Mailing-List:	contact cygwin-developers-help AT sourceware DOT cygnus DOT com; run by ezmlm
List-Subscribe:	<mailto:cygwin-developers-subscribe AT sources DOT redhat DOT com>
List-Archive:	<http://sources.redhat.com/ml/cygwin-developers/>
List-Post:	<mailto:cygwin-developers AT sources DOT redhat DOT com>
List-Help:	<mailto:cygwin-developers-help AT sources DOT redhat DOT com>, <http://sources.redhat.com/ml/#faqs>
Sender:	cygwin-developers-owner AT sources DOT redhat DOT com
Delivered-To:	mailing list cygwin-developers AT sources DOT redhat DOT com
Subject:	Re: Checking input parameters of syscalls
From:	Robert Collins <robert DOT collins AT itdomain DOT com DOT au>
To:	egor duda <cygwin-developers AT cygwin DOT com>
In-Reply-To:	<1297610035.20010911125836@logos-m.ru>
References:	<20010910154431 DOT A792 AT dothill DOT com>
	<20010910222228 DOT X937 AT cygbert DOT vinschen DOT de>
	<1297610035 DOT 20010911125836 AT logos-m DOT ru>
X-Mailer:	Evolution/0.13 (Preview Release)
Date:	11 Sep 2001 21:52:18 +1000
Message-Id:	<1000209139.7266.183.camel@lifelesswks>
Mime-Version:	1.0
X-OriginalArrivalTime:	11 Sep 2001 11:39:09.0558 (UTC) FILETIME=[5F031160:01C13AB6]

On Tue, 2001-09-11 at 18:58, egor duda wrote:
> Hi!
> 
> Tuesday, 11 September, 2001 Corinna Vinschen cygwin-patches AT cygwin DOT com wrote:
> 
> CV> Frankly, I don't know.  My first guess is to prioritize correctness
> CV> over speed ...
> 
> [...]
> 
> Speaking about correctness. I'm planning yet another update to
> testsuite and found that many cygwin functions still crash when
> application passes an invalid pointer to them, instead of returning
> EFAULT. Should we leave this as it is (to avoid possible slowdown) or
> sprinkle check_null_empty_str and check_null_invalid_struct all
> around? quick testing shows that IsBadWritePtr() function is
> quite fast (~10 microsecinds per call on my K6/400MHz), so i think
> the latter is the right way to go.
> 

I created check_valid_pointer for jsut this reason in thread.cc I cannot
recall _why_ I didn't use the other two, but I recall some reason :}...

And yes, IMO such trivial checks are essential for any function that can
crash or corrupt cygwin - particularly with the daemon scenario :].

Rob

- Raw text -