Backport 45d2d7757feb386186f69af6ef57bde7b5adc2db to release/18.x #79839
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This is only valid on targets with architected SGPRs.
|
@llvm/pr-subscribers-llvm-ir @llvm/pr-subscribers-backend-amdgpu Author: Jay Foad (jayfoad) ChangesThis just missed the branch creation and is the last piece of functionality required to get AMDGPU GFX12 support working in the 18.x release. Full diff: https://github.com/llvm/llvm-project/pull/79839.diff 6 Files Affected:
diff --git a/llvm/include/llvm/IR/IntrinsicsAMDGPU.td b/llvm/include/llvm/IR/IntrinsicsAMDGPU.td
index 9eb1ac8e27befb1..c5f43d17d1c1481 100644
--- a/llvm/include/llvm/IR/IntrinsicsAMDGPU.td
+++ b/llvm/include/llvm/IR/IntrinsicsAMDGPU.td
@@ -2777,6 +2777,10 @@ class AMDGPULoadTr<LLVMType ptr_ty>:
def int_amdgcn_global_load_tr : AMDGPULoadTr<global_ptr_ty>;
+// i32 @llvm.amdgcn.wave.id()
+def int_amdgcn_wave_id :
+ DefaultAttrsIntrinsic<[llvm_i32_ty], [], [IntrNoMem, IntrSpeculatable]>;
+
//===----------------------------------------------------------------------===//
// Deep learning intrinsics.
//===----------------------------------------------------------------------===//
diff --git a/llvm/lib/Target/AMDGPU/AMDGPULegalizerInfo.cpp b/llvm/lib/Target/AMDGPU/AMDGPULegalizerInfo.cpp
index 615685822f91eeb..e98ede88a7e2db9 100644
--- a/llvm/lib/Target/AMDGPU/AMDGPULegalizerInfo.cpp
+++ b/llvm/lib/Target/AMDGPU/AMDGPULegalizerInfo.cpp
@@ -6883,6 +6883,23 @@ bool AMDGPULegalizerInfo::legalizeStackSave(MachineInstr &MI,
return true;
}
+bool AMDGPULegalizerInfo::legalizeWaveID(MachineInstr &MI,
+ MachineIRBuilder &B) const {
+ // With architected SGPRs, waveIDinGroup is in TTMP8[29:25].
+ if (!ST.hasArchitectedSGPRs())
+ return false;
+ LLT S32 = LLT::scalar(32);
+ Register DstReg = MI.getOperand(0).getReg();
+ Register TTMP8 =
+ getFunctionLiveInPhysReg(B.getMF(), B.getTII(), AMDGPU::TTMP8,
+ AMDGPU::SReg_32RegClass, B.getDebugLoc(), S32);
+ auto LSB = B.buildConstant(S32, 25);
+ auto Width = B.buildConstant(S32, 5);
+ B.buildUbfx(DstReg, TTMP8, LSB, Width);
+ MI.eraseFromParent();
+ return true;
+}
+
bool AMDGPULegalizerInfo::legalizeIntrinsic(LegalizerHelper &Helper,
MachineInstr &MI) const {
MachineIRBuilder &B = Helper.MIRBuilder;
@@ -7005,6 +7022,8 @@ bool AMDGPULegalizerInfo::legalizeIntrinsic(LegalizerHelper &Helper,
case Intrinsic::amdgcn_workgroup_id_z:
return legalizePreloadedArgIntrin(MI, MRI, B,
AMDGPUFunctionArgInfo::WORKGROUP_ID_Z);
+ case Intrinsic::amdgcn_wave_id:
+ return legalizeWaveID(MI, B);
case Intrinsic::amdgcn_lds_kernel_id:
return legalizePreloadedArgIntrin(MI, MRI, B,
AMDGPUFunctionArgInfo::LDS_KERNEL_ID);
diff --git a/llvm/lib/Target/AMDGPU/AMDGPULegalizerInfo.h b/llvm/lib/Target/AMDGPU/AMDGPULegalizerInfo.h
index 56aabd4f6ab71b6..ecbe42681c6690c 100644
--- a/llvm/lib/Target/AMDGPU/AMDGPULegalizerInfo.h
+++ b/llvm/lib/Target/AMDGPU/AMDGPULegalizerInfo.h
@@ -212,6 +212,7 @@ class AMDGPULegalizerInfo final : public LegalizerInfo {
bool legalizeFPTruncRound(MachineInstr &MI, MachineIRBuilder &B) const;
bool legalizeStackSave(MachineInstr &MI, MachineIRBuilder &B) const;
+ bool legalizeWaveID(MachineInstr &MI, MachineIRBuilder &B) const;
bool legalizeImageIntrinsic(
MachineInstr &MI, MachineIRBuilder &B,
diff --git a/llvm/lib/Target/AMDGPU/SIISelLowering.cpp b/llvm/lib/Target/AMDGPU/SIISelLowering.cpp
index d60f511302613e1..c5ad9da88ec2b31 100644
--- a/llvm/lib/Target/AMDGPU/SIISelLowering.cpp
+++ b/llvm/lib/Target/AMDGPU/SIISelLowering.cpp
@@ -7920,6 +7920,18 @@ SDValue SITargetLowering::lowerSBuffer(EVT VT, SDLoc DL, SDValue Rsrc,
return Loads[0];
}
+SDValue SITargetLowering::lowerWaveID(SelectionDAG &DAG, SDValue Op) const {
+ // With architected SGPRs, waveIDinGroup is in TTMP8[29:25].
