Some simple examples, intro to senders 3. The lifecycle of an async operation 4. Under the hood of a concurrent operation 5. Implementing a simple algorithm 6. Senders and coroutines Part 2: 1. Structured compute_intensive(1); }), ex::schedule(sched) | ex::then([] { return compute_intensive(2); }) );11 SENDERS ARE EXPRESSION TEMPLATES ex::sender auto work = ex::when_all( ex::schedule(sched) | ex::then([] return compute_intensive(2); }) ); when_all then then then schedule schedule schedule12 SENDERS ARE EXPRESSION TEMPLATES ex::sender auto work = ex::when_all( ex::schedule(sched) | ex::then([]

Baremetal Senders https://github.com/intel/cpp-baremetal-senders-and-receivers 17just just_result_of just_error just_error_result_of just_stopped Sender Factories Functions that return senders. auto just_result_of just_error just_error_result_of just_stopped Sender Factories Functions that return senders. auto sndr = async::just_result_of( [] { return 42; }, [] { do_something(); [] { return 17; }); 1 2 3 4 19schedule Sender Factories Functions that return senders. auto sndr = sched.schedule(); 1 20continue_on start_on let_value let_error let_stopped then sequence

end-points: the Sender and the Receiver. • Channel Type: SPSC, SPMC, MPMC, MPSC. Senders Receive Senders Senders Receive Receivers Perf Gaps: CAS(Compare And Swap) can't perfectly scale with latency • No memory/cache latency • No CPU cycles DLB-Assist Channel Intro Hardware Senders Receive Senders Senders Receive Receivers Channel/Queue DLB-Assist Channel removes the penalty of CAS to achieve being dropped. Message-ID Reclaiming • Reclaim Message-ID • Transfers Message-ID through HW Senders Receivers Channel Hardware Hardware Producer Consumer Give Ownership Take Ownership Message-ID

105.1ASYNCHRONOUS EXECUTION ASYNCHRONOUS EXECUTION Let's talk about senders and receivers for a minute. Let's talk about senders and receivers for a minute. I'm pretty sure the authors of P2300 made passes the result to a function that returns a sender bind when_all completes when all the input senders complete (<*>) just sends the provided values pure 109P2300 P2300 This is how we know (in part) fmap-generalization -generalization when_all when_all can clearly run all its input senders in parallel, calling can clearly run all its input senders in parallel, calling the (multi-argument) function when all arguments

system to signal the availability of buffer space from receivers to senders. ??? Vasiliki Kalavri | Boston University 2020 27 • Senders maintain a credit balance for all their receivers and receivers regularly and is implemented in Apache Flink. • Each task informs its senders of its buffer availability via credit messages. • This way, senders always know whether receivers have the required capacity to

sending values to a closed channel will panic, so it is dangerous to send values to a channel if the senders don't know whether or not the channel is closed. The criticisms look reasonable (in fact not). Yes channel from the receiver side and don't close a channel if the channel has multiple concurrent senders. In other words, we should only close a channel in a sender goroutine if the sender is the only sender goroutine can close the channel safely. However, making such guarantees by a receiver or by one of many senders of a channel usually needs much effort, and often makes code complicated. On the contrary, it is

sending values to a closed channel will panic, so it is dangerous to send values to a channel if the senders don't know whether or not the channel is closed. The criticisms look reasonable (in fact not). Yes channel from the receiver side and don't close a channel if the channel has multiple concurrent senders. In other words, we should only close a channel in a sender goroutine if the sender is the only sender goroutine can close the channel safely. However, making such guarantees by a receiver or by one of many senders of a channel usually needs much effort, and often makes code complicated. On the contrary, it is

sending values to a closed channel will panic, so it is dangerous to send values to a channel if the senders don't know whether or not the channel is closed. The criticisms look reasonable (in fact not). Yes channel from the receiver side and don't close a channel if the channel has multiple concurrent senders. In other words, we should only close a channel in a sender goroutine if the sender is the only goroutine can close the channel safely. However, making such guarantees by a receiver or by one of many senders of a channel usually needs much effort, and often makes code complicated. On the contrary, it is

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256 41.7. Senders and Receivers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . much higher performance applications. Messaging systems decouple the senders of messages from the consumers of messages. The senders and consumers of messages are completely independent and know nothing consumers on the queue but a particular message will only ever be consumed by a maximum of one of them. Senders (also known as producers) to the queue are completely decoupled from receivers (also known as consumers)