[clang-tidy] New check to warn when storing dispatch_once_t in non-static, non-global storage

[mwyman]

Summary

Creates a new darwin ClangTidy module and adds the darwin-dispatch-once-nonstatic check that warns about dispatch_once_t variables not in static or global storage. This catches a missing static for local variables in e.g. singleton initialization behavior, and also warns on storing dispatch_once_t values in Objective-C instance variables. C/C++ struct/class instances may potentially live in static/global storage, and are ignored for this check.

The osx.API static analysis checker can find the non-static storage use of dispatch_once_t; I thought it useful to also catch this issue in clang-tidy when possible.

[lebedev.ri]

1. Please split each check into separate review.
2. Is dispatch_once_t OSX-specific thing? Should those checks be in osx module?

[mwyman]

Moved the assignment check to a separate review.

[mwyman]

In D67567#1669866, @lebedev.ri wrote:

1. Please split each check into separate review.
2. Is dispatch_once_t OSX-specific thing? Should those checks be in osx module?

1. I split the review.
2. I don't see an osx module—am I somehow missing seeing it? dispatch_once_t is something from libdispatchwhich is an Apple-created API, but is OSS and available outside Apple. It's also a C API, which is why I didn't feel objc was an appropriate module either. I'm open to a better home module for this check, it's just not clear to me where that might be.

[lebedev.ri]

In D67567#1670017, @mwyman wrote:

In D67567#1669866, @lebedev.ri wrote:

1. Please split each check into separate review.
2. Is dispatch_once_t OSX-specific thing? Should those checks be in osx module?

1. I split the review.

2. I don't see an osx module—am I somehow missing seeing it?

One can be added.

dispatch_once_t is something from libdispatchwhich is an Apple-created API, but is OSS and available outside Apple. It's also a C API, which is why I didn't feel objc was an appropriate module either. I'm open to a better home module for this check, it's just not clear to me where that might be.

No opinion on my question, just thought it should be asked,

[benhamilton]

It definitely shouldn't be in an osx directory since it's available on iOS and other Darwin-like operating systems.

Probably darwin would be OK. Adding a new subdirectory isn't trivial, there's a lot of places to update last I looked.

Should we tackle moving things around in a separate diff?

[mwyman]

In D67567#1670035, @lebedev.ri wrote:

In D67567#1670017, @mwyman wrote:

In D67567#1669866, @lebedev.ri wrote:

1. Please split each check into separate review.
2. Is dispatch_once_t OSX-specific thing? Should those checks be in osx module?

1. I split the review.

2. I don't see an osx module—am I somehow missing seeing it?

One can be added.

dispatch_once_t is something from libdispatchwhich is an Apple-created API, but is OSS and available outside Apple. It's also a C API, which is why I didn't feel objc was an appropriate module either. I'm open to a better home module for this check, it's just not clear to me where that might be.

No opinion on my question, just thought it should be asked,

As Ben commented, I don't feel this should be in an osx module, as libdispatch is highly used on iOS and other Apple platforms. I know it's available elsewhere, but it may be so seldom used that I think a darwinmodule might be reasonable. I looked at the process for creating a new module, and can do that if we'd like, but should that be part of this same review?

[NoQ]

FTR, we already have a similar Static Analyzer check, eg.:
https://github.com/llvm-mirror/clang/blob/release_80/test/Analysis/dispatch-once.m#L15
https://github.com/llvm-mirror/clang/blob/release_80/test/Analysis/dispatch-once.m#L26

Your check is a bit more aggressive but i don't see why didn't we do it that way in the first place :) Though you won't be able to warn on the heap example.

[mwyman]

In D67567#1670264, @NoQ wrote:

FTR, we already have a similar Static Analyzer check, eg.:
https://github.com/llvm-mirror/clang/blob/release_80/test/Analysis/dispatch-once.m#L15
https://github.com/llvm-mirror/clang/blob/release_80/test/Analysis/dispatch-once.m#L26

Your check is a bit more aggressive but i don't see why didn't we do it that way in the first place :) Though you won't be able to warn on the heap example.

