Rollup of 5 pull requests by matthiaskrgr · Pull Request #153245 · rust-lang/rust

matthiaskrgr · 2026-03-01T06:21:10Z

Successful merges:

Suggest async block instead of async closure when possible #152042 (Suggest async block instead of async closure when possible)
Introduce --ci flag in tidy #152949 (Introduce --ci flag in tidy)
Various small query cleanups #153169 (Various small query cleanups)
Disable debug_assert_not_in_new_nodes for multiple threads #152655 (Disable debug_assert_not_in_new_nodes for multiple threads)
rustfmt: add test for field representing type builtin syntax #153229 (rustfmt: add test for field representing type builtin syntax)

Failed merges:

Clean up QueryVTable::hash_result into hash_value_fn #153209 (Clean up QueryVTable::hash_result into hash_value_fn)

* add --ci flag in tidy This commit introduces --ci flag in tidy because currently bootstrap can't pass its ci env information to tidy. It also modifies how CiInfo initialize its ci_env variable. tidy codes which uses CiEnv::is_ci for checking ci are now using ci_env in CiInfo. * address review - Fix comment - Use Option for ci flag in order to have true/false explicitly or unspecified (implicit false) * integrate CiInfo into TidyCtx * remove CiInfo * CiEnv::current() should be called when ci flag is not added * extract base_commit() to a separate function * use &TidyCtx instead of clone

It's unused.

It has a single use.

This makes it clear the `CycleError` is not used after the call.

When there are two cases of equal size and importance, I find an if/else expression easier to read than an early return.

It reads better that way.

It's an unnecessary level of indirection.

Calling `abort_if_errors` after emitting an error is guaranteed to call `raise_fatal`, so just do that directly instead.

This exact comment block also appears in `compiler/rustc_middle/src/queries.rs`, which is a better place for it.

`Value` is an unhelpfully generic name. Standard naming procedure for a trait with a single method is for the trait name to match the method name, which is what this commit does. Likewise, the enclosing module is renamed from `values` to `from_cycle_error`. Also add a comment about some non-obvious behaviour.

It has no effect. `symbol_name` is the only query that produces a `SymbolName`. If it was marked with `cycle_delayed_bug`/`cycle_stash` then this `FromCycleError` impl would make sense, but that's not the case. Maybe it was the case in the past.

@estebank

…-closure-suggestion, r=JohnTitor Suggest async block instead of async closure when possible Fixesrust-lang#140265 r? @estebank

Introduce --ci flag in tidy ## Context Currently tidy doesn't have `--ci` flag. Even if bootstrap has --ci=true, it isn't passed to tidy. As a result, some parts of tidy where we need to turn on `ci mode` are never executed on our local. ## Change This PR introduces --ci flag in tidy, and modify CiInfo and other parts of tidy where currently use `CiEnv::is_ci()` for checking if the running environment is ci or not.

@oli-obk

…ochenkov Various small query cleanups I found these while doing a close read of the query code. r? @oli-obk

…51509, r=nnethercote Disable debug_assert_not_in_new_nodes for multiple threads Fixesrust-lang#152654 This debug assertion is no longer intended to succeed for the parallel compiler since rust-lang#151509 doesn't panic on a red-green query graph coloring race, generating both red and green nodes (rust-lang#151509 (comment))

@ytmimi

rustfmt: add test for field representing type builtin syntax As requested by @ytmimi in rust-lang#152730 (comment). Not sure if you want/need a more complex example. Just let me know.

matthiaskrgr · 2026-03-01T06:21:27Z

@bors r+ rollup=never p=5

rust-bors · 2026-03-01T06:21:30Z

📌 Commit 8e1ec75 has been approved by matthiaskrgr

It is now in the queue for this repository.

rust-bors · 2026-03-01T08:31:30Z

PR #153169, which is a member of this rollup, was unapproved.
This rollup was thus also unapproved.

rust-bors · 2026-03-01T08:31:32Z

☔ The latest upstream changes (presumably #153114) made this pull request unmergeable. Please resolve the merge conflicts.

matthiaskrgr · 2026-03-01T06:21:10Z

Successful merges:

