Folkert de Vries

Right, that's the issue we've been having. `cfg(bootstrap)` at least makes it possible to make changes, but it's a very tedious process. Seems like there is no real way around...

This was merged in https://github.com/rust-lang/rust/pull/141992 over in the rust repo itself.

at this point, let's wait for the audit to be done so that we can incorporate any improvements. That should give us a version that we can make the default...

It would be extremely surprising if `zlib-rs` caused this: we produce byte-for-byte equivalent output to `zlib-ng` for each compression level.

If the previous default was `miniz_oxide`, that does make sense: `zlib-rs` and `zlib-ng` have a much faster level 1, but it also compresses much worse. Generally, `zlib-rs` compression level 2...

@taiki-e this is not done, but does it look like a good direction to explore further?

in either case the alloca must be outside of the loop otherwise the stack grows on every iteration. I do think the problem is that the "next" variable is never...

Given that there is a `nextest` subcommand already, would it not make more sense to add a `cargo llvmcov fuzz` subcommand directly?

Sure well `cargo llvm-cov fuzz-coverage` could work does make more sense. It would run `cargo fuzz coverage` and then report the code coverage in the usual way. From a UX...

Yeah that seems independently useful because it is more general. But integration with other rust tools also seems valuable to me. This comment says https://github.com/taiki-e/cargo-llvm-cov/pull/144#pullrequestreview-913381236 > Ideally, I would like...