The Static Analyzer check was pointed out by a colleague; unfortunately our build environment doesn't currently play nice with running the static analyzer (so many devs don't end up running it) but ClangTidy gets run as part of our code review process. Given libdispatch's documented requirements, it seemed reasonable to be aggressive with a ClangTidy check when we can reasonably identify non-static/global storage.

[lebedev.ri]

In D67567#1671499, @mwyman wrote:

In D67567#1670264, @NoQ wrote:

FTR, we already have a similar Static Analyzer check, eg.:
https://github.com/llvm-mirror/clang/blob/release_80/test/Analysis/dispatch-once.m#L15
https://github.com/llvm-mirror/clang/blob/release_80/test/Analysis/dispatch-once.m#L26

Your check is a bit more aggressive but i don't see why didn't we do it that way in the first place :) Though you won't be able to warn on the heap example.

The Static Analyzer check was pointed out by a colleague; unfortunately our build environment doesn't currently play nice with running the static analyzer (so many devs don't end up running it) but ClangTidy gets run as part of our code review process. Given libdispatch's documented requirements, it seemed reasonable to be aggressive with a ClangTidy check when we can reasonably identify non-static/global storage.

You can run static analyzer checks as normal clang-tidy checks.

[mwyman]

Migrated check to new darwin module.

[stephanemoore]

In D67567#1670144, @mwyman wrote:

In D67567#1670035, @lebedev.ri wrote:

In D67567#1670017, @mwyman wrote:

In D67567#1669866, @lebedev.ri wrote:

2. Is dispatch_once_t OSX-specific thing? Should those checks be in osx module?

2. I don't see an osx module—am I somehow missing seeing it?

One can be added.

dispatch_once_t is something from libdispatchwhich is an Apple-created API, but is OSS and available outside Apple. It's also a C API, which is why I didn't feel objc was an appropriate module either. I'm open to a better home module for this check, it's just not clear to me where that might be.

No opinion on my question, just thought it should be asked,

As Ben commented, I don't feel this should be in an osx module, as libdispatch is highly used on iOS and other Apple platforms. I know it's available elsewhere, but it may be so seldom used that I think a darwinmodule might be reasonable. I looked at the process for creating a new module, and can do that if we'd like, but should that be part of this same review?

As alluded to, technically Grand Central Dispatch is documented to support FreeBSD. With that said, I would be okay with putting this in a darwin module because GCD is most prevalent on Darwin platforms.

clang-tools-extra/test/clang-tidy/darwin-dispatch-once-nonstatic.mm

45 ↗

(On Diff #220357)

Maybe add a CHECK-FIXES to make sure that no fix was recommended?

[gribozavr]

FWIW, I like this approach better than the one in Static Analyzer because it warns on the variable declaration -- that's where the root of the issue is, not at the call to dispatch_once(). Like, what can you in principle do with a local dispatch_once_t? not much -- therefore produce a warning right there.

clang-tools-extra/docs/clang-tidy/checks/darwin-dispatch-once-nonstatic.rst

10 ↗

(On Diff #220357)

s/paradigm/pattern/

[mwyman]

Addressed review comments.

[gribozavr]

Thanks! Do you need me to commit the patch for you?

[mwyman]

In D67567#1683905, @gribozavr wrote:

Thanks! Do you need me to commit the patch for you?

Yes, thank you. I don't have commit access—and also wasn't sure if anyone else had further comment.

[gribozavr]

Sorry, could you rebase the patch to apply cleanly to master? Seems like someone else edited ReleaseNotes.rst in the meanwhile.

$ arc patch D67567
...
Checking patch clang-tools-extra/docs/ReleaseNotes.rst...
error: while searching for:
  Finds instances where variables with static storage are initialized
  dynamically in header files.

