Remove most uses of the ERROR type

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

Remove most uses of the ERROR type

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.

show more ...

Convert "cannot add element" warning to exception

Use cheaper code for scalar type check

Change ZEND_RECV and ZEND_RECV_VARIADIC to use extended_value for cache slot (instead of op2), to be consistent with ZEND_RECV_INIT.

Change representation of zend_type from type code to MAY_BE_* mask

This switches zend_type from storing a single IS_* type code to
storing a MAY_BE_* type mask. Right now most code still

Change representation of zend_type from type code to MAY_BE_* mask

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.

show more ...

Fixed 32-bit JIT

Don't make argument nullable based on AST null initializer

Closes GH-4720.

Optimize access to thread local cache.

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(cur

Optimize access to thread local cache.

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.

show more ...

clenup

Avoid hash value recalculation on each counter update

Detect calls to abstract methods in get_method() already

Instead of checking for this during DO_FCALL, already detect this
case during get_method()/get_static_method(), similar to visibi

Detect calls to abstract methods in get_method() already

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).

show more ...

JIT: Fix missing free on JMPZ_EX with op1 == res

We need to free the operand before we overwrite it with the result.

Drop support for JIT without SSE

Closes GH-4388.

Read from original address

References are derefed into FCARG1, which is later clobbered by the
undef handling code. Make sure we load from the original address
instead.

Implement array access notice in JIT

Reduce cost for GC references to strings and resources in JIT

Always generate interrupt check in jit

Even if zend_interrupt_function is NULL, we still need to perform
the interrupt check for timeouts (which do not use
zend_interrupt_function).

Undef opline result on mod/shift error in jit

Fix out of bounds read in jit_fetch_obj_read

Fixed JIT failure, on ZTS build without global register vatriables

Attempt to fix JIT on Mac OSX ZTS build

Don't verify arginfo types for internal functions

To avoid duplicate type checks. In debug builds arginfo is still
checked and will generate an assertions if the function doesn't
sub

Don't verify arginfo types for internal functions

To avoid duplicate type checks. In debug builds arginfo is still
checked and will generate an assertions if the function doesn't
subsequently throw an exception.

Some test results change due to differences in zpp and arginfo
error messages.

show more ...

Fixed incorect register allocation in ZTS build

Fix helper call on 32-bit

This is using the fast-call calling convention. Use FCARG1a and
same code for 32-bit and 64-bit.

Fix bug #78154 with JIT