From mboxrd@z Thu Jan 1 00:00:00 1970 From: Michael Riesch Subject: Re: [PATCH] usbnet: fix oops in usbnet_start_xmit Date: Mon, 07 Nov 2011 14:29:13 +0100 Message-ID: <1320672553.2909.41.camel@malcolm> References: <20111106183337.5379.4356.stgit@zurg> Reply-To: michael@riesch.at Mime-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: 7bit Cc: Oliver Neukum , netdev@vger.kernel.org, "David S. Miller" , devel@openvz.org To: Konstantin Khlebnikov Return-path: Received: from smtprelay02.ispgateway.de ([80.67.31.29]:44726 "EHLO smtprelay02.ispgateway.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754691Ab1KGNoJ (ORCPT ); Mon, 7 Nov 2011 08:44:09 -0500 In-Reply-To: <20111106183337.5379.4356.stgit@zurg> Sender: netdev-owner@vger.kernel.org List-ID: On Sun, 2011-11-06 at 22:33 +0300, Konstantin Khlebnikov wrote: > This patch fixes the bug added in commit v3.1-rc7-1055-gf9b491e > SKB can be NULL at this point, at least for cdc-ncm. OK, I didn't think of that, but... > Let's call skb_tx_timestamp() after driver specific tx-fixup hacks. ... the reason I put the skb_tx_timestamp() call before the tx_fixup is that these hacks often perform skb_push/skb_pull or any other kind of framing. This may result (at least in the case of the asix drivers) in perfectly correct PTP packets being not recognized as such by the packet filter. Can we do a check like this: if(skb) skb_tx_timestamp() tx_fixup() ? Regards, Michael