[clang] Make getAttr, getSpecificAttr, hasAttr, and hasSpecificAttr variadic (PR #78518)
Erich Keane via cfe-commits
cfe-commits at lists.llvm.org
Thu Jan 18 08:13:17 PST 2024
https://github.com/erichkeane updated https://github.com/llvm/llvm-project/pull/78518
>From e3d667796b903d35405014db67b130ae20816bab Mon Sep 17 00:00:00 2001
From: erichkeane <ekeane at nvidia.com>
Date: Wed, 17 Jan 2024 10:26:22 -0800
Subject: [PATCH 1/2] Make getAttr, getSpecificAttr, hasAttr, and
hasSpecificAttr variadic
Discussed elsewhere, it makes sense to have these functions be
variadic. They work as an 'or', that is, 'has*Attr' is "one of the set",
and "get*Attr" means "first of the set".
This is accomplished by extracting specific_attr_iterator into a _impl
version, and creating a version + partial specialization for the
_iterator version that inherits from the _impl. The result is that all
of the 'existing' code 'just works'.
One exception (see the change in SemaOverload.cpp) could be solved with
a conversion function, but instead I put 'auto' in. Else, I could
specify the 'AttrVec' if it were preferential.
---
clang/include/clang/AST/AttrIterator.h | 108 ++++++++++++++++---------
clang/include/clang/AST/DeclBase.h | 28 ++++---
clang/lib/Sema/SemaOverload.cpp | 9 +--
3 files changed, 90 insertions(+), 55 deletions(-)
diff --git a/clang/include/clang/AST/AttrIterator.h b/clang/include/clang/AST/AttrIterator.h
index 66571e1cf0b8ec..51732d9a3334b2 100644
--- a/clang/include/clang/AST/AttrIterator.h
+++ b/clang/include/clang/AST/AttrIterator.h
@@ -27,12 +27,24 @@ class Attr;
/// AttrVec - A vector of Attr, which is how they are stored on the AST.
using AttrVec = SmallVector<Attr *, 4>;
-/// specific_attr_iterator - Iterates over a subrange of an AttrVec, only
-/// providing attributes that are of a specific type.
-template <typename SpecificAttr, typename Container = AttrVec>
-class specific_attr_iterator {
+/// Iterates over a subrange of container, only providing attributes that are of
+/// a specific type/s.
+template <typename Container, typename... SpecificAttrs>
+class specific_attr_iterator_impl {
using Iterator = typename Container::const_iterator;
+ /// Helper class to get either the singular 'specific-attr', or Attr,
+ /// depending on how many are specified.
+ template <typename... Ts> struct type_helper {
+ using type = Attr;
+ };
+ template <typename T> struct type_helper<T> {
+ using type = T;
+ };
+
+ /// The pointee type of the value_type, used for internal implementation.
+ using base_type = typename type_helper<SpecificAttrs...>::type;
+
/// Current - The current, underlying iterator.
/// In order to ensure we don't dereference an invalid iterator unless
/// specifically requested, we don't necessarily advance this all the
@@ -43,46 +55,46 @@ class specific_attr_iterator {
mutable Iterator Current;
void AdvanceToNext() const {
- while (!isa<SpecificAttr>(*Current))
+ while (!isa<SpecificAttrs...>(*Current))
++Current;
}
void AdvanceToNext(Iterator I) const {
- while (Current != I && !isa<SpecificAttr>(*Current))
+ while (Current != I && !isa<SpecificAttrs...>(*Current))
++Current;
}
public:
- using value_type = SpecificAttr *;
- using reference = SpecificAttr *;
- using pointer = SpecificAttr *;
+ using value_type = base_type *;
+ using reference = value_type;
+ using pointer = value_type;
using iterator_category = std::forward_iterator_tag;
using difference_type = std::ptrdiff_t;
- specific_attr_iterator() = default;
- explicit specific_attr_iterator(Iterator i) : Current(i) {}
