Remove //@ compare-output-lines-by-subset
#133725
Conversation
There was only ever one test which used this flag, and it was removed in rust-lang#132244. I think this is a bad flag that should never have been added; comparing by subset makes the test failures extremely hard to debug. Any test that needs complicated output filtering like this should just use run-make instead. Note that this does not remove the underlying comparison code, because it's still used if `runner` is set. I don't quite understand what's going on there, but since we still test on other platforms and in CI that the full output is accurate, I think it will be easier to debug than a test that uses compare-by-subset unconditionally.
|
r? @wesleywiser rustbot has assigned @wesleywiser. Use |
rustbot
added
A-compiletest
|
r? bootstrap |
|
r? jieyouxu |
|
You can r=me after PR CI is green. |
|
@bors r+ rollup |
bors
added
S-waiting-on-bors
|
This was added for ferrocene in #110444, back then it wasn't used by the project at implementation time either. More specifically this was used mainly for testing CLI flag output which frequently changes, I don't think modeling those a Though given the actual implementation logic stays (for now) I guess we can revert this PR in ferrocene given the diff itself is small. Though the story changes of course if the implementation also goes away |
I don't foresee us removing the implementation because it's still needed to handle test runner output. However, I don't want to accept tests in tree with |
run_make_support intentionally has support for diffing output: https://github.com/rust-lang/rust/blob/master/src/tools/run-make-support/src/diff/mod.rs |
…llaumeGomez Rollup of 13 pull requests Successful merges: - rust-lang#133603 (Eliminate magic numbers from expression precedence) - rust-lang#133715 (rustdoc-json: Include safety of `static`s) - rust-lang#133721 (rustdoc-json: Add test for `impl Trait for dyn Trait`) - rust-lang#133725 (Remove `//@ compare-output-lines-by-subset`) - rust-lang#133730 (Add pretty-printer parenthesis insertion test) - rust-lang#133736 (Add `needs-target-has-atomic` directive) - rust-lang#133739 (Re-add myself to rotation) - rust-lang#133743 (Fix docs for `<[T]>::as_array`.) - rust-lang#133744 (Fix typo README.md) - rust-lang#133745 (Remove static HashSet for default IDs list) - rust-lang#133749 (mir validator: don't store mir phase) - rust-lang#133751 (remove `Ty::is_copy_modulo_regions`) - rust-lang#133757 (`impl Default for EarlyDiagCtxt`) r? `@ghost` `@rustbot` modify labels: rollup
Rollup merge of rust-lang#133725 - jyn514:remove-compare-output-subset, r=jieyouxu Remove `//@ compare-output-lines-by-subset` There was only ever one test which used this flag, and it was removed in rust-lang#132244. I think this is a bad flag that should never have been added; comparing by subset makes the test failures extremely hard to debug. Any test that needs complicated output filtering like this should just use run-make instead. Note that this does not remove the underlying comparison code, because it's still used if `runner` is set. I don't quite understand what's going on there, but since we still test on other platforms and in CI that the full output is accurate, I think it will be easier to debug than a test that uses compare-by-subset unconditionally. rustc-dev-guide update PR: rust-lang/rustc-dev-guide#2151
There was only ever one test which used this flag, and it was removed in #132244. I think this is a bad flag that should never have been added; comparing by subset makes the test failures extremely hard to debug. Any test that needs complicated output filtering like this should just use run-make instead.
Note that this does not remove the underlying comparison code, because it's still used if
runneris set. I don't quite understand what's going on there, but since we still test on other platforms and in CI that the full output is accurate, I think it will be easier to debug than a test that uses compare-by-subset unconditionally.rustc-dev-guide update PR: rust-lang/rustc-dev-guide#2151