From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 30414 invoked by alias); 14 Jun 2010 15:59:13 -0000 Received: (qmail 30393 invoked by uid 22791); 14 Jun 2010 15:59:12 -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, 14 Jun 2010 15:59:08 +0000 Received: (qmail 20035 invoked from network); 14 Jun 2010 15:59:07 -0000 Received: from unknown (HELO orlando.localnet) (pedro@127.0.0.2) by mail.codesourcery.com with ESMTPA; 14 Jun 2010 15:59:07 -0000 From: Pedro Alves To: gdb-patches@sourceware.org Subject: Re: [rfc] Handle lack of non-stop support more gracefully Date: Mon, 14 Jun 2010 15:59:00 -0000 User-Agent: KMail/1.13.2 (Linux/2.6.32-22-generic; KDE/4.4.2; x86_64; ; ) Cc: "Ulrich Weigand" References: <201006141549.o5EFnuGE029348@d12av02.megacenter.de.ibm.com> In-Reply-To: <201006141549.o5EFnuGE029348@d12av02.megacenter.de.ibm.com> MIME-Version: 1.0 Content-Type: Text/Plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Message-Id: <201006141659.03332.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/msg00311.txt.bz2 On Monday 14 June 2010 16:49:56, Ulrich Weigand wrote: > > I'm still a bit confused over how the non-stop MI handle this. If (for the record, I meant "the all-stop MI tests" here.) > > mi_gdb_target_cmd fails to connect, it seems to return 0 anyway, so > > the following tests will just cascade in FAILs. For other, non-remote > > targets, mi_gdb_target_load will call perror on connection fail, but > > the gdbserver branch at the top doesn't. > > I would suggest that just about any use of perror is wrong here. If > the underlying library routine detects any condition that makes the > rest of the test execution impossible, it should itself issue an > appropriate test status, which would usually be FAIL (if the condition > is due to a GDB bug), UNSUPPORTED (if it is due to some feature not > available on the platform), or UNRESOLVED (if it is due to some setup > or other external issue, like the target connection failing). > > Then, the library should return an error code that causes the main > test case to silently stop any further test execution. Thanks. Sounds like a good plan. -- Pedro Alves