adamrk opened PR #14280 from adamrk:abk/guest-host-async-call-benchmarks to bytecodealliance:main:
Adds a benchmark for async lifted and async lowered nop calls between the guest and host. I'm getting:
component-async/host-to-guest time: [1.4025 µs 1.4068 µs 1.4122 µs] component-async/guest-to-host time: [279.31 ns 280.36 ns 281.67 ns]versus
sync/no-hook/core - host-to-wasm - typed - nop time: [27.718 ns 30.766 ns 33.759 ns] sync/no-hook/core - wasm-to-host - typed - nop time: [7.2183 ns 7.4901 ns 7.8316 ns]for sync core calls.
adamrk requested alexcrichton for a review on PR #14280.
adamrk requested wasmtime-default-reviewers for a review on PR #14280.
:thumbs_up: alexcrichton submitted PR review.
alexcrichton added PR #14280 bench: Add benchmark for guest <-> host async calls with CM-async to the merge queue.
:check: alexcrichton merged PR #14280.
alexcrichton removed PR #14280 bench: Add benchmark for guest <-> host async calls with CM-async from the merge queue.
Last updated: Sep 20 2026 at 18:08 UTC