Run a single huge par_body_owners instead of many small ones after each other.

[oli-obk]

This improves parallel rustc parallelism by avoiding the bottleneck after each individual par_body_owners (because it needs to wait for queries to finish, so if there is one long running one, a lot of cores will be idle while waiting for the single query).

[rustbot]

r? @davidtwco

rustbot has assigned @davidtwco.
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

[oli-obk]

@bors try @rust-timer queue

[bors]

⌛ Trying commit 164473b with merge c0e49b5...

[bors]

☀️ Try build successful - checks-actions
Build commit: c0e49b5 (c0e49b5e16f7b562786f83d7702b40f242942aee)

[rust-timer]

Finished benchmarking commit (c0e49b5): 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%	[0.2%, 0.3%]	3
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-0.5%	[-0.8%, -0.2%]	5
Improvements ✅ (secondary)	-0.3%	[-0.5%, -0.2%]	12
All ❌✅ (primary)	-0.2%	[-0.8%, 0.3%]	8

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)	2.4%	[2.4%, 2.4%]	1
Improvements ✅ (primary)	-1.9%	[-2.5%, -0.9%]	8
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-1.9%	[-2.5%, -0.9%]	8

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)	4.3%	[1.1%, 14.8%]	65
Regressions ❌ (secondary)	12.3%	[4.8%, 18.6%]	18
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	4.3%	[1.1%, 14.8%]	65

Binary size

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

Bootstrap: 648.501s -> 652.39s (0.60%)
Artifact size: 172.69 MiB -> 172.63 MiB (-0.03%)

[oli-obk]

@bors try @rust-timer queue

[bors]

⌛ Trying commit c531269 with merge 702bd6d...

[bors]

☀️ Try build successful - checks-actions
Build commit: 702bd6d (702bd6de43d33c2a791c1a2f6acccd68f55d3afd)

[rust-timer]

Finished benchmarking commit (702bd6d): 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)	0.3%	[0.3%, 0.3%]	2
Regressions ❌ (secondary)	1.7%	[1.7%, 1.7%]	1
Improvements ✅ (primary)	-0.3%	[-0.3%, -0.3%]	1
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	0.1%	[-0.3%, 0.3%]	3

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)	1.1%	[1.1%, 1.1%]	1
Regressions ❌ (secondary)	5.7%	[5.1%, 6.2%]	2
Improvements ✅ (primary)	-1.7%	[-2.4%, -0.7%]	5
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-1.2%	[-2.4%, 1.1%]	6

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)	3.9%	[1.1%, 13.7%]	71
Regressions ❌ (secondary)	11.7%	[2.9%, 20.6%]	20
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	3.9%	[1.1%, 13.7%]	71

Binary size

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

Bootstrap: 648.792s -> 653.885s (0.78%)
Artifact size: 172.56 MiB -> 172.55 MiB (-0.01%)

[davidtwco]

r=me if this is ready

[oli-obk]

@bors r=davidtwco

[bors]

📌 Commit 55ea944 has been approved by davidtwco

It is now in the queue for this repository.

[bors]

⌛ Testing commit 55ea944 with merge 65cd843...

[bors]

☀️ Test successful - checks-actions
Approved by: davidtwco
Pushing 65cd843 to master...

[rust-timer]

Finished benchmarking commit (65cd843): comparison URL.

Overall result: ❌✅ regressions and 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)	1.7%	[1.7%, 1.7%]	1
Improvements ✅ (primary)	-0.4%	[-0.6%, -0.3%]	2
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-0.4%	[-0.6%, -0.3%]	2

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)	-	-	0
Improvements ✅ (primary)	-1.8%	[-2.3%, -0.9%]	4
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-1.8%	[-2.3%, -0.9%]	4

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)	4.2%	[1.3%, 13.8%]	68
Regressions ❌ (secondary)	12.0%	[2.1%, 20.9%]	19
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-5.9%	[-9.0%, -3.6%]	8
All ❌✅ (primary)	4.2%	[1.3%, 13.8%]	68

Binary size

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

Bootstrap: 647.596s -> 652.28s (0.72%)
Artifact size: 310.01 MiB -> 309.94 MiB (-0.02%)

[Kobzol]

Umm, did this help for the parallel frontend (and how much?). Otherwise, this looks like a severe regression for the single-threaded version.

[bjorn3]

(for those that got confused, the linked perf report is for the cycles count rather than the instructions count)

[Kobzol]

Ah yes, sorry, forgot to mention that. The same regression is for walltime.

[oli-obk]

That's super weird. Did I trash CPU caching of the query tables or sth? (Edit: Jup, I did https://perf.rust-lang.org/compare.html?start=d255c6a57c393db6221b1ff700daea478436f1cd&end=65cd843ae06ad00123c131a431ed5304e4cd577a&stat=cache-misses) Before this PR we invoked all queries of the same kind right after each other, now we invoke all queries for the same id together.

Let's revert and revisit.