From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 19896 invoked by alias); 6 Sep 2010 19:45:48 -0000 Mailing-List: contact archer-help@sourceware.org; run by ezmlm Sender: Precedence: bulk List-Post: List-Help: List-Subscribe: List-Id: Received: (qmail 19885 invoked by uid 22791); 6 Sep 2010 19:45:46 -0000 X-SWARE-Spam-Status: No, hits=-6.5 required=5.0 tests=AWL,BAYES_00,RCVD_IN_DNSWL_HI,SPF_HELO_PASS,T_RP_MATCHES_RCVD X-Spam-Check-By: sourceware.org Date: Mon, 06 Sep 2010 19:45:00 -0000 From: Oleg Nesterov To: archer@sourceware.org, utrace-devel@redhat.com Subject: Re: gdbstub initial code, v8 && ptrace Message-ID: <20100906194229.GA27405@redhat.com> References: <20100903224047.GA8917@redhat.com> <20100906182359.GB22839@redhat.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20100906182359.GB22839@redhat.com> User-Agent: Mutt/1.5.18 (2008-05-17) X-SW-Source: 2010-q3/txt/msg00152.txt.bz2 On 09/06, Oleg Nesterov wrote: > > On 09/04, Oleg Nesterov wrote: > > > > - implement continue with signal. > > There is something wrong. I tried to debug the ptraced task via ugdb, > and the result is not good. I am not surprised ugdb confuses ptrace, > but I didn't expect that ptrace can break ugdb. > > I am still trying to understand what is wrong, it should be something > simple but somehow I can't explain the problem so far... OK, I am stupid. Indeed, say, "return UTRACE_STOP | UTRACE_SIGNAL_IGN" from under "case UTRACE_SIGNAL_REPORT" changes utrace_report->result, and confuses other tracers. Probably we can ignore ptrace, but this also means ugdb conflicts with itself and should be fixed. Oleg.