jianmu
/
lammps
forked from lijiext/lammps
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

git-svn-id: svn://svn.icms.temple.edu/lammps-ro/trunk@12132 f3b2605a-c512-4ea7-a41b-209d697bcdaa

This commit is contained in:
sjplimp 2014-06-14 19:34:19 +00:00
parent 7b5846a386
commit 2ec1442e97
5 changed files with 8 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
lib/cuda/Makefile.defaults
View File

@ -1,6 +1,6 @@
#precision setting: 1 single, 2 double, 4 mixed #precision setting: 1 single, 2 double, 4 mixed
precision ?= 2 precision ?= 4
#verbose setting: 0 no, 1 yes #verbose setting: 0 no, 1 yes
verbose ?= 1 verbose ?= 1

2
lib/cuda/Makefile.lammps
View File

@ -1,5 +1,5 @@
# Settings that the LAMMPS build will import when this package library is used # Settings that the LAMMPS build will import when this package library is used
CUDA_FLAGS := -I/usr/local/cuda/include -DUNIX -DFFT_CUFFT -DCUDA_PRECISION=2 -DCUDA_ARCH=20 CUDA_FLAGS := -I/usr/local/cuda/include -DUNIX -DFFT_CUFFT -DCUDA_PRECISION=1 -DX_PRECISION=2 -DV_PRECISION=2 -DCUDA_ARCH=20
CUDA_USRLIB_CONDITIONAL := -L/usr/local/cuda/lib -L/usr/local/cuda/lib64 -lcufft CUDA_USRLIB_CONDITIONAL := -L/usr/local/cuda/lib -L/usr/local/cuda/lib64 -lcufft
user-cuda_SYSINC = ${CUDA_FLAGS} user-cuda_SYSINC = ${CUDA_FLAGS}

2
lib/cuda/cuda_wrapper.cu
View File

@ -254,7 +254,7 @@ void cuda_check_error(char* comment)
printf("ERROR-CUDA %s %s\n", comment, cudaGetErrorString(cudaGetLastError())); printf("ERROR-CUDA %s %s\n", comment, cudaGetErrorString(cudaGetLastError()));
} }
int CudaWrapper_CheckMemUseage() int CudaWrapper_CheckMemUsage()
{ {
size_t free, total; size_t free, total;
cudaMemGetInfo(&free, &total); cudaMemGetInfo(&free, &total);

2
lib/cuda/cuda_wrapper_cu.h
View File

@ -36,7 +36,7 @@ extern "C" void CudaWrapper_CopyData(void* dev_dest, void* dev_source, unsigned
extern "C" void* CudaWrapper_AllocPinnedHostData(unsigned nbytes, bool mapped = false, bool writeCombind = false); extern "C" void* CudaWrapper_AllocPinnedHostData(unsigned nbytes, bool mapped = false, bool writeCombind = false);
extern "C" void CudaWrapper_FreePinnedHostData(void* dev_data); extern "C" void CudaWrapper_FreePinnedHostData(void* dev_data);
extern "C" void cuda_check_error(char* comment); extern "C" void cuda_check_error(char* comment);
extern "C" int CudaWrapper_CheckMemUseage(); extern "C" int CudaWrapper_CheckMemUsage();
extern "C" double CudaWrapper_CheckUploadTime(bool reset = false); extern "C" double CudaWrapper_CheckUploadTime(bool reset = false);
extern "C" double CudaWrapper_CheckDownloadTime(bool reset = false); extern "C" double CudaWrapper_CheckDownloadTime(bool reset = false);
extern "C" double CudaWrapper_CheckCPUBufUploadTime(bool reset = false); extern "C" double CudaWrapper_CheckCPUBufUploadTime(bool reset = false);

4
lib/gpu/lal_device.cpp
View File

@ -492,6 +492,7 @@ void DeviceT::output_times(UCL_Timer &time_pair, Answer<numtyp,acctyp> &ans,
const double driver_overhead, const double driver_overhead,
const int threads_per_atom, FILE *screen) { const int threads_per_atom, FILE *screen) {
double single[9], times[9]; double single[9], times[9];
int post_final=0;
single[0]=atom.transfer_time()+ans.transfer_time(); single[0]=atom.transfer_time()+ans.transfer_time();
single[1]=nbor.time_nbor.total_seconds()+nbor.time_hybrid1.total_seconds()+ single[1]=nbor.time_nbor.total_seconds()+nbor.time_hybrid1.total_seconds()+
@ -504,6 +505,9 @@ void DeviceT::output_times(UCL_Timer &time_pair, Answer<numtyp,acctyp> &ans,
single[7]=ans.cpu_idle_time(); single[7]=ans.cpu_idle_time();
single[8]=nbor.bin_time(); single[8]=nbor.bin_time();
MPI_Finalized(&post_final);
if (post_final) return;
MPI_Reduce(single,times,9,MPI_DOUBLE,MPI_SUM,0,_comm_replica); MPI_Reduce(single,times,9,MPI_DOUBLE,MPI_SUM,0,_comm_replica);
double my_max_bytes=max_bytes+atom.max_gpu_bytes(); double my_max_bytes=max_bytes+atom.max_gpu_bytes();