void test_vmulQ_nf32 (void)
{
float32x4_t out_float32x4_t;
float32x4_t arg0_float32x4_t;
float32_t arg1_float32_t;
out_float32x4_t = vmulq_n_f32 (arg0_float32x4_t, arg1_float32_t);
}
void dotProd_i16_neon(const float *dataf, const float *weightsf, float *vals, const int n, const int len, const float *istd) {
const int16_t *data = (const int16_t *)dataf;
const int16_t *weights = (const int16_t *)weightsf;
weightsf += n * len / 2; // sizeof(float) / sizeof(int16_t)
for (int i = 0; i < n; i += 4) {
int32x4_t accum0 = { 0, 0, 0, 0 };
int32x4_t accum1 = accum0;
int32x4_t accum2 = accum0;
int32x4_t accum3 = accum0;
for (int j = 0; j < len; j += 8) {
int16x4x2_t d0 = vld2_s16(data + j);
int16x4x2_t w0 = vld2_s16(weights);
int16x4x2_t w1 = vld2_s16(weights + 8);
int16x4x2_t w2 = vld2_s16(weights + 16);
int16x4x2_t w3 = vld2_s16(weights + 24);
accum0 = vmlal_s16(accum0, d0.val[0], w0.val[0]);
accum0 = vmlal_s16(accum0, d0.val[1], w0.val[1]);
accum1 = vmlal_s16(accum1, d0.val[0], w1.val[0]);
accum1 = vmlal_s16(accum1, d0.val[1], w1.val[1]);
accum2 = vmlal_s16(accum2, d0.val[0], w2.val[0]);
accum2 = vmlal_s16(accum2, d0.val[1], w2.val[1]);
accum3 = vmlal_s16(accum3, d0.val[0], w3.val[0]);
accum3 = vmlal_s16(accum3, d0.val[1], w3.val[1]);
weights += 32;
}
int32x2_t sum0 = vpadd_s32(vget_low_s32(accum0), vget_high_s32(accum0));
int32x2_t sum1 = vpadd_s32(vget_low_s32(accum1), vget_high_s32(accum1));
int32x2_t sum2 = vpadd_s32(vget_low_s32(accum2), vget_high_s32(accum2));
int32x2_t sum3 = vpadd_s32(vget_low_s32(accum3), vget_high_s32(accum3));
sum0 = vpadd_s32(sum0, sum1);
sum1 = vpadd_s32(sum2, sum3);
int32x4_t sum = vcombine_s32(sum0, sum1);
float32x4_t val = vcvtq_f32_s32(sum);
val = vmulq_f32(val, vld1q_f32(weightsf + i*2));
val = vmulq_n_f32(val, istd[0]);
val = vaddq_f32(val, vld1q_f32(weightsf + i*2 + 4));
vst1q_f32(vals + i, val);
}
}
/**
* @brief vector scale & accu: A[] = alpha * B[] + beta * A[].
*
* @param dst[out] the accumulating matrix A.
* src[in] the input matrix B.
* alpha[in] scale of B.
* beta[in] scale of A.
* elemCnt[in] number of elements to calc.
*
* @return void.
*/
void neon_axpby(float *dst,
const float *src,
const float alpha,
const float beta,
const int elemCnt)
{
int i;
for (i = 0; i <= elemCnt - 16; i += 16)
{
float32x4_t q0 = vld1q_f32(src + i);
float32x4_t q1 = vld1q_f32(src + i + 4);
float32x4_t q2 = vld1q_f32(src + i + 8);
float32x4_t q3 = vld1q_f32(src + i + 12);
float32x4_t q4 = vld1q_f32(dst + i);
float32x4_t q5 = vld1q_f32(dst + i + 4);
float32x4_t q6 = vld1q_f32(dst + i + 8);
float32x4_t q7 = vld1q_f32(dst + i + 12);
q0 = vmulq_n_f32(q0, alpha);
q1 = vmulq_n_f32(q1, alpha);
q2 = vmulq_n_f32(q2, alpha);
q3 = vmulq_n_f32(q3, alpha);
q0 = vmlaq_n_f32(q0, q4, beta);
q1 = vmlaq_n_f32(q1, q5, beta);
q2 = vmlaq_n_f32(q2, q6, beta);
q3 = vmlaq_n_f32(q3, q7, beta);
vst1q_f32(dst + i, q0);
vst1q_f32(dst + i + 4, q1);
vst1q_f32(dst + i + 8, q2);
vst1q_f32(dst + i + 12, q3);
}
for (; i < elemCnt; i++)
{
float a = src[i] * alpha + dst[i] * beta;
dst[i] = a;
}
}
void dotProd_neon(const float *data, const float *weights, float *vals, const int n, const int len, const float *istd) {
for (int i = 0; i < n; i += 4) {
float32x4_t accum0 = { 0.0f, 0.0f, 0.0f, 0.0f };
float32x4_t accum1 = accum0;
float32x4_t accum2 = accum0;
float32x4_t accum3 = accum0;
for (int j = 0; j < len; j += 4) {
float32x4_t d0 = vld1q_f32(data + j);
float32x4_t d1 = d0;
float32x4_t d2 = d0;
float32x4_t d3 = d0;
float32x4_t w0 = vld1q_f32(weights);
float32x4_t w1 = vld1q_f32(weights + 4);
float32x4_t w2 = vld1q_f32(weights + 8);
float32x4_t w3 = vld1q_f32(weights + 12);
accum0 = vaddq_f32(accum0, vmulq_f32(d0, w0));
accum1 = vaddq_f32(accum1, vmulq_f32(d1, w1));
accum2 = vaddq_f32(accum2, vmulq_f32(d2, w2));
accum3 = vaddq_f32(accum3, vmulq_f32(d3, w3));
weights += 16;
}
float32x2_t sum0 = vpadd_f32(vget_low_f32(accum0), vget_high_f32(accum0));
float32x2_t sum1 = vpadd_f32(vget_low_f32(accum1), vget_high_f32(accum1));
float32x2_t sum2 = vpadd_f32(vget_low_f32(accum2), vget_high_f32(accum2));
float32x2_t sum3 = vpadd_f32(vget_low_f32(accum3), vget_high_f32(accum3));
sum0 = vpadd_f32(sum0, sum1);
sum1 = vpadd_f32(sum2, sum3);
float32x4_t sum = vcombine_f32(sum0, sum1);
sum = vmulq_n_f32(sum, istd[0]);
sum = vaddq_f32(sum, vld1q_f32(weights + n*len + i));
vst1q_f32(vals + i, sum);
}
}
void x(float32x4_t v0, float32_t f)
{
float32x4_t vv = vmulq_n_f32( v0, f);
}
inline void lxt_v4_scale( lxt_v4* a_result, const lxt_v4* a_v, float a_scale )
{
a_result->m_i = vmulq_n_f32( a_v->m_i, a_scale );
}
// Updates the following smoothed Power Spectral Densities (PSD):
// - sd : near-end
// - se : residual echo
// - sx : far-end
// - sde : cross-PSD of near-end and residual echo
// - sxd : cross-PSD of near-end and far-end
//
// In addition to updating the PSDs, also the filter diverge state is determined
// upon actions are taken.
static void SmoothedPSD(AecCore* aec,
float efw[2][PART_LEN1],
float dfw[2][PART_LEN1],
float xfw[2][PART_LEN1],
int* extreme_filter_divergence) {
// Power estimate smoothing coefficients.
