Commit Graph

72965 Commits

Author SHA1 Message Date
Anatol Belski
b89c326578 Merge branch 'PHP-5.4' into PHP-5.5
* PHP-5.4:
  updated NEWS
  This commit was manufactured by cvs2svn to create tag 'NEWS'.
2013-03-03 19:15:37 +01:00
Anatol Belski
00ecc12c0f updated NEWS 2013-03-03 19:14:50 +01:00
Anatol Belski
faf14e9b2e Merge commit 'NEWS' into PHP-5.4
* commit 'NEWS':
  This commit was manufactured by cvs2svn to create tag 'NEWS'.
2013-03-03 19:14:17 +01:00
Anatol Belski
1cda84a986 Merge branch 'PHP-5.4' into PHP-5.5
* PHP-5.4:
  merged PCRE 8.32
2013-03-03 19:10:12 +01:00
Anatol Belski
ca4a3e6138 Merge branch 'PHP-5.3' into PHP-5.4
* PHP-5.3:
  merged PCRE 8.32
2013-03-03 19:07:49 +01:00
Anatol Belski
357ab3cbad merged PCRE 8.32
- defined HAVE_CONFIG_H to simplify the future merges
- PCRE NEWS http://pcre.org/news.txt
2013-03-03 19:06:22 +01:00
Derick Rethans
c596850af6 Merge branch 'PHP-5.4' into PHP-5.5
Conflicts:
	NEWS
2013-03-03 12:58:42 -05:00
Derick Rethans
a3f020a4be Merge branch 'PHP-5.3' into PHP-5.4 2013-03-03 12:53:23 -05:00
Derick Rethans
c134f2a1a5 - Updated to version 2013.1 (2013a) 2013-03-03 12:53:22 -05:00
Xinchen Hui
46e612ada4 Merge branch 'PHP-5.4' into PHP-5.5 2013-03-03 11:25:29 +08:00
Xinchen Hui
fe5ce5cb91 Merge branch 'PHP-5.4' of git.php.net:php-src into PHP-5.4 2013-03-03 11:25:08 +08:00
Xinchen Hui
9bda6a25a5 Fix test (skip on unable to create socket) 2013-03-03 11:24:45 +08:00
Xinchen Hui
2413fa6fbb Fix test after the convert_to_long be improved, see #64290 2013-03-03 11:13:52 +08:00
Xinchen Hui
505c6ca486 Fixed bug #64290 (behavior change of converting to long)
convert_to_long behavior has been improved
As Cataphrace said in IRC, the test should be fixed.
2013-03-03 10:59:15 +08:00
Xinchen Hui
f2246f352f Fixed Bug #64340 (sockets tests failed) Patch by Reeze
Add __APPLE_USE_RFC_3542 for OS X 10.7 because IPv6 options are incompatible with RFC 2292.
2013-03-03 10:51:04 +08:00
Dmitry Stogov
ecf0cbdf3b Fixed timestamp update on Phar contents modification. 2013-03-01 19:17:25 +04:00
Dmitry Stogov
b6656d0ec9 Fixed timestamp update on Phar contents modification. 2013-03-01 19:17:08 +04:00
Stanislav Malyshev
99b373da25 Merge branch 'PHP-5.4' into PHP-5.5
* PHP-5.4:
  5.4.24-dev
  5.3.23RC1
2013-02-28 21:30:02 -08:00
Stanislav Malyshev
03dfa08da9 Merge branch 'PHP-5.3' into PHP-5.4
* PHP-5.3:
  5.4.24-dev
  5.3.23RC1

Conflicts:
	configure.in
	main/php_version.h
2013-02-28 21:29:40 -08:00
Stanislav Malyshev
5c5bc7cf1b 5.4.24-dev 2013-02-28 21:25:36 -08:00
Stanislav Malyshev
8f533d2dfd 5.3.23RC1 2013-02-28 21:22:06 -08:00
Stanislav Malyshev
0d65a85cc0 Merge branch 'PHP-5.4' into PHP-5.5
* PHP-5.4:
  5.4.13 branched, now it's 5.4.14

