From mboxrd@z Thu Jan  1 00:00:00 1970
From: Ian Jackson <Ian.Jackson@eu.citrix.com>
Subject: Re: [PATCH 24/29] libxl: Introduce FILLZERO
Date: Tue, 31 Mar 2015 19:51:16 +0100
Message-ID: <21786.60580.482169.516456@mariner.uk.xensource.com>
References: <1423599016-32639-1-git-send-email-ian.jackson@eu.citrix.com>
	<1423599016-32639-25-git-send-email-ian.jackson@eu.citrix.com>
	<1427198621.21742.372.camel@citrix.com>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
Return-path: <xen-devel-bounces@lists.xen.org>
In-Reply-To: <1427198621.21742.372.camel@citrix.com>
List-Unsubscribe: <http://lists.xen.org/cgi-bin/mailman/options/xen-devel>,
	<mailto:xen-devel-request@lists.xen.org?subject=unsubscribe>
List-Post: <mailto:xen-devel@lists.xen.org>
List-Help: <mailto:xen-devel-request@lists.xen.org?subject=help>
List-Subscribe: <http://lists.xen.org/cgi-bin/mailman/listinfo/xen-devel>,
	<mailto:xen-devel-request@lists.xen.org?subject=subscribe>
Sender: xen-devel-bounces@lists.xen.org
Errors-To: xen-devel-bounces@lists.xen.org
To: Ian Campbell <ian.campbell@citrix.com>
Cc: xen-devel@lists.xensource.com, Euan Harris <euan.harris@citrix.com>
List-Id: xen-devel@lists.xenproject.org

Ian Campbell writes ("Re: [Xen-devel] [PATCH 24/29] libxl: Introduce FILLZERO"):
> On Tue, 2015-02-10 at 20:10 +0000, Ian Jackson wrote:
> > FILLZERO is a macro for memset(&foo,0,sizeof(foo)).  It eliminates the
> > possiblity to make the error memset(&foo,0,sizeof(&foo)).
> 
> but not:
>         foo *p = allocate_a_foo()
>      memset(p, 0, sizeof(p))
> although that's probably less likely to go wrong and I don't think it
> can be avoided by the sorts of tricks used here.

Well, we could make a macro that you _have_ to pass the pointer to.
The result is that passing a non-pointer object gets you an error.

> > No callers yet, but document it in CODING_STYLE.  (In accordance with
> > existing libxl policy, I haven't gone through all existing possible
> > call sites.)
> 
> We don't usually expose such helpers in the public API, but I suppose
> you have a good reason to do so here, could you mention it in the commit
> log please.

It should be in the public API because xl*.c is full of uses of memset
which ought to be replaced with FILLZERO.  Do you really think I need
to mention this in the commit message ?

> > +#define LIBXL_FILLZERO(object) (memset(&(object), 0, sizeof((object))))
> 
> Evaluates object twice, so LIBXL_FILEZERO(*(p++)), would behave
> surprisingly. I'm not sure if this can be resolved though, so this might
> be a Don't Do That Then situation.

sizeof() does not evaluate its argument.

Ian.