caio/clang-p2996
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
Files
be1294de9dd93ae02002bd1aa4b4a8d62e562c21
clang-p2996/llvm/test/Analysis/BasicAA/atomic-memory-intrinsics.ll
Nikita Popov 57d57b1afd [AAEval] Make compatible with opaque pointers 
With opaque pointers, we cannot use the pointer element type to
determine the LocationSize for the AA query. Instead, -aa-eval
tests are now required to have an explicit load or store for any
pointer they want to compute alias results for, and the load/store
types are used to determine the location size.

This may affect ordering of results, and sorting within one result,
as the type is not considered part of the sorted string anymore.

To somewhat minimize the churn, printing still uses faux typed
pointer notation.
2022-03-16 10:02:11 +01:00

136 lines
8.3 KiB
LLVM
Raw Blame History

; RUN: opt -aa-pipeline=basic-aa -passes=aa-eval -print-all-alias-modref-info -disable-output 2>&1 %s | FileCheck %s
declare void @llvm.memset.element.unordered.atomic.p0i8.i32(i8*, i8, i64, i32)
define void @test_memset_element_unordered_atomic_const_size(i8* noalias %a) {
; CHECK-LABEL: Function: test_memset_element_unordered_atomic_const_size
; CHECK: Just Mod (MustAlias): Ptr: i8* %a <-> call void @llvm.memset.element.unordered.atomic.p0i8.i64(i8* align 1 %a, i8 0, i64 4, i32 1)
; CHECK-NEXT: Just Mod: Ptr: i8* %a.gep.1 <-> call void @llvm.memset.element.unordered.atomic.p0i8.i64(i8* align 1 %a, i8 0, i64 4, i32 1)
; CHECK-NEXT: NoModRef: Ptr: i8* %a.gep.5 <-> call void @llvm.memset.element.unordered.atomic.p0i8.i64(i8* align 1 %a, i8 0, i64 4, i32 1)
;
entry:
load i8, i8* %a
call void @llvm.memset.element.unordered.atomic.p0i8.i32(i8* align 1 %a, i8 0, i64 4, i32 1)
%a.gep.1 = getelementptr i8, i8* %a, i32 1
store i8 0, i8* %a.gep.1
%a.gep.5 = getelementptr i8, i8* %a, i32 5
store i8 1, i8* %a.gep.5
ret void
}
define void @test_memset_element_unordered_atomic_variable_size(i8* noalias %a, i64 %n) {
; CHECK-LABEL: Function: test_memset_element_unordered_atomic_variable_size
; CHECK: Just Mod (MustAlias): Ptr: i8* %a <-> call void @llvm.memset.element.unordered.atomic.p0i8.i64(i8* align 1 %a, i8 0, i64 %n, i32 1)
; CHECK-NEXT: Just Mod: Ptr: i8* %a.gep.1 <-> call void @llvm.memset.element.unordered.atomic.p0i8.i64(i8* align 1 %a, i8 0, i64 %n, i32 1)
; CHECK-NEXT: Just Mod: Ptr: i8* %a.gep.5 <-> call void @llvm.memset.element.unordered.atomic.p0i8.i64(i8* align 1 %a, i8 0, i64 %n, i32 1)
;
entry:
load i8, i8* %a
call void @llvm.memset.element.unordered.atomic.p0i8.i32(i8* align 1 %a, i8 0, i64 %n, i32 1)
%a.gep.1 = getelementptr i8, i8* %a, i32 1
store i8 0, i8* %a.gep.1
%a.gep.5 = getelementptr i8, i8* %a, i32 5
store i8 1, i8* %a.gep.5
ret void
}
declare void @llvm.memcpy.element.unordered.atomic.p0i8.p0i8.i64(i8* nocapture writeonly, i8* nocapture readonly, i64, i32)
define void @test_memcpy_element_unordered_atomic_const_size(i8* noalias %a, i8* noalias %b) {
; CHECK-LABEL: Function: test_memcpy_element_unordered_atomic_const_size
; CHECK: Just Ref: Ptr: i8* %a <-> call void @llvm.memcpy.element.unordered.atomic.p0i8.p0i8.i64(i8* align 1 %b, i8* align 1 %a, i64 4, i32 1)
; CHECK-NEXT: Just Mod: Ptr: i8* %b <-> call void @llvm.memcpy.element.unordered.atomic.p0i8.p0i8.i64(i8* align 1 %b, i8* align 1 %a, i64 4, i32 1)
; CHECK-NEXT: Just Ref: Ptr: i8* %a.gep.1 <-> call void @llvm.memcpy.element.unordered.atomic.p0i8.p0i8.i64(i8* align 1 %b, i8* align 1 %a, i64 4, i32 1)
; CHECK-NEXT: NoModRef: Ptr: i8* %a.gep.5 <-> call void @llvm.memcpy.element.unordered.atomic.p0i8.p0i8.i64(i8* align 1 %b, i8* align 1 %a, i64 4, i32 1)
; CHECK-NEXT: Just Mod: Ptr: i8* %b.gep.1 <-> call void @llvm.memcpy.element.unordered.atomic.p0i8.p0i8.i64(i8* align 1 %b, i8* align 1 %a, i64 4, i32 1)
; CHECK-NEXT: NoModRef: Ptr: i8* %b.gep.5 <-> call void @llvm.memcpy.element.unordered.atomic.p0i8.p0i8.i64(i8* align 1 %b, i8* align 1 %a, i64 4, i32 1)
;
entry:
load i8, i8* %a
load i8, i8* %b
%a.gep.1 = getelementptr i8, i8* %a, i32 1
store i8 0, i8* %a.gep.1
%a.gep.5 = getelementptr i8, i8* %a, i32 5
store i8 1, i8* %a.gep.5
call void @llvm.memcpy.element.unordered.atomic.p0i8.p0i8.i64(i8* align 1 %b, i8* align 1 %a, i64 4, i32 1)
%b.gep.1 = getelementptr i8, i8* %b, i32 1
store i8 0, i8* %b.gep.1
%b.gep.5 = getelementptr i8, i8* %b, i32 5
store i8 1, i8* %b.gep.5
ret void
}
define void @test_memcpy_element_unordered_atomic_variable_size(i8* noalias %a, i8* noalias %b, i64 %n) {
; CHECK-LABEL: Function: test_memcpy_element_unordered_atomic_variable_size
; CHECK: Just Ref: Ptr: i8* %a <-> call void @llvm.memcpy.element.unordered.atomic.p0i8.p0i8.i64(i8* align 1 %b, i8* align 1 %a, i64 %n, i32 1)
; CHECK-NEXT: Just Mod: Ptr: i8* %b <-> call void @llvm.memcpy.element.unordered.atomic.p0i8.p0i8.i64(i8* align 1 %b, i8* align 1 %a, i64 %n, i32 1)
; CHECK-NEXT: Just Ref: Ptr: i8* %a.gep.1 <-> call void @llvm.memcpy.element.unordered.atomic.p0i8.p0i8.i64(i8* align 1 %b, i8* align 1 %a, i64 %n, i32 1)
; CHECK-NEXT: Just Ref: Ptr: i8* %a.gep.5 <-> call void @llvm.memcpy.element.unordered.atomic.p0i8.p0i8.i64(i8* align 1 %b, i8* align 1 %a, i64 %n, i32 1)
; CHECK-NEXT: Just Mod: Ptr: i8* %b.gep.1 <-> call void @llvm.memcpy.element.unordered.atomic.p0i8.p0i8.i64(i8* align 1 %b, i8* align 1 %a, i64 %n, i32 1)
; CHECK-NEXT: Just Mod: Ptr: i8* %b.gep.5 <-> call void @llvm.memcpy.element.unordered.atomic.p0i8.p0i8.i64(i8* align 1 %b, i8* align 1 %a, i64 %n, i32 1)
;
entry:
load i8, i8* %a
load i8, i8* %b
%a.gep.1 = getelementptr i8, i8* %a, i32 1
store i8 0, i8* %a.gep.1
%a.gep.5 = getelementptr i8, i8* %a, i32 5
store i8 1, i8* %a.gep.5
call void @llvm.memcpy.element.unordered.atomic.p0i8.p0i8.i64(i8* align 1 %b, i8* align 1 %a, i64 %n, i32 1)
%b.gep.1 = getelementptr i8, i8* %b, i32 1
store i8 0, i8* %b.gep.1
%b.gep.5 = getelementptr i8, i8* %b, i32 5
store i8 1, i8* %b.gep.5
ret void
}
declare void @llvm.memmove.element.unordered.atomic.p0i8.p0i8.i64(i8* nocapture writeonly, i8* nocapture readonly, i64, i32)
define void @test_memmove_element_unordered_atomic_const_size(i8* noalias %a, i8* noalias %b) {
; CHECK-LABEL: Function: test_memmove_element_unordered_atomic_const_size
; CHECK: Just Ref: Ptr: i8* %a <-> call void @llvm.memmove.element.unordered.atomic.p0i8.p0i8.i64(i8* align 1 %b, i8* align 1 %a, i64 4, i32 1)
; CHECK-NEXT: Just Mod: Ptr: i8* %b <-> call void @llvm.memmove.element.unordered.atomic.p0i8.p0i8.i64(i8* align 1 %b, i8* align 1 %a, i64 4, i32 1)
; CHECK-NEXT: Just Ref: Ptr: i8* %a.gep.1 <-> call void @llvm.memmove.element.unordered.atomic.p0i8.p0i8.i64(i8* align 1 %b, i8* align 1 %a, i64 4, i32 1)
; CHECK-NEXT: NoModRef: Ptr: i8* %a.gep.5 <-> call void @llvm.memmove.element.unordered.atomic.p0i8.p0i8.i64(i8* align 1 %b, i8* align 1 %a, i64 4, i32 1)
; CHECK-NEXT: Just Mod: Ptr: i8* %b.gep.1 <-> call void @llvm.memmove.element.unordered.atomic.p0i8.p0i8.i64(i8* align 1 %b, i8* align 1 %a, i64 4, i32 1)
; CHECK-NEXT: NoModRef: Ptr: i8* %b.gep.5 <-> call void @llvm.memmove.element.unordered.atomic.p0i8.p0i8.i64(i8* align 1 %b, i8* align 1 %a, i64 4, i32 1)
;
entry:
load i8, i8* %a
load i8, i8* %b
%a.gep.1 = getelementptr i8, i8* %a, i32 1
store i8 0, i8* %a.gep.1
%a.gep.5 = getelementptr i8, i8* %a, i32 5
store i8 1, i8* %a.gep.5
call void @llvm.memmove.element.unordered.atomic.p0i8.p0i8.i64(i8* align 1 %b, i8* align 1 %a, i64 4, i32 1)
%b.gep.1 = getelementptr i8, i8* %b, i32 1
store i8 0, i8* %b.gep.1
%b.gep.5 = getelementptr i8, i8* %b, i32 5
store i8 1, i8* %b.gep.5
ret void
}
define void @test_memmove_element_unordered_atomic_variable_size(i8* noalias %a, i8* noalias %b, i64 %n) {
; CHECK-LABEL: Function: test_memmove_element_unordered_atomic_variable_size
; CHECK: Just Ref: Ptr: i8* %a <-> call void @llvm.memmove.element.unordered.atomic.p0i8.p0i8.i64(i8* align 1 %b, i8* align 1 %a, i64 %n, i32 1)
; CHECK-NEXT: Just Mod: Ptr: i8* %b <-> call void @llvm.memmove.element.unordered.atomic.p0i8.p0i8.i64(i8* align 1 %b, i8* align 1 %a, i64 %n, i32 1)
; CHECK-NEXT: Just Ref: Ptr: i8* %a.gep.1 <-> call void @llvm.memmove.element.unordered.atomic.p0i8.p0i8.i64(i8* align 1 %b, i8* align 1 %a, i64 %n, i32 1)
; CHECK-NEXT: Just Ref: Ptr: i8* %a.gep.5 <-> call void @llvm.memmove.element.unordered.atomic.p0i8.p0i8.i64(i8* align 1 %b, i8* align 1 %a, i64 %n, i32 1)
; CHECK-NEXT: Just Mod: Ptr: i8* %b.gep.1 <-> call void @llvm.memmove.element.unordered.atomic.p0i8.p0i8.i64(i8* align 1 %b, i8* align 1 %a, i64 %n, i32 1)
; CHECK-NEXT: Just Mod: Ptr: i8* %b.gep.5 <-> call void @llvm.memmove.element.unordered.atomic.p0i8.p0i8.i64(i8* align 1 %b, i8* align 1 %a, i64 %n, i32 1)
;
entry:
load i8, i8* %a
load i8, i8* %b
%a.gep.1 = getelementptr i8, i8* %a, i32 1
store i8 0, i8* %a.gep.1
%a.gep.5 = getelementptr i8, i8* %a, i32 5
store i8 1, i8* %a.gep.5
call void @llvm.memmove.element.unordered.atomic.p0i8.p0i8.i64(i8* align 1 %b, i8* align 1 %a, i64 %n, i32 1)
%b.gep.1 = getelementptr i8, i8* %b, i32 1
store i8 0, i8* %b.gep.1
%b.gep.5 = getelementptr i8, i8* %b, i32 5
store i8 1, i8* %b.gep.5
ret void
}
Reference in New Issue View Git Blame Copy Permalink