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

Checkpoint Z4C CUDA resident sync progress

Browse Source
This commit is contained in:
CGH0S7
2026-05-02 10:53:52 +08:00
parent ba61702fc0
commit 52beb4d153
4 changed files with 1075 additions and 134 deletions
Download Patch File Download Diff File Expand all files Collapse all files

104
AMSS_NCKU_source/z4c_rhs_cuda.h
View File

@@ -53,6 +53,14 @@ int z4c_cuda_pack_state_batch_to_host_buffer(void *block_tag,
int i0, int j0, int k0,
int sx, int sy, int sz);
int z4c_cuda_pack_state_batch_to_host_buffer_for_host_views(void *block_tag,
double **state_host_key,
int state_count,
double *host_buffer,
int *ex,
int i0, int j0, int k0,
int sx, int sy, int sz);
int z4c_cuda_unpack_state_batch_from_host_buffer(void *block_tag,
int state_count,
double *host_buffer,
@@ -60,6 +68,14 @@ int z4c_cuda_unpack_state_batch_from_host_buffer(void *block_tag,
int i0, int j0, int k0,
int sx, int sy, int sz);
int z4c_cuda_unpack_state_batch_from_host_buffer_for_host_views(void *block_tag,
double **state_host_key,
int state_count,
double *host_buffer,
int *ex,
int i0, int j0, int k0,
int sx, int sy, int sz);
int z4c_cuda_pack_state_batch_to_device_buffer(void *block_tag,
int state_count,
double *device_buffer,
@@ -67,6 +83,14 @@ int z4c_cuda_pack_state_batch_to_device_buffer(void *block_tag,
int i0, int j0, int k0,
int sx, int sy, int sz);
int z4c_cuda_pack_state_batch_to_device_buffer_for_host_views(void *block_tag,
double **state_host_key,
int state_count,
double *device_buffer,
int *ex,
int i0, int j0, int k0,
int sx, int sy, int sz);
int z4c_cuda_unpack_state_batch_from_device_buffer(void *block_tag,
int state_count,
double *device_buffer,
@@ -74,6 +98,14 @@ int z4c_cuda_unpack_state_batch_from_device_buffer(void *block_tag,
int i0, int j0, int k0,
int sx, int sy, int sz);
int z4c_cuda_unpack_state_batch_from_device_buffer_for_host_views(void *block_tag,
double **state_host_key,
int state_count,
double *device_buffer,
int *ex,
int i0, int j0, int k0,
int sx, int sy, int sz);
int z4c_cuda_pack_state_segments_to_device_buffer(void *block_tag,
int state_count,
double *device_buffer,
@@ -81,6 +113,14 @@ int z4c_cuda_pack_state_segments_to_device_buffer(void *block_tag,
int segment_count,
const int *segment_meta);
int z4c_cuda_pack_state_segments_to_device_buffer_for_host_views(void *block_tag,
double **state_host_key,
int state_count,
double *device_buffer,
int *ex,
int segment_count,
const int *segment_meta);
int z4c_cuda_unpack_state_segments_from_device_buffer(void *block_tag,
int state_count,
double *device_buffer,
@@ -88,6 +128,14 @@ int z4c_cuda_unpack_state_segments_from_device_buffer(void *block_tag,
int segment_count,
const int *segment_meta);
int z4c_cuda_unpack_state_segments_from_device_buffer_for_host_views(void *block_tag,
double **state_host_key,
int state_count,
double *device_buffer,
int *ex,
int segment_count,
const int *segment_meta);
int z4c_cuda_restrict_state_segments_to_device_buffer(void *block_tag,
int state_count,
double *device_buffer,
@@ -96,6 +144,15 @@ int z4c_cuda_restrict_state_segments_to_device_buffer(void *block_tag,
const int *segment_meta,
const double *state_soa);
int z4c_cuda_restrict_state_segments_to_device_buffer_for_host_views(void *block_tag,
double **state_host_key,
int state_count,
double *device_buffer,
int *ex,
int segment_count,
const int *segment_meta,
const double *state_soa);
int z4c_cuda_prolong_state_segments_to_device_buffer(void *block_tag,
int state_count,
double *device_buffer,
@@ -104,6 +161,15 @@ int z4c_cuda_prolong_state_segments_to_device_buffer(void *block_tag,
const int *segment_meta,
const double *state_soa);
int z4c_cuda_prolong_state_segments_to_device_buffer_for_host_views(void *block_tag,
double **state_host_key,
int state_count,
double *device_buffer,
int *ex,
int segment_count,
const int *segment_meta,
const double *state_soa);
int z4c_cuda_restrict_state_batch_to_device_buffer(void *block_tag,
int state_count,
double *device_buffer,
@@ -112,6 +178,15 @@ int z4c_cuda_restrict_state_batch_to_device_buffer(void *block_tag,
int fi0, int fj0, int fk0,
const double *state_soa);
int z4c_cuda_restrict_state_batch_to_device_buffer_for_host_views(void *block_tag,
double **state_host_key,
int state_count,
double *device_buffer,
int *ex,
int sx, int sy, int sz,
int fi0, int fj0, int fk0,
const double *state_soa);
int z4c_cuda_prolong_state_batch_to_device_buffer(void *block_tag,
int state_count,
double *device_buffer,
@@ -121,6 +196,16 @@ int z4c_cuda_prolong_state_batch_to_device_buffer(void *block_tag,
int lbc_i, int lbc_j, int lbc_k,
const double *state_soa);
int z4c_cuda_prolong_state_batch_to_device_buffer_for_host_views(void *block_tag,
double **state_host_key,
int state_count,
double *device_buffer,
int *ex,
int sx, int sy, int sz,
int ii0, int jj0, int kk0,
int lbc_i, int lbc_j, int lbc_k,
const double *state_soa);
int z4c_cuda_download_state_subset(void *block_tag,
int *ex,
int subset_count,
@@ -138,6 +223,25 @@ int z4c_cuda_compute_constraints_resident(void *block_tag,
int Symmetry, double eps, int co,
double **constraint_host_out);
int z4c_cuda_interp_state_point3(void *block_tag,
int *ex,
int state0,
int state1,
int state2,
double x0,
double y0,
double z0,
double dx,
double dy,
double dz,
double px,
double py,
double pz,
int ordn,
int symmetry,
const double *soa3,
double *out3);
int z4c_cuda_download_constraint_outputs(int *ex,
double **constraint_host_out);