Commit graph

322 commits

Author SHA1 Message Date
Sergey Fedoseev 86b89916d1 Fix upsizing of marks stack in pickle module. (GH-8860)
Previously marks stack was upsized even there was space for additional item.
2018-08-25 10:54:40 +03:00
Alexey Izbyshev f8c06b0280 bpo-34456: pickle: Add missing NULL check to save_global(). (GH-8851)
Reported by Svace static analyzer.
2018-08-22 07:51:25 +03:00
Sergey Fedoseev 67b9cc8e60 bpo-34395: Fix memory leaks caused by incautious usage of PyMem_Resize(). (GH-8756) 2018-08-16 07:27:50 +03:00
Serhiy Storchaka 5d4cb54800
bpo-34141: Optimized pickling simple non-recursive values. (GH-8318) 2018-07-18 10:10:49 +03:00
Łukasz Langa c51d8c9ba6
bpo-23403: Bump pickle.DEFAULT_PROTOCOL to 4 (#6355)
This makes performance better and produces shorter pickles. This change is backwards compatible up to the oldest currently supported version of Python (3.4).
2018-04-03 23:06:53 -07:00
Serhiy Storchaka c869529ea9 bpo-33209: End framing at the end of C implementation of pickle.Pickler.dump(). (GH-6363) 2018-04-03 14:11:27 -07:00
Leo Arias c3d9508ff2 bpo-32746: Fix multiple typos (GH-5144)
Fix typos found by codespell in docs, docstrings, and comments.
2018-02-03 19:36:10 -05:00
Serhiy Storchaka f320be77ff bpo-32571: Avoid raising unneeded AttributeError and silencing it in C code (GH-5222)
Add two new private APIs: _PyObject_LookupAttr() and _PyObject_LookupAttrId()
2018-01-25 17:49:40 +09:00
Serhiy Storchaka 1211c9a989
bpo-32503: Avoid creating too small frames in pickles. (#5127) 2018-01-20 16:42:44 +02:00
Serhiy Storchaka 5b76bdba07
bpo-31993: Do not use memoryview when pickle large strings. (#5154)
PyMemoryView_FromMemory() created a memoryview referring to
the internal data of the string.  When the string is destroyed
the memoryview become referring to a freed memory.
2018-01-13 00:28:31 +02:00
Serhiy Storchaka 0a2da50e18
bpo-31993: Do not create frames for large bytes and str objects (#5114)
when serialize into memory buffer with C pickle implementations.

This optimization already is performed when serialize into memory
with Python pickle implementations or into a file with both
implementations.
2018-01-11 13:03:20 +02:00
Olivier Grisel 3cd7c6e6eb bpo-31993: Do not allocate large temporary buffers in pickle dump. (#4353)
The picklers do no longer allocate temporary memory when dumping large
bytes and str objects into a file object. Instead the data is
directly streamed into the underlying file object.

Previously the C implementation would buffer all content and issue a
single call to file.write() at the end of the dump. With protocol 4
this behavior has changed to issue one call to file.write() per frame.

The Python pickler with protocol 4 now dumps each frame content as a
memoryview to an IOBytes instance that is never reused and the
memoryview is no longer released after the call to write. This makes it
possible for the file object to delay access to the memoryview of
previous frames without forcing any additional memory copy as was
already possible with the C pickler.
2018-01-06 17:18:54 +02:00
Serhiy Storchaka 986375ebde
bpo-28416: Break reference cycles in Pickler and Unpickler subclasses (#4080)
with the persistent_id() and persistent_load() methods.
2017-11-30 22:48:31 +02:00
Serhiy Storchaka 3daaafb700
bpo-32037: Use the INT opcode for 32-bit integers in protocol 0 pickles. (#4407) 2017-11-16 09:44:43 +02:00
Mat M f76231f89a bpo-32013: _pickle: Add missing Py_DECREF in error case in fast_save_enter() (#4384) 2017-11-13 09:50:16 +02:00
Serhiy Storchaka 04e36af9b8 bpo-31572: Get rid of using _PyObject_HasAttrId() in pickle. (#3729) 2017-10-22 21:31:34 +03:00
Eric Snow 3f9eee6eb4 bpo-28411: Support other mappings in PyInterpreterState.modules. (#3593)
The concrete PyDict_* API is used to interact with PyInterpreterState.modules in a number of places. This isn't compatible with all dict subclasses, nor with other Mapping implementations. This patch switches the concrete API usage to the corresponding abstract API calls.

