Return-path: Received: from mail.klingt.org ([86.59.21.178]:34247 "EHLO klingt.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S934873AbZLGJc2 (ORCPT ); Mon, 7 Dec 2009 04:32:28 -0500 Message-ID: <4B1CCB81.5060104@klingt.org> Date: Mon, 07 Dec 2009 10:31:45 +0100 From: Tim Blechmann MIME-Version: 1.0 To: Gertjan van Wingerde CC: Andi Kleen , rjw@sisk.pl, IvDoorn@gmail.com, linville@tuxdriver.com, linux-wireless@vger.kernel.org Subject: Re: regression: rt2561 frequent "Arrived at non-free entry" errors in 2.6.32 References: <20091204233202.GA13658@basil.fritz.box> <4B1BB44E.90706@gmail.com> In-Reply-To: <4B1BB44E.90706@gmail.com> Content-Type: multipart/signed; micalg=pgp-sha1; protocol="application/pgp-signature"; boundary="------------enigE4600044CA413AFFC818098B" Sender: linux-wireless-owner@vger.kernel.org List-ID: This is an OpenPGP/MIME signed message (RFC 2440 and 3156) --------------enigE4600044CA413AFFC818098B Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: quoted-printable On 12/06/2009 02:40 PM, Gertjan van Wingerde wrote: > On 12/05/09 00:32, Andi Kleen wrote: >> [Rafael, something for your regression list] >> >> I upgraded a system which was running 2.6.30 to 2.6.32. >> It has a=20 >> >> 06:02.0 Network controller: RaLink RT2561/RT61 802.11g PCI >> >> wireless PCI card. Now regularly even under moderate traffic >> I see messages like >> >> phy0 -> rt2x00queue_write_tx_frame: Error - Arrived at non-free entry = in the non-full queue 2. >> Please file bug report to http://rt2x00.serialmonkey.com. >> >> and loss of connectivity, often until the wireless connection >> is restarted. This wasn't the case in 2.6.30, there the driver >> ran stable without any problems. The problem currently >> happens every few minutes. >> >=20 > Andi, Tim, >=20 > Both of you have reported this problem on 2.6.32. >=20 > In the past this always has occurred due to queue locking problems. Thi= s led me > to audit the queue locking code, and that certainly looked suspicious t= o me. >=20 > Would you be able to test whether the attached test patch fixes the pro= blem for > you. The patch basically applies proper queue locking to the code, alth= ough in=20 > a very course manner. The patch is relative to 2.6.32. >=20 > Note: I am not 100% sure that this is where the problem is, but at leas= t the test > patch should confirm whether I am searching in the right direction. the patch didn't cleanly apply to v2.6.32 and the patched kernel didn't boot ... which tree should it be applied on? tim --=20 tim@klingt.org http://tim.klingt.org Im =C3=BCbrigen ist es gescheiter, sich warm zuzudecken als sich zu betrinken. Werner Schwab --------------enigE4600044CA413AFFC818098B Content-Type: application/pgp-signature; name="signature.asc" Content-Description: OpenPGP digital signature Content-Disposition: attachment; filename="signature.asc" -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.9 (GNU/Linux) iEYEARECAAYFAkscy4UACgkQdL+4qsZfVsta/QCeOOTc/lP7Rz0AIywd0BfQLoey fsAAnAvv1KMkd05OnVbQSKJ1OQY4HToN =TtYp -----END PGP SIGNATURE----- --------------enigE4600044CA413AFFC818098B--