Skip to content

Commit 96d7a93

Browse files
jmmartinezjmmartinez
committed
[Review] mem-to-lds -> vmem-to-lds
1 parent ee2123f commit 96d7a93

File tree

10 files changed

+21
-21
lines changed

10 files changed

+21
-21
lines changed

‎clang/include/clang/Basic/BuiltinsAMDGPU.def

+1-1
Original file line numberDiff line numberDiff line change
@@ -254,7 +254,7 @@ TARGET_BUILTIN(__builtin_amdgcn_flat_atomic_fadd_v2bf16, "V2sV2s*0V2s", "t", "at
254254
TARGET_BUILTIN(__builtin_amdgcn_global_atomic_fadd_v2bf16, "V2sV2s*1V2s", "t", "atomic-global-pk-add-bf16-inst")
255255
TARGET_BUILTIN(__builtin_amdgcn_ds_atomic_fadd_v2bf16, "V2sV2s*3V2s", "t", "atomic-ds-pk-add-16-insts")
256256
TARGET_BUILTIN(__builtin_amdgcn_ds_atomic_fadd_v2f16, "V2hV2h*3V2h", "t", "atomic-ds-pk-add-16-insts")
257-
TARGET_BUILTIN(__builtin_amdgcn_global_load_lds, "vv*1v*3IUiIiIUi", "t", "mem-to-lds-load-insts")
257+
TARGET_BUILTIN(__builtin_amdgcn_global_load_lds, "vv*1v*3IUiIiIUi", "t", "vmem-to-lds-load-insts")
258258

259259
//===----------------------------------------------------------------------===//
260260
// Deep learning builtins.

‎clang/lib/Basic/Targets/AMDGPU.cpp

+1-1
Original file line numberDiff line numberDiff line change
@@ -260,7 +260,7 @@ AMDGPUTargetInfo::AMDGPUTargetInfo(const llvm::Triple &Triple,
260260

261261
MaxAtomicPromoteWidth = MaxAtomicInlineWidth = 64;
262262
CUMode = !(GPUFeatures & llvm::AMDGPU::FEATURE_WGP);
263-
for (auto F : {"image-insts", "gws", "mem-to-lds-load-insts"})
263+
for (auto F : {"image-insts", "gws", "vmem-to-lds-load-insts"})
264264
ReadOnlyFeatures.insert(F);
265265
HalfArgsAndReturns = true;
266266
}

‎clang/test/CodeGen/link-builtin-bitcode.c

+3-3
Original file line numberDiff line numberDiff line change
@@ -44,6 +44,6 @@ int bar() { return no_attr() + attr_in_target() + attr_not_in_target() + attr_in
4444
// CHECK-SAME: () #[[ATTR_INCOMPATIBLE:[0-9]+]] {
4545

4646
// CHECK: attributes #[[ATTR_BAR]] = { {{.*}} "target-cpu"="gfx90a" "target-features"="+16-bit-insts,+atomic-buffer-global-pk-add-f16-insts,+atomic-fadd-rtn-insts,+ci-insts,+dl-insts,+dot1-insts,+dot10-insts,+dot2-insts,+dot3-insts,+dot4-insts,+dot5-insts,+dot6-insts,+dot7-insts,+dpp,+gfx8-insts,+gfx9-insts,+gfx90a-insts,+mai-insts,+s-memrealtime,+s-memtime-inst,+wavefrontsize64" }
47-
// CHECK: attributes #[[ATTR_COMPATIBLE]] = { {{.*}} "target-cpu"="gfx90a" "target-features"="+16-bit-insts,+atomic-buffer-global-pk-add-f16-insts,+atomic-fadd-rtn-insts,+ci-insts,+dl-insts,+dot1-insts,+dot10-insts,+dot2-insts,+dot3-insts,+dot4-insts,+dot5-insts,+dot6-insts,+dot7-insts,+dpp,+gfx8-insts,+gfx9-insts,+gfx90a-insts,+gws,+image-insts,+mai-insts,+mem-to-lds-load-insts,+s-memrealtime,+s-memtime-inst,+wavefrontsize64" }
48-
// CHECK: attributes #[[ATTR_EXTEND]] = { {{.*}} "target-cpu"="gfx90a" "target-features"="+16-bit-insts,+atomic-buffer-global-pk-add-f16-insts,+atomic-fadd-rtn-insts,+ci-insts,+dl-insts,+dot1-insts,+dot10-insts,+dot2-insts,+dot3-insts,+dot4-insts,+dot5-insts,+dot6-insts,+dot7-insts,+dpp,+extended-image-insts,+gfx8-insts,+gfx9-insts,+gfx90a-insts,+gws,+image-insts,+mai-insts,+mem-to-lds-load-insts,+s-memrealtime,+s-memtime-inst,+wavefrontsize64" }
49-
// CHECK: attributes #[[ATTR_INCOMPATIBLE]] = { {{.*}} "target-cpu"="gfx90a" "target-features"="+16-bit-insts,+atomic-buffer-global-pk-add-f16-insts,+atomic-fadd-rtn-insts,+ci-insts,+dl-insts,+dot1-insts,+dot10-insts,+dot2-insts,+dot3-insts,+dot4-insts,+dot5-insts,+dot6-insts,+dot7-insts,+dpp,+gfx8-insts,+gfx90a-insts,+gws,+image-insts,+mai-insts,+mem-to-lds-load-insts,+s-memrealtime,+s-memtime-inst,+wavefrontsize64,-gfx9-insts" }
47+
// CHECK: attributes #[[ATTR_COMPATIBLE]] = { {{.*}} "target-cpu"="gfx90a" "target-features"="+16-bit-insts,+atomic-buffer-global-pk-add-f16-insts,+atomic-fadd-rtn-insts,+ci-insts,+dl-insts,+dot1-insts,+dot10-insts,+dot2-insts,+dot3-insts,+dot4-insts,+dot5-insts,+dot6-insts,+dot7-insts,+dpp,+gfx8-insts,+gfx9-insts,+gfx90a-insts,+gws,+image-insts,+mai-insts,+s-memrealtime,+s-memtime-inst,+vmem-to-lds-load-insts,+wavefrontsize64" }
48+
// CHECK: attributes #[[ATTR_EXTEND]] = { {{.*}} "target-cpu"="gfx90a" "target-features"="+16-bit-insts,+atomic-buffer-global-pk-add-f16-insts,+atomic-fadd-rtn-insts,+ci-insts,+dl-insts,+dot1-insts,+dot10-insts,+dot2-insts,+dot3-insts,+dot4-insts,+dot5-insts,+dot6-insts,+dot7-insts,+dpp,+extended-image-insts,+gfx8-insts,+gfx9-insts,+gfx90a-insts,+gws,+image-insts,+mai-insts,+s-memrealtime,+s-memtime-inst,+vmem-to-lds-load-insts,+wavefrontsize64" }
49+
// CHECK: attributes #[[ATTR_INCOMPATIBLE]] = { {{.*}} "target-cpu"="gfx90a" "target-features"="+16-bit-insts,+atomic-buffer-global-pk-add-f16-insts,+atomic-fadd-rtn-insts,+ci-insts,+dl-insts,+dot1-insts,+dot10-insts,+dot2-insts,+dot3-insts,+dot4-insts,+dot5-insts,+dot6-insts,+dot7-insts,+dpp,+gfx8-insts,+gfx90a-insts,+gws,+image-insts,+mai-insts,+s-memrealtime,+s-memtime-inst,+vmem-to-lds-load-insts,+wavefrontsize64,-gfx9-insts" }

‎clang/test/CodeGenCXX/dynamic-cast-address-space.cpp

+2-2
Original file line numberDiff line numberDiff line change
@@ -111,9 +111,9 @@ const B& f(A *a) {
111111
// CHECK: attributes #[[ATTR3]] = { nounwind }
112112
// CHECK: attributes #[[ATTR4]] = { noreturn }
113113
//.
114-
// WITH-NONZERO-DEFAULT-AS: attributes #[[ATTR0]] = { mustprogress noinline optnone "no-trapping-math"="true" "stack-protector-buffer-size"="8" "target-features"="+16-bit-insts,+ashr-pk-insts,+atomic-buffer-global-pk-add-f16-insts,+atomic-buffer-pk-add-bf16-inst,+atomic-ds-pk-add-16-insts,+atomic-fadd-rtn-insts,+atomic-flat-pk-add-16-insts,+atomic-global-pk-add-bf16-inst,+bf8-cvt-scale-insts,+bitop3-insts,+ci-insts,+dl-insts,+dot1-insts,+dot10-insts,+dot11-insts,+dot12-insts,+dot13-insts,+dot2-insts,+dot3-insts,+dot4-insts,+dot5-insts,+dot6-insts,+dot7-insts,+dot8-insts,+dot9-insts,+dpp,+f16bf16-to-fp6bf6-cvt-scale-insts,+f32-to-f16bf16-cvt-sr-insts,+fp4-cvt-scale-insts,+fp6bf6-cvt-scale-insts,+fp8-conversion-insts,+fp8-cvt-scale-insts,+fp8-insts,+gfx10-3-insts,+gfx10-insts,+gfx11-insts,+gfx12-insts,+gfx8-insts,+gfx9-insts,+gfx90a-insts,+gfx940-insts,+gfx950-insts,+gws,+image-insts,+mai-insts,+mem-to-lds-load-insts,+permlane16-swap,+permlane32-swap,+prng-inst,+s-memrealtime,+s-memtime-inst,+wavefrontsize32,+wavefrontsize64" }
114+
// WITH-NONZERO-DEFAULT-AS: attributes #[[ATTR0]] = { mustprogress noinline optnone "no-trapping-math"="true" "stack-protector-buffer-size"="8" "target-features"="+16-bit-insts,+ashr-pk-insts,+atomic-buffer-global-pk-add-f16-insts,+atomic-buffer-pk-add-bf16-inst,+atomic-ds-pk-add-16-insts,+atomic-fadd-rtn-insts,+atomic-flat-pk-add-16-insts,+atomic-global-pk-add-bf16-inst,+bf8-cvt-scale-insts,+bitop3-insts,+ci-insts,+dl-insts,+dot1-insts,+dot10-insts,+dot11-insts,+dot12-insts,+dot13-insts,+dot2-insts,+dot3-insts,+dot4-insts,+dot5-insts,+dot6-insts,+dot7-insts,+dot8-insts,+dot9-insts,+dpp,+f16bf16-to-fp6bf6-cvt-scale-insts,+f32-to-f16bf16-cvt-sr-insts,+fp4-cvt-scale-insts,+fp6bf6-cvt-scale-insts,+fp8-conversion-insts,+fp8-cvt-scale-insts,+fp8-insts,+gfx10-3-insts,+gfx10-insts,+gfx11-insts,+gfx12-insts,+gfx8-insts,+gfx9-insts,+gfx90a-insts,+gfx940-insts,+gfx950-insts,+gws,+image-insts,+mai-insts,+permlane16-swap,+permlane32-swap,+prng-inst,+s-memrealtime,+s-memtime-inst,+vmem-to-lds-load-insts,+wavefrontsize32,+wavefrontsize64" }
115115
// WITH-NONZERO-DEFAULT-AS: attributes #[[ATTR1:[0-9]+]] = { nounwind willreturn memory(read) }
116-
// WITH-NONZERO-DEFAULT-AS: attributes #[[ATTR2:[0-9]+]] = { "no-trapping-math"="true" "stack-protector-buffer-size"="8" "target-features"="+16-bit-insts,+ashr-pk-insts,+atomic-buffer-global-pk-add-f16-insts,+atomic-buffer-pk-add-bf16-inst,+atomic-ds-pk-add-16-insts,+atomic-fadd-rtn-insts,+atomic-flat-pk-add-16-insts,+atomic-global-pk-add-bf16-inst,+bf8-cvt-scale-insts,+bitop3-insts,+ci-insts,+dl-insts,+dot1-insts,+dot10-insts,+dot11-insts,+dot12-insts,+dot13-insts,+dot2-insts,+dot3-insts,+dot4-insts,+dot5-insts,+dot6-insts,+dot7-insts,+dot8-insts,+dot9-insts,+dpp,+f16bf16-to-fp6bf6-cvt-scale-insts,+f32-to-f16bf16-cvt-sr-insts,+fp4-cvt-scale-insts,+fp6bf6-cvt-scale-insts,+fp8-conversion-insts,+fp8-cvt-scale-insts,+fp8-insts,+gfx10-3-insts,+gfx10-insts,+gfx11-insts,+gfx12-insts,+gfx8-insts,+gfx9-insts,+gfx90a-insts,+gfx940-insts,+gfx950-insts,+gws,+image-insts,+mai-insts,+mem-to-lds-load-insts,+permlane16-swap,+permlane32-swap,+prng-inst,+s-memrealtime,+s-memtime-inst,+wavefrontsize32,+wavefrontsize64" }
116+
// WITH-NONZERO-DEFAULT-AS: attributes #[[ATTR2:[0-9]+]] = { "no-trapping-math"="true" "stack-protector-buffer-size"="8" "target-features"="+16-bit-insts,+ashr-pk-insts,+atomic-buffer-global-pk-add-f16-insts,+atomic-buffer-pk-add-bf16-inst,+atomic-ds-pk-add-16-insts,+atomic-fadd-rtn-insts,+atomic-flat-pk-add-16-insts,+atomic-global-pk-add-bf16-inst,+bf8-cvt-scale-insts,+bitop3-insts,+ci-insts,+dl-insts,+dot1-insts,+dot10-insts,+dot11-insts,+dot12-insts,+dot13-insts,+dot2-insts,+dot3-insts,+dot4-insts,+dot5-insts,+dot6-insts,+dot7-insts,+dot8-insts,+dot9-insts,+dpp,+f16bf16-to-fp6bf6-cvt-scale-insts,+f32-to-f16bf16-cvt-sr-insts,+fp4-cvt-scale-insts,+fp6bf6-cvt-scale-insts,+fp8-conversion-insts,+fp8-cvt-scale-insts,+fp8-insts,+gfx10-3-insts,+gfx10-insts,+gfx11-insts,+gfx12-insts,+gfx8-insts,+gfx9-insts,+gfx90a-insts,+gfx940-insts,+gfx950-insts,+gws,+image-insts,+mai-insts,+permlane16-swap,+permlane32-swap,+prng-inst,+s-memrealtime,+s-memtime-inst,+vmem-to-lds-load-insts,+wavefrontsize32,+wavefrontsize64" }
117117
// WITH-NONZERO-DEFAULT-AS: attributes #[[ATTR3]] = { nounwind }
118118
// WITH-NONZERO-DEFAULT-AS: attributes #[[ATTR4]] = { noreturn }
119119
//.

‎flang/test/Lower/OpenMP/target_cpu_features.f90

+1-1
Original file line numberDiff line numberDiff line change
@@ -12,7 +12,7 @@
1212
!AMDGCN-SAME: "+dl-insts", "+dot1-insts", "+dot10-insts", "+dot2-insts", "+dot3-insts",
1313
!AMDGCN-SAME: "+dot4-insts", "+dot5-insts", "+dot6-insts", "+dot7-insts", "+dpp",
1414
!AMDGCN-SAME: "+gfx8-insts", "+gfx9-insts", "+gws", "+image-insts", "+mai-insts",
15-
!AMDGCN-SAME: "+mem-to-lds-load-insts", "+s-memrealtime", "+s-memtime-inst", "+wavefrontsize64"]>
15+
!AMDGCN-SAME: "+s-memrealtime", "+s-memtime-inst", "+vmem-to-lds-load-insts", "+wavefrontsize64"]>
1616

1717
!NVPTX: module attributes {
1818
!NVPTX-SAME: fir.target_cpu = "sm_80"

‎llvm/lib/Target/AMDGPU/AMDGPU.td

+2-2
Original file line numberDiff line numberDiff line change
@@ -1273,8 +1273,8 @@ def FeatureLshlAddU64Inst
12731273
: SubtargetFeature<"lshl-add-u64-inst", "HasLshlAddU64Inst", "true",
12741274
"Has v_lshl_add_u64 instruction">;
12751275

1276-
def FeatureMemToLDSLoad : SubtargetFeature<"mem-to-lds-load-insts",
1277-
"HasMemToLDSLoad",
1276+
def FeatureMemToLDSLoad : SubtargetFeature<"vmem-to-lds-load-insts",
1277+
"HasVMemToLDSLoad",
12781278
"true",
12791279
"The platform has memory to lds instructions (global_load w/lds bit set, buffer_load w/lds bit set or global_load_lds. This does not include scratch_load_lds."
12801280
>;

‎llvm/lib/Target/AMDGPU/AMDGPUInstructionSelector.cpp

+2-2
Original file line numberDiff line numberDiff line change
@@ -3369,7 +3369,7 @@ bool AMDGPUInstructionSelector::selectG_INSERT_VECTOR_ELT(
33693369
}
33703370

33713371
bool AMDGPUInstructionSelector::selectBufferLoadLds(MachineInstr &MI) const {
3372-
if (!Subtarget->hasMemToLDSLoad())
3372+
if (!Subtarget->hasVMemToLDSLoad())
33733373
return false;
33743374
unsigned Opc;
33753375
unsigned Size = MI.getOperand(3).getImm();
@@ -3506,7 +3506,7 @@ static Register matchZeroExtendFromS32(MachineRegisterInfo &MRI, Register Reg) {
35063506
}
35073507

35083508
bool AMDGPUInstructionSelector::selectGlobalLoadLds(MachineInstr &MI) const{
3509-
if (!Subtarget->hasMemToLDSLoad())
3509+
if (!Subtarget->hasVMemToLDSLoad())
35103510
return false;
35113511

35123512
unsigned Opc;

‎llvm/lib/Target/AMDGPU/GCNSubtarget.h

+2-2
Original file line numberDiff line numberDiff line change
@@ -193,7 +193,7 @@ class GCNSubtarget final : public AMDGPUGenSubtargetInfo,
193193
bool SupportsSRAMECC = false;
194194
bool DynamicVGPR = false;
195195
bool DynamicVGPRBlockSize32 = false;
196-
bool HasMemToLDSLoad = false;
196+
bool HasVMemToLDSLoad = false;
197197

198198
// This should not be used directly. 'TargetID' tracks the dynamic settings
199199
// for SRAMECC.
@@ -1320,7 +1320,7 @@ class GCNSubtarget final : public AMDGPUGenSubtargetInfo,
13201320
return hasGFX950Insts();
13211321
}
13221322

1323-
bool hasMemToLDSLoad() const { return HasMemToLDSLoad; }
1323+
bool hasVMemToLDSLoad() const { return HasVMemToLDSLoad; }
13241324

13251325
bool hasSALUFloatInsts() const { return HasSALUFloatInsts; }
13261326

‎llvm/lib/Target/AMDGPU/SIISelLowering.cpp

+2-2
Original file line numberDiff line numberDiff line change
@@ -10104,7 +10104,7 @@ SDValue SITargetLowering::LowerINTRINSIC_VOID(SDValue Op,
1010410104
case Intrinsic::amdgcn_raw_ptr_buffer_load_lds:
1010510105
case Intrinsic::amdgcn_struct_buffer_load_lds:
1010610106
case Intrinsic::amdgcn_struct_ptr_buffer_load_lds: {
10107-
if (!Subtarget->hasMemToLDSLoad())
10107+
if (!Subtarget->hasVMemToLDSLoad())
1010810108
return SDValue();
1010910109
unsigned Opc;
1011010110
bool HasVIndex =
@@ -10212,7 +10212,7 @@ SDValue SITargetLowering::LowerINTRINSIC_VOID(SDValue Op,
1021210212
return SDValue(Load, 0);
1021310213
}
1021410214
case Intrinsic::amdgcn_global_load_lds: {
10215-
if (!Subtarget->hasMemToLDSLoad())
10215+
if (!Subtarget->hasVMemToLDSLoad())
1021610216
return SDValue();
1021710217

1021810218
unsigned Opc;

‎llvm/lib/TargetParser/TargetParser.cpp

+5-5
Original file line numberDiff line numberDiff line change
@@ -374,7 +374,7 @@ void AMDGPU::fillAMDGPUFeatureMap(StringRef GPU, const Triple &T,
374374
Features["prng-inst"] = true;
375375
Features["wavefrontsize32"] = true;
376376
Features["wavefrontsize64"] = true;
377-
Features["mem-to-lds-load-insts"] = true;
377+
Features["vmem-to-lds-load-insts"] = true;
378378
} else if (T.isAMDGCN()) {
379379
AMDGPU::GPUKind Kind = parseArchAMDGCN(GPU);
380380
switch (Kind) {
@@ -460,7 +460,7 @@ void AMDGPU::fillAMDGPUFeatureMap(StringRef GPU, const Triple &T,
460460
Features["s-memrealtime"] = true;
461461
Features["s-memtime-inst"] = true;
462462
Features["gws"] = true;
463-
Features["mem-to-lds-load-insts"] = true;
463+
Features["vmem-to-lds-load-insts"] = true;
464464
break;
465465
case GK_GFX1012:
466466
case GK_GFX1011:
@@ -485,7 +485,7 @@ void AMDGPU::fillAMDGPUFeatureMap(StringRef GPU, const Triple &T,
485485
Features["s-memrealtime"] = true;
486486
Features["s-memtime-inst"] = true;
487487
Features["gws"] = true;
488-
Features["mem-to-lds-load-insts"] = true;
488+
Features["vmem-to-lds-load-insts"] = true;
489489
break;
490490
case GK_GFX950:
491491
Features["bitop3-insts"] = true;
@@ -536,7 +536,7 @@ void AMDGPU::fillAMDGPUFeatureMap(StringRef GPU, const Triple &T,
536536
Features["ci-insts"] = true;
537537
Features["s-memtime-inst"] = true;
538538
Features["gws"] = true;
539-
Features["mem-to-lds-load-insts"] = true;
539+
Features["vmem-to-lds-load-insts"] = true;
540540
break;
541541
case GK_GFX90A:
542542
Features["gfx90a-insts"] = true;
@@ -589,7 +589,7 @@ void AMDGPU::fillAMDGPUFeatureMap(StringRef GPU, const Triple &T,
589589
Features["image-insts"] = true;
590590
Features["s-memtime-inst"] = true;
591591
Features["gws"] = true;
592-
Features["mem-to-lds-load-insts"] = true;
592+
Features["vmem-to-lds-load-insts"] = true;
593593
break;
594594
case GK_NONE:
595595
break;

0 commit comments

Comments
 (0)