So we really don't want this fix. If there are tests that are expecting "char" to show up, those tests should be fixed to deal unsigned chars. We really want any variable that is actually unsigned to be represented as unsigned in the clang::ASTContext objects we create for each shared library otherwise if you do expressions with those values, they might behave differently from how they would behave if you compiled them. There is no data in the DWARF that tells us that "char" should be treated as unsigned, so we can't create our ASTContext objects that way. So we need to represent variables correctly. So the tests should be fixed or conditional based on the platform and this fix should either turn into fixing the tests, or abandoned.
http://reviews.llvm.org/D8636 EMAIL PREFERENCES http://reviews.llvm.org/settings/panel/emailpreferences/ _______________________________________________ lldb-commits mailing list [email protected] http://lists.cs.uiuc.edu/mailman/listinfo/lldb-commits
