Commit message (Collapse) | Author | Age | Files | Lines | ||
---|---|---|---|---|---|---|
... | ||||||
* | | | | revert the local speedup approach | Anatol Belski | 2014-10-01 | 1 | -5/+2 | |
| | | | | ||||||
* | | | | approach local pointers in performance sensitive areas | krakjoe | 2014-09-27 | 1 | -2/+5 | |
| | | | | ||||||
* | | | | cleanup TSRMLS_FETCH | Anatol Belski | 2014-09-26 | 1 | -30/+1 | |
| | | | | ||||||
* | | | | reworked the patch, less new stuff but worky | Anatol Belski | 2014-09-25 | 1 | -2/+2 | |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | TLS is already used in TSRM, the way exporting the tsrm cache through a thread local variable is not portable. Additionally, the current patch suffers from bugs which are hard to find, but prevent it to be worky with apache. What is done here is mainly uses the idea from the RFC patch, but - __thread variable is removed - offset math and declarations are removed - extra macros and definitions are removed What is done merely is - use an inline function to access the tsrm cache. The function uses the portable tsrm_tls_get macro which is cheap - all the TSRM_* macros are set to placebo. Thus this opens the way remove them later Except that, the logic is old. TSRMLS_FETCH will have to be done once per thread, then tsrm_get_ls_cache() can be used. Things seeming to be worky are cli, cli server and apache. I also tried to enable bz2 shared and it has worked out of the box. The change is yet minimal diffing to the current master bus is a worky start, IMHO. Though will have to recheck the other previously done SAPIs - embed and cgi. The offsets can be added to the tsrm_resource_type struct, then it'll not be needed to declare them in the userspace. Even the "done" member type can be changed to int16 or smaller, then adding the offset as int16 will not change the struct size. As well on the todo might be removing the hashed storage, thread_id != thread_id and linked list logic in favour of the explicit TLS operations. | |||||
* | | | | native tls initial patch | krakjoe | 2014-09-20 | 1 | -2/+2 | |
|/ / / | ||||||
* | | | make the return value more logic | Anatol Belski | 2014-09-20 | 1 | -1/+1 | |
| | | | | | | | | | | | | as if clz() were returning 32 on empty size | |||||
* | | | make use of intrinsics available with VS for bitset lookups | Anatol Belski | 2014-09-20 | 1 | -0/+35 | |
| | | | ||||||
* | | | Expose zend_safe_address() and use it in zend_arena_calloc() | Dmitry Stogov | 2014-09-18 | 1 | -111/+5 | |
| | | | ||||||
* | | | better error messages in ZMM on Windows | Anatol Belski | 2014-09-16 | 1 | -2/+30 | |
| | | | ||||||
* | | | emalloc() must not return NULL | Dmitry Stogov | 2014-09-16 | 1 | -0/+14 | |
| | | | ||||||
* | | | Fixed comments and removed useless code. | Dmitry Stogov | 2014-09-08 | 1 | -2/+1 | |
| | | | ||||||
* | | | Revert "Move safe_address into zend_multiply.h" | Nikita Popov | 2014-09-06 | 1 | -4/+124 | |
| | | | | | | | | | | | | | | | | | | | | | This reverts commit 43b02e169c718335a82b8c90c3252194f988659a. I'm getting symbol resolution issues for zend_error_noreturn in the opcache.so, so reverting this for now. | |||||
* | | | Move safe_address into zend_multiply.h | Nikita Popov | 2014-09-06 | 1 | -124/+4 | |
| | | | | | | | | | | | | | | | | | | | | | This allows reuse in zend_arena_calloc. Also renamed the function to zend_safe_address as it's no longer private to the allocator. | |||||
* | | | fix number format | Anatol Belski | 2014-08-28 | 1 | -2/+2 | |
| | | | ||||||
* | | | fixed bitset computing on win64 build | Anatol Belski | 2014-08-26 | 1 | -13/+12 | |
| | | | ||||||
* | | | Fixed zero-size allocation in debug build | Dmitry Stogov | 2014-08-26 | 1 | -0/+2 | |
| | | | ||||||
* | | | temporary fix for win64 | Anatol Belski | 2014-08-26 | 1 | -1/+2 | |
| | | | ||||||
* | | | Replaced Zend Memory Manager | Dmitry Stogov | 2014-08-26 | 1 | -2107/+1690 | |
| | | | ||||||
* | | | several signature and data type fixes | Anatol Belski | 2014-08-26 | 1 | -2/+2 | |
| | | | ||||||
* | | | master renames phase 3 | Anatol Belski | 2014-08-25 | 1 | -4/+4 | |
| | | | ||||||
* | | | master renames phase 2 | Anatol Belski | 2014-08-25 | 1 | -1/+1 | |
| | | | ||||||
* | | | master renames phase 1 | Anatol Belski | 2014-08-25 | 1 | -23/+23 | |
| | | | ||||||
* | | | further fixes on core | Anatol Belski | 2014-08-16 | 1 | -5/+5 | |
| | | | ||||||
* | | | first shot on merging the core fro the int64 branch | Anatol Belski | 2014-08-16 | 1 | -22/+22 | |
| | | | ||||||
* | | | Revert "stash" (pushed by accident) | Xinchen Hui | 2014-07-18 | 1 | -2007/+2008 | |
| | | | | | | | | | | | | This reverts commit a18bf5fc9303e2739ee7026bdb3bbf272bf4baa8. | |||||
* | | | stash | Xinchen Hui | 2014-07-16 | 1 | -2008/+2007 | |
| | | | ||||||
* | | | Simplify call-frame handling | Dmitry Stogov | 2014-07-07 | 1 | -1/+1 | |
| | | | ||||||
* | | | Removed EG(in_execution). If EG(currentent_execute_data) is not NULL we are ↵ | Dmitry Stogov | 2014-07-03 | 1 | -1/+1 | |
| | | | | | | | | | | | | executing something. | |||||
* | | | Removed EG(active_op_array) and use corresponding value from ↵ | Dmitry Stogov | 2014-07-02 | 1 | -3/+13 | |
| | | | | | | | | | | | | EG(current_execute_data) | |||||
* | | | Use better data structures (incomplete) | Dmitry Stogov | 2014-02-10 | 1 | -2/+3 | |
| | | | ||||||
* | | | Merge branch 'PHP-5.6' | Xinchen Hui | 2014-01-03 | 1 | -1/+1 | |
|\ \ \ | ||||||
| * | | | Bump year | Xinchen Hui | 2014-01-03 | 1 | -1/+1 | |
| |/ / | ||||||
* | | | Merge branch 'PHP-5.6' | Dmitry Stogov | 2013-12-26 | 1 | -1/+1 | |
|\ \ \ | |/ / | | | | | | | | | | * PHP-5.6: Fixed ZEND_MM_MEM_TYPE=mmap_zero | |||||
| * | | Fixed ZEND_MM_MEM_TYPE=mmap_zero | Dmitry Stogov | 2013-12-26 | 1 | -1/+1 | |
| | | | ||||||
* | | | Eliminated the TSRMLS_FETCH() in zend_set_memory_limit() | Kalle Sommer Nielsen | 2013-11-18 | 1 | -3/+1 | |
|/ / | ||||||
* | | Fix assembly of safe_address() for aarch64 | Ard Biesheuvel | 2013-07-04 | 1 | -1/+1 | |
| | | ||||||
* | | fix assembly of safe_address() for x86 and x86_64 | Nuno Lopes | 2013-07-03 | 1 | -2/+2 | |
| | | ||||||
* | | Merge branch 'PHP-5.4' into PHP-5.5 | Gustavo Lopes | 2013-04-30 | 1 | -4/+13 | |
|\ \ | |/ | | | | | | | | | | | | | * PHP-5.4: Fix bug #64729: compilation failure on x32 Conflicts: Zend/zend_alloc.c | |||||
| * | Fix bug #64729: compilation failure on x32 | Gustavo Lopes | 2013-04-30 | 1 | -4/+14 | |
| | | ||||||
| * | Happy New Year | Xinchen Hui | 2013-01-01 | 1 | -1/+1 | |
| | | ||||||
* | | Low-level ARM optimizations | Ard Biesheuvel | 2013-02-11 | 1 | -0/+45 | |
| | | | | | | | | | | - added ARM versions of zend_mm_high_bit()/zend_mm_low_bit() - improved safe_address() | |||||
* | | Improve x86 inline assembler | Ard Biesheuvel | 2013-02-11 | 1 | -5/+5 | |
| | | | | | | | | | | | | | | - added cc annotation to inline asm that clobbers the condition flags - remove hardcoded constants (IS_LONG,IS_DOUBLE) - remove hardcoded offsets (zval->value, zval->type) | |||||
* | | Happy New Year | Xinchen Hui | 2013-01-01 | 1 | -1/+1 | |
| | | ||||||
* | | Merge branch 'PHP-5.4' | Stanislav Malyshev | 2012-09-17 | 1 | -3/+3 | |
|\ \ | |/ | | | | | | | | | | | * PHP-5.4: update news Support building PHP with the native client toolchain. test should fail | |||||
| * | Support building PHP with the native client toolchain. | Stuart Langley | 2012-09-17 | 1 | -3/+3 | |
| | | | | | | | | | | | | | | | | | | | | The native client compiler defines the C macro __x86_64__, but the size of an unsigned int is only 4 bytes. This causes the compile to fail because the inline asm code compiled is for an 8 byte unsigned int. Add the correct defines to fix this. As detailed in: http://code.google.com/p/nativeclient/issues/detail?id=2255 | |||||
| * | - Year++ | Felipe Pena | 2012-01-01 | 1 | -1/+1 | |
| | | ||||||
| * | Fixed ZE specific compile warnings (Bug #55629) | Dmitry Stogov | 2011-09-13 | 1 | -4/+4 | |
| | | ||||||
| * | Fixed bug #55509 (segfault on x86_64 using more than 2G memory). (Laruence) | Dmitry Stogov | 2011-09-13 | 1 | -1/+1 | |
| | | ||||||
| * | Zend Signal Handling | Ilia Alshanetsky | 2011-06-22 | 1 | -14/+50 | |
| | | ||||||
| * | MFH: Fixed bug #54957 (Memleak in alloc_globals_ctor) | Dmitry Stogov | 2011-05-31 | 1 | -0/+3 | |
| | |