Mailing-List: contact cygwin-help AT cygwin DOT com; run by ezmlm
Sender: cygwin-owner AT cygwin DOT com
Mail-Followup-To: cygwin AT cygwin DOT com
Message-ID: <4249BCDF.1040409@dunslane.net>
Date: Tue, 29 Mar 2005 15:38:55 -0500
From: Andrew Dunstan <andrew AT dunslane DOT net>
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.7.2) Gecko/20040906
MIME-Version: 1.0
To: cygwin AT cygwin DOT com
Subject: apparent scheduler problem
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit


Greetings Cygwin people!

I do some work with PostgreSQL, including running the build farm. 
Recently after upgrading my installation of Cygwin on XP-Pro, I noticed 
that PostgreSQL started failing one of its regression tests. Another 
user with a completely fresh Cygwin installation has noticed the same 
thing. The test is for the stats collector, and essentially it does some 
work, waits a couple of seconds in a fairly brain dead busy/wait loop, 
and then checks to see if the stats collector has done the work we 
expect it to have done. Nothing much has changed in this area of 
Postgres, certainly not since we know a successful test was done a few 
weeks ago. Some experimentation has shown that the expected result now 
appears when we ratchet the sleep right up to around 1 minute.

We are wondering if anything is known to have changed in Cygwin recently 
that could account for this. I regret to tell you I have no exact idea 
when I previously upgraded my Cygwin installation - certainly some time 
within the last 6 months. So, this would be something that happened not 
too long ago.

Or as Tom Lane, PostgreSQL uber-hacker, put it:

"It sounds to me like the problem is that the backend executing the test
script is in a tight loop (due to the half-baked implementation of sleep())
and for some reason this prevents the stats processes from running ---
for a far longer period than it by rights ought to.  Ask about recent
changes in process scheduling policy.  (I suppose that actually it's
Windows doing the scheduling, but what we want to know about is cygwin
changes that might have affected Windows scheduling parameters.)"

Any light you can shine on this would be helpful. We're certainly keen to continue to have Cygwin as a supported platform for PostgreSQL (and we'd also love it if we could do pltcl on Cygwin too, but tcl seems to be currently broken ;-( )

cheers

andrew


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