+ specific_attr_iterator_impl() = default;
+ explicit specific_attr_iterator_impl(Iterator i) : Current(i) {}
reference operator*() const {
AdvanceToNext();
- return cast<SpecificAttr>(*Current);
+ return cast<base_type>(*Current);
}
pointer operator->() const {
AdvanceToNext();
- return cast<SpecificAttr>(*Current);
+ return cast<base_type>(*Current);
}
- specific_attr_iterator& operator++() {
+ specific_attr_iterator_impl &operator++() {
++Current;
return *this;
}
- specific_attr_iterator operator++(int) {
- specific_attr_iterator Tmp(*this);
+ specific_attr_iterator_impl operator++(int) {
+ specific_attr_iterator_impl Tmp(*this);
++(*this);
return Tmp;
}
- friend bool operator==(specific_attr_iterator Left,
- specific_attr_iterator Right) {
+ friend bool operator==(specific_attr_iterator_impl Left,
+ specific_attr_iterator_impl Right) {
assert((Left.Current == nullptr) == (Right.Current == nullptr));
if (Left.Current < Right.Current)
Left.AdvanceToNext(Right.Current);
@@ -90,33 +102,55 @@ class specific_attr_iterator {
Right.AdvanceToNext(Left.Current);
return Left.Current == Right.Current;
}
- friend bool operator!=(specific_attr_iterator Left,
- specific_attr_iterator Right) {
+ friend bool operator!=(specific_attr_iterator_impl Left,
+ specific_attr_iterator_impl Right) {
return !(Left == Right);
}
};
-template <typename SpecificAttr, typename Container>
-inline specific_attr_iterator<SpecificAttr, Container>
- specific_attr_begin(const Container& container) {
- return specific_attr_iterator<SpecificAttr, Container>(container.begin());
+/// Iterates over a subrange of a collection, only providing attributes that are
+/// of a specific type/s.
+template <typename Container, typename... SpecificAttrs>
+class specific_attr_iterator;
+
+template <typename SpecificAttr>
+class specific_attr_iterator<SpecificAttr>
+ : public specific_attr_iterator_impl<AttrVec, SpecificAttr> {
+ using specific_attr_iterator_impl<AttrVec,
+ SpecificAttr>::specific_attr_iterator_impl;
+};
+
+template <typename Container, typename... SpecificAttrs>
+class specific_attr_iterator
+ : public specific_attr_iterator_impl<Container, SpecificAttrs...> {
+ using specific_attr_iterator_impl<
+ Container, SpecificAttrs...>::specific_attr_iterator_impl;
+};
+
+template <typename... SpecificAttrs, typename Container>
+inline specific_attr_iterator<Container, SpecificAttrs...>
+specific_attr_begin(const Container &container) {
+ return specific_attr_iterator<Container, SpecificAttrs...>(container.begin());
}
-template <typename SpecificAttr, typename Container>
-inline specific_attr_iterator<SpecificAttr, Container>
- specific_attr_end(const Container& container) {
- return specific_attr_iterator<SpecificAttr, Container>(container.end());
+
+template <typename... SpecificAttrs, typename Container>
+inline specific_attr_iterator<Container, SpecificAttrs...>
+specific_attr_end(const Container &container) {
+ return specific_attr_iterator<Container, SpecificAttrs...>(container.end());
}
-template <typename SpecificAttr, typename Container>
-inline bool hasSpecificAttr(const Container& container) {
- return specific_attr_begin<SpecificAttr>(container) !=
- specific_attr_end<SpecificAttr>(container);
+template <typename... SpecificAttrs, typename Container>
+inline bool hasSpecificAttr(const Container &container) {
+ return specific_attr_begin<SpecificAttrs...>(container) !=
+ specific_attr_end<SpecificAttrs...>(container);
}
-template <typename SpecificAttr, typename Container>
