Skip to content

Commit 9b12959

Browse files
shiltianshiltian
committed
[Attributor][AMD] Enable AAIndirectCallInfo for AMDAttributorPass
1 parent 9f75270 commit 9b12959

File tree

7 files changed

+97
-13
lines changed

7 files changed

+97
-13
lines changed

‎llvm/include/llvm/Transforms/IPO/Attributor.h‎

Lines changed: 5 additions & 4 deletions
Original file line numberDiff line numberDiff line change
@@ -1448,7 +1448,7 @@ struct AttributorConfig {
14481448
/// Callback function to determine if an indirect call targets should be made
14491449
/// direct call targets (with an if-cascade).
14501450
std::function<bool(Attributor &A, const AbstractAttribute &AA, CallBase &CB,
1451-
Function &AssummedCallee)>
1451+
Function &AssummedCallee, unsigned NumAssummedCallee)>
14521452
IndirectCalleeSpecializationCallback = nullptr;
14531453

14541454
/// Helper to update an underlying call graph and to delete functions.
@@ -1718,10 +1718,11 @@ struct Attributor {
17181718
/// Return true if we should specialize the call site \b CB for the potential
17191719
/// callee \p Fn.
17201720
bool shouldSpecializeCallSiteForCallee(const AbstractAttribute &AA,
1721-
CallBase &CB, Function &Callee) {
1721+
CallBase &CB, Function &Callee,
1722+
unsigned NumAssummedCallee) {
17221723
return Configuration.IndirectCalleeSpecializationCallback
1723-
? Configuration.IndirectCalleeSpecializationCallback(*this, AA,
1724-
CB, Callee)
1724+
? Configuration.IndirectCalleeSpecializationCallback(
1725+
*this, AA, CB, Callee, NumAssummedCallee)
17251726
: true;
17261727
}
17271728

‎llvm/lib/Target/AMDGPU/AMDGPUAttributor.cpp‎

Lines changed: 14 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -14,6 +14,7 @@
1414
#include "GCNSubtarget.h"
1515
#include "Utils/AMDGPUBaseInfo.h"
1616
#include "llvm/Analysis/CycleAnalysis.h"
17+
#include "llvm/Analysis/TargetTransformInfo.h"
1718
#include "llvm/CodeGen/TargetPassConfig.h"
1819
#include "llvm/IR/IntrinsicsAMDGPU.h"
1920
#include "llvm/IR/IntrinsicsR600.h"
@@ -1038,12 +1039,24 @@ static bool runImpl(Module &M, AnalysisGetter &AG, TargetMachine &TM) {
10381039
&AAPotentialValues::ID, &AAAMDFlatWorkGroupSize::ID,
10391040
&AAAMDWavesPerEU::ID, &AAAMDGPUNoAGPR::ID, &AACallEdges::ID,
10401041
&AAPointerInfo::ID, &AAPotentialConstantValues::ID,
1041-
&AAUnderlyingObjects::ID});
1042+
&AAUnderlyingObjects::ID, &AAIndirectCallInfo::ID, &AAInstanceInfo::ID});
10421043

10431044
AttributorConfig AC(CGUpdater);
10441045
AC.Allowed = &Allowed;
10451046
AC.IsModulePass = true;
10461047
AC.DefaultInitializeLiveInternals = false;
1048+
AC.IndirectCalleeSpecializationCallback =
1049+
[&TM](Attributor &A, const AbstractAttribute &AA, CallBase &CB,
1050+
Function &Callee, unsigned NumAssummedCallee) {
1051+
if (AMDGPU::isEntryFunctionCC(Callee.getCallingConv()))
1052+
return false;
1053+
// Singleton functions can be specialized.
1054+
if (NumAssummedCallee == 1)
1055+
return true;
1056+
// Otherwise specialize uniform values.
1057+
const auto &TTI = TM.getTargetTransformInfo(*CB.getCaller());
1058+
return TTI.isAlwaysUniform(CB.getCalledOperand());
1059+
};
10471060
AC.IPOAmendableCB = [](const Function &F) {
10481061
return F.getCallingConv() == CallingConv::AMDGPU_KERNEL;
10491062
};

‎llvm/lib/Transforms/IPO/Attributor.cpp‎

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -3836,7 +3836,7 @@ static bool runAttributorOnFunctions(InformationCache &InfoCache,
38363836
if (MaxSpecializationPerCB.getNumOccurrences()) {
38373837
AC.IndirectCalleeSpecializationCallback =
38383838
[&](Attributor &, const AbstractAttribute &AA, CallBase &CB,
3839-
Function &Callee) {
3839+
Function &Callee, unsigned) {
38403840
if (MaxSpecializationPerCB == 0)
38413841
return false;
38423842
auto &Set = IndirectCalleeTrackingMap[&CB];

‎llvm/lib/Transforms/IPO/AttributorAttributes.cpp‎

Lines changed: 2 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -12347,7 +12347,8 @@ struct AAIndirectCallInfoCallSite : public AAIndirectCallInfo {
1234712347
SmallVector<Function *, 8> SkippedAssumedCallees;
1234812348
SmallVector<std::pair<CallInst *, Instruction *>> NewCalls;
1234912349
for (Function *NewCallee : AssumedCallees) {
12350-
if (!A.shouldSpecializeCallSiteForCallee(*this, *CB, *NewCallee)) {
12350+
if (!A.shouldSpecializeCallSiteForCallee(*this, *CB, *NewCallee,
12351+
AssumedCallees.size())) {
1235112352
SkippedAssumedCallees.push_back(NewCallee);
1235212353
SpecializedForAllCallees = false;
1235312354
continue;

‎llvm/test/CodeGen/AMDGPU/amdgpu-attributor-no-agpr.ll‎

Lines changed: 3 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -231,7 +231,7 @@ define amdgpu_kernel void @indirect_calls_none_agpr(i1 %cond) {
231231
; CHECK-LABEL: define amdgpu_kernel void @indirect_calls_none_agpr(
232232
; CHECK-SAME: i1 [[COND:%.*]]) #[[ATTR0]] {
233233
; CHECK-NEXT: [[FPTR:%.*]] = select i1 [[COND]], ptr @empty, ptr @also_empty
234-
; CHECK-NEXT: call void [[FPTR]]()
234+
; CHECK-NEXT: call void [[FPTR]](), !callees [[META0:![0-9]+]]
235235
; CHECK-NEXT: ret void
236236
;
237237
%fptr = select i1 %cond, ptr @empty, ptr @also_empty
@@ -253,3 +253,5 @@ attributes #0 = { "amdgpu-no-agpr" }
253253
; CHECK: attributes #[[ATTR8:[0-9]+]] = { nocallback nofree nounwind willreturn memory(argmem: readwrite) "target-cpu"="gfx90a" }
254254
; CHECK: attributes #[[ATTR9]] = { "amdgpu-no-agpr" }
255255
;.
256+
; CHECK: [[META0]] = !{ptr @also_empty, ptr @empty}
257+
;.

‎llvm/test/CodeGen/AMDGPU/simple-indirect-call-2.ll‎

Lines changed: 63 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,63 @@
1+
; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --function-signature --check-globals
2+
; RUN: opt -S -mtriple=amdgcn-amd-amdhsa -passes=amdgpu-attributor %s | FileCheck -check-prefix=ATTRIBUTOR_GCN %s
3+
4+
target datalayout = "A5"
5+
6+
@global_ptr = local_unnamed_addr global ptr null, align 8
7+
8+
;.
9+
; ATTRIBUTOR_GCN: @global_ptr = local_unnamed_addr global ptr null, align 8
10+
;.
11+
define void @callee1() {
12+
; ATTRIBUTOR_GCN-LABEL: define {{[^@]+}}@callee1
13+
; ATTRIBUTOR_GCN-SAME: () #[[ATTR0:[0-9]+]] {
14+
; ATTRIBUTOR_GCN-NEXT: ret void
15+
;
16+
ret void
17+
}
18+
19+
define void @callee2() {
20+
; ATTRIBUTOR_GCN-LABEL: define {{[^@]+}}@callee2
21+
; ATTRIBUTOR_GCN-SAME: () #[[ATTR0]] {
22+
; ATTRIBUTOR_GCN-NEXT: ret void
23+
;
24+
ret void
25+
}
26+
27+
define ptr @take_addr_1() {
28+
; ATTRIBUTOR_GCN-LABEL: define {{[^@]+}}@take_addr_1
29+
; ATTRIBUTOR_GCN-SAME: () #[[ATTR0]] {
30+
; ATTRIBUTOR_GCN-NEXT: ret ptr @callee1
31+
;
32+
ret ptr @callee1
33+
}
34+
35+
define ptr @take_addr_2() {
36+
; ATTRIBUTOR_GCN-LABEL: define {{[^@]+}}@take_addr_2
37+
; ATTRIBUTOR_GCN-SAME: () #[[ATTR0]] {
38+
; ATTRIBUTOR_GCN-NEXT: ret ptr @callee2
39+
;
40+
ret ptr @callee2
41+
}
42+
43+
define amdgpu_kernel void @test_simple_indirect_call() {
44+
; ATTRIBUTOR_GCN-LABEL: define {{[^@]+}}@test_simple_indirect_call
45+
; ATTRIBUTOR_GCN-SAME: () #[[ATTR1:[0-9]+]] {
46+
; ATTRIBUTOR_GCN-NEXT: [[FP:%.*]] = load ptr, ptr @global_ptr, align 8
47+
; ATTRIBUTOR_GCN-NEXT: call void [[FP]]()
48+
; ATTRIBUTOR_GCN-NEXT: ret void
49+
;
50+
%fp = load ptr, ptr @global_ptr
51+
call void %fp()
52+
ret void
53+
}
54+
55+
56+
!llvm.module.flags = !{!0}
57+
!0 = !{i32 1, !"amdhsa_code_object_version", i32 500}
58+
;.
59+
; ATTRIBUTOR_GCN: attributes #[[ATTR0]] = { "amdgpu-no-agpr" "amdgpu-no-completion-action" "amdgpu-no-default-queue" "amdgpu-no-dispatch-id" "amdgpu-no-dispatch-ptr" "amdgpu-no-heap-ptr" "amdgpu-no-hostcall-ptr" "amdgpu-no-implicitarg-ptr" "amdgpu-no-lds-kernel-id" "amdgpu-no-multigrid-sync-arg" "amdgpu-no-queue-ptr" "amdgpu-no-workgroup-id-x" "amdgpu-no-workgroup-id-y" "amdgpu-no-workgroup-id-z" "amdgpu-no-workitem-id-x" "amdgpu-no-workitem-id-y" "amdgpu-no-workitem-id-z" "amdgpu-waves-per-eu"="4,10" "uniform-work-group-size"="false" }
60+
; ATTRIBUTOR_GCN: attributes #[[ATTR1]] = { "uniform-work-group-size"="false" }
61+
;.
62+
; ATTRIBUTOR_GCN: [[META0:![0-9]+]] = !{i32 1, !"amdhsa_code_object_version", i32 500}
63+
;.

‎llvm/test/CodeGen/AMDGPU/simple-indirect-call.ll‎

Lines changed: 9 additions & 5 deletions
Original file line numberDiff line numberDiff line change
@@ -37,7 +37,7 @@ define amdgpu_kernel void @test_simple_indirect_call() {
3737
; ATTRIBUTOR_GCN-NEXT: [[FPTR_CAST:%.*]] = addrspacecast ptr addrspace(5) [[FPTR]] to ptr
3838
; ATTRIBUTOR_GCN-NEXT: store ptr @indirect, ptr [[FPTR_CAST]], align 8
3939
; ATTRIBUTOR_GCN-NEXT: [[FP:%.*]] = load ptr, ptr [[FPTR_CAST]], align 8
40-
; ATTRIBUTOR_GCN-NEXT: call void [[FP]]()
40+
; ATTRIBUTOR_GCN-NEXT: call void @indirect()
4141
; ATTRIBUTOR_GCN-NEXT: ret void
4242
;
4343
; GFX9-LABEL: test_simple_indirect_call:
@@ -75,12 +75,16 @@ define amdgpu_kernel void @test_simple_indirect_call() {
7575
ret void
7676
}
7777

78+
79+
!llvm.module.flags = !{!0}
80+
!0 = !{i32 1, !"amdhsa_code_object_version", i32 500}
7881
;.
7982
; AKF_GCN: attributes #[[ATTR0]] = { "amdgpu-calls" "amdgpu-stack-objects" }
8083
;.
8184
; ATTRIBUTOR_GCN: attributes #[[ATTR0]] = { "amdgpu-no-agpr" "amdgpu-no-completion-action" "amdgpu-no-default-queue" "amdgpu-no-dispatch-id" "amdgpu-no-dispatch-ptr" "amdgpu-no-heap-ptr" "amdgpu-no-hostcall-ptr" "amdgpu-no-implicitarg-ptr" "amdgpu-no-lds-kernel-id" "amdgpu-no-multigrid-sync-arg" "amdgpu-no-queue-ptr" "amdgpu-no-workgroup-id-x" "amdgpu-no-workgroup-id-y" "amdgpu-no-workgroup-id-z" "amdgpu-no-workitem-id-x" "amdgpu-no-workitem-id-y" "amdgpu-no-workitem-id-z" "amdgpu-waves-per-eu"="4,10" "uniform-work-group-size"="false" }
82-
; ATTRIBUTOR_GCN: attributes #[[ATTR1]] = { "uniform-work-group-size"="false" }
85+
; ATTRIBUTOR_GCN: attributes #[[ATTR1]] = { "amdgpu-no-completion-action" "amdgpu-no-default-queue" "amdgpu-no-dispatch-id" "amdgpu-no-dispatch-ptr" "amdgpu-no-heap-ptr" "amdgpu-no-hostcall-ptr" "amdgpu-no-lds-kernel-id" "amdgpu-no-multigrid-sync-arg" "amdgpu-no-queue-ptr" "amdgpu-no-workgroup-id-x" "amdgpu-no-workgroup-id-y" "amdgpu-no-workgroup-id-z" "amdgpu-no-workitem-id-x" "amdgpu-no-workitem-id-y" "amdgpu-no-workitem-id-z" "uniform-work-group-size"="false" }
86+
;.
87+
; AKF_GCN: [[META0:![0-9]+]] = !{i32 1, !"amdhsa_code_object_version", i32 500}
88+
;.
89+
; ATTRIBUTOR_GCN: [[META0:![0-9]+]] = !{i32 1, !"amdhsa_code_object_version", i32 500}
8390
;.
84-
85-
!llvm.module.flags = !{!0}
86-
!0 = !{i32 1, !"amdhsa_code_object_version", i32 500}

0 commit comments

Comments
 (0)