Suggest async block instead of async closure when possible #152042 (Suggest async block instead of async closure when possible)
Introduce --ci flag in tidy #152949 (Introduce --ci flag in tidy)
Various small query cleanups #153169 (Various small query cleanups)
Disable debug_assert_not_in_new_nodes for multiple threads #152655 (Disable debug_assert_not_in_new_nodes for multiple threads)
rustfmt: add test for field representing type builtin syntax #153229 (rustfmt: add test for field representing type builtin syntax)

Failed merges:

Clean up QueryVTable::hash_result into hash_value_fn #153209 (Clean up QueryVTable::hash_result into hash_value_fn)

r? @ghost

Create a similar rollup

* add --ci flag in tidy This commit introduces --ci flag in tidy because currently bootstrap can't pass its ci env information to tidy. It also modifies how CiInfo initialize its ci_env variable. tidy codes which uses CiEnv::is_ci for checking ci are now using ci_env in CiInfo. * address review - Fix comment - Use Option for ci flag in order to have true/false explicitly or unspecified (implicit false) * integrate CiInfo into TidyCtx * remove CiInfo * CiEnv::current() should be called when ci flag is not added * extract base_commit() to a separate function * use &TidyCtx instead of clone

It's unused.

It has a single use.

This makes it clear the `CycleError` is not used after the call.

When there are two cases of equal size and importance, I find an if/else expression easier to read than an early return.

It reads better that way.

It's an unnecessary level of indirection.

Calling `abort_if_errors` after emitting an error is guaranteed to call `raise_fatal`, so just do that directly instead.

This exact comment block also appears in `compiler/rustc_middle/src/queries.rs`, which is a better place for it.

`Value` is an unhelpfully generic name. Standard naming procedure for a trait with a single method is for the trait name to match the method name, which is what this commit does. Likewise, the enclosing module is renamed from `values` to `from_cycle_error`. Also add a comment about some non-obvious behaviour.

It has no effect. `symbol_name` is the only query that produces a `SymbolName`. If it was marked with `cycle_delayed_bug`/`cycle_stash` then this `FromCycleError` impl would make sense, but that's not the case. Maybe it was the case in the past.

@estebank

…-closure-suggestion, r=JohnTitor Suggest async block instead of async closure when possible Fixesrust-lang#140265 r? @estebank

Introduce --ci flag in tidy ## Context Currently tidy doesn't have `--ci` flag. Even if bootstrap has --ci=true, it isn't passed to tidy. As a result, some parts of tidy where we need to turn on `ci mode` are never executed on our local. ## Change This PR introduces --ci flag in tidy, and modify CiInfo and other parts of tidy where currently use `CiEnv::is_ci()` for checking if the running environment is ci or not.

@oli-obk

…ochenkov Various small query cleanups I found these while doing a close read of the query code. r? @oli-obk

…51509, r=nnethercote Disable debug_assert_not_in_new_nodes for multiple threads Fixesrust-lang#152654 This debug assertion is no longer intended to succeed for the parallel compiler since rust-lang#151509 doesn't panic on a red-green query graph coloring race, generating both red and green nodes (rust-lang#151509 (comment))

@ytmimi

rustfmt: add test for field representing type builtin syntax As requested by @ytmimi in rust-lang#152730 (comment). Not sure if you want/need a more complex example. Just let me know.

matthiaskrgr · 2026-03-01T06:21:27Z

@bors r+ rollup=never p=5

rust-bors · 2026-03-01T06:21:30Z

📌 Commit 8e1ec75 has been approved by matthiaskrgr

It is now in the queue for this repository.

rust-bors · 2026-03-01T08:31:30Z

PR #153169, which is a member of this rollup, was unapproved.
This rollup was thus also unapproved.

rust-bors · 2026-03-01T08:31:32Z

☔ The latest upstream changes (presumably #153114) made this pull request unmergeable. Please resolve the merge conflicts.

matthiaskrgr · 2026-03-01T06:21:10Z

Successful merges:

Suggest async block instead of async closure when possible #152042 (Suggest async block instead of async closure when possible)
Introduce --ci flag in tidy #152949 (Introduce --ci flag in tidy)
Various small query cleanups #153169 (Various small query cleanups)
Disable debug_assert_not_in_new_nodes for multiple threads #152655 (Disable debug_assert_not_in_new_nodes for multiple threads)
rustfmt: add test for field representing type builtin syntax #153229 (rustfmt: add test for field representing type builtin syntax)

Failed merges:

Clean up QueryVTable::hash_result into hash_value_fn #153209 (Clean up QueryVTable::hash_result into hash_value_fn)

r? @ghost

Create a similar rollup

* add --ci flag in tidy This commit introduces --ci flag in tidy because currently bootstrap can't pass its ci env information to tidy. It also modifies how CiInfo initialize its ci_env variable. tidy codes which uses CiEnv::is_ci for checking ci are now using ci_env in CiInfo. * address review - Fix comment - Use Option for ci flag in order to have true/false explicitly or unspecified (implicit false) * integrate CiInfo into TidyCtx * remove CiInfo * CiEnv::current() should be called when ci flag is not added * extract base_commit() to a separate function * use &TidyCtx instead of clone

It's unused.

It has a single use.

This makes it clear the `CycleError` is not used after the call.

When there are two cases of equal size and importance, I find an if/else expression easier to read than an early return.

It reads better that way.

It's an unnecessary level of indirection.

Calling `abort_if_errors` after emitting an error is guaranteed to call `raise_fatal`, so just do that directly instead.

This exact comment block also appears in `compiler/rustc_middle/src/queries.rs`, which is a better place for it.

`Value` is an unhelpfully generic name. Standard naming procedure for a trait with a single method is for the trait name to match the method name, which is what this commit does. Likewise, the enclosing module is renamed from `values` to `from_cycle_error`. Also add a comment about some non-obvious behaviour.

It has no effect. `symbol_name` is the only query that produces a `SymbolName`. If it was marked with `cycle_delayed_bug`/`cycle_stash` then this `FromCycleError` impl would make sense, but that's not the case. Maybe it was the case in the past.

@estebank

…-closure-suggestion, r=JohnTitor Suggest async block instead of async closure when possible Fixesrust-lang#140265 r? @estebank

Introduce --ci flag in tidy ## Context Currently tidy doesn't have `--ci` flag. Even if bootstrap has --ci=true, it isn't passed to tidy. As a result, some parts of tidy where we need to turn on `ci mode` are never executed on our local. ## Change This PR introduces --ci flag in tidy, and modify CiInfo and other parts of tidy where currently use `CiEnv::is_ci()` for checking if the running environment is ci or not.

@oli-obk

…ochenkov Various small query cleanups I found these while doing a close read of the query code. r? @oli-obk

…51509, r=nnethercote Disable debug_assert_not_in_new_nodes for multiple threads Fixesrust-lang#152654 This debug assertion is no longer intended to succeed for the parallel compiler since rust-lang#151509 doesn't panic on a red-green query graph coloring race, generating both red and green nodes (rust-lang#151509 (comment))

@ytmimi

rustfmt: add test for field representing type builtin syntax As requested by @ytmimi in rust-lang#152730 (comment). Not sure if you want/need a more complex example. Just let me know.

matthiaskrgr · 2026-03-01T06:21:27Z

@bors r+ rollup=never p=5

rust-bors · 2026-03-01T06:21:30Z

📌 Commit 8e1ec75 has been approved by matthiaskrgr

It is now in the queue for this repository.

rust-bors · 2026-03-01T08:31:30Z

PR #153169, which is a member of this rollup, was unapproved.
This rollup was thus also unapproved.

rust-bors · 2026-03-01T08:31:32Z

☔ The latest upstream changes (presumably #153114) made this pull request unmergeable. Please resolve the merge conflicts.

matthiaskrgr · 2026-03-01T06:21:10Z

Successful merges:

Suggest async block instead of async closure when possible #152042 (Suggest async block instead of async closure when possible)
Introduce --ci flag in tidy #152949 (Introduce --ci flag in tidy)
Various small query cleanups #153169 (Various small query cleanups)
Disable debug_assert_not_in_new_nodes for multiple threads #152655 (Disable debug_assert_not_in_new_nodes for multiple threads)
rustfmt: add test for field representing type builtin syntax #153229 (rustfmt: add test for field representing type builtin syntax)

Failed merges:

Clean up QueryVTable::hash_result into hash_value_fn #153209 (Clean up QueryVTable::hash_result into hash_value_fn)

r? @ghost

Create a similar rollup

* add --ci flag in tidy This commit introduces --ci flag in tidy because currently bootstrap can't pass its ci env information to tidy. It also modifies how CiInfo initialize its ci_env variable. tidy codes which uses CiEnv::is_ci for checking ci are now using ci_env in CiInfo. * address review - Fix comment - Use Option for ci flag in order to have true/false explicitly or unspecified (implicit false) * integrate CiInfo into TidyCtx * remove CiInfo * CiEnv::current() should be called when ci flag is not added * extract base_commit() to a separate function * use &TidyCtx instead of clone

It's unused.

