Merge remote-tracking branch 'upstream/master'

46 files changed, 711 insertions, 526 deletions

diff --git a/numpy/core/__init__.py b/numpy/core/__init__.py
index e8d3a381b..f22c86f59 100644
--- a/numpy/core/__init__.py
+++ b/numpy/core/__init__.py
@@ -75,7 +75,7 @@ from . import fromnumeric
 from .fromnumeric import *
 from . import defchararray as char
 from . import records as rec
-from .records import *
+from .records import record, recarray, format_parser
 from .memmap import *
 from .defchararray import chararray
 from . import function_base
@@ -106,7 +106,7 @@ from . import _methods
 __all__ = ['char', 'rec', 'memmap']
 __all__ += numeric.__all__
 __all__ += fromnumeric.__all__
-__all__ += rec.__all__
+__all__ += ['record', 'recarray', 'format_parser']
 __all__ += ['chararray']
 __all__ += function_base.__all__
 __all__ += machar.__all__
diff --git a/numpy/core/_add_newdocs.py b/numpy/core/_add_newdocs.py
index 2cbfe52be..6073166a0 100644
--- a/numpy/core/_add_newdocs.py
+++ b/numpy/core/_add_newdocs.py
@@ -377,7 +377,7 @@ add_newdoc('numpy.core', 'nditer',
     ...    while not it.finished:
     ...        it[0] = lamdaexpr(*it[1:])
     ...        it.iternext()
-    ...        return it.operands[0]
+    ...    return it.operands[0]
 
     >>> a = np.arange(5)
     >>> b = np.ones(5)
@@ -821,7 +821,7 @@ add_newdoc('numpy.core.multiarray', 'array',
         ===== ========= ===================================================
 
         When ``copy=False`` and a copy is made for other reasons, the result is
-        the same as if ``copy=True``, with some exceptions for `A`, see the
+        the same as if ``copy=True``, with some exceptions for 'A', see the
         Notes section. The default order is 'K'.
     subok : bool, optional
         If True, then sub-classes will be passed-through, otherwise
diff --git a/numpy/core/_add_newdocs_scalars.py b/numpy/core/_add_newdocs_scalars.py
index b9b151224..d31f0037d 100644
--- a/numpy/core/_add_newdocs_scalars.py
+++ b/numpy/core/_add_newdocs_scalars.py
@@ -6,6 +6,7 @@ platform-dependent information.
 from numpy.core import dtype
 from numpy.core import numerictypes as _numerictypes
 from numpy.core.function_base import add_newdoc
+import platform
 
 ##############################################################################
 #
@@ -49,6 +50,8 @@ possible_aliases = numeric_type_aliases([
     ])
 
 
+
+
 def add_newdoc_for_scalar_type(obj, fixed_aliases, doc):
     # note: `:field: value` is rST syntax which renders as field lists.
     o = getattr(_numerictypes, obj)
@@ -56,7 +59,7 @@ def add_newdoc_for_scalar_type(obj, fixed_aliases, doc):
     character_code = dtype(o).char
     canonical_name_doc = "" if obj == o.__name__ else ":Canonical name: `numpy.{}`\n    ".format(obj)
     alias_doc = ''.join(":Alias: `numpy.{}`\n    ".format(alias) for alias in fixed_aliases)
-    alias_doc += ''.join(":Alias on this platform: `numpy.{}`: {}.\n    ".format(alias, doc)
+    alias_doc += ''.join(":Alias on this platform ({} {}): `numpy.{}`: {}.\n    ".format(platform.system(), platform.machine(), alias, doc)
                          for (alias_type, alias, doc) in possible_aliases if alias_type is o)
     docstring = """
     {doc}
diff --git a/numpy/core/arrayprint.py b/numpy/core/arrayprint.py
index 94ec8ed34..5c1d6cb63 100644
--- a/numpy/core/arrayprint.py
+++ b/numpy/core/arrayprint.py
@@ -41,6 +41,7 @@ from .numeric import concatenate, asarray, errstate
 from .numerictypes import (longlong, intc, int_, float_, complex_, bool_,
                            flexible)
 from .overrides import array_function_dispatch, set_module
+import operator
 import warnings
 import contextlib
 
@@ -78,6 +79,7 @@ def _make_options_dict(precision=None, threshold=None, edgeitems=None,
     if legacy not in [None, False, '1.13']:
         warnings.warn("legacy printing option can currently only be '1.13' or "
                       "`False`", stacklevel=3)
+
     if threshold is not None:
         # forbid the bad threshold arg suggested by stack overflow, gh-12351
         if not isinstance(threshold, numbers.Number):
@@ -85,6 +87,14 @@ def _make_options_dict(precision=None, threshold=None, edgeitems=None,
         if np.isnan(threshold):
             raise ValueError("threshold must be non-NAN, try "
                              "sys.maxsize for untruncated representation")
+
+    if precision is not None:
+        # forbid the bad precision arg as suggested by issue #18254
+        try:
+            options['precision'] = operator.index(precision)
+        except TypeError as e:
+            raise TypeError('precision must be an integer') from e
+
     return options
 
 
@@ -538,7 +548,7 @@ def array2string(a, max_line_width=None, precision=None,
     separator : str, optional
         Inserted between elements.
     prefix : str, optional
-    suffix: str, optional
+    suffix : str, optional
         The length of the prefix and suffix strings are used to respectively
         align and wrap the output. An array is typically printed as::
 
diff --git a/numpy/core/arrayprint.pyi b/numpy/core/arrayprint.pyi
index 6aaae0320..d2a5fdef9 100644
--- a/numpy/core/arrayprint.pyi
+++ b/numpy/core/arrayprint.pyi
@@ -21,12 +21,12 @@ from numpy import (
     longdouble,
     clongdouble,
 )
-from numpy.typing import ArrayLike, _CharLike, _FloatLike
+from numpy.typing import ArrayLike, _CharLike_co, _FloatLike_co
 
 if sys.version_info > (3, 8):
-    from typing import Literal, TypedDict
+    from typing import Literal, TypedDict, SupportsIndex
 else:
-    from typing_extensions import Literal, TypedDict
+    from typing_extensions import Literal, TypedDict, SupportsIndex
 
 _FloatMode = Literal["fixed", "unique", "maxprec", "maxprec_equal"]
 
@@ -40,13 +40,13 @@ class _FormatDict(TypedDict, total=False):
     complexfloat: Callable[[complexfloating[Any, Any]], str]
     longcomplexfloat: Callable[[clongdouble], str]
     void: Callable[[void], str]
-    numpystr: Callable[[_CharLike], str]
+    numpystr: Callable[[_CharLike_co], str]
     object: Callable[[object], str]
     all: Callable[[object], str]
     int_kind: Callable[[integer[Any]], str]
     float_kind: Callable[[floating[Any]], str]
     complex_kind: Callable[[complexfloating[Any, Any]], str]
-    str_kind: Callable[[_CharLike], str]
+    str_kind: Callable[[_CharLike_co], str]
 
 class _FormatOptions(TypedDict):
     precision: int
@@ -62,7 +62,7 @@ class _FormatOptions(TypedDict):
     legacy: Literal[False, "1.13"]
 
 def set_printoptions(
-    precision: Optional[int] = ...,
+    precision: Optional[SupportsIndex] = ...,
     threshold: Optional[int] = ...,
     edgeitems: Optional[int] = ...,
     linewidth: Optional[int] = ...,
@@ -79,7 +79,7 @@ def get_printoptions() -> _FormatOptions: ...
 def array2string(
     a: ndarray[Any, Any],
     max_line_width: Optional[int] = ...,
-    precision: Optional[int] = ...,
+    precision: Optional[SupportsIndex] = ...,
     suppress_small: Optional[bool] = ...,
     separator: str = ...,
     prefix: str = ...,
@@ -96,7 +96,7 @@ def array2string(
     legacy: Optional[Literal[False, "1.13"]] = ...,
 ) -> str: ...
 def format_float_scientific(
-    x: _FloatLike,
+    x: _FloatLike_co,
     precision: Optional[int] = ...,
     unique: bool = ...,
     trim: Literal["k", ".", "0", "-"] = ...,
@@ -105,7 +105,7 @@ def format_float_scientific(
     exp_digits: Optional[int] = ...,
 ) -> str: ...
 def format_float_positional(
-    x: _FloatLike,
+    x: _FloatLike_co,
     precision: Optional[int] = ...,
     unique: bool = ...,
     fractional: bool = ...,
@@ -117,20 +117,20 @@ def format_float_positional(
 def array_repr(
     arr: ndarray[Any, Any],
     max_line_width: Optional[int] = ...,
-    precision: Optional[int] = ...,
+    precision: Optional[SupportsIndex] = ...,
     suppress_small: Optional[bool] = ...,
 ) -> str: ...
 def array_str(
     a: ndarray[Any, Any],
     max_line_width: Optional[int] = ...,
-    precision: Optional[int] = ...,
+    precision: Optional[SupportsIndex] = ...,
     suppress_small: Optional[bool] = ...,
 ) -> str: ...
 def set_string_function(
     f: Optional[Callable[[ndarray[Any, Any]], str]], repr: bool = ...
 ) -> None: ...
 def printoptions(
-    precision: Optional[int] = ...,
+    precision: Optional[SupportsIndex] = ...,
     threshold: Optional[int] = ...,
     edgeitems: Optional[int] = ...,
     linewidth: Optional[int] = ...,
diff --git a/numpy/core/defchararray.py b/numpy/core/defchararray.py
index 9d7b54a1a..ab1166ad2 100644
--- a/numpy/core/defchararray.py
+++ b/numpy/core/defchararray.py
@@ -273,7 +273,7 @@ def str_len(a):
     out : ndarray
         Output array of integers
 
-    See also
+    See Also
     --------
     builtins.len
     """
@@ -368,7 +368,7 @@ def mod(a, values):
     out : ndarray
         Output array of str or unicode, depending on input types
 
-    See also
+    See Also
     --------
     str.__mod__
 
@@ -398,7 +398,7 @@ def capitalize(a):
         Output array of str or unicode, depending on input
         types
 
-    See also
+    See Also
     --------
     str.capitalize
 
@@ -443,7 +443,7 @@ def center(a, width, fillchar=' '):
         Output array of str or unicode, depending on input
         types
 
-    See also
+    See Also
     --------
     str.center
 
@@ -485,7 +485,7 @@ def count(a, sub, start=0, end=None):
     out : ndarray
         Output array of ints.
 
-    See also
+    See Also
     --------
     str.count
 
@@ -534,7 +534,7 @@ def decode(a, encoding=None, errors=None):
     -------
     out : ndarray
 
-    See also
+    See Also
     --------
     str.decode
 
@@ -580,7 +580,7 @@ def encode(a, encoding=None, errors=None):
     -------
     out : ndarray
 
-    See also
+    See Also
     --------
     str.encode
 
@@ -620,7 +620,7 @@ def endswith(a, suffix, start=0, end=None):
     out : ndarray
         Outputs an array of bools.
 
-    See also
+    See Also
     --------
     str.endswith
 
@@ -672,7 +672,7 @@ def expandtabs(a, tabsize=8):
     out : ndarray
         Output array of str or unicode, depending on input type
 
-    See also
+    See Also
     --------
     str.expandtabs
 
@@ -708,7 +708,7 @@ def find(a, sub, start=0, end=None):
     out : ndarray or int
         Output array of ints.  Returns -1 if `sub` is not found.
 
-    See also
+    See Also
     --------
     str.find
 
@@ -737,7 +737,7 @@ def index(a, sub, start=0, end=None):
     out : ndarray
         Output array of ints.  Returns -1 if `sub` is not found.
 
-    See also
+    See Also
     --------
     find, str.find
 
@@ -765,7 +765,7 @@ def isalnum(a):
     out : ndarray
         Output array of str or unicode, depending on input type
 
-    See also
+    See Also
     --------
     str.isalnum
     """
@@ -791,7 +791,7 @@ def isalpha(a):
     out : ndarray
         Output array of bools
 
-    See also
+    See Also
     --------
     str.isalpha
     """
@@ -817,7 +817,7 @@ def isdigit(a):
     out : ndarray
         Output array of bools
 
-    See also
+    See Also
     --------
     str.isdigit
     """
@@ -844,7 +844,7 @@ def islower(a):
     out : ndarray
         Output array of bools
 
-    See also
+    See Also
     --------
     str.islower
     """
@@ -871,7 +871,7 @@ def isspace(a):
     out : ndarray
         Output array of bools
 
-    See also
+    See Also
     --------
     str.isspace
     """
@@ -897,7 +897,7 @@ def istitle(a):
     out : ndarray
         Output array of bools
 
-    See also
+    See Also
     --------
     str.istitle
     """
@@ -924,7 +924,7 @@ def isupper(a):
     out : ndarray
         Output array of bools
 
-    See also
+    See Also
     --------
     str.isupper
     """
@@ -953,7 +953,7 @@ def join(sep, seq):
     out : ndarray
         Output array of str or unicode, depending on input types
 
-    See also
+    See Also
     --------
     str.join
     """
@@ -988,7 +988,7 @@ def ljust(a, width, fillchar=' '):
     out : ndarray
         Output array of str or unicode, depending on input type
 
-    See also
+    See Also
     --------
     str.ljust
 
@@ -1021,7 +1021,7 @@ def lower(a):
     out : ndarray, {str, unicode}
         Output array of str or unicode, depending on input type
 
-    See also
+    See Also
     --------
     str.lower
 
@@ -1066,7 +1066,7 @@ def lstrip(a, chars=None):
     out : ndarray, {str, unicode}
         Output array of str or unicode, depending on input type
 
-    See also
+    See Also
     --------
     str.lstrip
 
@@ -1127,7 +1127,7 @@ def partition(a, sep):
         The output array will have an extra dimension with 3
         elements per input element.
 
-    See also
+    See Also
     --------
     str.partition
 
@@ -1163,7 +1163,7 @@ def replace(a, old, new, count=None):
     out : ndarray
         Output array of str or unicode, depending on input type
 
-    See also
+    See Also
     --------
     str.replace
 
@@ -1197,7 +1197,7 @@ def rfind(a, sub, start=0, end=None):
     out : ndarray
        Output array of ints.  Return -1 on failure.
 
-    See also
+    See Also
     --------
     str.rfind
 
@@ -1227,7 +1227,7 @@ def rindex(a, sub, start=0, end=None):
     out : ndarray
        Output array of ints.
 
-    See also
+    See Also
     --------
     rfind, str.rindex
 
@@ -1258,7 +1258,7 @@ def rjust(a, width, fillchar=' '):
     out : ndarray
         Output array of str or unicode, depending on input type
 
-    See also
+    See Also
     --------
     str.rjust
 
@@ -1299,7 +1299,7 @@ def rpartition(a, sep):
         type.  The output array will have an extra dimension with
         3 elements per input element.
 
