The test-util feature flag is only intended to be used with tests.
However, it is possible to enable it in release mode accidentally. This
patch reduces the overhead of `Instant::now()` when the `test-util`
feature flag is enabled but `time::pause()` is not called.
The optimization is implemented by adding a static atomic flag that
tracks if `time::pause()` has ever been called. In `Instant::now()`, the
atomic flag is first checked before the thread-local and mutex are
accessed.
Some users mentioned that the behavior of a channel when the receivers
and/or senders are dropped isn't explicitly documented.
This change adds wording to the documentation for each channel in the
sync module, explaining under which conditions messages in a channel are
dropped with respect to dropping the senders and the receivers.
Refs: #5490
These flags were previously only needed due to a bug in the `cargo-semver-checks` CLI logic.
The correct behavior (available as of v0.18.3) for `cargo-semver-checks` is to ignore `publish = false` crates when scanning a workspace, *unless* those crates are specifically selected for checking.
All the crates being excluded here are `publish = false` so they are already excluded by the default behavior, so all `--exclude` flags are no-ops.
This additional benchmark exercises a common request/reply pattern using an MPSC for requests along with a oneshot payload as a reply mechanism. When used in a current threaded scenario, the bench is 17 times faster on my machine than when using the multi-threaded runtime and one worker thread. Not only that, but if I increase the number of worker threads to 6, performance degrades further.
Does this suggest a scheduling problem with the multi-threaded runtime?
No matter what, hopefully the benchmarks are a useful addition.
This patch removes `Arc` from Tokio's internal clock source. Instead of
cloning `Clock` when needed, a reference is passed into functions that
need to get the current instant.
Noah Kennedy