thread_local seems to be highly controversial according to the discussion in #14953, so remove it again from the tests.
Also remove boost::thread_group in the test that uses it, since I am touching it anyway.
thread_local seems to be highly controversial according to the discussion in #14953, so remove it again from the tests.
Also remove boost::thread_group in the test that uses it, since I am touching it anyway.
25 | @@ -26,7 +26,7 @@ std::ostream& operator<<(typename std::enable_if<std::is_enum<T>::value, std::os 26 | return stream << static_cast<typename std::underlying_type<T>::type>(e); 27 | } 28 | 29 | -thread_local extern FastRandomContext g_insecure_rand_ctx; 30 | +extern FastRandomContext g_insecure_rand_ctx; 31 | 32 | static inline void SeedInsecureRand(bool deterministic = false)
Would make sense to add a warning comment for these helper functions to only use them from the main thread.
utACK fa61202cae6979c754b11842c5ed8d90f1cdcc45
utACK fa61202.
151 | @@ -152,12 +152,13 @@ BOOST_AUTO_TEST_CASE(processnewblock_signals_ordering) 152 | // create a bunch of threads that repeatedly process a block generated above at random 153 | // this will create parallelism and randomness inside validation - the ValidationInterface 154 | // will subscribe to events generated during block validation and assert on ordering invariance 155 | - boost::thread_group threads;
👋
Milestone
0.18.0