Commit Graph

80 Commits

Author SHA1 Message Date
Victor Stinner f966e5397e
bpo-29564:_PyMem_DumpTraceback() suggests enabling tracemalloc (GH-10510)
If tracemalloc is not tracing Python memory allocations,
_PyMem_DumpTraceback() now suggests to enable tracemalloc
to get the traceback where the memory block has been allocated.
2018-11-13 15:14:58 +01:00
Victor Stinner 626bff8568
bpo-9263: Dump Python object on GC assertion failure (GH-10062)
Changes:

* Add _PyObject_AssertFailed() function.
* Add _PyObject_ASSERT() and _PyObject_ASSERT_WITH_MSG() macros.
* gc_decref(): replace assert() with _PyObject_ASSERT_WITH_MSG() to
  dump the faulty object if the assertion fails.

_PyObject_AssertFailed() calls:

* _PyMem_DumpTraceback(): try to log the traceback where the object
  memory has been allocated if tracemalloc is enabled.
* _PyObject_Dump(): log repr(obj).
* Py_FatalError(): log the current Python traceback.

_PyObject_AssertFailed() uses _PyObject_IsFreed() heuristic to check
if the object memory has been freed by a debug hook on Python memory
allocators.

Initial patch written by David Malcolm.

Co-Authored-By: David Malcolm <dmalcolm@redhat.com>
2018-10-25 17:31:10 +02:00
Victor Stinner 9e00e80e21
bpo-35053: Enhance tracemalloc to trace free lists (GH-10063)
tracemalloc now tries to update the traceback when an object is
reused from a "free list" (optimization for faster object creation,
used by the builtin list type for example).

Changes:

* Add _PyTraceMalloc_NewReference() function which tries to update
  the Python traceback of a Python object.
* _Py_NewReference() now calls _PyTraceMalloc_NewReference().
* Add an unit test.
2018-10-25 13:31:16 +02:00
Victor Stinner a7368ac636
bpo-32030: Enhance Py_Main() (#4412)
Parse more env vars in Py_Main():

* Add more options to _PyCoreConfig:

  * faulthandler
  * tracemalloc
  * importtime

* Move code to parse environment variables from _Py_InitializeCore()
  to Py_Main(). This change fixes a regression from Python 3.6:
  PYTHONUNBUFFERED is now read before calling pymain_init_stdio().
* _PyFaulthandler_Init() and _PyTraceMalloc_Init() now take an
  argument to decide if the module has to be enabled at startup.
* tracemalloc_start() is now responsible to check the maximum number
  of frames.

Other changes:

* Cleanup Py_Main():

  * Rename some pymain_xxx() subfunctions
  * Add pymain_run_python() subfunction

* Cleanup Py_NewInterpreter()
* _PyInterpreterState_Enable() now reports failure
* init_hash_secret() now considers pyurandom() failure as an "user
  error": don't fail with abort().
* pymain_optlist_append() and pymain_strdup() now sets err on memory
  allocation failure.
2017-11-15 18:11:45 -08:00
Masayuki Yamamoto 731e189014 bpo-25658: Implement PEP 539 for Thread Specific Storage (TSS) API (GH-1362)
See PEP 539 for details.

Highlights of changes:

- Add Thread Specific Storage (TSS) API
- Document the Thread Local Storage (TLS) API as deprecated
- Update code that used TLS API to use TSS API
2017-10-06 20:41:34 +10:00
Barry Warsaw b2e5794870 bpo-31338 (#3374)
* Add Py_UNREACHABLE() as an alias to abort().
* Use Py_UNREACHABLE() instead of assert(0)
* Convert more unreachable code to use Py_UNREACHABLE()
* Document Py_UNREACHABLE() and a few other macros.
2017-09-14 18:13:16 -07:00
Antoine Pitrou a6a4dc816d bpo-31370: Remove support for threads-less builds (#3385)
* Remove Setup.config
* Always define WITH_THREAD for compatibility.
2017-09-07 18:56:24 +02:00
Segev Finer 39243779f4 bpo-31018: Switch to #pragma pack from __declspec(align) (#2848) 2017-07-25 10:47:43 +02:00
Xiang Zhang 4ed5ad79ec bpo-30961: Fix decrementing a borrowed reference in tracemalloc. (#2747) 2017-07-19 00:33:29 +08:00
Victor Stinner 5ea4c06773 bpo-30054: Expose tracemalloc C API (#1236)
* Make PyTraceMalloc_Track() and PyTraceMalloc_Untrack() functions
  public (remove the "_" prefix)
* Remove the _PyTraceMalloc_domain_t type: use directly unsigned
  int.
* Document methods

Note: methods are already tested in test_tracemalloc.
2017-06-20 17:46:36 +02:00
Serhiy Storchaka 97353845f8 Issue #20186: Regenerated Argument Clinic. 2017-02-05 22:58:46 +02:00
Serhiy Storchaka b451f91786 Issue #20186: Converted the tracemalloc module to Argument Clinic.
Based on patch by Georg Brandl.
2017-02-04 12:18:38 +02:00
Benjamin Peterson 2f8bfef158 replace PY_SIZE_MAX with SIZE_MAX 2016-09-07 09:26:18 -07:00
Benjamin Peterson ca47063998 replace Py_(u)intptr_t with the c99 standard types 2016-09-06 13:47:26 -07:00
Raymond Hettinger 15f44ab043 Issue #27895: Spelling fixes (Contributed by Ville Skyttä). 2016-08-30 10:47:49 -07:00
doko@ubuntu.com bc7315068f - make some internal symbols static 2016-05-18 01:06:01 +02:00
Serhiy Storchaka b6a9c9761c Issue #26778: Fixed "a/an/and" typos in code comment, documentation and error
messages.
2016-04-17 09:39:28 +03:00
Serhiy Storchaka 6a7b3a77b4 Issue #26778: Fixed "a/an/and" typos in code comment and documentation. 2016-04-17 08:32:47 +03:00
Martin Panter cda80940ed Issue #15984: Merge PyUnicode doc from 3.5 2016-04-15 02:27:11 +00:00
Martin Panter 6245cb3c01 Correct “an” → “a” with “Unicode”, “user”, “UTF”, etc
This affects documentation, code comments, and a debugging messages.
2016-04-15 02:14:19 +00:00
Victor Stinner 5e14a38e8e _tracemalloc: use compact key for traces
Issue #26588: Optimize memory footprint of _tracemalloc before non-zero domain
is used. Start with compact key (Py_uintptr_t) and also switch to pointer_t key
when the first memory block with a non-zero domain is tracked.
2016-03-23 22:03:55 +01:00
Victor Stinner 5dacbd4c42 Cleanup hashtable.h
Issue #26588:

* Pass the hash table rather than the key size to hash and compare functions
* _Py_HASHTABLE_READ_KEY() and _Py_HASHTABLE_ENTRY_READ_KEY() macros now expect
  the hash table as the first parameter, rather than the key size
* tracemalloc_get_traces_fill(): use _Py_HASHTABLE_ENTRY_READ_DATA() rather
  than pointer dereference
* Remove the _Py_HASHTABLE_ENTRY_WRITE_PKEY() macro
* Move "PKEY" and "PDATA" macros inside hashtable.c
2016-03-23 09:52:13 +01:00
Victor Stinner ca79ccd9e6 Issue #26588:
* Optimize tracemalloc_add_trace(): modify hashtable entry data (trace) if the
  memory block is already tracked, rather than trying to remove the old trace
  and then add a new trace.
* Add _Py_HASHTABLE_ENTRY_WRITE_DATA() macro
2016-03-23 09:38:54 +01:00
Victor Stinner e8c6b2fd1b Issue #26588:
* _Py_HASHTABLE_ENTRY_DATA: change type from "char *" to "const void *"
* Add _Py_HASHTABLE_ENTRY_WRITE_PKEY() macro
* Rename _Py_HASHTABLE_ENTRY_WRITE_DATA() macro to
  _Py_HASHTABLE_ENTRY_WRITE_PDATA()
* Add _Py_HASHTABLE_ENTRY_WRITE_DATA() macro
2016-03-23 09:25:01 +01:00
Victor Stinner 42bcf37fcf Issue #26588: Optimize tracemalloc_realloc()
No need to remove the old trace if the memory block didn't move.
2016-03-23 09:08:08 +01:00
Victor Stinner 35b0fd74dd Merge 3.5 2016-03-23 00:17:54 +01:00
Victor Stinner d5871e62dd Enhance _tracemalloc debug mode
Issue #26588: Enhance assertion in set_reentrant()
2016-03-23 00:17:04 +01:00
Victor Stinner 2f49e09110 Fix _tracemalloc start/stop
Issue #26588: Fix _tracemalloc start/stop: don't play with the reentrant flag.

