linux-mips
[Top] [All Lists]

Re: [PATCH]: fix possible buffer overflow problem in promlib

To: Ralf Baechle <ralf@linux-mips.org>
Subject: Re: [PATCH]: fix possible buffer overflow problem in promlib
From: Thiemo Seufer <ica2_ts@csv.ica.uni-stuttgart.de>
Date: Fri, 3 Jan 2003 16:58:32 +0100
Cc: Geert Uytterhoeven <geert@linux-m68k.org>, Juan Quintela <quintela@mandrakesoft.com>, mipslist <linux-mips@linux-mips.org>
In-reply-to: <20030103164832.A11536@linux-mips.org>
Original-recipient: rfc822;linux-mips@linux-mips.org
References: <m2hecrbb3p.fsf@demo.mitica> <Pine.GSO.4.21.0301022047090.4873-100000@vervain.sonytel.be> <20030103135422.A7796@linux-mips.org> <20030103142956.GI31256@rembrandt.csv.ica.uni-stuttgart.de> <20030103164832.A11536@linux-mips.org>
Sender: linux-mips-bounce@linux-mips.org
User-agent: Mutt/1.4i
Ralf Baechle wrote:
> On Fri, Jan 03, 2003 at 03:29:56PM +0100, Thiemo Seufer wrote:
> 
> > > > What about making ppbuf static, to reduce stack usage?
> > > 
> > > By the time when prom_printf() is called stack overflow is not really a
> > > consideration anymore, something fatal has happened before.
> > 
> > I don't think printing e.g. "LINUX started..." indicates something fatal. 
> > :-)
> 
> So use printk :-)

printk might not be ready at that time. I grepped the example from
arch/mips/mips-boards/generic/init.c, and there are some more
occurences in arch/mips which seem to be non-fatal.


Thiemo

<Prev in Thread] Current Thread [Next in Thread>