+ if (!Subtarget->hasArchitectedSGPRs())
+ return {};
+ SDLoc SL(Op);
+ MVT VT = MVT::i32;
+ SDValue TTMP8 = CreateLiveInRegister(DAG, &AMDGPU::SReg_32RegClass,
+ AMDGPU::TTMP8, VT, SL);
+ return DAG.getNode(AMDGPUISD::BFE_U32, SL, VT, TTMP8,
+ DAG.getConstant(25, SL, VT), DAG.getConstant(5, SL, VT));
+}
+
SDValue SITargetLowering::lowerWorkitemID(SelectionDAG &DAG, SDValue Op,
unsigned Dim,
const ArgDescriptor &Arg) const {
@@ -8090,6 +8102,8 @@ SDValue SITargetLowering::LowerINTRINSIC_WO_CHAIN(SDValue Op,
case Intrinsic::amdgcn_workgroup_id_z:
return getPreloadedValue(DAG, *MFI, VT,
AMDGPUFunctionArgInfo::WORKGROUP_ID_Z);
+ case Intrinsic::amdgcn_wave_id:
+ return lowerWaveID(DAG, Op);
case Intrinsic::amdgcn_lds_kernel_id: {
if (MFI->isEntryFunction())
return getLDSKernelId(DAG, DL);
diff --git a/llvm/lib/Target/AMDGPU/SIISelLowering.h b/llvm/lib/Target/AMDGPU/SIISelLowering.h
index d66ba0b59ba906b..e436c23af5bcac0 100644
--- a/llvm/lib/Target/AMDGPU/SIISelLowering.h
+++ b/llvm/lib/Target/AMDGPU/SIISelLowering.h
@@ -80,6 +80,7 @@ class SITargetLowering final : public AMDGPUTargetLowering {
SDValue lowerStructBufferAtomicIntrin(SDValue Op, SelectionDAG &DAG,
unsigned NewOpcode) const;
+ SDValue lowerWaveID(SelectionDAG &DAG, SDValue Op) const;
SDValue lowerWorkitemID(SelectionDAG &DAG, SDValue Op, unsigned Dim,
const ArgDescriptor &ArgDesc) const;
diff --git a/llvm/test/CodeGen/AMDGPU/llvm.amdgcn.wave.id.ll b/llvm/test/CodeGen/AMDGPU/llvm.amdgcn.wave.id.ll
new file mode 100644
index 000000000000000..df5533b6295023b
--- /dev/null
+++ b/llvm/test/CodeGen/AMDGPU/llvm.amdgcn.wave.id.ll
@@ -0,0 +1,61 @@
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 4
+; RUN: not --crash llc -mtriple=amdgcn-amd-amdpal -mcpu=gfx900 -global-isel=0 -verify-machineinstrs < %s 2>&1 | FileCheck -check-prefix=GFX9-SDAG-ERR %s
+; RUN: not --crash llc -mtriple=amdgcn-amd-amdpal -mcpu=gfx900 -global-isel=1 -verify-machineinstrs < %s 2>&1 | FileCheck -check-prefix=GFX9-GISEL-ERR %s
+; RUN: llc -mtriple=amdgcn-amd-amdpal -mcpu=gfx900 -mattr=+architected-sgprs -global-isel=0 -verify-machineinstrs < %s | FileCheck -check-prefix=GFX9 %s
+; RUN: llc -mtriple=amdgcn-amd-amdpal -mcpu=gfx900 -mattr=+architected-sgprs -global-isel=1 -verify-machineinstrs < %s | FileCheck -check-prefix=GFX9 %s
+; RUN: llc -mtriple=amdgcn-amd-amdpal -mcpu=gfx1200 -global-isel=0 -verify-machineinstrs < %s | FileCheck -check-prefix=GFX12 %s
+; RUN: llc -mtriple=amdgcn-amd-amdpal -mcpu=gfx1200 -global-isel=1 -verify-machineinstrs < %s | FileCheck -check-prefix=GFX12 %s
+
+; GFX9-SDAG-ERR: LLVM ERROR: Cannot select: intrinsic %llvm.amdgcn.wave.id
