From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 35443 invoked by alias); 12 Feb 2019 19:35:07 -0000 Mailing-List: contact gdb-help@sourceware.org; run by ezmlm Precedence: bulk List-Id: List-Subscribe: List-Archive: List-Post: List-Help: , Sender: gdb-owner@sourceware.org Received: (qmail 35418 invoked by uid 89); 12 Feb 2019 19:35:06 -0000 Authentication-Results: sourceware.org; auth=none X-Spam-SWARE-Status: No, score=-1.9 required=5.0 tests=BAYES_00,RCVD_IN_DNSWL_NONE autolearn=ham version=3.3.2 spammy=HTo:U*gdb, offer X-HELO: mail-wm1-f51.google.com Received: from mail-wm1-f51.google.com (HELO mail-wm1-f51.google.com) (209.85.128.51) by sourceware.org (qpsmtpd/0.93/v0.84-503-g423c35a) with ESMTP; Tue, 12 Feb 2019 19:35:05 +0000 Received: by mail-wm1-f51.google.com with SMTP id b11so4338288wmj.1 for ; Tue, 12 Feb 2019 11:35:05 -0800 (PST) Return-Path: Received: from ?IPv6:2001:8a0:f913:f700:75e6:857f:3506:a1f4? ([2001:8a0:f913:f700:75e6:857f:3506:a1f4]) by smtp.gmail.com with ESMTPSA id i13sm18324035wrm.86.2019.02.12.11.35.02 (version=TLS1_2 cipher=ECDHE-RSA-AES128-GCM-SHA256 bits=128/128); Tue, 12 Feb 2019 11:35:02 -0800 (PST) Subject: Re: GDB is broken on systems with readline built without -fexceptions To: Brian Vandenberg , gdb@sourceware.org References: From: Pedro Alves Message-ID: Date: Tue, 12 Feb 2019 19:35:00 -0000 User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:52.0) Gecko/20100101 Thunderbird/52.9.1 MIME-Version: 1.0 In-Reply-To: Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 8bit X-SW-Source: 2019-02/txt/msg00014.txt.bz2 On 02/09/2019 01:28 AM, Brian Vandenberg wrote: > On 04/21/2016 03:31 PM, Pedro Alves wrote:  > > Forgot to qualify "broken".  The symptom is:  > >    > >  (gdb) whatever-command-that-causes-an-error  > >  terminate called after throwing an instance of 'gdb_exception_RETURN_MASK_ERROR'  > >  Aborted  > >  $  > >  > A proper fix is turning out to be more complicated than I'd hope,  > so in the mean time, to get things working again, I flipped gdb's  > TRY/CATCH mechanism back to setjmp/longjmp instead of C++ try/catch,  > again.  > You should no longer need the --enable-build-with-cxx=no workaround.  > Thanks,  > Pedro Alves > > Note that the original problems that led to that "broken" above have been fixed, a long while ago. > I'm still waiting for IT to get me the 8.2 source so I can try that version, but in the meantime: I'm seeing this error in Solaris 10 (sparc) with 8.1.1.  I built gdb configured like this: > > /path/to/src/configure \ >   --prefix=/install/path \ >   --enable-plugins \ >   --enable-tui \ >   --with-python=yes \ >   --disable-largefile \ >   --without-system-readline \ >   --enable-64-bit-bfd \ >   --with-expat \ >   --with-libexpat-prefix=/opt/csw \ >   --with-x \ >   LDFLAGS="$(echo -n -Wl,-{L,R}/opt/csw/lib)" \ >   {CFLAGS,CPPFLAGS,CXXFLAGS}=-isystem/opt/csw/include > > I don't recall the rationale behind all of the above options.  Some of them came from how the folks at opencsw.org build GDB, others were me just banging on it until it compiled successfully (perhaps to my own detriment). > > If you can offer any suggestions I would appreciate it. I'd suspect that this happens because GDB is built/linked with "-static-libstdc++ -static-libgcc", and that doesn't work correctly on Solaris. ISTR that we observed that on other ports as well. Thanks, Pedro Alves