From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752557AbbJHG3n (ORCPT ); Thu, 8 Oct 2015 02:29:43 -0400 Received: from mga09.intel.com ([134.134.136.24]:38129 "EHLO mga09.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751650AbbJHG3m (ORCPT ); Thu, 8 Oct 2015 02:29:42 -0400 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.17,653,1437462000"; d="scan'208";a="576379437" Subject: Re: [PATCH] perf tools: get version from uname(2), not /proc To: Jiri Olsa , acme@kernel.org, Vinson Lee , Peter Zijlstra , Ingo Molnar , Namhyung Kim , Kan Liang , Andi Kleen , linux-kernel@vger.kernel.org References: <1444171996-24030-1-git-send-email-mmullins@twopensource.com> <20151007081854.GA20302@krava.landal.opennet> <5614D585.4010707@intel.com> <20151007171107.GA8650@whiteoak.sf.office.twttr.net> From: Adrian Hunter Organization: Intel Finland Oy, Registered Address: PL 281, 00181 Helsinki, Business Identity Code: 0357606 - 4, Domiciled in Helsinki Message-ID: <56160CA3.6010703@intel.com> Date: Thu, 8 Oct 2015 09:26:43 +0300 User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:38.0) Gecko/20100101 Thunderbird/38.3.0 MIME-Version: 1.0 In-Reply-To: <20151007171107.GA8650@whiteoak.sf.office.twttr.net> Content-Type: text/plain; charset=windows-1252 Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On 07/10/15 20:11, Matt Mullins wrote: > On Wed, Oct 07, 2015 at 11:19:17AM +0300, Adrian Hunter wrote: >> On 07/10/15 11:18, Jiri Olsa wrote: >>> On Tue, Oct 06, 2015 at 03:53:14PM -0700, Matt Mullins wrote: >>>> Tools in kmod (e.g. modprobe) compose the module path from the release >>>> from uname(2). Because we use the UNAME26 personality, we need perf to >>>> find modules located at the same path as the system tools. >>> >>> I guess it's easy to google this up, but could you >>> please state in the changelog what's the difference >>> between the current version and the UNAME26 one? >>> >>> Also state (and check) this change wouldn't affect other >>> parts of the code that use this version (if there's any). > > This is the only caller of get_kernel_version. > >> Isn't the machine root dir for guests? uname() won't work for them. > > Aye, this would break --guestmount. Would it make sense to use uname() > iff root_dir is the empty string, or is that too much special-casing for > good taste? There is machine__is_host(), but what about /proc/sys/kernel/osrelease or does that have the same problem?