Avoid need of insertion NOP opcoes between unrelated SMART BRANCH instruction and following JMPZ/JMPNZ.
Now instead of checking the opcode of following instruction, the same information is encoded into SMART BRANH result_type.
Many of these now have reflection types,
making the old signatures redundant or inaccurate.
(Some methods also throw Errors instead of returning null/false)
Closes GH-4785.
preg_quote and curl_error now have a non-null return type of `string`.
zend_func_info now duplicates many real reflection types
(some of which are now non-null)
Closes GH-4788.
E.g. hash_equals is outdated - it has a non-null reflection type of bool.
Remove redundant signatures for types that are obviously not reference
counted (`F0`).
Closes GH-4787.
This is limited to types that aren't reference counted.
I assume the info that there is exactly one reference
should be preserved (e.g. for bcsqrt)
Closes GH-4786.
Some of these functions now have real reflection types
and will throw instead of return null/false.
Others provide the exact same reflection type information
that zend_func_info does. (or would after updating outdated types)
Closes GH-4790.
- all rules from pass2 moved to pass1
- all JMP unrelated rules from pass3 moved to pass1
- pass3 keeps only JMP optimization rules
- pass2.c is removed
- pass1_5.c remaned to pass1.c ("_5" was related to PHP 5)
This reverts commit b09bddcaa5.
This causes test failures on macos -- presumably the relevant
difference is that it has no global regs. I suspect that this
is related to the fact that SAVE_VALID_OPLINE may use r0 as
a scratch register, which is already in use. Reverting for now.
The verify arg functions may throw an exception, so we need to save
the opline beforehand. Asan failures could be observed in
Zend/tests/type_declarations/scalar_constant_defaults.phpt
with a ZTS build.
It is now only used to signal exceptions for property reads. ERROR
zvals are never returned back to the VM anymore, so there's no
need to check for them when receiving a VAR.
Also return MAY_BE_ERROR, as ERROR is now no longer relevant for
inference.
Resources used as array keys are generally handled by throwing a
notice and converting the resource to the resource handle. The only
exception is the [$resource => null] syntax, where this was treated
as an illegal offset type instead. However, this also only happened
for VM evaluations, the AST evaluator did handle resources correctly.
This removes object auto-vivification support.
This also means that we can remove the corresponding special
handling for typed properites: We no longer need to check that a
property is convertible to stdClass if such a conversion might
take place indirectly due to a nested property write.
Additionally OBJ_W style operations now no longer modify the
object operand, and as such we no longer need to treat op1 as a
def in SSA form.
The next step would be to actually compile the whole LHS of OBJ_W
operations in R rather than W mode, but that causes issues with
SimpleXML, whose object handlers depend on the current compilation
structure.
Part of https://wiki.php.net/rfc/engine_warnings.
See UPGRADING on the master branch.
```
- GD:
. The deprecated function image2wbmp() has been removed.
RFC: https://wiki.php.net/rfc/image2wbmp
. The deprecated functions png2wbmp() and jpeg2wbmp() have been removed.
RFC: https://wiki.php.net/rfc/deprecate-png-jpeg-2wbmp
- XML:
. xml_parser_create(_ns) will now return an XmlParser object rather than a
resource.
```
This switches zend_type from storing a single IS_* type code to
storing a MAY_BE_* type mask. Right now most code still assumes
that there is only a single type in the mask (or two together
with MAY_BE_NULL). But this will make it a lot simpler to introduce
union types.
An additional advantage (and why I'm doing this separately), is
that a number of special cases no longer need to be handled
separately: We can do a single mask & (1 << type) check to handle
all simple types, booleans (true|false) and null.
This patch saves one CPU instruction on each "_tsrm_ls_cache" access in ZTS CLI/CGI/FPM builds.
This reduce typical instruction sequence for EG(current_execute_data) access from 4 to 3 CPU instructions.
This is a fix for symfony/symfony#32995.
The behavior is:
* Throwing exception when loading parent/interface is allowed
(and we will also throw one if the class is simply not found).
* If this happens, the bucket key for the class is reset, so
it's possibly to try registering the same class again.
* However, if the class has already been used due to a variance
obligation, the exception is upgraded to a fatal error, as we
cannot safely unregister the class stub anymore.
microtime() doesn't return an array,
and gettimeofday() doesn't return a string.
See _php_gettimeofday in microtime.c (mode is non-zero for gettimeofday)
Instead of checking for this during DO_FCALL, already detect this
case during get_method()/get_static_method(), similar to visibility
checks.
This causes a minor difference in behavior, in that arguments will
no longer be evaluated. I think this is correct though (and consistent
with visibility errors).
(excluding spl_autoload)
spl_object_id() is of the most interest to me,
since I frequently call it in an application.
This includes false/null types caused by wrong argument types and wrong argument
counts.
I can't rule out iterator_to_array returning null in spl_iterator_apply,
so leave MAY_BE_NULL in.
With review comments by nikic:
Co-Authored-By: Nikita Popov <nikita.ppv@googlemail.com>
Fixes my earlier PR #4617
If I remember correctly,
The F0 macro is used for return values that are guaranteed to not be
reference counted.
The F1 macro is used for return values that may have up to 1 reference
(i.e. MAY_BE_RC1).
I didn't notice that time_nanosleep needed to be F1 since it could
return an array, and that array is reference counted.
It can return false if the resource type is wrong.
