]>
granicus.if.org Git - php/log
Joe Watkins [Tue, 11 Jun 2019 08:45:41 +0000 (10:45 +0200)]
Merge branch 'PHP-7.4' of git://github.com/php/php-src into PHP-7.4
Joe Watkins [Tue, 11 Jun 2019 08:42:25 +0000 (10:42 +0200)]
checkout depth of one doesn't always work with batched builds, it seems
Nikita Popov [Tue, 14 May 2019 12:27:10 +0000 (14:27 +0200)]
Add get_mangled_object_vars() function
Joe Watkins [Tue, 11 Jun 2019 08:23:59 +0000 (10:23 +0200)]
Merge branch 'i386-azure' into PHP-7.4
* i386-azure:
i386 build
Joe Watkins [Sun, 9 Jun 2019 13:31:19 +0000 (15:31 +0200)]
i386 build
Tyson Andre [Sun, 9 Jun 2019 20:20:20 +0000 (16:20 -0400)]
Fix typo in opcache function flag info (mage->image)
ext/gd/gd.c seems to return null for wrong parameter count, false for
error, and long otherwise. See https://php.net/imagecolorallocate
In php 8.0, the MAY_BE_NULL can probably be removed.
Tyson Andre [Sun, 9 Jun 2019 21:58:50 +0000 (17:58 -0400)]
Clean up opcache func flags for set_socket_blocking
It was removed in php 7.0. See https://php.net/set_socket_blocking
Tyson Andre [Sun, 9 Jun 2019 19:13:44 +0000 (15:13 -0400)]
Fix a typo in opcache func info for "get_parrent_class"
This should only have one r.
http://php.net/get_parent_class mentions that it returns string|false
Also, I'm not 100% sure if this is right - should this include
MAY_BE_NULL to handle the case of too many parameters?
(then again, get_called_class() also returns null in the global scope)
```
php > var_export(get_parent_class(2,3));
Warning: get_parent_class() expects at most 1 parameter, 2 given in php
shell code on line 1
NULL
```
Hugh McMaster [Fri, 7 Jun 2019 12:29:55 +0000 (22:29 +1000)]
ext/pgsql: Use PKG_CHECK_MODULES to detect the pq library
Christoph M. Becker [Tue, 11 Jun 2019 06:44:54 +0000 (08:44 +0200)]
Merge branch 'PHP-7.3' into PHP-7.4
* PHP-7.3:
Fix #77937: preg_match failed
Christoph M. Becker [Tue, 11 Jun 2019 06:43:20 +0000 (08:43 +0200)]
Merge branch 'PHP-7.2' into PHP-7.3
* PHP-7.2:
Fix #77937: preg_match failed
Christoph M. Becker [Tue, 30 Apr 2019 14:10:04 +0000 (16:10 +0200)]
Fix #77937: preg_match failed
On some recent Windows systems, ext\pcre\tests\locales.phpt fails,
because 'pt_PT' is accepted by `setlocale()`, but not properly
supported by the ctype functions, which are used internally by PCRE2 to
build the localized character tables.
Since there appears to be no way to properly check whether a given
locale is fully supported, but we want to minimize BC impact, we filter
out typical Unix locale names, except for a few cases which have
already been properly supported on Windows. This way code like
setlocale(LC_ALL, 'de_DE.UTF-8', 'de_DE', 'German_Germany.1252');
should work like on older Windows systems.
It should be noted that the locale names causing trouble are not (yet)
documented as valid names anyway, see
<https://docs.microsoft.com/en-us/cpp/c-runtime-library/locale-names-languages-and-country-region-strings?view=vs-2019>.
Dmitry Stogov [Mon, 10 Jun 2019 13:43:20 +0000 (16:43 +0300)]
micro-optimization
Joe Watkins [Mon, 10 Jun 2019 12:23:37 +0000 (14:23 +0200)]
watchpoints have flaws in implementation
Joe Watkins [Mon, 10 Jun 2019 12:19:39 +0000 (14:19 +0200)]
BIND_LEXICAL has a meaningless line number
Dmitry Stogov [Mon, 10 Jun 2019 09:53:04 +0000 (12:53 +0300)]
Merge branch 'PHP-7.3' into PHP-7.4
* PHP-7.3:
Fixed possible misalignment in 32-bit build.
Dmitry Stogov [Mon, 10 Jun 2019 09:52:50 +0000 (12:52 +0300)]
Merge branch 'PHP-7.2' into PHP-7.3
* PHP-7.2:
Fixed possible misalignment in 32-bit build.
