X-Recipient: archive-cygwin AT delorie DOT com
X-SWARE-Spam-Status: No, hits=-2.1 required=5.0	tests=AWL,BAYES_00,KHOP_THREADED,SPF_NEUTRAL,T_RP_MATCHES_RCVD
X-Spam-Check-By: sourceware.org
Message-ID: <4FD53FA9.2040105@cornell.edu>
Date: Sun, 10 Jun 2012 20:45:29 -0400
From: Ken Brown <kbrown AT cornell DOT edu>
User-Agent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:12.0) Gecko/20120428 Thunderbird/12.0.1
MIME-Version: 1.0
To: cygwin AT cygwin DOT com
Subject: Re: Performance problems with emacs-X11 in current cygwin
References: <4FC7D9E6 DOT 5050609 AT alice DOT it> <4FCA1FF0 DOT 8090703 AT alice DOT it>	<4FCA2CA9 DOT 7080704 AT cornell DOT edu> <4FCA634D DOT 1080206 AT cornell DOT edu>	<4FCB2991 DOT 3010701 AT users DOT sourceforge DOT net>	<4FCB5438 DOT 7080903 AT cornell DOT edu> <4FCB9872 DOT 5010506 AT cornell DOT edu>	<loom DOT 20120606T123651-460 AT post DOT gmane DOT org>	<4FD1F709 DOT 4050107 AT cornell DOT edu> <87k3zhbyyk DOT fsf AT Rainer DOT invalid> <4FD22C39 DOT 6070107 AT cornell DOT edu>
In-Reply-To: <4FD22C39.6070107@cornell.edu>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-PMX-CORNELL-SPAM-CHECKED: Pawpaw
X-Original-Sender: kbrown AT cornell DOT edu - Sun Jun 10 20:45:31 2012
X-PMX-CORNELL-REASON: CU_White_List_Override
X-IsSubscribed: yes
Mailing-List: contact cygwin-help AT cygwin DOT com; run by ezmlm
List-Id: <cygwin.cygwin.com>
List-Subscribe: <mailto:cygwin-subscribe AT cygwin DOT com>
List-Archive: <http://sourceware.org/ml/cygwin/>
List-Post: <mailto:cygwin AT cygwin DOT com>
List-Help: <mailto:cygwin-help AT cygwin DOT com>, <http://sourceware.org/ml/#faqs>
Sender: cygwin-owner AT cygwin DOT com
Mail-Followup-To: cygwin AT cygwin DOT com
Delivered-To: mailing list cygwin AT cygwin DOT com

On 6/8/2012 12:45 PM, Ken Brown wrote:
> On 6/8/2012 11:33 AM, Achim Gratz wrote:
>> Ken Brown writes:
>>> As I said earlier, I don't understand very well how git branches work,
>>> but I *think* this means we have to look in the 2-32 branch, prior to
>>> the 2.31.0 tag, to find the problematic commit. I've checked out the
>>> 2-32 branch, and I guess the next step is to find a problem-free
>>> revision of that branch, and then bisect between it and the 2.31.0
>>> tag. I'm in the process of reading the git documentation to figure out
>>> how to do that, but I wouldn't object if someone would save me some
>>> time by giving me the appropriate git commands.
>>
>> I've had a quick look at how the GNOME folks use their release branches:
>> they are tagged in master and then only some version bumping and a few
>> quickfixes. There are no odd numbered releases, so I assume they start
>> the disruptive changes right after a release, tag the unstable version
>> in master with an odd number and then work out the kinks until the new
>> release is done.
>>
>> So, you can indeed start on the 2.32 branch and then bisect down to the
>> 2.30 tag. Don't bother with the run-up between 2.31 and 2.32, just
>> bisect it whole, the bisect sequence will be just one build longer if at
>> all.
>>
>> git checkout glib-2-32
>> git bisect start bad
>> git bisect good 2.30.3
>>
>> If any of the intermediate versions doesn't build, say
>>
>> git bisect skip
>>
>> with the offending commit still checked out.
>
> Thanks, Achim. That helps a lot. The only thing I might have to change
> is the starting point for the bisection, since the tag 2.30.3 represents
> a fairly recent commit. But I think starting with 2.30.1 should work.
> I'll give it a try.

The bisection shows that the first problematic commit is this one:

http://git.gnome.org/browse/glib/commit/?h=glib-2-32&id=7eae486179e2799c369ed9ffcea663bf9161ce79

Author: Ryan Lortie <desrt AT desrt DOT ca>
Date:   Wed Aug 31 22:07:02 2011 -0400

     GMain: simplify logic for g_wakeup_acknowledge()

     Instead of messing around with context->poll_waiting, just look at 
the GPollFD to see if the GWakeup needs to be acknowledged.

In case anyone else wants to confirm this, you can get my glib builds by 
running

   setup.exe -K http://sanibeltranquility.com/cygwin/kbrown.gpg

and adding http://sanibeltranquility.com/cygwin to the list of mirrors. 
  The problematic version is

   libglib2.0_0-2.30.90_7eae4861-1

and the preceding version (without the problem) is

   libglib2.0_0-2.30.90_87880df-1

I've tested the latter with emacs-23, emacs-24, and gvim.

Ken

--
Problem reports:       http://cygwin.com/problems.html
FAQ:                   http://cygwin.com/faq/
Documentation:         http://cygwin.com/docs.html
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple