Use FxIndexMap instead FxHashMap to stabilize iteration order in EffectiveVisibilities

[michaelwoerister]

Part of MCP 533.

[rustbot]

r? @cjgillot

rustbot has assigned @cjgillot.
They will have a look at your PR within the next two weeks and either review your PR or reassign to another reviewer.

Use r? to explicitly pick a reviewer

[michaelwoerister]

@bors try @rust-timer queue

[bors]

⌛ Trying commit a50490c with merge 20a1b6d...

[bors]

☀️ Try build successful - checks-actions
Build commit: 20a1b6d (20a1b6db223806bd9fa3126be395eb2e7396680f)

[rust-timer]

Finished benchmarking commit (20a1b6d): comparison URL.

Overall result: ❌✅ regressions and improvements - 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.

Next Steps: If you can justify the regressions found in this try perf run, please indicate this with @rustbot label: +perf-regression-triaged along with sufficient written justification. If you cannot justify the regressions please fix the regressions and do another perf run. If the next run shows neutral or positive results, the label will be automatically removed.

@bors rollup=never
@rustbot label: -S-waiting-on-perf +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)	0.2%	[0.1%, 0.3%]	5
Regressions ❌ (secondary)	0.5%	[0.4%, 0.6%]	5
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-0.1%	[-0.1%, -0.1%]	4
All ❌✅ (primary)	0.2%	[0.1%, 0.3%]	5

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)	8.7%	[5.7%, 11.0%]	5
Improvements ✅ (primary)	-2.3%	[-2.3%, -2.3%]	1
Improvements ✅ (secondary)	-1.6%	[-1.8%, -1.4%]	2
All ❌✅ (primary)	-2.3%	[-2.3%, -2.3%]	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)	-	-	0
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-1.1%	[-1.7%, -0.6%]	2
Improvements ✅ (secondary)	-4.1%	[-5.7%, -2.1%]	6
All ❌✅ (primary)	-1.1%	[-1.7%, -0.6%]	2

Binary size

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

Bootstrap: 651.828s -> 653.326s (0.23%)
Artifact size: 311.10 MiB -> 311.15 MiB (0.02%)

[michaelwoerister]

I'm going to run a callgrind diff on the libc-check-full case to if the small regression is real.

[michaelwoerister]

set_bindings_effective_visibilities (which uses the map in question) does show up with an instruction count increase for the libc and unused-warnings benchmarks. The code seems to do a bit more work with FxIndexMap instead of FxHashMap. However, I don't see how that can be changed without a bigger refactoring of code that otherwise does not look like it's in need of refactoring. Using UnordMap instead of FxIndexMap is non-trivial too, since EffectiveVisibilities public exposes an iter method that gives access to internal map.

I think the small regression is acceptable, since this is a correctness/incr-comp-hardening fix.

[michaelwoerister]

@rustbot label: +perf-regression-triaged
(please let me know if there's disagreement on that)

[cjgillot]

cc @petrochenkov as this touches effective visibilities.

[michaelwoerister]

I just noticed that this is now the last instance of a HashMap occurring in a query result, so I pushed 7e4bc4a which removes and disallows the HashStable impl for HashMap 🎉

[petrochenkov]

Effective visibility changes look ok.
unused-warnings and libc are stress tests for effective visibilities, so they will regress a bit if FxIndexMap is any slower than FxHashMap.

[cjgillot]

@bors r+

[bors]

📌 Commit 7e4bc4a has been approved by cjgillot

It is now in the queue for this repository.

[bors]

⌛ Testing commit 7e4bc4a with merge 1e98e80...

[rust-log-analyzer]

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

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

[bors]

💔 Test failed - checks-actions

[cjgillot]

@bors retry

[klensy]

Gh gone crazy? Few first posts dated 1 mar, but written as "commented now".

[bors]

⌛ Testing commit 7e4bc4a with merge bf71dae...

[bors]

☀️ Test successful - checks-actions
Approved by: cjgillot
Pushing bf71dae to master...

[rust-timer]

Finished benchmarking commit (bf71dae): comparison URL.

Overall result: ❌ regressions - ACTION NEEDED

Next Steps: If you can justify the regressions found in this perf run, please indicate this with @rustbot label: +perf-regression-triaged along with sufficient written justification. If you cannot justify the regressions please open an issue or create a new PR that fixes the regressions, add a comment linking to the newly created issue or PR, and then add the perf-regression-triaged label to this PR.

@rustbot label: +perf-regression
cc @rust-lang/wg-compiler-performance

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)	0.3%	[0.2%, 0.4%]	13
Regressions ❌ (secondary)	0.4%	[0.3%, 0.4%]	3
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	0.3%	[0.2%, 0.4%]	13

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)	1.5%	[1.5%, 1.5%]	1
Regressions ❌ (secondary)	6.8%	[2.0%, 10.2%]	4
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	1.5%	[1.5%, 1.5%]	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)	-	-	0
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-1.8%	[-1.9%, -1.7%]	3
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-1.8%	[-1.9%, -1.7%]	3

Binary size

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

Bootstrap: 669.417s -> 666.819s (-0.39%)
Artifact size: 315.68 MiB -> 315.68 MiB (0.00%)

[Mark-Simulacrum]

See #121851 (comment) for justification.