From owner-freebsd-questions Tue Oct 30 9:46:22 2001 Delivered-To: freebsd-questions@freebsd.org Received: from dan.emsphone.com (dan.emsphone.com [199.67.51.101]) by hub.freebsd.org (Postfix) with ESMTP id 6AACE37B408 for ; Tue, 30 Oct 2001 09:46:20 -0800 (PST) Received: (from dan@localhost) by dan.emsphone.com (8.11.6/8.11.6) id f9UHkG568519; Tue, 30 Oct 2001 11:46:16 -0600 (CST) (envelope-from dan) Date: Tue, 30 Oct 2001 11:46:16 -0600 From: Dan Nelson To: Mihai Chelaru Cc: freebsd-questions@FreeBSD.ORG Subject: Re: Pthread implementation Message-ID: <20011030114616.A55278@dan.emsphone.com> References: <3BDE7927.66CB49FA@romtelenet.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <3BDE7927.66CB49FA@romtelenet.org> User-Agent: Mutt/1.3.23i X-OS: FreeBSD 5.0-CURRENT X-message-flag: Outlook Error Sender: owner-freebsd-questions@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.ORG In the last episode (Oct 30), Mihai Chelaru said: > I have a problem using freebsd pthreads. After i launch a number of > aprox. 10000 threads (not concomitent but in a period of 1 day) the > process suddenly goes up to 99% of CPU. i did a truss on that process > and look what i got: > > gettimeofday(0x28125188,0x0) = 0 (0x0) > poll(0x8070000,0x4,0x253a) = 1 (0x1) > gettimeofday(0x28125188,0x0) = 0 (0x0) > poll(0x8070000,0x4,0x253a) = 1 (0x1) See if you can strip your program down the minimal amount required to trigger the bug. It could be a pthreads bug, or it could be a bug in your program. Also try running strace (in ports); it'll tell you which filedescriptor is ready according to poll. -- Dan Nelson dnelson@allantgroup.com To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-questions" in the body of the message