diff --git a/timing/timing.c b/timing/timing.c
index 30a843b372652d30e7428bcf1bd0f25352e3a448..cd08caab5302c292f699f86107ca926ec98519ba 100644
--- a/timing/timing.c
+++ b/timing/timing.c
@@ -359,58 +359,57 @@ static void
show_help(char *prog_name) {
printf( "Usage:\n%s [options]\n\n", prog_name );
printf( "Options are:\n"
- " -h --help Show this help\n"
+ " -h --help Show this help\n"
"\n"
- " -t x\n"
- " --threads x Number of CPU workers (default: _SC_NPROCESSORS_ONLN)\n"
- " -g x\n"
- " --gpus X Number of GPU workers (default: 0)\n"
+ " Machine parameters:\n"
+ " -t, --threads=x Number of CPU workers (default: automatic detection through runtime)\n"
+ " -g, --gpus=x Number of GPU workers (default: 0)\n"
+ " -P, --P=x Rows (P) in the PxQ process grid (deafult: 1)\n"
+ " --nocpu All GPU kernels are exclusively executed on GPUs (default: 0)\n"
"\n"
- " -s --sync Enable synchronous calls in wrapper function such as POTRI\n"
- " -b --nobigmat Allocating one big mat or plenty of small (default: bigmat)\n"
- " -c --check Check result\n"
- " -P --progress Display progress indicator\n"
- " -G --gemm3m Use gemm3m complex method\n"
- " -i --inv Check on inverse\n"
- " -w --nowarmup Perform a warmup run to pre-load libraries (default: warmup)\n"
- " -T --trace Enable trace generation\n"
- " -d --dag Enable DAG generation\n"
- " Generates a dot_dag_file.dot.\n"
- " -5 --profile Print profiling informations (default: noprofile)\n"
- " -C --nocpu All GPU kernels are exclusively executed on GPUs (default: 0)\n"
-/* " --inplace Enable layout conversion inplace for lapack interface timers (default: enable)\n" */
-/* " --outplace Enable layout conversion out of place for lapack interface timers (default: disable)\n" */
-/* " --[no]atun Activate autotuning (default: noatun)\n" */
+ " Matrix parameters:\n"
+ " -m, --m, --M=x Dimension (M) of the matrices (default: N)\n"
+ " -n, --n, --N=x Dimension (N) of the matrices\n"
+ " -N, --n_range=R Range of N values\n"
+ " with R=Start:Stop:Step (default: 500:5000:500)\n"
+ " -k, --k, --K, --nrhs=x Dimension (K) of the matrices or number of right-hand size(default: 1)\n"
+ " -b, --nb=x Nb size. (default: 320)\n"
+ " -i, --ib=x IB size. (default: 32)\n"
+ " -x, --mx=x ?\n" //todo
+ " -X, --nx=x ?\n" //todo
"\n"
- " -n R\n"
- " --n_range R Range of N values\n"
- " with R=Start:Stop:Step (default: 500:5000:500)\n"
- " -m x\n"
- " --m x dimension (M) of the matrices (default: N)\n"
- " -k x\n"
- " --k x dimension (K) of the matrices (default: 1)\n"
- " --nrhs X Number of right-hand size (default: 1)\n"
- " --nb N Nb size. (default: 128)\n"
- " --ib N IB size. (default: 32)\n"
+ " Check/prints:\n"
+ " --niter=x Number of iterations performed for each test (default: 1)\n"
+ " -W, --nowarnings Do not show warnings\n"
+ " -w, --nowarmup Cancel the warmup run to pre-load libraries\n"
+ " -c, --check Check result\n"
+ " -C, --inv Check on inverse\n"
"\n"
- " -N x\n"
- " --niter x Number of iterations performed for each test (default: 1)\n"
+ " Profiling:\n"
+ " -T, --trace Enable trace generation\n"
+ " --progress Display progress indicator\n"
+ " -d, --dag Enable DAG generation\n"
+ " Generates a dot_dag_file.dot.\n"
