]> asedeno.scripts.mit.edu Git - linux.git/commit
nds32: add new emulations for floating point instruction
authorVincent Chen <vincentc@andestech.com>
Mon, 20 May 2019 01:21:13 +0000 (09:21 +0800)
committerGreentime Hu <greentime@andestech.com>
Fri, 31 May 2019 07:23:26 +0000 (15:23 +0800)
commit932296120543149e3397af252e7daee7af37eb05
treead3c0a00a16cae8cab3d85c3b74060893003ee83
parented32949e0acb73e7bc054bb02e0453d4d561ceda
nds32: add new emulations for floating point instruction

The existing floating point emulations is only available for floating
instruction that possibly issue denormalized input and underflow
exceptions. These existing FPU emulations are not sufficient when IEx
Trap is enabled because some floating point instructions only issue inexact
exception. This patch adds the emulations of such floating point
instructions.

Signed-off-by: Vincent Chen <vincentc@andestech.com>
Acked-by: Greentime Hu <greentime@andestech.com>
Signed-off-by: Greentime Hu <greentime@andestech.com>
15 files changed:
arch/nds32/include/asm/fpuemu.h
arch/nds32/math-emu/Makefile
arch/nds32/math-emu/fd2si.c [new file with mode: 0644]
arch/nds32/math-emu/fd2siz.c [new file with mode: 0644]
arch/nds32/math-emu/fd2ui.c [new file with mode: 0644]
arch/nds32/math-emu/fd2uiz.c [new file with mode: 0644]
arch/nds32/math-emu/fpuemu.c
arch/nds32/math-emu/fs2si.c [new file with mode: 0644]
arch/nds32/math-emu/fs2siz.c [new file with mode: 0644]
arch/nds32/math-emu/fs2ui.c [new file with mode: 0644]
arch/nds32/math-emu/fs2uiz.c [new file with mode: 0644]
arch/nds32/math-emu/fsi2d.c [new file with mode: 0644]
arch/nds32/math-emu/fsi2s.c [new file with mode: 0644]
arch/nds32/math-emu/fui2d.c [new file with mode: 0644]
arch/nds32/math-emu/fui2s.c [new file with mode: 0644]