```
php > var_export(hash_update_stream(hash_init('md5'),
imagecreate(1,1)));
Warning: hash_update_stream(): supplied resource is not a valid stream
resource in php shell code on line 1
false
```
The return types were initially added in
c88ffa9a56
This can also return an array. See
https://www.php.net/manual/en/function.time-nanosleep.php#refsect1-function.time-nanosleep-returnvalues
> If the delay was interrupted by a signal, an associative array will be
returned with the components:
>
> - seconds - number of seconds remaining in the delay
> - nanoseconds - number of nanoseconds remaining in the delay
Sending a SIGUSR1 to the below program would trigger this behavior.
```
pcntl_signal(\SIGUSR1, function ($signo, $signinfo) {
echo "Handling a signal $signo\n";
});
echo "Sleeping for 100 seconds\n";
var_export(time_nanosleep(100, 0));
```
The incomplete signature existed since c88ffa9a5.
No phpt tests existed for time_nanosleep returning an array
There are a few parts here:
* opcache should not be blocking signals while invoking compile_file,
otherwise signals may remain blocked on a compile error. While at
it, also protect SHM memory during compile_file.
* We should deactivate Zend signals at the end of the request, to make
sure that we gracefully recover from a missing unblock and signals
don't remain blocked forever.
* We don't use a critical section in deactivation, because it should
not be necessary. Additionally we want to clean up the signal queue,
if it is non-empty.
* Enable SIGG(check) in debug builds so we notice issues in the future.
The smart branch logic assumed b->start refers to the old offsets,
while b->start was already adjusted to the new offsets at this
point. Delay the change until later.
Formerly, there was at most a single OPcache instance per user and the
so called system ID (which is determined from the PHP version).
Sometimes multiple OPcaches might be desired, though, particularly for
unrelated CLI scripts, which may even be necessary (e.g. for our test
suite in parallel mode).
We therefore introduce a new INI directive `opcache.cache_id` which
allows to configure independent OPcache instances for the same user.
We also use `GetUserNameW()` instead of `php_win32_get_username()`,
because the latter retrieves the user name encoded in the
`default_charset`, which can obviously yield different results for
different charsets, leading to OPcache "incompatibilities". Slightly
worse, some characters may not even be encodeable in the
`default_charset` and would be replaced by question marks, which could
result in different users sharing the same OPcache.
We also refactor, and re-use existing APIs to avoid duplicated code.
New opcache directives have been added recently which are returned
if using `ini_get_all('zend opcache')` but are not listed in the
directives if using `opcache_get_configuration()`. This fix adds
those missing directives as well as if `opcache.mmap_base` is used
instead of `opcache.lockfile_path`. Also adds a test to ensure the
directives match with both methods of fetching.
When cleaning nops in the dfa pass, we were always keeping the
smart branch inhibiting nop that occurs directly before the jump
instruction. However, as we skip unreachable blocks entirely, it
may happen that we need to keep a nop that occurs further back,
prior to the unreachable blocks. Account for that case now.
We should really do something about the smart branch situation,
this is very fragile...
Conditional function declaration may be referenced by functions or
classes that have been preloaded, so we should not move them back
to scripts.
What we probably should be doing though is to discard conditional
functions that are not used in the optimizer. This is probably
reasonably common for polyfills, where we will be able to const-eval
the conditions and drop the BBs declaring the functions, but won't
delete the function declarations themselves.
LITERAL_VALUE with related literals cannot be merged, avoid inserting
them in the literals hash in the first place. Otherwise we may end up
inserting the same key multiple times. We may also miss merging
opportunities due to the duplicate keys.
Check whether there is a parent/interface/trait method with the same
name and only then require the type to be known. This reduces the
number of cases where this triggers in practice a lot.
opcache incorrectly handles PHAR files when opcache.validate_permission
option enabled, because it calls
access("phar://path-to/file.phar/path/inside.php", R_OK);
rather than
access("path-to/file.phar", R_OK)
The SKIPIF sections are executed in the directory where run-tests.php
is located; therefore a relative path like `../skipif.inc` won't work
as desired.
This test is easily tripped by former test runs with other PHP
versions. To avoid such false positives, we check that there is at
least one respective OPcache file, and that all found OPcache user ID
folders have exactly 32 hexadecimal digits.
It would be great if this were fully initialized, but it's not
really a problem either (as long as we don't care about reproducible
file cache), so ignore this for now.
Removed unused checks:
- mbsinit check removed, HAVE_MBSINIT removed (not used in php-src)
- mempcpy check removed, HAVE_MEMPCPY removed (not used in php-src anymore since
560ed89bfb which uses PHP's own implementation)
- strpncpy check removed, added via a8c9e893b6 and
not used.
- setpgid check removed since HAVE_SETPGID is not used
Moved to a central configure.ac:
- fpclass
- mbrlen moved to configure.ac (since the HAVE_MBRLEN is used accross the php-src)
- sigprocmask
- getcwd
- getwd
- glob
- strfmon
- nice
Duplicated checks removed:
- gethostname
- getlogin
- getpwuid_r
- socketpair
- mprotect check simplified
Also adds an --asan flag to run-tests.php to setup all the necessary
environment variables. Some tests are marked as skipped because they
are incompatible with asan or too slow.
I'm basing this on the DEBUG_ZTS build, which seems to give us the
most mileage.
The strcoll function is defined in the C89 standard and should be
on today's systems always available via the <string.h> header.
https://port70.net/~nsz/c/c89/c89-draft.html#4.11.4.3
- Remove also SKIPIF strcoll check in test
Some headers were checked multiple times in the main configure.ac file
and in the bundled extensions or SAPIs themselves. Also many of these
checks are then used accross other extensions or SAPIs so a central
configure.ac makes most sense for these checks.
