Add -Z small-data-threshold #117465

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Merged

bors merged 1 commit into rust-lang:master from paulmenage:small-data-limit

Sep 12, 2024

Contributor

paulmenage commented Oct 31, 2023 •

edited

Loading

This flag allows specifying the threshold size above which LLVM should not consider placing small objects in a .sdata or .sbss section.

Support is indicated in the target options via the
small-data-threshold-support target option, which can indicate either an
LLVM argument or an LLVM module flag. To avoid duplicate specifications
in a large number of targets, the default value for support is
DefaultForArch, which is translated to a concrete value according to the
target's architecture.

rustbot assigned b-naber

Collaborator

rustbot commented Oct 31, 2023

Thanks for the pull request, and welcome! The Rust team is excited to review your changes, and you should hear from @b-naber (or someone else) soon.

Please see the contribution instructions for more information. Namely, in order to ensure the minimum review times lag, PR authors and assigned reviewers should ensure that the review label (S-waiting-on-review and S-waiting-on-author) stays updated, invoking these commands when appropriate:

@rustbot author: the review is finished, PR author should check the comments and take action accordingly
@rustbot review: the author is ready for a review, this PR will be queued again in the reviewer's queue

rustbot added S-waiting-on-review T-compiler labels

paulmenage force-pushed the small-data-limit branch from 358ff44 to 42150ae Compare

October 31, 2023 21:39

Contributor Author

paulmenage commented Oct 31, 2023 •

edited

Loading

This is a more targeted version of what I was trying to achieve with #116555 (which would add -Z llvm-module-flag to enable the same goal on RISCV).

This comment has been minimized.

Sign in to view

paulmenage force-pushed the small-data-limit branch from 42150ae to 54eada2 Compare

October 31, 2023 22:24

This comment has been minimized.

Sign in to view

paulmenage force-pushed the small-data-limit branch from 54eada2 to 8a6e656 Compare

October 31, 2023 22:36

This comment has been minimized.

Sign in to view

paulmenage force-pushed the small-data-limit branch from 8a6e656 to 9f35e5f Compare

October 31, 2023 22:46

This comment has been minimized.

Sign in to view

paulmenage force-pushed the small-data-limit branch 3 times, most recently from 9511b44 to f4f0492 Compare

November 1, 2023 06:30

paulmenage force-pushed the small-data-limit branch from f4f0492 to aefc127 Compare

November 12, 2023 03:51

Contributor

b-naber commented Nov 17, 2023

Sorry for the late reply. I think this needs an MCP before it can be merged.

paulmenage mentioned this pull request

Add -Z small-data-threshold rust-lang/compiler-team#689

Closed

3 tasks

Collaborator

bors commented Nov 22, 2023

☔ The latest upstream changes (presumably #118178) made this pull request unmergeable. Please resolve the merge conflicts.

Contributor

b-naber commented Dec 9, 2023

This could be merged now. Can you rebase?

Contributor Author

paulmenage commented Dec 12, 2023

There's still some lack of clarity on the MCP zulip stream whether this is the right way to solve the problem, versus a target feature. I'll push the rebase but it's probably best to not merge it yet.

paulmenage force-pushed the small-data-limit branch from aefc127 to 4660911 Compare

December 12, 2023 13:05

paulmenage force-pushed the small-data-limit branch from 4660911 to a618f74 Compare

January 21, 2024 18:10

Contributor Author

paulmenage commented Jan 21, 2024

OK, the conclusion on the zulip chat appeared to be that a target feature wasn't the right approach, so I think this is ready.

Contributor

apiraino commented Jan 25, 2024

@b-naber can you r+? If I understand this patch is ripe for merging. thanks.

compiler-errors approved these changes

View reviewed changes

Member

compiler-errors commented Feb 5, 2024

r? compiler-errors @bors r+

rustbot assigned compiler-errors

Collaborator

bors commented Feb 5, 2024

📌 Commit a618f74 has been approved by compiler-errors

It is now in the queue for this repository.

This comment has been minimized.

Sign in to view

paulmenage force-pushed the small-data-limit branch 2 times, most recently from c338170 to d470615 Compare

September 10, 2024 18:16

compiler-errors approved these changes

View reviewed changes

Contributor Author

paulmenage commented Sep 10, 2024

In the last update, I tweaked it to include a default based on the architecture, to avoid having to duplicate the support in many different target files.

klensy reviewed

View reviewed changes

compiler/rustc_codegen_llvm/src/context.rs Outdated

    
                      // Set up the small-data optimization limit for architectures that use

                      // an LLVM module flag to control this.

                      (Some(threshold), SmallDataThresholdSupport::LlvmModuleFlag(flag)) => {

                          let flag = format!("{flag}\0");

Contributor

klensy Sep 10, 2024

There is a CString for that, no need to manually add nulls.

Contributor Author

paulmenage Sep 10, 2024

Something like let flag = CString::new(flag).unwrap()? Or should SmallDataThresholdSupport::LllvmModuleFlag contain a StaticCow<Cstr> rather than StaticCow<str>?

Contributor

klensy Sep 10, 2024

No, changing type to StaticCow<Cstr> is probably no needed.

Contributor Author

paulmenage Sep 10, 2024

OK, tweaked to use let flag = SmallCStr::new(flag.as_ref());

klensy reviewed

View reviewed changes

tests/assembly/small_data_threshold.rs Outdated

    
              //@ RISCV-NOT: .section

              //@ RISCV: U:

              //@ RISCV: .section .sbss

              //@ RISV-NOT: .section

Contributor

klensy Sep 10, 2024

typo

Contributor Author

paulmenage Sep 10, 2024

Oops. A bunch of copy-pasted typos, sorry. :-(

klensy reviewed

View reviewed changes

tests/assembly/small_data_threshold.rs Outdated

    
              //@ RISV-NOT: .section

              //@ RISCV: V:

              //@ RISCV .section .sdata

              //@ RISV-NOT: .section

Contributor

klensy Sep 10, 2024

here too

klensy reviewed

View reviewed changes

tests/assembly/small_data_threshold.rs Outdated

    
              //@ RISV-NOT: .section

              //@ RISCV: W:

              //@ RISCV: .section .sbss

              //@ RISV-NOT: .section

Contributor

klensy Sep 10, 2024

...

klensy reviewed

View reviewed changes

tests/assembly/small_data_threshold.rs Outdated

    
              //@ MIPS-NOT: .section

              //@ MIPS: U:

              //@ MIPS: .section .sbss

              //@ RISV-NOT: .section

Contributor

klensy Sep 10, 2024

This and 2 next looks like should be MIPS-NOT?

klensy reviewed

View reviewed changes

tests/assembly/small_data_threshold.rs Outdated

    
              //@ RISCV: .section .sbss

              //@ RISV-NOT: .section

              //@ RISCV: V:

              //@ RISCV .section .sdata

Contributor

klensy Sep 10, 2024

:

klensy reviewed

View reviewed changes

tests/assembly/small_data_threshold.rs Outdated

    
              //@ MIPS: .section .sbss

              //@ RISV-NOT: .section

              //@ MIPS: V:

              //@ MIPS .section .sdata

Contributor

klensy Sep 10, 2024

:


          Add -Z small-data-threshold

This flag allows specifying the threshold size above which LLVM should
not consider placing small objects in a .sdata or .sbss section.

Support is indicated in the target options via the
small-data-threshold-support target option, which can indicate either an
LLVM argument or an LLVM module flag.  To avoid duplicate specifications
in a large number of targets, the default value for support is
DefaultForArch, which is translated to a concrete value according to the
target's architecture.

paulmenage force-pushed the small-data-limit branch from d470615 to 3810386 Compare

September 10, 2024 19:20

Contributor Author

paulmenage commented Sep 10, 2024

I fixed the typos in the tests.

Member

compiler-errors commented Sep 11, 2024

Collaborator

bors commented Sep 11, 2024

📌 Commit 3810386 has been approved by compiler-errors

It is now in the queue for this repository.

bors added S-waiting-on-bors and removed S-waiting-on-review labels

Member

compiler-errors commented Sep 11, 2024

Thanks @klensy for your thorough reviews as always

Member

workingjubilee commented Sep 12, 2024

@bors rollup=iffy

Collaborator

bors commented Sep 12, 2024

⌛ Testing commit 3810386 with merge 1f51450...

bors mentioned this pull request

Update compiler-builtins to 0.1.125 #129992

Merged

Collaborator

bors commented Sep 12, 2024

☀️ Test successful - checks-actions
Approved by: compiler-errors
Pushing 1f51450 to master...

bors added the merged-by-bors label

bors merged commit 1f51450 into rust-lang:master

rustbot added this to the 1.83.0 milestone

This was referenced Sep 12, 2024

sess: stabilize -C stack-protector=all #121742

Closed

mark some target features as 'forbidden' so they cannot be (un)set with -Ctarget-feature #129884

Merged

Collaborator

rust-timer commented Sep 12, 2024

Finished benchmarking commit (1f51450): comparison URL.

Overall result: no relevant changes - no action needed

@rustbot label: -perf-regression

Instruction count

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

Max RSS (memory usage)

Results (primary 1.5%)

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)	1.5%	[0.6%, 2.4%]	2
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	1.5%	[0.6%, 2.4%]	2

Cycles

Results (primary 6.5%)

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)	6.5%	[6.5%, 6.5%]	1
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	6.5%	[6.5%, 6.5%]	1

Binary size

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

Bootstrap: 756.297s -> 757.698s (0.19%)
Artifact size: 341.30 MiB -> 341.34 MiB (0.01%)

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

merged-by-bors S-waiting-on-bors T-compiler