1 $NetBSD: softfloat-history.txt,v 1.1 2000/06/06 08:15:08 bjh21 Exp $
4 History of Major Changes to SoftFloat, up to Release 2a
10 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
11 Release 2a (1998 December)
13 -- Added functions to convert between 64-bit integers (int64) and all
14 supported floating-point formats.
16 -- Fixed a bug in all 64-bit-version square root functions except
17 `float32_sqrt' that caused the result sometimes to be off by 1 unit in
18 the last place (1 ulp) from what it should be. (Bug discovered by Paul
21 -- Improved the makefiles.
23 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
26 -- Created the 64-bit (bits64) version, adding the floatx80 and float128
29 -- Changed the source directory structure, splitting the sources into a
30 `bits32' and a `bits64' version. Renamed `environment.h' to `milieu.h'
31 (to avoid confusion with environment variables).
33 -- Fixed a small error that caused `float64_round_to_int' often to round the
34 wrong way in nearest/even mode when the operand was between 2^20 and 2^21
35 and halfway between two integers.
37 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
38 Release 1a (1996 July)
40 -- Corrected a mistake that caused borderline underflow cases not to raise
41 the underflow flag when they should have. (Problem reported by Doug
44 -- Added the `float_detect_tininess' variable to control whether tininess is
45 detected before or after rounding.
47 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
52 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -