CopyProp: Propagate moves of constants by clubby789 · Pull Request #120689 · rust-lang/rust

clubby789 · 2024-02-05T23:33:42Z

Yet another try of #120650 - this time, keep track of SSA locals with a constant value and propagate it. Then, re-run SimplifyConstCondition to take advantage of it (simply reordering the pass regressed some tests).

r? @ghost

clubby789 · 2024-02-05T23:33:48Z

@bors try @rust-timer queue

bors · 2024-02-05T23:34:57Z

⌛ Trying commit 18f27f4 with merge 25ba6e3...

CopyProp: Propagate moves of constants Yet another try of rust-lang#120650 - this time, keep track of SSA locals with a constant value and propagate it. Then, re-run `SimplifyConstCondition` to take advantage of it (simply reordering the pass regressed some tests). r? `@ghost`

cjgillot · 2024-02-05T23:40:23Z

I suggest you take a look at the discussion about determinism of MIR constants in gvn.rs. There are case where the same MIR constant will yield a different value each time it is evaluated.
Meanwhile, I take from this PR:

we could have a better ordering of passes;
we may be able to mark more constants as deterministic, for instance if they have integer type.

bors · 2024-02-06T01:02:29Z

☀️ Try build successful - checks-actions
Build commit: 25ba6e3 (25ba6e3cb867376959985fb0cad7a13bec6530eb)

rust-timer · 2024-02-06T13:20:34Z

Finished benchmarking commit (25ba6e3): 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)	0.4%	[0.4%, 0.4%]	2
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-0.7%	[-1.3%, -0.3%]	4
Improvements ✅ (secondary)	-0.6%	[-1.1%, -0.2%]	2
All ❌✅ (primary)	-0.4%	[-1.3%, 0.4%]	6

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)	5.9%	[0.2%, 13.3%]	6
Regressions ❌ (secondary)	4.3%	[0.8%, 7.2%]	7
Improvements ✅ (primary)	-5.2%	[-7.6%, -2.8%]	2
Improvements ✅ (secondary)	-3.3%	[-3.3%, -3.3%]	1
All ❌✅ (primary)	3.1%	[-7.6%, 13.3%]	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)	2.0%	[2.0%, 2.0%]	1
Regressions ❌ (secondary)	2.3%	[2.3%, 2.3%]	1
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	2.0%	[2.0%, 2.0%]	1

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%, 0.4%]	5
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-0.3%	[-0.6%, -0.0%]	42
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-0.2%	[-0.6%, 0.4%]	47

Bootstrap: 660.166s -> 663.073s (0.44%)
Artifact size: 308.14 MiB -> 308.13 MiB (-0.00%)

clubby789 · 2024-02-06T22:34:29Z

Removed the pass ordering part. If perf looks happier, perhaps we can use this to simplify #120650 - rather than needing to codegen_operand, we can just check if our discrim is a const scalar.

@bors try @rust-timer queue

bors · 2024-02-06T22:35:39Z

⌛ Trying commit 7a6abd5 with merge 22f542c...

CopyProp: Propagate moves of constants Yet another try of rust-lang#120650 - this time, keep track of SSA locals with a constant value and propagate it. Then, re-run `SimplifyConstCondition` to take advantage of it (simply reordering the pass regressed some tests). r? `@ghost`

bors · 2024-02-07T00:04:40Z

☀️ Try build successful - checks-actions
Build commit: 22f542c (22f542c3e4b78858e75d4296f49b7aa6134122a9)

bors · 2024-02-07T00:04:40Z

☀️ Try build successful - checks-actions
Build commit: 22f542c (22f542c3e4b78858e75d4296f49b7aa6134122a9)

rust-timer · 2024-02-07T03:51:22Z

Finished benchmarking commit (22f542c): comparison URL.

Overall result: ❌ regressions - no 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.

@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
Regressions ❌ (secondary)	1.4%	[1.4%, 1.4%]	1
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-	-	0

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)	3.7%	[0.3%, 6.6%]	5
Regressions ❌ (secondary)	3.9%	[3.9%, 3.9%]	1
Improvements ✅ (primary)	-3.2%	[-6.0%, -1.4%]	4
Improvements ✅ (secondary)	-10.2%	[-10.2%, -10.2%]	1
All ❌✅ (primary)	0.7%	[-6.0%, 6.6%]	9

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)	1.6%	[1.6%, 1.6%]	1
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	1.6%	[1.6%, 1.6%]	1

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%, 0.4%]	11
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-0.1%	[-0.3%, -0.1%]	15
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-0.0%	[-0.3%, 0.4%]	26

Bootstrap: 663.272s -> 663.514s (0.04%)
Artifact size: 308.23 MiB -> 308.20 MiB (-0.01%)

clubby789 · 2024-02-07T16:21:37Z

Small net improvements if I run perf locally, so I think just noise

clubby789 · 2024-02-10T23:32:23Z

It looks like the previous slight improvements to MIR in tests were also done by #120594. That said, this seems to be basically perf-neutral and a bit more general so may be worth having.

r? compiler

rustbot · 2024-02-10T23:32:26Z

Some changes occurred to MIR optimizations

cc @rust-lang/wg-mir-opt

clubby789 · 2024-02-11T01:32:05Z

-        let (Rvalue::Use(Operand::Copy(place) | Operand::Move(place))
-        | Rvalue::CopyForDeref(place)) = rvalue
-        else {
+        let (Rvalue::Use(Operand::Copy(place)) | Rvalue::CopyForDeref(place)) = rvalue else {


