#28740 [Com]: PHP has encountered an Access Violation
ID: 28740 Comment by: sekmun at mypostbox dot com Reported By: colorprint at mail dot ru Status: Bogus Bug Type: IIS related Operating System: Windows Server 2003 PHP Version: 4.3.7 New Comment: upgraded from 4.3.6 to 4.3.7, platform: win2k server/IIS (ISAPI)/Mysql Seeing this PHP Access Violation error now intermittently. Didn't use to get his problem. Might go back to 4.3.6 for the time being. Previous Comments: [2004-06-20 09:21:59] webmaster at yes-hosting dot com I get the same problem. I'm using win2k3/iis6 upgrade from php 4.3.4 to 4.3.7 error show at the end of php pages. [2004-06-16 15:44:51] jmertic at ncscredit dot com We have the same problem with our IIS 5/Win2k/MySQL. This is a 4.3.7 regression that needs to be addressed. I'm moving us back to 4.3.6. [2004-06-14 11:02:08] orangeasklessapple at cherrygmx dot net We just upgraded from php 4.3.4 to 4.3.7 and are having the same problem under Windows 2000, IIS 5.0, MS SQL Server 2000. A reset of the iis solves the problem just temporarliy. [2004-06-11 09:10:59] [EMAIL PROTECTED] 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. Thank you for your interest in PHP. . [2004-06-11 08:37:17] colorprint at mail dot ru Description: I get error at the end of all php-pages very often: PHP has encountered an Access Violation. This error is only when PHP 4.3.7 or 5.0 RC3 installed. With PHP 4.3.6 or 5.0 RC2 all right. Any idea how to fix it? -- Edit this bug report at http://bugs.php.net/?id=28740edit=1
#28788 [Opn]: Bogus form data when forms contains a single input field
ID: 28788 User updated by: lsjoberg at aland dot net Reported By: lsjoberg at aland dot net Status: Open Bug Type: Apache2 related Operating System: Redhat Enterprise Linux PHP Version: 5.0.0RC3 New Comment: I created a simple test page (http://www.vallnet.se/test.htm) which posts forms to both a PHP script which executes print_r($_POST); and a CGI script which prints its stdin. When run on a webserver with PHP as a CGI everything executes as expected. The CGI-script prints 'foo=5' and 'foo=5bar=5' and the PHP-script '[foo] = 5' and '[foo] = 5'/'[bar] = 5'. The CGI output from the Apache server is the same, but for the form with a single element the PHP output is Array ( [foo] = 5foo=5 ) The PHP output for the 2-field form is however correct, hence it seems to be some part of the apache/php interface that is failing... The same browser(s) and exactly the same scripts were used during testing. Previous Comments: [2004-06-19 11:56:19] lsjoberg at aland dot net The original code was fully W3C compliant, the testcase below was stripped down before submitted. Browsers used was Mozilla and Opera under FreeBSD and Internet Explorer in some Windows flavour. It feels like a browser bug but since it occured when upgrading from PHP4 to PHP5 without code changes it seems suspicious. I will do some more testing during next week to see if I can isolate the problem further. [2004-06-18 20:33:52] imprestavel at gameguru dot com dot br It looks more like a browser issue. Have you tryed writing conformant html? I tested with Mozilla/rv1.6 and IE/6.0 on Apache/2.0.49, PHP/5.0.0RC3 (register_globals=On and Off), under Windows2000 and didn't have that problem (with the same no- html). What browser are you using? [2004-06-15 11:31:36] lsjoberg at aland dot net Description: When a HTML form contains a single input field (this could be plain INPUT, HIDDEN, SELECT or just about any time) with an associated value the value is not passed correctly to the receiving PHP script. This only happends when the form contains exactly one field and when PHP is run as an Apache2 module. (Could not reproduce with PHP running as CGI under caudium). Reproduce code: --- test.htm: htmlbody form method=POST action=test.php input name=id value=5 input type=submit /form/body/html test.php: ? echo $id; ? Expected result: The expected result is that the php script should print a 5 and nothing else. (We are using register_globals=On). Actual result: -- The output of the php-script is '5id=5' instead of '5'. If a input type=hidden name=foo is added to the html form the correct output of 5 is displayed. -- Edit this bug report at http://bugs.php.net/?id=28788edit=1
#28863 [Opn]: Weird Bug with imagepng (including special chars)
ID: 28863 User updated by: jsgoupil at lookstrike dot com Reported By: jsgoupil at lookstrike dot com Status: Open Bug Type: GD related Operating System: WinXP PHP Version: 5.0.0RC3 New Comment: the reproduce code... supposed to be the same test file... so Open test1.php... and include(test1.php); :) Previous Comments: [2004-06-21 08:53:01] jsgoupil at lookstrike dot com Description: You will say that i'm really weird to do that, but i don't know how that happen : I check into one file, i found the following chars : chr(239) + chr(187) + chr(191) Keep the order ! it's important ! (I found that in one of my file, but dont ask me why !) If you try to include this kind of file before doing a PNG image, you will get a WRONG png format. (Starting with ?PNG...) The chars above can't be read under Notepad, neither under IE (perhaps because i'm under WinXP). It can't be deleted from NotePad (if you clear the file completly, the file will still have 3bytes undeletable) but you can delete it with wordpad... I search really a lot of time to found that problem... because I was including a file that had that kind of chars... Reproduce code: --- VBScript to put chr into file Open test1.php For Output As #1 Print #1, Chr(239) + Chr(187) + Chr(191) Close #1 PHP Script : ?php include(test3.php); $im = imagecreate(1024, 1024) or die(Can't Initialize the GD Libraty); imagecolorallocate($im,255,255,255); imagepng($im); ? Expected result: Supposed return some chars in IE and after the PNG format like : PNG . Because normally, if you do a echo a; imagepng($im); You will receive the a and after the PNG format in text. Actual result: -- Under WinXP I receive directly the wrong PNG format... like : ?PNG IHDRE#1273;?PLTE?#283;#512;?IDATx?#53313;? [EMAIL PROTECTED]#61828; ?!#1190;IEND?B`? (Without ) -- Edit this bug report at http://bugs.php.net/?id=28863edit=1
#28864 [NEW]: function to check if POST data is complete
From: xuefer at 21cn dot com Operating system: all PHP version: 4.3.7 PHP Bug Type: Feature/Change Request Bug description: function to check if POST data is complete Description: due to network issue, or stopped by user the content submitted by user maybe incomplete; there should be a way to check it. it's serious if the admins submit a incomplete post data. that will do unrecoverable damage. i'd like if there's a function: int content_is_complete() which return strlen(content) == Content-Length; yes, we can do 'input type=hidden value=1 name=complete/form', and do isset($_POST['complete']) but this is not easy for those html-designers to add it each form by hand 'input type=submit name=submit', and do isset($_POST['submit']), but not all submit is the last element of form -- Edit bug report at http://bugs.php.net/?id=28864edit=1 -- Try a CVS snapshot (php4): http://bugs.php.net/fix.php?id=28864r=trysnapshot4 Try a CVS snapshot (php5): http://bugs.php.net/fix.php?id=28864r=trysnapshot5 Fixed in CVS: http://bugs.php.net/fix.php?id=28864r=fixedcvs Fixed in release: http://bugs.php.net/fix.php?id=28864r=alreadyfixed Need backtrace: http://bugs.php.net/fix.php?id=28864r=needtrace Need Reproduce Script: http://bugs.php.net/fix.php?id=28864r=needscript Try newer version: http://bugs.php.net/fix.php?id=28864r=oldversion Not developer issue:http://bugs.php.net/fix.php?id=28864r=support Expected behavior: http://bugs.php.net/fix.php?id=28864r=notwrong Not enough info:http://bugs.php.net/fix.php?id=28864r=notenoughinfo Submitted twice:http://bugs.php.net/fix.php?id=28864r=submittedtwice register_globals: http://bugs.php.net/fix.php?id=28864r=globals PHP 3 support discontinued: http://bugs.php.net/fix.php?id=28864r=php3 Daylight Savings: http://bugs.php.net/fix.php?id=28864r=dst IIS Stability: http://bugs.php.net/fix.php?id=28864r=isapi Install GNU Sed:http://bugs.php.net/fix.php?id=28864r=gnused Floating point limitations: http://bugs.php.net/fix.php?id=28864r=float
#28865 [NEW]: Functions ftp_*list
From: 336764 at 163 dot com Operating system: MandRake 10 PHP version: 5.0.0RC3 PHP Bug Type: FTP related Bug description: Functions ftp_*list Description: Warning: ftp_nlist() [function.ftp-nlist]: Unable to create temporary file. Check permissions in temporary files directory. in /web/browse.inc.php on line 181 Warning: ftp_rawlist() [function.ftp-nlist]: Unable to create temporary file. Check permissions in temporary files directory. in /web/browse.inc.php on line 181 /tmp is chmod 777 and other temporary directory is 777 too. Reproduce code: --- $rawlist = ftp_rawlist($conn_id, -a); $rawlist = ftp_rawlist($conn_id, ); $rawlist = ftp_nlist($conn_id, -a); $rawlist = ftp_nlist($conn_id, ); Expected result: Warning: ftp_nlist() [function.ftp-nlist]: Unable to create temporary file. Check permissions in temporary files directory. /web/browse.inc.php on line 181 Actual result: -- list all files to array. -- Edit bug report at http://bugs.php.net/?id=28865edit=1 -- Try a CVS snapshot (php4): http://bugs.php.net/fix.php?id=28865r=trysnapshot4 Try a CVS snapshot (php5): http://bugs.php.net/fix.php?id=28865r=trysnapshot5 Fixed in CVS: http://bugs.php.net/fix.php?id=28865r=fixedcvs Fixed in release: http://bugs.php.net/fix.php?id=28865r=alreadyfixed Need backtrace: http://bugs.php.net/fix.php?id=28865r=needtrace Need Reproduce Script: http://bugs.php.net/fix.php?id=28865r=needscript Try newer version: http://bugs.php.net/fix.php?id=28865r=oldversion Not developer issue:http://bugs.php.net/fix.php?id=28865r=support Expected behavior: http://bugs.php.net/fix.php?id=28865r=notwrong Not enough info:http://bugs.php.net/fix.php?id=28865r=notenoughinfo Submitted twice:http://bugs.php.net/fix.php?id=28865r=submittedtwice register_globals: http://bugs.php.net/fix.php?id=28865r=globals PHP 3 support discontinued: http://bugs.php.net/fix.php?id=28865r=php3 Daylight Savings: http://bugs.php.net/fix.php?id=28865r=dst IIS Stability: http://bugs.php.net/fix.php?id=28865r=isapi Install GNU Sed:http://bugs.php.net/fix.php?id=28865r=gnused Floating point limitations: http://bugs.php.net/fix.php?id=28865r=float
#28435 [Com]: array_count_values() problem
ID: 28435
Comment by: vladb at pseudo-infinity dot ro
Reported By: ofirin at yahoo dot com
Status: Open
Bug Type: Arrays related
Operating System: Any
PHP Version: 4CVS-2004-05-18 (stable)
New Comment:
The php5 result seems more accurate than the php4 one. You got strings
in the $books array, no?
Changing this in php4 may break some code out there so maybe it'd be
best to just mention it as a 4-5 inconsistency.
Previous Comments:
[2004-06-18 14:24:41] programmer at bardware dot de
I noticed this problem with PHP5 RC2 on a Win2k box as an
Apache2-module. I have some HMTL-Checkboxes and on the server side want
to check the selected values. My checkboxes are equally named
name=choice2[] what lets PHP generate an array
$_POST[choice2][0]
$_POST[choice2][1]
$_POST[choice2][2] etc. according to the selected values.
Each checkbox has it's unique value attribute.
To test for a certain value I reverse this array with
$arrTmp=array_count_values($_POST[choice2]);
I now want to access $arrTmp[1] to check if the checkbox with the
attribute value=1 was selected. The respective value was - if
selected - 1, otherwise it's not existent in the
$_POST[choice2]-Array. This did not work. $arrTmp[1] didn't work
either. It was no way possible to access a member of the array.
The other poster mentioned the indexes are generated as strings, I want
to point they cannot be accessed at all.
It worked best on PHP 4.3.6
Best,
Bernhard
[2004-05-18 19:04:43] ofirin at yahoo dot com
Description:
I don't know if this is a php5 or php4 bug, but I'm sure something's
wrong here.
Whenever in php5 I do an array_count_values() on an array that
contains numeric values as strings the result array uses string keys
instead of the numeric values as indexes. This doesn't happen in php4.
I don't know what behaivor is correct.
I'm using the latest cvs versions of both php4 and php5 btw.
Reproduce code:
---
?php
$books = Array('10', '3', '6', '10');
$quantities = array_count_values($books);
var_dump($books);
var_dump($quantities);
?
Expected result:
This is what happens in php4:
array(4) {
[0]=
string(2) 10
[1]=
string(1) 3
[2]=
string(1) 6
[3]=
string(2) 10
}
array(3) {
[10]=
int(2)
[3]=
int(1)
[6]=
int(1)
}
Actual result:
--
This is what happens if php5:
array(4) {
[0]=
string(2) 10
[1]=
string(1) 3
[2]=
string(1) 6
[3]=
string(2) 10
}
array(3) {
[10]=
int(2)
[3]=
int(1)
[6]=
int(1)
}
--
Edit this bug report at http://bugs.php.net/?id=28435edit=1
#28113 [Com]: Memory is not freed with apache2handler SAPI
ID: 28113
Comment by: alex dot pagnoni at solarix dot it
Reported By: [EMAIL PROTECTED]
Status: Open
Bug Type: Apache2 related
Operating System: Linux
PHP Version: 5CVS-2004-04-23 (dev)
New Comment:
I'm experiencing huge memory leaks on Apache1 too.
Previous Comments:
[2004-04-23 00:33:00] [EMAIL PROTECTED]
Description:
PHP does not free the memory it's using during and after a request when
using apache2handler SAPI. apache2filter SAPI works fine.
The memory it's using is however reused, but if I would send 500MB data
and only have 256MB RAM, the OOM killer would start killing processes
and apache when there are no more processes to kill (depending on OOM
killer used).
Reproduce code:
---
?php
$string = '';
for ($i = 0; $i 1024; $i++)
$string .= 'A';
for ($i = 0; $i (1024 * 200); $i++) {
echo $string;
flush();
}
?
Expected result:
Low memory usage
Actual result:
--
PID USER PR NI VIRT RES SHR S %CPU %MEMTIME+ COMMAND
8434 nobody25 0 284m 238m 52m S 0.0 31.6 0:25.14 httpd
--
Edit this bug report at http://bugs.php.net/?id=28113edit=1
#28335 [Com]: wddx_serialize_value returns empty string
ID: 28335
Comment by: clemens at gutweiler dot net
Reported By: antonr at game dot permonline dot ru
Status: Open
Bug Type: WDDX related
Operating System: Win32
PHP Version: 5CVS-2004-05-09 (dev)
New Comment:
This example segfaults with PHP-5-RC3.
It contains recursive objects and should be solved like the 'r:123'-way
in serialize().
Previous Comments:
[2004-05-09 09:16:16] antonr at game dot permonline dot ru
Description:
In some cases wddx_serialize_value returns empty string.
Reproduce code:
---
?php
class MyClass1
{ public $myclass2;
}
class MyClass2
{ public $myclass1;
}
$a = new MyClass1;
$a-number = 1;
$a-myclass2 = new MyClass2;
$a-myclass2-myclass1 = $a;
echo wddx_serialize_value($a);
?
