Re: Standard IO
On Thu, 15 Mar 2001, Theodore Tso wrote:
> I'm not a glibc expert (and I don't play one on TV), but it looks like
> that on a Debian/testing system using glibc 2.2.2, stdout gets mangled
> using symbol versioning to become stdout@@GLIBC_2.0.
I need to clarify a misconception that I opriginally had, and have
probably spread. stdout@@GLIBC_2.0 is not as badly mangled as originally
though. In fact, the symbol name is correct (stdout), but it has a version
associated with it (GLIBC_2.0).
The symtax with the @@ in the middle is simply nm's way of expressing this
info. It screws up script that aren't aware of this. You have to study the
ELF structure to figure this out though.
The result it instead of having to deal with hideous mangles, we have to
explicitely define the version of every symbol instead. The current version
of the spec covers this now.
This isn't to say that we don't have interesting mangle to deal with, just
that they aren't as many as we originally though.
Stuart
Stuart R. Anderson anderson@metrolink.com
Metro Link Incorporated South Carolina Office
5807 North Andrews Way 129 Secret Cove Drive
Fort Lauderdale, Florida 33309 Lexington, SC 29072
voice: 954.660.2500 voice: 803.951.3630
http://www.metrolink.com/ XFree86 Core Team
Creative Applications Lab Chair - SIGGRAPH 2001
Reply to: