Chuck Hagenbuch [Thu, 27 Jan 2000 16:38:45 +0000 (16:38 +0000)]
Mcal cleanup, stage 1: it compiles again (the php3 syntax to php4/zend
syntax conversion left a lot of: PHP_FUNCTION void ... definitions, giving
two types to every function).
However, it's still segfaulting. I'll look into that next.
Sam Ruby [Thu, 27 Jan 2000 13:44:45 +0000 (13:44 +0000)]
PHP now starts up in CGI mode with Kaffe on RedHat Linux 6.1.
status: segfault atexit, npe in jawt - but otherwise both samples work!
Installation needs more work
Uwe Steinmann [Tue, 25 Jan 2000 17:00:09 +0000 (17:00 +0000)]
- resumed work on domxml and found a way to circumwent a segm fault
(see the source). No solution to fix it, yet. Probably needs some
understanding of zend, which I don't have.
- added more functions
Andi Gutmans [Mon, 24 Jan 2000 19:04:07 +0000 (19:04 +0000)]
- Make foreach() now copy the array but use the original array. It can
still be optimized A LOT but it's only a performance issue and not
a feature issue.
Sam Ruby [Wed, 19 Jan 2000 14:45:26 +0000 (14:45 +0000)]
First step to restoring the ability to build Java support on Unix:
1) Restore the ability to build CGI as a shared library
2) Correct make dependency rule so that the jar files are included
3) Search for jar executable, and use it, if found
Zeev Suraski [Sun, 16 Jan 2000 21:03:49 +0000 (21:03 +0000)]
- Remove the pre-request-shutdown hooks, they should no longer be necessary.
The session module may not work due to an unknown (and until now, IMO, misdiagnosed) problem,
if it does, please report it!
Zeev Suraski [Sun, 16 Jan 2000 20:59:03 +0000 (20:59 +0000)]
- Make zend_hash_apply() (and friends) reentrant and much, much quicker
- Introduce zend_hash_graceful_destroy(), which allows the destructor functions to
use zend_hash_apply() and/or zend_hash_graceful_destroy()
- Switch to zend_hash_graceful_destroy() in the resource list shutdowns
Zeev Suraski [Sun, 16 Jan 2000 19:43:34 +0000 (19:43 +0000)]
Remove the post-startup callback support. Modules are now started only after
the entire engine is started. People - please make sure that the session module
works properly for you.
Thies C. Arntzen [Sun, 16 Jan 2000 13:30:57 +0000 (13:30 +0000)]
@- OCI8 now supports binding of NULL-values. Module cleanups. (Thies)
lots of cleanup - all invisible to the user (i hope;-)
- full refcounting of connections & statements
- cleanup of the descriptor-code (descriptors are now real resources)
- we now use SQLT_CHR instead of SQLT_STR (no trailing \0) - smarter & faster
- ocifetchinto(OCI_BOTH) will use references.
Andrei Zmievski [Sat, 15 Jan 2000 19:41:17 +0000 (19:41 +0000)]
Added ability to set timeout on socket read operations. New function
set_socket_timeout(socket, seconds, microseconds) sets the timeout.
@- Added ability to set timeout on socket read operations through
@ set_socket_timeout() function. (Andrei)
Thies C. Arntzen [Sat, 15 Jan 2000 13:40:17 +0000 (13:40 +0000)]
if ZEND_DEBUG mode is on we'll now see warnings when a HashTable is accessed
while it's inconsistent.
Zeev, Andi - you welcome to revert this patch if you don't like it - i find it
useful! accesssing inconsistent hashtables is one of the hardest things to track!
Thies C. Arntzen [Sat, 15 Jan 2000 13:02:54 +0000 (13:02 +0000)]
- don't set php_errormsg on errors that will cause a zend_bailout().
using zend_hash_update() can make things worse in this situation.
- new function php_register_pre_request_shutdown(). this way modules
can register callbacks that will be called as soon as execution of
the script is done but *before* any cleanup (global symbol_table etc)
has taken place.
Zeev Suraski [Thu, 13 Jan 2000 17:37:25 +0000 (17:37 +0000)]
- Added flush() support to SAPI
- Got rid of the old flush() implemenetation in favour of the new one
- Added implicit_flush() support to the output buffering layer.
@- Added implicit_flush() to control whether flush() should be called
@ implicitly after any output (Zeev)