It has a single use.

This makes it clear the `CycleError` is not used after the call.

When there are two cases of equal size and importance, I find an if/else expression easier to read than an early return.

It reads better that way.

It's an unnecessary level of indirection.

Calling `abort_if_errors` after emitting an error is guaranteed to call `raise_fatal`, so just do that directly instead.

This exact comment block also appears in `compiler/rustc_middle/src/queries.rs`, which is a better place for it.

`Value` is an unhelpfully generic name. Standard naming procedure for a trait with a single method is for the trait name to match the method name, which is what this commit does. Likewise, the enclosing module is renamed from `values` to `from_cycle_error`. Also add a comment about some non-obvious behaviour.

It has no effect. `symbol_name` is the only query that produces a `SymbolName`. If it was marked with `cycle_delayed_bug`/`cycle_stash` then this `FromCycleError` impl would make sense, but that's not the case. Maybe it was the case in the past.

@estebank

…-closure-suggestion, r=JohnTitor Suggest async block instead of async closure when possible Fixesrust-lang#140265 r? @estebank

Introduce --ci flag in tidy ## Context Currently tidy doesn't have `--ci` flag. Even if bootstrap has --ci=true, it isn't passed to tidy. As a result, some parts of tidy where we need to turn on `ci mode` are never executed on our local. ## Change This PR introduces --ci flag in tidy, and modify CiInfo and other parts of tidy where currently use `CiEnv::is_ci()` for checking if the running environment is ci or not.

@oli-obk

…ochenkov Various small query cleanups I found these while doing a close read of the query code. r? @oli-obk

…51509, r=nnethercote Disable debug_assert_not_in_new_nodes for multiple threads Fixesrust-lang#152654 This debug assertion is no longer intended to succeed for the parallel compiler since rust-lang#151509 doesn't panic on a red-green query graph coloring race, generating both red and green nodes (rust-lang#151509 (comment))

@ytmimi

rustfmt: add test for field representing type builtin syntax As requested by @ytmimi in rust-lang#152730 (comment). Not sure if you want/need a more complex example. Just let me know.

matthiaskrgr · 2026-03-01T06:21:27Z

@bors r+ rollup=never p=5

rust-bors · 2026-03-01T06:21:30Z

📌 Commit 8e1ec75 has been approved by matthiaskrgr

It is now in the queue for this repository.

rust-bors · 2026-03-01T08:31:30Z

PR #153169, which is a member of this rollup, was unapproved.
This rollup was thus also unapproved.

rust-bors · 2026-03-01T08:31:32Z

☔ The latest upstream changes (presumably #153114) made this pull request unmergeable. Please resolve the merge conflicts.

matthiaskrgr · 2026-03-01T06:21:10Z

Successful merges:

Suggest async block instead of async closure when possible #152042 (Suggest async block instead of async closure when possible)
Introduce --ci flag in tidy #152949 (Introduce --ci flag in tidy)
Various small query cleanups #153169 (Various small query cleanups)
Disable debug_assert_not_in_new_nodes for multiple threads #152655 (Disable debug_assert_not_in_new_nodes for multiple threads)
rustfmt: add test for field representing type builtin syntax #153229 (rustfmt: add test for field representing type builtin syntax)

Failed merges:

Clean up QueryVTable::hash_result into hash_value_fn #153209 (Clean up QueryVTable::hash_result into hash_value_fn)

r? @ghost

Create a similar rollup

* add --ci flag in tidy This commit introduces --ci flag in tidy because currently bootstrap can't pass its ci env information to tidy. It also modifies how CiInfo initialize its ci_env variable. tidy codes which uses CiEnv::is_ci for checking ci are now using ci_env in CiInfo. * address review - Fix comment - Use Option for ci flag in order to have true/false explicitly or unspecified (implicit false) * integrate CiInfo into TidyCtx * remove CiInfo * CiEnv::current() should be called when ci flag is not added * extract base_commit() to a separate function * use &TidyCtx instead of clone

It's unused.

