From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 25145 invoked by alias); 25 Feb 2008 19:39:49 -0000 Received: (qmail 25137 invoked by uid 22791); 25 Feb 2008 19:39:48 -0000 X-Spam-Check-By: sourceware.org Received: from rock.gnat.com (HELO rock.gnat.com) (205.232.38.15) by sourceware.org (qpsmtpd/0.31) with ESMTP; Mon, 25 Feb 2008 19:39:14 +0000 Received: from localhost (localhost.localdomain [127.0.0.1]) by filtered-rock.gnat.com (Postfix) with ESMTP id 5F0752AA4DC; Mon, 25 Feb 2008 14:39:12 -0500 (EST) Received: from rock.gnat.com ([127.0.0.1]) by localhost (rock.gnat.com [127.0.0.1]) (amavisd-new, port 10024) with LMTP id TmfBXs7czsTG; Mon, 25 Feb 2008 14:39:12 -0500 (EST) Received: from joel.gnat.com (localhost.localdomain [127.0.0.1]) by rock.gnat.com (Postfix) with ESMTP id 27ECB2AA499; Mon, 25 Feb 2008 14:39:12 -0500 (EST) Received: by joel.gnat.com (Postfix, from userid 1000) id 30314E7ACB; Mon, 25 Feb 2008 11:39:10 -0800 (PST) Date: Tue, 26 Feb 2008 00:53:00 -0000 From: Joel Brobecker To: Michael Snyder Cc: Vladimir Prus , gdb-patches@sources.redhat.com Subject: Re: [RFA] Try2: Ignore breakpoints when reading memory. Message-ID: <20080225193910.GB20216@adacore.com> References: <200802241750.41264.vladimir@codesourcery.com> <1203967172.19253.273.camel@localhost.localdomain> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <1203967172.19253.273.camel@localhost.localdomain> User-Agent: Mutt/1.4.2.2i 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: 2008-02/txt/msg00376.txt.bz2 > On Sun, 2008-02-24 at 17:50 +0300, Vladimir Prus wrote: > > This patch is the immediate prerequisite for the > > patch that makes breakpoints always inserted. > > Hmmm? Did I miss this discussion? I think this is part of: Non-stop multi-threaded debugging http://sourceware.org/ml/gdb/2007-11/msg00198.html You need to leave the breakpoints inserted if you are to allow the other threads to continue running. > I'm not sure leaving breakpoints always inserted is a good idea. > If gdb becomes unexpectedly disconnected from the target/child, > these breakpoints become an instant problem, and even if you > can manage to re-connect, it's not a given that you can then > remove them. That's indeed a risk, but I believe that this behavior would not be the default. Incidentally, I think this could also be very useful in the context of very very slow debugging lines. We very recently had a customer who asked about the reason for removing the breakpoints at every stop, because apparently with his JTAG probe, inserting 30 breakpoints and then removing them takes a long time (~10secs). We have also verified that writing memory takes equally long. The customer is still waiting to hear from the manufacturer about this lack of performance which seems a bit suspicious, but having the option of keeping the breakpoints inserted in this case would be a nice work-around. -- Joel