We also add a PyImport_GetModule() function (and some other helpers) to reduce a bunch of code duplication.
2017-09-15 16:35:20 -06:00
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
Eric Snow 93c92f7d1d bpo-31404: Revert "remove modules from Py_InterpreterState (#1638)" (#3565)
PR #1638, for bpo-28411, causes problems in some (very) edge cases. Until that gets sorted out, we're reverting the merge. PR #3506, a fix on top of #1638, is also getting reverted.
2017-09-13 23:46:04 -07:00
Eric Snow fc1bf872e9 bpo-30860: Move windows.h include out of internal/*.h. (#3458)
PR #3397 introduced a large number of warnings to the Windows build. This patch fixes them.
2017-09-11 18:30:43 -07:00
Eric Snow 2ebc5ce42a bpo-30860: Consolidate stateful runtime globals. (#3397)
* group the (stateful) runtime globals into various topical structs
* consolidate the topical structs under a single top-level _PyRuntimeState struct
* add a check-c-globals.py script that helps identify runtime globals

Other globals are excluded (see globals.txt and check-c-globals.py).
2017-09-07 23:51:28 -06:00
Victor Stinner af46eb8d5f _pickle: Fix whichmodule() (#3358)
_PyUnicode_FromId() can return NULL: replace Py_INCREF() with
Py_XINCREF().

Fix coverity report: CID 1417269.
2017-09-05 14:30:16 -07:00
Eric Snow 86b7afdfee bpo-28411: Remove "modules" field from Py_InterpreterState. (#1638)
sys.modules is the one true source.
2017-09-04 17:54:09 -06:00
Serhiy Storchaka fff9a31a91 bpo-29865: Use PyXXX_GET_SIZE macros rather than Py_SIZE for concrete types. (#748) 2017-03-21 08:53:25 +02:00
Victor Stinner fd6d0d2a18 Issue #29368: Fix _Pickle_FastCall() usage in do_append()
_Pickle_FastCall() has a surprising API: it decrements the reference counter of
its second argument.
2017-02-02 10:56:47 +01:00
Serhiy Storchaka bee09aecc2 Issue #29368: The extend() method is now called instead of the append()
method when unpickle collections.deque and other list-like objects.
This can speed up unpickling to 2 times.
2017-02-02 11:12:47 +02:00
Serhiy Storchaka 32999ab48a Issue #29190: Fixed possible errors in comparing strings in the pickle module. 2017-01-09 10:10:07 +02:00
Serhiy Storchaka f0f35a6720 Issue #29190: Fixed possible errors in comparing strings in the pickle module. 2017-01-09 10:09:43 +02:00
Serhiy Storchaka 9937d90ee8 Issue #29190: Fixed possible errors in comparing strings in the pickle module. 2017-01-09 10:04:34 +02:00
Serhiy Storchaka 5ab81d787f Issue #28959: Added private macro PyDict_GET_SIZE for retrieving the size of dict. 2016-12-16 16:18:57 +02:00
Victor Stinner b110dad9ab Initialize variables to fix compiler warnings
Warnings seen on the "AMD64 Debian PGO 3.x" buildbot. Warnings are false
positive, but variable initialization should not harm performances.
2016-12-09 17:06:43 +01:00
Victor Stinner 55ba38a480 Use _PyObject_CallMethodIdObjArgs()
Issue #28915: Replace _PyObject_CallMethodId() with
_PyObject_CallMethodIdObjArgs() in various modules when the format string was
only made of "O" formats, PyObject* arguments.

