CLaSH.Sized.Index

Documentation

data Index n

Arbitrary-bounded unsigned integer represented by ceil(log_2(n)) bits.

Given an upper bound n, an Index n number has a range of: [0 .. n-1]

>>> maxBound :: Index 8
7
>>> minBound :: Index 8
0
>>> read (show (maxBound :: Index 8)) :: Index 8
7
>>> 1 + 2 :: Index 8
3
>>> 2 + 6 :: Index 8
*** Exception: CLaSH.Sized.Index: result 8 is out of bounds: [0..7]
>>> 1 - 3 :: Index 8
*** Exception: CLaSH.Sized.Index: result -2 is out of bounds: [0..7]
>>> 2 * 3 :: Index 8
6
>>> 2 * 4 :: Index 8
*** Exception: CLaSH.Sized.Index: result 8 is out of bounds: [0..7]

Instances

Resize Index

KnownNat n => Bounded (Index n)

KnownNat n => Enum (Index n)

The functions: enumFrom, enumFromThen, enumFromTo, and enumFromThenTo, are not synthesisable.

Eq (Index n)

KnownNat n => Integral (Index n)

Typeable Nat n => Data (Index n)

KnownNat n => Num (Index n)

Operators report an error on overflow and underflow

Ord (Index n)

KnownNat n => Read (Index n)

None of the Read class' methods are synthesisable.

KnownNat n => Real (Index n)

Show (Index n)

KnownNat n => Arbitrary (Index n)

KnownNat n => CoArbitrary (Index n)

KnownNat n => Default (Index n)

NFData (Index n)

KnownNat n => Lift (Index n)

KnownNat n => BitPack (Index n)

Bundle (Index n)

ExtendingNum (Index m) (Index n)

type Unbundled' clk (Index n) = Signal' clk (Index n)

type BitSize (Index n) = CLog 2 n

type AResult (Index m) (Index n) = Index ((-) ((+) m n) 1)

type MResult (Index m) (Index n) = Index ((+) (* ((-) m 1) ((-) n 1)) 1)

bv2i :: KnownNat (2 ^ n) => BitVector n -> Index (2 ^ n)

An alternative implementation of unpack for the Index data type; for when you know the size of the BitVector and want to determine the size of the Index.

That is, the type of unpack is:

unpack :: BitVector (CLog 2 n) -> Index n

And is useful when you know the size of the Index, and want to get a value from a BitVector that is large enough (CLog 2 n) enough to hold an Index. Note that unpack can fail at run-time when the value inside the BitVector is higher than 'n-1'.

bv2i on the other hand will never fail at run-time, because the BitVector argument determines the size.