Changes in warn-weak-vtables.cpp

807e2f12 - Revert "Remove -Wweak-template-vtables"

Mon, 07 Feb 2022 15:47:14 +0000

Revert "Remove -Wweak-template-vtables"Turns out there's still some code referencing this. No harm in keeping it in abit longer.> as it was planned for removal in clang 15 and we're now past the branch point>> See https://github.com/llvm/llvm-project/issues/19107>> Differential revision: https://reviews.llvm.org/D118762This reverts commit 564f9be11c9cb8d131f48df07538fab7a19b41a7. List of files: /llvm-project-15.0.7/clang/test/SemaCXX/warn-weak-vtables.cpp

564f9be1 - Remove -Wweak-template-vtables

Wed, 02 Feb 2022 09:57:39 +0000

Remove -Wweak-template-vtablesas it was planned for removal in clang 15 and we're now past the branch pointSee https://github.com/llvm/llvm-project/issues/19107Differential revision: https://reviews.llvm.org/D118762 List of files: /llvm-project-15.0.7/clang/test/SemaCXX/warn-weak-vtables.cpp

4d2765e9 - Re-instate -Wweak-template-vtables as a no-op flag

Thu, 28 Oct 2021 11:37:27 +0000

Re-instate -Wweak-template-vtables as a no-op flagFollow-up to 8c136805242014b6ad9ff1afcac9d7f4a18bec3f to allow a lessabrupt migration for users.Differential revision: https://reviews.llvm.org/D112704 List of files: /llvm-project-15.0.7/clang/test/SemaCXX/warn-weak-vtables.cpp

8c136805 - PR18733: Remove -Wweak-template-vtables

Sun, 25 Jul 2021 18:47:01 +0000

PR18733: Remove -Wweak-template-vtablesIt isn't really pulling its weight and I think splitting it out from-Wweak-vtables was the wrong call: I think it was just a bug in theoriginal warning, which was trying to not diagnose templateinstantiations, implicit or explicit. List of files: /llvm-project-15.0.7/clang/test/SemaCXX/warn-weak-vtables.cpp

b792e069 - Disable -Wweak-vtables when there are no key functions

Tue, 06 Dec 2016 21:44:41 +0000

Disable -Wweak-vtables when there are no key functionsOur -Wweak-vtables diagnostic is powered by our key functioncalculation, which checks if key functions are enabled. We won't findany key functions in C++ ABIs that lack key functions, so -Wweak-vtableswas warning on every dynamic class before this change. So, turn off thiswarning in ABIs without key functions.Addresses PR31220llvm-svn: 288850 List of files: /llvm-project-15.0.7/clang/test/SemaCXX/warn-weak-vtables.cpp

b3a9978d - Don't let virtual calls and dynamic casts call Sema::MarkVTableUsed().

Mon, 26 Jan 2015 06:23:36 +0000

Don't let virtual calls and dynamic casts call Sema::MarkVTableUsed().clang currently calls MarkVTableUsed() for classes that get their virtualmethods called or that participate in a dynamic_cast. This is unnecessary,since CodeGen only emits vtables when it generates constructor, destructor, andvtt code. (*)Note that Sema::MarkVTableUsed() doesn't cause the emission of a vtable.Its main user-visible effect is that it instantiates virtual member functionsof template classes, to make sure that if codegen decides to write a vtableall the entries in the vtable are defined.While this shouldn't change the behavior of codegen (other than being faster),it does make clang more permissive: virtual methods of templates (in particulardestructors) end up being instantiated less often. In particular, classes thathave members that are smart pointers to incomplete types will now get theirimplicit virtual destructor instantiated less frequently. For example, thisused to not compile but does now compile: template struct OwnPtr { ~OwnPtr() { static_assert((sizeof(T) > 0), "TypeMustBeComplete"); } }; class ScriptLoader; struct Base { virtual ~Base(); }; struct Sub : public Base { virtual void someFun() const {} OwnPtr m_loader; }; void f(Sub *s) { s->someFun(); }The more permissive behavior matches both gcc (where this is not oftenobservable, since in practice most things with virtual methods have a keyfunction, and Sema::DefineUsedVTables() skips vtables for classes with keyfunctions) and cl (which is my motivation for this change) – this fixesPR20337. See this issue and the review thread for some discussions aboutoptimizations.This is similar to r213109 in spirit. r225761 was a prerequisite for thischange.Various tests relied on "a->f()" marking a's vtable as used (in the semasense), switch these to just construct a on the stack. This forcesinstantiation of the implicit constructor, which will mark the vtable as used.(*) The exception is -fapple-kext mode: In this mode, qualified calls tovirtual functions (`a->Base::f()`) still go through the vtable, and since thevtable pointer off this doesn't point to Base's vtable, this needs to referenceBase's vtable directly. To keep this working, keep referencing the vtable forvirtual calls in apple kext mode.llvm-svn: 227073 List of files: /llvm-project-15.0.7/clang/test/SemaCXX/warn-weak-vtables.cpp

c9bd88e6 - Remove the -cxx-abi command-line flag.

