Get rid of the hir_owner query.

[cjgillot]

This query was meant as a firewall between hir_owner_nodes which is supposed to change often, and the queries that only depend on the item signature. That firewall was inefficient, leaking the contents of the HIR body through HirIds.

hir_owner incurs a significant cost, as we need to hash HIR twice in multiple modes. This PR proposes to remove it, and simplify the hashing scheme.

For the future, def_kind, def_span... are much more efficient for incremental decoupling, and should be preferred.

[cjgillot]

@bors try @rust-timer queue

[bors]

⌛ Trying commit d63eab3 with merge a14c977...

[bors]

☀️ Try build successful - checks-actions
Build commit: a14c977 (a14c9771908861d1d12466d4089bcc3d8d189d56)

[rust-timer]

Finished benchmarking commit (a14c977): comparison URL.

Overall result: ✅ improvements - 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-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
Regressions ❌ (secondary)	0.2%	[0.2%, 0.3%]	3
Improvements ✅ (primary)	-1.8%	[-5.4%, -0.2%]	127
Improvements ✅ (secondary)	-2.2%	[-4.2%, -0.3%]	42
All ❌✅ (primary)	-1.8%	[-5.4%, -0.2%]	127

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)	2.0%	[2.0%, 2.0%]	1
Regressions ❌ (secondary)	2.8%	[1.5%, 5.4%]	6
Improvements ✅ (primary)	-2.6%	[-13.5%, -0.6%]	10
Improvements ✅ (secondary)	-4.9%	[-11.3%, -1.4%]	7
All ❌✅ (primary)	-2.2%	[-13.5%, 2.0%]	11

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)	1.6%	[1.2%, 1.9%]	5
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-1.9%	[-2.9%, -1.0%]	35
Improvements ✅ (secondary)	-2.1%	[-2.7%, -0.7%]	9
All ❌✅ (primary)	-1.5%	[-2.9%, 1.9%]	40

Binary size

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

Bootstrap: 667.719s -> 666.383s (-0.20%)
Artifact size: 308.26 MiB -> 308.17 MiB (-0.03%)

[cjgillot]

r? compiler

[wesleywiser]

These are really nice wins, both for the perf improvements and also removing an additional layer of complexity from the code!

Great work @cjgillot 🎉

@bors r+

[bors]

📌 Commit d59968b has been approved by wesleywiser

It is now in the queue for this repository.

[bors]

⌛ Testing commit d59968b with merge 685da38...

[bors]

💥 Test timed out

[rust-log-analyzer]

A job failed! Check out the build log: (web) (plain)

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

[cjgillot]

@bors retry

[bors]

⌛ Testing commit d59968b with merge d3c9082...

[bors]

☀️ Test successful - checks-actions
Approved by: wesleywiser
Pushing d3c9082 to master...

[rust-timer]

Finished benchmarking commit (d3c9082): comparison URL.

Overall result: ✅ improvements - 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)	0.3%	[0.2%, 0.4%]	3
Improvements ✅ (primary)	-1.8%	[-5.5%, -0.3%]	129
Improvements ✅ (secondary)	-2.1%	[-4.4%, -0.4%]	46
All ❌✅ (primary)	-1.8%	[-5.5%, -0.3%]	129

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)	1.3%	[1.3%, 1.3%]	1
Regressions ❌ (secondary)	1.2%	[1.2%, 1.2%]	1
Improvements ✅ (primary)	-4.5%	[-13.7%, -0.6%]	6
Improvements ✅ (secondary)	-5.5%	[-10.8%, -1.9%]	6
All ❌✅ (primary)	-3.7%	[-13.7%, 1.3%]	7

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)	1.5%	[1.5%, 1.5%]	1
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-2.2%	[-3.5%, -1.1%]	34
Improvements ✅ (secondary)	-3.3%	[-4.4%, -2.0%]	10
All ❌✅ (primary)	-2.1%	[-3.5%, 1.5%]	35

Binary size

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)	-0.5%	[-0.5%, -0.5%]	1
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-0.5%	[-0.5%, -0.5%]	1

Bootstrap: 666.403s -> 665.653s (-0.11%)
Artifact size: 308.34 MiB -> 308.28 MiB (-0.02%)

[nnethercote]

The icount and cycle results are fantastic, but the wall-time results are pretty mixed. Looking at the graphs a lot of the wall-time regressions look like they might be temporary blips, though.