From mboxrd@z Thu Jan 1 00:00:00 1970 From: Johan Andersson Subject: Re: e1000e: Hardware CRC stripping doesn't work with 82566DM-2 Date: Thu, 07 Feb 2008 07:19:16 +0100 Message-ID: <47AAA2E4.5010007@transmode.se> References: <47AA1BB4.8010802@transmode.se> <47AA2A03.2000109@intel.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Cc: netdev@vger.kernel.org To: "Kok, Auke" Return-path: Received: from mail.transmode.se ([83.241.175.147]:63895 "EHLO tmnt04.transmode.se" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1759421AbYBGGT1 (ORCPT ); Thu, 7 Feb 2008 01:19:27 -0500 In-Reply-To: <47AA2A03.2000109@intel.com> Sender: netdev-owner@vger.kernel.org List-ID: Kok, Auke wrote: > Johan Andersson wrote: > >> Hi, >> >> In commit 140a74802894e9db57e5cd77ccff77e590ece5f3 a workaround for crc >> stripping was removed. >> However, the hardware supported crc stripping now in use doesn't seem to >> work with the following chip: >> Intel Corporation 82566DM-2 Gigabit Network Connection (rev 02) (8086:10bd) >> >> I discovered this while trying to make this nic member of a bridge. >> Since some frames get too large for the bridge they are dropped, making >> it impossible to use this nic in a bridge. >> >> Reverting this commit makes the nic and brigde work again. >> >> Are there any more chips in this family suffering from this? >> > > I have only heard from this issue for 82566dm-2 chipsets, but I'm looking into it > atm (time permitting). It has been reported by Herbert Xu to cause issues in the > same matter last week. > > I'm not sure what the real issue is right now, I hope to get an answer on this later. > > Auke > > Ok, thanks. Tell me if I can help out in any way. /Johan