gmane.comp.php.devel

[PHP-DEV] APC benchmark

Mohammad Saleh — 2012-05-22T15:18:13

All,

I was looking for a standard benchmark script that I could run with and without apc caching to see the general gains.
Is there something that is used by the internals team for such tests?
If not, are there any recommendations?


Thanks,
Mohammad

[PHP-DEV] memory usage ouchy

Rasmus Schultz — 2012-05-21T18:47:51

I just realized something that never occurred to me before - every
property is actually stored as a hash.

This test-script will demonstrate:

  <?php

  define('NUM_TESTS', 1000);

  $before = memory_get_usage(true);

  $test = array();

  class Foo
  {
    public $aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa;
    public $bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb;
    public $cccccccccccccccccccccccccccccccccccccccccccccccccccc;
    public $dddddddddddddddddddddddddddddddddddddddddddddddddddd;
  }

  $bytes = 0;

  for ($i=0; $i<NUM_TESTS; $i++) {
    $foo = new Foo;

    $foo->aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa = 'a';
    $foo->bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb = 'b';
    $foo->cccccccccccccccccccccccccccccccccccccccccccccccccccc = 'c';
    $foo->dddddddddddddddddddddddddddddddddddddddddddddddddddd = 'd';

    $bytes += 4;

    $test[] = $foo;
  }

  $after = memory_get_usage(true);

  header('Content-type: text/plain');

  echo ($after-$before).' byt

[PHP-DEV] zend_execute_internal hook missing from PHP 5

Stefan Esser — 2012-05-21T10:41:24

Hi,

it recently came to my attention that the function whitelist and blacklist feature inside Suhosin is easily bypassable since PHP 5.0.

The reason for this is that PHP is no longer calling the zend_execute_internal() hook if a function is called from another function (via zend_call_function)

In the days of PHP 4 the code looked like this:

if (!zend_execute_internal) {
((zend_internal_function *) EX(function_state).function)->handler(EX(opline)->extended_value, EX(Ts)[EX(opline)->result.u.var].var.ptr, EX(object).ptr, 1 TSRMLS_CC);
} else {
zend_execute_internal(&execute_data, 1 TSRMLS_CC);
}

Nowadays (since PHP 5.0) the code was moved from call_user_function_ex to zend_call_function and just looks like this:

((zend_internal_function *) EX(function_state).function)->handler(fci->param_count, *fci->retval_ptr_ptr, fci->retval_ptr_ptr, fci->object_ptr, 1 TSRMLS_CC);


While this has no immediate impact for average PHP users, it basically kills the possibility for an extension like Suhosin

[PHP-DEV] Old bug 40459: Stat and Dir stream wrapper methods do not call constructor

Jorrit Schippers - nCode — 2012-05-18T10:42:58

Hi all,

I encountered bug #40459 [1] recently. It seems the bug has been ignored
because not many people encounter it and the original author did not
give feedback.

The issue is assigned to Wez. According to the git checkin history, he
doesn't seem to be a contributor anymore. I mailed him to see if he can
recommend someone to handle this issue, no response.

So my question is: can anyone familiar with the Stream Wrapper code take
a look at this issue? My C is not of sufficient level to fix this
myself. My idea would be to abstract creating a stream wrapper instance
and invoking the constructor to a separate function and calling that
function from the various functions that use stream wrappers, as the
amount of code needed to construct an instance and call the constructor
seems quite high to copy everywhere.


Kind regards,

Jorrit Schippers




[1] https://bugs.php.net/bug.php?id=40459

[PHP-DEV] PHP governance question?

lijiu zhang — 2012-05-18T09:07:15

Hi everyone:

I am doing a study report about PHP, could anyone tell PHP governance? Such
as, who is project manager, package manager, how php make a decision?  And
finance support situation?

I cannot find these from php.net, wiki and google. Please give any
information or link is all right.

Thanks,
Lijiu Zhang

[PHP-DEV] IntlCalendar/IntlTimeZone and ext/date

Gustavo Lopes — 2012-05-17T21:34:37

I've made some modifications to ext/intl's IntlCalendar/IntlTimeZone to  
improve interoperability.

* DateTimeZone is now accepted in all places time zones are expected in  
IntlCalendar.
* Added IntlTimeZone::toDateTimeZone() and  
IntlTimeZone::fromDateTimeZone().
* Added IntlCalendar::toDateTime() and IntlCalendar::fromDateTime().

Unfortunately, ext/date doesn't expose a lot of interfaces to manipulate  
its objects, so in some places I ended up having to call PHP_FUNCTIONs (in  
one case even twice doing to a bug; see  
https://github.com/cataphract/php-src/commit/3a81f90ebc51140484c617539b1a5fb0e117f4e2#L1R1264  
).

The conversion from DateTimeZone to IntlTimeZone is made through the ids  
only because ext/date doesn't expose interfaces to efficiently implement  
ICU's TimeZone. This may lead to some different behavior for the same time  
zone id if the databases are out of sync.

