eaccelerator-developers — list to eAccelerator developers, discuss and control the project and code.

[eAccelerator-developers] More one test case. PHP 5.1.2 + cvs200602141812

[Marcus A. G. <ma...@co...>]

More one segmentation fault

Test case:
--
<?php

class ExceptionError extends Exception {
}

?>
--

PHP 5.1.2 + EA cvs200602141812

Regards

-- 
Marcus Alves Grando
marcus(at)corp.grupos.com.br  |  Grupos Internet S/A
   mnag(at)FreeBSD.org         |  FreeBSD.org

RE: [eAccelerator-developers] EA + PHP 5.1 (Status)

[Rodolfo G. G. <ro...@eq...>]

On Tue, 14 Feb 2006, Soenke Ruempler wrote:
> eac...@li... <> wrote on :
>> Which the status of EA + PHP 5.1?
> check this:
> http://snapshots.eaccelerator.net/
> and please test it.


PHP 5.1.3dev (CVS snapshot)
Apache 1.3
eAccelerator (the latest for 5.1 from snapshots)
Prado (CVS snapshot, http://xisc.com)

I get:

      Fatal error: Invalid opcode 99/4/1. in
      /gnt/lib/prado/framework/TApplication.php on line 197

Line 197 is:

      if($this->state!==self::STATE_OFF && $this->state!==self::STATE_ON)

Regards,
Rodolfo.

Re: [eAccelerator-developers] EA + PHP 5.1 (Status)

[Bart V. <bar...@zo...>]

>
> @Bart: Should I write some testing guide? Your testing suite is in my
> revision-que ;-)
>  =20
You're welcome to do so. Extra documentation never hurts.
I've already written big parts of the testing suite. I'll try to put
most of it online this week. But it needs a bit of cleaning up first.

Bart

--=20
Bart Vanbrabant <bar...@zo...>
PGP fingerprint: 093C BB84 17F6 3AA6 6D5E  FC4F 84E1 FED1 E426 64D1

RE: [eAccelerator-developers] EA + PHP 5.1 (Status)

[Soenke R. <rue...@to...>]

eac...@li... <> wrote on :

> Which the status of EA + PHP 5.1?

check this:

http://snapshots.eaccelerator.net/

and please test it.

@Bart: Should I write some testing guide? Your testing suite is in my
revision-que ;-)

[eAccelerator-developers] EA + PHP 5.1 (Status)

[Marcus A. G. <ma...@co...>]

Dev,

Which the status of EA + PHP 5.1?

Regards

-- 
Marcus Alves Grando
marcus(at)corp.grupos.com.br  |  Grupos Internet S/A
   mnag(at)FreeBSD.org         |  FreeBSD.org

[eAccelerator-developers] eAccelerator 0.9.4-rc2

[Bart V. <bar...@zo...>]

I just released 0.9.4-rc2. I hope this will be the final version for
0.9.4 so we can all concentrate on php 5.1 and join hans in his quest
for php 5.1 support!

Release Notes
=3D=3D=3D=3D=3D=3D=3D=3D=3D

Some small problems were fixed since rc1.
This release only contains some bugfixes and code cleanups. This release =
is still the old codebase. It doesn't contain any new code yet.
    - Some fixed were done to the encoder. The format changed a bit so I =
bumped
      up the encoder version. Because the encoder still isn't functional =
for
      php version other than php 4 you should only use this encoder if yo=
u want
      to have line numbers encoded.
      In next versions we will try to fix the encoder for php 5 and 5.1
    - eA should be php 5 compatible except some very unusual cases.
    - This version DOESN'T support php 5.1


gr,

Bart

--=20
Bart Vanbrabant <bar...@zo...>
PGP fingerprint: 093C BB84 17F6 3AA6 6D5E  FC4F 84E1 FED1 E426 64D1

[eAccelerator-developers] eAccelerator test system

[Bart V. <bar...@zo...>]

Hello everyone,

I get a lot of emails and questions on irc from people who want to help
coding on eAccelerator but they only know php. Now we need you're help.
I've been thinking on this system for quite some time but I lacked the
time to implement it. I've written it down in this email and I hope some
people are interested in developing this test system for eAccelerator.

Description
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D

eAccelerator needs a system that enables us to check if the
caching/encoding works for a particular php version. This also allows
us to do regression testing for each release. It would also be a
great help for development.


Requirements
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D

- It should be possible to import the php tests included with the php
  source distribution
- Because we target quite some different php version these tests should
  be versioned in some sort of sets
- These sets should be available online and automaticly downloaded by
  the test script. To save bandwith on the client and on the server
  hosting the tests only updated or new tests in a set should be
  downloaded.
- The test suite should be ran from the commandline. Because
  eAccelerator doesn't work on the commandline the tests should be
  executed on the webserver.
- Caching AND encoding should be tested.
- Because we don't need to test if the language functions correctly we
  don't need to define the required output for a test to pass. This is
  a serious advantage over the php tests because it requires a lot less
  time to write tests.
- The test program should run on all platform that php support and
  require no extra depencies so it's pretty obvious this program is
  written in php. This scripts must run on php 4 to 5.1
- It should only use pear libraries that are distributed with the php
  source.
- Generation of a report of the tests. This report should include
  detailed information about the php version, eAccelerator version.
  Also provide information about the environment (like os, build
  system, ...)
- Test with and without optimizer
- Create sets of tests. A set can contain tests and other sets. A set
  should contain information about the php and the eA versions it's
  intended to run on. An example:
    A set with OO-tests
    A set for PHP51 test which also contains the OO test set
- Being able to run a specified test, this is especcialy useful for
  development.


