Under high thread concurrency in free-threaded builds, `_Py_Specialize_LoadGlobal` suffers from lock contention when acquiring the critical section mutexes for the `globals` and `builtins` dictionaries during bytecode specialization.
This PR skips LOAD_GLOBAL bytecode specialization if acquiring the two object mutexes would block.
Revert commits:
gh-151593: Fix dead lock in PyDict insert_split_key() (#152200)
gh-150490: Raise PyType_Modified for insertion into split dictionary (#150489)
For gh-150489, it violates locking discipline and results in deadlocks,
gh-151593 is an example of it being hit in CI. The attempted fix
gh-152200 avoids the deadlock but introduces a data-race. The race
window is small but can be triggered with pure Python code.
* SEND specialization. Adds 2 new specialized instructions:
* SEND_VIRTUAL: for sends to virtual iterators e.g lists and tuples
* SEND_ASYNC_GEN: for sends to async generators
Tweak FOR_ITER_VIRTUAL so that SEND_VIRTUAL and FOR_ITER_VIRTUAL use equivalent guards
* Add FOR_ITER_VIRTUAL to specialize FOR_ITER for virtual iterators
* Add GET_ITER_SELF to specialize GET_ITER for iterators (including generators)
* Add GET_ITER_VIRTUAL to specialize GET_ITER for iterables as virtual iterators
* Add new (internal) _tp_iteritem function slot to PyTypeObject
* Put limited RESUME at start of genexpr for free-threading. Fix up exception handling in genexpr
Remove spurious Py_DECREF on borrowed ref in LOAD_GLOBAL specialization
_PyDict_LookupIndexAndValue() returns a borrowed reference via
_Py_dict_lookup(), but specialize_load_global_lock_held() called
Py_DECREF(value) on it when bailing out for lazy imports. Each time
the adaptive counter fired while a lazy import was still in globals,
this stole one reference from the dict's object. With 8+ threads
racing through LOAD_GLOBAL during concurrent lazy import resolution,
enough triggers accumulated to drive the refcount to zero while the
dict and other threads still referenced the object, causing
use-after-free.
Now that the specializing interpreter works with free threading,
replace ENABLE_SPECIALIZATION_FT with ENABLE_SPECIALIZATION and
replace requires_specialization_ft with requires_specialization.
Also limit the uniquely referenced check to FOR_ITER_RANGE. It's not
necessary for FOR_ITER_GEN and would cause test_for_iter_gen to fail.
This reverts gh-144055 and fixes the bug in a different way. Deferred
reference counting relies on the object being tracked by the GC,
otherwise the object will live until interpreter shutdown. So, take
care that we do not enable deferred reference counting for objects that
are untracked. Also, if a tuple has deferred reference counting
enabled, don't untrack it.
If we are specializing to `LOAD_GLOBAL_MODULE` or `LOAD_ATTR_MODULE`, try
to enable deferred reference counting for the value, if the object is owned by
a different thread. This applies to the free-threaded build only and should
improve scaling of multi-threaded programs.
Allow the --enable-pystats build option to be used with free-threading. The
stats are now stored on a per-interpreter basis, rather than process global.
For free-threaded builds, the stats structure is allocated per-thread and
then periodically merged into the per-interpreter stats structure (on thread
exit or when the reporting function is called). Most of the pystats related
code has be moved into the file Python/pystats.c.
* FOR_ITER now pushes either the iterator and NULL or leaves the iterable and pushes tagged zero
* NEXT_ITER uses the tagged int as the index into the sequence or, if TOS is NULL, iterates as before.
Sumit Kumar Das