[mwyman]

Rebased patch to apply to current master.

[mwyman]

In D67567#1684785, @gribozavr wrote:

Sorry, could you rebase the patch to apply cleanly to master? Seems like someone else edited ReleaseNotes.rst in the meanwhile.

$ arc patch D67567
...
Checking patch clang-tools-extra/docs/ReleaseNotes.rst...
error: while searching for:
  Finds instances where variables with static storage are initialized
  dynamically in header files.

Just rebased, resolved the conflict and uploaded, should be good.

[stephanemoore]

Looks good! A couple nits/questions.

Does the check allow variables in anonymous namespaces?

namespace {
dispatch_once_t onceToken;
}

I think such variables should satisfy initialization requirements.

If not, can we update the check to not trigger on variables? Either way, can we add a test to verify behavior for variables in anonymous namespaces?

clang-tools-extra/clang-tidy/darwin/CMakeLists.txt

13 ↗

(On Diff #222016)

Is this library necessary?

16 ↗

(On Diff #222016)

Is this library necessary?

clang-tools-extra/clang-tidy/darwin/DispatchOnceNonstaticCheck.cpp

44 ↗

(On Diff #222016)

"dispatch_once_t variables" to be consistent.

clang-tools-extra/docs/clang-tidy/checks/darwin-dispatch-once-nonstatic.rst

15 ↗

(On Diff #222016)

Maybe rephrase to this:

``dispatch_once_t`` variables

16 ↗

(On Diff #222016)

"structs or classes"

[stephanemoore]

Does the check allow variables in anonymous namespaces?

namespace {
dispatch_once_t onceToken;
}

I think such variables should satisfy initialization requirements.

If not, can we update the check to not trigger on variables? Either way, can we add a test to verify behavior for variables in anonymous namespaces?

Nevermind, I overlooked that there is already a variable in an anonymous namespace in the test 😅 Thanks for that!

[mwyman]

Addressed Stephane's review feedback.

[stephanemoore]

It looks like everything is in order. I will proceed with landing the patch on your behalf 👍

[gribozavr]

Sorry, I reverted it in r373032 because the test fails on Linux: http://lab.llvm.org:8011/builders/clang-x86_64-debian-fast/builds/18323 . Could you please take a look? Thanks!

[thakis]

I had a fix, but I was too slow. Here's what you need for the reland: http://codepad.org/3klmw1JV

The commit also broke sphinx with this:

`
6.290 [0/1/1] Generating html Sphinx documentation for clang-tools into "/home/buildbot/llvm-build-dir/clang-tools-sphinx-docs/llvm/build/tools/clang/tools/extra/docs/html"
FAILED: tools/clang/tools/extra/docs/CMakeFiles/docs-clang-tools-html 
cd /home/buildbot/llvm-build-dir/clang-tools-sphinx-docs/llvm/build/tools/clang/tools/extra/docs && /usr/bin/sphinx-build -b html -d /home/buildbot/llvm-build-dir/clang-tools-sphinx-docs/llvm/build/tools/clang/tools/extra/docs/_doctrees-clang-tools-html -q -W /home/buildbot/llvm-build-dir/clang-tools-sphinx-docs/llvm/src/tools/clang/tools/extra/docs /home/buildbot/llvm-build-dir/clang-tools-sphinx-docs/llvm/build/tools/clang/tools/extra/docs/html

Warning, treated as error:
/home/buildbot/llvm-build-dir/clang-tools-sphinx-docs/llvm/src/tools/clang/tools/extra/docs/ReleaseNotes.rst:82: WARNING: Inline interpreted text or phrase reference start-string without end-string.

Please try to fix that for the reland as well :)

[mwyman]

In D67567#1685036, @gribozavr wrote:

Sorry, I reverted it in r373032 because the test fails on Linux: http://lab.llvm.org:8011/builders/clang-x86_64-debian-fast/builds/18323 . Could you please take a look? Thanks!

