Commit Graph

680 Commits

Author SHA1 Message Date
Sam Gross d23be3947c
gh-121794: Don't set `ob_tid` to zero in fast-path dealloc (#121799)
We should maintain the invariant that a zero `ob_tid` implies the
refcount fields are merged.

* Move the assignment in `_Py_MergeZeroLocalRefcount` to immediately
  before the refcount merge.
* Update `_PyTrash_thread_destroy_chain` to set `ob_ref_shared` to
  `_Py_REF_MERGED` when setting `ob_tid` to zero.

Also check this invariant with assertions in the GC in debug builds.
That uncovered a bug when running out of memory during GC.
2024-07-15 17:50:10 -04:00
Irit Katriel 6f7acaab50
gh-120686: remove unused internal c api functions (#120687) 2024-06-27 11:09:30 +01:00
Petr Viktorin 6f1d448bc1
gh-113993: Allow interned strings to be mortal, and fix related issues (GH-120520)
* Add an InternalDocs file describing how interning should work and how to use it.

* Add internal functions to *explicitly* request what kind of interning is done:
  - `_PyUnicode_InternMortal`
  - `_PyUnicode_InternImmortal`
  - `_PyUnicode_InternStatic`

* Switch uses of `PyUnicode_InternInPlace` to those.

* Disallow using `_Py_SetImmortal` on strings directly.
  You should use `_PyUnicode_InternImmortal` instead:
  - Strings should be interned before immortalization, otherwise you're possibly
    interning a immortalizing copy.
  - `_Py_SetImmortal` doesn't handle the `SSTATE_INTERNED_MORTAL` to
    `SSTATE_INTERNED_IMMORTAL` update, and those flags can't be changed in
    backports, as they are now part of public API and version-specific ABI.

* Add private `_only_immortal` argument for `sys.getunicodeinternedsize`, used in refleak test machinery.

* Make sure the statically allocated string singletons are unique. This means these sets are now disjoint:
  - `_Py_ID`
  - `_Py_STR` (including the empty string)
  - one-character latin-1 singletons

  Now, when you intern a singleton, that exact singleton will be interned.

* Add a `_Py_LATIN1_CHR` macro, use it instead of `_Py_ID`/`_Py_STR` for one-character latin-1 singletons everywhere (including Clinic).

* Intern `_Py_STR` singletons at startup.

* For free-threaded builds, intern `_Py_LATIN1_CHR` singletons at startup.

* Beef up the tests. Cover internal details (marked with `@cpython_only`).

* Add lots of assertions

Co-Authored-By: Eric Snow <ericsnowcurrently@gmail.com>
2024-06-21 17:19:31 +02:00
Sam Gross 3af7263037
gh-117511: Make PyMutex public in the non-limited API (#117731) 2024-06-20 11:29:08 -04:00
Victor Stinner 16f8e22e7c
gh-120600: Make Py_TYPE() opaque in limited C API 3.14 (#120601)
In the limited C API 3.14 and newer, Py_TYPE() is now implemented as
an opaque function call to hide implementation details.
2024-06-18 14:28:48 +00:00
Sam Gross 4055577221
gh-119999: Fix potential race condition in `_Py_ExplicitMergeRefcount` (#120000)
We need to write to `ob_ref_local` and `ob_tid` before `ob_ref_shared`.
Once we mark `ob_ref_shared` as merged, some other thread may free the
object because the caller also passes in `-1` as `extra` to give up its
only reference.
2024-06-04 15:26:26 -04:00
Eric Snow 105f22ea46
gh-117398: Use Per-Interpreter State for the _datetime Static Types (gh-119929)
We make use of the same mechanism that we use for the static builtin types.  This required a few tweaks.

The relevant code could use some cleanup but I opted to avoid the significant churn in this change.  I'll tackle that separately.

This change is the final piece needed to make _datetime support multiple interpreters.  I've updated the module slot accordingly.
2024-06-03 17:09:18 -06:00
Sam Gross 47fb4327b5
gh-117657: Fix race involving immortalizing objects (#119927)
The free-threaded build currently immortalizes objects that use deferred
reference counting (see gh-117783). This typically happens once the
first non-main thread is created, but the behavior can be suppressed for
tests, in subinterpreters, or during a compile() call.