Tue, 14 Jan 2014 19:35:09 +0000

Remove the -cxx-abi command-line flag.This makes the C++ ABI depend entirely on the target: MS ABI for -win32 triples,Itanium otherwise. It's no longer possible to do weird combinations.To be able to run a test with a specific ABI without constraining it to aspecific triple, new substitutions are added to lit: %itanium_abi_triple and%ms_abi_triple can be used to get the current target triple adjusted to thedesired ABI. For example, if the test suite is running with the i686-pc-win32target, %itanium_abi_triple will expand to i686-pc-mingw32.Differential Revision: http://llvm-reviews.chandlerc.com/D2545llvm-svn: 199250 List of files: /llvm-project-15.0.7/clang/test/SemaCXX/warn-weak-vtables.cpp

9125b08b - Update tests in preparation for using the MS ABI for Win32 targets

Mon, 13 Jan 2014 19:48:13 +0000

Update tests in preparation for using the MS ABI for Win32 targetsIn preparation for making the Win32 triple imply MS ABI mode,make all tests pass in this mode, or make them use the Itaniummode explicitly.Differential Revision: http://llvm-reviews.chandlerc.com/D2401llvm-svn: 199130 List of files: /llvm-project-15.0.7/clang/test/SemaCXX/warn-weak-vtables.cpp

72b61203 - Provide a separate warning for weak vtables in explicit template instantiations. There's no (current) way to fix such templates to emit strong symbols/vtables, but perhaps users want to know about the cost being incurred anyway.

Fri, 09 Dec 2011 18:32:50 +0000

Provide a separate warning for weak vtables in explicit template instantiations. There's no (current) way to fix such templates to emit strong symbols/vtables, but perhaps users want to know about the cost being incurred anyway.llvm-svn: 146265 List of files: /llvm-project-15.0.7/clang/test/SemaCXX/warn-weak-vtables.cpp

34bc6e5e - When checking for weak vtables, check whether the actual definition of

Fri, 23 Sep 2011 19:04:03 +0000

When checking for weak vtables, check whether the actual definition ofthe key function is inline, rather than the originaldeclaration. Perhaps FunctionDecl::isInlined() is poorly named. Fixes. llvm-svn: 140400 List of files: /llvm-project-15.0.7/clang/test/SemaCXX/warn-weak-vtables.cpp

88d292cc - Rework when and how vtables are emitted, by tracking where vtables are

Thu, 13 May 2010 16:44:06 +0000

Rework when and how vtables are emitted, by tracking where vtables are"used" (e.g., we will refer to the vtable in the generated code) andwhen they are defined (i.e., because we've seen the key functiondefinition). Previously, we were effectively tracking "potentialdefinitions" rather than uses, so we were a bit too eager about emittingvtables for classes without key functions. The new scheme: - For every use of a vtable, Sema calls MarkVTableUsed() to indicate the use. For example, this occurs when calling a virtual member function of the class, defining a constructor of that class type, dynamic_cast'ing from that type to a derived class, casting to/through a virtual base class, etc. - For every definition of a vtable, Sema calls MarkVTableUsed() to indicate the definition. This happens at the end of the translation unit for classes whose key function has been defined (so we can delay computation of the key function; see PR6564), and will also occur with explicit template instantiation definitions. - For every vtable defined/used, we mark all of the virtual member functions of that vtable as defined/used, unless we know that the key function is in another translation unit. This instantiates virtual member functions when needed. - At the end of the translation unit, Sema tells CodeGen (via the ASTConsumer) which vtables must be defined (CodeGen will define them) and which may be used (for which CodeGen will define the vtables lazily). From a language perspective, both the old and the new schemes arepermissible: we're allowed to instantiate virtual member functionswhenever we want per the standard. However, all other C++ compilerswere more lazy than we were, and our eagerness was both a performanceissue (we instantiated too much) and a portability problem (we brokeBoost test cases, which now pass).Notes: (1) There's a ton of churn in the tests, because the order in which vtables get emitted to IR has changed. I've tried to isolate some of the larger tests from these issues. (2) Some diagnostics related to implicitly-instantiated/implicitly-defined virtual member functions have moved to the point of first use/definition. It's better this way. (3) I could use a review of the places where we MarkVTableUsed, to see if I missed any place where the language effectively requires a vtable.Fixes PR7114 and PR6564.llvm-svn: 103718 List of files: /llvm-project-15.0.7/clang/test/SemaCXX/warn-weak-vtables.cpp

4b73cfab - rename llvm::llvm_report_error -> llvm::report_fatal_error

Wed, 07 Apr 2010 22:58:06 +0000

rename llvm::llvm_report_error -> llvm::report_fatal_errorllvm-svn: 100708 List of files: /llvm-project-15.0.7/clang/test/SemaCXX/warn-weak-vtables.cpp

0da714a3 - Implement a warning diagnostic for weak vtables. Fixes PR6116.

Sat, 06 Feb 2010 02:27:10 +0000

Implement a warning diagnostic for weak vtables. Fixes PR6116.llvm-svn: 95472 List of files: /llvm-project-15.0.7/clang/test/SemaCXX/warn-weak-vtables.cpp