Repository navigation
Conversation
This comment has been minimized.
This comment has been minimized.
cee6920 to
e1d9080
Compare
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
88a3eb5 to
ababbd4
Compare
| //@ needs-target-std | ||
| //@ ignore-backends: gcc | ||
|
|
||
| //@ normalize-stderr: "dev release channel" -> "nightly release channel" |
There was a problem hiding this comment.
Wait, do we actually need this normalization rule? That seems strange?
I don't think we need this for ui tests, run-make tests may have had to do their own channel normalizations, but ui tests should already have this done at a test suite level?
EDIT: apparently we do, I thought this was already normalized but it wasn't, which is curious 🤔
There was a problem hiding this comment.
Oh, it's because only https://doc.rust-lang.org/error_codes/E0554.html mentions the actual channel in the diagnostics, intriguing.
There was a problem hiding this comment.
Thanks, you can r=me after addressing #164802 (comment).
|
@bors r+ rollup |
Migrate stable-specific run-make tests to UI tests Enabled by rust-lang#163664. I still kept the `stable_bare_rustc` function, as it could be genuinely useful (just not in the way it was used in run-make tests so far). Should help unblock rust-lang#162848. Note that the comment for the unstable target feature test was wrong (rust-lang#163428 (comment)). r? jieyouxu
Rollup of 12 pull requests Successful merges: - #164834 (miri subtree update) - #164869 (rustc_codegen_gcc subtree update) - #157123 (library: optimize truncate method) - #159916 (Optimize `str.escape_default().to_string()`) - #164508 (ping me for layout invariant changes) - #164758 (Update dependency jinja2 to v3.1.6 [SECURITY]) - #164802 (Migrate stable-specific run-make tests to UI tests) - #164836 (Add regression test for #162109) - #164857 (Do not checkout Cargo for bootstrap tests) - #164868 (Fix skipping library build with `--keep-stage[-std]`) - #164872 (cg_llvm: clarify Send / Sync story for modules) - #164881 (Add FIXME comments to `from_ascii_bytes_radix_impl`)
|
Hm, |
| LL | trait Bar: [const] Foo {} | ||
| | ^^^^^^^^^^^^^^^^^^^^^^^^^ | ||
|
|
||
| error[E0554]: `#![feature]` may not be used on the nightly release channel |
There was a problem hiding this comment.
This would be wrong / produce the diff as nightly because we pipe the release channel to the diagnostics 🤔 In some sense it's orthogonal to whether #![feature] can be used when we //@ act-as-stable
There was a problem hiding this comment.
Yeah, you can still have "stable-like" behavior, even though the actual channel string is whatever.
|
@bors r- |
|
This pull request was unapproved. |
|
@bors try jobs=test-x86_64-gnu-stable |
This comment has been minimized.
This comment has been minimized.
Migrate stable-specific run-make tests to UI tests try-job: test-x86_64-gnu-stable
This comment has been minimized.
This comment has been minimized.
|
💔 Test for 106dc8e failed: CI. Failed job:
|
This comment has been minimized.
This comment has been minimized.
c71bbdf to
cbdd3f3
Compare
|
@bors try jobs=test-x86_64-gnu-stable |
This comment has been minimized.
This comment has been minimized.
Migrate stable-specific run-make tests to UI tests try-job: test-x86_64-gnu-stable
|
The channel is hardcoded into the compiler binary, so unless we add some testing only dynamic override, we likely can't normalize much better than just turning stable and dev into nightly.. So I think that's good enough for now. @bors r=jieyouxu |
Migrate stable-specific run-make tests to UI tests Enabled by rust-lang#163664. I still kept the `stable_bare_rustc` function, as it could be genuinely useful (just not in the way it was used in run-make tests so far). Should help unblock rust-lang#162848. Note that the comment for the unstable target feature test was wrong (rust-lang#163428 (comment)). r? jieyouxu
Migrate stable-specific run-make tests to UI tests Enabled by rust-lang#163664. I still kept the `stable_bare_rustc` function, as it could be genuinely useful (just not in the way it was used in run-make tests so far). Should help unblock rust-lang#162848. Note that the comment for the unstable target feature test was wrong (rust-lang#163428 (comment)). r? jieyouxu
View all comments
Enabled by #163664.
I still kept the
stable_bare_rustcfunction, as it could be genuinely useful (just not in the way it was used in run-make tests so far).Should help unblock #162848.
Note that the comment for the unstable target feature test was wrong (#163428 (comment)).
r? jieyouxu