Re: Patch: catch_errors() cleanup chain restore

[Andrew Cagney <ac131313 at cygnus dot com>]

nsd@cygnus.com wrote:
> 
> >I can
> >(just) live with GDB occasionally leaking memory
> 
> The memory leak happened to be in one of the first RETURN_MASK_ERROR
> cases I checked, so I stopped checking.  There might be worse bugs
> lurking in other cases.
> 
> >This makes me think that the catch_errors() and return_to_top_level()'s
> >current unhealthy obsession with those error_return and quit_return
> >buffers should be cured.
> 
> Yes, please!
[...]
> >BTW, as far as I can tell the FIXME comment I made is still true -
> >fixing the problem involves examining every catch_errors call :-(

Actually, the fixme comment contains a bug!

      val = (*func) (args);
      /* FIXME: cagney/1999-11-05: A correct FUNC implementaton will
         clean things up (restoring the cleanup chain) to the state
         they were just prior to the call.  Technically, this means
         that the below restore_cleanups call is redundant.
         Unfortunatly, many FUNC's are not that well behaved.
         restore_cleanups should either be replaced with a do_cleanups
         call (to cover the problem) or an assertion check to detect
         bad FUNCs code. */

The restore_cleanups() call is always needed.  However, before that call
there should either be a call to do_cleanups() or a check that the
cleanup chain has been emptied.  Without that GDB could end up leaking
memory.  Problem is that I suspect that there is code abusing the
current behavour :-(

	Andrew