On Thu, 10 Aug 2023 11:23:02 GMT, Pavel Rappo <pra...@openjdk.org> wrote:
>> AFAIK, if you need reproducible randoms in tests, you should add the tags: >> >> * @key randomness >> * @library /test/lib >> >> and initialize your random generator with >> >> import jdk.test.lib.RandomFactory; >> ... >> Random rnd = RandomFactory.getRandom(); >> >> This prints the seed to STDOUT. > >> AFAIK, if you need reproducible randoms in tests, you should add the tags: >> >> ``` >> * @key randomness >> * @library /test/lib >> ``` >> >> and initialize your random generator with >> >> ``` >> import jdk.test.lib.RandomFactory; >> ... >> Random rnd = RandomFactory.getRandom(); >> ``` >> >> This prints the seed to STDOUT. > > It'd be applicable for tests, but not for benchmarks. Also, the point that > @cl4es seems to make is that randomness is captured and set up to reduce > noise, not to be able to reproduce "benchmark failures". My bad. Forget my note ------------- PR Review Comment: https://git.openjdk.org/jdk/pull/14630#discussion_r1289981733