Actual result:
--
empty string
--
Edit this bug report at http://bugs.php.net/?id=28335edit=1
#20110 [Com]: input in flex scanner failed in directory :)
ID: 20110 Comment by: bugme at gmx dot net Reported By: jeroen at unfix dot org Status: Closed Bug Type: Scripting Engine problem Operating System: Linux, Net|Free|OpenBSD, others? PHP Version: 4.2.3 Assigned To: derick New Comment: I recived the bug when i did an include after a db-query. After that i begann whit the query and then the include which would work. I'dont know why but maybe this helps to find the bug. Previous Comments: [2003-11-26 23:27:33] alan at bullmarket dot com dot au Hi, I have received that type of error (2 actually). Please go to http://www.bullmarket.com.au/affiliates/user/signup.php I am very new to PHP and I am having trouble isolating the issue. Any help you can offer would be great. Cheers, Alan [2003-08-06 09:02:36] rsemmens at names dot co dot uk We get the problem intermittently. We run our websites from NFS and if the file read fails it comes up with this message. Only when doing includes though. I agree with the better error message comment :) Btw, we are using PHP 4.3.2 so it's still a current problem. Drop me an e-mail if you want any further details. [2003-03-20 16:59:41] bostjan dot skufca at domenca dot com you can view the bug at: http://www.avtomobilizem.com/dev/modules/gallery2 this project is using Smarty template engine and bug only occurs when i set $smarty-force_compile = false; email me to provide you relevant source files [2002-12-24 20:27:44] gunnar at start dot no This error also appears with other functions. More specifically I've experienced it with the parse_ini_file() function. I do not have access to test it with PHP CVS version, do I don't know wether it has been fixed for just the include function, or all functions affected by this problem. [2002-10-27 18:28:18] [EMAIL PROTECTED] This bug has been fixed in CVS. In case this was a PHP problem, snapshots of the sources are packaged every three hours; this change will be in the next snapshot. You can grab the snapshot at http://snaps.php.net/. In case this was a documentation problem, the fix will show up soon at http://www.php.net/manual/. In case this was a PHP.net website problem, the change will show up on the PHP.net site and on the mirror sites in short time. Thank you for the report, and for helping us make PHP better. 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/20110 -- Edit this bug report at http://bugs.php.net/?id=20110edit=1
#28866 [NEW]: array of objs filling slowness
From: dmirand at abelia-decors dot com
Operating system: Linux 2.4 / glibc 2.3
PHP version: 5.0.0RC3
PHP Bug Type: Performance problem
Bug description: array of objs filling slowness
Description:
When running a portion of script which fills an array with
objects, it is easy to notice a significant slowness depending
on what has already run before in the script., even if that
pre-processing is totally independant . The more load that
runs before, the slower the filling will be...
Under 4.3.6 almost no differences between :
- a just filling script
- a big load followed by a filling part
Both 4.3.6 and 5.0.0 RC3 compiled from source.
Reproduce code:
---
$big_load = new BigLoad ;
$big_load-go() ;
unset( $big_load ) ;
/* Filling start */
$arr_obj_orders = array() ;
foreach( $arr_no_order as $no_order )
{
$obj_order = new Order ;
$obj_order-load( $no_order ) ;
// to show filling avancement
echo $no_order ;
$arr_obj_orders[$no_order] = $obj_order ;
}
/* Filling end */
Expected result:
The expected behavior is of course no slowness with the
filling part of the script, ie the same behavior as if there was
no big load before the filling part.
--
Edit bug report at http://bugs.php.net/?id=28866edit=1
--
Try a CVS snapshot (php4): http://bugs.php.net/fix.php?id=28866r=trysnapshot4
Try a CVS snapshot (php5): http://bugs.php.net/fix.php?id=28866r=trysnapshot5
Fixed in CVS: http://bugs.php.net/fix.php?id=28866r=fixedcvs
Fixed in release: http://bugs.php.net/fix.php?id=28866r=alreadyfixed
Need backtrace: http://bugs.php.net/fix.php?id=28866r=needtrace
Need Reproduce Script: http://bugs.php.net/fix.php?id=28866r=needscript
Try newer version: http://bugs.php.net/fix.php?id=28866r=oldversion
Not developer issue:http://bugs.php.net/fix.php?id=28866r=support
Expected behavior: http://bugs.php.net/fix.php?id=28866r=notwrong
Not enough info:http://bugs.php.net/fix.php?id=28866r=notenoughinfo
Submitted twice:http://bugs.php.net/fix.php?id=28866r=submittedtwice
register_globals: http://bugs.php.net/fix.php?id=28866r=globals
PHP 3 support discontinued: http://bugs.php.net/fix.php?id=28866r=php3
Daylight Savings: http://bugs.php.net/fix.php?id=28866r=dst
IIS Stability: http://bugs.php.net/fix.php?id=28866r=isapi
Install GNU Sed:http://bugs.php.net/fix.php?id=28866r=gnused
Floating point limitations: http://bugs.php.net/fix.php?id=28866r=float
#28767 [Com]: mysqli doesn't compile (link fails)
ID: 28767 Comment by: francois at nor-web dot ca Reported By: Franck at e-Dot dot nl Status: Open Bug Type: Compile Failure Operating System: Linux 2.4.19C13V (Sun Cobalt) PHP Version: 5.0.0RC3 New Comment: I cannot compile PHP5 RC3 either. I compiled PHP5Rc1, PHP5Rc2 without problem. Config: ./configure --with-apxs2=/usr/local/apache2/bin/apxs --with-gd --with-zlib --enable-sockets --enable-soap --enable-shmop --with-mysqli MySQL version: MySQL-server-4.1.1-1 MySQL-shared-4.1.1-0 MySQL-client-4.1.1-0 MySQL-devel-4.1.1-0 Linux version: Mandrake Linux 10.0 official kernel 2.6.3-7mdksmp In file included from /opt/php-5.0.0RC3/ext/mysqli/mysqli.c:31: /opt/php-5.0.0RC3/ext/mysqli/php_mysqli.h:86: error: `LOCAL_INFILE_ERROR_LEN' undeclared here (not in a function) /opt/php-5.0.0RC3/ext/mysqli/mysqli.c: In function `php_local_infile_init': /opt/php-5.0.0RC3/ext/mysqli/mysqli.c:744: error: `LOCAL_INFILE_ERROR_LEN' undeclared (first use in this function) /opt/php-5.0.0RC3/ext/mysqli/mysqli.c:744: error: (Each undeclared identifier is reported only once /opt/php-5.0.0RC3/ext/mysqli/mysqli.c:744: error: for each function it appears in.) /opt/php-5.0.0RC3/ext/mysqli/mysqli.c: In function `php_local_infile_read': /opt/php-5.0.0RC3/ext/mysqli/mysqli.c:825: error: `LOCAL_INFILE_ERROR_LEN' undeclared (first use in this function) make: *** [ext/mysqli/mysqli.lo] Error 1 Previous Comments: [2004-06-14 11:31:58] Franck at e-Dot dot nl Ow forgot to say: Without the --with-mysqli line in the configuration, everything works like a charm! [2004-06-14 08:37:31] Franck at e-Dot dot nl Description: Hi, I'm trying to build php5 with MySQLi support. But I was not able to do so. Tried PHP5.0.0RC2, PHP5.0.0RC3 and PHP5-200406140430 with MySQL 4.1.1alpha 4.1.2alpha. All give me the same error's. MySQL install: I've build my own rpm's from the source rpm. and installed client, devel, server and shared packages. PHP Configuration: ./configure \ --prefix=/usr \ --with-apxs=/usr/sbin/apxs \ --with-config-file-path=/etc/httpd \ --with-zlib \ --enable-magic-quotes \ --with-regex=system \ --enable-track-vars \ --with-iconv \ --enable-xml \ --disable-debug \ --with-gd \ --enable-mbstring \ --enable-mbstr-enc-trans \ --with-interbase=shared \ --with-mysql=/usr \ --with-mysqli=/usr/bin/mysql_config \ --with-ldap \ --with-openssl=/usr \ --with-jpeg-dir=/usr \ --with-png-dir=/usr \ --enable-calender \ --with-curl \ --enable-bcmath \ --with-dom \ --with-dom-xslt \ --enable-exit \ --enable-ftp \ --with-gettext \ --enable-mbregex \ --with-mcrypt \ --with-pear \ --enable-sockets \ --with-freetype-dir \ --enable-soap \ --enable-simplexml \ --with-inifile \ --enable-wddx \ --enable-pcntl This all goes fine, then it all stops on the 'make'. It seems that compiling is not the problem, only when linking the whole stuff it goes wrong. Error message: /usr/lib/mysql/libmysqlclient.a(libmysql.o)(.data+0x0): multiple definition of `net_buffer_length' /usr/lib/mysql/libmysqlclient.a(libmysql.o)(.data+0x0): first defined here /usr/lib/mysql/libmysqlclient.a(libmysql.o)(.data+0x4): multiple definition of `max_allowed_packet' /usr/lib/mysql/libmysqlclient.a(libmysql.o)(.data+0x4): first defined here /usr/lib/mysql/libmysqlclient.a(libmysql.o)(.data+0x8): multiple definition of `net_read_timeout' /usr/lib/mysql/libmysqlclient.a(libmysql.o)(.data+0x8): first defined here /usr/lib/mysql/libmysqlclient.a(libmysql.o)(.data+0xc): multiple definition of `net_write_timeout' /usr/lib/mysql/libmysqlclient.a(libmysql.o)(.data+0xc): first defined here /usr/lib/mysql/libmysqlclient.a(libmysql.o): In function `mysql_server_init': libmysql.o(.text+0x0): multiple definition of `mysql_server_init' /usr/lib/mysql/libmysqlclient.a(libmysql.o)(.text+0x0): first defined here /usr/lib/mysql/libmysqlclient.a(libmysql.o): In function `mysql_debug': libmysql.o(.text+0x1d0): multiple definition of `mysql_debug' /usr/lib/mysql/libmysqlclient.a(libmysql.o)(.text+0x1d0): first defined here /usr/lib/mysql/libmysqlclient.a(libmysql.o): In function `mysql_server_end': libmysql.o(.text+0xf0): multiple definition of `mysql_server_end' /usr/lib/mysql/libmysqlclient.a(libmysql.o)(.text+0xf0): first defined here /usr/lib/mysql/libmysqlclient.a(libmysql.o): In function `mysql_thread_end': libmysql.o(.text+0x150): multiple definition of `mysql_thread_end' /usr/lib/mysql/libmysqlclient.a(libmysql.o)(.text+0x150): first defined here /usr/lib/mysql/libmysqlclient.a(libmysql.o): In function `mysql_get_parameters': libmysql.o(.text+0x130): multiple definition of `mysql_get_parameters' /usr/lib/mysql/libmysqlclient.a(libmysql.o)(.text+0x130): first defined here /usr/lib/mysql/libmysqlclient.a(libmysql.o): In function `mysql_thread_init': libmysql.o(.text+0x140): multiple
#28740 [Com]: PHP has encountered an Access Violation
ID: 28740 Comment by: magnus2 at nova dot nu Reported By: colorprint at mail dot ru Status: Bogus Bug Type: IIS related Operating System: Windows Server 2003 PHP Version: 4.3.7 New Comment: Same here. I've always gotten this error (PHP has encountered an Access Violation at xxx) when running PHP 4.3.x (any version and for both CGI and ISAPI) on Windows 2000 or Windows 2003 that connects to MySQL. I have other sites on same machines connecting to MSSQL that works fine. In Windows 2003 I've isolated the servers in different applications pools (to be able to restart them seperately) and it's only sites using MySQL where this error occur. It can be a few hours between but sometimes it's days. For example on same machine I've set up two phpbb (http://www.phpbb.com) forums, one using a MySQL 4.0.16 database and the other a MSSQL 2000 SP3 database. Both sites in different application pools. Rest of the config is identical (Win 2003, PHP 4.3.5, ISAPI) and it's only the forum site using MySQL that will get this error every couple of days. I've read on other posts that it could be related to pconnect() or that PHP gets unloaded but I don't know what that is supposed to mean. Seems to be an issue when you combine IIS,PHP,MySQL on Windows. I've never seen a final soultion for this. Previous Comments: [2004-06-21 08:03:09] sekmun at mypostbox dot com upgraded from 4.3.6 to 4.3.7, platform: win2k server/IIS (ISAPI)/Mysql Seeing this PHP Access Violation error now intermittently. Didn't use to get his problem. Might go back to 4.3.6 for the time being. [2004-06-20 09:21:59] webmaster at yes-hosting dot com I get the same problem. I'm using win2k3/iis6 upgrade from php 4.3.4 to 4.3.7 error show at the end of php pages. [2004-06-16 15:44:51] jmertic at ncscredit dot com We have the same problem with our IIS 5/Win2k/MySQL. This is a 4.3.7 regression that needs to be addressed. I'm moving us back to 4.3.6. [2004-06-14 11:02:08] orangeasklessapple at cherrygmx dot net We just upgraded from php 4.3.4 to 4.3.7 and are having the same problem under Windows 2000, IIS 5.0, MS SQL Server 2000. A reset of the iis solves the problem just temporarliy. [2004-06-11 09:10:59] [EMAIL PROTECTED] 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. Thank you for your interest in PHP. . 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/28740 -- Edit this bug report at http://bugs.php.net/?id=28740edit=1
#28780 [Opn]: Web Server crashes connecting upon OCILogon