_PyObject_CallMethodIdObjArgs() avoids the creation of a temporary tuple and
doesn't have to parse a format string.
2016-12-09 16:09:30 +01:00
Victor Stinner 7bfb42d5b7 Issue #28858: Remove _PyObject_CallArg1() macro
Replace
   _PyObject_CallArg1(func, arg)
with
   PyObject_CallFunctionObjArgs(func, arg, NULL)

Using the _PyObject_CallArg1() macro increases the usage of the C stack, which
was unexpected and unwanted. PyObject_CallFunctionObjArgs() doesn't have this
issue.
2016-12-05 17:04:32 +01:00
Serhiy Storchaka 85b0f5beb1 Added the const qualifier to char* variables that refer to readonly internal
UTF-8 represenatation of Unicode objects.
2016-11-20 10:16:47 +02:00
Serhiy Storchaka 06515833fe Replaced outdated macros _PyUnicode_AsString and _PyUnicode_AsStringAndSize
with PyUnicode_AsUTF8 and PyUnicode_AsUTF8AndSize.
2016-11-20 09:13:07 +02:00
Serhiy Storchaka 3b73ea1278 Issue #28701: Replace PyUnicode_CompareWithASCIIString with _PyUnicode_EqualToASCIIString.
The latter function is more readable, faster and doesn't raise exceptions.
2016-11-16 10:19:20 +02:00
Serhiy Storchaka f4934ea77d Issue #28701: Replace PyUnicode_CompareWithASCIIString with _PyUnicode_EqualToASCIIString.
The latter function is more readable, faster and doesn't raise exceptions.
2016-11-16 10:17:58 +02:00
Serhiy Storchaka 90493ab30c Issue #25761: Improved error reporting about truncated pickle data in
C implementation of unpickler.  UnpicklingError is now raised instead of
AttributeError and ValueError in some cases.
2016-09-06 23:55:11 +03:00
Victor Stinner ad8c83ad6b Avoid inefficient way to call functions without argument
Don't pass "()" format to PyObject_CallXXX() to call a function without
argument: pass NULL as the format string instead. It avoids to have to parse a
string to produce 0 argument.
2016-09-05 17:53:15 -07:00
Raymond Hettinger 15f44ab043 Issue #27895: Spelling fixes (Contributed by Ville Skyttä). 2016-08-30 10:47:49 -07:00
Victor Stinner b98b927e72 _pickle: remove outdated comment
_Pickle_FastCall() is now fast again!

The optimization was introduced in Python 3.2, removed in Python 3.4 and
reintroduced in Python 3.6 (thanks to the new generic fastcall functions).
2016-08-25 00:58:58 +02:00
Victor Stinner 559bb6a713 Rename _PyObject_FastCall() to _PyObject_FastCallDict()
Issue #27809:

