#34216 [Opn->Asn]: Segfault with autoload

2005-08-22 Thread sebastian 
 ID:   34216
 Updated by:   [EMAIL PROTECTED]
 Reported By:  [EMAIL PROTECTED]
-Status:   Open
+Status:   Assigned
 Bug Type: SPL related
 Operating System: Linux
 PHP Version:  5.1.0RC1
-Assigned To:  
+Assigned To:  helly


Previous Comments:


[2005-08-23 07:27:09] [EMAIL PROTECTED]

Description:

I am experiencing a segfault with spl_autoload in the
PHP_5_1 branch with the upcoming PHPUnit 2.3.0.

The nature of PHPUnit's complexity makes it almost
impossible for me to come up with a simple, reproducing
script (I already invested hours into this, to no avail so
far).

It is weird that the segfault only occurs in one out of
two common code paths of PHPUnit's TextUI test runner.

Furthermore, small code changes (see below) seem to make
the problem go away:

Reproduce code:
---
Segfault:

   function PHPUnit2_Util_Classloader($className) {
   print "Loading $className.\n";

   if (strpos($className, 'PHPUnit2_') !== FALSE) {
   @include(str_replace('_', '/', $className) . '.php');
   }
   }

   spl_autoload_register('PHPUnit2_Util_Classloader');

Segfault:

   function PHPUnit2_Util_Classloader($className) {
   $bt = debug_backtrace();
   print "Loading $className.\n";

   if (strpos($className, 'PHPUnit2_') !== FALSE) {
   @include(str_replace('_', '/', $className) . '.php');
   }
   }

   spl_autoload_register('PHPUnit2_Util_Classloader');

No Segfault:

   function PHPUnit2_Util_Classloader($className) {
   $bt = debug_backtrace();
   print_r($bt[1]);
   print "Loading $className.\n";

   if (strpos($className, 'PHPUnit2_') !== FALSE) {
   @include(str_replace('_', '/', $className) . '.php');
   }
   }

   spl_autoload_register('PHPUnit2_Util_Classloader');

Actual result:
--
Below are the links to the results of GDB and Valgrind.

Interestingly, when run through GDB it segfaults, when run
through Valgrind it does not segfault.

GDB:
http://www.sebastian-bergmann.de/stuff/phpunit2-gdb.txt
Valgrind:
http://www.sebastian-bergmann.de/stuff/phpunit2-valgrind.txt





-- 
Edit this bug report at http://bugs.php.net/?id=34216&edit=1

#34216 [NEW]: Segfault with autoload

2005-08-22 Thread [EMAIL PROTECTED] 
From: [EMAIL PROTECTED]
Operating system: Linux
PHP version:  5.1.0RC1
PHP Bug Type: SPL related
Bug description:  Segfault with autoload

Description:

I am experiencing a segfault with spl_autoload in the
PHP_5_1 branch with the upcoming PHPUnit 2.3.0.

The nature of PHPUnit's complexity makes it almost
impossible for me to come up with a simple, reproducing
script (I already invested hours into this, to no avail so
far).

It is weird that the segfault only occurs in one out of
two common code paths of PHPUnit's TextUI test runner.

Furthermore, small code changes (see below) seem to make
the problem go away:

Reproduce code:
---
Segfault:

   function PHPUnit2_Util_Classloader($className) {
   print "Loading $className.\n";

   if (strpos($className, 'PHPUnit2_') !== FALSE) {
   @include(str_replace('_', '/', $className) . '.php');
   }
   }

   spl_autoload_register('PHPUnit2_Util_Classloader');

Segfault:

   function PHPUnit2_Util_Classloader($className) {
   $bt = debug_backtrace();
   print "Loading $className.\n";

   if (strpos($className, 'PHPUnit2_') !== FALSE) {
   @include(str_replace('_', '/', $className) . '.php');
   }
   }

   spl_autoload_register('PHPUnit2_Util_Classloader');

No Segfault:

   function PHPUnit2_Util_Classloader($className) {
   $bt = debug_backtrace();
   print_r($bt[1]);
   print "Loading $className.\n";

   if (strpos($className, 'PHPUnit2_') !== FALSE) {
   @include(str_replace('_', '/', $className) . '.php');
   }
   }

   spl_autoload_register('PHPUnit2_Util_Classloader');

Actual result:
--
Below are the links to the results of GDB and Valgrind.

Interestingly, when run through GDB it segfaults, when run
through Valgrind it does not segfault.

GDB:
http://www.sebastian-bergmann.de/stuff/phpunit2-gdb.txt
Valgrind:
http://www.sebastian-bergmann.de/stuff/phpunit2-valgrind.txt

-- 
Edit bug report at http://bugs.php.net/?id=34216&edit=1
-- 
Try a CVS snapshot (php4):   http://bugs.php.net/fix.php?id=34216&r=trysnapshot4
Try a CVS snapshot (php5.0): 
http://bugs.php.net/fix.php?id=34216&r=trysnapshot50
Try a CVS snapshot (php5.1): 
http://bugs.php.net/fix.php?id=34216&r=trysnapshot51
Fixed in CVS:http://bugs.php.net/fix.php?id=34216&r=fixedcvs
Fixed in release:http://bugs.php.net/fix.php?id=34216&r=alreadyfixed
Need backtrace:  http://bugs.php.net/fix.php?id=34216&r=needtrace
Need Reproduce Script:   http://bugs.php.net/fix.php?id=34216&r=needscript
Try newer version:   http://bugs.php.net/fix.php?id=34216&r=oldversion
Not developer issue: http://bugs.php.net/fix.php?id=34216&r=support
Expected behavior:   http://bugs.php.net/fix.php?id=34216&r=notwrong
Not enough info: 
http://bugs.php.net/fix.php?id=34216&r=notenoughinfo
Submitted twice: 
http://bugs.php.net/fix.php?id=34216&r=submittedtwice
register_globals:http://bugs.php.net/fix.php?id=34216&r=globals
PHP 3 support discontinued:  http://bugs.php.net/fix.php?id=34216&r=php3
Daylight Savings:http://bugs.php.net/fix.php?id=34216&r=dst
IIS Stability:   http://bugs.php.net/fix.php?id=34216&r=isapi
Install GNU Sed: http://bugs.php.net/fix.php?id=34216&r=gnused
Floating point limitations:  http://bugs.php.net/fix.php?id=34216&r=float
No Zend Extensions:  http://bugs.php.net/fix.php?id=34216&r=nozend
MySQL Configuration Error:   http://bugs.php.net/fix.php?id=34216&r=mysqlcfg

#34199 [Ctl]: SimpleXMLElement is Schroedinger's Cat

2005-08-22 Thread sniper 
 ID:   34199
 Updated by:   [EMAIL PROTECTED]
 Reported By:  [EMAIL PROTECTED]
 Status:   Critical
 Bug Type: Scripting Engine problem
 Operating System: *
 PHP Version:  5CVS-2005-08-20 (CVS)
 New Comment:

This outputs nothing with PHP 5, but does give "yes" with 4.4:



So in PHP5, an "empty" object is still true.
(then again, empty array() is considered false)
Was the change intentional? And if so, why?



Previous Comments:


[2005-08-20 22:53:04] [EMAIL PROTECTED]

Marking it critical.  I think we need to get this cleaned up before 5.1
goes out.  Having 2 different opcodes handling this that can potentially
return conflicting results causes all sorts of confusion.  Took me about
an hour working backwards through assumptions before I finally figured
out that if() and if(!) were to blame.



[2005-08-20 22:33:31] [EMAIL PROTECTED]

if ($kids) if handled by ZEND_JMPZ_SPEC_CV_HANDLER which ends up
calling i_zend_is_true and returning TRUE because $kids is an object.

if (! $kids) is handled by ZEND_BOOL_NOT_SPEC_CV_HANDLER which calls
convert_to_boolean and ends up calling the objects cast_object handler,
which in this case is sxe_object_cast and returns FALSE because object
is really empty.



[2005-08-20 21:13:15] [EMAIL PROTECTED]

Description:

An empty SimpleXMLElement object is both true and false

Reproduce code:
---
$xml = "";
$xml = simplexml_load_string($xml);
$kids = $xml->children();
if($kids) echo "True";
if(!$kids) echo "False";

Expected result:

False

Actual result:
--
TrueFalse





-- 
Edit this bug report at http://bugs.php.net/?id=34199&edit=1

#34089 [Fbk->Opn]: Configure fails build test for libxml2

2005-08-22 Thread [EMAIL PROTECTED] 
 ID:   34089
 User updated by:  [EMAIL PROTECTED]
 Reported By:  [EMAIL PROTECTED]
-Status:   Feedback
+Status:   Open
 Bug Type: *Configuration Issues
 Operating System: Mac OS X 10.4.2
 PHP Version:  6CVS-2005-08-12
 New Comment:

So, does this mean I should downgrade my system? ;-)

My system could be newer, but Andrei was apparently able to get this
thing to compile on a Mac and work. I wonder how he was able to do it?


