64-BitBrainstorm_2026/AMSS-NCKU
8
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Replace nested OutBdLow2Hi loops with batch calls in RestrictProlong

Browse Source 
The 8 nested while(Ppc){while(Pp){OutBdLow2Hi(single,single,...)}}
loops across RestrictProlong (3 overloads) and ProlongRestrict each
produced N_c × N_f separate transfer() → MPI_Waitall barriers.
Replace with the existing batch OutBdLow2Hi(MyList<Patch>*,...) which
merges all patch pairs into a single transfer() call with 1 MPI_Waitall.

Also add Restrict_cached, OutBdLow2Hi_cached, OutBdLow2Himix_cached
to Parallel (unused for now — kept as infrastructure for future use).

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
This commit is contained in:
CGH0S7
2026-02-11 16:09:08 +08:00
parent e09ae438a2
commit 5c1790277b
3 changed files with 222 additions and 96 deletions
Download Patch File Download Diff File Expand all files Collapse all files

112
AMSS_NCKU_source/bssn_class.C
View File

@@ -5819,21 +5819,11 @@ void bssn_class::RestrictProlong(int lev, int YN, bool BB,
#endif
#if (RPB == 0)
Ppc = GH->PatL[lev - 1];
while (Ppc)
{
Pp = GH->PatL[lev];
while (Pp)
{
#if (MIXOUTB == 0)
Parallel::OutBdLow2Hi(Ppc->data, Pp->data, SynchList_pre, SL, Symmetry);
Parallel::OutBdLow2Hi(GH->PatL[lev - 1], GH->PatL[lev], SynchList_pre, SL, Symmetry);
#elif (MIXOUTB == 1)
Parallel::OutBdLow2Himix(Ppc->data, Pp->data, SynchList_pre, SL, Symmetry);
Parallel::OutBdLow2Himix(GH->PatL[lev - 1], GH->PatL[lev], SynchList_pre, SL, Symmetry);
#endif
Pp = Pp->next;
}
Ppc = Ppc->next;
}
#elif (RPB == 1)
// Parallel::OutBdLow2Hi_bam(GH->PatL[lev-1],GH->PatL[lev],SynchList_pre,SL,Symmetry);
Parallel::OutBdLow2Hi_bam(GH->PatL[lev - 1], GH->PatL[lev], SynchList_pre, SL, GH->bdsul[lev], Symmetry);
@@ -5880,21 +5870,11 @@ void bssn_class::RestrictProlong(int lev, int YN, bool BB,
#endif
#if (RPB == 0)
Ppc = GH->PatL[lev - 1];
while (Ppc)
{
Pp = GH->PatL[lev];
while (Pp)
{
#if (MIXOUTB == 0)
Parallel::OutBdLow2Hi(Ppc->data, Pp->data, SL, SL, Symmetry);
Parallel::OutBdLow2Hi(GH->PatL[lev - 1], GH->PatL[lev], SL, SL, Symmetry);
#elif (MIXOUTB == 1)
Parallel::OutBdLow2Himix(Ppc->data, Pp->data, SL, SL, Symmetry);
Parallel::OutBdLow2Himix(GH->PatL[lev - 1], GH->PatL[lev], SL, SL, Symmetry);
#endif
Pp = Pp->next;
}
Ppc = Ppc->next;
}
#elif (RPB == 1)
// Parallel::OutBdLow2Hi_bam(GH->PatL[lev-1],GH->PatL[lev],SL,SL,Symmetry);
Parallel::OutBdLow2Hi_bam(GH->PatL[lev - 1], GH->PatL[lev], SL, SL, GH->bdsul[lev], Symmetry);
@@ -5969,21 +5949,11 @@ void bssn_class::RestrictProlong_aux(int lev, int YN, bool BB,
