Commit Graph

287 Commits

Author SHA1 Message Date
Mark Dickinson 0672a6c23b
Revert "gh-89381: Fix invalid signatures of math/cmath.log (#101404)" (#101580)
This reverts commit 0ef92d9793.
2023-02-05 16:36:33 +00:00
Raymond Hettinger 5a2b984568
GH-100485: Create an alternative code path when an accurate fma() implementation is not available (#101567) 2023-02-04 17:54:44 -06:00
Sergey B Kirpichev 0ef92d9793
gh-89381: Fix invalid signatures of math/cmath.log (#101404) 2023-01-29 11:50:10 -08:00
Raymond Hettinger 84483aacc0
GH-100485: Add extended accuracy test. Switch to faster fma() based variant. GH-101383) 2023-01-28 06:29:21 -06:00
Raymond Hettinger 7956e0c300
Speed-up and improve accuracy with Rump Algorithms (3.1) and (5.10) (GH-101366) 2023-01-27 01:56:19 -06:00
Raymond Hettinger 997073c28b
Sumprod(): Update citation. Reorder functions. Add final twosum() call. Improve comments. (#101249) 2023-01-22 17:07:52 -06:00
Nikita Sobolev 36f2329367
gh-100873: Fix "‘lo’ may be used uninitialized in this function" warning in `mathmodule.c` (#100881) 2023-01-09 21:21:24 +05:30
Mark Dickinson 87d3bd0e02
gh-100833: Remove 'volatile' qualifiers in fsum algorithm (#100845)
This PR removes the `volatile` qualifier on various intermediate quantities
in the `math.fsum` implementation, and updates the notes preceding the
algorithm accordingly (as well as fixing some of the exsting notes). See
the linked issue #100833 for discussion.
2023-01-08 19:40:15 +00:00
Raymond Hettinger b139bcd892
GH-100485: Tweaks to sumprod() (GH-100857) 2023-01-08 13:38:24 -06:00
Raymond Hettinger df3851fe4a
GH-100485: Convert from Fast2Sum to 2Sum (GH-100836) 2023-01-07 21:37:08 -06:00
Raymond Hettinger 47b9f83a83
GH-100485: Add math.sumprod() (GH-100677) 2023-01-07 12:46:35 -06:00
Victor Stinner 81f7359f67
gh-99537: Use Py_SETREF(var, NULL) in C code (#99687)
Replace "Py_DECREF(var); var = NULL;" with "Py_SETREF(var, NULL);".
2022-11-23 14:57:50 +01:00
Victor Stinner 7e3f09cad9
gh-99537: Use Py_SETREF() function in C code (#99656)
Fix potential race condition in code patterns:

* Replace "Py_DECREF(var); var = new;" with "Py_SETREF(var, new);"
* Replace "Py_XDECREF(var); var = new;" with "Py_XSETREF(var, new);"
* Replace "Py_CLEAR(var); var = new;" with "Py_XSETREF(var, new);"

Other changes:

* Replace "old = var; var = new; Py_DECREF(var)"
  with "Py_SETREF(var, new);"
* Replace "old = var; var = new; Py_XDECREF(var)"
  with "Py_XSETREF(var, new);"
* And remove the "old" variable.
2022-11-22 14:22:22 +01:00
Victor Stinner 3e2f7135e6
gh-99300: Use Py_NewRef() in Modules/ directory (#99469)
Replace Py_INCREF() and Py_XINCREF() with Py_NewRef() and
Py_XNewRef() in test C files of the Modules/ directory.
2022-11-14 16:21:23 +01:00
Kumar Aditya ab57505070
GH-98897: fix memory leak if `math.dist` raises exception (GH-98898) 2022-10-31 21:18:32 -05:00
Dong-hee Na 23c9febdc6
Remove usage of _Py_IDENTIFIER from math module (#93739) 2022-06-12 18:45:02 +09:00
Pieter Eendebak 5a80e8580e
remove redundant argument to log_helper (GH-93440) 2022-06-03 08:40:05 +01:00
Victor Stinner 804f2529d8
gh-91320: Use _PyCFunction_CAST() (#92251)
Replace "(PyCFunction)(void(*)(void))func" cast with
_PyCFunction_CAST(func).

Change generated by the command:

sed -i -e \
  's!(PyCFunction)(void(\*)(void)) *\([A-Za-z0-9_]\+\)!_PyCFunction_CAST(\1)!g' \
  $(find -name "*.c")
2022-05-03 21:42:14 +02:00
Raymond Hettinger 1ba82d4419
Change parameter name from *x* for reals to *n* for integers. (GH-32377) 2022-04-06 14:35:05 -05:00
Victor Stinner 1b2611eb02
bpo-46656: Remove Py_NO_NAN macro (GH-31160)
Building Python now requires support for floating point Not-a-Number
(NaN): remove the Py_NO_NAN macro.
2022-02-25 01:32:57 +01:00
Victor Stinner 9bbdde2180
bpo-45412: Add _PY_SHORT_FLOAT_REPR macro (GH-31171)
Remove the HAVE_PY_SET_53BIT_PRECISION macro (moved to the internal
C API).

* Move HAVE_PY_SET_53BIT_PRECISION macro to pycore_pymath.h.
* Replace PY_NO_SHORT_FLOAT_REPR macro with _PY_SHORT_FLOAT_REPR
  macro which is always defined. gcc -Wundef emits a warning when
  using _PY_SHORT_FLOAT_REPR but the macro is not defined, if
  pycore_pymath.h include was forgotten.
2022-02-23 18:16:23 +01:00
Eric Snow 81c72044a1
bpo-46541: Replace core use of _Py_IDENTIFIER() with statically initialized global objects. (gh-30928)
We're no longer using _Py_IDENTIFIER() (or _Py_static_string()) in any core CPython code.  It is still used in a number of non-builtin stdlib modules.

The replacement is: PyUnicodeObject (not pointer) fields under _PyRuntimeState, statically initialized as part of _PyRuntime.  A new _Py_GET_GLOBAL_IDENTIFIER() macro facilitates lookup of the fields (along with _Py_GET_GLOBAL_STRING() for non-identifier strings).

https://bugs.python.org/issue46541#msg411799 explains the rationale for this change.

The core of the change is in:

* (new) Include/internal/pycore_global_strings.h - the declarations for the global strings, along with the macros
* Include/internal/pycore_runtime_init.h - added the static initializers for the global strings
* Include/internal/pycore_global_objects.h - where the struct in pycore_global_strings.h is hooked into _PyRuntimeState
* Tools/scripts/generate_global_objects.py - added generation of the global string declarations and static initializers

I've also added a --check flag to generate_global_objects.py (along with make check-global-objects) to check for unused global strings.  That check is added to the PR CI config.

The remainder of this change updates the core code to use _Py_GET_GLOBAL_IDENTIFIER() instead of _Py_IDENTIFIER() and the related _Py*Id functions (likewise for _Py_GET_GLOBAL_STRING() instead of _Py_static_string()).  This includes adding a few functions where there wasn't already an alternative to _Py*Id(), replacing the _Py_Identifier * parameter with PyObject *.

The following are not changed (yet):

* stop using _Py_IDENTIFIER() in the stdlib modules
* (maybe) get rid of _Py_IDENTIFIER(), etc. entirely -- this may not be doable as at least one package on PyPI using this (private) API
* (maybe) intern the strings during runtime init

https://bugs.python.org/issue46541
2022-02-08 13:39:07 -07:00
Mark Dickinson d02c5e9b55
bpo-46258: Streamline isqrt fast path (#30333) 2022-01-15 09:58:04 +00:00
Serhiy Storchaka 2d787971c6
bpo-37295: Use constant-time comb() and perm() for larger n depending on k (GH-30305) 2022-01-09 15:32:25 +02:00
Mark Dickinson 0b58bac3e7
bpo-37295: More direct computation of power-of-two factor in math.comb (GH-30313)
Co-authored-by: Serhiy Storchaka <storchaka@gmail.com>
2021-12-31 19:52:27 +00:00
Mark Dickinson 02b5417f11
bpo-37295: Speed up math.comb(n, k) for 0 <= k <= n <= 67 (GH-30275) 2021-12-28 12:26:40 +00:00
Steve Dower 3363e1cb05
bpo-46018: Ensure that math.expm1 does not raise on underflow (GH-29997) 2021-12-09 18:31:54 +00:00
Serhiy Storchaka 60c320c38e
bpo-37295: Optimize math.comb() and math.perm() (GH-29090)
For very large numbers use divide-and-conquer algorithm for getting
benefit of Karatsuba multiplication of large numbers.

Do calculations completely in C unsigned long long instead of Python
integers if possible.
2021-12-05 22:26:10 +02:00
Gideon 6266e4af87
bpo-45917: Add math.exp2() method - return 2 raised to the power of x (GH-29829)
Co-authored-by: Mark Dickinson <dickinsm@gmail.com>
2021-11-29 18:55:43 +00:00
Christian Heimes fa26245a1c
bpo-45548: Remove _math.c workarounds for pre-C99 libm (GH-29179)
The :mod:`math` and :mod:`cmath` implementation now require a C99 compatible
``libm`` and no longer ship with workarounds for missing acosh, asinh,
expm1, and log1p functions.

The changeset also removes ``_math.c`` and moves the last remaining
workaround into ``_math.h``. This simplifies static builds with
``Modules/Setup`` and resolves symbol conflicts.

Co-authored-by: Mark Dickinson <mdickinson@enthought.com>
Co-authored-by: Brett Cannon <brett@python.org>
Signed-off-by: Christian Heimes <christian@python.org>
2021-10-25 01:25:27 -07:00
Christian Heimes 03e9f5dc75
bpo-43974: Move Py_BUILD_CORE_MODULE into module code (GH-29157)
setup.py no longer defines Py_BUILD_CORE_MODULE. Instead every
module defines the macro before #include "Python.h" unless
Py_BUILD_CORE_BUILTIN is already defined.