Previous Comments:


[2005-08-18 02:36:46] [EMAIL PROTECTED]

I'm definately no Macosx expert but this seems likely cause:
  
  http://www.xdarwin.org/forum/read.php?f=3&i=1172&t=1171

Either your system is newer or the prebuild icu lib is newer and
incompatible?




[2005-08-17 20:21:11] [EMAIL PROTECTED]

That configure line works.

I tried:
# ./configure --with-layout=PHP --prefix=/usr/local/php/6.0.0
--disable-all --with-icu-dir=/usr/local/icu
# make
# make install

All worked just fine, but now I'm getting a "Library not loaded"
error.

ramsey:~ ramsey$ /usr/local/php/6.0.0/bin/php -m
dyld: Library not loaded: libicui18n.dylib.34
  Referenced from: /usr/local/php/6.0.0/bin/php
  Reason: image not found
Trace/BPT trap



[2005-08-17 19:21:23] [EMAIL PROTECTED]

Does this work:

# ./configure --disable-all --with-icu-dir=/usr/local/icu




[2005-08-16 21:09:17] [EMAIL PROTECTED]

Updated my copy of HEAD today at 2005-08-16 18:54:06 GMT. This problem
still occurs.



[2005-08-16 01:43:57] [EMAIL PROTECTED]

Try the latest CVS HEAD. 



The remainder of the comments for this report are too long. To view
the rest of the comments, please view the bug report online at
http://bugs.php.net/34089

-- 
Edit this bug report at http://bugs.php.net/?id=34089&edit=1

#34050 [Opn->Fbk]: safe_mode_include_dir isn't honored

2005-08-22 Thread sniper 
 ID:   34050
 Updated by:   [EMAIL PROTECTED]
 Reported By:  wolfram at schlich dot org
-Status:   Open
+Status:   Feedback
 Bug Type: Safe Mode/open_basedir
 Operating System: Linux 2.6.12-gentoo-r4 x86_64
 PHP Version:  5CVS-2005-08-09 (dev)
 New Comment:

I can't reproduce this either. Try this on command line:

# php -r 'include("/usr/lib/php/Smarty/Smarty.class.php");'

Also, make sure that PHP really is loading the right php.ini file!!
(e.g. if you happen to have one in / that one will be loaded instead)



Previous Comments:


[2005-08-12 01:09:27] wolfram at schlich dot org

no symlink...



[2005-08-12 00:30:26] [EMAIL PROTECTED]

Can't reproduce.
Make sure that /usr/lib/php/Smarty (or any of it's parent dirs) is not
a symbolic link. 



[2005-08-09 18:16:23] wolfram at schlich dot org

then the comment above safe_mode_include_dir is silly:

; When safe_mode is on, UID/GID checks are bypassed when
; including files from this directory and its subdirectories.
; (directory must also be in include_path or full path must
; be used when including)

it says "UID/GID checks are bypassed", so why should I enable
safe_mod_gid then?

I even tried that, result:
--8<--
Warning: main() [function.main]: SAFE MODE Restriction in effect. The
script whose uid/gid is 667/888 is not allowed to access
/usr/lib/php/Smarty/Smarty.class.php owned by uid/gid 0/0 in
/home/wschlich/public_html/smartytest/class.MySmarty.inc.php on line
27
--8<--



[2005-08-09 18:09:19] [EMAIL PROTECTED]

; By default, Safe Mode does a UID compare check when
; opening files. If you want to relax this to a GID compare,
; then turn on safe_mode_gid.




[2005-08-09 16:56:52] wolfram at schlich dot org

Description:

when safe_mode is active, I want to be able to include files from
certain directories without checking for their UID/GID.

Reproduce code:
---
--8<--[ php.ini ]--8<--
safe_mode = On
safe_mode_gid = Off
safe_mode_include_dir = "/usr/share/php/:/usr/lib/php/"
safe_mode_allowed_env_vars = PHP_,LANG,LC_
--8<--[ apache vhost config ]--8<--
php_value include_path ".:/usr/share/php:/usr/lib/php"
--8<--[ sample php script ]--8<--
require_once('/usr/lib/php/Smarty/Smarty.class.php');
--8<--

Expected result:

included file, no error message.

Actual result:
--
Warning: main() [function.main]: SAFE MODE Restriction in effect. The
script whose uid is 667 is not allowed to access
/usr/lib/php/Smarty/Smarty.class.php owned by uid 0 in
/home/wschlich/public_html/smartytest/smartytest.php on line 19





-- 
Edit this bug report at http://bugs.php.net/?id=34050&edit=1

#33529 [Opn->Fbk]: __FILE__ is relative while running Apache SAPI

2005-08-22 Thread sniper 
 ID:   33529
 Updated by:   [EMAIL PROTECTED]
 Reported By:  [EMAIL PROTECTED]
-Status:   Open
+Status:   Feedback
 Bug Type: Scripting Engine problem
 Operating System: Mac OS X
 PHP Version:  5CVS, 4CVS (2005-07-03)
 New Comment:

Can you explain me why I can't reproduce this on same Macosx version,
same apache version and PHP 4.4.0 ??
(yes, I tested within a virtualhost and a subdir in it)

Does your setup have some symlinks somewhere in the path?



Previous Comments:


[2005-07-03 01:58:32] [EMAIL PROTECTED]

Yes, same result with 5.1.0-dev (indeed)



[2005-07-02 15:45:37] [EMAIL PROTECTED]

Does this happen with PHP 5.1-dev?

http://snaps.php.net/php5-latest.tar.gz




[2005-07-01 05:31:19] [EMAIL PROTECTED]

If you try with 2 scripts:

/index.php
/something/index.php

and /index.php contains print __FILE__ and /something/index.php
contains include('../index.php'), if you point your browser to
/index.php the content of __FILE__ is correct.

If you try /something/index.php, __FILE__ will contains '../index.php'



[2005-06-30 22:56:48] [EMAIL PROTECTED]

Description:

The content of the __FILE__ constant is relative while running through
a Virtual Host on OS X.

The bug is not seen while working with the CLI version or if the script
is not in a VirtualHost directive.

Configure is as follow (majorly taken from OS X compile flags):
'./configure' '--prefix=/usr' '--mandir=/usr/share/man'
'--infodir=/usr/share/info' '--with-apxs' '--enable-cli' '--with-zlib'
'--with-xml' '--enable-exif' '--enable-ftp' '--enable-sockets'
'--with-curl=/usr' '--with-config-file-path=/etc'
'--sysconfdir=/private/etc'

Apache is the one distributed with Tiger 10.4.1 (1.3.33).

The same code runs fine on Linux and Windows with the same config
(vhosts settings).

Reproduce code:
---
http://bugs.php.net/?id=33529&edit=1

#34214 [Opn->Bgs]: base64_decode does not properly ignore whitespace

2005-08-22 Thread sniper 
 ID:   34214
 Updated by:   [EMAIL PROTECTED]
 Reported By:  dev at worldlinked dot net
-Status:   Open
+Status:   Bogus
 Bug Type: *Mail Related
 Operating System: all
 PHP Version:  5.0.4
 New Comment:

Works fine with using PHP 4.4.0, PHP 5.1-dev, PHP 6.0-dev..
There's something really wrong either with your script or the machine
running the script.



Previous Comments:


[2005-08-23 00:08:48] dev at worldlinked dot net

This bug was first reported in
http://bugs.php.net/bug.php?id=1596 
however the patch listed there was either never applied or does not
work.



[2005-08-22 23:39:49] dev at worldlinked dot net

Description:

RFC 2045 says:
"All line breaks or other characters not found in Table 1 must be
ignored by decoding software."

However if a caracter such as a space is in the encoded data (which
some encoders such as newegg.com do) then the base64_decode does not
work.

Reproduce code:
---
http://www.worldlinked.net/blog/files/testbase64.php
or
http://www.worldlinked.net/blog/files/testbase64.phps


Expected result:

Hello, There seems to be a problem with php's base64_decode() function.
Whitespace needs to be trimmed from the email before it will properly
decode.