It has a single use.

This makes it clear the `CycleError` is not used after the call.

When there are two cases of equal size and importance, I find an if/else expression easier to read than an early return.

It reads better that way.

It's an unnecessary level of indirection.

Calling `abort_if_errors` after emitting an error is guaranteed to call `raise_fatal`, so just do that directly instead.

This exact comment block also appears in `compiler/rustc_middle/src/queries.rs`, which is a better place for it.

`Value` is an unhelpfully generic name. Standard naming procedure for a trait with a single method is for the trait name to match the method name, which is what this commit does. Likewise, the enclosing module is renamed from `values` to `from_cycle_error`. Also add a comment about some non-obvious behaviour.

It has no effect. `symbol_name` is the only query that produces a `SymbolName`. If it was marked with `cycle_delayed_bug`/`cycle_stash` then this `FromCycleError` impl would make sense, but that's not the case. Maybe it was the case in the past.

@estebank

…-closure-suggestion, r=JohnTitor Suggest async block instead of async closure when possible Fixesrust-lang#140265 r? @estebank

Introduce --ci flag in tidy ## Context Currently tidy doesn't have `--ci` flag. Even if bootstrap has --ci=true, it isn't passed to tidy. As a result, some parts of tidy where we need to turn on `ci mode` are never executed on our local. ## Change This PR introduces --ci flag in tidy, and modify CiInfo and other parts of tidy where currently use `CiEnv::is_ci()` for checking if the running environment is ci or not.

@oli-obk

…ochenkov Various small query cleanups I found these while doing a close read of the query code. r? @oli-obk

…51509, r=nnethercote Disable debug_assert_not_in_new_nodes for multiple threads Fixesrust-lang#152654 This debug assertion is no longer intended to succeed for the parallel compiler since rust-lang#151509 doesn't panic on a red-green query graph coloring race, generating both red and green nodes (rust-lang#151509 (comment))

@ytmimi

rustfmt: add test for field representing type builtin syntax As requested by @ytmimi in rust-lang#152730 (comment). Not sure if you want/need a more complex example. Just let me know.

matthiaskrgr · 2026-03-01T06:21:27Z

@bors r+ rollup=never p=5

rust-bors · 2026-03-01T06:21:30Z

📌 Commit 8e1ec75 has been approved by matthiaskrgr

It is now in the queue for this repository.

rust-bors · 2026-03-01T08:31:30Z

PR #153169, which is a member of this rollup, was unapproved.
This rollup was thus also unapproved.

rust-bors · 2026-03-01T08:31:32Z

☔ The latest upstream changes (presumably #153114) made this pull request unmergeable. Please resolve the merge conflicts.

matthiaskrgr · 2026-03-01T06:21:10Z

Successful merges:

Suggest async block instead of async closure when possible #152042 (Suggest async block instead of async closure when possible)
Introduce --ci flag in tidy #152949 (Introduce --ci flag in tidy)
Various small query cleanups #153169 (Various small query cleanups)
Disable debug_assert_not_in_new_nodes for multiple threads #152655 (Disable debug_assert_not_in_new_nodes for multiple threads)
rustfmt: add test for field representing type builtin syntax #153229 (rustfmt: add test for field representing type builtin syntax)

Failed merges:

Clean up QueryVTable::hash_result into hash_value_fn #153209 (Clean up QueryVTable::hash_result into hash_value_fn)

r? @ghost

Create a similar rollup

* add --ci flag in tidy This commit introduces --ci flag in tidy because currently bootstrap can't pass its ci env information to tidy. It also modifies how CiInfo initialize its ci_env variable. tidy codes which uses CiEnv::is_ci for checking ci are now using ci_env in CiInfo. * address review - Fix comment - Use Option for ci flag in order to have true/false explicitly or unspecified (implicit false) * integrate CiInfo into TidyCtx * remove CiInfo * CiEnv::current() should be called when ci flag is not added * extract base_commit() to a separate function * use &TidyCtx instead of clone

It's unused.

It has a single use.

This makes it clear the `CycleError` is not used after the call.

When there are two cases of equal size and importance, I find an if/else expression easier to read than an early return.

It reads better that way.

It's an unnecessary level of indirection.

Calling `abort_if_errors` after emitting an error is guaranteed to call `raise_fatal`, so just do that directly instead.

This exact comment block also appears in `compiler/rustc_middle/src/queries.rs`, which is a better place for it.

`Value` is an unhelpfully generic name. Standard naming procedure for a trait with a single method is for the trait name to match the method name, which is what this commit does. Likewise, the enclosing module is renamed from `values` to `from_cycle_error`. Also add a comment about some non-obvious behaviour.

It has no effect. `symbol_name` is the only query that produces a `SymbolName`. If it was marked with `cycle_delayed_bug`/`cycle_stash` then this `FromCycleError` impl would make sense, but that's not the case. Maybe it was the case in the past.

@estebank

…-closure-suggestion, r=JohnTitor Suggest async block instead of async closure when possible Fixesrust-lang#140265 r? @estebank

Introduce --ci flag in tidy ## Context Currently tidy doesn't have `--ci` flag. Even if bootstrap has --ci=true, it isn't passed to tidy. As a result, some parts of tidy where we need to turn on `ci mode` are never executed on our local. ## Change This PR introduces --ci flag in tidy, and modify CiInfo and other parts of tidy where currently use `CiEnv::is_ci()` for checking if the running environment is ci or not.

@oli-obk

…ochenkov Various small query cleanups I found these while doing a close read of the query code. r? @oli-obk

…51509, r=nnethercote Disable debug_assert_not_in_new_nodes for multiple threads Fixesrust-lang#152654 This debug assertion is no longer intended to succeed for the parallel compiler since rust-lang#151509 doesn't panic on a red-green query graph coloring race, generating both red and green nodes (rust-lang#151509 (comment))

@ytmimi

rustfmt: add test for field representing type builtin syntax As requested by @ytmimi in rust-lang#152730 (comment). Not sure if you want/need a more complex example. Just let me know.

matthiaskrgr · 2026-03-01T06:21:27Z

@bors r+ rollup=never p=5

rust-bors · 2026-03-01T06:21:30Z

📌 Commit 8e1ec75 has been approved by matthiaskrgr

It is now in the queue for this repository.

rust-bors · 2026-03-01T08:31:30Z

PR #153169, which is a member of this rollup, was unapproved.
This rollup was thus also unapproved.

rust-bors · 2026-03-01T08:31:32Z

☔ The latest upstream changes (presumably #153114) made this pull request unmergeable. Please resolve the merge conflicts.

matthiaskrgr · 2026-03-01T06:21:10Z

Successful merges:

Suggest async block instead of async closure when possible #152042 (Suggest async block instead of async closure when possible)
Introduce --ci flag in tidy #152949 (Introduce --ci flag in tidy)
Various small query cleanups #153169 (Various small query cleanups)
Disable debug_assert_not_in_new_nodes for multiple threads #152655 (Disable debug_assert_not_in_new_nodes for multiple threads)
rustfmt: add test for field representing type builtin syntax #153229 (rustfmt: add test for field representing type builtin syntax)

Failed merges:

Clean up QueryVTable::hash_result into hash_value_fn #153209 (Clean up QueryVTable::hash_result into hash_value_fn)

r? @ghost

Create a similar rollup

* add --ci flag in tidy This commit introduces --ci flag in tidy because currently bootstrap can't pass its ci env information to tidy. It also modifies how CiInfo initialize its ci_env variable. tidy codes which uses CiEnv::is_ci for checking ci are now using ci_env in CiInfo. * address review - Fix comment - Use Option for ci flag in order to have true/false explicitly or unspecified (implicit false) * integrate CiInfo into TidyCtx * remove CiInfo * CiEnv::current() should be called when ci flag is not added * extract base_commit() to a separate function * use &TidyCtx instead of clone

It's unused.

It has a single use.

This makes it clear the `CycleError` is not used after the call.

When there are two cases of equal size and importance, I find an if/else expression easier to read than an early return.

It reads better that way.

It's an unnecessary level of indirection.

Calling `abort_if_errors` after emitting an error is guaranteed to call `raise_fatal`, so just do that directly instead.

This exact comment block also appears in `compiler/rustc_middle/src/queries.rs`, which is a better place for it.

