| 1 |
6 |
jidan |
|
| 2 |
|
|
History of Major Changes to SoftFloat, up to Release 2b
|
| 3 |
|
|
|
| 4 |
|
|
John R. Hauser
|
| 5 |
|
|
2002 May 27
|
| 6 |
|
|
|
| 7 |
|
|
|
| 8 |
|
|
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
|
| 9 |
|
|
Release 2b (2002 May)
|
| 10 |
|
|
|
| 11 |
|
|
-- Made minor updates to the documentation, including improved wording of
|
| 12 |
|
|
the legal restrictions on using SoftFloat.
|
| 13 |
|
|
|
| 14 |
|
|
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
|
| 15 |
|
|
Release 2a (1998 December)
|
| 16 |
|
|
|
| 17 |
|
|
-- Added functions to convert between 64-bit integers (int64) and all
|
| 18 |
|
|
supported floating-point formats.
|
| 19 |
|
|
|
| 20 |
|
|
-- Fixed a bug in all 64-bit-version square root functions except
|
| 21 |
|
|
`float32_sqrt' that caused the result sometimes to be off by 1 unit in
|
| 22 |
|
|
the last place (1 ulp) from what it should be. (Bug discovered by Paul
|
| 23 |
|
|
Donahue.)
|
| 24 |
|
|
|
| 25 |
|
|
-- Improved the makefiles.
|
| 26 |
|
|
|
| 27 |
|
|
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
|
| 28 |
|
|
Release 2 (1997 June)
|
| 29 |
|
|
|
| 30 |
|
|
-- Created the 64-bit (bits64) version, adding the floatx80 and float128
|
| 31 |
|
|
formats.
|
| 32 |
|
|
|
| 33 |
|
|
-- Changed the source directory structure, splitting the sources into a
|
| 34 |
|
|
`bits32' and a `bits64' version. Renamed `environment.h' to `milieu.h'
|
| 35 |
|
|
to avoid confusion with environment variables.
|
| 36 |
|
|
|
| 37 |
|
|
-- Fixed a small error that caused `float64_round_to_int' often to round the
|
| 38 |
|
|
wrong way in nearest/even mode when the operand was between 2^20 and 2^21
|
| 39 |
|
|
and halfway between two integers.
|
| 40 |
|
|
|
| 41 |
|
|
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
|
| 42 |
|
|
Release 1a (1996 July)
|
| 43 |
|
|
|
| 44 |
|
|
-- Corrected a mistake that caused borderline underflow cases not to raise
|
| 45 |
|
|
the underflow flag when they should have. (Problem reported by Doug
|
| 46 |
|
|
Priest.)
|
| 47 |
|
|
|
| 48 |
|
|
-- Added the `float_detect_tininess' variable to control whether tininess is
|
| 49 |
|
|
detected before or after rounding.
|
| 50 |
|
|
|
| 51 |
|
|
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
|
| 52 |
|
|
Release 1 (1996 July)
|
| 53 |
|
|
|
| 54 |
|
|
-- Original release.
|
| 55 |
|
|
|
| 56 |
|
|
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
|
| 57 |
|
|
|