* Rename _PyObject_FastCall() function to _PyObject_FastCallDict()
* Add _PyObject_FastCall(), _PyObject_CallNoArg() and _PyObject_CallArg1()
  macros calling _PyObject_FastCallDict()
2016-08-22 22:48:54 +02:00
Victor Stinner 75210697ec Issue #27128: _pickle uses fast call
Use _PyObject_FastCall() to avoid the creation of temporary tuple.
2016-08-19 18:59:15 +02:00
Serhiy Storchaka 3410c01d83 Issue #17711: Fixed unpickling by the persistent ID with protocol 0.
Original patch by Alexandre Vassalotti.
2016-07-17 11:35:35 +03:00
Serhiy Storchaka dec25afab1 Issue #17711: Fixed unpickling by the persistent ID with protocol 0.
Original patch by Alexandre Vassalotti.
2016-07-17 11:24:17 +03:00
Serhiy Storchaka 2954f83999 - Issue #27332: Fixed the type of the first argument of module-level functions
generated by Argument Clinic.  Patch by Petr Viktorin.
2016-07-07 18:20:03 +03:00
Serhiy Storchaka 1a2b24f02d Issue #27332: Fixed the type of the first argument of module-level functions
generated by Argument Clinic.  Patch by Petr Viktorin.
2016-07-07 17:35:15 +03:00
Martin Panter e514093a2f Issue #27125: Merge typo fixes from 3.5 2016-05-30 05:24:49 +00:00
Martin Panter a90a4a9651 Issue #27125: Remove duplicated words from documentation and comments 2016-05-30 04:04:50 +00:00
Martin Panter 3e04d5b306 Issue #27076: Merge spelling from 3.5 2016-05-26 06:03:19 +00:00
Martin Panter 46f50726a0 Issue #27076: Doc, comment and tests spelling fixes
Most fixes to Doc/ and Lib/ directories by Ville Skyttä.
2016-05-26 05:35:26 +00:00
Victor Stinner da23056a3e Issue #27056: Fix _Unpickler_Read() to avoid integer overflow 2016-05-20 21:16:59 +02:00
Victor Stinner 19ed27ec2b Optimize pickle.load() and pickle.loads()
Issue #27056: Optimize pickle.load() and pickle.loads(), up to 10% faster to
deserialize a lot of small objects.
2016-05-20 11:42:37 +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
Serhiy Storchaka ec39756960 Issue #22570: Renamed Py_SETREF to Py_XSETREF. 2016-04-06 09:50:03 +03:00
Serhiy Storchaka 48842714b9 Issue #22570: Renamed Py_SETREF to Py_XSETREF. 2016-04-06 09:45:48 +03:00
Victor Stinner 21b47117ac _pickle: Fix load_counted_tuple(), use Py_ssize_t for size
Fix a warning on Windows 64-bit.
2016-03-14 18:09:39 +01:00
Serhiy Storchaka 576f132b98 Issue #20440: Cleaning up the code by using Py_SETREF. 2016-01-05 21:27:54 +02:00
Serhiy Storchaka bdb908ea54 Issue #20440: Applied yet one patch for using Py_SETREF.
The patch is automatically generated, it replaces the code that uses Py_CLEAR.
2015-12-27 12:38:28 +02:00
Serhiy Storchaka 4a1e70fc31 Issue #20440: Applied yet one patch for using Py_SETREF.
The patch is automatically generated, it replaces the code that uses Py_CLEAR.
2015-12-27 12:36:18 +02:00
Serhiy Storchaka ef1585eb9a Issue #25923: Added more const qualifiers to signatures of static and private functions. 2015-12-25 20:01:53 +02:00
Serhiy Storchaka a9406e77fa Issue #25421: __sizeof__ methods of builtin types now use dynamic basic size.
This allows sys.getsize() to work correctly with their subclasses with
__slots__ defined.
2015-12-19 20:07:11 +02:00
Serhiy Storchaka 5c4064e8bd Issue #25421: __sizeof__ methods of builtin types now use dynamic basic size.
This allows sys.getsize() to work correctly with their subclasses with
__slots__ defined.
2015-12-19 20:05:25 +02:00
Serhiy Storchaka 3ac5380d24 Issue #25761: Fixed reference leak added in previous changeset (5c670af0100f). 2015-12-07 11:32:00 +02:00
Serhiy Storchaka 59fb6342a4 Issue #25761: Improved detecting errors in broken pickle data. 2015-12-06 22:01:35 +02:00
Serhiy Storchaka fbc3e374a6 Fixed reference leak when read truncated pickle. 2015-12-01 00:20:49 +02:00
Serhiy Storchaka 77a61d292a Fixed reference leak when read truncated pickle. 2015-12-01 00:20:36 +02:00
Serhiy Storchaka ca28eba3d3 Fixed reference leak when read truncated pickle. 2015-12-01 00:18:23 +02:00
Serhiy Storchaka 14128d8bc5 Issue #25725: Fixed a reference leak in pickle.loads() when unpickling
invalid data including tuple instructions.
2015-11-25 15:07:36 +02:00
