Skip to content

[WIP] Implement RFC 41: lib.fixed #1578

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Draft
wants to merge 1 commit into
base: main
Choose a base branch
from
Draft
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
328 changes: 328 additions & 0 deletions amaranth/lib/fixed.py
Original file line number Diff line number Diff line change
@@ -0,0 +1,328 @@
# Based on latest iteration of fixed point types RFC, which
# is an effort undertaken by the Amaranth community, as well
# as an early (incomplete) RFC implementation by zyp@
#
# RFC (community): https://github.com/amaranth-lang/rfcs/pull/41
# Early implementation (zyp@): https://github.com/amaranth-lang/amaranth/pull/1005
#
# SPDX-License-Identifier: BSD-3-Clause

from .. import hdl, Mux
from ..utils import bits_for

__all__ = ["Shape", "SQ", "UQ", "Value", "Const"]

class Shape(hdl.ShapeCastable):

def __init__(self, shape, f_bits=0):
self._storage_shape = shape
self.i_bits, self.f_bits = shape.width-f_bits, f_bits
if self.i_bits < 0 or self.f_bits < 0:
raise TypeError(f"fixed.Shape may not be created with negative bit widths (i_bits={self.i_bits}, f_bits={self.f_bits})")
if shape.signed and self.i_bits == 0:
raise TypeError(f"A signed fixed.Shape cannot be created with i_bits=0")
if self.i_bits + self.f_bits == 0:
raise TypeError(f"fixed.Shape may not be created with zero width")

@property
def signed(self):
return self._storage_shape.signed

@staticmethod
def cast(shape, f_bits=0):
if not isinstance(shape, hdl.Shape):
raise TypeError(f"Object {shape!r} cannot be converted to a fixed.Shape")
return Shape(shape, f_bits)

def const(self, value):
if value is None:
value = 0
return Const(value, self)._target

def as_shape(self):
return self._storage_shape

def __call__(self, target):
return Value(self, target)

def min(self):
c = Const(0, self)
c._value = c._min_value()
return c

def max(self):
c = Const(0, self)
c._value = c._max_value()
return c

def from_bits(self, raw):
c = Const(0, self)
c._value = raw
if self.signed and raw > c._max_value():
# 2s complement signed value, but `raw` was unsigned.
c._value = c._min_value() + c._value - c._max_value() - 1
if c._value < c._min_value() or c._value > c._max_value():
raise ValueError(
f"{raw} outside expected range {c._min_value()}, {c._max_value()}")
return c

def __repr__(self):
return f"fixed.Shape({self._storage_shape}, f_bits={self.f_bits})"


class SQ(Shape):
def __init__(self, i_bits, f_bits):
super().__init__(hdl.Shape(i_bits + f_bits, signed=True), f_bits)


class UQ(Shape):
def __init__(self, i_bits, f_bits):
super().__init__(hdl.Shape(i_bits + f_bits, signed=False), f_bits)


class Value(hdl.ValueCastable):
def __init__(self, shape, target):
self._shape = shape
if self.signed and not target.shape().signed:
# When methods bit-pick or concatenate to
# the _target of a Value, and then use this
# to reconstruct a Value, we may lose the
# signedness of its underlying _target.
self._target = target.as_signed()
else:
self._target = target

@property
def signed(self):
return self._shape.signed

@staticmethod
def cast(value, f_bits=0):
return Shape.cast(value.shape(), f_bits)(value)

@property
def i_bits(self):
return self._shape.i_bits

@property
def f_bits(self):
return self._shape.f_bits

def shape(self):
return self._shape

def as_value(self):
return self._target

def eq(self, other):
if isinstance(other, hdl.Value):
return self.as_value().eq(other)
elif isinstance(other, int) or isinstance(other, float):
other = Const(other, self.shape())
elif not isinstance(other, Value):
raise TypeError(f"Object {other!r} cannot be converted to a fixed.Value")
other = other.reshape(self.f_bits)
return self.as_value().eq(other.as_value())

def reshape(self, f_bits):
# If we're increasing precision, extend with more fractional bits. If we're
# reducing precision, truncate bits.
shape = hdl.Shape(self.i_bits + f_bits, signed=self.signed)
if f_bits > self.f_bits:
result = Shape(shape, f_bits)(hdl.Cat(hdl.Const(0, f_bits - self.f_bits), self.as_value()))
else:
result = Shape(shape, f_bits)(self.as_value()[self.f_bits - f_bits:])
return result

def truncate(self, f_bits=0):
if f_bits > self.f_bits:
raise ValueError(
f"`.truncate(f_bits={f_bits}) exceeds the underlying type's f_bits={self.f_bits}. "
"Use `.reshape()` to instead extend `f_bits`."
)
return self.reshape(f_bits)

def clamp(self, lo, hi):
if not isinstance(lo, Value) or not isinstance(hi, Value):
raise TypeError(f"Cannot `clamp` as lo, hi are not fixed.Value")
lo = lo.reshape(self.f_bits)
hi = hi.reshape(self.f_bits)
return Value(self.shape(), Mux(
self > hi, hi,
Mux(self < lo, lo, self)
))

def saturate(self, shape):
if not isinstance(shape, Shape):
raise TypeError(f"Cannot `saturate` to bounds of {shape!r} as it is not a fixed.Shape")
if not shape.i_bits <= self.i_bits:
raise ValueError(f"Cannot `saturate`: shape.i_bits={shape.i_bits} > self.i_bits={self.i_bits} would have no effect.")
clamped = self.reshape(shape.f_bits).clamp(shape.min(), shape.max())
return Value(shape, clamped.as_value())

def _binary_op(self, rhs, operator, callable_f_bits = lambda a, b: max(a, b), pre_reshape=True, post_cast=True):
if isinstance(rhs, hdl.Value):
rhs = Value.cast(rhs)
elif isinstance(rhs, int):
rhs = Const(rhs)
elif not isinstance(rhs, Value):
raise TypeError(f"Object {rhs!r} cannot be converted to a fixed.Value")
f_bits = callable_f_bits(self.f_bits, rhs.f_bits)
if pre_reshape:
lhs = self.reshape(f_bits)
rhs = rhs.reshape(f_bits)
else:
lhs = self
value = getattr(lhs.as_value(), operator)(rhs.as_value())
return Value.cast(value, f_bits) if post_cast else value

def __mul__(self, other):
return self._binary_op(other, '__mul__', lambda a, b: a + b, pre_reshape=False)

__rmul__ = __mul__

def __add__(self, other):
return self._binary_op(other, '__add__')

__radd__ = __add__

def __sub__(self, other):
return self._binary_op(other, '__sub__')

def __rsub__(self, other):
return -self.__sub__(other)

def __pos__(self):
return self

def __neg__(self):
return Value.cast(-self.as_value(), self.f_bits)

def __abs__(self):
return Value.cast(abs(self.as_value()), self.f_bits)

def __lshift__(self, other):
if isinstance(other, int):
if other < 0:
raise ValueError("Shift amount cannot be negative")

if other > self.f_bits:
value = hdl.Cat(hdl.Const(0, other - self.f_bits), self.as_value())
return Value.cast(value.as_signed() if self.signed else value)
else:
return Value.cast(self.as_value(), self.f_bits - other)
elif not isinstance(other, hdl.Value):
raise TypeError("Shift amount must be an integer value")
if other.signed:
raise TypeError("Shift amount must be unsigned")
return Value.cast(self.as_value() << other, self.f_bits)

def __rshift__(self, other):
if isinstance(other, int):
if other < 0:
raise ValueError("Shift amount cannot be negative")
# Extend f_bits by fixed shift amount.
i_bits = self.i_bits - other
f_bits = self.f_bits + other
numerator = self.as_value()
elif isinstance(other, hdl.Value):
if other.shape().signed:
raise TypeError("Shift amount must be unsigned")
# Extend by maximum possible shift represented by hdl.Value.
f_bits = self.f_bits + 2**other.shape().width - 1
i_bits = self.i_bits - (f_bits - self.f_bits)
numerator = self.reshape(f_bits).as_value() >> other
else:
raise TypeError("Shift amount must be an integer value")
# Always keep at least 1 sign bit and prohibit negative i_bits.
# TODO: should we concat to _target for sign extension? (likely unnecessary)
if self.signed:
return SQ(max(1, i_bits), f_bits)(numerator)
else:
return UQ(max(0, i_bits), f_bits)(numerator)

def _binary_compare(self, other, operator):
return self._binary_op(other, operator, post_cast=False)

def __lt__(self, other):
return self._binary_compare(other, '__lt__')

def __ge__(self, other):
return self._binary_compare(other, '__ge__')

def __gt__(self, other):
return self._binary_compare(other, '__gt__')

def __le__(self, other):
return self._binary_compare(other, '__le__')

def __eq__(self, other):
return self._binary_compare(other, '__eq__')

def __repr__(self):
return f"fixed.{'SQ' if self.signed else 'UQ'}({self.i_bits}, {self.f_bits}) {self._target!r}"


class Const(Value):
def __init__(self, value, shape=None, clamp=False):

if isinstance(value, float) or isinstance(value, int):
num, den = value.as_integer_ratio()
elif isinstance(value, Const):
# FIXME: Memory inits seem to construct a fixed.Const with fixed.Const
self._shape = value._shape
self._value = value._value
return
else:
raise TypeError(f"Object {value!r} cannot be converted to a fixed.Const")

# Determine smallest possible shape if not already selected.
if shape is None:
signed = num < 0
f_bits = bits_for(den) - 1
i_bits = max(0, bits_for(abs(num)) - f_bits)
shape = SQ(i_bits+1, f_bits) if signed else UQ(i_bits, f_bits)

# Scale value to given precision.
if 2**shape.f_bits > den:
num *= 2**shape.f_bits // den
elif 2**shape.f_bits < den:
num = round(num / (den // 2**shape.f_bits))
value = num

self._shape = shape

if value > self._max_value():
if clamp:
value = self._max_value()
else:
raise ValueError(f"Constant {value!r} does not fit in {shape!r}.")

if value < self._min_value():
if clamp:
value = self._min_value()
else:
raise ValueError(f"Constant {value!r} does not fit in {shape!r}. ")

self._value = value

def _max_value(self):
return 2**(self._shape.i_bits +
self._shape.f_bits - (1 if self.signed else 0)) - 1

def _min_value(self):
if self._shape.signed:
return -1 * 2**(self._shape.i_bits +
self._shape.f_bits - 1)
else:
return 0

@property
def _target(self):
return hdl.Const(self._value, self._shape.as_shape())

def as_integer_ratio(self):
return self._value, 2**self.f_bits

def as_float(self):
return self._value / 2**self.f_bits
Loading
Loading