-    See also
+    See Also
     --------
     str.rpartition
 
@@ -1339,7 +1339,7 @@ def rsplit(a, sep=None, maxsplit=None):
     out : ndarray
        Array of list objects
 
-    See also
+    See Also
     --------
     str.rsplit, split
 
@@ -1378,7 +1378,7 @@ def rstrip(a, chars=None):
     out : ndarray
         Output array of str or unicode, depending on input type
 
-    See also
+    See Also
     --------
     str.rstrip
 
@@ -1423,7 +1423,7 @@ def split(a, sep=None, maxsplit=None):
     out : ndarray
         Array of list objects
 
-    See also
+    See Also
     --------
     str.split, rsplit
 
@@ -1459,7 +1459,7 @@ def splitlines(a, keepends=None):
     out : ndarray
         Array of list objects
 
-    See also
+    See Also
     --------
     str.splitlines
 
@@ -1495,7 +1495,7 @@ def startswith(a, prefix, start=0, end=None):
     out : ndarray
         Array of booleans
 
-    See also
+    See Also
     --------
     str.startswith
 
@@ -1528,7 +1528,7 @@ def strip(a, chars=None):
     out : ndarray
         Output array of str or unicode, depending on input type
 
-    See also
+    See Also
     --------
     str.strip
 
@@ -1569,7 +1569,7 @@ def swapcase(a):
     out : ndarray, {str, unicode}
         Output array of str or unicode, depending on input type
 
-    See also
+    See Also
     --------
     str.swapcase
 
@@ -1609,7 +1609,7 @@ def title(a):
     out : ndarray
         Output array of str or unicode, depending on input type
 
-    See also
+    See Also
     --------
     str.title
 
@@ -1654,7 +1654,7 @@ def translate(a, table, deletechars=None):
     out : ndarray
         Output array of str or unicode, depending on input type
 
-    See also
+    See Also
     --------
     str.translate
 
@@ -1687,7 +1687,7 @@ def upper(a):
     out : ndarray, {str, unicode}
         Output array of str or unicode, depending on input type
 
-    See also
+    See Also
     --------
     str.upper
 
@@ -1726,7 +1726,7 @@ def zfill(a, width):
     out : ndarray, {str, unicode}
         Output array of str or unicode, depending on input type
 
-    See also
+    See Also
     --------
     str.zfill
 
@@ -1760,7 +1760,7 @@ def isnumeric(a):
     out : ndarray, bool
         Array of booleans of same shape as `a`.
 
-    See also
+    See Also
     --------
     unicode.isnumeric
 
@@ -1792,7 +1792,7 @@ def isdecimal(a):
     out : ndarray, bool
         Array of booleans identical in shape to `a`.
 
-    See also
+    See Also
     --------
     unicode.isdecimal
 
@@ -2004,7 +2004,7 @@ class chararray(ndarray):
         """
         Return (self == other) element-wise.
 
-        See also
+        See Also
         --------
         equal
         """
@@ -2014,7 +2014,7 @@ class chararray(ndarray):
         """
         Return (self != other) element-wise.
 
-        See also
+        See Also
         --------
         not_equal
         """
@@ -2024,7 +2024,7 @@ class chararray(ndarray):
         """
         Return (self >= other) element-wise.
 
-        See also
+        See Also
         --------
         greater_equal
         """
@@ -2034,7 +2034,7 @@ class chararray(ndarray):
         """
         Return (self <= other) element-wise.
 
-        See also
+        See Also
         --------
         less_equal
         """
@@ -2044,7 +2044,7 @@ class chararray(ndarray):
         """
         Return (self > other) element-wise.
 
-        See also
+        See Also
         --------
         greater
         """
@@ -2054,7 +2054,7 @@ class chararray(ndarray):
         """
         Return (self < other) element-wise.
 
-        See also
+        See Also
         --------
         less
         """
@@ -2065,7 +2065,7 @@ class chararray(ndarray):
         Return (self + other), that is string concatenation,
         element-wise for a pair of array_likes of str or unicode.
 
-        See also
+        See Also
         --------
         add
         """
@@ -2076,7 +2076,7 @@ class chararray(ndarray):
         Return (other + self), that is string concatenation,
         element-wise for a pair of array_likes of `string_` or `unicode_`.
 
-        See also
+        See Also
         --------
         add
         """
@@ -2087,7 +2087,7 @@ class chararray(ndarray):
         Return (self * i), that is string multiple concatenation,
         element-wise.
 
-        See also
+        See Also
         --------
         multiply
         """
@@ -2098,7 +2098,7 @@ class chararray(ndarray):
         Return (self * i), that is string multiple concatenation,
         element-wise.
 
-        See also
+        See Also
         --------
         multiply
         """
@@ -2110,7 +2110,7 @@ class chararray(ndarray):
         (interpolation), element-wise for a pair of array_likes of `string_`
         or `unicode_`.
 
-        See also
+        See Also
         --------
         mod
         """
@@ -2145,7 +2145,7 @@ class chararray(ndarray):
         Return a copy of `self` with only the first character of each element
         capitalized.
 
-        See also
+        See Also
         --------
         char.capitalize
 
@@ -2157,7 +2157,7 @@ class chararray(ndarray):
         Return a copy of `self` with its elements centered in a
         string of length `width`.
 
-        See also
+        See Also
         --------
         center
         """
@@ -2168,7 +2168,7 @@ class chararray(ndarray):
         Returns an array with the number of non-overlapping occurrences of
         substring `sub` in the range [`start`, `end`].
 
-        See also
+        See Also
         --------
         char.count
 
@@ -2179,7 +2179,7 @@ class chararray(ndarray):
         """
         Calls `str.decode` element-wise.
 
-        See also
+        See Also
         --------
         char.decode
 
@@ -2190,7 +2190,7 @@ class chararray(ndarray):
         """
         Calls `str.encode` element-wise.
 
-        See also
+        See Also
         --------
         char.encode
 
@@ -2202,7 +2202,7 @@ class chararray(ndarray):
         Returns a boolean array which is `True` where the string element
         in `self` ends with `suffix`, otherwise `False`.
 
-        See also
+        See Also
         --------
         char.endswith
 
@@ -2214,7 +2214,7 @@ class chararray(ndarray):
         Return a copy of each string element where all tab characters are
         replaced by one or more spaces.
 
-        See also
+        See Also
         --------
         char.expandtabs
 
@@ -2226,7 +2226,7 @@ class chararray(ndarray):
         For each element, return the lowest index in the string where
         substring `sub` is found.
 
-        See also
+        See Also
         --------
         char.find
 
@@ -2237,7 +2237,7 @@ class chararray(ndarray):
         """
         Like `find`, but raises `ValueError` when the substring is not found.
 
-        See also
+        See Also
         --------
         char.index
 
@@ -2250,7 +2250,7 @@ class chararray(ndarray):
         are alphanumeric and there is at least one character, false
         otherwise.
 
-        See also
+        See Also
         --------
         char.isalnum
 
@@ -2263,7 +2263,7 @@ class chararray(ndarray):
         are alphabetic and there is at least one character, false
         otherwise.
 
-        See also
+        See Also
         --------
         char.isalpha
 
@@ -2275,7 +2275,7 @@ class chararray(ndarray):
         Returns true for each element if all characters in the string are
         digits and there is at least one character, false otherwise.
 
-        See also
+        See Also
         --------
         char.isdigit
 
@@ -2288,7 +2288,7 @@ class chararray(ndarray):
         string are lowercase and there is at least one cased character,
         false otherwise.
 
-        See also
+        See Also
         --------
         char.islower
 
@@ -2301,7 +2301,7 @@ class chararray(ndarray):
         characters in the string and there is at least one character,
         false otherwise.
 
-        See also
+        See Also
         --------
         char.isspace
 
@@ -2313,7 +2313,7 @@ class chararray(ndarray):
         Returns true for each element if the element is a titlecased
         string and there is at least one character, false otherwise.
 
-        See also
+        See Also
         --------
         char.istitle
 
@@ -2326,7 +2326,7 @@ class chararray(ndarray):
         string are uppercase and there is at least one character, false
         otherwise.
 
-        See also
+        See Also
         --------
         char.isupper
 
@@ -2338,7 +2338,7 @@ class chararray(ndarray):
         Return a string which is the concatenation of the strings in the
         sequence `seq`.
 
-        See also
+        See Also
         --------
         char.join
 
@@ -2350,7 +2350,7 @@ class chararray(ndarray):
         Return an array with the elements of `self` left-justified in a
         string of length `width`.
 
-        See also
+        See Also
         --------
         char.ljust
 
@@ -2362,7 +2362,7 @@ class chararray(ndarray):
         Return an array with the elements of `self` converted to
         lowercase.
 
-        See also
+        See Also
         --------
         char.lower
 
@@ -2374,7 +2374,7 @@ class chararray(ndarray):
         For each element in `self`, return a copy with the leading characters
         removed.
 
-        See also
+        See Also
         --------
         char.lstrip
 
@@ -2385,7 +2385,7 @@ class chararray(ndarray):
         """
         Partition each element in `self` around `sep`.
 
-        See also
+        See Also
         --------
         partition
         """
@@ -2396,7 +2396,7 @@ class chararray(ndarray):
         For each element in `self`, return a copy of the string with all
         occurrences of substring `old` replaced by `new`.
 
-        See also
+        See Also
         --------
         char.replace
 
@@ -2409,7 +2409,7 @@ class chararray(ndarray):
         where substring `sub` is found, such that `sub` is contained
         within [`start`, `end`].
 
-        See also
+        See Also
         --------
         char.rfind
 
@@ -2421,7 +2421,7 @@ class chararray(ndarray):
         Like `rfind`, but raises `ValueError` when the substring `sub` is
         not found.
 
-        See also
+        See Also
         --------
         char.rindex
 
@@ -2433,7 +2433,7 @@ class chararray(ndarray):
         Return an array with the elements of `self`
         right-justified in a string of length `width`.
 
-        See also
+        See Also
         --------
         char.rjust
 
@@ -2444,7 +2444,7 @@ class chararray(ndarray):
         """
         Partition each element in `self` around `sep`.
 
-        See also
+        See Also
         --------
         rpartition
         """
@@ -2455,7 +2455,7 @@ class chararray(ndarray):
         For each element in `self`, return a list of the words in
         the string, using `sep` as the delimiter string.
 
-        See also
+        See Also
         --------
         char.rsplit
 
@@ -2467,7 +2467,7 @@ class chararray(ndarray):
         For each element in `self`, return a copy with the trailing
         characters removed.
 
-        See also
+        See Also
         --------
         char.rstrip
 
@@ -2479,7 +2479,7 @@ class chararray(ndarray):
         For each element in `self`, return a list of the words in the
         string, using `sep` as the delimiter string.
 
-        See also
+        See Also
         --------
         char.split
 
@@ -2491,7 +2491,7 @@ class chararray(ndarray):
         For each element in `self`, return a list of the lines in the
         element, breaking at line boundaries.
 
-        See also
+        See Also
         --------
         char.splitlines
 
@@ -2503,7 +2503,7 @@ class chararray(ndarray):
         Returns a boolean array which is `True` where the string element
         in `self` starts with `prefix`, otherwise `False`.
 
-        See also
+        See Also
         --------
         char.startswith
 
@@ -2515,7 +2515,7 @@ class chararray(ndarray):
         For each element in `self`, return a copy with the leading and
         trailing characters removed.
 
-        See also
+        See Also
         --------
         char.strip
 
@@ -2527,7 +2527,7 @@ class chararray(ndarray):
         For each element in `self`, return a copy of the string with
         uppercase characters converted to lowercase and vice versa.
 
-        See also
+        See Also
         --------
         char.swapcase
 
@@ -2540,7 +2540,7 @@ class chararray(ndarray):
         string: words start with uppercase characters, all remaining cased
         characters are lowercase.
 
-        See also
+        See Also
         --------
         char.title
 
@@ -2554,7 +2554,7 @@ class chararray(ndarray):
         `deletechars` are removed, and the remaining characters have
         been mapped through the given translation table.
 
-        See also
+        See Also
         --------
         char.translate
 
@@ -2566,7 +2566,7 @@ class chararray(ndarray):
         Return an array with the elements of `self` converted to
         uppercase.
 
-        See also
+        See Also
         --------
         char.upper
 
@@ -2578,7 +2578,7 @@ class chararray(ndarray):
         Return the numeric string left-filled with zeros in a string of
         length `width`.
 
-        See also
+        See Also
         --------
         char.zfill
 
@@ -2590,7 +2590,7 @@ class chararray(ndarray):
         For each element in `self`, return True if there are only
         numeric characters in the element.
 
-        See also
+        See Also
         --------
         char.isnumeric
 
@@ -2602,7 +2602,7 @@ class chararray(ndarray):
         For each element in `self`, return True if there are only
         decimal characters in the element.
 
-        See also
+        See Also
         --------
         char.isdecimal
 
diff --git a/numpy/core/einsumfunc.py b/numpy/core/einsumfunc.py
index e0942beca..18157641a 100644
--- a/numpy/core/einsumfunc.py
+++ b/numpy/core/einsumfunc.py
@@ -327,7 +327,7 @@ def _greedy_path(input_sets, output_set, idx_dict, memory_limit):
         Set that represents the rhs side of the overall einsum subscript
     idx_dict : dictionary
         Dictionary of index sizes
-    memory_limit_limit : int
+    memory_limit : int
         The maximum number of elements in a temporary array
 
     Returns
@@ -1061,14 +1061,12 @@ def einsum(*operands, out=None, optimize=False, **kwargs):
     See Also
     --------
     einsum_path, dot, inner, outer, tensordot, linalg.multi_dot
-
-    einops:
+    einops :
         similar verbose interface is provided by
         `einops <https://github.com/arogozhnikov/einops>`_ package to cover
         additional operations: transpose, reshape/flatten, repeat/tile,
         squeeze/unsqueeze and reductions.
-
-    opt_einsum:
+    opt_einsum :
         `opt_einsum <https://optimized-einsum.readthedocs.io/en/stable/>`_
         optimizes contraction order for einsum-like expressions
         in backend-agnostic manner.
diff --git a/numpy/core/fromnumeric.py b/numpy/core/fromnumeric.py
index 52df1aad9..658b1aca5 100644
--- a/numpy/core/fromnumeric.py
+++ b/numpy/core/fromnumeric.py
@@ -319,34 +319,34 @@ def choose(a, choices, out=None, mode='raise'):
 
     But this omits some subtleties.  Here is a fully general summary:
 
-    Given an "index" array (`a`) of integers and a sequence of `n` arrays
+    Given an "index" array (`a`) of integers and a sequence of ``n`` arrays
     (`choices`), `a` and each choice array are first broadcast, as necessary,
     to arrays of a common shape; calling these *Ba* and *Bchoices[i], i =
     0,...,n-1* we have that, necessarily, ``Ba.shape == Bchoices[i].shape``
-    for each `i`.  Then, a new array with shape ``Ba.shape`` is created as
+    for each ``i``.  Then, a new array with shape ``Ba.shape`` is created as
     follows:
 
-    * if ``mode=raise`` (the default), then, first of all, each element of
-      `a` (and thus `Ba`) must be in the range `[0, n-1]`; now, suppose that
-      `i` (in that range) is the value at the `(j0, j1, ..., jm)` position
-      in `Ba` - then the value at the same position in the new array is the
-      value in `Bchoices[i]` at that same position;
+    * if ``mode='raise'`` (the default), then, first of all, each element of
+      ``a`` (and thus ``Ba``) must be in the range ``[0, n-1]``; now, suppose
+      that ``i`` (in that range) is the value at the ``(j0, j1, ..., jm)``
+      position in ``Ba`` - then the value at the same position in the new array
+      is the value in ``Bchoices[i]`` at that same position;
 
-    * if ``mode=wrap``, values in `a` (and thus `Ba`) may be any (signed)
+    * if ``mode='wrap'``, values in `a` (and thus `Ba`) may be any (signed)
       integer; modular arithmetic is used to map integers outside the range
       `[0, n-1]` back into that range; and then the new array is constructed
       as above;
 
-    * if ``mode=clip``, values in `a` (and thus `Ba`) may be any (signed)
-      integer; negative integers are mapped to 0; values greater than `n-1`
-      are mapped to `n-1`; and then the new array is constructed as above.
+    * if ``mode='clip'``, values in `a` (and thus ``Ba``) may be any (signed)
+      integer; negative integers are mapped to 0; values greater than ``n-1``
+      are mapped to ``n-1``; and then the new array is constructed as above.
 
     Parameters
     ----------
     a : int array
-        This array must contain integers in `[0, n-1]`, where `n` is the number
-        of choices, unless ``mode=wrap`` or ``mode=clip``, in which cases any
-        integers are permissible.
+        This array must contain integers in ``[0, n-1]``, where ``n`` is the
+        number of choices, unless ``mode=wrap`` or ``mode=clip``, in which
+        cases any integers are permissible.
     choices : sequence of arrays
         Choice arrays. `a` and all of the choices must be broadcastable to the
         same shape.  If `choices` is itself an array (not recommended), then
@@ -355,12 +355,12 @@ def choose(a, choices, out=None, mode='raise'):
     out : array, optional
         If provided, the result will be inserted into this array. It should
         be of the appropriate shape and dtype. Note that `out` is always
-        buffered if `mode='raise'`; use other modes for better performance.
+        buffered if ``mode='raise'``; use other modes for better performance.
     mode : {'raise' (default), 'wrap', 'clip'}, optional
-        Specifies how indices outside `[0, n-1]` will be treated:
+        Specifies how indices outside ``[0, n-1]`` will be treated:
 
           * 'raise' : an exception is raised
-          * 'wrap' : value becomes value mod `n`
+          * 'wrap' : value becomes value mod ``n``
           * 'clip' : values < 0 are mapped to 0, values > n-1 are mapped to n-1
 
     Returns
@@ -1381,7 +1381,7 @@ def resize(a, new_shape):
     --------
     np.reshape : Reshape an array without changing the total size.
     np.pad : Enlarge and pad an array.
-    np.repeat: Repeat elements of an array.
+    np.repeat : Repeat elements of an array.
     ndarray.resize : resize an array in-place.
 
     Notes
@@ -2007,7 +2007,7 @@ def compress(condition, a, axis=None, out=None):
     --------
     take, choose, diag, diagonal, select
     ndarray.compress : Equivalent method in ndarray
-    extract: Equivalent method when working on 1-D arrays
+    extract : Equivalent method when working on 1-D arrays
     :ref:`ufuncs-output-type`
 
     Examples
@@ -2475,14 +2475,11 @@ def cumsum(a, axis=None, dtype=None, out=None):
         result has the same size as `a`, and the same shape as `a` if
         `axis` is not None or `a` is a 1-d array.
 
-
     See Also
     --------
     sum : Sum array elements.
-
     trapz : Integration of array values using the composite trapezoidal rule.
-
-    diff :  Calculate the n-th discrete difference along given axis.
+    diff : Calculate the n-th discrete difference along given axis.
 
     Notes
     -----
diff --git a/numpy/core/fromnumeric.pyi b/numpy/core/fromnumeric.pyi
index 3b147e1d7..fc7f28a59 100644
--- a/numpy/core/fromnumeric.pyi
+++ b/numpy/core/fromnumeric.pyi
@@ -23,9 +23,8 @@ from numpy.typing import (
     ArrayLike,
     _ShapeLike,
     _Shape,
-    _IntLike,
-    _BoolLike,
-    _NumberLike,
+    _IntLike_co,
+    _NumberLike_co,
 )
 
 if sys.version_info >= (3, 8):
@@ -98,7 +97,7 @@ def choose(
 ) -> _ScalarIntOrBool: ...
 @overload
 def choose(
-    a: Union[_IntLike, _BoolLike], choices: ArrayLike, out: Optional[ndarray] = ..., mode: _ModeKind = ...
+    a: _IntLike_co, choices: ArrayLike, out: Optional[ndarray] = ..., mode: _ModeKind = ...
 ) -> Union[integer, bool_]: ...
 @overload
 def choose(
@@ -250,7 +249,7 @@ def sum(
     dtype: DTypeLike = ...,
     out: Optional[ndarray] = ...,
     keepdims: bool = ...,
-    initial: _NumberLike = ...,
+    initial: _NumberLike_co = ...,
     where: _ArrayLikeBool = ...,
 ) -> _Number: ...
 @overload
@@ -260,7 +259,7 @@ def sum(
     dtype: DTypeLike = ...,
     out: Optional[ndarray] = ...,
     keepdims: bool = ...,
-    initial: _NumberLike = ...,
+    initial: _NumberLike_co = ...,
     where: _ArrayLikeBool = ...,
 ) -> Union[number, ndarray]: ...
 @overload
@@ -324,7 +323,7 @@ def amax(
     axis: Optional[_ShapeLike] = ...,
     out: Optional[ndarray] = ...,
     keepdims: bool = ...,
-    initial: _NumberLike = ...,
+    initial: _NumberLike_co = ...,
     where: _ArrayLikeBool = ...,
 ) -> _Number: ...
 @overload
@@ -333,7 +332,7 @@ def amax(
     axis: None = ...,
     out: Optional[ndarray] = ...,
     keepdims: Literal[False] = ...,
-    initial: _NumberLike = ...,
+    initial: _NumberLike_co = ...,
     where: _ArrayLikeBool = ...,
 ) -> number: ...
 @overload
@@ -342,7 +341,7 @@ def amax(
     axis: Optional[_ShapeLike] = ...,
     out: Optional[ndarray] = ...,
     keepdims: bool = ...,
-    initial: _NumberLike = ...,
+    initial: _NumberLike_co = ...,
     where: _ArrayLikeBool = ...,
 ) -> Union[number, ndarray]: ...
 @overload
@@ -351,7 +350,7 @@ def amin(
     axis: Optional[_ShapeLike] = ...,
     out: Optional[ndarray] = ...,
     keepdims: bool = ...,
-    initial: _NumberLike = ...,
+    initial: _NumberLike_co = ...,
     where: _ArrayLikeBool = ...,
 ) -> _Number: ...
 @overload
@@ -360,7 +359,7 @@ def amin(
     axis: None = ...,
     out: Optional[ndarray] = ...,
     keepdims: Literal[False] = ...,
-    initial: _NumberLike = ...,
+    initial: _NumberLike_co = ...,
     where: _ArrayLikeBool = ...,
 ) -> number: ...
 @overload
@@ -369,7 +368,7 @@ def amin(
     axis: Optional[_ShapeLike] = ...,
     out: Optional[ndarray] = ...,
     keepdims: bool = ...,
-    initial: _NumberLike = ...,
+    initial: _NumberLike_co = ...,
     where: _ArrayLikeBool = ...,
 ) -> Union[number, ndarray]: ...
 
@@ -387,7 +386,7 @@ def prod(
     dtype: DTypeLike = ...,
     out: None = ...,
     keepdims: bool = ...,
-    initial: _NumberLike = ...,
+    initial: _NumberLike_co = ...,
     where: _ArrayLikeBool = ...,
 ) -> _Number: ...
 @overload
@@ -397,7 +396,7 @@ def prod(
     dtype: DTypeLike = ...,
     out: None = ...,
     keepdims: Literal[False] = ...,
-    initial: _NumberLike = ...,
+    initial: _NumberLike_co = ...,
     where: _ArrayLikeBool = ...,
 ) -> number: ...
 @overload
@@ -407,7 +406,7 @@ def prod(
     dtype: DTypeLike = ...,
     out: Optional[ndarray] = ...,
     keepdims: bool = ...,
-    initial: _NumberLike = ...,
+    initial: _NumberLike_co = ...,
     where: _ArrayLikeBool = ...,
 ) -> Union[number, ndarray]: ...
 def cumprod(
@@ -424,7 +423,7 @@ def around(
 ) -> _Number: ...
 @overload
 def around(
-    a: _NumberLike, decimals: int = ..., out: Optional[ndarray] = ...
+    a: _NumberLike_co, decimals: int = ..., out: Optional[ndarray] = ...
 ) -> number: ...
 @overload
 def around(
diff --git a/numpy/core/function_base.pyi b/numpy/core/function_base.pyi
index 1490bed4a..d4543f281 100644
--- a/numpy/core/function_base.pyi
+++ b/numpy/core/function_base.pyi
@@ -2,20 +2,17 @@ import sys
 from typing import overload, Tuple, Union, Sequence, Any
 
 from numpy import ndarray, inexact
-from numpy.typing import ArrayLike, DTypeLike, _SupportsArray, _NumberLike
+from numpy.typing import ArrayLike, DTypeLike, _SupportsArray, _NumberLike_co
 
 if sys.version_info >= (3, 8):
     from typing import SupportsIndex, Literal
 else:
-    from typing_extensions import Literal, Protocol
-
-    class SupportsIndex(Protocol):
-        def __index__(self) -> int: ...
+    from typing_extensions import SupportsIndex, Literal
 
 # TODO: wait for support for recursive types
 _ArrayLikeNested = Sequence[Sequence[Any]]
 _ArrayLikeNumber = Union[
-    _NumberLike, Sequence[_NumberLike], ndarray, _SupportsArray, _ArrayLikeNested
+    _NumberLike_co, Sequence[_NumberLike_co], ndarray, _SupportsArray, _ArrayLikeNested
 ]
 @overload
 def linspace(
diff --git a/numpy/core/include/numpy/random/distributions.h b/numpy/core/include/numpy/random/distributions.h
index c474c4d14..3ffacc8f9 100644
--- a/numpy/core/include/numpy/random/distributions.h
+++ b/numpy/core/include/numpy/random/distributions.h
@@ -1,6 +1,10 @@
 #ifndef _RANDOMDGEN__DISTRIBUTIONS_H_
 #define _RANDOMDGEN__DISTRIBUTIONS_H_
 
+#ifdef __cplusplus
+extern "C" {
+#endif
+
 #include "Python.h"
 #include "numpy/npy_common.h"
 #include <stddef.h>
@@ -197,4 +201,8 @@ static NPY_INLINE double next_double(bitgen_t *bitgen_state) {
     return bitgen_state->next_double(bitgen_state->state);
 }
 
+#ifdef __cplusplus
+}
+#endif
+
 #endif
diff --git a/numpy/core/multiarray.py b/numpy/core/multiarray.py
index 07179a627..b7277ac24 100644
--- a/numpy/core/multiarray.py
+++ b/numpy/core/multiarray.py
@@ -1441,7 +1441,7 @@ def is_busday(dates, weekmask=None, holidays=None, busdaycal=None, out=None):
 
     See Also
     --------
-    busdaycalendar: An object that specifies a custom set of valid days.
+    busdaycalendar : An object that specifies a custom set of valid days.
     busday_offset : Applies an offset counted in valid days.
     busday_count : Counts how many valid days are in a half-open date range.
 
@@ -1516,7 +1516,7 @@ def busday_offset(dates, offsets, roll=None, weekmask=None, holidays=None,
 
     See Also
     --------
-    busdaycalendar: An object that specifies a custom set of valid days.
+    busdaycalendar : An object that specifies a custom set of valid days.
     is_busday : Returns a boolean array indicating valid days.
     busday_count : Counts how many valid days are in a half-open date range.
 
@@ -1598,7 +1598,7 @@ def busday_count(begindates, enddates, weekmask=None, holidays=None,
 
     See Also
     --------
-    busdaycalendar: An object that specifies a custom set of valid days.
+    busdaycalendar : An object that specifies a custom set of valid days.
     is_busday : Returns a boolean array indicating valid days.
     busday_offset : Applies an offset counted in valid days.
 
diff --git a/numpy/core/numeric.py b/numpy/core/numeric.py
index c95c48d71..086439656 100644
--- a/numpy/core/numeric.py
+++ b/numpy/core/numeric.py
@@ -299,7 +299,7 @@ def full(shape, fill_value, dtype=None, order='C', *, like=None):
         Fill value.
     dtype : data-type, optional
         The desired data-type for the array  The default, None, means
-         `np.array(fill_value).dtype`.
+         ``np.array(fill_value).dtype``.
     order : {'C', 'F'}, optional
         Whether to store multidimensional data in C- or Fortran-contiguous
         (row- or column-wise) order in memory.
@@ -1427,12 +1427,11 @@ def moveaxis(a, source, destination):
 
     See Also
     --------
-    transpose: Permute the dimensions of an array.
-    swapaxes: Interchange two axes of an array.
+    transpose : Permute the dimensions of an array.
+    swapaxes : Interchange two axes of an array.
 
