Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754149AbYJCRX1 (ORCPT ); Fri, 3 Oct 2008 13:23:27 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1753190AbYJCRXR (ORCPT ); Fri, 3 Oct 2008 13:23:17 -0400 Received: from victor.provo.novell.com ([137.65.250.26]:50944 "EHLO victor.provo.novell.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752332AbYJCRXQ (ORCPT ); Fri, 3 Oct 2008 13:23:16 -0400 Message-ID: <48E65606.6040409@novell.com> Date: Fri, 03 Oct 2008 13:27:34 -0400 From: Gregory Haskins User-Agent: Thunderbird 2.0.0.17 (X11/20080922) MIME-Version: 1.0 To: Gilles Carry , Gregory Haskins , linux-rt-users@vger.kernel.org, linux-kernel@vger.kernel.org, rostedt@goodmis.org, dvhltc@us.ibm.com, dino@in.ibm.com Subject: Re: [PATCH 2/2] RT: remove "paranoid" limit in push_rt_task References: <20081003123745.17387.61782.stgit@dev.haskins.net> <20081003124305.17387.90233.stgit@dev.haskins.net> <48E62253.1090000@bull.net> <20081003154501.GB5032@linux.vnet.ibm.com> In-Reply-To: <20081003154501.GB5032@linux.vnet.ibm.com> X-Enigmail-Version: 0.95.7 OpenPGP: id=D8195319 Content-Type: multipart/signed; micalg=pgp-sha1; protocol="application/pgp-signature"; boundary="------------enig2E2421D04FB8E75D32C66AE1" Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1878 Lines: 59 This is an OpenPGP/MIME signed message (RFC 2440 and 3156) --------------enig2E2421D04FB8E75D32C66AE1 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable Chirag Jog wrote: > * Gilles Carry [2008-10-03 15:46:59]: > > =20 >> Sorry Greg, >> >> Neither PPC64 nor Intel64 make it with this patch. >> At boot time, it stops at the BUG_ON you added: >> 0xc00000000004eca4 is in push_rt_task (kernel/sched_rt.c:1102) >> =20 > I am also confirming this issue gilles reported. > > Although, i have a question: > When we enable group scheduling (CONFIG_RT_GROUP_SCHED), > all the problems disappear. > After skimming through the rt scheduler code, I don't feel group=20 > scheduling alters the behavior of push/pull strategies in any way. > > So I am wonder whether enabling group scheduling > actually solves the problem or just makes it tough=20 > to recreate? > =20 Hi Chirag, The issue that Gilles pointed me at is a race condition. I do not suspect GROUP_SCHED itself has anything to do with the problem other than it changes the timing. HTH! -Greg --------------enig2E2421D04FB8E75D32C66AE1 Content-Type: application/pgp-signature; name="signature.asc" Content-Description: OpenPGP digital signature Content-Disposition: attachment; filename="signature.asc" -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkjmVgYACgkQlOSOBdgZUxmCHwCeON4kWz3qBkueaSNOsVLdNDQY xVIAn1+tRmWW1t2s9gD3ceWrIiBC5h0a =ywZW -----END PGP SIGNATURE----- --------------enig2E2421D04FB8E75D32C66AE1-- -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/