Ensure that THIR unsafety check is done before stealing it

[Zoxc]

This ensures that THIR unsafety check is done before stealing it by running it on the typeck root instead of on a closure, which does nothing.

Fixes #111520

[rustbot]

r? @b-naber

(rustbot has picked a reviewer for you, use r? to override)

[compiler-errors]

Unrelated to this PR, but why is this ensure_with_value 🤔

[cjgillot]

ensure verifies that the dependency is green, without computing anything if not required.
This means that a query that is ensured may need to be computed later if some code needs the result value.

ensure_with_value verifies that the result is available, computed or loadable from on-disk cache.

[Zoxc]

Hm.. this means that query side effects can be emitted twice? Once when marking a query green and once when executing the query later to recover the value. I guess we need to make sure that's benign.

[cjgillot]

This is ok for diagnostics, as they are deduplicated. But this is a blocker for replaying any other side-effect. I've been trying to fix this to re-create DefId, but I'm not satisfied by my design.

[Zoxc]

There's an option to turn it off though. There's also extra logic for the parallel compiler to avoid emitting side effects multiple times, which is kind of unnecessary then.

How does DefId relate to this?

Representing side effects as forcible queries would avoid this issue, but does require work to support forcing arbitrary keys.

[cjgillot]

@bors r+

[bors]

📌 Commit 4170ca4 has been approved by cjgillot

It is now in the queue for this repository.

[bors]

⌛ Testing commit 4170ca4 with merge ba7e4ce028604a12acc8fc2632661f115083bff3...

[rust-log-analyzer]

The job aarch64-gnu failed! Check out the build log: (web) (plain)

Click to see the possible cause of the failure (guessed by this bot)

[RUSTC-TIMING] rustc_metadata test:false 62.101
   Compiling rustc_borrowck v0.0.0 (/checkout/compiler/rustc_borrowck)
[RUSTC-TIMING] rustc_traits test:false 15.571
   Compiling rustc_hir_typeck v0.1.0 (/checkout/compiler/rustc_hir_typeck)
##[error]The runner has received a shutdown signal. This can happen when the runner service is stopped, or a manually started runner is canceled.
##[group]Clock drift check
##[group]Clock drift check
Session terminated, killing shell... ...killed.
time="2023-08-23T00:02:03Z" level=error msg="error waiting for container: unexpected EOF"
##[error]The operation was canceled.
Cleaning up orphan processes

[bors]

💔 Test failed - checks-actions

[Zoxc]

Retry? Looks to be some CI issue?

[cjgillot]

@bors retry

[bors]

⌛ Testing commit 4170ca4 with merge c9db1f8...

[bors]

☀️ Test successful - checks-actions
Approved by: cjgillot
Pushing c9db1f8 to master...

[rust-timer]

Finished benchmarking commit (c9db1f8): comparison URL.

Overall result: no relevant changes - no action needed

@rustbot label: -perf-regression

Instruction count

This benchmark run did not return any relevant results for this metric.

Max RSS (memory usage)

Results

This is a less reliable metric that may be of interest but was not used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	5.8%	[5.8%, 5.8%]	1
Regressions ❌ (secondary)	3.9%	[3.0%, 4.7%]	2
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	5.8%	[5.8%, 5.8%]	1

Cycles

Results

This is a less reliable metric that may be of interest but was not used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	4.6%	[3.9%, 5.0%]	6
Regressions ❌ (secondary)	2.0%	[2.0%, 2.0%]	1
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	4.6%	[3.9%, 5.0%]	6

Binary size

This benchmark run did not return any relevant results for this metric.

Bootstrap: 635.876s -> 636.596s (0.11%)
Artifact size: 347.03 MiB -> 347.02 MiB (-0.00%)