You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Now that I think about it, cargo proper probably doesn't know about the +toolchain either; it's handled by the rustup stub. But since rustup doesn't know about xargo, but xargo knows about rustup, it seems xargo should be the one interpreting and redirecting the flag: xargo +toolchain -> rustup run toolchain xargo, which works fine.
The thing is that Xargo does nothing (it doesn't build a sysroot) when invoked with a stable / beta toolchain. The problem here is that Xargo uses the rustc_version crate, which knows nothing about rustup, to get information (e.g. what channel is being used) about the toolchain.
To fix this we could special case the +$channel argument and use the rustc_version::version_meta_for function with the output of rustc +$channel -Vv instead of vanilla rustc_version::version_meta.
Running
cargo +nightly build
builds with rustup's "nightly" toolchain.However, xargo seems to ignore this option; when it's passed, xargo doesn't even attempt to build libcore:
The text was updated successfully, but these errors were encountered: