83 lines
2.8 KiB
C
83 lines
2.8 KiB
C
/******************************************************************************\
|
|
* Authors: Iconoclast *
|
|
* Release: 2013.11.26 *
|
|
* License: CC0 Public Domain Dedication *
|
|
* *
|
|
* To the extent possible under law, the author(s) have dedicated all copyright *
|
|
* and related and neighboring rights to this software to the public domain *
|
|
* worldwide. This software is distributed without any warranty. *
|
|
* *
|
|
* You should have received a copy of the CC0 Public Domain Dedication along *
|
|
* with this software. *
|
|
* If not, see <http://creativecommons.org/publicdomain/zero/1.0/>. *
|
|
\******************************************************************************/
|
|
#include "vu.h"
|
|
|
|
INLINE static void do_lt(usf_state_t * state, short* VD, short* VS, short* VT)
|
|
{
|
|
|
|
#ifdef ARCH_MIN_ARM_NEON
|
|
|
|
int16x8_t zero = vdupq_n_s16(0);
|
|
|
|
int16x8_t vs = vld1q_s16((const int16_t *)VS);
|
|
int16x8_t vt = vld1q_s16((const int16_t *)VT);
|
|
int16x8_t v_ne = vld1q_s16((const int16_t *)state->ne);
|
|
int16x8_t v_co = vld1q_s16((const int16_t *)state->co);
|
|
|
|
uint16x8_t v_eq_u = vceqq_s16(vs,vt);
|
|
int16x8_t v_cn = vandq_s16(v_ne,v_co);
|
|
v_eq_u = vandq_u16(v_eq_u,(uint16x8_t)v_cn);
|
|
|
|
vst1q_s16(state->clip, zero);
|
|
|
|
uint16x8_t v_comp = vcltq_s16(vs, vt);
|
|
int16x8_t v_comp_s = vnegq_s16((int16x8_t)v_comp);
|
|
v_comp_s = vorrq_s16(v_comp_s, (int16x8_t)v_eq_u);
|
|
|
|
vst1q_s16(state->comp, v_comp_s);
|
|
|
|
merge(VACC_L, state->comp, VS, VT);
|
|
vector_copy(VD, VACC_L);
|
|
|
|
vst1q_s16(state->ne, zero);
|
|
vst1q_s16(state->co, zero);
|
|
return;
|
|
|
|
#else
|
|
|
|
ALIGNED short cn[N];
|
|
ALIGNED short eq[N];
|
|
register int i;
|
|
|
|
for (i = 0; i < N; i++)
|
|
eq[i] = (VS[i] == VT[i]);
|
|
for (i = 0; i < N; i++)
|
|
cn[i] = state->ne[i] & state->co[i];
|
|
for (i = 0; i < N; i++)
|
|
eq[i] = eq[i] & cn[i];
|
|
for (i = 0; i < N; i++)
|
|
state->clip[i] = 0;
|
|
for (i = 0; i < N; i++)
|
|
state->comp[i] = (VS[i] < VT[i]); /* less than */
|
|
for (i = 0; i < N; i++)
|
|
state->comp[i] = state->comp[i] | eq[i]; /* ... or equal (uncommonly) */
|
|
|
|
merge(VACC_L, state->comp, VS, VT);
|
|
vector_copy(VD, VACC_L);
|
|
for (i = 0; i < N; i++)
|
|
state->ne[i] = 0;
|
|
for (i = 0; i < N; i++)
|
|
state->co[i] = 0;
|
|
return;
|
|
#endif
|
|
}
|
|
|
|
static void VLT(usf_state_t * state, int vd, int vs, int vt, int e)
|
|
{
|
|
ALIGNED short ST[N];
|
|
|
|
SHUFFLE_VECTOR(ST, state->VR[vt], e);
|
|
do_lt(state, state->VR[vd], state->VR[vs], ST);
|
|
return;
|
|
}
|