Jeroen van Wolffelaar
c033288573
Back-substitute for Z_* macro's. If it breaks some extension (the script isn't optimal, it parses for example var->zval.value incorrect) please let me know.
2001-09-25 21:58:48 +00:00
Sascha Schumann
9002e37e4e
Always seed the LCG from the request-init hook, otherwise the seed
...
would be shared among the threads which is quite pointless. Also
use a function of the current time as one factor.
Use gettimeofday() instead of time(), because it is faster on some
operating systems.
2001-09-21 13:38:44 +00:00
Sascha Schumann
919e7e9f69
revert last commit. Guys, discuss such things first, then commit. Mkay?
2001-09-11 10:16:01 +00:00
Sterling Hughes
e302d6a5e8
Add an API function for setting the session id from a session module...
...
# The SRM folks wanted/needed this
2001-09-11 10:13:44 +00:00
Derick Rethans
78747bd2df
- Don't wrap lines... this is annoying while coding.
2001-09-09 13:29:31 +00:00
Zeev Suraski
f7e7b42d4b
Restore session_adapt_url()
2001-08-31 20:03:09 +00:00
Zeev Suraski
b080ddb867
Cleanup and clarify
2001-08-31 19:26:12 +00:00
Zeev Suraski
642eab8b46
define_sid updates
2001-08-31 19:22:08 +00:00
Zeev Suraski
b2b50ab525
Fix trans_sid handler registration logic
2001-08-31 15:22:10 +00:00
Zeev Suraski
d7149ba8fc
Cleanup
2001-08-27 02:06:32 +00:00
Zeev Suraski
b781720b36
More fixes and refinements
2001-08-27 00:23:55 +00:00
Zeev Suraski
d697a558ac
Fix my session updates.
...
trans_sid handling will no longer slow down performance unless a session
is actually in use.
2001-08-26 11:49:43 +00:00
Zeev Suraski
dc9769ea96
Fix trans-sid.
...
Still need to figure out how to turn trans-sid only when sessions
are actually used.
2001-08-26 11:14:35 +00:00
Zeev Suraski
7b73aee40f
Make the session module more independent
2001-08-26 01:09:20 +00:00
Zeev Suraski
d5e5e376ab
This option is no longer needed
2001-08-25 16:19:57 +00:00
Zeev Suraski
b8763a6f8c
Get rid of TRANS_SID - it's always possible to use it now.
2001-08-25 16:09:23 +00:00
Zeev Suraski
13ff183225
Move the trans-sid mechanism to use the output buffering mechanism.
...
Advantages:
- Would work with the output buffering mechanism in general, and with
output compression in particular
- Should yield better performance (untested as of yet)
2001-08-25 15:57:42 +00:00
Zeev Suraski
bc596544a1
add struct name
2001-08-25 12:47:12 +00:00
Zeev Suraski
e140b35b04
API update
2001-08-21 12:57:53 +00:00
Rasmus Lerdorf
4d11d90880
Track down a few more functions that don't check for 0 args and use
...
faster mechanism
2001-08-13 07:55:39 +00:00
Rasmus Lerdorf
8c497f05c4
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.
2001-08-13 06:43:47 +00:00
foobar
3619c6eb6a
Added $_SESSION to the $_* family.
2001-08-11 10:49:03 +00:00
Sascha Schumann
23b9300fd1
more tsrm cleanup
2001-08-06 03:50:52 +00:00
Sascha Schumann
a19563c8dc
more tsrm cleanup
2001-08-05 16:52:29 +00:00
Sascha Schumann
480ffb3d75
more tsrm cleanup
2001-08-05 16:43:57 +00:00
Sascha Schumann
373b3e101e
more tsrm cleanup -- output.c is not doing any fetches anymore
2001-08-05 15:55:43 +00:00
Sascha Schumann
8aef193056
more tsrm cleanup
2001-08-05 15:29:47 +00:00
Sascha Schumann
aafdfe82cb
further tsrm cleanup
2001-08-05 14:36:51 +00:00
Zeev Suraski
1159c84ab7
- TSRMLS_FETCH work
...
- whitespace fixes
2001-08-05 01:43:02 +00:00
Sascha Schumann
723bbcbea0
Do the nul'ifiying of php_var_serialize's result at the right place
2001-08-03 11:00:16 +00:00
Sascha Schumann
4a1943c542
Pass on nul'ified data
2001-08-03 10:39:48 +00:00
Sascha Schumann
60024a469d
Add some casts
2001-08-03 08:51:53 +00:00
Sascha Schumann
4dfa91543c
Convert serializer to smart_str.. avoids lots of sprintf's and
...
copying of data.
2001-08-03 07:25:27 +00:00
Zeev Suraski
c43806f415
Zend compatibility patch
2001-07-30 08:24:42 +00:00
Zeev Suraski
797a079a95
More TSRMLS_FETCH work, and a bit of cleanup
2001-07-30 06:18:13 +00:00
Zeev Suraski
d87cc976e1
Redesigned thread safety mechanism - nua nua
2001-07-28 11:36:37 +00:00
Zeev Suraski
fe6f8712a4
- Get rid of ELS_*(), and use TSRMLS_*() instead.
...
- Move to the new ts_allocate_id() API
This patch is *bound* to break some files, as I must have had typos somewhere.
If you use any uncommon extension, please try to build it...
2001-07-27 10:16:41 +00:00
Sascha Schumann
00596b414e
Get rid of memchr
2001-07-25 19:02:13 +00:00
Hartmut Holzgraefe
d3fdc92ff5
these tow fail with "Constant sid already defined" i
...
too if E_NOTICE error reporting is active
2001-07-22 21:46:39 +00:00
Zeev Suraski
6f297a370a
Fix Windows build
2001-07-20 14:40:30 +00:00
Sascha Schumann
ef42a25b42
Fix a serious long-standing bug which deleted the shared memory segment
...
when an Apache children died.
Dynamically resize hash-table.
The hash key is now stored as part of the hash entry structure which
saves a mm_malloc.
ps_sd_new has been restructured to avoid code duplication.
A bug concerning moving existing entries to the top of a linked list
has been fixed.
2001-07-19 04:16:54 +00:00
Sascha Schumann
664df24d11
Fix GC bug which did not increment the proper variable.
...
Fix warning regarding the initial hash value.
Use php_uint32 everywhere.
Speedup linked-list traversal by using the hash value.
Degrade linked-list to singledly linked-list.
Allocate a new shared memory buffer, if the data does not fit into the
current one. Prior to this change, we always allocated a new buffer.
Fix debug prints to be compatible with non-GCC compilers.
NUL-terminate string return values.
2001-07-18 23:18:58 +00:00
Sascha Schumann
9576ab24dd
Instead of using the poor system rand() algorithm and the
...
system-time seed, rely on the long-living LCG to determine
whether to start the GC process.
2001-07-18 22:56:56 +00:00
Sascha Schumann
eef0dda4da
Retire old hash algorithm and use the FNV-1 algorithm.
2001-07-18 21:39:02 +00:00
Thies C. Arntzen
249c010d92
make regession test of session work without the ugly hack in session.c
...
and removed that ugly hack.
2001-07-13 10:32:37 +00:00
Hartmut Holzgraefe
319149fd8b
workaround to make ext/session well behaved with 'make test' and E_ALL
2001-07-11 13:46:11 +00:00
Hartmut Holzgraefe
e15d0193d9
cleanup in preparation for session_readonly()
2001-07-11 13:27:27 +00:00
Sascha Schumann
ef13ab5750
Improve url scanner speed (up to 40% for large chunks of data)
...
and handle some corner cases better. The scanner has been changed
to the format as proposed in "RE2C - A More Versatile Scanner Generator"
by Cowan et al.
2001-07-04 15:30:21 +00:00
Hartmut Holzgraefe
7f08918b87
suppress sending of cookies if session id already came
...
from a non-cookie source
2001-06-29 05:27:57 +00:00
foobar
2c95a1d690
The extension name seems to be usually totally lower-cased.
2001-06-24 22:18:51 +00:00