Parallel::Sync_cached(GH->PatL[lev - 1], SynchList_pre, Symmetry, sync_cache_rp_coarse[lev]);
#if (RPB == 0)
Ppc = GH->PatL[lev - 1];
while (Ppc)
{
Pp = GH->PatL[lev];
while (Pp)
{
#if (MIXOUTB == 0)
Parallel::OutBdLow2Hi(Ppc->data, Pp->data, SynchList_pre, SL, Symmetry);
Parallel::OutBdLow2Hi(GH->PatL[lev - 1], GH->PatL[lev], SynchList_pre, SL, Symmetry);
#elif (MIXOUTB == 1)
Parallel::OutBdLow2Himix(Ppc->data, Pp->data, SynchList_pre, SL, Symmetry);
Parallel::OutBdLow2Himix(GH->PatL[lev - 1], GH->PatL[lev], SynchList_pre, SL, Symmetry);
#endif
Pp = Pp->next;
}
Ppc = Ppc->next;
}
#elif (RPB == 1)
// Parallel::OutBdLow2Hi_bam(GH->PatL[lev-1],GH->PatL[lev],SynchList_pre,SL,Symmetry);
Parallel::OutBdLow2Hi_bam(GH->PatL[lev - 1], GH->PatL[lev], SynchList_pre, SL, GH->bdsul[lev], Symmetry);
@@ -6001,21 +5971,11 @@ void bssn_class::RestrictProlong_aux(int lev, int YN, bool BB,
Parallel::Sync_cached(GH->PatL[lev - 1], SL, Symmetry, sync_cache_rp_coarse[lev]);
#if (RPB == 0)
Ppc = GH->PatL[lev - 1];
while (Ppc)
{
Pp = GH->PatL[lev];
while (Pp)
{
#if (MIXOUTB == 0)
Parallel::OutBdLow2Hi(Ppc->data, Pp->data, SL, SL, Symmetry);
Parallel::OutBdLow2Hi(GH->PatL[lev - 1], GH->PatL[lev], SL, SL, Symmetry);
#elif (MIXOUTB == 1)
Parallel::OutBdLow2Himix(Ppc->data, Pp->data, SL, SL, Symmetry);
Parallel::OutBdLow2Himix(GH->PatL[lev - 1], GH->PatL[lev], SL, SL, Symmetry);
#endif
Pp = Pp->next;
}
Ppc = Ppc->next;
}
#elif (RPB == 1)
// Parallel::OutBdLow2Hi_bam(GH->PatL[lev-1],GH->PatL[lev],SL,SL,Symmetry);
Parallel::OutBdLow2Hi_bam(GH->PatL[lev - 1], GH->PatL[lev], SL, SL, GH->bdsul[lev], Symmetry);
@@ -6076,21 +6036,11 @@ void bssn_class::RestrictProlong(int lev, int YN, bool BB)
Parallel::Sync_cached(GH->PatL[lev - 1], SynchList_pre, Symmetry, sync_cache_rp_coarse[lev]);
#if (RPB == 0)
Ppc = GH->PatL[lev - 1];
while (Ppc)
{
Pp = GH->PatL[lev];
while (Pp)
{
#if (MIXOUTB == 0)
Parallel::OutBdLow2Hi(Ppc->data, Pp->data, SynchList_pre, SynchList_cor, Symmetry);
Parallel::OutBdLow2Hi(GH->PatL[lev - 1], GH->PatL[lev], SynchList_pre, SynchList_cor, Symmetry);
#elif (MIXOUTB == 1)
Parallel::OutBdLow2Himix(Ppc->data, Pp->data, SynchList_pre, SynchList_cor, Symmetry);
Parallel::OutBdLow2Himix(GH->PatL[lev - 1], GH->PatL[lev], SynchList_pre, SynchList_cor, Symmetry);
#endif
Pp = Pp->next;
}
Ppc = Ppc->next;
}
#elif (RPB == 1)
// Parallel::OutBdLow2Hi_bam(GH->PatL[lev-1],GH->PatL[lev],SynchList_pre,SynchList_cor,Symmetry);
Parallel::OutBdLow2Hi_bam(GH->PatL[lev - 1], GH->PatL[lev], SynchList_pre, SynchList_cor, GH->bdsul[lev], Symmetry);
@@ -6110,21 +6060,11 @@ void bssn_class::RestrictProlong(int lev, int YN, bool BB)
