Sean Fertile via Phabricator <revi...@reviews.llvm.org> writes: > sfertile added inline comments. > > > ================ > Comment at: clang/lib/CodeGen/CodeGenModule.cpp:750 > + // If we can use a plt entry as the symbol address we can assume it > + // is local. > + if (isa<FunctionDecl>(D) && !CGOpts.NoPLT) > ---------------- > I don't think this is the case. I think this would break ppc, where we need > to restore the toc pointer after the plt stubs returns to the original call > site.
Excellent catch. I will update the patch to handle the ppc special case. Thanks, Rafael _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
