| ; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --check-globals |
| ; RUN: opt < %s -passes=simplifycfg -simplifycfg-require-and-preserve-domtree=1 -switch-to-lookup -S | FileCheck %s |
| ; RUN: opt < %s -passes='simplifycfg<switch-to-lookup>' -S | FileCheck %s |
| |
| target datalayout = "e-n32" |
| |
| ;. |
| ; CHECK: @switch.table.test1 = private unnamed_addr constant [4 x i32] [i32 11984, i32 1143, i32 99783, i32 99783], align 4 |
| ; CHECK: @switch.table.test3 = private unnamed_addr constant [3 x i32] [i32 11984, i32 1143, i32 99783], align 4 |
| ; CHECK: @switch.table.test6 = private unnamed_addr constant [4 x i32] [i32 99783, i32 99783, i32 1143, i32 11984], align 4 |
| ; CHECK: @switch.table.test8 = private unnamed_addr constant [5 x i32] [i32 11984, i32 1143, i32 99783, i32 8867, i32 99783], align 4 |
| ; CHECK: @switch.table.test9 = private unnamed_addr constant [28 x i32] [i32 99783, i32 8867, i32 99783, i32 8867, i32 8867, i32 8867, i32 11984, i32 8867, i32 8867, i32 8867, i32 8867, i32 8867, i32 8867, i32 8867, i32 8867, i32 8867, i32 8867, i32 1143, i32 8867, i32 8867, i32 8867, i32 8867, i32 8867, i32 8867, i32 8867, i32 8867, i32 8867, i32 99783], align 4 |
| ;. |
| define i32 @test1(i32 %a) !prof !0 { |
| ; CHECK-LABEL: @test1( |
| ; CHECK-NEXT: [[TMP1:%.*]] = sub i32 [[A:%.*]], 97 |
| ; CHECK-NEXT: [[TMP2:%.*]] = call i32 @llvm.fshl.i32(i32 [[TMP1]], i32 [[TMP1]], i32 28) |
| ; CHECK-NEXT: [[TMP3:%.*]] = icmp ult i32 [[TMP2]], 4 |
| ; CHECK-NEXT: br i1 [[TMP3]], label [[SWITCH_LOOKUP:%.*]], label [[COMMON_RET:%.*]], !prof [[PROF1:![0-9]+]] |
| ; CHECK: switch.lookup: |
| ; CHECK-NEXT: [[TMP4:%.*]] = zext nneg i32 [[TMP2]] to i64 |
| ; CHECK-NEXT: [[SWITCH_GEP:%.*]] = getelementptr inbounds [4 x i32], ptr @switch.table.test1, i64 0, i64 [[TMP4]] |
| ; CHECK-NEXT: [[SWITCH_LOAD:%.*]] = load i32, ptr [[SWITCH_GEP]], align 4 |
| ; CHECK-NEXT: br label [[COMMON_RET]] |
| ; CHECK: common.ret: |
| ; CHECK-NEXT: [[COMMON_RET_OP:%.*]] = phi i32 [ [[SWITCH_LOAD]], [[SWITCH_LOOKUP]] ], [ 8867, [[TMP0:%.*]] ] |
| ; CHECK-NEXT: ret i32 [[COMMON_RET_OP]] |
| ; |
| switch i32 %a, label %def [ |
| i32 97, label %one |
| i32 113, label %two |
| i32 129, label %three |
| i32 145, label %three |
| ], !prof !1 |
| |
| def: |
| ret i32 8867 |
| |
| one: |
| ret i32 11984 |
| two: |
| ret i32 1143 |
| three: |
| ret i32 99783 |
| } |
| |
| ; Optimization shouldn't trigger; bitwidth > 64 |
| define i128 @test2(i128 %a) { |
| ; CHECK-LABEL: @test2( |
| ; CHECK-NEXT: switch i128 [[A:%.*]], label [[COMMON_RET:%.*]] [ |
| ; CHECK-NEXT: i128 97, label [[ONE:%.*]] |
| ; CHECK-NEXT: i128 113, label [[TWO:%.*]] |
| ; CHECK-NEXT: i128 129, label [[THREE:%.*]] |
| ; CHECK-NEXT: i128 145, label [[THREE]] |
| ; CHECK-NEXT: ] |
| ; CHECK: common.ret: |
| ; CHECK-NEXT: [[COMMON_RET_OP:%.*]] = phi i128 [ 99783, [[THREE]] ], [ 11984, [[ONE]] ], [ 1143, [[TWO]] ], [ 8867, [[TMP0:%.*]] ] |
| ; CHECK-NEXT: ret i128 [[COMMON_RET_OP]] |
| ; CHECK: one: |
| ; CHECK-NEXT: br label [[COMMON_RET]] |
| ; CHECK: two: |
| ; CHECK-NEXT: br label [[COMMON_RET]] |
| ; CHECK: three: |
| ; CHECK-NEXT: br label [[COMMON_RET]] |
| ; |
| switch i128 %a, label %def [ |
| i128 97, label %one |
| i128 113, label %two |
| i128 129, label %three |
| i128 145, label %three |
| ] |
| |
| def: |
| ret i128 8867 |
| |
| one: |
| ret i128 11984 |
| two: |
| ret i128 1143 |
| three: |
| ret i128 99783 |
| } |
| |
| ; Optimization shouldn't trigger; no holes present |
| define i32 @test3(i32 %a) { |
| ; CHECK-LABEL: @test3( |
| ; CHECK-NEXT: [[SWITCH_TABLEIDX:%.*]] = sub i32 [[A:%.*]], 97 |
| ; CHECK-NEXT: [[TMP1:%.*]] = icmp ult i32 [[SWITCH_TABLEIDX]], 3 |
| ; CHECK-NEXT: br i1 [[TMP1]], label [[SWITCH_LOOKUP:%.*]], label [[COMMON_RET:%.*]] |
| ; CHECK: switch.lookup: |
| ; CHECK-NEXT: [[TMP2:%.*]] = zext nneg i32 [[SWITCH_TABLEIDX]] to i64 |
| ; CHECK-NEXT: [[SWITCH_GEP:%.*]] = getelementptr inbounds [3 x i32], ptr @switch.table.test3, i64 0, i64 [[TMP2]] |
| ; CHECK-NEXT: [[SWITCH_LOAD:%.*]] = load i32, ptr [[SWITCH_GEP]], align 4 |
| ; CHECK-NEXT: br label [[COMMON_RET]] |
| ; CHECK: common.ret: |
| ; CHECK-NEXT: [[COMMON_RET_OP:%.*]] = phi i32 [ [[SWITCH_LOAD]], [[SWITCH_LOOKUP]] ], [ 8867, [[TMP0:%.*]] ] |
| ; CHECK-NEXT: ret i32 [[COMMON_RET_OP]] |
| ; |
| switch i32 %a, label %def [ |
| i32 97, label %one |
| i32 98, label %two |
| i32 99, label %three |
| ] |
| |
| def: |
| ret i32 8867 |
| |
| one: |
| ret i32 11984 |
| two: |
| ret i32 1143 |
| three: |
| ret i32 99783 |
| } |
| |
| ; Optimization shouldn't trigger; not an arithmetic progression |
| define i32 @test4(i32 %a) { |
| ; CHECK-LABEL: @test4( |
| ; CHECK-NEXT: switch i32 [[A:%.*]], label [[COMMON_RET:%.*]] [ |
| ; CHECK-NEXT: i32 97, label [[ONE:%.*]] |
| ; CHECK-NEXT: i32 114, label [[TWO:%.*]] |
| ; CHECK-NEXT: i32 129, label [[THREE:%.*]] |
| ; CHECK-NEXT: i32 145, label [[THREE]] |
| ; CHECK-NEXT: ] |
| ; CHECK: common.ret: |
| ; CHECK-NEXT: [[COMMON_RET_OP:%.*]] = phi i32 [ 99783, [[THREE]] ], [ 11984, [[ONE]] ], [ 1143, [[TWO]] ], [ 8867, [[TMP0:%.*]] ] |
| ; CHECK-NEXT: ret i32 [[COMMON_RET_OP]] |
| ; CHECK: one: |
| ; CHECK-NEXT: br label [[COMMON_RET]] |
| ; CHECK: two: |
| ; CHECK-NEXT: br label [[COMMON_RET]] |
| ; CHECK: three: |
| ; CHECK-NEXT: br label [[COMMON_RET]] |
| ; |
| switch i32 %a, label %def [ |
| i32 97, label %one |
| i32 114, label %two |
| i32 129, label %three |
| i32 145, label %three |
| ] |
| |
| def: |
| ret i32 8867 |
| |
| one: |
| ret i32 11984 |
| two: |
| ret i32 1143 |
| three: |
| ret i32 99783 |
| } |
| |
| ; Optimization shouldn't trigger; not a power of two |
| define i32 @test5(i32 %a) { |
| ; CHECK-LABEL: @test5( |
| ; CHECK-NEXT: switch i32 [[A:%.*]], label [[COMMON_RET:%.*]] [ |
| ; CHECK-NEXT: i32 97, label [[ONE:%.*]] |
| ; CHECK-NEXT: i32 112, label [[TWO:%.*]] |
| ; CHECK-NEXT: i32 127, label [[THREE:%.*]] |
| ; CHECK-NEXT: i32 142, label [[THREE]] |
| ; CHECK-NEXT: ] |
| ; CHECK: common.ret: |
| ; CHECK-NEXT: [[COMMON_RET_OP:%.*]] = phi i32 [ 99783, [[THREE]] ], [ 11984, [[ONE]] ], [ 1143, [[TWO]] ], [ 8867, [[TMP0:%.*]] ] |
| ; CHECK-NEXT: ret i32 [[COMMON_RET_OP]] |
| ; CHECK: one: |
| ; CHECK-NEXT: br label [[COMMON_RET]] |
| ; CHECK: two: |
| ; CHECK-NEXT: br label [[COMMON_RET]] |
| ; CHECK: three: |
| ; CHECK-NEXT: br label [[COMMON_RET]] |
| ; |
| switch i32 %a, label %def [ |
| i32 97, label %one |
| i32 112, label %two |
| i32 127, label %three |
| i32 142, label %three |
| ] |
| |
| def: |
| ret i32 8867 |
| |
| one: |
| ret i32 11984 |
| two: |
| ret i32 1143 |
| three: |
| ret i32 99783 |
| } |
| |
| define i32 @test6(i32 %a) optsize { |
| ; CHECK-LABEL: @test6( |
| ; CHECK-NEXT: [[TMP1:%.*]] = sub i32 [[A:%.*]], -109 |
| ; CHECK-NEXT: [[TMP2:%.*]] = call i32 @llvm.fshl.i32(i32 [[TMP1]], i32 [[TMP1]], i32 30) |
| ; CHECK-NEXT: [[TMP3:%.*]] = icmp ult i32 [[TMP2]], 4 |
| ; CHECK-NEXT: br i1 [[TMP3]], label [[SWITCH_LOOKUP:%.*]], label [[COMMON_RET:%.*]] |
| ; CHECK: switch.lookup: |
| ; CHECK-NEXT: [[TMP4:%.*]] = zext nneg i32 [[TMP2]] to i64 |
| ; CHECK-NEXT: [[SWITCH_GEP:%.*]] = getelementptr inbounds [4 x i32], ptr @switch.table.test6, i64 0, i64 [[TMP4]] |
| ; CHECK-NEXT: [[SWITCH_LOAD:%.*]] = load i32, ptr [[SWITCH_GEP]], align 4 |
| ; CHECK-NEXT: br label [[COMMON_RET]] |
| ; CHECK: common.ret: |
| ; CHECK-NEXT: [[COMMON_RET_OP:%.*]] = phi i32 [ [[SWITCH_LOAD]], [[SWITCH_LOOKUP]] ], [ 8867, [[TMP0:%.*]] ] |
| ; CHECK-NEXT: ret i32 [[COMMON_RET_OP]] |
| ; |
| switch i32 %a, label %def [ |
| i32 -97, label %one |
| i32 -101, label %two |
| i32 -105, label %three |
| i32 -109, label %three |
| ] |
| |
| def: |
| ret i32 8867 |
| |
| one: |
| ret i32 11984 |
| two: |
| ret i32 1143 |
| three: |
| ret i32 99783 |
| } |
| |
| define i8 @test7(i8 %a) optsize { |
| ; CHECK-LABEL: @test7( |
| ; CHECK-NEXT: common.ret: |
| ; CHECK-NEXT: [[TMP0:%.*]] = sub i8 [[A:%.*]], -36 |
| ; CHECK-NEXT: [[TMP1:%.*]] = call i8 @llvm.fshl.i8(i8 [[TMP0]], i8 [[TMP0]], i8 6) |
| ; CHECK-NEXT: [[TMP2:%.*]] = icmp ult i8 [[TMP1]], 4 |
| ; CHECK-NEXT: [[SWITCH_CAST:%.*]] = zext i8 [[TMP1]] to i32 |
| ; CHECK-NEXT: [[SWITCH_SHIFTAMT:%.*]] = mul nuw nsw i32 [[SWITCH_CAST]], 8 |
| ; CHECK-NEXT: [[SWITCH_DOWNSHIFT:%.*]] = lshr i32 -943228976, [[SWITCH_SHIFTAMT]] |
| ; CHECK-NEXT: [[SWITCH_MASKED:%.*]] = trunc i32 [[SWITCH_DOWNSHIFT]] to i8 |
| ; CHECK-NEXT: [[COMMON_RET_OP:%.*]] = select i1 [[TMP2]], i8 [[SWITCH_MASKED]], i8 -93 |
| ; CHECK-NEXT: ret i8 [[COMMON_RET_OP]] |
| ; |
| switch i8 %a, label %def [ |
| i8 220, label %one |
| i8 224, label %two |
| i8 228, label %three |
| i8 232, label %three |
| ] |
| |
| def: |
| ret i8 8867 |
| |
| one: |
| ret i8 11984 |
| two: |
| ret i8 1143 |
| three: |
| ret i8 99783 |
| } |
| |
| define i32 @test8(i32 %a) optsize { |
| ; CHECK-LABEL: @test8( |
| ; CHECK-NEXT: [[TMP1:%.*]] = sub i32 [[A:%.*]], 97 |
| ; CHECK-NEXT: [[TMP2:%.*]] = call i32 @llvm.fshl.i32(i32 [[TMP1]], i32 [[TMP1]], i32 30) |
| ; CHECK-NEXT: [[TMP3:%.*]] = icmp ult i32 [[TMP2]], 5 |
| ; CHECK-NEXT: br i1 [[TMP3]], label [[SWITCH_LOOKUP:%.*]], label [[COMMON_RET:%.*]] |
| ; CHECK: switch.lookup: |
| ; CHECK-NEXT: [[TMP4:%.*]] = zext nneg i32 [[TMP2]] to i64 |
| ; CHECK-NEXT: [[SWITCH_GEP:%.*]] = getelementptr inbounds [5 x i32], ptr @switch.table.test8, i64 0, i64 [[TMP4]] |
| ; CHECK-NEXT: [[SWITCH_LOAD:%.*]] = load i32, ptr [[SWITCH_GEP]], align 4 |
| ; CHECK-NEXT: br label [[COMMON_RET]] |
| ; CHECK: common.ret: |
| ; CHECK-NEXT: [[COMMON_RET_OP:%.*]] = phi i32 [ [[SWITCH_LOAD]], [[SWITCH_LOOKUP]] ], [ 8867, [[TMP0:%.*]] ] |
| ; CHECK-NEXT: ret i32 [[COMMON_RET_OP]] |
| ; |
| switch i32 %a, label %def [ |
| i32 97, label %one |
| i32 101, label %two |
| i32 105, label %three |
| i32 113, label %three |
| ] |
| |
| def: |
| ret i32 8867 |
| |
| one: |
| ret i32 11984 |
| two: |
| ret i32 1143 |
| three: |
| ret i32 99783 |
| } |
| |
| define i32 @test9(i32 %a) { |
| ; CHECK-LABEL: @test9( |
| ; CHECK-NEXT: [[TMP1:%.*]] = sub i32 [[A:%.*]], 6 |
| ; CHECK-NEXT: [[TMP2:%.*]] = call i32 @llvm.fshl.i32(i32 [[TMP1]], i32 [[TMP1]], i32 31) |
| ; CHECK-NEXT: [[TMP3:%.*]] = icmp ult i32 [[TMP2]], 28 |
| ; CHECK-NEXT: br i1 [[TMP3]], label [[SWITCH_LOOKUP:%.*]], label [[COMMON_RET:%.*]] |
| ; CHECK: switch.lookup: |
| ; CHECK-NEXT: [[TMP4:%.*]] = zext nneg i32 [[TMP2]] to i64 |
| ; CHECK-NEXT: [[SWITCH_GEP:%.*]] = getelementptr inbounds [28 x i32], ptr @switch.table.test9, i64 0, i64 [[TMP4]] |
| ; CHECK-NEXT: [[SWITCH_LOAD:%.*]] = load i32, ptr [[SWITCH_GEP]], align 4 |
| ; CHECK-NEXT: br label [[COMMON_RET]] |
| ; CHECK: common.ret: |
| ; CHECK-NEXT: [[COMMON_RET_OP:%.*]] = phi i32 [ [[SWITCH_LOAD]], [[SWITCH_LOOKUP]] ], [ 8867, [[TMP0:%.*]] ] |
| ; CHECK-NEXT: ret i32 [[COMMON_RET_OP]] |
| ; |
| switch i32 %a, label %def [ |
| i32 18, label %one |
| i32 40, label %two |
| i32 6, label %three |
| i32 10, label %three |
| i32 60, label %three |
| ] |
| |
| def: |
| ret i32 8867 |
| |
| one: |
| ret i32 11984 |
| two: |
| ret i32 1143 |
| three: |
| ret i32 99783 |
| } |
| |
| ; Base=0 is profitable here: shifting out the common low zero bits produces a |
| ; dense range without subtracting the local minimum first. SimplifyCFG does not |
| ; form a lookup table here, so the reduced switch is tested directly. |
| define void @test10(i32 %a) { |
| ; CHECK-LABEL: @test10( |
| ; CHECK-NEXT: [[TMP:%.*]] = call i32 @llvm.fshl.i32(i32 [[A:%.*]], i32 [[A]], i32 27) |
| ; CHECK-NEXT: switch i32 [[TMP]], label [[COMMON_RET:%.*]] [ |
| ; CHECK-NEXT: i32 1, label [[ONE:%.*]] |
| ; CHECK-NEXT: i32 2, label [[TWO:%.*]] |
| ; CHECK-NEXT: i32 3, label [[THREE:%.*]] |
| ; CHECK-NEXT: i32 4, label [[FOUR:%.*]] |
| ; CHECK-NEXT: ] |
| ; CHECK: common.ret: |
| ; CHECK-NEXT: ret void |
| ; CHECK: one: |
| ; CHECK-NEXT: call void @side_effect(i32 32) |
| ; CHECK-NEXT: br label [[COMMON_RET]] |
| ; CHECK: two: |
| ; CHECK-NEXT: call void @side_effect(i32 64) |
| ; CHECK-NEXT: br label [[COMMON_RET]] |
| ; CHECK: three: |
| ; CHECK-NEXT: call void @side_effect(i32 96) |
| ; CHECK-NEXT: br label [[COMMON_RET]] |
| ; CHECK: four: |
| ; CHECK-NEXT: call void @side_effect(i32 128) |
| ; CHECK-NEXT: br label [[COMMON_RET]] |
| ; |
| switch i32 %a, label %def [ |
| i32 32, label %one |
| i32 64, label %two |
| i32 96, label %three |
| i32 128, label %four |
| ] |
| |
| def: |
| ret void |
| |
| one: |
| call void @side_effect(i32 32) |
| ret void |
| two: |
| call void @side_effect(i32 64) |
| ret void |
| three: |
| call void @side_effect(i32 96) |
| ret void |
| four: |
| call void @side_effect(i32 128) |
| ret void |
| } |
| |
| ; Base=0 is not dense after shifting, so this should still subtract the local |
| ; minimum before reducing the range. |
| define void @test11(i32 %a) { |
| ; CHECK-LABEL: @test11( |
| ; CHECK-NEXT: [[TMP1:%.*]] = sub i32 [[A:%.*]], 40 |
| ; CHECK-NEXT: [[TMP2:%.*]] = call i32 @llvm.fshl.i32(i32 [[TMP1]], i32 [[TMP1]], i32 27) |
| ; CHECK-NEXT: switch i32 [[TMP2]], label [[COMMON_RET:%.*]] [ |
| ; CHECK-NEXT: i32 0, label [[ONE:%.*]] |
| ; CHECK-NEXT: i32 1, label [[TWO:%.*]] |
| ; CHECK-NEXT: i32 2, label [[THREE:%.*]] |
| ; CHECK-NEXT: i32 3, label [[FOUR:%.*]] |
| ; CHECK-NEXT: ] |
| ; CHECK: common.ret: |
| ; CHECK-NEXT: ret void |
| ; CHECK: one: |
| ; CHECK-NEXT: call void @side_effect(i32 40) |
| ; CHECK-NEXT: br label [[COMMON_RET]] |
| ; CHECK: two: |
| ; CHECK-NEXT: call void @side_effect(i32 72) |
| ; CHECK-NEXT: br label [[COMMON_RET]] |
| ; CHECK: three: |
| ; CHECK-NEXT: call void @side_effect(i32 104) |
| ; CHECK-NEXT: br label [[COMMON_RET]] |
| ; CHECK: four: |
| ; CHECK-NEXT: call void @side_effect(i32 136) |
| ; CHECK-NEXT: br label [[COMMON_RET]] |
| ; |
| switch i32 %a, label %def [ |
| i32 40, label %one |
| i32 72, label %two |
| i32 104, label %three |
| i32 136, label %four |
| ] |
| |
| def: |
| ret void |
| |
| one: |
| call void @side_effect(i32 40) |
| ret void |
| two: |
| call void @side_effect(i32 72) |
| ret void |
| three: |
| call void @side_effect(i32 104) |
| ret void |
| four: |
| call void @side_effect(i32 136) |
| ret void |
| } |
| |
| declare void @side_effect(i32) |
| |
| !0 = !{!"function_entry_count", i32 100} |
| !1 = !{!"branch_weights", i32 5, i32 7, i32 11, i32 13, i32 17} |
| ;. |
| ; CHECK: attributes #[[ATTR0:[0-9]+]] = { optsize } |
| ; CHECK: attributes #[[ATTR1:[0-9]+]] = { nocallback nocreateundeforpoison nofree nosync nounwind speculatable willreturn memory(none) } |
| ;. |
| ; CHECK: [[META0:![0-9]+]] = !{!"function_entry_count", i32 100} |
| ; CHECK: [[PROF1]] = !{!"branch_weights", i32 48, i32 5} |
| ;. |