From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 12865 invoked by alias); 14 Dec 2006 19:50:51 -0000 Received: (qmail 12655 invoked by uid 22791); 14 Dec 2006 19:50:49 -0000 X-Spam-Check-By: sourceware.org Received: from mail.codesourcery.com (HELO mail.codesourcery.com) (65.74.133.4) by sourceware.org (qpsmtpd/0.31) with ESMTP; Thu, 14 Dec 2006 19:50:38 +0000 Received: (qmail 32010 invoked from network); 14 Dec 2006 19:50:37 -0000 Received: from unknown (HELO localhost) (jimb@127.0.0.2) by mail.codesourcery.com with ESMTPA; 14 Dec 2006 19:50:37 -0000 To: qinwei Cc: gdb@sourceware.org Subject: Re: Some testpatterns seems have problems References: From: Jim Blandy Date: Thu, 14 Dec 2006 19:50:00 -0000 In-Reply-To: (qinwei@sunnorth.com.cn's message of "Thu, 14 Dec 2006 18:27:06 +0800") Message-ID: User-Agent: Gnus/5.11 (Gnus v5.11) Emacs/22.0.50 (gnu/linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-IsSubscribed: yes 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 X-SW-Source: 2006-12/txt/msg00131.txt.bz2 Hi, Qinwei. (I haven't forgotten about your S+Core patch --- I'll give it a new review in a bit.) qinwei writes: > 1, long_long.exp, > when using p/a, if the address has instructions, gdb will print out the > offset, > > p/a *(char *)c > $32 = 0x1 <_start+1> > (gdb) FAIL: gdb.base/long_long.exp: p/a *(char *)c > > p/a *(short *)s > $40 = 0x123 > (gdb) FAIL: gdb.base/long_long.exp: p/a *(short *)s > > so, > gdb_test_char "p/a *(char *)c" "0x1" > gdb_test_short "p/a *(short *)s" "" "0x123" "" > should be > gdb_test_char "p/a *(char *)c" "0x1.*" > gdb_test_short "p/a *(short *)s" "" "0x123.*" "" Yes, this is right. Although I would prefer if you used a tighter pattern that didn't permit arbitrary garbage: say, "0x123( <.*>)?". These are definitely the sorts of changes which you should post as patches. Describing the fix in English, as you've done, makes it harder to review and test your patch. Generate diffs from the top of the source tree --- the directory containing 'gdb', 'bfd', 'sim', etc.) with 'cvs diff' (or 'cvs diff -N', if you have added files). Don't include the ChangeLog in the diff, because such patches are very likely to conflict needlessly; just put the entry above the patch. And as I've said, don't include diffs to generated files like 'configure' or 'aclocal.m4'. > 2, mi-basics.exp, mi2-basics.exp: > gdb will print out the full path. > > 202-environment-directory ../../../gdb-6.6/gdb/testsuite/gdb.mi > 202^done,source-path="/home/qinwei/gnu/GJ275_CVS/code/build/gdb/testsuite/../../../gdb-6.6/gdb/testsuite/gdb.mi > :$cdir:$cwd" > (gdb) > FAIL: gdb.mi/mi-basics.exp: environment-directory arg operation > > 203-environment-directory > 203^done,source-path="/home/qinwei/gnu/GJ275_CVS/code/build/gdb/testsuite/../../../gdb-6.6/gdb/testsuite/gdb.mi > :$cdir:$cwd" > (gdb) > FAIL: gdb.mi/mi-basics.exp: environment-directory empty-string operation > > 208-environment-path ../../../gdb-6.6/gdb/testsuite > /home/qinwei/gnu/GJ275_CVS/code/build/gdb/testsuite > 208^done,path="/home/qinwei/gnu/GJ275_CVS/code/build/gdb/testsuite/../../../gdb-6.6/gdb/testsuite:/home/qinwei/ > gnu/GJ275_CVS/code/build/gdb/testsuite:/usr/kerberos/bin:/usr/local/bin:/bin:/usr/bin:/usr/X11R6/bin:/home/qinw > ei/gcc/bin:/home/qinwei/usr/bin:/home/qinwei/usr/lib/gcc-lib/unsp-sgi-linux/sgicc-1.0:/home/qinwei/gnu/install- > score-elf/bin:/home/qinwei/bin:/home/qinwei/daVinci_V2.1" > (gdb) > FAIL: gdb.mi/mi-basics.exp: environment-path dir1 dir2 operation Yes, I see these failures too; I think they've been present for a while. Do you have a patch for this?