Remove the std workspace patch for compiler-builtins
#145622
Conversation
All dependencies of `std` have dropped the crates.io dependency on `compiler-builtins`, so this patch is no longer needed. Closes: RUST-142265
|
rustbot has assigned @Mark-Simulacrum. Use |
rustbot
added
the
S-waiting-on-review
|
@bors r+ |
bors
added
S-waiting-on-bors
Rollup of 13 pull requests Successful merges: - #139357 (Fix parameter order for `_by()` variants of `min` / `max`/ `minmax` in `std::cmp`) - #140314 (Rustdoc: typecheck scrape-examples.js) - #140794 (mention lint group in default level lint note) - #145006 (Clarify EOF handling for `BufRead::skip_until`) - #145252 (Demote x86_64-apple-darwin to Tier 2 with host tools) - #145359 (Fix bug where `rustdoc-js` tester would not pick the right `search.js` file if there is more than one) - #145381 (Implement feature `int_lowest_highest_one` for integer and NonZero types) - #145417 (std_detect: RISC-V platform guide documentation) - #145531 (Add runtime detection for APX-F and AVX10) - #145619 (`std_detect`: Use `rustc-std-workspace-*` to pull in `compiler-builtins`) - #145622 (Remove the std workspace patch for `compiler-builtins`) - #145623 (Pretty print the name of an future from calling async closure) - #145626 (add a fallback implementation for the `prefetch_*` intrinsics ) r? `@ghost` `@rustbot` modify labels: rollup
Rollup merge of #145622 - tgross35:remove-builtins-patch, r=Mark-Simulacrum Remove the std workspace patch for `compiler-builtins` All dependencies of `std` have dropped the crates.io dependency on `compiler-builtins`, so this patch is no longer needed. Closes: RUST-142265
This looks like a JIRA link or something, so GH didn't autoclose the issue :3 |
|
Huh, I thought that r-l/r was configured to use this kind of links (it is used to avoid spamming issues with links from commits). |
|
Argh, I’ve switched to that style link to shut triagebot up. I guess we get both autoclose and backlinks or neither of them. |
Rollup of 13 pull requests Successful merges: - rust-lang/rust#139357 (Fix parameter order for `_by()` variants of `min` / `max`/ `minmax` in `std::cmp`) - rust-lang/rust#140314 (Rustdoc: typecheck scrape-examples.js) - rust-lang/rust#140794 (mention lint group in default level lint note) - rust-lang/rust#145006 (Clarify EOF handling for `BufRead::skip_until`) - rust-lang/rust#145252 (Demote x86_64-apple-darwin to Tier 2 with host tools) - rust-lang/rust#145359 (Fix bug where `rustdoc-js` tester would not pick the right `search.js` file if there is more than one) - rust-lang/rust#145381 (Implement feature `int_lowest_highest_one` for integer and NonZero types) - rust-lang/rust#145417 (std_detect: RISC-V platform guide documentation) - rust-lang/rust#145531 (Add runtime detection for APX-F and AVX10) - rust-lang/rust#145619 (`std_detect`: Use `rustc-std-workspace-*` to pull in `compiler-builtins`) - rust-lang/rust#145622 (Remove the std workspace patch for `compiler-builtins`) - rust-lang/rust#145623 (Pretty print the name of an future from calling async closure) - rust-lang/rust#145626 (add a fallback implementation for the `prefetch_*` intrinsics ) r? `@ghost` `@rustbot` modify labels: rollup
|
So now we have https://github.com/rust-lang/rust/tree/master/library/compiler-builtins but that's not actually used by the sysroot build? That seems like it could be really confusing. EDIT: Ah no there is a path dependency... I don't think I understand how this patch worked.^^ This seems to just overwrite a path dependency with... itself? |
|
Some other dependencies of stdlib depended on compiler-builts from crates.io (IIRC), the patch was used to redirect those indirect dependencies to the in-tree version. |
|
I just commented on rust-lang/miri#4534 but yeah, the patch was only needed while std had dependencies that used compiler-builtins via crates.io. This used to be required but since rustc-std-workspace-core now pulls in the in-tree compiler-builtins, the patch was needed to make sure we don't wind up with both in-tree and crates.io builtins. The patch was just a migration path to give std's depsendencies a chance to delete their dep on I don't know what would have caused the Miri failure though, unless an old depsendencies was getting used somewhere. I thought Cargo.toml was updated for everything, but maybe it came via an indirect dep that we don't set a minimum version for? Edit: nevermind, Ralf cleared that up rust-lang/miri#4534 (comment) |
…Mark-Simulacrum Remove the std workspace patch for `compiler-builtins` All dependencies of `std` have dropped the crates.io dependency on `compiler-builtins`, so this patch is no longer needed. Closes: RUST-142265
All dependencies of
stdhave dropped the crates.io dependency oncompiler-builtins, so this patch is no longer needed.Closes: RUST-142265