Rollup of 11 pull requests #141739

GuillaumeGomez · 2025-05-29T15:03:04Z

Successful merges:

Make std/src/num mirror core/src/num #137574 (Make std/src/num mirror core/src/num)
Enable review queue tracking #141384 (Enable review queue tracking)
A variety of improvements to the codegen backends #141448 (A variety of improvements to the codegen backends)
avoid some usages of &mut P<T> in AST visitors #141636 (avoid some usages of &mut P<T> in AST visitors)
float: Disable total_cmp sNaN tests for f16 #141676 (float: Disable total_cmp sNaN tests for f16)
Add eslint as part of tidy run #141705 (Add eslint as part of tidy run)
Add loongarch64 with d feature to f32::midpoint fast path #141715 (Add loongarch64 with d feature to f32::midpoint fast path)
Provide secrets to try builds with new bors #141723 (Provide secrets to try builds with new bors)
Fix false documentation of FnCtxt::diverges #141728 (Fix false documentation of FnCtxt::diverges)
resolve target-libdir directly from rustc #141729 (resolve target-libdir directly from rustc)
creader: Remove extraenous String::clone #141732 (creader: Remove extraenous String::clone)

r? @ghost
@rustbot modify labels: rollup

There is an LLVM bug with lowering of basic `f16` operations that mean a round trip via `__extendhfsf2` and `__truncsfhf2` may happen for simple `abs` calls or bitcasts [1]. This is problematic because the round trip quiets signaling NaNs. For most operations this is acceptable, but it is causing `total_cmp` tests to fail unless optimizations are enabled. Disable `total_cmp` tests involving signaling NaNs until this issue is resolved. Fixes: rust-lang/rustc_codegen_cranelift#1578 Fixes: rust-lang#141503 [1]: llvm/llvm-project#104915

It is only relevant when using cg_ssa for driving compilation.

…a itself

There is no safety contract and I don't think any of them can actually cause UB in more ways than passing malicious source code to rustc can. While LtoModuleCodegen::optimize says that the returned ModuleCodegen points into the LTO module, the LTO module has already been dropped by the time this function returns, so if the returned ModuleCodegen indeed points into the LTO module, we would have seen crashes on every LTO compilation, which we don't. As such the comment is outdated.

This patch enables the optimized implementation of `f32::midpoint` for `loongarch64` targets that support the `d`feature. Targets with reliable 64-bit float support can safely use the faster and more accurate computation via `f64`, avoiding the fallback branchy version.

Leaving stage0 target-libdir resolution to rustc. This should also fix the issue with hard-coding `$sysroot/lib` which fails on systems that use `$sysroot/lib64` or `$sysroot/lib32`. Signed-off-by: onur-ozkan <work@onurozkan.dev>

The float modules in `std` are currently top-level but for `core`, they are nested within the `num` directory and referenced by `#[path = ...]`. For consistency, adjust `std` to use the same structure as `core`. Also change the `f16` and `f128` gates from outer attributes to inner attributes like `core` has.

… r=workingjubilee Make `std/src/num` mirror `core/src/num` The float modules in `std` are currently top-level but for `core`, they are nested within the `num` directory and referenced by `#[path = ...]`. For consistency, adjust `std` to use the same structure as `core`. Also change the `f16` and `f128` gates from outer attributes to inner attributes like `core` has.

…ouxu Enable review queue tracking This PR enables the new review queue tracking assignment logic implemented in triagebot. It is documented in rust-lang/rust-forge#853.

…Lapkin A variety of improvements to the codegen backends Some are just general improvements to cg_ssa or cg_llvm, while others will make it slightly easier to use cg_ssa in cg_clif in the future.

…x, r=petrochenkov avoid some usages of `&mut P<T>` in AST visitors It's a double indirection, and is also complicating our efforts at rust-lang#127615. r? `@ghost`

…workingjubilee float: Disable `total_cmp` sNaN tests for `f16` There is an LLVM bug with lowering of basic `f16` operations that mean a round trip via `__extendhfsf2` and `__truncsfhf2` may happen for simple `abs` calls or bitcasts [1]. This is problematic because the round trip quiets signaling NaNs. For most operations this is acceptable, but it is causing `total_cmp` tests to fail unless optimizations are enabled. Disable `total_cmp` tests involving signaling NaNs until this issue is resolved. Fixes: rust-lang/rustc_codegen_cranelift#1578 Fixes: rust-lang#141503 [1]: llvm/llvm-project#104915

