[PHP-BUG] Bug #52230 [NEW]: namespace & __autoload clash
From:
Operating system: gentoo & ubuntu
PHP version: 5.3.2
Package: Class/Object related
Bug Type: Bug
Bug description:namespace & __autoload clash
Description:
when you use __autoload($var) it will get the class needed depending on the
file.
by using namespace and getting the class by a full path name
ex
new \class\name
__autload will then try to load classname
Test script:
---
/** index.php **/
function __autoload($var) {
require_once($var.'.php');
}
//It will try to load the file Coreinit.php and not Core/init.php;
new Core\init
/** Core/init.php **/
namespace Core;
class init{
function __construct(){
echo 'Bingo..!';
}
}
Expected result:
that autoload loads Core/init.php
Actual result:
--
autoload loads Coreinit.php
--
Edit bug report at http://bugs.php.net/bug.php?id=52230&edit=1
--
Try a snapshot (PHP 5.2):
http://bugs.php.net/fix.php?id=52230&r=trysnapshot52
Try a snapshot (PHP 5.3):
http://bugs.php.net/fix.php?id=52230&r=trysnapshot53
Try a snapshot (trunk):
http://bugs.php.net/fix.php?id=52230&r=trysnapshottrunk
Fixed in SVN:
http://bugs.php.net/fix.php?id=52230&r=fixed
Fixed in SVN and need be documented:
http://bugs.php.net/fix.php?id=52230&r=needdocs
Fixed in release:
http://bugs.php.net/fix.php?id=52230&r=alreadyfixed
Need backtrace:
http://bugs.php.net/fix.php?id=52230&r=needtrace
Need Reproduce Script:
http://bugs.php.net/fix.php?id=52230&r=needscript
Try newer version:
http://bugs.php.net/fix.php?id=52230&r=oldversion
Not developer issue:
http://bugs.php.net/fix.php?id=52230&r=support
Expected behavior:
http://bugs.php.net/fix.php?id=52230&r=notwrong
Not enough info:
http://bugs.php.net/fix.php?id=52230&r=notenoughinfo
Submitted twice:
http://bugs.php.net/fix.php?id=52230&r=submittedtwice
register_globals:
http://bugs.php.net/fix.php?id=52230&r=globals
PHP 4 support discontinued: http://bugs.php.net/fix.php?id=52230&r=php4
Daylight Savings:http://bugs.php.net/fix.php?id=52230&r=dst
IIS Stability:
http://bugs.php.net/fix.php?id=52230&r=isapi
Install GNU Sed:
http://bugs.php.net/fix.php?id=52230&r=gnused
Floating point limitations:
http://bugs.php.net/fix.php?id=52230&r=float
No Zend Extensions:
http://bugs.php.net/fix.php?id=52230&r=nozend
MySQL Configuration Error:
http://bugs.php.net/fix.php?id=52230&r=mysqlcfg
Bug #50400 [Com]: Compile fails generating phar.phar
Edit report at http://bugs.php.net/bug.php?id=50400&edit=1 ID: 50400 Comment by: omars1234 at gmail dot com Reported by: lepage at grm dot polymtl dot ca Summary: Compile fails generating phar.phar Status: Assigned Type: Bug Package: PHAR related Operating System: Solaris 10 PHP Version: 5.3.1 Assigned To: dsp New Comment: As a work-around, pass "--disable-phar" to configure. I don't do anything that explicitly needs phar archives (so far), so I did this to build PHP5.3.2 on Solaris10 after running into the same problem. Previous Comments: [2010-01-26 17:09:36] ekcheu at uncg dot edu I've had this issue before. It appears that this error occurs depending upon which version of gcc you are using. After continuing to fail to compile on a version of gcc (4.2.1).. I used blastwave's gcc, and it was able to get past this issue. Don't ask why it compiles find on some versions of gcc and not others. [2009-12-07 20:13:01] lepage at grm dot polymtl dot ca Description: Cannot gcc compile php 5.3.1 on Solaris 10 (sparc) since it's failing with phar/phar.php error. It was the same with php 5.3.0. SunStudio is not an option since many libs are done with gcc and there is incompatibilities with some dependencies. php 515 does compile on Solaris 10. thanks. Reproduce code: --- php-5.3.1% make Generating phar.phar Parse error: syntax error, unexpected $end in /share/concorde/xta3511/install/web/php-5.3.1/ext/phar/phar.php on line 19 *** Error code 255 The following command caused the error: ` if test -x "/share/concorde/xta3511/install/web/php-5.3.1/sapi/cli/php"; then /share/concorde/xta3511/install/web/php-5.3.1/build/shtool echo -n -- "/share/concorde/xta3511/install/web/php-5.3.1/sapi/cli/php -n"; if test "x" != "x"; then /share/concorde/xta3511/install/web/php-5.3.1/build/shtool echo -n -- " -d extension_dir=/share/concorde/xta3511/install/web/php-5.3.1/modules"; for i in bz2 zlib phar; do if test -f "/share/concorde/xta3511/install/web/php-5.3.1/modules/$i.la"; then . /share/concorde/xta3511/install/web/php-5.3.1/modules/$i.la; /share/concorde/xta3511/install/web/php-5.3.1/build/shtool echo -n -- " -d extension=$dlname"; fi; done; fi; else /share/concorde/xta3511/install/web/php-5.3.1/build/shtool echo -n -- ""; fi;` -d 'open_basedir=' -d 'output_buffering=0' -d 'memory_limit=-1' -d phar.readonly=0 -d 'safe_mode=0' ext/phar/phar.php pack -f ext/phar/phar.phar -a pharcommand -c auto -x \\.svn -p 0 -s /share/concorde/xta3511/install/web/php-5.3.1/ext/phar/phar/phar.php -h sha1 -b "`if test -x ""; then /share/concorde/xta3511/install/web/php-5.3.1/build/shtool echo -n -- ""; else /share/concorde/xta3511/install/web/php-5.3.1/build/shtool echo -n -- "/usr/local_10/opt/php-531/bin/php"; fi; `" /share/concorde/xta3511/install/web/php-5.3.1/ext/phar/phar/ make: Fatal error: Command failed for target `ext/phar/phar.phar' Expected result: to compile. Actual result: -- failed compiling. -- Edit this bug report at http://bugs.php.net/bug.php?id=50400&edit=1
Req #7028 [Com]: xml=shared and wddx do not work together
Edit report at http://bugs.php.net/bug.php?id=7028&edit=1
ID: 7028
Comment by: gmblar+php at gmail dot com
Reported by: ipv4 at mail dot ru
Summary: xml=shared and wddx do not work together
Status: Analyzed
Type: Feature/Change Request
Package: Feature/Change Request
Operating System: Linux
PHP Version: 4.0.3RC2
New Comment:
$ php -r "wddx_deserialize('foo');"
dyld: lazy symbol binding failed: Symbol not found: _XML_ParserCreate
Referenced from:
/usr/lib/php/extensions/no-debug-non-zts-20090626/wddx.so
Expected in: flat namespace
dyld: Symbol not found: _XML_ParserCreate
Referenced from:
/usr/lib/php/extensions/no-debug-non-zts-20090626/wddx.so
Expected in: flat namespace
Trace/BPT trap
Previous Comments:
[2000-11-29 08:39:07] s...@php.net
I guess yes, this is by design. If you build WDDX, session
module uses it, so you have to build it statically. If you
have WDDX as static, you need XML to be static too, since
WDDX uses XML.
