From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 25383 invoked by alias); 7 Jun 2010 18:37:53 -0000 Received: (qmail 25375 invoked by uid 22791); 7 Jun 2010 18:37:52 -0000 X-SWARE-Spam-Status: No, hits=-2.1 required=5.0 tests=AWL,BAYES_00,T_RP_MATCHES_RCVD X-Spam-Check-By: sourceware.org Received: from mail.codesourcery.com (HELO mail.codesourcery.com) (38.113.113.100) by sourceware.org (qpsmtpd/0.43rc1) with ESMTP; Mon, 07 Jun 2010 18:37:46 +0000 Received: (qmail 8308 invoked from network); 7 Jun 2010 18:37:45 -0000 Received: from unknown (HELO orlando.localnet) (pedro@127.0.0.2) by mail.codesourcery.com with ESMTPA; 7 Jun 2010 18:37:45 -0000 From: Pedro Alves To: Michael Snyder Subject: Re: [RFA] fix crasher on detach command Date: Mon, 07 Jun 2010 18:37:00 -0000 User-Agent: KMail/1.13.2 (Linux/2.6.32-22-generic; KDE/4.4.2; x86_64; ; ) Cc: "gdb-patches@sourceware.org" References: <4C0D3636.8040206@vmware.com> <201006071928.19374.pedro@codesourcery.com> <4C0D3AFE.9020400@vmware.com> In-Reply-To: <4C0D3AFE.9020400@vmware.com> MIME-Version: 1.0 Content-Type: Text/Plain; charset="iso-8859-15" Content-Transfer-Encoding: 7bit Message-Id: <201006071937.41182.pedro@codesourcery.com> 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: 2010-06/txt/msg00184.txt.bz2 On Monday 07 June 2010 19:31:26, Michael Snyder wrote: > Pedro Alves wrote: > > On Monday 07 June 2010 19:11:02, Michael Snyder wrote: > >> Hi, > >> > >> The circumstances are, detach from a remote target that doesn't have > >> threads. Remote.c leaves the PID arbitrarily as "42000", and > >> target_detach calls remove_breakpoints_pid, which crashes because > >> find_inferior_pid returns NULL. > > > > It shouldn't matter that the PID is arbitrary; there should be an > > inferior with that PID in the inferior list. This probably means > > that the inferior got its PID cleared to 0 already when you get here? > > How? > > It was a bad connect, which aborted part way through. > So yes, we're in an inconsistent internal state. Then we need to fix that, instead of adding workarounds in other areas. > > What's the backtrace like at the time of the crash? I assume > > the remove_breakpoints_pid call is coming from within target_detach? > > It's fairly normal, target_detach is called by detach_command. Okay. What I was interested was in seeing a paste of the backtrace, to confirm that's the where the call to remove_breakpoints_pid is being done, but what you say seems to confirm it. -- Pedro Alves