Skip to content
GitLab
Commit 1d0da278 authored by James Willis's avatar James Willis
Browse files

Use 2 vectors to find neighbours and process interactions.

parent 686314cf
Hide whitespace changes
Inline Side-by-side
src/runner_doiact_vec.c
View file @ 1d0da278
......@@ -518,7 +518,7 @@ __attribute__((always_inline)) INLINE static void storeForceInteractions(
#endif /* defined(HAVE_AVX2) || defined(HAVE_AVX512_F) */
/* Flush the c2 cache if it has reached capacity. */
if (*icount >= (C2_CACHE_SIZE - (1 * VEC_SIZE))) {
if (*icount >= (C2_CACHE_SIZE - (2 * VEC_SIZE))) {
error("Flushing interaction cache...");
......@@ -528,37 +528,25 @@ __attribute__((always_inline)) INLINE static void storeForceInteractions(
calcRemForceInteractions(int_cache, *icount, a_hydro_xSum, a_hydro_ySum, a_hydro_zSum,
h_dtSum, v_sigSum, entropy_dtSum, v_hi_inv,
v_vix, v_viy, v_viz, *v_rhoi, *v_grad_hi, *v_pOrhoi2, *v_balsara_i, *v_ci,
&icount_align, 1);
&icount_align, 2);
vector int_mask;//, int_mask2;
vector int_mask, int_mask2;
int_mask.m = vec_setint1(0xFFFFFFFF);
//int_mask2.m = vec_setint1(0xFFFFFFFF);
int_mask2.m = vec_setint1(0xFFFFFFFF);
/* Perform interactions. */
for (int pjd = 0; pjd < icount_align; pjd += (1 * VEC_SIZE)) {
vector r2, dx, dy, dz;
vector vjx, vjy, vjz, mj, rhoj, grad_hj, pOrhoj2, balsara_j, cj, hj_inv;
r2.v = vec_load(&int_cache->r2q[pjd]);
dx.v = vec_load(&int_cache->dxq[pjd]);
dy.v = vec_load(&int_cache->dyq[pjd]);
dz.v = vec_load(&int_cache->dzq[pjd]);
vjx.v = vec_load(&int_cache->vxq[pjd]);
vjy.v = vec_load(&int_cache->vyq[pjd]);
vjz.v = vec_load(&int_cache->vzq[pjd]);
mj.v = vec_load(&int_cache->mq[pjd]);
rhoj.v = vec_load(&int_cache->rhoq[pjd]);
grad_hj.v = vec_load(&int_cache->grad_hq[pjd]);
pOrhoj2.v = vec_load(&int_cache->pOrho2q[pjd]);
balsara_j.v = vec_load(&int_cache->balsaraq[pjd]);
cj.v = vec_load(&int_cache->soundspeedq[pjd]);
hj_inv.v = vec_load(&int_cache->h_invq[pjd]);
runner_iact_nonsym_1_vec_force_2(
&r2, &dx, &dy, &dz, &v_vix, &v_viy, &v_viz, v_rhoi, v_grad_hi, v_pOrhoi2, v_balsara_i, v_ci,
&vjx, &vjy, &vjz, &rhoj, &grad_hj, &pOrhoj2, &balsara_j, &cj, &mj, v_hi_inv, hj_inv,
a_hydro_xSum, a_hydro_ySum, a_hydro_zSum, h_dtSum, v_sigSum, entropy_dtSum, int_mask);
for (int pjd = 0; pjd < icount_align; pjd += (2 * VEC_SIZE)) {
runner_iact_nonsym_2_vec_force(
&int_cache->r2q[pjd], &int_cache->dxq[pjd], &int_cache->dyq[pjd], &int_cache->dzq[pjd], &v_vix, &v_viy, &v_viz, v_rhoi, v_grad_hi, v_pOrhoi2, v_balsara_i, v_ci,
