Remove generics from `DepGraph` by Zoxc · Pull Request #151982 · rust-lang/rust

Zoxc · 2026-02-02T04:59:39Z

This removes generics from DepGraph by predefining some properties of dep kinds in rustc_query_system. DEP_KIND_NULL, DEP_KIND_RED, DEP_KIND_SIDE_EFFECT and DEP_KIND_ANON_ZERO_DEPS get their indices assigned in rustc_query_system and the bits needed to encode dep kinds is also hardcoded there. These are checked later in rustc_middle.

read_index, assert_ignored and with_ignore are redefined on DepContext to avoid manual passing of the Deps type.

Zoxc · 2026-02-02T10:09:48Z

@bors try @rust-timer queue

Remove generics from `DepGraph`

rust-bors · 2026-02-02T12:20:31Z

☀️ Try build successful (CI)
Build commit: 1f642ee (1f642eeae16567707264119f787cef8c5546ff24, parent: a60d12cbccfbeaf153f3cecb90454aa696ea4b3b)

rust-timer · 2026-02-02T13:01:44Z

Finished benchmarking commit (1f642ee): comparison URL.

Overall result: ❌ regressions - please read the text below

Benchmarking this pull request means it may be perf-sensitive – we'll automatically label it not fit for rolling up. You can override this, but we strongly advise not to, due to possible changes in compiler perf.

Next Steps: If you can justify the regressions found in this try perf run, please do so in sufficient writing along with @rustbot label: +perf-regression-triaged. If not, please fix the regressions and do another perf run. If its results are neutral or positive, the label will be automatically removed.

@bors rollup=never
@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%	[0.2%, 0.3%]	6
Regressions ❌ (secondary)	0.4%	[0.2%, 0.5%]	9
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	0.2%	[0.2%, 0.3%]	6

Max RSS (memory usage)

Results (primary -1.3%)

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

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-1.3%	[-1.3%, -1.3%]	1
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-1.3%	[-1.3%, -1.3%]	1

Cycles

Results (primary 8.9%, secondary 1.1%)

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

	mean	range	count
Regressions ❌ (primary)	8.9%	[2.1%, 16.3%]	12
Regressions ❌ (secondary)	5.7%	[5.7%, 5.7%]	1
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-3.5%	[-3.5%, -3.5%]	1
All ❌✅ (primary)	8.9%	[2.1%, 16.3%]	12

Binary size

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

Bootstrap: 473.737s -> 471.042s (-0.57%)
Artifact size: 397.70 MiB -> 397.78 MiB (0.02%)

Zoxc · 2026-02-02T20:04:34Z

@bors try @rust-timer queue

Remove generics from `DepGraph`

rust-bors · 2026-02-02T20:19:47Z

💔 Test for c718310 failed: CI

Zoxc · 2026-02-04T07:09:20Z

@bors try @rust-timer queue

Remove generics from `DepGraph`

rust-bors · 2026-02-04T09:19:39Z

☀️ Try build successful (CI)
Build commit: 3bc6aa4 (3bc6aa4a93660ead71a27b267705ece987ff1e49, parent: 1d05e3c131d7181eb3df1a8c261f43135c99200d)

rust-timer · 2026-02-04T10:41:26Z

Finished benchmarking commit (3bc6aa4): comparison URL.

Overall result: ❌ regressions - please read the text below

Benchmarking this pull request means it may be perf-sensitive – we'll automatically label it not fit for rolling up. You can override this, but we strongly advise not to, due to possible changes in compiler perf.

Next Steps: If you can justify the regressions found in this try perf run, please do so in sufficient writing along with @rustbot label: +perf-regression-triaged. If not, please fix the regressions and do another perf run. If its results are neutral or positive, the label will be automatically removed.

@bors rollup=never
@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)	0.3%	[0.2%, 0.3%]	6
Regressions ❌ (secondary)	0.3%	[0.0%, 0.5%]	15
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-0.0%	[-0.1%, -0.0%]	2
All ❌✅ (primary)	0.3%	[0.2%, 0.3%]	6

Max RSS (memory usage)

Results (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)	-	-	0
Regressions ❌ (secondary)	2.4%	[2.4%, 2.4%]	1
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-	-	0

Cycles

Results (primary -2.1%, 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)	-	-	0
Regressions ❌ (secondary)	4.5%	[4.1%, 4.8%]	2
Improvements ✅ (primary)	-2.1%	[-2.1%, -2.1%]	1
Improvements ✅ (secondary)	-4.8%	[-6.3%, -3.6%]	4
All ❌✅ (primary)	-2.1%	[-2.1%, -2.1%]	1

Binary size

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

Bootstrap: 476.238s -> 472.74s (-0.73%)
Artifact size: 397.76 MiB -> 397.85 MiB (0.02%)

Zalathar · 2026-02-04T10:58:22Z

I'm surprised to see a perf difference at all.

Perhaps removing the generics has accidentally resulted in less inlining, and we would need some inlining attributes to compensate?

Zalathar · 2026-02-04T11:00:25Z

compiler/rustc_query_system/src/dep_graph/dep_node.rs

+/// This must match the use of the define_dep_nodes! macro.
+#[repr(u16)]
+enum BuiltinDepKinds {
+    /// We use this for most things when incr. comp. is turned off.
+    Null,
+
+    /// We use this to create a forever-red node.
+    Red,
+
+    /// We use this to create a side effect node.
+    SideEffect,
+
+    /// We use this to create the anon node with zero dependencies.
+    AnonZeroDeps,
+}


Suggestion: Perhaps we could match exhaustively on this enum in the downstream-crate assertions, to statically double-check that we've asserted on all of the constants that actually need to match.

rust-bors · 2026-02-05T11:26:04Z

☔ The latest upstream changes made this pull request unmergeable. Please resolve the merge conflicts.