Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1756574AbcCDBxd (ORCPT ); Thu, 3 Mar 2016 20:53:33 -0500 Received: from mail-qg0-f48.google.com ([209.85.192.48]:35325 "EHLO mail-qg0-f48.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755678AbcCDBxc (ORCPT ); Thu, 3 Mar 2016 20:53:32 -0500 Date: Thu, 3 Mar 2016 20:53:27 -0500 (EST) From: Nicolas Pitre To: Michal Marek cc: linux-kbuild@vger.kernel.org, linux-kernel@vger.kernel.org Subject: Re: [PATCH v4 6/8] create/adjust generated/autoksyms.h In-Reply-To: <56D8B74F.3040903@suse.com> Message-ID: References: <1456717691-28298-1-git-send-email-nicolas.pitre@linaro.org> <1456717691-28298-7-git-send-email-nicolas.pitre@linaro.org> <56D8B74F.3040903@suse.com> User-Agent: Alpine 2.20 (LFD 67 2015-01-07) MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1598 Lines: 48 On Thu, 3 Mar 2016, Michal Marek wrote: > Hi Nicolas, > > sorry for the late review. > > Dne 29.2.2016 v 04:48 Nicolas Pitre napsal(a): > > +set -e > > + > > +cur_ksyms_file="include/generated/autoksyms.h" > > +new_ksyms_file="include/generated/autoksyms.h.tmpnew" > > + > > +info() { [ "$quiet" != "silent_" ] && printf " %-7s %s\n" "$1" "$2"; } > > Using && as a shorthand for if-statements does not work when running in > -e mode. This particular line causes the script to fail in silent mode. > There other uses of && in the script are inside sub-shells, so they do > work, but it would be good to convert them to avoid breakage if the > script is edited later. OK. > > +# Extract changes between old and new list and touch corresponding > > +# dependency files. > > +# Note: sort -m doesn't work well with underscore prefixed symbols so we > > +# use 'cat ... | sort' instead. > > +changed=$( > > +count=0 > > +cat "$cur_ksyms_file" "$new_ksyms_file" | sort | uniq -u | > > sort "$cur_ksyms_file" "$new_ksyms_file" ? Ah, indeed. Somehow I understood that you could use two files only with -m. > sort -m probably does not work, because the input files are not > sorted. They used to at some point as as the content comes from the earlier "sort -u". I had test samples but I can't find them anymore. Anyway I prefer not to rely on a pre-existing ordering if possible, and with the addition of the file header and the CONFIG_MODVERSIONS case that came afterwards the content is no longer 100% sorted. With "sort file1 file2" the comment is not very relevant anymore. Nicolas