Index: llvm/lib/Target/AMDGPU/AMDGPUCallLowering.cpp
===================================================================
--- llvm/lib/Target/AMDGPU/AMDGPUCallLowering.cpp
+++ llvm/lib/Target/AMDGPU/AMDGPUCallLowering.cpp
@@ -958,10 +958,14 @@
 }
 
 static unsigned getCallOpcode(const MachineFunction &CallerF, bool IsIndirect,
-                              bool IsTailCall) {
+                              bool IsTailCall, CallingConv::ID CC) {
   assert(!(IsIndirect && IsTailCall) && "Indirect calls can't be tail calls, "
                                         "because the address can be divergent");
-  return IsTailCall ? AMDGPU::SI_TCRETURN : AMDGPU::G_SI_CALL;
+  if (!IsTailCall)
+    return AMDGPU::G_SI_CALL;
+
+  return CC == CallingConv::AMDGPU_Gfx ? AMDGPU::SI_TCRETURN_GFX :
+                                         AMDGPU::SI_TCRETURN;
 }
 
 // Add operands to call instruction to track the callee.
@@ -1186,7 +1190,7 @@
   if (!IsSibCall)
     CallSeqStart = MIRBuilder.buildInstr(AMDGPU::ADJCALLSTACKUP);
 
-  unsigned Opc = getCallOpcode(MF, Info.Callee.isReg(), true);
+  unsigned Opc = getCallOpcode(MF, Info.Callee.isReg(), true, CalleeCC);
   auto MIB = MIRBuilder.buildInstrNoInsert(Opc);
   if (!addCallTargetOperands(MIB, MIRBuilder, Info))
     return false;
@@ -1350,7 +1354,7 @@
 
   // Create a temporarily-floating call instruction so we can add the implicit
   // uses of arg registers.
-  unsigned Opc = getCallOpcode(MF, Info.Callee.isReg(), false);
+  unsigned Opc = getCallOpcode(MF, Info.Callee.isReg(), false, Info.CallConv);
 
   auto MIB = MIRBuilder.buildInstrNoInsert(Opc);
   MIB.addDef(TRI->getReturnAddressReg(MF));
Index: llvm/lib/Target/AMDGPU/AMDGPUISelLowering.h
===================================================================
--- llvm/lib/Target/AMDGPU/AMDGPUISelLowering.h
+++ llvm/lib/Target/AMDGPU/AMDGPUISelLowering.h
@@ -367,6 +367,7 @@
   // Function call.
   CALL,
   TC_RETURN,
+  TC_RETURN_GFX,
   TRAP,
 
   // Masked control flow nodes.
Index: llvm/lib/Target/AMDGPU/AMDGPUISelLowering.cpp
===================================================================
--- llvm/lib/Target/AMDGPU/AMDGPUISelLowering.cpp
+++ llvm/lib/Target/AMDGPU/AMDGPUISelLowering.cpp
@@ -4588,6 +4588,7 @@
   NODE_NAME_CASE(LOOP)
   NODE_NAME_CASE(CALL)
   NODE_NAME_CASE(TC_RETURN)
+  NODE_NAME_CASE(TC_RETURN_GFX)
   NODE_NAME_CASE(TRAP)
   NODE_NAME_CASE(RET_GLUE)
   NODE_NAME_CASE(RETURN_TO_EPILOG)
Index: llvm/lib/Target/AMDGPU/AMDGPUInstrInfo.td
===================================================================
--- llvm/lib/Target/AMDGPU/AMDGPUInstrInfo.td
+++ llvm/lib/Target/AMDGPU/AMDGPUInstrInfo.td
@@ -90,6 +90,11 @@
   [SDNPHasChain, SDNPOptInGlue, SDNPVariadic]
 >;
 
