Received: from SOUTH-STATION-ANNEX.MIT.EDU by po7.MIT.EDU (5.61/4.7) id AA17024; Thu, 7 Dec 95 22:14:30 EST
Received: from uniwa.uwa.edu.au by MIT.EDU with SMTP
	id AA26438; Thu, 7 Dec 95 22:13:08 EST
Received: from ee.uwa.edu.au (swanee.ee.uwa.edu.au [130.95.208.1]) by uniwa.uwa.edu.au (8.6.11/8.6.9) with SMTP id LAA07705 for <proven@MIT.EDU>; Fri, 8 Dec 1995 11:14:12 +0800
Received: by s15.ee.uwa.edu.au (4.1/SMI-4.1)
	id AA00381; Fri, 8 Dec 95 11:16:34 WST
Date: Fri, 8 Dec 95 11:16:34 WST
From: chua_dy@ee.uwa.edu.au (darren chua)
Message-Id: <9512080316.AA00381@s15.ee.uwa.edu.au>
To: proven@MIT.EDU
Subject: Re: pthread_cond_wait

Hi Chris,

About the benchmarking routine I was talking about.... I've found out what I was doing wrong.

I wasn't implementing the pthread_joins properly, and I was eating up CPU.

In fact, after more testing, your pthread_cond_wait is taking up about 
10milliseconds, compared to 200milliseconds. That's a lot better, but it 
is still vary large for a real-time application.

(I'm using beta4_1, running on linux kernel 1.2.13)

Thanks, 
Darren Chua.

PS: Really sorry for wasting your time earlier.

