Commit message (Collapse) | Author | Age | Files | Lines | |
---|---|---|---|---|---|
* | Disable _PyStack_AsTuple() inlining | Victor Stinner | 2017-01-11 | 2 | -2/+17 |
| | | | | | | | | | | | | | | | Issue #29234: Inlining _PyStack_AsTuple() into callers increases their stack consumption, Disable inlining to optimize the stack consumption. Add _Py_NO_INLINE: use __attribute__((noinline)) of GCC and Clang. It reduces the stack consumption, bytes per call, before => after: test_python_call: 1040 => 976 (-64 B) test_python_getitem: 976 => 912 (-64 B) test_python_iterator: 1120 => 1056 (-64 B) => total: 3136 => 2944 (- 192 B) | ||||
* | Inline call_function() | Victor Stinner | 2017-01-10 | 1 | -2/+4 |
| | | | | | | | | | | | | | Issue #29227: Inline call_function() into _PyEval_EvalFrameDefault() using Py_LOCAL_INLINE to reduce the stack consumption. It reduces the stack consumption, bytes per call, before => after: test_python_call: 1152 => 1040 (-112 B) test_python_getitem: 1008 => 976 (-32 B) test_python_iterator: 1232 => 1120 (-112 B) => total: 3392 => 3136 (- 256 B) | ||||
* | call_method() now uses _PyObject_FastCall() | Victor Stinner | 2017-01-10 | 3 | -52/+66 |
| | | | | | | | | | | | | | | | | | | | Issue #29233: Replace the inefficient _PyObject_VaCallFunctionObjArgs() with _PyObject_FastCall() in call_method() and call_maybe(). Only a few functions call call_method() and call it with a fixed number of arguments. Avoid the complex and expensive _PyObject_VaCallFunctionObjArgs() function, replace it with an array allocated on the stack with the exact number of argumlents. It reduces the stack consumption, bytes per call, before => after: test_python_call: 1168 => 1152 (-16 B) test_python_getitem: 1344 => 1008 (-336 B) test_python_iterator: 1568 => 1232 (-336 B) Remove the _PyObject_VaCallFunctionObjArgs() function which became useless. Rename it to object_vacall() and make it private. | ||||
* | Issue #29217: Merge 3.6. | Xiang Zhang | 2017-01-10 | 1 | -1/+1 |
|\ | |||||
| * | Issue #29217: Merge 3.5. | Xiang Zhang | 2017-01-10 | 1 | -1/+1 |
| |\ | |||||
| | * | Issue #29217: Fix the wrong type description of UUID.variant. | Xiang Zhang | 2017-01-10 | 1 | -1/+1 |
| | | | |||||
* | | | Issue #29145: Merge 3.6. | Xiang Zhang | 2017-01-10 | 2 | -7/+9 |
|\ \ \ | |/ / | |||||
| * | | Issue #29145: Merge 3.5. | Xiang Zhang | 2017-01-10 | 1 | -6/+8 |
| |\ \ | | |/ | |||||
| | * | Issue #29145: Fix overflow checks in str.replace() and str.join(). | Xiang Zhang | 2017-01-10 | 1 | -6/+8 |
| | | | | | | | | | | | | Based on patch by Martin Panter. | ||||
* | | | Closes #29177: Merged fix from 3.6. | Vinay Sajip | 2017-01-09 | 1 | -24/+62 |
|\ \ \ | |/ / | |||||
| * | | Fixes #29177: Improved resilience of logging tests which use socket servers. | Vinay Sajip | 2017-01-09 | 1 | -24/+62 |
| | | | | | | | | | | | | Thanks to Xavier de Gaye for the report and patch improvements. | ||||
* | | | Closes #29133: merged update from 3.6. | Vinay Sajip | 2017-01-09 | 1 | -15/+11 |
|\ \ \ | |/ / | | | | | | | Thanks to Evan_ for the report and Marco Buttu for the patch. | ||||
| * | | Fixes #29133: clarified shlex documentation. | Vinay Sajip | 2017-01-09 | 1 | -15/+11 |
| | | | |||||
* | | | Add test for ea064ff3c10f | Raymond Hettinger | 2017-01-09 | 1 | -0/+9 |
| | | | |||||
* | | | Backed out changeset ea064ff3c10f | Raymond Hettinger | 2017-01-09 | 1 | -6/+0 |
| | | | |||||
* | | | While a speedup of 1% is measurable, contexts aren't created that often, | Stefan Krah | 2017-01-09 | 1 | -4/+4 |
| | | | | | | | | | | | | so let's defer this until 3.7, 3.8, ... all have this new function. | ||||
* | | | Merge 3.6. | Stefan Krah | 2017-01-09 | 1 | -7/+7 |
|\ \ \ | |/ / | |||||
| * | | Merge 3.5. | Stefan Krah | 2017-01-09 | 1 | -7/+7 |
| |\ \ | | |/ | |||||
| | * | Issue #28701: Revert part of 5bdc8e1a50c8 for the following reasons: | Stefan Krah | 2017-01-09 | 1 | -7/+7 |
| | | | | | | | | | | | | | | | | | | - There was no real problem to begin with. - The hypothetical problem has been fixed by 5bdc8e1a50c8. | ||||
* | | | Merge 3.6 | Victor Stinner | 2017-01-09 | 0 | -0/+0 |
|\ \ \ | |/ / | |||||
| * | | Null merge 3.5 | Victor Stinner | 2017-01-09 | 0 | -0/+0 |