Given that the Operand::Move case does nothing (due to the check below), we can remove that here and remove the duplicated check below

compiler-errors · 2024-02-12T15:58:34Z

r? cjgillot

cjgillot · 2024-02-13T13:22:36Z

From the results of the #120650 (comment), do we still need this PR? Or should we merge #120650 on its own?

clubby789 · 2024-04-16T10:58:03Z

First lets see if this still has any perf impact:

@bors try @rust-timer queue

I suspect not - if this comes up neutral I'll probably close

CopyProp: Propagate moves of constants Yet another try of rust-lang#120650 - this time, keep track of SSA locals with a constant value and propagate it. Then, re-run `SimplifyConstCondition` to take advantage of it (simply reordering the pass regressed some tests). r? `@ghost`

bors · 2024-04-16T11:03:02Z

⌛ Trying commit cf02fa9 with merge 1515cb7...

bors · 2024-04-16T12:35:33Z

☀️ Try build successful - checks-actions
Build commit: 1515cb7 (1515cb760c8a0555b9712621f65c11a283b0eaf4)

rust-timer · 2024-04-16T13:53:42Z

Finished benchmarking commit (1515cb7): comparison URL.

Overall result: no relevant changes - no 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.

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

Instruction count

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

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)	3.8%	[2.5%, 5.0%]	3
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-2.1%	[-2.1%, -2.1%]	1
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	2.3%	[-2.1%, 5.0%]	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)	-	-	0
Regressions ❌ (secondary)	1.1%	[1.1%, 1.1%]	1
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-0.8%	[-0.8%, -0.8%]	1
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)	0.1%	[0.0%, 0.5%]	9
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-0.1%	[-0.4%, -0.0%]	7
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	0.0%	[-0.4%, 0.5%]	16

Bootstrap: 678.471s -> 677.873s (-0.09%)
Artifact size: 316.03 MiB -> 315.98 MiB (-0.01%)

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 Feb 5, 2024

This comment has been minimized.

Sign in to view

rustbot added the S-waiting-on-perf Status: Waiting on a perf run to be completed. label Feb 5, 2024

This comment has been minimized.

Sign in to view

rustbot added perf-regression Performance regression. and removed S-waiting-on-perf Status: Waiting on a perf run to be completed. labels Feb 6, 2024

rust-cloud-vms Bot force-pushed the copy-prop-consts branch from 18f27f4 to 7a6abd5 Compare February 6, 2024 22:31

This comment has been minimized.

Sign in to view

rustbot added the S-waiting-on-perf Status: Waiting on a perf run to be completed. label Feb 6, 2024

This comment has been minimized.

Sign in to view

rustbot removed S-waiting-on-perf Status: Waiting on a perf run to be completed. perf-regression Performance regression. labels Feb 7, 2024

clubby789 mentioned this pull request Feb 7, 2024

Use br instead of a conditional when switching on a constant boolean #120650

Merged

CopyProp: Propagate moves of constants

0b0011c

rust-cloud-vms Bot force-pushed the copy-prop-consts branch from 7a6abd5 to 0b0011c Compare February 10, 2024 23:29

clubby789 marked this pull request as ready for review February 10, 2024 23:32

rustbot assigned compiler-errors Feb 10, 2024

CopyProp cleanups

cf02fa9

rust-cloud-vms Bot force-pushed the copy-prop-consts branch from 3be2167 to cf02fa9 Compare February 11, 2024 01:31

clubby789 commented Feb 11, 2024

View reviewed changes

rustbot assigned cjgillot and unassigned compiler-errors Feb 12, 2024

cjgillot 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 Feb 18, 2024

This comment has been minimized.

Sign in to view

rustbot added the S-waiting-on-perf Status: Waiting on a perf run to be completed. label Apr 16, 2024

This comment has been minimized.

Sign in to view

rustbot removed the S-waiting-on-perf Status: Waiting on a perf run to be completed. label Apr 16, 2024

clubby789 closed this Apr 16, 2024

Uh oh!

Conversation

clubby789 commented Feb 5, 2024

Uh oh!

clubby789 commented Feb 5, 2024

Uh oh!

This comment has been minimized.

bors commented Feb 5, 2024

Uh oh!

cjgillot commented Feb 5, 2024 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

bors commented Feb 6, 2024

Uh oh!

This comment has been minimized.

rust-timer commented Feb 6, 2024

Overall result: ❌✅ regressions and improvements - ACTION NEEDED

Uh oh!

clubby789 commented Feb 6, 2024

Uh oh!

This comment has been minimized.

bors commented Feb 6, 2024

Uh oh!

bors commented Feb 7, 2024

Uh oh!

bors commented Feb 7, 2024

Uh oh!

This comment has been minimized.

rust-timer commented Feb 7, 2024

Overall result: ❌ regressions - no action needed

Uh oh!

clubby789 commented Feb 7, 2024

Uh oh!

clubby789 commented Feb 10, 2024 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

rustbot commented Feb 10, 2024

Uh oh!

Choose a reason for hiding this comment

Uh oh!

compiler-errors commented Feb 12, 2024

Uh oh!

cjgillot commented Feb 13, 2024

Uh oh!

clubby789 commented Apr 16, 2024

Uh oh!

This comment has been minimized.

bors commented Apr 16, 2024

Uh oh!

bors commented Apr 16, 2024

Uh oh!

This comment has been minimized.

rust-timer commented Apr 16, 2024

Overall result: no relevant changes - no action needed

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

6 participants

cjgillot commented Feb 5, 2024 •

edited

Loading

clubby789 commented Feb 10, 2024 •

edited

Loading