Mailman 3 November 2016 - Python-checkins

cpython (3.6): Add TCP_CONGESTION and TCP_USER_TIMEOUT
by victor.stinner 29 Nov '16

29 Nov '16

https://hg.python.org/cpython/rev/6d69da76be6a changeset: 105390:6d69da76be6a branch: 3.6 parent: 105387:6b8f7d1e2ba4 user: Victor Stinner <victor.stinner(a)gmail.com> date: Tue Nov 29 16:55:04 2016 +0100 summary: Add TCP_CONGESTION and TCP_USER_TIMEOUT Issue #26273: Add new socket.TCP_CONGESTION (Linux 2.6.13) and socket.TCP_USER_TIMEOUT (Linux 2.6.37) constants. Patch written by Omar Sandoval. files: Misc/NEWS | 4 ++++ Modules/socketmodule.c | 6 ++++++ 2 files changed, 10 insertions(+), 0 deletions(-) diff --git a/Misc/NEWS b/Misc/NEWS --- a/Misc/NEWS +++ b/Misc/NEWS @@ -74,6 +74,10 @@ Library ------- +- Issue #26273: Add new :data:`socket.TCP_CONGESTION` (Linux 2.6.13) and + :data:`socket.TCP_USER_TIMEOUT` (Linux 2.6.37) constants. Patch written by + Omar Sandoval. + - Issue #28752: Restored the __reduce__() methods of datetime objects. - Issue #28727: Regular expression patterns, _sre.SRE_Pattern objects created diff --git a/Modules/socketmodule.c b/Modules/socketmodule.c --- a/Modules/socketmodule.c +++ b/Modules/socketmodule.c @@ -7512,6 +7512,12 @@ #ifdef TCP_FASTOPEN PyModule_AddIntMacro(m, TCP_FASTOPEN); #endif +#ifdef TCP_CONGESTION + PyModule_AddIntMacro(m, TCP_CONGESTION); +#endif +#ifdef TCP_USER_TIMEOUT + PyModule_AddIntMacro(m, TCP_USER_TIMEOUT); +#endif /* IPX options */ #ifdef IPX_TYPE -- Repository URL: https://hg.python.org/cpython

1 0

cpython: Add TCP_CONGESTION and TCP_USER_TIMEOUT
by victor.stinner 29 Nov '16

29 Nov '16

https://hg.python.org/cpython/rev/674fb9644eaa changeset: 105389:674fb9644eaa user: Victor Stinner <victor.stinner(a)gmail.com> date: Tue Nov 29 16:55:04 2016 +0100 summary: Add TCP_CONGESTION and TCP_USER_TIMEOUT Issue #26273: Add new socket.TCP_CONGESTION (Linux 2.6.13) and socket.TCP_USER_TIMEOUT (Linux 2.6.37) constants. Patch written by Omar Sandoval. files: Misc/NEWS | 4 ++++ Modules/socketmodule.c | 6 ++++++ 2 files changed, 10 insertions(+), 0 deletions(-) diff --git a/Misc/NEWS b/Misc/NEWS --- a/Misc/NEWS +++ b/Misc/NEWS @@ -160,6 +160,10 @@ Library ------- +- Issue #26273: Add new :data:`socket.TCP_CONGESTION` (Linux 2.6.13) and + :data:`socket.TCP_USER_TIMEOUT` (Linux 2.6.37) constants. Patch written by + Omar Sandoval. + - Issue #28752: Restored the __reduce__() methods of datetime objects. - Issue #28727: Regular expression patterns, _sre.SRE_Pattern objects created diff --git a/Modules/socketmodule.c b/Modules/socketmodule.c --- a/Modules/socketmodule.c +++ b/Modules/socketmodule.c @@ -7540,6 +7540,12 @@ #ifdef TCP_FASTOPEN PyModule_AddIntMacro(m, TCP_FASTOPEN); #endif +#ifdef TCP_CONGESTION + PyModule_AddIntMacro(m, TCP_CONGESTION); +#endif +#ifdef TCP_USER_TIMEOUT + PyModule_AddIntMacro(m, TCP_USER_TIMEOUT); +#endif /* IPX options */ #ifdef IPX_TYPE -- Repository URL: https://hg.python.org/cpython

1 0

GOOD Benchmark Results for Python Default 2016-11-29
by lp_benchmark_robot＠intel.com 29 Nov '16

29 Nov '16

Results for project Python default, build date 2016-11-29 03:02:18 +0000 commit: 9ded2433dc2c previous commit: 5c1bb72c0f5d revision date: 2016-11-28 18:56:37 +0000 environment: Haswell-EP cpu: Intel(R) Xeon(R) CPU E5-2699 v3 @ 2.30GHz 2x18 cores, stepping 2, LLC 45 MB mem: 128 GB os: CentOS 7.1 kernel: Linux 3.10.0-229.4.2.el7.x86_64 Baseline results were generated using release v3.4.3, with hash b4cbecbc0781 from 2015-02-25 12:15:33+00:00 ---------------------------------------------------------------------------------- benchmark relative change since change since current rev run std_dev* last run baseline with PGO ---------------------------------------------------------------------------------- :-) django_v2 0.20% -0.82% 5.16% 13.76% :-) pybench 0.26% 0.14% 5.61% 5.33% :-( regex_v8 3.77% -0.13% -3.12% 3.26% :-) nbody 0.10% 2.07% 1.46% 4.56% :-( json_dump_v2 0.29% 1.46% -2.82% 8.44% :-| normal_startup 0.54% -0.20% 0.81% 6.86% ---------------------------------------------------------------------------------- * Relative Standard Deviation (Standard Deviation/Average) If this is not displayed properly please visit our results page here: http://languagesperformance.intel.com/good-benchmark-results-for-python-def… Note: Benchmark results are measured in seconds. Subject Label Legend: Attributes are determined based on the performance evolution of the workloads compared to the previous measurement iteration. NEUTRAL: performance did not change by more than 1% for any workload GOOD: performance improved by more than 1% for at least one workload and there is no regression greater than 1% BAD: performance dropped by more than 1% for at least one workload and there is no improvement greater than 1% UGLY: performance improved by more than 1% for at least one workload and also dropped by more than 1% for at least one workload Our lab does a nightly source pull and build of the Python project and measures performance changes against the previous stable version and the previous nightly measurement. This is provided as a service to the community so that quality issues with current hardware can be identified quickly. Intel technologies' features and benefits depend on system configuration and may require enabled hardware, software or service activation. Performance varies depending on system configuration.

1 0

GOOD Benchmark Results for Python 2.7 2016-11-29
by lp_benchmark_robot＠intel.com 29 Nov '16

29 Nov '16

No new revisions. Here are the previous results: Results for project Python 2.7, build date 2016-11-29 07:56:58 +0000 commit: 59bd48afa1bc previous commit: 6eefdf0fade6 revision date: 2016-11-26 11:43:39 +0000 environment: Haswell-EP cpu: Intel(R) Xeon(R) CPU E5-2699 v3 @ 2.30GHz 2x18 cores, stepping 2, LLC 45 MB mem: 128 GB os: CentOS 7.1 kernel: Linux 3.10.0-229.4.2.el7.x86_64 Baseline results were generated using release v2.7.10, with hash 15c95b7d81dc from 2015-05-23 16:02:14+00:00 ---------------------------------------------------------------------------------- benchmark relative change since change since current rev run std_dev* last run baseline with PGO ---------------------------------------------------------------------------------- :-) django_v2 0.14% 1.76% 5.07% 8.38% :-) pybench 0.24% -0.01% 5.70% 4.55% :-( regex_v8 0.57% 0.06% -2.26% 10.27% :-) nbody 0.85% 0.01% 7.90% 2.67% :-) json_dump_v2 0.34% -0.26% 3.54% 10.10% :-| normal_startup 1.12% 0.06% -0.24% 2.36% :-) ssbench 0.20% -0.56% 2.06% 1.80% ---------------------------------------------------------------------------------- * Relative Standard Deviation (Standard Deviation/Average) If this is not displayed properly please visit our results page here: http://languagesperformance.intel.com/good-benchmark-results-for-python-2-7… Note: Benchmark results for ssbench are measured in requests/second while all other are measured in seconds. Subject Label Legend: Attributes are determined based on the performance evolution of the workloads compared to the previous measurement iteration. NEUTRAL: performance did not change by more than 1% for any workload GOOD: performance improved by more than 1% for at least one workload and there is no regression greater than 1% BAD: performance dropped by more than 1% for at least one workload and there is no improvement greater than 1% UGLY: performance improved by more than 1% for at least one workload and also dropped by more than 1% for at least one workload Our lab does a nightly source pull and build of the Python project and measures performance changes against the previous stable version and the previous nightly measurement. This is provided as a service to the community so that quality issues with current hardware can be identified quickly. Intel technologies' features and benefits depend on system configuration and may require enabled hardware, software or service activation. Performance varies depending on system configuration.

1 0

Daily reference leaks (9ded2433dc2c): sum=12
by solipsis＠pitrou.net 29 Nov '16

29 Nov '16

results for 9ded2433dc2c on branch "default" -------------------------------------------- test_collections leaked [7, -7, 8] memory blocks, sum=8 test_functools leaked [0, 3, 1] memory blocks, sum=4 Command line was: ['./python', '-m', 'test.regrtest', '-uall', '-R', '3:3:/home/psf-users/antoine/refleaks/reflogbAVG8X', '--timeout', '7200']

1 0

cpython (3.6): Issue #28797: Modifying the class __dict__ inside the __set_name__ method of
by serhiy.storchaka 29 Nov '16

29 Nov '16

https://hg.python.org/cpython/rev/6b8f7d1e2ba4 changeset: 105387:6b8f7d1e2ba4 branch: 3.6 parent: 105382:52038705827d user: Serhiy Storchaka <storchaka(a)gmail.com> date: Tue Nov 29 09:54:17 2016 +0200 summary: Issue #28797: Modifying the class __dict__ inside the __set_name__ method of a descriptor that is used inside that class no longer prevents calling the __set_name__ method of other descriptors. files: Lib/test/test_subclassinit.py | 16 ++++++++++++++++ Misc/NEWS | 4 ++++ Objects/typeobject.c | 18 +++++++++++++----- 3 files changed, 33 insertions(+), 5 deletions(-) diff --git a/Lib/test/test_subclassinit.py b/Lib/test/test_subclassinit.py --- a/Lib/test/test_subclassinit.py +++ b/Lib/test/test_subclassinit.py @@ -198,6 +198,22 @@ self.assertIs(B.meta_owner, B) self.assertEqual(B.name, 'd') + def test_set_name_modifying_dict(self): + notified = [] + class Descriptor: + def __set_name__(self, owner, name): + setattr(owner, name + 'x', None) + notified.append(name) + + class A: + a = Descriptor() + b = Descriptor() + c = Descriptor() + d = Descriptor() + e = Descriptor() + + self.assertCountEqual(notified, ['a', 'b', 'c', 'd', 'e']) + def test_errors(self): class MyMeta(type): pass diff --git a/Misc/NEWS b/Misc/NEWS --- a/Misc/NEWS +++ b/Misc/NEWS @@ -10,6 +10,10 @@ Core and Builtins ----------------- +- Issue #28797: Modifying the class __dict__ inside the __set_name__ method of + a descriptor that is used inside that class no longer prevents calling the + __set_name__ method of other descriptors. + - Issue #28782: Fix a bug in the implementation ``yield from`` when checking if the next instruction is YIELD_FROM. Regression introduced by WORDCODE (issue #26647). diff --git a/Objects/typeobject.c b/Objects/typeobject.c --- a/Objects/typeobject.c +++ b/Objects/typeobject.c @@ -7004,10 +7004,14 @@ static int set_names(PyTypeObject *type) { - PyObject *key, *value, *set_name, *tmp; + PyObject *names_to_set, *key, *value, *set_name, *tmp; Py_ssize_t i = 0; - while (PyDict_Next(type->tp_dict, &i, &key, &value)) { + names_to_set = PyDict_Copy(type->tp_dict); + if (names_to_set == NULL) + return -1; + + while (PyDict_Next(names_to_set, &i, &key, &value)) { set_name = lookup_maybe(value, &PyId___set_name__); if (set_name != NULL) { tmp = PyObject_CallFunctionObjArgs(set_name, type, key, NULL); @@ -7017,15 +7021,19 @@ "Error calling __set_name__ on '%.100s' instance %R " "in '%.100s'", value->ob_type->tp_name, key, type->tp_name); + Py_DECREF(names_to_set); return -1; } else Py_DECREF(tmp); } - else if (PyErr_Occurred()) + else if (PyErr_Occurred()) { + Py_DECREF(names_to_set); return -1; - } - + } + } + + Py_DECREF(names_to_set); return 0; } -- Repository URL: https://hg.python.org/cpython

