This patch canonicalizes getelementptr instructions with constant indices to use the `i8` source element type. This makes it easier for optimizations to recognize that two GEPs are identical, because they don't need to see past many different ways to express the same offset. This is a first step towards https://discourse.llvm.org/t/rfc-replacing-getelementptr-with-ptradd/68699. This is limited to constant GEPs only for now, as they have a clear canonical form, while we're not yet sure how exactly to deal with variable indices. The test llvm/test/Transforms/PhaseOrdering/switch_with_geps.ll gives two representative examples of the kind of optimization improvement we expect from this change. In the first test SimplifyCFG can now realize that all switch branches are actually the same. In the second test it can convert it into simple arithmetic. These are representative of common optimization failures we see in Rust. Fixes https://github.com/llvm/llvm-project/issues/69841.
194 lines
7.7 KiB
LLVM
194 lines
7.7 KiB
LLVM
; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
|
|
; This testcase tests for various features the basicaa test should be able to
|
|
; determine, as noted in the comments.
|
|
|
|
; RUN: opt < %s -aa-pipeline=basic-aa -passes=gvn,instcombine,dce -S | FileCheck %s --check-prefixes=CHECK,NO_ASSUME
|
|
; RUN: opt < %s -aa-pipeline=basic-aa -passes=gvn,instcombine,dce --enable-knowledge-retention -S | FileCheck %s --check-prefixes=CHECK,USE_ASSUME
|
|
target datalayout = "E-p:64:64:64-a0:0:8-f32:32:32-f64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:32:64-v64:64:64-v128:128:128"
|
|
|
|
@Global = external global { i32 }
|
|
|
|
declare void @external(ptr)
|
|
declare void @llvm.assume(i1)
|
|
|
|
; Array test: Test that operations on one local array do not invalidate
|
|
; operations on another array. Important for scientific codes.
|
|
;
|
|
define i32 @different_array_test(i64 %A, i64 %B) {
|
|
; NO_ASSUME-LABEL: @different_array_test(
|
|
; NO_ASSUME-NEXT: [[ARRAY11:%.*]] = alloca [100 x i32], align 4
|
|
; NO_ASSUME-NEXT: [[ARRAY22:%.*]] = alloca [200 x i32], align 4
|
|
; NO_ASSUME-NEXT: call void @llvm.assume(i1 true) [ "align"(ptr [[ARRAY11]], i32 4) ]
|
|
; NO_ASSUME-NEXT: call void @external(ptr nonnull [[ARRAY11]])
|
|
; NO_ASSUME-NEXT: call void @external(ptr nonnull [[ARRAY22]])
|
|
; NO_ASSUME-NEXT: [[POINTER2:%.*]] = getelementptr i32, ptr [[ARRAY22]], i64 [[B:%.*]]
|
|
; NO_ASSUME-NEXT: store i32 7, ptr [[POINTER2]], align 4
|
|
; NO_ASSUME-NEXT: ret i32 0
|
|
;
|
|
; USE_ASSUME-LABEL: @different_array_test(
|
|
; USE_ASSUME-NEXT: [[ARRAY11:%.*]] = alloca [100 x i32], align 4
|
|
; USE_ASSUME-NEXT: [[ARRAY22:%.*]] = alloca [200 x i32], align 4
|
|
; USE_ASSUME-NEXT: call void @external(ptr nonnull [[ARRAY11]])
|
|
; USE_ASSUME-NEXT: call void @external(ptr nonnull [[ARRAY22]])
|
|
; USE_ASSUME-NEXT: [[POINTER2:%.*]] = getelementptr i32, ptr [[ARRAY22]], i64 [[B:%.*]]
|
|
; USE_ASSUME-NEXT: store i32 7, ptr [[POINTER2]], align 4
|
|
; USE_ASSUME-NEXT: ret i32 0
|
|
;
|
|
%Array1 = alloca i32, i32 100
|
|
%Array2 = alloca i32, i32 200
|
|
call void @llvm.assume(i1 true) ["align"(ptr %Array1, i32 4)]
|
|
|
|
call void @external(ptr %Array1)
|
|
call void @external(ptr %Array2)
|
|
|
|
%pointer = getelementptr i32, ptr %Array1, i64 %A
|
|
%val = load i32, ptr %pointer
|
|
|
|
%pointer2 = getelementptr i32, ptr %Array2, i64 %B
|
|
store i32 7, ptr %pointer2
|
|
|
|
%REMOVE = load i32, ptr %pointer ; redundant with above load
|
|
%retval = sub i32 %REMOVE, %val
|
|
ret i32 %retval
|
|
}
|
|
|
|
; Constant index test: Constant indexes into the same array should not
|
|
; interfere with each other. Again, important for scientific codes.
