Index: lib/Target/AMDGPU/AMDGPUISelDAGToDAG.cpp
===================================================================
--- lib/Target/AMDGPU/AMDGPUISelDAGToDAG.cpp
+++ lib/Target/AMDGPU/AMDGPUISelDAGToDAG.cpp
@@ -253,7 +253,7 @@
 static unsigned selectSGPRVectorRegClassID(unsigned NumVectorElts) {
   switch (NumVectorElts) {
   case 1:
-    return AMDGPU::SReg_32RegClassID;
+    return AMDGPU::SReg_32_XM0RegClassID;
   case 2:
     return AMDGPU::SReg_64RegClassID;
   case 4:
Index: lib/Target/AMDGPU/SIISelLowering.cpp
===================================================================
--- lib/Target/AMDGPU/SIISelLowering.cpp
+++ lib/Target/AMDGPU/SIISelLowering.cpp
@@ -59,7 +59,7 @@
   addRegisterClass(MVT::i1, &AMDGPU::VReg_1RegClass);
   addRegisterClass(MVT::i64, &AMDGPU::SReg_64RegClass);
 
-  addRegisterClass(MVT::i32, &AMDGPU::SReg_32RegClass);
+  addRegisterClass(MVT::i32, &AMDGPU::SReg_32_XM0RegClass);
   addRegisterClass(MVT::f32, &AMDGPU::VGPR_32RegClass);
 
   addRegisterClass(MVT::f64, &AMDGPU::VReg_64RegClass);
@@ -79,8 +79,8 @@
   addRegisterClass(MVT::v16f32, &AMDGPU::VReg_512RegClass);
 
   if (Subtarget->has16BitInsts()) {
-    addRegisterClass(MVT::i16, &AMDGPU::SReg_32RegClass);
-    addRegisterClass(MVT::f16, &AMDGPU::SReg_32RegClass);
+    addRegisterClass(MVT::i16, &AMDGPU::SReg_32_XM0RegClass);
+    addRegisterClass(MVT::f16, &AMDGPU::SReg_32_XM0RegClass);
   }
 
   computeRegisterProperties(STI.getRegisterInfo());
@@ -941,25 +941,25 @@
   // Start adding system SGPRs.
   if (Info->hasWorkGroupIDX()) {
     unsigned Reg = Info->addWorkGroupIDX();
-    MF.addLiveIn(Reg, &AMDGPU::SReg_32RegClass);
+    MF.addLiveIn(Reg, &AMDGPU::SReg_32_XM0RegClass);
     CCInfo.AllocateReg(Reg);
   }
 
   if (Info->hasWorkGroupIDY()) {
     unsigned Reg = Info->addWorkGroupIDY();
-    MF.addLiveIn(Reg, &AMDGPU::SReg_32RegClass);
+    MF.addLiveIn(Reg, &AMDGPU::SReg_32_XM0RegClass);
     CCInfo.AllocateReg(Reg);
   }
 
   if (Info->hasWorkGroupIDZ()) {
     unsigned Reg = Info->addWorkGroupIDZ();
-    MF.addLiveIn(Reg, &AMDGPU::SReg_32RegClass);
+    MF.addLiveIn(Reg, &AMDGPU::SReg_32_XM0RegClass);
     CCInfo.AllocateReg(Reg);
   }
 
   if (Info->hasWorkGroupInfo()) {
     unsigned Reg = Info->addWorkGroupInfo();
-    MF.addLiveIn(Reg, &AMDGPU::SReg_32RegClass);
+    MF.addLiveIn(Reg, &AMDGPU::SReg_32_XM0RegClass);
     CCInfo.AllocateReg(Reg);
   }
 
