Mailing-List: contact cygwin-help@cygwin.com; run by ezmlm List-Subscribe: List-Archive: List-Post: List-Help: , Sender: cygwin-owner@cygwin.com Mail-Followup-To: cygwin@cygwin.com Delivered-To: mailing list cygwin@cygwin.com To: cygwin@cygwin.com From: Shankar Unni Subject: Re: cygwin, libtool, dlpreopen, and .rdata Date: Thu, 23 Sep 2004 10:11:16 -0700 Lines: 21 Message-ID: References: <41511C3F.7080003@cwilson.fastmail.fm> <4151E85A.5020305@cwilson.fastmail.fm> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: ppp-67-124-90-144.dsl.pltn13.pacbell.net User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; rv:1.7.3) Gecko/20040913 Thunderbird/0.8 Mnenhy/0.6.0.101 In-Reply-To: <4151E85A.5020305@cwilson.fastmail.fm> X-IsSubscribed: yes Charles Wilson wrote: > Actually, *function* addresses are fine; they don't get relocated (at > least, not by the runtime-pseudo-reloc gobbledygook). It's only the > addresses of DATA items exported by DLLs that get relocated in this way. > > AND, they get relocated IF AND ONLY IF you are linking to a DLL. On x86/PE, anyway. The rules for other platforms and object-file formats vary. Leaving this as "implementation-defined" (may or may not work - documented, but how easy is it to track down this documentation?) makes the programmer's task more difficult - they can never be sure where they can use 'const' on an initialized structure definition. What if function addresses are also 'bad' on some other platform? Gcc should either warn about this usage, or it should silently "do the right thing" on each platform based on its capabilities. But I'm just being picky, being an old compiler fart.. -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/