Pad cache with better fake particles so that they won't be in range of real...

Pad cache with better fake particles so that they won't be in range of real particles. Sedov blast now runs to completion but timesteps don't match serial output.

src/cache.h

@@ -409,12 +409,15 @@ __attribute__((always_inline)) INLINE void cache_read_two_partial_cells_sorted(
 
   /* Pad cache with fake particles that exist outside the cell so will not
    * interact.*/
-  float fake_pix = 2.0f * parts_i[sort_i[ci->count - 1].i].x[0];
+  const float max_dx = max(ci->dx_max_part, cj->dx_max_part);
+  const float pos_padded[3] = {-(2. * ci->width[0] + max_dx), -(2. * ci->width[1] + max_dx),
+                               -(2. * ci->width[2] + max_dx)};
+
   for (int i = ci->count - first_pi_align;
        i < ci->count - first_pi_align + VEC_SIZE; i++) {
-    x[i] = fake_pix;
-    y[i] = 1.f;
-    z[i] = 1.f;
+    x[i] = pos_padded[0];
+    y[i] = pos_padded[1];
+    z[i] = pos_padded[2];
     h[i] = 1.f;
 
     m[i] = 1.f;
@@ -479,11 +482,12 @@ __attribute__((always_inline)) INLINE void cache_read_two_partial_cells_sorted(
 
   /* Pad cache with fake particles that exist outside the cell so will not
    * interact.*/
-  float fake_pjx = 2.0f * cj->parts[sort_j[cj->count - 1].i].x[0];
+  const float pos_padded_j[3] = {-(2. * cj->width[0] + max_dx), -(2. * cj->width[1] + max_dx),
+                               -(2. * cj->width[2] + max_dx)};
   for (int i = last_pj_align + 1; i < last_pj_align + 1 + VEC_SIZE; i++) {
-    xj[i] = fake_pjx;
-    yj[i] = 1.f;
-    zj[i] = 1.f;
+    xj[i] = pos_padded_j[0];
+    yj[i] = pos_padded_j[1];
+    zj[i] = pos_padded_j[2];
     hj[i] = 1.f;
 
     mj[i] = 1.f;
@@ -593,12 +597,15 @@ cache_read_two_partial_cells_sorted_force(
 
   /* Pad cache with fake particles that exist outside the cell so will not
    * interact.*/
-  float fake_pix = 2.0f * ci->parts[sort_i[ci->count - 1].i].x[0];
+  const float max_dx = max(ci->dx_max_part, cj->dx_max_part);
+  const float pos_padded[3] = {-(2. * ci->width[0] + max_dx), -(2. * ci->width[1] + max_dx),
+                               -(2. * ci->width[2] + max_dx)};
+
   for (int i = ci->count - first_pi_align;
        i < ci->count - first_pi_align + VEC_SIZE; i++) {
-    x[i] = fake_pix;
-    y[i] = 1.f;
-    z[i] = 1.f;
+    x[i] = pos_padded[0];
+    y[i] = pos_padded[1];
+    z[i] = pos_padded[2];
     h[i] = FLT_MIN;
 
     m[i] = 1.f;
@@ -654,11 +661,13 @@ cache_read_two_partial_cells_sorted_force(
 
   /* Pad cache with fake particles that exist outside the cell so will not
    * interact.*/
-  float fake_pjx = 2.0f * cj->parts[sort_j[cj->count - 1].i].x[0];
+  const float pos_padded_j[3] = {-(2. * cj->width[0] + max_dx), -(2. * cj->width[1] + max_dx),
+                               -(2. * cj->width[2] + max_dx)};
+
   for (int i = last_pj_align + 1; i < last_pj_align + 1 + VEC_SIZE; i++) {
-    xj[i] = fake_pjx;
-    yj[i] = 1.f;
-    zj[i] = 1.f;
+    xj[i] = pos_padded_j[0];
+    yj[i] = pos_padded_j[1];
+    zj[i] = pos_padded_j[2];
     hj[i] = FLT_MIN;
 
     mj[i] = 1.f;