+ " -p, --profile Print profiling informations\n"
"\n"
- " -r N\n"
- " --rhblk N If N > 0, enable Householder mode for QR and LQ factorization\n"
- " N is the size of each subdomain (default: 0)\n"
-/* "\n" */
-/* " Options specific to the conversion format timings xgetri and xgecfi:\n" */
-/* " --ifmt Input format. (default: 0)\n" */
-/* " --ofmt Output format. (default: 1)\n" */
-/* " The possible values are:\n" */
-/* " 0 - MorseCM, Column major\n" */
-/* " 1 - MorseCCRB, Column-Colum rectangular block\n" */
-/* " 2 - MorseCRRB, Column-Row rectangular block\n" */
-/* " 3 - MorseRCRB, Row-Colum rectangular block\n" */
-/* " 4 - MorseRRRB, Row-Row rectangular block\n" */
-/* " 5 - MorseRM, Row Major\n" */
-/* " --thrdbypb Number of threads per subproblem for inplace transformation (default: 1)\n" */
+ " HQR options:\n"
+ " -a, --qr_a, --rhblk=N If N > 0, enable Householder mode for QR and LQ factorization\n"
+ " N is the size of each subdomain (default: 0)\n"
+ " -l, --llvl=x Tree used for low level reduction inside nodes (default: -1)\n"
+ " -L, --hlvl=x Tree used for high level reduction between nodes, only if P > 1 (default: -1).\n"
+ " (0: Flat, 1: Greedy, 2: Fibonacci, 3: Binary, 4: Replicated greedy)\n"
+ " -D, --domino Enable the domino between upper and lower trees.\n"
+ "\n"
+ " Advanced options\n"
+ " -M, --mode=x ?\n"//todo
+ " --nobigmat Allocating one big mat or plenty of small (default: bigmat)\n"
+ " -s, --sync Enable synchronous calls in wrapper function such as POTRI\n"
+ " -o, --ooc \n"
+ " -G, --gemm3m Use gemm3m complex method\n"
+ " --peak ?\n"//todo
+ " --bound ?\n"//todo
+ " --bounddeps ?\n"//todo
+ " --bounddepsprio ?\n"//todo
"\n");
}
@@ -458,10 +457,11 @@ print_header(char *prog_name, int * iparam) {
return;
}
-#define GETOPT_STRING "t:g:P:8m:n:N:k:b:i:x:X:1:WwcCT2dpa:M:l:L:D3soG4567"
+#define GETOPT_STRING "ht:g:P:8m:n:N:k:b:i:x:X:1:WwcCT2dpa:M:l:L:D3soG4567"
#if defined(CHAMELEON_HAVE_GETOPT_LONG)
static struct option long_options[] =
{
+ {"help", no_argument, 0, 'h'},
// Configuration
{"threads", required_argument, 0, 't'},
{"gpus", required_argument, 0, 'g'},
@@ -494,11 +494,11 @@ static struct option long_options[] =
// HQR options
{"rhblk", required_argument, 0, 'a'},
{"qr_a", required_argument, 0, 'a'},
- {"mode", required_argument, 0, 'M'},
{"llvl", required_argument, 0, 'l'},
{"hlvl", required_argument, 0, 'L'},
{"domino", no_argument, 0, 'D'},
// Other
+ {"mode", required_argument, 0, 'M'},
{"nobigmat", no_argument, 0, '3'},
{"sync", no_argument, 0, 's'},
{"ooc", no_argument, 0, 'o'},
@@ -524,8 +524,8 @@ set_iparam_default(int *iparam){
iparam[IPARAM_LDA ] = -1;
iparam[IPARAM_LDB ] = -1;
iparam[IPARAM_LDC ] = -1;
- iparam[IPARAM_MB ] = 128;
- iparam[IPARAM_NB ] = 128;
+ iparam[IPARAM_MB ] = 320;
+ iparam[IPARAM_NB ] = 320;
iparam[IPARAM_IB ] = 32;
iparam[IPARAM_NITER ] = 1;
iparam[IPARAM_WARMUP ] = 1;
@@ -566,47 +566,60 @@ parse_arguments(int *_argc, char ***_argv, int *iparam, int *start, int *stop, i
switch(c)
{
+ // Configuration
+ case 't' : iparam[IPARAM_THRDNBR ] = atoi(optarg); break;
+ case 'g' : iparam[IPARAM_NCUDAS ] = atoi(optarg); break;
+ case 'P' : iparam[IPARAM_P ] = atoi(optarg); break;
+ case '8' : iparam[IPARAM_NO_CPU ] = 1; break;
