| Commit message (Collapse) | Author | Age | Files | Lines |
|---|
| | |
|
| | |
|
| | |
|
| | |
|
| |
|
|
|
|
| |
https://github.com/tony2001/php-src"
This reverts commit a6be0f3fd6cdd59ac00ecd76630c6c04fee03417.
|
| | |
|
| | |
|
| |
|
|
| |
Simplify zval comparion API.
|
| | |
|
| |
|
|
|
| |
Given it's an issue with 32-bit gcc-4.8, asm is disabled for the
breaking parts.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Detecting overflow with the XER is slow, partially because we have to
clear it before use.
PHP already has a fast way of detecting overflow in its fallback
c implementation. Overflow only occurs if the signs of the two
operands are the same and the sign of the result is different.
Furthermore, leaving it in c allows gcc to schedule the instructions
better.
This is 9% faster on a POWER8 running a simple testcase:
<?php
function testcase($count = 100000000) {
$x = 1;
for ($i = 0; $i < $count; $i++) {
$x = $x + 1;
$x = $x + 1;
$x = $x + 1;
$x = $x + 1;
$x = $x + 1;
}
}
testcase();
?>
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Detecting overflow with the XER is slow, partially because we have to
clear it before use.
gcc does a better job of detecting overflow of an increment or decrement
than we can with inline assembly. It knows that an increment will only
overflow if it is one less than the overflow value. This means we end
up with a simple compare/branch. Furthermore, leaving it in c allows gcc
to schedule the instructions better.
This is 6% faster on a POWER8 running a simple testcase:
<?php
function testcase($count = 100000000) {
$x = 1;
for ($i = 0; $i < $count; $i++) {
$x++;
$x++;
$x++;
$x++;
$x++;
}
}
testcase();
?>
|
| |
|
|
| |
Also added extra const annotation to zend_memnstr.
|
| | |
|
| |
|
|
|
| |
compare_function() now has to be compatible with binary_op_type (use fastcall convention).
Introduced new zval_compare_function() to be used as zval comparison callback instead of compare_function().
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
This only involves switching zval_dtor to zval_ptr_dtor for arrays
and making the convert_to_object for arrays a bit more generic.
All the other changes outside zend_operators.c just make use of
this new ability (use COPY instead of DUP).
What's still missing: Proper references handling. I've seen many
convert_to* calls that will break when a reference is used.
Also fixes bug #69788.
|
| | |
|
| |
|
|
|
|
| |
takeing additional arguments.
Pair INSTANCEOF with the following JMPZ/JMPNZ.
|
| |
|
|
| |
returns +/-INF, modulo by zero and intdid() throws an exception, shifts by negative offset throw exceptions. Compile-time evaluation of division by zero is disabled.
|
| |
|
|
| |
fix for Windows 9x that Anatol changed a while back
|
| | |
|
| |
|
|
| |
with additional optimizations
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| |
|
|
| |
eliminate repeatable checks on each loop iteration in in_array() function.
|
| |
|
|
| |
It simplifies code and avoids unnecessary allocation and copying if string is already in lower case.
|
| | |
|
| | |
|
| | |
|
| | |
|
| |\
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
* zppFailOnOverflow:
Fix MySQLi tests
Fixed gd test
Refactor ZEND_LONG_MAX/MIN checks into ZEND_DOUBLE_FITS_LONG()
Fixed copy-and-paste error
Fix more 32-bit tests
Skip buncha tests on 32-bit
skip simplexml
skip posix 32-bit
skip tests on 32-bit
Fixes simplexml test
Fixes posix tests
Fixes iconv tests
Marked tests as 32-bit
Fixed more 32-bit tests
Fixed some 32-bit tests
Mark said ext/date tests as 32-bit only
Fixed ext/date tests broken by zpp error on overflow
Fixed broken tests
Make zpp fail if NaN passed for int, or out-of-range float for non-capping int
Conflicts:
ext/date/tests/getdate_variation7.phpt
ext/date/tests/localtime_variation3.phpt
|
| | | |
|
| | |
| |
| |
| |
| |
| |
| |
| | |
This adds fast architecture-specific implementations of the
following functions for the ppc64:
* fast_add_function
* fast_sub_function
|
| | |
| |
| |
| |
| |
| |
| |
| | |
This adds fast architecture-specific implementations of
the following functions for the ppc64:
* fast_increment_function
* fast_decrement_function
|
| |/
|
|
|
|
|
| |
zend_operators.h/zend_operators.c.
Splited the most expensive part of inline i_zend_is_true() into a separate zend_object_is_true().
Replaced zendi_convert_to_long() with cals to zend_is_true().
|
| |
|
|
| |
of corresponding operators
|
| | |
|
| | |
|
| |\ |
|
| | |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Now that zend_ulong is 64bit on 64bit platforms, it should be
sufficient to always use it, rather than supporting multiple
types.
API changes:
* _zend_print_unsigned_to_buf and _zend_print_signed_to_buf
no longer exist.
* smart_str(ing)_print_long and smart_str(ing)_print_unsigned
no longer exist.
* Instead of all these, zend_print_ulong_to_buf and
zend_print_long_to_buf should be used.
* smart_str_append_generic_ex no longer exists.
* smart_str(ing)_append_off_t(_ex) no longer exists, use
smart_str(ing)_append_long(_ex) instead.
|
| | | |
|
| | | |
|
| |\ \
| |/
| |
| |
| | |
Conflicts:
Zend/zend_operators.h
|
| | | |
|
| | | |
|
