Pass `alloc-variant-zeroed` to LLVM #144086

clubby789 · 2025-07-17T20:51:56Z

Makes use of llvm/llvm-project#138299 (once we pull in a version of LLVM with this attribute). ~~Unfortunately also requires llvm/llvm-project#149336 to work.~~

Closes #104847

bors · 2025-07-22T23:05:02Z

☔ The latest upstream changes (presumably #144249) made this pull request unmergeable. Please resolve the merge conflicts.

rustbot · 2025-08-13T13:30:35Z

r? @fee1-dead

rustbot has assigned @fee1-dead.
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

clubby789 · 2025-08-13T13:30:38Z

r? compiler

edit: oops I guess that's automatic 😅

nikic · 2025-08-13T14:08:33Z

@bors try @rust-timer queue

Pass `alloc-variant-zeroed` to LLVM

rust-bors · 2025-08-13T14:36:24Z

💔 Test for f993fda failed: CI. Failed jobs:

try - dist-x86_64-linux (web logs, extended logs)

nikic · 2025-08-13T16:44:07Z

I think the alloc symbols are mangled nowadays, so we need to pass the symbols with the proper mangling.

clubby789 · 2025-08-20T00:35:01Z

@bors retry

Pass `alloc-variant-zeroed` to LLVM Makes use of llvm/llvm-project#138299 (once we pull in a version of LLVM with this attribute). ~~Unfortunately also requires llvm/llvm-project#149336 to work.~~ Closes #104847

bors · 2025-08-20T03:34:16Z

⌛ Testing commit 3e9316a with merge 3372b83...

bors · 2025-08-20T04:34:36Z

💔 Test failed - checks-actions

nikic · 2025-08-20T10:52:31Z

Looks like the call to @"_ZN62_$LT$T$u20$as$u20$alloc..vec..spec_from_elem..SpecFromElem$GT$9from_elem17h22fb7d61224751b4E" in @vec_zero_tuple doesn't get inlined on wasm-wasip1 for whatever reason. Should probably restore the only-x86_64.

clubby789 · 2025-08-20T16:10:23Z

@bors retry

nikic · 2025-08-20T16:13:26Z

@bors r+

(retry is only if you make no changes)

bors · 2025-08-20T16:13:29Z

📌 Commit 8ea3b09 has been approved by nikic

It is now in the queue for this repository.

bors · 2025-08-20T17:16:38Z

⌛ Testing commit 8ea3b09 with merge 040a98a...

bors · 2025-08-20T20:28:54Z

☀️ Test successful - checks-actions
Approved by: nikic
Pushing 040a98a to master...

github-actions · 2025-08-20T20:32:07Z

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 e8a792d (parent) -> 040a98a (this PR)

Test differences

Show 18 test diffs

Stage 1

[codegen] tests/codegen-llvm/vec-calloc.rs#llvm21: [missing] -> ignore (ignored when the LLVM version 20.1.2 is older than 21.0.0) (J1)
[codegen] tests/codegen-llvm/vec-calloc.rs: pass -> [missing] (J4)
[codegen] tests/codegen-llvm/vec-calloc.rs#normal: [missing] -> pass (J4)
[codegen] tests/codegen-llvm/vec-calloc.rs#llvm21: [missing] -> ignore (ignored when the LLVM version 19.1.1 is older than 21.0.0) (J7)

Stage 2

[codegen] tests/codegen-llvm/vec-calloc.rs#llvm21: [missing] -> ignore (ignored when the LLVM version 20.1.2 is older than 21.0.0) (J0)
[codegen] tests/codegen-llvm/vec-calloc.rs#llvm21: [missing] -> ignore (ignored when the LLVM version 19.1.1 is older than 21.0.0) (J2)
[codegen] tests/codegen-llvm/vec-calloc.rs#llvm21: [missing] -> pass (J3)
[codegen] tests/codegen-llvm/vec-calloc.rs: ignore (only executed when the architecture is x86_64) -> [missing] (J5)
[codegen] tests/codegen-llvm/vec-calloc.rs#llvm21: [missing] -> ignore (only executed when the architecture is x86_64) (J5)
[codegen] tests/codegen-llvm/vec-calloc.rs#normal: [missing] -> ignore (only executed when the architecture is x86_64) (J5)
[codegen] tests/codegen-llvm/vec-calloc.rs: pass -> [missing] (J6)
[codegen] tests/codegen-llvm/vec-calloc.rs#normal: [missing] -> pass (J6)

Additionally, 6 doctest diffs were found. These are ignored, as they are noisy.

