From 8183fe17ba78d8a15a11af5ca220ab93a8f4acba Mon Sep 17 00:00:00 2001 From: Sylvester Kaczmarek <16242628+sylvesterkaczmarek@users.noreply.github.com> Date: Sat, 22 Aug 2026 21:16:09 +0100 Subject: [PATCH] Document benchmark configuration CSV --- Testing/README.md | 15 +++++++++++++-- 1 file changed, 13 insertions(+), 2 deletions(-) diff --git a/Testing/README.md b/Testing/README.md index c94854c4c..f1b7ad236 100644 --- a/Testing/README.md +++ b/Testing/README.md @@ -477,6 +477,17 @@ The results generated in previous step can be processed with a Python script. The `-f` option should be used to tell the script where to find the `Output.pickle` file if the script is not run from the `Testing` folder. +When the output contains benchmarks, `processResult.py` reads `currentConfig.csv` from the same +directory as the result file passed with `-r`. This file is not generated automatically. Create it +as a two-line CSV containing a header and the configuration used for the benchmark. For example: + + OPTIMIZED,HARDFP,FASTMATH,NEON,HELIUM,UNROLL,ROUNDING,PLATFORM,CORE,COMPILER,VERSION + 1,1,1,0,1,1,0,FVP_M55,M55,AC6,6.21 + +Use `1` or `0` for the first seven fields to show whether each option was enabled. Replace the +platform, core or configuration identifier, compiler, and version values with those used for the +benchmark run. + python processResult.py -f Output.pickle -e -r result.txt -e option is needed if the mode -e was used with processTests because the output has a different @@ -496,8 +507,8 @@ The script summaryBench can parse those results and compute regression formula. python summaryBench.py -r build\result.txt The file result.txt must be placed inside the build folder for this script to work. -Indeed, this script is using the path to result.txt to also find the file currentConfig.csv which has -been created by the cmake command. +Indeed, this script uses the path to result.txt to also find the currentConfig.csv file described +above. The Output.pickle file is used by default. It can be changed with -f option.