`Value` is an unhelpfully generic name. Standard naming procedure for a trait with a single method is for the trait name to match the method name, which is what this commit does. Likewise, the enclosing module is renamed from `values` to `from_cycle_error`. Also add a comment about some non-obvious behaviour.

It has no effect. `symbol_name` is the only query that produces a `SymbolName`. If it was marked with `cycle_delayed_bug`/`cycle_stash` then this `FromCycleError` impl would make sense, but that's not the case. Maybe it was the case in the past.

@estebank

…-closure-suggestion, r=JohnTitor Suggest async block instead of async closure when possible Fixesrust-lang#140265 r? @estebank

Introduce --ci flag in tidy ## Context Currently tidy doesn't have `--ci` flag. Even if bootstrap has --ci=true, it isn't passed to tidy. As a result, some parts of tidy where we need to turn on `ci mode` are never executed on our local. ## Change This PR introduces --ci flag in tidy, and modify CiInfo and other parts of tidy where currently use `CiEnv::is_ci()` for checking if the running environment is ci or not.

@oli-obk

…ochenkov Various small query cleanups I found these while doing a close read of the query code. r? @oli-obk

…51509, r=nnethercote Disable debug_assert_not_in_new_nodes for multiple threads Fixesrust-lang#152654 This debug assertion is no longer intended to succeed for the parallel compiler since rust-lang#151509 doesn't panic on a red-green query graph coloring race, generating both red and green nodes (rust-lang#151509 (comment))

@ytmimi

rustfmt: add test for field representing type builtin syntax As requested by @ytmimi in rust-lang#152730 (comment). Not sure if you want/need a more complex example. Just let me know.

matthiaskrgr · 2026-03-01T06:21:27Z

@bors r+ rollup=never p=5

rust-bors · 2026-03-01T06:21:30Z

📌 Commit 8e1ec75 has been approved by matthiaskrgr

It is now in the queue for this repository.

rust-bors · 2026-03-01T08:31:30Z

PR #153169, which is a member of this rollup, was unapproved.
This rollup was thus also unapproved.

rust-bors · 2026-03-01T08:31:32Z

☔ The latest upstream changes (presumably #153114) made this pull request unmergeable. Please resolve the merge conflicts.

matthiaskrgr · 2026-03-01T06:21:10Z

Successful merges:

Suggest async block instead of async closure when possible #152042 (Suggest async block instead of async closure when possible)
Introduce --ci flag in tidy #152949 (Introduce --ci flag in tidy)
Various small query cleanups #153169 (Various small query cleanups)
Disable debug_assert_not_in_new_nodes for multiple threads #152655 (Disable debug_assert_not_in_new_nodes for multiple threads)
rustfmt: add test for field representing type builtin syntax #153229 (rustfmt: add test for field representing type builtin syntax)

Failed merges:

Clean up QueryVTable::hash_result into hash_value_fn #153209 (Clean up QueryVTable::hash_result into hash_value_fn)

r? @ghost

Create a similar rollup

* add --ci flag in tidy This commit introduces --ci flag in tidy because currently bootstrap can't pass its ci env information to tidy. It also modifies how CiInfo initialize its ci_env variable. tidy codes which uses CiEnv::is_ci for checking ci are now using ci_env in CiInfo. * address review - Fix comment - Use Option for ci flag in order to have true/false explicitly or unspecified (implicit false) * integrate CiInfo into TidyCtx * remove CiInfo * CiEnv::current() should be called when ci flag is not added * extract base_commit() to a separate function * use &TidyCtx instead of clone

It's unused.

It has a single use.

This makes it clear the `CycleError` is not used after the call.

When there are two cases of equal size and importance, I find an if/else expression easier to read than an early return.

It reads better that way.

It's an unnecessary level of indirection.

Calling `abort_if_errors` after emitting an error is guaranteed to call `raise_fatal`, so just do that directly instead.

This exact comment block also appears in `compiler/rustc_middle/src/queries.rs`, which is a better place for it.

`Value` is an unhelpfully generic name. Standard naming procedure for a trait with a single method is for the trait name to match the method name, which is what this commit does. Likewise, the enclosing module is renamed from `values` to `from_cycle_error`. Also add a comment about some non-obvious behaviour.

It has no effect. `symbol_name` is the only query that produces a `SymbolName`. If it was marked with `cycle_delayed_bug`/`cycle_stash` then this `FromCycleError` impl would make sense, but that's not the case. Maybe it was the case in the past.

