From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753243Ab1EEUhW (ORCPT ); Thu, 5 May 2011 16:37:22 -0400 Received: from moutng.kundenserver.de ([212.227.17.10]:59824 "EHLO moutng.kundenserver.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750925Ab1EEUhS (ORCPT ); Thu, 5 May 2011 16:37:18 -0400 From: Arnd Bergmann To: Geert Uytterhoeven Subject: Re: [PATCH] m68k: Merge mmu and non-mmu versions of sys_call_table Date: Thu, 5 May 2011 22:37:11 +0200 User-Agent: KMail/1.13.5 (Linux/2.6.39-rc4+; KDE/4.5.1; x86_64; ; ) Cc: "Linux/m68k" , Greg Ungerer , Linux Kernel Development , uClinux list References: <201105050824.34056.arnd@arndb.de> In-Reply-To: MIME-Version: 1.0 Content-Type: Text/Plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Message-Id: <201105052237.11656.arnd@arndb.de> X-Provags-ID: V02:K0:vKX68s592UJIZTtooTnsPt2jptIVobmUgqRnAZ2lQZI yKPlqw8tjaM1y0zdSkG2kp0VqbNg49loDbCjy7BddxoyGiVLLW I4Z2GD1MoDCpdzzWWhw4nENs/2hc4zdhGzH3YgCboYPBmC3EOO hNX6qXMOTYFsqGrXq/42HVhk9ytu6/dvrgyPUUfop/WPQXiN0I kekB/YB3pTAlfPNnzHo4g== Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Thursday 05 May 2011 20:44:20 Geert Uytterhoeven wrote: > On Thu, 5 May 2011, Arnd Bergmann wrote: > > Did you check the macros in unistd to see if they match the > > actual table? I guess it would be a good time to comment out > > the ones that are not implemented in either of the two ABIs. > > Like this? Looks good! > warning: #warning syscall pselect6 not implemented > warning: #warning syscall ppoll not implemented > warning: #warning syscall recvmmsg not implemented > > Do we need pselect6 and ppoll? I have vague memories not requiring it. I'm pretty sure you need them. Glibc provides wrappers for them when they are not there, but the wrappers not completely reliable. > recvmmsg is a false positive, as we set __ARCH_WANT_SYS_SOCKETCALL. This looks really strange. The commit that introduced recvmmsg (a2e27255) added it to both socketcall and as a separate syscall in a number of architectures, without a good reason for it. I guess it's too late to change that now, but we should at least fix the script so we don't report it missing when socketcall is set. However, if your unistd.h has defined __NR_recvmmsg before, you should probably add it to the syscall table, just in case that someone built a binary with that number. Arnd From mboxrd@z Thu Jan 1 00:00:00 1970 From: Arnd Bergmann Subject: Re: [PATCH] m68k: Merge mmu and non-mmu versions of sys_call_table Date: Thu, 5 May 2011 22:37:11 +0200 Message-ID: <201105052237.11656.arnd@arndb.de> References: <201105050824.34056.arnd@arndb.de> Mime-Version: 1.0 Content-Type: Text/Plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: Sender: linux-m68k-owner@vger.kernel.org List-Id: linux-m68k@vger.kernel.org To: Geert Uytterhoeven Cc: Linux/m68k , Greg Ungerer , Linux Kernel Development , uClinux list On Thursday 05 May 2011 20:44:20 Geert Uytterhoeven wrote: > On Thu, 5 May 2011, Arnd Bergmann wrote: > > Did you check the macros in unistd to see if they match the > > actual table? I guess it would be a good time to comment out > > the ones that are not implemented in either of the two ABIs. > > Like this? Looks good! > warning: #warning syscall pselect6 not implemented > warning: #warning syscall ppoll not implemented > warning: #warning syscall recvmmsg not implemented > > Do we need pselect6 and ppoll? I have vague memories not requiring it. I'm pretty sure you need them. Glibc provides wrappers for them when they are not there, but the wrappers not completely reliable. > recvmmsg is a false positive, as we set __ARCH_WANT_SYS_SOCKETCALL. This looks really strange. The commit that introduced recvmmsg (a2e27255) added it to both socketcall and as a separate syscall in a number of architectures, without a good reason for it. I guess it's too late to change that now, but we should at least fix the script so we don't report it missing when socketcall is set. However, if your unistd.h has defined __NR_recvmmsg before, you should probably add it to the syscall table, just in case that someone built a binary with that number. Arnd