From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Kok, Auke" Subject: Re: [patch] e1000=y && e1000e=m regression fix Date: Wed, 09 Apr 2008 13:12:21 -0700 Message-ID: <47FD2325.2030705@intel.com> References: <20080408083606.GA20863@elte.hu> <47FB9ABB.9080403@intel.com> <20080408183921.GA20803@elte.hu> <20080408193245.GG11962@parisc-linux.org> <20080408195123.GA28148@elte.hu> <47FBCE00.2020309@garzik.org> <20080408200652.GC28148@elte.hu> <47FBD34A.6080508@garzik.org> <20080408203314.GA28952@elte.hu> <47FBDBE9.9040700@garzik.org> <20080409193850.GA11763@elte.hu> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Cc: Jeff Garzik , Matthew Wilcox , Linux Kernel Mailing List , NetDev , e1000-list , linux-pci maillist , Andrew Morton , "David S. Miller" , Linus Torvalds , Jesse Brandeburg , "Ronciak, John" , "Allan, Bruce W" , Greg KH , Arjan van de Ven , "Rafael J. Wysocki" To: Ingo Molnar Return-path: Received: from mga02.intel.com ([134.134.136.20]:18373 "EHLO mga02.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752216AbYDIUND (ORCPT ); Wed, 9 Apr 2008 16:13:03 -0400 In-Reply-To: <20080409193850.GA11763@elte.hu> Sender: netdev-owner@vger.kernel.org List-ID: Ingo Molnar wrote: > * Jeff Garzik wrote: > >>> We've got three thousand Kconfig options - it is clearly not >>> realistic for users to keep such details in mind to avoid pitfalls. >> Agreed -- hence the multiple announcements, including in this thread, >> to put said details into mind. > > which part of "it took a kernel developer more than an hour to figure > out why his laptop had a dead network interface" did you not understand? > Whatever you did, it was not apparent to me. I dont follow every tiny > detail of the e1000 driver family, nor do 99%+ [*] of our users. > > find the fix below, against current -git. > > the current upstream behavior is the worst possible one and is just a > plain bug, and the solution is dead-simple. If this makes people happy then I am happy to ack this. Acked-by: Auke Kok > > Ingo > > [*] guesstimate > > ---------------> > Subject: e1000=y && e1000e=m regression fix > From: Ingo Molnar > Date: Wed Apr 09 21:09:35 CEST 2008 > > fix a regression from v2.6.24: do not transfer the e1000e PCI IDs from > e1000 to e1000e if e1000 is built-in and e1000e is a module. > > Built-in drivers take precedence over modules in many ways - and in this > case it's clear that the user intended the e1000 driver to be the > primary one. "Silently change behavior and break existing configs" is > never a good migration strategy. Most users will use distro kernels that > are not affected by this problem at all - nor are they affected by this > patch - but this problem can hit users and developers who build their > kernels themselves and migrate from v2.6.24 to v2.6.25. > > this fixes: http://bugzilla.kernel.org/show_bug.cgi?id=10427 > > Signed-off-by: Ingo Molnar > --- > drivers/net/Kconfig | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > Index: linux-x86.q/drivers/net/Kconfig > =================================================================== > --- linux-x86.q.orig/drivers/net/Kconfig > +++ linux-x86.q/drivers/net/Kconfig > @@ -2022,7 +2022,7 @@ config E1000E > will be called e1000e. > > config E1000E_ENABLED > - def_bool E1000E != n > + def_bool E1000E = y || ((E1000E != n) && (E1000 = E1000E)) > > config IP1000 > tristate "IP1000 Gigabit Ethernet support"