From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754110Ab0JKX32 (ORCPT ); Mon, 11 Oct 2010 19:29:28 -0400 Received: from iona.labri.fr ([147.210.8.143]:57697 "EHLO iona.labri.fr" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753449Ab0JKX31 (ORCPT ); Mon, 11 Oct 2010 19:29:27 -0400 X-Greylist: delayed 1743 seconds by postgrey-1.27 at vger.kernel.org; Mon, 11 Oct 2010 19:29:26 EDT Date: Tue, 12 Oct 2010 01:00:17 +0200 From: Samuel Thibault To: Tracey Dent Cc: greg@kroah.com, w.d.hubbs@gmail.com, chris@the-brannons.com, kirk@braille.uwo.ca, linux-kernel@vger.kernel.org Subject: Re: [RFC PATCH] Staging: speakup: Added EXPORT_SYMBOL_GPL() markings. Message-ID: <20101011230017.GT4721@const.famille.thibault.fr> Mail-Followup-To: Samuel Thibault , Tracey Dent , greg@kroah.com, w.d.hubbs@gmail.com, chris@the-brannons.com, kirk@braille.uwo.ca, linux-kernel@vger.kernel.org References: <1286836590-6708-1-git-send-email-tdent48227@gmail.com> MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Disposition: inline Content-Transfer-Encoding: 8bit In-Reply-To: <1286836590-6708-1-git-send-email-tdent48227@gmail.com> User-Agent: Mutt/1.5.12-2006-07-14 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Tracey Dent, le Mon 11 Oct 2010 18:36:30 -0400, a écrit : > -void speakup_clear_selection(void) > -{ > - sel_start = -1; > -} Mmm, this won't be so simple actually, as speakup_set_selection and speakup_paste_selection will still be using speakup/selection.c's sel_start variable. I guess all speakup/selection.c will need to be migrated to using drivers/char/selection.c functions at the same time. speakup_paste_selection is basically like paste_selection, except that it can be called with in_atomic() being true. Yes, it's a hack. The calls to speakup_past_selection should first be moved to a helper thread, and then the main paste_selection function can be used instead. speakup_set_selection is basically like set_selection, except that it doesn't take a userland structure, but gets the values from kernelland speakup. A kernel-land variant of set_selection should probably be defined that takes a kernelland structure. Apart from that, I don't think speakup_set_selection has anything particular compared to the main version. Kirk, can you confirm? Samuel