Date: Sat, 28 Apr 2001 12:59:29 +0200 From: Erik Trulsson <ertr1013@student.uu.se> To: stable@FreeBSD.ORG Subject: Re: Trouble with 4.3-RELEASE compiler Message-ID: <20010428125928.A5681@student.uu.se> In-Reply-To: <20010428093802.59828860.steveo@eircom.net>; from steveo@eircom.net on Sat, Apr 28, 2001 at 09:38:02AM %2B0200 References: <20010427194022.A18639@roma.coe.ufrj.br> <200104280127330789.011EEEF5@tensor.xs4all.nl> <20010428093802.59828860.steveo@eircom.net>
next in thread | previous in thread | raw e-mail | index | archive | help
On Sat, Apr 28, 2001 at 09:38:02AM +0200, Steve O'Hara-Smith wrote: > On Sat, 28 Apr 2001 01:27:33 +0200 > "Dimitry Andric" <dim@xs4all.nl> wrote: > > DA> Squid with gcc 2.95.2 and optimization (both -O, -O2 and -O666), and > DA> I can assure you it bombed out with inexplicable null pointer > DA> accesses. Yet when you compile with -O0, no such thing happens... > > I have been working getting swish++ set up as a port (4.3-STABLE) and > I've found that the search program only works if compiled with no -O setting, > -O3 (the original) and -O cause segmentation violation, while -O2 gave an > illegal instruction trap. The index program OTOH appears to work with all > optimisations settings. > > It makes me wonder just how safe -O is :( > Although it is quite possible that gcc generates incorrect code in some cases when invoked with -O it is not very likely. I would say that it is much more likely that the code which is being compiled contains a bug that is exposed by the optimization and that the code just happens to work when comiled with -O0. Generally I would say that -O is *safer* than -O0 for the simple reason that it is used more and therefore gets more testing. -- <Insert your favourite quote here.> Erik Trulsson ertr1013@student.uu.se To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-stable" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20010428125928.A5681>