Avoid wrapping constant allocations in packed structs when not necessary #138503

bjorn3 · 2025-03-14T15:56:53Z

This way LLVM will set the string merging flag if the alloc is a nul terminated string, reducing binary sizes.

rustbot · 2025-03-14T15:56:58Z

rustbot has assigned @estebank.
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

rustbot · 2025-03-14T15:56:59Z

Some changes occurred in compiler/rustc_codegen_gcc

cc @antoyo, @GuillaumeGomez

bjorn3 · 2025-03-14T15:57:03Z

@bors try @rust-timer queue

bors · 2025-03-14T15:58:16Z

⌛ Trying commit 1688182 with merge 3da9ce8...

Avoid wrapping constant allocations in packed structs when not necessary This way LLVM will set the string merging flag if the alloc is a nul terminated string, reducing binary sizes.

bjorn3 · 2025-03-14T16:05:44Z

tests/codegen/external-no-mangle-statics.rs

    #[no_mangle]
    pub static mut L: u8 = 0;
 };

 fn x() {
-    // CHECK: @M = {{(dso_local )?}}local_unnamed_addr constant
+    // CHECK-DAG: @M = {{(dso_local )?}}local_unnamed_addr constant


LLVM decided to put this definition before A. Using CHECK-DAG everywhere allows filecheck to match independent of the order.

…r=<try> Nul terminate rust string literals This allows taking advantage of the C string merging functionality of linkers, reducing code size. Marked as draft to see if this actually has much of an effect. The disadvantage of this is that people may start to rely on string literals getting nul terminated. A potential solution for that would be to put a byte that is not part of a valid UTF-8 character right before the nul terminator. Builds on rust-lang#138503

bors · 2025-03-14T18:01:24Z

☀️ Try build successful - checks-actions
Build commit: 3da9ce8 (3da9ce89fe5dc7ed237607149c0cfc7e3f490ce1)

rust-timer · 2025-03-14T22:03:41Z

Finished benchmarking commit (3da9ce8): comparison URL.

Overall result: ❌✅ regressions and improvements - please read the text below

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 the most reliable metric that we have; it was used to determine the overall result at the top of this comment. However, even this metric can sometimes exhibit noise.

	mean	range	count
Regressions ❌ (primary)	1.7%	[1.7%, 1.7%]	1
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-1.7%	[-4.4%, -0.3%]	5
Improvements ✅ (secondary)	-2.3%	[-3.3%, -1.9%]	8
All ❌✅ (primary)	-1.1%	[-4.4%, 1.7%]	6

Max RSS (memory usage)

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

Cycles

Results (primary -2.2%, secondary -3.7%)

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)	-2.2%	[-4.2%, -1.3%]	4
Improvements ✅ (secondary)	-3.7%	[-4.7%, -3.2%]	8
All ❌✅ (primary)	-2.2%	[-4.2%, -1.3%]	4

Binary size

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

Bootstrap: 775.364s -> 773.305s (-0.27%)
Artifact size: 365.01 MiB -> 365.01 MiB (0.00%)

bjorn3 · 2025-03-14T22:08:21Z

No statistically significant binary size changes, which is not all that surprising as most strings are not nul terminated. Compilation gets a bit faster though.

This way LLVM will set the string merging flag if the alloc is a nul terminated string, reducing binary sizes.

bjorn3 · 2025-03-17T10:57:58Z

Seems that cg_gcc doesn't handle changing the type of the static to the actual type. Reverted the change for cg_gcc.

antoyo · 2025-03-17T12:39:54Z

Seems that cg_gcc doesn't handle changing the type of the static to the actual type. Reverted the change for cg_gcc.

Perhaps this was missing a bitcast?

bjorn3 · 2025-03-17T12:45:48Z

I think it was

rust/compiler/rustc_codegen_gcc/src/consts.rs

Lines 89 to 92 in 9bad8ac

    
           #[cfg(feature = "master")] 
        
           if global.to_rvalue().get_type() != val_llty { 
        
               global.to_rvalue().set_type(val_llty); 
        
           }

being configured out.

antoyo · 2025-03-17T13:56:45Z

I think it was

rust/compiler/rustc_codegen_gcc/src/consts.rs

Lines 89 to 92 in 9bad8ac

#[cfg(feature = "master")]

if global.to_rvalue().get_type() != val_llty {

global.to_rvalue().set_type(val_llty);

}

being configured out.

Oh, I meant an actual call to new_bitcast(), but yeah, maybe that set_type() call should have done it.
I'll check that on my side when this PR is merged.

tmiasko · 2025-03-25T11:31:37Z

Can you add a regression test with C-string literals? An end-to-end test would be nice to ensure that everything works as expected, but LLVM code generation test with an explanation also seems sufficient. Can you also include a reference to https://github.com/rust-lang/llvm-project/blob/acaea3d2bb8f351b740db7ebce7d7a40b9e21488/llvm/lib/Target/TargetLoweringObjectFile.cpp#L249-L280.

r? tmiasko

rustbot assigned estebank Mar 14, 2025

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 Mar 14, 2025

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 Mar 14, 2025

bjorn3 commented Mar 14, 2025

View reviewed changes

bjorn3 mentioned this pull request Mar 14, 2025

Nul terminate rust string literals #138504

Draft

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 Mar 14, 2025

Avoid wrapping constant allocations in packed structs when not necessary

adeb113

This way LLVM will set the string merging flag if the alloc is a nul terminated string, reducing binary sizes.

bjorn3 force-pushed the string_merging branch from 1688182 to adeb113 Compare March 17, 2025 10:56

bjorn3 mentioned this pull request Mar 17, 2025

Add Location::file_with_nul #135054

Open

rustbot assigned tmiasko and unassigned estebank Mar 25, 2025

Add test and comment

a84b2c6

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Avoid wrapping constant allocations in packed structs when not necessary #138503

Avoid wrapping constant allocations in packed structs when not necessary #138503

bjorn3 commented Mar 14, 2025

rustbot commented Mar 14, 2025

rustbot commented Mar 14, 2025

bjorn3 commented Mar 14, 2025

This comment has been minimized.

bors commented Mar 14, 2025

bjorn3 Mar 14, 2025

This comment has been minimized.

bors commented Mar 14, 2025

This comment has been minimized.

rust-timer commented Mar 14, 2025

bjorn3 commented Mar 14, 2025

bjorn3 commented Mar 17, 2025

antoyo commented Mar 17, 2025

bjorn3 commented Mar 17, 2025

antoyo commented Mar 17, 2025

tmiasko commented Mar 25, 2025

Avoid wrapping constant allocations in packed structs when not necessary #138503

Are you sure you want to change the base?

Avoid wrapping constant allocations in packed structs when not necessary #138503

Conversation

bjorn3 commented Mar 14, 2025

rustbot commented Mar 14, 2025

rustbot commented Mar 14, 2025

bjorn3 commented Mar 14, 2025

This comment has been minimized.

bors commented Mar 14, 2025

bjorn3 Mar 14, 2025

Choose a reason for hiding this comment

This comment has been minimized.

bors commented Mar 14, 2025

This comment has been minimized.

rust-timer commented Mar 14, 2025

Overall result: ❌✅ regressions and improvements - please read the text below

Instruction count

Max RSS (memory usage)

Cycles

Binary size

bjorn3 commented Mar 14, 2025

bjorn3 commented Mar 17, 2025

antoyo commented Mar 17, 2025

bjorn3 commented Mar 17, 2025

antoyo commented Mar 17, 2025

tmiasko commented Mar 25, 2025