     Examples
     --------
-
     >>> x = np.zeros((3, 4, 5))
     >>> np.moveaxis(x, 0, -1).shape
     (4, 5, 3)
diff --git a/numpy/core/records.py b/numpy/core/records.py
index 00d456658..a626a0589 100644
--- a/numpy/core/records.py
+++ b/numpy/core/records.py
@@ -45,7 +45,10 @@ from numpy.core.overrides import set_module
 from .arrayprint import get_printoptions
 
 # All of the functions allow formats to be a dtype
-__all__ = ['record', 'recarray', 'format_parser']
+__all__ = [
+    'record', 'recarray', 'format_parser',
+    'fromarrays', 'fromrecords', 'fromstring', 'fromfile', 'array',
+]
 
 
 ndarray = sb.ndarray
@@ -962,16 +965,16 @@ def array(obj, dtype=None, shape=None, offset=0, strides=None, formats=None,
 
     Parameters
     ----------
-    obj: any
+    obj : any
         Input object. See Notes for details on how various input types are
         treated.
-    dtype: data-type, optional
+    dtype : data-type, optional
         Valid dtype for array.
-    shape: int or tuple of ints, optional
+    shape : int or tuple of ints, optional
         Shape of each array.
-    offset: int, optional
+    offset : int, optional
         Position in the file or buffer to start reading from.
-    strides: tuple of ints, optional
+    strides : tuple of ints, optional
         Buffer (`buf`) is interpreted according to these strides (strides
         define how many bytes each array element, row, column, etc.
         occupy in memory).
@@ -979,7 +982,7 @@ def array(obj, dtype=None, shape=None, offset=0, strides=None, formats=None,
         If `dtype` is ``None``, these arguments are passed to
         `numpy.format_parser` to construct a dtype. See that function for
         detailed documentation.
-    copy: bool, optional
+    copy : bool, optional
         Whether to copy the input object (True), or to use a reference instead.
         This option only applies when the input is an ndarray or recarray.
         Defaults to True.
diff --git a/numpy/core/setup_common.py b/numpy/core/setup_common.py
index 2d85e0718..378d93c06 100644
--- a/numpy/core/setup_common.py
+++ b/numpy/core/setup_common.py
@@ -317,8 +317,8 @@ def pyod(filename):
     out : seq
         list of lines of od output
 
-    Note
-    ----
+    Notes
+    -----
     We only implement enough to get the necessary information for long double
     representation, this is not intended as a compatible replacement for od.
     """
diff --git a/numpy/core/shape_base.py b/numpy/core/shape_base.py
index e90358ba5..89e98ab30 100644
--- a/numpy/core/shape_base.py
+++ b/numpy/core/shape_base.py
@@ -607,7 +607,7 @@ def _block_info_recursion(arrays, max_depth, result_ndim, depth=0):
         The arrays to check
     max_depth : list of int
         The number of nested lists
-    result_ndim: int
+    result_ndim : int
         The number of dimensions in thefinal array.
 
     Returns
diff --git a/numpy/core/shape_base.pyi b/numpy/core/shape_base.pyi
index b20598b1a..ec40a8814 100644
--- a/numpy/core/shape_base.pyi
+++ b/numpy/core/shape_base.pyi
@@ -7,9 +7,7 @@ from numpy.typing import ArrayLike
 if sys.version_info >= (3, 8):
     from typing import SupportsIndex
 else:
-    from typing_extensions import Protocol
-    class SupportsIndex(Protocol):
-        def __index__(self) -> int: ...
+    from typing_extensions import SupportsIndex
 
 _ArrayType = TypeVar("_ArrayType", bound=ndarray)
 
diff --git a/numpy/core/src/_simd/_simd.dispatch.c.src b/numpy/core/src/_simd/_simd.dispatch.c.src
index af42192a9..e5b58a8d2 100644
--- a/numpy/core/src/_simd/_simd.dispatch.c.src
+++ b/numpy/core/src/_simd/_simd.dispatch.c.src
@@ -23,7 +23,8 @@
  * #mul_sup   = 1,  1,  1,   1,   1,   1,   0,   0,   1,   1#
  * #div_sup   = 0,  0,  0,   0,   0,   0,   0,   0,   1,   1#
  * #fused_sup = 0,  0,  0,   0,   0,   0,   0,   0,   1,   1#
- * #sum_sup   = 0,  0,  0,   0,   1,   0,   0,   0,   1,   1#
+ * #sumup_sup = 1,  0,  1,   0,   0,   0,   0,   0,   0,   0#
+ * #sum_sup   = 0,  0,  0,   0,   1,   0,   1,   0,   1,   1#
  * #rev64_sup = 1,  1,  1,   1,   1,   1,   0,   0,   1,   0#
  * #ncont_sup = 0,  0,  0,   0,   1,   1,   1,   1,   1,   1#
  * #shl_imm   = 0,  0,  15,  15,  31,  31,  63,  63,  0,   0#
@@ -365,6 +366,10 @@ SIMD_IMPL_INTRIN_3(@intrin@_@sfx@, v@sfx@, v@sfx@, v@sfx@, v@sfx@)
 SIMD_IMPL_INTRIN_1(sum_@sfx@, @sfx@, v@sfx@)
 #endif // sum_sup
 
+#if @sumup_sup@
+SIMD_IMPL_INTRIN_1(sumup_@sfx@, @esfx@, v@sfx@)
+#endif // sumup_sup
+
 /***************************
  * Math
  ***************************/
@@ -452,7 +457,8 @@ static PyMethodDef simd__intrinsics_methods[] = {
  * #mul_sup   = 1,  1,  1,   1,   1,   1,   0,   0,   1,   1#
  * #div_sup   = 0,  0,  0,   0,   0,   0,   0,   0,   1,   1#
  * #fused_sup = 0,  0,  0,   0,   0,   0,   0,   0,   1,   1#
- * #sum_sup   = 0,  0,  0,   0,   1,   0,   0,   0,   1,   1#
+ * #sumup_sup = 1,  0,  1,   0,   0,   0,   0,   0,   0,   0#
+ * #sum_sup   = 0,  0,  0,   0,   1,   0,   1,   0,   1,   1#
  * #rev64_sup = 1,  1,  1,   1,   1,   1,   0,   0,   1,   0#
  * #ncont_sup = 0,  0,  0,   0,   1,   1,   1,   1,   1,   1#
  * #shl_imm   = 0,  0,  15,  15,  31,  31,  63,  63,  0,   0#
@@ -574,6 +580,9 @@ SIMD_INTRIN_DEF(@intrin@_@sfx@)
 SIMD_INTRIN_DEF(sum_@sfx@)
 #endif // sum_sup
 
+#if @sumup_sup@
+SIMD_INTRIN_DEF(sumup_@sfx@)
+#endif // sumup_sup
 /***************************
  * Math
  ***************************/
diff --git a/numpy/core/src/common/simd/avx2/arithmetic.h b/numpy/core/src/common/simd/avx2/arithmetic.h
index 3a3a82798..4b8258759 100644
--- a/numpy/core/src/common/simd/avx2/arithmetic.h
+++ b/numpy/core/src/common/simd/avx2/arithmetic.h
@@ -5,6 +5,7 @@
 #ifndef _NPY_SIMD_AVX2_ARITHMETIC_H
 #define _NPY_SIMD_AVX2_ARITHMETIC_H
 
+#include "../sse/utils.h"
 /***************************
  * Addition
  ***************************/
@@ -117,8 +118,11 @@
     }
 #endif // !NPY_HAVE_FMA3
 
-// Horizontal add: Calculates the sum of all vector elements.
-NPY_FINLINE npy_uint32 npyv_sum_u32(__m256i a)
+/***************************
+ * Summation
+ ***************************/
+// reduce sum across vector
+NPY_FINLINE npy_uint32 npyv_sum_u32(npyv_u32 a)
 {
     __m256i s0 = _mm256_hadd_epi32(a, a);
             s0 = _mm256_hadd_epi32(s0, s0);
@@ -127,7 +131,14 @@ NPY_FINLINE npy_uint32 npyv_sum_u32(__m256i a)
     return _mm_cvtsi128_si32(s1);
 }
 
-NPY_FINLINE float npyv_sum_f32(__m256 a)
+NPY_FINLINE npy_uint64 npyv_sum_u64(npyv_u64 a)
+{
+    __m256i two = _mm256_add_epi64(a, _mm256_shuffle_epi32(a, _MM_SHUFFLE(1, 0, 3, 2)));
+    __m128i one = _mm_add_epi64(_mm256_castsi256_si128(two), _mm256_extracti128_si256(two, 1));
+    return (npy_uint64)npyv128_cvtsi128_si64(one);
+}
+
+NPY_FINLINE float npyv_sum_f32(npyv_f32 a)
 {
     __m256 sum_halves = _mm256_hadd_ps(a, a);
     sum_halves = _mm256_hadd_ps(sum_halves, sum_halves);
@@ -137,7 +148,7 @@ NPY_FINLINE float npyv_sum_f32(__m256 a)
     return _mm_cvtss_f32(sum);
 }
 
-NPY_FINLINE double npyv_sum_f64(__m256d a)
+NPY_FINLINE double npyv_sum_f64(npyv_f64 a)
 {
     __m256d sum_halves = _mm256_hadd_pd(a, a);
     __m128d lo = _mm256_castpd256_pd128(sum_halves);
@@ -146,6 +157,24 @@ NPY_FINLINE double npyv_sum_f64(__m256d a)
     return _mm_cvtsd_f64(sum);
 }
 
+// expand the source vector and performs sum reduce
+NPY_FINLINE npy_uint16 npyv_sumup_u8(npyv_u8 a)
+{
+    __m256i four = _mm256_sad_epu8(a, _mm256_setzero_si256());
+    __m128i two  = _mm_add_epi16(_mm256_castsi256_si128(four), _mm256_extracti128_si256(four, 1));
+    __m128i one  = _mm_add_epi16(two, _mm_unpackhi_epi64(two, two));
+    return (npy_uint16)_mm_cvtsi128_si32(one);
+}
+
+NPY_FINLINE npy_uint32 npyv_sumup_u16(npyv_u16 a)
+{
+    const npyv_u16 even_mask = _mm256_set1_epi32(0x0000FFFF);
+    __m256i even  = _mm256_and_si256(a, even_mask);
+    __m256i odd   = _mm256_srli_epi32(a, 16);
+    __m256i eight = _mm256_add_epi32(even, odd);
+    return npyv_sum_u32(eight);
+}
+
 #endif // _NPY_SIMD_AVX2_ARITHMETIC_H
 
 
diff --git a/numpy/core/src/common/simd/avx512/arithmetic.h b/numpy/core/src/common/simd/avx512/arithmetic.h
index 6f668f439..450da7ea5 100644
--- a/numpy/core/src/common/simd/avx512/arithmetic.h
+++ b/numpy/core/src/common/simd/avx512/arithmetic.h
@@ -6,7 +6,7 @@
 #define _NPY_SIMD_AVX512_ARITHMETIC_H
 
 #include "../avx2/utils.h"
-
+#include "../sse/utils.h"
 /***************************
  * Addition
  ***************************/
@@ -130,7 +130,7 @@ NPY_FINLINE __m512i npyv_mul_u8(__m512i a, __m512i b)
 #define npyv_nmulsub_f64 _mm512_fnmsub_pd
 
 /***************************
- * Reduce Sum: Calculates the sum of all vector elements.
+ * Summation: Calculates the sum of all vector elements.
  * there are three ways to implement reduce sum for AVX512:
  * 1- split(256) /add /split(128) /add /hadd /hadd /extract
  * 2- shuff(cross) /add /shuff(cross) /add /shuff /add /shuff /add /extract
@@ -144,19 +144,29 @@ NPY_FINLINE __m512i npyv_mul_u8(__m512i a, __m512i b)
  * The third one is almost the same as the second one but only works for
  * intel compiler/GCC 7.1/Clang 4, we still need to support older GCC.
  ***************************/
-
-NPY_FINLINE npy_uint32 npyv_sum_u32(npyv_u32 a)
-{
-    __m256i half = _mm256_add_epi32(npyv512_lower_si256(a), npyv512_higher_si256(a));
-    __m128i quarter = _mm_add_epi32(_mm256_castsi256_si128(half), _mm256_extracti128_si256(half, 1));
-    quarter = _mm_hadd_epi32(quarter, quarter);
-    return _mm_cvtsi128_si32(_mm_hadd_epi32(quarter, quarter));
-}
-
+// reduce sum across vector
 #ifdef NPY_HAVE_AVX512F_REDUCE
+    #define npyv_sum_u32 _mm512_reduce_add_epi32
+    #define npyv_sum_u64 _mm512_reduce_add_epi64
     #define npyv_sum_f32 _mm512_reduce_add_ps
     #define npyv_sum_f64 _mm512_reduce_add_pd
 #else
+    NPY_FINLINE npy_uint32 npyv_sum_u32(npyv_u32 a)
+    {
+        __m256i half = _mm256_add_epi32(npyv512_lower_si256(a), npyv512_higher_si256(a));
+        __m128i quarter = _mm_add_epi32(_mm256_castsi256_si128(half), _mm256_extracti128_si256(half, 1));
+        quarter = _mm_hadd_epi32(quarter, quarter);
+        return _mm_cvtsi128_si32(_mm_hadd_epi32(quarter, quarter));
+    }
+
+    NPY_FINLINE npy_uint64 npyv_sum_u64(npyv_u64 a)
+    {
+        __m256i four = _mm256_add_epi64(npyv512_lower_si256(a), npyv512_higher_si256(a));
+        __m256i two = _mm256_add_epi64(four, _mm256_shuffle_epi32(four, _MM_SHUFFLE(1, 0, 3, 2)));
+        __m128i one = _mm_add_epi64(_mm256_castsi256_si128(two), _mm256_extracti128_si256(two, 1));
+        return (npy_uint64)npyv128_cvtsi128_si64(one);
+    }
+
     NPY_FINLINE float npyv_sum_f32(npyv_f32 a)
     {
         __m512 h64   = _mm512_shuffle_f32x4(a, a, _MM_SHUFFLE(3, 2, 3, 2));
@@ -169,6 +179,7 @@ NPY_FINLINE npy_uint32 npyv_sum_u32(npyv_u32 a)
         __m512 sum4  = _mm512_add_ps(sum8, h4);
         return _mm_cvtss_f32(_mm512_castps512_ps128(sum4));
     }
+
     NPY_FINLINE double npyv_sum_f64(npyv_f64 a)
     {
         __m512d h64   = _mm512_shuffle_f64x2(a, a, _MM_SHUFFLE(3, 2, 3, 2));
@@ -181,4 +192,29 @@ NPY_FINLINE npy_uint32 npyv_sum_u32(npyv_u32 a)
     }
 #endif
 
