From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 5023 invoked by alias); 10 Nov 2004 19:43:17 -0000 Mailing-List: contact gdb-patches-help@sources.redhat.com; run by ezmlm Precedence: bulk List-Subscribe: List-Archive: List-Post: List-Help: , Sender: gdb-patches-owner@sources.redhat.com Received: (qmail 4724 invoked from network); 10 Nov 2004 19:43:05 -0000 Received: from unknown (HELO balder.inter.net.il) (192.114.186.15) by sourceware.org with SMTP; 10 Nov 2004 19:43:05 -0000 Received: from zaretski ([80.230.154.117]) by balder.inter.net.il (Mirapoint Messaging Server MOS 3.3.7-GR) with ESMTP id DVZ88241 (AUTH halo1); Wed, 10 Nov 2004 21:32:39 +0200 (IST) Date: Wed, 10 Nov 2004 19:43:00 -0000 From: "Eli Zaretskii" To: Daniel Jacobowitz Message-ID: <01c4c75b$Blat.v2.2.2$5af9fae0@zahav.net.il> Content-Transfer-Encoding: 7BIT Content-Type: text/plain; charset=ISO-8859-1 CC: jjohnstn@redhat.com, cagney@gnu.org, gdb-patches@sources.redhat.com In-reply-to: <20041109193124.GA4085@nevyn.them.org> (message from Daniel Jacobowitz on Tue, 9 Nov 2004 14:31:24 -0500) Subject: Re: [RFA]: Watchpoints per thread patch Reply-to: Eli Zaretskii References: <01c4bca9$Blat.v2.2.2$adcffb00@zahav.net.il> <418A741C.4080306@redhat.com> <20041105044917.GA13554@nevyn.them.org> <418BAFC9.6050705@gnu.org> <20041105182850.GA22533@nevyn.them.org> <418FE5E7.3070501@gnu.org> <20041109010425.GA31431@nevyn.them.org> <4190292D.5070103@gnu.org> <20041109023306.GA1797@nevyn.them.org> <4191151F.6070607@redhat.com> <20041109193124.GA4085@nevyn.them.org> X-SW-Source: 2004-11/txt/msg00207.txt.bz2 > Date: Tue, 9 Nov 2004 14:31:24 -0500 > From: Daniel Jacobowitz > Cc: Andrew Cagney , Eli Zaretskii , > gdb-patches@sources.redhat.com > > We could use a Linux native specific observer, or handle this through > the target stack. I think handling it through the target stack makes > more sense, but I haven't sketched out what the target method would > look like. If other GDB developers think that the precedent of a > native-code-only observer isn't a bad one, then maybe we should go back > to your previous placement of the observer and give it a Linux specific > name. Is there any significant difference between native-code-only observers and the other kind? Could you elaborate? Anyway, I said in the past several times that I don't really like to use the observers too much. The reason for that is that with a mechanism such as this, which is like hooks in Emacs or interrupt handlers in the old DOS days, you often get in trouble once more than one observer is hooked to some event: the order of the the observers' invocation might matter, and AFAIK we do not have any way to control that (nor would we know what order is ``right'', even if we had a way to control it). So in general, if there's a reasonably good design that avoids using observers, I'd favor that.