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.
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
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.
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>
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>
* 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().
Fix typo in the private _PyObject_CallNoArg() function name: rename
it to _PyObject_CallNoArgs() to be consistent with the public
function PyObject_CallNoArgs().
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.
Use _PyLong_GetZero() and _PyLong_GetOne() in Modules/ directory.
_cursesmodule.c and zoneinfo.c are now built with
Py_BUILD_CORE_MODULE macro defined.
In GH-2866, _Py_Bit_Length() was added to pymath.h for lack of a better
location. GH-20518 added a more appropriate header file for bit utilities. It
also shows how to properly use intrinsics. This allows reconsidering bpo-29782.
* Move the function to the new header.
* Changed return type to match __builtin_clzl() and reviewed usage.
* Use intrinsics where available.
* Pick a fallback implementation suitable for inlining.
Previously, the result could have been an instance of a subclass of int.
Also revert bpo-26202 and make attributes start, stop and step of the range
object having exact type int.
Add private function _PyNumber_Index() which preserves the old behavior
of PyNumber_Index() for performance to use it in the conversion functions
like PyLong_AsLong().
In math_2(), the first PyFloat_AsDouble() call should be checked
for failure before the second call.
Co-authored-by: Mark Dickinson <dickinsm@gmail.com>
* bpo-39648: Expand math.gcd() and math.lcm() to handle multiple arguments.
* Simplify fast path.
* Difine lcm() without arguments returning 1.
* Apply suggestions from code review
Co-Authored-By: Mark Dickinson <dickinsm@gmail.com>
Co-authored-by: Mark Dickinson <dickinsm@gmail.com>
Move the dtoa.h header file to the internal C API as pycore_dtoa.h:
it only contains private functions (prefixed by "_Py").
The math and cmath modules must now be compiled with the
Py_BUILD_CORE macro defined.
The PyFPE_START_PROTECT() and PyFPE_END_PROTECT() macros are empty:
they have been doing nothing for the last year (since commit
735ae8d139), so stop using them.
* Add math.isqrt function computing the integer square root.
* Code cleanup: remove redundant comments, rename some variables.
* Tighten up code a bit more; use Py_XDECREF to simplify error handling.
* Update Modules/mathmodule.c
Co-Authored-By: Serhiy Storchaka <storchaka@gmail.com>
* Update Modules/mathmodule.c
Use real argument clinic type instead of an alias
Co-Authored-By: Serhiy Storchaka <storchaka@gmail.com>
* Add proof sketch
* Updates from review.
* Correct and expand documentation.
* Fix bad reference handling on error; make some variables block-local; other tidying.
* Style and consistency fixes.
* Add missing error check; don't try to DECREF a NULL a
* Simplify some error returns.
* Another two test cases:
- clarify that floats are rejected even if they happen to be
squares of small integers
- TypeError beats ValueError for a negative float
* Add fast path for small inputs. Needs tests.
* Speed up isqrt for n >= 2**64 as well; add extra tests.
* Reduce number of test-cases to avoid dominating the run-time of test_math.
* Don't perform unnecessary extra iterations when computing c_bit_length.
* Abstract common uint64_t code out into a separate function.
* Cleanup.
* Add a missing Py_DECREF in an error branch. More cleanup.
* Update Modules/mathmodule.c
Add missing `static` declaration to helper function.
Co-Authored-By: Serhiy Storchaka <storchaka@gmail.com>
* Add missing backtick.
* Add math.isqrt function computing the integer square root.
* Code cleanup: remove redundant comments, rename some variables.
* Tighten up code a bit more; use Py_XDECREF to simplify error handling.
* Update Modules/mathmodule.c
Co-Authored-By: Serhiy Storchaka <storchaka@gmail.com>
* Update Modules/mathmodule.c
Use real argument clinic type instead of an alias
Co-Authored-By: Serhiy Storchaka <storchaka@gmail.com>
* Add proof sketch
* Updates from review.
* Correct and expand documentation.
* Fix bad reference handling on error; make some variables block-local; other tidying.
* Style and consistency fixes.
* Add missing error check; don't try to DECREF a NULL a
* Simplify some error returns.
* Another two test cases:
- clarify that floats are rejected even if they happen to be
squares of small integers
- TypeError beats ValueError for a negative float
* Documentation and markup improvements; thanks Serhiy for the suggestions!
* Cleaner Misc/NEWS entry wording.
* Clean up (with one fix) to the algorithm explanation and proof.
The overflow check was relying on undefined behaviour as it was using the result of the multiplication to do the check, and once the overflow has already happened, any operation on the result is undefined behaviour.
Some extra checks that exercise code paths related to this are also added.
The standard math library (libm) may follow IEEE-754 recommendation to
include an implementation of sinPi(), i.e. sinPi(x):=sin(pi*x).
And this triggers a name clash, found by FreeBSD developer
Steve Kargl, who worken on putting sinpi into libm used on FreeBSD
(it has to be named "sinpi", not "sinPi", cf. e.g.
https://en.cppreference.com/w/c/experimental/fpext4).
Use the fast call convention for math functions atan2(),
copysign(), hypot() and remainder() and inline unpacking
arguments. This sped up them by 1.3--2.5 times.
The *max* value is no longer treated as a special case in the main loop. Besides making the main loop simpler and branchless, this also lets us relax the input restriction of *vec* to contain only non-negative values.
* 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.
* Implement math.remainder.
* Fix markup for arguments; use double spaces after period.
* Mark up function reference in what's new entry.
* Add comment explaining the calculation in the final branch.
* Fix out-of-order entry in whatsnew.
* Add comment explaining why it's good enough to compare m with c, in spite of possible rounding error.
* bpo-26121: Use C library implementation for math functions:
tgamma(), lgamma(), erf() and erfc().
* Don't use tgamma() and lgamma() from libc on OS X.
Issue #28858: The change b9c9691c72c5 introduced a regression. It seems like
_PyObject_CallArg1() uses more stack memory than
PyObject_CallFunctionObjArgs().
* PyObject_CallFunctionObjArgs(func, NULL) => _PyObject_CallNoArg(func)
* PyObject_CallFunctionObjArgs(func, arg, NULL) => _PyObject_CallArg1(func, arg)
PyObject_CallFunctionObjArgs() allocates 40 bytes on the C stack and requires
extra work to "parse" C arguments to build a C array of PyObject*.
_PyObject_CallNoArg() and _PyObject_CallArg1() are simpler and don't allocate
memory on the C stack.
This change is part of the fastcall project. The change on listsort() is
related to the issue #23507.
Victor Stinner
Eric Snow
Mark Dickinson
Serhiy Storchaka
Steve Dower
Gideon
Christian Heimes
Pablo Galindo Salgado
Ajith Ramachandran
Alperen Serkan Aksöz
Raymond Hettinger
Niklas Fiekas
Dong-hee Na
Zackery Spytz
ananthan-123
Mark Dickinson
Giovanni Cappellotto
Min ho Kim
Yash Aggarwal
Ned Batchelder
Dima Pasechnik
Ammar Askar
Barry Warsaw
Louie Lu
Victor Stinner
Raymond Hettinger
Guido van Rossum
Martin Panter