Mailing-List: contact lucy-dev-help@incubator.apache.org; run by ezmlm
Precedence: bulk
Reply-To: lucy-dev@incubator.apache.org
Received-SPF: neutral (nike.apache.org: local policy)
Message-ID: <4EBA95BA.5020205@peknet.com>
Date: Wed, 09 Nov 2011 09:01:14 -0600
From: Peter Karman <peter@peknet.com>
Reply-To: peter@peknet.com
User-Agent: Thunderbird 2.0.0.24 (X11/20101027)
MIME-Version: 1.0
To: Marvin Humphrey <marvin@rectangular.com>
CC: lucy-dev@incubator.apache.org
References: <20111109013246.GA26278@rectangular.com>
 <4EB9DCFF.4040706@peknet.com> <20111109053348.GA26543@rectangular.com>
In-Reply-To: <20111109053348.GA26543@rectangular.com>
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 7bit
Subject: Re: [lucy-dev] Non-deterministic destruction in Perl 5.15

Marvin Humphrey wrote on 11/08/2011 11:33 PM:

> It's not surprising that SWISH::3 would pass its tests.  You haven't set up
> tests to validate behavior under refcount leakage, have you?

I do, actually.

http://cpansearch.perl.org/src/KARMAN/SWISH-3-1.000000/t/07-refcnt.t
among others.

I use a ref_cnt counter on the underlying C struct, not relying on the
Perl ref counting. So my DESTROY blocks are conditional on that ref_cnt.

void
DESTROY(self)
    SV *self;

    PREINIT:
        swish_3 *s3;

    CODE:
        s3 = (swish_3*)sp_extract_ptr(self);
        s3->ref_cnt--;

        if (s3->ref_cnt < 1) {
            sp_Stash_destroy( s3->stash );
        }


but SWISH::3 is far less complex than Lucy in terms of the number of
classes and code generation etc, so I was willing to hand-code all the
ref counting etc of internal objects, etc.

> 
> What I'm suggesting if a user writes a program which leaks Lucy objects, and
> if we disconnect that "tripwire" exception without implementing the
> PERL_PHASE_DESTRUCT wrapping, that program may segfault during global
> destruction in Perl 5.15.
> 
> Most small programs don't leak, because most small programs don't have
> circular references.  Circular refs are pretty common in big, complex
> programs, though.
> 
> If SWISH::3 -- or Perl/Tk, or whatever else -- has complex objects which count
> on deterministic order of destruction, I believe that the potential for
> segfaulting during global destruction exists in Perl 5.15.
> 

I agree with all that. I think your patch is the right choice.

-- 
Peter Karman  .  http://peknet.com/  .  peter@peknet.com