The fix was correct but break some code (at least in Horde)
This is a temporary workaround to fix regressioni in 5.4, 5.5 and 5.6
This make php_date_initialize more consistent
- on success return 1 + time initiliazed
- on failure return 0 + time = zero
which is check by DATE_CHECK_INITIALIZED by later method call
Will restore consistency with other date classes in master.
XML doesn't allow most of the characters < 0x20 and binary output
breaks XML parsers. Such characters are not allowed as entities
either, so the generated entities are escaped in order to be printed
as text.
The version test that we had in run-tests.php assumed that the major and minor
version numbers were always single digits. This removes that assumption and
also uses version_compare() for the comparison instead of naively converting it
to an integer.
The actual issue lays in the unserializer code which doesn't honor
the unserialize callback. By contrast, the serialize callback is
respected. This leads to the situation that even if a class has
disabled the serialization explicitly, user could still construct
a vulnerable string which would result bad things when trying
to unserialize.
This conserns also the classes implementing Serializable as well
as some core classes disabling serialize/unserialize callbacks
explicitly (PDO, SimpleXML, SplFileInfo and co). As of now, the
flow is first to call the unserialize callback (if available),
then call __wakeup. If the unserialize callback returns with no
success, no object is instantiated. This makes the scheme used
by internal classes effective, to disable unserialize just assign
zend_class_unserialize_deny as callback.
curl_exec returns an empty string when data is received from a domain that returns zero content. curl_multi_getcontent
returned null. Now it returns an empty string fixing the incosistency
On little endian:
map->p == php_magic_database
map->magic[i] = pointer into the map
map->p == NULL
map->magic[i] = pointer to allocated memory
On big endian (ppc64, s390x, ...):
map->p != php_magic_database and map->p != NULL
map->magic[i] = pointer into a copy of the map
Trying to efree pointer in the later cause memory corruption
Thanks to dkatulek / Red Hat for the report.
Switch to SHA1, which match internal openssl hardcoded algo.
In most case, won't even be noticed
- priority on user input (default_md)
- fallback on system config
- fallback on this default value
Recent system reject MD5 digest, noticed in bug36732.phpt failure.
While SHA1 is better than MD5, SHA256 is recommenced,
and defined as default algo in provided configuration on
recent system (Fedora 21, RHEL-7, ...). But the idea is to
keep in sync with openssl internal value for PHP internal value.
# By Pierre Joye
# Via Pierre Joye
* 'PHP-5.4' of https://git.php.net/repository/php-src:
fix#66872, invalid argument crashes gmp_testbit
fix#66872, invalid argument crashes gmp_testbit
add vc12 (2013)
# By Dmitry Panin (1) and Remi Collet (1)
# Via Nikita Popov
* 'PHP-5.4' of git.php.net:php-src:
Fix HTML entity table generation
Set default Digest Message to use SHA1 instead of MD5 in openssl tests as MD5 signature are now rejected by newer openssl Version.
# By Anatol Belski (22) and others
# Via Derick Rethans (5) and others
* 'PHP-5.4' of git.php.net:php-src: (176 commits)
NEWS
Fixed Bug #66820 out-of-bounds memory access in fileinfo
Improves fix for memory leak, keep in sync with upstream.
Updated news for #60602
proc_open(): separate environment values that aren't strings
Updated NEWS for #66535
Fixed test case title
[bug 66535] X-PHP-Originating-Script adds newline if no custom headers are given
man page: long option name is --strip, not --stripped
--global have be removed in 5.2
NEWS
test for bug #66762
Fixed Bug #66762 Segfault in mysqli_stmt::bind_result() when link closed
fix tests broken by 633f898f15
add news entry
add clear_env option to FPM config
Reduce test noise on cross Oracle client <-> server version tests. This fix is already in PHP 5.6+
Reduce test noise in cross Oracle client <-> server version testing. This change is already in PHP 5.6+
fixed macro
Make sure value is initialized
...
A newline is added to the mail headers when mail.add_x_header is used and no other headers are passed to mail().
The scenario in which custom headers are used was already fixed in #48620, back in 2009.
Each new mysqli_stmt now increase the refcount of the link object.
So the link is really destroy after all statements.
Only implemented with libmysqlclient, as mysqlnd already implement
this internally.
So, libmysqlclient and mysqlnd have the same behavior.
This makes it possible to leave the envoronment as is on startup and
pass all the variables to the workers.
The default value of clear_env is "yes", preserving previous behaviour.
Patch by Paul Annesley.
* PHP-5.3:
fix the primary qa email address and update the list of available people being able to add moderators to our lists
Revert "Fix #62479: Some chars not parsed in passwords"
Fix#62479: Some chars not parsed in passwords
There are situations where mysqlnd dupliates zvals while freeing result
sets. If the memory_limit is reached during this operation the engine
will bailout. This patch makes sure that a later attempt (during
RSHIUTDOWN) won't cause a double free, instead we rely on the engine to
free emalloc()ed memory after bailout.
If the previous value of session.name was NULL then any call to
session_name($string) would result in a segmentation fault.
This changes the behaviour to set the value of session.name to
"PHPSESSID" if a blank value is given in php.ini or via -d on the
command line. There is already protection against setting it to NULL via
session_name() or ini_set().
std_compare_objects immidiately returned 0 if the property tables
of both objects contain NULL at some index. Thus it would report
objects as equal even though properties following after that
differ.
This piece of code seems to retain compatibility with some windows
versions before 2000. Today it's not relevant anymore and only
produces a warning C4273 because of the conflicting prototype. In
VC9 the declaration is present in winbase.h:4142, but available
is it already in winxp. Here's the doc link
http://msdn.microsoft.com/en-us/library/windows/desktop/ms680345(v=vs.85).aspx
That way extensions etc. don't have to maintain their own lists anymore.
Also as phpdbg is included and needs such a map, it'd be counterproductive to need to change two things on every new opcode.
This fixes GD compilation against libfreetype 2.5.1 and later after they made
the rather interesting decision to change their include directory layout in a
point release.
The original suggestion in the bug was to use pkg-config, but my inclination is
to use freetype-config instead: we should be able to get the same configuration
information without actually needing pkg-config installed, since pkg-config is
by no means guaranteed to exist on many Unices and distros, whereas
freetype-config should always be present if a libfreetype build environment is
installed. Let's try it out and see what happens.
Fixes bug #64405 (Use freetype-config for determining freetype2 dir(s)).
The 'result' argument of fast_add_function() may alias with either
of its operands (or both). Take care not to write to 'result' before
reading op1 and op2.
Functions registered using zend_register_functions instead of zend_module_entry.functions are not seen on reflection.
Ex: additional_functions from api_module_entry.
Ex: in CLI, dl, cli_set_process_title and cli_get_process_title
Note:
- also affects functions overrided in extension
(should be be reported in extension, where overrided, not in original extension)
- also allow extension to call zend_register_functions for various list
(instead of having a single bug list)
if a mode like "rn" was passed to fopen(), then
php_stream_parse_fopen_modes() would assign O_WRONLY to
flags, because O_NONBLOCK tainted flags for the r/w/+ check
These were returned to the general allocation pool by RFC 3330, and hence
shouldn't cause an IP address validation failure due to being reserved. At
least 128.0.0.0/16 is in use on the public Internet today.
Fixes bug #66229 (128.0.0.0/16 isn't reserved any longer).
Passing DOMDocumentFragment to DOMDocument::saveHTML()
produces invalid markup, because a DocumentFragment is just a container
for child nodes and not a real node itself.
This fixes the issue in
https://wiki.php.net/rfc/datetime_and_daylight_saving_time#forward_transitions
There is a period during transition to DST where a time (such as 02:30) does
not exist. PHP already calculated the correct timestamp for this, but failed to
"rounded forward" to the existing correct hour value.
previous codes: "#define php_mblen(ptr, len) ((ptr) == NULL ? mbsinit(&BG(mblen_state)):
(int)mbrlen(ptr, len, &BG(mblen_state)))#
it use mbsinit there, seems try to initialize the mblen_state, but:
"This function does not change the state identified by ps. Typical ways
to make the state pointed by ps an initial state are:
memset (ps,0,sizeof(*ps)); // ps points to zero-valued object
"
http://www.cplusplus.com/reference/cwchar/mbsinit/?kw=mbsinit
* 'PHP-5.4' of git.php.net:php-src: (65 commits)
Add a couple more test cases to parse_url() tests
fix missing change from 'tcp_socket' to the more common 'server'
fix many parallel test issues
Cleanup temp test file
Fixed Bug #66034 (Segmentation Fault when constructor of PDO statement throws an exception)
Typo fix: umknown -> unknown
Fix bug #66008
5.4.23-dev
Update NEWS
Fixed Bug 64760 var_export() does not use full precision for floating-point numbers
add bundled libzip LICENSE, as required by BSD License terms
- Updated to version 2013.8 (2013h)
remove "PHP 6" staff
Fixed bug #65950 Field name truncation if the field name is bigger than 32 characters
- Updated to version 2013.7 (2013g)
Fix Coverity issue reporting wrong sizeof()
exif NEWS
add tests for bug #62523
Merged PR #293 (Exif crash on unknown encoding was fixed) By: Draal Conflicts: configure.in main/php_version.h
Just SKIP that test on travis
...
While running these on HHVM I've run into a lot of parallelism issues.
I'm backporting all the fixes I had to do in
https://github.com/facebook/hiphop-php/blob/master/hphp/tools/import_zend_test.py#L650
to php core.
Most of these changes were just filenames that were shared between
tests, but I did more surgery on the fixed ports. I can apreciate port
31337 as much as the next nerd, but random ports are better for tests.
Changed json_encode() so that when the JSON_PRETTY_PRINT option is specified,
the pair of linefeeds immediately after an opening bracket and before the
corresponding closing bracket is omitted when the array or object contains no
elements or accessible properties (and hence would have a blank line between
the brackets).
* 'PHP-5.4' of git.php.net:php-src:
- Updated to version 2013.8 (2013h)
Fixed bug #65950 Field name truncation if the field name is bigger than 32 characters
- Updated to version 2013.7 (2013g)
* 'PHP-5.4' of git.php.net:php-src: (101 commits)
exif NEWS
add tests for bug #62523
Merged PR #293 (Exif crash on unknown encoding was fixed) By: Draal Conflicts: configure.in main/php_version.h
Just SKIP that test on travis
fix memory leak on error (from Coverity scan)
Fix coverity issue with -1 returned by findOffset not being handled by getPreferredTag
5.4.21 release date
When src->src is null this doesn't get initialized but it is still used, so the passed in *ze will point to unitialized memory. Hopefully src->src is never null, but just in case this initialization doesn't hurt.
Fix typo
Clean up this weird safe_emalloc() call
Minor Coverity tweaks
- Moved NULL check before dereferencing
- Fixed possible NULL ptr dereference
- Fixed possible uninitialized scalar variable usage (spotted by Coverity)
Remove senseless check here
- Fix extern declaration according to definition
- Fix possible memory leak
- Moved allocation to if block to make Coverity happy
- Fixed possible memory leak
Fix unitialized opened_path here - found by Coverity
...
but it is still used, so the passed in *ze will point
to unitialized memory. Hopefully src->src is never
null, but just in case this initialization doesn't hurt.
In context of static accesses like classname::$this, the string
"$this" should not be handled like a $this variable, but as an
identifier for a static variable.
# By Daniel Lowrey (1) and Michael Wallner (1)
# Via Michael Wallner
* 'PHP-5.4' of https://git.php.net/repository/php-src:
C89 compatibility
Fixed segfault when built with OpenSSL >= 1.0.1
* 'PHP-5.4' of https://git.php.net/push/php-src:
- Fixed bug #65818 (Segfault with built-in webserver and chunked transfer encoding)
- BFN
Fix bug #65667: ftp_nb_continue produces segfault
fix bug #64146 (serialize incorrectly saving objects when they are cloned)
such a weird hack probably helps in finding regressions in the future
Fixed bug #64230 (XMLReader does not suppress errors)
typo: really fix bug #51936 Crash with clone xmlreader
fix bug #59613 (Crash with clone XMLReader)
fix bug #65808 the socket_connect() won't work with IPv6 address
5.4.22-dev now
fix bug #62396 'make test' crashes starting with 5.3.14 (missing gzencode())
Fixed bug #61548
fix test
fix memleak on resetting rebind_proc
Fix bug #65322: compile time errors won't trigger auto loading
5.4.20 release date
Add information about which INI file is which inside respective files
- Updated to version 2013.6 (2013f)
* 'PHP-5.4' of https://git.php.net/repository/php-src:
Fix bug #65667: ftp_nb_continue produces segfault
fix bug #64146 (serialize incorrectly saving objects when they are cloned)
such a weird hack probably helps in finding regressions in the future
the idea behind ftp_nb_get is for it to be followed by multiple calls
to ftp_nb_continue in order to download a file piece-by-piece.
As such, it's unwise to close the stream used to write the downloaded
data to when the file hasn't been completely downloaded within the first
call to ftp_nb_get.
This regression was added in a93a462dce
and this patch restores the behavior that was seen pre-patch.
# By Michael Wallner (2) and others
# Via Michael Wallner
* 'PHP-5.4' of https://git.php.net/repository/php-src:
fix test
fix memleak on resetting rebind_proc
Fix bug #65322: compile time errors won't trigger auto loading
5.4.20 release date
Also fixes duplicate bugs #54054 and #42098.
Furthermore this fixes incorrect error messages thrown from code
running inside an error handler when a compilation is in progress.
The error file and line are now correctly associated with the
file/line of the executor, rather than the compiler.
# By Derick Rethans (1) and others
# Via Lior Kaplan (1) and Stanley Sufficool (1)
* 'PHP-5.4' of https://git.php.net/repository/php-src:
Add information about which INI file is which inside respective files
- Updated to version 2013.6 (2013f)
FIX BUG #48539 - Disable TEXTLIMIT for FreeTDS driver
* 'PHP-5.4' of https://git.php.net/push/php-src: (140 commits)
Copy dba_*() keys before converting to string.
Fix the broken sh syntax in ext/imap/config.m4.
Revert "EmptyIterator now implements Countable; fixes bug 60577"
RFC 6598 reserved ip range starts at 100.64.0.0
fix a very rare case of use of uninitialized value combined with a memleak
NEWS for added reserved ip addresses according to RFC 6598
Add RFC 6598 IPs to reserved addresses
NEWS for #60577
NEWS for bug #64441
Fix bug #64441 (FILTER_VALIDATE_URL rejects fully qualified domain names)
EmptyIterator now implements Countable; fixes bug 60577
News for bugfix #64157
Bug 64157 Changed error message to make sense
Tinker with the wording of the short_open_tag description.
Handle CLI server request headers case insensitively.
5.4.21 now
Typo....
Add a XFAIL test for #64896
Fixed Bug #65564 stack-buffer-overflow in DateTimeZone stuff caught by AddressSanitizer
Fixed bug #60598 (cli/apache sapi segfault on objects manipulation)
...
A nice Sunday afternoon project for somebody would be to refactor the dba
functions to use zend_parse_parameters() reliably and try to untangle some of
the macros in dba.c. Sadly, it is not a nice Sunday afternoon here.
Fixes bug #65708 (dba functions cast $key param to string in-place, bypassing
copy on write).
# By David Soria Parra (6) and others
# Via David Soria Parra
* 'PHP-5.4' of https://git.php.net/repository/php-src:
Revert "EmptyIterator now implements Countable; fixes bug 60577"
RFC 6598 reserved ip range starts at 100.64.0.0
fix a very rare case of use of uninitialized value combined with a memleak
NEWS for added reserved ip addresses according to RFC 6598
Add RFC 6598 IPs to reserved addresses
NEWS for #60577
NEWS for bug #64441
Fix bug #64441 (FILTER_VALIDATE_URL rejects fully qualified domain names)
EmptyIterator now implements Countable; fixes bug 60577
News for bugfix #64157
Bug 64157 Changed error message to make sense
# By Adam Harvey
# Via Adam Harvey
* 'PHP-5.4' of https://git.php.net/repository/php-src:
Tinker with the wording of the short_open_tag description.
Handle CLI server request headers case insensitively.
* 'PHP-5.4' of git.php.net:php-src:
bump version
Update NEWS
fix bug #65481 (shutdown segfault due to serialize)
Track created curl_slist structs by option so they can be updated in situ.
Fixed bug #64503 (Compilation fails with error: conflicting types for 'zendparse').
added new glob() test
fix using wrong buffer pointer
Fix bug #65470 Segmentation fault in zend_error() with --enable-dtrace
Fix for php bug #64802 includes test case
new for fix#65225Fixed#65225: PHP_BINARY incorrectly set
Use pkg-config to detect iodbc
Add -P option to use the current binary
Create test to the extension xmlrpc
Fixbug: phpize --clean will delete include/*.h
Loaded 100 of 1239 files, more files were not shown because too many files have changed in this diff.
Show more
Reference in new issue
Block a user
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.