Comment on attachment 8819703 [details] [diff] [review] Port the must-override check to the clang-tidy compatible framework Review of attachment 8819703 [details] [diff] [review]: ----------------------------------------------------------------- ::: build/clang-plugin/MustOverrideChecker.cpp @@ +9,5 @@ > + AstMatcher->addMatcher(cxxRecordDecl(isDefinition()).bind("class"), this); > +} > + > +void MustOverrideChecker::registerPPCallbacks(CompilerInstance& CI) { > + this->CI = &CI; I really don't like this, especially considering that in our "normal" use-case of being directly inside of clang instead of inside of clang-tidy this isn't a problem. I can't think of any other way to get access to the Sema object off of my head though, and re-implementing Sema::IsOverload in tree would be kinda nuts.

I share the same feeling. :(

Pushed by eakhgari@mozilla.com: https://hg.mozilla.org/integration/mozilla-inbound/rev/7d0b2be36233 Port the must-override check to the clang-tidy compatible framework; r=mystor

Pushed by eakhgari@mozilla.com: https://hg.mozilla.org/integration/mozilla-inbound/rev/70a3d7e03d55 follow-up: Fix a mistake that breaks the build

https://hg.mozilla.org/mozilla-central/rev/7d0b2be36233 https://hg.mozilla.org/mozilla-central/rev/70a3d7e03d55