I know I said I wasn't going to submit any more PRs for awhile, but I just couldn't help myself. :)
I updated the tests to include future and sharedfuture, per your request. I also added tests for the trylock_*() functions that shared_mutex provides that the other mutex's don't. And since upgrade_mutex is just a typedef of shared_mutex, I removed the tests for upgrade_mutex.
I know I said I wasn't going to submit any more PRs for awhile, but I just couldn't help myself. :)
I updated the tests to include future and sharedfuture, per your request. I also added tests for the trylock_*() functions that shared_mutex provides that the other mutex's don't. And since upgrade_mutex is just a typedef of shared_mutex, I removed the tests for upgrade_mutex.
test_boost_sleep.cpp.txt Makefile.linux.txt Makefile.windows.txt
Latest results from the develop branch (for comparison to see how far we've come): linux_results_orig_no_monotonic.txt linux_results_orig_yes_monotonic.txt windows_results_orig_no_monotonic.txt
Latest results from the timespec_clocks branch: linux_results_fixed_no_monotonic.txt linux_results_fixed_yes_monotonic.txt windows_results_fixed_no_monotonic.txt
Results with these PR changes: linux_results_austin_no_monotonic.txt linux_results_austin_yes_monotonic.txt windows_results_austin_no_monotonic.txt