From mboxrd@z Thu Jan 1 00:00:00 1970 From: "H. Peter Anvin" Subject: Re: [PATCH 21/21] posix_types: Remove fd_set macros Date: Wed, 08 Feb 2012 13:30:55 -0800 Message-ID: <4F32E98F.2090909@zytor.com> References: <4F32A98F.7030400@zytor.com> <1328677745-20121-22-git-send-email-hpa@zytor.com> <1328677745-20121-1-git-send-email-hpa@zytor.com> <8328.1328703626@redhat.com> <16325.1328736262@redhat.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <16325.1328736262@redhat.com> Sender: linux-kernel-owner@vger.kernel.org To: David Howells Cc: linux-arch@vger.kernel.org, linux-kernel@vger.kernel.org, Linus Torvalds , Arnd Bergmann List-Id: linux-arch.vger.kernel.org On 02/08/2012 01:24 PM, David Howells wrote: > H. Peter Anvin wrote: > >> Anyway, select() paraphernalia goes in (as opposed to >> ), poll() paraphernalia goes in, and the linux/* >> files mimics that. > > That doesn't seem to be so - at least not for my copy of glibc: > > /usr/include/sys/select.h:/* fd_set for select and pselect. */ > /usr/include/sys/select.h: } fd_set; > /usr/include/sys/select.h:#define FD_ISSET(fd, fdsetp) __FD_ISSET (fd, fdsetp) > > Besides, this is inside the kernel. The I/O polling architecture definitions > (which includes select) ought to be in one place there, not spread out all > over the place. > I think is more recent, but anyway, look at the existing and you see why it's there. It's fine to say it shouldn't be, but then ALL the definitions should move (and that should be a separate change from this patchset)... but do note that some of those macro definitions (which I did NOT change) are exported to user space so make those changes with care. -hpa From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from terminus.zytor.com ([198.137.202.10]:49893 "EHLO mail.zytor.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1756342Ab2BHVbG (ORCPT ); Wed, 8 Feb 2012 16:31:06 -0500 Message-ID: <4F32E98F.2090909@zytor.com> Date: Wed, 08 Feb 2012 13:30:55 -0800 From: "H. Peter Anvin" MIME-Version: 1.0 Subject: Re: [PATCH 21/21] posix_types: Remove fd_set macros References: <4F32A98F.7030400@zytor.com> <1328677745-20121-22-git-send-email-hpa@zytor.com> <1328677745-20121-1-git-send-email-hpa@zytor.com> <8328.1328703626@redhat.com> <16325.1328736262@redhat.com> In-Reply-To: <16325.1328736262@redhat.com> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Sender: linux-arch-owner@vger.kernel.org List-ID: To: David Howells Cc: linux-arch@vger.kernel.org, linux-kernel@vger.kernel.org, Linus Torvalds , Arnd Bergmann Message-ID: <20120208213055.VFBsh4uFzi4GwBEbILYKl3qQJC12efsRCvvemRvGxvk@z> On 02/08/2012 01:24 PM, David Howells wrote: > H. Peter Anvin wrote: > >> Anyway, select() paraphernalia goes in (as opposed to >> ), poll() paraphernalia goes in, and the linux/* >> files mimics that. > > That doesn't seem to be so - at least not for my copy of glibc: > > /usr/include/sys/select.h:/* fd_set for select and pselect. */ > /usr/include/sys/select.h: } fd_set; > /usr/include/sys/select.h:#define FD_ISSET(fd, fdsetp) __FD_ISSET (fd, fdsetp) > > Besides, this is inside the kernel. The I/O polling architecture definitions > (which includes select) ought to be in one place there, not spread out all > over the place. > I think is more recent, but anyway, look at the existing and you see why it's there. It's fine to say it shouldn't be, but then ALL the definitions should move (and that should be a separate change from this patchset)... but do note that some of those macro definitions (which I did NOT change) are exported to user space so make those changes with care. -hpa