caio/clang-p2996
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
Files
aef0877b1ba5f14eaeece956aa69df6c208b94de
clang-p2996/clang/test/CodeGenCXX/cxx11-initializer-array-new.cpp
Roman Lebedev 3dd5a298bf [clang] Annotating C++'s operator new with more attributes 
Summary:
Right now we annotate C++'s `operator new` with `noalias` attribute,
which very much is healthy for optimizations.

However as per [[ http://eel.is/c++draft/basic.stc.dynamic.allocation | `[basic.stc.dynamic.allocation]` ]],
there are more promises on global `operator new`, namely:
* non-`std::nothrow_t` `operator new` *never* returns `nullptr`
* If `std::align_val_t align` parameter is taken, the pointer will also be `align`-aligned
* ~~global `operator new`-returned pointer is `__STDCPP_DEFAULT_NEW_ALIGNMENT__`-aligned ~~ It's more caveated than that.

Supplying this information may not cause immediate landslide effects
on any specific benchmarks, but it for sure will be healthy for optimizer
in the sense that the IR will better reflect the guarantees provided in the source code.

The caveat is `-fno-assume-sane-operator-new`, which currently prevents emitting `noalias`
attribute, and is automatically passed by Sanitizers ([[ https://bugs.llvm.org/show_bug.cgi?id=16386 | PR16386 ]]) - should it also cover these attributes?
The problem is that the flag is back-end-specific, as seen in `test/Modules/explicit-build-flags.cpp`.
But while it is okay to add `noalias` metadata in backend, we really should be adding at least
the alignment metadata to the AST, since that allows us to perform sema checks on it.

Reviewers: erichkeane, rjmccall, jdoerfert, eugenis, rsmith

Reviewed By: rsmith

Subscribers: xbolva00, jrtc27, atanasyan, nlopes, cfe-commits

Tags: #llvm, #clang

Differential Revision: https://reviews.llvm.org/D73380
2020-02-26 01:37:17 +03:00

160 lines
6.4 KiB
C++
Raw Blame History

// RUN: %clang_cc1 -triple x86_64-linux-gnu -std=c++11 %s -emit-llvm -o - | FileCheck %s
// PR10878
struct S { S(); S(int); ~S(); int n; };
void *p = new S[2][3]{ { 1, 2, 3 }, { 4, 5, 6 } };
// CHECK-LABEL: define
// CHECK: %[[ALLOC:.*]] = call noalias nonnull i8* @_Znam(i64 32)
// CHECK: %[[COOKIE:.*]] = bitcast i8* %[[ALLOC]] to i64*
// CHECK: store i64 6, i64* %[[COOKIE]]
// CHECK: %[[START_AS_i8:.*]] = getelementptr inbounds i8, i8* %[[ALLOC]], i64 8
// CHECK: %[[START_AS_S:.*]] = bitcast i8* %[[START_AS_i8]] to %[[S:.*]]*
//
// Explicit initializers:
//
// { 1, 2, 3 }
//
// CHECK: %[[S_0:.*]] = bitcast %[[S]]* %[[START_AS_S]] to [3 x %[[S]]]*
//
// CHECK: %[[S_0_0:.*]] = getelementptr inbounds [3 x %[[S]]], [3 x %[[S]]]* %[[S_0]], i64 0, i64 0
// CHECK: call void @_ZN1SC1Ei(%[[S]]* %[[S_0_0]], i32 1)
// CHECK: %[[S_0_1:.*]] = getelementptr inbounds %[[S]], %[[S]]* %[[S_0_0]], i64 1
// CHECK: call void @_ZN1SC1Ei(%[[S]]* %[[S_0_1]], i32 2)
// CHECK: %[[S_0_2:.*]] = getelementptr inbounds %[[S]], %[[S]]* %[[S_0_1]], i64 1
// CHECK: call void @_ZN1SC1Ei(%[[S]]* %[[S_0_2]], i32 3)
//
// { 4, 5, 6 }
//
// CHECK: %[[S_1:.*]] = getelementptr inbounds [3 x %[[S]]], [3 x %[[S]]]* %[[S_0]], i64 1
//
// CHECK: %[[S_1_0:.*]] = getelementptr inbounds [3 x %[[S]]], [3 x %[[S]]]* %[[S_1]], i64 0, i64 0
// CHECK: call void @_ZN1SC1Ei(%[[S]]* %[[S_1_0]], i32 4)
// CHECK: %[[S_1_1:.*]] = getelementptr inbounds %[[S]], %[[S]]* %[[S_1_0]], i64 1
// CHECK: call void @_ZN1SC1Ei(%[[S]]* %[[S_1_1]], i32 5)
// CHECK: %[[S_1_2:.*]] = getelementptr inbounds %[[S]], %[[S]]* %[[S_1_1]], i64 1
// CHECK: call void @_ZN1SC1Ei(%[[S]]* %[[S_1_2]], i32 6)
//
// CHECK-NOT: br i1
// CHECK-NOT: call
// CHECK: }
int n;
void *q = new S[n][3]{ { 1, 2, 3 }, { 4, 5, 6 } };
// CHECK-LABEL: define
//
// CHECK: load i32, i32* @n
// CHECK: call {{.*}} @llvm.umul.with.overflow.i64(i64 %[[N:.*]], i64 12)
// CHECK: %[[ELTS:.*]] = mul i64 %[[N]], 3
// CHECK: call {{.*}} @llvm.uadd.with.overflow.i64(i64 %{{.*}}, i64 8)
// CHECK: %[[ALLOC:.*]] = call noalias nonnull i8* @_Znam(i64 %{{.*}})
//
// CHECK: %[[COOKIE:.*]] = bitcast i8* %[[ALLOC]] to i64*
// CHECK: store i64 %[[ELTS]], i64* %[[COOKIE]]
// CHECK: %[[START_AS_i8:.*]] = getelementptr inbounds i8, i8* %[[ALLOC]], i64 8
// CHECK: %[[START_AS_S:.*]] = bitcast i8* %[[START_AS_i8]] to %[[S]]*
//
// Explicit initializers:
//
// { 1, 2, 3 }
//
// CHECK: %[[S_0:.*]] = bitcast %[[S]]* %[[START_AS_S]] to [3 x %[[S]]]*
//
// CHECK: %[[S_0_0:.*]] = getelementptr inbounds [3 x %[[S]]], [3 x %[[S]]]* %[[S_0]], i64 0, i64 0
// CHECK: call void @_ZN1SC1Ei(%[[S]]* %[[S_0_0]], i32 1)
// CHECK: %[[S_0_1:.*]] = getelementptr inbounds %[[S]], %[[S]]* %[[S_0_0]], i64 1
// CHECK: call void @_ZN1SC1Ei(%[[S]]* %[[S_0_1]], i32 2)
// CHECK: %[[S_0_2:.*]] = getelementptr inbounds %[[S]], %[[S]]* %[[S_0_1]], i64 1
// CHECK: call void @_ZN1SC1Ei(%[[S]]* %[[S_0_2]], i32 3)
//
// { 4, 5, 6 }
//
// CHECK: %[[S_1:.*]] = getelementptr inbounds [3 x %[[S]]], [3 x %[[S]]]* %[[S_0]], i64 1
//
// CHECK: %[[S_1_0:.*]] = getelementptr inbounds [3 x %[[S]]], [3 x %[[S]]]* %[[S_1]], i64 0, i64 0
// CHECK: call void @_ZN1SC1Ei(%[[S]]* %[[S_1_0]], i32 4)
// CHECK: %[[S_1_1:.*]] = getelementptr inbounds %[[S]], %[[S]]* %[[S_1_0]], i64 1
// CHECK: call void @_ZN1SC1Ei(%[[S]]* %[[S_1_1]], i32 5)
// CHECK: %[[S_1_2:.*]] = getelementptr inbounds %[[S]], %[[S]]* %[[S_1_1]], i64 1
// CHECK: call void @_ZN1SC1Ei(%[[S]]* %[[S_1_2]], i32 6)
//
// And the rest.
//
// CHECK: %[[S_2:.*]] = getelementptr inbounds [3 x %[[S]]], [3 x %[[S]]]* %[[S_1]], i64 1
// CHECK: %[[S_2_AS_S:.*]] = bitcast [3 x %[[S]]]* %[[S_2]] to %[[S]]*
//
// CHECK: %[[REST:.*]] = sub i64 %[[ELTS]], 6
// CHECK: icmp eq i64 %[[REST]], 0
// CHECK: br i1
//
// CHECK: %[[END:.*]] = getelementptr inbounds %[[S]], %[[S]]* %[[S_2_AS_S]], i64 %[[REST]]
// CHECK: br label
//
// CHECK: %[[CUR:.*]] = phi %[[S]]* [ %[[S_2_AS_S]], {{.*}} ], [ %[[NEXT:.*]], {{.*}} ]
// CHECK: call void @_ZN1SC1Ev(%[[S]]* %[[CUR]])
// CHECK: %[[NEXT]] = getelementptr inbounds %[[S]], %[[S]]* %[[CUR]], i64 1
// CHECK: icmp eq %[[S]]* %[[NEXT]], %[[END]]
// CHECK: br i1
//
// CHECK: }
struct T { int a; };
void *r = new T[n][3]{ { 1, 2, 3 }, { 4, 5, 6 } };
// CHECK-LABEL: define
//
// CHECK: load i32, i32* @n
// CHECK: call {{.*}} @llvm.umul.with.overflow.i64(i64 %[[N:.*]], i64 12)
// CHECK: %[[ELTS:.*]] = mul i64 %[[N]], 3
//
// No cookie.
// CHECK-NOT: @llvm.uadd.with.overflow
//
// CHECK: %[[ALLOC:.*]] = call noalias nonnull i8* @_Znam(i64 %{{.*}})
//
// CHECK: %[[START_AS_T:.*]] = bitcast i8* %[[ALLOC]] to %[[T:.*]]*
//
// Explicit initializers:
//
// { 1, 2, 3 }
//
// CHECK: %[[T_0:.*]] = bitcast %[[T]]* %[[START_AS_T]] to [3 x %[[T]]]*
//
// CHECK: %[[T_0_0:.*]] = getelementptr inbounds [3 x %[[T]]], [3 x %[[T]]]* %[[T_0]], i64 0, i64 0
// CHECK: %[[T_0_0_0:.*]] = getelementptr inbounds %[[T]], %[[T]]* %[[T_0_0]], i32 0, i32 0
// CHECK: store i32 1, i32* %[[T_0_0_0]]
// CHECK: %[[T_0_1:.*]] = getelementptr inbounds %[[T]], %[[T]]* %[[T_0_0]], i64 1
// CHECK: %[[T_0_1_0:.*]] = getelementptr inbounds %[[T]], %[[T]]* %[[T_0_1]], i32 0, i32 0
// CHECK: store i32 2, i32* %[[T_0_1_0]]
// CHECK: %[[T_0_2:.*]] = getelementptr inbounds %[[T]], %[[T]]* %[[T_0_1]], i64 1
// CHECK: %[[T_0_2_0:.*]] = getelementptr inbounds %[[T]], %[[T]]* %[[T_0_2]], i32 0, i32 0
// CHECK: store i32 3, i32* %[[T_0_2_0]]
//
// { 4, 5, 6 }
//
// CHECK: %[[T_1:.*]] = getelementptr inbounds [3 x %[[T]]], [3 x %[[T]]]* %[[T_0]], i64 1
//
// CHECK: %[[T_1_0:.*]] = getelementptr inbounds [3 x %[[T]]], [3 x %[[T]]]* %[[T_1]], i64 0, i64 0
// CHECK: %[[T_1_0_0:.*]] = getelementptr inbounds %[[T]], %[[T]]* %[[T_1_0]], i32 0, i32 0
// CHECK: store i32 4, i32* %[[T_1_0_0]]
// CHECK: %[[T_1_1:.*]] = getelementptr inbounds %[[T]], %[[T]]* %[[T_1_0]], i64 1
// CHECK: %[[T_1_1_0:.*]] = getelementptr inbounds %[[T]], %[[T]]* %[[T_1_1]], i32 0, i32 0
// CHECK: store i32 5, i32* %[[T_1_1_0]]
// CHECK: %[[T_1_2:.*]] = getelementptr inbounds %[[T]], %[[T]]* %[[T_1_1]], i64 1
// CHECK: %[[T_1_2_0:.*]] = getelementptr inbounds %[[T]], %[[T]]* %[[T_1_2]], i32 0, i32 0
// CHECK: store i32 6, i32* %[[T_1_2_0]]
//
// And the rest gets memset to 0.
//
// CHECK: %[[T_2:.*]] = getelementptr inbounds [3 x %[[T]]], [3 x %[[T]]]* %[[T_1]], i64 1
// CHECK: %[[T_2_AS_T:.*]] = bitcast [3 x %[[T]]]* %[[T_2]] to %[[T]]*
//
// CHECK: %[[SIZE:.*]] = sub i64 %{{.*}}, 24
// CHECK: %[[REST:.*]] = bitcast %[[T]]* %[[T_2_AS_T]] to i8*
// CHECK: call void @llvm.memset.p0i8.i64(i8* align 4 %[[REST]], i8 0, i64 %[[SIZE]], i1 false)
//
// CHECK: }
Reference in New Issue View Git Blame Copy Permalink