ID: 28780 User updated by: reinhard dot e dot voglmaier at gsk dot com Reported By: reinhard dot e dot voglmaier at gsk dot com Status: Open Bug Type: OCI8 related Operating System: Sun Solaris 9 PHP Version: 4.3.7 Assigned To: tony2001 New Comment: Finally I have the trace: Starting program: /opt/apache/itweb/sbin/httpd_php/httpd -X Program received signal SIGSEGV, Segmentation fault. 0xff0b347c in strlen () from /usr/lib/libc.so.1 (gdb) bt #0 0xff0b347c in strlen () from /usr/lib/libc.so.1 #1 0xfe79f8f8 in snauca_check_adapter () from /opt/oracl/ product/9.2.0/lib32/libclntsh.so.9.0 #2 0xfe79d8f0 in nau_viat () from /opt/oracl/ product/9.2.0/lib32/libclntsh.so.9.0 #3 0xfe7944f4 in nau_gettab () from /opt/oracl/ product/9.2.0/lib32/libclntsh.so.9.0 #4 0xfe7922b8 in nau_ini () from /opt/oracl/ product/9.2.0/lib32/libclntsh.so.9.0 #5 0xfe783f8c in nainit () from /opt/oracl/product/9.2.0/ lib32/libclntsh.so.9.0 #6 0xfe734c3c in nsnainit () from /opt/oracl/ product/9.2.0/lib32/libclntsh.so.9.0 #7 0xfe7271dc in nsopen () from /opt/oracl/product/9.2.0/ lib32/libclntsh.so.9.0 #8 0xfe70d188 in nscall1 () from /opt/oracl/ product/9.2.0/lib32/libclntsh.so.9.0 #9 0xfe70c7a8 in nscall () from /opt/oracl/product/9.2.0/ lib32/libclntsh.so.9.0 #10 0xfe7b8df0 in niotns () from /opt/oracl/product/9.2.0/ lib32/libclntsh.so.9.0 #11 0xfe7b2954 in nigcall () from /opt/oracl/ product/9.2.0/lib32/libclntsh.so.9.0 #12 0xfe744de0 in osncon () from /opt/oracl/product/9.2.0/ lib32/libclntsh.so.9.0 #13 0xfe55f190 in kpuadef () from /opt/oracl/ product/9.2.0/lib32/libclntsh.so.9.0 #14 0xfe55108c in upiini () from /opt/oracl/product/9.2.0/ lib32/libclntsh.so.9.0 #15 0xfe55108c in upiini () from /opt/oracl/product/9.2.0/ lib32/libclntsh.so.9.0 Previous frame identical to this frame (corrupt stack?) (gdb) Previous Comments: [2004-06-16 11:30:35] reinhard dot e dot voglmaier at gsk dot com Maybe there are problems with the 64 bit Environment ? Is PHP + Oracle 64bit safe ? ( Pointer casted into int or unsigned int ? conversions between int ? adress arithmetic used ? ) I am using 64bit Oracle. [2004-06-15 12:42:06] reinhard dot e dot voglmaier at gsk dot com I'm using Oracle 9.2.0 same product as I am connecting to. Oracle is 64 bit ( -m64 switch for gcc ). I started it under truss, here the last few lines: lstat(/opt/apache/itweb/htdocs, 0x7FFF80F0) = 0 getuid()= 60001 [60001] door_info(4, 0x7FFF5D68)= 0 door_call(4, 0x7FFF5D38)= 0 sigprocmask(SIG_SETMASK, 0x7C91D500, 0x7FFF7C90) = 0 sigprocmask(SIG_SETMASK, 0x7FFF7C7C, 0x) = 0 sigaction(SIGINT, 0x7FFF7AF0, 0x7FFF7CE8) = 0 sigprocmask(SIG_SETMASK, 0x7C91D500, 0x7FFF7C90) = 0 sigprocmask(SIG_SETMASK, 0x7FFF7C7C, 0x) = 0 brk(0x1004C6210)= 0 brk(0x1004CA210)= 0 brk(0x1004CA210)= 0 brk(0x1004CE210)= 0 brk(0x1004CE210)= 0 brk(0x1004D2210)= 0 door_info(4, 0x7FFF4538)= 0 door_call(4, 0x7FFF4508)= 0 brk(0x1004D2210)= 0 brk(0x1004D6210)= 0 so_socket(PF_INET, SOCK_STREAM, IPPROTO_IP, , 1) = 7 connect(7, 0x7FFF67F4, 16, 1) = 0 getsockname(7, 0x7FFF67F4, 0x7FFF6804, 1) = 0 setsockopt(7, tcp, TCP_NODELAY, 0x7FFF6964, 4, 1) = 0 fcntl(7, F_SETFD, 0x0001) = 0 brk(0x1004D6210)= 0 brk(0x1004DA210)= 0 brk(0x1004DA210)= 0 brk(0x1004DE210)= 0 brk(0x1004DE210)= 0 brk(0x1004FE210)= 0 Incurred fault #6, FLTBOUNDS %pc = 0x7D63C1AC siginfo: SIGSEGV SEGV_MAPERR addr=0x Received signal #11, SIGSEGV [default] siginfo: SIGSEGV SEGV_MAPERR addr=0x [2004-06-15 12:31:26] [EMAIL PROTECTED] Seems to be a Solaris-specific GDB bug. Have a look at these messages: http://www.mail-archive.com/[EMAIL PROTECTED]/msg00818.html http://www.mail-archive.com/[EMAIL PROTECTED]/msg00452.html Btw, what Oracle client version do you use and does it correspond to Oracle server version? [2004-06-15 11:59:21] reinhard dot e dot
#28848 [Opn-Bgs]: PHP 5 RC3 probs with mysql
ID: 28848 Updated by: [EMAIL PROTECTED] Reported By: PatrickLehmann at freenet dot de -Status: Open +Status: Bogus Bug Type: MySQL related Operating System: Microsoft Windows Server 2003 PHP Version: 5.0.0RC3 New Comment: Mysql extension works fine. You probably forgot to copy libmysql.dll to a directory in your path (such as c:\windows\system32). Another option is to add your php folder to the path. Previous Comments: [2004-06-21 00:03:58] PatrickLehmann at freenet dot de ²uncledave where have you installed PHP? PHP 4.3.7 doesn't run with Apache2, too. my load commands: LoadModule php4_module C:/Programme/PHP4/sapi/php4apache2.dll AddModule mod_php4.c AddType application/x-httpd-php .php Apache error: module couldn't be loaded, because file doesn't exist *lol* it's still there!!! I remembered that PHP 4.0.4 has a bug with it's install-dir, PHP had to be installed to C:\php\ otherwise it doesn't run is this bug still present? [2004-06-20 23:42:33] PatrickLehmann at freenet dot de include_path is set to this: include_path = .;C:\Programme\PHP5\ext PHP was unpacked to C:\Programme\PHP5 and under extensions this: extension=php_mysql.dll I think this is done like the description in the install manual [2004-06-20 12:17:41] uncledave at xmail dot net I've experienced the same problem. I'm running XP with Apache 2.04.3 PHP 5.0 RC3. I cannot run Apache/PHP with the mysql.dll extension. I get PHP Startup: Unable to load dyanmic library 'c:\PHP\ext\php_mysql.dll - The specified module could not be found.' I have reset to run 4.x and it runs perfectly.. back to 5.0 and the same error. [2004-06-20 00:22:03] PatrickLehmann at freenet dot de Description: Hi, my system: Microsoft Windows Server 2003, Enterprise Edition Apache 2.0.49 PHP 5.0 RC 3 MySQL 4.0.18 apache with php is running successfully if mysql extension ist disabled. when i enable the extension php_mysql.dll apache coundn't be started - no error message - port 80 is used for apache (a 2. start stops with an error that port 80 is used yet) - apache is loaded into the RAM libmysql.dll was copyed to C:\Windows php runs perfectliy if mysql is disabled, but with this extension ames no reaction ok, than the 2. problem IIS 6.0 doesn't work with the ISAPI extension of php 5 setting: ISAPI Filter is set to C:\Programme\PHP5\php5isapi.dll Filter name: PHP 5 RC 3 the error message is an 404 error in the browser (index.php with phpinfo command) -- Edit this bug report at http://bugs.php.net/?id=28848edit=1
#28868 [NEW]: Thread safty issue with stream_wrapper_register?
From: cleong at nflc dot org
Operating system: Windows 2000
PHP version: 4.3.7
PHP Bug Type: Filesystem function related
Bug description: Thread safty issue with stream_wrapper_register?
Description:
On Windows 2000, with PHP running as an Apache module,
stream_wrapper_register() would fail when there are concurent scripts
using that function. It would either fail without an error message or
causes the Apache process to crash. When there's just one script running
the function works fine.
To reproduce this bug, open a couple browser windows and go to the
reproduce code in each.
I can get around the bug by generating a random class name and protocol
name everytime (using eval for the class definition).
Reproduce code:
---
?
// example class taken from manual
class VariableStream {
var $position;
var $varname;
function stream_open($path, $mode, $options, $opened_path)
{
$url = parse_url($path);
$this-varname = $url[host];
$this-position = 0;
return true;
}
function stream_read($count)
{
$ret = substr($GLOBALS[$this-varname], $this-position, $count);
$this-position += strlen($ret);
return $ret;
}
function stream_write($data)
{
$left = substr($GLOBALS[$this-varname], 0, $this-position);
$right = substr($GLOBALS[$this-varname], $this-position +
strlen($data));
$GLOBALS[$this-varname] = $left . $data . $right;
$this-position += strlen($data);
return strlen($data);
}
function stream_tell()
{
return $this-position;
}
function stream_eof()
{
return $this-position = strlen($GLOBALS[$this-varname]);
}
function stream_seek($offset, $whence)
{
switch ($whence) {
case SEEK_SET:
if ($offset strlen($GLOBALS[$this-varname]) $offset
= 0) {
$this-position = $offset;
return true;
} else {
return false;
}
break;
case SEEK_CUR:
if ($offset = 0) {
$this-position += $offset;
return true;
} else {
return false;
}
break;
case SEEK_END:
if (strlen($GLOBALS[$this-varname]) + $offset = 0) {
$this-position = strlen($GLOBALS[$this-varname]) +
$offset;
return true;
} else {
return false;
}
break;
default:
return false;
}
}
}
stream_wrapper_register(var, VariableStream)
or die(Failed to register protocol);
$myvar = ;
$fp = fopen(var://myvar, r+);
fwrite($fp, line1\n);
fwrite($fp, line2\n);
fwrite($fp, line3\n);
rewind($fp);
while (!feof($fp)) {
echo fgets($fp);
}
session_write_close(); // in case auto-session is on
for($i = 0; $i 30; $i++) {
// make the script run for 30 seconds
sleep(1);
}
fclose($fp);
var_dump($myvar);
?
Expected result:
Same result in each window.
Actual result:
--
While the first script is still running, the second script would say
Failed to register protocol. If you refresh the second window, sometimes
it works, sometimes it causes a general protection error.
--
Edit bug report at http://bugs.php.net/?id=28868edit=1
--
Try a CVS snapshot (php4): http://bugs.php.net/fix.php?id=28868r=trysnapshot4
Try a CVS snapshot (php5): http://bugs.php.net/fix.php?id=28868r=trysnapshot5
Fixed in CVS: http://bugs.php.net/fix.php?id=28868r=fixedcvs
Fixed in release: http://bugs.php.net/fix.php?id=28868r=alreadyfixed
Need backtrace: http://bugs.php.net/fix.php?id=28868r=needtrace
Need Reproduce Script: http://bugs.php.net/fix.php?id=28868r=needscript
Try newer version: http://bugs.php.net/fix.php?id=28868r=oldversion
Not developer issue:http://bugs.php.net/fix.php?id=28868r=support
Expected behavior: http://bugs.php.net/fix.php?id=28868r=notwrong
Not enough info:http://bugs.php.net/fix.php?id=28868r=notenoughinfo
Submitted twice:http://bugs.php.net/fix.php?id=28868r=submittedtwice
register_globals: http://bugs.php.net/fix.php?id=28868r=globals
PHP 3 support discontinued: http://bugs.php.net/fix.php?id=28868r=php3
Daylight Savings: http://bugs.php.net/fix.php?id=28868r=dst
IIS Stability: http://bugs.php.net/fix.php?id=28868r=isapi
Install GNU Sed:http://bugs.php.net/fix.php?id=28868r=gnused
Floating point limitations: http://bugs.php.net/fix.php?id=28868r=float
#28869 [NEW]: possible wrong behavior of fgetcsv (?)
From: laph at gmx dot net
Operating system: All
PHP version: 4.3.6
PHP Bug Type: Filesystem function related
Bug description: possible wrong behavior of fgetcsv (?)
Description:
Assume a file with 1 line of text (toImport.csv) and a [newline] after the
last char in that line.
The code below will echo 2 (Lines) using either fgets or fgetcsv.
While this is right for fgets, it's wrong for fgetcsv (I think), because
in CSV-Files the [newline] (logically) terminates a row, it doesn't start
a new one - at least if you think of a CSV-File as a Recordset from a
SQL-Statement (If you read the last line of a SQL-Result, move the
Row-Pointer to the next row and try to read it, you'll get false/EOF and
not a row with 0 elements).
If you return the Array $data in each iteration you get a array with
(int) 0 elements as last row, what makes using a loop until EOF
impossible, because you (additionally) have to check the number of
elements returned in $data.
It's quite anoying, that fgetcsv makes difference if a csv-file is
terminated by a newline or not.
Reproduce code:
---
$fh = fopen(dirname(__FILE__)./toImport.csv,r);
$row = 0;
while (!feof($fh)) {
$data = fgetcsv($fh,1000,;);
// $data = fgets($fh,1000);
$row++;
}
echo($row);
Expected result:
fgetcsv should echo 1, while fgets should echo 2
--
Edit bug report at http://bugs.php.net/?id=28869edit=1
--
Try a CVS snapshot (php4): http://bugs.php.net/fix.php?id=28869r=trysnapshot4
Try a CVS snapshot (php5): http://bugs.php.net/fix.php?id=28869r=trysnapshot5
Fixed in CVS: http://bugs.php.net/fix.php?id=28869r=fixedcvs
Fixed in release: http://bugs.php.net/fix.php?id=28869r=alreadyfixed
Need backtrace: http://bugs.php.net/fix.php?id=28869r=needtrace
Need Reproduce Script: http://bugs.php.net/fix.php?id=28869r=needscript
Try newer version: http://bugs.php.net/fix.php?id=28869r=oldversion
Not developer issue:http://bugs.php.net/fix.php?id=28869r=support
Expected behavior: http://bugs.php.net/fix.php?id=28869r=notwrong
Not enough info:http://bugs.php.net/fix.php?id=28869r=notenoughinfo
Submitted twice:http://bugs.php.net/fix.php?id=28869r=submittedtwice
register_globals: http://bugs.php.net/fix.php?id=28869r=globals
PHP 3 support discontinued: http://bugs.php.net/fix.php?id=28869r=php3
Daylight Savings: http://bugs.php.net/fix.php?id=28869r=dst
IIS Stability: http://bugs.php.net/fix.php?id=28869r=isapi
Install GNU Sed:http://bugs.php.net/fix.php?id=28869r=gnused
Floating point limitations: http://bugs.php.net/fix.php?id=28869r=float
#28435 [Opn-Csd]: array_count_values() problem
ID: 28435
User updated by: ofirin at yahoo dot com
Reported By: ofirin at yahoo dot com
-Status: Open
+Status: Closed
Bug Type: Arrays related
Operating System: Any
PHP Version: 4CVS-2004-05-18 (stable)
New Comment:
Ok, I think you're right, we all should be looking forward working with
php5, rather than trying to fix old bugs in php4.
Previous Comments:
[2004-06-21 11:31:59] vladb at pseudo-infinity dot ro
The php5 result seems more accurate than the php4 one. You got strings
in the $books array, no?
Changing this in php4 may break some code out there so maybe it'd be
best to just mention it as a 4-5 inconsistency.
[2004-06-18 14:24:41] programmer at bardware dot de
I noticed this problem with PHP5 RC2 on a Win2k box as an
Apache2-module. I have some HMTL-Checkboxes and on the server side want
to check the selected values. My checkboxes are equally named
name=choice2[] what lets PHP generate an array
$_POST[choice2][0]
$_POST[choice2][1]
$_POST[choice2][2] etc. according to the selected values.
Each checkbox has it's unique value attribute.
To test for a certain value I reverse this array with
$arrTmp=array_count_values($_POST[choice2]);
I now want to access $arrTmp[1] to check if the checkbox with the
attribute value=1 was selected. The respective value was - if
selected - 1, otherwise it's not existent in the
$_POST[choice2]-Array. This did not work. $arrTmp[1] didn't work
either. It was no way possible to access a member of the array.
The other poster mentioned the indexes are generated as strings, I want
to point they cannot be accessed at all.
It worked best on PHP 4.3.6
Best,
Bernhard
[2004-05-18 19:04:43] ofirin at yahoo dot com
Description:
I don't know if this is a php5 or php4 bug, but I'm sure something's
wrong here.
Whenever in php5 I do an array_count_values() on an array that
contains numeric values as strings the result array uses string keys
instead of the numeric values as indexes. This doesn't happen in php4.
I don't know what behaivor is correct.
I'm using the latest cvs versions of both php4 and php5 btw.
Reproduce code:
---
?php
$books = Array('10', '3', '6', '10');
$quantities = array_count_values($books);
var_dump($books);
var_dump($quantities);
?
Expected result:
This is what happens in php4:
array(4) {
[0]=
string(2) 10
[1]=
string(1) 3
[2]=
string(1) 6
[3]=
string(2) 10
}
array(3) {
[10]=
int(2)
[3]=
int(1)
[6]=
int(1)
}
Actual result:
--
This is what happens if php5:
array(4) {
[0]=
string(2) 10
[1]=
string(1) 3
[2]=
string(1) 6
[3]=
string(2) 10
}
array(3) {
[10]=
int(2)
[3]=
int(1)
[6]=
int(1)
}
--
Edit this bug report at http://bugs.php.net/?id=28435edit=1
#28780 [Opn-Fbk]: Web Server crashes connecting upon OCILogon
ID: 28780 Updated by: [EMAIL PROTECTED] Reported By: reinhard dot e dot voglmaier at gsk dot com -Status: Open +Status: Feedback Bug Type: OCI8 related Operating System: Sun Solaris 9 PHP Version: 4.3.7 Assigned To: tony2001 New Comment: The backtrace shows, that the bug has nothing to do with PHP, but most likely is a Oracle Client problem. Could you please search Oracle bug database ? Previous Comments: [2004-06-21 14:30:39] reinhard dot e dot voglmaier at gsk dot com Finally I have the trace: Starting program: /opt/apache/itweb/sbin/httpd_php/httpd -X Program received signal SIGSEGV, Segmentation fault. 0xff0b347c in strlen () from /usr/lib/libc.so.1 (gdb) bt #0 0xff0b347c in strlen () from /usr/lib/libc.so.1 #1 0xfe79f8f8 in snauca_check_adapter () from /opt/oracl/ product/9.2.0/lib32/libclntsh.so.9.0 #2 0xfe79d8f0 in nau_viat () from /opt/oracl/ product/9.2.0/lib32/libclntsh.so.9.0 #3 0xfe7944f4 in nau_gettab () from /opt/oracl/ product/9.2.0/lib32/libclntsh.so.9.0 #4 0xfe7922b8 in nau_ini () from /opt/oracl/ product/9.2.0/lib32/libclntsh.so.9.0 #5 0xfe783f8c in nainit () from /opt/oracl/product/9.2.0/ lib32/libclntsh.so.9.0 #6 0xfe734c3c in nsnainit () from /opt/oracl/ product/9.2.0/lib32/libclntsh.so.9.0 #7 0xfe7271dc in nsopen () from /opt/oracl/product/9.2.0/ lib32/libclntsh.so.9.0 #8 0xfe70d188 in nscall1 () from /opt/oracl/ product/9.2.0/lib32/libclntsh.so.9.0 #9 0xfe70c7a8 in nscall () from /opt/oracl/product/9.2.0/ lib32/libclntsh.so.9.0 #10 0xfe7b8df0 in niotns () from /opt/oracl/product/9.2.0/ lib32/libclntsh.so.9.0 #11 0xfe7b2954 in nigcall () from /opt/oracl/ product/9.2.0/lib32/libclntsh.so.9.0 #12 0xfe744de0 in osncon () from /opt/oracl/product/9.2.0/ lib32/libclntsh.so.9.0 #13 0xfe55f190 in kpuadef () from /opt/oracl/ product/9.2.0/lib32/libclntsh.so.9.0 #14 0xfe55108c in upiini () from /opt/oracl/product/9.2.0/ lib32/libclntsh.so.9.0 #15 0xfe55108c in upiini () from /opt/oracl/product/9.2.0/ lib32/libclntsh.so.9.0 Previous frame identical to this frame (corrupt stack?) (gdb) [2004-06-16 11:30:35] reinhard dot e dot voglmaier at gsk dot com Maybe there are problems with the 64 bit Environment ? Is PHP + Oracle 64bit safe ? ( Pointer casted into int or unsigned int ? conversions between int ? adress arithmetic used ? ) I am using 64bit Oracle. [2004-06-15 12:42:06] reinhard dot e dot voglmaier at gsk dot com I'm using Oracle 9.2.0 same product as I am connecting to. Oracle is 64 bit ( -m64 switch for gcc ). I started it under truss, here the last few lines: lstat(/opt/apache/itweb/htdocs, 0x7FFF80F0) = 0 getuid()= 60001 [60001] door_info(4, 0x7FFF5D68)= 0 door_call(4, 0x7FFF5D38)= 0 sigprocmask(SIG_SETMASK, 0x7C91D500, 0x7FFF7C90) = 0 sigprocmask(SIG_SETMASK, 0x7FFF7C7C, 0x) = 0 sigaction(SIGINT, 0x7FFF7AF0, 0x7FFF7CE8) = 0 sigprocmask(SIG_SETMASK, 0x7C91D500, 0x7FFF7C90) = 0 sigprocmask(SIG_SETMASK, 0x7FFF7C7C, 0x) = 0 brk(0x1004C6210)= 0 brk(0x1004CA210)= 0 brk(0x1004CA210)= 0 brk(0x1004CE210)= 0 brk(0x1004CE210)= 0 brk(0x1004D2210)= 0 door_info(4, 0x7FFF4538)= 0 door_call(4, 0x7FFF4508)= 0 brk(0x1004D2210)= 0 brk(0x1004D6210)= 0 so_socket(PF_INET, SOCK_STREAM, IPPROTO_IP, , 1) = 7 connect(7, 0x7FFF67F4, 16, 1) = 0 getsockname(7, 0x7FFF67F4, 0x7FFF6804, 1) = 0 setsockopt(7, tcp, TCP_NODELAY, 0x7FFF6964, 4, 1) = 0 fcntl(7, F_SETFD, 0x0001) = 0 brk(0x1004D6210)= 0 brk(0x1004DA210)= 0 brk(0x1004DA210)= 0 brk(0x1004DE210)= 0 brk(0x1004DE210)= 0 brk(0x1004FE210)= 0 Incurred fault #6, FLTBOUNDS %pc = 0x7D63C1AC siginfo: SIGSEGV SEGV_MAPERR addr=0x Received signal #11, SIGSEGV [default] siginfo: SIGSEGV SEGV_MAPERR addr=0x [2004-06-15 12:31:26] [EMAIL PROTECTED] Seems to be a Solaris-specific GDB bug. Have a look at these messages: http://www.mail-archive.com/[EMAIL
#28868 [Opn-Ana]: Thread safty issue with stream_wrapper_register?
ID: 28868
Updated by: [EMAIL PROTECTED]
Reported By: cleong at nflc dot org
-Status: Open
+Status: Analyzed
Bug Type: Filesystem function related
Operating System: Windows 2000
PHP Version: 4.3.7
New Comment:
There is absolutely a thread safety issue here. I assume you're
running Apache2?
For the moment, I'd suggest Apache1 or Apache2-pre-fork. In the mean
time we'll be working on fixing this asap.
Previous Comments:
[2004-06-21 16:53:43] cleong at nflc dot org
Description:
On Windows 2000, with PHP running as an Apache module,
stream_wrapper_register() would fail when there are concurent scripts
using that function. It would either fail without an error message or
causes the Apache process to crash. When there's just one script
running the function works fine.
To reproduce this bug, open a couple browser windows and go to the
reproduce code in each.
I can get around the bug by generating a random class name and protocol
name everytime (using eval for the class definition).
Reproduce code:
---
?
// example class taken from manual
class VariableStream {
var $position;
var $varname;
function stream_open($path, $mode, $options, $opened_path)
{
$url = parse_url($path);
$this-varname = $url[host];
$this-position = 0;
return true;
}
function stream_read($count)
{
$ret = substr($GLOBALS[$this-varname], $this-position,
$count);
$this-position += strlen($ret);
return $ret;
}
function stream_write($data)
{
$left = substr($GLOBALS[$this-varname], 0, $this-position);
$right = substr($GLOBALS[$this-varname], $this-position +
strlen($data));
$GLOBALS[$this-varname] = $left . $data . $right;
$this-position += strlen($data);
return strlen($data);
}
function stream_tell()
{
return $this-position;
}
function stream_eof()
{
return $this-position = strlen($GLOBALS[$this-varname]);
}
function stream_seek($offset, $whence)
{
switch ($whence) {
case SEEK_SET:
if ($offset strlen($GLOBALS[$this-varname])
$offset = 0) {
$this-position = $offset;
return true;
} else {
return false;
}
break;
case SEEK_CUR:
if ($offset = 0) {
$this-position += $offset;
return true;
} else {
return false;
}
break;
case SEEK_END:
if (strlen($GLOBALS[$this-varname]) + $offset = 0) {
$this-position = strlen($GLOBALS[$this-varname])
+ $offset;
return true;
} else {
return false;
}
break;
default:
return false;
}
}
}
stream_wrapper_register(var, VariableStream)
or die(Failed to register protocol);
$myvar = ;
$fp = fopen(var://myvar, r+);
fwrite($fp, line1\n);
fwrite($fp, line2\n);
fwrite($fp, line3\n);
rewind($fp);
while (!feof($fp)) {
echo fgets($fp);
}
session_write_close(); // in case auto-session is on
for($i = 0; $i 30; $i++) {
// make the script run for 30 seconds
sleep(1);
}
fclose($fp);
var_dump($myvar);
?
Expected result:
Same result in each window.
Actual result:
--
While the first script is still running, the second script would say
Failed to register protocol. If you refresh the second window,
sometimes it works, sometimes it causes a general protection error.
--
Edit this bug report at http://bugs.php.net/?id=28868edit=1
#28871 [NEW]: mysql_afftected_rows returns 0 even when updates record
From: nav at nc dot ru
Operating system: Windows XP
PHP version: 4.3.6
PHP Bug Type: MySQL related
Bug description: mysql_afftected_rows returns 0 even when updates record
Description:
I've got the problem similar to already reported here.
mysql_afftected_rows sometimes returns 0, sometimes 1. Although actually
it does UPDATES the record. This query changes user login (password,
e-mail) in the table user (see below. If it matters, MySQL is
3.23.53-max, Apache is 1.3.27):
Reproduce code:
---
mysql_query(UPDATE user SET .$setRequest. WHERE login =
'$userCurrentLogin' AND password = md5('$userCurrentPass'));
/* $setRequest is the string compiled conditionally, but I've just printed
whole query string and it looked correct:
UPDATE user SET login = 'aaa' WHERE login = 'bbb' AND password =
md5('ccc') */
$num = mysql_affected_rows();
if ($num == 0) $errorMsg = Could not complete your request. Check the
data;
else if ($num == -1) $errorMsg = Database error. Try again later;
else $changeMsg = Your data has been successfully changed;
Expected result:
Of course, I expect correct work of mysql_affected_rows :)
Actual result:
--
BUT the result ($num) is sometimes 0, sometimes 1, even if the record is
actually updated.
--
Edit bug report at http://bugs.php.net/?id=28871edit=1
--
Try a CVS snapshot (php4): http://bugs.php.net/fix.php?id=28871r=trysnapshot4
Try a CVS snapshot (php5): http://bugs.php.net/fix.php?id=28871r=trysnapshot5
Fixed in CVS: http://bugs.php.net/fix.php?id=28871r=fixedcvs
Fixed in release: http://bugs.php.net/fix.php?id=28871r=alreadyfixed
Need backtrace: http://bugs.php.net/fix.php?id=28871r=needtrace
Need Reproduce Script: http://bugs.php.net/fix.php?id=28871r=needscript
Try newer version: http://bugs.php.net/fix.php?id=28871r=oldversion
Not developer issue:http://bugs.php.net/fix.php?id=28871r=support
Expected behavior: http://bugs.php.net/fix.php?id=28871r=notwrong
Not enough info:http://bugs.php.net/fix.php?id=28871r=notenoughinfo
Submitted twice:http://bugs.php.net/fix.php?id=28871r=submittedtwice
register_globals: http://bugs.php.net/fix.php?id=28871r=globals
PHP 3 support discontinued: http://bugs.php.net/fix.php?id=28871r=php3
Daylight Savings: http://bugs.php.net/fix.php?id=28871r=dst
IIS Stability: http://bugs.php.net/fix.php?id=28871r=isapi
Install GNU Sed:http://bugs.php.net/fix.php?id=28871r=gnused
Floating point limitations: http://bugs.php.net/fix.php?id=28871r=float
#27517 [Com]: SimpleXML dont get CDATA
ID: 27517 Comment by: schmich at uiuc dot edu Reported By: cyppe at telia dot com Status: No Feedback Bug Type: XML related Operating System: linux PHP Version: 5.0.0b4 (beta4) New Comment: Same behavior confirmed with PHP5 RC3 on linux Previous Comments: [2004-03-31 19:51:12] jaba at inbox dot lv Just wanted to add that this is PHP5 RC1 running on Windows XP [2004-03-31 19:49:57] jaba at inbox dot lv SimpleXML does support CDATA, but it fails to show nodes consisting of CDATA when you call print_r() or var_dump(). It only says [tag_name] = simplexml_element Object When you call $xmlobject-[tag_name] then you get what you expect, but with print_r() - no! [2004-03-13 05:01:41] [EMAIL PROTECTED] No feedback was provided. The bug is being suspended because we assume that you are no longer experiencing the problem. If this is not the case and you are able to provide the information that was requested earlier, please do so and change the status of the bug back to Open. Thank you. [2004-03-07 09:41:42] [EMAIL PROTECTED] Not enough information was provided for us to be able to handle this bug. Please re-read the instructions at http://bugs.php.net/how-to-report.php If you can provide more information, feel free to add it to this bug and change the status back to Open. Thank you for your interest in PHP. [2004-03-07 09:30:07] cyppe at telia dot com Description: SimpleXML seems to work all good, except for getting a CDATA node, it returns nothing. Really nothing more to say. -- Edit this bug report at http://bugs.php.net/?id=27517edit=1
#28872 [NEW]: accessing a __get (getter) method from object value fails
From: nick dot rich at textmarketer dot co dot uk
Operating system: Linux
PHP version: 5.0.0RC3
PHP Bug Type: Class/Object related
Bug description: accessing a __get (getter) method from object value fails
Description:
Trying to access a propery (via the get method) using a
value from another object as the method name fails.
See code::
if the line /// fails
is
replaced with
$localVar = $config-methodnames[1];
$test-$localVar
It works correctly
Reproduce code:
---
class config
{
public $methodnames = array(dog,cat);
}
class testClass
{
function Cat(){
echo cat;
}
function Dog(){
echo dog;
}
function __get($method){
eval(\$this-\$method(););
}
}
class main
{
function __construct(){
$config = new config();
$test = new testClass();
$test-$config-methodnames[1]; /// fails
}
}
$test = new main();
Expected result:
cat
Actual result:
--
Fatal error: Call to undefined method testClass::
Object() in /Users/nicholasrich/Sites/bits/
testclass.php(18) : eval()'d code on line 1
--
Edit bug report at http://bugs.php.net/?id=28872edit=1
--
Try a CVS snapshot (php4): http://bugs.php.net/fix.php?id=28872r=trysnapshot4
Try a CVS snapshot (php5): http://bugs.php.net/fix.php?id=28872r=trysnapshot5
Fixed in CVS: http://bugs.php.net/fix.php?id=28872r=fixedcvs
Fixed in release: http://bugs.php.net/fix.php?id=28872r=alreadyfixed
Need backtrace: http://bugs.php.net/fix.php?id=28872r=needtrace
Need Reproduce Script: http://bugs.php.net/fix.php?id=28872r=needscript
Try newer version: http://bugs.php.net/fix.php?id=28872r=oldversion
Not developer issue:http://bugs.php.net/fix.php?id=28872r=support
Expected behavior: http://bugs.php.net/fix.php?id=28872r=notwrong
Not enough info:http://bugs.php.net/fix.php?id=28872r=notenoughinfo
Submitted twice:http://bugs.php.net/fix.php?id=28872r=submittedtwice
register_globals: http://bugs.php.net/fix.php?id=28872r=globals
PHP 3 support discontinued: http://bugs.php.net/fix.php?id=28872r=php3
Daylight Savings: http://bugs.php.net/fix.php?id=28872r=dst
IIS Stability: http://bugs.php.net/fix.php?id=28872r=isapi
Install GNU Sed:http://bugs.php.net/fix.php?id=28872r=gnused
Floating point limitations: http://bugs.php.net/fix.php?id=28872r=float
#28868 [Ana-Asn]: Thread safty issue with stream_wrapper_register?
ID: 28868
Updated by: [EMAIL PROTECTED]
Reported By: cleong at nflc dot org
-Status: Analyzed
+Status: Assigned
Bug Type: Filesystem function related
Operating System: Windows 2000
PHP Version: 4.3.7
-Assigned To:
+Assigned To: pollita
Previous Comments:
[2004-06-21 18:34:55] [EMAIL PROTECTED]
There is absolutely a thread safety issue here. I assume you're
running Apache2?
For the moment, I'd suggest Apache1 or Apache2-pre-fork. In the mean
time we'll be working on fixing this asap.