Conflicts:
	configure.in
	main/php_version.h
2013-02-26 22:26:36 -08:00
Stanislav Malyshev
31a6f8b7f4 5.4.13 branched, now it's 5.4.14 2013-02-26 22:25:33 -08:00
Stanislav Malyshev
84b65cf7ab Merge branch 'PHP-5.4' into PHP-5.5
* PHP-5.4:
  add news for SOAP fixes
2013-02-26 22:20:50 -08:00
Stanislav Malyshev
7ffb77d243 add news for SOAP fixes 2013-02-26 22:20:21 -08:00
Stanislav Malyshev
207d0ee08a Merge branch 'PHP-5.4' into PHP-5.5
* PHP-5.4:
  Bug #52861: unset fails with ArrayObject and deep arrays
2013-02-26 22:13:01 -08:00
Stanislav Malyshev
7c08232509 Merge branch 'PHP-5.3' into PHP-5.4
* PHP-5.3:
  Bug #52861: unset fails with ArrayObject and deep arrays
2013-02-26 22:12:36 -08:00
Mike Willbanks
61099f8585 Bug #52861: unset fails with ArrayObject and deep arrays
When checking to make into a reference write, readwrite are checked but not unset
2013-02-26 22:11:52 -08:00
Remi Collet
7a3933a93b Merge branch 'PHP-5.4' into PHP-5.5
* PHP-5.4:
  add support for ppc64p7 arch (Power7 optimized)
2013-02-26 14:35:47 +01:00
Remi Collet
52d1add0fe add support for ppc64p7 arch (Power7 optimized) 2013-02-26 14:35:34 +01:00
Xinchen Hui
e75804c34e Merge branch 'PHP-5.4' into PHP-5.5 2013-02-24 12:59:10 +08:00
Xinchen Hui
d24ac6953e Merge branch 'PHP-5.3' into PHP-5.4
Conflicts:
	ext/spl/spl_fixedarray.c
2013-02-24 12:58:39 +08:00
Xinchen Hui
1b58bd39a6 Fixed bug #64264 (SPLFixedArray toArray problem) 2013-02-24 12:55:53 +08:00
Gustavo Lopes
59ec22b370 sendrecvmsg_shutdown function moved to mshutdown
The function php_socket_sendrecvmsg_shutdown() should have been called in
MSHUTDOWN, not RSHUTDOWN.

Bug only on TSRM builds.

Should fix bug #64287.
2013-02-24 03:50:38 +01:00
Gustavo Lopes
189fbfd53b Fix ext/sockets build with phpize
Fixes bug #64285
2013-02-24 00:49:25 +01:00
Gustavo Lopes
97f22702c4 Fix ext/intl build with phpize
Fixes bug #64284.
2013-02-24 00:49:01 +01:00
Gustavo Lopes
fa3cb8608c NEWS for 8718755 2013-02-23 18:07:01 +01:00
Gustavo Lopes
12745594a4 Remove trailing whitespace; break long lines 2013-02-23 18:01:52 +01:00
Gustavo Lopes
8718755c9f Merge branch 'dval_to_lval' into PHP-5.5
* dval_to_lval:
  Fix rounding of zend_dval_to_lval
  Fix zend_dval_to_lval outside 64bit integers range
2013-02-23 17:59:28 +01:00
Gustavo Lopes
a86fcfbc1d Fix rounding of zend_dval_to_lval
Now rounding is always towards zero; the rounding can be though as if
occurring on the original double. Only relevant for the 32-bit long
variant.
2013-02-23 16:55:50 +00:00
Gustavo Lopes
77566edbaf Fix zend_dval_to_lval outside 64bit integers range
PHP should preserve the least significant bits when casting from double
to long. Zend.m4 contains this:

AC_DEFINE([ZEND_DVAL_TO_LVAL_CAST_OK], 1, [Define if double cast to long preserves least significant bits])

If ZEND_DVAL_TO_LVAL_CAST_OK is not defined, zend_operators.h had an
inline implementation of zend_dval_to_lval() that would do a cast to an
int64_t (when sizeof(long) == 4), then a cast to unsigned long and
finally the cast to long.

While this works well for doubles inside the range of values of the type
used in the first cast (int64_t in the 32-bit version and unsigned long
in the 64-bit version), if outside the range, it is undefined behavior
that WILL give varying and not particularly useful results.

This commit uses fmod() to first put the double in a range that can
safely be cast to unsigned long and then casts this unsigned long to
long. This last cast is implementation defined, but it's very likely
that this gives the expected result (i.e. the internal 2's complement
representation is unchanged) on all platforms that PHP supports. In any
case, the previous implementationa already had this assumption.

This alternative code path is indeed significantly slower than simply
casting the double (almost an order of magnitude), but that should not
matter because casting doubles with a very high absolute value is a
rare event.
2013-02-23 17:23:49 +01:00
Andrey Hristov
64a2a8a753 relax the check 2013-02-22 13:50:04 +01:00
Andrey Hristov
1643d4ed56 Add some parameter checks 2013-02-22 13:17:46 +01:00
Johannes Schlüter
b479c2bffb Merge branch 'PHP-5.4' into PHP-5.5 2013-02-21 16:50:59 +01:00
Johannes Schlüter
064c62e4cf Fix #60840 (undefined symbol: mysqlnd_debug_std_no_trace_funcs) 2013-02-21 16:49:28 +01:00
Johannes Schlüter
7b719235b3 Merge branch 'PHP-5.4' into PHP-5.5 2013-02-21 15:10:47 +01:00
Johannes Schlüter
1e9bc56775 Merge branch 'PHP-5.3' into PHP-5.4 2013-02-21 15:10:15 +01:00
Johannes Schlüter
fcd4b5335a Fix TSRM (after afc1debb) 2013-02-21 15:08:58 +01:00
Dmitry Stogov
8b7222e22f Merge branch 'PHP-5.4' into PHP-5.5
* PHP-5.4:
  PHAR should modify entry timestamp when it's modified
2013-02-21 16:08:07 +04:00
Dmitry Stogov
b5d267aa95 Merge branch 'PHP-5.3' into PHP-5.4
* PHP-5.3:
  PHAR should modify entry timestamp when it's modified
2013-02-21 15:57:26 +04:00