Do not register placeholder RegionOutlives obligations when considering_regions is false

[compiler-errors]

NOTE: I'm kinda just putting this up for discussion. I'm not certain this is correct...?

This was introduced in 608625d.

Interestingly, we only check data.has_placeholders() for RegionOutlives, and not for TypeOutlives... why? For the record, that different treatment between RegionOutlives and TypeOutlives is why the fix "The compiling succeeds when all 'a : 'b are replaced with &'a () : 'b" in #100689 "works", but it seems like an implementation detail considering this.

Also, why do we care about placeholder regions being registered if considering_regions is false? It doesn't seem to affect any UI tests, for example.

r? @lcnr

Fixes #102899
Fixes #100689

[matthiaskrgr]

Should we eventually beta-nominate the fix as well? The ice-code doesn't use any nightly only features as far as I seen and I was able to reproduce #102899 on beta too.

[lcnr]

i think i added the special-case for placeholders because of the leak-check for which we have to care about placeholders, even when not doing regionck. Even when rewriting rustc to replace the leak-check, we will still have to care about placeholders.

However, fulfill currently doesn't rely on the leak-check right now, because it's never used in any place where the leak-check is used.

And even if we did use it, currently evaluate uses the leak check but it doesn't add region constraints for Outlives obligations, only when directly relating regions ✨ so anyways. So even if we were to use fulfill in some leak-checked snapshot, this would still differ from the behavior of evaluate.

So yeah, the current setup sucks but your PR doesn't make stuff worse. cc @nikomatsakis for the above comment.

r=me
@bors try @rust-timer queue

[rust-timer]

Awaiting bors try build completion.

@rustbot label: +S-waiting-on-perf

[bors]

⌛ Trying commit 3021598 with merge 25fac99eb1c69846b65127999dd084d21bcfd8f8...

[bors]

☀️ Try build successful - checks-actions
Build commit: 25fac99eb1c69846b65127999dd084d21bcfd8f8 (25fac99eb1c69846b65127999dd084d21bcfd8f8)

[rust-timer]

Queued 25fac99eb1c69846b65127999dd084d21bcfd8f8 with parent 1536ab1, future comparison URL.

[rust-timer]

Finished benchmarking commit (25fac99eb1c69846b65127999dd084d21bcfd8f8): comparison URL.

Overall result: no relevant changes - no action needed

Benchmarking this pull request likely means that it is perf-sensitive, so we're automatically marking it as not fit for rolling up. While you can manually mark this PR as fit for rollup, we strongly recommend not doing so since this PR may lead to changes in compiler perf.

@bors rollup=never
@rustbot label: +S-waiting-on-review -S-waiting-on-perf -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.

	mean1	range	count2
Regressions ❌ (primary)	3.7%	[3.7%, 3.7%]	1
Regressions ❌ (secondary)	2.9%	[2.3%, 3.4%]	14
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	3.7%	[3.7%, 3.7%]	1

Cycles

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

Footnotes

1. the arithmetic mean of the percent change ↩

2. number of relevant changes ↩

[lcnr]

@bors r+ rollup=always

[bors]

📌 Commit 3021598 has been approved by lcnr

It is now in the queue for this repository.