[2004-06-21 16:53:43] cleong at nflc dot org
Description:
On Windows 2000, with PHP running as an Apache module,
stream_wrapper_register() would fail when there are concurent scripts
using that function. It would either fail without an error message or
causes the Apache process to crash. When there's just one script
running the function works fine.
To reproduce this bug, open a couple browser windows and go to the
reproduce code in each.
I can get around the bug by generating a random class name and protocol
name everytime (using eval for the class definition).
Reproduce code:
---
?
// example class taken from manual
class VariableStream {
var $position;
var $varname;
function stream_open($path, $mode, $options, $opened_path)
{
$url = parse_url($path);
$this-varname = $url[host];
$this-position = 0;
return true;
}
function stream_read($count)
{
$ret = substr($GLOBALS[$this-varname], $this-position,
$count);
$this-position += strlen($ret);
return $ret;
}
function stream_write($data)
{
$left = substr($GLOBALS[$this-varname], 0, $this-position);
$right = substr($GLOBALS[$this-varname], $this-position +
strlen($data));
$GLOBALS[$this-varname] = $left . $data . $right;
$this-position += strlen($data);
return strlen($data);
}
function stream_tell()
{
return $this-position;
}
function stream_eof()
{
return $this-position = strlen($GLOBALS[$this-varname]);
}
function stream_seek($offset, $whence)
{
switch ($whence) {
case SEEK_SET:
if ($offset strlen($GLOBALS[$this-varname])
$offset = 0) {
$this-position = $offset;
return true;
} else {
return false;
}
break;
case SEEK_CUR:
if ($offset = 0) {
$this-position += $offset;
return true;
} else {
return false;
}
break;
case SEEK_END:
if (strlen($GLOBALS[$this-varname]) + $offset = 0) {
$this-position = strlen($GLOBALS[$this-varname])
+ $offset;
return true;
} else {
return false;
}
break;
default:
return false;
}
}
}
stream_wrapper_register(var, VariableStream)
or die(Failed to register protocol);
$myvar = ;
$fp = fopen(var://myvar, r+);
fwrite($fp, line1\n);
fwrite($fp, line2\n);
fwrite($fp, line3\n);
rewind($fp);
while (!feof($fp)) {
echo fgets($fp);
}
session_write_close(); // in case auto-session is on
for($i = 0; $i 30; $i++) {
// make the script run for 30 seconds
sleep(1);
}
fclose($fp);
var_dump($myvar);
?
Expected result:
Same result in each window.
Actual result:
--
While the first script is still running, the second script would say
Failed to register protocol. If you refresh the second window,
sometimes it works, sometimes it causes a general protection error.
--
Edit this bug report at http://bugs.php.net/?id=28868edit=1
#28856 [Opn]: session_start() causes apache process to hang
ID: 28856 Updated by: [EMAIL PROTECTED] Reported By: pubnelle at megaphone dot ch Status: Open Bug Type: Session related Operating System: Mac OS X server 10.3.4 PHP Version: 4.3.6 New Comment: The suggestion there to use include/require was intended to mean a local include/require not an external one. In a typical Apache setup any files accessible to one virtual host is going to be accessible to another. Not directly via http of course, but PHP's include uses the filesystem so it would be something like: include '/var/www/example.com/file.php'; as opposed to: include 'http://example.com/file.php'; The former is an order of magnitude faster than the latter. Previous Comments: [2004-06-21 07:52:05] pubnelle at megaphone dot ch That definitely makes sense ! There are plenty of times when you need to include a file that is on the same server ! Here, it is to use a Content Management System, which has more rights on the databases, and which is not directly accessible by our clients. When it is not localhost, it can simply be another virtual host, by the way. We use it often. In the virtual() doc : As of PHP 4.0.6, you can use virtual() on PHP files. However, it is typically better to use include() or require() if you need to include another PHP file. It is not precised if this is for http include or local include, that's true. [2004-06-20 23:03:09] [EMAIL PROTECTED] Why are you using an http include to localhost? That makes no sense. You have fast direct access to this file already. Just include it directly. If you really need it to act like a completely separate request (which you shouldn't) use a sub-request via virtual(). [2004-06-20 19:57:22] pubnelle at megaphone dot ch Description: session_start() causes apache process to hang when a call is made, via http include, to another php script which stores the session files at the same location. It create an http get error in the log files eventually - I had a hard time to find the problem... I upgraded the version of PHP from 4.3.1 built in OS X server to 4.3.6, same problem. I upgraded Apache from 1.3.28 built in OS X server to 1.3.31, same problem. We really need this, to link a shop to a content management system, using the same session information to follow up the caddie stuff ... It helps to use the workaround of bug #22526, to close the session, but is this really the only solution ? this is painfull ... Reproduce code: --- ? session_start(); $urltoinclude=http://localhost/index.php;; include($cmsurl/index.php? . session_name() . = . session_id()); ? Expected result: The file included ;-) To reproduce it, you need to put this code in a page that registers the sessions at the same location than the included URL. Actual result: -- White page hanging ... something appearing eventually (without the include) after some time, exactly 2 minutes. -- Edit this bug report at http://bugs.php.net/?id=28856edit=1
#28873 [NEW]: PHP Crashes. But not in 4.3.6.
From: sayez at fucam dot ac dot be Operating system: Windows 2003 Server PHP version: 4.3.7 PHP Bug Type: *General Issues Bug description: PHP Crashes. But not in 4.3.6. Description: Impossible to tell how to produce the bug for the moment. At the end of the first page of the site : PHP Encountered error and all the other pages are inaccessible. Reset IIS is the only way to solve. We use MSSQL, GD, LDAP Library and ISAPI module. -- Edit bug report at http://bugs.php.net/?id=28873edit=1 -- Try a CVS snapshot (php4): http://bugs.php.net/fix.php?id=28873r=trysnapshot4 Try a CVS snapshot (php5): http://bugs.php.net/fix.php?id=28873r=trysnapshot5 Fixed in CVS: http://bugs.php.net/fix.php?id=28873r=fixedcvs Fixed in release: http://bugs.php.net/fix.php?id=28873r=alreadyfixed Need backtrace: http://bugs.php.net/fix.php?id=28873r=needtrace Need Reproduce Script: http://bugs.php.net/fix.php?id=28873r=needscript Try newer version: http://bugs.php.net/fix.php?id=28873r=oldversion Not developer issue:http://bugs.php.net/fix.php?id=28873r=support Expected behavior: http://bugs.php.net/fix.php?id=28873r=notwrong Not enough info:http://bugs.php.net/fix.php?id=28873r=notenoughinfo Submitted twice:http://bugs.php.net/fix.php?id=28873r=submittedtwice register_globals: http://bugs.php.net/fix.php?id=28873r=globals PHP 3 support discontinued: http://bugs.php.net/fix.php?id=28873r=php3 Daylight Savings: http://bugs.php.net/fix.php?id=28873r=dst IIS Stability: http://bugs.php.net/fix.php?id=28873r=isapi Install GNU Sed:http://bugs.php.net/fix.php?id=28873r=gnused Floating point limitations: http://bugs.php.net/fix.php?id=28873r=float
#28856 [Opn]: session_start() causes apache process to hang
ID: 28856 User updated by: pubnelle at megaphone dot ch Reported By: pubnelle at megaphone dot ch Status: Open Bug Type: Session related Operating System: Mac OS X server 10.3.4 PHP Version: 4.3.6 New Comment: Yes, of course, you are right. But to start with, we use the maximum of what php offers us to secure our scripts, so that files that are on our server but not owned by the same user that the one of the current script cannot be accessed, and files that are not in the folder of the virtual host cannot be accessed also. Then, when you need a functionality that is available on this server, but without giving out the configuration of this functionality (passwords to access databases, for example), the way to do it is to access it through http. Evey time that this is possible we avoid the include of http url, of course. But we are extensively using a Content Management System, developed in php, that works great, and that requires those includes to have all in one the correct behavior and the design of the client host. Those includes, once again, are also a protection to scripts that have more rights on databases than we want to give to the virtual host that is including it. By the way, what you do of it is not a problem to me. But the precise information that this problem, that has already be one for other users, occurs exactly because of this session point and precisely when we access it, same session id, twice on the same server, saved in the same session files, but once through an http include, seemed of interest. Sure enough, this could be a security feature, and not a bug - but only if you are aware of it. ;-) Previous Comments: [2004-06-21 20:15:51] [EMAIL PROTECTED] The suggestion there to use include/require was intended to mean a local include/require not an external one. In a typical Apache setup any files accessible to one virtual host is going to be accessible to another. Not directly via http of course, but PHP's include uses the filesystem so it would be something like: include '/var/www/example.com/file.php'; as opposed to: include 'http://example.com/file.php'; The former is an order of magnitude faster than the latter. [2004-06-21 07:52:05] pubnelle at megaphone dot ch That definitely makes sense ! There are plenty of times when you need to include a file that is on the same server ! Here, it is to use a Content Management System, which has more rights on the databases, and which is not directly accessible by our clients. When it is not localhost, it can simply be another virtual host, by the way. We use it often. In the virtual() doc : As of PHP 4.0.6, you can use virtual() on PHP files. However, it is typically better to use include() or require() if you need to include another PHP file. It is not precised if this is for http include or local include, that's true. [2004-06-20 23:03:09] [EMAIL PROTECTED] Why are you using an http include to localhost? That makes no sense. You have fast direct access to this file already. Just include it directly. If you really need it to act like a completely separate request (which you shouldn't) use a sub-request via virtual(). [2004-06-20 19:57:22] pubnelle at megaphone dot ch Description: session_start() causes apache process to hang when a call is made, via http include, to another php script which stores the session files at the same location. It create an http get error in the log files eventually - I had a hard time to find the problem... I upgraded the version of PHP from 4.3.1 built in OS X server to 4.3.6, same problem. I upgraded Apache from 1.3.28 built in OS X server to 1.3.31, same problem. We really need this, to link a shop to a content management system, using the same session information to follow up the caddie stuff ... It helps to use the workaround of bug #22526, to close the session, but is this really the only solution ? this is painfull ... Reproduce code: --- ? session_start(); $urltoinclude=http://localhost/index.php;; include($cmsurl/index.php? . session_name() . = . session_id()); ? Expected result: The file included ;-) To reproduce it, you need to put this code in a page that registers the sessions at the same location than the included URL. Actual result: -- White page hanging ... something appearing eventually (without the include) after some time, exactly 2 minutes. -- Edit this bug report at http://bugs.php.net/?id=28856edit=1
#28788 [Com]: Bogus form data when forms contains a single input field
ID: 28788 Comment by: imprestavel at gameguru dot com dot br Reported By: lsjoberg at aland dot net Status: Open Bug Type: Apache2 related Operating System: Redhat Enterprise Linux PHP Version: 5.0.0RC3 New Comment: Yeah... this is odd. I can't reproduce it here... Tryed running php as module and cgi. Let's wait for a word from php team... Previous Comments: [2004-06-21 08:46:41] lsjoberg at aland dot net I created a simple test page (http://www.vallnet.se/test.htm) which posts forms to both a PHP script which executes print_r($_POST); and a CGI script which prints its stdin. When run on a webserver with PHP as a CGI everything executes as expected. The CGI-script prints 'foo=5' and 'foo=5bar=5' and the PHP-script '[foo] = 5' and '[foo] = 5'/'[bar] = 5'. The CGI output from the Apache server is the same, but for the form with a single element the PHP output is Array ( [foo] = 5foo=5 ) The PHP output for the 2-field form is however correct, hence it seems to be some part of the apache/php interface that is failing... The same browser(s) and exactly the same scripts were used during testing. [2004-06-19 11:56:19] lsjoberg at aland dot net The original code was fully W3C compliant, the testcase below was stripped down before submitted. Browsers used was Mozilla and Opera under FreeBSD and Internet Explorer in some Windows flavour. It feels like a browser bug but since it occured when upgrading from PHP4 to PHP5 without code changes it seems suspicious. I will do some more testing during next week to see if I can isolate the problem further. [2004-06-18 20:33:52] imprestavel at gameguru dot com dot br It looks more like a browser issue. Have you tryed writing conformant html? I tested with Mozilla/rv1.6 and IE/6.0 on Apache/2.0.49, PHP/5.0.0RC3 (register_globals=On and Off), under Windows2000 and didn't have that problem (with the same no- html). What browser are you using? [2004-06-15 11:31:36] lsjoberg at aland dot net Description: When a HTML form contains a single input field (this could be plain INPUT, HIDDEN, SELECT or just about any time) with an associated value the value is not passed correctly to the receiving PHP script. This only happends when the form contains exactly one field and when PHP is run as an Apache2 module. (Could not reproduce with PHP running as CGI under caudium). Reproduce code: --- test.htm: htmlbody form method=POST action=test.php input name=id value=5 input type=submit /form/body/html test.php: ? echo $id; ? Expected result: The expected result is that the php script should print a 5 and nothing else. (We are using register_globals=On). Actual result: -- The output of the php-script is '5id=5' instead of '5'. If a input type=hidden name=foo is added to the html form the correct output of 5 is displayed. -- Edit this bug report at http://bugs.php.net/?id=28788edit=1
#28872 [Opn]: accessing a __get (getter) method from object value fails
ID: 28872
Updated by: [EMAIL PROTECTED]
Reported By: nick dot rich at textmarketer dot co dot uk
Status: Open
Bug Type: Class/Object related
-Operating System: Linux
+Operating System: *
PHP Version: 5.0.0RC3
New Comment:
first of all __get() is supposed to return something.
Previous Comments:
[2004-06-21 19:35:32] nick dot rich at textmarketer dot co dot uk
Description:
Trying to access a propery (via the get method) using a
value from another object as the method name fails.
See code::
if the line /// fails
is
replaced with
$localVar = $config-methodnames[1];
$test-$localVar
It works correctly
Reproduce code:
---
class config
{
public $methodnames = array(dog,cat);
}
class testClass
{
function Cat(){
echo cat;
}
function Dog(){
echo dog;
}
function __get($method){
eval(\$this-\$method(););
}
}
class main
{
function __construct(){
$config = new config();
$test = new testClass();
$test-$config-methodnames[1]; /// fails
}
}
$test = new main();
Expected result:
cat
Actual result:
--
Fatal error: Call to undefined method testClass::
Object() in /Users/nicholasrich/Sites/bits/
testclass.php(18) : eval()'d code on line 1
--
Edit this bug report at http://bugs.php.net/?id=28872edit=1
#28874 [NEW]: Parse error in if (!empty($this-MyFunction())) echo 'Some one';
From: leadaxe at yandex dot ru
Operating system: Win XP
PHP version: 4.3.7
PHP Bug Type: Class/Object related
Bug description: Parse error in if (!empty($this-MyFunction())) echo 'Some one';
Description:
sorry, that this is repeat.
It is as #18274, #8702 bags and we wont solution this problem.
We wont use this construction becouse this have common sense.
We wont use result as argument if it possible.
I create same object and use in metod this code, but it create error.
class MyClass {
function Some_One_Function() {
if (!empty($this-MyFunction())) echo 'Some one';
//this error?
}
function MyFunction() { return true;}
}
But in all changes I very like PHP. Thank you for all.
Reproduce code:
---
class MyClass {
function Some_One_Function() {
if (!empty($this-MyFunction())) echo 'Some one';
//this error?
}
function MyFunction() { return true;}
}
Expected result:
Parse error: parse error, unexpected '(', expecting ')' in ...
And I now use:
class MyClass {
function Some_One_Function() {
$em=empty($this-MyFunction();
if (!$em)) echo 'Some one';
}
function MyFunction() { return true;}
}
But it isn't logic!
Actual result:
--
I think, that if it worked, it would be very convenient.
--
Edit bug report at http://bugs.php.net/?id=28874edit=1
--
Try a CVS snapshot (php4): http://bugs.php.net/fix.php?id=28874r=trysnapshot4
Try a CVS snapshot (php5): http://bugs.php.net/fix.php?id=28874r=trysnapshot5
Fixed in CVS: http://bugs.php.net/fix.php?id=28874r=fixedcvs
Fixed in release: http://bugs.php.net/fix.php?id=28874r=alreadyfixed
Need backtrace: http://bugs.php.net/fix.php?id=28874r=needtrace
Need Reproduce Script: http://bugs.php.net/fix.php?id=28874r=needscript
Try newer version: http://bugs.php.net/fix.php?id=28874r=oldversion
Not developer issue:http://bugs.php.net/fix.php?id=28874r=support
Expected behavior: http://bugs.php.net/fix.php?id=28874r=notwrong
Not enough info:http://bugs.php.net/fix.php?id=28874r=notenoughinfo
Submitted twice:http://bugs.php.net/fix.php?id=28874r=submittedtwice
register_globals: http://bugs.php.net/fix.php?id=28874r=globals
PHP 3 support discontinued: http://bugs.php.net/fix.php?id=28874r=php3
Daylight Savings: http://bugs.php.net/fix.php?id=28874r=dst
IIS Stability: http://bugs.php.net/fix.php?id=28874r=isapi
Install GNU Sed:http://bugs.php.net/fix.php?id=28874r=gnused
Floating point limitations: http://bugs.php.net/fix.php?id=28874r=float
#28868 [Asn]: Thread safty issue with stream_wrapper_register?
ID: 28868
Updated by: [EMAIL PROTECTED]
Reported By: cleong at nflc dot org
Status: Assigned
Bug Type: Filesystem function related
Operating System: Windows 2000
PHP Version: 4.3.7
Assigned To: pollita
New Comment:
The userwrappers part of this bug has been fixed in CVS, please try the
next snapshot available from http://snaps.php.net .
However this particular bug highlights a similar problem with userspace
filters, so I'm going to leave this bugreport open while I address that
side of things. Fortunately since userspace filters are not available
in PHP4, this will only require fixing in this PHP5 branch.
Previous Comments:
[2004-06-21 18:34:55] [EMAIL PROTECTED]
There is absolutely a thread safety issue here. I assume you're
running Apache2?
For the moment, I'd suggest Apache1 or Apache2-pre-fork. In the mean
time we'll be working on fixing this asap.
[2004-06-21 16:53:43] cleong at nflc dot org
Description:
On Windows 2000, with PHP running as an Apache module,
stream_wrapper_register() would fail when there are concurent scripts
using that function. It would either fail without an error message or
causes the Apache process to crash. When there's just one script
running the function works fine.
To reproduce this bug, open a couple browser windows and go to the
reproduce code in each.
I can get around the bug by generating a random class name and protocol
name everytime (using eval for the class definition).
Reproduce code:
---
?
// example class taken from manual
class VariableStream {
var $position;
var $varname;
function stream_open($path, $mode, $options, $opened_path)
{
$url = parse_url($path);
$this-varname = $url[host];
$this-position = 0;
return true;
}
function stream_read($count)
{
$ret = substr($GLOBALS[$this-varname], $this-position,
$count);
$this-position += strlen($ret);
return $ret;
}
function stream_write($data)
{
$left = substr($GLOBALS[$this-varname], 0, $this-position);
$right = substr($GLOBALS[$this-varname], $this-position +
strlen($data));
$GLOBALS[$this-varname] = $left . $data . $right;
$this-position += strlen($data);
return strlen($data);
}
function stream_tell()
{
return $this-position;
}
function stream_eof()
{
return $this-position = strlen($GLOBALS[$this-varname]);
}
function stream_seek($offset, $whence)
{
switch ($whence) {
case SEEK_SET:
if ($offset strlen($GLOBALS[$this-varname])
$offset = 0) {
$this-position = $offset;
return true;
} else {
return false;
}
break;
case SEEK_CUR:
if ($offset = 0) {
$this-position += $offset;
return true;
} else {
return false;
}
break;
case SEEK_END:
if (strlen($GLOBALS[$this-varname]) + $offset = 0) {
$this-position = strlen($GLOBALS[$this-varname])
+ $offset;
return true;
} else {
return false;
}
break;
default:
return false;
}
}
}
stream_wrapper_register(var, VariableStream)
or die(Failed to register protocol);
$myvar = ;
$fp = fopen(var://myvar, r+);
fwrite($fp, line1\n);
fwrite($fp, line2\n);
fwrite($fp, line3\n);
rewind($fp);
while (!feof($fp)) {
echo fgets($fp);
}
session_write_close(); // in case auto-session is on
for($i = 0; $i 30; $i++) {
// make the script run for 30 seconds
sleep(1);
}
fclose($fp);
var_dump($myvar);
?
Expected result:
Same result in each window.
Actual result:
--
While the first script is still running, the second script would say
Failed to register protocol. If you refresh the second window,
sometimes it works, sometimes it causes a general protection error.
--
Edit this bug report at http://bugs.php.net/?id=28868edit=1
#10720 [Com]: Error Parsing Browscap.ini
ID: 10720 Comment by: michael83_2 at hotmail dot com Reported By: istvan at siposs dot com Status: Bogus Bug Type: IIS related Operating System: Windows 2000 Professional PHP Version: 4.0.5 New Comment: I get the same error and it's not fixed with the test.zip file. It makes a empty browscap.ini file. Previous Comments: [2004-05-06 00:20:40] ivan dot roth at free dot fr Maybe you will read this if you were patient for answers! :) www.ivan-roth.com/test.zip this could help you! Extract it in the same directory that browscap.ini and run it. It will creat a browscap.ini.new file. delete browscap.ini and rename browscap.ini.new into browscap.ini. This file will then be compatible with PHP. It worked for me. I hope it'll work for you too. But I hope that you are not blocked by this problem since 3 years! :) Bye, Ivan. [2002-04-06 22:22:20] katsuey at katsuey dot com I also received a browscap/browscap.ini file in my download of PHP Triad and I received the same error notice about line 8206 when attempting to use Komodo (ActiveState.com). I'm on WinXP [2001-05-10 10:35:53] istvan at siposs dot com That's odd. I downloaded the Win32 Zip version from the download page here. I unzipped the package to c:\php, and under this tree a directory called browscap was created with a browscap.ini file. In fact, I'm just looking at the contents of the zip file php-4.0.5-Win32.zip and it has a browscap\browscap.ini in it. This was the browscap.ini file that I was referring to. Istvan Siposs [2001-05-10 05:11:39] [EMAIL PROTECTED] There are no browscap.ini files bundled with PHP. --Jani [2001-05-07 21:08:51] istvan at siposs dot com When I modify my php.ini file and point the browscap configuration setting to the browscap.ini provided with PHP (under c:\php\browscap\browscap.ini) I get the following error at the end of every php page: PHP: Error parsing c:\php\browscap\browscap.ini on line 8206 If I use the default browscap.ini installed by Windows itself, the error does not happen. -- Edit this bug report at http://bugs.php.net/?id=10720edit=1
#28875 [NEW]: Wrong regex crashes server
From: amoljak at cox dot net Operating system: Windows 2000 Server PHP version: 4.3.6 PHP Bug Type: Reproducible crash Bug description: Wrong regex crashes server Description: Note the extra ] in the regex. Try to load this page three times and it will crash the third time. Reproduce code: --- ?php $filecontent = my name is amol; $regex = /(^|[^a-zA-Z0-9-])([Nn]on[-s? []]]degrees? [Ss]tudents?)([^a-zA-Z0-9-]|$)/es; //preg_match_all($regex, $filecontent , $matches, PREG_OFFSET_CAPTURE); preg_match($regex, $filecontent , $matches, PREG_OFFSET_CAPTURE); print matches = $matchesbr; ? Expected result: This is from end user's code and I don't know what he expected. But he was able to crash the server... thats bad. -- Edit bug report at http://bugs.php.net/?id=28875edit=1 -- Try a CVS snapshot (php4): http://bugs.php.net/fix.php?id=28875r=trysnapshot4 Try a CVS snapshot (php5): http://bugs.php.net/fix.php?id=28875r=trysnapshot5 Fixed in CVS: http://bugs.php.net/fix.php?id=28875r=fixedcvs Fixed in release: http://bugs.php.net/fix.php?id=28875r=alreadyfixed Need backtrace: http://bugs.php.net/fix.php?id=28875r=needtrace Need Reproduce Script: http://bugs.php.net/fix.php?id=28875r=needscript Try newer version: http://bugs.php.net/fix.php?id=28875r=oldversion Not developer issue:http://bugs.php.net/fix.php?id=28875r=support Expected behavior: http://bugs.php.net/fix.php?id=28875r=notwrong Not enough info:http://bugs.php.net/fix.php?id=28875r=notenoughinfo Submitted twice:http://bugs.php.net/fix.php?id=28875r=submittedtwice register_globals: http://bugs.php.net/fix.php?id=28875r=globals PHP 3 support discontinued: http://bugs.php.net/fix.php?id=28875r=php3 Daylight Savings: http://bugs.php.net/fix.php?id=28875r=dst IIS Stability: http://bugs.php.net/fix.php?id=28875r=isapi Install GNU Sed:http://bugs.php.net/fix.php?id=28875r=gnused Floating point limitations: http://bugs.php.net/fix.php?id=28875r=float
#28263 [Bgs]: PNG image generation causes Segmentation Fault
ID: 28263 User updated by: Rich dot West at wesmo dot com Reported By: Rich dot West at wesmo dot com Status: Bogus Bug Type: GD related Operating System: Fedora Core 1 PHP Version: 4.3.6 Assigned To: pajoye New Comment: I tried the same process against 4.3.7 today, and the result was the same. :( No PNG image. :( Yes, the libpng and libpng-devel packages are in sync with one another, and yes, the system has the GD libraries installed, and, yes, PHP was compiled it use the bundled GD library rather than the system-wide one. I can get this to work when building 4.3.3, but not with 4.3.5/6/7. I've moved libpdf out to a DSO rather than compiling it in, but it made little difference. :( I'm trying a complete package rebuild using the SRPM from the FC1 updates to see if that will help.. Previous Comments: [2004-05-28 05:23:58] Rich dot West at wesmo dot com I'll check the latest HEAD from CVS. libpdf was compiled in with --with-libpdf. Not sure why it would show up in that particular traceback, though. Possible sources of problems: - You did not update the -devel package for libpng - --with-png-dir=DIR, DIR is not optionnal. Same for jpeg or freetype - Be sure you do not have the GD lib installed in your system. Use only the bundled one. From the first posting, it clearly shows --with-png-dir=/usr and --with-gd. I thought we got past that whole part. I'll test it from the latest CVS HEAD branch as you originally suggested on May 27th, but it is unlikely an installation issue since, as the thread shows, it works with the same libpng libraries if compiled with 4.3.3. Anything later, such as 4.3.5 and 4.3.6, does not work. [2004-05-27 22:01:15] [EMAIL PROTECTED] #1 0x00f1d850 in png_create_struct_2 () from /usr/lib/libpdf.so.1 I'm wondering what the libpdf does here... Please check your install. The original script works with php4 and php5 HEAD and latest RC. Possible sources of problems: - You did not update the -devel package for libpng - --with-png-dir=DIR, DIR is not optionnal. Same for jpeg or freetype - Be sure you do not have the GD lib installed in your system. Use only the bundled one. I set it back to bogus. As it works on 3 different FC1 (and on FC2) without problems. I feel your problem as an install issue only. --Pierre [2004-05-27 13:00:24] duh at dowebwedo dot com I am experiencing the same bug since we upgraded from php 4.3.5 tot 4.3.6, 4.3.7 and the latest snapshot (php4-STABLE-200405270830). I have backtraced the point to where php crashes and that is the imagecreatefrompng function: ... eval(sprintf(\$this-image = imagecreatefrom%s(\$this-image_file);,$this-ext2func[$image_type])); ... (note: i use eval in my class to dynamically open images; however without the evaluation the function still kills my application) However downgrading back to php-4.3.5 does not seem to solve the problem. We are running debian stable with all recent upgrades (libpng etc) with apache (1.3.29) and php and compiled php with the following configure command: './configure' '--with-pgsql' '--with-apxs=/www/bin/apxs' '--with-gd' '--with-png-dir' '--with-freetype-dir' '--with-jpeg-dir' '--with-pear' '--with-zlib-dir' '--enable-track-vars' '--enable-trans-sid' '--disable-posix-threads' '--enable-shared' '--enable-exif' '--enable-memory-limit' I hope this bug will be fixed soon. Grtz, Jeroen [2004-05-23 01:00:09] 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. [2004-05-15 06:30:06] [EMAIL PROTECTED] Change status when you give the feedback asked for. 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/28263 -- Edit this bug report at http://bugs.php.net/?id=28263edit=1
#28854 [Com]: Class and Vars
ID: 28854
Comment by: rodolfo at rodsoft dot org
Reported By: toppi at kacke dot de
Status: Open
Bug Type: *General Issues
Operating System: linux/unix
PHP Version: 5.0.0RC3
New Comment:
The correct way to declare the member variables would be:
var $a. This way, your example would be:
class a
{
var $a;
var $b;
var $a;
}
which would give the correct fatal error:
Fatal error: Cannot redeclare a::$a in /var/www/htdocs/tst.php on line
5
Previous Comments:
[2004-06-20 16:27:55] toppi at kacke dot de
Description:
Class and Vars
Declare a var twice
Script breaks without error when including
Reproduce code:
---
class a {
var a;
var b;
var a;
function a(){
//
}
}
Expected result:
any warning/error
Actual result:
--
script ends/break regular @ including-point
--
Edit this bug report at http://bugs.php.net/?id=28854edit=1
#28870 [Opn]: mysqli_bind_result doesn't work with cloned objects
ID: 28870
User updated by: rodolfo at rodsoft dot org
Reported By: rodolfo at rodsoft dot org
Status: Open
Bug Type: *Database Functions
Operating System: linux 2.6.7
PHP Version: 5.0.0RC3
New Comment:
A possible 'hack' that I've found is to bind the result to another
variable before cloning, and just after that restore the bind to the
original variable. This way the result is the expected one. It seems
that there's a problem of how mysqli_bind_result stores the variable
information while it is being cloned.
Previous Comments:
[2004-06-21 18:35:56] rodolfo at rodsoft dot org
Description:
If you pass to mysqli_bind_result a member variable of an object that
will be cloned and modified multiple times, each modification happens
in all cloned objects, as if they weren't cloned copies, but a single
copy of the object.
Reproduce code:
---
?
/* Say you have a database called 'test' with one table
called 'test' that has one column called 'a' */
$db = new mysqli('localhost');
$db-select_db('test');
$stmt = $db-prepare('SELECT a FROM test');
$stmt-execute();
$result = new stdclass;
// Comment the next line to have the 'right' behaviour
$stmt-bind_result($result-a);
for($i = 0; $i10; ++$i)
{
$result-a = $i;
$results[] = clone $result;
}
foreach($results as $result)
echo $result-a,\n;
$db-close();
?
Expected result:
0
1
2
3
4
5
6
7
8
9
Actual result:
--
9
9
9
9
9
9
9
9
9
9
--
Edit this bug report at http://bugs.php.net/?id=28870edit=1
#28837 [Opn-Fbk]: White page
ID: 28837 Updated by: [EMAIL PROTECTED] Reported By: flotosello at hotmail dot com -Status: Open +Status: Feedback Bug Type: *General Issues Operating System: Linux PHP Version: 4.3.7 New Comment: Not enough information was provided for us to be able to handle this bug. Please re-read the instructions at http://bugs.php.net/how-to-report.php If you can provide more information, feel free to add it to this bug and change the status back to Open. Thank you for your interest in PHP. Previous Comments: [2004-06-18 20:12:16] flotosello at hotmail dot com Description: Hello, After sending a message on my forum, or just clicking on a link, my visitors get a white page. Some people told me this comes from the PHP 4.3.7. Is that true ? Regards Expected result: A white page -- Edit this bug report at http://bugs.php.net/?id=28837edit=1
#28846 [Opn-Fbk]: Apache2 child killed by Illegal instruction (4)
ID: 28846 Updated by: [EMAIL PROTECTED] Reported By: trond at sausjord dot com -Status: Open +Status: Feedback Bug Type: Reproducible crash Operating System: Solaris 9 PHP Version: 4.3.7 New Comment: Not enough information was provided for us to be able to handle this bug. Please re-read the instructions at http://bugs.php.net/how-to-report.php If you can provide more information, feel free to add it to this bug and change the status back to Open. Thank you for your interest in PHP. Please provide a shortest possible script that can be used to replicate the bug. Previous Comments: [2004-06-19 20:13:01] trond at sausjord dot com Description: The apache child dies when I try to run Horde 2.2.5 on Apache 2.0.49. It happens both with prefork and worker. From Apache error log: [Sat Jun 19 18:21:35 2004] [notice] child pid 15489 exit signal Illegal instruction (4) Php build: './configure' '--prefix=/usr/local' '--with-mysql' '--with-apxs2=/usr/local/apache/bin/apxs' '--with-openssl' '--with-gettext' '--with-xml' '--with-imap=../imap-2004/' '--enable-debug' Apache2 build: CFLAGS=-mcpu=athlon-mp -march=athlon-mp -O3; export CFLAGS ./configure \ --prefix=/usr/local/apache \ --enable-ssl \ --enable-auth-dbm \ --enable-so \ --enable-nonportable-atomics=yes \ --with-mpm=worker Reproduce code: --- See index.php in Horde-2.2.5 (www.horde.org) Actual result: -- GDB Backtrace: (gdb) bt #0 0xdd58f262 in zend_highlight (syntax_highlighter_ini=0x81c09a8, tsrm_ls=0x81f6a58) at /usr/local/php-4.3.7/Zend/zend_highlight.c:105 #1 0xddade18d in apr_pool_destroy () from /usr/local/apache/lib/libapr-0.so.0 #2 0xddade248 in apr_pool_destroy () from /usr/local/apache/lib/libapr-0.so.0 #3 0x080ca667 in main () Zend/zend_highlight.c: 105:zend_printf(code); -- Edit this bug report at http://bugs.php.net/?id=28846edit=1
#28876 [Opn-Bgs]: MySQL Stored Procedure
ID: 28876 Updated by: [EMAIL PROTECTED] Reported By: mdf092 at yahoo dot com -Status: Open +Status: Bogus Bug Type: MySQL related Operating System: MS XP PHP Version: 4.3.7 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. Thank you for your interest in PHP. . Previous Comments: [2004-06-21 23:30:43] mdf092 at yahoo dot com Description: Why PHP Doesn't support MySQL Stored Procedure? -- Edit this bug report at http://bugs.php.net/?id=28876edit=1
#28850 [Opn-Bgs]: Passing 0 as expire in setcookie is not the same as passing nothing
ID: 28850 Updated by: [EMAIL PROTECTED] Reported By: fromphpnet at blaksaga dot com -Status: Open +Status: Bogus Bug Type: Session related Operating System: Linux PHP Version: 4.3.7 New Comment: Thank you for taking the time to write to us, but this is not a bug. Please double-check the documentation available at http://www.php.net/manual/ and the instructions on how to report a bug at http://bugs.php.net/how-to-report.php If you want the value to default to whatever PHP defaults to pass NULL. Previous Comments: [2004-06-20 03:36:08] fromphpnet at blaksaga dot com Description: The overall jist of the problem is this. I want to be able to set a cookie and specify the domain and path WITHOUT setting an expiration (making the cookie get killed when the browser is closed). This is what the documentation for setcookie states: All the arguments except the name argument are optional. You may also replace an argument with an empty string () in order to skip that argument. Because the expire and secure arguments are integers, they cannot be skipped with an empty string, use a zero (0) instead. Unfortunately, using a 0 for expire has different results than when passing nothing. The cookie is not killed at session end when using 0 like it should be. These statements should produce exactly the same results but do not: setcookie ( blaksaga_cookies , enabled, 0 ); setcookie ( blaksaga_cookies2 , enabled ); This is either a bug in the setcookie function or in your documentation. None-the-less, there needs to be some way to skip the expire argument while passing domain and path arguments. -- Edit this bug report at http://bugs.php.net/?id=28850edit=1
#28875 [Opn-Fbk]: Wrong regex crashes server
ID: 28875 Updated by: [EMAIL PROTECTED] Reported By: amoljak at cox dot net -Status: Open +Status: Feedback Bug Type: Reproducible crash Operating System: Windows 2000 Server PHP Version: 4.3.6 New Comment: Please try using this CVS snapshot: http://snaps.php.net/php4-STABLE-latest.tar.gz For Windows: http://snaps.php.net/win32/php4-win32-STABLE-latest.zip Previous Comments: [2004-06-21 22:58:29] amoljak at cox dot net Description: Note the extra ] in the regex. Try to load this page three times and it will crash the third time. Reproduce code: --- ?php $filecontent = my name is amol; $regex = /(^|[^a-zA-Z0-9-])([Nn]on[-s? []]]degrees? [Ss]tudents?)([^a-zA-Z0-9-]|$)/es; //preg_match_all($regex, $filecontent , $matches, PREG_OFFSET_CAPTURE); preg_match($regex, $filecontent , $matches, PREG_OFFSET_CAPTURE); print matches = $matchesbr; ? Expected result: This is from end user's code and I don't know what he expected. But he was able to crash the server... thats bad. -- Edit this bug report at http://bugs.php.net/?id=28875edit=1
#28772 [Com]: No input file specified.
ID: 28772 Comment by: jaoshar at yahoo dot fr Reported By: pieter at amazed dot nl Status: Bogus Bug Type: iPlanet related Operating System: win nt 4 PHP Version: 4.3.7 New Comment: No, it is a bug. I use PHP v4.3.5 and it works fine in cmd, but returns the unremovable No input file specified with Apache (2.0.48 or 1.3.39). I've been months on this problem, read all the support/discussion places, but it seems there's no way out of this problem. My personal conclusion is that PHP guys intended to put such a bug in the windows binary (maybe to support products like Sokkit). I stopped trying. When it will be solved, I'll be back to PHP ;o), but learnd lots of things with Perl in the meanwhile ! Previous Comments: [2004-06-14 15:45:43] [EMAIL PROTECTED] 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. Thank you for your interest in PHP. . [2004-06-14 15:29:36] pieter at amazed dot nl Description: I recently installed the latest php 437 win 32 version after associating works fine test.php with cmd want to work under netscape 3.6x enterprise shellcgi all i get is No input file specified Reproduce code: --- no input file specified -- Edit this bug report at http://bugs.php.net/?id=28772edit=1
#28763 [Fbk-NoF]: Tidy extension causing aborted MySQL connections
ID: 28763
Updated by: [EMAIL PROTECTED]
Reported By: johnfivealive at hotmail dot com
-Status: Feedback
+Status: No Feedback
Bug Type: Unknown/Other Function
Operating System: Fedora Core 2
PHP Version: 5.0.0RC3
New Comment:
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.
Previous Comments:
[2004-06-14 01:39:11] [EMAIL PROTECTED]
Check you mysql settings:
http://dev.mysql.com/doc/mysql/en/Communication_errors.html
[2004-06-13 18:02:35] johnfivealive at hotmail dot com
Description:
Whenever I execute a MySQL query I call my escapeString function to
escape strings properly before inserting them into the database. When I
call the tidy_repair_string function from my escapeString function, I
start seeing the following error in the MySQL error log:
Aborted connection 199 to db: 'logik' user: 'logik' host: `localhost'
(Got an error reading communication packets)
This error only seems to occur when I'm attempting to execute a lot of
queries in succession, for example importing a large XML file into my
database.
I think there is some bug between PHP and the Tidy extension. Perhaps
Tidy is taking too long to return or their are some thread
synchronization issues which are causing MySQL to malfunction. I'm
pretty sure it is the Tidy extension that is causing problems here
because when I comment out the call to tidy_repair_string my program
works with no problems whatsoever.
Tidy info as reported by phpinfo():
tidy
Tidy supportenabled
libTidy Release 1st June 2004
Extension Version 2.0-dev ($Id: tidy.c,v 1.53 2004/05/28 20:32:52 john
Exp $)
Reproduce code:
---
This is how I'm calling tidy_repair_string:
public static function tidyString( $string )
{
$tidyOptions = array( bare = true,
break-before-br = true,
char-encoding = latin1,
clean = true,
doctype = omit,
drop-proprietary-attributes = true,
logical-emphasis = false,
output-xhtml = true,
quote-ampersand = true,
show-body-only = true,
word-2000 = true,
wrap = 0 );
// Get rid of all funky characters
return tidy_repair_string( $string, $tidyOptions ); // PHP 5
}
Expected result:
Tidy should work in sequence with MySQL queries.
Actual result:
--
The correct string is returned, but when queries need to be executed
quickly and in succession things start getting erratic and MySQL says:
Aborted connection 199 to db: 'logik' user: 'logik' host: `localhost'
(Got an error reading communication packets)
Also PHP outputs the following warning:
Warning: Unknown list entry type in request shutdown (18545152) in
Unknown on line 0
Other times, the script never finishes executing, thus no output is
produced, warnings or otherwise.
There are also MySQL warnings, but these are a result of Tidy not
returning the correct string, or perhaps not returning in time.
When I don't call the tidy function, there are not warnings or
problems.
--
Edit this bug report at http://bugs.php.net/?id=28763edit=1
#28875 [Fbk-Csd]: Wrong regex crashes server
ID: 28875 User updated by: amoljak at cox dot net Reported By: amoljak at cox dot net -Status: Feedback +Status: Closed Bug Type: Reproducible crash Operating System: Windows 2000 Server PHP Version: 4.3.6 New Comment: I tried PHP 4.3.8-dev and it is not crashing. So I am closing the bug. I will reopen if it occures on the machine I initially tried. Thanks Previous Comments: [2004-06-22 00:45:31] [EMAIL PROTECTED] Please try using this CVS snapshot: http://snaps.php.net/php4-STABLE-latest.tar.gz For Windows: http://snaps.php.net/win32/php4-win32-STABLE-latest.zip [2004-06-21 22:58:29] amoljak at cox dot net Description: Note the extra ] in the regex. Try to load this page three times and it will crash the third time. Reproduce code: --- ?php $filecontent = my name is amol; $regex = /(^|[^a-zA-Z0-9-])([Nn]on[-s? []]]degrees? [Ss]tudents?)([^a-zA-Z0-9-]|$)/es; //preg_match_all($regex, $filecontent , $matches, PREG_OFFSET_CAPTURE); preg_match($regex, $filecontent , $matches, PREG_OFFSET_CAPTURE); print matches = $matchesbr; ? Expected result: This is from end user's code and I don't know what he expected. But he was able to crash the server... thats bad. -- Edit this bug report at http://bugs.php.net/?id=28875edit=1
#28263 [Bgs-Csd]: PNG image generation causes Segmentation Fault
ID: 28263 User updated by: Rich dot West at wesmo dot com Reported By: Rich dot West at wesmo dot com -Status: Bogus +Status: Closed Bug Type: GD related Operating System: Fedora Core 1 PHP Version: 4.3.6 Assigned To: pajoye New Comment: Fixed! I rebuilt the whole package based upon the 4.3.6 SRPM, updated it for 4.3.7, and viola! It works! Previous Comments: [2004-06-21 23:20:23] Rich dot West at wesmo dot com I tried the same process against 4.3.7 today, and the result was the same. :( No PNG image. :( Yes, the libpng and libpng-devel packages are in sync with one another, and yes, the system has the GD libraries installed, and, yes, PHP was compiled it use the bundled GD library rather than the system-wide one. I can get this to work when building 4.3.3, but not with 4.3.5/6/7. I've moved libpdf out to a DSO rather than compiling it in, but it made little difference. :( I'm trying a complete package rebuild using the SRPM from the FC1 updates to see if that will help.. [2004-05-28 05:23:58] Rich dot West at wesmo dot com I'll check the latest HEAD from CVS. libpdf was compiled in with --with-libpdf. Not sure why it would show up in that particular traceback, though. Possible sources of problems: - You did not update the -devel package for libpng - --with-png-dir=DIR, DIR is not optionnal. Same for jpeg or freetype - Be sure you do not have the GD lib installed in your system. Use only the bundled one. From the first posting, it clearly shows --with-png-dir=/usr and --with-gd. I thought we got past that whole part. I'll test it from the latest CVS HEAD branch as you originally suggested on May 27th, but it is unlikely an installation issue since, as the thread shows, it works with the same libpng libraries if compiled with 4.3.3. Anything later, such as 4.3.5 and 4.3.6, does not work. [2004-05-27 22:01:15] [EMAIL PROTECTED] #1 0x00f1d850 in png_create_struct_2 () from /usr/lib/libpdf.so.1 I'm wondering what the libpdf does here... Please check your install. The original script works with php4 and php5 HEAD and latest RC. Possible sources of problems: - You did not update the -devel package for libpng - --with-png-dir=DIR, DIR is not optionnal. Same for jpeg or freetype - Be sure you do not have the GD lib installed in your system. Use only the bundled one. I set it back to bogus. As it works on 3 different FC1 (and on FC2) without problems. I feel your problem as an install issue only. --Pierre [2004-05-27 13:00:24] duh at dowebwedo dot com I am experiencing the same bug since we upgraded from php 4.3.5 tot 4.3.6, 4.3.7 and the latest snapshot (php4-STABLE-200405270830). I have backtraced the point to where php crashes and that is the imagecreatefrompng function: ... eval(sprintf(\$this-image = imagecreatefrom%s(\$this-image_file);,$this-ext2func[$image_type])); ... (note: i use eval in my class to dynamically open images; however without the evaluation the function still kills my application) However downgrading back to php-4.3.5 does not seem to solve the problem. We are running debian stable with all recent upgrades (libpng etc) with apache (1.3.29) and php and compiled php with the following configure command: './configure' '--with-pgsql' '--with-apxs=/www/bin/apxs' '--with-gd' '--with-png-dir' '--with-freetype-dir' '--with-jpeg-dir' '--with-pear' '--with-zlib-dir' '--enable-track-vars' '--enable-trans-sid' '--disable-posix-threads' '--enable-shared' '--enable-exif' '--enable-memory-limit' I hope this bug will be fixed soon. Grtz, Jeroen [2004-05-23 01:00:09] 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. 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/28263 -- Edit this bug report at http://bugs.php.net/?id=28263edit=1
#28877 [Com]: The First 4 letters missing When Post
ID: 28877 Comment by: scottm at spamcop dot net Reported By: jli at pegasuspharm dot com Status: Open Bug Type: Feature/Change Request Operating System: MS Windows 2000 PHP Version: 4.3.6 New Comment: Can you confirm your php version as this sounds like a bug in php 4.2.3 See http://bugs.php.net/bug.php?id=19404 Previous Comments: [2004-06-22 03:01:48] jli at pegasuspharm dot com Description: The First 4 letters missing with the following codes When Post:(Note: the text fileds below is an array like desc[]) form name=service method=post action=?php $_SERVER[PHP_SELF]; ? onSubmit=return checkEnter(); input name=Desc[] type=text size=40 maxlength=50 value=?php echo $_POST[Desc][0]; ? input name=Desc[] type=text size=40 maxlength=50 value=?php echo $_POST[Desc][1]; ? input name=Desc[] type=text size=40 maxlength=50 value=?php echo $_POST[Desc][2]; ? /form Reproduce code: --- form name=service method=post action=?php $_SERVER[PHP_SELF]; ? onSubmit=return checkEnter(); input name=Desc[] type=text size=40 maxlength=50 value=?php echo $_POST[Desc][0]; ? input name=Desc[] type=text size=40 maxlength=50 value=?php echo $_POST[Desc][1]; ? input name=Desc[] type=text size=40 maxlength=50 value=?php echo $_POST[Desc][2]; ? /form Actual result: -- The first 4 letters provided in a field missing. -- Edit this bug report at http://bugs.php.net/?id=28877edit=1
#28850 [Bgs]: Passing 0 as expire in setcookie is not the same as passing nothing
ID: 28850 User updated by: fromphpnet at blaksaga dot com Reported By: fromphpnet at blaksaga dot com Status: Bogus Bug Type: Session related Operating System: Linux PHP Version: 4.3.7 New Comment: I'm sorry, but in my eyes this is a bug. If nothing else, it is a documentation bug. - All the arguments except the name argument are optional. You may also replace an argument with an empty string () in order to skip that argument. Because the expire and secure arguments are integers, they cannot be skipped with an empty string, use a zero (0) instead. - Unless I am reading it wrong, the documentation is incorrect. But thank you for your time. I will post a comment on the doc page about this. If you could include/change it in the docs it would be appreciated. Thanks, blaksaga Previous Comments: [2004-06-22 00:44:59] [EMAIL PROTECTED] Thank you for taking the time to write to us, but this is not a bug. Please double-check the documentation available at http://www.php.net/manual/ and the instructions on how to report a bug at http://bugs.php.net/how-to-report.php If you want the value to default to whatever PHP defaults to pass NULL. [2004-06-20 03:36:08] fromphpnet at blaksaga dot com Description: The overall jist of the problem is this. I want to be able to set a cookie and specify the domain and path WITHOUT setting an expiration (making the cookie get killed when the browser is closed). This is what the documentation for setcookie states: All the arguments except the name argument are optional. You may also replace an argument with an empty string () in order to skip that argument. Because the expire and secure arguments are integers, they cannot be skipped with an empty string, use a zero (0) instead. Unfortunately, using a 0 for expire has different results than when passing nothing. The cookie is not killed at session end when using 0 like it should be. These statements should produce exactly the same results but do not: setcookie ( blaksaga_cookies , enabled, 0 ); setcookie ( blaksaga_cookies2 , enabled ); This is either a bug in the setcookie function or in your documentation. None-the-less, there needs to be some way to skip the expire argument while passing domain and path arguments. -- Edit this bug report at http://bugs.php.net/?id=28850edit=1
#22328 [Com]: Erroneous behaviour of fetchRow method
ID: 22328
Comment by: tko at lostway dot org
Reported By: miki_fossati at libero dot it
Status: Bogus
Bug Type: MySQL related
Operating System: linux suse 7.2
PHP Version: 4.3.3RC2-dev
Assigned To: georg
New Comment:
mysql bug #706 is not a bug. It is a specification.