So, while it'd be nice to have such a possibility, currently
it doesn't work by design.
Thus I reclassify it as feature request.
[2000-11-29 04:41:24] sni...@php.net
reclassified
[2000-10-05 01:46:43] ipv4 at mail dot ru
configure ... --with-xml=shared --enable-wddx
or
--with-xml=shared --enable-wddx=shared
doesn't work. wddx support is always compiled in statically
and resulting module (or standalone php) becomes unusable as it contains
unresolved symbols from xml module.
Currently the only way to build php with xml and wddx is to build xml
support statically. :( Is that by design?
--
Edit this bug report at http://bugs.php.net/bug.php?id=7028&edit=1
Bug #52228 [Opn->Bgs]: fseek failure to seek
Edit report at http://bugs.php.net/bug.php?id=52228&edit=1 ID: 52228 Updated by: paj...@php.net Reported by: VJTD3 at VJTD3 dot com Summary: fseek failure to seek -Status: Open +Status: Bogus Type: Bug Package: Filesystem function related Operating System: * PHP Version: Irrelevant New Comment: We already have feature requests for that. Previous Comments: [2010-07-01 22:00:17] VJTD3 at VJTD3 dot com Description: 2147483647 or PHP_INT_MAX on the system is the max integer. fseek fails with files over PHP_INT_MAX+2 in size. binary math would resolve the fseek issue internally. Test script: --- reading a file of 20 gigabytes (or anything larger then PHP_INT_MAX+2) is successful with: however fseek fails: Expected result: echo the byte from position PHP_INT_MAX+1 Actual result: -- crash/fail -- Edit this bug report at http://bugs.php.net/bug.php?id=52228&edit=1
[PHP-BUG] Bug #52228 [NEW]: fseek failure to seek
From: Operating system: * PHP version: Irrelevant Package: Filesystem function related Bug Type: Bug Bug description:fseek failure to seek Description: 2147483647 or PHP_INT_MAX on the system is the max integer. fseek fails with files over PHP_INT_MAX+2 in size. binary math would resolve the fseek issue internally. Test script: --- reading a file of 20 gigabytes (or anything larger then PHP_INT_MAX+2) is successful with: however fseek fails: Expected result: echo the byte from position PHP_INT_MAX+1 Actual result: -- crash/fail -- Edit bug report at http://bugs.php.net/bug.php?id=52228&edit=1 -- Try a snapshot (PHP 5.2): http://bugs.php.net/fix.php?id=52228&r=trysnapshot52 Try a snapshot (PHP 5.3): http://bugs.php.net/fix.php?id=52228&r=trysnapshot53 Try a snapshot (trunk): http://bugs.php.net/fix.php?id=52228&r=trysnapshottrunk Fixed in SVN: http://bugs.php.net/fix.php?id=52228&r=fixed Fixed in SVN and need be documented: http://bugs.php.net/fix.php?id=52228&r=needdocs Fixed in release: http://bugs.php.net/fix.php?id=52228&r=alreadyfixed Need backtrace: http://bugs.php.net/fix.php?id=52228&r=needtrace Need Reproduce Script: http://bugs.php.net/fix.php?id=52228&r=needscript Try newer version: http://bugs.php.net/fix.php?id=52228&r=oldversion Not developer issue: http://bugs.php.net/fix.php?id=52228&r=support Expected behavior: http://bugs.php.net/fix.php?id=52228&r=notwrong Not enough info: http://bugs.php.net/fix.php?id=52228&r=notenoughinfo Submitted twice: http://bugs.php.net/fix.php?id=52228&r=submittedtwice register_globals: http://bugs.php.net/fix.php?id=52228&r=globals PHP 4 support discontinued: http://bugs.php.net/fix.php?id=52228&r=php4 Daylight Savings:http://bugs.php.net/fix.php?id=52228&r=dst IIS Stability: http://bugs.php.net/fix.php?id=52228&r=isapi Install GNU Sed: http://bugs.php.net/fix.php?id=52228&r=gnused Floating point limitations: http://bugs.php.net/fix.php?id=52228&r=float No Zend Extensions: http://bugs.php.net/fix.php?id=52228&r=nozend MySQL Configuration Error: http://bugs.php.net/fix.php?id=52228&r=mysqlcfg
[PHP-BUG] Bug #52227 [NEW]: parse_ini_* does not parse namespaced constants
From: Operating system: Linux, Mac, Windows PHP version: 5.3.2 Package: Filesystem function related Bug Type: Bug Bug description:parse_ini_* does not parse namespaced constants Description: The parse_ini_file/parse_ini_string pair of functions do not expand namespaced constants. Test script: --- 'parse_ini_file_namespaced', 'project.constant.namspaced' => 'my\\PROJECT_NAME', ) Actual result: -- array ( 'project.constant.global' => 'PROJECT_NAME', 'project.constant.namspaced' => 'my\\PROJECT_NAME', ) -- Edit bug report at http://bugs.php.net/bug.php?id=52227&edit=1 -- Try a snapshot (PHP 5.2): http://bugs.php.net/fix.php?id=52227&r=trysnapshot52 Try a snapshot (PHP 5.3): http://bugs.php.net/fix.php?id=52227&r=trysnapshot53 Try a snapshot (trunk): http://bugs.php.net/fix.php?id=52227&r=trysnapshottrunk Fixed in SVN: http://bugs.php.net/fix.php?id=52227&r=fixed Fixed in SVN and need be documented: http://bugs.php.net/fix.php?id=52227&r=needdocs Fixed in release: http://bugs.php.net/fix.php?id=52227&r=alreadyfixed Need backtrace: http://bugs.php.net/fix.php?id=52227&r=needtrace Need Reproduce Script: http://bugs.php.net/fix.php?id=52227&r=needscript Try newer version: http://bugs.php.net/fix.php?id=52227&r=oldversion Not developer issue: http://bugs.php.net/fix.php?id=52227&r=support Expected behavior: http://bugs.php.net/fix.php?id=52227&r=notwrong Not enough info: http://bugs.php.net/fix.php?id=52227&r=notenoughinfo Submitted twice: http://bugs.php.net/fix.php?id=52227&r=submittedtwice register_globals: http://bugs.php.net/fix.php?id=52227&r=globals PHP 4 support discontinued: http://bugs.php.net/fix.php?id=52227&r=php4 Daylight Savings:http://bugs.php.net/fix.php?id=52227&r=dst IIS Stability: http://bugs.php.net/fix.php?id=52227&r=isapi Install GNU Sed: http://bugs.php.net/fix.php?id=52227&r=gnused Floating point limitations: http://bugs.php.net/fix.php?id=52227&r=float No Zend Extensions: http://bugs.php.net/fix.php?id=52227&r=nozend MySQL Configuration Error: http://bugs.php.net/fix.php?id=52227&r=mysqlcfg
[PHP-BUG] Req #52225 [NEW]: __get() and __set() for static classes
From:
Operating system: Any
PHP version: 5.3.2
Package: Class/Object related
Bug Type: Feature/Change Request
Bug description:__get() and __set() for static classes
Description:
Would it be possible to have magic setter/getter functions available on
static
context ?
The documentation says "Property overloading only works in object context.
These
magic methods will not be triggered in static context."
I would find it useful on a static context where I have a class A extending
another class B, to access
Test script:
---
Example:
Class A {
static $data = array('bar' => 99, 'foo' => 101);
function __get($name) {
if (isset(self::$data[$name]))
return self::$data[$name];
return null;
}
}
Class B extends A {
}
echo B::$foo;
Expected result:
echo B::$foo;
In the example should return 101.
Actual result:
--
In the example it triggers a fatal error.
--
Edit bug report at http://bugs.php.net/bug.php?id=52225&edit=1
--
Try a snapshot (PHP 5.2):
http://bugs.php.net/fix.php?id=52225&r=trysnapshot52
Try a snapshot (PHP 5.3):
http://bugs.php.net/fix.php?id=52225&r=trysnapshot53
Try a snapshot (trunk):
http://bugs.php.net/fix.php?id=52225&r=trysnapshottrunk
Fixed in SVN:
http://bugs.php.net/fix.php?id=52225&r=fixed
Fixed in SVN and need be documented:
http://bugs.php.net/fix.php?id=52225&r=needdocs
Fixed in release:
http://bugs.php.net/fix.php?id=52225&r=alreadyfixed
Need backtrace:
http://bugs.php.net/fix.php?id=52225&r=needtrace
Need Reproduce Script:
http://bugs.php.net/fix.php?id=52225&r=needscript
Try newer version:
http://bugs.php.net/fix.php?id=52225&r=oldversion
Not developer issue:
http://bugs.php.net/fix.php?id=52225&r=support
Expected behavior:
http://bugs.php.net/fix.php?id=52225&r=notwrong
Not enough info:
http://bugs.php.net/fix.php?id=52225&r=notenoughinfo
Submitted twice:
http://bugs.php.net/fix.php?id=52225&r=submittedtwice
register_globals:
http://bugs.php.net/fix.php?id=52225&r=globals
PHP 4 support discontinued: http://bugs.php.net/fix.php?id=52225&r=php4
Daylight Savings:http://bugs.php.net/fix.php?id=52225&r=dst
IIS Stability:
http://bugs.php.net/fix.php?id=52225&r=isapi
Install GNU Sed:
http://bugs.php.net/fix.php?id=52225&r=gnused
Floating point limitations:
http://bugs.php.net/fix.php?id=52225&r=float
No Zend Extensions:
http://bugs.php.net/fix.php?id=52225&r=nozend
MySQL Configuration Error:
http://bugs.php.net/fix.php?id=52225&r=mysqlcfg
Bug #52037 [Asn]: Concurrent builds fail in install-programs
Edit report at http://bugs.php.net/bug.php?id=52037&edit=1 ID: 52037 User updated by: seanius at debian dot org Reported by: seanius at debian dot org Summary: Concurrent builds fail in install-programs Status: Assigned Type: Bug Package: Compile Failure Operating System: Debian GNU/Linux PHP Version: 5.3.2 Assigned To: kalle New Comment: hi, yes, i believe so. at least from a quick visual review it looks like it has the same problem. Previous Comments: [2010-06-21 07:52:31] ka...