From mboxrd@z Thu Jan 1 00:00:00 1970 From: Luc Van Oostenryck Subject: Re: [PATCH 1/2] give a type to builtin functions Date: Wed, 3 Jan 2018 17:41:01 +0100 Message-ID: <20180103164057.qmhyx6eb47c26mm2@ltop.local> References: <20180101225817.3cb135dc04ea0195f28d8512@redchan.it> <20180102151116.22060-1-luc.vanoostenryck@gmail.com> <20180102151116.22060-2-luc.vanoostenryck@gmail.com> <20180103155743.240cb67f567d3fb4ae970ef7@redchan.it> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Received: from mail-wm0-f51.google.com ([74.125.82.51]:43273 "EHLO mail-wm0-f51.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751097AbeACQlE (ORCPT ); Wed, 3 Jan 2018 11:41:04 -0500 Received: by mail-wm0-f51.google.com with SMTP id n138so3673653wmg.2 for ; Wed, 03 Jan 2018 08:41:04 -0800 (PST) Content-Disposition: inline In-Reply-To: <20180103155743.240cb67f567d3fb4ae970ef7@redchan.it> Sender: linux-sparse-owner@vger.kernel.org List-Id: linux-sparse@vger.kernel.org To: foobar Cc: linux-sparse@vger.kernel.org On Wed, Jan 03, 2018 at 03:57:43PM +0000, foobar wrote: > On Tue, 2 Jan 2018 16:11:15 +0100 > Luc Van Oostenryck wrote: > > > When creating builtin functions via init-builtins(), it shouldn't > > really be needed to give them a type because their purpose is only > > as a placeholder for the symbol->op->{evaluate,expand,..}. Also, > > they should be part of the builtin_scope and never be visible. > > yeah, it seems odd (and annoying at the same time) that ctags emits internal symbols __builtin_bswap16 and similar stuff, unlike exuberant ctags. There is several ways to change this but I need to think a bit about it. There is also that sparse's ctags work on the preprocessed source while, I think, other ctags work purely on the lexical level, thus on non-preprocessed sources. Ciao and hanks for the bug repport -- Luc