What's the process for re-landing a change? New Differential, or update this one?

[thakis]

In D67567#1685077, @mwyman wrote:

In D67567#1685036, @gribozavr wrote:

Sorry, I reverted it in r373032 because the test fails on Linux: http://lab.llvm.org:8011/builders/clang-x86_64-debian-fast/builds/18323 . Could you please take a look? Thanks!

What's the process for re-landing a change? New Differential, or update this one?

Either works; I think most people update the existing one.

[Eugene.Zelenko]

Please take a look on PR43827.

clang-tools-extra/trunk/clang-tidy/CMakeLists.txt

 endif()
 
 add_subdirectory(android)
 add_subdirectory(abseil)
 add_subdirectory(boost)
 add_subdirectory(bugprone)
 add_subdirectory(cert)
 add_subdirectory(cppcoreguidelines)
+add_subdirectory(darwin)
 add_subdirectory(fuchsia)
 add_subdirectory(google)
 add_subdirectory(hicpp)
 add_subdirectory(linuxkernel)
 add_subdirectory(llvm)
 add_subdirectory(misc)
 add_subdirectory(modernize)
 if(CLANG_ENABLE_STATIC_ANALYZER)

clang-tools-extra/trunk/clang-tidy/ClangTidyForceLinker.h

 static int LLVM_ATTRIBUTE_UNUSED LLVMModuleAnchorDestination =
     LLVMModuleAnchorSource;
 
 // This anchor is used to force the linker to link the CppCoreGuidelinesModule.
 extern volatile int CppCoreGuidelinesModuleAnchorSource;
 static int LLVM_ATTRIBUTE_UNUSED CppCoreGuidelinesModuleAnchorDestination =
     CppCoreGuidelinesModuleAnchorSource;
 
+// This anchor is used to force the linker to link the DarwinModule.
+extern volatile int DarwinModuleAnchorSource;
+static int LLVM_ATTRIBUTE_UNUSED DarwinModuleAnchorDestination =
+    DarwinModuleAnchorSource;
+
 // This anchor is used to force the linker to link the FuchsiaModule.
 extern volatile int FuchsiaModuleAnchorSource;
 static int LLVM_ATTRIBUTE_UNUSED FuchsiaModuleAnchorDestination =
     FuchsiaModuleAnchorSource;
 
 // This anchor is used to force the linker to link the GoogleModule.
 extern volatile int GoogleModuleAnchorSource;
 static int LLVM_ATTRIBUTE_UNUSED GoogleModuleAnchorDestination =

clang-tools-extra/trunk/clang-tidy/darwin/CMakeLists.txt

+set(LLVM_LINK_COMPONENTS support)
+
+add_clang_library(clangTidyDarwinModule
+  DarwinTidyModule.cpp
+  DispatchOnceNonstaticCheck.cpp
+
+  LINK_LIBS
+  clangAnalysis
+  clangAST
+  clangASTMatchers
+  clangBasic
+  clangLex
+  clangTidy
+  clangTidyUtils
+  )

clang-tools-extra/trunk/clang-tidy/darwin/DarwinTidyModule.cpp

+//===--- MiscTidyModule.cpp - clang-tidy ----------------------------------===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+
+#include "../ClangTidy.h"
+#include "../ClangTidyModule.h"
+#include "../ClangTidyModuleRegistry.h"
+#include "DispatchOnceNonstaticCheck.h"
+
+namespace clang {
+namespace tidy {
+namespace darwin {
+
+class DarwinModule : public ClangTidyModule {
+public:
+  void addCheckFactories(ClangTidyCheckFactories &CheckFactories) override {
+    CheckFactories.registerCheck<DispatchOnceNonstaticCheck>(
+        "darwin-dispatch-once-nonstatic");
+  }
+};
+
+} // namespace darwin
+
+// Register the DarwinTidyModule using this statically initialized variable.
+static ClangTidyModuleRegistry::Add<darwin::DarwinModule>
+    X("misc-module", "Adds miscellaneous lint checks.");
+
+// This anchor is used to force the linker to link in the generated object file
+// and thus register the DarwinModule.
+volatile int DarwinModuleAnchorSource = 0;
+
+} // namespace tidy
+} // namespace clang

