From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 25413 invoked by alias); 6 Jul 2004 18:30:32 -0000 Mailing-List: contact gdb-patches-help@sources.redhat.com; run by ezmlm Precedence: bulk List-Subscribe: List-Archive: List-Post: List-Help: , Sender: gdb-patches-owner@sources.redhat.com Received: (qmail 25405 invoked from network); 6 Jul 2004 18:30:31 -0000 Received: from unknown (HELO nevyn.them.org) (66.93.172.17) by sourceware.org with SMTP; 6 Jul 2004 18:30:31 -0000 Received: from drow by nevyn.them.org with local (Exim 4.34 #1 (Debian)) id 1Bhuho-0000hC-CZ; Tue, 06 Jul 2004 14:30:12 -0400 Date: Tue, 06 Jul 2004 18:30:00 -0000 From: Daniel Jacobowitz To: Michael Elizabeth Chastain Cc: ac131313@redhat.com, gdb-patches@sources.redhat.com Subject: Re: [rfa] New test sigbpt.{c,exp} Message-ID: <20040706183012.GA2583@nevyn.them.org> Mail-Followup-To: Michael Elizabeth Chastain , ac131313@redhat.com, gdb-patches@sources.redhat.com References: <20040706181221.153B24B104@berman.michael-chastain.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20040706181221.153B24B104@berman.michael-chastain.com> User-Agent: Mutt/1.5.5.1+cvs20040105i X-SW-Source: 2004-07/txt/msg00050.txt.bz2 On Tue, Jul 06, 2004 at 02:12:21PM -0400, Michael Chastain wrote: > drow> 0x080483f2 in bowler () at /opt/src/gdb/src/gdb/testsuite/gdb.base/sigbpt.c:38 > drow> 38 return *(char *) (v1 + v2 + v3); > drow> 1: x/i $pc 0x80483f2 : mov $0x55c35d00,%esi > drow> (gdb) FAIL: gdb.base/sigbpt.exp: stepi out of handler; stepi bp at segv > drow> > drow> That's not the instruction that should be there; there's a multi-byte > drow> instruction at . > > How about: > > (1) before running the program, 'disassemble bowler' > (2) at the breakpoint, 'x/i $pc' > check that the output of (2) matches one of the lines in (1) That's not a good generic fix; it's just a peculiarity of x86 that we get this particular problem. If that hadn't been a multibyte opcode something different would have happened. I think just executing another instruction here after the faulting one is enough to match the kfail, whether it's the size of a breakpoint or not. -- Daniel Jacobowitz