[Bug middle-end/38002] gcc crash using -fvisibility-ms-compat

2010-07-30 Thread binocs38149 at mypacks dot net
--- Comment #6 from binocs38149 at mypacks dot net 2010-07-30 19:01 --- (In reply to comment #5) Apple seems to have fixed it a different way: http://www.opensource.apple.com/source/gcc/gcc-5659/gcc/cp/decl2.c { tree underlying_type = TREE_TYPE (DECL_NAME (decl)); int

[Bug middle-end/38002] gcc crash using -fvisibility-ms-compat

2009-10-27 Thread boz_gnu at boz dot org dot uk
--- Comment #5 from boz_gnu at boz dot org dot uk 2009-10-27 16:22 --- Created an attachment (id=18917) --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=18917&action=view) Proposed patch Looks like all we were missing was a check on CLASS_TYPE_P before calling CLASSTYPE_VISIBILITY_SP

[Bug middle-end/38002] gcc crash using -fvisibility-ms-compat

2009-10-26 Thread boz_gnu at boz dot org dot uk
--- Comment #4 from boz_gnu at boz dot org dot uk 2009-10-26 17:39 --- (In reply to comment #3) > Reduced testcase, not a regression. Here's a further reduction which still reproduces the crash: namespace std __attribute__ ((__visibility__ ("default"))) { class type_info { }; };

[Bug middle-end/38002] gcc crash using -fvisibility-ms-compat

2008-11-03 Thread rguenth at gcc dot gnu dot org
--- Comment #3 from rguenth at gcc dot gnu dot org 2008-11-03 16:42 --- Reduced testcase, not a regression. namespace std __attribute__ ((__visibility__ ("default"))) { class type_info { }; }; template void throw_exception(E const & e); class bad_lexical_cast { public: bad_le