@php.net Does this apply to 5.2 aswell? [2010-06-09 20:58:28] seanius at debian dot org Description: originally reported as http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=584348 which was assumed to be a problem in the build environment of the bug reporter, but after his persistence i investigated further and found that there is indeed a problem with parallel builds. namely, the install-programs target assumes that INSTALL_ROOT/bindir exists and doesn't check/create it iself. if executed serially after the install-build target, the directory does exist since it is created by that target, but if called in parallel (or before calling install-build), the error mentioned in the above debian bug will be seen: > make[1]: Entering directory `/build/user-php5_5.3.2-1-amd64-JiIO8n/php5-5.3.2/apache2-build' > Installing build environment: /build/user-php5_5.3.2-1-amd64-JiIO8n/php5-5.3.2/debian/libapache2-mod-php5/usr/lib/php5/build/ > Installing helper programs: /build/user-php5_5.3.2-1-amd64-JiIO8n/php5-5.3.2/debian/libapache2-mod-php5/usr/bin/ > program: phpize > cp: cannot create regular file `/build/user-php5_5.3.2-1-amd64-JiIO8n/php5-5.3.2/debian/libapache2-mod-php5/usr/bin/#i...@31393#': No such file or directory > program: php-config (the strange filename is a shtool temp file, but the error is that the directory does not exist) the patch for this is very simple, attached. -- Edit this bug report at http://bugs.php.net/bug.php?id=52037&edit=1
Bug #52223 [Opn->Bgs]: switch case shows unexpected behaviour in particular case
Edit report at http://bugs.php.net/bug.php?id=52223&edit=1 ID: 52223 Updated by: degeb...@php.net Reported by: webmaster at cephario dot com Summary: switch case shows unexpected behaviour in particular case -Status: Open +Status: Bogus Type: Bug Package: Unknown/Other Function Operating System: Win32Server2003SE PHP Version: 5.3.2 New Comment: Sorry, but your problem does not imply a bug in PHP itself. For a list of more appropriate places to ask for help using PHP, please visit http://www.php.net/support.php as this bug system is not the appropriate forum for asking support questions. Due to the volume of reports we can not explain in detail here why your report is not a bug. The support channels will be able to provide an explanation for you. Thank you for your interest in PHP. 0 == 'N', so the case passes. Previous Comments: [2010-07-01 16:01:54] webmaster at cephario dot com Description: if the given value is a numeric 0 (AND YES APPLIES ONLY FOR THE 0 negative or positive int values are working as expected) the first string expression (case) is matched. Test script: --- Expected result: DATANEW Actual result: -- NOW() -- Edit this bug report at http://bugs.php.net/bug.php?id=52223&edit=1
Bug #52218 [Bgs]: php module not loaded when php module is enabled and then restarted
Edit report at http://bugs.php.net/bug.php?id=52218&edit=1
ID: 52218
User updated by: php at spam dot wizbit dot be
Reported by: php at spam dot wizbit dot be
Summary: php module not loaded when php module is enabled and
then restarted
Status: Bogus
Type: Bug
Package: Apache2 related
Operating System: Linux
-PHP Version: Irrelevant
+PHP Version: 5.3.3-RC1
New Comment:
Since you insisted: Retested with 5.3.3-RC1. (even tho I checked the
code in trunk and said it was the same.)
No surprise: the same problem exist.
First start with php disabled:
Thu Jul 01 18:02:07 2010] [notice] Digest: generating secret for digest
authentication ...
[Thu Jul 01 18:02:07 2010] [notice] Digest: done
[Thu Jul 01 18:02:07 2010] [notice] Apache/2.2.15 (Unix) DAV/2
configured -- resuming normal operations
First restart with php enabled:
[Thu Jul 01 18:02:13 2010] [notice] SIGHUP received. Attempting to
restart
[Thu Jul 01 18:02:13 2010] [notice] Digest: generating secret for digest
authentication ...
[Thu Jul 01 18:02:13 2010] [notice] Digest: done
[Thu Jul 01 18:02:13 2010] [notice] Apache/2.2.15 (Unix) DAV/2
configured -- resuming normal operations
=> PHP not loaded but it is enabled in the code
Second restart with PHP enabled:
[Thu Jul 01 18:02:15 2010] [notice] SIGHUP received. Attempting to
restart
[Thu Jul 01 18:02:16 2010] [notice] Digest: generating secret for digest
authentication ...
[Thu Jul 01 18:02:16 2010] [notice] Digest: done
[Thu Jul 01 18:02:16 2010] [notice] Apache/2.2.15 (Unix) DAV/2
PHP/5.3.3-RC1 configured -- resuming normal operations
=> PHP now loaded.
Note that this never was a problem debian builds. I said I experienced
this first with a stock php-4.4.9 and that I then tested with debian
since that was easier then compiling a stock 5.3.3 build.
Previous Comments:
[2010-07-01 16:03:35] johan...@php.net
Both 4.4.9 and 5.2.6 are old. If there is an issue with the Debian
builds please talk to them.
[2010-07-01 12:10:27] php at spam dot wizbit dot be
Description:
When apache is restarted after enabling php then it does not load PHP.
When apache is restarted again then it does load PHP.
I first noticed this with (stock) php-4.4.9 and apache 2.2.15 where it
resulted in segmentation faults on every request.
I then tested this with php-5.2.6 that comes with debian
(PHP/5.2.6-1+lenny8 with Suhosin-Patch) and it shows the same problem
but does not segfault.
