mirror of https://git.ffmpeg.org/ffmpeg.git
113 lines
3.5 KiB
C
113 lines
3.5 KiB
C
/*
|
|
* Alpha optimized DSP utils
|
|
* Copyright (c) 2002 Falk Hueffner <falk@debian.org>
|
|
*
|
|
* This file is part of FFmpeg.
|
|
*
|
|
* FFmpeg is free software; you can redistribute it and/or
|
|
* modify it under the terms of the GNU Lesser General Public
|
|
* License as published by the Free Software Foundation; either
|
|
* version 2.1 of the License, or (at your option) any later version.
|
|
*
|
|
* FFmpeg 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
|
|
* Lesser General Public License for more details.
|
|
*
|
|
* You should have received a copy of the GNU Lesser General Public
|
|
* License along with FFmpeg; if not, write to the Free Software
|
|
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
|
|
*/
|
|
|
|
#include "libavutil/attributes.h"
|
|
#include "libavcodec/dsputil.h"
|
|
#include "dsputil_alpha.h"
|
|
#include "asm.h"
|
|
|
|
#if 0
|
|
/* These functions were the base for the optimized assembler routines,
|
|
and remain here for documentation purposes. */
|
|
static void put_pixels_clamped_mvi(const int16_t *block, uint8_t *pixels,
|
|
ptrdiff_t line_size)
|
|
{
|
|
int i = 8;
|
|
uint64_t clampmask = zap(-1, 0xaa); /* 0x00ff00ff00ff00ff */
|
|
|
|
do {
|
|
uint64_t shorts0, shorts1;
|
|
|
|
shorts0 = ldq(block);
|
|
shorts0 = maxsw4(shorts0, 0);
|
|
shorts0 = minsw4(shorts0, clampmask);
|
|
stl(pkwb(shorts0), pixels);
|
|
|
|
shorts1 = ldq(block + 4);
|
|
shorts1 = maxsw4(shorts1, 0);
|
|
shorts1 = minsw4(shorts1, clampmask);
|
|
stl(pkwb(shorts1), pixels + 4);
|
|
|
|
pixels += line_size;
|
|
block += 8;
|
|
} while (--i);
|
|
}
|
|
|
|
void add_pixels_clamped_mvi(const int16_t *block, uint8_t *pixels,
|
|
ptrdiff_t line_size)
|
|
{
|
|
int h = 8;
|
|
/* Keep this function a leaf function by generating the constants
|
|
manually (mainly for the hack value ;-). */
|
|
uint64_t clampmask = zap(-1, 0xaa); /* 0x00ff00ff00ff00ff */
|
|
uint64_t signmask = zap(-1, 0x33);
|
|
signmask ^= signmask >> 1; /* 0x8000800080008000 */
|
|
|
|
do {
|
|
uint64_t shorts0, pix0, signs0;
|
|
uint64_t shorts1, pix1, signs1;
|
|
|
|
shorts0 = ldq(block);
|
|
shorts1 = ldq(block + 4);
|
|
|
|
pix0 = unpkbw(ldl(pixels));
|
|
/* Signed subword add (MMX paddw). */
|
|
signs0 = shorts0 & signmask;
|
|
shorts0 &= ~signmask;
|
|
shorts0 += pix0;
|
|
shorts0 ^= signs0;
|
|
/* Clamp. */
|
|
shorts0 = maxsw4(shorts0, 0);
|
|
shorts0 = minsw4(shorts0, clampmask);
|
|
|
|
/* Next 4. */
|
|
pix1 = unpkbw(ldl(pixels + 4));
|
|
signs1 = shorts1 & signmask;
|
|
shorts1 &= ~signmask;
|
|
shorts1 += pix1;
|
|
shorts1 ^= signs1;
|
|
shorts1 = maxsw4(shorts1, 0);
|
|
shorts1 = minsw4(shorts1, clampmask);
|
|
|
|
stl(pkwb(shorts0), pixels);
|
|
stl(pkwb(shorts1), pixels + 4);
|
|
|
|
pixels += line_size;
|
|
block += 8;
|
|
} while (--h);
|
|
}
|
|
#endif
|
|
|
|
av_cold void ff_dsputil_init_alpha(DSPContext *c, AVCodecContext *avctx)
|
|
{
|
|
/* amask clears all bits that correspond to present features. */
|
|
if (amask(AMASK_MVI) == 0) {
|
|
c->sad[0] = pix_abs16x16_mvi_asm;
|
|
c->sad[1] = pix_abs8x8_mvi;
|
|
c->pix_abs[0][0] = pix_abs16x16_mvi_asm;
|
|
c->pix_abs[1][0] = pix_abs8x8_mvi;
|
|
c->pix_abs[0][1] = pix_abs16x16_x2_mvi;
|
|
c->pix_abs[0][2] = pix_abs16x16_y2_mvi;
|
|
c->pix_abs[0][3] = pix_abs16x16_xy2_mvi;
|
|
}
|
|
|
|
}
|