AMDGPU/SI: Add support for triples with the mesa3d operating system

Summary:
mesa3d will use the same kernel calling convention as amdhsa, but it will
handle everything else like the default 'unknown' OS type.

Reviewers: arsenm

Subscribers: arsenm, llvm-commits, kzhuravl

Differential Revision: https://reviews.llvm.org/D22783

llvm-svn: 281779

10 files changed, 89 insertions, 77 deletions

diff --git a/llvm/lib/Target/AMDGPU/AMDGPUAnnotateKernelFeatures.cpp b/llvm/lib/Target/AMDGPU/AMDGPUAnnotateKernelFeatures.cpp
index 7ee2dc71fce..60afba65694 100644
--- a/llvm/lib/Target/AMDGPU/AMDGPUAnnotateKernelFeatures.cpp
+++ b/llvm/lib/Target/AMDGPU/AMDGPUAnnotateKernelFeatures.cpp
@@ -202,7 +202,7 @@ bool AMDGPUAnnotateKernelFeatures::runOnModule(Module &M) {
   // always initialized.
 
   bool Changed = addAttrsForIntrinsics(M, IntrinsicToAttr);
-  if (TT.getOS() == Triple::AMDHSA) {
+  if (TT.getOS() == Triple::AMDHSA || TT.getOS() == Triple::Mesa3D) {
     Changed |= addAttrsForIntrinsics(M, HSAIntrinsicToAttr);
 
     for (Function &F : M) {
diff --git a/llvm/lib/Target/AMDGPU/AMDGPUAsmPrinter.cpp b/llvm/lib/Target/AMDGPU/AMDGPUAsmPrinter.cpp
index 6be53b11d0e..f2b9fc8477d 100644
--- a/llvm/lib/Target/AMDGPU/AMDGPUAsmPrinter.cpp
+++ b/llvm/lib/Target/AMDGPU/AMDGPUAsmPrinter.cpp
@@ -123,7 +123,7 @@ void AMDGPUAsmPrinter::EmitStartOfAsmFile(Module &M) {
 void AMDGPUAsmPrinter::EmitFunctionBodyStart() {
   const AMDGPUSubtarget &STM = MF->getSubtarget<AMDGPUSubtarget>();
   SIProgramInfo KernelInfo;
-  if (STM.isAmdHsaOS()) {
+  if (STM.isAmdCodeObjectV2()) {
     getSIProgramInfo(KernelInfo, *MF);
     EmitAmdKernelCodeT(*MF, KernelInfo);
   }
@@ -132,7 +132,7 @@ void AMDGPUAsmPrinter::EmitFunctionBodyStart() {
 void AMDGPUAsmPrinter::EmitFunctionEntryLabel() {
   const SIMachineFunctionInfo *MFI = MF->getInfo<SIMachineFunctionInfo>();
   const AMDGPUSubtarget &STM = MF->getSubtarget<AMDGPUSubtarget>();
-  if (MFI->isKernel() && STM.isAmdHsaOS()) {
+  if (MFI->isKernel() && STM.isAmdCodeObjectV2()) {
     AMDGPUTargetStreamer *TS =
         static_cast<AMDGPUTargetStreamer *>(OutStreamer->getTargetStreamer());
     TS->EmitAMDGPUSymbolType(CurrentFnSym->getName(),
diff --git a/llvm/lib/Target/AMDGPU/AMDGPUSubtarget.h b/llvm/lib/Target/AMDGPU/AMDGPUSubtarget.h
index b1cadeb0459..6b953eadeed 100644
--- a/llvm/lib/Target/AMDGPU/AMDGPUSubtarget.h
+++ b/llvm/lib/Target/AMDGPU/AMDGPUSubtarget.h
@@ -138,6 +138,10 @@ public:
     return TargetTriple.getOS() == Triple::AMDHSA;
   }
 
+  bool isMesa3DOS() const {
+    return TargetTriple.getOS() == Triple::Mesa3D;
+  }
+
   Generation getGeneration() const {
     return Gen;
   }
@@ -270,10 +274,14 @@ public:
     return EnableXNACK;
   }
 
