This issue is properly fixed by GH-7121 on master. For older
branches, disable the use of range information in SCCP, to
reduce impact of potentially incorrect ranges.
Literal compaction was incorrectly assuming that literals with
the same base literal and the same number of related literals
would be equal. Maybe that was the case historically, but at
least it isn't true in PHP 8, where FETCH_CONSTANT and INIT_METHOD
have distinct literals at the second position.
Fix this by making the cache key a concatenation of all literals,
rather than just the base literal. We still distinguish the number
of related literals based on a bias added to the string hash.
Make sure that the previous opline is part of the same block,
otherwise it may be non-dominating.
The test case does not fail on PHP-7.4, but I think the general
problem can appear on 7.4 as well, so I'm applying the patch to
that branch.
When encountering the following SSA graph:
BB1:
#2.T1 [string] = COALESCE #1.CV0($str) [null, string] BB2
BB2:
#5.T1 [string] = QM_ASSIGN string("")
BB3:
#7.X1 [string] = Phi(#2.X1 [string], #5.X1 [string])
FREE #7.T1 [string]
We would currently determine that #7, #5 are dead, and eliminate
the FREE and QM_ASSIGN. However, we cannot eliminate #2, as
COALESCE is also responsible for control flow.
Fix this my marking all non-CV phis as live to start with. This
can be relaxed to check the kind of the source instruction, but
I couldn't immediately come up with a case where it would be
useful.
Don't use r0 as temporary register in math_double_long if it is
already used for a memory result.
This was already done in one branch, but not the other.
Replacing the result type in the general case is dangerous,
because not all opcodes support both VAR and TMP. One common case
is the in_array() result being passed to SEND_VAR, which would
have to be changed to SEND_VAL.
Rather than complicating this logic, reduce the scope to only
doing the type replacement for JMPZ and JMPNZ. The only reason
we're doing this in the first place is to enable the smart branch
optimization, so we can limit it to the relevant opcodes. Replacing
the result type may be marginally useful in other cases as well
(as it may avoid reference checks), but not worth the bother.
We should clear the exception *before* we destroy the execute_data.
Add a variation of the test that indirects through another file,
and would crash otherwise.
If there is a previous use of the new variable in the phi, we need
to NULL out the use chain of the new source we're adding.
Test case is reduced from an assertion failure in the Symfony Demo.
This cherry-picks 33969c2252 and
2effbfd871 from PHP-8.0.
The issues these commits fix could also manifest in PHP 7.4, and
a commenter on bug #80307 reports this this might indeed be
happening.
If a trait method is inherited, preloading trait fixup might be
performed on it multiple times. Usually this is fine, because
the opcodes pointer will have already been updated, and will thus
not be found in the xlat table.
However, it can happen that the new opcodes pointer is the same
as one of the old opcodes pointers, if the pointer has been reused
by the allocator. In this case we will look up the wrong op array
and overwrite the trait method with an unrelated trait method.
We fix this by indexing the xlat table not by the opcodes pointer,
but by the refcount pointer. The refcount pointer is not changed
during optimization, and accurately represents which op arrays
should use the same opcodes.
Fixes bug #80307. The test case does not reproduce the bug, because
this depends on a lot of "luck" with the allocator. The test case
merely illustrates a case where orig_op_array would have been NULL
in the original code.
We should use normal function renaming if the function is declared
during preloading itself, rather than afterwards.
This fixes a regression introduced by
68f80be9d1.
We should only disable early binding during the opcache_compile_file()
calls, not inside the preloading script or anything it includes.
The right condition to check for is whether we compile the file
without execution, as declaring classes is "execution".
This is a bit annoying: When preloading is used, types might be
resolved during inheritance checks, so we need to deal with CE
types rather than just NAME types everywhere.
When preloading, it's fine to make use of internal class information,
as we do not support Windows. It is also necessary to allow proper
variance checks against internal classes.
Dmitry Stogov