c/2160: signal 11 when compiling large generated program
Joseph S. Myers
jsm28@cam.ac.uk
Mon Jul 23 03:06:00 GMT 2001
The following reply was made to PR c/2160; it has been noted by GNATS.
From: "Joseph S. Myers" <jsm28@cam.ac.uk>
To: <anton@mips.complang.tuwien.ac.at>
Cc: <gcc-bugs@gcc.gnu.org>, <gcc-gnats@gcc.gnu.org>
Subject: Re: c/2160: signal 11 when compiling large generated program
Date: Mon, 23 Jul 2001 11:05:32 +0100 (BST)
On Mon, 23 Jul 2001, Anton Ertl wrote:
> > (not sent to gcc-gnats, so not filed with this PR).
>
> Should I do something else than replying to all recipients of your mail?
It needs to go to gcc-gnats@gcc.gnu.org to be filed with the PR. GNATS
status change messages used to include this address in the headers, but it
seems they don't any more. This is a bug in the bug tracking system
configuration.
> > Excess memory consumption is still a bug; changed from
> > ice-on-legal-code to sw-bug.
>
> I finally managed to compile this file on an Alpha with 1GB with
> gcc-3.0 with
>
> gcc -mieee -c engine.i
>
> Top showed about 1550MB virtual memory consumption during most of the
> time. Top showed that the CPU was mostly in system or idle mode, with
> only about 0.5% user time; took about 30 hours. The compiler produced
> a file engine.o, so I assume the compilation went ok (unfortunately I
> was not smart enough to log the output and status).
>
> - anton
--
Joseph S. Myers
jsm28@cam.ac.uk
More information about the Gcc-prs
mailing list