91cdd35008
This is a major change on how we represent nested name qualifications in the AST. * The nested name specifier itself and how it's stored is changed. The prefixes for types are handled within the type hierarchy, which makes canonicalization for them super cheap, no memory allocation required. Also translating a type into nested name specifier form becomes a no-op. An identifier is stored as a DependentNameType. The nested name specifier gains a lightweight handle class, to be used instead of passing around pointers, which is similar to what is implemented for TemplateName. There is still one free bit available, and this handle can be used within a PointerUnion and PointerIntPair, which should keep bit-packing aficionados happy. * The ElaboratedType node is removed, all type nodes in which it could previously apply to can now store the elaborated keyword and name qualifier, tail allocating when present. * TagTypes can now point to the exact declaration found when producing these, as opposed to the previous situation of there only existing one TagType per entity. This increases the amount of type sugar retained, and can have several applications, for example in tracking module ownership, and other tools which care about source file origins, such as IWYU. These TagTypes are lazily allocated, in order to limit the increase in AST size. This patch offers a great performance benefit. It greatly improves compilation time for [stdexec](https://github.com/NVIDIA/stdexec). For one datapoint, for `test_on2.cpp` in that project, which is the slowest compiling test, this patch improves `-c` compilation time by about 7.2%, with the `-fsyntax-only` improvement being at ~12%. This has great results on compile-time-tracker as well:  This patch also further enables other optimziations in the future, and will reduce the performance impact of template specialization resugaring when that lands. It has some other miscelaneous drive-by fixes. About the review: Yes the patch is huge, sorry about that. Part of the reason is that I started by the nested name specifier part, before the ElaboratedType part, but that had a huge performance downside, as ElaboratedType is a big performance hog. I didn't have the steam to go back and change the patch after the fact. There is also a lot of internal API changes, and it made sense to remove ElaboratedType in one go, versus removing it from one type at a time, as that would present much more churn to the users. Also, the nested name specifier having a different API avoids missing changes related to how prefixes work now, which could make existing code compile but not work. How to review: The important changes are all in `clang/include/clang/AST` and `clang/lib/AST`, with also important changes in `clang/lib/Sema/TreeTransform.h`. The rest and bulk of the changes are mostly consequences of the changes in API. PS: TagType::getDecl is renamed to `getOriginalDecl` in this patch, just for easier to rebasing. I plan to rename it back after this lands. Fixes #136624 Fixes https://github.com/llvm/llvm-project/issues/43179 Fixes https://github.com/llvm/llvm-project/issues/68670 Fixes https://github.com/llvm/llvm-project/issues/92757
108 lines
4.3 KiB
Objective-C
108 lines
4.3 KiB
Objective-C
// RUN: rm -rf %t
|
|
// RUN: split-file %s %t
|
|
|
|
|
|
// Expect no crash
|
|
// RUN: %clang_cc1 -fmodules -fmodules-cache-path=%t/modcache -fmodule-map-file=%t/module.modulemap %t/source.m
|
|
|
|
// Add -ast-dump-all to check that the AST nodes are merged correctly.
|
|
// RUN: %clang_cc1 -fmodules -fmodules-cache-path=%t/modcache -fmodule-map-file=%t/module.modulemap %t/source.m -ast-dump-all 2>&1 | FileCheck %s
|
|
|
|
|
|
//--- shared.h
|
|
// This header is shared between two modules, but it's not a module itself.
|
|
// The enums defined here are parsed in both modules, and merged while building ModB.
|
|
|
|
typedef enum MyEnum1 { MyVal_A } MyEnum1;
|
|
// CHECK: |-EnumDecl 0x{{.*}} imported in ModA.ModAFile1 <undeserialized declarations> MyEnum1
|
|
// CHECK-NEXT: | |-also in ModB
|
|
// CHECK-NEXT: | `-EnumConstantDecl 0x{{.*}} imported in ModA.ModAFile1 referenced MyVal_A 'int'
|
|
// CHECK-NEXT: |-TypedefDecl 0x{{.*}} imported in ModA.ModAFile1 hidden MyEnum1 'enum MyEnum1'
|
|
// CHECK-NEXT: | `-EnumType 0x{{.*}} 'enum MyEnum1' imported
|
|
// CHECK-NEXT: | `-Enum 0x{{.*}} 'MyEnum1'
|
|
|
|
|
|
enum MyEnum2 { MyVal_B };
|
|
// CHECK: |-EnumDecl 0x{{.*}} imported in ModA.ModAFile1 <undeserialized declarations> MyEnum2
|
|
// CHECK-NEXT: | |-also in ModB
|
|
// CHECK-NEXT: | `-EnumConstantDecl 0x{{.*}} imported in ModA.ModAFile1 referenced MyVal_B 'int'
|
|
|
|
|
|
typedef enum { MyVal_C } MyEnum3;
|
|
// CHECK: |-EnumDecl 0x{{.*}} imported in ModA.ModAFile1 <undeserialized declarations>
|
|
// CHECK-NEXT: | |-also in ModB
|
|
// CHECK-NEXT: | `-EnumConstantDecl 0x{{.*}} imported in ModA.ModAFile1 referenced MyVal_C 'int'
|
|
// CHECK-NEXT: |-TypedefDecl 0x{{.*}} imported in ModA.ModAFile1 hidden MyEnum3 'enum MyEnum3'
|
|
// CHECK-NEXT: | `-EnumType 0x{{.*}} 'enum MyEnum3' imported
|
|
// CHECK-NEXT: | `-Enum 0x{{.*}}
|
|
|
|
struct MyStruct {
|
|
enum MyEnum5 { MyVal_D } Field;
|
|
};
|
|
|
|
// CHECK: |-RecordDecl 0x{{.*}} imported in ModA.ModAFile1 <undeserialized declarations> struct MyStruct definition
|
|
// CHECK-NEXT: | |-also in ModB
|
|
// CHECK-NEXT: | |-EnumDecl 0x{{.*}} imported in ModA.ModAFile1 <undeserialized declarations> MyEnum5
|
|
// CHECK-NEXT: | | |-also in ModB
|
|
// CHECK-NEXT: | | `-EnumConstantDecl 0x{{.*}} imported in ModA.ModAFile1 referenced MyVal_D 'int'
|
|
// CHECK-NEXT: | `-FieldDecl 0x{{.*}} imported in ModA.ModAFile1 hidden Field 'enum MyEnum5'
|
|
|
|
// In this case, no merging happens on the EnumDecl in Objective-C, and ASTWriter writes both EnumConstantDecls when building ModB.
|
|
enum { MyVal_E };
|
|
// CHECK: |-EnumDecl 0x{{.*}} imported in ModA.ModAFile1 hidden <undeserialized declarations>
|
|
// CHECK-NEXT: | `-EnumConstantDecl 0x{{.*}} imported in ModA.ModAFile1 hidden MyVal_E 'int'
|
|
|
|
|
|
// Redeclarations coming from ModB.
|
|
// CHECK: |-TypedefDecl 0x{{.*}} prev 0x{{.*}} imported in ModB MyEnum1 'enum MyEnum1'
|
|
// CHECK-NEXT: | `-EnumType 0x{{.*}} 'enum MyEnum1' imported
|
|
// CHECK-NEXT: | `-Enum 0x{{.*}} 'MyEnum1'
|
|
|
|
// CHECK: |-EnumDecl 0x{{.*}} prev 0x{{.*}} imported in ModB <undeserialized declarations>
|
|
// CHECK-NEXT: | |-also in ModB
|
|
// CHECK-NEXT: | `-EnumConstantDecl 0x{{.*}} imported in ModB MyVal_C 'int'
|
|
// CHECK-NEXT: |-TypedefDecl 0x{{.*}} prev 0x{{.*}} imported in ModB MyEnum3 'enum MyEnum3'
|
|
// CHECK-NEXT: | `-EnumType 0x{{.*}} 'enum MyEnum3' imported
|
|
// CHECK-NEXT: | `-Enum 0x{{.*}}
|
|
|
|
// CHECK: |-EnumDecl 0x{{.*}} imported in ModB <undeserialized declarations>
|
|
// CHECK-NEXT: | `-EnumConstantDecl 0x{{.*}} first 0x{{.*}} imported in ModB referenced MyVal_E 'int'
|
|
|
|
|
|
|
|
//--- module.modulemap
|
|
module ModA {
|
|
module ModAFile1 {
|
|
header "ModAFile1.h"
|
|
export *
|
|
}
|
|
module ModAFile2 {
|
|
header "ModAFile2.h"
|
|
export *
|
|
}
|
|
}
|
|
// The goal of writing ModB is to test that ASTWriter can handle the merged AST nodes correctly.
|
|
// For example, a stale declaration in IdResolver can cause an assertion failure while writing the identifier table.
|
|
module ModB {
|
|
header "ModBFile.h"
|
|
export *
|
|
}
|
|
|
|
//--- ModAFile1.h
|
|
#include "shared.h"
|
|
|
|
//--- ModAFile2.h
|
|
// Including this file, triggers loading of the module ModA with nodes coming ModAFile1.h being hidden.
|
|
|
|
//--- ModBFile.h
|
|
// ModBFile depends on ModAFile2.h only.
|
|
#include "ModAFile2.h"
|
|
// Including shared.h here causes Sema to merge the AST nodes from shared.h with the hidden ones from ModA.
|
|
#include "shared.h"
|
|
|
|
|
|
//--- source.m
|
|
#include "ModBFile.h"
|
|
|
|
int main() { return MyVal_A + MyVal_B + MyVal_C + MyVal_D + MyVal_E; }
|