Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id ; Fri, 26 Jul 2002 05:08:54 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id ; Fri, 26 Jul 2002 05:08:54 -0400 Received: from [195.63.194.11] ([195.63.194.11]:52999 "EHLO mail.stock-world.de") by vger.kernel.org with ESMTP id ; Fri, 26 Jul 2002 05:08:53 -0400 Message-ID: <3D411134.60904@evision.ag> Date: Fri, 26 Jul 2002 11:07:00 +0200 From: Marcin Dalecki Reply-To: martin@dalecki.de User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.1b) Gecko/20020722 X-Accept-Language: en-us, en, pl, ru MIME-Version: 1.0 To: Alan Cox CC: martin@dalecki.de, Linus Torvalds , Kernel Mailing List Subject: Re: [PATCH] IDE 104 References: <3D40F8F9.1050507@evision.ag> <1027678411.13428.3.camel@irongate.swansea.linux.org.uk> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1559 Lines: 38 Alan Cox wrote: > On Fri, 2002-07-26 at 08:23, Marcin Dalecki wrote: > >>- Make the bit-sliced data types in hdreg.h use the bit-slice data types >> instead of the generic ones. This makes clear that those are supposed >> to be register masks. > > > Because its an external API it needs to remain constant. Swapping ??? The only thing which is really 'external API' there is the declaration of the HDIO_XXX ioctl and among them in reality only HDIO_GETGEO is really used outside the scope of the dreddy hdparm application. And 99% of times its usage is bogous anyway. Or do you know any better examples I'm not aware of? The remainings will be moved away from there soon becouse it doesn't make any sense to include this at every single place out there where HDIO_GETGEO is the needed declaration. If some application needs ATA command macros - it should come with they own header declaring them instead of peeing in to the kernel becouse they don't depend on the kernel in question. They are cut in stone at www.t13.org ;-). > uchar/uint looks fine and the structure is off the drive anyway. However > they should be __u16 __u8 __u32 so that C libraries and apps can still > use the header (its __u16 because 'u16' cannot be exposed to user space > directly or via libc) I know - see above. - 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/