Open zamazan4ik opened 1 year ago
I don't believe so, but I also would not expect much gain here. There's not much CPU-intensive work in the entire application, last I knew, and most of what there is is just calculating hashes, which is already a highly-optimized operation. 99% of what sccache does is sitting around waiting for I/O (reading/writing cache entries) and waiting for compilers to finish.
Hi!
Did anyone try to apply Profile-Guided Optimization (PGO) to
sccache
itself? It could optimize the performance "automatically" in CPU-intensive parts of the application. Since it's written in Rust it will be pretty easy to test with cargo-pgo.