, $sgpr0_sgpr1_sgpr2_sgpr3, $sgpr4, 0, 0, 0, 0, implicit $exec
     $vgpr1 = FLAT_LOAD_DWORD $vgpr2_vgpr3, 0, 0, 0, implicit $exec, implicit $flat_scr
 
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 
@@ -474,11 +474,11 @@
     ; GCN: $vgpr2 = FLAT_LOAD_DWORD $vgpr4_vgpr5, 0, 0, 0, implicit $exec, implicit $flat_scr
     ; XNACK-NEXT: S_NOP 0
     ; GCN-NEXT: $vgpr4 = FLAT_ATOMIC_ADD_RTN $vgpr5_vgpr6, $vgpr7, 0, 0, implicit $exec, implicit $flat_scr
-    ; GCN-NEXT: S_ENDPGM
+    ; GCN-NEXT: S_ENDPGM 0
 
     $vgpr2 = FLAT_LOAD_DWORD $vgpr4_vgpr5, 0, 0, 0, implicit $exec, implicit $flat_scr
     $vgpr4 = FLAT_ATOMIC_ADD_RTN $vgpr5_vgpr6, $vgpr7, 0, 0, implicit $exec, implicit $flat_scr
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 name: break_clause_atomic_nortn_ptr_load_flat4
@@ -488,11 +488,11 @@
     ; GCN-LABEL: name: break_clause_atomic_nortn_ptr_load_flat4
     ; GCN: FLAT_ATOMIC_ADD $vgpr0_vgpr1, $vgpr2, 0, 0, implicit $exec, implicit $flat_scr
     ; GCN-NEXT: $vgpr2 = FLAT_LOAD_DWORD $vgpr3_vgpr4, 0, 0, 0, implicit $exec, implicit $flat_scr
-    ; GCN-NEXT: S_ENDPGM
+    ; GCN-NEXT: S_ENDPGM 0
 
     FLAT_ATOMIC_ADD $vgpr0_vgpr1, $vgpr2, 0, 0, implicit $exec, implicit $flat_scr
     $vgpr2 = FLAT_LOAD_DWORD $vgpr3_vgpr4, 0, 0, 0, implicit $exec, implicit $flat_scr
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 
@@ -504,11 +504,11 @@
     ; GCN: $vgpr1 = BUFFER_LOAD_DWORD_OFFEN $vgpr2, $sgpr0_sgpr1_sgpr2_sgpr3, $sgpr4, 0, 0, 0, 0, implicit $exec
     ; XNACK-NEXT: S_NOP 0
     ; GCN-NEXT: $vgpr2 = BUFFER_ATOMIC_ADD_OFFEN_RTN $vgpr2, $vgpr5, $sgpr0_sgpr1_sgpr2_sgpr3, $sgpr4, 0, 0, implicit $exec
-    ; GCN-NEXT: S_ENDPGM
+    ; GCN-NEXT: S_ENDPGM 0
 
     $vgpr1 = BUFFER_LOAD_DWORD_OFFEN $vgpr2, $sgpr0_sgpr1_sgpr2_sgpr3, $sgpr4, 0, 0, 0, 0, implicit $exec
     $vgpr2 = BUFFER_ATOMIC_ADD_OFFEN_RTN $vgpr2, $vgpr5, $sgpr0_sgpr1_sgpr2_sgpr3, $sgpr4, 0, 0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 
@@ -519,11 +519,11 @@
     ; GCN-LABEL: name: break_clause_atomic_nortn_ptr_load_mubuf4
     ; GCN: BUFFER_ATOMIC_ADD_OFFEN $vgpr0, $vgpr1, $sgpr0_sgpr1_sgpr2_sgpr3, $sgpr4, 0, 0, implicit $exec
     ; GCN-NEXT: $vgpr1 = BUFFER_LOAD_DWORD_OFFEN $vgpr2, $sgpr0_sgpr1_sgpr2_sgpr3, $sgpr4, 0, 0, 0, 0, implicit $exec
-    ; GCN-NEXT: S_ENDPGM
+    ; GCN-NEXT: S_ENDPGM 0
 
     BUFFER_ATOMIC_ADD_OFFEN $vgpr0, $vgpr1, $sgpr0_sgpr1_sgpr2_sgpr3, $sgpr4, 0, 0, implicit $exec
     $vgpr1 = BUFFER_LOAD_DWORD_OFFEN $vgpr2, $sgpr0_sgpr1_sgpr2_sgpr3, $sgpr4, 0, 0, 0, 0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # Make sure there is no assert on mubuf instructions which do not have
@@ -535,10 +535,10 @@
     ; GCN-LABEL: name: no_break_clause_mubuf_load_novaddr
     ; GCN: $vgpr1 = BUFFER_LOAD_DWORD_OFFSET $sgpr0_sgpr1_sgpr2_sgpr3, $sgpr4, 0, 0, 0, 0, implicit $exec
     ; GCN-NEXT: $vgpr3 = BUFFER_LOAD_DWORD_OFFSET $sgpr0_sgpr1_sgpr2_sgpr3, $sgpr4, 0, 0, 0, 0, implicit $exec
-    ; GCN-NEXT: S_ENDPGM
+    ; GCN-NEXT: S_ENDPGM 0
     $vgpr1 = BUFFER_LOAD_DWORD_OFFSET $sgpr0_sgpr1_sgpr2_sgpr3, $sgpr4, 0, 0, 0, 0, implicit $exec
     $vgpr3 = BUFFER_LOAD_DWORD_OFFSET $sgpr0_sgpr1_sgpr2_sgpr3, $sgpr4, 0, 0, 0, 0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # Loads and stores using different addresses theoretically does not
@@ -557,7 +557,7 @@
     $vgpr10 = FLAT_LOAD_DWORD $vgpr2_vgpr3, 0, 0, 0, implicit $exec, implicit $flat_scr
     FLAT_STORE_DWORD $vgpr2_vgpr3, $vgpr6, 0, 0, 0, implicit $exec, implicit $flat_scr
     $vgpr11 = FLAT_LOAD_DWORD $vgpr4_vgpr5, 0, 0, 0, implicit $exec, implicit $flat_scr
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # Loads and stores using the same address needs a nop.
@@ -576,5 +576,5 @@
     $vgpr10 = FLAT_LOAD_DWORD $vgpr2_vgpr3, 0, 0, 0, implicit $exec, implicit $flat_scr
     FLAT_STORE_DWORD $vgpr0_vgpr1, $vgpr6, 0, 0, 0, implicit $exec, implicit $flat_scr
     $vgpr11 = FLAT_LOAD_DWORD $vgpr4_vgpr5, 0, 0, 0, implicit $exec, implicit $flat_scr
-    S_ENDPGM
+    S_ENDPGM 0
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/clamp-omod-special-case.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/clamp-omod-special-case.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/clamp-omod-special-case.mir
@@ -59,7 +59,7 @@
     %20 = V_ADD_F32_e64 0, killed %17, 0, 1065353216, 0, 0, implicit $exec
     %21 = V_MAX_F32_e64 0, killed %20, 0, killed %20, 0, 0, implicit $exec
     BUFFER_STORE_DWORD_ADDR64 killed %21, %26, killed %16, 0, 0, 0, 0, 0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -121,7 +121,7 @@
     %20 = V_ADD_F32_e64 0, killed %17, 0, 1065353216, 0, 0, implicit $exec
     %21 = V_MAX_F32_e64 0, killed %20, 0, killed %20, 1, 3, implicit $exec
     BUFFER_STORE_DWORD_ADDR64 killed %21, %26, killed %16, 0, 0, 0, 0, 0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # Don't fold a mul that looks like an omod if itself has omod set
@@ -184,7 +184,7 @@
     %20 = V_ADD_F32_e64 0, killed %17, 0, 1065353216, 0, 0, implicit $exec
     %21 = V_MUL_F32_e64 0, killed %20, 0, 1056964608, 0, 3, implicit $exec
     BUFFER_STORE_DWORD_ADDR64 killed %21, %26, killed %16, 0, 0, 0, 0, 0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -249,7 +249,7 @@
     %20 = V_ADD_F32_e64 0, killed %17, 0, 1065353216, 0, 0, implicit $exec
     %21 = V_MUL_F32_e64 0, killed %20, 0, 1056964608, 1, 0, implicit $exec
     BUFFER_STORE_DWORD_ADDR64 killed %21, %26, killed %16, 0, 0, 0, 0, 0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 
@@ -326,7 +326,7 @@
     %20 = V_ADD_F32_e64 0, killed %17, 0, 1065353216, 0, 0, implicit $exec
     %21 = V_ADD_F32_e64 0, killed %20, 0, killed %20, 0, 3, implicit $exec
     BUFFER_STORE_DWORD_ADDR64 killed %21, %26, killed %16, 0, 0, 0, 0, 0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -391,7 +391,7 @@
     %20 = V_ADD_F32_e64 0, killed %17, 0, 1065353216, 0, 0, implicit $exec
     %21 = V_ADD_F32_e64 0, killed %20, 0, killed %20, 1, 0, implicit $exec
     BUFFER_STORE_DWORD_ADDR64 killed %21, %26, killed %16, 0, 0, 0, 0, 0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
Index: llvm/trunk/test/CodeGen/AMDGPU/cluster-flat-loads-postra.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/cluster-flat-loads-postra.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/cluster-flat-loads-postra.mir
@@ -26,6 +26,6 @@
     $vgpr1 = V_ADDC_U32_e32 $vgpr3, killed $vgpr6, implicit-def dead $vcc, implicit $vcc, implicit $exec
     FLAT_STORE_DWORD $vgpr2_vgpr3, killed $vgpr0, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4)
     FLAT_STORE_DWORD $vgpr0_vgpr1, killed $vgpr4, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/coalescer-extend-pruned-subrange.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/coalescer-extend-pruned-subrange.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/coalescer-extend-pruned-subrange.mir
@@ -31,7 +31,7 @@
     %15:sreg_64_xexec = V_CMP_EQ_U32_e64 0, %14, implicit $exec
     %16:vgpr_32 = V_CNDMASK_B32_e64 0, 1, %15, implicit $exec
     BUFFER_STORE_DWORD_OFFEN_exact %16, undef %17:vgpr_32, undef %18:sreg_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable store 4 into constant-pool, align 1, addrspace 4)
-    S_ENDPGM
+    S_ENDPGM 0
 
   bb.2:
     successors: %bb.3, %bb.4
Index: llvm/trunk/test/CodeGen/AMDGPU/coalescer-subranges-another-copymi-not-live.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/coalescer-subranges-another-copymi-not-live.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/coalescer-subranges-another-copymi-not-live.mir
@@ -129,5 +129,5 @@
     %29:vgpr_32 = V_ADD_F32_e32 0, killed %28, implicit $exec
     $m0 = S_MOV_B32 -1
     DS_WRITE_B32 undef %30:vgpr_32, killed %29, 0, 0, implicit $m0, implicit $exec :: (store 4 into `i32 addrspace(3)* undef`, addrspace 3)
-    S_ENDPGM
+    S_ENDPGM 0
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/coalescer-subranges-another-prune-error.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/coalescer-subranges-another-prune-error.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/coalescer-subranges-another-prune-error.mir
@@ -264,5 +264,5 @@
     %115:vgpr_32 = V_MAX_F32_e32 0, killed %114, implicit $exec
     %116:vgpr_32 = V_CVT_PKRTZ_F16_F32_e64 0, killed %115, 0, 1065353216, 0, 0, implicit $exec
     EXP 0, undef %117:vgpr_32, killed %116, undef %118:vgpr_32, undef %119:vgpr_32, -1, -1, 15, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/coalescer-subregjoin-fullcopy.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/coalescer-subregjoin-fullcopy.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/coalescer-subregjoin-fullcopy.mir
@@ -156,7 +156,7 @@
     %54:vreg_128 = COPY killed %38
     %55:vgpr_32 = V_FMA_F32 0, killed %54.sub1, 0, target-flags(amdgpu-gotprel32-lo) 1056964608, 0, 1056964608, 0, 0, implicit $exec
     EXP 1, undef %56:vgpr_32, killed %55, undef %57:vgpr_32, undef %58:vgpr_32, -1, 0, 15, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 
   bb.18:
     successors: %bb.7(0x80000000)
Index: llvm/trunk/test/CodeGen/AMDGPU/coalescer-with-subregs-bad-identical.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/coalescer-with-subregs-bad-identical.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/coalescer-with-subregs-bad-identical.mir
@@ -180,5 +180,5 @@
     successors: %bb.21(0x80000000)
 
   bb.21:
-    S_ENDPGM
+    S_ENDPGM 0
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/coalescing-with-subregs-in-loop-bug.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/coalescing-with-subregs-in-loop-bug.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/coalescing-with-subregs-in-loop-bug.mir
@@ -93,6 +93,6 @@
     undef %32.sub0:vreg_128 = COPY killed %31.sub0
     %32.sub2:vreg_128 = COPY %33
     $vgpr0_vgpr1_vgpr2_vgpr3 = COPY %32:vreg_128
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/constant-fold-imm-immreg.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/constant-fold-imm-immreg.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/constant-fold-imm-immreg.mir
@@ -55,7 +55,7 @@
     %9 = S_AND_B32 killed %7, killed %8, implicit-def dead $scc
     %10 = COPY %9
     BUFFER_STORE_DWORD_OFFSET killed %10, killed %6, 0, 0, 0, 0, 0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -158,7 +158,7 @@
     %31 = V_AND_B32_e64 %34, %34, implicit $exec
     FLAT_STORE_DWORD %37, %31, 0, 0, 0, implicit $exec, implicit $flat_scr
 
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -220,7 +220,7 @@
     %12 = S_LSHL_B32 killed %5, 12, implicit-def dead $scc
     %13 = COPY %12
     BUFFER_STORE_DWORD_OFFSET killed %13, killed %10, 0, 0, 0, 0, 0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -361,7 +361,7 @@
     %28 = V_LSHL_B32_e32 %27, %6, implicit $exec
     FLAT_STORE_DWORD %20, %28, 0, 0, 0, implicit $exec, implicit $flat_scr
 
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -420,7 +420,7 @@
     %12 = S_ASHR_I32 killed %5, 12, implicit-def dead $scc
     %13 = COPY %12
     BUFFER_STORE_DWORD_OFFSET killed %13, killed %10, 0, 0, 0, 0, 0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 
@@ -569,7 +569,7 @@
     %28 = V_ASHR_I32_e32 %27, %35, implicit $exec
     FLAT_STORE_DWORD %20, %28, 0, 0, 0, implicit $exec, implicit $flat_scr
 
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -628,7 +628,7 @@
     %12 = S_LSHR_B32 killed %5, 12, implicit-def dead $scc
     %13 = COPY %12
     BUFFER_STORE_DWORD_OFFSET killed %13, killed %10, 0, 0, 0, 0, 0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -778,7 +778,7 @@
     %28 = V_LSHR_B32_e32 %27, %35, implicit $exec
     FLAT_STORE_DWORD %20, %28, 0, 0, 0, implicit $exec, implicit $flat_scr
 
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -788,7 +788,7 @@
 # GCN-LABEL: name: undefined_vreg_operand{{$}}
 # GCN: bb.0
 # GCN-NEXT: FLAT_STORE_DWORD undef %3:vreg_64, undef %1:vgpr_32,
-# GCN-NEXT: S_ENDPGM
+# GCN-NEXT: S_ENDPGM 0
 name: undefined_vreg_operand
 tracksRegLiveness: true
 registers:
@@ -801,7 +801,7 @@
     %0 = V_MOV_B32_e32 0, implicit $exec
     %2 = V_XOR_B32_e64 killed %0, undef %1, implicit $exec
     FLAT_STORE_DWORD undef %3, %2, 0, 0, 0, implicit $exec, implicit $flat_scr
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -828,13 +828,13 @@
   bb.3:
     liveins: $vcc
     SI_END_CF %0, implicit-def dead $exec, implicit-def dead $scc, implicit $exec
-    S_ENDPGM implicit $vcc
+    S_ENDPGM 0, implicit $vcc
 
 ...
 ---
 # GCN-LABEL: name: constant_fold_lshl_or_reg0_immreg_reg{{$}}
 # GCN: %2:vgpr_32 = COPY $vgpr0, implicit $exec
-# GCN-NEXT: S_ENDPGM implicit %2
+# GCN-NEXT: S_ENDPGM 0, implicit %2
 
 name: constant_fold_lshl_or_reg0_immreg_reg
 alignment:       0
@@ -851,7 +851,7 @@
   %0:vgpr_32 = V_MOV_B32_e32 0, implicit $exec
   %1:vgpr_32 = V_MOV_B32_e32 16, implicit $exec
   %2:vgpr_32 = V_LSHL_OR_B32 %0,%1, $vgpr0, implicit $exec
-  S_ENDPGM implicit %2
+  S_ENDPGM 0, implicit %2
 
 ...
 
@@ -859,7 +859,7 @@
 
 # GCN-LABEL: name: constant_fold_lshl_or_reg0_immreg_imm{{$}}
 # GCN: %2:vgpr_32 = V_MOV_B32_e32 10, implicit $exec
-# GCN-NEXT: S_ENDPGM implicit %2
+# GCN-NEXT: S_ENDPGM 0, implicit %2
 
 name: constant_fold_lshl_or_reg0_immreg_imm
 alignment:       0
@@ -875,7 +875,7 @@
   %0:vgpr_32 = V_MOV_B32_e32 0, implicit $exec
   %1:vgpr_32 = V_MOV_B32_e32 16, implicit $exec
   %2:vgpr_32 = V_LSHL_OR_B32 %0, %1, 10, implicit $exec
-  S_ENDPGM implicit %2
+  S_ENDPGM 0, implicit %2
 
 ...
 