This fixes a race condition involving the tracking of whether the
behavior is suppressed.
2024-06-03 20:58:41 +00:00
mpage cb6f75a32c
gh-117657: Fix data races when writing / reading `ob_gc_bits` (#118292)
Use relaxed atomics when reading / writing to the field. There are still a
few places in the GC where we do not use atomics. Those should be safe as
the world is stopped.
2024-05-08 16:03:39 -04:00
Jelle Zijlstra 3c079a0203
gh-118767: Make bool(NotImplemented) raise TypeError (#118775) 2024-05-08 11:12:00 -07:00
Dino Viehland 636b8d94c9
gh-112075: Fix race in constructing dict for instance (#118499) 2024-05-06 23:31:09 +00:00
Dino Viehland 5a1618a2c8
gh-118362: Fix thread safety around lookups from the type cache in the face of concurrent mutators (#118454)
Add _PyType_LookupRef and use incref before setting attribute on type
Makes setting an attribute on a class and signaling type modified atomic
Avoid adding re-entrancy exposing the type cache in an inconsistent state by decrefing after type is updated
2024-05-06 10:50:35 -07:00
Tian Gao b034f14a4b
gh-74929: Implement PEP 667 (GH-115153) 2024-05-04 12:12:10 +01:00
Pablo Galindo Salgado 6bcbee09df
gh-93502: Add new C-API functions to trace object creation and destruction (#115945) 2024-05-02 19:30:00 +02:00
Guido van Rossum 7d83f7bcc4
gh-118335: Configure Tier 2 interpreter at build time (#118339)
The code for Tier 2 is now only compiled when configured
with `--enable-experimental-jit[=yes|interpreter]`.

We drop support for `PYTHON_UOPS` and -`Xuops`,
but you can disable the interpreter or JIT
at runtime by setting `PYTHON_JIT=0`.
You can also build it without enabling it by default
using `--enable-experimental-jit=yes-off`;
enable with `PYTHON_JIT=1`.

On Windows, the `build.bat` script supports
`--experimental-jit`, `--experimental-jit-off`,
`--experimental-interpreter`.

In the C code, `_Py_JIT` is defined as before
when the JIT is enabled; the new variable
`_Py_TIER2` is defined when the JIT *or* the
interpreter is enabled. It is actually a bitmask:
1: JIT; 2: default-off; 4: interpreter.
2024-04-30 18:26:34 -07:00
Sam Gross 7ccacb220d
gh-117783: Immortalize objects that use deferred reference counting (#118112)
Deferred reference counting is not fully implemented yet. As a temporary
measure, we immortalize objects that would use deferred reference
counting to avoid multi-threaded scaling bottlenecks.

This is only performed in the free-threaded build once the first
non-main thread is started. Additionally, some tests, including refleak
tests, suppress this behavior.
2024-04-29 14:36:02 -04:00
Dino Viehland 8b541c017e
gh-112075: Make instance attributes stored in inline "dict" thread safe (#114742)
Make instance attributes stored in inline "dict" thread safe on free-threaded builds
2024-04-21 22:57:05 -07:00
Dino Viehland b45af00bad
[gh-117657] _Py_MergeZeroLocalRefcount isn't loading ob_ref_shared with strong enough semantics (#118111)
Use acquire for load of ob_ref_shared
2024-04-19 14:21:01 -07:00
Irit Katriel c179c0e6cb
gh-117680: make _PyInstructionSequence a PyObject and use it in tests (#117629) 2024-04-17 16:42:04 +01:00
Mark Shannon 147cd0581e
GH-117760: Streamline the trashcan mechanism (GH-117763) 2024-04-17 11:08:05 +01:00
Sam Gross 4ad8f090cc
gh-117376: Partial implementation of deferred reference counting (#117696)
This marks objects as using deferred refrence counting using the
`ob_gc_bits` field in the free-threaded build and collects those objects
during GC.
2024-04-12 17:36:20 +00:00
Serhiy Storchaka 2d3d9b4461
gh-117764: Add docstrings and signatures for the types of None, Ellipsis and NotImplemented (GH-117813) 2024-04-12 15:45:23 +03:00
Sam Gross 1a6594f661
gh-117439: Make refleak checking thread-safe without the GIL (#117469)
This keeps track of the per-thread total reference count operations in
PyThreadState in the free-threaded builds. The count is merged into the
interpreter's total when the thread exits.
2024-04-08 12:11:36 -04:00
Mark Shannon c32dc47aca
GH-115776: Embed the values array into the object, for "normal" Python objects. (GH-116115) 2024-04-02 11:59:21 +01:00
Eric Snow 617158e078
gh-76785: Drop PyInterpreterID_Type (gh-117101)
I added it quite a while ago as a strategy for managing interpreter lifetimes relative to the PEP 554 (now 734) implementation.  Relatively recently I refactored that implementation to no longer rely on InterpreterID objects.  Thus now I'm removing it.
2024-03-21 17:15:02 +00:00
Victor Stinner 8bea6c411d
gh-115754: Add Py_GetConstant() function (#116883)
Add Py_GetConstant() and Py_GetConstantBorrowed() functions.

In the limited C API version 3.13, getting Py_None, Py_False,
Py_True, Py_Ellipsis and Py_NotImplemented singletons is now
implemented as function calls at the stable ABI level to hide
implementation details. Getting these constants still return borrowed
references.

Add _testlimitedcapi/object.c and test_capi/test_object.py to test
Py_GetConstant() and Py_GetConstantBorrowed() functions.
2024-03-21 16:07:00 +00:00
Mark Shannon 15309329b6
GH-108362: Incremental Cycle GC (GH-116206) 2024-03-20 08:54:42 +00:00
Sam Gross 347acded84
gh-115491: Keep some fields valid across allocations in obmalloc (free-threading) (#115745) 2024-02-21 10:00:08 -08:00
Donghee Na f15795c9a0
gh-111968: Rename freelist related struct names to Eric's suggestion (gh-115329) 2024-02-14 00:32:51 +00:00
Donghee Na d4d5bae147
gh-111968: Refactor _PyXXX_Fini to integrate with _PyObject_ClearFreeLists (gh-114899) 2024-02-10 00:57:04 +00:00
Sam Gross a3af3cb4f4
gh-110481: Implement inter-thread queue for biased reference counting (#114824)
Biased reference counting maintains two refcount fields in each object:
`ob_ref_local` and `ob_ref_shared`. The true refcount is the sum of these two
fields. In some cases, when refcounting operations are split across threads,
the ob_ref_shared field can be negative (although the total refcount must be
at least zero). In this case, the thread that decremented the refcount
requests that the owning thread give up ownership and merge the refcount
fields.
2024-02-09 17:08:32 -05:00
Sam Gross 31633f4473
gh-115184: Fix refleak tracking issues in free-threaded build (#115188)
Fixes a few issues related to refleak tracking in the free-threaded build:

- Count blocks in abandoned segments
- Call `_mi_page_free_collect` earlier during heap traversal in order to get an accurate count of blocks in use.
- Add missing refcount tracking in `_Py_DecRefSharedDebug` and `_Py_ExplicitMergeRefcount`.
- Pause threads in  `get_num_global_allocated_blocks` to ensure that traversing the mimalloc heaps is safe.
2024-02-09 09:23:12 -05:00
Mark Shannon 8a3c499ffe
GH-108362: Revert "GH-108362: Incremental GC implementation (GH-108038)" (#115132)
Revert "GH-108362: Incremental GC implementation (GH-108038)"

This reverts commit 36518e69d7.
2024-02-07 12:38:34 +00:00
Mark Shannon 36518e69d7
GH-108362: Incremental GC implementation (GH-108038) 2024-02-05 18:28:51 +00:00
Sam Gross 587d480203
gh-112529: Remove PyGC_Head from object pre-header in free-threaded build (#114564)
* gh-112529: Remove PyGC_Head from object pre-header in free-threaded build

This avoids allocating space for PyGC_Head in the free-threaded build.
The GC implementation for free-threaded CPython does not use the
PyGC_Head structure.

 * The trashcan mechanism uses the `ob_tid` field instead of `_gc_prev`
   in the free-threaded build.
 * The GDB libpython.py file now determines the offset of the managed
   dict field based on whether the running process is a free-threaded
   build. Those are identified by the `ob_ref_local` field in PyObject.
 * Fixes `_PySys_GetSizeOf()` which incorrectly incorrectly included the
   size of `PyGC_Head` in the size of static `PyTypeObject`.
2024-02-01 12:29:19 -08:00
Sam Gross d0f0308a37
gh-113750: Fix object resurrection in free-threaded builds (gh-113751)
gh-113750: Fix object resurrection on free-threaded builds

This avoids the undesired re-initializing of fields like `ob_gc_bits`,
`ob_mutex`, and `ob_tid` when an object is resurrected due to its
finalizer being called.

This change has no effect on the default (with GIL) build.
2024-01-06 12:12:26 +09:00
Christopher Chavez a545a86ec6
gh-111178: Make slot functions in typeobject.c have compatible types (GH-112752) 2023-12-20 15:13:44 +01:00
andrewluotechnologies 9c3458e058
gh-112125: Fix None.__ne__(None) returning NotImplemented instead of False (#112504) 2023-12-07 13:56:01 +01:00
Hugo van Kemenade 3b3ec0d77f
gh-111863: Rename `Py_NOGIL` to `Py_GIL_DISABLED` (#111864)
Rename Py_NOGIL to Py_GIL_DISABLED
2023-11-20 15:52:00 +02:00
Serhiy Storchaka 18203a6bc9
gh-111789: Use PyDict_GetItemRef() in Objects/ (GH-111827) 2023-11-14 11:25:39 +02:00
Sam Gross fa84e5fe0a
gh-110481: fix 'unused function' warning for `is_shared_refcnt_dead`. (gh-111974)
Fix 'unused function' warning for `is_shared_refcnt_dead`.