+  bool isAmdCodeObjectV2() const {
+    return isAmdHsaOS() || isMesa3DOS();
+  }
+
   /// \brief Returns the offset in bytes from the start of the input buffer
   ///        of the first explicit kernel argument.
   unsigned getExplicitKernelArgOffset() const {
-    return isAmdHsaOS() ? 0 : 36;
+    return isAmdCodeObjectV2() ? 0 : 36;
   }
 
   unsigned getAlignmentForImplicitArgPtr() const {
diff --git a/llvm/lib/Target/AMDGPU/SIFrameLowering.cpp b/llvm/lib/Target/AMDGPU/SIFrameLowering.cpp
index d184e1b9620..0c0d9c9f077 100644
--- a/llvm/lib/Target/AMDGPU/SIFrameLowering.cpp
+++ b/llvm/lib/Target/AMDGPU/SIFrameLowering.cpp
@@ -221,7 +221,7 @@ void SIFrameLowering::emitPrologue(MachineFunction &MF,
     MF, SIRegisterInfo::PRIVATE_SEGMENT_WAVE_BYTE_OFFSET);
 
   unsigned PreloadedPrivateBufferReg = AMDGPU::NoRegister;
-  if (ST.isAmdHsaOS()) {
+  if (ST.isAmdCodeObjectV2()) {
     PreloadedPrivateBufferReg = TRI->getPreloadedValue(
       MF, SIRegisterInfo::PRIVATE_SEGMENT_BUFFER);
   }
@@ -240,7 +240,7 @@ void SIFrameLowering::emitPrologue(MachineFunction &MF,
   MRI.addLiveIn(PreloadedScratchWaveOffsetReg);
   MBB.addLiveIn(PreloadedScratchWaveOffsetReg);
 
-  if (ST.isAmdHsaOS()) {
+  if (ST.isAmdCodeObjectV2()) {
     MRI.addLiveIn(PreloadedPrivateBufferReg);
     MBB.addLiveIn(PreloadedPrivateBufferReg);
   }
@@ -265,7 +265,7 @@ void SIFrameLowering::emitPrologue(MachineFunction &MF,
       .addReg(PreloadedScratchWaveOffsetReg, RegState::Kill);
   }
 
-  if (ST.isAmdHsaOS()) {
+  if (ST.isAmdCodeObjectV2()) {
     // Insert copies from argument register.
     assert(
       !TRI->isSubRegisterEq(PreloadedPrivateBufferReg, ScratchRsrcReg) &&
diff --git a/llvm/lib/Target/AMDGPU/SIISelLowering.cpp b/llvm/lib/Target/AMDGPU/SIISelLowering.cpp
index 4dd2932f56b..6b94333bd41 100644
--- a/llvm/lib/Target/AMDGPU/SIISelLowering.cpp
+++ b/llvm/lib/Target/AMDGPU/SIISelLowering.cpp
@@ -872,12 +872,12 @@ SDValue SITargetLowering::LowerFormalArguments(
   if (HasStackObjects)
     Info->setHasNonSpillStackObjects(true);
 
-  if (ST.isAmdHsaOS()) {
+  if (ST.isAmdCodeObjectV2()) {
     // TODO: Assume we will spill without optimizations.
     if (HasStackObjects) {
       // If we have stack objects, we unquestionably need the private buffer
-      // resource. For the HSA ABI, this will be the first 4 user SGPR
-      // inputs. We can reserve those and use them directly.
+      // resource. For the Code Object V2 ABI, this will be the first 4 user
+      // SGPR inputs. We can reserve those and use them directly.
 
       unsigned PrivateSegmentBufferReg = TRI->getPreloadedValue(
         MF, SIRegisterInfo::PRIVATE_SEGMENT_BUFFER);
@@ -1994,7 +1994,7 @@ SDValue SITargetLowering::LowerINTRINSIC_WO_CHAIN(SDValue Op,
   switch (IntrinsicID) {
   case Intrinsic::amdgcn_dispatch_ptr:
   case Intrinsic::amdgcn_queue_ptr: {
-    if (!Subtarget->isAmdHsaOS()) {
+    if (!Subtarget->isAmdCodeObjectV2()) {
       DiagnosticInfoUnsupported BadIntrin(
           *MF.getFunction(), "unsupported hsa intrinsic without hsa target",
           DL.getDebugLoc());
diff --git a/llvm/lib/Target/AMDGPU/SIMachineFunctionInfo.cpp b/llvm/lib/Target/AMDGPU/SIMachineFunctionInfo.cpp
index 10f73a7f0ee..e911817c451 100644
--- a/llvm/lib/Target/AMDGPU/SIMachineFunctionInfo.cpp
+++ b/llvm/lib/Target/AMDGPU/SIMachineFunctionInfo.cpp
@@ -114,7 +114,7 @@ SIMachineFunctionInfo::SIMachineFunctionInfo(const MachineFunction &MF)
   if (HasStackObjects || MaySpill)
     PrivateSegmentWaveByteOffset = true;
 