+def AMDGPUtc_return_gfx: SDNode<"AMDGPUISD::TC_RETURN_GFX",
+  SDTypeProfile<0, 3, [SDTCisPtrTy<0>]>,
+  [SDNPHasChain, SDNPOptInGlue, SDNPVariadic]
+>;
+
 def AMDGPUtrap : SDNode<"AMDGPUISD::TRAP",
   SDTypeProfile<0, -1, [SDTCisVT<0, i16>]>,
     [SDNPHasChain, SDNPVariadic, SDNPSideEffect, SDNPInGlue]
Index: llvm/lib/Target/AMDGPU/AMDGPUMCInstLower.cpp
===================================================================
--- llvm/lib/Target/AMDGPU/AMDGPUMCInstLower.cpp
+++ llvm/lib/Target/AMDGPU/AMDGPUMCInstLower.cpp
@@ -134,7 +134,8 @@
     OutMI.addOperand(Dest);
     OutMI.addOperand(Src);
     return;
-  } else if (Opcode == AMDGPU::SI_TCRETURN) {
+  } else if (Opcode == AMDGPU::SI_TCRETURN ||
+             Opcode == AMDGPU::SI_TCRETURN_GFX) {
     // TODO: How to use branch immediate and avoid register+add?
     Opcode = AMDGPU::S_SETPC_B64;
   }
Index: llvm/lib/Target/AMDGPU/SIISelLowering.cpp
===================================================================
--- llvm/lib/Target/AMDGPU/SIISelLowering.cpp
+++ llvm/lib/Target/AMDGPU/SIISelLowering.cpp
@@ -3396,7 +3396,9 @@
   // actual call instruction.
   if (IsTailCall) {
     MFI.setHasTailCall();
-    return DAG.getNode(AMDGPUISD::TC_RETURN, DL, NodeTys, Ops);
+    unsigned OPC = CallConv == CallingConv::AMDGPU_Gfx ?
+                   AMDGPUISD::TC_RETURN_GFX : AMDGPUISD::TC_RETURN;
+    return DAG.getNode(OPC, DL, NodeTys, Ops);
   }
 
   // Returns a chain and a flag for retval copy to use.
Index: llvm/lib/Target/AMDGPU/SIInstructions.td
===================================================================
--- llvm/lib/Target/AMDGPU/SIInstructions.td
+++ llvm/lib/Target/AMDGPU/SIInstructions.td
@@ -614,12 +614,34 @@
   let isConvergent = 1;
 }
 
+// Tail call handling pseudo for AMDGPU_gfx
+def SI_TCRETURN_GFX : SPseudoInstSI <(outs),
+  (ins Gfx_CCR_SGPR_64:$src0, unknown:$callee, i32imm:$fpdiff),
+  [(AMDGPUtc_return_gfx i64:$src0, tglobaladdr:$callee, i32:$fpdiff)]> {
+  let Size = 4;
+  let FixedSize = 1;
+  let isCall = 1;
+  let isTerminator = 1;
+  let isReturn = 1;
+  let isBarrier = 1;
+  let UseNamedOperandTable = 1;
+  let SchedRW = [WriteBranch];
+  // TODO: Should really base this on the call target
+  let isConvergent = 1;
+}
+
 // Handle selecting indirect tail calls
 def : GCNPat<
   (AMDGPUtc_return i64:$src0, (i64 0), (i32 timm:$fpdiff)),
   (SI_TCRETURN CCR_SGPR_64:$src0, (i64 0), i32imm:$fpdiff)
 >;
 
+// Handle selecting indirect tail calls for AMDGPU_gfx
+def : GCNPat<
+  (AMDGPUtc_return_gfx i64:$src0, (i64 0), (i32 timm:$fpdiff)),
+  (SI_TCRETURN_GFX Gfx_CCR_SGPR_64:$src0, (i64 0), i32imm:$fpdiff)
+>;
+
 def ADJCALLSTACKUP : SPseudoInstSI<
   (outs), (ins i32imm:$amt0, i32imm:$amt1),
   [(callseq_start timm:$amt0, timm:$amt1)],
Index: llvm/lib/Target/AMDGPU/SIRegisterInfo.td
===================================================================
--- llvm/lib/Target/AMDGPU/SIRegisterInfo.td
+++ llvm/lib/Target/AMDGPU/SIRegisterInfo.td
@@ -817,12 +817,21 @@
   let HasSGPR = 1;
 }
 
-// CCR (call clobbered registers) SGPR 64-bit registers, currently used for tail call
-// return address only.
-def CCR_SGPR_64 : SIRegisterClass<"AMDGPU", SGPR_64.RegTypes, 32, (add (trunc SGPR_64, 16))> {
+// CCR (call clobbered registers) SGPR 64-bit registers
+def CCR_SGPR_64 : SIRegisterClass<"AMDGPU", SGPR_64.RegTypes, 32,
+                                (add (trunc SGPR_64, 16))> {
   let CopyCost = SGPR_64.CopyCost;
   let AllocationPriority = SGPR_64.AllocationPriority;
-  let HasSGPR = SGPR_64.HasSGPR;
+  let HasSGPR = 1;
+}
+
+// Call clobbered 64-bit SGPRs for AMDGPU_Gfx CC
+def Gfx_CCR_SGPR_64 : SIRegisterClass<"AMDGPU", SGPR_64.RegTypes, 32,
+                                (add (trunc (shl SGPR_64, 15), 1), // s[30:31]
+                                     (trunc (shl SGPR_64, 18), 14))> { // s[36:37]-s[s62:63]
+  let CopyCost = SGPR_64.CopyCost;
+  let AllocationPriority = SGPR_64.AllocationPriority;
+  let HasSGPR = 1;
 }
 
 def TTMP_64 : SIRegisterClass<"AMDGPU", [v2i32, i64, f64, v4i16, v4f16], 32,
Index: llvm/test/CodeGen/AMDGPU/inline-asm.i128.ll
===================================================================
--- llvm/test/CodeGen/AMDGPU/inline-asm.i128.ll
+++ llvm/test/CodeGen/AMDGPU/inline-asm.i128.ll
@@ -8,15 +8,15 @@
 define amdgpu_kernel void @s_input_output_i128() {
   ; GFX908-LABEL: name: s_input_output_i128
   ; GFX908: bb.0 (%ir-block.0):
-  ; GFX908-NEXT:   INLINEASM &"; def $0", 1 /* sideeffect attdialect */, 6946826 /* regdef:SGPR_128 */, def %4
+  ; GFX908-NEXT:   INLINEASM &"; def $0", 1 /* sideeffect attdialect */, 7274506 /* regdef:SGPR_128 */, def %4
   ; GFX908-NEXT:   [[COPY:%[0-9]+]]:sgpr_128 = COPY %4
-  ; GFX908-NEXT:   INLINEASM &"; use $0", 1 /* sideeffect attdialect */, 6946825 /* reguse:SGPR_128 */, [[COPY]]
+  ; GFX908-NEXT:   INLINEASM &"; use $0", 1 /* sideeffect attdialect */, 7274505 /* reguse:SGPR_128 */, [[COPY]]
   ; GFX908-NEXT:   S_ENDPGM 0
   ; GFX90A-LABEL: name: s_input_output_i128
   ; GFX90A: bb.0 (%ir-block.0):
-  ; GFX90A-NEXT:   INLINEASM &"; def $0", 1 /* sideeffect attdialect */, 6946826 /* regdef:SGPR_128 */, def %4
+  ; GFX90A-NEXT:   INLINEASM &"; def $0", 1 /* sideeffect attdialect */, 7274506 /* regdef:SGPR_128 */, def %4
   ; GFX90A-NEXT:   [[COPY:%[0-9]+]]:sgpr_128 = COPY %4
-  ; GFX90A-NEXT:   INLINEASM &"; use $0", 1 /* sideeffect attdialect */, 6946825 /* reguse:SGPR_128 */, [[COPY]]
+  ; GFX90A-NEXT:   INLINEASM &"; use $0", 1 /* sideeffect attdialect */, 7274505 /* reguse:SGPR_128 */, [[COPY]]
   ; GFX90A-NEXT:   S_ENDPGM 0
   %val = tail call i128 asm sideeffect "; def $0", "=s"()
   call void asm sideeffect "; use $0", "s"(i128 %val)
@@ -26,15 +26,15 @@
 define amdgpu_kernel void @v_input_output_i128() {
   ; GFX908-LABEL: name: v_input_output_i128
   ; GFX908: bb.0 (%ir-block.0):
-  ; GFX908-NEXT:   INLINEASM &"; def $0", 1 /* sideeffect attdialect */, 5898250 /* regdef:VReg_128 */, def %4
+  ; GFX908-NEXT:   INLINEASM &"; def $0", 1 /* sideeffect attdialect */, 6225930 /* regdef:VReg_128 */, def %4
   ; GFX908-NEXT:   [[COPY:%[0-9]+]]:vreg_128 = COPY %4
-  ; GFX908-NEXT:   INLINEASM &"; use $0", 1 /* sideeffect attdialect */, 5898249 /* reguse:VReg_128 */, [[COPY]]
+  ; GFX908-NEXT:   INLINEASM &"; use $0", 1 /* sideeffect attdialect */, 6225929 /* reguse:VReg_128 */, [[COPY]]
   ; GFX908-NEXT:   S_ENDPGM 0
   ; GFX90A-LABEL: name: v_input_output_i128
   ; GFX90A: bb.0 (%ir-block.0):
-  ; GFX90A-NEXT:   INLINEASM &"; def $0", 1 /* sideeffect attdialect */, 6225930 /* regdef:VReg_128_Align2 */, def %4
+  ; GFX90A-NEXT:   INLINEASM &"; def $0", 1 /* sideeffect attdialect */, 6553610 /* regdef:VReg_128_Align2 */, def %4
   ; GFX90A-NEXT:   [[COPY:%[0-9]+]]:vreg_128_align2 = COPY %4
-  ; GFX90A-NEXT:   INLINEASM &"; use $0", 1 /* sideeffect attdialect */, 6225929 /* reguse:VReg_128_Align2 */, [[COPY]]
+  ; GFX90A-NEXT:   INLINEASM &"; use $0", 1 /* sideeffect attdialect */, 6553609 /* reguse:VReg_128_Align2 */, [[COPY]]
   ; GFX90A-NEXT:   S_ENDPGM 0
   %val = tail call i128 asm sideeffect "; def $0", "=v"()
   call void asm sideeffect "; use $0", "v"(i128 %val)