The changes are here:  
https://github.com/cataphract/php-src/compare/intl_calendar

If no one objects, I'll merge it soo

[PHP-DEV] PHP 5.3.14RC1 and PHP 5.4.4RC1 Released for Testing

Johannes Schlüter — 2012-05-17T14:16:52

Hi,

The first release candidates for 5.3.14 and PHP 5.4.4 were
released for testing and can be downloaded here:

    PHP 5.3.13RC1:
    http://downloads.php.net/johannes/php-5.3.14RC1.tar.bz2
    http://downloads.php.net/johannes/php-5.3.14RC1.tar.gz

    PHP 5.4.4RC1:
    http://downloads.php.net/stas/php-5.4.4RC1.tar.bz2
    http://downloads.php.net/stas/php-5.4.4RC1.tar.gz

The Windows binaries for both are available at:

    http://windows.php.net/qa/

Please ensure that these releases are solid and all things behave
as expected! Please test these RC against your code base and report any
problems you encounter or successful tests you've run.

Depending on the feedback the final version, or another RC, is scheduled
to be released in two weeks time, on May 31st.

Regards,
  Stas, David and Johannes

[PHP-DEV] having an exception instance

Yader Hernandez — 2012-05-16T03:45:05

Hi,

I'm calling this function:

static zend_class_entry *create_class_entry_runtime_exception(TSRMLS_C) {
    zend_class_entry **pce;

    if (zend_hash_find(CG(class_table), "runtimeexception",
sizeof("RuntimeException"), (void **) &pce) == SUCCESS) {
        return *pce;
    }
}

but it's causing a segfault.

The function is being called in PHP_MINIT_FUNCTION and assigning the return
value to a static zend_class_entry *runtime_exception_ptr

My final goal is to be able to easily call zend_throw_exception and
pass runtime_exception_ptr pointer.

[PHP-DEV] Internals books (c) 2007+ ?

Sanford Whiteman — 2012-05-15T05:00:44

Hi All,

Trying to ready myself for some possible work w/the core (after I
resurrect all my never-that-great C, heh), I went looking for a recent
book. (I still like old-school supplements.)

I see Sara's from 2006 on Amazon, but nothing after that under 'PHP
internals'. I'm sure that one's not totally obsolete, but I don't know
if programming styles and patterns have changed even if the bulk of
the code has not. At the risk of fanning some flames I don't know
about... is there a more recent book that's generally liked?

Thanks,

S.

[PHP-DEV] Persist context across threads?

David Rueter — 2012-05-14T18:47:28

I am interested in preserving the complete PHP context for a thread (globals, 
variables, interpreter, etc.--everything) for later access from a different 
thread.

What would be involved in this?  It seems like:

1) Avoid calling ts_free_thread
2) Call tsrm_set_interpreter_context from new thread

Is it safe to assume that the context of a terminated thread will not be 
destroyed or overwritten as long as ts_free_thread is not called?

Thanks in advance,

David R.

[PHP-DEV] Low hanging fruit / entry point for new developers to the core

Miah Gregory — 2012-05-14T11:27:16

Hi there,

The recent suhosin 'discussions' and subsequent browsing of various
mailing lists prompted me into looking at doing some core
development/bug fixing/whatever, since I have a vested interest as a
user.

I had a look around the various php.net sites but didn't really find
much in the way of guidelines to get involved, code wise, or suggestions
as to low hanging fruit that wouldn't be too painful to start on. If
someone could point me in the right direction it would be much
appreciated.

Would also be of use to know what automated (or otherwise) testing
procedures are available for pre-commit checking?

[PHP-DEV] [RFC] [DISCUSSION] array_part()

Gustavo Lopes — 2012-05-14T10:21:27

Hi

I have a proposal to add a new function to ext/standard: array_part():

https://wiki.php.net/rfc/array_part

Comments would be very welcome, especially the constructive kind.

Please keep this on topic. In particular, avoid proposing new syntax as  
I'm not interested in that. You can of course create your own RFC for that  
purpose.

[PHP-DEV] Array decomposition, more than list()

Clint Priest — 2012-05-13T16:41:22

I've been using array decomposition in other languages such as Javascript recently and find it very useful.  I did not find any conversation about it in the archives, has this topic been discussed/vetted/shot down already?

Example use case: Indexed database results being iterated:

Input:
$tValues[$dbRow['ID1']][$dbRow['ID2']][$dbRow['ID3']] = $dbRow['Value'];

Output:
foreach($tValues as $ID1 => $tSubValues1) {
    foreach($tSubValues1 as $ID2 => $tSubValues2) {
        foreach($tSubValues2 as $ID3 => $Value) {
            /* Do something, such as build an SQL insert */
        }
    }
}

The above is a semi-common piece of code in our production application, I could see list decomposition as a very convenient alternative.

New functionality would support:

foreach($tValues as $ID1 => [$ID2 => [$ID3 => $Value] ]) {
    /* Do something */
}

The above also would indicate that associative array key decomposition would also be allowed.  I don't believe anything like the above is possible, even outside of a fo

[PHP-DEV] bug 54547

Stas Malyshev — 2012-05-13T02:39:33

Hi!

