Re: [Php-java-bridge-users] Having some trouble with hung apache children and an older version of java-bridge

[<php...@li...>]

php...@li... wrote:
> Hi
> 
> please see our FAQ "How do I enable the Java extension in my php.ini file?"
> (http://php-java-bridge.sourceforge.net/pjb/FAQ.html)
> 
> 

I'm confused how that is relevant to my question. I must not have 
explained my problem sufficiently in my previous post.

We are running 4.3.3 fine now. We were previously running a version 
around 4.0.1.

When we upgraded to apache 2.2.9 using the same java.so we found that 
apache processes were hanging and not exiting. I got a backtrace from 
them that looked like:

(gdb) backtrace
#0  0x00002b800b4ed38f in read () from /lib/libc.so.6
#1  0x00002aaaad00e380 in java_shutdown_library () from 
/usr/lib/php5/20060613/java.so
#2  0x00002aaaad001e54 in zm_shutdown_java () from 
/usr/lib/php5/20060613/java.so
#3  0x00002aaaac0a17b1 in module_destructor () from 
/usr/lib/apache2/modules/libphp5.so
#4  0x00002aaaac0a7f32 in zend_hash_quick_find () from 
/usr/lib/apache2/modules/libphp5.so
#5  0x00002aaaac0a81a8 in zend_hash_graceful_reverse_destroy () from 
/usr/lib/apache2/modules/libphp5.so
#6  0x00002aaaac09db27 in zend_shutdown () from 
/usr/lib/apache2/modules/libphp5.so
#7  0x00002aaaac05b68a in php_module_shutdown () from 
/usr/lib/apache2/modules/libphp5.so
#8  0x00002aaaac05b729 in php_module_shutdown_wrapper () from 
/usr/lib/apache2/modules/libphp5.so
#9  0x00002aaaac11e911 in zend_get_zval_ptr () from 
/usr/lib/apache2/modules/libphp5.so
#10 0x00002b800b20a1ad in apr_pool_cleanup_run () from 
/usr/lib/libapr-1.so.0
#11 0x00002b800b20a7bd in apr_pool_destroy () from /usr/lib/libapr-1.so.0
#12 0x00000000004462be in ap_graceful_stop_signalled ()
#13 0x000000000044671a in ap_graceful_stop_signalled ()
#14 0x0000000000446904 in ap_graceful_stop_signalled ()
#15 0x0000000000446d83 in ap_mpm_run ()
#16 0x0000000000421027 in main ()

This seems to be identical to the discussion I previously referenced.

Things seem to have changed significantly in the 5 stream. We will 
change our code to work with the new stream, eventually but it doesn't 
seem trivial.