clang-tools-extra/trunk/clang-tidy/darwin/DispatchOnceNonstaticCheck.h

+//===--- DispatchOnceNonstaticCheck.h - clang-tidy --------------*- C++ -*-===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+
+#ifndef LLVM_CLANG_TOOLS_EXTRA_CLANG_TIDY_DARWIN_DISPATCHONCENONSTATICCHECK_H
+#define LLVM_CLANG_TOOLS_EXTRA_CLANG_TIDY_DARWIN_DISPATCHONCENONSTATICCHECK_H
+
+#include "../ClangTidyCheck.h"
+
+namespace clang {
+namespace tidy {
+namespace darwin {
+
+/// Finds variables of type dispatch_once_t that do not have static or global
+/// storage duration, as required by the libdispatch documentation.
+///
+/// For the user-facing documentation see:
+/// http://clang.llvm.org/extra/clang-tidy/checks/darwin-dispatch-once-nonstatic.html
+class DispatchOnceNonstaticCheck : public ClangTidyCheck {
+public:
+  DispatchOnceNonstaticCheck(StringRef Name, ClangTidyContext *Context)
+      : ClangTidyCheck(Name, Context) {}
+  void registerMatchers(ast_matchers::MatchFinder *Finder) override;
+  void check(const ast_matchers::MatchFinder::MatchResult &Result) override;
+};
+
+} // namespace darwin
+} // namespace tidy
+} // namespace clang
+
+#endif // LLVM_CLANG_TOOLS_EXTRA_CLANG_TIDY_DARWIN_DISPATCHONCENONSTATICCHECK_H

clang-tools-extra/trunk/clang-tidy/darwin/DispatchOnceNonstaticCheck.cpp

+//===--- DispatchOnceNonstaticCheck.cpp - clang-tidy ----------------------===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+
+#include "DispatchOnceNonstaticCheck.h"
+#include "clang/AST/ASTContext.h"
+#include "clang/AST/Decl.h"
+#include "clang/AST/DeclObjC.h"
+#include "clang/ASTMatchers/ASTMatchFinder.h"
+#include "clang/ASTMatchers/ASTMatchers.h"
+#include "clang/Basic/Diagnostic.h"
+
+using namespace clang::ast_matchers;
+
+namespace clang {
+namespace tidy {
+namespace darwin {
+
+void DispatchOnceNonstaticCheck::registerMatchers(MatchFinder *Finder) {
+  // Find variables without static or global storage. VarDecls do not include
+  // struct/class members, which are FieldDecls.
+  Finder->addMatcher(
+      varDecl(hasLocalStorage(), hasType(asString("dispatch_once_t")))
+          .bind("non-static-var"),
+      this);
+
+  // Members of structs or classes might be okay, if the use is at static or
+  // global scope. These will be ignored for now. But ObjC ivars can be
+  // flagged immediately, since they cannot be static.
+  Finder->addMatcher(
+      objcIvarDecl(hasType(asString("dispatch_once_t"))).bind("ivar"), this);
+}
+
+void DispatchOnceNonstaticCheck::check(const MatchFinder::MatchResult &Result) {
+  if (const auto *VD = Result.Nodes.getNodeAs<VarDecl>("non-static-var")) {
+    if (const auto *PD = dyn_cast<ParmVarDecl>(VD)) {
+      // Catch function/method parameters, as any dispatch_once_t should be
+      // passed by pointer instead.
+      diag(PD->getTypeSpecStartLoc(),
+           "dispatch_once_t variables must have static or global storage "
+           "duration; function parameters should be pointer references");
+    } else {
+      diag(VD->getTypeSpecStartLoc(), "dispatch_once_t variables must have "
+                                      "static or global storage duration")
+          << FixItHint::CreateInsertion(VD->getTypeSpecStartLoc(), "static ");
+    }
+  }
+
+  if (const auto *D = Result.Nodes.getNodeAs<ObjCIvarDecl>("ivar")) {
+    diag(D->getTypeSpecStartLoc(),
+         "dispatch_once_t variables must have static or global storage "
+         "duration and cannot be Objective-C instance variables");
+  }
+}
+
+} // namespace darwin
+} // namespace tidy
+} // namespace clang