&int_cache->vxq[pjd], &int_cache->vyq[pjd], &int_cache->vzq[pjd], &int_cache->rhoq[pjd], &int_cache->grad_hq[pjd], &int_cache->pOrho2q[pjd], &int_cache->balsaraq[pjd], &int_cache->soundspeedq[pjd], &int_cache->mq[pjd], v_hi_inv, &int_cache->h_invq[pjd],
a_hydro_xSum, a_hydro_ySum, a_hydro_zSum, h_dtSum, v_sigSum, entropy_dtSum, int_mask, int_mask2
#ifdef HAVE_AVX512_F
knl_mask, knl_mask2);
#else
);
#endif
}
......@@ -1761,7 +1749,7 @@ __attribute__((always_inline)) INLINE void runner_doself2_force_vec_3(
int doi_mask;
struct part *restrict pi;
int count_align;
int num_vec_proc = 1;//NUM_VEC_PROC;
int num_vec_proc = 2;//NUM_VEC_PROC;
struct part *restrict parts = c->parts;
const int count = c->count;
......@@ -1850,14 +1838,17 @@ __attribute__((always_inline)) INLINE void runner_doself2_force_vec_3(
cell_cache->x[i] = pix.f[0];
cell_cache->y[i] = piy.f[0];
cell_cache->z[i] = piz.f[0];
cell_cache->h[i] = 1.f;
}
}
vector pjx, pjy, pjz;
vector pjvx, pjvy, pjvz, mj, hj, hjg2;
vector v_rhoj, v_grad_hj, v_pOrhoj2, v_balsara_j, v_cj;
//vector pjx2, pjy2, pjz2;
//vector pjvx2, pjvy2, pjvz2, mj2, hj_2, hjg2_2;
vector pjx2, pjy2, pjz2;
vector pjvx2, pjvy2, pjvz2, mj2, hj_2, hjg2_2;
vector v_rhoj_2, v_grad_hj_2, v_pOrhoj2_2, v_balsara_j_2, v_cj_2;
/* Find all of particle pi's interacions and store needed values in the
* secondary cache.*/
......@@ -1875,44 +1866,53 @@ __attribute__((always_inline)) INLINE void runner_doself2_force_vec_3(
hj.v = vec_load(&cell_cache->h[pjd]);
hjg2.v = vec_mul(vec_mul(hj.v,hj.v), kernel_gamma2_vec.v);
/* TODO: Don't load unneeded quantities until you have to in storeInteractions()!!! */
v_rhoj.v = vec_load(&cell_cache->rho[pjd]);
v_grad_hj.v = vec_load(&cell_cache->grad_h[pjd]);
v_pOrhoj2.v = vec_load(&cell_cache->pOrho2[pjd]);
v_balsara_j.v = vec_load(&cell_cache->balsara[pjd]);
v_cj.v = vec_load(&cell_cache->soundspeed[pjd]);
//pjx2.v = vec_load(&cell_cache->x[pjd + VEC_SIZE]);
//pjy2.v = vec_load(&cell_cache->y[pjd + VEC_SIZE]);
//pjz2.v = vec_load(&cell_cache->z[pjd + VEC_SIZE]);
//pjvx2.v = vec_load(&cell_cache->vx[pjd + VEC_SIZE]);
//pjvy2.v = vec_load(&cell_cache->vy[pjd + VEC_SIZE]);
//pjvz2.v = vec_load(&cell_cache->vz[pjd + VEC_SIZE]);
//mj2.v = vec_load(&cell_cache->m[pjd + VEC_SIZE]);
pjx2.v = vec_load(&cell_cache->x[pjd + VEC_SIZE]);
pjy2.v = vec_load(&cell_cache->y[pjd + VEC_SIZE]);
pjz2.v = vec_load(&cell_cache->z[pjd + VEC_SIZE]);
pjvx2.v = vec_load(&cell_cache->vx[pjd + VEC_SIZE]);
pjvy2.v = vec_load(&cell_cache->vy[pjd + VEC_SIZE]);
pjvz2.v = vec_load(&cell_cache->vz[pjd + VEC_SIZE]);