Add eslint as part of `tidy` run Rustdoc uses `eslint` to run lints on the JS files. Currently you need to run it by hand since it's not part of any `x.py` command. This PR makes it part of `test tidy`. However, to prevent having all rust developers to install `npm` and `eslint`, I made it optional: if `eslint` is not installed, then the check is simply skipped (but will tell that it is being skipped). The second commit removes the manual checks from the docker file since `eslint` is run as part of tidy. cc `@lolbinarycat,` [#t-rustdoc > eslint seems to only be run in CI](https://rust-lang.zulipchat.com/#narrow/channel/266220-t-rustdoc/topic/eslint.20seems.20to.20only.20be.20run.20in.20CI/with/520761477)

…8472 Add `loongarch64` with `d` feature to `f32::midpoint` fast path This patch enables the optimized implementation of `f32::midpoint` for `loongarch64` targets that support the `d`feature. Targets with reliable 64-bit float support can safely use the faster and more accurate computation via `f64`, avoiding the fallback branchy version.

bors · 2025-05-29T15:03:17Z

📌 Commit 18646a8 has been approved by GuillaumeGomez

It is now in the queue for this repository.

bors · 2025-05-29T15:20:35Z

⌛ Testing commit 18646a8 with merge de6db15...

Rollup of 11 pull requests Successful merges: - #137574 (Make `std/src/num` mirror `core/src/num`) - #141384 (Enable review queue tracking) - #141448 (A variety of improvements to the codegen backends) - #141636 (avoid some usages of `&mut P<T>` in AST visitors) - #141676 (float: Disable `total_cmp` sNaN tests for `f16`) - #141705 (Add eslint as part of `tidy` run) - #141715 (Add `loongarch64` with `d` feature to `f32::midpoint` fast path) - #141723 (Provide secrets to try builds with new bors) - #141728 (Fix false documentation of FnCtxt::diverges) - #141729 (resolve target-libdir directly from rustc) - #141732 (creader: Remove extraenous String::clone) r? `@ghost` `@rustbot` modify labels: rollup

Kobzol · 2025-05-29T19:39:02Z

@bors ping

bors · 2025-05-29T19:39:04Z

😪 I'm awake I'm awake

Kobzol · 2025-05-29T19:39:45Z

@bors r-

Kobzol · 2025-05-29T19:44:14Z

Looks like bors forgot to merge this PR. Now it at least started testing another one. Reapproving.

@bors r=GuillaumeGomez

bors · 2025-05-29T19:44:17Z

📌 Commit 18646a8 has been approved by GuillaumeGomez

It is now in the queue for this repository.

bors · 2025-05-29T23:02:34Z

⌛ Testing commit 18646a8 with merge 1ac1950...

bors · 2025-05-30T02:20:09Z

☀️ Test successful - checks-actions
Approved by: GuillaumeGomez
Pushing 1ac1950 to master...

github-actions · 2025-05-30T02:22:59Z

What is this?

This is an experimental post-merge analysis report that shows differences in test outcomes between the merged PR and its parent PR.

Comparing 1bbd62e (parent) -> 1ac1950 (this PR)

Test differences

Show 744 test diffs

744 doctest diffs were found. These are ignored, as they are noisy.

Test dashboard

Run

cargo run --manifest-path src/ci/citool/Cargo.toml -- \
    test-dashboard 1ac1950c337039add1a83113ed6d1bd64bcb1142 --output-dir test-dashboard

And then open test-dashboard/index.html in your browser to see an overview of all executed tests.

Job duration changes

x86_64-apple-2: 5131.0s -> 6373.4s (24.2%)
dist-apple-various: 6106.8s -> 7396.2s (21.1%)
dist-aarch64-linux: 6553.9s -> 5413.8s (-17.4%)
aarch64-gnu: 8338.7s -> 6889.5s (-17.4%)
dist-x86_64-apple: 9516.4s -> 8151.3s (-14.3%)
x86_64-gnu-llvm-19-2: 6193.9s -> 6906.7s (11.5%)
aarch64-apple: 5657.1s -> 5092.4s (-10.0%)
dist-aarch64-apple: 5353.8s -> 5880.6s (9.8%)
dist-x86_64-netbsd: 5333.1s -> 4932.9s (-7.5%)
dist-riscv64-linux: 5434.2s -> 5057.8s (-6.9%)

How to interpret the job duration changes?

