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
Date: Mon, 15 Jan 2001 14:16:13 -0500
From: Christopher Faylor <cgf AT redhat DOT com>
To: cygwin-developers AT cygwin DOT com
Subject: Re: Chris change to read
Message-ID: <20010115141613.A28484@redhat.com>
Reply-To: cygwin-developers AT cygwin DOT com
Mail-Followup-To: cygwin-developers AT cygwin DOT com
References: <3A633012 DOT 3D1713E1 AT yahoo DOT com> <20010115121919 DOT A27350 AT redhat DOT com> <3A634285 DOT A26BDBAA AT yahoo DOT com>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
User-Agent: Mutt/1.3.11i
In-Reply-To: <3A634285.A26BDBAA@yahoo.com>; from earnie_boyd@yahoo.com on Mon, Jan 15, 2001 at 01:33:41PM -0500

On Mon, Jan 15, 2001 at 01:33:41PM -0500, Earnie Boyd wrote:
>Christopher Faylor wrote:
>> 
>> On Mon, Jan 15, 2001 at 12:14:58PM -0500, Earnie Boyd wrote:
>> >I just wanted to praise you for speeding up Cygwin.  It's at least 15
>> >percent faster or more.  In my timing tests I currently get a faster
>> >initial output from `ls --color -l /bin' than I do in repeating the
>> >command.  I once had an initial timing of .65 seconds user time on an
>> >initial display; this was after having exited Cygwin processes, did some
>> >Netscape work, read some email and then restarted a Cygwin process.
>> 
>> Wow, thanks.  Where is cygwin faster?  Is this a recent change is the
>> snapshot faster than 1.1.7?  Or is 1.1.7 faster than 1.1.4?  I did
>> add some code to read to bypass signal interruption if there were
>> no signal handlers present which should have some speed improvements
>> but that was added in 1.1.6.
>> 
>
>I'm noticing the improvement in the current CVS as compared to 1.1.7.
>
>> I also just implemented SA_RESTART but I don't that would have an effect
>> on anything since it requires a reconfiguration and recompilation.
>> 
>
>A recompilation and reconfiguration of what, Cygwin?  I do that with 
>  make clean && make
>so if that is what you mean that is what I'm testing.

It requires a reconfiguration and recompilation of anything that would use
the SA_RESTART setting in sigaction, i.e., applications using cygwin.

>> DJ also did some performance analysis and targetted some areas that
>> needed tweaking.  This should have shown up in 1.1.6.  I was never
>> certain that there was a clear improvement, though, or I would have
>> announced this as a 1.1.6 change.
>> 
>
>Well then, it wasn't that.  Perhaps some of Corinna's mmap changes have
>something to do with it and the thanks belongs to her.  I was just
>assuming that your change to the read function was what the cause and
>affect was but maybe it's elsewhere.  However, the change is definitely
>noticeable.

Hmm.  Interesting.

cgf