[soft-fp] Add support for various half-precision conversion routines.
commit87ab10d6524fe4faabd7eb3eac5868165ecfb323
authorJames Greenhalgh <james.greenhalgh@arm.com>
Wed, 21 Sep 2016 21:02:54 +0000 (21 21:02 +0000)
committerJoseph Myers <joseph@codesourcery.com>
Wed, 21 Sep 2016 21:02:54 +0000 (21 21:02 +0000)
tree31f378666f9bfc6f0793611bfa0f0a5c6aae8ce9
parentd942e95cd7e493efbbff58b78e63013c79e634cd
[soft-fp] Add support for various half-precision conversion routines.

This patch adds conversion routines required for _Float16 support in
AArch64.

These are one-step conversions to and from TImode and TFmode. We need
these on AArch64 regardless of presence of the ARMv8.2-A 16-bit
floating-point extensions.

In the patch, soft-fp/half.h is derived from soft-fp/single.h .  The
conversion routines are derivatives of their respective SFmode
variants.

* soft-fp/extendhftf2.c: New.
* soft-fp/fixhfti.c: Likewise.
* soft-fp/fixunshfti.c: Likewise.
* soft-fp/floattihf.c: Likewise.
* soft-fp/floatuntihf.c: Likewise.
* soft-fp/half.h: Likewise.
* soft-fp/trunctfhf2.c: Likewise.
ChangeLog
soft-fp/extendhftf2.c [new file with mode: 0644]
soft-fp/fixhfti.c [new file with mode: 0644]
soft-fp/fixunshfti.c [new file with mode: 0644]
soft-fp/floattihf.c [new file with mode: 0644]
soft-fp/floatuntihf.c [new file with mode: 0644]
soft-fp/half.h [new file with mode: 0644]
soft-fp/trunctfhf2.c [new file with mode: 0644]