Py_BUILD_CORE_BUILTIN is defined for every module that is built by
Modules/Setup.

The PR also simplifies Modules/Setup. Makefile and makesetup
already define Py_BUILD_CORE_BUILTIN and include Modules/internal
for us.

Signed-off-by: Christian Heimes <christian@python.org>
2021-10-22 15:36:28 +02:00
Victor Stinner d943d19172
bpo-45439: Move _PyObject_CallNoArgs() to pycore_call.h (GH-28895)
* Move _PyObject_CallNoArgs() to pycore_call.h (internal C API).
* _ssl, _sqlite and _testcapi extensions now call the public
  PyObject_CallNoArgs() function, rather than _PyObject_CallNoArgs().
* _lsprof extension is now built with Py_BUILD_CORE_MODULE macro
  defined to get access to internal _PyObject_CallNoArgs().
2021-10-12 08:38:19 +02:00
Victor Stinner ce3489cfdb
bpo-45439: Rename _PyObject_CallNoArg() to _PyObject_CallNoArgs() (GH-28891)
Fix typo in the private _PyObject_CallNoArg() function name: rename
it to _PyObject_CallNoArgs() to be consistent with the public
function PyObject_CallNoArgs().
2021-10-12 00:42:23 +02:00
Pablo Galindo Salgado 84975146a7
bpo-35606: Fix math.prod tests using 'start' as keyword parameter (GH-28595) 2021-09-28 13:32:43 +01:00
Mark Dickinson 4a42cebf6d
bpo-44339: Fix math.pow corner case to comply with IEEE 754 (GH-26606)
Change the behaviour of `math.pow(0.0, -math.inf)` and `math.pow(-0.0, -math.inf)` to return positive infinity instead of raising `ValueError`. This makes `math.pow` consistent with the built-in `pow` (and the `**` operator) for this particular special case, and brings the `math.pow` special-case handling into compliance with IEEE 754.
2021-06-12 10:23:02 +01:00
Ajith Ramachandran ac867f10b4
bpo-44357:Add `math.cbrt()` function: Cube Root (GH-26622)
* Add math.cbrt() function: Cube Root

Co-authored-by: blurb-it[bot] <43283697+blurb-it[bot]@users.noreply.github.com>
Co-authored-by: Mark Dickinson <dickinsm@gmail.com>
2021-06-10 17:42:09 +01:00
Victor Stinner 3e7ee02327
bpo-42161: mathmodule.c: move _PyLong_GetOne() loop invariant (GH-26391)
Move _PyLong_GetZero() and _PyLong_GetOne() loop invariants outside
loops in functions:

* math.comb()
* math.gcd()
* math.lcm()
* math.perm()
2021-05-27 00:51:07 +02:00
Alperen Serkan Aksöz 09605ad726
missing multiply symbol in the documentation (GH-24686) 2021-03-03 14:59:52 +01:00
Victor Stinner 0837f99d33
bpo-42323: Fix math.nextafter() on AIX (GH-24381)
math_nextafter_impl() must return a Python object, not a C double.
2021-01-29 23:04:50 +01:00
Victor Stinner c1c3493fb7
bpo-42323: Fix math.nextafter() for NaN on AIX (GH-24265) 2021-01-20 15:20:13 +01:00
Victor Stinner 37834136d0
bpo-42161: Modules/ uses _PyLong_GetZero() and _PyLong_GetOne() (GH-22998)
Use _PyLong_GetZero() and _PyLong_GetOne() in Modules/ directory.

_cursesmodule.c and zoneinfo.c are now built with
Py_BUILD_CORE_MODULE macro defined.
2020-10-27 17:12:53 +01:00
Raymond Hettinger 497126f7ea
Update link to supporting references (GH-22488) 2020-10-01 19:30:54 -07:00
Raymond Hettinger 438e9fc66f
bpo-41513: Improve order of adding fractional values. Improve variable names. (GH-22368) 2020-09-22 20:01:12 -07:00
Raymond Hettinger ec8a15b034
Make fractional value accumulation consistent inside and outside the loop. (GH-22315) 2020-09-18 17:57:28 -07:00
Raymond Hettinger 457d4e97de
bpo-41513: Add docs and tests for hypot() (GH-22238) 2020-09-13 23:33:41 -07:00
Raymond Hettinger 67c998de24
bpo-41513: Expand comments and add references for a better understanding (GH-22123) 2020-09-06 15:10:07 -07:00
Raymond Hettinger 5b24d1592a
Improve hypot() accuracy with three separate accumulators (GH-22032) 2020-09-01 22:00:50 -07:00
Raymond Hettinger 92c38164a4
Further improve accuracy of math.hypot() (GH-22013) 2020-08-30 10:00:11 -07:00
Raymond Hettinger 27de28607a
bpo-41513: Save unnecessary steps in the hypot() calculation (#21994) 2020-08-29 09:11:04 -07:00
Raymond Hettinger 82e79480d6
Fix typos in comment (GH-21966) 2020-08-26 13:09:40 -07:00