From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 19912 invoked by alias); 16 Mar 2011 19:59:52 -0000 Received: (qmail 19904 invoked by uid 22791); 16 Mar 2011 19:59:51 -0000 X-SWARE-Spam-Status: No, hits=-1.9 required=5.0 tests=AWL,BAYES_00,T_RP_MATCHES_RCVD X-Spam-Check-By: sourceware.org Received: from mail.codesourcery.com (HELO mail.codesourcery.com) (38.113.113.100) by sourceware.org (qpsmtpd/0.43rc1) with ESMTP; Wed, 16 Mar 2011 19:59:47 +0000 Received: (qmail 17940 invoked from network); 16 Mar 2011 19:59:46 -0000 Received: from unknown (HELO macbook-2.local) (stan@127.0.0.2) by mail.codesourcery.com with ESMTPA; 16 Mar 2011 19:59:46 -0000 Message-ID: <4D8116B1.90605@codesourcery.com> Date: Wed, 16 Mar 2011 20:03:00 -0000 From: Stan Shebs User-Agent: Mozilla/5.0 (Macintosh; U; Intel Mac OS X 10.5; en-US; rv:1.9.2.15) Gecko/20110303 Thunderbird/3.1.9 MIME-Version: 1.0 To: gdb-patches@sourceware.org Subject: Re: [RFA] make first parameter of to_lookup_symbol const char * References: <58CFFAD9-1C02-4AD8-B416-FC48D1A7F006@adacore.com> <20110314134616.GA2680@host1.jankratochvil.net> <20110316135810.GA6590@adacore.com> In-Reply-To: <20110316135810.GA6590@adacore.com> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Mailing-List: contact gdb-patches-help@sourceware.org; run by ezmlm Precedence: bulk List-Id: List-Subscribe: List-Archive: List-Post: List-Help: , Sender: gdb-patches-owner@sourceware.org X-SW-Source: 2011-03/txt/msg00874.txt.bz2 On 3/16/11 6:58 AM, Joel Brobecker wrote: >>> BTW, it looks like no target defines this operation... >> Unless you use it for some new patch of yours it should be removed instead. > This is what it looks like to remove the target_ops method. It feels > a little like excising a potentially useful feature, so I'm not going > to commit without review, although there is no sign that we'll ever > need it any time soon. But I added a comment explaining what we used > to do, to give us a clue later on, if we encounter a target where > we might need something of this kind. Yeah, it looks like the last use quietly evaporated with the deletion of remote-vx.c in 2004. Features without any means of exercise are likely to bitrot semantically even if they continue to compile/run (witness tracepoints), so it's in our interest to be ruthless. I used to have fun using gcov with the GDB testsuite to find segments of code that were never exercised - gcov results are cumulative, so the post-testing coverage display is good for ideas as to what test cases ought to be written, and also suggests code that might turn out to be intrinsically unreachable. Perhaps msnyder can play with that when he gets tired of coverity. :-) Stan