DEV TOOLS Signal 124
B-level scheduler shows promise for Cargo builds in replay experiments
↗
Illustration only Photo by Quilia on Unsplash
A blog post analyzes Cargo's build scheduling by replaying dependency graphs and finds a b-level priority scheduler promising.
For engineers who build Rust projects, Cargo's scheduling directly affects build times. The post suggests a b-level priority approach could improve on Cargo's default, but the results are preliminary and based on replay experiments, not production builds.
Written by elseif from the cluster below · every claim links back to a sourceThe three things worth knowing
The author records dependency graphs from Cargo builds using syscall tracing, splitting each rustc invocation into frontend and rest nodes.
A b-level scheduler, prioritizing tasks on the critical path, performed best among the simple schedulers tested.
The analysis is based on replay experiments, not on modifying Cargo itself, and does not compare against Cargo's own scheduler.
THE CLUSTER