const float* ptrGCoh = aec->extended_filter_enabled
? WebRtcAec_kExtendedSmoothingCoefficients[aec->mult - 1]
: WebRtcAec_kNormalSmoothingCoefficients[aec->mult - 1];
int i;
float sdSum = 0, seSum = 0;
const float32x4_t vec_15 = vdupq_n_f32(WebRtcAec_kMinFarendPSD);
float32x4_t vec_sdSum = vdupq_n_f32(0.0f);
float32x4_t vec_seSum = vdupq_n_f32(0.0f);
for (i = 0; i + 3 < PART_LEN1; i += 4) {
const float32x4_t vec_dfw0 = vld1q_f32(&dfw[0][i]);
const float32x4_t vec_dfw1 = vld1q_f32(&dfw[1][i]);
const float32x4_t vec_efw0 = vld1q_f32(&efw[0][i]);
const float32x4_t vec_efw1 = vld1q_f32(&efw[1][i]);
const float32x4_t vec_xfw0 = vld1q_f32(&xfw[0][i]);
const float32x4_t vec_xfw1 = vld1q_f32(&xfw[1][i]);
float32x4_t vec_sd = vmulq_n_f32(vld1q_f32(&aec->sd[i]), ptrGCoh[0]);
float32x4_t vec_se = vmulq_n_f32(vld1q_f32(&aec->se[i]), ptrGCoh[0]);
float32x4_t vec_sx = vmulq_n_f32(vld1q_f32(&aec->sx[i]), ptrGCoh[0]);
float32x4_t vec_dfw_sumsq = vmulq_f32(vec_dfw0, vec_dfw0);
float32x4_t vec_efw_sumsq = vmulq_f32(vec_efw0, vec_efw0);
float32x4_t vec_xfw_sumsq = vmulq_f32(vec_xfw0, vec_xfw0);
vec_dfw_sumsq = vmlaq_f32(vec_dfw_sumsq, vec_dfw1, vec_dfw1);
vec_efw_sumsq = vmlaq_f32(vec_efw_sumsq, vec_efw1, vec_efw1);
vec_xfw_sumsq = vmlaq_f32(vec_xfw_sumsq, vec_xfw1, vec_xfw1);
vec_xfw_sumsq = vmaxq_f32(vec_xfw_sumsq, vec_15);
vec_sd = vmlaq_n_f32(vec_sd, vec_dfw_sumsq, ptrGCoh[1]);
vec_se = vmlaq_n_f32(vec_se, vec_efw_sumsq, ptrGCoh[1]);
vec_sx = vmlaq_n_f32(vec_sx, vec_xfw_sumsq, ptrGCoh[1]);
vst1q_f32(&aec->sd[i], vec_sd);
vst1q_f32(&aec->se[i], vec_se);
vst1q_f32(&aec->sx[i], vec_sx);
{
float32x4x2_t vec_sde = vld2q_f32(&aec->sde[i][0]);
float32x4_t vec_dfwefw0011 = vmulq_f32(vec_dfw0, vec_efw0);
float32x4_t vec_dfwefw0110 = vmulq_f32(vec_dfw0, vec_efw1);
vec_sde.val[0] = vmulq_n_f32(vec_sde.val[0], ptrGCoh[0]);
vec_sde.val[1] = vmulq_n_f32(vec_sde.val[1], ptrGCoh[0]);
vec_dfwefw0011 = vmlaq_f32(vec_dfwefw0011, vec_dfw1, vec_efw1);
vec_dfwefw0110 = vmlsq_f32(vec_dfwefw0110, vec_dfw1, vec_efw0);
vec_sde.val[0] = vmlaq_n_f32(vec_sde.val[0], vec_dfwefw0011, ptrGCoh[1]);
vec_sde.val[1] = vmlaq_n_f32(vec_sde.val[1], vec_dfwefw0110, ptrGCoh[1]);
vst2q_f32(&aec->sde[i][0], vec_sde);
}
{
float32x4x2_t vec_sxd = vld2q_f32(&aec->sxd[i][0]);
float32x4_t vec_dfwxfw0011 = vmulq_f32(vec_dfw0, vec_xfw0);
float32x4_t vec_dfwxfw0110 = vmulq_f32(vec_dfw0, vec_xfw1);
vec_sxd.val[0] = vmulq_n_f32(vec_sxd.val[0], ptrGCoh[0]);
vec_sxd.val[1] = vmulq_n_f32(vec_sxd.val[1], ptrGCoh[0]);
vec_dfwxfw0011 = vmlaq_f32(vec_dfwxfw0011, vec_dfw1, vec_xfw1);
vec_dfwxfw0110 = vmlsq_f32(vec_dfwxfw0110, vec_dfw1, vec_xfw0);
vec_sxd.val[0] = vmlaq_n_f32(vec_sxd.val[0], vec_dfwxfw0011, ptrGCoh[1]);
vec_sxd.val[1] = vmlaq_n_f32(vec_sxd.val[1], vec_dfwxfw0110, ptrGCoh[1]);
vst2q_f32(&aec->sxd[i][0], vec_sxd);
}
vec_sdSum = vaddq_f32(vec_sdSum, vec_sd);
vec_seSum = vaddq_f32(vec_seSum, vec_se);
}
{
float32x2_t vec_sdSum_total;
float32x2_t vec_seSum_total;
// A B C D
vec_sdSum_total = vpadd_f32(vget_low_f32(vec_sdSum),
vget_high_f32(vec_sdSum));
vec_seSum_total = vpadd_f32(vget_low_f32(vec_seSum),
vget_high_f32(vec_seSum));
// A+B C+D
vec_sdSum_total = vpadd_f32(vec_sdSum_total, vec_sdSum_total);
vec_seSum_total = vpadd_f32(vec_seSum_total, vec_seSum_total);
// A+B+C+D A+B+C+D
sdSum = vget_lane_f32(vec_sdSum_total, 0);
seSum = vget_lane_f32(vec_seSum_total, 0);
}
// scalar code for the remaining items.
for (; i < PART_LEN1; i++) {
aec->sd[i] = ptrGCoh[0] * aec->sd[i] +
ptrGCoh[1] * (dfw[0][i] * dfw[0][i] + dfw[1][i] * dfw[1][i]);
aec->se[i] = ptrGCoh[0] * aec->se[i] +
ptrGCoh[1] * (efw[0][i] * efw[0][i] + efw[1][i] * efw[1][i]);
// We threshold here to protect against the ill-effects of a zero farend.
// The threshold is not arbitrarily chosen, but balances protection and
// adverse interaction with the algorithm's tuning.
// TODO(bjornv): investigate further why this is so sensitive.
aec->sx[i] =
ptrGCoh[0] * aec->sx[i] +
ptrGCoh[1] * WEBRTC_SPL_MAX(
xfw[0][i] * xfw[0][i] + xfw[1][i] * xfw[1][i],
WebRtcAec_kMinFarendPSD);
aec->sde[i][0] =
ptrGCoh[0] * aec->sde[i][0] +
ptrGCoh[1] * (dfw[0][i] * efw[0][i] + dfw[1][i] * efw[1][i]);
aec->sde[i][1] =
ptrGCoh[0] * aec->sde[i][1] +
ptrGCoh[1] * (dfw[0][i] * efw[1][i] - dfw[1][i] * efw[0][i]);
aec->sxd[i][0] =
ptrGCoh[0] * aec->sxd[i][0] +
ptrGCoh[1] * (dfw[0][i] * xfw[0][i] + dfw[1][i] * xfw[1][i]);
aec->sxd[i][1] =
ptrGCoh[0] * aec->sxd[i][1] +
ptrGCoh[1] * (dfw[0][i] * xfw[1][i] - dfw[1][i] * xfw[0][i]);
sdSum += aec->sd[i];
seSum += aec->se[i];
}
// Divergent filter safeguard update.
aec->divergeState = (aec->divergeState ? 1.05f : 1.0f) * seSum > sdSum;
// Signal extreme filter divergence if the error is significantly larger
// than the nearend (13 dB).
*extreme_filter_divergence = (seSum > (19.95f * sdSum));
}
/**
* @brief matrix dot mul vector: C[] = alpha * A[] * B[] + beta * C[].
*
* @param dst[out] the result matrix C.
* src1[in] the input matrix A.
* src2[in] the input vector B.
* alpha[in] scale of A * B.
* beta[in] scale of C.
* dimM[in] row number.
