From mboxrd@z Thu Jan 1 00:00:00 1970 From: Peter Ujfalusi Subject: Re: linux-next: workqueues tree build failure Date: Thu, 26 Nov 2009 15:23:40 +0200 Message-ID: <200911261523.40952.peter.ujfalusi@nokia.com> References: <20091126190050.3f9d7fef.sfr@canb.auug.org.au> <200911261344.32051.peter.ujfalusi@nokia.com> <1259239796.3062.20.camel@palomino.walls.org> Mime-Version: 1.0 Content-Type: Text/Plain; charset=iso-8859-1 Content-Transfer-Encoding: QUOTED-PRINTABLE Return-path: Received: from smtp.nokia.com ([192.100.105.134]:19662 "EHLO mgw-mx09.nokia.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753463AbZKZNZB convert rfc822-to-8bit (ORCPT ); Thu, 26 Nov 2009 08:25:01 -0500 In-Reply-To: <1259239796.3062.20.camel@palomino.walls.org> Sender: linux-next-owner@vger.kernel.org List-ID: To: ext Andy Walls Cc: ext Mark Brown , ext Tejun Heo , Stephen Rothwell , "linux-next@vger.kernel.org" , "linux-kernel@vger.kernel.org" , Takashi Iwai On Thursday 26 November 2009 14:49:56 ext Andy Walls wrote: >=20 > Peter, >=20 > I would suspect using a single-threaded workqueue is better than a > wake_up() of another thread. IIRC, after queuing work, the workqueue= 's > single thread may run almost immediately on the same processor. With > waking up sleeping threads, I've run into scheduler delays around 10 = ms > on a dual core x86_64 desktop system. Hello Andy, I have sent the patch which changes from rt to singlethread, I hope it = fixes the=20 breakage in linux-next. In short testing, when there is virtually no load on the system I can n= ot see=20 any difference, which might change later. But I'll keep in mind that probably I'm not going to better off with th= e waking=20 up of a sleeping thread. >=20 > Regards, > Andy Thanks, P=E9ter