Dmitry Stogov [Mon, 10 Jun 2019 09:51:59 +0000 (12:51 +0300)]
Fixed possible misalignment in 32-bit build.
Joe Watkins [Mon, 10 Jun 2019 08:35:31 +0000 (10:35 +0200)]
fix flaky ipv4 loopback test
Peter Kokot [Sun, 9 Jun 2019 20:55:21 +0000 (22:55 +0200)]
Merge branch 'PHP-7.3' into PHP-7.4
* PHP-7.3:
Extend wildcard files section in EditorConfig [ci skip]
Peter Kokot [Sun, 9 Jun 2019 20:54:59 +0000 (22:54 +0200)]
Merge branch 'PHP-7.2' into PHP-7.3
* PHP-7.2:
Extend wildcard files section in EditorConfig [ci skip]
Peter Kokot [Sun, 9 Jun 2019 20:53:58 +0000 (22:53 +0200)]
Extend wildcard files section in EditorConfig [ci skip]
Changes:
- Trim trailing whitespace for all files except patches. There isn't
really any practical reason to not trim the trailing whitespace in all
other files. Binary files or phpt files that include trailing
whitespace as part of the test should be manually set in editors/IDEs
or by disabling the editorconfig for particular editing.
- Add *.ac, *.d, *.l, *.skl, *.re, *.wsdl, *.dtd, *.html, *.rng, *.xml,
*.xsd, *.xsl, buildconf, and Makefile* files settings.
Closes #4156
Peter Kokot [Sat, 11 May 2019 02:43:40 +0000 (04:43 +0200)]
Simplify LiteSpeed SAPI module version
This patch syncs versioning of the LiteSpeed SAPI module by making it in
sync with the rest of the bundled PHP extensions and SAPIs.
Closes #4145
Peter Kokot [Sun, 19 May 2019 15:18:29 +0000 (17:18 +0200)]
Use PHP_VERSION for zend modules
Closes #4147
Jan Altensen [Thu, 6 Jun 2019 15:37:04 +0000 (17:37 +0200)]
azure macos
Change-Id: I3465dcc865c254cea02feb619f6706fa0d164757
Joe Watkins [Sun, 9 Jun 2019 11:39:35 +0000 (13:39 +0200)]
azure image update [not skipping on purpose to test new configuration]
Joe Watkins [Sat, 8 Jun 2019 19:47:07 +0000 (21:47 +0200)]
Fix #78132 wrong line number on ZEND_BIND_STATIC for lexical vars
Dmitry Stogov [Fri, 7 Jun 2019 09:39:24 +0000 (12:39 +0300)]
Merge branch 'PHP-7.3' into PHP-7.4
* PHP-7.3:
Restored NEWS entry
Dmitry Stogov [Fri, 7 Jun 2019 09:38:37 +0000 (12:38 +0300)]
Merge branch 'PHP-7.2' into PHP-7.3
* PHP-7.2:
Restored NEWS entry
Dmitry Stogov [Fri, 7 Jun 2019 09:37:49 +0000 (12:37 +0300)]
Restored NEWS entry
Nikita Popov [Fri, 7 Jun 2019 08:57:51 +0000 (10:57 +0200)]
Merge branch 'PHP-7.3' into PHP-7.4
Nikita Popov [Fri, 7 Jun 2019 08:57:43 +0000 (10:57 +0200)]
Merge branch 'PHP-7.2' into PHP-7.3
Nikita Popov [Fri, 7 Jun 2019 08:26:37 +0000 (10:26 +0200)]
Enable STRICT_TRANS_TABLES in new test
The part testing error cases relies on this.