Steps to reproduce (on debian):
1. install apache2 and php
2. stop apache: apache2ctl -k stop
3. disable the php module: rm /etc/apache2/mods-enabled/php.*
4. start apache: apache2ctl -k start
5. enable the php module: ln -s /etc/apache2/mods-available/php.*
/etc/apache2/mods-enabled/
6. restart apache: apache2ctl -k restart
7. Check the log file => PHP not loaded. (on php-4.4.9 every request
from this moment on caused a segmentation fault)
8. restart apache again: apache2ctl -k restart
9. Check the log file => PHP is loaded
NOTE: do not use the debian init script (/etc/init.d/apache2).
I do not know what the severity of this issue should be.
It does not appear to be segfaulting in recent versions but that does
not nessesarly mean that the segfault is fixed (it could be hiding
somewhere - I did not try very hard to get it to segfault).
The logfile/traces/a guess where the problem is:
An example log file:
PHP disabled: apache2ctl -k start:
[Thu Jul 01 10:55:03 2010] [notice] Apache/2.2.9 (Debian) configured
-- resuming normal operations
PHP enabled; apache2ctl -k restart:
[Thu Jul 01 10:55:10 2010] [notice] SIGHUP received. Attempting to
restart
[Thu Jul 01 10:55:10 2010] [notice] Apache/2.2.9 (Debian) configured
-- resuming normal operations
==> PHP not loaded
Another restart: apache2ctl -k restart:
[Thu Jul 01 10:55:15 2010] [notice] SIGHUP received. Attempting to
restart
[Thu Jul 01 10:55:15 2010] [notice] Apache/2.2.9 (Debian)
PHP/5.2.6-1+lenny8 with Suhosin-Patch configured -- resuming normal
operations
==> PHP now loaded
debugging this with php 4.4.9 may points to a problem in
sapi/apache2handler/sapi_apache2.c: php_apache_server_startup.
(I did not debug this with php 5 but I did check the code in trunk and
it looks the same).
The start of the code:
static int
php_apache_server_startup(apr_pool_t *pconf, apr_pool_t *plog,
apr_pool_t *ptemp, server_rec *s)
{
void *data = NULL;
const char *userdata_key = "apache2hook_post_config";
/* Apache will load, unload and then reload a DSO module. This
* prevents us from starting PHP until the second load. */
The comment makes me
Bug #52218 [Opn->Bgs]: php module not loaded when php module is enabled and then restarted
Edit report at http://bugs.php.net/bug.php?id=52218&edit=1
ID: 52218
Updated by: johan...@php.net
Reported by: php at spam dot wizbit dot be
Summary: php module not loaded when php module is enabled and
then restarted
-Status: Open
+Status: Bogus
Type: Bug
Package: Apache2 related
Operating System: Linux
PHP Version: Irrelevant
New Comment:
Both 4.4.9 and 5.2.6 are old. If there is an issue with the Debian
builds please talk to them.
Previous Comments:
[2010-07-01 12:10:27] php at spam dot wizbit dot be
Description:
When apache is restarted after enabling php then it does not load PHP.
When apache is restarted again then it does load PHP.
I first noticed this with (stock) php-4.4.9 and apache 2.2.15 where it
resulted in segmentation faults on every request.
I then tested this with php-5.2.6 that comes with debian
(PHP/5.2.6-1+lenny8 with Suhosin-Patch) and it shows the same problem
but does not segfault.
Steps to reproduce (on debian):
1. install apache2 and php
2. stop apache: apache2ctl -k stop
3. disable the php module: rm /etc/apache2/mods-enabled/php.*
4. start apache: apache2ctl -k start
5. enable the php module: ln -s /etc/apache2/mods-available/php.*
/etc/apache2/mods-enabled/
6. restart apache: apache2ctl -k restart
7. Check the log file => PHP not loaded. (on php-4.4.9 every request
from this moment on caused a segmentation fault)
8. restart apache again: apache2ctl -k restart
9. Check the log file => PHP is loaded
NOTE: do not use the debian init script (/etc/init.d/apache2).
I do not know what the severity of this issue should be.
It does not appear to be segfaulting in recent versions but that does
not nessesarly mean that the segfault is fixed (it could be hiding
somewhere - I did not try very hard to get it to segfault).
The logfile/traces/a guess where the problem is:
An example log file:
PHP disabled: apache2ctl -k start:
[Thu Jul 01 10:55:03 2010] [notice] Apache/2.2.9 (Debian) configured
-- resuming normal operations
PHP enabled; apache2ctl -k restart:
[Thu Jul 01 10:55:10 2010] [notice] SIGHUP received. Attempting to
restart
[Thu Jul 01 10:55:10 2010] [notice] Apache/2.2.9 (Debian) configured
-- resuming normal operations
==> PHP not loaded
Another restart: apache2ctl -k restart:
[Thu Jul 01 10:55:15 2010] [notice] SIGHUP received. Attempting to
restart
[Thu Jul 01 10:55:15 2010] [notice] Apache/2.2.9 (Debian)
PHP/5.2.6-1+lenny8 with Suhosin-Patch configured -- resuming normal
operations
==> PHP now loaded
debugging this with php 4.4.9 may points to a problem in
sapi/apache2handler/sapi_apache2.c: php_apache_server_startup.
(I did not debug this with php 5 but I did check the code in trunk and
it looks the same).
The start of the code:
static int
php_apache_server_startup(apr_pool_t *pconf, apr_pool_t *plog,
apr_pool_t *ptemp, server_rec *s)
{
void *data = NULL;
const char *userdata_key = "apache2hook_post_config";
/* Apache will load, unload and then reload a DSO module. This
* prevents us from starting PHP until the second load. */
The comment makes me believe that it is the expectation that
php_apache_server_startup is called twice but that does not appear to be
the case.
gdb session:
Breakpoint 1 (php_apache_server_startup) pending.
(gdb) c
Continuing.
Program received signal SIGHUP, Hangup.
0xb7c2cfb8 in select () from /ub/lib/libc.so.6
(gdb) c
Continuing.
Program received signal SIGHUP, Hangup.
0xb7b9df51 in kill () from /ub/lib/libc.so.6
# NOTE: this is the first restart after enabling the module
(gdb) c
Continuing.
Breakpoint 2 at 0xb78de20f: file
/root/work/php-4.4.9/sapi/apache2handler/sapi_apache2.c, line 367.
Pending breakpoint "php_apache_server_startup" resolved
Breakpoint 2, php_apache_server_startup (pconf=0x80e80a8,
plog=0x81301c8, ptemp=0x80f20d0, s=0x80ed0c0) at
/root/work/php-4.4.9/sapi/apache2handler/sapi_apache2.c:367
367 void *data = NULL;
(gdb) n
368 const char *userdata_key = "apache2hook_post_config";
(gdb) n
372 apr_pool_userdata_get(&data, userdata_key, s->process->pool);
(gdb) n
373 if (data == NULL) {
(gdb) p data
$1 = (void *) 0x0
(gdb) c
Continuing.
# NOTE: PHP is not loaded in apache even tho it is enabled
Program received signal SIGHUP, Hangup.
0xb7c2cfb8 in select () from /ub/lib/libc.so.6
(gdb) c
Continuing.
Program received signal SIGHUP, Hangup.
0xb7b9df51 in kill () from /ub/lib/libc.so.6
(gdb) c
Continuing.
# NOTE: this is the second restart after enabling the module
warning: Temporarily disabling breakpoints for unloaded shared library
"/ub/pkg/httpd/module
[PHP-BUG] Bug #52223 [NEW]: switch case shows unexpected behaviour in particular case
From: Operating system: Win32Server2003SE PHP version: 5.3.2 Package: Unknown/Other Function Bug Type: Bug Bug description:switch case shows unexpected behaviour in particular case Description: if the given value is a numeric 0 (AND YES APPLIES ONLY FOR THE 0 negative or positive int values are working as expected) the first string expression (case) is matched. Test script: --- Expected result: DATANEW Actual result: -- NOW() -- Edit bug report at http://bugs.php.net/bug.php?id=52223&edit=1 -- Try a snapshot (PHP 5.2): http://bugs.php.net/fix.php?id=52223&r=trysnapshot52 Try a snapshot (PHP 5.3): http://bugs.php.net/fix.php?id=52223&r=trysnapshot53 Try a snapshot (trunk): http://bugs.php.net/fix.php?id=52223&r=trysnapshottrunk Fixed in SVN: http://bugs.php.net/fix.php?id=52223&r=fixed Fixed in SVN and need be documented: http://bugs.php.net/fix.php?id=52223&r=needdocs Fixed in release: http://bugs.php.net/fix.php?id=52223&r=alreadyfixed Need backtrace: http://bugs.php.net/fix.php?id=52223&r=needtrace Need Reproduce Script: http://bugs.php.net/fix.php?id=52223&r=needscript Try newer version: http://bugs.php.net/fix.php?id=52223&r=oldversion Not developer issue: http://bugs.php.net/fix.php?id=52223&r=support Expected behavior: http://bugs.php.net/fix.php?id=52223&r=notwrong Not enough info: http://bugs.php.net/fix.php?id=52223&r=notenoughinfo Submitted twice: http://bugs.php.net/fix.php?id=52223&r=submittedtwice register_globals: http://bugs.php.net/fix.php?id=52223&r=globals PHP 4 support discontinued: http://bugs.php.net/fix.php?id=52223&r=php4 Daylight Savings:http://bugs.php.net/fix.php?id=52223&r=dst IIS Stability: http://bugs.php.net/fix.php?id=52223&r=isapi Install GNU Sed: http://bugs.php.net/fix.php?id=52223&r=gnused Floating point limitations: http://bugs.php.net/fix.php?id=52223&r=float No Zend Extensions: http://bugs.php.net/fix.php?id=52223&r=nozend MySQL Configuration Error: http://bugs.php.net/fix.php?id=52223&r=mysqlcfg
Req #52215 [Opn]: Add var_ref_count()
Edit report at http://bugs.php.net/bug.php?id=52215&edit=1
ID: 52215
Updated by: der...@php.net
Reported by: dev at alepe dot com
Summary: Add var_ref_count()
Status: Open
Type: Feature/Change Request
Package: Variables related
Operating System: Any
PHP Version: 5.3.3RC1
New Comment:
This won't work, for the same reason that debug_zval_dump() is flawed.
By passing a variable to a function, you mess with the refcount/is_ref
values. Xdebug's xdebug_debug_zval()
(http://xdebug.org/docs/all_functions#xdebug_debug_zval) goes around
that by looking up the symbol directly (but it doesn't support array
elements directly).
Previous Comments:
[2010-07-01 03:49:32] dtajchre...@php.net
http://www.php.net/manual/en/function.debug-zval-dump.php
[2010-07-01 03:34:42] dev at alepe dot com
Please excuse me if I have some theoretical misconceptions or if my
English is not very good.
[2010-07-01 03:33:03] dev at alepe dot com
Description:
It seems there is no easy way to know if an object/array/... is
reference.
Looking at the source code of ext/standard/var.c it seems it may be not
so hard to add that function. As debug_zval_dump already outputs the
reference count, it would be better to obtain that value in order to
determine if an object is referenced or not. Maybe something like (I'm
not C programmer):
PHPAPI void php_var_ref_count(zval **struc)
{
return Z_REFCOUNT_PP(struc);
}
Knowing the reference count may be helpful to:
- make copies of a structure without references
- remove variables that have more than 1 reference (safe remove)
- remove variables that are not referenced (unused values)
- prevent modifying a variable that is referenced
I believe there must be more applications but these are the ones I can
think of.
