1234567891011121314151617181920212223242526272829303132333435363738394041424344454647484950515253545556575859606162636465666768697071727374757677 |
- /* Signed 32 bit division optimized for Epiphany.
- Copyright (C) 2009-2022 Free Software Foundation, Inc.
- Contributed by Embecosm on behalf of Adapteva, Inc.
- This file is part of GCC.
- This file is free software; you can redistribute it and/or modify it
- under the terms of the GNU General Public License as published by the
- Free Software Foundation; either version 3, or (at your option) any
- later version.
- This file is distributed in the hope that it will be useful, but
- WITHOUT ANY WARRANTY; without even the implied warranty of
- MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
- General Public License for more details.
- Under Section 7 of GPL version 3, you are granted additional
- permissions described in the GCC Runtime Library Exception, version
- 3.1, as published by the Free Software Foundation.
- You should have received a copy of the GNU General Public License and
- a copy of the GCC Runtime Library Exception along with this program;
- see the files COPYING3 and COPYING.RUNTIME respectively. If not, see
- <http://www.gnu.org/licenses/>. */
- #include "epiphany-asm.h"
- FSTAB (__divsi3,T_UINT)
- .global SYM(__divsi3)
- .balign 4
- HIDDEN_FUNC(__divsi3)
- SYM(__divsi3):
- float TMP2,r0
- mov TMP4,0
- float TMP1,r1
- sub TMP0,TMP4,r0
- beq .Lret_r0
- movgt r0,TMP0
- sub TMP0,TMP4,r1
- movgt r1,TMP0
- mov TMP0,1
- sub TMP2,TMP2,TMP1
- asr TMP3,TMP2,31 ; save sign
- lsl TMP2,TMP2,1
- blt .Lret0
- sub TMP1,TMP2,1 ; rounding compensation, avoid overflow
- movgte TMP2,TMP1
- lsr TMP2,TMP2,24
- lsl r1,r1,TMP2
- lsl TMP0,TMP0,TMP2
- sub TMP1,r0,r1
- movgteu r0,TMP1
- movgteu TMP4,TMP0
- lsl TMP5,TMP0,1
- sub TMP1,r0,r1
- movgteu r0,TMP1
- movgteu TMP4,TMP5
- sub TMP1,r1,1
- mov r1,%low(.L0step)
- movt r1,%high(.L0step)
- lsl TMP2,TMP2,3
- sub r1,r1,TMP2
- jr r1
- .rep 30
- lsl r0,r0,1
- sub.l r1,r0,TMP1
- movgteu r0,r1
- .endr
- .L0step:sub r1,TMP0,1 ; mask result bits from steps ...
- and r0,r0,r1
- orr r0,r0,TMP4 ; ... and combine with first bit.
- eor r0,r0,TMP3 ; restore sign
- sub r0,r0,TMP3
- .Lret_r0:rts
- .Lret0: mov r0,0
- rts
- ENDFUNC(__divsi3)
|