Return-path: Received: from tex.lwn.net ([70.33.254.29]:40849 "EHLO vena.lwn.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752101Ab3EEUiG (ORCPT ); Sun, 5 May 2013 16:38:06 -0400 Date: Sun, 5 May 2013 14:38:03 -0600 From: Jake Edge To: linux-wireless@vger.kernel.org Cc: Johannes Berg , lkml Subject: Bisected 3.9 regression for iwl4965 connection problem to 1672c0e3 Message-ID: <20130505143803.7e46e4c6@chukar.edge2.net> (sfid-20130505_223826_138840_12AE724B) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Sender: linux-wireless-owner@vger.kernel.org List-ID: After building 3.9 for my HP/Compaq 2510p laptop, the wireless (iwl4965) would seemingly no longer connect. Actually, it seems that it *will* eventually, where that is somewhere between 4 and 21+ retries (from KDE network manager). Going back to 3.8 (and earlier) and the problem goes away. (All on Fedora 18, fwiw, but I don't think that's significant). I bisected the problem to: commit 1672c0e31917f49d31d30d79067103432bc20cc7 Author: Johannes Berg Date: Tue Jan 29 15:02:27 2013 +0100 mac80211: start auth/assoc timeout on frame status but I can't (easily) revert that in 3.9 (maybe I can in -rc1 or something? haven't tried that) What more information is needed from me? I may still mess around with trying to revert that patch just to nail it down for sure, but two separate bisection exercises ended up at the same place. There are a few traps for the unwary if anyone else ends up bisecting in through here -- the first bisection (3.8 - 3.9-rc1) ran into an IPv6 build error along the way, which I "fixed" by configuring it out -- also, the amount of retries goes from *many* (15-21+) to relatively few (4) somewhere in there, which confused me and led to the second bisection (3.8-3.9) ... jake -- Jake Edge - LWN - jake@lwn.net - http://lwn.net