-  if (ST.isAmdHsaOS()) {
+  if (ST.isAmdCodeObjectV2()) {
     if (HasStackObjects || MaySpill)
       PrivateSegmentBuffer = true;
 
diff --git a/llvm/lib/Target/AMDGPU/SIRegisterInfo.cpp b/llvm/lib/Target/AMDGPU/SIRegisterInfo.cpp
index 6804c0f6bb7..bcee76f5299 100644
--- a/llvm/lib/Target/AMDGPU/SIRegisterInfo.cpp
+++ b/llvm/lib/Target/AMDGPU/SIRegisterInfo.cpp
@@ -801,7 +801,8 @@ unsigned SIRegisterInfo::getPreloadedValue(const MachineFunction &MF,
   case SIRegisterInfo::PRIVATE_SEGMENT_WAVE_BYTE_OFFSET:
     return MFI->PrivateSegmentWaveByteOffsetSystemSGPR;
   case SIRegisterInfo::PRIVATE_SEGMENT_BUFFER:
-    assert(ST.isAmdHsaOS() && "Non-HSA ABI currently uses relocations");
+    assert(ST.isAmdCodeObjectV2() &&
+           "Non-CodeObjectV2 ABI currently uses relocations");
     assert(MFI->hasPrivateSegmentBuffer());
     return MFI->PrivateSegmentBufferUserSGPR;
   case SIRegisterInfo::KERNARG_SEGMENT_PTR:
diff --git a/llvm/test/CodeGen/AMDGPU/llvm.amdgcn.kernarg.segment.ptr.ll b/llvm/test/CodeGen/AMDGPU/llvm.amdgcn.kernarg.segment.ptr.ll
index c79cf8dffe0..c108be5cd8a 100644
--- a/llvm/test/CodeGen/AMDGPU/llvm.amdgcn.kernarg.segment.ptr.ll
+++ b/llvm/test/CodeGen/AMDGPU/llvm.amdgcn.kernarg.segment.ptr.ll
@@ -1,11 +1,12 @@
-; RUN: llc -mtriple=amdgcn--amdhsa -mcpu=kaveri -verify-machineinstrs < %s | FileCheck -check-prefix=HSA -check-prefix=ALL %s
-; RUN: llc -mtriple=amdgcn-mesa-mesa3d -verify-machineinstrs < %s | FileCheck -check-prefix=MESA -check-prefix=ALL %s
+; RUN: llc -mtriple=amdgcn--amdhsa -mcpu=kaveri -verify-machineinstrs < %s | FileCheck -check-prefixes=CO-V2,HSA,ALL %s
+; RUN: llc -mtriple=amdgcn-mesa-mesa3d -verify-machineinstrs < %s | FileCheck -check-prefixes=CO-V2,OS-MESA3D,MESA,ALL %s
+; RUN: llc -mtriple=amdgcn-mesa-unknown -verify-machineinstrs < %s | FileCheck -check-prefixes=OS-UNKNOWN,MESA,ALL %s
 
 ; ALL-LABEL: {{^}}test:
-; HSA: enable_sgpr_kernarg_segment_ptr = 1
-; HSA: s_load_dword s{{[0-9]+}}, s[4:5], 0xa
+; CO-V2: enable_sgpr_kernarg_segment_ptr = 1
+; CO-V2: s_load_dword s{{[0-9]+}}, s[4:5], 0xa
 
-; MESA: s_load_dword s{{[0-9]+}}, s[0:1], 0xa
+; OS-UNKNOWN: s_load_dword s{{[0-9]+}}, s[0:1], 0xa
 define void @test(i32 addrspace(1)* %out) #1 {
   %kernarg.segment.ptr = call noalias i8 addrspace(2)* @llvm.amdgcn.kernarg.segment.ptr()
   %header.ptr = bitcast i8 addrspace(2)* %kernarg.segment.ptr to i32 addrspace(2)*
@@ -17,7 +18,7 @@ define void @test(i32 addrspace(1)* %out) #1 {
 