+// expand the source vector and performs sum reduce
+NPY_FINLINE npy_uint16 npyv_sumup_u8(npyv_u8 a)
+{
+#ifdef NPY_HAVE_AVX512BW
+    __m512i eight = _mm512_sad_epu8(a, _mm512_setzero_si512());
+    __m256i four  = _mm256_add_epi16(npyv512_lower_si256(eight), npyv512_higher_si256(eight));
+#else
+    __m256i lo_four = _mm256_sad_epu8(npyv512_lower_si256(a), _mm256_setzero_si256());
+    __m256i hi_four = _mm256_sad_epu8(npyv512_higher_si256(a), _mm256_setzero_si256());
+    __m256i four    = _mm256_add_epi16(lo_four, hi_four);
+#endif
+    __m128i two     = _mm_add_epi16(_mm256_castsi256_si128(four), _mm256_extracti128_si256(four, 1));
+    __m128i one     = _mm_add_epi16(two, _mm_unpackhi_epi64(two, two));
+    return (npy_uint16)_mm_cvtsi128_si32(one);
+}
+
+NPY_FINLINE npy_uint32 npyv_sumup_u16(npyv_u16 a)
+{
+    const npyv_u16 even_mask = _mm512_set1_epi32(0x0000FFFF);
+    __m512i even = _mm512_and_si512(a, even_mask);
+    __m512i odd  = _mm512_srli_epi32(a, 16);
+    __m512i ff   = _mm512_add_epi32(even, odd);
+    return npyv_sum_u32(ff);
+}
+
 #endif // _NPY_SIMD_AVX512_ARITHMETIC_H
diff --git a/numpy/core/src/common/simd/neon/arithmetic.h b/numpy/core/src/common/simd/neon/arithmetic.h
index 1c8bde15a..69a49f571 100644
--- a/numpy/core/src/common/simd/neon/arithmetic.h
+++ b/numpy/core/src/common/simd/neon/arithmetic.h
@@ -131,12 +131,21 @@
     { return vfmsq_f64(vnegq_f64(c), a, b); }
 #endif // NPY_SIMD_F64
 
-// Horizontal add: Calculates the sum of all vector elements.
+/***************************
+ * Summation
+ ***************************/
+// reduce sum across vector
 #if NPY_SIMD_F64
     #define npyv_sum_u32 vaddvq_u32
+    #define npyv_sum_u64 vaddvq_u64
     #define npyv_sum_f32 vaddvq_f32
     #define npyv_sum_f64 vaddvq_f64
 #else
+    NPY_FINLINE npy_uint64 npyv_sum_u64(npyv_u64 a)
+    {
+        return vget_lane_u64(vadd_u64(vget_low_u64(a), vget_high_u64(a)),0);
+    }
+
     NPY_FINLINE npy_uint32 npyv_sum_u32(npyv_u32 a)
     {
         uint32x2_t a0 = vpadd_u32(vget_low_u32(a), vget_high_u32(a)); 
@@ -150,4 +159,24 @@
     }
 #endif
 
+// expand the source vector and performs sum reduce
+#if NPY_SIMD_F64
+    #define npyv_sumup_u8  vaddlvq_u8
+    #define npyv_sumup_u16 vaddlvq_u16
+#else
+    NPY_FINLINE npy_uint16 npyv_sumup_u8(npyv_u8 a)
+    {
+        uint32x4_t t0 = vpaddlq_u16(vpaddlq_u8(a));
+        uint32x2_t t1 = vpadd_u32(vget_low_u32(t0), vget_high_u32(t0));
+        return vget_lane_u32(vpadd_u32(t1, t1), 0);
+    }
+
+    NPY_FINLINE npy_uint32 npyv_sumup_u16(npyv_u16 a)
+    {
+        uint32x4_t t0 = vpaddlq_u16(a);
+        uint32x2_t t1 = vpadd_u32(vget_low_u32(t0), vget_high_u32(t0));
+        return vget_lane_u32(vpadd_u32(t1, t1), 0);
+    }
+#endif
+
 #endif // _NPY_SIMD_NEON_ARITHMETIC_H
diff --git a/numpy/core/src/common/simd/sse/arithmetic.h b/numpy/core/src/common/simd/sse/arithmetic.h
index faf5685d9..c21b7da2d 100644
--- a/numpy/core/src/common/simd/sse/arithmetic.h
+++ b/numpy/core/src/common/simd/sse/arithmetic.h
@@ -148,16 +148,24 @@ NPY_FINLINE __m128i npyv_mul_u8(__m128i a, __m128i b)
     }
 #endif // !NPY_HAVE_FMA3
 
-// Horizontal add: Calculates the sum of all vector elements.
-
-NPY_FINLINE npy_uint32 npyv_sum_u32(__m128i a)
+/***************************
+ * Summation
+ ***************************/
+// reduce sum across vector
+NPY_FINLINE npy_uint32 npyv_sum_u32(npyv_u32 a)
 {
     __m128i t = _mm_add_epi32(a, _mm_srli_si128(a, 8));
     t = _mm_add_epi32(t, _mm_srli_si128(t, 4));
     return (unsigned)_mm_cvtsi128_si32(t);
 }
 
-NPY_FINLINE float npyv_sum_f32(__m128 a)
+NPY_FINLINE npy_uint64 npyv_sum_u64(npyv_u64 a)
+{
+    __m128i one = _mm_add_epi64(a, _mm_unpackhi_epi64(a, a));
+    return (npy_uint64)npyv128_cvtsi128_si64(one);
+}
+
+NPY_FINLINE float npyv_sum_f32(npyv_f32 a)
 {
 #ifdef NPY_HAVE_SSE3
     __m128 sum_halves = _mm_hadd_ps(a, a);
@@ -171,7 +179,7 @@ NPY_FINLINE float npyv_sum_f32(__m128 a)
 #endif
 }
 
-NPY_FINLINE double npyv_sum_f64(__m128d a)
+NPY_FINLINE double npyv_sum_f64(npyv_f64 a)
 {
 #ifdef NPY_HAVE_SSE3
     return _mm_cvtsd_f64(_mm_hadd_pd(a, a));
@@ -180,6 +188,23 @@ NPY_FINLINE double npyv_sum_f64(__m128d a)
 #endif
 }
 
+// expand the source vector and performs sum reduce
+NPY_FINLINE npy_uint16 npyv_sumup_u8(npyv_u8 a)
+{
+    __m128i two = _mm_sad_epu8(a, _mm_setzero_si128());
+    __m128i one = _mm_add_epi16(two, _mm_unpackhi_epi64(two, two));
+    return (npy_uint16)_mm_cvtsi128_si32(one);
+}
+
+NPY_FINLINE npy_uint32 npyv_sumup_u16(npyv_u16 a)
+{
+    const __m128i even_mask = _mm_set1_epi32(0x0000FFFF);
+    __m128i even = _mm_and_si128(a, even_mask);
+    __m128i odd  = _mm_srli_epi32(a, 16);
+    __m128i four = _mm_add_epi32(even, odd);
+    return npyv_sum_u32(four);
+}
+
 #endif // _NPY_SIMD_SSE_ARITHMETIC_H
 
 
diff --git a/numpy/core/src/common/simd/sse/sse.h b/numpy/core/src/common/simd/sse/sse.h
index dc0b62f73..0bb404312 100644
--- a/numpy/core/src/common/simd/sse/sse.h
+++ b/numpy/core/src/common/simd/sse/sse.h
@@ -62,6 +62,7 @@ typedef struct { __m128d val[3]; } npyv_f64x3;
 #define npyv_nlanes_f32 4
 #define npyv_nlanes_f64 2
 
+#include "utils.h"
 #include "memory.h"
 #include "misc.h"
 #include "reorder.h"
diff --git a/numpy/core/src/common/simd/sse/utils.h b/numpy/core/src/common/simd/sse/utils.h
new file mode 100644
index 000000000..c23def11d
--- /dev/null
+++ b/numpy/core/src/common/simd/sse/utils.h
@@ -0,0 +1,19 @@
+#ifndef NPY_SIMD
+    #error "Not a standalone header"
+#endif
+
+#ifndef _NPY_SIMD_SSE_UTILS_H
+#define _NPY_SIMD_SSE_UTILS_H
+
+#if !defined(__x86_64__) && !defined(_M_X64)
+NPY_FINLINE npy_int64 npyv128_cvtsi128_si64(__m128i a)
+{
+    npy_int64 NPY_DECL_ALIGNED(16) idx[2];
+    _mm_store_si128((__m128i *)idx, a);
+    return idx[0];
+}
+#else
+    #define npyv128_cvtsi128_si64 _mm_cvtsi128_si64
+#endif
+
+#endif // _NPY_SIMD_SSE_UTILS_H
diff --git a/numpy/core/src/common/simd/vsx/arithmetic.h b/numpy/core/src/common/simd/vsx/arithmetic.h
index 1288a52a7..7c4e32f27 100644
--- a/numpy/core/src/common/simd/vsx/arithmetic.h
+++ b/numpy/core/src/common/simd/vsx/arithmetic.h
@@ -116,7 +116,14 @@
 #define npyv_nmulsub_f32 vec_nmadd // equivalent to -(a*b + c)
 #define npyv_nmulsub_f64 vec_nmadd
 
-// Horizontal add: Calculates the sum of all vector elements.
+/***************************
+ * Summation
+ ***************************/
+// reduce sum across vector
+NPY_FINLINE npy_uint64 npyv_sum_u64(npyv_u64 a)
+{
+    return vec_extract(vec_add(a, vec_mergel(a, a)), 0);
+}
 
 NPY_FINLINE npy_uint32 npyv_sum_u32(npyv_u32 a)
 {
@@ -135,4 +142,22 @@ NPY_FINLINE double npyv_sum_f64(npyv_f64 a)
     return vec_extract(a, 0) + vec_extract(a, 1);
 }
 
+// expand the source vector and performs sum reduce
+NPY_FINLINE npy_uint16 npyv_sumup_u8(npyv_u8 a)
+{
+    const npyv_u32 zero = npyv_zero_u32();
+    npyv_u32 four = vec_sum4s(a, zero);
+    npyv_s32 one  = vec_sums((npyv_s32)four, (npyv_s32)zero);
+    return (npy_uint16)vec_extract(one, 3);
+}
+
+NPY_FINLINE npy_uint32 npyv_sumup_u16(npyv_u16 a)
+{
+    const npyv_s32 zero = npyv_zero_s32();
+    npyv_u32x2 eight = npyv_expand_u32_u16(a);
+    npyv_u32   four  = vec_add(eight.val[0], eight.val[1]);
+    npyv_s32   one   = vec_sums((npyv_s32)four, zero);
+    return (npy_uint32)vec_extract(one, 3);
+}
+
 #endif // _NPY_SIMD_VSX_ARITHMETIC_H
diff --git a/numpy/core/src/multiarray/compiled_base.c b/numpy/core/src/multiarray/compiled_base.c
index fa5d7db75..de793f87c 100644
--- a/numpy/core/src/multiarray/compiled_base.c
+++ b/numpy/core/src/multiarray/compiled_base.c
@@ -1037,7 +1037,7 @@ arr_ravel_multi_index(PyObject *self, PyObject *args, PyObject *kwds)
 
     NpyIter *iter = NULL;
 
-    char *kwlist[] = {"multi_index", "dims", "mode", "order", NULL};
+    static char *kwlist[] = {"multi_index", "dims", "mode", "order", NULL};
 
     memset(op, 0, sizeof(op));
     dtype[0] = NULL;
@@ -1232,7 +1232,7 @@ arr_unravel_index(PyObject *self, PyObject *args, PyObject *kwds)
     int i, ret_ndim;
     npy_intp ret_dims[NPY_MAXDIMS], ret_strides[NPY_MAXDIMS];
 
-    char *kwlist[] = {"indices", "shape", "order", NULL};
+    static char *kwlist[] = {"indices", "shape", "order", NULL};
 
     if (!PyArg_ParseTupleAndKeywords(args, kwds, "OO&|O&:unravel_index",
                     kwlist,
diff --git a/numpy/core/src/multiarray/ctors.c b/numpy/core/src/multiarray/ctors.c
index 58571b678..ef105ff2d 100644
--- a/numpy/core/src/multiarray/ctors.c
+++ b/numpy/core/src/multiarray/ctors.c
@@ -2124,7 +2124,16 @@ PyArray_FromInterface(PyObject *origin)
 
     if (iface == NULL) {
         if (PyErr_Occurred()) {
-            return NULL;
+            if (PyErr_ExceptionMatches(PyExc_RecursionError) ||
+                    PyErr_ExceptionMatches(PyExc_MemoryError)) {
+                /* RecursionError and MemoryError are considered fatal */
+                return NULL;
+            }
+            /*
+             * This probably be deprecated, but at least shapely raised
+             * a NotImplementedError expecting it to be cleared (gh-17965)
+             */
+            PyErr_Clear();
         }
         return Py_NotImplemented;
     }
@@ -2392,7 +2401,13 @@ PyArray_FromArrayAttr(PyObject *op, PyArray_Descr *typecode, PyObject *context)
     array_meth = PyArray_LookupSpecial_OnInstance(op, "__array__");
     if (array_meth == NULL) {
         if (PyErr_Occurred()) {
-            return NULL;
+            if (PyErr_ExceptionMatches(PyExc_RecursionError) ||
+                PyErr_ExceptionMatches(PyExc_MemoryError)) {
+                /* RecursionError and MemoryError are considered fatal */
+                return NULL;
+            }
+            /* This probably be deprecated. */
+            PyErr_Clear();
         }
         return Py_NotImplemented;
     }
diff --git a/numpy/core/src/multiarray/datetime_busday.c b/numpy/core/src/multiarray/datetime_busday.c
index 2cf157551..f0564146d 100644
--- a/numpy/core/src/multiarray/datetime_busday.c
+++ b/numpy/core/src/multiarray/datetime_busday.c
@@ -934,8 +934,8 @@ NPY_NO_EXPORT PyObject *
 array_busday_offset(PyObject *NPY_UNUSED(self),
                       PyObject *args, PyObject *kwds)
 {
-    char *kwlist[] = {"dates", "offsets", "roll",
-                      "weekmask", "holidays", "busdaycal", "out", NULL};
+    static char *kwlist[] = {"dates", "offsets", "roll",
+                             "weekmask", "holidays", "busdaycal", "out", NULL};
 
     PyObject *dates_in = NULL, *offsets_in = NULL, *out_in = NULL;
 
@@ -1065,8 +1065,8 @@ NPY_NO_EXPORT PyObject *
 array_busday_count(PyObject *NPY_UNUSED(self),
                       PyObject *args, PyObject *kwds)
 {
-    char *kwlist[] = {"begindates", "enddates",
-                      "weekmask", "holidays", "busdaycal", "out", NULL};
+    static char *kwlist[] = {"begindates", "enddates",
+                             "weekmask", "holidays", "busdaycal", "out", NULL};
 
     PyObject *dates_begin_in = NULL, *dates_end_in = NULL, *out_in = NULL;
 
@@ -1210,8 +1210,8 @@ NPY_NO_EXPORT PyObject *
 array_is_busday(PyObject *NPY_UNUSED(self),
                       PyObject *args, PyObject *kwds)
 {
-    char *kwlist[] = {"dates",
-                      "weekmask", "holidays", "busdaycal", "out", NULL};
+    static char *kwlist[] = {"dates",
+                             "weekmask", "holidays", "busdaycal", "out", NULL};
 
