From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 12277 invoked by alias); 9 Oct 2003 20:52:53 -0000 Mailing-List: contact gdb-help@sources.redhat.com; run by ezmlm Precedence: bulk List-Subscribe: List-Archive: List-Post: List-Help: , Sender: gdb-owner@sources.redhat.com Received: (qmail 12270 invoked from network); 9 Oct 2003 20:52:52 -0000 Received: from unknown (HELO localhost.redhat.com) (207.219.125.105) by sources.redhat.com with SMTP; 9 Oct 2003 20:52:52 -0000 Received: from redhat.com (localhost [127.0.0.1]) by localhost.redhat.com (Postfix) with ESMTP id D0F192B89; Thu, 9 Oct 2003 16:52:49 -0400 (EDT) Message-ID: <3F85CAA1.9080302@redhat.com> Date: Thu, 09 Oct 2003 20:52:00 -0000 From: Andrew Cagney User-Agent: Mozilla/5.0 (X11; U; NetBSD macppc; en-US; rv:1.0.2) Gecko/20030820 X-Accept-Language: en-us, en MIME-Version: 1.0 To: Jim Blandy Cc: gdb@sources.redhat.com Subject: Re: "struct target_ops" -> "struct gdbtarg" || "struct target" References: <3F856516.9020904@redhat.com> <3F8567D2.3020902@redhat.com> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-SW-Source: 2003-10/txt/msg00167.txt.bz2 > Andrew Cagney writes: > > >> > Hello, >> > The current "target_ops" structure appeared with GDB 4. The >> > original implementation containing only methods. Since then the >> > target_ops have evolved to include data vis: >> > struct section_table >> > *to_sections; >> > struct section_table >> > *to_sections_end; >> > I think, the vector should be re-named to "struct target" or "struct >> > gdbtarg" (consistent with gdbarch, and more name space proof) so >> > that it correctly reflects its current implementation. >> > I'd like to do this now, before the target methods start being >> > explicitly parameterized with their target vector. > >> >> I should note that an alternative is to have "struct gdbarch" as the >> object and "struct target_ops" as the methods vis: >> >> struct target >> { >> .. data elements ...; >> const struct target_ops *ops; >> }; > > > You mean 'struct target' not 'struct gdbarch', right? Obviously. > The second > alternative seems better to me, since it separates the static stuff > from the dynamic stuff. (Except, of course, that the 'static' stuff > isn't actually static, because of update_current_target...) While it seems better, it's significantly harder. It will involve a period of rope jumping (which will likely trip up everyone) during which code wanting to modify things like "target_ops.to_section_end" will co-exist with code wanting to modify "target.to_section_end". There are likely ways to ease the pain, but lets not ignore that there will be pain. Given this, I think the most straight forward step is a straight: "struct target_ops" -> "struct gdbtarg" transformation. Moving the "ops" to a separate "static" vector being a second pass. Andrew