Commits

Author Commit Message Labels Comments Date
Michael Hudson-Doyle
Allow specifying the list of benchmarks to run in a file.
Michael Hudson-Doyle
Be more tolerant of whitespace when parsing the --benchmarks command line option.
Armin Rigo
merge heads
Armin Rigo
Fix for runs that only end up doing 1 iteration (e.g. hexiom2 with --fast).
Maciej Fijalkowski
improve reporting even more
Maciej Fijalkowski
merge
Maciej Fijalkowski
improve reporting
Armin Rigo
Improve the output in case of error
Brian Kearns
fix --niceness 0
Brian Kearns
have runner.py support --niceness option
Maciej Fijalkowski
another try
Maciej Fijalkowski
readd pypy, maybe better this time
Maciej Fijalkowski
like this
Maciej Fijalkowski
add PYTHONPATH to translate
Maciej Fijalkowski
rexport pypy repo here
Hakan Ardo
cleanup
Hakan Ardo
increase the scimark cycle count again and dont run it without the jit
Hakan Ardo
reenable the scimark benchmarks and lower their cycle count to give resonable runtimes without the jit
Maciej Fijalkowski
disable scimark for now
Hakan Ardo
skip the large flavour
Hakan Ardo
add scimark benchmark
Maciej Fijalkowski
add a benchmark that parses E (the language) contributed by Allan Short (dash)
Maciej Fijalkowski
merge
Maciej Fijalkowski
(__stian__) add pidigits benchmark to the nightly run
Armin Rigo
Add the benchmark "hexiom2" from Laurent Vaucher.
Maciej Fijalkowski
disable the web benchmark, again
Maciej Fijalkowski
add a LICENSE file
Maciej Fijalkowski
merge
Maciej Fijalkowski
Cut down on iterations here. Otherwise we run out of TCP connections and we get inconsistent bad results.
Lukas Diekmann
check track_memory capability with own process, so we don't have to run it with root anymore
  1. Prev
  2. Next