Job group index

J0: x86_64-gnu-llvm-20-2
J1: x86_64-gnu-llvm-20-3
J2: x86_64-gnu-llvm-19-2
J3: test-various, x86_64-gnu, x86_64-gnu-nopt, x86_64-gnu-stable, x86_64-mingw-1, x86_64-msvc-1
J4: x86_64-gnu-llvm-19-3, x86_64-gnu-llvm-20-3
J5: aarch64-apple, aarch64-gnu, aarch64-gnu-llvm-19-1, aarch64-msvc-1, arm-android, armhf-gnu, dist-i586-gnu-i586-i686-musl, i686-gnu-1, i686-gnu-nopt-1, i686-msvc-1
J6: test-various, x86_64-gnu, x86_64-gnu-llvm-19-2, x86_64-gnu-llvm-20-2, x86_64-gnu-nopt, x86_64-gnu-stable, x86_64-mingw-1, x86_64-msvc-1
J7: x86_64-gnu-llvm-19-3

Test dashboard

Run

cargo run --manifest-path src/ci/citool/Cargo.toml -- \
    test-dashboard 040a98af70f0a7da03f3d5356531b28a2a7a77e4 --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

dist-aarch64-linux: 6499.1s -> 8945.1s (37.6%)
i686-msvc-2: 7878.1s -> 9200.7s (16.8%)
x86_64-gnu-llvm-19: 2500.1s -> 2889.0s (15.6%)
dist-aarch64-msvc: 6223.6s -> 5256.9s (-15.5%)
pr-check-1: 1396.0s -> 1583.7s (13.4%)
i686-gnu-2: 5456.9s -> 6097.5s (11.7%)
dist-apple-various: 4804.6s -> 4248.7s (-11.6%)
x86_64-gnu-tools: 3409.6s -> 3803.3s (11.5%)
aarch64-gnu-llvm-19-2: 2177.7s -> 2413.3s (10.8%)
aarch64-gnu: 6585.3s -> 7219.1s (9.6%)

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 · 2025-08-20T22:46:36Z

Finished benchmarking commit (040a98a): comparison URL.

Overall result: ✅ improvements - no action needed

@rustbot label: -perf-regression

Instruction count

Our most reliable metric. Used to determine the overall result above. However, even this metric can be noisy.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-0.3%	[-0.4%, -0.2%]	7
All ❌✅ (primary)	-	-	0

Max RSS (memory usage)

Results (secondary 2.4%)

A less reliable metric. May be of interest, but not used to determine the overall result above.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	2.4%	[2.4%, 2.4%]	1
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-	-	0

Cycles

Results (secondary -0.8%)

A less reliable metric. May be of interest, but not used to determine the overall result above.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	3.6%	[3.2%, 4.0%]	2
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-3.7%	[-4.4%, -3.1%]	3
All ❌✅ (primary)	-	-	0

Binary size

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

Bootstrap: 471.66s -> 471.047s (-0.13%)
Artifact size: 378.20 MiB -> 378.31 MiB (0.03%)

Pass `alloc-variant-zeroed` to LLVM Makes use of llvm/llvm-project#138299 (once we pull in a version of LLVM with this attribute). ~~Unfortunately also requires llvm/llvm-project#149336 to work.~~ Closes rust-lang#104847

aeubanks · 2025-08-28T20:54:28Z

I believe this causing https://crbug.com/441524277. In summary, the rust calloc function _RNvCsfo6nMSlQo1l_7___rustc19___rust_alloc_zeroed definition does not have the alloc-family attribute, but user modules of it have the alloc-family attribute on the calloc declaration. however, if ThinLTO decides to import the function, we lose the alloc-family attribute and the verifier complains 'alloc-variant-zeroed' must name a function belonging to the same 'alloc-family'.

