From mboxrd@z Thu Jan 1 00:00:00 1970 From: Sarah Sharp Subject: Re: Help testing for USB ethernet/xHCI regression Date: Wed, 29 Jan 2014 13:54:08 -0800 Message-ID: <20140129215408.GC5991@xanatos> References: <22081260.RNay0J72dY@athas> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: linux-usb@vger.kernel.org, Mark Lord , Greg Kroah-Hartman , netdev@vger.kernel.org, David Laight To: renevant@internode.on.net Return-path: Received: from mga02.intel.com ([134.134.136.20]:11098 "EHLO mga02.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750942AbaA2VyK (ORCPT ); Wed, 29 Jan 2014 16:54:10 -0500 Content-Disposition: inline In-Reply-To: <22081260.RNay0J72dY@athas> Sender: netdev-owner@vger.kernel.org List-ID: On Wed, Jan 29, 2014 at 04:21:00PM +1100, renevant@internode.on.net wrote: > Hello, > > I am someone who has been struggling to get an ax88179 net adapter working > reliably. > > I have an integrated Asmedia 1042 xhci controller that is reportedly version > 0.96 on an ASUS M5A99FX PRO R2.0, BIOS 2201 motherboard based on the AMD 990FX > chipset. > > The big issue I am currently facing is that I cannot get the device to work at > all with 3.13 and current 3.14 mainline. This does not occur with 3.12. With the working kernel, were you using vanilla 3.12, or a later 3.12 stable release, like 3.12.5? > Current issue is when plugging in the ax88179 there is lag when bringing the > interface up and a bunch of kernel messages: With which kernel? > [ 63.389440] ax88179_178a 2-1:1.0 eth0: register 'ax88179_178a' at > usb-0000:07:00.0-1, ASIX AX88179 USB 3.0 Gigabit Ethernet, 80:3f:5d:08:0c:65 > [ 63.389500] usbcore: registered new interface driver ax88179_178a > [ 63.423942] systemd-udevd[560]: renamed network interface eth0 to enp7s0u1 > [ 79.481028] IPv6: ADDRCONF(NETDEV_UP): enp7s0u1: link is not ready > [ 82.210721] ax88179_178a 2-1:1.0 enp7s0u1: ax88179 - Link status is: 1 > [ 82.338947] ax88179_178a 2-1:1.0 enp7s0u1: ax88179 - Link status is: 1 > [ 82.467148] ax88179_178a 2-1:1.0 enp7s0u1: kevent 4 may have been dropped > [ 82.470028] ax88179_178a 2-1:1.0 enp7s0u1: ax88179 - Link status is: 1 > [ 82.595364] ax88179_178a 2-1:1.0 enp7s0u1: kevent 4 may have been dropped > [ 82.598312] ax88179_178a 2-1:1.0 enp7s0u1: ax88179 - Link status is: 1 > [ 82.723580] ax88179_178a 2-1:1.0 enp7s0u1: kevent 4 may have been dropped > [ 82.726487] ax88179_178a 2-1:1.0 enp7s0u1: ax88179 - Link status is: 1 > [ 82.851796] ax88179_178a 2-1:1.0 enp7s0u1: kevent 4 may have been dropped > [ 82.854642] ax88179_178a 2-1:1.0 enp7s0u1: ax88179 - Link status is: 1 > [snip] > [ 87.218379] ax88179_178a 2-1:1.0 enp7s0u1: Failed to write reg index > 0x0002: -110 Can you enable xHCI debugging as well, and send dmesg? You'll need to have CONFIG_USB_DEBUG turned on, and also run this command as root: echo -n 'module xhci_hcd =p' > /sys/kernel/debug/dynamic_debug/control > Any traffic sent to the nic isn't seen by Linux. > > > I can confirm the following problems for me with 3.14: > > reverting the writeq patch made USB 3 work for me again. > > http://marc.info/?t=139093294600002&r=1&w=2 So what was the issue that caused you to revert that patch? The traffic not going to the device, or the device not being enumerated at all? > and > > http://marc.info/?l=linux-usb&m=139092084714232&w=2 > > got rid of transiever errors > > > > > I cannot really test stability for the ax88179 until I can get it to work when > plugging it in as above. I've tried the frag reversion and it's made no > difference to this issue. What about also reverting the writeq patch, on top of the other patches I asked Mark to try? Sarah Sharp