     PyObject *dates_in = NULL, *out_in = NULL;
 
diff --git a/numpy/core/src/multiarray/dtypemeta.c b/numpy/core/src/multiarray/dtypemeta.c
index 2931977c2..b2f36d794 100644
--- a/numpy/core/src/multiarray/dtypemeta.c
+++ b/numpy/core/src/multiarray/dtypemeta.c
@@ -407,6 +407,19 @@ string_unicode_common_dtype(PyArray_DTypeMeta *cls, PyArray_DTypeMeta *other)
         Py_INCREF(Py_NotImplemented);
         return (PyArray_DTypeMeta *)Py_NotImplemented;
     }
+    if (other->type_num != NPY_STRING && other->type_num != NPY_UNICODE) {
+        /* Deprecated 2020-12-19, NumPy 1.21. */
+        if (DEPRECATE_FUTUREWARNING(
+                "Promotion of numbers and bools to strings is deprecated. "
+                "In the future, code such as `np.concatenate((['string'], [0]))` "
+                "will raise an error, while `np.asarray(['string', 0])` will "
+                "return an array with `dtype=object`.  To avoid the warning "
+                "while retaining a string result use `dtype='U'` (or 'S').  "
+                "To get an array of Python objects use `dtype=object`. "
+                "(Warning added in NumPy 1.21)") < 0) {
+            return NULL;
+        }
+    }
     /*
      * The builtin types are ordered by complexity (aside from object) here.
      * Arguably, we should not consider numbers and strings "common", but
diff --git a/numpy/core/src/multiarray/einsum_sumprod.c.src b/numpy/core/src/multiarray/einsum_sumprod.c.src
index d1b76de4e..333b8e188 100644
--- a/numpy/core/src/multiarray/einsum_sumprod.c.src
+++ b/numpy/core/src/multiarray/einsum_sumprod.c.src
@@ -20,28 +20,6 @@
 #include "simd/simd.h"
 #include "common.h"
 
-#ifdef NPY_HAVE_SSE_INTRINSICS
-#define EINSUM_USE_SSE1 1
-#else
-#define EINSUM_USE_SSE1 0
-#endif
-
-#ifdef NPY_HAVE_SSE2_INTRINSICS
-#define EINSUM_USE_SSE2 1
-#else
-#define EINSUM_USE_SSE2 0
-#endif
-
-#if EINSUM_USE_SSE1
-#include <xmmintrin.h>
-#endif
-
-#if EINSUM_USE_SSE2
-#include <emmintrin.h>
-#endif
-
-#define EINSUM_IS_SSE_ALIGNED(x) ((((npy_intp)x)&0xf) == 0)
-
 // ARM/Neon don't have instructions for aligned memory access
 #ifdef NPY_HAVE_NEON
     #define EINSUM_IS_ALIGNED(x) 0
@@ -311,6 +289,77 @@ finish_after_unrolled_loop:
 
 #elif @nop@ == 2 && !@complex@
 
+// calculate the multiply and add operation such as dataout = data*scalar+dataout
+static NPY_GCC_OPT_3 void
+@name@_sum_of_products_muladd(@type@ *data, @type@ *data_out, @temptype@ scalar, npy_intp count)
+{
+#if @NPYV_CHK@ // NPYV check for @type@
+    /* Use aligned instructions if possible */
+    const int is_aligned = EINSUM_IS_ALIGNED(data) && EINSUM_IS_ALIGNED(data_out);
+    const int vstep = npyv_nlanes_@sfx@;
+    const npyv_@sfx@ v_scalar = npyv_setall_@sfx@(scalar);
+    /**begin repeat2
+     * #cond = if(is_aligned), else#
+     * #ld = loada, load#
+     * #st = storea, store#
+     */
+    @cond@ {
+        const npy_intp vstepx4 = vstep * 4;
+        for (; count >= vstepx4; count -= vstepx4, data += vstepx4, data_out += vstepx4) {
+            /**begin repeat3
+             * #i = 0, 1, 2, 3#
+             */
+            npyv_@sfx@ b@i@ = npyv_@ld@_@sfx@(data + vstep * @i@);
+            npyv_@sfx@ c@i@ = npyv_@ld@_@sfx@(data_out + vstep * @i@);
+            /**end repeat3**/
+            /**begin repeat3
+             * #i = 0, 1, 2, 3#
+             */
+            npyv_@sfx@ abc@i@ = npyv_muladd_@sfx@(v_scalar, b@i@, c@i@);
+            /**end repeat3**/
+            /**begin repeat3
+             * #i = 0, 1, 2, 3#
+             */
+            npyv_@st@_@sfx@(data_out + vstep * @i@, abc@i@);
+            /**end repeat3**/
+        }
+    }
+    /**end repeat2**/
+    for (; count > 0; count -= vstep, data += vstep, data_out += vstep) {
+        npyv_@sfx@ a = npyv_load_tillz_@sfx@(data, count);
+        npyv_@sfx@ b = npyv_load_tillz_@sfx@(data_out, count);
+        npyv_store_till_@sfx@(data_out, count, npyv_muladd_@sfx@(a, v_scalar, b));
+    }
+    npyv_cleanup();
+#else
+#ifndef NPY_DISABLE_OPTIMIZATION
+    for (; count >= 4; count -= 4, data += 4, data_out += 4) {
+        /**begin repeat2
+         * #i = 0, 1, 2, 3#
+         */
+        const @type@ b@i@ = @from@(data[@i@]);
+        const @type@ c@i@ = @from@(data_out[@i@]);
+        /**end repeat2**/
+        /**begin repeat2
+         * #i = 0, 1, 2, 3#
+         */
+        const @type@ abc@i@ = scalar * b@i@ + c@i@;
+        /**end repeat2**/
+        /**begin repeat2
+         * #i = 0, 1, 2, 3#
+         */
+        data_out[@i@] = @to@(abc@i@);
+        /**end repeat2**/
+    }
+#endif // !NPY_DISABLE_OPTIMIZATION
+    for (; count > 0; --count, ++data, ++data_out) {
+        const @type@ b = @from@(*data);
+        const @type@ c = @from@(*data_out);
+        *data_out = @to@(scalar * b + c);
+    }
+#endif // NPYV check for @type@
+}
+
 static void
 @name@_sum_of_products_contig_two(int nop, char **dataptr,
                                 npy_intp const *NPY_UNUSED(strides), npy_intp count)
@@ -403,242 +452,23 @@ static void
     @type@ *data1 = (@type@ *)dataptr[1];
     @type@ *data_out = (@type@ *)dataptr[2];
 
-#if EINSUM_USE_SSE1 && @float32@
-    __m128 a, b, value0_sse;
-#elif EINSUM_USE_SSE2 && @float64@
-    __m128d a, b, value0_sse;
-#endif
-
     NPY_EINSUM_DBG_PRINT1("@name@_sum_of_products_stride0_contig_outcontig_two (%d)\n",
                                                     (int)count);
-
-/* This is placed before the main loop to make small counts faster */
-finish_after_unrolled_loop:
-    switch (count) {
-/**begin repeat2
- * #i = 6, 5, 4, 3, 2, 1, 0#
- */
-        case @i@+1:
-            data_out[@i@] = @to@(value0 *
-                                 @from@(data1[@i@]) +
-                                 @from@(data_out[@i@]));
-/**end repeat2**/
-        case 0:
-            return;
-    }
-
-#if EINSUM_USE_SSE1 && @float32@
-    value0_sse = _mm_set_ps1(value0);
-
-    /* Use aligned instructions if possible */
-    if (EINSUM_IS_SSE_ALIGNED(data1) && EINSUM_IS_SSE_ALIGNED(data_out)) {
-        /* Unroll the loop by 8 */
-        while (count >= 8) {
-            count -= 8;
-
-/**begin repeat2
- * #i = 0, 4#
- */
-            a = _mm_mul_ps(value0_sse, _mm_load_ps(data1+@i@));
-            b = _mm_add_ps(a, _mm_load_ps(data_out+@i@));
-            _mm_store_ps(data_out+@i@, b);
-/**end repeat2**/
-            data1 += 8;
-            data_out += 8;
-        }
-
-        /* Finish off the loop */
-        if (count > 0) {
-            goto finish_after_unrolled_loop;
-        }
-        else {
-            return;
-        }
-    }
-#elif EINSUM_USE_SSE2 && @float64@
-    value0_sse = _mm_set1_pd(value0);
-
-    /* Use aligned instructions if possible */
-    if (EINSUM_IS_SSE_ALIGNED(data1) && EINSUM_IS_SSE_ALIGNED(data_out)) {
-        /* Unroll the loop by 8 */
-        while (count >= 8) {
-            count -= 8;
-
-/**begin repeat2
- * #i = 0, 2, 4, 6#
- */
-            a = _mm_mul_pd(value0_sse, _mm_load_pd(data1+@i@));
-            b = _mm_add_pd(a, _mm_load_pd(data_out+@i@));
-            _mm_store_pd(data_out+@i@, b);
-/**end repeat2**/
-            data1 += 8;
-            data_out += 8;
-        }
-
-        /* Finish off the loop */
-        if (count > 0) {
-            goto finish_after_unrolled_loop;
-        }
-        else {
-            return;
-        }
-    }
-#endif
-
-    /* Unroll the loop by 8 */
-    while (count >= 8) {
-        count -= 8;
-
-#if EINSUM_USE_SSE1 && @float32@
-/**begin repeat2
- * #i = 0, 4#
- */
-        a = _mm_mul_ps(value0_sse, _mm_loadu_ps(data1+@i@));
-        b = _mm_add_ps(a, _mm_loadu_ps(data_out+@i@));
-        _mm_storeu_ps(data_out+@i@, b);
-/**end repeat2**/
-#elif EINSUM_USE_SSE2 && @float64@
-/**begin repeat2
- * #i = 0, 2, 4, 6#
- */
-        a = _mm_mul_pd(value0_sse, _mm_loadu_pd(data1+@i@));
-        b = _mm_add_pd(a, _mm_loadu_pd(data_out+@i@));
-        _mm_storeu_pd(data_out+@i@, b);
-/**end repeat2**/
-#else
-/**begin repeat2
- * #i = 0, 1, 2, 3, 4, 5, 6, 7#
- */
-        data_out[@i@] = @to@(value0 *
-                             @from@(data1[@i@]) +
-                             @from@(data_out[@i@]));
-/**end repeat2**/
-#endif
-        data1 += 8;
-        data_out += 8;
-    }
-
-    /* Finish off the loop */
-    if (count > 0) {
-        goto finish_after_unrolled_loop;
-    }
+    @name@_sum_of_products_muladd(data1, data_out, value0, count);
+    
 }
 
 static void
 @name@_sum_of_products_contig_stride0_outcontig_two(int nop, char **dataptr,
                                 npy_intp const *NPY_UNUSED(strides), npy_intp count)
 {
-    @type@ *data0 = (@type@ *)dataptr[0];
     @temptype@ value1 = @from@(*(@type@ *)dataptr[1]);
+    @type@ *data0 = (@type@ *)dataptr[0];
     @type@ *data_out = (@type@ *)dataptr[2];
 
-#if EINSUM_USE_SSE1 && @float32@
-    __m128 a, b, value1_sse;
-#elif EINSUM_USE_SSE2 && @float64@
-    __m128d a, b, value1_sse;
-#endif
-
     NPY_EINSUM_DBG_PRINT1("@name@_sum_of_products_contig_stride0_outcontig_two (%d)\n",
                                                     (int)count);
-
-/* This is placed before the main loop to make small counts faster */
-finish_after_unrolled_loop:
-    switch (count) {
-/**begin repeat2
- * #i = 6, 5, 4, 3, 2, 1, 0#
- */
-        case @i@+1:
-            data_out[@i@] = @to@(@from@(data0[@i@])*
-                                 value1  +
-                                 @from@(data_out[@i@]));
-/**end repeat2**/
-        case 0:
-            return;
-    }
-
-#if EINSUM_USE_SSE1 && @float32@
-    value1_sse = _mm_set_ps1(value1);
-
-    /* Use aligned instructions if possible */
-    if (EINSUM_IS_SSE_ALIGNED(data0) && EINSUM_IS_SSE_ALIGNED(data_out)) {
-        /* Unroll the loop by 8 */
-        while (count >= 8) {
-            count -= 8;
-
-/**begin repeat2
- * #i = 0, 4#
- */
-            a = _mm_mul_ps(_mm_load_ps(data0+@i@), value1_sse);
-            b = _mm_add_ps(a, _mm_load_ps(data_out+@i@));
-            _mm_store_ps(data_out+@i@, b);
-/**end repeat2**/
-            data0 += 8;
-            data_out += 8;
-        }
-
-        /* Finish off the loop */
-        goto finish_after_unrolled_loop;
-    }
-#elif EINSUM_USE_SSE2 && @float64@
-    value1_sse = _mm_set1_pd(value1);
-
-    /* Use aligned instructions if possible */
-    if (EINSUM_IS_SSE_ALIGNED(data0) && EINSUM_IS_SSE_ALIGNED(data_out)) {
-        /* Unroll the loop by 8 */
-        while (count >= 8) {
-            count -= 8;
-
-/**begin repeat2
- * #i = 0, 2, 4, 6#
- */
-            a = _mm_mul_pd(_mm_load_pd(data0+@i@), value1_sse);
-            b = _mm_add_pd(a, _mm_load_pd(data_out+@i@));
-            _mm_store_pd(data_out+@i@, b);
-/**end repeat2**/
-            data0 += 8;
-            data_out += 8;
-        }
-
-        /* Finish off the loop */
-        goto finish_after_unrolled_loop;
-    }
-#endif
-
-    /* Unroll the loop by 8 */
-    while (count >= 8) {
-        count -= 8;
-
-#if EINSUM_USE_SSE1 && @float32@
-/**begin repeat2
- * #i = 0, 4#
- */
-        a = _mm_mul_ps(_mm_loadu_ps(data0+@i@), value1_sse);
-        b = _mm_add_ps(a, _mm_loadu_ps(data_out+@i@));
-        _mm_storeu_ps(data_out+@i@, b);
-/**end repeat2**/
-#elif EINSUM_USE_SSE2 && @float64@
-/**begin repeat2
- * #i = 0, 2, 4, 6#
- */
-        a = _mm_mul_pd(_mm_loadu_pd(data0+@i@), value1_sse);
-        b = _mm_add_pd(a, _mm_loadu_pd(data_out+@i@));
-        _mm_storeu_pd(data_out+@i@, b);
-/**end repeat2**/
-#else
-/**begin repeat2
- * #i = 0, 1, 2, 3, 4, 5, 6, 7#
- */
-        data_out[@i@] = @to@(@from@(data0[@i@])*
-                             value1  +
-                             @from@(data_out[@i@]));
-/**end repeat2**/
-#endif
-        data0 += 8;
-        data_out += 8;
-    }
-
-    /* Finish off the loop */
-    goto finish_after_unrolled_loop;
+    @name@_sum_of_products_muladd(data0, data_out, value1, count);
 }
 
 static NPY_GCC_OPT_3 void
diff --git a/numpy/core/src/multiarray/mapping.c b/numpy/core/src/multiarray/mapping.c
index d64962f87..8b9b67387 100644
--- a/numpy/core/src/multiarray/mapping.c
+++ b/numpy/core/src/multiarray/mapping.c
@@ -2328,7 +2328,7 @@ PyArray_MapIterNext(PyArrayMapIterObject *mit)
  * @param Number of indices
  * @param The array that is being iterated
  *
- * @return 0 on success -1 on failure
+ * @return 0 on success -1 on failure (broadcasting or too many fancy indices)
  */
 static int
 mapiter_fill_info(PyArrayMapIterObject *mit, npy_index_info *indices,
@@ -2369,6 +2369,17 @@ mapiter_fill_info(PyArrayMapIterObject *mit, npy_index_info *indices,
             }
         }
 
