Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S932356AbWALSSt (ORCPT ); Thu, 12 Jan 2006 13:18:49 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S932620AbWALSSt (ORCPT ); Thu, 12 Jan 2006 13:18:49 -0500 Received: from hera.kernel.org ([140.211.167.34]:28560 "EHLO hera.kernel.org") by vger.kernel.org with ESMTP id S932356AbWALSSs (ORCPT ); Thu, 12 Jan 2006 13:18:48 -0500 To: linux-kernel@vger.kernel.org From: Stephen Hemminger Subject: Re: sk98lin Date: Thu, 12 Jan 2006 10:18:43 -0800 Organization: OSDL Message-ID: <20060112101843.0b0e159f@dxpl.pdx.osdl.net> References: <20060112180048.8A18EBC32@mx.dtiltas.lt> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit X-Trace: build.pdx.osdl.net 1137089925 2217 10.8.0.74 (12 Jan 2006 18:18:45 GMT) X-Complaints-To: abuse@osdl.org NNTP-Posting-Date: Thu, 12 Jan 2006 18:18:45 +0000 (UTC) X-Newsreader: Sylpheed-Claws 1.9.100 (GTK+ 2.6.10; x86_64-redhat-linux-gnu) Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1897 Lines: 50 On Thu, 12 Jan 2006 20:01:43 +0200 Nerijus Baliunas wrote: > Hello, > > Why a newer driver (http://www.marvell.com/drivers/upload/install-8_28.tar.bz2, > it's v8.28.1.3, while kernel has v6.23) is not integrated into kernel? > Because of no one submitted it or does it have some problems? It was submitted, but has several problems: * wasn't done as small pieces; too much whole sale replacement * ignored all the bugfixes and work that went into the mainline kernel * merges support for two kinds of hardware in one driver Also, it increases the amount of vendor ugly code; the sk98lin driver would probably not be accepted today. While developing the skge and sky2 driver I discovered more problems and those got fixed in the mainline sk98lin driver. The vendor version has issues like: * does NAPI but has interrupts disabled * has a watchdog routine to mask off all the bugs they never managed to fix. * sets PCI-express parameters to benchmark values that cause random hangs and data corruption That is why I wouldn't recommend the vendor version for any production systems. > Because w/o it some newer cards are not recognized > (Marvell Technology Group Ltd. 88E8050 Gigabit Ethernet Controller (rev 17) > for example). > > Regards, > Nerijus > - > To unsubscribe from this list: send the line "unsubscribe linux-kernel" in > the body of a message to majordomo@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html > Please read the FAQ at http://www.tux.org/lkml/ -- Stephen Hemminger OSDL http://developer.osdl.org/~shemminger - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/