The `is_shared_refcnt_dead` function is only used if `Py_REF_DEBUG` is set.
2023-11-11 08:54:35 +09:00
Sam Gross 31c90d5838
gh-111569: Implement Python critical section API (gh-111571)
Critical sections are helpers to replace the global interpreter lock
with finer grained locking.  They provide similar guarantees to the GIL
and avoid the deadlock risk that plain locking involves.  Critical
sections are implicitly ended whenever the GIL would be released.  They
are resumed when the GIL would be acquired.  Nested critical sections
behave as if the sections were interleaved.
2023-11-08 15:39:29 -07:00
Serhiy Storchaka f55cb44359
gh-106672: C API: Report indiscriminately ignored errors (GH-106674)
Functions which indiscriminately ignore all errors now report them as
unraisable errors.
2023-11-07 15:58:04 +02:00
Victor Stinner 20cfab903d
gh-111506: Implement Py_SET_REFCNT() as opaque function in limited C API (#111508)
In the limited C API version 3.13, Py_SET_REFCNT() function is now
implemented as an opaque function call.

Add _Py_SetRefcnt() to the stable ABI.
2023-11-03 18:18:57 +01:00
Sam Gross 6dfb8fe023
gh-110481: Implement biased reference counting (gh-110764) 2023-10-30 16:06:09 +00:00
Savannah Ostrowski 4a929d432b
GH-111339: Fix initialization and finalization of static optimizer types (GH-111430) 2023-10-29 13:53:25 -07:00
Victor Stinner 8b626a47ba
gh-110079: Remove extern "C" { ...} in C code (#110080) 2023-09-29 10:56:49 +02:00
Serhiy Storchaka add16f1a5e
gh-108511: Add C API functions which do not silently ignore errors (GH-109025)
Add the following functions:

* PyObject_HasAttrWithError()
* PyObject_HasAttrStringWithError()
* PyMapping_HasKeyWithError()
* PyMapping_HasKeyStringWithError()
2023-09-17 14:23:31 +03:00
Victor Stinner b936cf4fe0
gh-108634: PyInterpreterState_New() no longer calls Py_FatalError() (#108748)
pycore_create_interpreter() now returns a status, rather than
calling Py_FatalError().

* PyInterpreterState_New() now calls Py_ExitStatusException() instead
  of calling Py_FatalError() directly.
* Replace Py_FatalError() with PyStatus in init_interpreter() and
  _PyObject_InitState().
* _PyErr_SetFromPyStatus() now raises RuntimeError, instead of
  ValueError. It can now call PyErr_NoMemory(), raise MemoryError,
  if it detects _PyStatus_NO_MEMORY() error message.
2023-09-01 12:43:30 +02:00
Victor Stinner 13a00078b8
gh-108634: Py_TRACE_REFS uses a hash table (#108663)
Python built with "configure --with-trace-refs" (tracing references)
is now ABI compatible with Python release build and debug build.
Moreover, it now also supports the Limited API.

Change Py_TRACE_REFS build:

* Remove _PyObject_EXTRA_INIT macro.
* The PyObject structure no longer has two extra members (_ob_prev
  and _ob_next).
* Use a hash table (_Py_hashtable_t) to trace references (all
  objects): PyInterpreterState.object_state.refchain.
* Py_TRACE_REFS build is now ABI compatible with release build and
  debug build.
* Limited C API extensions can now be built with Py_TRACE_REFS:
  xxlimited, xxlimited_35, _testclinic_limited.
* No longer rename PyModule_Create2() and PyModule_FromDefAndSpec2()
  functions to PyModule_Create2TraceRefs() and
  PyModule_FromDefAndSpec2TraceRefs().
* _Py_PrintReferenceAddresses() is now called before
  finalize_interp_delete() which deletes the refchain hash table.
* test_tracemalloc find_trace() now also filters by size to ignore
  the memory allocated by _PyRefchain_Trace().

Test changes for Py_TRACE_REFS:

* Add test.support.Py_TRACE_REFS constant.
* Add test_sys.test_getobjects() to test sys.getobjects() function.
* test_exceptions skips test_recursion_normalizing_with_no_memory()
  and test_memory_error_in_PyErr_PrintEx() if Python is built with
  Py_TRACE_REFS.
* test_repl skips test_no_memory().
* test_capi skisp test_set_nomemory().
2023-08-31 18:33:34 +02:00