Prevent using #[target_feature] on lang item functions

[eduardosm]

Fixes #109411 and also prevents from using #[target_feature] on other fn lang items to mitigate the concerns from #109411 (comment).

[rustbot]

r? @wesleywiser

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

[cjgillot]

This check should probably happen in rustc_passes::check_attr, which is meant to verify that attributes are well-formed.

[eduardosm]

This check should probably happen in rustc_passes::check_attr, which is meant to verify that attributes are well-formed.

I'm not so sure. Moving the check to check_attr would require figuring there both if the item is a lang item and if it is has target features. In collect_item_extended we already know we are visiting a lang item (and which one in particular), so it only needs to check for the target_feature attribute. Maybe it could be improved by using codegen_fn_attrs to avoid looping over the attributes.

In general, I've seen that check_attr are more like syntactic checks, e.g. that attributes are used on the correct target kind.

[cjgillot]

What about just check for a lang attribute?

[eduardosm]

What about just check for a lang attribute?

I'll try that with hir::lang_items::extract, which will also take care of #[panic_handler].

[cjgillot]

Thanks!
@bors r+

[bors]

📌 Commit 9102816 has been approved by cjgillot

It is now in the queue for this repository.

[bors]

⌛ Testing commit 9102816 with merge aace2df...

[bors]

☀️ Test successful - checks-actions
Approved by: cjgillot
Pushing aace2df to master...

[rust-timer]

Finished benchmarking commit (aace2df): comparison URL.

Overall result: ❌ regressions - no action needed

@rustbot label: -perf-regression

Instruction count

This is a highly reliable metric that was used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	0.7%	[0.7%, 0.7%]	1
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-	-	0

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)	-	-	0
Regressions ❌ (secondary)	1.4%	[0.8%, 2.0%]	2
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-	-	0

Cycles

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

Binary size

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

Bootstrap: 631.519s -> 632.249s (0.12%)
Artifact size: 317.47 MiB -> 317.50 MiB (0.01%)