rust-lang/rust#144086 is causing crashes in ThinLTO builds, temporarily revert while we figure out what to do. Bug: 441524277 Change-Id: I8f68aeac9d582678443376c9195475e79b49c233 Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/6896856 Reviewed-by: Zequan Wu <zequanwu@google.com> Commit-Queue: Arthur Eubanks <aeubanks@google.com> Cr-Commit-Position: refs/heads/main@{#1508072}

nikic · 2025-09-01T08:28:45Z

@aeubanks Where does the definition without the attributes come from? Is this rustc failing to generate them, or is this something your specific setup is injecting?

clubby789 added the S-blocked Status: Blocked on something else such as an RFC or other implementation work. label Jul 17, 2025

clubby789 force-pushed the alloc-zeroed branch from 124411e to a3712ef Compare July 23, 2025 14:40

rust-cloud-vms bot force-pushed the alloc-zeroed branch from a3712ef to 4b8de0d Compare August 13, 2025 13:27

clubby789 removed the S-blocked Status: Blocked on something else such as an RFC or other implementation work. label Aug 13, 2025

clubby789 marked this pull request as ready for review August 13, 2025 13:30

rustbot assigned fee1-dead Aug 13, 2025

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 Aug 13, 2025

rustbot assigned lcnr and unassigned fee1-dead Aug 13, 2025

rust-cloud-vms bot force-pushed the alloc-zeroed branch from 4b8de0d to 5cf5610 Compare August 13, 2025 13:46

This comment has been minimized.

Sign in to view

rust-bors bot added a commit that referenced this pull request Aug 13, 2025

Auto merge of #144086 - clubby789:alloc-zeroed, r=<try>

f993fda

Pass `alloc-variant-zeroed` to LLVM

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 Aug 13, 2025

This comment has been minimized.

Sign in to view

clubby789 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 Aug 13, 2025

rust-cloud-vms bot force-pushed the alloc-zeroed branch from 5cf5610 to 3e9316a Compare August 13, 2025 17:25

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 Aug 20, 2025

This comment has been minimized.

Sign in to view

bors added S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. and removed S-waiting-on-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion. labels Aug 20, 2025

Pass alloc-variant-zeroed to LLVM

8ea3b09

clubby789 force-pushed the alloc-zeroed branch from 3e9316a to 8ea3b09 Compare August 20, 2025 16:09

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 Aug 20, 2025

bors added the merged-by-bors This PR was explicitly merged by bors. label Aug 20, 2025

bors merged commit 040a98a into rust-lang:master Aug 20, 2025
11 checks passed

rustbot added this to the 1.91.0 milestone Aug 20, 2025

rustbot removed the perf-regression Performance regression. label Aug 20, 2025

clubby789 mentioned this pull request Aug 29, 2025

alloc-variant-zeroed causing issues with LTO #145995

Open

Pass alloc-variant-zeroed to LLVM #144086

Pass alloc-variant-zeroed to LLVM #144086

Conversation

clubby789 commented Jul 17, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

bors commented Jul 22, 2025

Uh oh!

rustbot commented Aug 13, 2025

Uh oh!

clubby789 commented Aug 13, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

nikic commented Aug 13, 2025

Uh oh!

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

rust-bors bot commented Aug 13, 2025

Uh oh!

nikic commented Aug 13, 2025

Uh oh!

clubby789 commented Aug 20, 2025

Uh oh!

bors commented Aug 20, 2025

Uh oh!

This comment has been minimized.

bors commented Aug 20, 2025

Uh oh!

nikic commented Aug 20, 2025

Uh oh!

clubby789 commented Aug 20, 2025

Uh oh!

nikic commented Aug 20, 2025

Uh oh!

bors commented Aug 20, 2025

Uh oh!

bors commented Aug 20, 2025

Uh oh!

bors commented Aug 20, 2025

Uh oh!

Uh oh!

github-actions bot commented Aug 20, 2025

Test differences

Stage 1

Stage 2

Job duration changes

Uh oh!

rust-timer commented Aug 20, 2025

Overall result: ✅ improvements - no action needed

Instruction count

Max RSS (memory usage)

Cycles

Binary size

Uh oh!

aeubanks commented Aug 28, 2025

Uh oh!

nikic commented Sep 1, 2025

Uh oh!

Uh oh!

Pass `alloc-variant-zeroed` to LLVM #144086

Pass `alloc-variant-zeroed` to LLVM #144086

clubby789 commented Jul 17, 2025 •

edited

Loading

clubby789 commented Aug 13, 2025 •

edited

Loading