set_reentrant(1) fails with an assertion error if tracemalloc_init() is called
first in a thread A and tracemalloc_start() is called second in a thread B. The
tracemalloc is imported in a thread A. Importing the module calls
tracemalloc_init(). tracemalloc.start() is called in a thread B.
2016-03-23 00:10:24 +01:00
Victor Stinner f9a71153e9 Issue #26588: remove debug traces from _tracemalloc. 2016-03-22 23:54:42 +01:00
Victor Stinner eddb074c53 Issue #26588: more debug traces 2016-03-22 21:06:07 +01:00
Victor Stinner b8d927266d Issue #26588: Don't call tracemalloc_init() at module initilization
So it's possible to get debug messages in test_tracemalloc.
2016-03-22 20:56:49 +01:00
Victor Stinner e31e35d34e Issue #26588: one more assertion 2016-03-22 18:48:50 +01:00
Victor Stinner 4a06647534 Add assertions on tracemalloc_reentrant_key
Issue #26588.
2016-03-22 17:45:09 +01:00
Victor Stinner 0cfc058d61 Issue #26588: more assertions 2016-03-22 17:40:07 +01:00
Victor Stinner 84aab09421 Issue #26588: add debug traces
Try to debug random failure on buildbots.
2016-03-22 16:13:31 +01:00
Victor Stinner 10b73e1748 Add C functions _PyTraceMalloc_Track()
Issue #26530:

* Add C functions _PyTraceMalloc_Track() and _PyTraceMalloc_Untrack() to track
  memory blocks using the tracemalloc module.
* Add _PyTraceMalloc_GetTraceback() to get the traceback of an object.
2016-03-22 13:39:05 +01:00
Victor Stinner e492ae50e2 tracemalloc now supports domains
Issue #26588:

* The _tracemalloc now supports tracing memory allocations of multiple address
  spaces (domains).
* Add domain parameter to tracemalloc_add_trace() and
  tracemalloc_remove_trace().
* tracemalloc_add_trace() now starts by removing the previous trace, if any.
* _tracemalloc._get_traces() now returns a list of (domain, size,
  traceback_frames): the domain is new.
* Add tracemalloc.DomainFilter
* tracemalloc.Filter: add an optional domain parameter to the constructor and a
  domain attribute
* Sublte change: use Py_uintptr_t rather than void* in the traces key.
* Add tracemalloc_config.use_domain, currently hardcoded to 1
2016-03-22 12:58:23 +01:00
Victor Stinner c9553876ae Simplify implementation of hashtable.c
Issue #26588: Remove copy_data, free_data and get_data_size callbacks from
hashtable.h. These callbacks are not used in Python and makes the code more
complex.

Remove also the _Py_HASHTABLE_ENTRY_DATA_AS_VOID_P() macro which uses an unsafe
pointer dereference (can cause memory alignment issue). Replace the macro usage
with _Py_HASHTABLE_ENTRY_READ_DATA() which is implemented with the safe
memcpy() function.
2016-03-22 12:13:01 +01:00
Victor Stinner b32a7edb22 Issue #26588: Fix compilation warning on Windows 2016-03-21 23:05:08 +01:00
Victor Stinner 285cf0a601 hashtable.h now supports keys of any size
Issue #26588: hashtable.h now supports keys of any size, not only
sizeof(void*). It allows to support key larger than sizeof(void*), but also to
use less memory for key smaller than sizeof(void*).
2016-03-21 22:00:58 +01:00
Victor Stinner 322bc12c31 Ooops, revert changeset ea9efa06c137
Change pushed by mistake, the patch is still under review :-/

"""
_tracemalloc: add domain to trace keys

* hashtable.h: key has now a variable size
* _tracemalloc uses (pointer: void*, domain: unsigned int) as key for traces
"""
2016-03-21 14:36:39 +01:00
Victor Stinner 51b846c47a _tracemalloc: add domain to trace keys
* hashtable.h: key has now a variable size
* _tracemalloc uses (pointer: void*, domain: unsigned int) as key for traces
2016-03-18 21:52:22 +01:00
Victor Stinner 0611c26a58 On memory error, dump the memory block traceback
Issue #26564: _PyObject_DebugDumpAddress() now dumps the traceback where a
memory block was allocated on memory block. Use the tracemalloc module to get
the traceback.
2016-03-15 22:22:13 +01:00
Victor Stinner af584a02a5 Merge 3.5 (_tracemalloc) 2016-03-15 21:57:23 +01:00
Victor Stinner 9528334e16 _tracemalloc: store lineno as unsigned int
Issue #26564. Cleanup the code, lineno is never negative.
2016-03-15 21:57:02 +01:00
Victor Stinner 7105e9f3de _tracemalloc: filename cannot be NULL 2016-03-15 14:28:04 +01:00
Victor Stinner d222653f8f Issue #26558: Remove useless check in tracemalloc
The first instruction of tracemalloc_add_trace() is traceback_new() which
already checks the GIL.
2016-03-14 17:01:32 +01:00
Victor Stinner cb7e5f6f08 Merge 3.4 (tracemalloc typo) 2015-03-18 16:05:32 +01:00
Victor Stinner 444f124fcf _tracemalloc.c: Fix typo 2015-03-18 16:05:18 +01:00
Serhiy Storchaka 56a6d855e2 Removed duplicated words in in comments and docs. 2014-12-01 18:28:43 +02:00