dcpomatic/src/lib/j2k_encoder.cc, branch gpu2 DCP-o-matic DCP tools https://git.carlh.net/cgit/dcpomatic/atom?h=gpu2 2020-08-24T07:44:30Z Add loop for gpu multithread; add some timing. 2020-08-24T07:44:30Z Carl Hetherington cth@carlh.net 2020-08-24T07:44:30Z urn:sha1:ec1c6875af12ce761f93dd8547a43227796715ad Fix up rebase and add timer. 2020-08-23T20:50:34Z Carl Hetherington cth@carlh.net 2020-08-23T20:50:28Z urn:sha1:180b1d27d0186526b87852709d1f70c7d2c949f4 Add stub GPU encode worker. 2020-08-23T18:32:49Z Carl Hetherington cth@carlh.net 2019-02-23T00:09:24Z urn:sha1:1d0b111ac9692dedfd4a23c268f362e522105bcf Make class hierarchy to handle different J2K encoder 'worker's. 2020-08-23T18:32:01Z Carl Hetherington cth@carlh.net 2019-02-22T23:14:26Z urn:sha1:90154a126c2b71b4a574ab21b366bd6c72ff6d17 Fix bugs in thread termination causing occasional pthread 2020-07-29T18:22:54Z Carl Hetherington cth@carlh.net 2020-07-29T18:22:54Z urn:sha1:e3c7656f9dc0acbaf518c051b847ee2e4eb7ba23 assertion failures. Before this, it was possible for J2KEncoder::terminate_threads() to finish without terminating all threads if the thread _running_ terminate_threads() was itself interrupt()ed. This is because the thread_group::join_all() in terminate_threads() is an interruption point, so it was possible it not to complete but instead to throw interrupted_exception. Then the owning J2KEncoder would be torn down but the threads would still be running, causing use-after-frees. This commit adds some boost::this_thread::disable_interruption objects to ensure that the owning thread is not interrupted while it is being destroyed. Also tidy up code that does this stuff, assuming that it's safe to not call thread::joinable but instead do thread.interrupt(); try { thread.join(); } catch (...) {} Various OSX warnings fixes. 2020-07-26T19:30:59Z Carl Hetherington cth@carlh.net 2020-07-26T19:29:43Z urn:sha1:6eba051dcbb8c56e3e2efea946ce0380d17a7b33 Use thread_group for improved exception safety (#1785). 2020-07-23T20:53:56Z Carl Hetherington cth@carlh.net 2020-07-23T20:53:43Z urn:sha1:73ebb92e9df01ba7afb97121b6e2cef6ca13a18e Try to give a better error when we fail to start J2K encoding 2020-04-13T22:44:33Z Carl Hetherington cth@carlh.net 2020-04-13T22:44:33Z urn:sha1:778a8b0e19715ed98969ce84cdf4859e1e0e7e52 (probably because of a lack of addressable RAM). See #1540. Can't put boost::thread in a std::list with our macOS compiler. 2020-02-19T20:55:04Z Carl Hetherington cth@carlh.net 2020-02-19T20:55:04Z urn:sha1:5b1b70c86df7225a2a102bdde3b38ea591a6dcbb Various thread cleanups. 2020-01-30T21:54:38Z Carl Hetherington cth@carlh.net 2020-01-30T21:54:38Z urn:sha1:ea6b2dae46caa1da829fbf499e83cd6ae3b3773a