@@ -883,7 +883,7 @@
 
 # GCN-LABEL: name: constant_fold_lshl_or_reg0_immreg_immreg{{$}}
 # GCN: %3:vgpr_32 = V_MOV_B32_e32 12345, implicit $exec, implicit $exec
-# GCN-NEXT: S_ENDPGM implicit %3
+# GCN-NEXT: S_ENDPGM 0, implicit %3
 
 name: constant_fold_lshl_or_reg0_immreg_immreg
 alignment:       0
@@ -900,6 +900,6 @@
   %1:vgpr_32 = V_MOV_B32_e32 16, implicit $exec
   %2:vgpr_32 = V_MOV_B32_e32 12345, implicit $exec
   %3:vgpr_32 = V_LSHL_OR_B32 %0, %1, %2, implicit $exec
-  S_ENDPGM implicit %3
+  S_ENDPGM 0, implicit %3
 
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/couldnt-join-subrange-3.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/couldnt-join-subrange-3.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/couldnt-join-subrange-3.mir
@@ -311,9 +311,9 @@
     %44:vgpr_32 = V_MAD_F32 0, killed %43, 0, 0, 0, 0, 0, 0, implicit $exec
     %45:vgpr_32 = V_CVT_PKRTZ_F16_F32_e64 0, killed %44, 0, undef %46:vgpr_32, 0, 0, implicit $exec
     EXP_DONE 0, killed %45, undef %47:vgpr_32, undef %48:vgpr_32, undef %49:vgpr_32, -1, -1, 15, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 
   bb.6.DummyReturnBlock:
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/endpgm-dce.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/endpgm-dce.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/endpgm-dce.mir
@@ -2,7 +2,7 @@
 
 # GCN-LABEL: name: kill_all
 # GCN:      bb.0:
-# GCN-NEXT: S_ENDPGM
+# GCN-NEXT: S_ENDPGM 0
 name: kill_all
 tracksRegLiveness: true
 registers:
@@ -20,13 +20,13 @@
     %1 = FLAT_LOAD_DWORD %0, 0, 0, 0, implicit $exec, implicit $flat_scr :: (load 4)
     %2 = V_ADD_F32_e64 0, killed %1, 0, 1, 0, 0, implicit $exec
     %4 = S_ADD_U32 %3, 1, implicit-def $scc
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # GCN-LABEL: name: load_without_memoperand
 # GCN:      $sgpr0_sgpr1 = S_OR_B64 $exec, killed $vcc, implicit-def $scc
 # GCN-NEXT: dead %1:vgpr_32 = FLAT_LOAD_DWORD %0, 0, 0, 0, implicit $exec, implicit $flat_scr
-# GCN-NEXT: S_ENDPGM
+# GCN-NEXT: S_ENDPGM 0
 name: load_without_memoperand
 tracksRegLiveness: true
 registers:
@@ -44,13 +44,13 @@
     %1 = FLAT_LOAD_DWORD %0, 0, 0, 0, implicit $exec, implicit $flat_scr
     %2 = V_ADD_F32_e64 0, killed %1, 0, 1, 0, 0, implicit $exec
     %4 = S_ADD_U32 %3, 1, implicit-def $scc
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # GCN-LABEL: name: load_volatile
 # GCN:      $sgpr0_sgpr1 = S_OR_B64 $exec, killed $vcc, implicit-def $scc
 # GCN-NEXT: dead %1:vgpr_32 = FLAT_LOAD_DWORD %0, 0, 0, 0, implicit $exec, implicit $flat_scr :: (volatile load 4)
-# GCN-NEXT: S_ENDPGM
+# GCN-NEXT: S_ENDPGM 0
 name: load_volatile
 tracksRegLiveness: true
 registers:
@@ -68,13 +68,13 @@
     %1 = FLAT_LOAD_DWORD %0, 0, 0, 0, implicit $exec, implicit $flat_scr :: (volatile load 4)
     %2 = V_ADD_F32_e64 0, killed %1, 0, 1, 0, 0, implicit $exec
     %4 = S_ADD_U32 %3, 1, implicit-def $scc
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # GCN-LABEL: name: store
 # GCN:      $sgpr0_sgpr1 = S_OR_B64 $exec, killed $vcc, implicit-def $scc
 # GCN-NEXT: FLAT_STORE_DWORD %0, %1, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4)
-# GCN-NEXT: S_ENDPGM
+# GCN-NEXT: S_ENDPGM 0
 name: store
 tracksRegLiveness: true
 registers:
@@ -87,13 +87,13 @@
     %1 = IMPLICIT_DEF
     $sgpr0_sgpr1 = S_OR_B64 $exec, killed $vcc, implicit-def $scc
     FLAT_STORE_DWORD %0, %1, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4)
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # GCN-LABEL: name: barrier
 # GCN:      $sgpr0_sgpr1 = S_OR_B64 $exec, killed $vcc, implicit-def $scc
 # GCN-NEXT: S_BARRIER
-# GCN-NEXT: S_ENDPGM
+# GCN-NEXT: S_ENDPGM 0
 name: barrier
 tracksRegLiveness: true
 body:             |
@@ -101,13 +101,13 @@
     $vcc = IMPLICIT_DEF
     $sgpr0_sgpr1 = S_OR_B64 $exec, killed $vcc, implicit-def $scc
     S_BARRIER
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # GCN-LABEL: name: call
 # GCN:      $sgpr0_sgpr1 = S_OR_B64 $exec, killed $vcc, implicit-def $scc
 # GCN-NEXT: $sgpr4_sgpr5 = S_SWAPPC_B64 $sgpr2_sgpr3
-# GCN-NEXT: S_ENDPGM
+# GCN-NEXT: S_ENDPGM 0
 name: call
 tracksRegLiveness: true
 body:             |
@@ -115,13 +115,13 @@
     $vcc = IMPLICIT_DEF
     $sgpr0_sgpr1 = S_OR_B64 $exec, killed $vcc, implicit-def $scc
     $sgpr4_sgpr5 = S_SWAPPC_B64 $sgpr2_sgpr3
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # GCN-LABEL: name: exp
 # GCN:      $sgpr0_sgpr1 = S_OR_B64 $exec, killed $vcc, implicit-def $scc
 # GCN-NEXT: EXP 32, undef %0:vgpr_32, undef %1:vgpr_32, %2, undef %3:vgpr_32, 0, 0, 15, implicit $exec
-# GCN-NEXT: S_ENDPGM
+# GCN-NEXT: S_ENDPGM 0
 name: exp
 tracksRegLiveness: true
 registers:
@@ -135,7 +135,7 @@
     %2 = IMPLICIT_DEF
     $sgpr0_sgpr1 = S_OR_B64 $exec, killed $vcc, implicit-def $scc
     EXP 32, undef %0, undef %1, killed %2, undef %3, 0, 0, 15, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # GCN-LABEL: name: return_to_epilog
@@ -156,7 +156,7 @@
 # GCN-NEXT:   successors: %bb.1
 # GCN-NOT:  S_OR_B64
 # GCN:      bb.1:
-# GCN-NEXT:   S_ENDPGM
+# GCN-NEXT:   S_ENDPGM 0
 name: split_block
 tracksRegLiveness: true
 registers:
@@ -174,7 +174,7 @@
     %2 = IMPLICIT_DEF
     %1 = V_ADD_F32_e64 0, killed %0, 0, 1, 0, 0, implicit $exec
     %3 = S_ADD_U32 %2, 1, implicit-def $scc
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # GCN-LABEL: name: split_block_empty_block
@@ -183,7 +183,7 @@
 # GCN-NOT:  S_OR_B64
 # GCN:      bb.1:
 # GCN:      bb.2:
-# GCN-NEXT:   S_ENDPGM
+# GCN-NEXT:   S_ENDPGM 0
 name: split_block_empty_block
 tracksRegLiveness: true
 body:             |
@@ -194,7 +194,7 @@
   bb.1:
 
   bb.2:
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # GCN-LABEL: name: split_block_uncond_branch
@@ -203,7 +203,7 @@
 # GCN:        S_BRANCH %bb.1
 # GCN-NOT:  S_OR_B64
 # GCN:      bb.1:
-# GCN-NEXT:   S_ENDPGM
+# GCN-NEXT:   S_ENDPGM 0
 name: split_block_uncond_branch
 tracksRegLiveness: true
 body:             |
@@ -213,7 +213,7 @@
     S_BRANCH %bb.1
 
   bb.1:
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # GCN-LABEL: name: split_block_cond_branch
@@ -223,7 +223,7 @@
 # GCN:        S_CBRANCH_VCCNZ %bb.2, implicit undef $vcc
 # GCN:      bb.1:
 # GCN:      bb.2:
-# GCN-NEXT:   S_ENDPGM
+# GCN-NEXT:   S_ENDPGM 0
 name: split_block_cond_branch
 tracksRegLiveness: true
 body:             |
@@ -235,7 +235,7 @@
   bb.1:
 
   bb.2:
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # GCN-LABEL: name: two_preds_both_dead
@@ -248,7 +248,7 @@
 # GCN-NOT:    S_AND
 # GCN:        S_BRANCH %bb.2
 # GCN:      bb.2:
-# GCN-NEXT:   S_ENDPGM
+# GCN-NEXT:   S_ENDPGM 0
 name: two_preds_both_dead
 tracksRegLiveness: true
 body:             |
@@ -263,7 +263,7 @@
     S_BRANCH %bb.2
 
   bb.2:
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # GCN-LABEL: name: two_preds_one_dead
@@ -277,7 +277,7 @@
 # GCN-NOT:    S_AND
 # GCN:        S_BRANCH %bb.2
 # GCN:      bb.2:
-# GCN-NEXT:   S_ENDPGM
+# GCN-NEXT:   S_ENDPGM 0
 name: two_preds_one_dead
 tracksRegLiveness: true
 body:             |
@@ -293,14 +293,14 @@
     S_BRANCH %bb.2
 
   bb.2:
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 
 # GCN-LABEL: name: implicit_use_on_s_endpgm
 # GCN: V_ADD_I32
 # GCN: COPY
 # GCN: V_ADDC_U32
-# GCN: S_ENDPGM implicit %3
+# GCN: S_ENDPGM 0, implicit %3
 name: implicit_use_on_s_endpgm
 tracksRegLiveness: true
 
@@ -309,6 +309,6 @@
     dead %0:vgpr_32 = V_ADD_I32_e32 12345, undef %1:vgpr_32, implicit-def $vcc, implicit $exec
     %2:sreg_64_xexec = COPY $vcc
     %3:vgpr_32, dead %4:sreg_64_xexec = V_ADDC_U32_e64 undef %5:vgpr_32, undef %6:vgpr_32, %2, implicit $exec
-    S_ENDPGM implicit %3
+    S_ENDPGM 0, implicit %3
 
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/flat-load-clustering.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/flat-load-clustering.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/flat-load-clustering.mir
@@ -72,6 +72,6 @@
     %10.sub1 = V_ADDC_U32_e32 %9.sub1, %2, implicit-def dead $vcc, implicit killed $vcc, implicit $exec
     FLAT_STORE_DWORD %9, %5, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4 into %ir.gep2)
     FLAT_STORE_DWORD %10, %6, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4 into %ir.gep4)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/fold-imm-f16-f32.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/fold-imm-f16-f32.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/fold-imm-f16-f32.mir
@@ -162,7 +162,7 @@
     %12 = V_MOV_B32_e32 1065353216, implicit $exec
     %13 = V_ADD_F16_e64 0, killed %11, 0, %12, 0, 0, implicit $exec
     BUFFER_STORE_SHORT_OFFSET killed %13, %10, 0, 0, 0, 0, 0, implicit $exec :: (volatile store 2 into `half addrspace(1)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -229,7 +229,7 @@
     %15 = V_ADD_F16_e64 0, killed %12, 0, killed %13, 0, 0, implicit $exec
     BUFFER_STORE_SHORT_OFFSET killed %14, %10, 0, 0, 0, 0, 0, implicit $exec :: (volatile store 2 into `half addrspace(1)* undef`)
     BUFFER_STORE_SHORT_OFFSET killed %15, %10, 0, 0, 0, 0, 0, implicit $exec :: (volatile store 2 into `half addrspace(1)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -297,7 +297,7 @@
     %16 = V_ADD_F32_e64 0, killed %13, 0, killed %14, 0, 0, implicit $exec
     BUFFER_STORE_SHORT_OFFSET killed %15, %10, 0, 0, 0, 0, 0, implicit $exec :: (volatile store 2 into `half addrspace(1)* undef`)
     BUFFER_STORE_DWORD_OFFSET killed %16, %10, 0, 0, 0, 0, 0, implicit $exec :: (volatile store 4 into `float addrspace(1)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -370,7 +370,7 @@
     BUFFER_STORE_SHORT_OFFSET killed %15, %10, 0, 0, 0, 0, 0, implicit $exec :: (volatile store 2 into `half addrspace(1)* undef`)
     BUFFER_STORE_SHORT_OFFSET killed %16, %10, 0, 0, 0, 0, 0, implicit $exec :: (volatile store 2 into `half addrspace(1)* undef`)
     BUFFER_STORE_DWORD_OFFSET killed %17, %10, 0, 0, 0, 0, 0, implicit $exec :: (volatile store 4 into `float addrspace(1)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -434,7 +434,7 @@
     %15 = V_ADD_F16_e64 0, killed %12, 0, killed %13, 0, 0, implicit $exec
     BUFFER_STORE_SHORT_OFFSET killed %14, %10, 0, 0, 0, 0, 0, implicit $exec :: (volatile store 2 into `half addrspace(1)* undef`)
     BUFFER_STORE_SHORT_OFFSET killed %15, %10, 0, 0, 0, 0, 0, implicit $exec :: (volatile store 2 into `half addrspace(1)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -504,7 +504,7 @@
     BUFFER_STORE_SHORT_OFFSET killed %15, %10, 0, 0, 0, 0, 0, implicit $exec :: (volatile store 2 into `half addrspace(1)* undef`)
     BUFFER_STORE_SHORT_OFFSET killed %16, %10, 0, 0, 0, 0, 0, implicit $exec :: (volatile store 2 into `half addrspace(1)* undef`)
     BUFFER_STORE_DWORD_OFFSET killed %17, %10, 0, 0, 0, 0, 0, implicit $exec :: (volatile store 4 into `float addrspace(1)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -571,7 +571,7 @@
     %15 = V_ADD_F32_e64 0, %12, 0, %13, 0, 0, implicit $exec
     BUFFER_STORE_DWORD_OFFSET killed %14, %10, 0, 0, 0, 0, 0, implicit $exec :: (volatile store 4 into `float addrspace(1)* undef`)
     BUFFER_STORE_DWORD_OFFSET killed %15, %10, 0, 0, 0, 0, 0, implicit $exec :: (volatile store 4 into `float addrspace(1)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -638,7 +638,7 @@
     %15 = V_ADD_F16_e64 0, %12, 0, %13, 0, 0, implicit $exec
     BUFFER_STORE_SHORT_OFFSET killed %14, %10, 0, 0, 0, 0, 0, implicit $exec :: (volatile store 2 into `half addrspace(1)* undef`)
     BUFFER_STORE_SHORT_OFFSET killed %15, %10, 0, 0, 0, 0, 0, implicit $exec :: (volatile store 2 into `half addrspace(1)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -704,6 +704,6 @@
     %15 = V_ADD_F16_e64 0, %12, 0, %13, 0, 0, implicit $exec
     BUFFER_STORE_DWORD_OFFSET killed %14, %10, 0, 0, 0, 0, 0, implicit $exec :: (volatile store 4 into `float addrspace(1)* undef`)
     BUFFER_STORE_SHORT_OFFSET killed %15, %10, 0, 0, 0, 0, 0, implicit $exec :: (volatile store 2 into `half addrspace(1)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/fold-immediate-operand-shrink-with-carry.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/fold-immediate-operand-shrink-with-carry.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/fold-immediate-operand-shrink-with-carry.mir
@@ -15,14 +15,14 @@
     ; GCN: [[DEF:%[0-9]+]]:vgpr_32 = IMPLICIT_DEF
     ; GCN: [[V_ADD_I32_e32_:%[0-9]+]]:vgpr_32 = V_ADD_I32_e32 [[S_MOV_B32_]], [[DEF]], implicit-def $vcc, implicit $exec
     ; GCN: [[COPY:%[0-9]+]]:sreg_64_xexec = COPY killed $vcc
-    ; GCN: S_ENDPGM implicit [[COPY]]
+    ; GCN: S_ENDPGM 0, implicit [[COPY]]
     %0:sreg_32_xm0 = S_MOV_B32 12345
     %1:vgpr_32 = IMPLICIT_DEF
     %2:vgpr_32 = IMPLICIT_DEF
     %3:vgpr_32 = IMPLICIT_DEF
 
     %4:vgpr_32, %5:sreg_64_xexec = V_ADD_I32_e64 %0, %1, implicit $exec
-    S_ENDPGM implicit %5
+    S_ENDPGM 0, implicit %5
 
 ...
 ---
@@ -38,7 +38,7 @@
     ; GCN: [[DEF1:%[0-9]+]]:vgpr_32 = IMPLICIT_DEF
     ; GCN: [[V_ADD_I32_e64_:%[0-9]+]]:vgpr_32, [[V_ADD_I32_e64_1:%[0-9]+]]:sreg_64_xexec = V_ADD_I32_e64 [[S_MOV_B32_]], [[DEF]], implicit $exec
     ; GCN: [[V_ADD_I32_e64_2:%[0-9]+]]:vgpr_32, [[V_ADD_I32_e64_3:%[0-9]+]]:sreg_64_xexec = V_ADD_I32_e64 [[S_MOV_B32_]], [[DEF1]], implicit $exec
-    ; GCN: S_ENDPGM implicit [[V_ADD_I32_e64_1]], implicit [[V_ADD_I32_e64_2]]
+    ; GCN: S_ENDPGM 0, implicit [[V_ADD_I32_e64_1]], implicit [[V_ADD_I32_e64_2]]
     %0:sreg_32_xm0 = S_MOV_B32 12345
     %1:vgpr_32 = IMPLICIT_DEF
     %2:vgpr_32 = IMPLICIT_DEF
@@ -47,7 +47,7 @@
 
     %5:vgpr_32, %6:sreg_64_xexec = V_ADD_I32_e64 %0, %1, implicit $exec
     %7:vgpr_32, %8:sreg_64_xexec = V_ADD_I32_e64 %0, %2, implicit $exec
-    S_ENDPGM implicit %6, implicit %7
+    S_ENDPGM 0, implicit %6, implicit %7
 
 ...
 ---
@@ -64,7 +64,7 @@
     ; GCN: [[DEF:%[0-9]+]]:vgpr_32 = IMPLICIT_DEF
     ; GCN: [[V_ADD_I32_e32_:%[0-9]+]]:vgpr_32 = V_ADD_I32_e32 [[S_MOV_B32_]], [[DEF]], implicit-def $vcc, implicit $exec
     ; GCN: DBG_VALUE %5:sreg_64_xexec, $noreg
-    ; GCN: S_ENDPGM implicit [[V_ADD_I32_e32_]]
+    ; GCN: S_ENDPGM 0, implicit [[V_ADD_I32_e32_]]
     %0:sreg_32_xm0 = S_MOV_B32 12345
     %1:vgpr_32 = IMPLICIT_DEF
     %2:vgpr_32 = IMPLICIT_DEF
@@ -72,7 +72,7 @@
 
     %4:vgpr_32, %5:sreg_64_xexec = V_ADD_I32_e64 %0, %1, implicit $exec
     DBG_VALUE %5, $noreg
-    S_ENDPGM implicit %4
+    S_ENDPGM 0, implicit %4
 
 ...
 
@@ -93,7 +93,7 @@
     ; GCN: [[V_ADD_I32_e32_:%[0-9]+]]:vgpr_32 = V_ADD_I32_e32 [[S_MOV_B32_]], [[DEF]], implicit-def $vcc, implicit $exec
     ; GCN: [[COPY:%[0-9]+]]:sreg_64_xexec = COPY killed $vcc
     ; GCN: [[V_ADDC_U32_e64_:%[0-9]+]]:vgpr_32, [[V_ADDC_U32_e64_1:%[0-9]+]]:sreg_64_xexec = V_ADDC_U32_e64 [[DEF1]], [[DEF2]], [[COPY]], implicit $exec
-    ; GCN: S_ENDPGM implicit [[V_ADDC_U32_e64_]]
+    ; GCN: S_ENDPGM 0, implicit [[V_ADDC_U32_e64_]]
     %0:sreg_32_xm0 = S_MOV_B32 12345
     %1:vgpr_32 = IMPLICIT_DEF
     %2:vgpr_32 = IMPLICIT_DEF
@@ -101,6 +101,6 @@
 
     %4:vgpr_32, %5:sreg_64_xexec = V_ADD_I32_e64 %0, %1, implicit $exec
     %6:vgpr_32, %7:sreg_64_xexec = V_ADDC_U32_e64 %2, %3, %5, implicit $exec
-    S_ENDPGM implicit %6
+    S_ENDPGM 0, implicit %6
 
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/fold-immediate-operand-shrink.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/fold-immediate-operand-shrink.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/fold-immediate-operand-shrink.mir
@@ -12,11 +12,11 @@
     ; GCN: [[S_MOV_B32_:%[0-9]+]]:sreg_32_xm0 = S_MOV_B32 12345
     ; GCN: [[DEF:%[0-9]+]]:vgpr_32 = IMPLICIT_DEF
     ; GCN: [[V_ADD_I32_e32_:%[0-9]+]]:vgpr_32 = V_ADD_I32_e32 [[S_MOV_B32_]], [[DEF]], implicit-def $vcc, implicit $exec
-    ; GCN: S_ENDPGM implicit [[V_ADD_I32_e32_]]
+    ; GCN: S_ENDPGM 0, implicit [[V_ADD_I32_e32_]]
     %0:sreg_32_xm0 = S_MOV_B32 12345
     %1:vgpr_32 = IMPLICIT_DEF
     %2:vgpr_32, %3:sreg_64 = V_ADD_I32_e64 %0, %1, implicit $exec
-    S_ENDPGM implicit %2
+    S_ENDPGM 0, implicit %2
 
 ...
 
@@ -31,11 +31,11 @@
     ; GCN: [[DEF:%[0-9]+]]:vgpr_32 = IMPLICIT_DEF
     ; GCN: [[S_MOV_B32_:%[0-9]+]]:sreg_32_xm0 = S_MOV_B32 12345
     ; GCN: [[V_ADD_I32_e32_:%[0-9]+]]:vgpr_32 = V_ADD_I32_e32 [[S_MOV_B32_]], [[DEF]], implicit-def $vcc, implicit $exec
-    ; GCN: S_ENDPGM implicit [[V_ADD_I32_e32_]]
+    ; GCN: S_ENDPGM 0, implicit [[V_ADD_I32_e32_]]
     %0:vgpr_32 = IMPLICIT_DEF
     %1:sreg_32_xm0 = S_MOV_B32 12345
     %2:vgpr_32, %3:sreg_64 = V_ADD_I32_e64 %0, %1, implicit $exec
-    S_ENDPGM implicit %2
+    S_ENDPGM 0, implicit %2
 
 ...
 ---
@@ -49,11 +49,11 @@
     ; GCN: [[S_MOV_B32_:%[0-9]+]]:sreg_32_xm0 = S_MOV_B32 12345
     ; GCN: [[DEF:%[0-9]+]]:vgpr_32 = IMPLICIT_DEF
     ; GCN: [[V_ADD_I32_e32_:%[0-9]+]]:vgpr_32 = V_ADD_I32_e32 [[S_MOV_B32_]], [[DEF]], implicit-def $vcc, implicit $exec
-    ; GCN: S_ENDPGM implicit [[V_ADD_I32_e32_]]
+    ; GCN: S_ENDPGM 0, implicit [[V_ADD_I32_e32_]]
     %0:sreg_32_xm0 = S_MOV_B32 12345
     %1:vgpr_32 = IMPLICIT_DEF
     %2:vgpr_32, %3:sreg_64 = V_ADD_I32_e64 %0, %1, implicit $exec
-    S_ENDPGM implicit %2
+    S_ENDPGM 0, implicit %2
 
 ...
 ---
@@ -71,11 +71,11 @@
     ; GCN: [[V_MOV_B32_e32_:%[0-9]+]]:vgpr_32 = V_MOV_B32_e32 12345, implicit $exec
     ; GCN: [[DEF:%[0-9]+]]:sreg_32_xm0 = IMPLICIT_DEF
     ; GCN: [[V_ADD_I32_e64_:%[0-9]+]]:vgpr_32, [[V_ADD_I32_e64_1:%[0-9]+]]:sreg_64 = V_ADD_I32_e64 [[DEF]], [[V_MOV_B32_e32_]], implicit $exec
-    ; GCN: S_ENDPGM implicit [[V_ADD_I32_e64_]]
+    ; GCN: S_ENDPGM 0, implicit [[V_ADD_I32_e64_]]
     %0:vgpr_32 = V_MOV_B32_e32 12345, implicit $exec
     %1:sreg_32_xm0 = IMPLICIT_DEF
     %2:vgpr_32, %3:sreg_64 = V_ADD_I32_e64 %0, %1, implicit $exec
-    S_ENDPGM implicit %2
+    S_ENDPGM 0, implicit %2
 
 ...
 
@@ -90,11 +90,11 @@
     ; GCN: [[DEF:%[0-9]+]]:sreg_32_xm0 = IMPLICIT_DEF
     ; GCN: [[V_MOV_B32_e32_:%[0-9]+]]:vgpr_32 = V_MOV_B32_e32 12345, implicit $exec
     ; GCN: [[V_ADD_I32_e64_:%[0-9]+]]:vgpr_32, [[V_ADD_I32_e64_1:%[0-9]+]]:sreg_64 = V_ADD_I32_e64 [[V_MOV_B32_e32_]], [[DEF]], implicit $exec
-    ; GCN: S_ENDPGM implicit [[V_ADD_I32_e64_]]
+    ; GCN: S_ENDPGM 0, implicit [[V_ADD_I32_e64_]]
     %0:sreg_32_xm0 = IMPLICIT_DEF
     %1:vgpr_32 = V_MOV_B32_e32 12345, implicit $exec
     %2:vgpr_32, %3:sreg_64 = V_ADD_I32_e64 %0, %1, implicit $exec
-    S_ENDPGM implicit %2
+    S_ENDPGM 0, implicit %2
 
 ...
 
@@ -110,12 +110,12 @@
     ; GCN: [[S_MOV_B32_:%[0-9]+]]:sreg_32_xm0 = S_MOV_B32 12345
     ; GCN: [[DEF:%[0-9]+]]:vgpr_32 = IMPLICIT_DEF
     ; GCN: [[V_ADD_I32_e64_:%[0-9]+]]:vgpr_32, [[V_ADD_I32_e64_1:%[0-9]+]]:sreg_64 = V_ADD_I32_e64 [[S_MOV_B32_]], [[DEF]], implicit $exec
-    ; GCN: S_ENDPGM implicit [[V_ADD_I32_e64_]], implicit $vcc
+    ; GCN: S_ENDPGM 0, implicit [[V_ADD_I32_e64_]], implicit $vcc
     $vcc = S_MOV_B64 -1
     %0:sreg_32_xm0 = S_MOV_B32 12345
     %1:vgpr_32 = IMPLICIT_DEF
     %2:vgpr_32, %3:sreg_64 = V_ADD_I32_e64 %0, %1, implicit $exec
-    S_ENDPGM implicit %2, implicit $vcc
+    S_ENDPGM 0, implicit %2, implicit $vcc
 
 ...
 
@@ -134,7 +134,7 @@
   ; GCN:   [[V_ADD_I32_e64_:%[0-9]+]]:vgpr_32, [[V_ADD_I32_e64_1:%[0-9]+]]:sreg_64 = V_ADD_I32_e64 [[S_MOV_B32_]], [[DEF]], implicit $exec
   ; GCN: bb.1:
   ; GCN:   liveins: $vcc
-  ; GCN:   S_ENDPGM implicit [[V_ADD_I32_e64_]], implicit $vcc
+  ; GCN:   S_ENDPGM 0, implicit [[V_ADD_I32_e64_]], implicit $vcc
   bb.0:
     successors: %bb.1
     $vcc = S_MOV_B64 -1
@@ -144,7 +144,7 @@
 
   bb.1:
     liveins: $vcc
-    S_ENDPGM implicit %2, implicit $vcc
+    S_ENDPGM 0, implicit %2, implicit $vcc
 
 ...
 ---
@@ -161,7 +161,7 @@
   ; GCN:   [[V_ADD_I32_e64_:%[0-9]+]]:vgpr_32, [[V_ADD_I32_e64_1:%[0-9]+]]:sreg_64 = V_ADD_I32_e64 [[S_MOV_B32_]], [[DEF]], implicit $exec
   ; GCN: bb.1:
   ; GCN:   liveins: $vcc_lo
-  ; GCN:   S_ENDPGM implicit [[V_ADD_I32_e64_]], implicit $vcc_lo
+  ; GCN:   S_ENDPGM 0, implicit [[V_ADD_I32_e64_]], implicit $vcc_lo
   bb.0:
     successors: %bb.1
     $vcc = S_MOV_B64 -1
@@ -171,7 +171,7 @@
 
   bb.1:
     liveins: $vcc_lo
-    S_ENDPGM implicit %2, implicit $vcc_lo
+    S_ENDPGM 0, implicit %2, implicit $vcc_lo
 
 ...
 ---
@@ -191,7 +191,7 @@
   ; GCN:   [[S_MOV_B32_:%[0-9]+]]:sreg_32_xm0 = S_MOV_B32 12345
   ; GCN:   [[DEF:%[0-9]+]]:vgpr_32 = IMPLICIT_DEF
   ; GCN:   [[V_ADD_I32_e64_:%[0-9]+]]:vgpr_32, [[V_ADD_I32_e64_1:%[0-9]+]]:sreg_64 = V_ADD_I32_e64 [[S_MOV_B32_]], [[DEF]], implicit $exec
-  ; GCN:   S_ENDPGM implicit [[V_ADD_I32_e64_]], implicit $vcc_lo
+  ; GCN:   S_ENDPGM 0, implicit [[V_ADD_I32_e64_]], implicit $vcc_lo
   bb.0:
     successors: %bb.1
     $vcc = S_MOV_B64 -1
@@ -201,7 +201,7 @@
     %0:sreg_32_xm0 = S_MOV_B32 12345
     %1:vgpr_32 = IMPLICIT_DEF
     %2:vgpr_32, %3:sreg_64 = V_ADD_I32_e64 %0, %1, implicit $exec
-    S_ENDPGM implicit %2, implicit $vcc_lo
+    S_ENDPGM 0, implicit %2, implicit $vcc_lo
 
 ...
 ---
@@ -222,7 +222,7 @@
   ; GCN:   [[V_ADD_I32_e64_:%[0-9]+]]:vgpr_32, [[V_ADD_I32_e64_1:%[0-9]+]]:sreg_64 = V_ADD_I32_e64 [[S_MOV_B32_]], [[DEF]], implicit $exec
   ; GCN: bb.2:
   ; GCN:   liveins: $vcc_hi
-  ; GCN:   S_ENDPGM implicit [[V_ADD_I32_e64_]], implicit $vcc_hi
+  ; GCN:   S_ENDPGM 0, implicit [[V_ADD_I32_e64_]], implicit $vcc_hi
   bb.0:
     successors: %bb.1
     $vcc_hi = S_MOV_B32 -1
@@ -236,7 +236,7 @@
   bb.2:
     liveins: $vcc_hi
 
-    S_ENDPGM implicit %2, implicit $vcc_hi
+    S_ENDPGM 0, implicit %2, implicit $vcc_hi
 
 ...
 
@@ -251,11 +251,11 @@
     ; GCN: [[S_MOV_B32_:%[0-9]+]]:sreg_32_xm0 = S_MOV_B32 12345
     ; GCN: [[DEF:%[0-9]+]]:vgpr_32 = IMPLICIT_DEF
     ; GCN: [[V_SUBREV_I32_e32_:%[0-9]+]]:vgpr_32 = V_SUBREV_I32_e32 [[S_MOV_B32_]], [[DEF]], implicit-def $vcc, implicit $exec
-    ; GCN: S_ENDPGM implicit [[V_SUBREV_I32_e32_]]
+    ; GCN: S_ENDPGM 0, implicit [[V_SUBREV_I32_e32_]]
     %0:sreg_32_xm0 = S_MOV_B32 12345
     %1:vgpr_32 = IMPLICIT_DEF
     %2:vgpr_32, %3:sreg_64 = V_SUB_I32_e64 %0, %1, implicit $exec
-    S_ENDPGM implicit %2
+    S_ENDPGM 0, implicit %2
 
 ...
 
@@ -270,11 +270,11 @@
     ; GCN: [[DEF:%[0-9]+]]:vgpr_32 = IMPLICIT_DEF
     ; GCN: [[S_MOV_B32_:%[0-9]+]]:sreg_32_xm0 = S_MOV_B32 12345
     ; GCN: [[V_SUB_I32_e32_:%[0-9]+]]:vgpr_32 = V_SUB_I32_e32 [[S_MOV_B32_]], [[DEF]], implicit-def $vcc, implicit $exec
-    ; GCN: S_ENDPGM implicit [[V_SUB_I32_e32_]]
+    ; GCN: S_ENDPGM 0, implicit [[V_SUB_I32_e32_]]
     %0:vgpr_32 = IMPLICIT_DEF
     %1:sreg_32_xm0 = S_MOV_B32 12345
     %2:vgpr_32, %3:sreg_64 = V_SUB_I32_e64 %0, %1, implicit $exec
-    S_ENDPGM implicit %2
+    S_ENDPGM 0, implicit %2
 
 ...
 
@@ -289,11 +289,11 @@
     ; GCN: [[S_MOV_B32_:%[0-9]+]]:sreg_32_xm0 = S_MOV_B32 12345
     ; GCN: [[DEF:%[0-9]+]]:vgpr_32 = IMPLICIT_DEF
     ; GCN: [[V_SUB_I32_e32_:%[0-9]+]]:vgpr_32 = V_SUB_I32_e32 [[S_MOV_B32_]], [[DEF]], implicit-def $vcc, implicit $exec
-    ; GCN: S_ENDPGM implicit [[V_SUB_I32_e32_]]
+    ; GCN: S_ENDPGM 0, implicit [[V_SUB_I32_e32_]]
     %0:sreg_32_xm0 = S_MOV_B32 12345
     %1:vgpr_32 = IMPLICIT_DEF
     %2:vgpr_32, %3:sreg_64 = V_SUBREV_I32_e64 %0, %1, implicit $exec
-    S_ENDPGM implicit %2
+    S_ENDPGM 0, implicit %2
 
 ...
 
@@ -308,11 +308,11 @@
     ; GCN: [[DEF:%[0-9]+]]:vgpr_32 = IMPLICIT_DEF
     ; GCN: [[S_MOV_B32_:%[0-9]+]]:sreg_32_xm0 = S_MOV_B32 12345
     ; GCN: [[V_SUBREV_I32_e32_:%[0-9]+]]:vgpr_32 = V_SUBREV_I32_e32 [[S_MOV_B32_]], [[DEF]], implicit-def $vcc, implicit $exec
-    ; GCN: S_ENDPGM implicit [[V_SUBREV_I32_e32_]]
+    ; GCN: S_ENDPGM 0, implicit [[V_SUBREV_I32_e32_]]
     %0:vgpr_32 = IMPLICIT_DEF
     %1:sreg_32_xm0 = S_MOV_B32 12345
     %2:vgpr_32, %3:sreg_64 = V_SUBREV_I32_e64 %0, %1, implicit $exec
-    S_ENDPGM implicit %2
+    S_ENDPGM 0, implicit %2
 
 ...
 
@@ -331,7 +331,7 @@
   ; GCN:   [[DEF:%[0-9]+]]:vgpr_32 = IMPLICIT_DEF
   ; GCN:   [[V_ADD_I32_e32_:%[0-9]+]]:vgpr_32 = V_ADD_I32_e32 [[S_MOV_B32_]], [[DEF]], implicit-def $vcc, implicit $exec
   ; GCN: bb.1:
-  ; GCN:   S_ENDPGM implicit [[V_ADD_I32_e32_]]
+  ; GCN:   S_ENDPGM 0, implicit [[V_ADD_I32_e32_]]
   bb.0:
     successors: %bb.1
 
@@ -372,7 +372,7 @@
     S_NOP 0
 
   bb.1:
-    S_ENDPGM implicit %2
+    S_ENDPGM 0, implicit %2
 
 ...
 
@@ -392,7 +392,7 @@
   ; GCN:   [[DEF:%[0-9]+]]:vgpr_32 = IMPLICIT_DEF
   ; GCN:   [[V_ADD_I32_e32_:%[0-9]+]]:vgpr_32 = V_ADD_I32_e32 [[S_MOV_B32_]], [[DEF]], implicit-def $vcc, implicit $exec
   ; GCN: bb.1:
-  ; GCN:   S_ENDPGM implicit [[V_ADD_I32_e32_]]
+  ; GCN:   S_ENDPGM 0, implicit [[V_ADD_I32_e32_]]
   bb.0:
     successors: %bb.1
 
@@ -404,7 +404,7 @@
     S_NOP 0
 
   bb.1:
-    S_ENDPGM implicit %2
+    S_ENDPGM 0, implicit %2
 
 ...
 ---
@@ -449,7 +449,7 @@
     ; GCN: DBG_VALUE $noreg, 0
     ; GCN: DBG_VALUE $noreg, 0
     ; GCN: [[V_ADD_I32_e32_:%[0-9]+]]:vgpr_32 = V_ADD_I32_e32 [[S_MOV_B32_]], [[DEF]], implicit-def $vcc, implicit $exec
-    ; GCN: S_ENDPGM implicit [[V_ADD_I32_e32_]]
+    ; GCN: S_ENDPGM 0, implicit [[V_ADD_I32_e32_]]
     %0:sreg_32_xm0 = S_MOV_B32 12345
     %1:vgpr_32 = IMPLICIT_DEF
     DBG_VALUE $noreg, 0
@@ -481,7 +481,7 @@
     DBG_VALUE $noreg, 0
     DBG_VALUE $noreg, 0
     %2:vgpr_32, %3:sreg_64 = V_ADD_I32_e64 %0, %1, implicit $exec
-    S_ENDPGM implicit %2
+    S_ENDPGM 0, implicit %2
 
 ...
 ---
@@ -526,7 +526,7 @@
     ; GCN: DBG_VALUE $noreg, 0
     ; GCN: DBG_VALUE $noreg, 0
     ; GCN: DBG_VALUE $noreg, 0
-    ; GCN: S_ENDPGM implicit [[V_ADD_I32_e32_]]
+    ; GCN: S_ENDPGM 0, implicit [[V_ADD_I32_e32_]]
     %0:sreg_32_xm0 = S_MOV_B32 12345
     %1:vgpr_32 = IMPLICIT_DEF
     S_NOP 0
@@ -587,6 +587,6 @@
     DBG_VALUE $noreg, 0
     DBG_VALUE $noreg, 0
     $vcc = S_MOV_B64 0
-    S_ENDPGM implicit %2
+    S_ENDPGM 0, implicit %2
 
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/fold-immediate-output-mods.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/fold-immediate-output-mods.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/fold-immediate-output-mods.mir
@@ -67,7 +67,7 @@
     %24 = V_MAC_F32_e64 0, killed %19, 0, killed %21, 0, %23, 1, 0, implicit $exec
     %26 = COPY %29
     BUFFER_STORE_DWORD_ADDR64 killed %24, %26, killed %18, 0, 0, 0, 0, 0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -138,7 +138,7 @@
     %24 = V_MAC_F32_e64 0, killed %19, 0, killed %21, 0, %23, 0, 2, implicit $exec
     %26 = COPY %29
     BUFFER_STORE_DWORD_ADDR64 killed %24, %26, killed %18, 0, 0, 0, 0, 0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -209,7 +209,7 @@
     %24 = V_MAD_F32 0, killed %19, 0, killed %21, 0, %23, 1, 0, implicit $exec
     %26 = COPY %29
     BUFFER_STORE_DWORD_ADDR64 killed %24, %26, killed %18, 0, 0, 0, 0, 0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -280,6 +280,6 @@
     %24 = V_MAD_F32 0, killed %19, 0, killed %21, 0, %23, 0, 1, implicit $exec
     %26 = COPY %29
     BUFFER_STORE_DWORD_ADDR64 killed %24, %26, killed %18, 0, 0, 0, 0, 0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/fold-implicit-operand.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/fold-implicit-operand.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/fold-implicit-operand.mir
@@ -4,11 +4,11 @@
 # implicit use
 
 # CHECK: %0:vgpr_32 = V_MOV_B32_e32 0, implicit $exec
-# CHECK-NEXT: S_ENDPGM implicit %0
+# CHECK-NEXT: S_ENDPGM 0, implicit %0
 name: fold_imm_implicit_operand
 body:             |
   bb.0:
     %0:vgpr_32 = V_MOV_B32_e32 0, implicit $exec
-    S_ENDPGM implicit %0
+    S_ENDPGM 0, implicit %0
 
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/fold-multiple.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/fold-multiple.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/fold-multiple.mir
@@ -35,6 +35,6 @@
     %4 = V_AND_B32_e64 killed %2, killed %3, implicit $exec
     %5 = IMPLICIT_DEF
     BUFFER_STORE_DWORD_OFFSET killed %4, killed %5, 0, 0, 0, 0, 0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/global-load-store-atomics.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/global-load-store-atomics.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/global-load-store-atomics.mir
@@ -245,5 +245,5 @@
     GLOBAL_STORE_DWORDX2 %11, %78, 0, 0, 0, implicit $exec :: (volatile store 4 into `i32 addrspace(1)* undef`, addrspace 1)
     GLOBAL_ATOMIC_CMPSWAP_X2 %11, %80, 16, 0, implicit $exec :: (volatile load store seq_cst 4, addrspace 1)
 
-    S_ENDPGM
+    S_ENDPGM 0
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/hazard-buffer-store-v-interp.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/hazard-buffer-store-v-interp.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/hazard-buffer-store-v-interp.mir
@@ -14,6 +14,6 @@
   
     BUFFER_STORE_DWORDX4_OFFSET_exact killed $vgpr7_vgpr8_vgpr9_vgpr10, $sgpr4_sgpr5_sgpr6_sgpr7, 0, 96, 0, 0, 0, implicit $exec
     $vgpr7 = V_INTERP_P1_F32 $vgpr0, 0, 0, implicit $m0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/hazard-inlineasm.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/hazard-inlineasm.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/hazard-inlineasm.mir
@@ -18,7 +18,7 @@
   bb.0:
    FLAT_STORE_DWORDX4 $vgpr49_vgpr50, $vgpr26_vgpr27_vgpr28_vgpr29, 0, 0, 0, implicit $exec, implicit $flat_scr
    INLINEASM &"v_mad_u64_u32 $0, $1, $2, $3, $4", 0, 2621450, def $vgpr26_vgpr27, 2818058, def dead $sgpr14_sgpr15, 589833, $sgpr12, 327689, killed $vgpr51, 2621449, $vgpr46_vgpr47
-   S_ENDPGM
+   S_ENDPGM 0
 ...
 
 
Index: llvm/trunk/test/CodeGen/AMDGPU/hazard-kill.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/hazard-kill.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/hazard-kill.mir
@@ -27,6 +27,6 @@
     renamable $vgpr0 = V_INTERP_MOV_F32 2, 0, 0, implicit $m0, implicit $exec
     renamable $sgpr0 = S_MOV_B32 0
 
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/hazard.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/hazard.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/hazard.mir
@@ -81,7 +81,7 @@
     $vgpr5 = IMPLICIT_DEF
     $vgpr6 = IMPLICIT_DEF
     S_SENDMSG 3, implicit $exec, implicit $m0
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 
 # GCN-LABEL: name: hazard-lookahead-dbg-value
@@ -102,7 +102,7 @@
     DBG_VALUE 5
     DBG_VALUE 6
     S_SENDMSG 3, implicit $exec, implicit $m0
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 
 # GCN-LABEL: name: hazard-lookahead-dbg-label
@@ -123,5 +123,5 @@
     DBG_LABEL 5
     DBG_LABEL 6
     S_SENDMSG 3, implicit $exec, implicit $m0
-    S_ENDPGM
+    S_ENDPGM 0
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/insert-skip-from-vcc.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/insert-skip-from-vcc.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/insert-skip-from-vcc.mir
@@ -17,7 +17,7 @@
     $sgpr0_sgpr1 = S_MOV_B64 -1
     $vcc = S_AND_B64 $exec, killed $sgpr0_sgpr1, implicit-def dead $scc
     S_CBRANCH_VCCZ %bb.1, implicit killed $vcc
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # GCN-LABEL: name: and_execz_imm_vccz
@@ -34,7 +34,7 @@
   bb.2:
     $vcc = S_AND_B64 $exec, -1, implicit-def dead $scc
     S_CBRANCH_VCCZ %bb.1, implicit killed $vcc
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # GCN-LABEL: name: and_execnz_imm_vccnz
@@ -51,7 +51,7 @@
   bb.2:
     $vcc = S_AND_B64 $exec, -1, implicit-def dead $scc
     S_CBRANCH_VCCNZ %bb.1, implicit killed $vcc
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # GCN-LABEL: name: and_execz_imm_vccz_live_scc
@@ -68,7 +68,7 @@
   bb.2:
     $vcc = S_AND_B64 $exec, -1, implicit-def $scc
     S_CBRANCH_VCCZ %bb.1, implicit killed $vcc
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # GCN-LABEL: name: and_execz_mov_vccz_live_scc
@@ -87,7 +87,7 @@
     $sgpr0_sgpr1 = S_MOV_B64 -1
     $vcc = S_AND_B64 $exec, killed $sgpr0_sgpr1, implicit-def $scc
     S_CBRANCH_VCCZ %bb.1, implicit killed $vcc
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # GCN-LABEL: name: and_execz_mov_vccz_live_sreg
@@ -105,7 +105,7 @@
     $sgpr0_sgpr1 = S_MOV_B64 -1
     $vcc = S_AND_B64 $exec, $sgpr0_sgpr1, implicit-def dead $scc
     S_CBRANCH_VCCZ %bb.1, implicit killed $vcc
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # GCN-LABEL: name: and_execz_mov_vccz_live_sreg_commute
@@ -123,7 +123,7 @@
     $sgpr0_sgpr1 = S_MOV_B64 -1
     $vcc = S_AND_B64 $sgpr0_sgpr1, $exec, implicit-def dead $scc
     S_CBRANCH_VCCZ %bb.1, implicit killed $vcc
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # GCN-LABEL: name: and_execz_mov_vccz_live_scc_commute
@@ -142,7 +142,7 @@
     $sgpr0_sgpr1 = S_MOV_B64 -1
     $vcc = S_AND_B64 killed $sgpr0_sgpr1, $exec, implicit-def $scc
     S_CBRANCH_VCCZ %bb.1, implicit killed $vcc
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # GCN-LABEL: name: and_execz_mov_vccz_commute
@@ -161,12 +161,12 @@
     $sgpr0_sgpr1 = S_MOV_B64 -1
     $vcc = S_AND_B64 killed $sgpr0_sgpr1, $exec, implicit-def dead $scc
     S_CBRANCH_VCCZ %bb.1, implicit killed $vcc
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # GCN-LABEL: name: and_execz_mov_exec_vccz
 # GCN:      $exec = S_MOV_B64 -1
-# GCN-NEXT: S_ENDPGM
+# GCN-NEXT: S_ENDPGM 0
 name:            and_execz_mov_exec_vccz
 body:             |
   bb.0:
@@ -179,7 +179,7 @@
     $exec = S_MOV_B64 -1
     $vcc = S_AND_B64 $exec, $exec, implicit-def dead $scc
     S_CBRANCH_VCCZ %bb.1, implicit killed $vcc
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # GCN-LABEL: name: and_execz_mov_exec_vccnz
@@ -197,7 +197,7 @@
     $exec = S_MOV_B64 -1
     $vcc = S_AND_B64 $exec, $exec, implicit-def dead $scc
     S_CBRANCH_VCCNZ %bb.1, implicit killed $vcc
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # GCN-LABEL: name: and_execz_mov_vccz_reads_sreg_early
@@ -217,7 +217,7 @@
     $sgpr2 = S_MOV_B32 $sgpr1
     $vcc = S_AND_B64 $exec, killed $sgpr0_sgpr1, implicit-def dead $scc
     S_CBRANCH_VCCZ %bb.1, implicit killed $vcc
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # GCN-LABEL: name: and_execz_mov_vccz_reads_sreg_late
@@ -237,7 +237,7 @@
     $vcc = S_AND_B64 $exec, $sgpr0_sgpr1, implicit-def dead $scc
     $sgpr2 = S_MOV_B32 $sgpr1
     S_CBRANCH_VCCZ %bb.1, implicit killed $vcc
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 # GCN-LABEL: name: and_execz_mov_vccz_reads_writes_sreg_early
 # GCN:      $sgpr0_sgpr1 = S_MOV_B64 -1
@@ -257,7 +257,7 @@
     $sgpr1 = S_MOV_B32 $sgpr0
     $vcc = S_AND_B64 $exec, killed $sgpr0_sgpr1, implicit-def dead $scc
     S_CBRANCH_VCCZ %bb.1, implicit killed $vcc
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # GCN-LABEL: name: and_execz_mov_vccz_reads_cond
@@ -277,7 +277,7 @@
     $vcc = S_AND_B64 $exec, killed $sgpr0_sgpr1, implicit-def dead $scc
     $sgpr2 = S_MOV_B32 $vcc_lo
     S_CBRANCH_VCCZ %bb.1, implicit killed $vcc
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # GCN-LABEL: name: and_execz_mov_vccz_modifies_sreg
@@ -298,13 +298,13 @@
     $sgpr0 = S_MOV_B32 0
     $vcc = S_AND_B64 $exec, killed $sgpr0_sgpr1, implicit-def dead $scc
     S_CBRANCH_VCCZ %bb.1, implicit killed $vcc
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # GCN-LABEL: name: and_execz_imm_vccz_liveout_scc
 # GCN:      $vcc = S_AND_B64 $exec, -1, implicit-def $scc
 # GCN-NEXT: S_CBRANCH_EXECZ %bb.1, implicit $exec
-# GCN-NEXT  S_ENDPGM implicit $scc
+# GCN-NEXT  S_ENDPGM 0, implicit $scc
 name:            and_execz_imm_vccz_liveout_scc
 body:             |
   bb.0:
@@ -316,5 +316,5 @@
   bb.2:
     $vcc = S_AND_B64 $exec, -1, implicit-def $scc
     S_CBRANCH_VCCZ %bb.1, implicit killed $vcc
-    S_ENDPGM implicit $scc
+    S_ENDPGM 0, implicit $scc
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/insert-skips-kill-uncond.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/insert-skips-kill-uncond.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/insert-skips-kill-uncond.mir
@@ -18,10 +18,10 @@
 
 # CHECK: bb.3:
 # CHECK-NEXT: EXP_DONE
-# CHECK: S_ENDPGM
+# CHECK: S_ENDPGM 0
 
 # CHECK: bb.2:
-# CHECK: S_ENDPGM
+# CHECK: S_ENDPGM 0
 
 name: kill_uncond_branch
 
@@ -37,4 +37,4 @@
     S_BRANCH %bb.2
 
   bb.2:
-    S_ENDPGM
+    S_ENDPGM 0
Index: llvm/trunk/test/CodeGen/AMDGPU/inserted-wait-states.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/inserted-wait-states.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/inserted-wait-states.mir
@@ -95,7 +95,7 @@
   bb.3:
     $vgpr4, $vcc = V_DIV_SCALE_F32 $vgpr1, $vgpr1, $vgpr3, implicit $exec
     $vgpr0 = V_DIV_FMAS_F32 0, $vgpr1, 0, $vgpr2, 0, $vgpr3, 0, 0, implicit $vcc, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 
@@ -146,7 +146,7 @@
   bb.3:
     S_SETREG_B32 $sgpr0, 0
     $sgpr1 = S_GETREG_B32 1
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 
 ...
@@ -185,7 +185,7 @@
   bb.2:
     S_SETREG_B32 $sgpr0, 1
     S_SETREG_B32 $sgpr1, 0
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 
 ...
@@ -257,7 +257,7 @@
     $vgpr3 = V_MOV_B32_e32 0, implicit $exec
     FLAT_ATOMIC_FCMPSWAP_X2 $vgpr0_vgpr1, $vgpr2_vgpr3_vgpr4_vgpr5, 0, 0, implicit $exec, implicit $flat_scr
     $vgpr3 = V_MOV_B32_e32 0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 
@@ -319,7 +319,7 @@
   bb.3:
     $vgpr0,implicit $vcc = V_ADD_I32_e32 $vgpr1, $vgpr2, implicit $vcc, implicit $exec
     $vgpr4 = V_WRITELANE_B32 $sgpr4, $vcc_lo, $vgpr4
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 
@@ -348,7 +348,7 @@
   bb.1:
     S_SETREG_B32 $sgpr0, 0
     S_RFE_B64 $sgpr2_sgpr3
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 
@@ -377,7 +377,7 @@
   bb.1:
     $sgpr0 = S_MOV_FED_B32 $sgpr0
     $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 
@@ -427,7 +427,7 @@
   bb.3:
     $m0 = S_MOV_B32 0
     $sgpr0_sgpr1 = S_MOVRELD_B64 $sgpr0_sgpr1, implicit $m0
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 
 ...
@@ -476,7 +476,7 @@
   bb.3:
     $m0 = S_MOV_B32 0
     $vgpr0 = V_INTERP_MOV_F32 0, 0, 0, implicit $m0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 
 ...
@@ -510,7 +510,7 @@
   bb.1:
     implicit $exec, implicit $vcc = V_CMPX_EQ_I32_e32 $vgpr0, $vgpr1, implicit $exec
     $vgpr3 = V_MOV_B32_dpp $vgpr3, $vgpr0, 0, 15, 15, 0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 name:            mov_fed_hazard_crash_on_dbg_value
@@ -557,6 +557,6 @@
     $sgpr8 = S_MOV_B32 $sgpr4, implicit killed $sgpr4_sgpr5
     $vgpr0 = V_MOV_B32_e32 killed $sgpr8, implicit $exec
     BUFFER_STORE_DWORD_OFFSET $vgpr0, $sgpr0_sgpr1_sgpr2_sgpr3, $sgpr9, 0, 0, 0, 0, implicit $exec :: (store 4 into %ir.A.addr)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/invert-br-undef-vcc.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/invert-br-undef-vcc.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/invert-br-undef-vcc.mir
@@ -81,6 +81,6 @@
     $sgpr3 = S_MOV_B32 61440
     $sgpr2 = S_MOV_B32 -1
     BUFFER_STORE_DWORD_OFFSET killed $vgpr0, killed $sgpr0_sgpr1_sgpr2_sgpr3, 0, 0, 0, 0, 0, implicit $exec :: (store 4 into %ir.out)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/memory-legalizer-atomic-insert-end.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/memory-legalizer-atomic-insert-end.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/memory-legalizer-atomic-insert-end.mir
@@ -110,7 +110,7 @@
     liveins: $sgpr2_sgpr3
 
     $exec = S_OR_B64 $exec, killed $sgpr2_sgpr3, implicit-def $scc
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 
Index: llvm/trunk/test/CodeGen/AMDGPU/memory-legalizer-invalid-addrspace.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/memory-legalizer-invalid-addrspace.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/memory-legalizer-invalid-addrspace.mir
@@ -15,7 +15,7 @@
     $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
     $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
     FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4 into `i32* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -31,7 +31,7 @@
     $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
     $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
     FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, 0, implicit $exec, implicit $flat_scr :: (volatile store syncscope("agent") seq_cst 4 into `i32 addrspace(42)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -48,7 +48,7 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr5, implicit $exec, implicit $exec
     $vgpr2 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
     FLAT_ATOMIC_CMPSWAP killed renamable $vgpr2_vgpr3, killed renamable $vgpr0_vgpr1, 0, 0, implicit $exec, implicit $flat_scr :: (volatile load store syncscope("workgroup") seq_cst seq_cst 4 on `i32 addrspace(42)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -64,6 +64,6 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr3, implicit $exec, implicit $sgpr2_sgpr3, implicit $exec
     $vgpr2 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     FLAT_ATOMIC_SWAP killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, implicit $exec, implicit $flat_scr :: (volatile load store syncscope("wavefront") seq_cst 4 on `i32 addrspace(42)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/memory-legalizer-local.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/memory-legalizer-local.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/memory-legalizer-local.mir
@@ -21,7 +21,7 @@
     $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
     $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
     FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4 into `i32* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -45,7 +45,7 @@
     $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
     $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
     FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4 into `i32* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -69,7 +69,7 @@
     $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
     $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
     FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4 into `i32* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -93,7 +93,7 @@
     $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
     $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
     FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4 into `i32* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -117,7 +117,7 @@
     $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
     $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
     FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4 into `i32* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -141,7 +141,7 @@
     $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
     $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
     FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4 into `i32* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -165,7 +165,7 @@
     $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
     $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
     FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4 into `i32* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -189,7 +189,7 @@
     $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
     $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
     FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4 into `i32* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -213,7 +213,7 @@
     $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
     $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
     FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4 into `i32* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -237,7 +237,7 @@
     $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
     $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
     FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4 into `i32* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -261,7 +261,7 @@
     $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
     $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
     FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4 into `i32* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -285,7 +285,7 @@
     $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
     $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
     FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4 into `i32* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -309,7 +309,7 @@
     $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
     $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
     FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4 into `i32* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -333,7 +333,7 @@
     $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
     $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
     FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4 into `i32* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -357,7 +357,7 @@
     $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
     $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
     FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4 into `i32* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -381,7 +381,7 @@
     $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
     $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
     FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4 into `i32* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -405,7 +405,7 @@
     $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
     $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
     FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4 into `i32* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -429,7 +429,7 @@
     $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
     $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
     FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4 into `i32* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -453,7 +453,7 @@
     $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
     $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
     FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4 into `i32* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -477,7 +477,7 @@
     $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
     $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
     FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4 into `i32* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -488,7 +488,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRITE_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            store_singlethread_unordered
 body:             |
@@ -499,7 +499,7 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     DS_WRITE_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 0, implicit $m0, implicit $exec :: (volatile store syncscope("singlethread") unordered 4 into `i32 addrspace(3)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -510,7 +510,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRITE_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            store_singlethread_monotonic
 body:             |
@@ -521,7 +521,7 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     DS_WRITE_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 0, implicit $m0, implicit $exec :: (volatile store syncscope("singlethread") monotonic 4 into `i32 addrspace(3)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -532,7 +532,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRITE_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            store_singlethread_release
 body:             |
@@ -543,7 +543,7 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     DS_WRITE_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 0, implicit $m0, implicit $exec :: (volatile store syncscope("singlethread") release 4 into `i32 addrspace(3)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -554,7 +554,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRITE_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            store_singlethread_seq_cst
 body:             |
@@ -565,7 +565,7 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     DS_WRITE_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 0, implicit $m0, implicit $exec :: (volatile store syncscope("singlethread") seq_cst 4 into `i32 addrspace(3)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -576,7 +576,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRITE_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            store_wavefront_unordered
 body:             |
@@ -587,7 +587,7 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     DS_WRITE_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 0, implicit $m0, implicit $exec :: (volatile store syncscope("wavefront") unordered 4 into `i32 addrspace(3)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -598,7 +598,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRITE_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            store_wavefront_monotonic
 body:             |
@@ -609,7 +609,7 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     DS_WRITE_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 0, implicit $m0, implicit $exec :: (volatile store syncscope("wavefront") monotonic 4 into `i32 addrspace(3)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -620,7 +620,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRITE_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            store_wavefront_release
 body:             |
@@ -631,7 +631,7 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     DS_WRITE_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 0, implicit $m0, implicit $exec :: (volatile store syncscope("wavefront") release 4 into `i32 addrspace(3)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -642,7 +642,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRITE_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            store_wavefront_seq_cst
 body:             |
@@ -653,7 +653,7 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     DS_WRITE_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 0, implicit $m0, implicit $exec :: (volatile store syncscope("wavefront") seq_cst 4 into `i32 addrspace(3)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -664,7 +664,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRITE_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            store_workgroup_unordered
 body:             |
@@ -675,7 +675,7 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     DS_WRITE_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 0, implicit $m0, implicit $exec :: (volatile store syncscope("workgroup") unordered 4 into `i32 addrspace(3)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -686,7 +686,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRITE_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            store_workgroup_monotonic
 body:             |
@@ -697,7 +697,7 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     DS_WRITE_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 0, implicit $m0, implicit $exec :: (volatile store syncscope("workgroup") monotonic 4 into `i32 addrspace(3)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -708,7 +708,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRITE_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            store_workgroup_release
 body:             |
@@ -719,7 +719,7 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     DS_WRITE_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 0, implicit $m0, implicit $exec :: (volatile store syncscope("workgroup") release 4 into `i32 addrspace(3)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -730,7 +730,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRITE_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            store_workgroup_seq_cst
 body:             |
@@ -741,7 +741,7 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     DS_WRITE_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 0, implicit $m0, implicit $exec :: (volatile store syncscope("workgroup") seq_cst 4 into `i32 addrspace(3)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -752,7 +752,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRITE_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            store_agent_unordered
 body:             |
@@ -763,7 +763,7 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     DS_WRITE_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 0, implicit $m0, implicit $exec :: (volatile store syncscope("agent") unordered 4 into `i32 addrspace(3)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -774,7 +774,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRITE_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            store_agent_monotonic
 body:             |
@@ -785,7 +785,7 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     DS_WRITE_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 0, implicit $m0, implicit $exec :: (volatile store syncscope("agent") monotonic 4 into `i32 addrspace(3)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -796,7 +796,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRITE_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            store_agent_release
 body:             |
@@ -807,7 +807,7 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     DS_WRITE_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 0, implicit $m0, implicit $exec :: (volatile store syncscope("agent") release 4 into `i32 addrspace(3)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -818,7 +818,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRITE_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            store_agent_seq_cst
 body:             |
@@ -829,7 +829,7 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     DS_WRITE_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 0, implicit $m0, implicit $exec :: (volatile store syncscope("agent") seq_cst 4 into `i32 addrspace(3)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -840,7 +840,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRITE_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            store_system_unordered
 body:             |
@@ -851,7 +851,7 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     DS_WRITE_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 0, implicit $m0, implicit $exec :: (volatile store unordered 4 into `i32 addrspace(3)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -862,7 +862,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRITE_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            store_system_monotonic
 body:             |
@@ -873,7 +873,7 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     DS_WRITE_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 0, implicit $m0, implicit $exec :: (volatile store monotonic 4 into `i32 addrspace(3)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -884,7 +884,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRITE_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            store_system_release
 body:             |
@@ -895,7 +895,7 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     DS_WRITE_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 0, implicit $m0, implicit $exec :: (volatile store release 4 into `i32 addrspace(3)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -906,7 +906,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRITE_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            store_system_seq_cst
 body:             |
@@ -917,7 +917,7 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     DS_WRITE_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 0, implicit $m0, implicit $exec :: (volatile store seq_cst 4 into `i32 addrspace(3)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -928,7 +928,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRXCHG_RTN_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            atomicrmw_singlethread_unordered
 body:             |
@@ -939,7 +939,7 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     $vgpr2 = DS_WRXCHG_RTN_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 0, implicit $m0, implicit $exec :: (volatile store syncscope("singlethread") unordered 4 into `i32 addrspace(3)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -950,7 +950,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRXCHG_RTN_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            atomicrmw_singlethread_monotonic
 body:             |
@@ -961,7 +961,7 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     $vgpr2 = DS_WRXCHG_RTN_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 0, implicit $m0, implicit $exec :: (volatile store syncscope("singlethread") monotonic 4 into `i32 addrspace(3)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -972,7 +972,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRXCHG_RTN_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            atomicrmw_singlethread_acquire
 body:             |
@@ -983,7 +983,7 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     $vgpr2 = DS_WRXCHG_RTN_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 0, implicit $m0, implicit $exec :: (volatile store syncscope("singlethread") acquire 4 into `i32 addrspace(3)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -994,7 +994,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRXCHG_RTN_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            atomicrmw_singlethread_release
 body:             |
@@ -1005,7 +1005,7 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     $vgpr2 = DS_WRXCHG_RTN_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 0, implicit $m0, implicit $exec :: (volatile store syncscope("singlethread") release 4 into `i32 addrspace(3)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -1016,7 +1016,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRXCHG_RTN_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            atomicrmw_singlethread_acq_rel
 body:             |
@@ -1027,7 +1027,7 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     $vgpr2 = DS_WRXCHG_RTN_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 0, implicit $m0, implicit $exec :: (volatile store syncscope("singlethread") acq_rel 4 into `i32 addrspace(3)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -1038,7 +1038,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRXCHG_RTN_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            atomicrmw_singlethread_seq_cst
 body:             |
@@ -1049,6 +1049,6 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     $vgpr2 = DS_WRXCHG_RTN_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 0, implicit $m0, implicit $exec :: (volatile store syncscope("singlethread") seq_cst 4 into `i32 addrspace(3)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/memory-legalizer-multiple-mem-operands-atomics.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/memory-legalizer-multiple-mem-operands-atomics.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/memory-legalizer-multiple-mem-operands-atomics.mir
@@ -60,6 +60,6 @@
     $vgpr2 = V_MOV_B32_e32 killed $sgpr5, implicit $exec, implicit $sgpr4_sgpr5, implicit $exec
     S_WAITCNT 3952
     FLAT_STORE_DWORD killed $vgpr1_vgpr2, killed $vgpr0, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4 into `i32 addrspace(1)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/memory-legalizer-multiple-mem-operands-nontemporal-1.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/memory-legalizer-multiple-mem-operands-nontemporal-1.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/memory-legalizer-multiple-mem-operands-nontemporal-1.mir
@@ -154,6 +154,6 @@
     $vgpr2 = V_MOV_B32_e32 killed $sgpr5, implicit $exec, implicit $sgpr4_sgpr5, implicit $exec
     S_WAITCNT 3952
     FLAT_STORE_DWORD killed $vgpr1_vgpr2, killed $vgpr0, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4 into %ir.out)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/memory-legalizer-multiple-mem-operands-nontemporal-2.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/memory-legalizer-multiple-mem-operands-nontemporal-2.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/memory-legalizer-multiple-mem-operands-nontemporal-2.mir
@@ -134,6 +134,6 @@
     $vgpr2 = V_MOV_B32_e32 killed $sgpr5, implicit $exec, implicit $sgpr4_sgpr5, implicit $exec
     S_WAITCNT 3952
     FLAT_STORE_DWORD killed $vgpr1_vgpr2, killed $vgpr0, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4 into %ir.out)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/memory-legalizer-region.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/memory-legalizer-region.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/memory-legalizer-region.mir
@@ -21,7 +21,7 @@
     $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
     $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
     FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4 into `i32* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -45,7 +45,7 @@
     $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
     $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
     FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4 into `i32* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -69,7 +69,7 @@
     $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
     $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
     FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4 into `i32* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -93,7 +93,7 @@
     $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
     $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
     FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4 into `i32* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -117,7 +117,7 @@
     $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
     $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
     FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4 into `i32* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -141,7 +141,7 @@
     $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
     $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
     FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4 into `i32* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -165,7 +165,7 @@
     $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
     $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
     FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4 into `i32* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -189,7 +189,7 @@
     $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
     $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
     FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4 into `i32* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -213,7 +213,7 @@
     $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
     $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
     FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4 into `i32* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -237,7 +237,7 @@
     $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
     $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
     FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4 into `i32* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -261,7 +261,7 @@
     $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
     $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
     FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4 into `i32* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -285,7 +285,7 @@
     $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
     $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
     FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4 into `i32* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -309,7 +309,7 @@
     $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
     $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
     FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4 into `i32* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -333,7 +333,7 @@
     $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
     $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
     FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4 into `i32* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -357,7 +357,7 @@
     $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
     $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
     FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4 into `i32* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -381,7 +381,7 @@
     $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
     $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
     FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4 into `i32* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -405,7 +405,7 @@
     $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
     $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
     FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4 into `i32* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -429,7 +429,7 @@
     $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
     $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
     FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4 into `i32* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -453,7 +453,7 @@
     $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
     $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
     FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4 into `i32* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -477,7 +477,7 @@
     $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
     $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
     FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4 into `i32* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -488,7 +488,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRITE_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            store_singlethread_unordered
 body:             |
@@ -499,7 +499,7 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     DS_WRITE_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 1, implicit $m0, implicit $exec :: (volatile store syncscope("singlethread") unordered 4 into `i32 addrspace(2)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -510,7 +510,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRITE_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            store_singlethread_monotonic
 body:             |
@@ -521,7 +521,7 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     DS_WRITE_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 1, implicit $m0, implicit $exec :: (volatile store syncscope("singlethread") monotonic 4 into `i32 addrspace(2)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -532,7 +532,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRITE_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            store_singlethread_release
 body:             |
@@ -543,7 +543,7 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     DS_WRITE_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 1, implicit $m0, implicit $exec :: (volatile store syncscope("singlethread") release 4 into `i32 addrspace(2)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -554,7 +554,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRITE_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            store_singlethread_seq_cst
 body:             |
@@ -565,7 +565,7 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     DS_WRITE_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 1, implicit $m0, implicit $exec :: (volatile store syncscope("singlethread") seq_cst 4 into `i32 addrspace(2)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -576,7 +576,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRITE_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            store_wavefront_unordered
 body:             |
@@ -587,7 +587,7 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     DS_WRITE_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 1, implicit $m0, implicit $exec :: (volatile store syncscope("wavefront") unordered 4 into `i32 addrspace(2)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -598,7 +598,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRITE_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            store_wavefront_monotonic
 body:             |
@@ -609,7 +609,7 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     DS_WRITE_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 1, implicit $m0, implicit $exec :: (volatile store syncscope("wavefront") monotonic 4 into `i32 addrspace(2)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -620,7 +620,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRITE_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            store_wavefront_release
 body:             |
@@ -631,7 +631,7 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     DS_WRITE_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 1, implicit $m0, implicit $exec :: (volatile store syncscope("wavefront") release 4 into `i32 addrspace(2)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -642,7 +642,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRITE_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            store_wavefront_seq_cst
 body:             |
@@ -653,7 +653,7 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     DS_WRITE_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 1, implicit $m0, implicit $exec :: (volatile store syncscope("wavefront") seq_cst 4 into `i32 addrspace(2)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -664,7 +664,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRITE_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            store_workgroup_unordered
 body:             |
@@ -675,7 +675,7 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     DS_WRITE_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 1, implicit $m0, implicit $exec :: (volatile store syncscope("workgroup") unordered 4 into `i32 addrspace(2)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -686,7 +686,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRITE_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            store_workgroup_monotonic
 body:             |
@@ -697,7 +697,7 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     DS_WRITE_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 1, implicit $m0, implicit $exec :: (volatile store syncscope("workgroup") monotonic 4 into `i32 addrspace(2)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -708,7 +708,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRITE_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            store_workgroup_release
 body:             |
@@ -719,7 +719,7 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     DS_WRITE_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 1, implicit $m0, implicit $exec :: (volatile store syncscope("workgroup") release 4 into `i32 addrspace(2)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -730,7 +730,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRITE_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            store_workgroup_seq_cst
 body:             |
@@ -741,7 +741,7 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     DS_WRITE_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 1, implicit $m0, implicit $exec :: (volatile store syncscope("workgroup") seq_cst 4 into `i32 addrspace(2)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -752,7 +752,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRITE_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            store_agent_unordered
 body:             |
@@ -763,8 +763,7 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     DS_WRITE_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 1, implicit $m0, implicit $exec :: (volatile store syncscope("agent") unordered 4 into `i32 addrspace(2)* undef`)
-    S_ENDPGM
-
+    S_ENDPGM 0
 ...
 ---
 
@@ -774,7 +773,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRITE_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            store_agent_monotonic
 body:             |
@@ -785,7 +784,7 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     DS_WRITE_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 1, implicit $m0, implicit $exec :: (volatile store syncscope("agent") monotonic 4 into `i32 addrspace(2)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -796,7 +795,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRITE_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            store_agent_release
 body:             |
@@ -807,7 +806,7 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     DS_WRITE_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 1, implicit $m0, implicit $exec :: (volatile store syncscope("agent") release 4 into `i32 addrspace(2)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -818,7 +817,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRITE_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            store_agent_seq_cst
 body:             |
@@ -829,7 +828,7 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     DS_WRITE_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 1, implicit $m0, implicit $exec :: (volatile store syncscope("agent") seq_cst 4 into `i32 addrspace(2)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -840,7 +839,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRITE_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            store_system_unordered
 body:             |
@@ -851,7 +850,7 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     DS_WRITE_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 1, implicit $m0, implicit $exec :: (volatile store unordered 4 into `i32 addrspace(2)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -862,7 +861,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRITE_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            store_system_monotonic
 body:             |
@@ -873,7 +872,7 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     DS_WRITE_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 1, implicit $m0, implicit $exec :: (volatile store monotonic 4 into `i32 addrspace(2)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -884,7 +883,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRITE_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            store_system_release
 body:             |
@@ -895,7 +894,7 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     DS_WRITE_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 1, implicit $m0, implicit $exec :: (volatile store release 4 into `i32 addrspace(2)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -906,7 +905,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRITE_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            store_system_seq_cst
 body:             |
@@ -917,7 +916,7 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     DS_WRITE_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 1, implicit $m0, implicit $exec :: (volatile store seq_cst 4 into `i32 addrspace(2)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -928,7 +927,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRXCHG_RTN_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            atomicrmw_singlethread_unordered
 body:             |
@@ -939,7 +938,7 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     $vgpr2 = DS_WRXCHG_RTN_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 1, implicit $m0, implicit $exec :: (volatile store syncscope("singlethread") unordered 4 into `i32 addrspace(2)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -950,7 +949,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRXCHG_RTN_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            atomicrmw_singlethread_monotonic
 body:             |
@@ -961,7 +960,7 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     $vgpr2 = DS_WRXCHG_RTN_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 1, implicit $m0, implicit $exec :: (volatile store syncscope("singlethread") monotonic 4 into `i32 addrspace(2)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -972,7 +971,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRXCHG_RTN_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            atomicrmw_singlethread_acquire
 body:             |
@@ -983,7 +982,7 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     $vgpr2 = DS_WRXCHG_RTN_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 1, implicit $m0, implicit $exec :: (volatile store syncscope("singlethread") acquire 4 into `i32 addrspace(2)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -994,7 +993,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRXCHG_RTN_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            atomicrmw_singlethread_release
 body:             |
@@ -1005,7 +1004,7 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     $vgpr2 = DS_WRXCHG_RTN_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 1, implicit $m0, implicit $exec :: (volatile store syncscope("singlethread") release 4 into `i32 addrspace(2)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -1016,7 +1015,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRXCHG_RTN_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            atomicrmw_singlethread_acq_rel
 body:             |
@@ -1027,7 +1026,7 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     $vgpr2 = DS_WRXCHG_RTN_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 1, implicit $m0, implicit $exec :: (volatile store syncscope("singlethread") acq_rel 4 into `i32 addrspace(2)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -1038,7 +1037,7 @@
 # GCN-NOT:   S_WAITCNT
 # GCN:       DS_WRXCHG_RTN_B32
 # GCN-NOT:   S_WAITCNT
-# GCN:       S_ENDPGM
+# GCN:       S_ENDPGM 0
 
 name:            atomicrmw_singlethread_seq_cst
 body:             |
@@ -1049,6 +1048,6 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
     $vgpr0 = V_MOV_B32_e32 killed $sgpr0, implicit $exec, implicit $exec
     $vgpr2 = DS_WRXCHG_RTN_B32 killed renamable $vgpr0, killed renamable $vgpr1, 0, 1, implicit $m0, implicit $exec :: (volatile store syncscope("singlethread") seq_cst 4 into `i32 addrspace(2)* undef`)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/memory_clause.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/memory_clause.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/memory_clause.mir
@@ -366,7 +366,7 @@
 # GCN-NEXT: dead %3:vgpr_32 = FLAT_ATOMIC_ADD_RTN %0, %1, 0, 0, implicit $exec, implicit $flat_scr
 # GCN-NEXT: FLAT_ATOMIC_ADD %0, %1, 0, 0, implicit $exec, implicit $flat_scr
 # GCN-NEXT: FLAT_ATOMIC_ADD %0, %1, 0, 0, implicit $exec, implicit $flat_scr
-# GCN-NEXT: S_ENDPGM
+# GCN-NEXT: S_ENDPGM 0
 
 ---
 name:            atomic
@@ -384,5 +384,5 @@
     %3:vgpr_32 = FLAT_ATOMIC_ADD_RTN %0, %1, 0, 0, implicit $exec, implicit $flat_scr
     FLAT_ATOMIC_ADD %0, %1, 0, 0, implicit $exec, implicit $flat_scr
     FLAT_ATOMIC_ADD %0, %1, 0, 0, implicit $exec, implicit $flat_scr
-    S_ENDPGM
+    S_ENDPGM 0
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/merge-load-store-physreg.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/merge-load-store-physreg.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/merge-load-store-physreg.mir
@@ -62,7 +62,7 @@
     %12:sgpr_32 = S_ADDC_U32 %10, 0, implicit-def dead $scc, implicit $scc
 
     %3:vgpr_32 = DS_READ_B32 %1, 64, 0, implicit $m0, implicit $exec :: (load 4 from %ir.ptr.64)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 
@@ -111,6 +111,6 @@
     %11:sgpr_32 = S_ADDC_U32 %10, 0, implicit-def dead $scc, implicit $scc
 
     %3:vgpr_32 = DS_READ_B32 %1, 64, 0, implicit $m0, implicit $exec :: (load 4 from %ir.ptr.64)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/merge-load-store-vreg.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/merge-load-store-vreg.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/merge-load-store-vreg.mir
@@ -73,7 +73,7 @@
     S_BRANCH %bb.2
 
   bb.1:
-    S_ENDPGM
+    S_ENDPGM 0
 
   bb.2:
     %1:sreg_64_xexec = V_CMP_NE_U32_e64 %0, 0, implicit $exec
@@ -108,7 +108,7 @@
     S_BRANCH %bb.2
 
   bb.1:
-    S_ENDPGM
+    S_ENDPGM 0
 
   bb.2:
     %1:sreg_64_xexec = V_CMP_NE_U32_e64 %0.sub0, 0, implicit $exec
@@ -139,7 +139,7 @@
     S_BRANCH %bb.2
 
   bb.1:
-    S_ENDPGM
+    S_ENDPGM 0
 
   bb.2:
     %1:sreg_64_xexec = V_CMP_NE_U32_e64 %0.sub0, 0, implicit $exec
Index: llvm/trunk/test/CodeGen/AMDGPU/merge-load-store.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/merge-load-store.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/merge-load-store.mir
@@ -100,7 +100,7 @@
     %4:vgpr_32 = DS_READ_B32 %1, 4, 0, implicit $m0, implicit $exec :: (load 4 from %ir.ptr.4)
     %5:vgpr_32 = V_ADD_I32_e32 killed %3, killed %4, implicit-def $vcc, implicit $exec
     DS_WRITE_B32 killed %1, %5, 0, 0, implicit killed $m0, implicit $exec :: (store 4 into %ir.ptr.0)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
Index: llvm/trunk/test/CodeGen/AMDGPU/misched-killflags.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/misched-killflags.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/misched-killflags.mir
@@ -23,7 +23,7 @@
     $vgpr2 = V_MOV_B32_e32 $sgpr10, implicit $exec, implicit $sgpr8_sgpr9_sgpr10_sgpr11
     $vgpr3 = V_MOV_B32_e32 $sgpr11, implicit $exec, implicit $sgpr8_sgpr9_sgpr10_sgpr11, implicit $exec
     S_NOP 0, implicit killed $sgpr6_sgpr7, implicit $sgpr0_sgpr1_sgpr2_sgpr3, implicit $sgpr4, implicit killed $vgpr0_vgpr1_vgpr2_vgpr3
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 # CHECK-LABEL: name: func0
 # CHECK-DAG: $sgpr10 = S_MOV_B32 5
@@ -42,4 +42,4 @@
 # CHECK: $vgpr2 = V_MOV_B32_e32 $sgpr10, implicit $exec, implicit $sgpr8_sgpr9_sgpr10_sgpr11
 # CHECK: $vgpr3 = V_MOV_B32_e32 killed $sgpr11, implicit $exec, implicit $sgpr8_sgpr9_sgpr10_sgpr11, implicit $exec
 # CHECK: S_NOP 0, implicit killed $sgpr6_sgpr7, implicit $sgpr0_sgpr1_sgpr2_sgpr3, implicit $sgpr4, implicit killed $vgpr0_vgpr1_vgpr2_vgpr3
-# CHECK: S_ENDPGM
+# CHECK: S_ENDPGM 0
Index: llvm/trunk/test/CodeGen/AMDGPU/mode-register.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/mode-register.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/mode-register.mir
@@ -23,7 +23,7 @@
     $vgpr1 = V_INTERP_P2_F16 0, $vgpr2, 2, 1, 0, killed $vgpr1, 0, 0, implicit $m0, implicit $exec
     $vgpr0 = V_INTERP_P2_F16 0, killed $vgpr2, 2, 1, 0, killed $vgpr0, -1, 0, implicit $m0, implicit $exec
     $vgpr0 = V_ADD_F16_e32 killed $vgpr1, killed $vgpr0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # check that the mode is not changed for interp f16 when the mode is already RTZ
@@ -49,7 +49,7 @@
     $vgpr1 = V_INTERP_P2_F16 0, $vgpr2, 2, 1, 0, killed $vgpr1, 0, 0, implicit $m0, implicit $exec
     $vgpr0 = V_INTERP_P2_F16 0, killed $vgpr2, 2, 1, 0, killed $vgpr0, -1, 0, implicit $m0, implicit $exec
     $vgpr0 = V_ADD_F16_e32 killed $vgpr1, killed $vgpr0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # check that explicit RTN mode change is registered
@@ -75,7 +75,7 @@
     $vgpr0 = V_INTERP_P2_F16 0, killed $vgpr2, 2, 1, 0, killed $vgpr0, -1, 0, implicit $m0, implicit $exec
     S_SETREG_IMM32_B32 0, 2177
     $vgpr0 = V_ADD_F16_e32 killed $vgpr1, killed $vgpr0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # check that the mode is unchanged from RTN for F64 instruction
@@ -90,7 +90,7 @@
   bb.0:
     liveins: $vgpr1_vgpr2
     $vgpr1_vgpr2 = V_FRACT_F64_e32 killed $vgpr1_vgpr2, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # check that the mode is changed from RTZ to RTN for F64 instruction
@@ -108,7 +108,7 @@
     liveins: $vgpr1_vgpr2
     S_SETREG_IMM32_B32 3, 2177
     $vgpr1_vgpr2 = V_FRACT_F64_e32 killed $vgpr1_vgpr2, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # CHECK-LABEL: name: rtz_from_rtn
@@ -127,7 +127,7 @@
 
   bb.1:
     $vgpr1 = V_INTERP_P1LL_F16 0, $vgpr0, 2, 1, 0, 0, 0, implicit $m0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # check that the mode is changed from RTZ to RTN for F64 instruction
@@ -157,7 +157,7 @@
     $vgpr3_vgpr4 = V_FRACT_F64_e32 killed $vgpr3_vgpr4, implicit $exec
     $vgpr0 = V_INTERP_P2_F16 0, killed $vgpr2, 2, 1, 0, killed $vgpr0, -1, 0, implicit $m0, implicit $exec
     $vgpr0 = V_ADD_F16_e32 killed $sgpr0, killed $vgpr0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # check that an explicit change to the single precision mode has no effect
@@ -187,7 +187,7 @@
     $vgpr3_vgpr4 = V_FRACT_F64_e32 killed $vgpr3_vgpr4, implicit $exec
     $vgpr0 = V_INTERP_P2_F16 0, killed $vgpr2, 2, 1, 0, killed $vgpr0, -1, 0, implicit $m0, implicit $exec
     $vgpr0 = V_ADD_F16_e32 killed $sgpr0, killed $vgpr0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # check that mode is propagated back to start of loop - first instruction is RTN but needs
@@ -223,7 +223,7 @@
     S_BRANCH %bb.3
 
   bb.3:
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # two back-edges to same node with different modes
@@ -272,7 +272,7 @@
     S_BRANCH %bb.6
 
   bb.6:
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # check that mode is propagated back to start of loop and through a block that
@@ -311,7 +311,7 @@
     S_BRANCH %bb.4
 
   bb.4:
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # check that multiple mode values are propagated to a block that uses the mode
@@ -346,7 +346,7 @@
     S_BRANCH %bb.4
 
   bb.4:
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # check that multiple mode values are propagated through a block that neither
@@ -387,7 +387,7 @@
     S_BRANCH %bb.5
 
   bb.5:
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # CHECK-LABEL: name: pass_through_blocks
@@ -420,7 +420,7 @@
 
   bb.4:
     $vgpr1 = V_INTERP_P1LL_F16 0, $vgpr0, 2, 1, 0, 0, 0, implicit $m0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # check that multiple mode values are propagated
@@ -455,5 +455,5 @@
     S_BRANCH %bb.4
 
   bb.4:
-    S_ENDPGM
+    S_ENDPGM 0
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/movrels-bug.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/movrels-bug.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/movrels-bug.mir
@@ -26,6 +26,6 @@
     $vgpr1 = V_MOVRELS_B32_e32 undef $vgpr1, implicit $m0, implicit $exec, implicit killed $vgpr1_vgpr2_vgpr3_vgpr4_vgpr5_vgpr6_vgpr7_vgpr8
     $vgpr4 = V_MAC_F32_e32 undef $vgpr0, undef $vgpr0, undef $vgpr4, implicit $exec
     EXP_DONE 15, undef $vgpr0, killed $vgpr1, killed $vgpr4, undef $vgpr0, 0, 0, 12, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/omod-nsz-flag.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/omod-nsz-flag.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/omod-nsz-flag.mir
@@ -21,7 +21,7 @@
 # GCN-LABEL: name: omod_inst_flag_nsz_src
 # GCN: %0:vgpr_32 = nsz V_ADD_F32_e64 0, $vgpr0, 0, $vgpr1, 0, 0, implicit $exec
 # GCN-NEXT: %1:vgpr_32 = V_MUL_F32_e64 0, %0, 0, 1073741824, 0, 0, implicit $exec
-# GCN-NEXT: S_ENDPGM implicit %1
+# GCN-NEXT: S_ENDPGM 0, implicit %1
 name: omod_inst_flag_nsz_src
 tracksRegLiveness: true
 
@@ -31,14 +31,14 @@
 
   %0:vgpr_32 = nsz V_ADD_F32_e64 0, $vgpr0, 0, $vgpr1, 0, 0, implicit $exec
   %1:vgpr_32 = V_MUL_F32_e64 0, %0, 0, 1073741824, 0, 0, implicit $exec
-  S_ENDPGM implicit %1
+  S_ENDPGM 0, implicit %1
 
 ...
 ---
 
 # GCN-LABEL: name: omod_inst_flag_nsz_result
 # GCN: %0:vgpr_32 = V_ADD_F32_e64 0, $vgpr0, 0, $vgpr1, 0, 1, implicit $exec
-# GCN-NEXT: S_ENDPGM implicit %0
+# GCN-NEXT: S_ENDPGM 0, implicit %0
 
 name: omod_inst_flag_nsz_result
 tracksRegLiveness: true
@@ -49,14 +49,14 @@
 
   %0:vgpr_32 = V_ADD_F32_e64 0, $vgpr0, 0, $vgpr1, 0, 0, implicit $exec
   %1:vgpr_32 = nsz V_MUL_F32_e64 0, %0, 0, 1073741824, 0, 0, implicit $exec
-  S_ENDPGM implicit %1
+  S_ENDPGM 0, implicit %1
 ...
 