Actual result:
--
Hello, There seems to be a problem with php's base64_decode() function.
Whitespace needs to be trimmed from the
emù…¥°‰•™½É”¥ÐÝ¥±°ÁɽÁ•É±ä‘•½‘”¸





-- 
Edit this bug report at http://bugs.php.net/?id=34214&edit=1

#34214 [Opn]: base64_decode does not properly ignore whitespace

2005-08-22 Thread dev at worldlinked dot net 
 ID:   34214
 User updated by:  dev at worldlinked dot net
 Reported By:  dev at worldlinked dot net
 Status:   Open
 Bug Type: *Mail Related
 Operating System: all
 PHP Version:  5.0.4
 New Comment:

This bug was first reported in
http://bugs.php.net/bug.php?id=1596 
however the patch listed there was either never applied or does not
work.


Previous Comments:


[2005-08-22 23:39:49] dev at worldlinked dot net

Description:

RFC 2045 says:
"All line breaks or other characters not found in Table 1 must be
ignored by decoding software."

However if a caracter such as a space is in the encoded data (which
some encoders such as newegg.com do) then the base64_decode does not
work.

Reproduce code:
---
http://www.worldlinked.net/blog/files/testbase64.php
or
http://www.worldlinked.net/blog/files/testbase64.phps


Expected result:

Hello, There seems to be a problem with php's base64_decode() function.
Whitespace needs to be trimmed from the email before it will properly
decode.

Actual result:
--
Hello, There seems to be a problem with php's base64_decode() function.
Whitespace needs to be trimmed from the
emù…¥°‰•™½É”¥ÐÝ¥±°ÁɽÁ•É±ä‘•½‘”¸





-- 
Edit this bug report at http://bugs.php.net/?id=34214&edit=1

#34214 [NEW]: base64_decode does not properly ignore whitespace

2005-08-22 Thread dev at worldlinked dot net 
From: dev at worldlinked dot net
Operating system: all
PHP version:  5.0.4
PHP Bug Type: *Mail Related
Bug description:  base64_decode does not properly ignore whitespace

Description:

RFC 2045 says:
"All line breaks or other characters not found in Table 1 must be ignored
by decoding software."

However if a caracter such as a space is in the encoded data (which some
encoders such as newegg.com do) then the base64_decode does not work.

Reproduce code:
---
http://www.worldlinked.net/blog/files/testbase64.php
or
http://www.worldlinked.net/blog/files/testbase64.phps


Expected result:

Hello, There seems to be a problem with php's base64_decode() function.
Whitespace needs to be trimmed from the email before it will properly
decode.

Actual result:
--
Hello, There seems to be a problem with php's base64_decode() function.
Whitespace needs to be trimmed from the
emù…¥°‰•™½É”¥ÐÝ¥±°ÁɽÁ•É±ä‘•½‘”¸

-- 
Edit bug report at http://bugs.php.net/?id=34214&edit=1
-- 
Try a CVS snapshot (php4):   http://bugs.php.net/fix.php?id=34214&r=trysnapshot4
Try a CVS snapshot (php5.0): 
http://bugs.php.net/fix.php?id=34214&r=trysnapshot50
Try a CVS snapshot (php5.1): 
http://bugs.php.net/fix.php?id=34214&r=trysnapshot51
Fixed in CVS:http://bugs.php.net/fix.php?id=34214&r=fixedcvs
Fixed in release:http://bugs.php.net/fix.php?id=34214&r=alreadyfixed
Need backtrace:  http://bugs.php.net/fix.php?id=34214&r=needtrace
Need Reproduce Script:   http://bugs.php.net/fix.php?id=34214&r=needscript
Try newer version:   http://bugs.php.net/fix.php?id=34214&r=oldversion
Not developer issue: http://bugs.php.net/fix.php?id=34214&r=support
Expected behavior:   http://bugs.php.net/fix.php?id=34214&r=notwrong
Not enough info: 
http://bugs.php.net/fix.php?id=34214&r=notenoughinfo
Submitted twice: 
http://bugs.php.net/fix.php?id=34214&r=submittedtwice
register_globals:http://bugs.php.net/fix.php?id=34214&r=globals
PHP 3 support discontinued:  http://bugs.php.net/fix.php?id=34214&r=php3
Daylight Savings:http://bugs.php.net/fix.php?id=34214&r=dst
IIS Stability:   http://bugs.php.net/fix.php?id=34214&r=isapi
Install GNU Sed: http://bugs.php.net/fix.php?id=34214&r=gnused
Floating point limitations:  http://bugs.php.net/fix.php?id=34214&r=float
No Zend Extensions:  http://bugs.php.net/fix.php?id=34214&r=nozend
MySQL Configuration Error:   http://bugs.php.net/fix.php?id=34214&r=mysqlcfg

#34211 [Opn->Asn]: ext/oci8: Allow for data type "TIMESTAMP(0) WITH LOCAL TIME ZONE"

2005-08-22 Thread sniper 
 ID:   34211
 Updated by:   [EMAIL PROTECTED]
-Summary:  Allow for data type "TIMESTAMP(0) WITH LOCAL TIME
   ZONE"
 Reported By:  kurtb149 at yahoo dot com
-Status:   Open
+Status:   Assigned
-Bug Type: OCI8 related
+Bug Type: Feature/Change Request
 Operating System: Linux
 PHP Version:  6CVS-2005-08-22 (CVS)
-Assigned To:  
+Assigned To:  tony2001
 New Comment:

reclassified



Previous Comments:


[2005-08-22 19:11:40] kurtb149 at yahoo dot com

Description:

In the file:

   ext/pdo_oci/oci_statement.c

the function oci_stmt_describe() does not allow for the 
data type "TIMESTAMP(0) WITH LOCAL TIME ZONE". Here is the diff of an
updated oci_statement.c that would allow for the data type:

$ cvs diff oci_statement.c 
Index: oci_statement.c
===
RCS file: /repository/php-src/ext/pdo_oci/oci_statement.c,v
retrieving revision 1.16
diff -r1.16 oci_statement.c
407a408,410
> #ifdef SQLT_TIMESTAMP_LTZ
>   || dtype == SQLT_TIMESTAMP_LTZ
> #endif







-- 
Edit this bug report at http://bugs.php.net/?id=34211&edit=1

#34189 [WFx]: After Configure and Make can load symbol ap_loaded_modules

2005-08-22 Thread david dot marek at mattel dot com 
 ID:   34189
 User updated by:  david dot marek at mattel dot com
 Reported By:  david dot marek at mattel dot com
 Status:   Wont fix
 Bug Type: Apache related
 Operating System: AIX 5.0
 PHP Version:  5.1.0b3
 New Comment:

Understood.  IBM and Oracle do have issues.  

I did modify php_apache.c in sapi/apache to remove the get_modules and
it now works (No the best answer)  Something on AIX doesn't like
get_modules and includes it in the library. Solaris doesn't do this.

Thanks!


Previous Comments:


[2005-08-22 22:30:41] [EMAIL PROTECTED]

Friendly advise: Don't use AIX. 
I changed this to "wont fix". Ask IBM or Oracle to fix their crappy
software.




[2005-08-22 19:18:12] david dot marek at mattel dot com

I am reopening as I have some questions you may be able to help with. I
hope!



[2005-08-22 16:35:25] david dot marek at mattel dot com

I do have question.  If we build this agains solaris all is good. 
Meaning we don't see the ap_Loaded_modules in the php library.  In AIX
we do.  This is the compile and build on the PHP side.  Is this
something we can remove or find a way not to include.  We can't move
forward with the migration without it so any help would be great.

Thanks!



[2005-08-19 11:46:40] [EMAIL PROTECTED]

If it works with Apache 1.3.33, it's no bug in PHP. Complain to Oracle.
Also, we have never said we support anything but the 'real' apache..




[2005-08-19 06:52:12] david dot marek at mattel dot com

It works with the build of Apache 1.3.33 but it still does not work
with Oracle's Apache.  We are using GCC and the dependencies have all
be recompiled.  When I build PHP on Solaris I don't get the symbol
referenced in the libphp5.so but on AIX I do.  Is there a reason the
make is added the symbol to the AIX build and not the Solaris build.
Everything else is the same.  I grep on ap_loaded_modules on libphp5.so
and I get a value back on AIX and that appears to be the culprit.



The remainder of the comments for this report are too long. To view
the rest of the comments, please view the bug report online at
http://bugs.php.net/34189

-- 
Edit this bug report at http://bugs.php.net/?id=34189&edit=1

#34208 [Opn->Csd]: crash using session_set_save_handler & throwing exception in own errorhandler

2005-08-22 Thread sniper 
 ID:   34208
 Updated by:   [EMAIL PROTECTED]
 Reported By:  carsten dot jendro at gmx dot net
-Status:   Open
+Status:   Closed
 Bug Type: Reproducible crash
 Operating System: win32 only
 PHP Version:  5.1.0RC1


Previous Comments:


[2005-08-22 22:40:50] carsten dot jendro at gmx dot net

works fine with actual
http://snaps.php.net/win32/php5-win32-latest.zip

what was the bugfix? looks like php sessions are disabled after fatal
error.



[2005-08-22 17:40:39] [EMAIL PROTECTED]

Please try using this CVS snapshot:

  http://snaps.php.net/php5-latest.tar.gz
 
For Windows:
 
  http://snaps.php.net/win32/php5-win32-latest.zip





[2005-08-22 11:09:16] [EMAIL PROTECTED]

Does NOT crash under Linux. (ZTS problem propably)




[2005-08-22 01:39:04] carsten dot jendro at gmx dot net

Description:

php5/apache crashes when using set_error_handler and
session_set_save_handler together and throwing an exception in error
handler.

Testet with php 5.0.4.4 and 5.1rc1, apache 1 & 2

Reproduce code:
---
Bar();

?>

Expected result:

no crash ;-)

Actual result:
--
php5ts.dll!00a2b37a()   
php5ts.dll!00aa1248()   
php5ts.dll!00aa1494()   
php5ts.dll!00aa59fe()   
php5ts.dll!00aa587d()   
php5ts.dll!00aa5beb()   
php5ts.dll!009f640c()   
php5ts.dll!00a2e552()   
php5ts.dll!009f23bd()   
php5ts.dll!00a5244e()   
msvcrt.dll!77bfc2e3()   
ntwdblib.dll!7333ab22() 
ntwdblib.dll!7333f33e() 
ntwdblib.dll!7333f3eb() 
php5apache2.dll!003f2f9f()  
php5apache2.dll!003f34e8()  
php5apache2.dll!003f3163()  
libhttpd.dll!6ff0155f() 
libhttpd.dll!6ff018a9() 
libhttpd.dll!6ff0d582() 
libhttpd.dll!6ff095e6() 
libhttpd.dll!6ff0411f() 
libhttpd.dll!6ff04372() 
libhttpd.dll!6ff1bc3a() 
msvcrt.dll!77c0a3b0()   
kernel32.dll!7c80b50b() 
kernel32.dll!7c8399f3() 






-- 
Edit this bug report at http://bugs.php.net/?id=34208&edit=1

#34208 [Fbk->Opn]: crash using session_set_save_handler & throwing exception in own errorhandler

2005-08-22 Thread carsten dot jendro at gmx dot net 
 ID:   34208
 User updated by:  carsten dot jendro at gmx dot net
 Reported By:  carsten dot jendro at gmx dot net
-Status:   Feedback
+Status:   Open
 Bug Type: Reproducible crash
 Operating System: win32 only
 PHP Version:  5.1.0RC1
 New Comment:

works fine with actual
http://snaps.php.net/win32/php5-win32-latest.zip

what was the bugfix? looks like php sessions are disabled after fatal
error.


Previous Comments:


[2005-08-22 17:40:39] [EMAIL PROTECTED]

Please try using this CVS snapshot:

  http://snaps.php.net/php5-latest.tar.gz
 
For Windows:
 
  http://snaps.php.net/win32/php5-win32-latest.zip





[2005-08-22 11:09:16] [EMAIL PROTECTED]

Does NOT crash under Linux. (ZTS problem propably)




[2005-08-22 01:39:04] carsten dot jendro at gmx dot net

Description:

php5/apache crashes when using set_error_handler and
session_set_save_handler together and throwing an exception in error
handler.

Testet with php 5.0.4.4 and 5.1rc1, apache 1 & 2

Reproduce code:
---
Bar();

?>

Expected result:

no crash ;-)

Actual result:
--
php5ts.dll!00a2b37a()   
php5ts.dll!00aa1248()   
php5ts.dll!00aa1494()   
php5ts.dll!00aa59fe()   
php5ts.dll!00aa587d()   
php5ts.dll!00aa5beb()   
php5ts.dll!009f640c()   
php5ts.dll!00a2e552()   
php5ts.dll!009f23bd()   
php5ts.dll!00a5244e()   
msvcrt.dll!77bfc2e3()   
ntwdblib.dll!7333ab22() 
ntwdblib.dll!7333f33e() 
ntwdblib.dll!7333f3eb() 
php5apache2.dll!003f2f9f()  
php5apache2.dll!003f34e8()  
php5apache2.dll!003f3163()  
libhttpd.dll!6ff0155f() 
libhttpd.dll!6ff018a9() 
libhttpd.dll!6ff0d582() 
libhttpd.dll!6ff095e6() 
libhttpd.dll!6ff0411f() 
libhttpd.dll!6ff04372() 
libhttpd.dll!6ff1bc3a() 
msvcrt.dll!77c0a3b0()   
kernel32.dll!7c80b50b() 
kernel32.dll!7c8399f3() 






-- 
Edit this bug report at http://bugs.php.net/?id=34208&edit=1

#34189 [Opn->WFx]: After Configure and Make can load symbol ap_loaded_modules

2005-08-22 Thread sniper 
 ID:   34189
 Updated by:   [EMAIL PROTECTED]
 Reported By:  david dot marek at mattel dot com
-Status:   Open
+Status:   Wont fix
 Bug Type: Apache related
 Operating System: AIX 5.0
 PHP Version:  5.1.0b3
 New Comment:

Friendly advise: Don't use AIX. 
I changed this to "wont fix". Ask IBM or Oracle to fix their crappy
software.



Previous Comments:


[2005-08-22 19:18:12] david dot marek at mattel dot com

I am reopening as I have some questions you may be able to help with. I
hope!



[2005-08-22 16:35:25] david dot marek at mattel dot com

I do have question.  If we build this agains solaris all is good. 
Meaning we don't see the ap_Loaded_modules in the php library.  In AIX
we do.  This is the compile and build on the PHP side.  Is this
something we can remove or find a way not to include.  We can't move
forward with the migration without it so any help would be great.

Thanks!



[2005-08-19 11:46:40] [EMAIL PROTECTED]

If it works with Apache 1.3.33, it's no bug in PHP. Complain to Oracle.
Also, we have never said we support anything but the 'real' apache..




[2005-08-19 06:52:12] david dot marek at mattel dot com

It works with the build of Apache 1.3.33 but it still does not work
with Oracle's Apache.  We are using GCC and the dependencies have all
be recompiled.  When I build PHP on Solaris I don't get the symbol
referenced in the libphp5.so but on AIX I do.  Is there a reason the
make is added the symbol to the AIX build and not the Solaris build.
Everything else is the same.  I grep on ap_loaded_modules on libphp5.so
and I get a value back on AIX and that appears to be the culprit.



[2005-08-18 23:46:14] [EMAIL PROTECTED]

Let's see first if this works any better with real Apache.
Don't reply before you get results.




The remainder of the comments for this report are too long. To view
the rest of the comments, please view the bug report online at
http://bugs.php.net/34189

-- 
Edit this bug report at http://bugs.php.net/?id=34189&edit=1

#34213 [Opn->Bgs]: if a file exists , mkdir() can't creat a folder as same name

2005-08-22 Thread derick 
 ID:   34213
 Updated by:   [EMAIL PROTECTED]
 Reported By:  js1 at qq dot com
-Status:   Open
+Status:   Bogus
 Bug Type: Directory function related
 Operating System: windows2003+IIS6
 PHP Version:  5.0.4
 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.

That's your operating system preventing this!


Previous Comments:


[2005-08-22 21:47:08] js1 at qq dot com

Description:

file "D:/web/test1" exists,can`t creat a folder "D:/web/test1/"

Reproduce code:
---


Expected result:

creat a folder "D:/web/test1/"






-- 
Edit this bug report at http://bugs.php.net/?id=34213&edit=1

#34213 [NEW]: if a file exists , mkdir() can't creat a folder as same name

2005-08-22 Thread js1 at qq dot com 
From: js1 at qq dot com
Operating system: windows2003+IIS6
PHP version:  5.0.4
PHP Bug Type: Directory function related
Bug description:  if a file exists , mkdir() can't creat a folder as same name

Description:

file "D:/web/test1" exists,can`t creat a folder "D:/web/test1/"

Reproduce code:
---


Expected result:

creat a folder "D:/web/test1/"


-- 
Edit bug report at http://bugs.php.net/?id=34213&edit=1
-- 
Try a CVS snapshot (php4):   http://bugs.php.net/fix.php?id=34213&r=trysnapshot4
Try a CVS snapshot (php5.0): 
http://bugs.php.net/fix.php?id=34213&r=trysnapshot50
Try a CVS snapshot (php5.1): 
http://bugs.php.net/fix.php?id=34213&r=trysnapshot51
Fixed in CVS:http://bugs.php.net/fix.php?id=34213&r=fixedcvs
Fixed in release:http://bugs.php.net/fix.php?id=34213&r=alreadyfixed
Need backtrace:  http://bugs.php.net/fix.php?id=34213&r=needtrace
Need Reproduce Script:   http://bugs.php.net/fix.php?id=34213&r=needscript
Try newer version:   http://bugs.php.net/fix.php?id=34213&r=oldversion
Not developer issue: http://bugs.php.net/fix.php?id=34213&r=support
Expected behavior:   http://bugs.php.net/fix.php?id=34213&r=notwrong
Not enough info: 
http://bugs.php.net/fix.php?id=34213&r=notenoughinfo
Submitted twice: 
http://bugs.php.net/fix.php?id=34213&r=submittedtwice
register_globals:http://bugs.php.net/fix.php?id=34213&r=globals
PHP 3 support discontinued:  http://bugs.php.net/fix.php?id=34213&r=php3
Daylight Savings:http://bugs.php.net/fix.php?id=34213&r=dst
IIS Stability:   http://bugs.php.net/fix.php?id=34213&r=isapi
Install GNU Sed: http://bugs.php.net/fix.php?id=34213&r=gnused
Floating point limitations:  http://bugs.php.net/fix.php?id=34213&r=float
No Zend Extensions:  http://bugs.php.net/fix.php?id=34213&r=nozend
MySQL Configuration Error:   http://bugs.php.net/fix.php?id=34213&r=mysqlcfg

#34189 [Bgs->Opn]: After Configure and Make can load symbol ap_loaded_modules

2005-08-22 Thread david dot marek at mattel dot com 
 ID:   34189
 User updated by:  david dot marek at mattel dot com
 Reported By:  david dot marek at mattel dot com
-Status:   Bogus
+Status:   Open
 Bug Type: Apache related
 Operating System: AIX 5.0
 PHP Version:  5.1.0b3
 New Comment:

I am reopening as I have some questions you may be able to help with. I
hope!


Previous Comments:


[2005-08-22 16:35:25] david dot marek at mattel dot com

I do have question.  If we build this agains solaris all is good. 
Meaning we don't see the ap_Loaded_modules in the php library.  In AIX
we do.  This is the compile and build on the PHP side.  Is this
something we can remove or find a way not to include.  We can't move
forward with the migration without it so any help would be great.

Thanks!



[2005-08-19 11:46:40] [EMAIL PROTECTED]

If it works with Apache 1.3.33, it's no bug in PHP. Complain to Oracle.
Also, we have never said we support anything but the 'real' apache..




[2005-08-19 06:52:12] david dot marek at mattel dot com

It works with the build of Apache 1.3.33 but it still does not work
with Oracle's Apache.  We are using GCC and the dependencies have all
be recompiled.  When I build PHP on Solaris I don't get the symbol
referenced in the libphp5.so but on AIX I do.  Is there a reason the
make is added the symbol to the AIX build and not the Solaris build.
Everything else is the same.  I grep on ap_loaded_modules on libphp5.so
and I get a value back on AIX and that appears to be the culprit.



[2005-08-18 23:46:14] [EMAIL PROTECTED]

Let's see first if this works any better with real Apache.
Don't reply before you get results.




[2005-08-18 23:28:57] david dot marek at mattel dot com

Starting to do the apache standalone as a test but we need to use iAS
plus we have it working on solaris. Is there some special AIX stuff we
need to do in the configure.  Any env variables that would differ.  I
haven't been able to find anything.



The remainder of the comments for this report are too long. To view
the rest of the comments, please view the bug report online at
http://bugs.php.net/34189

-- 
Edit this bug report at http://bugs.php.net/?id=34189&edit=1

#34211 [NEW]: Allow for data type "TIMESTAMP(0) WITH LOCAL TIME ZONE"

2005-08-22 Thread kurtb149 at yahoo dot com 
From: kurtb149 at yahoo dot com
Operating system: Linux
PHP version:  6CVS-2005-08-22 (CVS)
PHP Bug Type: OCI8 related
Bug description:  Allow for data type "TIMESTAMP(0) WITH LOCAL TIME ZONE"

Description:

In the file:

   ext/pdo_oci/oci_statement.c

the function oci_stmt_describe() does not allow for the 
data type "TIMESTAMP(0) WITH LOCAL TIME ZONE". Here is the diff of an
updated oci_statement.c that would allow for the data type:

$ cvs diff oci_statement.c 
Index: oci_statement.c
===
RCS file: /repository/php-src/ext/pdo_oci/oci_statement.c,v
retrieving revision 1.16
diff -r1.16 oci_statement.c
407a408,410
> #ifdef SQLT_TIMESTAMP_LTZ
>   || dtype == SQLT_TIMESTAMP_LTZ
> #endif



-- 
Edit bug report at http://bugs.php.net/?id=34211&edit=1
-- 
Try a CVS snapshot (php4):   http://bugs.php.net/fix.php?id=34211&r=trysnapshot4
Try a CVS snapshot (php5.0): 
http://bugs.php.net/fix.php?id=34211&r=trysnapshot50
Try a CVS snapshot (php5.1): 
http://bugs.php.net/fix.php?id=34211&r=trysnapshot51
Fixed in CVS:http://bugs.php.net/fix.php?id=34211&r=fixedcvs
Fixed in release:http://bugs.php.net/fix.php?id=34211&r=alreadyfixed
Need backtrace:  http://bugs.php.net/fix.php?id=34211&r=needtrace
Need Reproduce Script:   http://bugs.php.net/fix.php?id=34211&r=needscript
Try newer version:   http://bugs.php.net/fix.php?id=34211&r=oldversion
Not developer issue: http://bugs.php.net/fix.php?id=34211&r=support
Expected behavior:   http://bugs.php.net/fix.php?id=34211&r=notwrong
Not enough info: 
http://bugs.php.net/fix.php?id=34211&r=notenoughinfo
Submitted twice: 
http://bugs.php.net/fix.php?id=34211&r=submittedtwice
register_globals:http://bugs.php.net/fix.php?id=34211&r=globals
PHP 3 support discontinued:  http://bugs.php.net/fix.php?id=34211&r=php3
Daylight Savings:http://bugs.php.net/fix.php?id=34211&r=dst
IIS Stability:   http://bugs.php.net/fix.php?id=34211&r=isapi
Install GNU Sed: http://bugs.php.net/fix.php?id=34211&r=gnused
Floating point limitations:  http://bugs.php.net/fix.php?id=34211&r=float
No Zend Extensions:  http://bugs.php.net/fix.php?id=34211&r=nozend
MySQL Configuration Error:   http://bugs.php.net/fix.php?id=34211&r=mysqlcfg

#22108 [Asn->WFx]: php doesn't ignore the utf-8 BOM

2005-08-22 Thread derick 
 ID:   22108
 Updated by:   [EMAIL PROTECTED]
 Reported By:  bugzilla at jellycan dot com
-Status:   Assigned
+Status:   Wont fix
 Bug Type: Feature/Change Request
 Operating System: *
 PHP Version:  *
 Assigned To:  moriyoshi
 New Comment:

This will come with Unicode support in PHP 6.0


Previous Comments:


[2005-08-22 18:32:38] jwagner at cc dot hut dot fi

PHP 5.0.4 for Windows /still/ does not seem to have it
(enable-zend-multibyte) enabled by default. For example session_start()
is broken for UTF-8 encoded php files. I would strongly suggest to make
enable-zend-multibyte a default for the windows release!



[2005-01-12 23:43:27] lapo at lapo dot it

> Is it possible to port this support for windows too?

Of course I quoted the wrong like, zend-multibyte support is POSSIBLE
(not DEFAULT) in the Windows version.



[2005-01-12 18:36:08] lapo at lapo dot it

> How about making this --enable-zend-multibyte default option?

It is already available on Windows. In fact, I'm using it on a
production server since june 2003, with no problems and with many
satisfactions.

Any reason this is still not in by default?
Someone else is encountering bugs with it?



[2005-01-06 21:08:05] [EMAIL PROTECTED]

How about making this --enable-zend-multibyte default option?
Is it possible to port this support for windows too?
And for 4.3.x branch?
Should it be marked open again?