* dimN[in] column number.
* leadingN[in]the aligned column number
*
* @return void.
*/
void neon_dotMulVec(float *dst,
const float *src1,
const float *src2,
const float alpha,
const float beta,
const int dimM,
const int dimN,
const int leadingN)
{
float *mat0 = (float *)src1;
float *mat1 = (float *)src1 + leadingN;
float *mat2 = (float *)src1 + leadingN*2;
float *mat3 = (float *)src1 + leadingN*3;
float *mat4 = (float *)src2;
float *mat8 = dst;
float *mat9 = dst + leadingN;
float *matA = dst + leadingN*2;
float *matB = dst + leadingN*3;
int i = 0;
for (i = 0; i <= dimM - 4; i += 4)
{
int j = 0;
for (j = 0; j <= dimN - 4; j += 4)
{
float32x4_t q0 = vld1q_f32(mat0 + j);
float32x4_t q1 = vld1q_f32(mat1 + j);
float32x4_t q2 = vld1q_f32(mat2 + j);
float32x4_t q3 = vld1q_f32(mat3 + j);
float32x4_t q4 = vld1q_f32(mat4 + j);
float32x4_t q8 = vld1q_f32(mat8 + j);
float32x4_t q9 = vld1q_f32(mat9 + j);
float32x4_t qA = vld1q_f32(matA + j);
float32x4_t qB = vld1q_f32(matB + j);
q0 = vmulq_n_f32(q0, alpha);
q1 = vmulq_n_f32(q1, alpha);
q2 = vmulq_n_f32(q2, alpha);
q3 = vmulq_n_f32(q3, alpha);
q0 = vmulq_f32(q0, q4);
q1 = vmulq_f32(q1, q4);
q2 = vmulq_f32(q2, q4);
q3 = vmulq_f32(q3, q4);
q0 = vmlaq_n_f32(q0, q8, beta);
q1 = vmlaq_n_f32(q1, q9, beta);
q2 = vmlaq_n_f32(q2, qA, beta);
q3 = vmlaq_n_f32(q3, qB, beta);
vst1q_f32(mat8 + j, q0);
vst1q_f32(mat9 + j, q1);
vst1q_f32(matA + j, q2);
vst1q_f32(matB + j, q3);
}
for (; j < dimN; j++)
{
float a0 = mat8[j] * beta;
float a1 = mat9[j] * beta;
float a2 = matA[j] * beta;
float a3 = matB[j] * beta;
a0 += mat0[j] * mat4[j] * alpha;
a1 += mat1[j] * mat4[j] * alpha;
a2 += mat2[j] * mat4[j] * alpha;
a3 += mat3[j] * mat4[j] * alpha;
mat8[j] = a0;
mat9[j] = a1;
matA[j] = a2;
matB[j] = a3;
}
mat0 += leadingN * 4;
mat1 += leadingN * 4;
mat2 += leadingN * 4;
mat3 += leadingN * 4;
mat8 += leadingN * 4;
mat9 += leadingN * 4;
matA += leadingN * 4;
matB += leadingN * 4;
}
for (; i < dimM; i++)
{
int j = 0;
for (j = 0; j <= dimN - 4; j += 4)
{
float32x4_t q0 = vld1q_f32(mat0 + j);
float32x4_t q4 = vld1q_f32(mat4 + j);
float32x4_t q8 = vld1q_f32(mat8 + j);
q0 = vmulq_n_f32(q0, alpha);
q0 = vmulq_f32(q0, q4);
q0 = vmlaq_n_f32(q0, q8, beta);
vst1q_f32(mat8 + j, q0);
}
for (; j < dimN; j++)
{
float a0 = mat0[j] * mat4[j] * alpha + mat8[j] * beta;
mat8[j] = a0;
}
mat0 += leadingN;
