Thies C. Arntzen [Wed, 15 Aug 2001 16:00:56 +0000 (16:00 +0000)]
mega patch to avoid recursive oci-calls as they might break oracle!
php will simply exit() (!!!) once it encounters an attemp to call oracle
while still in an oracle-call. this can only happen when the script gets
interrupted (timeout, apache-restart) and php tries to run the cleanups even
though we are in an unclean state. i will add an -ini flag soon (not before
4.0.7) which will (optional) make php not run the cleanups when in an unclean
state!
Sterling Hughes [Wed, 15 Aug 2001 05:21:43 +0000 (05:21 +0000)]
Removed select(), fd_set(), etc. (wasn't present in the first place, and
these are implemented in the sockwasn't present in the first place, and
these are implemented in the sockets extension))
Stig Bakken [Tue, 14 Aug 2001 21:11:33 +0000 (21:11 +0000)]
* expected error codes are stored on a stack now, PEAR::expectError()
pushes into it, PEAR::popExpect() pops off it
* made pear_error test work on other boxes than my own :-)
foobar [Tue, 14 Aug 2001 11:07:18 +0000 (11:07 +0000)]
@- Fixed a bug in dbase_get_record() and dbase_get_record_with_names().
@ boolean fields are now returned correctly.
@ Patch by Lawrence E. Widman <widman@cardiothink.com> (Jani)
Alan Brown [Tue, 14 Aug 2001 03:52:32 +0000 (03:52 +0000)]
Whenever typelibrary constants were loaded, there was a GP Fault at process exit as shutdown_memory_manager seemed to delete constants that were supposed to have been deleted earlier. It seems as though CONST_PERSISTENT is a bad thing to use. Resetting that bit on the constant creation still seems to work and the GP Fault at exit is gone.
Sascha Schumann [Mon, 13 Aug 2001 21:55:21 +0000 (21:55 +0000)]
improved thttpd-2.21b patch.. I got tired of applying my patches for
these bugs.
fixes poll(2) issue
fixes hanging cgi issue
fixes off-by-one in scanning input buffers in case of EAGAIN/EWOULDBLOCK
fixes potential bug in managing write buffers
add "index.php" to default files to look for
Zeev Suraski [Mon, 13 Aug 2001 19:31:18 +0000 (19:31 +0000)]
Fix crashes in parse_parameters calls
HEADS UP: Make sure you supply TSRMLS_CC for this function!
I'll try to think of a way that'd allow us to find this issue using
the compiler.
Rasmus Lerdorf [Mon, 13 Aug 2001 06:43:47 +0000 (06:43 +0000)]
We don't consistently check for args passed to functions that don't
take any args. In some cases we probably want to skip the check for
performance reasons, but in other cases where performance is unlikely
to be a factor, not throwing a warning on the wrong number of args passed
to a function is at best inconsistent, and at worst it could hide a bug.
So, add a few such checks. There are still lots of cases out there.
Thies C. Arntzen [Sun, 12 Aug 2001 18:44:42 +0000 (18:44 +0000)]
tags need to be UTF8-decoded as well.
guys, case_folding can only work when the parser target encoding equal (or
compatible) with the locale setting of the system as we use toupper to "do it"