mj2.v = vec_load(&cell_cache->m[pjd + VEC_SIZE]);
//hj_2.v = vec_load(&cell_cache->h[pjd + VEC_SIZE]);
//hjg2_2.v = vec_mul(vec_mul(hj_2.v,hj_2.v), kernel_gamma2_vec.v);
hj_2.v = vec_load(&cell_cache->h[pjd + VEC_SIZE]);
hjg2_2.v = vec_mul(vec_mul(hj_2.v,hj_2.v), kernel_gamma2_vec.v);
vector v_hj_inv;
v_rhoj_2.v = vec_load(&cell_cache->rho[pjd + VEC_SIZE]);
v_grad_hj_2.v = vec_load(&cell_cache->grad_h[pjd + VEC_SIZE]);
v_pOrhoj2_2.v = vec_load(&cell_cache->pOrho2[pjd + VEC_SIZE]);
v_balsara_j_2.v = vec_load(&cell_cache->balsara[pjd + VEC_SIZE]);
v_cj_2.v = vec_load(&cell_cache->soundspeed[pjd + VEC_SIZE]);
vector v_hj_inv, v_hj_inv_2;
v_hj_inv = vec_reciprocal(hj);
v_hj_inv_2 = vec_reciprocal(hj_2);
/* Compute the pairwise distance. */
vector v_dx_tmp, v_dy_tmp, v_dz_tmp;
//vector v_dx_tmp2, v_dy_tmp2, v_dz_tmp2, v_r2_2;
vector v_dx_tmp2, v_dy_tmp2, v_dz_tmp2, v_r2_2;
v_dx_tmp.v = vec_sub(pix.v, pjx.v);
//v_dx_tmp2.v = vec_sub(pix.v, pjx2.v);
v_dx_tmp2.v = vec_sub(pix.v, pjx2.v);
v_dy_tmp.v = vec_sub(piy.v, pjy.v);
//v_dy_tmp2.v = vec_sub(piy.v, pjy2.v);
v_dy_tmp2.v = vec_sub(piy.v, pjy2.v);
v_dz_tmp.v = vec_sub(piz.v, pjz.v);
//v_dz_tmp2.v = vec_sub(piz.v, pjz2.v);
v_dz_tmp2.v = vec_sub(piz.v, pjz2.v);
v_r2.v = vec_mul(v_dx_tmp.v, v_dx_tmp.v);
//v_r2_2.v = vec_mul(v_dx_tmp2.v, v_dx_tmp2.v);
v_r2_2.v = vec_mul(v_dx_tmp2.v, v_dx_tmp2.v);
v_r2.v = vec_fma(v_dy_tmp.v, v_dy_tmp.v, v_r2.v);
//v_r2_2.v = vec_fma(v_dy_tmp2.v, v_dy_tmp2.v, v_r2_2.v);
v_r2_2.v = vec_fma(v_dy_tmp2.v, v_dy_tmp2.v, v_r2_2.v);
v_r2.v = vec_fma(v_dz_tmp.v, v_dz_tmp.v, v_r2.v);
//v_r2_2.v = vec_fma(v_dz_tmp2.v, v_dz_tmp2.v, v_r2_2.v);
v_r2_2.v = vec_fma(v_dz_tmp2.v, v_dz_tmp2.v, v_r2_2.v);
/* Form a mask from r2 < hig2 and r2 > 0.*/
#ifdef HAVE_AVX512_F
......@@ -1930,8 +1930,8 @@ __attribute__((always_inline)) INLINE void runner_doself2_force_vec_3(
#else
vector v_doi_mask, v_doi_mask_check, v_doi_N3_mask;
//vector v_doi_mask2, v_doi_mask2_check, v_doi_N3_mask2;
//int doi_mask2;
vector v_doi_mask2, v_doi_mask2_check, v_doi_N3_mask2;
int doi_mask2;
/* Form r2 > 0 mask, r2 < hig2 mask and r2 < hjg2 mask. */
v_doi_mask_check.v = vec_cmp_gt(v_r2.v, vec_setzero());
......@@ -1939,15 +1939,16 @@ __attribute__((always_inline)) INLINE void runner_doself2_force_vec_3(
v_doi_N3_mask.v = vec_cmp_lt(v_r2.v, hjg2.v);
/* Form r2 > 0 mask and r2 < hig2 mask. */
//v_doi_mask2_check.v = vec_cmp_gt(v_r2_2.v, vec_setzero());
//v_doi_mask2.v = vec_cmp_lt(v_r2_2.v, v_hig2.v);
//v_doi_N3_mask2.v = vec_cmp_lt(v_r2_2.v, v_hjg2_2.v);
v_doi_mask2_check.v = vec_cmp_gt(v_r2_2.v, vec_setzero());
v_doi_mask2.v = vec_cmp_lt(v_r2_2.v, v_hig2.v);
v_doi_N3_mask2.v = vec_cmp_lt(v_r2_2.v, hjg2_2.v);
v_doi_mask.v = vec_and(vec_add(v_doi_mask.v, v_doi_N3_mask.v), v_doi_mask_check.v);
v_doi_mask2.v = vec_and(vec_add(v_doi_mask2.v, v_doi_N3_mask2.v), v_doi_mask2_check.v);
/* Combine two masks and form integer mask. */
doi_mask = vec_cmp_result(v_doi_mask.v);
//doi_mask2 = vec_cmp_result(vec_add(vec_and(v_doi_mask2.v, v_doi_mask2_check.v), v_doi_N3_mask2.v));
doi_mask2 = vec_cmp_result(v_doi_mask2.v);
#endif /* HAVE_AVX512_F */
......@@ -1955,15 +1956,20 @@ __attribute__((always_inline)) INLINE void runner_doself2_force_vec_3(
* cache. */
if (doi_mask) {
//for(int k=0; k<VEC_SIZE; k++) {
// if( v_r2.f[k] == 0.f) v_r2.f[k] = 1.f;
//}
storeForceInteractions(doi_mask, pjd, &v_r2, &v_dx_tmp, &v_dy_tmp, &v_dz_tmp,
&mj, &pjvx, &pjvy, &pjvz, &v_rhoj, &v_grad_hj, &v_pOrhoj2, &v_balsara_j, &v_cj, &v_hj_inv, cell_cache, &int_cache,
&icount, &a_hydro_xSum, &a_hydro_ySum, &a_hydro_zSum,
&h_dtSum, &v_sigSum, &entropy_dtSum,
v_hi_inv, v_vix, v_viy, v_viz, &v_rhoi, &v_grad_hi, &v_pOrhoi2, &v_balsara_i, &v_ci);
}
if (doi_mask2) {
storeForceInteractions(doi_mask2, pjd + VEC_SIZE, &v_r2_2, &v_dx_tmp2, &v_dy_tmp2, &v_dz_tmp2,
&mj2, &pjvx2, &pjvy2, &pjvz2, &v_rhoj_2, &v_grad_hj_2, &v_pOrhoj2_2, &v_balsara_j_2, &v_cj_2, &v_hj_inv_2, cell_cache, &int_cache,
&icount, &a_hydro_xSum, &a_hydro_ySum, &a_hydro_zSum,
&h_dtSum, &v_sigSum, &entropy_dtSum,
v_hi_inv, v_vix, v_viy, v_viz, &v_rhoi, &v_grad_hi, &v_pOrhoi2, &v_balsara_i, &v_ci);
}
} /* Loop over all other particles. */
......@@ -1991,7 +1997,7 @@ __attribute__((always_inline)) INLINE void runner_doself2_force_vec_3(
/* Perform interaction with 2 vectors. */
for (int pjd = 0; pjd < icount_align; pjd += (2 * VEC_SIZE)) {
runner_iact_nonsym_2_vec_force(
runner_iact_nonsym_2_vec_force(
&int_cache.r2q[pjd], &int_cache.dxq[pjd], &int_cache.dyq[pjd], &int_cache.dzq[pjd], &v_vix, &v_viy, &v_viz, &v_rhoi, &v_grad_hi, &v_pOrhoi2, &v_balsara_i, &v_ci,
&int_cache.vxq[pjd], &int_cache.vyq[pjd], &int_cache.vzq[pjd], &int_cache.rhoq[pjd], &int_cache.grad_hq[pjd], &int_cache.pOrho2q[pjd], &int_cache.balsaraq[pjd], &int_cache.soundspeedq[pjd], &int_cache.mq[pjd], v_hi_inv, &int_cache.h_invq[pjd],
&a_hydro_xSum, &a_hydro_ySum, &a_hydro_zSum, &h_dtSum, &v_sigSum, &entropy_dtSum, int_mask, int_mask2
......
Supports Markdown
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment