* Check for iODBC and unixODBC with pkg-config in PDO_ODBC
PDO_ODBC required that these backends had their path specified manually,
which was clumsy and contrary to how procedural ODBC checked it. This
adds a pkg-config based path to check for these backends that ignores
the 'dir' part of the flag, so i.e. --with-pdo-odbc=unixODBC should pick
it up from the correct location.
Generic and the special ibm-db2 usecase should be unaffected. The header
situation is unfortunately ugly, and has a workaround; this should also
be cleaned up.
* Move check for valid headers to after
* Use existing CFLAGS for PDO_ODBC header check
...instead of a separate funny variable. It does mean we have to save
and restore the value of CPPFLAGS, as AC_CHECK_HEADERS and friends rely
on that variable instead of CFLAGS.
Co-authored-by: Peter Kokot <peterkokot@gmail.com>
* Move PDO_ODBC_TYPE to AC_DEFINE, simplify CFLAGS handling
The variable PDO_ODBC_INCLUDE becomes redundant, as is the CFLAGS
override for PHP_NEW_EXTENSION if we call PHP_EVAL_INCLINE in the
generic case.
Co-authored-by: Peter Kokot <peterkokot@gmail.com>
* Use same variable names so evals can be combined
* Fix identation
* Suggested shell syntax cleanups
---------
Co-authored-by: Peter Kokot <peterkokot@gmail.com>
* ext/standard: change `highlight_string()` return type from `string|bool` to `string|true`
* ext/standard: change `print_r` return type from `string|bool` to `string|true`
The template element in HTML 5 is special in the sense that it does not
add its contents into the DOM tree, but instead keeps them in a separate
shadow DOM document fragment. Interacting with the DOM tree cannot touch
the elements in the document fragment.
Closes GH-14906.
The test failure is not particularly related to Travis, but rather is
caused by the GD font file to only be suitable for platforms where
`int` stores 32bit values in little endian byte order. This platform
dependence is documented in the source code[1]. Thus we fix the skip
condition and skip reason accordingly.
An alternative would be to dynamically create the font file just before
running the test, but that appears to be overkill.
[1] <d59691c02f/ext/gd/gd.c (L545-L556)>
Closes GH-14922.
Besides demonstrating the new behavior, this test also ensures that the
bundled and external libgd now behave the same. It has to be noted,
though, that we only test one of the five code paths.
Closes GH-14945.
- CS synced
- When checking for libcurl linked against old OpenSSL the LIBS can be
used instead of LDFLAGS to put -lcurl to proper place. Also, flags
manipulation variables are wrapped in the AC_CACHE_CHECK commands
because there is also OpenSSL setup done later in the code which
changes LDFLAGS, LIBS and/or CFLAGS.
- CFLAGS added to the check to have edge case of -I flags of custom
installation paths taken into consideration
- All macro arguments quoted
- SSL check simplified a bit
- The HAVE_CURL symbol help text synced
- Shebang added so it can be called with ./generate_patch.sh
- Script changes its working directory to ext/fileinfo (where it is
located) so it can be also called as ./ext/fileinfo/generate_patch.sh
- Syntax adjusted a bit for using Bourne or a compatible shell (/bin/sh)
- Downloaded files added to .gitignore
According to the docs (https://www.php.net/manual/en/function.imagecreatefromwebp.php and https://www.php.net/manual/en/function.imagewebp.php), `false` should be returned on errors (similar to other functions of the `gd` extension), but actually all errors result in a `Fatal Error`. It doesn't look normal when trying to read an empty file or a file in the wrong format causes the program to stop. The problem seems to be related to a mega-patch that replaced `zend_error` with `zend_error_noreturn` almost everywhere. My patch fixes this behavior by switching from `zend_error_noerror` to `gd_error` (i.e. to `E_WARNING` level). All necessary memory cleanup is already in the code (as it was before the "zend_error_noreturn" patch).
Close GH-13774
- Macro renamed to PHP_PDO_ODBC_CHECK_HEADER
- AC_DEFINE_UNQUOTED used instead of PHP_DEF_HAVE
- help texts added to CPP macro definitions
- CS synced a bit
LDFLAGS is by convention used more for linker options like -s and -L.
The -l option should go to LIBS instead. This does the same check but
is more understandable in the logs and M4 code.
The problem is that this line in the VM: `ZVAL_NULL(result);` changes the type
of arg1 as well, because after the DFA pass the result and input both use
CV0($result).
We should not contract assignments with CVs in frameless calls with
arguments.
An older attempt is found at GH-14876 that tried to modify the VM/JIT.
Closes GH-14903.
- Fixed warning: cast from pointer to integer of different size
using the intptr_t instead of int type case
- CS synced
- Check wrapped in AC_CACHE_CHECK
- Over-quoted arguments reduced
- Also the include flags added to the check when building with oniguruma
installed on custom places
We don't need to reconcile when we clone into the same document because
the namespace mapper is the same. Only when cloning into another
document is the namespace mapper different and do we need a
reconciliation.
In the original specification, the scale of bc_num was directly changed
and compared.
This becomes a problem when objects are supported, so we will modify it
to compare without changing bc_num.
The LICENSE file of libmagic had been inadvertently removed when the
lib had been upgraded to 5.43. So we add the file `COPYING` from that
release[1], and rename it to `LICENSE`.
[1] <https://github.com/file/file/releases/tag/FILE5_43>
Closes GH-14917.
The create_obj handler of InternalIterator is overwritten, but not the
clone_obj handler. This is not allowed.
In PHP 8.2 this didn't cause a segfault because the standard object
handler was used for the clone instead of the internal handler.
So then it allocates and frees the object using the standard object handlers.
In 8.3 however, the object is created using the standard object handler and
freed using the custom handler, resulting in the buffer overflow.
Even though bisect points to 1e1ea4f this only reveals the bug.
Closes GH-14882.
This creates a single M4 macro PHP_CHECK_BUILTIN and removes other
PHP_CHECK_BUILTIN_* macros. Checks are wrapped in AC_CACHE_CHECK and
PHP_HAVE_BUILTIN_* CPP macro definitions are defined to 1 if builtin
is found and undefined if not.
This also changes all PHP_HAVE_BUILTIN_ symbols to be either undefined
or defined (to value 1) and syncs all #if/ifdef/defined usages of them
in the php-src code. This way it is simpler to use them because they
don't need to be defined to value 0 on Windows, for example. This is
done as previous usages in php-src were mixed and on many places they
were only checked with ifdef.
