This is the mail archive of the
libstdc++@gcc.gnu.org
mailing list for the libstdc++ project.
Re: [PATCH] Builtins and C++ and such (take 3)
- From: Roger Sayle <roger at eyesopen dot com>
- To: Benjamin Kosnik <bkoz at redhat dot com>
- Cc: Mark Mitchell <mark at codesourcery dot com>, "gcc-patches at gcc dot gnu dot org" <gcc-patches at gcc dot gnu dot org>, "libstdc++ at gcc dot gnu dot org" <libstdc++ at gcc dot gnu dot org>, Jason Merrill <jason at redhat dot com>
- Date: Mon, 1 Apr 2002 06:21:16 -0700 (MST)
- Subject: Re: [PATCH] Builtins and C++ and such (take 3)
Hi Ben,
> Roger's patch fixes the library bugs, but then gives:
>
> FAIL: g++.dg/inherit/template-as-base.C (test for excess errors)
> FAIL: g++.benjamin/tem03.C caused compiler crash
> FAIL: g++.benjamin/typedef01.C caused compiler crash
> FAIL: g++.brendan/line1.C caused compiler crash
> FAIL: g++.jason/crash7.C caused compiler crash
> FAIL: g++.mike/net44.C caused compiler crash
> FAIL: g++.other/typedef5.C caused compiler crash
There's something suspicious going on here. My one line change to
"decl.c:duplicate_decls" is very unlikely the cause of these regressions.
For example, during the compilation of g++.dg/inherit/template-as-base.C
using cc1plus duplicate_decls is never called! Which would also explain
why deleting that line altogether has no effect on the regression.
The other supporting evidence is that "-fno-builtin" also has no
effect on this regression, which disables almost all of the code in
my builtins patch. I'm investigating further...
It looks from the gcc-testresults mailing list that atleast one of the
above tests has started failing in an unpatched tree. Ben, can you
try reverting my one line fix in your tree, to determine which (if
any) of the above regressions are really caused by my fix.
Many thanks,
Roger
--
Roger Sayle, E-mail: roger@eyesopen.com
OpenEye Scientific Software, WWW: http://www.eyesopen.com/
Suite 1107, 3600 Cerrillos Road, Tel: (+1) 505-473-7385
Santa Fe, New Mexico, 87507. Fax: (+1) 505-473-0833