Optimize DefaultHasher siphash

[Noratrieb]

let's see how re-applying #69152 again goes.
imo this is a huge speedup that would be worth some compile time regressions, but i wanna see first. probably won't have the time and energy to argue for it though, if there are significant regressions ^^'

cc @nnethercote

[Noratrieb]

@bors try @rust-timer queue

[rustbot]

r? @cuviper

rustbot has assigned @cuviper.
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

[bors]

⌛ Trying commit 1bb82de with merge b6cd5d6...

[bors]

☀️ Try build successful - checks-actions
Build commit: b6cd5d6 (b6cd5d61b5c53b0513dac3cb0ae01e4742c328ef)

[rust-timer]

Finished benchmarking commit (b6cd5d6): 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)	1.1%	[0.3%, 3.3%]	10
Regressions ❌ (secondary)	2.8%	[0.3%, 7.1%]	4
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-0.5%	[-0.6%, -0.4%]	2
All ❌✅ (primary)	1.1%	[0.3%, 3.3%]	10

Max RSS (memory usage)

Results (primary 1.4%, secondary 5.6%)

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)	5.5%	[4.5%, 6.5%]	2
Regressions ❌ (secondary)	5.6%	[5.6%, 5.6%]	1
Improvements ✅ (primary)	-2.6%	[-3.6%, -1.5%]	2
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	1.4%	[-3.6%, 6.5%]	4

Cycles

Results (primary 1.3%, secondary 6.8%)

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%	[0.7%, 3.1%]	5
Regressions ❌ (secondary)	6.8%	[6.8%, 6.8%]	1
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	1.3%	[0.7%, 3.1%]	5

Binary size

Results (primary 0.1%)

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)	0.2%	[0.0%, 1.0%]	33
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-0.1%	[-0.2%, -0.0%]	21
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	0.1%	[-0.2%, 1.0%]	54

Bootstrap: 756.685s -> 757.535s (0.11%)
Artifact size: 341.18 MiB -> 341.13 MiB (-0.01%)

[nnethercote]

The perf CI results show the slowdown this causes in compile times, which aren't great, but also aren't terrible. But they don't provide any measurements about the benefits of this change, because rustc doesn't use SipHasher. This comment explains how rustc can be modified to use SipHasher. Having CI perf runs of "rustc"1 and "rustc2" in this PR would be useful.