+; GFX9-GISEL-ERR: LLVM ERROR: unable to legalize instruction: {{.*}} = G_INTRINSIC intrinsic(@llvm.amdgcn.wave.id)
+
+define amdgpu_cs void @test_wave_id(ptr addrspace(1) %out) {
+; GFX9-LABEL: test_wave_id:
+; GFX9: ; %bb.0:
+; GFX9-NEXT: s_bfe_u32 s0, ttmp8, 0x50019
+; GFX9-NEXT: v_mov_b32_e32 v2, s0
+; GFX9-NEXT: global_store_dword v[0:1], v2, off
+; GFX9-NEXT: s_endpgm
+;
+; GFX12-LABEL: test_wave_id:
+; GFX12: ; %bb.0:
+; GFX12-NEXT: s_bfe_u32 s0, ttmp8, 0x50019
+; GFX12-NEXT: s_delay_alu instid0(SALU_CYCLE_1)
+; GFX12-NEXT: v_mov_b32_e32 v2, s0
+; GFX12-NEXT: global_store_b32 v[0:1], v2, off
+; GFX12-NEXT: s_nop 0
+; GFX12-NEXT: s_sendmsg sendmsg(MSG_DEALLOC_VGPRS)
+; GFX12-NEXT: s_endpgm
+ %waveid = call i32 @llvm.amdgcn.wave.id()
+ store i32 %waveid, ptr addrspace(1) %out
+ ret void
+}
+
+define amdgpu_gfx void @test_wave_id_callable(ptr addrspace(1) %out) {
+; GFX9-LABEL: test_wave_id_callable:
+; GFX9: ; %bb.0:
+; GFX9-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
+; GFX9-NEXT: s_bfe_u32 s34, ttmp8, 0x50019
+; GFX9-NEXT: v_mov_b32_e32 v2, s34
+; GFX9-NEXT: global_store_dword v[0:1], v2, off
+; GFX9-NEXT: s_waitcnt vmcnt(0)
+; GFX9-NEXT: s_setpc_b64 s[30:31]
+;
+; GFX12-LABEL: test_wave_id_callable:
+; GFX12: ; %bb.0:
+; GFX12-NEXT: s_wait_loadcnt_dscnt 0x0
+; GFX12-NEXT: s_wait_expcnt 0x0
+; GFX12-NEXT: s_wait_samplecnt 0x0
+; GFX12-NEXT: s_wait_bvhcnt 0x0
+; GFX12-NEXT: s_wait_kmcnt 0x0
+; GFX12-NEXT: s_bfe_u32 s0, ttmp8, 0x50019
+; GFX12-NEXT: s_delay_alu instid0(SALU_CYCLE_1)
+; GFX12-NEXT: v_mov_b32_e32 v2, s0
+; GFX12-NEXT: global_store_b32 v[0:1], v2, off
+; GFX12-NEXT: s_setpc_b64 s[30:31]
+ %waveid = call i32 @llvm.amdgcn.wave.id()
+ store i32 %waveid, ptr addrspace(1) %out
+ ret void
+}
+
+declare i32 @llvm.amdgcn.wave.id()
|
arsenm
approved these changes
Jan 29, 2024
| LLT S32 = LLT::scalar(32); | ||
| Register DstReg = MI.getOperand(0).getReg(); | ||
| Register TTMP8 = | ||
| getFunctionLiveInPhysReg(B.getMF(), B.getTII(), AMDGPU::TTMP8, |
There was a problem hiding this comment.
This avoids the live in in the later patch?
There was a problem hiding this comment.
True, 66c710e improved this to avoid adding liveins. I wasn't going to bother backporting that since I didn't think it was required for correctness. But I have cherry-picked it into this PR now.
Tweak the implementation of llvm.amdgcn.wave.id to not add TTMP8 to the function liveins.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This just missed the branch creation and is the last piece of functionality required to get AMDGPU GFX12 support working in the 18.x release.