MIR: opt-in normalization of `BasicBlock` and `Local` numbering #111813

scottmcm · 2023-05-21T08:41:48Z

This doesn't matter at all for actual codegen, but after spending some time reading pre-codegen MIR, I was wishing I didn't have to jump around so much in reading post-inlining code.

So this add two passes that are off by default for every mir level, but can be enabled (-Zmir-enable-passes=+ReorderBasicBlocks,+ReorderLocals) for humans.

rustbot · 2023-05-21T08:41:53Z

r? @cjgillot

(rustbot has picked a reviewer for you, use r? to override)

rustbot · 2023-05-21T08:41:56Z

Some changes occurred to MIR optimizations

cc @rust-lang/wg-mir-opt

cjgillot · 2023-05-21T08:43:40Z

tests/mir-opt/inline/unchecked_shifts.rs

@@ -4,6 +4,7 @@

 // ignore-debug: the debug assertions prevent the inlining we are testing for
 // compile-flags: -Zmir-opt-level=2 -Zinline-mir
+// compile-flags: -Zmir-enable-passes=+ReorderBasicBlocks,+ReorderLocals


Should this flag be passed by default for mir-opt tests?

Ooh, I like it, since it'll only affect tests that are using the PreCodegen MIR, not all the intermediate stuff.

cjgillot · 2023-05-21T08:48:01Z

compiler/rustc_mir_transform/src/prettify.rs

+use rustc_middle::mir::visit::{MutVisitor, PlaceContext, Visitor};
+use rustc_session::Session;
+
+pub struct ReorderBasicBlocks;


Could you doc-comment in which order the blocks are ordered?

cjgillot · 2023-05-21T08:48:09Z

compiler/rustc_mir_transform/src/prettify.rs

+    }
+}
+
+pub struct ReorderLocals;


cjgillot · 2023-05-21T08:50:19Z

compiler/rustc_mir_transform/src/prettify.rs

+            map: finder.map.invert_bijective_mapping(),
+            tcx
+        };
+        debug_assert_eq!(updater.map[RETURN_PLACE], RETURN_PLACE);


Should we expand this assertion to arguments too?

cjgillot · 2023-05-21T09:19:38Z

compiler/rustc_mir_transform/src/prettify.rs

+
+impl<'tcx> Visitor<'tcx> for LocalFinder {
+    fn visit_local(&mut self, l: Local, context: PlaceContext, _location: Location) {
+        if context.is_use() {


Why exclude non-uses?

Because otherwise StorageLive ends up putting locals much earlier in the list than makes sense to me. I'll comment it (and do the other suggestions).

@rustbot author

Since this only affects `PreCodegen MIR, and it would be nice for that to be resilient to permutations of things that don't affect the actual semantic behaviours.

scottmcm · 2023-05-22T03:10:28Z

@rustbot ready

cjgillot · 2023-05-24T17:22:53Z

tests/ui/issues/issue-52049.stderr

@@ -2,12 +2,10 @@ error[E0716]: temporary value dropped while borrowed
  --> $DIR/issue-52049.rs:6:10
   |
 LL |     foo(&unpromotable(5u32));
-   |     -----^^^^^^^^^^^^^^^^^^-
+   |     -----^^^^^^^^^^^^^^^^^^-- temporary value is freed at the end of this statement


Because I changed the RPO of

A / \ B C \ / D

from A C B D to A B C D, and apparently that changed some error locations.

I can pull that change out, if you'd prefer.

cjgillot · 2023-05-28T07:00:52Z

Thanks! Those error are placed a bit better now anyway.
@bors r+

bors · 2023-05-28T07:00:54Z

📌 Commit d69725d has been approved by cjgillot

It is now in the queue for this repository.

bors · 2023-05-28T21:53:59Z

⌛ Testing commit d69725d with merge 089677e...

bors · 2023-05-29T00:25:47Z

☀️ Test successful - checks-actions
Approved by: cjgillot
Pushing 089677e to master...

rust-timer · 2023-05-29T02:33:32Z

Finished benchmarking commit (089677e): comparison URL.

Overall result: ❌✅ regressions and improvements - ACTION NEEDED

Next Steps: If you can justify the regressions found in this perf run, please indicate this with @rustbot label: +perf-regression-triaged along with sufficient written justification. If you cannot justify the regressions please open an issue or create a new PR that fixes the regressions, add a comment linking to the newly created issue or PR, and then add the perf-regression-triaged label to this PR.