| |\ \ | | |/ | | | | | | | Python/random.c is more then in the 3.6 branch. | ||||
| | * | Issue #29157: Prefer getrandom() over getentropy() | Victor Stinner | 2017-01-09 | 1 | -200/+294 |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Copy and then adapt Python/random.c from default branch. Difference between 3.5 and default branches: * Python 3.5 only uses getrandom() in non-blocking mode: flags=GRND_NONBLOCK * If getrandom() fails with EAGAIN: py_getrandom() immediately fails and remembers that getrandom() doesn't work. * Python 3.5 has no _PyOS_URandomNonblock() function: _PyOS_URandom() works in non-blocking mode on Python 3.5 | ||||
* | | | Issue #29190: Fixed possible errors in comparing strings in the pickle module. | Serhiy Storchaka | 2017-01-09 | 1 | -12/+6 |
|\ \ \ | |/ / | |||||
| * | | Issue #29190: Fixed possible errors in comparing strings in the pickle module. | Serhiy Storchaka | 2017-01-09 | 1 | -12/+6 |
| |\ \ | | |/ | |||||
| | * | Issue #29190: Fixed possible errors in comparing strings in the pickle module. | Serhiy Storchaka | 2017-01-09 | 1 | -12/+6 |
| | | | |||||
* | | | Issue #29142: Merge 3.6. | Xiang Zhang | 2017-01-09 | 3 | -1/+7 |
|\ \ \ | |/ / | |||||
| * | | Issue #29142: Merge 3.5. | Xiang Zhang | 2017-01-09 | 3 | -1/+7 |
| |\ \ | | |/ | |||||
| | * | Issue #29142: Fix suffixes in no_proxy handling in urllib. | Xiang Zhang | 2017-01-09 | 3 | -1/+6 |
| | | | | | | | | | | | | | | | | | | In urllib.request, suffixes in no_proxy environment variable with leading dots could match related hostnames again (e.g. .b.c matches a.b.c). Patch by Milan Oberkirch. | ||||
* | | | Sync-up lru_cache() C code with space saving feature in the Python version. | Raymond Hettinger | 2017-01-09 | 1 | -0/+6 |
| | | | |||||
* | | | merge | Raymond Hettinger | 2017-01-09 | 0 | -0/+0 |
|\ \ \ | |/ / | |||||
| * | | Sync-up with 3.7 by backporting minor lru_cache code beautification | Raymond Hettinger | 2017-01-09 | 1 | -3/+4 |
| | | | |||||
* | | | Complete the merge for issue #29203 | Raymond Hettinger | 2017-01-09 | 1 | -31/+18 |
| | | | |||||
* | | | merge | Raymond Hettinger | 2017-01-09 | 2 | -4/+13 |
|\ \ \ | |/ / | |||||
| * | | Issue #29203: functools.lru_cache() now respects PEP 468 | Raymond Hettinger | 2017-01-09 | 4 | -35/+35 |
| | | | |||||
* | | | Issue #29034: Merge 3.6. | Xiang Zhang | 2017-01-08 | 2 | -53/+58 |
|\ \ \ | |/ / | |||||
| * | | Issue #29034: Fix memory leak and use-after-free in path_converter. | Xiang Zhang | 2017-01-08 | 2 | -53/+58 |
| | | | |||||
* | | | merge | Raymond Hettinger | 2017-01-08 | 1 | -0/+8 |
|\ \ \ | |/ / | |||||
| * | | Add OrderedDict test for PEP 468 (guaranteed ordered of kwargs) | Raymond Hettinger | 2017-01-08 | 1 | -0/+8 |
| | | | |||||
* | | | merge | Raymond Hettinger | 2017-01-08 | 1 | -4/+5 |
|\ \ \ | |/ / | |||||
| * | | Update OrderedDict docs to reflect acceptance of PEP 468 | Raymond Hettinger | 2017-01-08 | 1 | -4/+5 |
| | | | |||||
* | | | Issue #29200: Fix test to use self.assertEqual instead of py.test style tests | Raymond Hettinger | 2017-01-08 | 1 | -12/+12 |
| | | | |||||
* | | | Issue #29200: Add test for lru cache only calling __hash__ once | Raymond Hettinger | 2017-01-08 | 1 | -0/+36 |
| | | | |||||
* | | | Issue #28815: Merge test tweak from 3.6 | Martin Panter | 2017-01-08 | 1 | -7/+3 |
|\ \ \ | |/ / | |||||
| * | | Issue #28815: Merge test tweak from 3.5 | Martin Panter | 2017-01-08 | 1 | -7/+3 |
| |\ \ | | |/ | |||||
| | * | Issue #28815: Use new exception subclasses | Martin Panter | 2017-01-08 | 1 | -7/+3 |
| | | | |||||
* | | | Merge 3.6. | Stefan Krah | 2017-01-08 | 1 | -0/+1 |
|\ \ \ | |/ / | |||||
| * | | Add comment why the change in d83884b3a427 wasn't necessary. | Stefan Krah | 2017-01-08 | 1 | -0/+1 |
| | | | |||||
* | | | Merge 3.6. | Stefan Krah | 2017-01-08 | 1 | -1/+1 |
|\ \ \ | |/ / | |||||
| * | | Revert part of 3cb3e224b692 in code that I maintain. | Stefan Krah | 2017-01-08 | 1 | -1/+1 |
| | | | |||||
* | | | Revert part of dbf72357cb4a that is in a rarely used path and causes | Stefan Krah | 2017-01-07 | 1 | -1/+1 |
| | | | | | | | | | | | | maintenance issues (cost/benefit). |