Serhiy Storchaka c5f3b4285a Issue #25725: Fixed a reference leak in pickle.loads() when unpickling
invalid data including tuple instructions.
2015-11-25 15:06:49 +02:00
Serhiy Storchaka a49de6be36 Issue #25725: Fixed a reference leak in pickle.loads() when unpickling
invalid data including tuple instructions.
2015-11-25 15:01:53 +02:00
Serhiy Storchaka b1ce30281d Issue #23914: Fixed SystemError raised by unpickler on broken pickle data. 2015-11-23 15:20:21 +02:00
Serhiy Storchaka 4409c6cfae Issue #23914: Fixed SystemError raised by unpickler on broken pickle data. 2015-11-23 15:19:23 +02:00
Serhiy Storchaka e9b3074cf9 Issue #23914: Fixed SystemError raised by unpickler on broken pickle data. 2015-11-23 15:17:43 +02:00
Victor Stinner 9ba97df69c Closes #25645: Fix a reference leak introduced by change bc5894a3a0e6 of the
issue #24164.
2015-11-17 12:15:07 +01:00
Serhiy Storchaka fad85aadb0 Issue #25558: Use compile-time asserts. 2015-11-07 15:42:38 +02:00
Martin Panter e56a919100 Issue #25523: Merge a-to-an corrections from 3.5 2015-11-02 04:27:17 +00:00
Martin Panter 2eb819f7a8 Issue #25523: Merge "a" to "an" fixes from 3.4 into 3.5 2015-11-02 04:04:57 +00:00
Martin Panter 7462b64911 Issue #25523: Correct "a" article to "an" article
This changes the main documentation, doc strings, source code comments, and a
couple error messages in the test suite. In some cases the word was removed
or edited some other way to fix the grammar.
2015-11-02 03:37:02 +00:00
Victor Stinner 358af13526 Issue #25353: Optimize unicode escape and raw unicode escape encoders to use
the new _PyBytesWriter API.
2015-10-12 22:36:57 +02:00
Serhiy Storchaka 0d554d7ef1 Issue #24164: Objects that need calling `__new__` with keyword arguments,
can now be pickled using pickle protocols older than protocol version 4.
2015-10-10 22:42:18 +03:00
Serhiy Storchaka 525faaeffc Issue #25262. Added support for BINBYTES8 opcode in Python implementation of
unpickler.  Highest 32 bits of 64-bit size for BINUNICODE8 and BINBYTES8
opcodes no longer silently ignored on 32-bit platforms in C implementation.
2015-09-29 22:12:29 +03:00
Serhiy Storchaka e060619d4b Issue #25262. Added support for BINBYTES8 opcode in Python implementation of
unpickler.  Highest 32 bits of 64-bit size for BINUNICODE8 and BINBYTES8
opcodes no longer silently ignored on 32-bit platforms in C implementation.
2015-09-29 22:10:07 +03:00
Benjamin Peterson bdb93c3e63 merge 3.4 (#25245) 2015-09-27 01:16:12 -07:00
Benjamin Peterson 6aa1564e9c initialize return value to NULL to avoid compiler compliants (closes #25245) 2015-09-27 01:16:03 -07:00
Benjamin Peterson 03c59b9bef merge 3.4 2015-09-26 00:09:32 -07:00
Benjamin Peterson e48cf7e729 prevent overflow in _Unpickler_Read 2015-09-26 00:08:34 -07:00
Yury Selivanov f488fb422a Issue #19235: Add new RecursionError exception. Patch by Georg Brandl. 2015-07-03 01:04:23 -04:00
Benjamin Peterson 42e913aa3a merge 3.4 (#24552) 2015-07-02 16:18:58 -05:00
Benjamin Peterson 80f78a3efc fix use after free (closes #24552) 2015-07-02 16:18:38 -05:00
Benjamin Peterson 7763c68dcd merge 3.4 2015-06-27 14:18:23 -05:00
Benjamin Peterson ff0f322edb merge 3.3 2015-06-27 13:56:46 -05:00
Benjamin Peterson 59b08c18a8 use safe allocation and reallocation macros 2015-06-27 13:41:33 -05:00
Larry Hastings 89964c48d1 Issue #23944: Argument Clinic now wraps long impl prototypes at column 78. 2015-04-14 18:07:59 -04:00
Serhiy Storchaka 1009bf18b3 Issue #23501: Argumen Clinic now generates code into separate files by default. 2015-04-03 23:53:51 +03:00
Serhiy Storchaka 58e4134a1c Issue #23611: Serializing more "lookupable" objects (such as unbound methods
or nested classes) now are supported with pickle protocols < 4.
2015-03-31 14:07:24 +03:00
Serhiy Storchaka d5d818d40b Issue #18473: Fixed 2to3 and 3to2 compatible pickle mappings.
Fixed ambigious reverse mappings.  Added many new mappings.  Import mapping
is no longer applied to modules already mapped with full name mapping.

Added tests for compatible pickling and unpickling and for consistency of
_compat_pickle mappings.
2015-03-31 13:17:10 +03:00