@estebank

…-closure-suggestion, r=JohnTitor Suggest async block instead of async closure when possible Fixesrust-lang#140265 r? @estebank

Introduce --ci flag in tidy ## Context Currently tidy doesn't have `--ci` flag. Even if bootstrap has --ci=true, it isn't passed to tidy. As a result, some parts of tidy where we need to turn on `ci mode` are never executed on our local. ## Change This PR introduces --ci flag in tidy, and modify CiInfo and other parts of tidy where currently use `CiEnv::is_ci()` for checking if the running environment is ci or not.

@oli-obk

…ochenkov Various small query cleanups I found these while doing a close read of the query code. r? @oli-obk

…51509, r=nnethercote Disable debug_assert_not_in_new_nodes for multiple threads Fixesrust-lang#152654 This debug assertion is no longer intended to succeed for the parallel compiler since rust-lang#151509 doesn't panic on a red-green query graph coloring race, generating both red and green nodes (rust-lang#151509 (comment))

@ytmimi

rustfmt: add test for field representing type builtin syntax As requested by @ytmimi in rust-lang#152730 (comment). Not sure if you want/need a more complex example. Just let me know.

matthiaskrgr · 2026-03-01T06:21:27Z

@bors r+ rollup=never p=5

rust-bors · 2026-03-01T06:21:30Z

📌 Commit 8e1ec75 has been approved by matthiaskrgr

It is now in the queue for this repository.

rust-bors · 2026-03-01T08:31:30Z

PR #153169, which is a member of this rollup, was unapproved.
This rollup was thus also unapproved.

rust-bors · 2026-03-01T08:31:32Z

☔ The latest upstream changes (presumably #153114) made this pull request unmergeable. Please resolve the merge conflicts.

zetanumbersand others added 20 commits February 15, 2026 14:41

Disable debug_assert_not_in_new_nodes for multiple threads
9fb9225

Suggest async block instead of async closure when possible
448097d

Fix async closure suggestion when no space between || and {
3736458

rustfmt: add test for field representing type builtin syntax
d41e16b

Remove DepKindVTable::is_anon.
d117486
It's unused.

Inline and remove handle_cycle_error.
f1789fa
It has a single use.

Make from_cycle_error consume the CycleError.
adf2d1e
This makes it clear the `CycleError` is not used after the call.

Avoid an early return in try_execute_query.
86e1af9
When there are two cases of equal size and importance, I find an if/else expression easier to read than an early return.

Merge two assertion blocks in execute_job_non_incr.
f1c39fe
It reads better that way.

Remove three single-use type synonyms.
e7dfd53
It's an unnecessary level of indirection.

Replace two abort_if_errors calls.
2322cca
Calling `abort_if_errors` after emitting an error is guaranteed to call `raise_fatal`, so just do that directly instead.

Remove a duplicated comment.
19b3617
This exact comment block also appears in `compiler/rustc_middle/src/queries.rs`, which is a better place for it.

Remove FromCycleError impl for SymbolName.
fa5138b
It has no effect. `symbol_name` is the only query that produces a `SymbolName`. If it was marked with `cycle_delayed_bug`/`cycle_stash` then this `FromCycleError` impl would make sense, but that's not the case. Maybe it was the case in the past.

Rollup merge ofrust-lang#152042- chenyukang:yukang-fix-140265-async…
8654366
…-closure-suggestion, r=JohnTitor Suggest async block instead of async closure when possible Fixesrust-lang#140265 r? @estebank

Rollup merge ofrust-lang#153169- nnethercote:rm-is_anon-etc, r=petr…
ca2655c
…ochenkov Various small query cleanups I found these while doing a close read of the query code. r? @oli-obk

Rollup merge ofrust-lang#153229- BennoLossin:frt-fmt, r=ytmimi
8e1ec75
rustfmt: add test for field representing type builtin syntax As requested by @ytmimi in rust-lang#152730 (comment). Not sure if you want/need a more complex example. Just let me know.

rust-borsBot added the rollup A PR which is a rollup label Mar 1, 2026

matthiaskrgr closed this Mar 1, 2026

rustbot removed the S-waiting-on-author Status: This is awaiting some action (such as code changes or more information) from the author. label Mar 1, 2026