archive/php-src
1
0
Fork 0
mirror of https://github.com/php/php-src.git synced 2025-08-20 09:24:05 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions 2
59918 commits 526 branches 1434 tags 881 MiB
Commit graph

82 commits

Author SHA1 Message Date
Andrey Hristov
9c0439e4bc Fix for Bug #52613 crash in mysqlnd 2010-08-17 18:08:25 +00:00
Andrey Hristov
c803b0024a Make this function static, as it is used only in this file. 
Add additional comment about how it works.
 2010-08-12 14:17:31 +00:00
Andrey Hristov
c974313aed Switch from using PHP_MAJOR_VERSION to separate define for 
unicode. Unicode is no more, but these are "bookmarks" where
to change mysqlnd, if Unicode becomes trendy again.
 2010-08-12 12:02:02 +00:00
Andrey Hristov
d27983e781 ws + cs 2010-08-12 11:38:08 +00:00
Andrey Hristov
da42d146ab Fix Request #52302 mysqli_fetch_all does not work with MYSQLI_USE_RESULT 2010-08-09 17:29:30 +00:00
Andrey Hristov
12f8599b86 Give an error when USE_RESULT is used with fetch_all 2010-07-13 13:02:00 +00:00
Andrey Hristov
baf72b7336 Fix sprintf modifiers all around the place. In most cases 
%u should be used. Where size_t is used then MYSQLND_SZ_T_SPEC should be
the modifier
 2010-06-21 15:32:26 +00:00
Andrey Hristov
c7f10efabc Fix for bug #52115 mysqli_result::fetch_all returns null, not an empty array 2010-06-18 17:25:41 +00:00
Andrey Hristov
d02e785d68 Handle problems in the row_decoder, return type used to be void, 
now it can return on problems.
 2010-05-31 17:57:03 +00:00
Andrey Hristov
ae9ac28ec2 Fix possible crashes, in case of OOM, due to half-baken 
objects.
 2010-05-27 12:05:02 +00:00
Andrey Hristov
433cb868d6 Handle the situation when MYSQLND_PROTOCOL's methods return NULL. 
mysqlnd should not crash but gracefully return with an error.
 2010-05-25 23:18:13 +00:00
Andrey Hristov
954e7db644 Handle OOM in the code that generates PS Execute requests 
to the server. Fixes crashes in case of OOM.
 2010-05-25 19:19:29 +00:00
Andrey Hristov
a25482105e Check set->data for validity before using it. 
Prevents crashes after OOM.
 2010-05-21 13:37:18 +00:00
Andrey Hristov
8a093b3811 Allow overloading of mysqlnd_result_meta_init(). 
Add free space for future hooks.
 2010-05-14 15:11:56 +00:00
Andrey Hristov
8691786ebd - Give possibility to overload mysqlnd_result_init(). 
- Always use conn->m->stmt_init instead of mysqlnd_stmt_init()
 2010-05-14 13:18:39 +00:00
Andrey Hristov
07619b7190 - initialize some variables 
- initialize the methods in a shorter way, use the compiler.
 = on structures copies them. Allows also exportability of
 the methods.
 2010-05-13 15:29:08 +00:00
Andrey Hristov
d53337bbe7 OOM stability fixes 2010-05-04 14:02:42 +00:00
Andrey Hristov
5ca71c00f5 Handle OOM in mysqlnd_res::store_result_fetch_data 2010-05-03 19:32:07 +00:00
Andrey Hristov
9c276eee6c Handle OOM in mysqlnd_fetch_row_buffered_c() 2010-05-03 19:09:45 +00:00
Andrey Hristov
2b00096004 Handle OOM in mysqlnd_fetch_row_unbuffered_c() 2010-05-03 19:02:48 +00:00
Andrey Hristov
2b18a3ff93 Handle OOM stemming from mysqlnd_result_init in the same 
function and up the stack.
 2010-05-03 18:50:47 +00:00
Andrey Hristov
7de2607c12 Handle gracefully OOM in mysqlnd_mempool_create and up the stack 2010-05-03 17:07:18 +00:00
Andrey Hristov
ec0fe858c6 Remove unused parameter 2010-05-03 16:09:05 +00:00
Andrey Hristov
1ed3da294b Remove dead code and add possibility to fine tune the size of 
a mempool (used for result sets).
 2010-04-22 15:07:40 +00:00
Felipe Pena
6d951e2ded - Fix copyrights 2010-04-20 13:50:34 +00:00
Andrey Hristov
f40779091c Fix some int/uint comparison warnings 2010-04-15 15:28:00 +00:00
Andrey Hristov
6f0adc06fa more WS 2010-04-15 15:16:29 +00:00
Andrey Hristov
a9880e78c6 CS + WS changes 2010-04-15 12:55:04 +00:00
Andrey Hristov
7868bca437 Fix tests that fail when the MySQL's socket file is not /tmp/mysql.sock 
as it is in when compiled from source and the default for mysqlnd.
SuSE for example uses /var/run/mysql/mysql.sock . Also, sql.safe_mode
(ext/mysql and ingres) needs the socket.
Fix possible crashes in mysqlnd. When packets are shorter, functions should
return error.
 2010-03-26 16:58:25 +00:00
Andrey Hristov
a8460553be decouple the methods in MYSQLND_STMT from the data, 
needed to move to a new structure MYSQLND_STMT. Makes
the code cleaner and less error-prone.
Also fix PDO/MySQL which directly touch mysqlnd internals
instead of using API calls.
 2010-03-16 12:36:57 +00:00
Andrey Hristov
564a1d07f8 allow persistency of PS 2010-03-12 13:03:46 +00:00
Andrey Hristov
055f558554 Pass tsrmls to all functions, which might need it in the future. 
We are anyway breaking the internal ABI in 5.3.2 so this won't hurt
and make us prepared for the future.
 2010-02-03 17:31:29 +00:00
Andrey Hristov
204f52f7e0 export more functionality for phpize builds 2010-01-13 18:00:41 +00:00
Andrey Hristov
a449a371a8 cosmetics. typedef-ed types should be in capital case 2010-01-11 20:10:10 +00:00
Andrey Hristov
a98a67940a improve the stats subsystem. Reduce the number of macro definitions 
in half by smartly introducing 2 new macros. Make MYSQLND::stats a pointer
from being aggregated and add triggers.
 2010-01-08 17:35:20 +00:00
Andrey Hristov
81b0630f5a Change of way the packet objects are created/initialised 
If the protocol gets changed, ever, we can decide at runtime
easily which protocol to use by instantiating the right protocol
object. But this is restricted to the structure of the packets, not
the flow.
 2010-01-07 19:44:59 +00:00
Andrey Hristov
311684e1e4 Add plugin storage for MYSQLND_STMT too 2009-12-10 11:18:05 +00:00
Andrey Hristov
3a7f8ea575 Remove stub code which never actually was used for anything. 
Things like this can be built on top of the core.
 2009-12-09 20:38:11 +00:00
Andrey Hristov
fd44827701 Remove the zval caching from mysqlnd. It was disabled versions ago 
due to problems on windows, which were not debugged. Better have 
code that is disabled not in the core.
 2009-12-09 20:21:05 +00:00
Andrey Hristov
157f1bddf1 Remove disabled code - threaded fetching. This can be implemented 
on a upper level and by offloading it we reduce the complexity of
the core.
 2009-12-09 16:57:03 +00:00
Pierre Joye
80e5ef6fa5 - void is NaN, you can't do math ops on void. Declarations go first. Fix TS build 2009-11-30 10:56:01 +00:00
Andrey Hristov
7674c942c7 Compressed protocol support + extensibility for mysqlnd 2009-11-20 08:12:14 +00:00
Ulf Wendel
bd64a66e1a Fix: log slow queries in stats regardless of mysqlnd.log_mask setting. 2009-10-23 12:44:05 +00:00
Andrey Hristov
bc7bbaa65a Use mnd_ wrappers wherever possible. Needed for 
statistics.
 2009-09-30 23:13:43 +00:00
Andrey Hristov
4af8d99eb6 Fix trace output. DBG_RETURN should be used in traced 
functions.
 2009-09-30 22:26:19 +00:00
Ulf Wendel
d79cba2afd Fixing a leak in mysqlnd when passing invalid fetch modes to mysqlnd. 2009-09-30 14:39:33 +00:00
Andrey Hristov
e8412fb61c Fix for bug#48745 
mysqlnd: mysql_num_fields returns wrong column count for mysql_list_fields
 2009-08-28 09:30:16 +00:00
David Soria Parra
565e6bdbdf Fix bug #48644 mysqlnd does not compile with '--enable-mysqlnd-threading' 2009-06-23 14:14:32 +00:00
Andrey Hristov
7289831fe0 Memory usage optimisation. mysqlnd is not libmysql. mysqlnd does use the 
Zend allocator, which means that is easier to hit memory_limit if you
have big stored (buffered) result sets. Before with libmysql you won't
hit memory_limit because libmysql uses libc's allocator and nothing is
checked. Now, with mysqlnd the situation is stricter and it is easier to
hit memory_limit. We try to optimize for big result sets. If a result set
is larger than 10 rows we will start freeing some data to keep memory usage
after 10 rows constant. This will help in the cases where a buffered result
set is scrolled forward only and just only once, or mysqlnd will need to
decode data from the network buffers again - yes, it is a trade-off between
CPU time and memory size. The best for big result sets is of course using
unbuffered queries - for comparison : 3 Million rows with buffered take
at least 180MB, with buffered you will stay at 3MB, and unbuffered will be
just 7-8% slower.
 2009-06-16 13:06:45 +00:00
Andrey Hristov
5d3ea92ee9 Hardwire function call instead of using callbacks. We don't actually need 
callbacks, it was done for making 2 functions static, not to pollute the
global functions space but that had its price of 8 bytes overheat per
allocation, which is just too much. Also making the app member 32b instead
of 64b, which should save additional 4 byte, to the total of 12 byte per
allocation of a row buffer.
 2009-06-16 09:15:09 +00:00