Newsgroups: php.internals
Path: news.php.net
Xref: news.php.net php.internals:32703
Mailing-List: contact internals-help@lists.php.net; run by ezmlm
Received-SPF: error (pb1.pair.com: domain tonybibbs.com from 209.191.119.231 cause and error)
Date: Mon, 8 Oct 2007 10:04:26 -0700 (PDT)
To: internals@lists.php.net
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable
Message-ID: <339257.69263.qm@web706.biz.mail.mud.yahoo.com>
Subject: Re: [PHP-DEV] Reference cycle collector patch
From: tony@tonybibbs.com (Tony Bibbs)

For the untrained eye, would this then fix these bugs?=0A=0Ahttp://bugs.php=
.net/bug.php?id=3D33595=0Ahttp://bugs.php.net/bug.php?id=3D33487=0A=0A--Ton=
y=0A=0A----- Original Message ----=0AFrom: David Wang <planetbeing@gmail.co=
m>=0ATo: internals@lists.php.net=0ASent: Monday, October 8, 2007 2:12:31 AM=
=0ASubject: [PHP-DEV] Reference cycle collector patch=0A=0AHey all,=0A=0ANo=
w that macros for manipulating refcount and is_ref have been implemented, I=
=E2=80=99m ready to submit patches for the GC. I know you all have been wai=
ting for this for a long time =E2=98=BA.=0A=0AI have made two sets of patch=
es, one for HEAD and one for PHP_5_3 (note that I had to rearrange the orde=
r of definitions and #includes in zend.h to get around a recursive dependen=
cy problem):=0A=0Ahttp://zoo.cs.yale.edu/~yw226/gc/gc6.diff.txt=0Ahttp://zo=
o.cs.yale.edu/~yw226/gc/gc5_3.diff.txt=0A=0AAfter applying those, these two=
 files should be placed in the Zend subdirectory:=0A=0Ahttp://zoo.cs.yale.e=
du/~yw226/gc/zend_gc.c=0Ahttp://zoo.cs.yale.edu/~yw226/gc/zend_gc.h=0A=0AFo=
r those who were following the progress of the GC through the incubator SVN=
, the GC implemented with these patches is DIFFERENT than the one currently=
 in the SVN tree. They are from an =E2=80=9Cexperimental=E2=80=9D branch th=
at I believed was more tolerant of the zval juggling PHP and its extensions=
 sometimes like to do. It turned out to be very successful, but I hadn=E2=
=80=99t bothered remerging it.=0A=0AI highly recommend these two patches in=
 addition (but they are optional for GC functioning):=0A=0Ahttp://zoo.cs.ya=
le.edu/~yw226/gc/alwaysinline6.diff.txt=0Ahttp://zoo.cs.yale.edu/~yw226/gc/=
alwaysinline5_3.diff.txt=0A=0AThese are meant to be applied after the previ=
ous patches. These simply change all the inline functions in the Zend engin=
e to =E2=80=9Calways_inline=E2=80=9D. Zend object files can become seriousl=
y oversized and this makes the compiler silently ignore the inline keyword.=
 Changing this to always_inline (defined in zend.h) forces the compiler to =
inline the function or at least give a warning if it cannot.=0A=0AThe reaso=
n I bring up the issue here is that my patch is known to sometimes push obj=
ect files that were just under the internal compiler limit over the edge an=
d make the compiler refuse to inline. This has the effect of making the who=
le Zend engine disproportionately slower.=0A=0AAnyway, please try these out=
 and tell me how it goes!=0A=0A=0ADavid=0A=0AP.S. This patch has the GC ena=
bled by default.=0A=0A--=0APHP Internals - PHP Runtime Development Mailing =
List=0ATo unsubscribe, visit: http://www.php.net/unsub.php=0A=0A=0A=0A