I know this was discussed a number of times here, but just to bring it
to a conclusion - I intend to apply patch in the bug report - which
removes conversion for strings that do not convert to integers - to 5.4.
If anybody sees anything that breaks because of this please tell. Not
sure what about 5.3 - Johannes, could you please comment?
I've run the tests and the patch does not seem to cause any breakage.

[PHP-DEV] the problem of php with gmail

lijiu zhang — 2012-05-11T14:34:08

Hi guys:

I have a problem when I use php send email by gmail. At the beginning, it
is fine, but the errors came after sending some emails.

Warning: fwrite() [function.fwrite]: SSL operation failed with code 1.
OpenSSL Error messages: error:140D00CF:SSL routines:SSL_write:protocol is
shutdown in

Could anyone know this problem?

Regards,
LijiuZhang

[PHP-DEV] Implicit Reference

Mathias Grimm — 2012-05-10T19:38:26

There is a suching extension to handle (intercepts) implicit (or no)
reference.

Ex.:

$a = new A():
$b = $a;

I want something like __clone(), something like __reference() to be invoked
in the second line of the given example.

Thanks.

[PHP-DEV] php interpreter

Xin Tong — 2012-05-09T22:05:15

Hello

I am new to php runtime. i am doing some research on runtime
interpreter. can anyone please tell me where the interpreter of the
php runtime is ? which file ? and does the php runtime has a JIT
compiler ?

Thanks

[PHP-DEV] [RFC] Custom Casting

Seva Lapsha — 2012-05-09T21:56:00

Hello All,

Please be kind to review and comment my proposal for custom casting in PHP
(or let me know if a similar proposal was already discussed).

http://pastebin.com/sPb0i8U6

IMHO, it's more simple and on the other hand more flexible than proposed in:

https://wiki.php.net/rfc/object_cast_to_types
https://wiki.php.net/rfc/class_casting_to_scalar

Thank you in advance,
Seva Lapsha

On Wed, Apr 11, 2012 at 6:14 PM, Yasuo Ohgaki <yohgaki< at >ohgaki.net> wrote:

[PHP-DEV] [rmtools][php-5.4] Build error between rf5936cceeb26d6094f0e142fcaed510ba5549c4f and 0b4d7a4

2012-05-08T18:45:01

This is an automatic mail from the rmtoool's build bots.

New build errors have been introduced between  f5936cceeb26d6094f0e142fcaed510ba5549c4f 
and 0b4d7a4.

The errors are:

php-5.4, build ts-windows-vc9-x86:
php_cli_server.c, 1359, error, C2065, 'prev_pach_len' : undeclared identifier, sapi\cli



Full logs are available at (htmlized versions too):

http://windows.php.net/downloads/snaps/php-5.4/r0b4d7a4

[PHP-DEV] PHP 5.4.3 and PHP 5.3.13 Released!

David Soria Parra — 2012-05-08T18:33:08

Hello!

The PHP Development Team announces the immediate availability of PHP
5.4.3 and PHP 5.3.13. The releases fix security related issues in a
previous attempt to fix a CGI vulnerability (CVE-2012-2311).

PHP 5.4.3 also contains a fix for a buffer overflow vulnernability in
apache_request_headers (CVE-2012-2329).  The PHP 5.3 series is not
affected.

All users are adviced to update to PHP 5.3.13 or PHP 5.4.3!

PHP 5.4.3:
 Release Announcement: http://www.php.net/releases/5_4_3.php
 Downloads:            http://php.net/downloads.php#v5.4.3
 ChangeLog:            http://www.php.net/ChangeLog-5.php#5.4.3

PHP 5.3.13:
 Release Announcement: http://www.php.net/releases/5_3_13.php
 Downloads:            http://php.net/downloads.php#v5.3.13
 ChangeLog:            http://www.php.net/ChangeLog-5.php#5.3.13

regards,
    Johannes, Stas and David

[PHP-DEV] Are feature request bugs ever looked at?

Dmitri Dmitrison — 2012-05-06T14:28:07

Hi all,

First time poster here... Just hoping to get some information.
It says on bugs.php.net that the best way to 'support' a request, is to
vote on it.
However, the most voted feature requests rarely get so much as a reply from
a PHP developer, let alone an implementation.

I'm not being critical, I just wanna understand what is the official PHP
team's stance on this issue?
It seems to me there are two ways to submit requests now days - one via
RFC, and one via the bug system.

On one hand, I've witnessed some good RFCs come to life rather quickly...
On the other, I've seen highly rated feature requests in the bug system lay
unattended for years and decades.

So that just makes me wonder.
Is anyone even checking the bug system for requests anymore?
If the team feels strongly not to implement certain features, could you
just reply to those tickets and close them?
Would it be better to create RFCs instead?

If so, how do we, mere users, go about doing it?
Since it is not a open wiki, is there a person we c