From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754227Ab3A2KuD (ORCPT ); Tue, 29 Jan 2013 05:50:03 -0500 Received: from mail.skyhub.de ([78.46.96.112]:42377 "EHLO mail.skyhub.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751675Ab3A2KuA (ORCPT ); Tue, 29 Jan 2013 05:50:00 -0500 Date: Tue, 29 Jan 2013 11:49:58 +0100 From: Borislav Petkov To: Ingo Molnar , Arnaldo Carvalho de Melo Cc: LKML , Borislav Petkov Subject: Re: [PATCH 2/2] perf: Make -lnuma optional Message-ID: <20130129104958.GC14302@pd.tnic> Mail-Followup-To: Borislav Petkov , Ingo Molnar , Arnaldo Carvalho de Melo , LKML , Borislav Petkov References: <1359456492-22156-2-git-send-email-bp@alien8.de> MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Disposition: inline In-Reply-To: <1359456492-22156-2-git-send-email-bp@alien8.de> User-Agent: Mutt/1.5.21 (2010-09-15) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Tue, Jan 29, 2013 at 11:48:12AM +0100, Borislav Petkov wrote: > From: Borislav Petkov > > a0c17eadf3bf9 ("perf: Add 'perf bench numa mem' NUMA performance > measurement suite") added -lnuma to the EXTLIBS but we use those when > checking for other libraries like libelf, libunwind, etc and the checks > failed even when the dev libraries are installed on the system: > > CHK libelf > /usr/bin/ld: cannot find -lnuma > collect2: error: ld returned 1 exit status > CHK glibc > Makefile:552: No libelf found, disables 'probe' tool, please install elfutils-libelf-devel/libelf-dev > CHK libunwind > /usr/bin/ld: cannot find -lnuma > collect2: error: ld returned 1 exit status > Makefile:586: No libunwind found, disabling post unwind support. Please install libunwind-dev[el] >= 0.99 > ... > > So, add glue to detect the presence of -lnuma and build perf bench numa > only then. While at it, I noticed another annoyance: if FLEX and BISON are needed unconditionally for the build, this needs to be checked at the beginning of the makefile and make shouldn't fail in-between. Oh well, I'll let someone beat me to it with a fix since I'm lazy :). -- Regards/Gruss, Boris. Sent from a fat crate under my desk. Formatting is fine. --