@@ -44,15 +44,15 @@
 define amdgpu_kernel void @a_input_output_i128() {
   ; GFX908-LABEL: name: a_input_output_i128
   ; GFX908: bb.0 (%ir-block.0):
-  ; GFX908-NEXT:   INLINEASM &"; def $0", 1 /* sideeffect attdialect */, 5832714 /* regdef:AReg_128 */, def %4
+  ; GFX908-NEXT:   INLINEASM &"; def $0", 1 /* sideeffect attdialect */, 6160394 /* regdef:AReg_128 */, def %4
   ; GFX908-NEXT:   [[COPY:%[0-9]+]]:areg_128 = COPY %4
-  ; GFX908-NEXT:   INLINEASM &"; use $0", 1 /* sideeffect attdialect */, 5832713 /* reguse:AReg_128 */, [[COPY]]
+  ; GFX908-NEXT:   INLINEASM &"; use $0", 1 /* sideeffect attdialect */, 6160393 /* reguse:AReg_128 */, [[COPY]]
   ; GFX908-NEXT:   S_ENDPGM 0
   ; GFX90A-LABEL: name: a_input_output_i128
   ; GFX90A: bb.0 (%ir-block.0):
-  ; GFX90A-NEXT:   INLINEASM &"; def $0", 1 /* sideeffect attdialect */, 6094858 /* regdef:AReg_128_Align2 */, def %4
+  ; GFX90A-NEXT:   INLINEASM &"; def $0", 1 /* sideeffect attdialect */, 6422538 /* regdef:AReg_128_Align2 */, def %4
   ; GFX90A-NEXT:   [[COPY:%[0-9]+]]:areg_128_align2 = COPY %4
-  ; GFX90A-NEXT:   INLINEASM &"; use $0", 1 /* sideeffect attdialect */, 6094857 /* reguse:AReg_128_Align2 */, [[COPY]]
+  ; GFX90A-NEXT:   INLINEASM &"; use $0", 1 /* sideeffect attdialect */, 6422537 /* reguse:AReg_128_Align2 */, [[COPY]]
   ; GFX90A-NEXT:   S_ENDPGM 0
   %val = call i128 asm sideeffect "; def $0", "=a"()
   call void asm sideeffect "; use $0", "a"(i128 %val)
Index: llvm/test/CodeGen/AMDGPU/partial-regcopy-and-spill-missed-at-regalloc.ll
===================================================================
--- llvm/test/CodeGen/AMDGPU/partial-regcopy-and-spill-missed-at-regalloc.ll
+++ llvm/test/CodeGen/AMDGPU/partial-regcopy-and-spill-missed-at-regalloc.ll
@@ -11,7 +11,7 @@
   ; REGALLOC-GFX908-NEXT:   liveins: $sgpr4_sgpr5
   ; REGALLOC-GFX908-NEXT: {{  $}}
   ; REGALLOC-GFX908-NEXT:   INLINEASM &"; use $0", 1 /* sideeffect attdialect */, 1703945 /* reguse:AGPR_32 */, undef %5:agpr_32
-  ; REGALLOC-GFX908-NEXT:   INLINEASM &"; def $0", 1 /* sideeffect attdialect */, 5898250 /* regdef:VReg_128 */, def %26
+  ; REGALLOC-GFX908-NEXT:   INLINEASM &"; def $0", 1 /* sideeffect attdialect */, 6225930 /* regdef:VReg_128 */, def %26
   ; REGALLOC-GFX908-NEXT:   [[COPY:%[0-9]+]]:av_128 = COPY %26
   ; REGALLOC-GFX908-NEXT:   INLINEASM &"; def $0", 1 /* sideeffect attdialect */, 3080202 /* regdef:VReg_64 */, def %23
   ; REGALLOC-GFX908-NEXT:   SI_SPILL_V64_SAVE %23, %stack.0, $sgpr32, 0, implicit $exec :: (store (s64) into %stack.0, align 4, addrspace 5)
@@ -35,7 +35,7 @@
   ; PEI-GFX908-NEXT:   $sgpr8 = S_ADD_U32 $sgpr8, $sgpr7, implicit-def $scc, implicit-def $sgpr8_sgpr9_sgpr10_sgpr11
   ; PEI-GFX908-NEXT:   $sgpr9 = S_ADDC_U32 $sgpr9, 0, implicit-def dead $scc, implicit $scc, implicit-def $sgpr8_sgpr9_sgpr10_sgpr11
   ; PEI-GFX908-NEXT:   INLINEASM &"; use $0", 1 /* sideeffect attdialect */, 1703945 /* reguse:AGPR_32 */, undef renamable $agpr0
-  ; PEI-GFX908-NEXT:   INLINEASM &"; def $0", 1 /* sideeffect attdialect */, 5898250 /* regdef:VReg_128 */, def renamable $vgpr0_vgpr1_vgpr2_vgpr3