Nice features
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D

- Some sort of speed information to see what the speed increase is.


Information
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D

To know if a test is successful we need to know what the output is
without eAccelerator enabled. This should be the output when a script
is ran for the first time. At this moment the script is cached so the
next request is cached one. By comparing the output we know if a test
was successful.
We need to check with and without the optimizer. This means we need
to remove the script from cache. This can be done by touching the
scripts so eAccelerator will reload it.


If you are interested in working on this please let us know, this would
be a great help. We really need this system to release php 5.1 support.


gr,

Bart

--=20
Bart Vanbrabant <bar...@zo...>
PGP fingerprint: 093C BB84 17F6 3AA6 6D5E  FC4F 84E1 FED1 E426 64D1

RE: [eAccelerator-developers] Error using __autoload and the last CVS version (20060206)

[Yvan T. <yva...@m4...>]

So far so good :)

tHANkS a lot !

I keep up running tests, phpmyadmin is running fine for the moment.

Best regards,
Yvan.

> -----Message d'origine-----
> De=A0: eac...@li...
> [mailto:eac...@li...] De la =
part de
> Hans Rakers
> Envoy=E9=A0: mardi 7 f=E9vrier 2006 17:33
> =C0=A0: eac...@li...
> Objet=A0: Re: [eAccelerator-developers] Error using __autoload and the =
last
> CVS version (20060206)
>=20
>=20
> This should be fixed in the latest snapshot
>=20
> See http://snapshots.eaccelerator.net/
>=20
> Regards,
>=20
> Hans
>=20
>=20
> Yvan Taviaud wrote:
>=20
> >Hi,
> >
> >
> >
> >I've just installed the latest CVS Version,
> >eaccelerator-5.1-cvs200602061812, but I'm facing a problem with
> __autoload()
> >that never occurred before.
> >
> >
> >
> >Here's a example script :
> >
> >
> >
> >
> >
> >Test.php :
> >
> ><?php
> >
> >function __autoload($class) {
> >
> >            echo 'The class <b>'.$class.'</b> can\'t be loaded<br =
/>';
> >
> >}
> >
> >
> >
> >include 'Test.class.php';
> >
> >
> >
> >$test =3D new Test();
> >
> >?>
> >
> >
> >
> >And the Test.class.php :
> >
> ><?php
> >
> >class Test {
> >
> >            function __construct() {
> >
> >                        echo 'Object created successfully';
> >
> >            }
> >
> >}
> >
> >?>
> >
> >
> >
> >Expected output :
> >
> >Object created successfully
> >
> >
> >
> >Current output :
> >
> >Object created successfully (for the first time the file is executed)
> >
> >
> >
> >The class Test can't be loaded
> >
> >Object created successfully (once the file is EA-compiled)
> >
> >
> >
> >
> >
> >The strange point : the class exists, is loaded, as the string =
"Object
> >created successfully" is echoed.
> >
> >
> >
> >Can you do something for this ?
> >
> >
> >
> >Best regards,
> >
> >Yvan.
> >
> >
> >
> >
>=20
>=20
>=20
> -------------------------------------------------------
> This SF.net email is sponsored by: Splunk Inc. Do you grep through log
> files
> for problems?  Stop!  Download the new AJAX search engine that makes
> searching your log files as easy as surfing the  web.  DOWNLOAD =
SPLUNK!
> =
http://sel.as-us.falkag.net/sel?cmd=3Dlnk&kid=3D103432&bid=3D230486&dat=3D=
121642
> _______________________________________________
> eAccelerator-developers mailing list
> eAc...@li...
> https://lists.sourceforge.net/lists/listinfo/eaccelerator-developers

Re: [eAccelerator-developers] Error using __autoload and the last CVS version (20060206)

[Hans R. <ha...@pa...>]

This should be fixed in the latest snapshot

See http://snapshots.eaccelerator.net/

Regards,

Hans


Yvan Taviaud wrote:

>Hi,
>
> 
>
>I've just installed the latest CVS Version,
>eaccelerator-5.1-cvs200602061812, but I'm facing a problem with __autoload()
>that never occurred before.
>
> 
>
>Here's a example script :
>
> 
>
> 
>
>Test.php :
>
><?php
>
>function __autoload($class) {
>
>            echo 'The class <b>'.$class.'</b> can\'t be loaded<br />';
>
>}
>
> 
>
>include 'Test.class.php';
>
> 
>
>$test = new Test();
>
>?>
>
> 
>
>And the Test.class.php :
>
><?php
>
>class Test {
>
>            function __construct() {
>
>                        echo 'Object created successfully';
>
>            }
>
>}
>
>?>
>
> 
>
>Expected output :
>
>Object created successfully
>
> 
>
>Current output :
>
>Object created successfully (for the first time the file is executed)
>
> 
>
>The class Test can't be loaded
>
>Object created successfully (once the file is EA-compiled)
>
> 
>
> 
>
>The strange point : the class exists, is loaded, as the string "Object
>created successfully" is echoed.
>
> 
>
>Can you do something for this ?
>
> 
>
>Best regards,
>
>Yvan.
>
>
>  
>

[eAccelerator-developers] Error using __autoload and the last CVS version (20060206)

[Yvan T. <yva...@m4...>]

Hi,

 

I've just installed the latest CVS Version,
eaccelerator-5.1-cvs200602061812, but I'm facing a problem with __autoload()
that never occurred before.

