Re: [RFA]: Fix for pending breakpoints in manually loaded/unloaded shlibs

[Michael Chastain <mec.gnu@mindspring.com>]

Jeff Johnston <jjohnstn@redhat.com> wrote:
> The test fails now because gdb is not issuing the message that is
> currently in the test script.  The message in the test script is a new
> message which does not currently exist and never did in any gdb
> release.

Okay, that unconfuses me.

> Thus, I was wondering if I still need to have checks for old 
> and new messages in the script.

If gdb 6.2 didn't print a message, then you don't need to have it in the
script.  What I care about is comparing gdb 6.2 and gdb 6.2.91 with the
test suite that's going to be in gdb 6.2.91.

On the last spin I did with gdb 6.2 and the current version
of unload.exp (from gdb HEAD 2004-08-18), I got this:

  (gdb) PASS: gdb.base/unload.exp: running program
  continue
  Continuing.
  y is 7
  
  Program exited normally.
  (gdb) FAIL: gdb.base/unload.exp: continuing to end of program
  run
  Starting program: /tmp/migbat-testgdb-8amaeH24/test/gdb.base/unload 
  Warning:
  Cannot insert breakpoint 2.
  Error accessing memory address 0x400136cb: Input/output error.
  
  (gdb) FAIL: gdb.base/unload.exp: rerun to shared library breakpoint

Which is cool.  The test script is doing its job, making gdb barf.
That will be a genuine "before" FAIL in the tables, not a fake "before".

Maybe someday, omeone can even mine this data source and add more
"bugs fixed in this release" sections to NEWS.