From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <gdb-patches-return-14263-listarch-gdb-patches=sourceware.cygnus.com@sources.redhat.com>
Received: (qmail 11017 invoked by alias); 13 Apr 2002 01:24:18 -0000
Mailing-List: contact gdb-patches-help@sources.redhat.com; run by ezmlm
Precedence: bulk
List-Subscribe: <mailto:gdb-patches-subscribe@sources.redhat.com>
List-Archive: <http://sources.redhat.com/ml/gdb-patches/>
List-Post: <mailto:gdb-patches@sources.redhat.com>
List-Help: <mailto:gdb-patches-help@sources.redhat.com>, <http://sources.redhat.com/ml/#faqs>
Sender: gdb-patches-owner@sources.redhat.com
Received: (qmail 11000 invoked from network); 13 Apr 2002 01:24:14 -0000
Received: from unknown (HELO duracef.shout.net) (204.253.184.12)
  by sources.redhat.com with SMTP; 13 Apr 2002 01:24:14 -0000
Received: (from mec@localhost)
	by duracef.shout.net (8.11.6/8.11.6) id g3D1ODV17848;
	Fri, 12 Apr 2002 20:24:13 -0500
Date: Fri, 12 Apr 2002 18:24:00 -0000
From: Michael Elizabeth Chastain <mec@shout.net>
Message-Id: <200204130124.g3D1ODV17848@duracef.shout.net>
To: gdb-patches@sources.redhat.com, kevinb@redhat.com
Subject: Re: [patch] gdb.c++/local.exp: add pr numbers
X-SW-Source: 2002-04/txt/msg00484.txt.bz2

Kevin Buettner asks:
> What happens as the bugs are fixed?  Do we then take the PR numbers
> out?

That brings up a good issue.

The test code that I have right now is resilient to this.  Viz:

  send_gdb "foo ..."
  gdb_expect {
    -re "bar\r\n$gdb_prompt $" {
      pass "foo"
    }
    -re "bletch\r\n$gdb_prompt $" {
      # setup_kfail "gnats:gdb/NNN"
      fail "foo (gnats:gdb/NNN)"
    }
    -re ".*$gdb_prompt $" {
      fail "foo"
    }
    timeout {
      fail "foo (timeout)"
    }

"bar" is a good response.  "bletch" is a particular known bad response.
So as long as gdb responds with "bletch", it will be a FAIL
(soon to be a KFAIL).

When gdb starts responding with "bar", a good response, the test will PASS.

There is no need to remove the case for "bletch" for a long time.
Eventually someone will do a cleanup pass and notice that the bug for
"bletch" got fixed 2 years ago.  It's never dangerous to remove the
"bletch" case because that will just give the ordinary ".*" FAIL again.

I was thinking about this today.  The real issue is where to represent
information about known failures.  The bug database holds one view of
this information, and now we are putting it into the test scripts as well,
via KFAIL.  That means significantly more activity in the test scripts.

Michael C