+++ /dev/null
-$NetBSD: softfloat-history.txt,v 1.1 2000/06/06 08:15:08 bjh21 Exp $\r
-\r
-History of Major Changes to SoftFloat, up to Release 2a\r
-\r
-John R. Hauser\r
-1998 December 16\r
-\r
-\r
-- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -\r
-Release 2a (1998 December)\r
-\r
--- Added functions to convert between 64-bit integers (int64) and all\r
- supported floating-point formats.\r
-\r
--- Fixed a bug in all 64-bit-version square root functions except\r
- `float32_sqrt' that caused the result sometimes to be off by 1 unit in\r
- the last place (1 ulp) from what it should be. (Bug discovered by Paul\r
- Donahue.)\r
-\r
--- Improved the makefiles.\r
-\r
-- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -\r
-Release 2 (1997 June)\r
-\r
--- Created the 64-bit (bits64) version, adding the floatx80 and float128\r
- formats.\r
-\r
--- Changed the source directory structure, splitting the sources into a\r
- `bits32' and a `bits64' version. Renamed `environment.h' to `milieu.h'\r
- (to avoid confusion with environment variables).\r
-\r
--- Fixed a small error that caused `float64_round_to_int' often to round the\r
- wrong way in nearest/even mode when the operand was between 2^20 and 2^21\r
- and halfway between two integers.\r
-\r
-- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -\r
-Release 1a (1996 July)\r
-\r
--- Corrected a mistake that caused borderline underflow cases not to raise\r
- the underflow flag when they should have. (Problem reported by Doug\r
- Priest.)\r
-\r
--- Added the `float_detect_tininess' variable to control whether tininess is\r
- detected before or after rounding.\r
-\r
-- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -\r
-Release 1 (1996 July)\r
-\r
--- Original release.\r
-\r
-- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -\r
-\r