This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: PATCH: Fix PR 8391
- From: Mark Mitchell <mark at codesourcery dot com>
- To: Geoff Keating <geoffk at geoffk dot org>
- Cc: "gcc-patches at gcc dot gnu dot org" <gcc-patches at gcc dot gnu dot org>
- Date: Fri, 01 Nov 2002 18:03:41 -0800
- Subject: Re: PATCH: Fix PR 8391
No, it's the reverse! The concepts involved (inlining, vtables,
linker features) are all known primarily to the backend, and so the
backend should be deciding what functions to output and how.
Sorry; I completely disagree.
It's not the back end's job to know what functions the C++ ABI might
require it to output. I don't in any way object to feedback from the
back end (such as TREE_SYMBOL_REFERENCED) that can be used to guide
the front end in deciding what's needed and what's not, but
rest_of_compilation should be the function the front end calls when
it wants to spit out code. It doesn't have to do at that instant,
but the back end should be obligated to emit code for that function
at some point.
If the back end tries to get clever, we can end up with problems like
the one I fixed, or others that simply manifest as missing symbols in
the .o files.
--
Mark Mitchell mark@codesourcery.com
CodeSourcery, LLC http://www.codesourcery.com