Dmitry Stogov [Fri, 7 Jun 2019 08:50:19 +0000 (11:50 +0300)]
Merge branch 'PHP-7.3' into PHP-7.4
* PHP-7.3:
Fixed bug #77135 (Extract with EXTR_SKIP should skip $this)
Dmitry Stogov [Fri, 7 Jun 2019 08:48:57 +0000 (11:48 +0300)]
Merge branch 'PHP-7.2' into PHP-7.3
* PHP-7.2:
Fixed bug #77135 (Extract with EXTR_SKIP should skip $this)
Dmitry Stogov [Fri, 7 Jun 2019 08:36:39 +0000 (11:36 +0300)]
Fixed bug #77135 (Extract with EXTR_SKIP should skip $this)
Nikita Popov [Fri, 7 Jun 2019 07:53:35 +0000 (09:53 +0200)]
Use batch mode on Azure
Nikita Popov [Fri, 7 Jun 2019 07:49:09 +0000 (09:49 +0200)]
Merge branch 'PHP-7.3' into PHP-7.4
Nikita Popov [Fri, 7 Jun 2019 07:49:01 +0000 (09:49 +0200)]
Merge branch 'PHP-7.2' into PHP-7.3
Cameron Porter [Wed, 6 Mar 2019 06:33:40 +0000 (00:33 -0600)]
Fixed bug #38546
Properly support binding boolean parameters with emulated prepared
statements disabled. Also add the necessary mysqlnd support for
MYSQL_TYPE_TINY.
Hugh McMaster [Thu, 6 Jun 2019 12:22:00 +0000 (22:22 +1000)]
ext/xsl: Use PKG_CHECK_MODULES to detect the EXSLT library
Hugh McMaster [Thu, 6 Jun 2019 11:19:19 +0000 (21:19 +1000)]
ext/xsl: Use PKG_CHECK_MODULES to detect the XSLT library
Peter Kokot [Thu, 6 Jun 2019 23:24:02 +0000 (01:24 +0200)]
Christoph M. Becker [Thu, 6 Jun 2019 14:18:25 +0000 (16:18 +0200)]
Use zero port in socket_(export|import)_stream-4-win.phpt
These regularly fail due to port clashes, though we don't seem to
use those ports in any other tests.
As we don't care about the specific port that is used here, pass
zero so that a random ephemeral port is used.
Christoph M. Becker [Thu, 6 Jun 2019 13:54:12 +0000 (15:54 +0200)]
Fix bug #78094: File Search Problem Excessive Time
Instead of checking GetBinaryType() for each file, we do a much cheaper
pre-check whether the filename extension matches .exe or .com, and call
GetBinaryType() only in this case. For BC we also report .bat and .cmd
files as executables again.
The patch has been provided by @weltling.
Christoph M. Becker [Thu, 6 Jun 2019 13:02:41 +0000 (15:02 +0200)]
Merge branch 'PHP-7.3' into PHP-7.4
* PHP-7.3:
Fix test regarding Unix Domain Sockets on Windows
Christoph M. Becker [Thu, 6 Jun 2019 13:00:35 +0000 (15:00 +0200)]
Merge branch 'PHP-7.2' into PHP-7.3
* PHP-7.2:
Fix test regarding Unix Domain Sockets on Windows
Christoph M. Becker [Thu, 6 Jun 2019 12:56:47 +0000 (14:56 +0200)]
Fix test regarding Unix Domain Sockets on Windows
Recent Windows versions actually support Unix Domain Sockets. Cf.
<https://github.com/curl/curl/pull/3939>.
Nikita Popov [Thu, 6 Jun 2019 12:56:42 +0000 (14:56 +0200)]
Try to fix Windows build
Dmitry Stogov [Thu, 6 Jun 2019 11:22:10 +0000 (14:22 +0300)]
Use ZEND_ASSUME() to perform the following check only on "slow" path.
Dmitry Stogov [Thu, 6 Jun 2019 11:09:42 +0000 (14:09 +0300)]
Support for exceptions thrown during "Array to string conversion" error processing
Dmitry Stogov [Thu, 6 Jun 2019 10:13:09 +0000 (13:13 +0300)]
Reduce over-specialization for quite seldom instructions
Nikita Popov [Thu, 6 Jun 2019 09:46:23 +0000 (11:46 +0200)]
php_zip_pcre: Match pattern before stating for directories
For two reasons: First, it's generally cheaper to match a regex than
perform a stat (especially on Windows). Second, it will not fail on
concurrent modification of a directory in parts that are not matched
by the pattern, such as the spurious failure in ext/zip/tests/bug72660.phpt.
Christoph M. Becker [Thu, 6 Jun 2019 08:01:12 +0000 (10:01 +0200)]
Merge branch 'PHP-7.3' into PHP-7.4
* PHP-7.3:
Fix #78114: segfault when calling sodium_* functions from eval
Christoph M. Becker [Thu, 6 Jun 2019 07:29:44 +0000 (09:29 +0200)]
Fix #78114: segfault when calling sodium_* functions from eval
We must not follow the NULL pointer.
Dmitry Stogov [Wed, 5 Jun 2019 22:47:22 +0000 (01:47 +0300)]
Cheaper checks for exceptions thrown from __toString()
Dmitry Stogov [Wed, 5 Jun 2019 22:40:40 +0000 (01:40 +0300)]
Fixed reference-counting
Joe Watkins [Wed, 5 Jun 2019 17:27:50 +0000 (19:27 +0200)]
junit testcase classname is used as filename on azure
Joe Watkins [Wed, 5 Jun 2019 17:24:41 +0000 (19:24 +0200)]
this test is flaky on azure, and can't see why from current output
so refactored, and restored an accidentally removed check (resource)
Dmitry Stogov [Wed, 5 Jun 2019 14:38:13 +0000 (17:38 +0300)]
Prevent generation of specialized ZEND_ASSIGN_OP_..._STATIC_PROP handlers, that call unspecialized helper, anyway.
Joe Watkins [Wed, 5 Jun 2019 14:35:55 +0000 (16:35 +0200)]
Merge branch 'PHP-7.3' into PHP-7.4
* PHP-7.3:
fix flaky posix test
Joe Watkins [Wed, 5 Jun 2019 14:35:43 +0000 (16:35 +0200)]
Merge branch 'PHP-7.2' into PHP-7.3
* PHP-7.2:
fix flaky posix test
Joe Watkins [Wed, 5 Jun 2019 14:34:50 +0000 (16:34 +0200)]
fix flaky posix test
Nikita Popov [Wed, 5 Jun 2019 13:33:52 +0000 (15:33 +0200)]
Deprecate ReflectionType::__toString()
We weren't able to do this in 7.1 because the deprecation notice
may be converted to an exception and __toString() can't throw,
which means that it ultimately become a fatal error. This issue
is resolved now, so we can mark the method as deprecated.
Nikita Popov [Tue, 26 Feb 2019 14:32:18 +0000 (15:32 +0100)]
Allow exceptions in __toString()
RFC: https://wiki.php.net/rfc/tostring_exceptions
And convert some object to string conversion related recoverable
fatal errors into Error exceptions.
Improve exception safety of internal code performing string
conversions.
Nikita Popov [Wed, 5 Jun 2019 09:55:57 +0000 (11:55 +0200)]
Use zero port in socket_(export|import)_stream-5.phpt
These regularly fail due to port clashes, though we don't seem to
use those ports in any other tests.
As we don't care about the specific port that is used here, pass
zero so that a random ephemeral port is used.
Remi Collet [Wed, 5 Jun 2019 09:48:48 +0000 (11:48 +0200)]
ignore first gets time which is usually a few ms only
in previous version, max_ms was "1" and "10" in travis/appveyor
having 10 by default make it fails on fast computer
Nikita Popov [Wed, 5 Jun 2019 09:26:20 +0000 (11:26 +0200)]
Make mysqli_connect arguments explicitly nullable
It should be possible to skip any of these (and use the ini configured
defaults) by passing null, independently of strict_types settings.
Noticed while working on GH-4227.
Remi Collet [Wed, 5 Jun 2019 08:56:05 +0000 (10:56 +0200)]
keep DIR option to --with-unixODBC for old versions without libodbc.pc
Peter Kokot [Wed, 5 Jun 2019 08:04:57 +0000 (10:04 +0200)]
Fixed bug #78102
Nikita Popov [Wed, 5 Jun 2019 07:58:59 +0000 (09:58 +0200)]
Avoid concurrent directory writes in FilesystemIterator test
The test directory is often used for temporary files, let's use
the source directory, which should be more stable. If that doesn't
work we need to create a temporary directory.
Joe Watkins [Tue, 4 Jun 2019 15:11:16 +0000 (17:11 +0200)]
add azure pipelines badge to readme [ci skip]
Nikita Popov [Tue, 4 Jun 2019 14:28:08 +0000 (16:28 +0200)]
Merge branch 'PHP-7.3' into PHP-7.4
Nikita Popov [Tue, 4 Jun 2019 14:27:45 +0000 (16:27 +0200)]
Merge branch 'PHP-7.2' into PHP-7.3
Nikita Popov [Tue, 4 Jun 2019 13:08:16 +0000 (15:08 +0200)]
Fixed bug #78050
This is a backport of
a9821255612a99f9773c3601ff1914de4e7a7e32 .
Nikita Popov [Tue, 4 Jun 2019 08:03:34 +0000 (10:03 +0200)]
Actually load opcache on Azure
Nikita Popov [Tue, 4 Jun 2019 14:14:26 +0000 (16:14 +0200)]
Merge branch 'PHP-7.3' into PHP-7.4
Nikita Popov [Tue, 4 Jun 2019 14:14:14 +0000 (16:14 +0200)]
Merge branch 'PHP-7.2' into PHP-7.3
Nikita Popov [Tue, 4 Jun 2019 13:23:10 +0000 (15:23 +0200)]
Reload tsrmls_id in release builds as well
If TSRM is shut down and started again (something that phpdbg does),
then tsrmls_id needs to be reloaded everywhere. As tsrmls_id
update is a rare operation, doing that shouldn't be a problem.
Nikita Popov [Tue, 4 Jun 2019 14:11:52 +0000 (16:11 +0200)]
Fix potential port conflicts
Hugh McMaster [Mon, 3 Jun 2019 10:22:03 +0000 (20:22 +1000)]
ext/zip: Replace $LIBZIP_LIBDIR with $LIBZIP_LIBS
Closes GH-4220.
Nikita Popov [Tue, 4 Jun 2019 13:08:16 +0000 (15:08 +0200)]
Try to fix phpdbg + opcache compatibility
Christoph M. Becker [Tue, 4 Jun 2019 12:41:25 +0000 (14:41 +0200)]
PHP 7.4 is built with Visual Studio 2019 on Windows
Nikita Popov [Tue, 4 Jun 2019 10:18:21 +0000 (12:18 +0200)]
Try to make 17result.php more robust
This is our most common intermittent failure. I don't understand
why it happens and can't reproduce it, but defining an explicit
sort order seems worth a shot...
Nikita Popov [Tue, 4 Jun 2019 08:18:00 +0000 (10:18 +0200)]
Azure: Don't build zend-test as shared
This is shared on Travis to perform a dl() sanity check. Revert to
building it statically here so we don't need to explicitly load it.
Hugh McMaster [Wed, 29 May 2019 13:09:43 +0000 (23:09 +1000)]
Use a common setup macro to detect the Expat library
Closes GH-4221.
Joe Watkins [Tue, 4 Jun 2019 06:23:43 +0000 (08:23 +0200)]
should be or
Joe Watkins [Tue, 4 Jun 2019 06:12:53 +0000 (08:12 +0200)]
always run
Joe Watkins [Tue, 4 Jun 2019 06:10:52 +0000 (08:10 +0200)]
run/publish test on success and failure
Joe Watkins [Tue, 4 Jun 2019 06:06:53 +0000 (08:06 +0200)]
always run/publish tests, even when failing
Joe Watkins [Tue, 4 Jun 2019 05:50:46 +0000 (07:50 +0200)]
always run/publish tests
Remi Collet [Tue, 4 Jun 2019 05:19:46 +0000 (07:19 +0200)]
Merge branch 'PHP-7.3' into PHP-7.4
* PHP-7.3:
this was in 7.3.5, not 7.3.6
Remi Collet [Tue, 4 Jun 2019 05:19:29 +0000 (07:19 +0200)]
this was in 7.3.5, not 7.3.6
Joe Watkins [Tue, 4 Jun 2019 03:04:24 +0000 (05:04 +0200)]
correct zts flags for 7.4
Joe Watkins [Mon, 3 Jun 2019 20:47:12 +0000 (22:47 +0200)]
get the name of the branch right
Joe Watkins [Wed, 29 May 2019 15:59:36 +0000 (17:59 +0200)]
Set up CI with Azure Pipelines
Nikita Popov [Mon, 3 Jun 2019 16:58:52 +0000 (18:58 +0200)]
Prevent use of TLS 1.3 in stream_server_reneg_limit.phpt
TLS 1.3 does not support renegotiation, make sure this test does
not use it.
Nikita Popov [Mon, 3 Jun 2019 13:15:22 +0000 (15:15 +0200)]
Merge branch 'PHP-7.3' into PHP-7.4
Nikita Popov [Mon, 3 Jun 2019 13:15:07 +0000 (15:15 +0200)]
Merge branch 'PHP-7.2' into PHP-7.3
Nikita Popov [Mon, 3 Jun 2019 13:14:01 +0000 (15:14 +0200)]
Clear errors after SSL_CTX_load_verify_locations()
We report our own errors here. Make sure these don't clog up the
error queue.