From 49f4459a0f528caa2057475507f2f3d3f56051b8 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Jakub=20Klinkovsk=C3=BD?= <klinkjak@fjfi.cvut.cz>
Date: Sat, 19 Mar 2016 16:45:09 +0100
Subject: [PATCH] Added parser of command line parameters to
tnl-cuda-benchmarks
---
tests/benchmarks/share/tnl-log-to-html.py | 15 +-
tests/benchmarks/tnl-cuda-benchmarks.h | 184 +++++++++++++---------
2 files changed, 125 insertions(+), 74 deletions(-)
diff --git a/tests/benchmarks/share/tnl-log-to-html.py b/tests/benchmarks/share/tnl-log-to-html.py
index 6a2c4c6315..38adaad400 100644
--- a/tests/benchmarks/share/tnl-log-to-html.py
+++ b/tests/benchmarks/share/tnl-log-to-html.py
@@ -3,6 +3,19 @@
import sys
import collections
+def getSortKey(value):
+ # try to convert to number if possible
+ try:
+ return int(value)
+ except ValueError:
+ try:
+ return float(value)
+ except ValueError:
+ if value:
+ return value
+ # None or empty string
+ return 0
+
class columnFormating:
def __init__( self, data ):
@@ -305,7 +318,7 @@ class logToHtmlConvertor:
# TODO: check this
# sort again (just in case, previous sorting might compare values from
# different columns)
- self.tableRows.sort(key=lambda row: list(row.values()))
+ self.tableRows.sort(key=lambda row: [getSortKey(value) for value in row.values()])
def countSubcolumns( self ):
for path, col in self.tableColumns.items():
diff --git a/tests/benchmarks/tnl-cuda-benchmarks.h b/tests/benchmarks/tnl-cuda-benchmarks.h
index 7d5f562c87..fa9f5bf9dd 100644
--- a/tests/benchmarks/tnl-cuda-benchmarks.h
+++ b/tests/benchmarks/tnl-cuda-benchmarks.h
@@ -18,6 +18,8 @@
#ifndef TNLCUDABENCHMARKS_H_
#define TNLCUDBENCHMARKS_H_
+#include <config/tnlConfigDescription.h>
+#include <config/tnlParameterContainer.h>
#include <core/tnlList.h>
#include <matrices/tnlCSRMatrix.h>
#include <matrices/tnlEllpackMatrix.h>
@@ -54,7 +56,7 @@ int setHostTestMatrix( Matrix& matrix,
matrix.setElement( row, col + element, element + 1 );
elements++;
}
- }
+ }
}
return elements;
}
@@ -67,13 +69,13 @@ __global__ void setCudaTestMatrixKernel( Matrix* matrix,
const int rowIdx = ( gridIdx * tnlCuda::getMaxGridSize() + blockIdx.x ) * blockDim.x + threadIdx.x;
if( rowIdx >= matrix->getRows() )
return;
- int col = rowIdx - elementsPerRow / 2;
+ int col = rowIdx - elementsPerRow / 2;
for( int element = 0; element < elementsPerRow; element++ )
{
if( col + element >= 0 &&
col + element < matrix->getColumns() )
matrix->setElementFast( rowIdx, col + element, element + 1 );
- }
+ }
}
template< typename Matrix >
@@ -186,79 +188,115 @@ benchmarkSpMV( Benchmark & benchmark,
return true;
}
-int main( int argc, char* argv[] )
+template< typename Real >
+void
+runCudaBenchmarks( Benchmark & benchmark,
+ Benchmark::MetadataMap metadata,
+ const unsigned & minSize,
+ const unsigned & maxSize,
+ const unsigned & loops,
+ const unsigned & elementsPerRow )
+{
+ const tnlString precision = getType< Real >();
+ metadata["precision"] = precision;
+
+ // Array operations
+ benchmark.newBenchmark( tnlString("Array operations (") + precision + ")",
+ metadata );
+ for( unsigned size = minSize; size <= maxSize; size *= 2 ) {
+ benchmark.setMetadataColumns( Benchmark::MetadataColumns({
+ {"size", size},
+ } ));
+ benchmarkArrayOperations< Real >( benchmark, loops, size );
+ }
+
+ // Vector operations
+ benchmark.newBenchmark( tnlString("Vector operations (") + precision + ")",
+ metadata );
+ for( unsigned size = minSize; size <= maxSize; size *= 2 ) {
+ benchmark.setMetadataColumns( Benchmark::MetadataColumns({
+ {"size", size},
+ } ));
+ benchmarkVectorOperations< Real >( benchmark, loops, size );
+ }
+
+ // Sparse matrix-vector multiplication
+ benchmark.newBenchmark( tnlString("Sparse matrix-vector multiplication (") + precision + ")",
+ metadata );
+ for( unsigned size = minSize; size <= maxSize; size *= 2 ) {
+ benchmark.setMetadataColumns( Benchmark::MetadataColumns({
+ {"rows", size},
+ {"columns", size},
+ {"elements per row", elementsPerRow},
+ } ));
+
+ // TODO: benchmark all formats from tnl-benchmark-spmv (different parameters of the base formats)
+ benchmarkSpMV< Real, tnlCSRMatrix >( benchmark, loops, size, elementsPerRow );
+ benchmarkSpMV< Real, tnlEllpackMatrix >( benchmark, loops, size, elementsPerRow );
+ benchmarkSpMV< Real, SlicedEllpackMatrix >( benchmark, loops, size, elementsPerRow );
+ benchmarkSpMV< Real, tnlChunkedEllpackMatrix >( benchmark, loops, size, elementsPerRow );
+ }
+}
+
+void
+setupConfig( tnlConfigDescription & config )
+{
+ config.addDelimiter( "Benchmark settings:" );
+ config.addEntry< tnlString >( "log-file", "Log file name.", "tnl-cuda-benchmarks.log");
+ config.addEntry< tnlString >( "precision", "Precision of the arithmetics.", "double" );
+ config.addEntryEnum( "float" );
+ config.addEntryEnum( "double" );
+ config.addEntryEnum( "all" );
+ config.addEntry< int >( "min-size", "Minimum size of arrays/vectors used in the benchmark (next size is 2*min-size and so on, up to max-size).", 100000 );
+ config.addEntry< int >( "max-size", "Minimum size of arrays/vectors used in the benchmark (next size is 2*min-size and so on, up to max-size).", 10000000 );
+ config.addEntry< int >( "loops", "Number of iterations for every computation.", 1.0 );
+ config.addEntry< int >( "elements-per-row", "Number of elements per row of the sparse matrix used in the matrix-vector multiplication benchmark.", 5 );
+ config.addEntry< int >( "verbose", "Verbose mode.", 1 );
+}
+
+int
+main( int argc, char* argv[] )
{
#ifdef HAVE_CUDA
-
- typedef double Real;
- tnlString precision = getType< Real >();
-
- /****
- * The first argument of this program is the size od data set to be reduced.