+  ; PEI-GFX908-NEXT:   INLINEASM &"; def $0", 1 /* sideeffect attdialect */, 6225930 /* regdef:VReg_128 */, def renamable $vgpr0_vgpr1_vgpr2_vgpr3
   ; PEI-GFX908-NEXT:   renamable $agpr0_agpr1_agpr2_agpr3 = COPY killed renamable $vgpr0_vgpr1_vgpr2_vgpr3, implicit $exec
   ; PEI-GFX908-NEXT:   INLINEASM &"; def $0", 1 /* sideeffect attdialect */, 3080202 /* regdef:VReg_64 */, def renamable $vgpr0_vgpr1
   ; PEI-GFX908-NEXT:   BUFFER_STORE_DWORD_OFFSET killed $vgpr0, $sgpr8_sgpr9_sgpr10_sgpr11, 0, 4, 0, 0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $vgpr0_vgpr1 :: (store (s32) into %stack.0, addrspace 5)
@@ -58,7 +58,7 @@
   ; REGALLOC-GFX90A-NEXT:   liveins: $sgpr4_sgpr5
   ; REGALLOC-GFX90A-NEXT: {{  $}}
   ; REGALLOC-GFX90A-NEXT:   INLINEASM &"; use $0", 1 /* sideeffect attdialect */, 1703945 /* reguse:AGPR_32 */, undef %5:agpr_32
-  ; REGALLOC-GFX90A-NEXT:   INLINEASM &"; def $0", 1 /* sideeffect attdialect */, 6225930 /* regdef:VReg_128_Align2 */, def %25
+  ; REGALLOC-GFX90A-NEXT:   INLINEASM &"; def $0", 1 /* sideeffect attdialect */, 6553610 /* regdef:VReg_128_Align2 */, def %25
   ; REGALLOC-GFX90A-NEXT:   [[COPY:%[0-9]+]]:av_128_align2 = COPY %25
   ; REGALLOC-GFX90A-NEXT:   INLINEASM &"; def $0", 1 /* sideeffect attdialect */, 3407882 /* regdef:VReg_64_Align2 */, def %23
   ; REGALLOC-GFX90A-NEXT:   SI_SPILL_V64_SAVE %23, %stack.0, $sgpr32, 0, implicit $exec :: (store (s64) into %stack.0, align 4, addrspace 5)
@@ -80,7 +80,7 @@
   ; PEI-GFX90A-NEXT:   $sgpr8 = S_ADD_U32 $sgpr8, $sgpr7, implicit-def $scc, implicit-def $sgpr8_sgpr9_sgpr10_sgpr11
   ; PEI-GFX90A-NEXT:   $sgpr9 = S_ADDC_U32 $sgpr9, 0, implicit-def dead $scc, implicit $scc, implicit-def $sgpr8_sgpr9_sgpr10_sgpr11
   ; PEI-GFX90A-NEXT:   INLINEASM &"; use $0", 1 /* sideeffect attdialect */, 1703945 /* reguse:AGPR_32 */, undef renamable $agpr0
-  ; PEI-GFX90A-NEXT:   INLINEASM &"; def $0", 1 /* sideeffect attdialect */, 6225930 /* regdef:VReg_128_Align2 */, def renamable $vgpr0_vgpr1_vgpr2_vgpr3
+  ; PEI-GFX90A-NEXT:   INLINEASM &"; def $0", 1 /* sideeffect attdialect */, 6553610 /* regdef:VReg_128_Align2 */, def renamable $vgpr0_vgpr1_vgpr2_vgpr3
   ; PEI-GFX90A-NEXT:   renamable $agpr0_agpr1_agpr2_agpr3 = COPY killed renamable $vgpr0_vgpr1_vgpr2_vgpr3, implicit $exec
   ; PEI-GFX90A-NEXT:   INLINEASM &"; def $0", 1 /* sideeffect attdialect */, 3407882 /* regdef:VReg_64_Align2 */, def renamable $vgpr0_vgpr1
   ; PEI-GFX90A-NEXT:   BUFFER_STORE_DWORD_OFFSET killed $vgpr0, $sgpr8_sgpr9_sgpr10_sgpr11, 0, 4, 0, 0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $vgpr0_vgpr1 :: (store (s32) into %stack.0, addrspace 5)
Index: llvm/test/CodeGen/AMDGPU/tail-call-amdgpu-gfx.ll
===================================================================
--- llvm/test/CodeGen/AMDGPU/tail-call-amdgpu-gfx.ll
+++ llvm/test/CodeGen/AMDGPU/tail-call-amdgpu-gfx.ll
@@ -2,8 +2,6 @@
 ; RUN: llc -mtriple=amdgcn--amdpal -mcpu=gfx900 -verify-machineinstrs < %s | FileCheck -check-prefixes=GCN -enable-var-scope %s
 ; RUN: llc -global-isel -mtriple=amdgcn--amdpal -mcpu=gfx900 -verify-machineinstrs < %s | FileCheck -check-prefixes=GCN -enable-var-scope %s
 