[2004-05-25 12:33:30] lapo at lapo dot it

Adding '--enable-zend-multibyte' to latest PHP5 port for FreeBSD for
sure solves the problem:

All files contain:


cyberx [~] $ php /usr/tmp/utf8-bom.php 
à èéìòù
cyberx [~] $ php /usr/tmp/utf8Y-bom.php 
àèéìòù
cyberx [~] $ php /usr/tmp/utf16-bom.php 
àèéìòù
cyberx [~] $ php /usr/tmp/utf16BE-bom.php 
àèéìòù
cyberx [~] $ php /usr/tmp/utf16LE-bom.php 
àèéìòù

Except for "UTF8 without BOM" that is, of course, not distinguishable
from ISO8859-15 (default here), all theother formats are correctly
interpreted and outputted.
(notice that the 'header' instruction prior of the 'echo' one would
stutter with a non-BOM-aware PHP compile).

I wonder if and when this great multibyte support would be available by
default in Win32 compiles, I would really use it for work and am not
willing to but VisualC just to compile that ;-)
(though I'm trying compiling it with cygwin's gcc using '-mno-cygwin'
option, we'll see...)



The remainder of the comments for this report are too long. To view
the rest of the comments, please view the bug report online at
http://bugs.php.net/22108

-- 
Edit this bug report at http://bugs.php.net/?id=22108&edit=1

#22108 [Com]: php doesn't ignore the utf-8 BOM

2005-08-22 Thread jwagner at cc dot hut dot fi 
 ID:   22108
 Comment by:   jwagner at cc dot hut dot fi
 Reported By:  bugzilla at jellycan dot com
 Status:   Assigned
 Bug Type: Feature/Change Request
 Operating System: *
 PHP Version:  *
 Assigned To:  moriyoshi
 New Comment:

PHP 5.0.4 for Windows /still/ does not seem to have it
(enable-zend-multibyte) enabled by default. For example session_start()
is broken for UTF-8 encoded php files. I would strongly suggest to make
enable-zend-multibyte a default for the windows release!


Previous Comments:


[2005-01-12 23:43:27] lapo at lapo dot it

> Is it possible to port this support for windows too?

Of course I quoted the wrong like, zend-multibyte support is POSSIBLE
(not DEFAULT) in the Windows version.



[2005-01-12 18:36:08] lapo at lapo dot it

> How about making this --enable-zend-multibyte default option?

It is already available on Windows. In fact, I'm using it on a
production server since june 2003, with no problems and with many
satisfactions.

Any reason this is still not in by default?
Someone else is encountering bugs with it?



[2005-01-06 21:08:05] [EMAIL PROTECTED]

How about making this --enable-zend-multibyte default option?
Is it possible to port this support for windows too?
And for 4.3.x branch?
Should it be marked open again?




[2004-05-25 12:33:30] lapo at lapo dot it

Adding '--enable-zend-multibyte' to latest PHP5 port for FreeBSD for
sure solves the problem:

All files contain:


cyberx [~] $ php /usr/tmp/utf8-bom.php 
à èéìòù
cyberx [~] $ php /usr/tmp/utf8Y-bom.php 
àèéìòù
cyberx [~] $ php /usr/tmp/utf16-bom.php 
àèéìòù
cyberx [~] $ php /usr/tmp/utf16BE-bom.php 
àèéìòù
cyberx [~] $ php /usr/tmp/utf16LE-bom.php 
àèéìòù

Except for "UTF8 without BOM" that is, of course, not distinguishable
from ISO8859-15 (default here), all theother formats are correctly
interpreted and outputted.
(notice that the 'header' instruction prior of the 'echo' one would
stutter with a non-BOM-aware PHP compile).

I wonder if and when this great multibyte support would be available by
default in Win32 compiles, I would really use it for work and am not
willing to but VisualC just to compile that ;-)
(though I'm trying compiling it with cygwin's gcc using '-mno-cygwin'
option, we'll see...)



[2003-11-09 16:12:50] a9c83cd8bb41db324db5b449352f183 at arcor dot de

Thought about it... Now I think it's better when the BOM isn't part of
the output because that would cause trouble if you want to output
images or PDF or something like that...



The remainder of the comments for this report are too long. To view
the rest of the comments, please view the bug report online at
http://bugs.php.net/22108

-- 
Edit this bug report at http://bugs.php.net/?id=22108&edit=1

#34208 [Opn->Fbk]: crash using session_set_save_handler & throwing exception in own errorhandler

2005-08-22 Thread sniper 
 ID:   34208
 Updated by:   [EMAIL PROTECTED]
 Reported By:  carsten dot jendro at gmx dot net
-Status:   Open
+Status:   Feedback
 Bug Type: Reproducible crash
 Operating System: win32 only
 PHP Version:  5.1.0RC1
 New Comment:

Please try using this CVS snapshot:

  http://snaps.php.net/php5-latest.tar.gz
 
For Windows:
 
  http://snaps.php.net/win32/php5-win32-latest.zip




Previous Comments:


[2005-08-22 11:09:16] [EMAIL PROTECTED]

Does NOT crash under Linux. (ZTS problem propably)




[2005-08-22 01:39:04] carsten dot jendro at gmx dot net

Description:

php5/apache crashes when using set_error_handler and
session_set_save_handler together and throwing an exception in error
handler.

Testet with php 5.0.4.4 and 5.1rc1, apache 1 & 2

Reproduce code:
---
Bar();

?>

Expected result:

no crash ;-)

Actual result:
--
php5ts.dll!00a2b37a()   
php5ts.dll!00aa1248()   
php5ts.dll!00aa1494()   
php5ts.dll!00aa59fe()   
php5ts.dll!00aa587d()   
php5ts.dll!00aa5beb()   
php5ts.dll!009f640c()   
php5ts.dll!00a2e552()   
php5ts.dll!009f23bd()   
php5ts.dll!00a5244e()   
msvcrt.dll!77bfc2e3()   
ntwdblib.dll!7333ab22() 
ntwdblib.dll!7333f33e() 
ntwdblib.dll!7333f3eb() 
php5apache2.dll!003f2f9f()  
php5apache2.dll!003f34e8()  
php5apache2.dll!003f3163()  
libhttpd.dll!6ff0155f() 
libhttpd.dll!6ff018a9() 
libhttpd.dll!6ff0d582() 
libhttpd.dll!6ff095e6() 
libhttpd.dll!6ff0411f() 
libhttpd.dll!6ff04372() 
libhttpd.dll!6ff1bc3a() 
msvcrt.dll!77c0a3b0()   
kernel32.dll!7c80b50b() 
kernel32.dll!7c8399f3() 






-- 
Edit this bug report at http://bugs.php.net/?id=34208&edit=1

#33220 [Asn]: Denial of service through imagecreatefromgif()

2005-08-22 Thread sniper 
 ID:   33220
 Updated by:   [EMAIL PROTECTED]
 Reported By:  phpbugs at pureftpd dot org
 Status:   Assigned
 Bug Type: GD related
 Operating System: *
 PHP Version:  5CVS-2005-08-17
 Assigned To:  pajoye
 New Comment:

Pierre, are you going to do something for this bug?



Previous Comments:


[2005-06-02 15:11:16] phpbugs at pureftpd dot org

Description:

The following (invalid) picture triggers an infinite loop  
in gdImageCreateFromGif() with 100% CPU usage.  
  
Tested on Linux and OpenBSD, PHP4 with GD 1 and PHP5 with 
built-in GD. 

Reproduce code:
---


Get kaboom-gd.gif from ftp://ftp.00f.net/misc/kaboom-gd.gif


Expected result:

No 100% CPU usage. 

Actual result:
--
100% CPU usage, infinite loop. 





-- 
Edit this bug report at http://bugs.php.net/?id=33220&edit=1

#34175 [Opn->Csd]: debug_backtrace shows wrong arguments when called from an own error handler

2005-08-22 Thread sniper 
 ID:   34175
 Updated by:   [EMAIL PROTECTED]
 Reported By:  flummi at everymail dot net
-Status:   Open
+Status:   Closed
 Bug Type: Scripting Engine problem
 Operating System: Redhat Linux
 PHP Version:  5.0.4


Previous Comments:


[2005-08-22 16:23:06] flummi at everymail dot net

I just tried the latest development version (php5-200508221230) - here
it seems to work quite well, so the bug can be closed, as I can say.
Thanx!



