Skip to content

Latest commit

 

History

History
102 lines (86 loc) · 4.63 KB

fpadd_example.md

File metadata and controls

102 lines (86 loc) · 4.63 KB

DSLX Example: Floating-point addition

This document explains how floating-point addition is implemented in DSLX, along with a discussion on the algorithm and how it maps to DSLX.

This document assumes familiarity with floating-point numbers in general (layout, precision, error, etc.).

[TOC]

Background

Floating-point addition, like any FP operation, is much more complicated than integer addition, and has many more steps.

  1. Expand significands: Floating-point operations are computed with bits beyond that in their normal representations for increased precision. For IEEE 754 numbers, there are three extra, called the guard, rounding and sticky bits. The first two behave normally, but the last, the "sticky" bit, is special. During shift operations (below), if a "1" value is ever shifted into the sticky bit, it "sticks" - the bit will remain "1" through any further shift operations. In this step, the significands are expanded by these three bits.
  2. Align significands: To ensure that significands are added with appropriate magnitudes, they must be aligned according to their exponents. To do so, the smaller significant needs to be shifted to the right (each right shift is equivalent to increasing the exponent by one).
    • The extra precision bits are populated in this shift.
    • As part of this step, the leading 1 bit... and a sign bit Note: The sticky bit is calculated and applied in this step.
  3. Sign-adjustment: if the significands differ in sign, then the significand with the smaller initial exponent needs to be (two's complement) negated.
  4. Add the significands and capture the carry bit. Note that, if the signs of the significands differs, then this could result in higher bits being cleared.
  5. Normalize the significands: Shift the result so that the leading '1' is present in the proper space. This means shifting right one place if the result set the carry bit, and to the left some number of places if high bits were cleared.
    • The sticky bit must be preserved in any of these shifts!
  6. Rounding: Here, the extra precision bits are examined to determine if the result significand's last bit should be rounded up. IEEE 754 supports five rounding modes:
    • Round towards 0: just chop off the extra precision bits.
    • Round towards +infinity: round up if any extra precision bits are set.
    • Round towards -infinity: round down if any extra precision bits are set.
    • Round to nearest, ties away from zero: Rounds to the nearest value. In cases where the extra precision bits are halfway between values, i.e., 0b100, then the result is rounded up for positive numbers and down for negative ones.
    • Round to nearest, ties to even: Rounds to the nearest value. In cases where the extra precision bits are halfway between values, then the result is rounded in whichever direction causes the LSB of the result significant to be 0.
      • This is the most commonly-used rounding mode.
      • This is [currently] the only supported mode by the DSLX implementation.
  7. Special case handling: The results are examined for special cases such as NaNs, infinities, or (optionally) subnormals.

DSLX implementation

With an understanding of the algorithm above, the DSLX implementation is relatively straightforward. "Interesting" chunks are described below.

Result sign determination

The sign of the result will normally be the same as the sign of the operand with the greater exponent, but there are two extra cases to consider. If the operands have the same exponent, then the sign will be that of the greater significand, and if the result is 0, then we favor positive 0 vs. negative 0.

  let sfd = (addend_x as s29) + (addend_y as s29);
  let sfd_is_zero = sfd == s29:0;
  let result_sign = match (sfd_is_zero, sfd < s29:0) {
    (true, _) => u1:0,
    (false, true) => !greater_exp.sign,
    _ => greater_exp.sign,
  };

Rounding

As complicated as rounding is to describe, its implementation is relatively straightforward.

  let normal_chunk = shifted_sfd[0:3];
  let half_way_chunk = shifted_sfd[2:4];
  let do_round_up =
      u1:1 if (normal_chunk > u3:0x4) | (half_way_chunk == u2:0x3)
      else u1:0;

  // We again need an extra bit for carry.
  let rounded_sfd = (shifted_sfd as u28) + u28:0x8 if do_round_up
      else (shifted_sfd as u28);
  let rounding_carry = rounded_sfd[-1:];

The behavior of logic descriptions - even in a higher level language such as DSLX - can be non-obvious to a new reader, so extensive comments, such as those here, are invaluable.