clang-tools-extra/trunk/clang-tidy/plugin/CMakeLists.txt

 add_clang_library(clangTidyPlugin
   ClangTidyPlugin.cpp
 
   LINK_LIBS
   clangAST
   clangASTMatchers
   clangBasic
   clangFrontend
   clangSema
   clangTidy
   clangTidyAbseilModule
   clangTidyAndroidModule
   clangTidyBoostModule
   clangTidyBugproneModule
   clangTidyCERTModule
   clangTidyCppCoreGuidelinesModule
+  clangTidyDarwinModule
   clangTidyFuchsiaModule
   clangTidyGoogleModule
   clangTidyHICPPModule
   clangTidyLinuxKernelModule
   clangTidyLLVMModule
   clangTidyMiscModule
   clangTidyModernizeModule
   clangTidyObjCModule

clang-tools-extra/trunk/clang-tidy/tool/CMakeLists.txt

target_link_libraries(clang-tidy
   clangBasic
   clangTidy
   clangTidyAndroidModule
   clangTidyAbseilModule
   clangTidyBoostModule
   clangTidyBugproneModule
   clangTidyCERTModule
   clangTidyCppCoreGuidelinesModule
+  clangTidyDarwinModule
   clangTidyFuchsiaModule
   clangTidyGoogleModule
   clangTidyHICPPModule
   clangTidyLinuxKernelModule
   clangTidyLLVMModule
   clangTidyMiscModule
   clangTidyModernizeModule
   clangTidyObjCModule

clang-tools-extra/trunk/clangd/CMakeLists.txt

add_clang_library(clangDaemon
   clangSerialization
   clangTidy
   clangTidyAndroidModule
   clangTidyAbseilModule
   clangTidyBoostModule
   clangTidyBugproneModule
   clangTidyCERTModule
   clangTidyCppCoreGuidelinesModule
+  clangTidyDarwinModule
   clangTidyFuchsiaModule
   clangTidyGoogleModule
   clangTidyHICPPModule
   clangTidyLinuxKernelModule
   clangTidyLLVMModule
   clangTidyMiscModule
   clangTidyModernizeModule
   clangTidyObjCModule

clang-tools-extra/trunk/docs/ReleaseNotes.rst

- New :doc:`bugprone-dynamic-static-initializers
   dynamically in header files.
 
 - New :doc:`bugprone-infinite-loop
   <clang-tidy/checks/bugprone-infinite-loop>` check.
 
   Finds obvious infinite loops (loops where the condition variable is not
   changed at all).
 
+- New :doc:`darwin-dispatch-once-nonstatic
+  <clang-tidy/checks/darwin-dispatch-once-nonstatic>`check.
+
+  Finds instances of ``dispatch_once_t`` variables not having static or global
+  storage.
+
 - New :doc:`linuxkernel-must-use-errs
   <clang-tidy/checks/linuxkernel-must-use-errs>` check.
 
   Checks Linux kernel code to see if it uses the results from the functions in
   ``linux/err.h``.
 