+ // Matrix parameters
+ case 'm' : iparam[IPARAM_M ] = atoi(optarg); break;
+ case 'n' : iparam[IPARAM_N ] = atoi(optarg); break;
+ case 'N' : get_range(optarg, start, stop, step); break;
+ case 'k' : iparam[IPARAM_K ] = atoi(optarg); break;
+ case 'b' : iparam[IPARAM_NB ] = atoi(optarg);
+ iparam[IPARAM_MB ] = atoi(optarg); break;
+ case 'i' : iparam[IPARAM_IB ] = atoi(optarg); break;
+ case 'x' : iparam[IPARAM_MX ] = atoi(optarg); break;
+ case 'X' : iparam[IPARAM_NX ] = atoi(optarg); break;
+ // Check/prints
+ case '1' : iparam[IPARAM_NITER ] = atoi(optarg); break;
+ case 'W' : iparam[IPARAM_PRINT_WARNINGS] = 0; break;
+ case 'w' : iparam[IPARAM_WARMUP ] = 0; break;
case 'c' : iparam[IPARAM_CHECK ] = 1; break;
- case '3' : iparam[IPARAM_BIGMAT ] = 0; break;
case 'C' : iparam[IPARAM_INVERSE ] = 1; break;
- case 'w' : iparam[IPARAM_WARMUP ] = 0; break;
+ // Profiling
case 'T' : iparam[IPARAM_TRACE ] = 1; break;
- case 'G' : iparam[IPARAM_GEMM3M ] = 1; break;
case '2' : iparam[IPARAM_PROGRESS ] = 1; break;
case 'd' : iparam[IPARAM_DAG ] = 1; break;
+ case 'p' : iparam[IPARAM_PROFILE ] = 1; break;
+ // HQR options
+ case 'a' : iparam[IPARAM_RHBLK ] = atoi(optarg); break;
+ case 'l' : iparam[IPARAM_LOWLVL_TREE ] = atoi(optarg); break;
+ case 'L' : iparam[IPARAM_HIGHLVL_TREE ] = atoi(optarg); break;
+ case 'D' : iparam[IPARAM_QR_DOMINO ] = 1; break;
+ //Other
+ case 'M' : iparam[IPARAM_MODE ] = atoi(optarg); break;
+ case '3' : iparam[IPARAM_BIGMAT ] = 0; break;
case 's' : iparam[IPARAM_ASYNC ] = 0; break;
case 'o' : iparam[IPARAM_OOC ] = 1; break;
+ case 'G' : iparam[IPARAM_GEMM3M ] = 1; break;
case '4' : iparam[IPARAM_PEAK ] = 1; break;
- case 'p' : iparam[IPARAM_PROFILE ] = 1; break;
- case 'W' : iparam[IPARAM_PRINT_WARNINGS] = 0; break;
- case '8' : iparam[IPARAM_NO_CPU ] = 1; break;
case '5' : iparam[IPARAM_BOUND ] = 1; break;
case '6' : iparam[IPARAM_BOUND ] = 1;
iparam[IPARAM_BOUNDDEPS ] = 1; break;
case '7' : iparam[IPARAM_BOUND ] = 1;
iparam[IPARAM_BOUNDDEPS ] = 1;
iparam[IPARAM_BOUNDDEPSPRIO ] = 1; break;
- case 't' : iparam[IPARAM_THRDNBR ] = atoi(optarg); break;
- case 'g' : iparam[IPARAM_NCUDAS ] = atoi(optarg); break;
- case 'm' : iparam[IPARAM_M ] = atoi(optarg); break;
- case 'n' : iparam[IPARAM_N ] = atoi(optarg); break;
- case 'k' : iparam[IPARAM_K ] = atoi(optarg); break;
- case 'i' : iparam[IPARAM_IB ] = atoi(optarg); break;
- case '1' : iparam[IPARAM_NITER ] = atoi(optarg); break;
- case 'x' : iparam[IPARAM_MX ] = atoi(optarg); break;
- case 'X' : iparam[IPARAM_NX ] = atoi(optarg); break;
- case 'a' : iparam[IPARAM_RHBLK ] = atoi(optarg); break;
- case 'P' : iparam[IPARAM_P ] = atoi(optarg); break;
- case 'M' : iparam[IPARAM_MODE ] = atoi(optarg); break;
- case 'b' : iparam[IPARAM_NB ] = atoi(optarg);
- iparam[IPARAM_MB ] = atoi(optarg); break;
- case 'N' : get_range(optarg, start, stop, step); break;
- case 'h' : show_help(argv[0]); break;
+ case 'h' :
+ case '?' :
+ show_help(argv[0]); exit(EXIT_FAILURE);
default:
break;
}
}while(-1 != c);
+ if(-'P' == iparam[IPARAM_QR_HLVL_SZE]) iparam[IPARAM_QR_HLVL_SZE] = iparam[IPARAM_P];
+ if(-'Q' == iparam[IPARAM_QR_HLVL_SZE]) iparam[IPARAM_QR_HLVL_SZE] = iparam[IPARAM_Q];
}
int