(background:
http://stackoverflow.com/questions/3148125/php-check-if-object-array-is-a-reference)
Test script:
---
"apple", "b" => "banana");
$es = array("a" => "manzana", "b" => "platano");
$dict = array(
"Eng" => $en,
"Esp" => $es,
"Non" => array("a" => "A", "b" => "B")
);
echo var_ref_count($dict["Eng"]);
echo " # ";
echo var_ref_count($dict["Non"]);
?>
Expected result:
2 # 1
Actual result:
--
None (inexistent)
--
Edit this bug report at http://bugs.php.net/bug.php?id=52215&edit=1
Req #52219 [Com]: More useful return values for SoapClient::__getFunctions() / __getTypes()
Edit report at http://bugs.php.net/bug.php?id=52219&edit=1
ID: 52219
Comment by: php_bugs at m-h-it dot de
Reported by: php_bugs at m-h-it dot de
Summary: More useful return values for
SoapClient::__getFunctions() / __getTypes()
Status: Open
Type: Feature/Change Request
Package: SOAP related
Operating System: irrelevant
PHP Version: 5.2.13
New Comment:
I'm sorry, i have typo in my suggestion. It should be:
array (
array(
'name'=>'loginCustomer',
'returnType'=>'LoginCustomerResponse',
'arguments'=>array(
array('name'=>'loginCustomer','type'=>'LoginCustomer')
)
),
array(
'name'=>'findCustomer',
'returnType'=>'FindCustomerResponse',
'arguments'=>array(
array('name'=>'findCustomer','type'=>'FindCustomerResponse'),
)
),
...
)
Previous Comments:
[2010-07-01 12:29:29] php_bugs at m-h-it dot de
Description:
SoapClient::__getFunctions() / SoapClient::__getTypes() now return both
an array
of strings. Each string contains a function signature or a structure
description
respectively. So it's hard to really use this information
programmatically. You
have to parse the strings into something useful first.
It would be more useful to get a nested array with all information from
the
WSDL. For example __getFunctions() now delivers this:
array (
0 => 'LoginCustomerResponse loginCustomer(LoginCustomer
$loginCustomer)',
1 => 'FindCustomerResponse findCustomer(FindCustomer $findCustomer)',
...
)
it's easier to use this:
array (
array(
'name'=>'loginCustomer',
'returnType'=>'LoginCustomerResponse',
'arguments'=>array(
array('name'=>'loginCustomer'),
array('type'=>'LoginCustomer')
)
),
array(
'name'=>'findCustomer',
'returnType'=>'FindCustomerResponse',
'arguments'=>array(
array('name'=>'findCustomer'),
array('type'=>'FindCustomerResponse')
)
),
...
)
To keep BC i suggest to use default argument:
SoapClient::__getFunctions($asStrings=true)
SoapClient::__getTypes($asStrings=true)
--
Edit this bug report at http://bugs.php.net/bug.php?id=52219&edit=1
[PHP-BUG] Req #52219 [NEW]: More useful return values for SoapClient::__getFunctions() / __getTypes()
From:
Operating system: irrelevant
PHP version: 5.2.13
Package: SOAP related
Bug Type: Feature/Change Request
Bug description:More useful return values for SoapClient::__getFunctions() /
__getTypes()
Description:
SoapClient::__getFunctions() / SoapClient::__getTypes() now return both an
array
of strings. Each string contains a function signature or a structure
description
respectively. So it's hard to really use this information programmatically.
You
have to parse the strings into something useful first.
It would be more useful to get a nested array with all information from the
WSDL. For example __getFunctions() now delivers this:
array (
0 => 'LoginCustomerResponse loginCustomer(LoginCustomer
$loginCustomer)',
1 => 'FindCustomerResponse findCustomer(FindCustomer $findCustomer)',
...
)
it's easier to use this:
array (
array(
'name'=>'loginCustomer',
'returnType'=>'LoginCustomerResponse',
'arguments'=>array(
array('name'=>'loginCustomer'),
array('type'=>'LoginCustomer')
)
),
array(
'name'=>'findCustomer',
'returnType'=>'FindCustomerResponse',
'arguments'=>array(
array('name'=>'findCustomer'),
array('type'=>'FindCustomerResponse')
)
),
...
)
To keep BC i suggest to use default argument:
SoapClient::__getFunctions($asStrings=true)
SoapClient::__getTypes($asStrings=true)
--
Edit bug report at http://bugs.php.net/bug.php?id=52219&edit=1
--
Try a snapshot (PHP 5.2):
http://bugs.php.net/fix.php?id=52219&r=trysnapshot52
Try a snapshot (PHP 5.3):
http://bugs.php.net/fix.php?id=52219&r=trysnapshot53
Try a snapshot (trunk):
http://bugs.php.net/fix.php?id=52219&r=trysnapshottrunk
Fixed in SVN:
http://bugs.php.net/fix.php?id=52219&r=fixed
Fixed in SVN and need be documented:
http://bugs.php.net/fix.php?id=52219&r=needdocs
Fixed in release:
http://bugs.php.net/fix.php?id=52219&r=alreadyfixed
Need backtrace:
http://bugs.php.net/fix.php?id=52219&r=needtrace
Need Reproduce Script:
http://bugs.php.net/fix.php?id=52219&r=needscript
Try newer version:
http://bugs.php.net/fix.php?id=52219&r=oldversion
Not developer issue:
http://bugs.php.net/fix.php?id=52219&r=support
Expected behavior:
http://bugs.php.net/fix.php?id=52219&r=notwrong
Not enough info:
http://bugs.php.net/fix.php?id=52219&r=notenoughinfo
Submitted twice:
http://bugs.php.net/fix.php?id=52219&r=submittedtwice
register_globals:
http://bugs.php.net/fix.php?id=52219&r=globals
PHP 4 support discontinued: http://bugs.php.net/fix.php?id=52219&r=php4
Daylight Savings:http://bugs.php.net/fix.php?id=52219&r=dst
IIS Stability:
http://bugs.php.net/fix.php?id=52219&r=isapi
Install GNU Sed:
http://bugs.php.net/fix.php?id=52219&r=gnused
Floating point limitations:
http://bugs.php.net/fix.php?id=52219&r=float
No Zend Extensions:
http://bugs.php.net/fix.php?id=52219&r=nozend
MySQL Configuration Error:
http://bugs.php.net/fix.php?id=52219&r=mysqlcfg
[PHP-BUG] Bug #52218 [NEW]: php module not loaded when php module is enabled and then restarted
From:
Operating system: Linux
PHP version: Irrelevant
Package: Apache2 related
Bug Type: Bug
Bug description:php module not loaded when php module is enabled and then
restarted
Description:
When apache is restarted after enabling php then it does not load PHP.