 - New :doc:`google-upgrade-googletest-case
   <clang-tidy/checks/google-upgrade-googletest-case>` check.

clang-tools-extra/trunk/docs/clang-tidy/checks/darwin-dispatch-once-nonstatic.rst

+.. title:: clang-tidy - darwin-dispatch-once-nonstatic
+
+darwin-dispatch-once-nonstatic
+==============================
+
+Finds declarations of ``dispatch_once_t`` variables without static or global
+storage. The behavior of using ``dispatch_once_t`` predicates with automatic or
+dynamic storage is undefined by libdispatch, and should be avoided.
+
+It is a common pattern to have functions initialize internal static or global
+data once when the function runs, but programmers have been known to miss the
+static on the ``dispatch_once_t`` predicate, leading to an uninitialized flag
+value at the mercy of the stack.
+
+Programmers have also been known to make ``dispatch_once_t`` variables be
+members of structs or classes, with the intent to lazily perform some expensive
+struct or class member initialization only once; however, this violates the
+libdispatch requirements.
+
+See the discussion section of
+`Apple's dispatch_once documentation <https://developer.apple.com/documentation/dispatch/1447169-dispatch_once>`_
+for more information.

clang-tools-extra/trunk/docs/clang-tidy/checks/list.rst

.. toctree::
    cppcoreguidelines-pro-type-cstyle-cast
    cppcoreguidelines-pro-type-member-init
    cppcoreguidelines-pro-type-reinterpret-cast
    cppcoreguidelines-pro-type-static-cast-downcast
    cppcoreguidelines-pro-type-union-access
    cppcoreguidelines-pro-type-vararg
    cppcoreguidelines-slicing
    cppcoreguidelines-special-member-functions
+   darwin-dispatch-once-nonstatic
    fuchsia-default-arguments-calls
    fuchsia-default-arguments-declarations
    fuchsia-header-anon-namespaces (redirects to google-build-namespaces) <fuchsia-header-anon-namespaces>
    fuchsia-multiple-inheritance
    fuchsia-overloaded-operator
    fuchsia-restrict-system-includes
    fuchsia-statically-constructed-objects
    fuchsia-trailing-return

clang-tools-extra/trunk/test/clang-tidy/darwin-dispatch-once-nonstatic.mm

+// RUN: %check_clang_tidy %s darwin-dispatch-once-nonstatic %t
+
+typedef int dispatch_once_t;
+extern void dispatch_once(dispatch_once_t *pred, void(^block)(void));
+
+
+void bad_dispatch_once(dispatch_once_t once, void(^block)(void)) {}
+// CHECK-MESSAGES: :[[@LINE-1]]:24: warning: dispatch_once_t variables must have static or global storage duration; function parameters should be pointer references [darwin-dispatch-once-nonstatic]
+
+// file-scope dispatch_once_ts have static storage duration.
+dispatch_once_t global_once;
+static dispatch_once_t file_static_once;
+namespace {
+dispatch_once_t anonymous_once;
+} // end anonymous namespace
+
+int Correct(void) {
+  static int value;
+  static dispatch_once_t once;
+  dispatch_once(&once, ^{
+    value = 1;
+  });
+  return value;
+}
+
+int Incorrect(void) {
+  static int value;
+  dispatch_once_t once;
+  // CHECK-MESSAGES: :[[@LINE-1]]:3: warning: dispatch_once_t variables must have static or global storage duration [darwin-dispatch-once-nonstatic]
+  // CHECK-FIXES: static dispatch_once_t once;
+  dispatch_once(&once, ^{
+    value = 1;
+  });
+  return value;
+}
+
+struct OnceStruct {
+  static dispatch_once_t staticOnce; // Allowed
+  int value;
+  dispatch_once_t once;  // Allowed (at this time)
+};
+
+@interface MyObject {
+  dispatch_once_t _once;
+  // CHECK-MESSAGES: :[[@LINE-1]]:3: warning: dispatch_once_t variables must have static or global storage duration and cannot be Objective-C instance variables [darwin-dispatch-once-nonstatic]
+  // CHECK-FIXES: dispatch_once_t _once;
+}
+@end