Parallel::Sync_cached(GH->PatL[lev - 1], StateList, Symmetry, sync_cache_rp_coarse[lev]);
#if (RPB == 0)
Ppc = GH->PatL[lev - 1];
while (Ppc)
{
Pp = GH->PatL[lev];
while (Pp)
{
#if (MIXOUTB == 0)
Parallel::OutBdLow2Hi(Ppc->data, Pp->data, StateList, SynchList_cor, Symmetry);
Parallel::OutBdLow2Hi(GH->PatL[lev - 1], GH->PatL[lev], StateList, SynchList_cor, Symmetry);
#elif (MIXOUTB == 1)
Parallel::OutBdLow2Himix(Ppc->data, Pp->data, StateList, SynchList_cor, Symmetry);
Parallel::OutBdLow2Himix(GH->PatL[lev - 1], GH->PatL[lev], StateList, SynchList_cor, Symmetry);
#endif
Pp = Pp->next;
}
Ppc = Ppc->next;
}
#elif (RPB == 1)
// Parallel::OutBdLow2Hi_bam(GH->PatL[lev-1],GH->PatL[lev],StateList,SynchList_cor,Symmetry);
Parallel::OutBdLow2Hi_bam(GH->PatL[lev - 1], GH->PatL[lev], StateList, SynchList_cor, GH->bdsul[lev], Symmetry);
@@ -6161,21 +6101,11 @@ void bssn_class::ProlongRestrict(int lev, int YN, bool BB)
}
#if (RPB == 0)
Ppc = GH->PatL[lev - 1];
while (Ppc)
{
Pp = GH->PatL[lev];
while (Pp)
{
#if (MIXOUTB == 0)
Parallel::OutBdLow2Hi(Ppc->data, Pp->data, SynchList_pre, SynchList_cor, Symmetry);
Parallel::OutBdLow2Hi(GH->PatL[lev - 1], GH->PatL[lev], SynchList_pre, SynchList_cor, Symmetry);
#elif (MIXOUTB == 1)
Parallel::OutBdLow2Himix(Ppc->data, Pp->data, SynchList_pre, SynchList_cor, Symmetry);
Parallel::OutBdLow2Himix(GH->PatL[lev - 1], GH->PatL[lev], SynchList_pre, SynchList_cor, Symmetry);
#endif
Pp = Pp->next;
}
Ppc = Ppc->next;
}
#elif (RPB == 1)
// Parallel::OutBdLow2Hi_bam(GH->PatL[lev-1],GH->PatL[lev],SynchList_pre,SynchList_cor,Symmetry);
Parallel::OutBdLow2Hi_bam(GH->PatL[lev - 1], GH->PatL[lev], SynchList_pre, SynchList_cor, GH->bdsul[lev], Symmetry);
@@ -6184,21 +6114,11 @@ void bssn_class::ProlongRestrict(int lev, int YN, bool BB)
else // no time refinement levels and for all same time levels
{
#if (RPB == 0)
Ppc = GH->PatL[lev - 1];
while (Ppc)
{
Pp = GH->PatL[lev];
while (Pp)
{
#if (MIXOUTB == 0)
Parallel::OutBdLow2Hi(Ppc->data, Pp->data, StateList, SynchList_cor, Symmetry);
Parallel::OutBdLow2Hi(GH->PatL[lev - 1], GH->PatL[lev], StateList, SynchList_cor, Symmetry);
#elif (MIXOUTB == 1)
Parallel::OutBdLow2Himix(Ppc->data, Pp->data, StateList, SynchList_cor, Symmetry);
Parallel::OutBdLow2Himix(GH->PatL[lev - 1], GH->PatL[lev], StateList, SynchList_cor, Symmetry);
#endif
Pp = Pp->next;
}
Ppc = Ppc->next;
}
#elif (RPB == 1)
// Parallel::OutBdLow2Hi_bam(GH->PatL[lev-1],GH->PatL[lev],StateList,SynchList_cor,Symmetry);
Parallel::OutBdLow2Hi_bam(GH->PatL[lev - 1], GH->PatL[lev], StateList, SynchList_cor, GH->bdsul[lev], Symmetry);