@rustbot label: +perf-regression
cc @rust-lang/wg-compiler-performance

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%]	1
Regressions ❌ (secondary)	0.7%	[0.3%, 1.2%]	8
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-1.0%	[-1.5%, -0.3%]	7
All ❌✅ (primary)	0.3%	[0.3%, 0.3%]	1

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)	0.1%	[0.1%, 0.1%]	1
Regressions ❌ (secondary)	2.3%	[2.3%, 2.3%]	1
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	0.1%	[0.1%, 0.1%]	1

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)	-	-	0
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-1.9%	[-1.9%, -1.9%]	2
All ❌✅ (primary)	-	-	0

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)	0.2%	[0.0%, 0.6%]	50
Regressions ❌ (secondary)	0.0%	[0.0%, 0.0%]	6
Improvements ✅ (primary)	-0.0%	[-0.0%, -0.0%]	8
Improvements ✅ (secondary)	-0.3%	[-1.0%, -0.0%]	9
All ❌✅ (primary)	0.2%	[-0.0%, 0.6%]	58

Bootstrap: 646.44s -> 646.979s (0.08%)

rustbot assigned cjgillot May 21, 2023

rustbot added S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. T-compiler Relevant to the compiler team, which will review and decide on the PR/issue. labels May 21, 2023

This comment has been minimized.

Sign in to view

scottmcm force-pushed the pretty-mir branch from 4533b0a to 52ce40d Compare May 21, 2023 09:00

cjgillot reviewed May 21, 2023

View reviewed changes

rustbot added S-waiting-on-author Status: This is awaiting some action (such as code changes or more information) from the author. and removed S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. labels May 21, 2023

This comment has been minimized.

Sign in to view

scottmcm force-pushed the pretty-mir branch from 99ccad8 to d8e4534 Compare May 21, 2023 23:16

rustbot added A-testsuite Area: The testsuite used to check the correctness of rustc T-bootstrap Relevant to the bootstrap subteam: Rust's build system (x.py and src/bootstrap) labels May 21, 2023

scottmcm force-pushed the pretty-mir branch from d8e4534 to 1313325 Compare May 22, 2023 00:06

This comment has been minimized.

Sign in to view

scottmcm added 4 commits May 21, 2023 17:48

Tweak the post-order for multi-successor blocks

57c5ac7

MIR: opt-in normalization of BasicBlock and Local numbering

977ac5b

PR feedback: better comments and debug asserts

05bf42b

Normalize block and local orders in mir-opt tests

d69725d

Since this only affects `PreCodegen MIR, and it would be nice for that to be resilient to permutations of things that don't affect the actual semantic behaviours.

scottmcm force-pushed the pretty-mir branch from 1313325 to d69725d Compare May 22, 2023 00:48

rustbot added S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. and removed S-waiting-on-author Status: This is awaiting some action (such as code changes or more information) from the author. labels May 22, 2023

cjgillot reviewed May 24, 2023

View reviewed changes

bors added S-waiting-on-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion. and removed S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. labels May 28, 2023

bors added the merged-by-bors This PR was explicitly merged by bors. label May 29, 2023

bors merged commit 089677e into rust-lang:master May 29, 2023

rustbot added this to the 1.72.0 milestone May 29, 2023

scottmcm deleted the pretty-mir branch May 29, 2023 00:26

scottmcm mentioned this pull request May 29, 2023

[mir-opt] SimplifyLocals should also clean up debuginfo #110702

Closed

rustbot added the perf-regression Performance regression. label May 29, 2023

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

MIR: opt-in normalization of `BasicBlock` and `Local` numbering #111813

MIR: opt-in normalization of `BasicBlock` and `Local` numbering #111813

scottmcm commented May 21, 2023

rustbot commented May 21, 2023

rustbot commented May 21, 2023

This comment has been minimized.

cjgillot May 21, 2023

scottmcm May 22, 2023

cjgillot May 21, 2023

cjgillot May 21, 2023

cjgillot May 21, 2023

cjgillot May 21, 2023

scottmcm May 21, 2023

This comment has been minimized.

This comment has been minimized.

scottmcm commented May 22, 2023

cjgillot May 24, 2023

scottmcm May 27, 2023 •

edited

Loading

cjgillot commented May 28, 2023

bors commented May 28, 2023

bors commented May 28, 2023

bors commented May 29, 2023

rust-timer commented May 29, 2023

MIR: opt-in normalization of BasicBlock and Local numbering #111813

MIR: opt-in normalization of BasicBlock and Local numbering #111813

Conversation

scottmcm commented May 21, 2023

rustbot commented May 21, 2023

rustbot commented May 21, 2023

This comment has been minimized.

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

This comment has been minimized.

This comment has been minimized.

scottmcm commented May 22, 2023

Choose a reason for hiding this comment

scottmcm May 27, 2023 • edited Loading

Choose a reason for hiding this comment

cjgillot commented May 28, 2023

bors commented May 28, 2023

bors commented May 28, 2023

bors commented May 29, 2023

rust-timer commented May 29, 2023

Overall result: ❌✅ regressions and improvements - ACTION NEEDED

MIR: opt-in normalization of `BasicBlock` and `Local` numbering #111813

MIR: opt-in normalization of `BasicBlock` and `Local` numbering #111813

scottmcm May 27, 2023 •

edited

Loading