So you should check mysql_errno().
A client user should reset errno somehow (I think of re-connecting db)
if he needs.
Previous Comments:
[2003-06-24 07:29:23] [EMAIL PROTECTED]
It's a bug in libmysql which is already reported (mysql bug #
706)
[2003-06-23 19:30:59] [EMAIL PROTECTED]
Happens with PHP 4.3.3RC2-dev, Georg? :)
[2003-06-21 11:40:51] [EMAIL PROTECTED]
Ok, I've removed the mysql_errno() check on fetchInto().
IMHO this is a bug in the mysql extension, which doesn't properly reset
the error. This problem doesn't ocurr with postgres or interbase.
Test script to reproduce the problem:
?php
$con = mysql_connect('localhost', 'root');
mysql_selectdb('test');
$res = mysql_query('SELECT * FROM table');
$continue = true;
do {
$row = mysql_fetch_row($res);
if (mysql_errno()) {
// Here there is no fetch_row error, it's the same
// from FAKE QUERY
echo Failed fetching with error: . mysql_errno() . \n;
// Don't want to continue on fetch errors
$continue = false;
} elseif (!$row) {
$continue = false;
} else {
$res2 = mysql_query('FAKE QUERY');
echo Failed to launch FAKE QUERY: . mysql_errno() . \n;
// Ok, I got the error but want to continue
$continue = true;
}
} while ($continue);
?
Output:
Failed to launch FAKE QUERY: 1064
Failed fetching with error: 1064
[2003-05-11 01:19:07] [EMAIL PROTECTED]
Actually lsmith's fix seems to be fine. Is there anything speaking
against applying it?
[2003-04-29 15:14:00] [EMAIL PROTECTED]
the problem is in the implementation of fetchInto:
function fetchInto($result, $arr, $fetchmode, $rownum=null)
{
if ($rownum !== null) {
if ([EMAIL PROTECTED]($result, $rownum)) {
return null;
}
}
if ($fetchmode DB_FETCHMODE_ASSOC) {
$arr = @mysql_fetch_array($result, MYSQL_ASSOC);
} else {
$arr = @mysql_fetch_row($result);
}
if (!$arr) {
$errno = @mysql_errno($this-connection);
if (!$errno) {
return NULL;
}
return $this-mysqlRaiseError($errno);
}
return DB_OK;
}
this line picks up the error that was meant for the FAKE QUERY and
think by mistake that therefore not the end of the result set was
reached but an error occured.
Since during a fetch only a connection error can occur it might be a
good idea to remove the mysql_errno() call in order to fix this.
Alternative we could check for a connection error:
if (!$arr) {
$errno = @mysql_errno($this-connection);
if ($errno == 2013) {
return $this-mysqlRaiseError($errno);
}
return NULL;
}
$errno = @mysql_errno($this-connection);
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/22328
--
Edit this bug report at http://bugs.php.net/?id=22328edit=1
#28878 [NEW]: Setting of inikey's in obj.conf fails
From: michaelw at webcentral dot com dot au
Operating system: Solaris 9
PHP version: 4.3.7
PHP Bug Type: iPlanet related
Bug description: Setting of inikey's in obj.conf fails
Description:
Attempting to set ini values within the obj.conf file (or vserver.obj.conf
if thats where your virtual hosts are pointed) fails silently.
This appears to be caused by the logic in sapi/nsapi/nsapi.c ( static void
nsapi_php_ini_entries(NSLS_D TSRMLS_DC) ) , in particular the following
snippet:
--- SNIP ---
ok=1;
for (j=0; nsapi_exclude_from_ini_entries[j]; j++) {
ok=(!strcasecmp(entry-param-name,nsapi_exclude_from_ini_entries[j]));
}
if (ok) {
--- SNIP ---
It appears that ok is never returning as TRUE, and thus the code never
enters the if (ok) block.
I modified it as follows (although I'm sure there are better ways..):
--- SNIP ---
ok=1;
for (j=0; nsapi_exclude_from_ini_entries[j]; j++) {
if (!strcasecmp(entry-param-name,nsapi_exclude_from_ini_entries[j]))
{
ok = 0;
}
}
if (ok) {
--- SNIP ---
Additionally, the code that actually updates the ini value in the same
method):
--- SNIP ---
if (ok) {
/* change the ini entry */
if
(zend_alter_ini_entry(entry-param-name,strlen(entry-param-name)+1,entry-param-value,strlen(entry-param-value),PHP_INI_USER,PHP_INI_STAGE_RUNTIME)==FAILURE)
{
log_error(LOG_WARN, pblock_findval(fn, NSG(pb)),NSG(sn), NSG(rq),
Cannot change php.ini key \%s\ to \%s\, entry-param-name,
entry-param-value);
}
}
--- SNIP ---
checks the method to see if it is PHP_INI_USER allowed, whilst since it is
in a system file (obj.conf or vserver.obj.conf) I would expect it to test
against PHP_INI_SYSTEM ?
--
Edit bug report at http://bugs.php.net/?id=28878edit=1
--
Try a CVS snapshot (php4): http://bugs.php.net/fix.php?id=28878r=trysnapshot4
Try a CVS snapshot (php5): http://bugs.php.net/fix.php?id=28878r=trysnapshot5
Fixed in CVS: http://bugs.php.net/fix.php?id=28878r=fixedcvs
Fixed in release: http://bugs.php.net/fix.php?id=28878r=alreadyfixed
Need backtrace: http://bugs.php.net/fix.php?id=28878r=needtrace
Need Reproduce Script: http://bugs.php.net/fix.php?id=28878r=needscript
Try newer version: http://bugs.php.net/fix.php?id=28878r=oldversion
Not developer issue:http://bugs.php.net/fix.php?id=28878r=support
Expected behavior: http://bugs.php.net/fix.php?id=28878r=notwrong
Not enough info:http://bugs.php.net/fix.php?id=28878r=notenoughinfo
Submitted twice:http://bugs.php.net/fix.php?id=28878r=submittedtwice
register_globals: http://bugs.php.net/fix.php?id=28878r=globals
PHP 3 support discontinued: http://bugs.php.net/fix.php?id=28878r=php3
Daylight Savings: http://bugs.php.net/fix.php?id=28878r=dst
IIS Stability: http://bugs.php.net/fix.php?id=28878r=isapi
Install GNU Sed:http://bugs.php.net/fix.php?id=28878r=gnused
Floating point limitations: http://bugs.php.net/fix.php?id=28878r=float
#28826 [Opn-Bgs]: SimpleXML expands entities but DOM + XSL does not
ID: 28826 User updated by: joe at joerags dot com Reported By: joe at joerags dot com -Status: Open +Status: Bogus Bug Type: XSLT related Operating System: Mac OS 10.2.8 PHP Version: 5CVS-2004-06-18 (dev) New Comment: I did the following and the test script I submitted works. Using Fink (http://fink.sourceforge.net), I had it build and install libxml2 2.5.10 and libxslt 1.0.32. I then recompiled PHP 5 RC3 with these configuration commands: ./configure --prefix=/usr/local --with-zlib --with-libxml-dir=/sw --with-xsl=/sw --with-apxs=/usr/ sbin/apxs The test script I submitted likely didn't work either because of user error on my part or a bug in something other than PHP. Previous Comments: [2004-06-18 11:09:05] joe at joerags dot com Description: When attempting to do an XSL transformation using the new DOM and XSL extensions available in PHP 5, I've noticed entities aren't expanded in the output. If I import the DOM object into SimpleXML, the entities are expanded. The first time I attempted to run the following code I was using PHP 5 RC3, libxml2 2.6.10, and libxslt 1.1.7. I am now using CVS versions of PHP 5, libxml2, and libxslt in order to see if this was a bug fixed in either PHP 5, libxml2, or libxslt. The CVS versions I'm using were downloaded June 18th. I run Apache 1.3.31 under Mac OS 10.2.8. I don't recall seeing any errors when I built and installed Apache, PHP, libxml2, or libxslt. Reproduce code: --- ?php $xml = EOD !DOCTYPE example [ !ENTITY entity1 Entity 1 !ENTITY entity2 Entity 2 ] example element1Example Document/element1 element2entity1;/element2 element3entity2;/element3 /example EOD; $xsl = EOD xsl:stylesheet version=1.0 xmlns:xsl=http://www.w3.org/1999/XSL/Transform; xsl:template match=example pXSL Element 1: xsl:value-of select=element1 //p pXSL Element 2: xsl:value-of select=element2 //p pXSL Element 3: xsl:value-of select=element3 //p /xsl:template /xsl:stylesheet EOD; $xmlDocument = new DomDocument; $xmlDocument-loadXML($xml); $xslDocument = new DomDocument; $xslDocument-loadXML($xsl); $processor = new XsltProcessor; $processor-importStyleSheet($xslDocument); print $processor-transformToXML($xmlDocument); $xmlDocument = simplexml_import_dom($xmlDocument); print 'pSimpleXML Element 1: ' . $xmlDocument-element1 . '/p'; print 'pSimpleXML Element 2: ' . $xmlDocument-element2 . '/p'; print 'pSimpleXML Element 3: ' . $xmlDocument-element3 . '/p'; ? Expected result: XSL Element 1: Example Document XSL Element 2: Entity 1 XSL Element 3: Entity 2 SimpleXML Element 1: Example Document SimpleXML Element 2: Entity 1 SimpleXML Element 3: Entity 2 Actual result: -- XSL Element 1: Example Document XSL Element 2: XSL Element 3: SimpleXML Element 1: Example Document SimpleXML Element 2: Entity 1 SimpleXML Element 3: Entity 2 -- Edit this bug report at http://bugs.php.net/?id=28826edit=1
#28263 [Csd-Bgs]: PNG image generation causes Segmentation Fault
ID: 28263 Updated by: [EMAIL PROTECTED] Reported By: Rich dot West at wesmo dot com -Status: Closed +Status: Bogus Bug Type: GD related Operating System: Fedora Core 1 PHP Version: 4.3.6 Assigned To: pajoye New Comment: Not a bug in PHP then, so marking it bogus. Previous Comments: [2004-06-22 01:35:28] Rich dot West at wesmo dot com Fixed! I rebuilt the whole package based upon the 4.3.6 SRPM, updated it for 4.3.7, and viola! It works! [2004-06-21 23:20:23] Rich dot West at wesmo dot com I tried the same process against 4.3.7 today, and the result was the same. :( No PNG image. :( Yes, the libpng and libpng-devel packages are in sync with one another, and yes, the system has the GD libraries installed, and, yes, PHP was compiled it use the bundled GD library rather than the system-wide one. I can get this to work when building 4.3.3, but not with 4.3.5/6/7. I've moved libpdf out to a DSO rather than compiling it in, but it made little difference. :( I'm trying a complete package rebuild using the SRPM from the FC1 updates to see if that will help.. [2004-05-28 05:23:58] Rich dot West at wesmo dot com I'll check the latest HEAD from CVS. libpdf was compiled in with --with-libpdf. Not sure why it would show up in that particular traceback, though. Possible sources of problems: - You did not update the -devel package for libpng - --with-png-dir=DIR, DIR is not optionnal. Same for jpeg or freetype - Be sure you do not have the GD lib installed in your system. Use only the bundled one. From the first posting, it clearly shows --with-png-dir=/usr and --with-gd. I thought we got past that whole part. I'll test it from the latest CVS HEAD branch as you originally suggested on May 27th, but it is unlikely an installation issue since, as the thread shows, it works with the same libpng libraries if compiled with 4.3.3. Anything later, such as 4.3.5 and 4.3.6, does not work. [2004-05-27 22:01:15] [EMAIL PROTECTED] #1 0x00f1d850 in png_create_struct_2 () from /usr/lib/libpdf.so.1 I'm wondering what the libpdf does here... Please check your install. The original script works with php4 and php5 HEAD and latest RC. Possible sources of problems: - You did not update the -devel package for libpng - --with-png-dir=DIR, DIR is not optionnal. Same for jpeg or freetype - Be sure you do not have the GD lib installed in your system. Use only the bundled one. I set it back to bogus. As it works on 3 different FC1 (and on FC2) without problems. I feel your problem as an install issue only. --Pierre [2004-05-27 13:00:24] duh at dowebwedo dot com I am experiencing the same bug since we upgraded from php 4.3.5 tot 4.3.6, 4.3.7 and the latest snapshot (php4-STABLE-200405270830). I have backtraced the point to where php crashes and that is the imagecreatefrompng function: ... eval(sprintf(\$this-image = imagecreatefrom%s(\$this-image_file);,$this-ext2func[$image_type])); ... (note: i use eval in my class to dynamically open images; however without the evaluation the function still kills my application) However downgrading back to php-4.3.5 does not seem to solve the problem. We are running debian stable with all recent upgrades (libpng etc) with apache (1.3.29) and php and compiled php with the following configure command: './configure' '--with-pgsql' '--with-apxs=/www/bin/apxs' '--with-gd' '--with-png-dir' '--with-freetype-dir' '--with-jpeg-dir' '--with-pear' '--with-zlib-dir' '--enable-track-vars' '--enable-trans-sid' '--disable-posix-threads' '--enable-shared' '--enable-exif' '--enable-memory-limit' I hope this bug will be fixed soon. Grtz, Jeroen 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/28263 -- Edit this bug report at http://bugs.php.net/?id=28263edit=1