Job durations can vary a lot, based on the actual runner instance
that executed the job, system noise, invalidated caches, etc. The table above is provided
mostly for t-infra members, for simpler debugging of potential CI slow-downs.

rust-timer · 2025-05-30T02:23:16Z

📌 Perf builds for each rolled up PR:

PR#	Message	Perf Build Sha
#137574	Make `std/src/num` mirror `core/src/num`	`c7aa99712f9f4c4332dba6e9d0fb90b126505179` (link)
#141384	Enable review queue tracking	`95bd96a18fb218363c6c9e056bdf77cc1b536b15` (link)
#141448	A variety of improvements to the codegen backends	`6816a1766be0c53ffb72a08e4f70b779723b48f9` (link)
#141636	avoid some usages of `&mut P<T>` in AST visitors	`30c3979ac213ffccc30fa2213965c471c87b7f32` (link)
#141676	float: Disable `total_cmp` sNaN tests for `f16`	`59d217e5cd704147c96946d6d612d1f9d88b5d33` (link)
#141705	Add eslint as part of `tidy` run	`7a6e778394eb9c57998c8ed28330b74e9ed074e9` (link)
#141715	Add `loongarch64` with `d` feature to `f32::midpoint` fast …	`6b13d836a552899718d48306136f6a9f058b0b6f` (link)
#141723	Provide secrets to try builds with new bors	`e55c7efbdaee0f09cd293645212bd13b71fad821` (link)
#141728	Fix false documentation of FnCtxt::diverges	`27aa7e4a6113bfdb5eb486eff7808155e74c68ce` (link)
#141729	resolve target-libdir directly from rustc	`b955083d54fe26a1d61f74cdda9d8bd0b925dc5b` (link)
#141732	creader: Remove extraenous String::clone	`717757832a0c1b3ab34d5df7be596e5ecfaa74df` (link)

previous master: 1bbd62e547

In the case of a perf regression, run the following command for each PR you suspect might be the cause: @rust-timer build $SHA

rust-timer · 2025-05-30T03:41:59Z

Finished benchmarking commit (1ac1950): comparison URL.

Overall result: ❌✅ regressions and improvements - BENCHMARK(S) FAILED

Our benchmarks found a performance regression caused by this PR.
This might be an actual regression, but it can also be just noise.

Next Steps:

If the regression was expected or you think it can be justified,
please write a comment with sufficient written justification, and add
@rustbot label: +perf-regression-triaged to it, to mark the regression as triaged.
If you think that you know of a way to resolve the regression, try to create
a new PR with a fix for the regression.
If you do not understand the regression or you think that it is just noise,
you can ask the @rust-lang/wg-compiler-performance working group for help (members of this group
were already notified of this PR).

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

❗ ❗ ❗ ❗ ❗
Warning ⚠️: The following benchmark(s) failed to build:

rustc

❗ ❗ ❗ ❗ ❗

cc @rust-lang/wg-compiler-performance

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)	0.1%	[0.1%, 0.1%]	4
Regressions ❌ (secondary)	0.5%	[0.4%, 0.6%]	2
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-0.4%	[-0.5%, -0.4%]	2
All ❌✅ (primary)	0.1%	[0.1%, 0.1%]	4

Max RSS (memory usage)

Results (primary 1.3%, secondary 1.4%)

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%	[1.3%, 3.1%]	3
Regressions ❌ (secondary)	1.9%	[0.4%, 4.7%]	24
Improvements ✅ (primary)	-0.9%	[-0.9%, -0.9%]	1
Improvements ✅ (secondary)	-1.7%	[-2.9%, -0.5%]	4
All ❌✅ (primary)	1.3%	[-0.9%, 3.1%]	4

Cycles

Results (secondary 0.4%)

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)	1.2%	[0.4%, 2.6%]	11
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-1.0%	[-2.5%, -0.5%]	6
All ❌✅ (primary)	-	-	0

Binary size

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

Bootstrap: missing data
Artifact size: 370.31 MiB -> 370.32 MiB (0.00%)

matthiaskrgr · 2025-05-30T05:05:23Z

looks like #141729 broke benchmarking cc @onur-ozkan

onur-ozkan · 2025-05-30T05:41:10Z

Could you post the build log?

matthiaskrgr · 2025-05-30T05:45:20Z

https://perf.rust-lang.org/compare.html?start=1bbd62e547ba5cc08ccb44c27def3d33195d2dd5&end=1ac1950c337039add1a83113ed6d1bd64bcb1142&stat=instructions:u

Kobzol and others added 30 commits May 22, 2025 11:21

Enable review queue tracking

8bed646

Add eslint as part of tidy run

0527f02

Remove checks that are run with tidy

9e5dd51

Remove codegen_unit from MiscCodegenMethods

c593c01

The personality function is a Function, not a Value

5b0ab2c

Reduce amount of types that need to be PartialEq

a4cb1c7

Remove a couple of uses of interior mutability around statics

0fd257d

Move supports_parallel from CodegenBackend to ExtraBackendMethods

0809b41

It is only relevant when using cg_ssa for driving compilation.

Make predefine methods take &mut self

669e2ea

Remove methods from StaticCodegenMethods that are not called in cg_ss…

d7c0bde

…a itself

Remove unused arg_memory_ty method

865c7b9

avoid some usages of &mut P<T> in AST visitors

367a877

Provide secrets to try builds with new bors

059bc38

Fix false documentation

8e5d579

creader: Remove extraenous String::clone

8c8d2c2

resolve target-libdir directly from rustc

0d9f25b

Leaving stage0 target-libdir resolution to rustc. This should also fix the issue with hard-coding `$sysroot/lib` which fails on systems that use `$sysroot/lib64` or `$sysroot/lib32`. Signed-off-by: onur-ozkan <work@onurozkan.dev>

Install eslint in host-x86_64 Dockerfile

f0f661d

Centralize the eslint version between tidy and docker

b1723fc

Fix npm install error

8645ef7

Rollup merge of rust-lang#141384 - Kobzol:enable-review-prefs, r=jiey…

08bbf24

…ouxu Enable review queue tracking This PR enables the new review queue tracking assignment logic implemented in triagebot. It is documented in rust-lang/rust-forge#853.

Rollup merge of rust-lang#141448 - bjorn3:codegen_refactors, r=Waffle…

161cf3e

…Lapkin A variety of improvements to the codegen backends Some are just general improvements to cg_ssa or cg_llvm, while others will make it slightly easier to use cg_ssa in cg_clif in the future.

Rollup merge of rust-lang#141636 - fee1-dead-contrib:push-ntqvvxwuvrv…

2b08e4d

…x, r=petrochenkov avoid some usages of `&mut P<T>` in AST visitors It's a double indirection, and is also complicating our efforts at rust-lang#127615. r? `@ghost`

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 29, 2025

bors 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-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion. labels May 29, 2025

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-author Status: This is awaiting some action (such as code changes or more information) from the author. labels May 29, 2025

bors added the merged-by-bors This PR was explicitly merged by bors. label May 30, 2025

bors merged commit 1ac1950 into rust-lang:master May 30, 2025
10 checks passed

rustbot added this to the 1.89.0 milestone May 30, 2025

This was referenced May 30, 2025

Enable Non-determinism of float operations in Miri and change std tests #138062

Open

Note atan2 can return -PI #140487

Open

Add const support for float rounding methods #141521

Merged

rustbot added the perf-regression Performance regression. label May 30, 2025

GuillaumeGomez deleted the rollup-ivboqwd branch May 30, 2025 07:51

Rollup of 11 pull requests #141739

Rollup of 11 pull requests #141739

Uh oh!

Conversation

GuillaumeGomez commented May 29, 2025 • edited by rustbot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

bors commented May 29, 2025

Uh oh!

bors commented May 29, 2025

Uh oh!

Kobzol commented May 29, 2025

Uh oh!

bors commented May 29, 2025

Uh oh!

Kobzol commented May 29, 2025

Uh oh!

Kobzol commented May 29, 2025

Uh oh!

bors commented May 29, 2025

Uh oh!

bors commented May 29, 2025

Uh oh!

bors commented May 30, 2025

Uh oh!

Uh oh!

github-actions bot commented May 30, 2025

Test differences

Job duration changes

Uh oh!

rust-timer commented May 30, 2025

Uh oh!

rust-timer commented May 30, 2025

Overall result: ❌✅ regressions and improvements - BENCHMARK(S) FAILED

Instruction count

Max RSS (memory usage)

Cycles

Binary size

Uh oh!

matthiaskrgr commented May 30, 2025

Uh oh!

onur-ozkan commented May 30, 2025

Uh oh!

matthiaskrgr commented May 30, 2025

Uh oh!

Uh oh!

GuillaumeGomez commented May 29, 2025 •

edited by rustbot

Loading