Remap syntax contexts and expansion ids for det. metadata encoding

[aerooneqq]

View all comments

#163495 which parallelizes AST -> HIR lowering breaks det. encoding of syntax contexts and expansions, however this PR with #162809 should fix the problem.

Now draft for perf and CI (dont look at the code). Remapping of those entities can cost us much, because unlike def ids there may be many expansions and contexts after we committed end of determinism.

r? @petrochenkov

[petrochenkov]

@bors try @rust-timer queue

[petrochenkov]

@bors try cancel

[rust-bors]

Try build cancelled. Cancelled workflows:

• https://github.com/rust-lang/rust/actions/runs/37029803876

Hint: if you want to run another try build, you do not need to manually cancel the previous one. Just run @bors try and bors will cancel the previous build automatically.

[petrochenkov]

@bors try @rust-timer queue

[aerooneqq]

@bors try @rust-timer queue

[rust-bors]

☀️ Try build successful (CI)
Build commit: a7c6732 (a7c67326654d086b506b6f55d82af110f520b88b)
Base parent: 602727f (602727f26878dcf9eb2999e2f58e5df1269ff47e)

[rust-timer]

Finished benchmarking commit (a7c6732): comparison URL.

Overall result: ❌✅ regressions and improvements - please read:

Benchmarking means the PR may be perf-sensitive. It's automatically marked not fit for rolling up. Overriding is possible but disadvised: it risks changing compiler perf.

Next, please: If you can, justify the regressions found in this try perf run in writing along with @rustbot label: +perf-regression-triaged. If not, fix the regressions and do another perf run. Neutral or positive results will clear the label automatically.

@bors rollup=never rustc-perf
@rustbot label: -S-waiting-on-perf +perf-regression

Instruction count

Our most reliable metric. Used to determine the overall result above. However, even this metric can be noisy.

	mean	range	count
Regressions ❌ (primary)	0.7%	[0.3%, 1.2%]	7
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-0.2%	[-1.2%, -0.1%]	57
Improvements ✅ (secondary)	-0.3%	[-0.7%, -0.1%]	28
All ❌✅ (primary)	-0.1%	[-1.2%, 1.2%]	64

Max RSS (memory usage)

Results (primary -2.9%, secondary -1.7%)

A less reliable metric. May be of interest, but not used to determine the overall result above.

	mean	range	count
Regressions ❌ (primary)	3.1%	[3.1%, 3.1%]	1
Regressions ❌ (secondary)	2.6%	[1.6%, 6.6%]	9
Improvements ✅ (primary)	-3.0%	[-8.3%, -0.9%]	112
Improvements ✅ (secondary)	-3.1%	[-5.4%, -1.2%]	29
All ❌✅ (primary)	-2.9%	[-8.3%, 3.1%]	113

Cycles

Results (primary 11.8%, secondary 2.4%)

A less reliable metric. May be of interest, but not used to determine the overall result above.

	mean	range	count
Regressions ❌ (primary)	11.8%	[2.3%, 27.3%]	15
Regressions ❌ (secondary)	3.8%	[3.0%, 4.5%]	5
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-4.4%	[-4.4%, -4.4%]	1
All ❌✅ (primary)	11.8%	[2.3%, 27.3%]	15

Binary size

This perf run didn't have relevant results for this metric.

Bootstrap: 490.152s -> 495.841s (1.16%)
Artifact size: 408.63 MiB -> 408.61 MiB (-0.00%)

[rustbot]

This PR was rebased onto a different main commit. Here's a range-diff highlighting what actually changed.

Rebasing is a normal part of keeping PRs up to date, so no action is needed—this note is just to help reviewers.

[aerooneqq]

@rustbot ready

[petrochenkov]

This can be derived, similarly to SyntaxContext::default.

View changes since the review

[aerooneqq]

SyntaxContext is a struct here, LocalExpnId is declared through newtype_index!, so we can't derive (I don't see how to do it, no #[default] in this macro).

[petrochenkov]

I assumed putting #[derive(Default)] on struct SyntaxContext would just work, because the macro would pass it through.

[petrochenkov]

#163668 (comment)
@bors try @rust-timer queue

[rust-bors]

☀️ Try build successful (CI)
Build commit: 689c066 (689c0668839e3c831abf44c8c9c9cf1e5244e43d)
Base parent: b26b4e7 (b26b4e75de6aac63fc6beeecebad5b3a5f84d2a4)

[rust-timer]

Finished benchmarking commit (689c066): comparison URL.

Overall result: ❌✅ regressions and improvements - please read:

Benchmarking means the PR may be perf-sensitive. It's automatically marked not fit for rolling up. Overriding is possible but disadvised: it risks changing compiler perf.

Next, please: If you can, justify the regressions found in this try perf run in writing along with @rustbot label: +perf-regression-triaged. If not, fix the regressions and do another perf run. Neutral or positive results will clear the label automatically.

@bors rollup=never rustc-perf
@rustbot label: -S-waiting-on-perf +perf-regression

Instruction count

Our most reliable metric. Used to determine the overall result above. However, even this metric can be noisy.

	mean	range	count
Regressions ❌ (primary)	0.2%	[0.1%, 0.6%]	60
Regressions ❌ (secondary)	0.3%	[0.1%, 0.5%]	44
Improvements ✅ (primary)	-0.2%	[-0.3%, -0.2%]	7
Improvements ✅ (secondary)	-0.2%	[-0.3%, -0.1%]	6
All ❌✅ (primary)	0.2%	[-0.3%, 0.6%]	67

Max RSS (memory usage)

Results (primary 0.2%, secondary 1.2%)

A less reliable metric. May be of interest, but not used to determine the overall result above.

	mean	range	count
Regressions ❌ (primary)	1.5%	[1.5%, 1.5%]	1
Regressions ❌ (secondary)	3.2%	[0.7%, 5.0%]	4
Improvements ✅ (primary)	-1.0%	[-1.0%, -1.0%]	1
Improvements ✅ (secondary)	-1.6%	[-3.4%, -0.5%]	3
All ❌✅ (primary)	0.2%	[-1.0%, 1.5%]	2

Cycles

Results (primary 0.9%, secondary 0.1%)

A less reliable metric. May be of interest, but not used to determine the overall result above.

	mean	range	count
Regressions ❌ (primary)	2.7%	[2.3%, 3.1%]	2
Regressions ❌ (secondary)	4.0%	[0.8%, 7.3%]	2
Improvements ✅ (primary)	-2.5%	[-2.5%, -2.5%]	1
Improvements ✅ (secondary)	-1.2%	[-3.2%, -0.4%]	6
All ❌✅ (primary)	0.9%	[-2.5%, 3.1%]	3

Binary size

This perf run didn't have relevant results for this metric.

Bootstrap: 489.846s -> 490.341s (0.10%)
Artifact size: 409.05 MiB -> 408.98 MiB (-0.02%)