Date: Fri, 13 May 2005 15:39:53 +0300 From: Giorgos Keramidas <keramida@freebsd.org> To: Andre Guibert de Bruet <andy@siliconlandmark.com> Cc: freebsd-current@freebsd.org Subject: Re: Trimming top's header to display width Message-ID: <20050513123953.GB26764@orion.daedalusnetworks.priv> In-Reply-To: <20050513083059.W820@lexi.siliconlandmark.com> References: <20050512123637.GA2966@orion.daedalusnetworks.priv> <20050513083059.W820@lexi.siliconlandmark.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On 2005-05-13 08:33, Andre Guibert de Bruet <andy@siliconlandmark.com> wrote:
> On Thu, 12 May 2005, Giorgos Keramidas wrote:
> >Does the diff below look like a good way to fix this?
> >
> >%%%
> >+char *
> >+trim_header(text)
> >+
> >+char *text;
> >+
> >+{
> >+ char *s;
> >+ int width;
> >+
> >+ s = NULL;
>
> Setting s to NULL is useless because malloc returns either NULL on failure
> or a pointer to the allocated memory segment.
Setting s to NULL unconditionally means that even if the if block right
below doesn't run, the pointer *WILL* be initialized to something
meaningful.
>
> >+ width = display_width;
> >+ header_length = strlen(text);
> >+ if (header_length >= width) {
This part won't run all the time. If it doesn't, then `s' could
potentially hold garbage from the stack and the callers of trim_header()
would find it very surprising that their non-NULL pointer causes a
segfault in free() :-)
> Other than that pet-peeve, the patch looks good to me...
Cool :)
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20050513123953.GB26764>