-inline SpecificAttr *getSpecificAttr(const Container& container) {
- specific_attr_iterator<SpecificAttr, Container> i =
- specific_attr_begin<SpecificAttr>(container);
- if (i != specific_attr_end<SpecificAttr>(container))
+
+template <typename... SpecificAttrs, typename Container>
+inline typename specific_attr_iterator_impl<Container,
+ SpecificAttrs...>::value_type
+getSpecificAttr(const Container &container) {
+ auto i = specific_attr_begin<SpecificAttrs...>(container);
+ if (i != specific_attr_end<SpecificAttrs...>(container))
return *i;
else
return nullptr;
diff --git a/clang/include/clang/AST/DeclBase.h b/clang/include/clang/AST/DeclBase.h
index d957ea24f6394a..cb111af9b17fc3 100644
--- a/clang/include/clang/AST/DeclBase.h
+++ b/clang/include/clang/AST/DeclBase.h
@@ -560,27 +560,29 @@ class alignas(8) Decl {
template <typename T> void dropAttr() { dropAttrs<T>(); }
- template <typename T>
- llvm::iterator_range<specific_attr_iterator<T>> specific_attrs() const {
- return llvm::make_range(specific_attr_begin<T>(), specific_attr_end<T>());
+ template <typename... Ts>
+ llvm::iterator_range<specific_attr_iterator<AttrVec, Ts...>>
+ specific_attrs() const {
+ return llvm::make_range(specific_attr_begin<Ts...>(),
+ specific_attr_end<Ts...>());
}
- template <typename T>
- specific_attr_iterator<T> specific_attr_begin() const {
- return specific_attr_iterator<T>(attr_begin());
+ template <typename... Ts>
+ specific_attr_iterator<AttrVec, Ts...> specific_attr_begin() const {
+ return specific_attr_iterator<AttrVec, Ts...>(attr_begin());
}
- template <typename T>
- specific_attr_iterator<T> specific_attr_end() const {
- return specific_attr_iterator<T>(attr_end());
+ template <typename... Ts>
+ specific_attr_iterator<AttrVec, Ts...> specific_attr_end() const {
+ return specific_attr_iterator<AttrVec, Ts...>(attr_end());
}
- template<typename T> T *getAttr() const {
- return hasAttrs() ? getSpecificAttr<T>(getAttrs()) : nullptr;
+ template <typename... Ts> auto *getAttr() const {
+ return hasAttrs() ? getSpecificAttr<Ts...>(getAttrs()) : nullptr;
}
- template<typename T> bool hasAttr() const {
- return hasAttrs() && hasSpecificAttr<T>(getAttrs());
+ template <typename... Ts> bool hasAttr() const {
+ return hasAttrs() && hasSpecificAttr<Ts...>(getAttrs());
}
/// getMaxAlignment - return the maximum alignment specified by attributes
diff --git a/clang/lib/Sema/SemaOverload.cpp b/clang/lib/Sema/SemaOverload.cpp
index 37c62b306b3cd3..c4c713efa7f5ab 100644
--- a/clang/lib/Sema/SemaOverload.cpp
+++ b/clang/lib/Sema/SemaOverload.cpp
@@ -1478,11 +1478,10 @@ static bool IsOverloadOrOverrideImpl(Sema &SemaRef, FunctionDecl *New,
return true;
// enable_if attributes are an order-sensitive part of the signature.
- for (specific_attr_iterator<EnableIfAttr>
- NewI = New->specific_attr_begin<EnableIfAttr>(),
- NewE = New->specific_attr_end<EnableIfAttr>(),
- OldI = Old->specific_attr_begin<EnableIfAttr>(),
- OldE = Old->specific_attr_end<EnableIfAttr>();
+ for (auto NewI = New->specific_attr_begin<EnableIfAttr>(),
+ NewE = New->specific_attr_end<EnableIfAttr>(),
+ OldI = Old->specific_attr_begin<EnableIfAttr>(),
+ OldE = Old->specific_attr_end<EnableIfAttr>();
NewI != NewE || OldI != OldE; ++NewI, ++OldI) {
if (NewI == NewE || OldI == OldE)
return true;
>From 1958942585cc1792cbc58d159bcb1d02b9ed00b8 Mon Sep 17 00:00:00 2001
From: erichkeane <ekeane at nvidia.com>
Date: Thu, 18 Jan 2024 07:35:56 -0800
Subject: [PATCH 2/2] MOdify uses in SemaDecl.cpp to use variadic hasAttr.