 

Here's a example script :

 

 

Test.php :

<?php

function __autoload($class) {

            echo 'The class <b>'.$class.'</b> can\'t be loaded<br />';

}

 

include 'Test.class.php';

 

$test = new Test();

?>

 

And the Test.class.php :

<?php

class Test {

            function __construct() {

                        echo 'Object created successfully';

            }

}

?>

 

Expected output :

Object created successfully

 

Current output :

Object created successfully (for the first time the file is executed)

 

The class Test can't be loaded

Object created successfully (once the file is EA-compiled)

 

 

The strange point : the class exists, is loaded, as the string "Object
created successfully" is echoed.

 

Can you do something for this ?

 

Best regards,

Yvan.

[eAccelerator-developers] Testing eAccelerator from CLI

[Soenke R. <rue...@to...>]

Hi,

is there any possibilty to test eAccelerator through CLI? That would be nice
so we could write some test scripts for common crash cases and run them
automatically (and maybe integrate them later into some `make test' later).

Thx.

--

Soenke Ruempler

RE: [eAccelerator-developers] PHP 5.1.2 testing

[Soenke R. <rue...@to...>]

Hi,

eac...@li... <> wrote on :

> A new php-5.1 compatibility development snapshot is available at
> http://snapshots.eaccelerator.net/eaccelerator-5.1-cvs20060206
> 1812.tar.gz 

After a short first test on our heavy loaded php5-oop system I received this
segfault:

(gdb) file /usr/sbin/httpd
Reading symbols from /usr/sbin/httpd...done.
Using host libthread_db library "/lib/libthread_db.so.1".
(gdb) set args -X
(gdb) run
Starting program: /usr/sbin/httpd -X
Reading symbols from shared object read from target memory...done.
Loaded system supplied DSO at 0xffffe000
[Thread debugging using libthread_db enabled]
[New Thread -1213073728 (LWP 6490)]

Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread -1213073728 (LWP 6490)]
0xb70a934c in restore_op_array (to=0x92e7474, from=0xb091c29c) at
/usr/src/eaccelerator-5.1-cvs200602061812/ea_restore.c:579
579                             ea_debug_printf(EA_DEBUG, "[%d]
class_entry->parent='%s' [%08x]\n",
(gdb) bt
#0  0xb70a934c in restore_op_array (to=0x92e7474, from=0xb091c29c) at
/usr/src/eaccelerator-5.1-cvs200602061812/ea_restore.c:579
#1  0xb70a9665 in restore_op_array_ptr (from=0xb091c29c) at
/usr/src/eaccelerator-5.1-cvs200602061812/ea_restore.c:685
#2  0xb70a8ccd in restore_hash (target=0x8fa3008, source=0xb091be00,
copy_bucket=0xb70a9648 <restore_op_array_ptr>)
    at /usr/src/eaccelerator-5.1-cvs200602061812/ea_restore.c:360
#3  0xb70a9c65 in restore_class_entry (to=0x8fa2fec, from=0xb091bdec) at
/usr/src/eaccelerator-5.1-cvs200602061812/ea_restore.c:939
#4  0xb70a9dbe in restore_class (p=0xb0910130) at
/usr/src/eaccelerator-5.1-cvs200602061812/ea_restore.c:999
#5  0xb70a0714 in eaccelerator_compile_file (file_handle=0xbfd72d34, type=8)
at /usr/src/eaccelerator-5.1-cvs200602061812/eaccelerator.c:954
#6  0xb79689ee in ZEND_INCLUDE_OR_EVAL_SPEC_CONST_HANDLER
(execute_data=0xbfd72e30) at /usr/src/php-5.1.2/Zend/zend_vm_execute.h:1871
#7  0xb7963f71 in execute (op_array=0x858fd9c) at
/usr/src/php-5.1.2/Zend/zend_vm_execute.h:92
#8  0xb7968775 in ZEND_INCLUDE_OR_EVAL_SPEC_CONST_HANDLER
(execute_data=0xbfd72fd0) at /usr/src/php-5.1.2/Zend/zend_vm_execute.h:1918
#9  0xb7963f71 in execute (op_array=0x92bd0b4) at
/usr/src/php-5.1.2/Zend/zend_vm_execute.h:92
#10 0xb794e2d7 in zend_execute_scripts (type=8, retval=Variable "retval" is
not available.
) at /usr/src/php-5.1.2/Zend/zend.c:1101
#11 0xb791aa32 in php_execute_script (primary_file=0xbfd75314) at
/usr/src/php-5.1.2/main/main.c:1720
#12 0xb79c652f in php_handler (r=0x9335e00) at
/usr/src/php-5.1.2/sapi/apache2handler/sapi_apache2.c:584
#13 0x080676ce in ap_run_handler (r=0x9335e00) at config.c:152
#14 0x08067a46 in ap_invoke_handler (r=0x9335e00) at config.c:364
#15 0x08064e21 in ap_internal_redirect (
    new_uri=0xb70af3e0 "[%d]", ' ' <repeats 19 times>, "[internal function
from=%08x,to=%08x] class_entry='%s' [%08x]\n", r=0x195a) at
http_request.c:465
#16 0xb7ae08ae in handler_redirect (r=Variable "r" is not available.
) at mod_rewrite.c:1735
#17 0x080676ce in ap_run_handler (r=0x836f378) at config.c:152
#18 0x08067a46 in ap_invoke_handler (r=0x836f378) at config.c:364
#19 0x08065199 in ap_process_request (r=0x836f378) at http_request.c:249
#20 0x08060bd9 in ap_process_http_connection (c=0x835fe18) at
http_core.c:251
#21 0x0806feda in ap_run_process_connection (c=0x835fe18) at connection.c:43
#22 0x08065e6e in child_main (child_num_arg=Variable "child_num_arg" is not
available.
) at prefork.c:610
#23 0x08066021 in make_child (s=0x809ce58, slot=0) at prefork.c:650
#24 0x080660ea in startup_children (number_to_start=5) at prefork.c:722
#25 0x080668ac in ap_mpm_run (_pconf=0x809b0a8, plog=0x80c5150, s=0x809ce58)
at prefork.c:941
#26 0x0806bb7f in main (argc=2, argv=0xbfd757e4) at main.c:618
(gdb) quit
The program is running.  Exit anyway? (y or n) y

Config is:

;zend_extension="/usr/lib/php/extensions/no-debug-non-zts-20050922/eaccelera
tor.so"
eaccelerator.shm_size="90"
eaccelerator.cache_dir="/tmp/eaccelerator"
eaccelerator.enable="1"
eaccelerator.optimizer="0"
eaccelerator.check_mtime="1"
eaccelerator.debug="0"
eaccelerator.filter=""
eaccelerator.shm_max="90"
eaccelerator.shm_ttl="0"
eaccelerator.shm_prune_period="0"
eaccelerator.shm_only="0"
eaccelerator.compress="1"
eaccelerator.compress_level="0"

./configure:

 --without-eaccelerator-content-caching --without-eaccelerator-sessions
--without-eaccelerator-encoder   --without-eaccelerator-optimizer

Re: [eAccelerator-developers] PHP 5.1.2 testing

[Matt <dat...@gm...>]

I have this snapshot running under FreeBSD 6, Apache 2.2 and PHP 5.1.2
and the initial testing looks promising.  Certain conditions in
MediaWiki, for example, that would always segfault before appear
stable now.

I will continue to test..

Thank you for all the hard work.

Matt

On 2/6/06, Ramon van Alteren <ra...@hy...> wrote:
> Hans,
>
> Hans Rakers wrote:
>
> > Ramon van Alteren wrote:
> >
> >>
> >> Update: I pushed the test-server update in straight away.
> >>
> >> In a quick test-run we're seeing immediate apache segfaulting with
> >> signal 11 after the new eaccelerator install.
> >> What can I do to provide you with debug info ?
> >
> > Did you empty your cache dir before restarting Apache? This is really
> > important!
>
> Nope, will do so now.
> And try again.....
>
> It seems to run now..... but I lack the time to fully test. I'm off
> home...... about time ;-)
>
> I'll get Boris (our main developer on this tomorrow and we'll report
> more thoroughly.
> If there's any way we can get meaningfull debug-output for you, don't
> hesitate to let us know.
>
> Grtz Ramon
>
> --
> mail:    ra...@hy...
> gsm:     +31 (0) 6 48261300
> msn:     rja...@ho...
> www:     http://ramon71.hyves.nl
>
>
>
> -------------------------------------------------------
> This SF.net email is sponsored by: Splunk Inc. Do you grep through log fi=
les
> for problems?  Stop!  Download the new AJAX search engine that makes
> searching your log files as easy as surfing the  web.  DOWNLOAD SPLUNK!
> http://sel.as-us.falkag.net/sel?cmd=3Dlnk&kid=3D103432&bid=3D230486&dat=
=3D121642
> _______________________________________________
> eAccelerator-developers mailing list
> eAc...@li...
> https://lists.sourceforge.net/lists/listinfo/eaccelerator-developers
>

Re: [eAccelerator-developers] PHP 5.1.2 testing

[Ramon v. A. <ra...@hy...>]

Hans,

Hans Rakers wrote:

> Ramon van Alteren wrote:
>
>>
>> Update: I pushed the test-server update in straight away.
>>
>> In a quick test-run we're seeing immediate apache segfaulting with 
>> signal 11 after the new eaccelerator install.
>> What can I do to provide you with debug info ?
>
> Did you empty your cache dir before restarting Apache? This is really 
> important!

Nope, will do so now.
And try again.....

It seems to run now..... but I lack the time to fully test. I'm off 
home...... about time ;-)

I'll get Boris (our main developer on this tomorrow and we'll report 
more thoroughly.
If there's any way we can get meaningfull debug-output for you, don't 
hesitate to let us know.

Grtz Ramon

-- 
mail:    ra...@hy...
gsm:     +31 (0) 6 48261300
msn:     rja...@ho...
www:     http://ramon71.hyves.nl

Re: [eAccelerator-developers] PHP 5.1.2 testing

[Hans R. <ha...@pa...>]

Ramon van Alteren wrote:

>
> Update: I pushed the test-server update in straight away.
>
> In a quick test-run we're seeing immediate apache segfaulting with 
> signal 11 after the new eaccelerator install.
> What can I do to provide you with debug info ?
>
> Grtz Ramon
>

Did you empty your cache dir before restarting Apache? This is really 
important!

Hans

Re: [eAccelerator-developers] PHP 5.1.2 testing

[Ramon v. A. <ra...@hy...>]

Ramon van Alteren wrote:

> Hans Rakers wrote:
>
>>
>> A new php-5.1 compatibility development snapshot is available at 
>> http://snapshots.eaccelerator.net/eaccelerator-5.1-cvs200602061812.tar.gz 
>>
>
>
> Great.
>
>> Please give it a spin. I'd love some feedback on the current state. 
>> Do note that this is a work in progress, so its not recommended for 
>> production use.
>
>
> I'll try and get it up and running tomorrow and will post results to 
> the list.
> Thanx again for the work :-)

Update: I pushed the test-server update in straight away.

In a quick test-run we're seeing immediate apache segfaulting with 
signal 11 after the new eaccelerator install.
What can I do to provide you with debug info ?

Grtz Ramon

-- 
mail:    ra...@hy...
gsm:     +31 (0) 6 48261300
msn:     rja...@ho...
www:     http://ramon71.hyves.nl

Re: [eAccelerator-developers] PHP 5.1.2 testing

[Ramon v. A. <ra...@hy...>]

Hans Rakers wrote:

>
> A new php-5.1 compatibility development snapshot is available at 
> http://snapshots.eaccelerator.net/eaccelerator-5.1-cvs200602061812.tar.gz

Great.

> Please give it a spin. I'd love some feedback on the current state. Do 
> note that this is a work in progress, so its not recommended for 
> production use.

I'll try and get it up and running tomorrow and will post results to the 
list.
Thanx again for the work :-)

Grtz Ramon

-- 
mail:    ra...@hy...
gsm:     +31 (0) 6 48261300
msn:     rja...@ho...
www:     http://ramon71.hyves.nl

Re: [eAccelerator-developers] PHP 5.1.2 testing

[Hans R. <ha...@pa...>]

A new php-5.1 compatibility development snapshot is available at 
http://snapshots.eaccelerator.net/eaccelerator-5.1-cvs200602061812.tar.gz

Please give it a spin. I'd love some feedback on the current state. Do 
note that this is a work in progress, so its not recommended for 
production use.

BTW: Don't bother testing the encoder/loader on php-5.1 since that 
hasn't been worked on yet.

Greetings,

Hans Rakers


Hans Rakers wrote:

>
> I will be hopefully be doing a major commit later today, which should 
> result in much better compatibility. Please wait for the next snapshot :)
>
> Greetings,
>
> Hans Rakers
>
>
> Soenke Ruempler wrote:
>
>> Hi,
>>
>> I just tried the latest snapshot from 
>> http://snapshots.eaccelerator.net/ on
>> a new development box.
>>
>> Unfortunately it segfaults in the shutdown process:
>>
>> Starting program: /usr/bin/php -v
>> Reading symbols from shared object read from target memory...done.
>> Loaded system supplied DSO at 0xd67000
>> [Thread debugging using libthread_db enabled]
>> [New Thread -1208288800 (LWP 15499)]
>> PHP 5.1.2 (cli) (built: Jan 27 2006 16:25:00)
>> Copyright (c) 1997-2006 The PHP Group
>> Zend Engine v2.1.0, Copyright (c) 1998-2006 Zend Technologies
>>    with eAccelerator v0.9.4-cvs200601121552-rc1, Copyright (c) 2004-2005
>> eAccelerator, by eAccelerator
>>
>> Program received signal SIGSEGV, Segmentation fault.
>> [Switching to Thread -1208288800 (LWP 15499)]
>> 0x0828de91 in shutdown_memory_manager (silent=0, full_shutdown=0) at
>> /usr/src/php-5.1.2/Zend/zend_alloc.c:511
>> 511                                     REMOVE_POINTER_FROM_LIST(ptr);
>> (gdb) bt
>> #0  0x0828de91 in shutdown_memory_manager (silent=0, full_shutdown=0) at
>> /usr/src/php-5.1.2/Zend/zend_alloc.c:511
>> #1  0x0826d4b9 in php_request_shutdown (dummy=0x0) at
>> /usr/src/php-5.1.2/main/main.c:1303
>> #2  0x08319335 in main (argc=2, argv=0xbfcc8214) at
>> /usr/src/php-5.1.2/sapi/cli/php_cli.c:1230
>>
>> Any hints?
>>
>>
>>
>> -------------------------------------------------------
>> This SF.net email is sponsored by: Splunk Inc. Do you grep through 
>> log files
>> for problems?  Stop!  Download the new AJAX search engine that makes
>> searching your log files as easy as surfing the  web.  DOWNLOAD SPLUNK!
>> http://sel.as-us.falkag.net/sel?cmd=lnk&kid=103432&bid=230486&dat=121642
>> _______________________________________________
>> eAccelerator-developers mailing list
>> eAc...@li...
>> https://lists.sourceforge.net/lists/listinfo/eaccelerator-developers
>>  
>>
>
>
>
> -------------------------------------------------------
> This SF.net email is sponsored by: Splunk Inc. Do you grep through log 
> files
> for problems?  Stop!  Download the new AJAX search engine that makes
> searching your log files as easy as surfing the  web.  DOWNLOAD SPLUNK!
> http://sel.as-us.falkag.net/sel?cmd=lnk&kid=103432&bid=230486&dat=121642
> _______________________________________________
> eAccelerator-developers mailing list
> eAc...@li...
> https://lists.sourceforge.net/lists/listinfo/eaccelerator-developers

Re: [eAccelerator-developers] PHP 5.1.2 testing

[Hans R. <ha...@pa...>]

I will be hopefully be doing a major commit later today, which should 
result in much better compatibility. Please wait for the next snapshot :)

Greetings,

Hans Rakers


Soenke Ruempler wrote:

>Hi,
>
>I just tried the latest snapshot from http://snapshots.eaccelerator.net/ on
>a new development box.
>
>Unfortunately it segfaults in the shutdown process:
>
>Starting program: /usr/bin/php -v
>Reading symbols from shared object read from target memory...done.
>Loaded system supplied DSO at 0xd67000
>[Thread debugging using libthread_db enabled]
>[New Thread -1208288800 (LWP 15499)]
>PHP 5.1.2 (cli) (built: Jan 27 2006 16:25:00)
>Copyright (c) 1997-2006 The PHP Group
>Zend Engine v2.1.0, Copyright (c) 1998-2006 Zend Technologies
>    with eAccelerator v0.9.4-cvs200601121552-rc1, Copyright (c) 2004-2005
>eAccelerator, by eAccelerator
>
>Program received signal SIGSEGV, Segmentation fault.
>[Switching to Thread -1208288800 (LWP 15499)]
>0x0828de91 in shutdown_memory_manager (silent=0, full_shutdown=0) at
>/usr/src/php-5.1.2/Zend/zend_alloc.c:511
>511                                     REMOVE_POINTER_FROM_LIST(ptr);
>(gdb) bt
>#0  0x0828de91 in shutdown_memory_manager (silent=0, full_shutdown=0) at
>/usr/src/php-5.1.2/Zend/zend_alloc.c:511
>#1  0x0826d4b9 in php_request_shutdown (dummy=0x0) at
>/usr/src/php-5.1.2/main/main.c:1303
>#2  0x08319335 in main (argc=2, argv=0xbfcc8214) at
>/usr/src/php-5.1.2/sapi/cli/php_cli.c:1230
>
>Any hints?
>
>
>
>-------------------------------------------------------
>This SF.net email is sponsored by: Splunk Inc. Do you grep through log files
>for problems?  Stop!  Download the new AJAX search engine that makes
>searching your log files as easy as surfing the  web.  DOWNLOAD SPLUNK!
>http://sel.as-us.falkag.net/sel?cmd=lnk&kid=103432&bid=230486&dat=121642
>_______________________________________________
>eAccelerator-developers mailing list
>eAc...@li...
>https://lists.sourceforge.net/lists/listinfo/eaccelerator-developers
>  
>

[eAccelerator-developers] PHP 5.1.2 testing

[Soenke R. <rue...@to...>]

Hi,

I just tried the latest snapshot from http://snapshots.eaccelerator.net/ on
a new development box.

Unfortunately it segfaults in the shutdown process:

Starting program: /usr/bin/php -v
Reading symbols from shared object read from target memory...done.
Loaded system supplied DSO at 0xd67000
[Thread debugging using libthread_db enabled]
[New Thread -1208288800 (LWP 15499)]
PHP 5.1.2 (cli) (built: Jan 27 2006 16:25:00)
Copyright (c) 1997-2006 The PHP Group
Zend Engine v2.1.0, Copyright (c) 1998-2006 Zend Technologies
    with eAccelerator v0.9.4-cvs200601121552-rc1, Copyright (c) 2004-2005
eAccelerator, by eAccelerator

Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread -1208288800 (LWP 15499)]
0x0828de91 in shutdown_memory_manager (silent=0, full_shutdown=0) at
/usr/src/php-5.1.2/Zend/zend_alloc.c:511
511                                     REMOVE_POINTER_FROM_LIST(ptr);
(gdb) bt
#0  0x0828de91 in shutdown_memory_manager (silent=0, full_shutdown=0) at
/usr/src/php-5.1.2/Zend/zend_alloc.c:511
#1  0x0826d4b9 in php_request_shutdown (dummy=0x0) at
/usr/src/php-5.1.2/main/main.c:1303
#2  0x08319335 in main (argc=2, argv=0xbfcc8214) at
/usr/src/php-5.1.2/sapi/cli/php_cli.c:1230

Any hints?

RE: [eAccelerator-developers] Eaccelerator doesn't work with PHP 5

[Soenke R. <rue...@to...>]

eac...@li... <> wrote on :

> But ther are a lot of messages like these :
> Jan 31 13:54:05 master kernel: pid 69654 (httpd), uid 80: exited on
> signal 11 Jan 31 13:54:22 master kernel: pid 69669 (httpd), uid 80:
> exited on signal 11 Jan 31 13:54:23 master kernel: pid 69641 (httpd),
> uid 80: exited on signal 11 Jan 31 13:54:40 master kernel: pid 69677
> (httpd), uid 80: exited on signal 11 Jan 31 13:54:42 master kernel:
> pid 69636 (httpd), uid 80: exited on signal 11 Jan 31 13:54:42 master
> kernel: pid 69667 (httpd), uid 80: exited on signal 11 
> 
> I think it's old problem - I've saw the same more then year ago even
> with PHP 4. 
> 
> Is it possible to fix ?

The version for 5.1 is still unstable and has some bugs mainly with
oop-scripts. Please read the list archives for more information.

Try to make a minimal reproduce script and then let apache run in debug mode
with gdb:

http://httpd.apache.org/dev/debugging.html#gdb

Don't forget the backtrace :)

Thx.

Re: [eAccelerator-developers] Eaccelerator doesn't work with PHP 5

[Krok <kr...@vo...>]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Thanks !
It's works !

But ther are a lot of messages like these :
Jan 31 13:54:05 master kernel: pid 69654 (httpd), uid 80: exited on
signal 11
Jan 31 13:54:22 master kernel: pid 69669 (httpd), uid 80: exited on
signal 11
Jan 31 13:54:23 master kernel: pid 69641 (httpd), uid 80: exited on
signal 11
Jan 31 13:54:40 master kernel: pid 69677 (httpd), uid 80: exited on
signal 11
Jan 31 13:54:42 master kernel: pid 69636 (httpd), uid 80: exited on
signal 11
Jan 31 13:54:42 master kernel: pid 69667 (httpd), uid 80: exited on
signal 11

I think it's old problem - I've saw the same more then year ago even
with PHP 4.

Is it possible to fix ?

Soenke Ruempler wrote:
| Hi,
|
| eac...@li... <> wrote on :
|
|
|>I've tried to install EAccelerator 0.9.2a, 0.9.3 and 0.9.4-rc1 on PHP
|>5.1.1 and PHP 5.1.2 (OS: FreeBSD 5.4-RELEASE-p6, Apache: 1.3.33).
|
|
| Try this snapshot:
|
| http://snapshots.eaccelerator.net/eaccelerator-5.1-cvs200601161947.tar.gz
|
| And please read the mailing list archives about the 5.1 compatiblity
state.
|
|
|
| -------------------------------------------------------
| This SF.net email is sponsored by: Splunk Inc. Do you grep through log
files
| for problems?  Stop!  Download the new AJAX search engine that makes
| searching your log files as easy as surfing the  web.  DOWNLOAD SPLUNK!
| http://sel.as-us.falkag.net/sel?cmd=lnk&kid=103432&bid=230486&dat=121642
| _______________________________________________
| eAccelerator-developers mailing list
| eAc...@li...
| https://lists.sourceforge.net/lists/listinfo/eaccelerator-developers

- --
With best regards,
Krok.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.0 (MingW32)
Comment: GnuPT 2.5.8.1 by EQUIPMENTE.DE
Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org

iD8DBQFD30NesXuomovtlAARAvbjAJwJ4Hpcmh4SXJqw6mXAEWy4OmJu8gCeLS+v
w06FFsmjTu6ZCkUA2g1DfZE=
=gTpn
-----END PGP SIGNATURE-----

RE: [eAccelerator-developers] Eaccelerator doesn't work with PHP 5

[Soenke R. <rue...@to...>]

Hi,

eac...@li... <> wrote on :

> I've tried to install EAccelerator 0.9.2a, 0.9.3 and 0.9.4-rc1 on PHP
> 5.1.1 and PHP 5.1.2 (OS: FreeBSD 5.4-RELEASE-p6, Apache: 1.3.33).

Try this snapshot:

http://snapshots.eaccelerator.net/eaccelerator-5.1-cvs200601161947.tar.gz

And please read the mailing list archives about the 5.1 compatiblity state.

[eAccelerator-developers] Eaccelerator doesn't work with PHP 5

[Krok <kr...@vo...>]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hello.

I've tried to install EAccelerator 0.9.2a, 0.9.3 and 0.9.4-rc1 on PHP
5.1.1 and PHP 5.1.2 (OS: FreeBSD 5.4-RELEASE-p6, Apache: 1.3.33).

1. Unable to build EAccelerator due to following changes in PHP 5.x:
Zend/ChangeLog:
...
line 756:
2005-06-16  Dmitry Stogov  <dm...@ze...>

~    * zend_execute.c
~      zend_vm_def.h
~      zend_vm_execute.h
~      zend_vm_opcodes.h:
~      Compilation warnings

~    * zend_compile.c
~      zend_compile.h
~      zend_vm_def.h
~      zend_vm_execute.h
~      zend_vm_opcodes.h:
- ->>>  ZEND_UNSET_DIM_OBJ is splitted to ZEND_UNSET_DIM and ZEND_UNSET_OBJ.
...
line 896:
~    * (php_5_1_0b1)
~      zend_compile.c
~      zend_vm_def.h
~      zend_vm_execute.h
~      zend_vm_opcodes.h:
- ->>>  Merge three opcodes (ZEND_NEW, ZEND_JMP_NO_CTOR, ZEND_INIT_CTOR)
into one (ZEND_NEW). There was no real reason for this anymore and API
should be changed before 5.1

2. As I can see in PHP ChangeLog garbage collector was removed from
public and we must use function zend_clean_garbage(TSRMLS_D) but not
cycle such as

~      while (EG(garbage_ptr)) {
~        zval_ptr_dtor(&EG(garbage)[--EG(garbage_ptr)]);
~      }

3. I'm not some C/C++ guru, so after I've fixed this bugs with my pure
skill, I've build EAccelerator but Apache crashe with signal 10 just
after startup.


Can you fix it correctly ?

Thanks.

- --
With best regards,
Krok.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.0 (MingW32)
Comment: GnuPT 2.5.8.1 by EQUIPMENTE.DE
Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org

iD8DBQFD3xqTsXuomovtlAARAqFxAKCFoveOLs8YZvahHk6QoODMTV7KiQCeOS1o
pI3QzYD2zcoKFSO/WJZmwo4=
=MoT/
-----END PGP SIGNATURE-----

[eAccelerator-developers] SegFault using eA 5.1 CVS

[Yvan T. <yva...@m4...>]

Sorry if I didn=92t follow the previous mail, I don=92t have it anymore. =
So this
show be considered as an answer for php5.1 + eA-5.1-cvs200601121553
(Segmentation fault) =A0

So here=92s a backtrace for the phpmyadmin segmentation fault :
EA : eaccelerator-5.1-cvs200601161947
PHP : 5.1.1
Apache : 2.0.55
Linux Debian Sarge 3.1, gcc 3.3