[2005-08-17 18:59:45] [EMAIL PROTECTED]

Please try using this CVS snapshot:

  http://snaps.php.net/php5-latest.tar.gz
 
For Windows:
 
  http://snaps.php.net/win32/php5-win32-latest.zip





[2005-08-17 18:27:21] flummi at everymail dot net

Description:

When calling debug_backtrace from an own error handler, it provides a
different output than when calling it directly. As you can see in the
code, I have two calls to debug_backtrace, one from my error handler,
the other from a function. As I raise an error shortly after calling
debug_backtrace the first time, I would expect to have the same first
few entries in the debug_backtrace array. I.e. the last entry shows an
error on line 18, where the function 'a_test' is called with arg
'another friend'. In real this function is called with the arg 'friend'
(arg 'another friend' is used in the next function call.
I hope I could explain the bug good enough as my natural language is
german. :-)

Reproduce code:
---





Expected result:

Hi: friend
Hi another friendArray
(
[0] => Array
(
[file] => /var/www/html/xiview/test/testerror2.php
[line] => 10
[function] => b_test
[args] => Array
(
[0] => another friend
)

)

[1] => Array
(
[file] => /var/www/html/xiview/test/testerror2.php
[line] => 18
[function] => a_test
[args] => Array
(
[0] => friend
)

)

)
Array
(
[0] => Array
(
[file] => /var/www/html/xiview/test/testerror2.php
[line] => 16
[function] => HandleError
)

[1] => Array
(
[file] => /var/www/html/xiview/test/testerror2.php
[line] => 16
[function] => b_test
)

[2] => Array
(
[file] => /var/www/html/xiview/test/testerror2.php
[line] => 10
[function] => b_test
[args] => Array
(
[str2] => another friend
)

)

[3] => Array
(
[file] => /var/www/html/xiview/test/testerror2.php
[line] => 18
[function] => a_test
[args] => Array
(
[0] => friend
)

)

)

Actual result:
--
Hi: friend
Hi another friendArray
(
[0] => Array
(
[file] => /var/www/html/xiview/test/testerror2.php
[line] => 10
[function] => b_test
[args] => Array
(
[0] => another friend
)

)

[1] => Array
(
[file] => /var/www/html/xiview/test/testerror2.php
[line] => 18
[function] => a_test
[args] => Array
(
[0] => friend
)

)

)
Array
(
[0] => Array
(
[file] => /var/www/html/xiview/test/testerror2.php
[line] => 16
[function] => HandleError
)

[1] => Array
(
[file] => /var/www/html/xiview/test/testerror2.php
[line] => 16
[function] => b_test
)

[2] => Array
(
[file] => /var/www/html/xiview/test/testerror2.php
[line] => 10
[function] => b_test
[args] => Array
(
[0] => 2
[1] => Division by zero
[2] => /var/www/html/xiview/test/testerror2.php
[3] => 16
[4] => Array
(
[str2] => another friend
)

)

)

[3] => Array
(
[file] => /var/www/html/xiview/test/testerror2.php
[line] => 18
[function] => a_test
[args] => Array
(
[0] => another friend
)

)

)





-- 
Edit this bug report at http://bu

#34189 [Bgs]: After Configure and Make can load symbol ap_loaded_modules

2005-08-22 Thread david dot marek at mattel dot com 
 ID:   34189
 User updated by:  david dot marek at mattel dot com
 Reported By:  david dot marek at mattel dot com
 Status:   Bogus
 Bug Type: Apache related
 Operating System: AIX 5.0
 PHP Version:  5.1.0b3
 New Comment:

I do have question.  If we build this agains solaris all is good. 
Meaning we don't see the ap_Loaded_modules in the php library.  In AIX
we do.  This is the compile and build on the PHP side.  Is this
something we can remove or find a way not to include.  We can't move
forward with the migration without it so any help would be great.

Thanks!


Previous Comments:


[2005-08-19 11:46:40] [EMAIL PROTECTED]

If it works with Apache 1.3.33, it's no bug in PHP. Complain to Oracle.
Also, we have never said we support anything but the 'real' apache..




[2005-08-19 06:52:12] david dot marek at mattel dot com

It works with the build of Apache 1.3.33 but it still does not work
with Oracle's Apache.  We are using GCC and the dependencies have all
be recompiled.  When I build PHP on Solaris I don't get the symbol
referenced in the libphp5.so but on AIX I do.  Is there a reason the
make is added the symbol to the AIX build and not the Solaris build.
Everything else is the same.  I grep on ap_loaded_modules on libphp5.so
and I get a value back on AIX and that appears to be the culprit.



[2005-08-18 23:46:14] [EMAIL PROTECTED]

Let's see first if this works any better with real Apache.
Don't reply before you get results.




[2005-08-18 23:28:57] david dot marek at mattel dot com

Starting to do the apache standalone as a test but we need to use iAS
plus we have it working on solaris. Is there some special AIX stuff we
need to do in the configure.  Any env variables that would differ.  I
haven't been able to find anything.



[2005-08-18 23:07:25] [EMAIL PROTECTED]

Try with stock apache 1.3.33 installation, not with some Oracle
crap..also, use GCC as compiler.





The remainder of the comments for this report are too long. To view
the rest of the comments, please view the bug report online at
http://bugs.php.net/34189

-- 
Edit this bug report at http://bugs.php.net/?id=34189&edit=1

#34175 [Fbk->Opn]: debug_backtrace shows wrong arguments when called from an own error handler

2005-08-22 Thread flummi at everymail dot net 
 ID:   34175
 User updated by:  flummi at everymail dot net
 Reported By:  flummi at everymail dot net
-Status:   Feedback
+Status:   Open
 Bug Type: Scripting Engine problem
 Operating System: Redhat Linux
 PHP Version:  5.0.4
 New Comment:

I just tried the latest development version (php5-200508221230) - here
it seems to work quite well, so the bug can be closed, as I can say.
Thanx!


Previous Comments:


[2005-08-17 18:59:45] [EMAIL PROTECTED]

Please try using this CVS snapshot:

  http://snaps.php.net/php5-latest.tar.gz
 
For Windows:
 
  http://snaps.php.net/win32/php5-win32-latest.zip





[2005-08-17 18:27:21] flummi at everymail dot net

Description:

When calling debug_backtrace from an own error handler, it provides a
different output than when calling it directly. As you can see in the
code, I have two calls to debug_backtrace, one from my error handler,
the other from a function. As I raise an error shortly after calling
debug_backtrace the first time, I would expect to have the same first
few entries in the debug_backtrace array. I.e. the last entry shows an
error on line 18, where the function 'a_test' is called with arg
'another friend'. In real this function is called with the arg 'friend'
(arg 'another friend' is used in the next function call.
I hope I could explain the bug good enough as my natural language is
german. :-)

Reproduce code:
---





Expected result:

Hi: friend
Hi another friendArray
(
[0] => Array
(
[file] => /var/www/html/xiview/test/testerror2.php
[line] => 10
[function] => b_test
[args] => Array
(
[0] => another friend
)

)

[1] => Array
(
[file] => /var/www/html/xiview/test/testerror2.php
[line] => 18
[function] => a_test
[args] => Array
(
[0] => friend
)

)

)
Array
(
[0] => Array
(
[file] => /var/www/html/xiview/test/testerror2.php
[line] => 16
[function] => HandleError
)

[1] => Array
(
[file] => /var/www/html/xiview/test/testerror2.php
[line] => 16
[function] => b_test
)

[2] => Array
(
[file] => /var/www/html/xiview/test/testerror2.php
[line] => 10
[function] => b_test
[args] => Array
(
[str2] => another friend
)

)

[3] => Array
(
[file] => /var/www/html/xiview/test/testerror2.php
[line] => 18
[function] => a_test
[args] => Array
(
[0] => friend
)

)

)

Actual result:
--
Hi: friend
Hi another friendArray
(
[0] => Array
(
[file] => /var/www/html/xiview/test/testerror2.php
[line] => 10
[function] => b_test
[args] => Array
(
[0] => another friend
)

)

[1] => Array
(
[file] => /var/www/html/xiview/test/testerror2.php
[line] => 18
[function] => a_test
[args] => Array
(
[0] => friend
)

)

)
Array
(
[0] => Array
(
[file] => /var/www/html/xiview/test/testerror2.php
[line] => 16
[function] => HandleError
)

[1] => Array
(
[file] => /var/www/html/xiview/test/testerror2.php
[line] => 16
[function] => b_test
)

[2] => Array
(
[file] => /var/www/html/xiview/test/testerror2.php
[line] => 10
[function] => b_test
[args] => Array
(
[0] => 2
[1] => Division by zero
[2] => /var/www/html/xiview/test/testerror2.php
[3] => 16
[4] => Array
(
[str2] => another friend
)

)

)

[3] => Array
(
[file] => /var/www/html/xiview/test/testerror2.php
[line] => 18
[function] => a_test
[args] => Array
(
[0] => another friend
)

)

)





-- 
Edit this bug report at http://bugs.php.net/?id=34175&edit=1

#34210 [NEW]: Feature to run function inline within the local scope

2005-08-22 Thread jdv at foobar dot be 
From: jdv at foobar dot be
Operating system: Linux
PHP version:  5.0.4
PHP Bug Type: Feature/Change Request
Bug description:  Feature to run function inline within the local scope

Description:

It would be nice to see a feature that allowed for functions to be
executed, inline, within the current local scope. The behaviour I intend
for is that of eval(). Calling eval($code) allows $code to access
variables from the local scope from which eval is called. It also allows
$code to set variables in this local scope. Both features, if accessible
to user-defined functions, would come in handy on occasions.

One particular useful function that could be built if this functionality
existed, would accept a string for argument and replace all %foo
occurences in it by the -- according to context -- properly escaped value
of $foo. Such a function could then be used for composing HTML and SQL
strings more easily.


-- 
Edit bug report at http://bugs.php.net/?id=34210&edit=1
-- 
Try a CVS snapshot (php4):   http://bugs.php.net/fix.php?id=34210&r=trysnapshot4
Try a CVS snapshot (php5.0): 
http://bugs.php.net/fix.php?id=34210&r=trysnapshot50
Try a CVS snapshot (php5.1): 
http://bugs.php.net/fix.php?id=34210&r=trysnapshot51
Fixed in CVS:http://bugs.php.net/fix.php?id=34210&r=fixedcvs
Fixed in release:http://bugs.php.net/fix.php?id=34210&r=alreadyfixed
Need backtrace:  http://bugs.php.net/fix.php?id=34210&r=needtrace
Need Reproduce Script:   http://bugs.php.net/fix.php?id=34210&r=needscript
Try newer version:   http://bugs.php.net/fix.php?id=34210&r=oldversion
Not developer issue: http://bugs.php.net/fix.php?id=34210&r=support
Expected behavior:   http://bugs.php.net/fix.php?id=34210&r=notwrong
Not enough info: 
http://bugs.php.net/fix.php?id=34210&r=notenoughinfo
Submitted twice: 
http://bugs.php.net/fix.php?id=34210&r=submittedtwice
register_globals:http://bugs.php.net/fix.php?id=34210&r=globals
PHP 3 support discontinued:  http://bugs.php.net/fix.php?id=34210&r=php3
Daylight Savings:http://bugs.php.net/fix.php?id=34210&r=dst
IIS Stability:   http://bugs.php.net/fix.php?id=34210&r=isapi
Install GNU Sed: http://bugs.php.net/fix.php?id=34210&r=gnused
Floating point limitations:  http://bugs.php.net/fix.php?id=34210&r=float
No Zend Extensions:  http://bugs.php.net/fix.php?id=34210&r=nozend
MySQL Configuration Error:   http://bugs.php.net/fix.php?id=34210&r=mysqlcfg

#17609 [Opn->Csd]: Ingres config file only works in unix-text-mode

2005-08-22 Thread grantc 
 ID:   17609
 Updated by:   [EMAIL PROTECTED]
 Reported By:  joerg dot schoppet at web dot de
-Status:   Open
+Status:   Closed
 Bug Type: Feature/Change Request
 Operating System: Win2k
 PHP Version:  4.2.1
 New Comment:

this is ingres bug 110563 - fixed with Ingres r3. If you require a fix
for an earlier supported release please contact your local support
office.


Previous Comments:


[2002-06-05 08:21:24] joerg dot schoppet at web dot de

Hi all,

I use the ingres-extension under win2k. When I include the
php_ingres.dll in the php.ini file and start php, it always reports a
"trashed symbol.tbl" and php hangs up. So I have to convert this file,
which is found in %II_SYSTEM%\ingres\files from dos-format to
unix-format. Then everything works fine. But after restarting the pc or
the database-service the symbol.tbl-file is set to dos-format and I have
to do the conversion again.

So I think there's a problem with the unix "lf" and the windows
"crlf".

Could that be fixed?

Thanxs in advance

Joerg Schoppet




-- 
Edit this bug report at http://bugs.php.net/?id=17609&edit=1

#29840 [Com]: is_executable does not honor the safe_mode_exec_dir setting

2005-08-22 Thread fogbank at fogbank dot org 
 ID:   29840
 Comment by:   fogbank at fogbank dot org
 Reported By:  markus at cultcom dot de
 Status:   No Feedback
 Bug Type: Filesystem function related
 Operating System: *
 PHP Version:  5CVS, 4CVS (2005-01-04)
 Assigned To:  tony2001
 New Comment:

This bug is present in version 4.4.0, too.
Please, fix it, since right now it breaks every package in PEAR that
relies on System::which().


Previous Comments:


[2005-08-20 01:00:03] php-bugs at lists dot php dot net

No feedback was provided for this bug for over a week, so it is
being suspended automatically. If you are able to provide the
information that was originally requested, please do so and change
the status of the bug back to "Open".



[2005-08-12 01:00:33] [EMAIL PROTECTED]

Please try this patch:
http://tony2001.phpclub.net/dev/tmp/bugs_29840_31618.diff
(with the latest snapshot/CVS).



[2004-12-12 11:36:05] [EMAIL PROTECTED]

Jani, I sent a patch for this problem to Wez long ago. 
The problem is that is_executable() indeed doesn't respect
safe_mode_exec_dir as it should, so I'm assigning this to Wez.



[2004-12-12 02:07:16] [EMAIL PROTECTED]

It works just fine. (it returns false always when in safe-mode..)




[2004-08-25 18:06:12] markus at cultcom dot de

Description:

Seems to be a common problem nobody complains about...
"is_executable()" does not work in safe_mode!

Some PHP-Projects check for sendmail using this function and don't work
in safe_mode even if sendmail acutally IS executable (i.e. PEAR:
Mail.php).

is_executable() should at least honor the safe_mode_exec_dir directive!

Reproduce code:
---


Try with PHP/CGI and suexec + safe_mode
where example-UID != sendmail-UID


Expected result:

true, what else?

Actual result:
--
false.





-- 
Edit this bug report at http://bugs.php.net/?id=29840&edit=1

#34208 [Opn]: crash using session_set_save_handler & throwing exception in own errorhandler

2005-08-22 Thread sniper 
 ID:   34208
 Updated by:   [EMAIL PROTECTED]
 Reported By:  carsten dot jendro at gmx dot net
 Status:   Open
 Bug Type: Reproducible crash
-Operating System: Windows
+Operating System: win32 only
 PHP Version:  5.1.0RC1
 New Comment:

Does NOT crash under Linux. (ZTS problem propably)



Previous Comments:


[2005-08-22 01:39:04] carsten dot jendro at gmx dot net

Description:

php5/apache crashes when using set_error_handler and
session_set_save_handler together and throwing an exception in error
handler.

Testet with php 5.0.4.4 and 5.1rc1, apache 1 & 2

Reproduce code:
---
Bar();

?>

Expected result:

no crash ;-)

Actual result:
--
php5ts.dll!00a2b37a()   
php5ts.dll!00aa1248()   
php5ts.dll!00aa1494()   
php5ts.dll!00aa59fe()   
php5ts.dll!00aa587d()   
php5ts.dll!00aa5beb()   
php5ts.dll!009f640c()   
php5ts.dll!00a2e552()   
php5ts.dll!009f23bd()   
php5ts.dll!00a5244e()   
msvcrt.dll!77bfc2e3()   
ntwdblib.dll!7333ab22() 
ntwdblib.dll!7333f33e() 
ntwdblib.dll!7333f3eb() 
php5apache2.dll!003f2f9f()  
php5apache2.dll!003f34e8()  
php5apache2.dll!003f3163()  
libhttpd.dll!6ff0155f() 
libhttpd.dll!6ff018a9() 
libhttpd.dll!6ff0d582() 
libhttpd.dll!6ff095e6() 
libhttpd.dll!6ff0411f() 
libhttpd.dll!6ff04372() 
libhttpd.dll!6ff1bc3a() 
msvcrt.dll!77c0a3b0()   
kernel32.dll!7c80b50b() 
kernel32.dll!7c8399f3() 






-- 
Edit this bug report at http://bugs.php.net/?id=34208&edit=1