+        /* Before contunuing, ensure that there are not too fancy indices */
+        if (indices[i].type & HAS_FANCY) {
+            if (NPY_UNLIKELY(j >= NPY_MAXDIMS)) {
+                PyErr_Format(PyExc_IndexError,
+                        "too many advanced (array) indices. This probably "
+                        "means you are indexing with too many booleans. "
+                        "(more than %d found)", NPY_MAXDIMS);
+                return -1;
+            }
+        }
+
         /* (iterating) fancy index, store the iterator */
         if (indices[i].type == HAS_FANCY) {
             mit->fancy_strides[j] = PyArray_STRIDE(arr, curr_dim);
@@ -2655,6 +2666,7 @@ PyArray_MapIterNew(npy_index_info *indices , int index_num, int index_type,
     /* For shape reporting on error */
     PyArrayObject *original_extra_op = extra_op;
 
+    /* NOTE: MAXARGS is the actual limit (2*NPY_MAXDIMS is index number one) */
     PyArrayObject *index_arrays[NPY_MAXDIMS];
     PyArray_Descr *intp_descr;
     PyArray_Descr *dtypes[NPY_MAXDIMS];  /* borrowed references */
diff --git a/numpy/core/src/multiarray/methods.c b/numpy/core/src/multiarray/methods.c
index 8bcf591a2..04ce53ed7 100644
--- a/numpy/core/src/multiarray/methods.c
+++ b/numpy/core/src/multiarray/methods.c
@@ -2289,7 +2289,7 @@ array_dot(PyArrayObject *self, PyObject *args, PyObject *kwds)
 {
     PyObject *a = (PyObject *)self, *b, *o = NULL;
     PyArrayObject *ret;
-    char* kwlist[] = {"b", "out", NULL };
+    static char* kwlist[] = {"b", "out", NULL};
 
 
     if (!PyArg_ParseTupleAndKeywords(args, kwds, "O|O:dot", kwlist, &b, &o)) {
diff --git a/numpy/core/src/multiarray/multiarraymodule.c b/numpy/core/src/multiarray/multiarraymodule.c
index dfd27a0bc..2c00c498b 100644
--- a/numpy/core/src/multiarray/multiarraymodule.c
+++ b/numpy/core/src/multiarray/multiarraymodule.c
@@ -2319,7 +2319,7 @@ array_matrixproduct(PyObject *NPY_UNUSED(dummy), PyObject *args, PyObject* kwds)
 {
     PyObject *v, *a, *o = NULL;
     PyArrayObject *ret;
-    char* kwlist[] = {"a", "b", "out", NULL };
+    static char* kwlist[] = {"a", "b", "out", NULL};
 
     if (!PyArg_ParseTupleAndKeywords(args, kwds, "OO|O:matrixproduct",
                                      kwlist, &a, &v, &o)) {
diff --git a/numpy/core/src/multiarray/scalartypes.c.src b/numpy/core/src/multiarray/scalartypes.c.src
index e480628e7..10f304fe7 100644
--- a/numpy/core/src/multiarray/scalartypes.c.src
+++ b/numpy/core/src/multiarray/scalartypes.c.src
@@ -2711,7 +2711,7 @@ static PyObject *
 
     /* TODO: include type name in error message, which is not @name@ */
     PyObject *obj = NULL;
-    char *kwnames[] = {"", NULL};  /* positional-only */
+    static char *kwnames[] = {"", NULL};  /* positional-only */
     if (!PyArg_ParseTupleAndKeywords(args, kwds, "|O", kwnames, &obj)) {
         return NULL;
     }
@@ -2799,7 +2799,7 @@ static PyObject *
 object_arrtype_new(PyTypeObject *NPY_UNUSED(type), PyObject *args, PyObject *kwds)
 {
     PyObject *obj = Py_None;
-    char *kwnames[] = {"", NULL};  /* positional-only */
+    static char *kwnames[] = {"", NULL};  /* positional-only */
     if (!PyArg_ParseTupleAndKeywords(args, kwds, "|O:object_", kwnames, &obj)) {
         return NULL;
     }
@@ -2825,7 +2825,7 @@ static PyObject *
     PyObject *obj = NULL, *meta_obj = NULL;
     Py@Name@ScalarObject *ret;
 
-    char *kwnames[] = {"", "", NULL};  /* positional-only */
+    static char *kwnames[] = {"", "", NULL};  /* positional-only */
     if (!PyArg_ParseTupleAndKeywords(args, kwds, "|OO", kwnames, &obj, &meta_obj)) {
         return NULL;
     }
@@ -2884,7 +2884,7 @@ bool_arrtype_new(PyTypeObject *NPY_UNUSED(type), PyObject *args, PyObject *kwds)
     PyObject *obj = NULL;
     PyArrayObject *arr;
 
-    char *kwnames[] = {"", NULL};  /* positional-only */
+    static char *kwnames[] = {"", NULL};  /* positional-only */
     if (!PyArg_ParseTupleAndKeywords(args, kwds, "|O:bool_", kwnames, &obj)) {
         return NULL;
     }
@@ -2995,7 +2995,7 @@ void_arrtype_new(PyTypeObject *type, PyObject *args, PyObject *kwds)
     PyObject *obj, *arr;
     PyObject *new = NULL;
 
-    char *kwnames[] = {"", NULL};  /* positional-only */
+    static char *kwnames[] = {"", NULL};  /* positional-only */
     if (!PyArg_ParseTupleAndKeywords(args, kwds, "O:void", kwnames, &obj)) {
         return NULL;
     }
diff --git a/numpy/core/src/multiarray/usertypes.c b/numpy/core/src/multiarray/usertypes.c
index a1ed46f13..15d46800c 100644
--- a/numpy/core/src/multiarray/usertypes.c
+++ b/numpy/core/src/multiarray/usertypes.c
@@ -235,7 +235,7 @@ PyArray_RegisterDataType(PyArray_Descr *descr)
             !PyDict_CheckExact(descr->fields)) {
             PyErr_Format(PyExc_ValueError,
                     "Failed to register dtype for %S: Legacy user dtypes "
-                    "using `NPY_ITEM_IS_POINTER` or `NPY_ITEM_REFCOUNT` are"
+                    "using `NPY_ITEM_IS_POINTER` or `NPY_ITEM_REFCOUNT` are "
                     "unsupported.  It is possible to create such a dtype only "
                     "if it is a structured dtype with names and fields "
                     "hardcoded at registration time.\n"
diff --git a/numpy/core/src/umath/ufunc_type_resolution.c b/numpy/core/src/umath/ufunc_type_resolution.c
index be48be079..c46346118 100644
--- a/numpy/core/src/umath/ufunc_type_resolution.c
+++ b/numpy/core/src/umath/ufunc_type_resolution.c
@@ -111,14 +111,18 @@ raise_no_loop_found_error(
         return -1;
     }
     for (i = 0; i < ufunc->nargs; ++i) {
-        Py_INCREF(dtypes[i]);
-        PyTuple_SET_ITEM(dtypes_tup, i, (PyObject *)dtypes[i]);
+        PyObject *tmp = Py_None;
+        if (dtypes[i] != NULL) {
+            tmp = (PyObject *)dtypes[i];
+        }
+        Py_INCREF(tmp);
+        PyTuple_SET_ITEM(dtypes_tup, i, tmp);
     }
 
     /* produce an error object */
     exc_value = PyTuple_Pack(2, ufunc, dtypes_tup);
     Py_DECREF(dtypes_tup);
-    if (exc_value == NULL){
+    if (exc_value == NULL) {
         return -1;
     }
     PyErr_SetObject(exc_type, exc_value);
@@ -329,10 +333,23 @@ PyUFunc_SimpleBinaryComparisonTypeResolver(PyUFuncObject *ufunc,
     }
 
     if (type_tup == NULL) {
-        /* Input types are the result type */
-        out_dtypes[0] = PyArray_ResultType(2, operands, 0, NULL);
-        if (out_dtypes[0] == NULL) {
-            return -1;
+        /*
+         * DEPRECATED NumPy 1.20, 2020-12.
+         * This check is required to avoid the FutureWarning that
+         * ResultType will give for number->string promotions.
+         * (We never supported flexible dtypes here.)
+         */
+        if (!PyArray_ISFLEXIBLE(operands[0]) &&
+                !PyArray_ISFLEXIBLE(operands[1])) {
+            out_dtypes[0] = PyArray_ResultType(2, operands, 0, NULL);
+            if (out_dtypes[0] == NULL) {
+                return -1;
+            }
+        }
+        else {
+            /* Not doing anything will lead to a loop no found error. */
+            out_dtypes[0] = PyArray_DESCR(operands[0]);
+            Py_INCREF(out_dtypes[0]);
         }
         out_dtypes[1] = out_dtypes[0];
         Py_INCREF(out_dtypes[1]);
@@ -488,6 +505,30 @@ PyUFunc_SimpleUniformOperationTypeResolver(
             out_dtypes[0] = ensure_dtype_nbo(PyArray_DESCR(operands[0]));
         }
         else {
+            int iop;
+            npy_bool has_flexible = 0;
+            npy_bool has_object = 0;
+            for (iop = 0; iop < ufunc->nin; iop++) {
+                if (PyArray_ISOBJECT(operands[iop])) {
+                    has_object = 1;
+                }
+                if (PyArray_ISFLEXIBLE(operands[iop])) {
+                    has_flexible = 1;
+                }
+            }
+            if (NPY_UNLIKELY(has_flexible && !has_object)) {
+                /*
+                 * DEPRECATED NumPy 1.20, 2020-12.
+                 * This check is required to avoid the FutureWarning that
+                 * ResultType will give for number->string promotions.
+                 * (We never supported flexible dtypes here.)
+                 */
+                for (iop = 0; iop < ufunc->nin; iop++) {
+                    out_dtypes[iop] = PyArray_DESCR(operands[iop]);
+                    Py_INCREF(out_dtypes[iop]);
+                }
+                return raise_no_loop_found_error(ufunc, out_dtypes);
+            }
             out_dtypes[0] = PyArray_ResultType(ufunc->nin, operands, 0, NULL);
         }
         if (out_dtypes[0] == NULL) {
diff --git a/numpy/core/tests/test_array_coercion.py b/numpy/core/tests/test_array_coercion.py
index 08b32dfcc..45c792ad2 100644
--- a/numpy/core/tests/test_array_coercion.py
+++ b/numpy/core/tests/test_array_coercion.py
@@ -234,6 +234,7 @@ class TestScalarDiscovery:
 
     # Additionally to string this test also runs into a corner case
     # with datetime promotion (the difference is the promotion order).
+    @pytest.mark.filterwarnings("ignore:Promotion of numbers:FutureWarning")
     def test_scalar_promotion(self):
         for sc1, sc2 in product(scalar_instances(), scalar_instances()):
             sc1, sc2 = sc1.values[0], sc2.values[0]
@@ -702,17 +703,19 @@ class TestArrayLikes:
 
     @pytest.mark.parametrize("attribute",
         ["__array_interface__", "__array__", "__array_struct__"])
-    def test_bad_array_like_attributes(self, attribute):
-        # Check that errors during attribute retrieval are raised unless
-        # they are Attribute errors.
+    @pytest.mark.parametrize("error", [RecursionError, MemoryError])
+    def test_bad_array_like_attributes(self, attribute, error):
+        # RecursionError and MemoryError are considered fatal. All errors
+        # (except AttributeError) should probably be raised in the future,
+        # but shapely made use of it, so it will require a deprecation.
 
         class BadInterface:
             def __getattr__(self, attr):
                 if attr == attribute:
-                    raise RuntimeError
+                    raise error
                 super().__getattr__(attr)
 
-        with pytest.raises(RuntimeError):
+        with pytest.raises(error):
             np.array(BadInterface())
 
     @pytest.mark.parametrize("error", [RecursionError, MemoryError])
diff --git a/numpy/core/tests/test_arrayprint.py b/numpy/core/tests/test_arrayprint.py
index a2703d81b..2c5f1577d 100644
--- a/numpy/core/tests/test_arrayprint.py
+++ b/numpy/core/tests/test_arrayprint.py
@@ -923,6 +923,9 @@ class TestPrintOptions:
         assert_raises(TypeError, np.set_printoptions, threshold='1')
         assert_raises(TypeError, np.set_printoptions, threshold=b'1')
 
+        assert_raises(TypeError, np.set_printoptions, precision='1')
+        assert_raises(TypeError, np.set_printoptions, precision=1.5)
+
 def test_unicode_object_array():
     expected = "array(['é'], dtype=object)"
     x = np.array([u'\xe9'], dtype=object)
diff --git a/numpy/core/tests/test_deprecations.py b/numpy/core/tests/test_deprecations.py
index 5498e1cf9..459a89eaa 100644
--- a/numpy/core/tests/test_deprecations.py
+++ b/numpy/core/tests/test_deprecations.py
@@ -687,16 +687,16 @@ class TestDeprecatedGlobals(_DeprecationTestCase):
         reason='module-level __getattr__ not supported')
     def test_type_aliases(self):
         # from builtins
-        self.assert_deprecated(lambda: np.bool)
-        self.assert_deprecated(lambda: np.int)
-        self.assert_deprecated(lambda: np.float)
-        self.assert_deprecated(lambda: np.complex)
-        self.assert_deprecated(lambda: np.object)
-        self.assert_deprecated(lambda: np.str)
+        self.assert_deprecated(lambda: np.bool(True))
+        self.assert_deprecated(lambda: np.int(1))
+        self.assert_deprecated(lambda: np.float(1))
+        self.assert_deprecated(lambda: np.complex(1))
+        self.assert_deprecated(lambda: np.object())
+        self.assert_deprecated(lambda: np.str('abc'))
 
         # from np.compat
-        self.assert_deprecated(lambda: np.long)
-        self.assert_deprecated(lambda: np.unicode)
+        self.assert_deprecated(lambda: np.long(1))
+        self.assert_deprecated(lambda: np.unicode('abc'))
 
 
 class TestMatrixInOuter(_DeprecationTestCase):
@@ -1100,3 +1100,41 @@ class TestNoseDecoratorsDeprecated(_DeprecationTestCase):
                 count += 1
             assert_(count == 3)
         self.assert_deprecated(_test_parametrize)
+
+
+class TestStringPromotion(_DeprecationTestCase):
+    # Deprecated 2020-12-19, NumPy 1.21
+    warning_cls = FutureWarning
+    message = "Promotion of numbers and bools to strings is deprecated."
+
+    @pytest.mark.parametrize("dtype", "?bhilqpBHILQPefdgFDG")
+    @pytest.mark.parametrize("string_dt", ["S", "U"])
+    def test_deprecated(self, dtype, string_dt):
+        self.assert_deprecated(lambda: np.promote_types(dtype, string_dt))
+
+        # concatenate has to be able to promote to find the result dtype:
+        arr1 = np.ones(3, dtype=dtype)
+        arr2 = np.ones(3, dtype=string_dt)
+        self.assert_deprecated(lambda: np.concatenate((arr1, arr2), axis=0))
+        self.assert_deprecated(lambda: np.concatenate((arr1, arr2), axis=None))
+
+        # coercing to an array is similar, but will fall-back to `object`
+        # (when raising the FutureWarning, this already happens)
+        self.assert_deprecated(lambda: np.array([arr1[0], arr2[0]]),
+                               exceptions=())
+
+    @pytest.mark.parametrize("dtype", "?bhilqpBHILQPefdgFDG")
+    @pytest.mark.parametrize("string_dt", ["S", "U"])
+    def test_not_deprecated(self, dtype, string_dt):
+        # The ufunc type resolvers run into this, but giving a futurewarning
+        # here is unnecessary (it ends up as an error anyway), so test that
+        # no warning is given:
+        arr1 = np.ones(3, dtype=dtype)
+        arr2 = np.ones(3, dtype=string_dt)
+
+        # Adding two arrays uses result_type normally, which would fail:
+        with pytest.raises(TypeError):
+            self.assert_not_deprecated(lambda: arr1 + arr2)
+        # np.equal uses a different type resolver:
+        with pytest.raises(TypeError):
+            self.assert_not_deprecated(lambda: np.equal(arr1, arr2))
diff --git a/numpy/core/tests/test_half.py b/numpy/core/tests/test_half.py
index 1b6fd21e1..449a01d21 100644
--- a/numpy/core/tests/test_half.py
+++ b/numpy/core/tests/test_half.py
@@ -71,8 +71,10 @@ class TestHalf:
     def test_half_conversion_to_string(self, string_dt):
         # Currently uses S/U32 (which is sufficient for float32)
         expected_dt = np.dtype(f"{string_dt}32")
-        assert np.promote_types(np.float16, string_dt) == expected_dt
-        assert np.promote_types(string_dt, np.float16) == expected_dt
+        with pytest.warns(FutureWarning):
+            assert np.promote_types(np.float16, string_dt) == expected_dt
+        with pytest.warns(FutureWarning):
+            assert np.promote_types(string_dt, np.float16) == expected_dt
 
         arr = np.ones(3, dtype=np.float16).astype(string_dt)
         assert arr.dtype == expected_dt
diff --git a/numpy/core/tests/test_indexing.py b/numpy/core/tests/test_indexing.py
index 667c49240..73dbc429c 100644
--- a/numpy/core/tests/test_indexing.py
+++ b/numpy/core/tests/test_indexing.py
@@ -3,6 +3,8 @@ import warnings
 import functools
 import operator
 
+import pytest
+
 import numpy as np
 from numpy.core._multiarray_tests import array_indexing
 from itertools import product
@@ -547,6 +549,21 @@ class TestIndexing:
         assert_array_equal(arr[0], np.array("asdfg", dtype="c"))
         assert arr[0, 1] == b"s"  # make sure not all were set to "a" for both
 
+    @pytest.mark.parametrize("index",
+            [True, False, np.array([0])])
+    @pytest.mark.parametrize("num", [32, 40])
+    @pytest.mark.parametrize("original_ndim", [1, 32])
+    def test_too_many_advanced_indices(self, index, num, original_ndim):
+        # These are limitations based on the number of arguments we can process.
+        # For `num=32` (and all boolean cases), the result is actually define;
+        # but the use of NpyIter (NPY_MAXARGS) limits it for technical reasons.
+        arr = np.ones((1,) * original_ndim)
+        with pytest.raises(IndexError):
+            arr[(index,) * num]
+        with pytest.raises(IndexError):
+            arr[(index,) * num] = 1.
+
+
 class TestFieldIndexing:
     def test_scalar_return_type(self):
         # Field access on an array should return an array, even if it
diff --git a/numpy/core/tests/test_nditer.py b/numpy/core/tests/test_nditer.py
index 5e6472ae5..94f61baca 100644
--- a/numpy/core/tests/test_nditer.py
+++ b/numpy/core/tests/test_nditer.py
@@ -1365,6 +1365,7 @@ def test_iter_copy():
 
 @pytest.mark.parametrize("dtype", np.typecodes["All"])
 @pytest.mark.parametrize("loop_dtype", np.typecodes["All"])
+@pytest.mark.filterwarnings("ignore::numpy.ComplexWarning")
 def test_iter_copy_casts(dtype, loop_dtype):
     # Ensure the dtype is never flexible:
     if loop_dtype.lower() == "m":
diff --git a/numpy/core/tests/test_numeric.py b/numpy/core/tests/test_numeric.py
index 6de9e3764..cdfecc0f5 100644
--- a/numpy/core/tests/test_numeric.py
+++ b/numpy/core/tests/test_numeric.py
@@ -847,10 +847,12 @@ class TestTypes:
         assert_equal(np.promote_types('<i8', '<i8'), np.dtype('i8'))
         assert_equal(np.promote_types('>i8', '>i8'), np.dtype('i8'))
 
-        assert_equal(np.promote_types('>i8', '>U16'), np.dtype('U21'))
-        assert_equal(np.promote_types('<i8', '<U16'), np.dtype('U21'))
-        assert_equal(np.promote_types('>U16', '>i8'), np.dtype('U21'))
-        assert_equal(np.promote_types('<U16', '<i8'), np.dtype('U21'))
+        with pytest.warns(FutureWarning,
+                match="Promotion of numbers and bools to strings"):
+            assert_equal(np.promote_types('>i8', '>U16'), np.dtype('U21'))
+            assert_equal(np.promote_types('<i8', '<U16'), np.dtype('U21'))
+            assert_equal(np.promote_types('>U16', '>i8'), np.dtype('U21'))
+            assert_equal(np.promote_types('<U16', '<i8'), np.dtype('U21'))
 
         assert_equal(np.promote_types('<S5', '<U8'), np.dtype('U8'))
         assert_equal(np.promote_types('>S5', '>U8'), np.dtype('U8'))
@@ -897,32 +899,38 @@ class TestTypes:
             promote_types = np.promote_types
 
         S = string_dtype
-        # Promote numeric with unsized string:
-        assert_equal(promote_types('bool', S), np.dtype(S+'5'))
-        assert_equal(promote_types('b', S), np.dtype(S+'4'))
-        assert_equal(promote_types('u1', S), np.dtype(S+'3'))
-        assert_equal(promote_types('u2', S), np.dtype(S+'5'))
-        assert_equal(promote_types('u4', S), np.dtype(S+'10'))
-        assert_equal(promote_types('u8', S), np.dtype(S+'20'))
-        assert_equal(promote_types('i1', S), np.dtype(S+'4'))
-        assert_equal(promote_types('i2', S), np.dtype(S+'6'))
-        assert_equal(promote_types('i4', S), np.dtype(S+'11'))
-        assert_equal(promote_types('i8', S), np.dtype(S+'21'))
-        # Promote numeric with sized string:
-        assert_equal(promote_types('bool', S+'1'), np.dtype(S+'5'))
-        assert_equal(promote_types('bool', S+'30'), np.dtype(S+'30'))
-        assert_equal(promote_types('b', S+'1'), np.dtype(S+'4'))
-        assert_equal(promote_types('b', S+'30'), np.dtype(S+'30'))
-        assert_equal(promote_types('u1', S+'1'), np.dtype(S+'3'))
-        assert_equal(promote_types('u1', S+'30'), np.dtype(S+'30'))
-        assert_equal(promote_types('u2', S+'1'), np.dtype(S+'5'))
-        assert_equal(promote_types('u2', S+'30'), np.dtype(S+'30'))
-        assert_equal(promote_types('u4', S+'1'), np.dtype(S+'10'))
-        assert_equal(promote_types('u4', S+'30'), np.dtype(S+'30'))
-        assert_equal(promote_types('u8', S+'1'), np.dtype(S+'20'))
-        assert_equal(promote_types('u8', S+'30'), np.dtype(S+'30'))
-        # Promote with object:
-        assert_equal(promote_types('O', S+'30'), np.dtype('O'))
+        
+        with pytest.warns(FutureWarning,
+                match="Promotion of numbers and bools to strings") as record:
+            # Promote numeric with unsized string:
+            assert_equal(promote_types('bool', S), np.dtype(S+'5'))
+            assert_equal(promote_types('b', S), np.dtype(S+'4'))
+            assert_equal(promote_types('u1', S), np.dtype(S+'3'))
+            assert_equal(promote_types('u2', S), np.dtype(S+'5'))
+            assert_equal(promote_types('u4', S), np.dtype(S+'10'))
+            assert_equal(promote_types('u8', S), np.dtype(S+'20'))
+            assert_equal(promote_types('i1', S), np.dtype(S+'4'))
+            assert_equal(promote_types('i2', S), np.dtype(S+'6'))
+            assert_equal(promote_types('i4', S), np.dtype(S+'11'))
+            assert_equal(promote_types('i8', S), np.dtype(S+'21'))
+            # Promote numeric with sized string:
+            assert_equal(promote_types('bool', S+'1'), np.dtype(S+'5'))
+            assert_equal(promote_types('bool', S+'30'), np.dtype(S+'30'))
+            assert_equal(promote_types('b', S+'1'), np.dtype(S+'4'))
+            assert_equal(promote_types('b', S+'30'), np.dtype(S+'30'))
+            assert_equal(promote_types('u1', S+'1'), np.dtype(S+'3'))
+            assert_equal(promote_types('u1', S+'30'), np.dtype(S+'30'))
+            assert_equal(promote_types('u2', S+'1'), np.dtype(S+'5'))
+            assert_equal(promote_types('u2', S+'30'), np.dtype(S+'30'))
+            assert_equal(promote_types('u4', S+'1'), np.dtype(S+'10'))
+            assert_equal(promote_types('u4', S+'30'), np.dtype(S+'30'))
+            assert_equal(promote_types('u8', S+'1'), np.dtype(S+'20'))
+            assert_equal(promote_types('u8', S+'30'), np.dtype(S+'30'))
+            # Promote with object:
+            assert_equal(promote_types('O', S+'30'), np.dtype('O'))
+
+        assert len(record) == 22  # each string promotion gave one warning
+
 
     @pytest.mark.parametrize(["dtype1", "dtype2"],
             [[np.dtype("V6"), np.dtype("V10")],
@@ -972,6 +980,7 @@ class TestTypes:
             assert res.isnative
 
     @pytest.mark.slow
+    @pytest.mark.filterwarnings('ignore:Promotion of numbers:FutureWarning')
     @pytest.mark.parametrize(["dtype1", "dtype2"],
             itertools.product(
                 list(np.typecodes["All"]) +
diff --git a/numpy/core/tests/test_regression.py b/numpy/core/tests/test_regression.py
index 831e48e8b..5faa9923c 100644
--- a/numpy/core/tests/test_regression.py
+++ b/numpy/core/tests/test_regression.py
@@ -782,7 +782,9 @@ class TestRegression:
         # Ticket #514
         s = "aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa"
         t = []
-        np.hstack((t, s))
+        with pytest.warns(FutureWarning,
+                match="Promotion of numbers and bools to strings"):
+            np.hstack((t, s))
 
     def test_arr_transpose(self):
         # Ticket #516
diff --git a/numpy/core/tests/test_shape_base.py b/numpy/core/tests/test_shape_base.py
index 9922c9173..a0c72f9d0 100644
--- a/numpy/core/tests/test_shape_base.py
+++ b/numpy/core/tests/test_shape_base.py
@@ -256,7 +256,7 @@ class TestConcatenate:
         r = np.concatenate((a, b), axis=None)
         assert_equal(r.size, a.size + len(b))
         assert_equal(r.dtype, a.dtype)
-        r = np.concatenate((a, b, c), axis=None)
+        r = np.concatenate((a, b, c), axis=None, dtype="U")
         d = array(['0.0', '1.0', '2.0', '3.0',
                    '0', '1', '2', 'x'])
         assert_array_equal(r, d)
@@ -377,7 +377,8 @@ class TestConcatenate:
         # Note that U0 and S0 should be deprecated eventually and changed to
         # actually give the empty string result (together with `np.array`)
         res = np.concatenate(arrs, axis=axis, dtype=string_dt, casting="unsafe")
-        assert res.dtype == np.promote_types("d", string_dt)
+        # The actual dtype should be identical to a cast (of a double array):
+        assert res.dtype == np.array(1.).astype(string_dt).dtype
 
     @pytest.mark.parametrize("axis", [None, 0])
     def test_string_dtype_does_not_inspect(self, axis):
diff --git a/numpy/core/tests/test_simd.py b/numpy/core/tests/test_simd.py
index 23a5bb6c3..1d1a111be 100644
--- a/numpy/core/tests/test_simd.py
+++ b/numpy/core/tests/test_simd.py
@@ -736,11 +736,9 @@ class _SIMD_ALL(_Test_Utility):
     def test_arithmetic_reduce_sum(self):
         """
         Test reduce sum intrinics:
-            npyv_sum_u32
-            npyv_sum_f32
-            npyv_sum_f64
+            npyv_sum_##sfx
         """
-        if self.sfx not in ("u32", "f32", "f64"):
+        if self.sfx not in ("u32", "u64", "f32", "f64"):
             return
         # reduce sum
         data = self._data()
@@ -750,6 +748,21 @@ class _SIMD_ALL(_Test_Utility):
         vsum = self.sum(vdata)
         assert vsum == data_sum
 
+    def test_arithmetic_reduce_sumup(self):
+        """
+        Test extend reduce sum intrinics:
+            npyv_sumup_##sfx
+        """
+        if self.sfx not in ("u8", "u16"):
+            return
+        rdata = (0, self.nlanes, self._int_min(), self._int_max()-self.nlanes)
+        for r in rdata:
+            data = self._data(r)
+            vdata = self.load(data)
+            data_sum = sum(data)
+            vsum = self.sumup(vdata)
+            assert vsum == data_sum
+
     def test_mask_conditional(self):
         """
         Conditional addition and subtraction for all supported data types.