From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 14367 invoked by alias); 8 Aug 2008 12:43:13 -0000 Received: (qmail 14359 invoked by uid 22791); 8 Aug 2008 12:43:13 -0000 X-Spam-Check-By: sourceware.org Received: from NaN.false.org (HELO nan.false.org) (208.75.86.248) by sourceware.org (qpsmtpd/0.31) with ESMTP; Fri, 08 Aug 2008 12:42:32 +0000 Received: from nan.false.org (localhost [127.0.0.1]) by nan.false.org (Postfix) with ESMTP id A480D98419; Fri, 8 Aug 2008 12:42:30 +0000 (GMT) Received: from caradoc.them.org (22.svnf5.xdsl.nauticom.net [209.195.183.55]) by nan.false.org (Postfix) with ESMTP id 905A79809F; Fri, 8 Aug 2008 12:42:30 +0000 (GMT) Received: from drow by caradoc.them.org with local (Exim 4.69) (envelope-from ) id 1KRRIb-00067O-AL; Fri, 08 Aug 2008 08:42:29 -0400 Date: Fri, 08 Aug 2008 12:43:00 -0000 From: Daniel Jacobowitz To: Eli Zaretskii Cc: Pedro Alves , gdb-patches@sourceware.org Subject: Re: go32-nat, always a thread Message-ID: <20080808124229.GA23208@caradoc.them.org> Mail-Followup-To: Eli Zaretskii , Pedro Alves , gdb-patches@sourceware.org References: <200808080439.17254.pedro@codesourcery.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.5.17 (2008-05-11) X-IsSubscribed: yes 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: 2008-08/txt/msg00193.txt.bz2 On Fri, Aug 08, 2008 at 11:00:47AM +0300, Eli Zaretskii wrote: > What do we need this for? DJGPP doesn't support threading; the ptid_t > variables in go32-nat.c are simply dummies devised to make the > compiler happy because go32-nat.c uses interfaces that pass thread > IDs. But there's no infrastructure behind this farce. The GDB core has a list of threads. Until recently, a single-threaded program would usually have no threads in the list (though in some cases, there would be a single thread). This makes it difficult to keep track of per-thread state in a uniform way; state that a single-threaded program should have one copy of but a multi-threaded program should have one copy per thread. > The code you added will now cause all kinds of notifications, like > "thread started", be displayed, which is just confusing when the > programmer knows there're no threads. Is this really justified? No notifications should be displayed; that's because Pedro used add_thread_silent and delete_thread_silent. -- Daniel Jacobowitz CodeSourcery