dcpomatic/src/lib/j2k_encoder.cc, branch disk-optDCP-o-matic DCP toolshttps://git.carlh.net/cgit/dcpomatic/atom?h=disk-opt2021-01-31T02:14:24ZMore enum class additions.2021-01-31T02:14:24ZCarl Hetheringtoncth@carlh.net2021-01-31T02:14:24Zurn:sha1:8fedaaa75c4586a4cc7ffb393bd71d1fdb091dc8More c++ tidying.2021-01-11T15:46:13ZCarl Hetheringtoncth@carlh.net2021-01-11T15:46:13Zurn:sha1:9b946fc5250eff5a5dd4a661896916fcd5d9bd4bBOOST_FOREACH.2021-01-07T22:27:56ZCarl Hetheringtoncth@carlh.net2021-01-04T20:32:13Zurn:sha1:a5d004b0773f633401528392fc28e66d70e13ac8std::shared_ptr2021-01-07T21:48:29ZCarl Hetheringtoncth@carlh.net2021-01-04T20:16:53Zurn:sha1:dd9be86db6cde0afa5da0d1d1ac43b42e05dca26Fix transcode finish time estimations when re-encoding a DCP (#1863).2020-12-05T19:24:18ZCarl Hetheringtoncth@carlh.net2020-12-05T19:24:18Zurn:sha1:a842e3bb5c9900ee14c02d8fcf0f614f32783b4bIt feels unsafe not to lock _threads_mutex between terminate_threads()2020-11-26T01:04:25ZCarl Hetheringtoncth@carlh.net2020-11-26T01:04:25Zurn:sha1:82fbf2b9e43fb234f843dc3352fecbd08eeed3f7
and _threads.reset(); move the lock.
Re-add mutex that was taken away in2020-11-04T22:36:56ZCarl Hetheringtoncth@carlh.net2020-11-04T22:33:10Zurn:sha1:0fc3ee1eac41d391ebb79044b0f46849a90f5138
73ebb92e9df01ba7afb97121b6e2cef6ca13a18e
I'm not sure why it was removed, and it seems necessary;
servers_list_changed() can be called from a variety of different places
(the constructor and two different EncodeServerFinder threads).
May help with #1854.
Pass around JPEG2000 data as a shared_ptr and hence avoid a copy2020-11-02T22:10:04ZCarl Hetheringtoncth@carlh.net2020-11-01T22:51:19Zurn:sha1:2ea3a0d0e4066a166c0700bd1d53daa7f1c50dff
of 4MB of data for every JPEG2000 frame we decode.
Remove support for Windows XP.2020-09-12T21:19:19ZCarl Hetheringtoncth@carlh.net2020-09-12T21:19:19Zurn:sha1:44532cd5f87df3ef10bbbd6b987492e36e4a0eb6Fix bugs in thread termination causing occasional pthread2020-07-29T18:22:54ZCarl Hetheringtoncth@carlh.net2020-07-29T18:22:54Zurn: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 (...) {}