 ---
 
 # GCN-LABEL: name: omod_inst_flag_nsz_both
 # GCN: %0:vgpr_32 = nsz V_ADD_F32_e64 0, $vgpr0, 0, $vgpr1, 0, 1, implicit $exec
-# GCN-NEXT: S_ENDPGM implicit %0
+# GCN-NEXT: S_ENDPGM 0, implicit %0
 
 name: omod_inst_flag_nsz_both
 tracksRegLiveness: true
@@ -67,5 +67,5 @@
 
   %0:vgpr_32 = nsz V_ADD_F32_e64 0, $vgpr0, 0, $vgpr1, 0, 0, implicit $exec
   %1:vgpr_32 = nsz V_MUL_F32_e64 0, %0, 0, 1073741824, 0, 0, implicit $exec
-  S_ENDPGM implicit %1
+  S_ENDPGM 0, implicit %1
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/opt-sgpr-to-vgpr-copy.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/opt-sgpr-to-vgpr-copy.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/opt-sgpr-to-vgpr-copy.mir
@@ -141,7 +141,7 @@
 
   bb.2.bb2:
     SI_END_CF %1, implicit-def dead $exec, implicit-def dead $scc, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -247,7 +247,7 @@
 
   bb.2.bb2:
     SI_END_CF %1, implicit-def dead $exec, implicit-def dead $scc, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -336,6 +336,6 @@
 
   bb.2.bb2:
     SI_END_CF %1, implicit-def dead $exec, implicit-def dead $scc, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/optimize-if-exec-masking.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/optimize-if-exec-masking.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/optimize-if-exec-masking.mir
@@ -160,7 +160,7 @@
     $sgpr3 = S_MOV_B32 61440
     $sgpr2 = S_MOV_B32 -1
     BUFFER_STORE_DWORD_OFFSET killed $vgpr0, $sgpr0_sgpr1_sgpr2_sgpr3, 0, 0, 0, 0, 0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -197,7 +197,7 @@
     $sgpr3 = S_MOV_B32 61440
     $sgpr2 = S_MOV_B32 -1
     BUFFER_STORE_DWORD_OFFSET killed $vgpr0, $sgpr0_sgpr1_sgpr2_sgpr3, 0, 0, 0, 0, 0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -234,7 +234,7 @@
     $sgpr3 = S_MOV_B32 61440
     $sgpr2 = S_MOV_B32 -1
     BUFFER_STORE_DWORD_OFFSET killed $vgpr0, $sgpr0_sgpr1_sgpr2_sgpr3, 0, 0, 0, 0, 0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -275,7 +275,7 @@
     $sgpr3 = S_MOV_B32 61440
     $sgpr2 = S_MOV_B32 -1
     BUFFER_STORE_DWORD_OFFSET killed $vgpr0, $sgpr0_sgpr1_sgpr2_sgpr3, 0, 0, 0, 0, 0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -313,7 +313,7 @@
     $sgpr3 = S_MOV_B32 61440
     $sgpr2 = S_MOV_B32 -1
     BUFFER_STORE_DWORD_OFFSET killed $vgpr0, $sgpr4_sgpr5_sgpr6_sgpr7, 0, 0, 0, 0, 0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -357,7 +357,7 @@
     $sgpr2 = S_MOV_B32 -1
     $sgpr3 = S_MOV_B32 61440
     BUFFER_STORE_DWORD_OFFSET killed $vgpr0, $sgpr0_sgpr1_sgpr2_sgpr3, 0, 0, 0, 0, 0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -396,7 +396,7 @@
     $sgpr3 = S_MOV_B32 61440
     $sgpr2 = S_MOV_B32 -1
     BUFFER_STORE_DWORD_OFFSET killed $vgpr0, $sgpr0_sgpr1_sgpr2_sgpr3, 0, 0, 0, 0, 0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 
@@ -435,7 +435,7 @@
     $sgpr3 = S_MOV_B32 61440
     $sgpr2 = S_MOV_B32 -1
     BUFFER_STORE_DWORD_OFFSET killed $vgpr0, $sgpr0_sgpr1_sgpr2_sgpr3, 0, 0, 0, 0, 0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -472,7 +472,7 @@
     $sgpr3 = S_MOV_B32 61440
     $sgpr2 = S_MOV_B32 -1
     BUFFER_STORE_DWORD_OFFSET killed $vgpr0, $sgpr0_sgpr1_sgpr2_sgpr3, 0, 0, 0, 0, 0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -509,7 +509,7 @@
     $sgpr3 = S_MOV_B32 61440
     $sgpr2 = S_MOV_B32 -1
     BUFFER_STORE_DWORD_OFFSET killed $vgpr0, $sgpr0_sgpr1_sgpr2_sgpr3, 0, 0, 0, 0, 0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -548,5 +548,5 @@
     $sgpr3 = S_MOV_B32 61440
     $sgpr2 = S_MOV_B32 -1
     BUFFER_STORE_DWORD_OFFSET killed $vgpr0, $sgpr0_sgpr1_sgpr2_sgpr3, 0, 0, 0, 0, 0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/optimize-negated-cond-exec-masking.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/optimize-negated-cond-exec-masking.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/optimize-negated-cond-exec-masking.mir
@@ -19,7 +19,7 @@
     S_BRANCH %bb.0
 
   bb.2:
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 
 # GCN: name: negated_cond_vop3
@@ -41,7 +41,7 @@
     S_BRANCH %bb.0
 
   bb.2:
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 
 # GCN: name: negated_cond_vop2_redef_vcc1
@@ -67,7 +67,7 @@
     S_BRANCH %bb.0
 
   bb.2:
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 
 # GCN: name: negated_cond_vop2_redef_vcc2
@@ -93,7 +93,7 @@
     S_BRANCH %bb.0
 
   bb.2:
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 
 # GCN: name: negated_cond_vop3_redef_cmp
@@ -119,7 +119,7 @@
     S_BRANCH %bb.0
 
   bb.2:
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 
 # GCN: name: negated_cond_undef_vcc
@@ -137,7 +137,7 @@
     S_BRANCH %bb.0
 
   bb.2:
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 
 # GCN: name: negated_cond_vop3_imp_vcc
@@ -159,7 +159,7 @@
     S_BRANCH %bb.0
 
   bb.2:
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 
 # GCN: name: negated_cond_vop2_imp_vcc
@@ -181,7 +181,7 @@
     S_BRANCH %bb.0
 
   bb.2:
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 
 # GCN: name: negated_cond_vop3_redef_sel
@@ -207,7 +207,7 @@
     S_BRANCH %bb.0
 
   bb.2:
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 
 # GCN: name: negated_cond_vop2_used_sel
@@ -231,7 +231,7 @@
 
   bb.2:
     $vgpr0 = COPY %1
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 
 # GCN: name: negated_cond_vop2_used_vcc
@@ -257,7 +257,7 @@
     S_BRANCH %bb.0
 
   bb.2:
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 
 # GCN: name: negated_cond_vop3_sel_wrong_subreg1
@@ -283,7 +283,7 @@
     S_BRANCH %bb.0
 
   bb.2:
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 
 # GCN: name: negated_cond_vop3_sel_wrong_subreg2
@@ -309,7 +309,7 @@
     S_BRANCH %bb.0
 
   bb.2:
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 
 # GCN: name: negated_cond_vop3_sel_right_subreg1
@@ -333,7 +333,7 @@
     S_BRANCH %bb.0
 
   bb.2:
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 
 # GCN: name: negated_cond_vop3_sel_right_subreg2
@@ -357,7 +357,7 @@
     S_BRANCH %bb.0
 
   bb.2:
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 
 # GCN: name: negated_cond_vop3_sel_subreg_overlap
@@ -383,7 +383,7 @@
     S_BRANCH %bb.0
 
   bb.2:
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 
 # GCN: name: negated_cond_vop2_dominated_blocks
@@ -407,7 +407,7 @@
     S_BRANCH %bb.1
 
   bb.3:
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 
 # GCN: name: negated_cond_vop2_different_blocks_cmp_and
@@ -431,7 +431,7 @@
     S_BRANCH %bb.1
 
   bb.3:
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 
 # GCN: name: negated_cond_vop2_not_dominated_blocks
@@ -461,5 +461,5 @@
     S_BRANCH %bb.2
 
   bb.4:
-    S_ENDPGM
+    S_ENDPGM 0
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/reduce-saveexec.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/reduce-saveexec.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/reduce-saveexec.mir
@@ -3,7 +3,7 @@
 ---
 # GCN-LABEL: name: reduce_and_saveexec
 # GCN:      $exec = S_AND_B64 $exec, killed $vcc
-# GCN-NEXT: S_ENDPGM
+# GCN-NEXT: S_ENDPGM 0
 name: reduce_and_saveexec
 tracksRegLiveness: true
 body:             |
@@ -11,12 +11,12 @@
     $vcc = IMPLICIT_DEF
     $sgpr0_sgpr1 = S_AND_B64 $exec, killed $vcc, implicit-def $scc
     $exec = COPY killed $sgpr0_sgpr1
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # GCN-LABEL: name: reduce_and_saveexec_commuted
 # GCN:      $exec = S_AND_B64 killed $vcc, $exec
-# GCN-NEXT: S_ENDPGM
+# GCN-NEXT: S_ENDPGM 0
 name: reduce_and_saveexec_commuted
 tracksRegLiveness: true
 body:             |
@@ -24,7 +24,7 @@
     $vcc = IMPLICIT_DEF
     $sgpr0_sgpr1 = S_AND_B64 killed $vcc, $exec, implicit-def $scc
     $exec = COPY killed $sgpr0_sgpr1
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # GCN-LABEL: name: reduce_and_saveexec_liveout
@@ -37,12 +37,12 @@
     $vcc = IMPLICIT_DEF
     $sgpr0_sgpr1 = S_AND_B64 $exec, killed $vcc, implicit-def $scc
     $exec = COPY $sgpr0_sgpr1
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # GCN-LABEL: name: and_saveexec
 # GCN:      $sgpr0_sgpr1 = S_AND_SAVEEXEC_B64 $vcc
-# GCN-NEXT: S_ENDPGM
+# GCN-NEXT: S_ENDPGM 0
 name: and_saveexec
 tracksRegLiveness: true
 body:             |
@@ -51,12 +51,12 @@
     $sgpr0_sgpr1 = COPY $exec
     $sgpr2_sgpr3 = S_AND_B64 $sgpr0_sgpr1, killed $vcc, implicit-def $scc
     $exec = S_MOV_B64_term $sgpr2_sgpr3
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # GCN-LABEL: name: reduce_or_saveexec
 # GCN:      $exec = S_OR_B64 $exec, killed $vcc
-# GCN-NEXT: S_ENDPGM
+# GCN-NEXT: S_ENDPGM 0
 name: reduce_or_saveexec
 tracksRegLiveness: true
 body:             |
@@ -64,12 +64,12 @@
     $vcc = IMPLICIT_DEF
     $sgpr0_sgpr1 = S_OR_B64 $exec, killed $vcc, implicit-def $scc
     $exec = COPY killed $sgpr0_sgpr1
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # GCN-LABEL: name: reduce_xor_saveexec
 # GCN:      $exec = S_XOR_B64 $exec, killed $vcc
-# GCN-NEXT: S_ENDPGM
+# GCN-NEXT: S_ENDPGM 0
 name: reduce_xor_saveexec
 tracksRegLiveness: true
 body:             |
@@ -77,12 +77,12 @@
     $vcc = IMPLICIT_DEF
     $sgpr0_sgpr1 = S_XOR_B64 $exec, killed $vcc, implicit-def $scc
     $exec = COPY killed $sgpr0_sgpr1
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # GCN-LABEL: name: reduce_andn2_saveexec
 # GCN:      $exec = S_ANDN2_B64 $exec, killed $vcc
-# GCN-NEXT: S_ENDPGM
+# GCN-NEXT: S_ENDPGM 0
 name: reduce_andn2_saveexec
 tracksRegLiveness: true
 body:             |
@@ -90,12 +90,12 @@
     $vcc = IMPLICIT_DEF
     $sgpr0_sgpr1 = S_ANDN2_B64 $exec, killed $vcc, implicit-def $scc
     $exec = COPY killed $sgpr0_sgpr1
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # GCN-LABEL: name: reduce_orn2_saveexec
 # GCN:      $exec = S_ORN2_B64 $exec, killed $vcc
-# GCN-NEXT: S_ENDPGM
+# GCN-NEXT: S_ENDPGM 0
 name: reduce_orn2_saveexec
 tracksRegLiveness: true
 body:             |
@@ -103,12 +103,12 @@
     $vcc = IMPLICIT_DEF
     $sgpr0_sgpr1 = S_ORN2_B64 $exec, killed $vcc, implicit-def $scc
     $exec = COPY killed $sgpr0_sgpr1
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # GCN-LABEL: name: reduce_nand_saveexec
 # GCN:      $exec = S_NAND_B64 $exec, killed $vcc
-# GCN-NEXT: S_ENDPGM
+# GCN-NEXT: S_ENDPGM 0
 name: reduce_nand_saveexec
 tracksRegLiveness: true
 body:             |
@@ -116,12 +116,12 @@
     $vcc = IMPLICIT_DEF
     $sgpr0_sgpr1 = S_NAND_B64 $exec, killed $vcc, implicit-def $scc
     $exec = COPY killed $sgpr0_sgpr1
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # GCN-LABEL: name: reduce_nor_saveexec
 # GCN:      $exec = S_NOR_B64 $exec, killed $vcc
-# GCN-NEXT: S_ENDPGM
+# GCN-NEXT: S_ENDPGM 0
 name: reduce_nor_saveexec
 tracksRegLiveness: true
 body:             |
@@ -129,12 +129,12 @@
     $vcc = IMPLICIT_DEF
     $sgpr0_sgpr1 = S_NOR_B64 $exec, killed $vcc, implicit-def $scc
     $exec = COPY killed $sgpr0_sgpr1
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # GCN-LABEL: name: reduce_xnor_saveexec
 # GCN:      $exec = S_XNOR_B64 $exec, killed $vcc
-# GCN-NEXT: S_ENDPGM
+# GCN-NEXT: S_ENDPGM 0
 name: reduce_xnor_saveexec
 tracksRegLiveness: true
 body:             |
@@ -142,6 +142,6 @@
     $vcc = IMPLICIT_DEF
     $sgpr0_sgpr1 = S_XNOR_B64 $exec, killed $vcc, implicit-def $scc
     $exec = COPY killed $sgpr0_sgpr1
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
Index: llvm/trunk/test/CodeGen/AMDGPU/regcoal-subrange-join-seg.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/regcoal-subrange-join-seg.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/regcoal-subrange-join-seg.mir
@@ -234,6 +234,6 @@
   bb.34:
 
   bb.35:
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/regcoal-subrange-join.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/regcoal-subrange-join.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/regcoal-subrange-join.mir
@@ -157,6 +157,6 @@
 
   bb.4:
     EXP 32, undef %53, undef %54, killed %46, undef %55, 0, 0, 15, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/regcoalesce-cannot-join-failures.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/regcoalesce-cannot-join-failures.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/regcoalesce-cannot-join-failures.mir
@@ -14,7 +14,7 @@
   ; CHECK:   %0.sub1:sreg_64_xexec = COPY %0.sub0
   ; CHECK:   S_BRANCH %bb.2
   ; CHECK: bb.2:
-  ; CHECK:   S_ENDPGM implicit %0
+  ; CHECK:   S_ENDPGM 0, implicit %0
   bb.0:
     successors: %bb.1
 
@@ -29,7 +29,7 @@
 
   bb.2:
     dead %2:sreg_32_xm0 = COPY %0.sub0:sreg_64_xexec
-    S_ENDPGM implicit killed %1
+    S_ENDPGM 0, implicit killed %1
 
 ...
 ---
@@ -40,11 +40,11 @@
     ; CHECK-LABEL: name: couldnt_join_subrange_no_implicit_def_inst
     ; CHECK: undef %0.sub0:sreg_64 = S_MOV_B32 0
     ; CHECK: %0.sub1:sreg_64 = COPY %0.sub0
-    ; CHECK: S_ENDPGM implicit %0.sub1
+    ; CHECK: S_ENDPGM 0, implicit %0.sub1
     undef %0.sub0:sreg_64 = S_MOV_B32 0
     %1:sreg_64 = COPY %0:sreg_64
     %0.sub1:sreg_64 = COPY %0.sub0:sreg_64
-    S_ENDPGM implicit %1.sub1:sreg_64
+    S_ENDPGM 0, implicit %1.sub1:sreg_64
 
 ...
 ---
@@ -59,7 +59,7 @@
   ; CHECK:   %0.sub0:sreg_64 = S_MOV_B32 0
   ; CHECK:   [[COPY:%[0-9]+]]:sreg_64 = COPY %0
   ; CHECK:   dead %0.sub1:sreg_64 = COPY %0.sub0
-  ; CHECK:   S_ENDPGM implicit [[COPY]].sub1
+  ; CHECK:   S_ENDPGM 0, implicit [[COPY]].sub1
   bb.0:
     successors: %bb.1
     undef %0.sub1:sreg_64 = S_MOV_B32 -1
@@ -68,7 +68,7 @@
     %0.sub0:sreg_64 = S_MOV_B32 0
     %1:sreg_64 = COPY %0:sreg_64
     dead %0.sub1:sreg_64 = COPY %0.sub0:sreg_64
-    S_ENDPGM implicit %1.sub1:sreg_64
+    S_ENDPGM 0, implicit %1.sub1:sreg_64
 
 ...
 ---
@@ -82,13 +82,13 @@
     ; CHECK: %0.sub1:sreg_64_xexec = S_MOV_B32 0
     ; CHECK: S_NOP 0, implicit %0.sub1
     ; CHECK: S_NOP 0, implicit %0
-    ; CHECK: S_ENDPGM
+    ; CHECK: S_ENDPGM 0
     undef %0.sub0:sreg_64_xexec = S_MOV_B32 0
     %1:sreg_64 = COPY %0
     %0.sub1:sreg_64_xexec = S_MOV_B32 0
     S_NOP 0, implicit %0.sub1
     S_NOP 0, implicit %1
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -102,7 +102,7 @@
   ; CHECK:   %0.sub1:sreg_64_xexec = COPY %0.sub0
   ; CHECK: bb.1:
   ; CHECK:   S_NOP 0, implicit %0.sub1
-  ; CHECK:   S_ENDPGM implicit %0
+  ; CHECK:   S_ENDPGM 0, implicit %0
   bb.0:
     successors: %bb.1
 