mat8 += leadingN;
}
}
void ConvexPolygonShapeSW::project_range(const Vector3& p_normal, const Transform& p_transform, real_t &r_min, real_t &r_max) const {
int vertex_count=mesh.vertices.size();
if (vertex_count==0)
return;
const Vector3 *vrts=&mesh.vertices[0];
#ifndef NEON
for (int i=0;i<vertex_count;i++) {
float d=p_normal.dot( p_transform.xform( vrts[i] ) );
if (i==0 || d > r_max)
r_max=d;
if (i==0 || d < r_min)
r_min=d;
}
#else
int i;
Matrix3 m = p_transform.get_basis();
Vector3 o = p_transform.get_origin();
float32x4_t vo[4] = {{o[0],o[0],o[0],o[0]} , {o[1],o[1],o[1],o[1]} , {o[2],o[2],o[2],o[2]} , {o[3],o[3],o[3],o[3]}};
for (i=0;i<vertex_count-4;i+=4) { // as long as 4 calculations at a time are possible
/*_FORCE_INLINE_ Vector3 Transform::xform(const Vector3& p_vector) const {
return Vector3(
basis[0].dot(p_vector)+origin.x,
basis[1].dot(p_vector)+origin.y,
basis[2].dot(p_vector)+origin.z
);
}*/
//print_line("yay");
//float d1, d2, d3, d4;
//float f1_1, f1_2, f1_3, f1_4, f2_1, f2_2, f2_3, f2_4, f3_1, f3_2, f3_3, f3_4;
float32x4_t f1, f2, f3;
float32x4_t d;
float32x4_t vrts_x = {vrts[i].x, vrts[i+1].x, vrts[i+2].x, vrts[i+3].x};
float32x4_t vrts_y = {vrts[i].y, vrts[i+1].y, vrts[i+2].y, vrts[i+3].y};
float32x4_t vrts_z = {vrts[i].z, vrts[i+1].z, vrts[i+2].z, vrts[i+3].z};
/*f1_1 = m[0][0]*vrts[i][0];
f1_2 = m[0][0]*vrts[i+1][0];
f1_3 = m[0][0]*vrts[i+2][0];
f1_4 = m[0][0]*vrts[i+3][0];*/
//f1 = vrts_x * m[0][0];
f1 = vmulq_n_f32(vrts_x, m[0][0]);
/*f2_1 = m[1][0]*vrts[i][0];
f2_2 = m[1][0]*vrts[i+1][0];
f2_3 = m[1][0]*vrts[i+2][0];
f2_4 = m[1][0]*vrts[i+3][0];*/
//f2 = m[1][0] * vrts_x;
f2 = vmulq_n_f32(vrts_x, m[1][0]);
/*f3_1 = m[2][0]*vrts[i][0];
f3_2 = m[2][0]*vrts[i+1][0];
f3_3 = m[2][0]*vrts[i+2][0];
f3_4 = m[2][0]*vrts[i+3][0];*/
//f3 = m[2][0] * vrts_x;
f3 = vmulq_n_f32(vrts_x, m[2][0]);
/*f1_1 += m[0][1]*vrts[i][1];
f1_2 += m[0][1]*vrts[i+1][1];
f1_3 += m[0][1]*vrts[i+2][1];
f1_4 += m[0][1]*vrts[i+3][1];*/
//f1 += m[0][1] * vrts_y;
f1 += vmulq_n_f32(vrts_y, m[0][1]);
/*f2_1 += m[1][1]*vrts[i][1];
f2_2 += m[1][1]*vrts[i+1][1];
f2_3 += m[1][1]*vrts[i+2][1];
f2_4 += m[1][1]*vrts[i+3][1];*/
//f2 += m[1][1] * vrts_y;
f2 += vmulq_n_f32(vrts_y, m[1][1]);
/*f3_1 += m[2][1]*vrts[i][1];
f3_2 += m[2][1]*vrts[i+1][1];
f3_3 += m[2][1]*vrts[i+2][1];
f3_4 += m[2][1]*vrts[i+3][1];*/
