Commit 85c2a393 authored by Aleksandar Markovic's avatar Aleksandar Markovic
Browse files

tests/tcg: target/mips: Add wrappers for MSA bit counting instructions



Add a header that contains wrappers around MSA instructions assembler
invocations. For now, only bit counting instructions (NLOC, NLZC, and
PCNT; each in four data format flavors) are supported.

Reviewed-by: default avatarAleksandar Rikalo <arikalo@wavecomp.com>
Signed-off-by: default avatarAleksandar Markovic <amarkovic@wavecomp.com>
parent 08857c88
Loading
Loading
Loading
Loading
+57 −0
Original line number Diff line number Diff line
/*
 *  Header file for wrappers around MSA instructions assembler invocations
 *
 *  Copyright (C) 2018  Wave Computing, Inc.
 *  Copyright (C) 2018  Aleksandar Markovic <amarkovic@wavecomp.com>
 *
 *  This program 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 2 of the License, or
 *  (at your option) any later version.
 *
 *  This program 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.
 *
 *  You should have received a copy of the GNU General Public License
 *  along with this program.  If not, see <https://www.gnu.org/licenses/>.
 *
 */

#ifndef WRAPPERS_MSA_H
#define WRAPPERS_MSA_H


#define DO_MSA__WD__WS(suffix, mnemonic)                               \
static inline void do_msa_##suffix(void *input, void *output)          \
{                                                                      \
   __asm__ volatile (                                                  \
      "move $t0, %0\n\t"                                               \
      "ld.d $w11, 0($t0)\n\t"                                          \
      #mnemonic " $w10, $w11\n\t"                                      \
      "move $t0, %1\n\t"                                               \
      "st.d $w10, 0($t0)\n\t"                                          \
      :                                                                \
      : "r" (input), "r" (output)                                      \
      : "t0", "memory"                                                 \
   );                                                                  \
}

DO_MSA__WD__WS(NLOC_B, nloc.b)
DO_MSA__WD__WS(NLOC_H, nloc.h)
DO_MSA__WD__WS(NLOC_W, nloc.w)
DO_MSA__WD__WS(NLOC_D, nloc.d)

DO_MSA__WD__WS(NLZC_B, nlzc.b)
DO_MSA__WD__WS(NLZC_H, nlzc.h)
DO_MSA__WD__WS(NLZC_W, nlzc.w)
DO_MSA__WD__WS(NLZC_D, nlzc.d)

DO_MSA__WD__WS(PCNT_B, pcnt.b)
DO_MSA__WD__WS(PCNT_H, pcnt.h)
DO_MSA__WD__WS(PCNT_W, pcnt.w)
DO_MSA__WD__WS(PCNT_D, pcnt.d)


#endif