From mboxrd@z Thu Jan 1 00:00:00 1970 From: James Bottomley Subject: Re: [PATCH 08/24] parisc/uapi: Use Kbuild logic to provide Date: Mon, 25 Nov 2013 14:22:35 +0400 Message-ID: <1385374955.2354.24.camel@dabdike> References: <1385369734-24893-1-git-send-email-geert@linux-m68k.org> <1385369734-24893-9-git-send-email-geert@linux-m68k.org> <1385373313.2354.16.camel@dabdike> Mime-Version: 1.0 Content-Type: text/plain; charset="ISO-8859-15" Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: Sender: linux-parisc-owner@vger.kernel.org To: Geert Uytterhoeven Cc: Arnd Bergmann , Linux-Arch , "linux-kernel@vger.kernel.org" , "James E.J. Bottomley" , Helge Deller , Parisc List List-Id: linux-arch.vger.kernel.org On Mon, 2013-11-25 at 11:12 +0100, Geert Uytterhoeven wrote: > On Mon, Nov 25, 2013 at 10:55 AM, James Bottomley > wrote: > > On Mon, 2013-11-25 at 09:55 +0100, Geert Uytterhoeven wrote: > >> Uapi just includes > > > > This really doesn't look like a good idea: Firstly the default of > > asm-generic/types.h should *not* be int-ll64, because that's for 32 bit > > only and most processors are moving towards 64 bit. Secondly parisc > > In kernelspace, both 32 and 64 bit are ll64. Not necessarily; the files exist because of variations in gcc. Some versions of gcc have long long being 128 bits and for those you have to use int-l64.h > > currently only has a 32 bit userspace, which is why the file is a simple > > include of int-ll64.h; however, people are working on 64 bit userspace > > Sorry, I didn't know you support 32 bit userspace only. > > > (albeit very slowly) and that file would then have to change. If we > > take the file away someone will forget to add the changed file back. > > However, is this a "new" 64 bit userspace ABI? If yes, you may want to > consider using ll64 in userspace on 64 bit too, like all "new" 64 bit ports > do (arm64, s390, sparc, tile, x86). Only legacy alpha, ia64, mips and > old ppc use l64. > > Or follow the ppc64 road: > > /* > * This is here because we used to use l64 for 64bit powerpc > * and we don't want to impact user mode with our change to ll64 > * in the kernel. > * > * However, some user programs are fine with this. They can > * flag __SANE_USERSPACE_TYPES__ to get int-ll64.h here. > */ > #ifndef __KERNEL__ > #if !defined(__SANE_USERSPACE_TYPES__) && defined(__powerpc64__) > # include > #else > # include > #endif > #endif That's the one I was thinking we might need. James From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from bedivere.hansenpartnership.com ([66.63.167.143]:53802 "EHLO bedivere.hansenpartnership.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750766Ab3KYKWk (ORCPT ); Mon, 25 Nov 2013 05:22:40 -0500 Message-ID: <1385374955.2354.24.camel@dabdike> Subject: Re: [PATCH 08/24] parisc/uapi: Use Kbuild logic to provide From: James Bottomley Date: Mon, 25 Nov 2013 14:22:35 +0400 In-Reply-To: References: <1385369734-24893-1-git-send-email-geert@linux-m68k.org> <1385369734-24893-9-git-send-email-geert@linux-m68k.org> <1385373313.2354.16.camel@dabdike> Content-Type: text/plain; charset="ISO-8859-15" Mime-Version: 1.0 Content-Transfer-Encoding: 7bit Sender: linux-arch-owner@vger.kernel.org List-ID: To: Geert Uytterhoeven Cc: Arnd Bergmann , Linux-Arch , "linux-kernel@vger.kernel.org" , "James E.J. Bottomley" , Helge Deller , Parisc List Message-ID: <20131125102235.stX-eKHzuo7rWb3FSUF0R78I9QHWClAUw-C5Set72HI@z> On Mon, 2013-11-25 at 11:12 +0100, Geert Uytterhoeven wrote: > On Mon, Nov 25, 2013 at 10:55 AM, James Bottomley > wrote: > > On Mon, 2013-11-25 at 09:55 +0100, Geert Uytterhoeven wrote: > >> Uapi just includes > > > > This really doesn't look like a good idea: Firstly the default of > > asm-generic/types.h should *not* be int-ll64, because that's for 32 bit > > only and most processors are moving towards 64 bit. Secondly parisc > > In kernelspace, both 32 and 64 bit are ll64. Not necessarily; the files exist because of variations in gcc. Some versions of gcc have long long being 128 bits and for those you have to use int-l64.h > > currently only has a 32 bit userspace, which is why the file is a simple > > include of int-ll64.h; however, people are working on 64 bit userspace > > Sorry, I didn't know you support 32 bit userspace only. > > > (albeit very slowly) and that file would then have to change. If we > > take the file away someone will forget to add the changed file back. > > However, is this a "new" 64 bit userspace ABI? If yes, you may want to > consider using ll64 in userspace on 64 bit too, like all "new" 64 bit ports > do (arm64, s390, sparc, tile, x86). Only legacy alpha, ia64, mips and > old ppc use l64. > > Or follow the ppc64 road: > > /* > * This is here because we used to use l64 for 64bit powerpc > * and we don't want to impact user mode with our change to ll64 > * in the kernel. > * > * However, some user programs are fine with this. They can > * flag __SANE_USERSPACE_TYPES__ to get int-ll64.h here. > */ > #ifndef __KERNEL__ > #if !defined(__SANE_USERSPACE_TYPES__) && defined(__powerpc64__) > # include > #else > # include > #endif > #endif That's the one I was thinking we might need. James