@@ -2415,15 +2415,15 @@
                                   SI::KernelInputOffsets::LOCAL_SIZE_Z);
   case Intrinsic::amdgcn_workgroup_id_x:
   case Intrinsic::r600_read_tgid_x:
-    return CreateLiveInRegister(DAG, &AMDGPU::SReg_32RegClass,
+    return CreateLiveInRegister(DAG, &AMDGPU::SReg_32_XM0RegClass,
       TRI->getPreloadedValue(MF, SIRegisterInfo::WORKGROUP_ID_X), VT);
   case Intrinsic::amdgcn_workgroup_id_y:
   case Intrinsic::r600_read_tgid_y:
-    return CreateLiveInRegister(DAG, &AMDGPU::SReg_32RegClass,
+    return CreateLiveInRegister(DAG, &AMDGPU::SReg_32_XM0RegClass,
       TRI->getPreloadedValue(MF, SIRegisterInfo::WORKGROUP_ID_Y), VT);
   case Intrinsic::amdgcn_workgroup_id_z:
   case Intrinsic::r600_read_tgid_z:
-    return CreateLiveInRegister(DAG, &AMDGPU::SReg_32RegClass,
+    return CreateLiveInRegister(DAG, &AMDGPU::SReg_32_XM0RegClass,
       TRI->getPreloadedValue(MF, SIRegisterInfo::WORKGROUP_ID_Z), VT);
   case Intrinsic::amdgcn_workitem_id_x:
   case Intrinsic::r600_read_tidig_x:
@@ -4181,7 +4181,7 @@
       default:
         return std::make_pair(0U, nullptr);
       case 32:
-        return std::make_pair(0U, &AMDGPU::SReg_32RegClass);
+        return std::make_pair(0U, &AMDGPU::SReg_32_XM0RegClass);
       case 64:
         return std::make_pair(0U, &AMDGPU::SGPR_64RegClass);
       case 128:
Index: lib/Target/AMDGPU/SIInstrInfo.cpp
===================================================================
--- lib/Target/AMDGPU/SIInstrInfo.cpp
+++ lib/Target/AMDGPU/SIInstrInfo.cpp
@@ -364,7 +364,8 @@
     return;
   }
 
-  if (RC == &AMDGPU::SReg_32RegClass) {
+  if (RC == &AMDGPU::SReg_32_XM0RegClass ||
+      RC == &AMDGPU::SReg_32RegClass) {
     if (SrcReg == AMDGPU::SCC) {
       BuildMI(MBB, MI, DL, get(AMDGPU::S_CSELECT_B32), DestReg)
           .addImm(-1)
Index: lib/Target/AMDGPU/SIRegisterInfo.cpp
===================================================================
--- lib/Target/AMDGPU/SIRegisterInfo.cpp
+++ lib/Target/AMDGPU/SIRegisterInfo.cpp
@@ -233,7 +233,7 @@
 
   MachineRegisterInfo &MRI = MF->getRegInfo();
   unsigned UnusedCarry = MRI.createVirtualRegister(&AMDGPU::SReg_64RegClass);
-  unsigned OffsetReg = MRI.createVirtualRegister(&AMDGPU::SReg_32RegClass);
+  unsigned OffsetReg = MRI.createVirtualRegister(&AMDGPU::SReg_32_XM0RegClass);
 
   BuildMI(*MBB, Ins, DL, TII->get(AMDGPU::S_MOV_B32), OffsetReg)
     .addImm(Offset);
Index: lib/Target/AMDGPU/SIRegisterInfo.td
===================================================================
--- lib/Target/AMDGPU/SIRegisterInfo.td
+++ lib/Target/AMDGPU/SIRegisterInfo.td
@@ -120,6 +120,11 @@
   let isAllocatable = 0;
 }
 
+def M0_CLASS : RegisterClass<"AMDGPU", [i32], 32, (add M0)> {
+  let CopyCost = 1;
+  let isAllocatable = 0;
+}
+
 // TODO: Do we need to set DwarfRegAlias on register tuples?
 
 // SGPR 32-bit registers
@@ -259,8 +264,9 @@
 
 // Register class for all scalar registers (SGPRs + Special Registers)
 def SReg_32 : RegisterClass<"AMDGPU", [i32, f32, i16, f16], 32,
-  (add SReg_32_XM0, M0, VCC_LO, VCC_HI, EXEC_LO, EXEC_HI, FLAT_SCR_LO, FLAT_SCR_HI)> {
+  (add SReg_32_XM0, M0_CLASS)> {
   let AllocationPriority = 1;
+  let isAllocatable = 0;
 }
 
 def SGPR_64 : RegisterClass<"AMDGPU", [v2i32, i64, f64], 32, (add SGPR_64Regs)> {
Index: lib/Target/AMDGPU/SIWholeQuadMode.cpp
===================================================================
--- lib/Target/AMDGPU/SIWholeQuadMode.cpp
+++ lib/Target/AMDGPU/SIWholeQuadMode.cpp
@@ -437,7 +437,7 @@
 MachineBasicBlock::iterator
 SIWholeQuadMode::saveSCC(MachineBasicBlock &MBB,
                          MachineBasicBlock::iterator Before) {
-  unsigned SaveReg = MRI->createVirtualRegister(&AMDGPU::SReg_32RegClass);
+  unsigned SaveReg = MRI->createVirtualRegister(&AMDGPU::SReg_32_XM0RegClass);
 
