Closed g-k closed 6 years ago
Changes:
Running the single timing test with repeated go test -run TestI2OSPTiming gets worse from ~1000 ns on master to 2000-3000ns on this branch
go test -run TestI2OSPTiming
However, go test -run TestI2OSP (which calls the correctness then the timing test) goes from ~500 ns on master to under ~200ns on this branch
go test -run TestI2OSP
Not sure if this fixes #40 or not.
Need to profile this, but we could:
r? @franziskuskiefer
Coverage increased (+0.8%) to 97.164% when pulling 55494f43b198c732da64fec064ab052324e42d2c on fix-i2osp into 3d08d63c9c1da50b3b4bf263e34acc36388a02e8 on master.
Changes:
Running the single timing test with repeated
go test -run TestI2OSPTiming
gets worse from ~1000 ns on master to 2000-3000ns on this branchHowever,
go test -run TestI2OSP
(which calls the correctness then the timing test) goes from ~500 ns on master to under ~200ns on this branchNot sure if this fixes #40 or not.
Need to profile this, but we could:
r? @franziskuskiefer