Newsgroups: php.internals
Path: news.php.net
Xref: news.php.net php.internals:87891
Mailing-List: contact internals-help@lists.php.net; run by ezmlm
Received-SPF: error (pb1.pair.com: domain belski.net from 85.214.73.107 cause and error)
To: "'Jakub Zelenka'" <bukka@php.net>
Cc: "'Dmitry Stogov'" <dmitry@php.net>,
	"'Xinchen Hui'" <xinchen.h@zend.com>,
	"'Nikita Popov'" <nikita.ppv@gmail.com>,
	"'Pierre Joye'" <pierre.php@gmail.com>,
	"'Bob Weinand'" <bobwei9@hotmail.com>,
	"'Matt Wilmas'" <php_lists@realplain.com>,
	"'PHP internals list'" <internals@lists.php.net>
References: <02a601d0dbed$2c828df0$8587a9d0$@belski.net> <CAEKnhAHgd29YA75xqMOvTm5cwYY3rR46Fou0M3gJpnjciwr_yw@mail.gmail.com>
In-Reply-To: <CAEKnhAHgd29YA75xqMOvTm5cwYY3rR46Fou0M3gJpnjciwr_yw@mail.gmail.com>
Date: Mon, 24 Aug 2015 15:57:41 +0200
Message-ID: <00ad01d0de74$db4773d0$91d65b70$@belski.net>
MIME-Version: 1.0
Content-Type: text/plain;
	charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Thread-Index: AQK9NdJN9egRUY1Umglkf8yuRLR3fgF4idffnDaM04A=
Content-Language: en-us
Subject: RE: [PHP-DEV] Re: Overflow checks and integral vars comparison
From: anatol.php@belski.net ("Anatol Belski")

Hi Jakub,

> -----Original Message-----
> From: jakub.php@gmail.com [mailto:jakub.php@gmail.com] On Behalf Of =
Jakub
> Zelenka
> Sent: Sunday, August 23, 2015 8:29 PM
> To: Anatol Belski <anatol.php@belski.net>
> Cc: Dmitry Stogov <dmitry@php.net>; Xinchen Hui <xinchen.h@zend.com>;
> Nikita Popov <nikita.ppv@gmail.com>; Pierre Joye =
<pierre.php@gmail.com>;
> Bob Weinand <bobwei9@hotmail.com>; Matt Wilmas
> <php_lists@realplain.com>; PHP internals list =
<internals@lists.php.net>
> Subject: [PHP-DEV] Re: Overflow checks and integral vars comparison
>=20
> Hi Anatol,
>=20
> On Fri, Aug 21, 2015 at 9:41 AM, Anatol Belski <anatol.php@belski.net>
> wrote:
>=20
> > Hi,
> >
> > Resending this as missed internals at the start.
> >
> > I was lately rethinking some part of the 64-bit RFC, and also seeing
> > now Jakub's work on catching overflows in ext/openssl and Matt
> > Williams suggestions on it (which was going a bit more global over
> > it). So I came up with these macros with two goals
> >
> > - standardize the overflow checks
> > - do actualy checks only on architectures where it's needed
> > - simplify the checks where external libs (openssl, libxml, etc.) =
use
> > firm datatypes like int
> >
> > #if SIZEOF_INT =3D=3D SIZEOF_ZEND_LONG
> > # define ZEND_LONG_INT_OVFL(zl) (0)
> > # define ZEND_LONG_INT_UDFL(zl) (0)
> > #else
> > # define ZEND_LONG_INT_OVFL(zlong) ((zlong) > (zend_long)INT_MAX) #
> > define
> > ZEND_LONG_INT_UDFL(zlong) ((zlong) < (zend_long)INT_MIN) #endif
> >
> > #define ZEND_SIZE_T_INT_OVFL(size) ((size) > (size_t)INT_MAX)
> >
> > So having it like
> >
> > If (ZEND_LONG_INT_OVFL(x)) {
> >         return;
> > }
> >
> >
> I think this makes and sense and it's a good start.
>=20
> What I think would be even more useful is to have such checks in ZPP. =
The thing
> is that usual use case for libs where the most used integer type is =
int (e.g.
> openssl) is following:
>=20
> 1. Get zend_long "l" resp. string "s" with size_t len from =
zend_parse_paramters
> 2. check if it doesn't overflow INT_MAX or when needed INT_MIN 3. cast =
it to
> int and pass the value to the lib function
>=20
> If we have such functionality in ZPP, then it would be much simpler. I =
would
> imaging adding "i" for int and something like "q" (maybe different =
letter :) )  for
> string with int size. It would be basically the same as "l" and "s" =
but there would
> be an extra int overflow check which would fail if it's bigger / =
smaller. That
> would be very useful IMHO.
>=20
Yeah, it is probably a good addition. If it is known that some API like =
libxml2 requires integer which is 32-bit on LLP64, LP64 or alike so it =
could overflow - ZPP could check it. It is useful in the case the exact =
value is passed directly to the before mentioned API. If some =
calculations, etc. need to be done before passing to that API, the =
macros can still be useful.

So maybe both a ZPP and macro change were useful and it would make sense =
to go for an RFC first. I've started a draft here =
https://wiki.php.net/rfc/range_checks_for_64_bit .

Regards

Anatol