On 1/28/2018 11:43 PM, Guido van Rossum wrote: > So why can't you just run "make test" if that's faster? Not a standard option on Windows ;-). > On Sun, Jan 28, 2018 at 8:30 PM, Nick Coghlan <ncoghlan at gmail.com > <mailto:ncoghlan at gmail.com>> wrote: > > On my current system, "make test" runs in around 3 minutes, while > "./python -m test" runs in around 16 minutes. And that's with "make > test" actually running more tests (since it enables several of the > "-u" options). > > The difference is that "make test" passes "-j0" and hence not only > uses all the available cores in the machines, but will also run other > tests while some tests are sleeping. > > How would folks feel about making "-j 0" the default in the test > suite, and then adjusted the handling of "-j 1" to switch back to the > current default single process mode? > > My rationale for that is to improve the default edit-test cycle in > local development, while still providing a way to force single-process > execution for failure investigation purposes. I would like this (though I could write a .bat file). I routinely pass -j14 or so on a 6 core x 2 processes/core machine and get about the same times. The speedup would be even better but for the last very long running test. I wish each test file was limited to about 30 seconds, or even a minute. -- Terry Jan Reedy
RetroSearch is an open source project built by @garambo | Open a GitHub Issue
Search and Browse the WWW like it's 1997 | Search results from DuckDuckGo
HTML:
3.2
| Encoding:
UTF-8
| Version:
0.7.4