j3d1/json
1
0
Fork 0
Code Issues Pull requests Releases Wiki Activity

Simplify the json/src/benchmarks.cpp to allow more optimal code gen.

Browse source 
o For some unknown reason, the complexity of the benchmark platform
  prevented some C++ compilers from generating optimal code, properly
  reflective of the real performance in actual deployment.
o Added the json_benchmarks_simple target, which performs the same
  suite of tests as json_benchmarks.
o Simplified the benchmark platform, and emit an "Average" TPS
  (Transactions Per Second) value reflective of aggregate parse/output
  performance.
This commit is contained in:
Perry Kundert 2017-10-07 15:50:19 -07:00
parent 23440eb86e
commit 0b803d0a5f
4 changed files with 190 additions and 7 deletions
Download patch file Download diff file Expand all files Collapse all files

18
benchmarks/Makefile
View file

@ -1,11 +1,21 @@
all: json_benchmarks
./json_benchmarks
json_benchmarks: src/benchmarks.cpp ../src/json.hpp number_jsons
#
# Build/run json.hpp benchmarks, eg. CXX=g++-7 make
#
# The existing json_benchmarks did not allow optimization under some compilers
#
all: json_benchmarks json_benchmarks_simple number_jsons
bash -c 'time ./json_benchmarks'
bash -c 'time ./json_benchmarks_simple'
json_benchmarks: src/benchmarks.cpp ../src/json.hpp
$(CXX) -std=c++11 -pthread $(CXXFLAGS) -DNDEBUG -O3 -flto -I thirdparty/benchpress -I thirdparty/cxxopts -I../src src/benchmarks.cpp $(LDFLAGS) -o $@
json_benchmarks_simple: src/benchmarks_simple.cpp ../src/json.hpp
$(CXX) -std=c++11 $(CXXFLAGS) -DNDEBUG -O3 -flto -I../src $(<) $(LDFLAGS) -o $@
number_jsons:
(test -e files/numbers/floats.json -a -e files/numbers/signed_ints.json -a -e files/numbers/unsigned_ints.json) || (cd files/numbers ; python generate.py)
clean:
rm -f json_benchmarks files/numbers/*.json
rm -f json_benchmarks json_benchmarks_simple files/numbers/*.json