Chuck Hagenbuch [Thu, 17 Aug 2000 16:01:14 +0000 (16:01 +0000)]
prevent mail_criteria from munging the variable passed in containing the
query - just had to add a cpystr(). if this isn't the right fix, let me
know, but it does work.
Ben Mansell [Thu, 17 Aug 2000 13:40:40 +0000 (13:40 +0000)]
(send_headers) Removed code freeing http_status_line, this is performed
in sapi_send_headers()
# Looks like a similar problem in sapi/pi3web/pi3web_sapi.c as well.
# I can't test this though, so I'm being paranoid and not changing that code.
# Could someone please check this?
Andi Gutmans [Wed, 16 Aug 2000 19:26:21 +0000 (19:26 +0000)]
- The beginning of an attempt to cleanup fopen-wrappers.
- I started with trying to localize the V_FOPEN's so that we can have a
- version which won't really open the file for include_once/require_once to
- work faster and have less chance of a race which would cause a fd leak.
- What I did will, therefore, change but I want to do this step by step
- because the code is extremley messy so first of all I want to make sure
- that the isolating of the V_FOPEN code doesn't break anything.
- How about moving URL stuff out of this file?
- php_fopen_url_wrapper() copy and pasted the second part of
- php_fopen_wrapper() (incorrectly). Please try not to copy&paste code but
- centralize functionality. Need to think of a nice way to nuke one of the
- copies and have both functions use the same one.
Jouni Ahto [Wed, 16 Aug 2000 01:03:20 +0000 (01:03 +0000)]
- Return NULL type when column is NULL, now that we have it in PHP4.
- Fixed a bug in handling NULL columns, PQgetvalue() doesn't return a NULL
pointer in this case, but a pointer to a string of 0 length...
David Croft [Tue, 15 Aug 2000 09:47:35 +0000 (09:47 +0000)]
renamed ini section, quoted default host and tabulated.
# it doesn't make sense to have the tab-width set to 4 when we're
# using tabs for alignment in this case, it just makes it look screwy
# in everything but emacs
David Croft [Tue, 15 Aug 2000 09:37:39 +0000 (09:37 +0000)]
display ini file entries on phpinfo(), unregister them on shutdown
numeric ini file entries stored as ints
removed unnecessary casts from PFPROG()
pval -> zval
Sascha Schumann [Mon, 14 Aug 2000 19:17:16 +0000 (19:17 +0000)]
Fix check for libmcrypt in non-system library paths.
Fix case where neither libmcrypt 2.4.x and 2.2.x were found.
Fix library order (scope limited to static libraries)
Fix another crash with long formats (#6145)
# From strncpy manual:
# In the case where the length of src is less than that of
# n, the remainder of dest will be padded with nulls.
Andi Gutmans [Thu, 10 Aug 2000 20:42:57 +0000 (20:42 +0000)]
@ Fix switch which only has a default rule (Andi, Zend Engine)
@ Change require_once() to use the same file list as include_once().
@ Patch includes making require() & include() to behave the same when it
@ comes to scoping. require() is now an include() which isn't allowed to fail.
@ require() caused too many memory reallocations which ended up being quite
@ slow for sites that required lots of files. (Andi & Zeev, Zend Engine)
- Fix switch() which only has default rule (bug #5879,
- Homogonized CCVS with the rest of the modules in the ext/ tree:
- All ZEND_* macro calls changed to PHP_* macros.
- All zval's converted to pval's.
# Per discussion with Hartmut Holzgraefe on [PHP-DEV] list.
Zeev Suraski [Wed, 9 Aug 2000 19:22:35 +0000 (19:22 +0000)]
The patch we promised - redesigned the compilation/execution API:
Advantages:
- Smaller memory footprint for the op arrays
- Slightly faster compilation times (due to saved erealloc() calls and faster zend_op
initialization)
- include_once() & require_once() share the same file list
- Consistency between include() and require() - this mostly means that return()
works inside require()'d files just as it does in include() files (it used to
be meaningless in require()'d files, most of the time (see below))
- Made require() consistent with itself. Before, if the argument was not a constant
string, require() took the include() behavior (with return()).
- Removed lots of duplicate code.
Bottom line - require() and include() are very similar now; require() is simply an include()
which isn't allowed to fail. Due to the erealloc() calls for large op arrays, require()
didn't end up being any faster than include() in the Zend engine.
Updated Supported CCVS Versions based on Conversation with Doug. All versions of CCVS 3.0 and greater should work without problems.
@ The CCVS module is now stable and compiling. It compiles as a CGI and into apache cleanly without warnings. (Brendan W. McAdams)
Looks like I found my problem.
I was predeclaring my functions in ccvs.h (stupid of me), so PHP was seeing the predec from internal_functions.h, the predec again, and then the actual functions and barfing. Compiles into apache now. Will test further.