cast to appropriate types.

branches: 1.2.6;
fix -Wsign-compare issue (on sparc64)

branches: 1.1.4; 1.1.46; 1.1.50;
Import the 32-bit version of SoftFloat 2a and separate it and the 64-bit
version into different directories. Which version a given port uses is
controlled by the SOFTFLOAT_BITS make variable. This is set to 64 (which
uses the same code we had before) by default. 32-bit platforms that don't
need extended precision support might get better performance by using 32.

Set the ARM port to use the 32-bit version of SoftFloat, since this is more
than a factor of two faster than the 64-bit version. This should get the
floating-point performance back to what it was in 1.5.

Fix thinko.

Change the remainder of missed INFs for floatx80.

Return the right value for infinity for floatx80.

There are only 5 failures and 1 expected failure left.

Fix return values for softfloat functions to appease gcc.

Not sure if it is gcc and m68k only so it wrapped up in defines.

Shift the high word of floatx80 accordingly on m68k.

This is because gcc will store floatx80 as signbit(1)exponent(15) padding(32)
as opposed to sign and exponent in a short(16 bit) word.

appease lint

Avoid warnings (signed/unsigned comparision and unused variable)

Fix a cast from the lint cleanup that made small exponents (i.e. values < 1)
sign extend wrong and overflow, causing an underflow in all 128 bit sqrt
calculations.

Add hooks to softfloat to support per-lwp exception state.

branches: 1.11.2;
Fix a bug introduced by lint cleanup.

cast to appropriate types.

Remove initialized but unused variable

branches: 1.8.2;
Fix so that float128_to_uint64_round_to_zero is defined for sparc64.
(isn't tha a namespace violation?)

Add SOFTFLOAT_NEED_FIXUNS condition around the float128 fixuns

Add __floatunsidf __floatunsisf __floatunsitf routines.
XXX i think they are correct but not sure.

branches: 1.5.20;
When creating unsigned int results, do not distinguish positive versus
negative overflow.

When converting long double values to integer types, explicitly use the
"round to zero" variants of the softfloat conversion functions.
Add a variant to convert long double to unsigned long - the "to long"
variant checked for overflows that do not apply to unsigned results.
This fixes the regress/lib/libc/convfp tests for sparc64.

branches: 1.3.10;
add128() wants bits64 * as it's 5th argument.

netbsd.org->NetBSD.org

branches: 1.1.4;
Import the 32-bit version of SoftFloat 2a and separate it and the 64-bit
version into different directories. Which version a given port uses is
controlled by the SOFTFLOAT_BITS make variable. This is set to 64 (which
uses the same code we had before) by default. 32-bit platforms that don't
need extended precision support might get better performance by using 32.

Set the ARM port to use the 32-bit version of SoftFloat, since this is more
than a factor of two faster than the 64-bit version. This should get the
floating-point performance back to what it was in 1.5.