|
|
;
|
|
define i32 @constant_array_index_test() {
|
|
; CHECK-LABEL: @constant_array_index_test(
|
|
; CHECK-NEXT: [[ARRAY1:%.*]] = alloca [100 x i32], align 4
|
|
; CHECK-NEXT: call void @external(ptr nonnull [[ARRAY1]])
|
|
; CHECK-NEXT: [[P2:%.*]] = getelementptr inbounds i8, ptr [[ARRAY1]], i64 24
|
|
; CHECK-NEXT: store i32 1, ptr [[P2]], align 4
|
|
; CHECK-NEXT: ret i32 0
|
|
;
|
|
%Array = alloca i32, i32 100
|
|
call void @external(ptr %Array)
|
|
|
|
%P1 = getelementptr i32, ptr %Array, i64 7
|
|
%P2 = getelementptr i32, ptr %Array, i64 6
|
|
|
|
%A = load i32, ptr %P1
|
|
store i32 1, ptr %P2 ; Should not invalidate load
|
|
%BREMOVE = load i32, ptr %P1
|
|
%Val = sub i32 %A, %BREMOVE
|
|
ret i32 %Val
|
|
}
|
|
|
|
; Test that if two pointers are spaced out by a constant getelementptr, that
|
|
; they cannot alias.
|
|
define i32 @gep_distance_test(ptr %A) {
|
|
; NO_ASSUME-LABEL: @gep_distance_test(
|
|
; NO_ASSUME-NEXT: [[B:%.*]] = getelementptr i8, ptr [[A:%.*]], i64 8
|
|
; NO_ASSUME-NEXT: store i32 7, ptr [[B]], align 4
|
|
; NO_ASSUME-NEXT: ret i32 0
|
|
;
|
|
; USE_ASSUME-LABEL: @gep_distance_test(
|
|
; USE_ASSUME-NEXT: [[B:%.*]] = getelementptr i8, ptr [[A:%.*]], i64 8
|
|
; USE_ASSUME-NEXT: store i32 7, ptr [[B]], align 4
|
|
; USE_ASSUME-NEXT: call void @llvm.assume(i1 true) [ "dereferenceable"(ptr [[A]], i64 4), "nonnull"(ptr [[A]]), "align"(ptr [[A]], i64 4) ]
|
|
; USE_ASSUME-NEXT: ret i32 0
|
|
;
|
|
%REMOVEu = load i32, ptr %A
|
|
%B = getelementptr i32, ptr %A, i64 2 ; Cannot alias A
|
|
store i32 7, ptr %B
|
|
%REMOVEv = load i32, ptr %A
|
|
%r = sub i32 %REMOVEu, %REMOVEv
|
|
ret i32 %r
|
|
}
|
|
|
|
; Test that if two pointers are spaced out by a constant offset, that they
|
|
; cannot alias, even if there is a variable offset between them...
|
|
define i32 @gep_distance_test2(ptr %A, i64 %distance) {
|
|
; NO_ASSUME-LABEL: @gep_distance_test2(
|
|
; NO_ASSUME-NEXT: [[B:%.*]] = getelementptr { i32, i32 }, ptr [[A:%.*]], i64 [[DISTANCE:%.*]], i32 1
|
|
; NO_ASSUME-NEXT: store i32 7, ptr [[B]], align 4
|
|
; NO_ASSUME-NEXT: ret i32 0
|
|
;
|
|
; USE_ASSUME-LABEL: @gep_distance_test2(
|
|
; USE_ASSUME-NEXT: [[B:%.*]] = getelementptr { i32, i32 }, ptr [[A:%.*]], i64 [[DISTANCE:%.*]], i32 1
|
|
; USE_ASSUME-NEXT: store i32 7, ptr [[B]], align 4
|
|
; USE_ASSUME-NEXT: call void @llvm.assume(i1 true) [ "dereferenceable"(ptr [[A]], i64 4), "nonnull"(ptr [[A]]), "align"(ptr [[A]], i64 4) ]
|
|
; USE_ASSUME-NEXT: ret i32 0
|
|
;
|
|
%A1 = getelementptr {i32,i32}, ptr %A, i64 0, i32 0
|
|
%REMOVEu = load i32, ptr %A1
|
|
%B = getelementptr {i32,i32}, ptr %A, i64 %distance, i32 1
|
|
store i32 7, ptr %B ; B cannot alias A, it's at least 4 bytes away
|
|
%REMOVEv = load i32, ptr %A1
|
|
%r = sub i32 %REMOVEu, %REMOVEv
|
|
ret i32 %r
|
|
}
|
|
|
|
; Test that we can do funny pointer things and that distance calc will still
|
|
; work.
|
|
define i32 @gep_distance_test3(ptr %A) {
|
|
; NO_ASSUME-LABEL: @gep_distance_test3(
|
|
; NO_ASSUME-NEXT: [[C:%.*]] = getelementptr i8, ptr [[A:%.*]], i64 4
|
|
; NO_ASSUME-NEXT: store i8 42, ptr [[C]], align 1
|
|
; NO_ASSUME-NEXT: ret i32 0
|
|
;
|
|
; USE_ASSUME-LABEL: @gep_distance_test3(
|
|
; USE_ASSUME-NEXT: [[C:%.*]] = getelementptr i8, ptr [[A:%.*]], i64 4
|
|
; USE_ASSUME-NEXT: store i8 42, ptr [[C]], align 1
|
|
; USE_ASSUME-NEXT: call void @llvm.assume(i1 true) [ "dereferenceable"(ptr [[A]], i64 4), "nonnull"(ptr [[A]]), "align"(ptr [[A]], i64 4) ]
|
|
; USE_ASSUME-NEXT: ret i32 0
|
|
;
|
|
%X = load i32, ptr %A
|
|
%C = getelementptr i8, ptr %A, i64 4
|
|
store i8 42, ptr %C
|
|
%Y = load i32, ptr %A
|
|
%R = sub i32 %X, %Y
|
|
ret i32 %R
|
|
}
|
|
|
|
; Test that we can disambiguate globals reached through constantexpr geps
|
|
define i32 @constexpr_test() {
|
|
; CHECK-LABEL: @constexpr_test(
|
|
; CHECK-NEXT: [[X:%.*]] = alloca i32, align 4
|
|
; CHECK-NEXT: call void @external(ptr nonnull [[X]])
|
|
; CHECK-NEXT: store i32 5, ptr @Global, align 4
|
|
; CHECK-NEXT: ret i32 0
|
|
;
|
|
%X = alloca i32
|
|
call void @external(ptr %X)
|
|
|
|
%Y = load i32, ptr %X
|
|
store i32 5, ptr @Global
|
|
%REMOVE = load i32, ptr %X
|
|
%retval = sub i32 %Y, %REMOVE
|
|
ret i32 %retval
|
|
}
|
|
|
|
|
|
|
|
; PR7589
|
|
; These two index expressions are different, this cannot be CSE'd.
|
|
define i16 @zext_sext_confusion(ptr %row2col, i5 %j) nounwind{
|
|
; CHECK-LABEL: @zext_sext_confusion(
|
|
; CHECK-NEXT: entry:
|
|
; CHECK-NEXT: [[SUM5_CAST:%.*]] = zext i5 [[J:%.*]] to i64
|
|
; CHECK-NEXT: [[P1:%.*]] = getelementptr i16, ptr [[ROW2COL:%.*]], i64 [[SUM5_CAST]]
|
|
; CHECK-NEXT: [[ROW2COL_LOAD_1_2:%.*]] = load i16, ptr [[P1]], align 1
|
|
; CHECK-NEXT: [[SUM13_CAST31:%.*]] = sext i5 [[J]] to i6
|
|
; CHECK-NEXT: [[SUM13_CAST:%.*]] = zext i6 [[SUM13_CAST31]] to i64
|
|
; CHECK-NEXT: [[P2:%.*]] = getelementptr i16, ptr [[ROW2COL]], i64 [[SUM13_CAST]]
|
|
; CHECK-NEXT: [[ROW2COL_LOAD_1_6:%.*]] = load i16, ptr [[P2]], align 1
|
|
; CHECK-NEXT: [[DOTRET:%.*]] = sub i16 [[ROW2COL_LOAD_1_6]], [[ROW2COL_LOAD_1_2]]
|
|
; CHECK-NEXT: ret i16 [[DOTRET]]
|
|
;
|
|
entry:
|
|
%sum5.cast = zext i5 %j to i64 ; <i64> [#uses=1]
|
|
%P1 = getelementptr i16, ptr %row2col, i64 %sum5.cast
|
|
%row2col.load.1.2 = load i16, ptr %P1, align 1 ; <i16> [#uses=1]
|
|
|
|
%sum13.cast31 = sext i5 %j to i6 ; <i6> [#uses=1]
|
|
%sum13.cast = zext i6 %sum13.cast31 to i64 ; <i64> [#uses=1]
|
|
%P2 = getelementptr i16, ptr %row2col, i64 %sum13.cast
|
|
%row2col.load.1.6 = load i16, ptr %P2, align 1 ; <i16> [#uses=1]
|
|
|
|
%.ret = sub i16 %row2col.load.1.6, %row2col.load.1.2 ; <i16> [#uses=1]
|
|
ret i16 %.ret
|
|
}
|