---
clang/lib/Sema/SemaDecl.cpp | 21 ++++++++-------------
1 file changed, 8 insertions(+), 13 deletions(-)
diff --git a/clang/lib/Sema/SemaDecl.cpp b/clang/lib/Sema/SemaDecl.cpp
index 5472b43aafd4f3..88f71e26c9abbf 100644
--- a/clang/lib/Sema/SemaDecl.cpp
+++ b/clang/lib/Sema/SemaDecl.cpp
@@ -2026,8 +2026,7 @@ static bool ShouldDiagnoseUnusedDecl(const LangOptions &LangOpts,
if (D->isPlaceholderVar(LangOpts))
return false;
- if (D->hasAttr<UnusedAttr>() || D->hasAttr<ObjCPreciseLifetimeAttr>() ||
- D->hasAttr<CleanupAttr>())
+ if (D->hasAttr<UnusedAttr, ObjCPreciseLifetimeAttr, CleanupAttr>())
return false;
if (isa<LabelDecl>(D))
@@ -7331,8 +7330,8 @@ static bool isIncompleteDeclExternC(Sema &S, const T *D) {
return false;
// So do CUDA's host/device attributes.
- if (S.getLangOpts().CUDA && (D->template hasAttr<CUDADeviceAttr>() ||
- D->template hasAttr<CUDAHostAttr>()))
+ if (S.getLangOpts().CUDA &&
+ (D->template hasAttr<CUDADeviceAttr, CUDAHostAttr>()))
return false;
}
return D->isExternC();
@@ -8035,8 +8034,7 @@ NamedDecl *Sema::ActOnVariableDeclarator(
// CUDA B.2.5: "__shared__ and __constant__ variables have implied static
// storage [duration]."
if (SC == SC_None && S->getFnParent() != nullptr &&
- (NewVD->hasAttr<CUDASharedAttr>() ||
- NewVD->hasAttr<CUDAConstantAttr>())) {
+ (NewVD->hasAttr<CUDASharedAttr, CUDAConstantAttr>())) {
NewVD->setStorageClass(SC_Static);
}
}
@@ -8805,8 +8803,7 @@ void Sema::CheckVariableDeclarationType(VarDecl *NewVD) {
}
bool isVM = T->isVariablyModifiedType();
- if (isVM || NewVD->hasAttr<CleanupAttr>() ||
- NewVD->hasAttr<BlocksAttr>())
+ if (isVM || NewVD->hasAttr<CleanupAttr, BlocksAttr>())
setFunctionHasBranchProtectedScope();
if ((isVM && NewVD->hasLinkage()) ||
@@ -10797,8 +10794,7 @@ Sema::ActOnFunctionDeclarator(Scope *S, Declarator &D, DeclContext *DC,
// in device-side CUDA code, unless someone passed
// -fcuda-allow-variadic-functions.
if (!getLangOpts().CUDAAllowVariadicFunctions && NewFD->isVariadic() &&
- (NewFD->hasAttr<CUDADeviceAttr>() ||
- NewFD->hasAttr<CUDAGlobalAttr>()) &&
+ (NewFD->hasAttr<CUDADeviceAttr, CUDAGlobalAttr>()) &&
!(II && II->isStr("printf") && NewFD->isExternC() &&
!D.isFunctionDefinition())) {
Diag(NewFD->getLocation(), diag::err_variadic_device_fn);
@@ -14617,8 +14613,7 @@ void Sema::CheckStaticLocalForDllExport(VarDecl *VD) {
// Find outermost function when VD is in lambda function.
while (FD && !getDLLAttr(FD) &&
- !FD->hasAttr<DLLExportStaticLocalAttr>() &&
- !FD->hasAttr<DLLImportStaticLocalAttr>()) {
+ !FD->hasAttr<DLLExportStaticLocalAttr, DLLImportStaticLocalAttr>()) {
FD = dyn_cast_or_null<FunctionDecl>(FD->getParentFunctionOrMethod());
}
@@ -16675,7 +16670,7 @@ void Sema::AddKnownFunctionAttributes(FunctionDecl *FD) {
if (Context.BuiltinInfo.isConst(BuiltinID) && !FD->hasAttr<ConstAttr>())
FD->addAttr(ConstAttr::CreateImplicit(Context, FD->getLocation()));
if (getLangOpts().CUDA && Context.BuiltinInfo.isTSBuiltin(BuiltinID) &&
- !FD->hasAttr<CUDADeviceAttr>() && !FD->hasAttr<CUDAHostAttr>()) {
+ !FD->hasAttr<CUDADeviceAttr, CUDAHostAttr>()) {
// Add the appropriate attribute, depending on the CUDA compilation mode
// and which target the builtin belongs to. For example, during host
// compilation, aux builtins are __device__, while the rest are __host__.
More information about the cfe-commits
mailing list