- * If no argument is given we use hardcoded default value.
- */
- int size = 1 << 22;
- if( argc > 1 )
- size = atoi( argv[ 1 ] );
- int loops = 10;
- if( argc > 2 )
- loops = atoi( argv[ 2 ] );
- int elementsPerRow = 5;
- if( argc > 3 )
- elementsPerRow = atoi( argv[ 3 ] );
-
- ofstream logFile( "tnl-cuda-benchmarks.log" );
- Benchmark benchmark( loops, true );
-// ostream & logFile = cout;
-// Benchmark benchmark( loops, false );
-
- // TODO: add hostname, CPU info, GPU info, date, ...
- Benchmark::MetadataMap metadata {
- {"precision", precision},
- };
- // TODO: loop over sizes
-
-
- // Array operations
- benchmark.newBenchmark( tnlString("Array operations (") + precision + ")",
- metadata );
- benchmark.setMetadataColumns( Benchmark::MetadataColumns({
- {"size", size},
- } ));
- benchmarkArrayOperations< Real >( benchmark, loops, size );
-
- // Vector operations
- benchmark.newBenchmark( tnlString("Vector operations (") + precision + ")",
- metadata );
- benchmark.setMetadataColumns( Benchmark::MetadataColumns({
- {"size", size},
- } ));
- benchmarkVectorOperations< Real >( benchmark, loops, size );
-
-
- // Sparse matrix-vector multiplication
- benchmark.newBenchmark( tnlString("Sparse matrix-vector multiplication (") + precision + ")",
- metadata );
- benchmark.setMetadataColumns( Benchmark::MetadataColumns({
- {"rows", size},
- {"columns", size},
- {"elements per row", elementsPerRow},
- } ));
-
- // TODO: benchmark all formats from tnl-benchmark-spmv (different parameters of the base formats)
- benchmarkSpMV< Real, tnlCSRMatrix >( benchmark, loops, size, elementsPerRow );
- benchmarkSpMV< Real, tnlEllpackMatrix >( benchmark, loops, size, elementsPerRow );
- benchmarkSpMV< Real, SlicedEllpackMatrix >( benchmark, loops, size, elementsPerRow );
- benchmarkSpMV< Real, tnlChunkedEllpackMatrix >( benchmark, loops, size, elementsPerRow );
-
-
- if( ! benchmark.save( logFile ) )
- return EXIT_FAILURE;
-
- return EXIT_SUCCESS;
+ tnlParameterContainer parameters;
+ tnlConfigDescription conf_desc;
+
+ setupConfig( conf_desc );
+
+ if( ! parseCommandLine( argc, argv, conf_desc, parameters ) ) {
+ conf_desc.printUsage( argv[ 0 ] );
+ return 1;
+ }
+
+ ofstream logFile( parameters.getParameter< tnlString >( "log-file" ).getString() );
+ const tnlString & precision = parameters.getParameter< tnlString >( "precision" );
+ const unsigned minSize = parameters.getParameter< unsigned >( "min-size" );
+ const unsigned maxSize = parameters.getParameter< unsigned >( "max-size" );
+ const unsigned loops = parameters.getParameter< unsigned >( "loops" );
+ const unsigned elementsPerRow = parameters.getParameter< unsigned >( "elements-per-row" );
+ const unsigned verbose = parameters.getParameter< unsigned >( "verbose" );
+
+ // init benchmark and common metadata
+ Benchmark benchmark( loops, verbose );
+ // TODO: add hostname, CPU info, GPU info, date, ...
+ Benchmark::MetadataMap metadata {
+// {"key", value},
+ };
+
+ if( precision == "all" || precision == "float" )
+ runCudaBenchmarks< float >( benchmark, metadata, minSize, maxSize, loops, elementsPerRow );
+ if( precision == "all" || precision == "double" )
+ runCudaBenchmarks< double >( benchmark, metadata, minSize, maxSize, loops, elementsPerRow );
+
+ if( ! benchmark.save( logFile ) ) {
+ cerr << "Failed to write the benchmark results to file '" << parameters.getParameter< tnlString >( "log-file" ) << "'." << endl;
+ return EXIT_FAILURE;
+ }
+
+ return EXIT_SUCCESS;
#else
- tnlCudaSupportMissingMessage;
- return EXIT_FAILURE;
+ tnlCudaSupportMissingMessage;
+ return EXIT_FAILURE;
#endif
}
--
GitLab