-
Notifications
You must be signed in to change notification settings - Fork 12.7k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Support non-stage0 check #81064
Support non-stage0 check #81064
Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This is a lot better than my approach I think, but I'm worried it will break when cross-compiling, because ensure(Std)
will build it for the target, not the host. Can you run x.py check --target x8_64-pc-windows-gnu
or something to make sure this works?
This comment has been minimized.
This comment has been minimized.
51a0478
to
6a7be27
Compare
This comment has been minimized.
This comment has been minimized.
6a7be27
to
97ddd41
Compare
5f35f50
to
21e16e5
Compare
Adjusted to support target checking (well, host checking, but it's not actually important). |
21e16e5
to
53989e4
Compare
This is great, thanks so much :) @bors r+ |
📌 Commit 53989e4 has been approved by |
…k, r=jyn514 Support non-stage0 check Seems to work locally - a full stage 1 check succeeds, building std (because we can't get away with checking it), and then checking the compiler and other tools. This ran into the problem that a unconditional x.py check in stage 1 *both* checks and builds stage 1 std, and then has to clean up because for some reason the rmeta and rlib artifacts conflict (though I'm not actually entirely sure why, but it doesn't seem worth digging in in too much detail). Ideally we wouldn't be building and checking like that but it's a minor worry as checking std is pretty fast and you can avoid it if you're aiming for speed by passing the compiler (e.g., compiler/rustc) explicitly. r? `@jyn514`
…k, r=jyn514 Support non-stage0 check Seems to work locally - a full stage 1 check succeeds, building std (because we can't get away with checking it), and then checking the compiler and other tools. This ran into the problem that a unconditional x.py check in stage 1 *both* checks and builds stage 1 std, and then has to clean up because for some reason the rmeta and rlib artifacts conflict (though I'm not actually entirely sure why, but it doesn't seem worth digging in in too much detail). Ideally we wouldn't be building and checking like that but it's a minor worry as checking std is pretty fast and you can avoid it if you're aiming for speed by passing the compiler (e.g., compiler/rustc) explicitly. r? ``@jyn514``
Rollup of 13 pull requests Successful merges: - rust-lang#79298 (correctly deal with late-bound lifetimes in anon consts) - rust-lang#80031 (resolve: Reject ambiguity built-in attr vs different built-in attr) - rust-lang#80201 (Add benchmark and fast path for BufReader::read_exact) - rust-lang#80635 (Improve diagnostics when closure doesn't meet trait bound) - rust-lang#80765 (resolve: Simplify collection of traits in scope) - rust-lang#80932 (Allow downloading LLVM on Windows and MacOS) - rust-lang#80983 (Remove is_dllimport_foreign_item definition from cg_ssa) - rust-lang#81064 (Support non-stage0 check) - rust-lang#81080 (Force vec![] to expression position only) - rust-lang#81082 (BTreeMap: clean up a few more comments) - rust-lang#81084 (Use Option::map instead of open-coding it) - rust-lang#81095 (Use Option::unwrap_or instead of open-coding it) - rust-lang#81107 (Add NonZeroUn::is_power_of_two) Failed merges: r? `@ghost` `@rustbot` modify labels: rollup
Stage 1 check has been supported since rust-lang#81064. rust-lang#81064 changed the error message for this, but I don't think there's any reason we should prevent using it. I tested locally and `keep-stage` works fine. Don't give a hard error when trying to use it.
…ulacrum Support `x --keep-stage 0 check` Stage 1 check has been supported since rust-lang#81064. rust-lang#81064 changed the error message for this, but I don't think there's any reason we should prevent using it. I tested locally and `keep-stage` works fine. r? `@Mark-Simulacrum`
Seems to work locally - a full stage 1 check succeeds, building std (because we can't get away with checking it), and then checking the compiler and other tools. This ran into the problem that a unconditional x.py check in stage 1 both checks and builds stage 1 std, and then has to clean up because for some reason the rmeta and rlib artifacts conflict (though I'm not actually entirely sure why, but it doesn't seem worth digging in in too much detail).
Ideally we wouldn't be building and checking like that but it's a minor worry as checking std is pretty fast and you can avoid it if you're aiming for speed by passing the compiler (e.g., compiler/rustc) explicitly.
r? @jyn514