When apache is restarted again then it does load PHP.
I first noticed this with (stock) php-4.4.9 and apache 2.2.15 where it
resulted in segmentation faults on every request.
I then tested this with php-5.2.6 that comes with debian
(PHP/5.2.6-1+lenny8 with Suhosin-Patch) and it shows the same problem but
does not segfault.
Steps to reproduce (on debian):
1. install apache2 and php
2. stop apache: apache2ctl -k stop
3. disable the php module: rm /etc/apache2/mods-enabled/php.*
4. start apache: apache2ctl -k start
5. enable the php module: ln -s /etc/apache2/mods-available/php.*
/etc/apache2/mods-enabled/
6. restart apache: apache2ctl -k restart
7. Check the log file => PHP not loaded. (on php-4.4.9 every request from
this moment on caused a segmentation fault)
8. restart apache again: apache2ctl -k restart
9. Check the log file => PHP is loaded
NOTE: do not use the debian init script (/etc/init.d/apache2).
I do not know what the severity of this issue should be.
It does not appear to be segfaulting in recent versions but that does not
nessesarly mean that the segfault is fixed (it could be hiding somewhere -
I did not try very hard to get it to segfault).
The logfile/traces/a guess where the problem is:
An example log file:
PHP disabled: apache2ctl -k start:
[Thu Jul 01 10:55:03 2010] [notice] Apache/2.2.9 (Debian) configured --
resuming normal operations
PHP enabled; apache2ctl -k restart:
[Thu Jul 01 10:55:10 2010] [notice] SIGHUP received. Attempting to
restart
[Thu Jul 01 10:55:10 2010] [notice] Apache/2.2.9 (Debian) configured --
resuming normal operations
==> PHP not loaded
Another restart: apache2ctl -k restart:
[Thu Jul 01 10:55:15 2010] [notice] SIGHUP received. Attempting to
restart
[Thu Jul 01 10:55:15 2010] [notice] Apache/2.2.9 (Debian)
PHP/5.2.6-1+lenny8 with Suhosin-Patch configured -- resuming normal
operations
==> PHP now loaded
debugging this with php 4.4.9 may points to a problem in
sapi/apache2handler/sapi_apache2.c: php_apache_server_startup.
(I did not debug this with php 5 but I did check the code in trunk and it
looks the same).
The start of the code:
static int
php_apache_server_startup(apr_pool_t *pconf, apr_pool_t *plog,
apr_pool_t
*ptemp, server_rec *s)
{
void *data = NULL;
const char *userdata_key = "apache2hook_post_config";
/* Apache will load, unload and then reload a DSO module. This
* prevents us from starting PHP until the second load. */
The comment makes me believe that it is the expectation that
php_apache_server_startup is called twice but that does not appear to be
the case.
gdb session:
Breakpoint 1 (php_apache_server_startup) pending.
(gdb) c
Continuing.
Program received signal SIGHUP, Hangup.
0xb7c2cfb8 in select () from /ub/lib/libc.so.6
(gdb) c
Continuing.
Program received signal SIGHUP, Hangup.
0xb7b9df51 in kill () from /ub/lib/libc.so.6
# NOTE: this is the first restart after enabling the module
(gdb) c
Continuing.
Breakpoint 2 at 0xb78de20f: file
/root/work/php-4.4.9/sapi/apache2handler/sapi_apache2.c, line 367.
Pending breakpoint "php_apache_server_startup" resolved
Breakpoint 2, php_apache_server_startup (pconf=0x80e80a8, plog=0x81301c8,
ptemp=0x80f20d0, s=0x80ed0c0) at
/root/work/php-4.4.9/sapi/apache2handler/sapi_apache2.c:367
367 void *data = NULL;
(gdb) n
368 const char *userdata_key = "apache2hook_post_config";
(gdb) n
372 apr_pool_userdata_get(&data, userdata_key, s->process->pool);
(gdb) n
373 if (data == NULL) {
(gdb) p data
$1 = (void *) 0x0
(gdb) c
Continuing.
# NOTE: PHP is not loaded in apache even tho it is enabled
Program received signal SIGHUP, Hangup.
0xb7c2cfb8 in select () from /ub/lib/libc.so.6
(gdb) c
Continuing.
Program received signal SIGHUP, Hangup.
0xb7b9df51 in kill () from /ub/lib/libc.so.6
(gdb) c
Continuing.
# NOTE: this is the second restart after enabling the module
warning: Temporarily disabling breakpoints for unloaded shared library
"/ub/pkg/httpd/modules/libphp4.so"
Breakpoint 2, php_apache_server_startup (pconf=0x80e80a8, plog=0x81301c8,
ptemp=0x81361e0, s=0x80ed0c0) at
/root/work/php-4.4.9/sapi/apache2handler/sapi_apache2.c:367
367 void *data = NULL;
(gdb) n
368 const char *userdata_key = "apache2hook_post_config";
(gdb) n
372 apr_pool_userdata_get(&data, userdata_key, s->process->pool);
(gdb) n
373 if (data == NULL) {
(gdb) p data
$2 = (void *) 0x1
(gdb) n
384
