Commit 0b0fb47c authored by James Willis's avatar James Willis
Browse files

Removed unneeded variables.

parent 4b3fc4a9
......@@ -321,8 +321,8 @@ __attribute__((always_inline)) INLINE void cache_read_two_partial_cells_sorted(
if (*last_pj + pad < cj->count) *last_pj += pad;
}
int first_pi_align = *first_pi;
int last_pj_align = *last_pj;
const int first_pi_align = *first_pi;
const int last_pj_align = *last_pj;
const struct part *restrict parts_i = ci->parts;
const struct part *restrict parts_j = cj->parts;
double loc[3];
......@@ -535,8 +535,8 @@ cache_read_two_partial_cells_sorted_force(
if (*last_pj + pad < cj->count) *last_pj += pad;
}
int first_pi_align = *first_pi;
int last_pj_align = *last_pj;
const int first_pi_align = *first_pi;
const int last_pj_align = *last_pj;
const struct part *restrict parts_i = ci->parts;
const struct part *restrict parts_j = cj->parts;
double loc[3];
......
......@@ -1002,14 +1002,12 @@ void runner_dopair1_density_vec(struct runner *r, struct cell *ci,
first_pi = min(first_pi, max_index_j[0]);
/* Read the needed particles into the two caches. */
int first_pi_align = first_pi;
int last_pj_align = last_pj;
cache_read_two_partial_cells_sorted(ci, cj, ci_cache, cj_cache, sort_i,
sort_j, shift, &first_pi_align,
&last_pj_align);
sort_j, shift, &first_pi,
&last_pj);
/* Get the number of particles read into the ci cache. */
int ci_cache_count = count_i - first_pi_align;
int ci_cache_count = count_i - first_pi;
if (active_ci) {
......@@ -1021,7 +1019,7 @@ void runner_dopair1_density_vec(struct runner *r, struct cell *ci,
if (!part_is_active_no_debug(pi, max_active_bin)) continue;
/* Set the cache index. */
int ci_cache_idx = pid - first_pi_align;
int ci_cache_idx = pid - first_pi;
/* Skip this particle if no particle in cj is within range of it. */
const float hi = ci_cache->h[ci_cache_idx];
......@@ -1071,7 +1069,7 @@ void runner_dopair1_density_vec(struct runner *r, struct cell *ci,
if (rem != 0) {
int pad = VEC_SIZE - rem;
if (exit_iteration_align + pad <= last_pj_align + 1)
if (exit_iteration_align + pad <= last_pj + 1)
exit_iteration_align += pad;
}
......@@ -1087,9 +1085,9 @@ void runner_dopair1_density_vec(struct runner *r, struct cell *ci,
#ifdef SWIFT_DEBUG_CHECKS
if (cj_cache_idx % VEC_SIZE != 0 || cj_cache_idx < 0 ||
cj_cache_idx + (VEC_SIZE - 1) > (last_pj_align + 1 + VEC_SIZE)) {
error("Unaligned read!!! cj_cache_idx=%d, last_pj_align=%d",
cj_cache_idx, last_pj_align);
cj_cache_idx + (VEC_SIZE - 1) > (last_pj + 1 + VEC_SIZE)) {
error("Unaligned read!!! cj_cache_idx=%d, last_pj=%d",
cj_cache_idx, last_pj);
}
#endif
......@@ -1198,7 +1196,7 @@ void runner_dopair1_density_vec(struct runner *r, struct cell *ci,
vector pix, piy, piz;
/* Convert exit iteration to cache indices. */
int exit_iteration_align = exit_iteration - first_pi_align;
int exit_iteration_align = exit_iteration - first_pi;
/* Pad the exit iteration align so cache reads are aligned. */
int rem = exit_iteration_align % VEC_SIZE;
......@@ -1214,11 +1212,11 @@ void runner_dopair1_density_vec(struct runner *r, struct cell *ci,
#ifdef SWIFT_DEBUG_CHECKS
if (ci_cache_idx % VEC_SIZE != 0 || ci_cache_idx < 0 ||
ci_cache_idx + (VEC_SIZE - 1) >
(count_i - first_pi_align + VEC_SIZE)) {
(count_i - first_pi + VEC_SIZE)) {
error(
"Unaligned read!!! ci_cache_idx=%d, first_pi_align=%d, "
"Unaligned read!!! ci_cache_idx=%d, first_pi=%d, "
"count_i=%d",
ci_cache_idx, first_pi_align, count_i);
ci_cache_idx, first_pi, count_i);
}
#endif
......@@ -1387,14 +1385,12 @@ void runner_dopair2_force_vec(struct runner *r, struct cell *ci,
first_pi = min(first_pi, max_index_j[0]);
/* Read the needed particles into the two caches. */
int first_pi_align = first_pi;
int last_pj_align = last_pj;
cache_read_two_partial_cells_sorted_force(ci, cj, ci_cache, cj_cache, sort_i,
sort_j, shift, &first_pi_align,
&last_pj_align);
sort_j, shift, &first_pi,
&last_pj);
/* Get the number of particles read into the ci cache. */
int ci_cache_count = count_i - first_pi_align;
int ci_cache_count = count_i - first_pi;
if (active_ci) {
......@@ -1406,7 +1402,7 @@ void runner_dopair2_force_vec(struct runner *r, struct cell *ci,
if (!part_is_active(pi, e)) continue;
/* Set the cache index. */
int ci_cache_idx = pid - first_pi_align;
int ci_cache_idx = pid - first_pi;
/* Skip this particle if no particle in cj is within range of it. */
const float hi = ci_cache->h[ci_cache_idx];
......@@ -1459,7 +1455,7 @@ void runner_dopair2_force_vec(struct runner *r, struct cell *ci,
if (rem != 0) {
int pad = VEC_SIZE - rem;
if (exit_iteration_align + pad <= last_pj_align + 1)
if (exit_iteration_align + pad <= last_pj + 1)
exit_iteration_align += pad;
}
......@@ -1475,9 +1471,9 @@ void runner_dopair2_force_vec(struct runner *r, struct cell *ci,
#ifdef SWIFT_DEBUG_CHECKS
if (cj_cache_idx % VEC_SIZE != 0 || cj_cache_idx < 0 ||
cj_cache_idx + (VEC_SIZE - 1) > (last_pj_align + 1 + VEC_SIZE)) {
error("Unaligned read!!! cj_cache_idx=%d, last_pj_align=%d",
cj_cache_idx, last_pj_align);
cj_cache_idx + (VEC_SIZE - 1) > (last_pj + 1 + VEC_SIZE)) {
error("Unaligned read!!! cj_cache_idx=%d, last_pj=%d",
cj_cache_idx, last_pj);
}
#endif
......@@ -1600,7 +1596,7 @@ void runner_dopair2_force_vec(struct runner *r, struct cell *ci,
entropy_dtSum.v = vec_setzero();
/* Convert exit iteration to cache indices. */
int exit_iteration_align = exit_iteration - first_pi_align;
int exit_iteration_align = exit_iteration - first_pi;
/* Pad the exit iteration align so cache reads are aligned. */
int rem = exit_iteration_align % VEC_SIZE;
......
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