Improve polymorphization of raw pointer formatting by hanna-kruppe · Pull Request #157714 · rust-lang/rust

hanna-kruppe · 2026-06-10T15:59:49Z

This reduces the amount of LLVM IR rustc generates for basically every use of fmt::Pointer by several hundred lines per pointee type. Most of the gain is from side-stepping a code size issue with the current implementation of Formatter::with_field. But even if those were fixed, it would still be an improvement:

The "is it a thin pointer?" check can't be eliminated during codegen, so rustc codegens both branches for every pointer type and LLVM has to clean it up later.
For wide pointers, the use of DebugStruct isn't a lot of code, but it's still a bit of extra code per wide pointer type that can easily be shared across all wide pointer types.

This reduces the amount of LLVM IR rustc generates for basically every use of `fmt::Pointer` by several hundred lines per pointee type. Most of the gain is from side-stepping a code size issue with the current implementation of `Formatter::with_field`. But even if those were fixed, it would still be an improvement: * The "is it a thin pointer?" check can't be eliminated during codegen, so rustc codegens *both* branches for every pointer type and LLVM has to clean it up later. * For wide pointers, the use of `DebugStruct` isn't a *lot* of code, but it's still a bit of extra code per wide pointer type that can easily be shared across all wide pointer types.

rustbot · 2026-06-10T15:59:55Z

Failed to set assignee to LawnGnome: req=https://api.github.com/repos/rust-lang/rust/issues/157714/assignees (x-github-request-id: 4B19:AA77B:90C467D:9232340:6A2989F8): {
"message": "Requires authentication",
"documentation_url": "https://docs.github.com/rest",
"status": "401"
}

Note: Only org members with at least the repository "read" role, users with write permissions, or people who have commented on the PR may be assigned.

rustbot · 2026-06-10T15:59:55Z

r? @LawnGnome

rustbot has assigned @LawnGnome.
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

Why was this reviewer chosen?

The reviewer was selected based on:

Owners of files modified in this PR: libs
libs expanded to 10 candidates
Random selection from LawnGnome, Mark-Simulacrum, clarfonthey, jhpratt

Kobzol · 2026-06-10T16:19:01Z

@bors try @rust-timer queue

rust-bors · 2026-06-10T16:19:30Z

❌ Encountered an error while executing command

Kobzol · 2026-06-12T08:35:40Z

@bors try

Improve polymorphization of raw pointer formatting

rust-bors · 2026-06-12T10:45:00Z

☀️ Try build successful (CI)
Build commit: 8a88eb6 (8a88eb633379915cf400b6d8ef34a80d1048e88b, parent: 09a371361240e42b0d69438fd1179efcf212e576)

rust-timer · 2026-06-12T11:25:28Z

Finished benchmarking commit (8a88eb6): comparison URL.

Overall result: ✅ improvements - no action needed

Benchmarking means the PR may be perf-sensitive. Consider adding rollup=never if this change is not fit for rolling up.

@rustbot label: -S-waiting-on-perf -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)	-0.2%	[-0.2%, -0.2%]	1
All ❌✅ (primary)	-	-	0

Max RSS (memory usage)

This perf run didn't have relevant results for this metric.

Cycles

Results (primary -2.3%, secondary -2.6%)

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)	-	-	0
Improvements ✅ (primary)	-2.3%	[-2.3%, -2.3%]	1
Improvements ✅ (secondary)	-2.6%	[-2.6%, -2.6%]	1
All ❌✅ (primary)	-2.3%	[-2.3%, -2.3%]	1

Binary size

This perf run didn't have relevant results for this metric.

Bootstrap: 517.481s -> 518.904s (0.27%)
Artifact size: 401.31 MiB -> 401.37 MiB (0.02%)

LawnGnome · 2026-06-12T15:17:54Z

Makes sense to me. Nice work!

@bors r+ rollup=never (given perf implications)

rust-bors · 2026-06-12T15:17:57Z

📌 Commit c043f23 has been approved by LawnGnome

It is now in the queue for this repository.

Improve polymorphization of raw pointer formatting This reduces the amount of LLVM IR rustc generates for basically every use of `fmt::Pointer` by several hundred lines per pointee type. Most of the gain is from side-stepping a code size issue with the current implementation of `Formatter::with_field`. But even if those were fixed, it would still be an improvement: * The "is it a thin pointer?" check can't be eliminated during codegen, so rustc codegens *both* branches for every pointer type and LLVM has to clean it up later. * For wide pointers, the use of `DebugStruct` isn't a *lot* of code, but it's still a bit of extra code per wide pointer type that can easily be shared across all wide pointer types.

rust-bors · 2026-06-13T06:50:36Z

💔 Test for d390a19 failed: CI. Failed job:

auto - test-various (web logs, enhanced plaintext logs)

rust-log-analyzer · 2026-06-13T07:00:35Z

A job failed! Check out the build log: (web) (plain enhanced) (plain)

Click to see the possible cause of the failure (guessed by this bot)

  IMAGE: test-various
##[endgroup]
    Updating crates.io index
error: failed to get `adler2` as a dependency of package `miniz_oxide v0.8.8`
    ... which satisfies dependency `miniz_oxide = "^0.8.5"` of package `flate2 v1.1.9`
    ... which satisfies dependency `flate2 = "^1.1.9"` of package `citool v0.1.0 (/home/runner/work/rust/rust/src/ci/citool)`

Caused by:
  failed to load source for dependency `adler2`

Caused by:
  unable to update registry `crates-io`

Caused by:
  download of ad/le/adler2 failed

Caused by:
  curl failed

Caused by:

JonathanBrouwer · 2026-06-13T07:48:31Z

@bors retry

JonathanBrouwer · 2026-06-13T07:49:18Z

@bors rollup=iffy
Perf is most likely noise

rust-bors · 2026-06-13T07:58:14Z

⌛ Testing commit c043f23 with merge be2c792...

Workflow: https://github.com/rust-lang/rust/actions/runs/27461008959

Improve polymorphization of raw pointer formatting This reduces the amount of LLVM IR rustc generates for basically every use of `fmt::Pointer` by several hundred lines per pointee type. Most of the gain is from side-stepping a code size issue with the current implementation of `Formatter::with_field`. But even if those were fixed, it would still be an improvement: * The "is it a thin pointer?" check can't be eliminated during codegen, so rustc codegens *both* branches for every pointer type and LLVM has to clean it up later. * For wide pointers, the use of `DebugStruct` isn't a *lot* of code, but it's still a bit of extra code per wide pointer type that can easily be shared across all wide pointer types.

jhpratt · 2026-06-13T08:06:49Z

yielding to encompassing rollup

@bors yield

rust-bors · 2026-06-13T08:06:52Z

Auto build was cancelled. