portelli/Grid
1
0
Fork 0
mirror of https://github.com/paboyle/Grid.git synced 2024-11-10 07:55:35 +00:00
Code Releases Activity

Pragma use for nvcc, warning elimination.

Browse Source
This commit is contained in:
paboyle 2018-01-24 13:15:43 +00:00
parent 22d137d4e5
commit bd08dc4f45
3 changed files with 33 additions and 42 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

73
lib/parallelIO/BinaryIO.h
View File

@ -104,27 +104,22 @@ public:
const uint64_t size32 = sizeof(fobj) / sizeof(uint32_t);
uint64_t lsites = grid->lSites();
if (fbuf.size() == 1)
{
lsites = 1;
}
if (fbuf.size() == 1) {
lsites = 1;
}
thread_region {
#pragma omp parallel
{
uint32_t nersc_csum_thr = 0;
#pragma omp for
for (uint64_t local_site = 0; local_site < lsites; local_site++)
{
thread_loop( (uint64_t local_site = 0; local_site < lsites; local_site++), {
uint32_t *site_buf = (uint32_t *)&fbuf[local_site];
for (uint64_t j = 0; j < size32; j++)
{
nersc_csum_thr = nersc_csum_thr + site_buf[j];
}
}
for (uint64_t j = 0; j < size32; j++) {
nersc_csum_thr = nersc_csum_thr + site_buf[j];
}
});
#pragma omp critical
{
thread_critical {
nersc_csum += nersc_csum_thr;
}
}
@ -132,9 +127,6 @@ public:
template<class fobj> static inline void ScidacChecksum(GridBase *grid,std::vector<fobj> &fbuf,uint32_t &scidac_csuma,uint32_t &scidac_csumb)
{
const uint64_t size32 = sizeof(fobj)/sizeof(uint32_t);
int nd = grid->_ndimension;
uint64_t lsites =grid->lSites();
@ -145,15 +137,14 @@ public:
std::vector<int> local_start =grid->LocalStarts();
std::vector<int> global_vol =grid->FullDimensions();
#pragma omp parallel
{
thread_region {
std::vector<int> coor(nd);
uint32_t scidac_csuma_thr=0;
uint32_t scidac_csumb_thr=0;
uint32_t site_crc=0;
#pragma omp for
for(uint64_t local_site=0;local_site<lsites;local_site++){
thread_loop( (uint64_t local_site=0;local_site<lsites;local_site++),{
uint32_t * site_buf = (uint32_t *)&fbuf[local_site];
@ -177,11 +168,11 @@ public:
// std::cout << "Site "<<local_site << " crc "<<std::hex<<site_crc<<std::dec<<std::endl;
// std::cout << "Site "<<local_site << std::hex<<site_buf[0] <<site_buf[1]<<std::dec <<std::endl;
scidac_csuma_thr ^= site_crc<<gsite29 | site_crc>>(32-gsite29);
scidac_csumb_thr ^= site_crc<<gsite31 | site_crc>>(32-gsite31);
}
scidac_csumb_thr ^= site_crc<<gsite31 | site_crc>>(32-gsite31);
#pragma omp critical
{
});
thread_critical {
scidac_csuma^= scidac_csuma_thr;
scidac_csumb^= scidac_csumb_thr;
}
@ -198,9 +189,9 @@ public:
{
uint32_t * f = (uint32_t *)file_object;
uint64_t count = bytes/sizeof(uint32_t);
parallel_for(uint64_t i=0;i<count;i++){
thread_loop( (uint64_t i=0;i<count;i++), {
f[i] = ntohl(f[i]);
}
});
}
// LE must Swap and switch to host
static inline void le32toh_v(void *file_object,uint64_t bytes)
@ -209,12 +200,12 @@ public:
uint32_t f;
uint64_t count = bytes/sizeof(uint32_t);
parallel_for(uint64_t i=0;i<count;i++){
thread_loop( (uint64_t i=0;i<count;i++), {
f = fp[i];
// got network order and the network to host
f = ((f&0xFF)<<24) | ((f&0xFF00)<<8) | ((f&0xFF0000)>>8) | ((f&0xFF000000UL)>>24) ;
fp[i] = ntohl(f);
}
});
}
// BE is same as network
@ -222,9 +213,9 @@ public:
{
uint64_t * f = (uint64_t *)file_object;
uint64_t count = bytes/sizeof(uint64_t);
parallel_for(uint64_t i=0;i<count;i++){
thread_loop( (uint64_t i=0;i<count;i++),{
f[i] = Grid_ntohll(f[i]);
}
});
}
// LE must swap and switch;
@ -234,7 +225,7 @@ public:
uint64_t f,g;
uint64_t count = bytes/sizeof(uint64_t);
parallel_for(uint64_t i=0;i<count;i++){
thread_loop( (uint64_t i=0;i<count;i++), {
f = fp[i];
// got network order and the network to host
g = ((f&0xFF)<<24) | ((f&0xFF00)<<8) | ((f&0xFF0000)>>8) | ((f&0xFF000000UL)>>24) ;
@ -242,7 +233,7 @@ public:
f = f >> 32;
g|= ((f&0xFF)<<24) | ((f&0xFF00)<<8) | ((f&0xFF0000)>>8) | ((f&0xFF000000UL)>>24) ;
fp[i] = Grid_ntohll(g);
}
});
}
/////////////////////////////////////////////////////////////////////////////
// Real action:
@ -543,7 +534,7 @@ public:
GridStopWatch timer;
timer.Start();
parallel_for(int x=0;x<lsites;x++) munge(iodata[x], scalardata[x]);
thread_loop( (int x=0;x<lsites;x++), { munge(iodata[x], scalardata[x]); });
vectorizeFromLexOrdArray(scalardata,Umu);
grid->Barrier();
@ -579,7 +570,7 @@ public:
GridStopWatch timer; timer.Start();
unvectorizeToLexOrdArray(scalardata,Umu);
parallel_for(int x=0;x<lsites;x++) munge(scalardata[x],iodata[x]);
thread_loop( (int x=0;x<lsites;x++), { munge(scalardata[x],iodata[x]);});
grid->Barrier();
timer.Stop();
@ -625,11 +616,11 @@ public:
nersc_csum,scidac_csuma,scidac_csumb);
timer.Start();
parallel_for(int lidx=0;lidx<lsites;lidx++){
thread_loop( (int lidx=0;lidx<lsites;lidx++),{
std::vector<RngStateType> tmp(RngStateCount);
std::copy(iodata[lidx].begin(),iodata[lidx].end(),tmp.begin());
parallel.SetState(tmp,lidx);
}
});
timer.Stop();
iodata.resize(1);
@ -683,11 +674,11 @@ public:
timer.Start();
std::vector<RNGstate> iodata(lsites);
parallel_for(int lidx=0;lidx<lsites;lidx++){
thread_loop( (int lidx=0;lidx<lsites;lidx++),COMMA_SAFE({
std::vector<RngStateType> tmp(RngStateCount);
parallel.GetState(tmp,lidx);
std::copy(tmp.begin(),tmp.end(),iodata[lidx].begin());
}
}));
timer.Stop();
IOobject(w,grid,iodata,file,offset,format,BINARYIO_WRITE|BINARYIO_LEXICOGRAPHIC,

1
lib/parallelIO/IldgIO.h
View File

@ -722,6 +722,7 @@ public:
assert(found_ildgBinary);
assert(found_ildgFormat);
assert(found_scidacChecksum);
assert(found_ildgLFN==0);
// Must find something with the lattice dimensions
assert(found_FieldMetaData||found_ildgFormat);

1
lib/parallelIO/NerscIO.h
View File

@ -56,7 +56,6 @@ public:
// for the header-reader
static inline int readHeader(std::string file,GridBase *grid, FieldMetaData &field)
{
int offset=0;
std::map<std::string,std::string> header;
std::string line;