 ; ALL-LABEL: {{^}}test_implicit:
 ; 10 + 9 (36 prepended implicit bytes) + 2(out pointer) = 21 = 0x15
-; MESA: s_load_dword s{{[0-9]+}}, s[0:1], 0x15
+; OS-UNKNOWN: s_load_dword s{{[0-9]+}}, s[0:1], 0x15
 define void @test_implicit(i32 addrspace(1)* %out) #1 {
   %implicitarg.ptr = call noalias i8 addrspace(2)* @llvm.amdgcn.implicitarg.ptr()
   %header.ptr = bitcast i8 addrspace(2)* %implicitarg.ptr to i32 addrspace(2)*
@@ -28,8 +29,9 @@ define void @test_implicit(i32 addrspace(1)* %out) #1 {
 }
 
 ; ALL-LABEL: {{^}}test_implicit_alignment
-; MESA: s_load_dword [[VAL:s[0-9]+]], s[{{[0-9]+:[0-9]+}}], 0xc
+; OS-UNKNOWN: s_load_dword [[VAL:s[0-9]+]], s[{{[0-9]+:[0-9]+}}], 0xc
 ; HSA: s_load_dword [[VAL:s[0-9]+]], s[{{[0-9]+:[0-9]+}}], 0x4
+; OS-MESA3D: s_load_dword [[VAL:s[0-9]+]], s[{{[0-9]+:[0-9]+}}], 0x3
 ; ALL: v_mov_b32_e32 [[V_VAL:v[0-9]+]], [[VAL]]
 ; MESA: buffer_store_dword [[V_VAL]]
 ; HSA: flat_store_dword v[{{[0-9]+:[0-9]+}}], [[V_VAL]]
diff --git a/llvm/test/CodeGen/AMDGPU/llvm.amdgcn.workgroup.id.ll b/llvm/test/CodeGen/AMDGPU/llvm.amdgcn.workgroup.id.ll
index 75a9ec91eca..58529b87444 100644
--- a/llvm/test/CodeGen/AMDGPU/llvm.amdgcn.workgroup.id.ll
+++ b/llvm/test/CodeGen/AMDGPU/llvm.amdgcn.workgroup.id.ll
@@ -1,7 +1,9 @@
-; RUN: llc -march=amdgcn -mtriple=amdgcn-unknown-amdhsa -mcpu=kaveri -verify-machineinstrs < %s | FileCheck -check-prefix=ALL -check-prefix=HSA -check-prefix=CI-HSA  %s
-; RUN: llc -march=amdgcn -mtriple=amdgcn-unknown-amdhsa -mcpu=carrizo -verify-machineinstrs < %s | FileCheck -check-prefix=ALL -check-prefix=HSA -check-prefix=VI-HSA  %s
-; RUN: llc -march=amdgcn -mcpu=tahiti -verify-machineinstrs < %s | FileCheck -check-prefix=ALL -check-prefix=MESA -check-prefix=SI-MESA %s
-; RUN: llc -march=amdgcn -mcpu=tonga -verify-machineinstrs < %s | FileCheck -check-prefix=ALL -check-prefix=MESA -check-prefix=VI-MESA %s
+; RUN: llc -march=amdgcn -mtriple=amdgcn-unknown-amdhsa -mcpu=kaveri -verify-machineinstrs < %s | FileCheck -check-prefix=ALL -check-prefix=CO-V2 -check-prefix=CI-HSA  %s
+; RUN: llc -march=amdgcn -mtriple=amdgcn-unknown-amdhsa -mcpu=carrizo -verify-machineinstrs < %s | FileCheck -check-prefix=ALL -check-prefix=CO-V2 -check-prefix=VI-HSA  %s
+; RUN: llc -march=amdgcn -mcpu=tahiti -verify-machineinstrs < %s | FileCheck -check-prefix=ALL -check-prefix=UNKNOWN-OS -check-prefix=SI-MESA %s
+; RUN: llc -march=amdgcn -mcpu=tonga -verify-machineinstrs < %s | FileCheck -check-prefix=ALL -check-prefix=UNKNOWN-OS -check-prefix=VI-MESA %s
+; RUN: llc -mtriple=amdgcn-unknown-mesa3d -mcpu=tahiti -verify-machineinstrs < %s | FileCheck -check-prefixes=ALL,CO-V2,SI-MESA %s
+; RUN: llc -mtriple=amdgcn-unknown-mesa3d -mcpu=tonga -verify-machineinstrs < %s | FileCheck -check-prefixes=ALL,CO-V2,VI-MESA %s
 
 declare i32 @llvm.amdgcn.workgroup.id.x() #0
 declare i32 @llvm.amdgcn.workgroup.id.y() #0
@@ -9,26 +11,25 @@ declare i32 @llvm.amdgcn.workgroup.id.z() #0
 
 ; ALL-LABEL {{^}}test_workgroup_id_x:
 
-; HSA: .amd_kernel_code_t
-; HSA: user_sgpr_count = 6
-; HSA: enable_sgpr_workgroup_id_x = 1
-; HSA: enable_sgpr_workgroup_id_y = 0
-; HSA: enable_sgpr_workgroup_id_z = 0
-; HSA: enable_sgpr_workgroup_info = 0
-; HSA: enable_vgpr_workitem_id = 0
-; HSA: enable_sgpr_grid_workgroup_count_x = 0
-; HSA: enable_sgpr_grid_workgroup_count_y = 0
-; HSA: enable_sgpr_grid_workgroup_count_z = 0
-; HSA: .end_amd_kernel_code_t
+; CO-V2: .amd_kernel_code_t
+; CO-V2: user_sgpr_count = 6
+; CO-V2: enable_sgpr_workgroup_id_x = 1
+; CO-V2: enable_sgpr_workgroup_id_y = 0
+; CO-V2: enable_sgpr_workgroup_id_z = 0
+; CO-V2: enable_sgpr_workgroup_info = 0
+; CO-V2: enable_vgpr_workitem_id = 0
+; CO-V2: enable_sgpr_grid_workgroup_count_x = 0
+; CO-V2: enable_sgpr_grid_workgroup_count_y = 0
+; CO-V2: enable_sgpr_grid_workgroup_count_z = 0
+; CO-V2: .end_amd_kernel_code_t
 
-; MESA: v_mov_b32_e32 [[VCOPY:v[0-9]+]], s2{{$}}
-; HSA: v_mov_b32_e32 [[VCOPY:v[0-9]+]], s6{{$}}
+; UNKNOWN-OS: v_mov_b32_e32 [[VCOPY:v[0-9]+]], s2{{$}}
+; CO-V2: v_mov_b32_e32 [[VCOPY:v[0-9]+]], s6{{$}}
 
-; ALL-NOT: [[VCOPY]]
 ; ALL: {{buffer|flat}}_store_dword {{.*}}[[VCOPY]]
 
-; HSA: COMPUTE_PGM_RSRC2:USER_SGPR: 6
-; ALL-NOHSA: COMPUTE_PGM_RSRC2:USER_SGPR: 2
+; CO-V2: COMPUTE_PGM_RSRC2:USER_SGPR: 6
+; ALL-NOCO-V2: COMPUTE_PGM_RSRC2:USER_SGPR: 2
 ; ALL: COMPUTE_PGM_RSRC2:TGID_X_EN: 1
 ; ALL: COMPUTE_PGM_RSRC2:TGID_Y_EN: 0
 ; ALL: COMPUTE_PGM_RSRC2:TGID_Z_EN: 0
@@ -40,23 +41,22 @@ define void @test_workgroup_id_x(i32 addrspace(1)* %out) #1 {
 }
 
 ; ALL-LABEL {{^}}test_workgroup_id_y:
-; HSA: user_sgpr_count = 6
-; HSA: enable_sgpr_workgroup_id_x = 1
-; HSA: enable_sgpr_workgroup_id_y = 1
-; HSA: enable_sgpr_workgroup_id_z = 0
-; HSA: enable_sgpr_workgroup_info = 0
-; HSA: enable_sgpr_grid_workgroup_count_x = 0
-; HSA: enable_sgpr_grid_workgroup_count_y = 0
-; HSA: enable_sgpr_grid_workgroup_count_z = 0
+; CO-V2: user_sgpr_count = 6
+; CO-V2: enable_sgpr_workgroup_id_x = 1
+; CO-V2: enable_sgpr_workgroup_id_y = 1
+; CO-V2: enable_sgpr_workgroup_id_z = 0
+; CO-V2: enable_sgpr_workgroup_info = 0
+; CO-V2: enable_sgpr_grid_workgroup_count_x = 0
+; CO-V2: enable_sgpr_grid_workgroup_count_y = 0
+; CO-V2: enable_sgpr_grid_workgroup_count_z = 0
 
-; MESA: v_mov_b32_e32 [[VCOPY:v[0-9]+]], s3{{$}}
+; UNKNOWN-OS: v_mov_b32_e32 [[VCOPY:v[0-9]+]], s3{{$}}
 ; HSA: v_mov_b32_e32 [[VCOPY:v[0-9]+]], s7{{$}}
 
-; ALL-NOT: [[VCOPY]]
 ; ALL: {{buffer|flat}}_store_dword {{.*}}[[VCOPY]]
 
-; HSA: COMPUTE_PGM_RSRC2:USER_SGPR: 6
-; ALL-NOHSA: COMPUTE_PGM_RSRC2:USER_SGPR: 2
+; CO-V2: COMPUTE_PGM_RSRC2:USER_SGPR: 6
+; ALL-NOCO-V2: COMPUTE_PGM_RSRC2:USER_SGPR: 2
 ; ALL: COMPUTE_PGM_RSRC2:TGID_X_EN: 1
 ; ALL: COMPUTE_PGM_RSRC2:TGID_Y_EN: 1
 ; ALL: COMPUTE_PGM_RSRC2:TGID_Z_EN: 0
@@ -68,31 +68,30 @@ define void @test_workgroup_id_y(i32 addrspace(1)* %out) #1 {
 }
 
 ; ALL-LABEL {{^}}test_workgroup_id_z:
-; HSA: user_sgpr_count = 6
-; HSA: enable_sgpr_workgroup_id_x = 1
-; HSA: enable_sgpr_workgroup_id_y = 0
-; HSA: enable_sgpr_workgroup_id_z = 1
-; HSA: enable_sgpr_workgroup_info = 0
-; HSA: enable_vgpr_workitem_id = 0
-; HSA: enable_sgpr_private_segment_buffer = 1
-; HSA: enable_sgpr_dispatch_ptr = 0
-; HSA: enable_sgpr_queue_ptr = 0
-; HSA: enable_sgpr_kernarg_segment_ptr = 1
-; HSA: enable_sgpr_dispatch_id = 0
-; HSA: enable_sgpr_flat_scratch_init = 0
-; HSA: enable_sgpr_private_segment_size = 0
-; HSA: enable_sgpr_grid_workgroup_count_x = 0
-; HSA: enable_sgpr_grid_workgroup_count_y = 0
-; HSA: enable_sgpr_grid_workgroup_count_z = 0
+; CO-V2: user_sgpr_count = 6
+; CO-V2: enable_sgpr_workgroup_id_x = 1
+; CO-V2: enable_sgpr_workgroup_id_y = 0
+; CO-V2: enable_sgpr_workgroup_id_z = 1
+; CO-V2: enable_sgpr_workgroup_info = 0
+; CO-V2: enable_vgpr_workitem_id = 0
+; CO-V2: enable_sgpr_private_segment_buffer = 1
+; CO-V2: enable_sgpr_dispatch_ptr = 0
+; CO-V2: enable_sgpr_queue_ptr = 0
+; CO-V2: enable_sgpr_kernarg_segment_ptr = 1
+; CO-V2: enable_sgpr_dispatch_id = 0
+; CO-V2: enable_sgpr_flat_scratch_init = 0
+; CO-V2: enable_sgpr_private_segment_size = 0
+; CO-V2: enable_sgpr_grid_workgroup_count_x = 0
+; CO-V2: enable_sgpr_grid_workgroup_count_y = 0
+; CO-V2: enable_sgpr_grid_workgroup_count_z = 0
 
-; MESA: v_mov_b32_e32 [[VCOPY:v[0-9]+]], s3{{$}}
+; UNKNOWN-OS: v_mov_b32_e32 [[VCOPY:v[0-9]+]], s3{{$}}
 ; HSA: v_mov_b32_e32 [[VCOPY:v[0-9]+]], s7{{$}}
 
-; ALL-NOT: [[VCOPY]]
 ; ALL: {{buffer|flat}}_store_dword {{.*}}[[VCOPY]]
 
-; HSA: COMPUTE_PGM_RSRC2:USER_SGPR: 6
-; ALL-NOHSA: COMPUTE_PGM_RSRC2:USER_SGPR: 2
+; CO-V2: COMPUTE_PGM_RSRC2:USER_SGPR: 6
+; ALL-NOCO-V2: COMPUTE_PGM_RSRC2:USER_SGPR: 2
 ; ALL: COMPUTE_PGM_RSRC2:TGID_X_EN: 1
 ; ALL: COMPUTE_PGM_RSRC2:TGID_Y_EN: 0
 ; ALL: COMPUTE_PGM_RSRC2:TGID_Z_EN: 1
diff --git a/llvm/test/CodeGen/AMDGPU/llvm.amdgcn.workitem.id.ll b/llvm/test/CodeGen/AMDGPU/llvm.amdgcn.workitem.id.ll
index 393a593fad0..8fc12890fad 100644
--- a/llvm/test/CodeGen/AMDGPU/llvm.amdgcn.workitem.id.ll
+++ b/llvm/test/CodeGen/AMDGPU/llvm.amdgcn.workitem.id.ll
@@ -1,7 +1,9 @@
-; RUN: llc -march=amdgcn -mtriple=amdgcn-unknown-amdhsa -mcpu=kaveri -verify-machineinstrs < %s | FileCheck -check-prefix=ALL -check-prefix=HSA -check-prefix=CI-HSA  %s
-; RUN: llc -march=amdgcn -mtriple=amdgcn-unknown-amdhsa -mcpu=carrizo -verify-machineinstrs < %s | FileCheck -check-prefix=ALL -check-prefix=HSA -check-prefix=VI-HSA  %s
+; RUN: llc -march=amdgcn -mtriple=amdgcn-unknown-amdhsa -mcpu=kaveri -verify-machineinstrs < %s | FileCheck -check-prefix=ALL -check-prefix=CO-V2 -check-prefix=CI-HSA  %s
+; RUN: llc -march=amdgcn -mtriple=amdgcn-unknown-amdhsa -mcpu=carrizo -verify-machineinstrs < %s | FileCheck -check-prefix=ALL -check-prefix=CO-V2 -check-prefix=VI-HSA  %s
 ; RUN: llc -march=amdgcn -mcpu=tahiti -verify-machineinstrs < %s | FileCheck -check-prefix=ALL -check-prefix=MESA -check-prefix=SI-MESA %s
 ; RUN: llc -march=amdgcn -mcpu=tonga -verify-machineinstrs < %s | FileCheck -check-prefix=ALL -check-prefix=MESA -check-prefix=VI-MESA %s
+; RUN: llc -mtriple=amdgcn-unknown-mesa3d -mcpu=tahiti -verify-machineinstrs < %s | FileCheck -check-prefixes=ALL,CO-V2,SI-MESA %s
+; RUN: llc -mtriple=amdgcn-unknown-mesa3d -mcpu=tonga -verify-machineinstrs < %s | FileCheck -check-prefixes=ALL,CO-V2,VI-MESA %s
 
 declare i32 @llvm.amdgcn.workitem.id.x() #0
 declare i32 @llvm.amdgcn.workitem.id.y() #0
@@ -12,7 +14,7 @@ declare i32 @llvm.amdgcn.workitem.id.z() #0
 ; MESA-NEXT: .long 132{{$}}
 
 ; ALL-LABEL {{^}}test_workitem_id_x:
-; HSA: enable_vgpr_workitem_id = 0
+; CO-V2: enable_vgpr_workitem_id = 0
 
 ; ALL-NOT: v0
 ; ALL: {{buffer|flat}}_store_dword {{.*}}v0
@@ -27,7 +29,7 @@ define void @test_workitem_id_x(i32 addrspace(1)* %out) #1 {
 ; MESA-NEXT: .long 2180{{$}}
 
 ; ALL-LABEL {{^}}test_workitem_id_y:
-; HSA: enable_vgpr_workitem_id = 1
+; CO-V2: enable_vgpr_workitem_id = 1
 
 ; ALL-NOT: v1
 ; ALL: {{buffer|flat}}_store_dword {{.*}}v1
@@ -42,7 +44,7 @@ define void @test_workitem_id_y(i32 addrspace(1)* %out) #1 {
 ; MESA-NEXT: .long 4228{{$}}
 
 ; ALL-LABEL {{^}}test_workitem_id_z:
-; HSA: enable_vgpr_workitem_id = 2
+; CO-V2: enable_vgpr_workitem_id = 2
 
 ; ALL-NOT: v2
 ; ALL: {{buffer|flat}}_store_dword {{.*}}v2