From mboxrd@z Thu Jan 1 00:00:00 1970 From: Arnd Bergmann Subject: Re: [RFC 26/72] arm: Move the ARM/ACORN drivers Date: Sat, 25 Jun 2011 18:59:38 +0200 Message-ID: <201106251859.38963.arnd@arndb.de> References: <1309010363-22750-1-git-send-email-jeffrey.t.kirsher@intel.com> <201106251803.53416.arnd@arndb.de> <20110625161227.GJ23234@n2100.arm.linux.org.uk> Mime-Version: 1.0 Content-Type: Text/Plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Cc: Jeff Kirsher , davem@davemloft.net, netdev@vger.kernel.org, Sriram , Vinay Hegde , Cyril Chemparathy , Wan ZongShun , Lennert Buytenhek , Krzysztof Halasa , Anant Gole , Chaithrika U S To: "Russell King - ARM Linux" Return-path: Received: from moutng.kundenserver.de ([212.227.126.187]:58923 "EHLO moutng.kundenserver.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751778Ab1FYQ7o (ORCPT ); Sat, 25 Jun 2011 12:59:44 -0400 In-Reply-To: <20110625161227.GJ23234@n2100.arm.linux.org.uk> Sender: netdev-owner@vger.kernel.org List-ID: On Saturday 25 June 2011, Russell King - ARM Linux wrote: > > How do you deal with NE2K or 8390 stuff, where you have a design which > has been cloned by many different manufacturers? Who's do you decide > to put it in? > > More interestingly, if you have one of the 8390 clones, which directory > do you look in? And shouldn't etherh.c live along side 8390's files > as etherh is 8390-derived? Jeff created a drivers/net/ethernet/8390/ directory for all of those, which sounds sensible to me, but apparently he missed some of the nonobvious ones. > ether1 is 82586, so presumably that should be Intel subdirectory. Yes, according to the new layout, it should at least live together with the other 82586-based drivers (eexpress.c, ni52.c, sun3_82586.c), which could be either a directory for all of these, or the generic intel directory. > ether3 is SEEQ8005, so that should be in the SEEQ subdirectory. Same here, it can be grouped together with sqiseeq.c (SEEQ8003) and seeq8005.c. I don't really care about these old ones too much though. Nobody will spend the effort to clean up and merge these ancient drivers any more, and it's also very unlikely that we want to add support for new boards with these. For the soc vendor specific ones, I think we should decide on the new location based on how easy it is to spot duplicates, e.g. to avoid adding an AVR32 copy of the at91 ethernet driver. For reference (since you probably didn't see the full 72 patch series on netdev), some of the related driver moves suggested by Jeff are: drivers/net/{ => ethernet/8390}/*8390*.c | 0 drivers/net/{ => ethernet/amd}/*lance.c | 0 drivers/net/{ => ethernet/atmel}/macb.c | 0 drivers/net/{ => ethernet/micrel}/ks8842.c | 0 drivers/net/{ => ethernet/micrel}/ks8851.c | 0 drivers/net/{ => ethernet/ti}/cpmac.c | 0 drivers/net/{ => ethernet/ti}/tlan.c | 0 drivers/net/{ => ethernet/isa}/eexpress.c | 0 drivers/net/{ => ethernet/sun}/sun3_82586.c | 0 drivers/net/{ => ethernet/isa}/seeq8005.c | 0 drivers/net/{ => ethernet/sgi}/sgiseeq.c | 0 Arnd