From mboxrd@z Thu Jan  1 00:00:00 1970
X-Spam-Checker-Version: SpamAssassin 3.4.4 (2020-01-24) on polar.synack.me
X-Spam-Level: 
X-Spam-Status: No, score=-1.3 required=5.0 tests=BAYES_00,INVALID_MSGID
	autolearn=no autolearn_force=no version=3.4.4
X-Google-Language: ENGLISH,ASCII-7-bit
X-Google-Thread: 103376,71a5620985deba52
X-Google-Attributes: gid103376,public
From: dkusu@atc-1s.hac.com (David Kusuda)
Subject: core Files with VADS 6.2.1(l)
Date: 1996/02/22
Message-ID: <4ggdu5$60h@hacgate2.hac.com>#1/1
X-Deja-AN: 140493977
organization: Hughes Aircraft Company
newsgroups: comp.lang.ada
Date: 1996-02-22T00:00:00+00:00
List-Id: <comp.lang.ada>


We're using VADS 6.2.1(l) running under HP-UX 9.05 and trying to use the VADS
V_I_Except package to generate core files for unhandled exceptions.  The problem
that we're encountering is that if the main procedure contains a generic
instantiation, the core file is useless...it just takes us to the end of the
main procedure.

For example:

  with Text_IO;
  with V_I_Except;

  procedure Test is

    package IIO is new Text_IO.Integer_IO (Integer);

  begin
    V_I_Except.Enable_Unhandled_Dump_Core;
    V_I_Except.Enable_Traceback_Regs;
    .
    .
    .
  end Test;

For the above code fragment, if we get an unhandled exception, the debugger
points to the "end Test;" line.  But if instead we had:

  with Text_IO;

  package Extra_Package is

    package IIO is new Text_IO.Integer_IO (Integer);

  end Extra_Package;

  with Extra_Package;
  with V_I_Except;

  procedure Test is

    package IIO renames Extra_Package.IIO;

  begin
    V_I_Except.Enable_Unhandled_Dump_Core;
    V_I_Except.Enable_Traceback_Regs;
    .
    .
    .
  end Test;

If we get an unhandled exception with this construct, the debugger takes us
right to the offending line.

We're trying to backfit the V_I_Except package into existing code so we really
don't want to have to create new packages, since it won't necessarily be as
trivial as the above examples.  Has anyone else experienced the same thing, and
if so, is there a different work-around?  Does this problem exist in later
releases?



Dave