   MachineInstr *Save =
       BuildMI(MBB, Before, DebugLoc(), TII->get(AMDGPU::COPY), SaveReg)
Index: test/CodeGen/AMDGPU/detect-dead-lanes.mir
===================================================================
--- test/CodeGen/AMDGPU/detect-dead-lanes.mir
+++ test/CodeGen/AMDGPU/detect-dead-lanes.mir
@@ -27,9 +27,9 @@
 # CHECK: S_NOP 0, implicit undef %5.sub0
 name: test0
 registers:
-  - { id: 0, class: sreg_32 }
-  - { id: 1, class: sreg_32 }
-  - { id: 2, class: sreg_32 }
+  - { id: 0, class: sreg_32_xm0 }
+  - { id: 1, class: sreg_32_xm0 }
+  - { id: 2, class: sreg_32_xm0 }
   - { id: 3, class: sreg_128 }
   - { id: 4, class: sreg_64 }
   - { id: 5, class: sreg_64 }
@@ -87,13 +87,13 @@
   - { id: 0, class: sreg_128 }
   - { id: 1, class: sreg_128 }
   - { id: 2, class: sreg_64 }
-  - { id: 3, class: sreg_32 }
+  - { id: 3, class: sreg_32_xm0 }
   - { id: 4, class: sreg_128 }
   - { id: 5, class: sreg_64 }
-  - { id: 6, class: sreg_32 }
-  - { id: 7, class: sreg_32 }
+  - { id: 6, class: sreg_32_xm0 }
+  - { id: 7, class: sreg_32_xm0 }
   - { id: 8, class: sreg_64 }
-  - { id: 9, class: sreg_32 }
+  - { id: 9, class: sreg_32_xm0 }
   - { id: 10, class: sreg_128 }
 body: |
   bb.0:
@@ -162,12 +162,12 @@
 
 name: test2
 registers:
-  - { id: 0, class: sreg_32 }
-  - { id: 1, class: sreg_32 }
+  - { id: 0, class: sreg_32_xm0 }
+  - { id: 1, class: sreg_32_xm0 }
   - { id: 2, class: sreg_64 }
   - { id: 3, class: sreg_128 }
-  - { id: 4, class: sreg_32 }
-  - { id: 5, class: sreg_32 }
+  - { id: 4, class: sreg_32_xm0 }
+  - { id: 5, class: sreg_32_xm0 }
   - { id: 6, class: sreg_64 }
   - { id: 7, class: sreg_128 }
   - { id: 8, class: sreg_64 }
@@ -260,7 +260,7 @@
 name: test5
 tracksRegLiveness: true
 registers:
-  - { id: 0, class: sreg_32 }
+  - { id: 0, class: sreg_32_xm0 }
   - { id: 1, class: sreg_64 }
 body: |
   bb.0:
@@ -286,9 +286,9 @@
 name: loop0
 tracksRegLiveness: true
 registers:
-  - { id: 0, class: sreg_32 }
-  - { id: 1, class: sreg_32 }
-  - { id: 2, class: sreg_32 }
+  - { id: 0, class: sreg_32_xm0 }
+  - { id: 1, class: sreg_32_xm0 }
+  - { id: 2, class: sreg_32_xm0 }
   - { id: 3, class: sreg_128 }
   - { id: 4, class: sreg_128 }
   - { id: 5, class: sreg_128 }
@@ -339,10 +339,10 @@
 name: loop1
 tracksRegLiveness: true
 registers:
-  - { id: 0, class: sreg_32 }
-  - { id: 1, class: sreg_32 }
-  - { id: 2, class: sreg_32 }
-  - { id: 3, class: sreg_32 }
+  - { id: 0, class: sreg_32_xm0 }
+  - { id: 1, class: sreg_32_xm0 }
+  - { id: 2, class: sreg_32_xm0 }
+  - { id: 3, class: sreg_32_xm0 }
   - { id: 4, class: sreg_128 }
   - { id: 5, class: sreg_128 }
   - { id: 6, class: sreg_128 }
@@ -390,7 +390,7 @@
 name: loop2
 tracksRegLiveness: true
 registers:
-  - { id: 0, class: sreg_32 }
+  - { id: 0, class: sreg_32_xm0 }
   - { id: 1, class: sreg_128 }
   - { id: 2, class: sreg_128 }
   - { id: 3, class: sreg_128 }
Index: test/CodeGen/AMDGPU/inline-constraints.ll
===================================================================
--- test/CodeGen/AMDGPU/inline-constraints.ll
+++ test/CodeGen/AMDGPU/inline-constraints.ll
@@ -22,10 +22,11 @@
   ret void
 }
 
