background threads Anthony Williams Just Software Solutions Ltd https://www.justsoftwaresolutions.co.uk September 2020Get off my thread: Techniques for moving work to background threads Why do we need thread? We don’t just need to move the work, we need to prevent blocking on our event-handling threads. void event_handler(){ auto handle=spawn_background_task(); handle.wait(); // no benefit }Aside: executorSpawning new threads There are lots of ways to spawn new threads: std::thread std::jthread std::async(std::launch::async,...) Platform-specific APIsSpawning new threads There are lots of ways

std::thread library. We will introduce topics with pragmatic code examples introducing the ideas of threads and locks, and showing how these programming primitives enable concurrency. In addition, we will conditional variables, promises/futures). Attendees will leave this session being able to use threads, locks, and start thinking about architecting multithreaded software. All materials and code samples time ○ The ‘main thread’ is where our program starts ■ We may then have 1 or more additional threads: ● executing a block of code ● executing other functions ● And overall--sharing the same code

happen on different threads https://en.cppreference.com/w/cpp/thread/counting_semaphoreCppCon 2023 9 David Olsen - Back to Basics: Concurrency CONCURRENCY “Multiple logical threads of execution with DefinitionCppCon 2023 10 David Olsen - Back to Basics: Concurrency CONCURRENCY “Multiple logical threads of execution with unknown inter-task dependencies” – Daisy Hollman, “A Unifying Abstraction for “Multiple logical threads of execution with [some] inter-task dependencies” DefinitionCppCon 2023 12 David Olsen - Back to Basics: Concurrency CONCURRENCY “Multiple logical threads of execution with

meticulously designed lock-free queues experience a significant drop in performance as the number of threads increases. ● Multiple sub-queues can be used to mitigate contention but this introduces a myriad of other shortcomings which further complicate issues ● By their nature, queues bring tasks to threads: ○ Requiring tasks to convey both data and logic ○ Often leading to task class hierarchies, pointers contracts in the group) ● Threads are brought to the “task” rather than the “task” being brought to the threads (as is done with Task Queues) ○ Minimize thread contention because threads do not compete for the

mental model for coroutines: Cooperative Threads void spawn_task () { // ... Result r = outer_function (); }16/55 A mental model for coroutines: Cooperative Threads void spawn_task () { // ... Result r Cooperative Threads PartialResult middle_function () { auto r = inner_function (); return PartialResult :: from_io_result (r); }16/55 A mental model for coroutines: Cooperative Threads PartialResult function() inner function() Task18/55 Threads - A straightforward solution std::future spawn_task () { return std:: async( outer_function ); }18/55 Threads - A straightforward solution main()

in Parallel or Consecutively When multiple tests are run, by default they run in parallel using threads. This means the tests will finish running faster, so that we can get faster feedback on whether or of threads used, you can send the --test-threads flag and the number of threads you want to use to the test binary. For example: $ cargo test -- --test-threads=1 We set the number of test threads to tests are running in parallel as we talked about in the previous section. Try using both the --test-threads=1 option and the --nocapture function and see what the output looks like then! Running a Subset

. . . . . . . . . . . . . . . . 297 24 Multi-Threading 299 24.1 Starting Julia with multiple threads . . . . . . . . . . . . . . . . . . . . . . . . . . . 299 24.2 Threadpools . . . . . . . . . . Communication and synchronization . . . . . . . . . . . . . . . . . . . . . . . . . . 301 24.4 The @threads Macro . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 302 24.5 Atomic Operations new method definition won't be visible to the current runtime environment, including Tasks and Threads (and any previously defined @generated functions). Let's start with an example to see what this

GCopyright © 2020 Bob Steagall K E W B C O M P U T I N G Overview • Sharing a container among multiple threads • A motivating problem • Some possible solutions • A solution based on strict timestamp ordering Steagall K E W B C O M P U T I N G Sharing a Container • Sometimes a container must be shared between threads • We desire to avoid race conditions during writes • Assume elements are themselves unsynchronized Bob Steagall K E W B C O M P U T I N G Sharing a Container – Avoiding Race Conditions • If all threads are readers… • No locking is required • If the number of reads is much larger than the number of

. . . . . . . . . . . . . . . . 297 24 Multi-Threading 299 24.1 Starting Julia with multiple threads . . . . . . . . . . . . . . . . . . . . . . . . . . . 299 24.2 Threadpools . . . . . . . . . . Communication and synchronization . . . . . . . . . . . . . . . . . . . . . . . . . . 301 24.4 The @threads Macro . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 302 24.5 Atomic Operations new method definition won't be visible to the current runtime environment, including Tasks and Threads (and any previously defined @generated functions). Let's start with an example to see what this

. . . . . . . . . . . . . . . . 297 24 Multi-Threading 299 24.1 Starting Julia with multiple threads . . . . . . . . . . . . . . . . . . . . . . . . . . . 299 24.2 Threadpools . . . . . . . . . . Communication and synchronization . . . . . . . . . . . . . . . . . . . . . . . . . . 301 24.4 The @threads Macro . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 302 24.5 Atomic Operations new method definition won't be visible to the current runtime environment, including Tasks and Threads (and any previously defined @generated functions). Let's start with an example to see what this