1 0

cpython (merge 3.6 -> default): Issue #28797: Modifying the class __dict__ inside the __set_name__ method of
by serhiy.storchaka 29 Nov '16

29 Nov '16

https://hg.python.org/cpython/rev/18ed518d2eef changeset: 105388:18ed518d2eef parent: 105386:9ded2433dc2c parent: 105387:6b8f7d1e2ba4 user: Serhiy Storchaka <storchaka(a)gmail.com> date: Tue Nov 29 09:56:07 2016 +0200 summary: Issue #28797: Modifying the class __dict__ inside the __set_name__ method of a descriptor that is used inside that class no longer prevents calling the __set_name__ method of other descriptors. files: Lib/test/test_subclassinit.py | 16 ++++++++++++++++ Misc/NEWS | 4 ++++ Objects/typeobject.c | 18 +++++++++++++----- 3 files changed, 33 insertions(+), 5 deletions(-) diff --git a/Lib/test/test_subclassinit.py b/Lib/test/test_subclassinit.py --- a/Lib/test/test_subclassinit.py +++ b/Lib/test/test_subclassinit.py @@ -198,6 +198,22 @@ self.assertIs(B.meta_owner, B) self.assertEqual(B.name, 'd') + def test_set_name_modifying_dict(self): + notified = [] + class Descriptor: + def __set_name__(self, owner, name): + setattr(owner, name + 'x', None) + notified.append(name) + + class A: + a = Descriptor() + b = Descriptor() + c = Descriptor() + d = Descriptor() + e = Descriptor() + + self.assertCountEqual(notified, ['a', 'b', 'c', 'd', 'e']) + def test_errors(self): class MyMeta(type): pass diff --git a/Misc/NEWS b/Misc/NEWS --- a/Misc/NEWS +++ b/Misc/NEWS @@ -10,6 +10,10 @@ Core and Builtins ----------------- +- Issue #28797: Modifying the class __dict__ inside the __set_name__ method of + a descriptor that is used inside that class no longer prevents calling the + __set_name__ method of other descriptors. + - Issue #28799: Remove the ``PyEval_GetCallStats()`` function and deprecate the untested and undocumented ``sys.callstats()`` function. Remove the ``CALL_PROFILE`` special build: use the :func:`sys.setprofile` function, diff --git a/Objects/typeobject.c b/Objects/typeobject.c --- a/Objects/typeobject.c +++ b/Objects/typeobject.c @@ -7004,10 +7004,14 @@ static int set_names(PyTypeObject *type) { - PyObject *key, *value, *set_name, *tmp; + PyObject *names_to_set, *key, *value, *set_name, *tmp; Py_ssize_t i = 0; - while (PyDict_Next(type->tp_dict, &i, &key, &value)) { + names_to_set = PyDict_Copy(type->tp_dict); + if (names_to_set == NULL) + return -1; + + while (PyDict_Next(names_to_set, &i, &key, &value)) { set_name = lookup_maybe(value, &PyId___set_name__); if (set_name != NULL) { tmp = PyObject_CallFunctionObjArgs(set_name, type, key, NULL); @@ -7017,15 +7021,19 @@ "Error calling __set_name__ on '%.100s' instance %R " "in '%.100s'", value->ob_type->tp_name, key, type->tp_name); + Py_DECREF(names_to_set); return -1; } else Py_DECREF(tmp); } - else if (PyErr_Occurred()) + else if (PyErr_Occurred()) { + Py_DECREF(names_to_set); return -1; - } - + } + } + + Py_DECREF(names_to_set); return 0; } -- Repository URL: https://hg.python.org/cpython

1 0

cpython: Issue #28823: Simplified compiling with opcode BUILD_MAP_UNPACK.
by serhiy.storchaka 28 Nov '16