-; FIXME: @caller uses s[4:5] to store the address of @callee.
-;        These registers are callee-save in the amdgpu_gfx calling convention, so they must not be clobbered, but they are clobbered here.
 
 ; Callee with VGPR arguments
 define hidden amdgpu_gfx float @callee(float %v.arg0, float %v.arg1) {
@@ -20,22 +18,12 @@
 ; GCN-LABEL: caller:
 ; GCN:       ; %bb.0:
 ; GCN-NEXT:    s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
-; GCN-NEXT:    s_xor_saveexec_b64 s[34:35], -1
-; GCN-NEXT:    buffer_store_dword v2, off, s[0:3], s32 ; 4-byte Folded Spill
-; GCN-NEXT:    s_mov_b64 exec, s[34:35]
-; GCN-NEXT:    v_writelane_b32 v2, s4, 0
-; GCN-NEXT:    v_writelane_b32 v2, s5, 1
-; GCN-NEXT:    s_getpc_b64 s[4:5]
-; GCN-NEXT:    s_add_u32 s4, s4, callee@rel32@lo+4
-; GCN-NEXT:    s_addc_u32 s5, s5, callee@rel32@hi+12
 ; GCN-NEXT:    v_add_f32_e32 v0, 1.0, v0
 ; GCN-NEXT:    v_mov_b32_e32 v1, 2.0
-; GCN-NEXT:    v_readlane_b32 s5, v2, 1
-; GCN-NEXT:    v_readlane_b32 s4, v2, 0
-; GCN-NEXT:    s_xor_saveexec_b64 s[34:35], -1
-; GCN-NEXT:    buffer_load_dword v2, off, s[0:3], s32 ; 4-byte Folded Reload
-; GCN-NEXT:    s_mov_b64 exec, s[34:35]
-; GCN-NEXT:    s_setpc_b64 s[4:5]
+; GCN-NEXT:    s_getpc_b64 s[36:37]
+; GCN-NEXT:    s_add_u32 s36, s36, callee@rel32@lo+4
+; GCN-NEXT:    s_addc_u32 s37, s37, callee@rel32@hi+12
+; GCN-NEXT:    s_setpc_b64 s[36:37]
   %add = fadd float %arg0, 1.0
   %call = tail call amdgpu_gfx float @callee(float %add, float 2.0)
   ret float %call