Return-path: Received: from mail-lb0-f170.google.com ([209.85.217.170]:56562 "EHLO mail-lb0-f170.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1758384AbbA2Jq6 (ORCPT ); Thu, 29 Jan 2015 04:46:58 -0500 MIME-Version: 1.0 Date: Thu, 29 Jan 2015 11:46:56 +0200 Message-ID: (sfid-20150129_104712_811763_14664A11) Subject: random failures with Intel devices starting from kernel 3.18 From: Emmanuel Grumbach To: linux-wireless , "linux-pci@vger.kernel.org" Cc: sait.a.umar@gmail.com, gredner@gmail.com, kichawa23@interia.pl, Emmanuel Grumbach Content-Type: text/plain; charset=UTF-8 Sender: linux-wireless-owner@vger.kernel.org List-ID: Hi, I recently got quite a few reports about failures of the Intel wireless devices. I looked at the logs and it turns out that we can't access our device anymore. We had bad issues around this and I fixed them in the iwlwifi driver in the end. What makes me feels that I need to add PCI folks is that I can feel that we have a regression in 3.18 (based on the number of reports) and we didn't change much there that could cause such a bug. Also, very surprisingly, I asked someone to disable ASPM as a boot parameter and that didn't actually disable it. You can find more details here [1]. What I see in the logs is that iwlwifi is trying to access the mapped memory of the device (I may have a bad terminology, but it is BAR + 0x24 - let me know if the term is not the right one), and we get 0xffffffff... Any thoughts? Thanks! [1] https://bugzilla.kernel.org/show_bug.cgi?id=91171 Emmanuel Grumbach egrumbach@gmail.com