+; FIXME: Should be able to avoid copy
 ; GCN-LABEL: {{^}}inline_sreg_constraint_m0:
 ; GCN: s_mov_b32 m0, -1
-; GCN-NOT: s_mov_b32 s{{[0-9]+}}, m0
-; GCN: ; use m0
+; GCN: s_mov_b32 [[COPY_M0:s[0-9]+]], m0
+; GCN: ; use [[COPY_M0]]
 define void @inline_sreg_constraint_m0() {
   %m0 = tail call i32 asm sideeffect "s_mov_b32 m0, -1", "={M0}"()
   tail call void asm sideeffect "; use $0", "s"(i32 %m0)
Index: test/CodeGen/AMDGPU/llvm.amdgcn.readfirstlane.ll
===================================================================
--- test/CodeGen/AMDGPU/llvm.amdgcn.readfirstlane.ll
+++ test/CodeGen/AMDGPU/llvm.amdgcn.readfirstlane.ll
@@ -22,7 +22,8 @@
 ; TODO: m0 should be folded.
 ; CHECK-LABEL: {{^}}test_readfirstlane_m0:
 ; CHECK: s_mov_b32 m0, -1
-; CHECK: v_mov_b32_e32 [[VVAL:v[0-9]]], m0
+; CHECK: s_mov_b32 [[COPY_M0:s[0-9]+]], m0
+; CHECK: v_mov_b32_e32 [[VVAL:v[0-9]]], [[COPY_M0]]
 ; CHECK: v_readfirstlane_b32 s{{[0-9]+}}, [[VVAL]]
 define void @test_readfirstlane_m0(i32 addrspace(1)* %out) #1 {
   %m0 = call i32 asm "s_mov_b32 m0, -1", "={M0}"()
Index: test/CodeGen/AMDGPU/llvm.amdgcn.readlane.ll
===================================================================
--- test/CodeGen/AMDGPU/llvm.amdgcn.readlane.ll
+++ test/CodeGen/AMDGPU/llvm.amdgcn.readlane.ll
@@ -22,7 +22,8 @@
 ; TODO: m0 should be folded.
 ; CHECK-LABEL: {{^}}test_readlane_m0_sreg:
 ; CHECK: s_mov_b32 m0, -1
-; CHECK: v_mov_b32_e32 [[VVAL:v[0-9]]], m0
+; CHECK: s_mov_b32 [[COPY_M0:s[0-9]+]], m0
+; CHECK: v_mov_b32_e32 [[VVAL:v[0-9]]], [[COPY_M0]]
 ; CHECK: v_readlane_b32 s{{[0-9]+}}, [[VVAL]], s{{[0-9]+}}
 define void @test_readlane_m0_sreg(i32 addrspace(1)* %out, i32 %src1) #1 {
   %m0 = call i32 asm "s_mov_b32 m0, -1", "={M0}"()
Index: test/CodeGen/AMDGPU/read_register.ll
===================================================================
--- test/CodeGen/AMDGPU/read_register.ll
+++ test/CodeGen/AMDGPU/read_register.ll
@@ -3,9 +3,11 @@
 declare i32 @llvm.read_register.i32(metadata) #0
 declare i64 @llvm.read_register.i64(metadata) #0
 
+; FIXME: Should be able to eliminate copy
 ; CHECK-LABEL: {{^}}test_read_m0:
 ; CHECK: s_mov_b32 m0, -1
-; CHECK: v_mov_b32_e32 [[COPY:v[0-9]+]], m0
+; CHECK: s_mov_b32 [[COPY_M0:s[0-9]+]], m0
+; CHECK: v_mov_b32_e32 [[COPY:v[0-9]+]], [[COPY_M0]]
 ; CHECK: buffer_store_dword [[COPY]]
 define void @test_read_m0(i32 addrspace(1)* %out) #0 {
   store volatile i32 0, i32 addrspace(3)* undef
Index: test/CodeGen/AMDGPU/spill-m0.ll
===================================================================
--- test/CodeGen/AMDGPU/spill-m0.ll
+++ test/CodeGen/AMDGPU/spill-m0.ll
@@ -9,38 +9,39 @@
 ; GCN-LABEL: {{^}}spill_m0:
 ; TOSMEM: s_mov_b32 s84, SCRATCH_RSRC_DWORD0
 
-; GCN: s_cmp_lg_u32
+; GCN-DAG: s_cmp_lg_u32
 
-; TOVGPR: s_mov_b32 vcc_hi, m0
-; TOVGPR: v_writelane_b32 [[SPILL_VREG:v[0-9]+]], vcc_hi, 0
+; TOVGPR-DAG: s_mov_b32 [[M0_COPY:s[0-9]+]], m0
+; TOVGPR: v_writelane_b32 [[SPILL_VREG:v[0-9]+]], [[M0_COPY]], 0
 
-; TOVMEM: v_mov_b32_e32 [[SPILL_VREG:v[0-9]+]], m0
+; TOVMEM-DAG: s_mov_b32 [[M0_COPY:s[0-9]+]], m0
+; TOVMEM-DAG: v_mov_b32_e32 [[SPILL_VREG:v[0-9]+]], [[M0_COPY]]
 ; TOVMEM: buffer_store_dword [[SPILL_VREG]], off, s{{\[[0-9]+:[0-9]+\]}}, s{{[0-9]+}} ; 4-byte Folded Spill
 ; TOVMEM: s_waitcnt vmcnt(0)
 
-; TOSMEM: s_mov_b32 vcc_hi, m0
+; TOSMEM-DAG: s_mov_b32 [[M0_COPY:s[0-9]+]], m0
 ; TOSMEM: s_mov_b32 m0, s3{{$}}
-; TOSMEM-NOT: vcc_hi
-; TOSMEM: s_buffer_store_dword vcc_hi, s[84:87], m0 ; 4-byte Folded Spill
+; TOSMEM-NOT: [[M0_COPY]]
+; TOSMEM: s_buffer_store_dword [[M0_COPY]], s[84:87], m0 ; 4-byte Folded Spill
 ; TOSMEM: s_waitcnt lgkmcnt(0)
 
 ; GCN: s_cbranch_scc1 [[ENDIF:BB[0-9]+_[0-9]+]]
 
 ; GCN: [[ENDIF]]:
-; TOVGPR: v_readlane_b32 vcc_hi, [[SPILL_VREG]], 0
-; TOVGPR: s_mov_b32 m0, vcc_hi
+; TOVGPR: v_readlane_b32 [[M0_RESTORE:s[0-9]+]], [[SPILL_VREG]], 0
+; TOVGPR: s_mov_b32 m0, [[M0_RESTORE]]
 
 ; TOVMEM: buffer_load_dword [[RELOAD_VREG:v[0-9]+]], off, s{{\[[0-9]+:[0-9]+\]}}, s{{[0-9]+}} ; 4-byte Folded Reload
 ; TOVMEM: s_waitcnt vmcnt(0)
-; TOVMEM: v_readfirstlane_b32 vcc_hi, [[RELOAD_VREG]]
-; TOVMEM: s_mov_b32 m0, vcc_hi
+; TOVMEM: v_readfirstlane_b32 [[M0_RESTORE:s[0-9]+]], [[RELOAD_VREG]]
+; TOVMEM: s_mov_b32 m0, [[M0_RESTORE]]
 
 ; TOSMEM: s_mov_b32 m0, s3{{$}}
-; TOSMEM: s_buffer_load_dword vcc_hi, s[84:87], m0 ; 4-byte Folded Reload
-; TOSMEM-NOT: vcc_hi
-; TOSMEM: s_mov_b32 m0, vcc_hi
+; TOSMEM: s_buffer_load_dword [[M0_RESTORE:s[0-9]+]], s[84:87], m0 ; 4-byte Folded Reload
+; TOSMEM-NOT: [[M0_RESTORE]]
+; TOSMEM: s_mov_b32 m0, [[M0_RESTORE]]
 
-; GCN: s_add_i32 m0, m0, 1
+; GCN: s_add_i32 s{{[0-9]+}}, m0, 1
 define void @spill_m0(i32 %cond, i32 addrspace(1)* %out) #0 {
 entry:
   %m0 = call i32 asm sideeffect "s_mov_b32 m0, 0", "={M0}"() #0
Index: test/CodeGen/MIR/AMDGPU/si-fix-sgpr-copies.mir
===================================================================
--- test/CodeGen/MIR/AMDGPU/si-fix-sgpr-copies.mir
+++ test/CodeGen/MIR/AMDGPU/si-fix-sgpr-copies.mir
@@ -6,14 +6,14 @@
 name: phi_visit_order
 tracksRegLiveness: true
 registers:
-  - { id: 0, class: sreg_32 }
+  - { id: 0, class: sreg_32_xm0 }
   - { id: 1, class: sreg_64 }
-  - { id: 2, class: sreg_32 }
+  - { id: 2, class: sreg_32_xm0 }
   - { id: 7, class: vgpr_32 }
-  - { id: 8, class: sreg_32 }
+  - { id: 8, class: sreg_32_xm0 }
   - { id: 9, class: vgpr_32 }
   - { id: 10, class: sreg_64 }
-  - { id: 11, class: sreg_32 }
+  - { id: 11, class: sreg_32_xm0 }
 
 body: |
   ; GCN-LABEL: name: phi_visit_order