initial papi instrumentation

[calebyhan]

Still needs LAPACK routine support, issues with counter not resetting

calebhan ~/slate/test/> ./tester --repeat 3 --papi y --dim 500:500 --nb 200 --check n --ref n gemm
% SLATE version 2023.06.00, id 30873c6e
% input: ./tester --repeat 3 --papi y --dim 500:500 --nb 200 --check n --ref n gemm
% 2023-07-28 17:28:29, 1 MPI ranks, CPU-only MPI, 256 OpenMP threads, 8 GPU devices per MPI rank
                                                                                                                                                                                                         
type  origin  target  gemm   go   A   B   C   transA   transB       m       n       k      alpha       beta    nb    p    q  la      error   time (s)       gflop/s  ref time (s)   ref gflop/s  status  
   d    host    task  auto  col   1   1   1  notrans  notrans     500     500     500   3.1+1.4i   2.7+1.7i   200    1    1   1         NA     0.0205        12.176            NA            NA  no check  
gemm( N, N, 200, 200, 100 ) count 4, flop count 3.20e+07
gemm( N, N, 200, 100, 200 ) count 4, flop count 3.20e+07
gemm( N, N, 100, 200, 100 ) count 2, flop count 8.00e+06
gemm( N, N, 200, 200, 200 ) count 8, flop count 1.28e+08
gemm( N, N, 100, 100, 200 ) count 2, flop count 8.00e+06
gemm( N, N, 100, 100, 100 ) count 1, flop count 2.00e+06
gemm( N, N, 100, 200, 200 ) count 4, flop count 3.20e+07
gemm( N, N, 200, 100, 100 ) count 2, flop count 8.00e+06
total BLAS flop count 2.50e+08

   d    host    task  auto  col   1   1   1  notrans  notrans     500     500     500   3.1+1.4i   2.7+1.7i   200    1    1   1         NA     0.0231        10.843            NA            NA  no check  
gemm( N, N, 200, 200, 100 ) count 8, flop count 6.40e+07
gemm( N, N, 200, 100, 200 ) count 8, flop count 6.40e+07
gemm( N, N, 100, 200, 100 ) count 4, flop count 1.60e+07
gemm( N, N, 200, 200, 200 ) count 16, flop count 2.56e+08
gemm( N, N, 100, 100, 200 ) count 4, flop count 1.60e+07
gemm( N, N, 100, 100, 100 ) count 2, flop count 4.00e+06
gemm( N, N, 100, 200, 200 ) count 8, flop count 6.40e+07
gemm( N, N, 200, 100, 100 ) count 4, flop count 1.60e+07
total BLAS flop count 5.00e+08

   d    host    task  auto  col   1   1   1  notrans  notrans     500     500     500   3.1+1.4i   2.7+1.7i   200    1    1   1         NA     0.0202        12.375            NA            NA  no check  
gemm( N, N, 200, 200, 100 ) count 12, flop count 9.60e+07
gemm( N, N, 200, 100, 200 ) count 12, flop count 9.60e+07
gemm( N, N, 100, 200, 100 ) count 6, flop count 2.40e+07
gemm( N, N, 200, 200, 200 ) count 24, flop count 3.84e+08
gemm( N, N, 100, 100, 200 ) count 6, flop count 2.40e+07
gemm( N, N, 100, 100, 100 ) count 3, flop count 6.00e+06
gemm( N, N, 100, 200, 200 ) count 12, flop count 9.60e+07
gemm( N, N, 200, 100, 100 ) count 6, flop count 2.40e+07
total BLAS flop count 7.50e+08



% Matrix kinds:
%  1: rand, cond unknown

% All tests passed: gemm