Fix copy-and-paste mistakes in the sodium extension.
crypto_scalarmult_BYTES and crypto_scalarmult_ristretto255_BYTES
have the same value, so behavior is unchanged.
RFC: https://wiki.php.net/rfc/partial_function_application_v2
For FCCs, the parser generates a normal function call AST node, the but argument
list is a ZEND_AST_CALLABLE_CONVERT / zend_ast_fcc node.
We extend this for PFAs so that zend_ast_fcc can represent arguments.
* Support PFA syntax in grammar
* Update zend_ast_fcc so that arguments can be represented
* Support serialization of zend_ast_fcc arguments in SHM / file cache
* Introduce zend_ast_arg_list_add(): Same as zend_ast_list_add(), but wraps the
list in a ZEND_AST_CALLABLE_CONVERT when adding any placeholder argument.
Technically the arg list wrapping is not required, but it results in simpler
code later as it will be very convenient in the compiler (determines whether a
function calls is a PFA/FCC), and for PFA-in-const-expr support. It also allows
to unify FCCs and PFAs in the grammar.
Closes GH-20717.
The remote resources don't work because remote streams don't have a stat
method.
Since the check is only here for a best-effort check to return
"directory" instead of "empty", we can try the stat and still execute
the magic_stream() code even if it failed. Unfortunately we can't
distinguish between a failed stat and an unimplemented stat. If we
could, then this code could be even more robust.
As the comment suggests, the return value of frameless calls is rarely not used.
When it isn't used (and isn't refcounted) the FREE is already elided by the
optimizer.
Closes GH-20819
zend_string_truncate() doesn't put a NUL byte.
Right now this doesn't matter as this code path is only hittable via the
equals() method, but if other extension (or future other code) starts
using this code path, then it can be problematic as all user-exposed
zend_strings need to end with a NUL byte.
The reason why freeing was not done yet is because the pointer in these
variables may be:
- Static data set by the readline/libedit library initially, not heap
data.
- Data set by another thread. Although the libraries appear to be not
thread-safe anyway.
To solve this, introduce some TLS variables to hold a pointer for us
when we override the settings, such that we can free them and are
certain they are allocated by us.
Closes GH-20794.
This removes the artificial limitation that is not necessary. The fact
that some streams can have some data buffered is not a problem because
the similar situation is already present for OpenSSL streams where
OpenSSL can internally buffer data for the unprocessed part of the
record.
Closes GH-20540
Ilija Tovilo