28 Nov '16

https://hg.python.org/cpython/rev/9ded2433dc2c changeset: 105386:9ded2433dc2c user: Serhiy Storchaka <storchaka(a)gmail.com> date: Mon Nov 28 20:56:37 2016 +0200 summary: Issue #28823: Simplified compiling with opcode BUILD_MAP_UNPACK. files: Python/compile.c | 7 ++----- 1 files changed, 2 insertions(+), 5 deletions(-) diff --git a/Python/compile.c b/Python/compile.c --- a/Python/compile.c +++ b/Python/compile.c @@ -3344,11 +3344,8 @@ /* If there is more than one dict, they need to be merged into a new * dict. If there is one dict and it's an unpacking, then it needs * to be copied into a new dict." */ - while (containers > 1 || is_unpacking) { - int oparg = containers < 255 ? containers : 255; - ADDOP_I(c, BUILD_MAP_UNPACK, oparg); - containers -= (oparg - 1); - is_unpacking = 0; + if (containers > 1 || is_unpacking) { + ADDOP_I(c, BUILD_MAP_UNPACK, containers); } return 1; } -- Repository URL: https://hg.python.org/cpython

1 0

cpython: call_function(): document PyMethod optimization
by victor.stinner 28 Nov '16

28 Nov '16

https://hg.python.org/cpython/rev/c085f2e0e1af changeset: 105385:c085f2e0e1af user: Victor Stinner <victor.stinner(a)gmail.com> date: Mon Nov 28 18:32:31 2016 +0100 summary: call_function(): document PyMethod optimization files: Python/ceval.c | 6 +++++- 1 files changed, 5 insertions(+), 1 deletions(-) diff --git a/Python/ceval.c b/Python/ceval.c --- a/Python/ceval.c +++ b/Python/ceval.c @@ -4736,7 +4736,11 @@ } else { if (PyMethod_Check(func) && PyMethod_GET_SELF(func) != NULL) { - /* optimize access to bound methods */ + /* Optimize access to bound methods. Reuse the Python stack + to pass 'self' as the first argument, replace 'func' + with 'self'. It avoids the creation of a new temporary tuple + for arguments (to replace func with self) when the method uses + FASTCALL. */ PyObject *self = PyMethod_GET_SELF(func); Py_INCREF(self); func = PyMethod_GET_FUNCTION(func); -- Repository URL: https://hg.python.org/cpython

1 0

cpython: Reintroduce Python2 support in generate_opcode_h.py
by victor.stinner 28 Nov '16

28 Nov '16

https://hg.python.org/cpython/rev/a6ea34c8b413 changeset: 105384:a6ea34c8b413 user: Victor Stinner <victor.stinner(a)gmail.com> date: Mon Nov 28 18:13:52 2016 +0100 summary: Reintroduce Python2 support in generate_opcode_h.py Issue #28821. Add also a message to show that the command did something :-) files: Tools/scripts/generate_opcode_h.py | 12 +++++++++--- 1 files changed, 9 insertions(+), 3 deletions(-) diff --git a/Tools/scripts/generate_opcode_h.py b/Tools/scripts/generate_opcode_h.py --- a/Tools/scripts/generate_opcode_h.py +++ b/Tools/scripts/generate_opcode_h.py @@ -1,6 +1,8 @@ # This script generates the opcode.h header file. import sys +import tokenize + header = """/* Auto-generated by Tools/scripts/generate_opcode_h.py */ #ifndef Py_OPCODE_H #define Py_OPCODE_H @@ -32,12 +34,14 @@ #endif /* !Py_OPCODE_H */ """ -import tokenize - def main(opcode_py, outfile='Include/opcode.h'): opcode = {} - with tokenize.open(opcode_py) as fp: + if hasattr(tokenize, 'open'): + fp = tokenize.open(opcode_py) # Python 3.2+ + else: + fp = open(opcode_py) # Python 2.7 + with fp: code = fp.read() exec(code, opcode) opmap = opcode['opmap'] @@ -51,6 +55,8 @@ ('HAVE_ARGUMENT', opcode['HAVE_ARGUMENT'])) fobj.write(footer) + print("%s regenerated from %s" % (outfile, opcode_py)) + if __name__ == '__main__': main(sys.argv[1], sys.argv[2]) -- Repository URL: https://hg.python.org/cpython

1 0