(gdb) bt
#0  0x405c21c4 in _zval_copy_ctor_func (zvalue=3D0xbfffd250,
__zend_filename=3D0x40801760 =
"/usr/local/src/php-5.1.1/Zend/zend_variables.h",
__zend_lineno=3D45)
    at /usr/local/src/php-5.1.1/Zend/zend_variables.c:119
#1  0x4063b9af in _zval_copy_ctor (zvalue=3D0xbfffd250,
__zend_filename=3D0x407fc680 =
"/usr/local/src/php-5.1.1/Zend/zend_execute.c",
__zend_lineno=3D806)
    at zend_variables.h:45
#2  0x4063da9d in zend_assign_to_variable (result=3D0x40c17f6c,
op1=3D0x40c17f80, op2=3D0x40c17f94, value=3D0x40c17f98, type=3D1, =
Ts=3D0xbfffd2e4)
    at /usr/local/src/php-5.1.1/Zend/zend_execute.c:806
#3  0x4062ba71 in ZEND_ASSIGN_SPEC_CV_CONST_HANDLER
(execute_data=3D0xbfffd3b0) at zend_vm_execute.h:20701
#4  0x405f14e9 in execute (op_array=3D0x829401c) at zend_vm_execute.h:88
#5  0x40bec24c in profile_execute (op_array=3D0x829401c) at
/usr/local/src/eaccelerator-5.1-cvs200601161947/eaccelerator.c:1573
#6  0x405c428d in zend_execute_scripts (type=3D8, retval=3D0x0, =
file_count=3D3) at
/usr/local/src/php-5.1.1/Zend/zend.c:1090
#7  0x40579b0f in php_execute_script (primary_file=3D0xbffff7a0) at
/usr/local/src/php-5.1.1/main/main.c:1704
#8  0x4063f45c in php_handler (r=3D0x8241988) at
/usr/local/src/php-5.1.1/sapi/apache2handler/sapi_apache2.c:584
#9  0x080921f5 in ap_run_handler (r=3D0x8241988) at config.c:152
#10 0x08092800 in ap_invoke_handler (r=3D0x8241988) at config.c:364
#11 0x0807734a in ap_process_request (r=3D0x8241988) at =
http_request.c:249
#12 0x0807262d in ap_process_http_connection (c=3D0x823b910) at
http_core.c:251
#13 0x0809d4e5 in ap_run_process_connection (c=3D0x823b910) at =
connection.c:43
#14 0x080907f4 in child_main (child_num_arg=3D137295940) at =
prefork.c:610
#15 0x08090947 in make_child (s=3D0xbfffd250, slot=3D4) at prefork.c:704
#16 0x08090a68 in startup_children (number_to_start=3D16) at =
prefork.c:722
#17 0x080912da in ap_mpm_run (_pconf=3D0x80d6038, plog=3D0x8112128, =
s=3D0x80dab48)
at prefork.c:941
#18 0x0809798d in main (argc=3D3, argv=3D0xbffffb64) at main.c:618

I must stress out that when I've compiled in debug mode (--enable-debug =
in
PHP and --with-eaccelerator-debug in EA) the EA web page (eaccelerator()
function in a PHP file) started to act really weird (Zend errors =
mentioning
"ZZZZZ" constants non terminated). Almost every page was dead since =
then.

OK, here's a BT in non-debug mode, same system :

#0  0x4016af3e in __strtol_internal () from /lib/libc.so.6
#1  0x40525fdc in compare_function (result=3D0xbfffb81c, =
op1=3D0x83626b4,
op2=3D0x0) at stdlib.h:317
#2  0x405265b2 in is_equal_function (result=3D0xbfffb81c, =
op1=3D0x400f70cc,
op2=3D0x400f70cc) at /usr/local/src/php-5.1.1/Zend/zend_operators.c:1474
#3  0x4057f258 in ZEND_IS_EQUAL_SPEC_CV_CONST_HANDLER
(execute_data=3D0xbfffc040) at
/usr/local/src/php-5.1.1/Zend/zend_execute.c:200
#4  0x4054c438 in execute (op_array=3D0x835da98) at zend_vm_execute.h:88
#5  0x4054c901 in zend_do_fcall_common_helper_SPEC =
(execute_data=3D0xbfffc410)
at zend_vm_execute.h:222
#6  0x4054c438 in execute (op_array=3D0x835f100) at zend_vm_execute.h:88
#7  0x4054c901 in zend_do_fcall_common_helper_SPEC =
(execute_data=3D0xbfffd460)
at zend_vm_execute.h:222
#8  0x4054c438 in execute (op_array=3D0x829f6c4) at zend_vm_execute.h:88
#9  0x405292e0 in zend_execute_scripts (type=3D8, retval=3D0x0, =
file_count=3D3) at
/usr/local/src/php-5.1.1/Zend/zend.c:1090
#10 0x404f06ed in php_execute_script (primary_file=3D0xbffff7b0) at
/usr/local/src/php-5.1.1/main/main.c:1704
#11 0x40597dcf in php_handler (r=3D0x8241048) at
/usr/local/src/php-5.1.1/sapi/apache2handler/sapi_apache2.c:584
#12 0x080921f5 in ap_run_handler (r=3D0x8241048) at config.c:152
#13 0x08092800 in ap_invoke_handler (r=3D0x8241048) at config.c:364
#14 0x0807734a in ap_process_request (r=3D0x8241048) at =
http_request.c:249
#15 0x0807262d in ap_process_http_connection (c=3D0x823af80) at
http_core.c:251
#16 0x0809d4e5 in ap_run_process_connection (c=3D0x823af80) at =
connection.c:43
#17 0x080907f4 in child_main (child_num_arg=3D1074753740) at =
prefork.c:610
#18 0x08090947 in make_child (s=3D0x40aac8a4, slot=3D17) at =
prefork.c:704
#19 0x08090a68 in startup_children (number_to_start=3D3) at =
prefork.c:722
#20 0x080912da in ap_mpm_run (_pconf=3D0x80d6038, plog=3D0x8112128, =
s=3D0x80dab48)
at prefork.c:941
#21 0x0809798d in main (argc=3D3, argv=3D0xbffffb44) at main.c:618

HTH,
Yvan.