//f3 += m[2][1] * vrts_y;
f3 += vmulq_n_f32(vrts_y, m[2][1]);
/*f1_1 += m[0][2]*vrts[i][2];
f1_2 += m[0][2]*vrts[i+1][2];
f1_3 += m[0][2]*vrts[i+2][2];
f1_4 += m[0][2]*vrts[i+3][2];*/
//f1 += m[0][2] * vrts_z;
f1 += vmulq_n_f32(vrts_z, m[0][2]);
/*f2_1 += m[1][2]*vrts[i][2];
f2_2 += m[1][2]*vrts[i+1][2];
f2_3 += m[1][2]*vrts[i+2][2];
f2_4 += m[1][2]*vrts[i+3][2];*/
//f2 += m[1][2] * vrts_z;
f2 += vmulq_n_f32(vrts_z, m[1][2]);
/*f3_1 += m[2][2]*vrts[i][2];
f3_2 += m[2][2]*vrts[i+1][2];
f3_3 += m[2][2]*vrts[i+2][2];
f3_4 += m[2][2]*vrts[i+3][2];*/
//f3 += m[2][2] * vrts_z;
f3 += vmulq_n_f32(vrts_z, m[2][2]);
/*f1_1 += o[0];
f1_2 += o[0];
f1_3 += o[0];
f1_4 += o[0];*/
f1 += vo[0];
/*f2_1 += o[1];
f2_2 += o[1];
f2_3 += o[1];
f2_4 += o[1];*/
f2 += vo[1];
/*f3_1 += o[2];
f3_2 += o[2];
f3_3 += o[2];
f3_4 += o[2];*/
f3 += vo[2];
/*d1 = f1_1*p_normal[0];
d2 = f1_2*p_normal[0];
d3 = f1_3*p_normal[0];
d4 = f1_4*p_normal[0];*/
d = vmulq_n_f32(f1 , p_normal[0]);
/*d1 += f2_1*p_normal[1];
d2 += f2_2*p_normal[1];
d3 += f2_3*p_normal[1];
d4 += f2_4*p_normal[1];*/
d += vmulq_n_f32(f2 , p_normal[1]);
/*d1 += f3_1*p_normal[2];
d2 += f3_2*p_normal[2];
d3 += f3_3*p_normal[2];
d4 += f3_4*p_normal[2];*/
d += vmulq_n_f32(f3 , p_normal[2]);
float *fd = (float *)&d;
if (i==0 || fd[0] > r_max)
r_max=fd[0];
if (i==0 || fd[0] < r_min)
r_min=fd[0];
if (i==0 || fd[1] > r_max)
r_max=fd[1];
if (i==0 || fd[1] < r_min)
r_min=fd[1];
if (i==0 || fd[2] > r_max)
r_max=fd[2];
if (i==0 || fd[2] < r_min)
r_min=fd[2];
if (i==0 || fd[3] > r_max)
r_max=fd[3];
if (i==0 || fd[3] < r_min)
r_min=fd[3];
}
for (i=i;i<vertex_count;i++) { // rest
/*_FORCE_INLINE_ Vector3 Transform::xform(const Vector3& p_vector) const {
return Vector3(
basis[0].dot(p_vector)+origin.x,
basis[1].dot(p_vector)+origin.y,
basis[2].dot(p_vector)+origin.z
);
}*/
float d;
Matrix3 m = p_transform.get_basis();
Vector3 o = p_transform.get_origin();
float f1, f2, f3;
f1 = m[0][0]*vrts[i][0];
f2 = m[1][0]*vrts[i][0];
f3 = m[2][0]*vrts[i][0];
f1 += m[0][1]*vrts[i][1];
f2 += m[1][1]*vrts[i][1];
f3 += m[2][1]*vrts[i][1];
f1 += m[0][2]*vrts[i][2];
f2 += m[1][2]*vrts[i][2];
f3 += m[2][2]*vrts[i][2];
f1 += o[0];
f2 += o[1];
f3 += o[2];
d = f1*p_normal[0];
d += f2*p_normal[1];
d += f3*p_normal[2];
if (i==0 || d > r_max)
r_max=d;
if (i==0 || d < r_min)
r_min=d;
}
#endif
}