This is the mail archive of the
gcc-bugs@gcc.gnu.org
mailing list for the GCC project.
Re: ICE during bootstrap with --enable-checking
- To: aj at suse dot de
- Subject: Re: ICE during bootstrap with --enable-checking
- From: "Martin v. Loewis" <martin at loewis dot home dot cs dot tu-berlin dot de>
- Date: Tue, 14 Mar 2000 12:12:08 +0100
- CC: gcc-bugs at gcc dot gnu dot org, <mark at codesourcery dot com>
- References: <hoem9e7gse.fsf@maclaurin.suse.de>
> Shall I send the preprossed source? It should be easy to reproduce
> this - otherwise tell me.
I was just thinking about my previous claim that it might be a GC
problem. Still without having reproduced the problem: What in gcc
currently guarantees that tree nodes that appear as semantic values of
bison are marked as GC roots?
Or, put it differently: We have
extension:
EXTENSION
{ $$ = build_int_2 (pedantic | (warn_pointer_arith << 1), 0);
pedantic = 0;
warn_pointer_arith = 0; }
;
$$ is the only reference to the integer, and that eventually lives in
yyparse:yyvsa. Is there any guarantee it won't be collected?
Martin