@@ -113,6 +113,6 @@
   bb.1:
 
     S_NOP 0, implicit %0.sub1
-    S_ENDPGM implicit %1
+    S_ENDPGM 0, implicit %1
 
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/regcoalesce-dbg.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/regcoalesce-dbg.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/regcoalesce-dbg.mir
@@ -71,6 +71,6 @@
     %20 = V_LSHL_B64 killed %19, 2, implicit $exec
     %16 = COPY killed %5
     BUFFER_STORE_DWORD_ADDR64 killed %16, killed %20, killed %13, 0, 0, 0, 0, 0, implicit $exec :: (store 4 into %ir.out)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/regcoalesce-keep-valid-lanes-implicit-def-bug39602.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/regcoalesce-keep-valid-lanes-implicit-def-bug39602.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/regcoalesce-keep-valid-lanes-implicit-def-bug39602.mir
@@ -16,7 +16,7 @@
   ; CHECK:   %0.sub0:sreg_64 = S_MOV_B32 0
   ; CHECK:   [[COPY:%[0-9]+]]:sreg_64 = COPY %0
   ; CHECK:   dead %0.sub1:sreg_64 = COPY %0.sub0
-  ; CHECK:   S_ENDPGM implicit [[COPY]].sub1
+  ; CHECK:   S_ENDPGM 0, implicit [[COPY]].sub1
   bb.0:
     successors: %bb.1
     undef %0.sub1:sreg_64 = IMPLICIT_DEF
@@ -25,7 +25,7 @@
     %0.sub0:sreg_64 = S_MOV_B32 0
     %1:sreg_64 = COPY %0:sreg_64
     dead %0.sub1:sreg_64 = COPY %0.sub0:sreg_64
-    S_ENDPGM implicit %1.sub1:sreg_64
+    S_ENDPGM 0, implicit %1.sub1:sreg_64
 
 ...
 ---
@@ -43,7 +43,7 @@
   ; CHECK:   %0.sub0:sreg_64 = S_MOV_B32 0
   ; CHECK:   [[COPY:%[0-9]+]]:sreg_64 = COPY %0
   ; CHECK:   dead %0.sub1:sreg_64 = COPY %0.sub0
-  ; CHECK:   S_ENDPGM implicit [[COPY]].sub1
+  ; CHECK:   S_ENDPGM 0, implicit [[COPY]].sub1
   bb.0:
     successors: %bb.1
     undef %0.sub1:sreg_64 = S_MOV_B32 -1
@@ -52,6 +52,6 @@
     %0.sub0:sreg_64 = S_MOV_B32 0
     %1:sreg_64 = COPY %0:sreg_64
     dead %0.sub1:sreg_64 = COPY %0.sub0:sreg_64
-    S_ENDPGM implicit %1.sub1:sreg_64
+    S_ENDPGM 0, implicit %1.sub1:sreg_64
 
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/regcoalesce-prune.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/regcoalesce-prune.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/regcoalesce-prune.mir
@@ -27,5 +27,5 @@
 
   bb.3:
     %3 : vgpr_32 = V_CVT_F32_I32_e32 killed %6.sub1, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/regcoalescing-remove-partial-redundancy-assert.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/regcoalescing-remove-partial-redundancy-assert.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/regcoalescing-remove-partial-redundancy-assert.mir
@@ -191,9 +191,9 @@
     S_BRANCH %bb.29
   
   bb.33:
-    S_ENDPGM
+    S_ENDPGM 0
   
   bb.34:
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/rename-independent-subregs-mac-operands.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/rename-independent-subregs-mac-operands.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/rename-independent-subregs-mac-operands.mir
@@ -64,7 +64,7 @@
     undef %15.sub0 = COPY killed %16
     %15.sub1 = COPY killed %14
     FLAT_STORE_DWORDX2 undef %11, killed %15, 0, 0, 0, implicit $exec, implicit $flat_scr
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
Index: llvm/trunk/test/CodeGen/AMDGPU/scalar-store-cache-flush.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/scalar-store-cache-flush.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/scalar-store-cache-flush.mir
@@ -49,7 +49,7 @@
 # CHECK: bb.0:
 # CHECK-NEXT: S_STORE_DWORD
 # CHECK-NEXT: S_DCACHE_WB
-# CHECK-NEXT: S_ENDPGM
+# CHECK-NEXT: S_ENDPGM 0
 
 name: basic_insert_dcache_wb
 tracksRegLiveness: false
@@ -57,7 +57,7 @@
 body: |
   bb.0:
     S_STORE_DWORD_SGPR undef $sgpr2, undef $sgpr0_sgpr1, undef $m0, 0
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # Already has an explicitly requested flush after the last store.
@@ -65,7 +65,7 @@
 # CHECK: bb.0:
 # CHECK-NEXT: S_STORE_DWORD
 # CHECK-NEXT: S_DCACHE_WB
-# CHECK-NEXT: S_ENDPGM
+# CHECK-NEXT: S_ENDPGM 0
 
 name: explicit_flush_after
 tracksRegLiveness: false
@@ -74,7 +74,7 @@
   bb.0:
     S_STORE_DWORD_SGPR undef $sgpr2, undef $sgpr0_sgpr1, undef $m0, 0
     S_DCACHE_WB
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # Already has an explicitly requested flush before the last store.
@@ -83,7 +83,7 @@
 # CHECK-NEXT: S_DCACHE_WB
 # CHECK-NEXT: S_STORE_DWORD
 # CHECK-NEXT: S_DCACHE_WB
-# CHECK-NEXT: S_ENDPGM
+# CHECK-NEXT: S_ENDPGM 0
 
 name: explicit_flush_before
 tracksRegLiveness: false
@@ -92,30 +92,30 @@
   bb.0:
     S_DCACHE_WB
     S_STORE_DWORD_SGPR undef $sgpr2, undef $sgpr0_sgpr1, undef $m0, 0
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # CHECK-LABEL: no_scalar_store
 # CHECK: bb.0
-# CHECK-NEXT: S_ENDPGM
+# CHECK-NEXT: S_ENDPGM 0
 name: no_scalar_store
 tracksRegLiveness: false
 
 body: |
   bb.0:
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 
 # CHECK-LABEL: name: multi_block_store
 # CHECK: bb.0:
 # CHECK-NEXT: S_STORE_DWORD
 # CHECK-NEXT: S_DCACHE_WB
-# CHECK-NEXT: S_ENDPGM
+# CHECK-NEXT: S_ENDPGM 0
 
 # CHECK: bb.1:
 # CHECK-NEXT: S_STORE_DWORD
 # CHECK-NEXT: S_DCACHE_WB
-# CHECK-NEXT: S_ENDPGM
+# CHECK-NEXT: S_ENDPGM 0
 
 name: multi_block_store
 tracksRegLiveness: false
@@ -123,11 +123,11 @@
 body: |
   bb.0:
     S_STORE_DWORD_SGPR undef $sgpr2, undef $sgpr0_sgpr1, undef $m0, 0
-    S_ENDPGM
+    S_ENDPGM 0
 
   bb.1:
     S_STORE_DWORD_SGPR undef $sgpr4, undef $sgpr6_sgpr7, undef $m0, 0
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ...
 
@@ -137,23 +137,23 @@
 # CHECK-LABEL: name: one_block_store
 # CHECK: bb.0:
 # CHECK-NEXT: S_DCACHE_WB
-# CHECK-NEXT: S_ENDPGM
+# CHECK-NEXT: S_ENDPGM 0
 
 # CHECK: bb.1:
 # CHECK-NEXT: S_STORE_DWORD
 # CHECK-NEXT: S_DCACHE_WB
-# CHECK-NEXT: S_ENDPGM
+# CHECK-NEXT: S_ENDPGM 0
 
 name: one_block_store
 tracksRegLiveness: false
 
 body: |
   bb.0:
-    S_ENDPGM
+    S_ENDPGM 0
 
   bb.1:
     S_STORE_DWORD_SGPR undef $sgpr4, undef $sgpr6_sgpr7, undef $m0, 0
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # CHECK-LABEL: name: si_return
Index: llvm/trunk/test/CodeGen/AMDGPU/sched-crash-dbg-value.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/sched-crash-dbg-value.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/sched-crash-dbg-value.mir
@@ -329,6 +329,6 @@
     dead $sgpr30_sgpr31 = SI_CALL %127, @func, csr_amdgpu_highregs, implicit $sgpr0_sgpr1_sgpr2_sgpr3, implicit $sgpr4, implicit $vgpr0, implicit $vgpr1_vgpr2, implicit killed $vgpr3
     ADJCALLSTACKDOWN 0, 0, implicit-def $sgpr32, implicit $sgpr32
     %128:vreg_64, dead %129:sreg_64 = V_MAD_I64_I32 %20, %34, 0, 0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/schedule-regpressure.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/schedule-regpressure.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/schedule-regpressure.mir
@@ -52,6 +52,6 @@
     %7 = COPY %5
     %6 = DS_READ_B32 %7, 0, 0, implicit $m0, implicit $exec
     DS_WRITE_B32 %7, %6, 4, 0, implicit killed $m0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/sdwa-preserve.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/sdwa-preserve.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/sdwa-preserve.mir
@@ -96,7 +96,7 @@
     %11:vgpr_32 = V_AND_B32_e64 %3, killed %10, implicit $exec
     %17:vgpr_32 = V_MOV_B32_sdwa 0, %4, 0, 5, 2, 4, implicit $exec, implicit %11(tied-def 0)
     FLAT_STORE_DWORD %0, %17, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -139,6 +139,6 @@
     %11:vgpr_32 = V_AND_B32_e64 %3, killed %10, implicit $exec
     %17:vgpr_32 = V_MOV_B32_sdwa 0, %4, 0, 5, 2, 4, implicit $exec, implicit %11(tied-def 0)
     FLAT_STORE_DWORD %0, %17, 0, 0, 0, implicit $exec, implicit $flat_scr :: (store 4)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/sdwa-scalar-ops.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/sdwa-scalar-ops.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/sdwa-scalar-ops.mir
@@ -210,7 +210,7 @@
     S_BRANCH %bb.2.bb2
 
   bb.1.bb1:
-    S_ENDPGM
+    S_ENDPGM 0
 
   bb.2.bb2:
     successors: %bb.1.bb1(0x04000000), %bb.2.bb2(0x7c000000)
@@ -373,7 +373,7 @@
     S_BRANCH %bb.2.bb2
 
   bb.1.bb1:
-    S_ENDPGM
+    S_ENDPGM 0
 
   bb.2.bb2:
     successors: %bb.1.bb1(0x04000000), %bb.2.bb2(0x7c000000)
Index: llvm/trunk/test/CodeGen/AMDGPU/sendmsg-m0-hazard.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/sendmsg-m0-hazard.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/sendmsg-m0-hazard.mir
@@ -15,7 +15,7 @@
   bb.0:
     $m0 = S_MOV_B32 -1
     S_SENDMSG 3, implicit $exec, implicit $m0
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 
@@ -30,7 +30,7 @@
   bb.0:
     $m0 = S_MOV_B32 -1
     S_SENDMSGHALT 3, implicit $exec, implicit $m0
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 
@@ -45,5 +45,5 @@
   bb.0:
     $m0 = S_MOV_B32 -1
     S_TTRACEDATA implicit $m0
-    S_ENDPGM
+    S_ENDPGM 0
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/shrink-vop3-carry-out.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/shrink-vop3-carry-out.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/shrink-vop3-carry-out.mir
@@ -86,7 +86,7 @@
     %29, %9 = V_ADD_I32_e64 %19, %17, implicit $exec
     %24 = V_CNDMASK_B32_e64 0, 1, killed %9, implicit $exec
     BUFFER_STORE_DWORD_ADDR64 %24, %28, killed %16, 0, 0, 0, 0, 0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -170,7 +170,7 @@
     %29, %9 = V_SUB_I32_e64 %19, %17, implicit $exec
     %24 = V_CNDMASK_B32_e64 0, 1, killed %9, implicit $exec
     BUFFER_STORE_DWORD_ADDR64 %24, %28, killed %16, 0, 0, 0, 0, 0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -254,7 +254,7 @@
     %29, %9 = V_SUBREV_I32_e64 %19, %17, implicit $exec
     %24 = V_CNDMASK_B32_e64 0, 1, killed %9, implicit $exec
     BUFFER_STORE_DWORD_ADDR64 %29, %28, killed %16, 0, 0, 0, 0, 0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -338,7 +338,7 @@
     %29, $vcc = V_ADDC_U32_e64 %19, %17, %9, implicit $exec
     %24 = V_CNDMASK_B32_e64 0, 1, killed $vcc, implicit $exec
     BUFFER_STORE_DWORD_ADDR64 %24, %28, killed %16, 0, 0, 0, 0, 0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -423,7 +423,7 @@
     %29, $vcc = V_ADDC_U32_e64 %19, %17, $vcc, implicit $exec
     %24 = V_CNDMASK_B32_e64 0, 1, killed $vcc, implicit $exec
     BUFFER_STORE_DWORD_ADDR64 %24, %28, killed %16, 0, 0, 0, 0, 0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 
@@ -507,6 +507,6 @@
     %29, $vcc = V_ADDC_U32_e64 %19, %17, undef $vcc, implicit $exec
     %24 = V_CNDMASK_B32_e64 0, 1, killed $vcc, implicit $exec
     BUFFER_STORE_DWORD_ADDR64 %24, %28, killed %16, 0, 0, 0, 0, 0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/si-fix-sgpr-copies.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/si-fix-sgpr-copies.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/si-fix-sgpr-copies.mir
@@ -42,7 +42,7 @@
 # GCN-LABEL: name: dead_illegal_virtreg_copy
 # GCN: %0:vgpr_32 = COPY $vgpr0
 # GCN: %1:sreg_32_xm0 = IMPLICIT_DEF
-# GCN: S_ENDPGM implicit %0
+# GCN: S_ENDPGM 0, implicit %0
 
 name: dead_illegal_virtreg_copy
 tracksRegLiveness: true
@@ -52,7 +52,7 @@
     liveins: $vgpr0
     %0:vgpr_32 = COPY $vgpr0
     %1:sreg_32_xm0 = COPY %0
-    S_ENDPGM implicit %1
+    S_ENDPGM 0, implicit %1
 ...
 
 ---
@@ -60,7 +60,7 @@
 # GCN-LABEL: name: dead_illegal_physreg_copy
 # GCN %2:vgpr_32 = COPY $vgpr0
 # GCN: %1:sreg_32_xm0 = IMPLICIT_DEF
-# GCN: S_ENDPGM implicit %2
+# GCN: S_ENDPGM 0, implicit %2
 
 name: dead_illegal_physreg_copy
 tracksRegLiveness: true
@@ -70,5 +70,5 @@
     liveins: $vgpr0
     %0:sreg_32_xm0 = COPY $vgpr0
     %1:sreg_32_xm0 = COPY %0
-    S_ENDPGM implicit %1
+    S_ENDPGM 0, implicit %1
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/si-lower-control-flow.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/si-lower-control-flow.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/si-lower-control-flow.mir
@@ -13,12 +13,12 @@
     ; GCN: [[S_LOAD_DWORD_IMM:%[0-9]+]]:sreg_32_xm0_xexec = S_LOAD_DWORD_IMM [[COPY]], 16, 0
     ; GCN: [[S_AND_B32_:%[0-9]+]]:sreg_32_xm0 = S_AND_B32 [[S_LOAD_DWORD_IMM]], 255, implicit-def $scc
     ; GCN: [[S_AND_B32_1:%[0-9]+]]:sreg_32_xm0 = S_AND_B32 65535, [[S_AND_B32_]], implicit-def $scc
-    ; GCN: S_ENDPGM
+    ; GCN: S_ENDPGM 0
     %0:sgpr_64 = COPY $sgpr4_sgpr5
     %1:sreg_32_xm0_xexec = S_LOAD_DWORD_IMM %0, 16, 0
     %2:sreg_32_xm0 = S_AND_B32 %1, 255, implicit-def $scc
     %3:sreg_32_xm0 = S_AND_B32 65535, %2, implicit-def $scc
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 
 ---
@@ -37,7 +37,7 @@
   ; GCN: bb.1:
   ; GCN:   successors: %bb.2(0x80000000)
   ; GCN: bb.2:
-  ; GCN:   S_ENDPGM
+  ; GCN:   S_ENDPGM 0
   bb.0:
     successors: %bb.1, %bb.2
 
@@ -48,6 +48,6 @@
     successors: %bb.2
 
   bb.2:
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/smem-no-clause-coalesced.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/smem-no-clause-coalesced.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/smem-no-clause-coalesced.mir
@@ -40,7 +40,7 @@
     %8:sreg_64_xexec = S_BUFFER_LOAD_DWORDX2_IMM %3, 640, 0 :: (dereferenceable invariant load 8)
     undef %9.sub0:vreg_128 = V_LSHL_ADD_U32 %6, 4, %4, implicit $exec
     %9.sub1:vreg_128 = V_LSHL_ADD_U32 %5, 4, %0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 
Index: llvm/trunk/test/CodeGen/AMDGPU/subreg-split-live-in-error.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/subreg-split-live-in-error.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/subreg-split-live-in-error.mir
@@ -278,5 +278,5 @@
     %68:vgpr_32 = V_MUL_F32_e32 0, %4, implicit $exec
     %69:vgpr_32 = V_CVT_PKRTZ_F16_F32_e64 0, undef %70:vgpr_32, 0, %68, 0, 0, implicit $exec
     EXP 0, undef %71:vgpr_32, %69, undef %72:vgpr_32, undef %73:vgpr_32, -1, -1, 15, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/undefined-physreg-sgpr-spill.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/undefined-physreg-sgpr-spill.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/undefined-physreg-sgpr-spill.mir
@@ -74,7 +74,7 @@
     liveins: $vgpr0, $vgpr1, $vgpr2, $sgpr2_sgpr3, $sgpr4_sgpr5
 
     $vcc = COPY $vgpr1
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -138,6 +138,6 @@
     liveins: $vgpr0, $vgpr1, $vgpr2, $sgpr2_sgpr3, $sgpr4_sgpr5
 
     $vcc = COPY $vgpr1
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/v_swap_b32.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/v_swap_b32.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/v_swap_b32.mir
@@ -93,7 +93,7 @@
     $vgpr0 = V_MOV_B32_e32 killed $vgpr1, implicit $exec
     $vgpr1 = V_MOV_B32_e32 0, implicit $exec
     $vgpr1 = V_MOV_B32_e32 killed $vgpr2, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 
 # GCN-LABEL: name: swap_virt_copy_condense
@@ -176,7 +176,7 @@
 # GCN-NEXT: %3:vgpr_32 = COPY %0
 # GCN-NEXT: %0:vgpr_32 = COPY %1
 # GCN-NEXT: %1:vgpr_32 = COPY %2
-# GCN-NEXT: S_ENDPGM
+# GCN-NEXT: S_ENDPGM 0
 
 ---
 name:            swap_virt_read_x
@@ -193,7 +193,7 @@
     %3 = COPY %0
     %0 = COPY %1
     %1 = COPY %2
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 
 # GCN-LABEL: name: swap_virt_read_t_twice
@@ -203,7 +203,7 @@
 # GCN-NEXT: %2:vgpr_32 = COPY %0
 # GCN-NEXT: %3:vgpr_32 = COPY %2
 # GCN-NEXT: %0:vgpr_32, %1:vgpr_32 = V_SWAP_B32 %1, %0, implicit $exec
-# GCN-NEXT: S_ENDPGM
+# GCN-NEXT: S_ENDPGM 0
 
 ---
 name:            swap_virt_read_t_twice
@@ -220,7 +220,7 @@
     %3 = COPY %2
     %0 = COPY %1
     %1 = COPY %2
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 
 # GCN-LABEL: name: swap_virt_clobber_y
@@ -231,7 +231,7 @@
 # GCN-NEXT: %0:vgpr_32 = COPY %1
 # GCN-NEXT: %1:vgpr_32 = IMPLICIT_DEF
 # GCN-NEXT: %1:vgpr_32 = COPY %2
-# GCN-NEXT: S_ENDPGM
+# GCN-NEXT: S_ENDPGM 0
 
 ---
 name:            swap_virt_clobber_y
@@ -247,7 +247,7 @@
     %0 = COPY %1
     %1 = IMPLICIT_DEF
     %1 = COPY %2
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 
 # GCN-LABEL: name: swap_virt_clobber_x1
@@ -258,7 +258,7 @@
 # GCN-NEXT: %0:vgpr_32 = COPY %1
 # GCN-NEXT: %0:vgpr_32 = IMPLICIT_DEF
 # GCN-NEXT: %1:vgpr_32 = COPY %2
-# GCN-NEXT: S_ENDPGM
+# GCN-NEXT: S_ENDPGM 0
 
 ---
 name:            swap_virt_clobber_x1
@@ -274,7 +274,7 @@
     %0 = COPY %1
     %0 = IMPLICIT_DEF
     %1 = COPY %2
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 
 # GCN-LABEL: name: swap_virt_clobber_x2
@@ -285,7 +285,7 @@
 # GCN-NEXT: %0:vgpr_32 = IMPLICIT_DEF
 # GCN-NEXT: %0:vgpr_32 = COPY %1
 # GCN-NEXT: %1:vgpr_32 = COPY %2
-# GCN-NEXT: S_ENDPGM
+# GCN-NEXT: S_ENDPGM 0
 
 ---
 name:            swap_virt_clobber_x2
@@ -301,7 +301,7 @@
     %0 = IMPLICIT_DEF
     %0 = COPY %1
     %1 = COPY %2
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 
 # GCN-LABEL: name: swap_virt_clobber_t
@@ -312,7 +312,7 @@
 # GCN-NEXT: %0:vgpr_32 = COPY %1
 # GCN-NEXT: %2:vgpr_32 = IMPLICIT_DEF
 # GCN-NEXT: %1:vgpr_32 = COPY %2
-# GCN-NEXT: S_ENDPGM
+# GCN-NEXT: S_ENDPGM 0
 
 ---
 name:            swap_virt_clobber_t
@@ -328,7 +328,7 @@
     %0 = COPY %1
     %2 = IMPLICIT_DEF
     %1 = COPY %2
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 
 # GCN-LABEL: name: swap_virt_copy_subreg_overlap_x_full
@@ -453,7 +453,7 @@
 # GCN-NEXT: %1:vreg_128 = IMPLICIT_DEF
 # GCN-NEXT: %0.sub0:vreg_128, %1.sub0:vreg_128 = V_SWAP_B32 %1.sub0, %0.sub0, implicit $exec
 # GCN-NEXT: %0.sub1:vreg_128, %1.sub1:vreg_128 = V_SWAP_B32 %1.sub1, %0.sub1, implicit $exec
-# GCN-NEXT: S_ENDPGM
+# GCN-NEXT: S_ENDPGM 0
 ---
 name:            swap_virt_b128_sub0_1
 registers:
@@ -467,7 +467,7 @@
     %2.sub0_sub1 = COPY %0.sub0_sub1
     %0.sub0_sub1 = COPY %1.sub0_sub1
     %1.sub0_sub1 = COPY %2.sub0_sub1
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 
 # GCN-LABEL: name: swap_virt_b128_sub2_3
@@ -476,7 +476,7 @@
 # GCN-NEXT: %1:vreg_128 = IMPLICIT_DEF
 # GCN-NEXT: %0.sub2:vreg_128, %1.sub2:vreg_128 = V_SWAP_B32 %1.sub2, %0.sub2, implicit $exec
 # GCN-NEXT: %0.sub3:vreg_128, %1.sub3:vreg_128 = V_SWAP_B32 %1.sub3, %0.sub3, implicit $exec
-# GCN-NEXT: S_ENDPGM
+# GCN-NEXT: S_ENDPGM 0
 ---
 name:            swap_virt_b128_sub2_3
 registers:
@@ -490,7 +490,7 @@
     %2.sub2_sub3 = COPY %0.sub2_sub3
     %0.sub2_sub3 = COPY %1.sub2_sub3
     %1.sub2_sub3 = COPY %2.sub2_sub3
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 
 
@@ -544,7 +544,7 @@
 # GCN-NEXT: %0.sub0:vreg_64 = COPY %1.sub0, implicit %0
 # GCN-NEXT: %0.sub1:vreg_64 = COPY %1.sub1
 # GCN-NEXT: %1.sub0:vreg_64 = COPY %2.sub0
-# GCN-NEXT: S_ENDPGM
+# GCN-NEXT: S_ENDPGM 0
 ---
 name:            swap_virt_copy_subreg_impuse_x
 registers:
@@ -560,5 +560,5 @@
     %0.sub0 = COPY %1.sub0, implicit %0
     %0.sub1 = COPY %1.sub1
     %1.sub0 = COPY %2.sub0
-    S_ENDPGM
+    S_ENDPGM 0
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/vccz-corrupt-bug-workaround.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/vccz-corrupt-bug-workaround.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/vccz-corrupt-bug-workaround.mir
@@ -105,7 +105,7 @@
     $sgpr3 = S_MOV_B32 61440
     $sgpr2 = S_MOV_B32 -1
     BUFFER_STORE_DWORD_OFFSET killed $vgpr0, killed $sgpr0_sgpr1_sgpr2_sgpr3, 0, 0, 0, 0, 0, implicit $exec :: (store 4 into %ir.out)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 ---
@@ -166,6 +166,6 @@
     $sgpr3 = S_MOV_B32 61440
     $sgpr2 = S_MOV_B32 -1
     BUFFER_STORE_DWORD_OFFSET killed $vgpr0, killed $sgpr0_sgpr1_sgpr2_sgpr3, 0, 0, 0, 0, 0, implicit $exec :: (store 4 into %ir.out)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/vop-shrink-frame-index.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/vop-shrink-frame-index.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/vop-shrink-frame-index.mir
@@ -51,7 +51,7 @@
     %0 = V_MOV_B32_e32 %stack.0.alloca, implicit $exec
     %1 = IMPLICIT_DEF
     %2, $vcc = V_ADD_I32_e64 %0, %1, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 # GCN-LABEL: name: fold_vgpr_fi{{$}}
@@ -72,7 +72,7 @@
     %0 = V_MOV_B32_e32 %stack.0.alloca, implicit $exec
     %1 = IMPLICIT_DEF
     %2, $vcc = V_ADD_I32_e64 %1, %0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 # GCN-LABEL: name: fold_sgpr_fi{{$}}
@@ -94,7 +94,7 @@
     %0 = V_MOV_B32_e32 %stack.0.alloca, implicit $exec
     %1 = IMPLICIT_DEF
     %2, $vcc = V_ADD_I32_e64 %1, %0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 # GCN-LABEL: name: fold_fi_sgpr{{$}}
@@ -116,7 +116,7 @@
     %0 = V_MOV_B32_e32 %stack.0.alloca, implicit $exec
     %1 = IMPLICIT_DEF
     %2, $vcc = V_ADD_I32_e64 %0, %1, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 # TODO: Should probably prefer folding immediate first
 # GCN-LABEL: name: fold_fi_imm{{$}}
@@ -137,7 +137,7 @@
     %0 = V_MOV_B32_e32 %stack.0.alloca, implicit $exec
     %1 = V_MOV_B32_e32 999, implicit $exec
     %2, $vcc = V_ADD_I32_e64 %0, %1, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 # GCN-LABEL: name: fold_imm_fi{{$}}
@@ -158,4 +158,4 @@
     %0 = V_MOV_B32_e32 %stack.0.alloca, implicit $exec
     %1 = V_MOV_B32_e32 999, implicit $exec
     %2, $vcc = V_ADD_I32_e64 %1, %0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
Index: llvm/trunk/test/CodeGen/AMDGPU/vop-shrink-non-ssa.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/vop-shrink-non-ssa.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/vop-shrink-non-ssa.mir
@@ -17,7 +17,7 @@
     %0 = V_MOV_B32_e32 123, implicit $exec
     %1 = V_MOV_B32_e32 456, implicit $exec
     %2, $vcc = V_ADD_I32_e64 %0, %1, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 # GCN-LABEL: name: fold_partially_defined_superreg{{$}}
@@ -35,6 +35,6 @@
     undef %3.sub0 = V_MOV_B32_e32 123, implicit $exec, implicit-def %3
     %1 = V_MOV_B32_e32 456, implicit $exec
     %2, $vcc = V_ADD_I32_e64 %3.sub0, %1, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/waitcnt-back-edge-loop.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/waitcnt-back-edge-loop.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/waitcnt-back-edge-loop.mir
@@ -55,7 +55,7 @@
 
     $vgpr4 = V_MAC_F32_e32 killed $vgpr0, killed $vgpr3, killed $vgpr4, implicit $exec
     EXP_DONE 12, killed $vgpr4, undef $vgpr0, undef $vgpr0, undef $vgpr0, 0, 0, 15, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 
@@ -89,5 +89,5 @@
 
   bb.6:
     S_CBRANCH_SCC1 %bb.0, implicit $scc
-    S_ENDPGM
+    S_ENDPGM 0
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/waitcnt-loop-irreducible.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/waitcnt-loop-irreducible.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/waitcnt-loop-irreducible.mir
@@ -48,7 +48,7 @@
 
   bb.4:
 
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
 
@@ -67,7 +67,7 @@
 # GCN: BUFFER_ATOMIC_ADD_OFFSET_RTN
 # GCN: S_WAITCNT 3952
 # GCN: FLAT_STORE_DWORD
-# GCN: S_ENDPGM
+# GCN: S_ENDPGM 0
 name: irreducible_loop_extended
 
 body: |
@@ -99,5 +99,5 @@
 
   bb.6:
     FLAT_STORE_DWORD $vgpr3_vgpr4, $vgpr2, 0, 0, 0, implicit $exec, implicit $flat_scr
-    S_ENDPGM
+    S_ENDPGM 0
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/waitcnt-loop-single-basic-block.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/waitcnt-loop-single-basic-block.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/waitcnt-loop-single-basic-block.mir
@@ -22,5 +22,5 @@
     $vgpr11 = GLOBAL_LOAD_DWORD $vgpr11_vgpr12, 0, 0, 0, implicit $exec
     S_CBRANCH_SCC1 %bb.1, implicit $scc
   bb.2:
-    S_ENDPGM
+    S_ENDPGM 0
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/waitcnt-preexisting.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/waitcnt-preexisting.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/waitcnt-preexisting.mir
@@ -33,5 +33,5 @@
     $vgpr2 = V_MOV_B32_e32 $vgpr1, implicit $exec, implicit $exec
     $vgpr3 = V_MOV_B32_e32 $vgpr1, implicit $exec, implicit $exec
     IMAGE_STORE_V4_V2 killed renamable $vgpr0_vgpr1_vgpr2_vgpr3, killed renamable $vgpr0_vgpr1, killed renamable $sgpr0_sgpr1_sgpr2_sgpr3_sgpr4_sgpr5_sgpr6_sgpr7, 15, -1, 1, 0, 0, 0, 0, 0, 0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 ...
Index: llvm/trunk/test/CodeGen/AMDGPU/waitcnt.mir
===================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/waitcnt.mir
+++ llvm/trunk/test/CodeGen/AMDGPU/waitcnt.mir
@@ -60,7 +60,7 @@
     $vgpr0 = FLAT_LOAD_DWORD $vgpr1_vgpr2, 0, 0, 0, implicit $exec, implicit $flat_scr :: (load 4 from %ir.flat4)
     $vgpr3_vgpr4_vgpr5_vgpr6 = FLAT_LOAD_DWORDX4 $vgpr7_vgpr8, 0, 0, 0, implicit $exec, implicit $flat_scr :: (load 16 from %ir.flat16)
     $vgpr0 = V_MOV_B32_e32 $vgpr1, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # There is only a single fallthrough successor block, so there's no
@@ -84,7 +84,7 @@
   bb.1:
     $vgpr3_vgpr4 = V_LSHLREV_B64 4, $vgpr7_vgpr8, implicit $exec
     FLAT_STORE_DWORD $vgpr3_vgpr4, $vgpr0, 0, 0, 0, implicit $exec, implicit $flat_scr
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 # The block has a single predecessor with a single successor, but it
@@ -95,7 +95,7 @@
 
 # CHECK: bb.1
 # CHECK-NEXT: FLAT_STORE_DWORD
-# CHECK-NEXT: S_ENDPGM
+# CHECK-NEXT: S_ENDPGM 0
 
 # CHECK: bb.2:
 # CHECK-NEXT: V_LSHLREV_B64
@@ -111,10 +111,10 @@
 
   bb.1:
     FLAT_STORE_DWORD $vgpr8_vgpr9, $vgpr10, 0, 0, 0, implicit $exec, implicit $flat_scr
-    S_ENDPGM
+    S_ENDPGM 0
 
   bb.2:
      $vgpr3_vgpr4 = V_LSHLREV_B64 4, $vgpr7_vgpr8, implicit $exec
     FLAT_STORE_DWORD $vgpr3_vgpr4, $vgpr0, 0, 0, 0, implicit $exec, implicit $flat_scr
-    S_ENDPGM
+    S_ENDPGM 0
 ...
Index: llvm/trunk/test/CodeGen/MIR/AMDGPU/expected-target-index-name.mir
===================================================================
--- llvm/trunk/test/CodeGen/MIR/AMDGPU/expected-target-index-name.mir
+++ llvm/trunk/test/CodeGen/MIR/AMDGPU/expected-target-index-name.mir
@@ -45,5 +45,5 @@
     $sgpr6 = S_MOV_B32 -1
     $vgpr0 = V_MOV_B32_e32 killed $sgpr2, implicit $exec
     BUFFER_STORE_DWORD_OFFSET killed $vgpr0, $sgpr4_sgpr5_sgpr6_sgpr7, 0, 0, 0, 0, 0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 ...
Index: llvm/trunk/test/CodeGen/MIR/AMDGPU/invalid-target-index-operand.mir
===================================================================
--- llvm/trunk/test/CodeGen/MIR/AMDGPU/invalid-target-index-operand.mir
+++ llvm/trunk/test/CodeGen/MIR/AMDGPU/invalid-target-index-operand.mir
@@ -45,5 +45,5 @@
     $sgpr6 = S_MOV_B32 -1
     $vgpr0 = V_MOV_B32_e32 killed $sgpr2, implicit $exec
     BUFFER_STORE_DWORD_OFFSET killed $vgpr0, $sgpr4_sgpr5_sgpr6_sgpr7, 0, 0, 0, 0, 0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 ...
Index: llvm/trunk/test/CodeGen/MIR/AMDGPU/mir-canon-multi.mir
===================================================================
--- llvm/trunk/test/CodeGen/MIR/AMDGPU/mir-canon-multi.mir
+++ llvm/trunk/test/CodeGen/MIR/AMDGPU/mir-canon-multi.mir
@@ -23,6 +23,6 @@
     %16:sgpr_128 = REG_SEQUENCE killed %vreg123_0, %subreg.sub0, %vreg123_1, %subreg.sub1, %vreg123_2, %subreg.sub2, %vreg123_3, %subreg.sub3
 
     BUFFER_STORE_DWORD_ADDR64 %vreg123_1, %27, killed %16, 0, 0, 0, 0, 0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
Index: llvm/trunk/test/CodeGen/MIR/AMDGPU/stack-id.mir
===================================================================
--- llvm/trunk/test/CodeGen/MIR/AMDGPU/stack-id.mir
+++ llvm/trunk/test/CodeGen/MIR/AMDGPU/stack-id.mir
@@ -31,5 +31,5 @@
 
 body: |
   bb.0:
-    S_ENDPGM
+    S_ENDPGM 0
 ...
Index: llvm/trunk/test/CodeGen/MIR/AMDGPU/syncscopes.mir
===================================================================
--- llvm/trunk/test/CodeGen/MIR/AMDGPU/syncscopes.mir
+++ llvm/trunk/test/CodeGen/MIR/AMDGPU/syncscopes.mir
@@ -95,6 +95,6 @@
     $vgpr1 = V_MOV_B32_e32 killed $sgpr5, implicit $exec, implicit killed $sgpr4_sgpr5, implicit $sgpr4_sgpr5, implicit $exec
     $vgpr2 = V_MOV_B32_e32 killed $sgpr8, implicit $exec, implicit $exec
     FLAT_STORE_DWORD killed $vgpr0_vgpr1, killed $vgpr2, 0, -1, 0, implicit $exec, implicit $flat_scr :: (volatile non-temporal store syncscope("wavefront") seq_cst 4 into %ir.wavefront_out)
-    S_ENDPGM
+    S_ENDPGM 0
 
 ...
Index: llvm/trunk/test/CodeGen/MIR/AMDGPU/target-flags.mir
===================================================================
--- llvm/trunk/test/CodeGen/MIR/AMDGPU/target-flags.mir
+++ llvm/trunk/test/CodeGen/MIR/AMDGPU/target-flags.mir
@@ -24,9 +24,9 @@
     ; CHECK-LABEL: name: flags
     ; CHECK: [[SI_PC_ADD_REL_OFFSET:%[0-9]+]]:sreg_64 = SI_PC_ADD_REL_OFFSET target-flags(amdgpu-rel32-lo) @foo + 4, target-flags(amdgpu-rel32-hi) @foo + 4, implicit-def dead $scc
     ; CHECK: [[S_MOV_B64_:%[0-9]+]]:sreg_64 = S_MOV_B64 target-flags(amdgpu-gotprel) @foo
-    ; CHECK: S_ENDPGM
+    ; CHECK: S_ENDPGM 0
     %0 = SI_PC_ADD_REL_OFFSET target-flags(amdgpu-rel32-lo) @foo + 4, target-flags(amdgpu-rel32-hi) @foo + 4, implicit-def dead $scc
   %1 = S_MOV_B64 target-flags(amdgpu-gotprel) @foo
 
-    S_ENDPGM
+    S_ENDPGM 0
 ...
Index: llvm/trunk/test/CodeGen/MIR/AMDGPU/target-index-operands.mir
===================================================================
--- llvm/trunk/test/CodeGen/MIR/AMDGPU/target-index-operands.mir
+++ llvm/trunk/test/CodeGen/MIR/AMDGPU/target-index-operands.mir
@@ -53,7 +53,7 @@
     $sgpr6 = S_MOV_B32 -1
     $vgpr0 = V_MOV_B32_e32 killed $sgpr2, implicit $exec
     BUFFER_STORE_DWORD_OFFSET killed $vgpr0, $sgpr4_sgpr5_sgpr6_sgpr7, 0, 0, 0, 0, 0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 ...
 ---
 name: float2
@@ -83,5 +83,5 @@
     $sgpr6 = S_MOV_B32 -1
     $vgpr0 = V_MOV_B32_e32 killed $sgpr2, implicit $exec
     BUFFER_STORE_DWORD_OFFSET killed $vgpr0, $sgpr4_sgpr5_sgpr6_sgpr7, 0, 0, 0, 0, 0, implicit $exec
-    S_ENDPGM
+    S_ENDPGM 0
 ...
Index: llvm/trunk/test/MC/AMDGPU/s_endpgm.s
===================================================================
--- llvm/trunk/test/MC/AMDGPU/s_endpgm.s
+++ llvm/trunk/test/MC/AMDGPU/s_endpgm.s
@@ -0,0 +1,17 @@
+// RUN: llvm-mc -arch=amdgcn -show-encoding %s | FileCheck %s --check-prefix=GCN
+// RUN: llvm-mc -arch=amdgcn -mcpu=gfx900 -filetype=obj %s | llvm-objcopy -S -K keep_symbol - | llvm-objdump -disassemble -mcpu=gfx900 - | FileCheck %s --check-prefix=BIN
+
+// GCN: s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
+// BIN: s_endpgm    // 000000000000: BF810000
+s_endpgm
+
+// GCN: s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
+// BIN: s_endpgm    // 000000000004: BF810000
+s_endpgm 0
+
+  
+// GCN: s_endpgm 1 ; encoding: [0x01,0x00,0x81,0xbf]
+// BIN: s_endpgm 1  // 000000000008: BF810001
+s_endpgm 1
+  
+  
Index: llvm/trunk/test/MachineVerifier/verifier-implicit-virtreg-invalid-physreg-liveness.mir
===================================================================
--- llvm/trunk/test/MachineVerifier/verifier-implicit-virtreg-invalid-physreg-liveness.mir
+++ llvm/trunk/test/MachineVerifier/verifier-implicit-virtreg-invalid-physreg-liveness.mir
@@ -5,8 +5,8 @@
 
 
 # ERROR: *** Bad machine code: Using an undefined physical register ***
-# ERROR: instruction: S_ENDPGM implicit %0:vgpr_32, implicit $vcc
-# ERROR: operand 1:   implicit $vcc
+# ERROR: instruction: S_ENDPGM 0, implicit %0:vgpr_32, implicit $vcc
+# ERROR: operand 2:   implicit $vcc
 
 ...
 
@@ -16,7 +16,7 @@
 body:             |
   bb.0:
     %0:vgpr_32 = IMPLICIT_DEF
-    S_ENDPGM implicit %0, implicit $vcc
+    S_ENDPGM 0, implicit %0, implicit $vcc
 
 ...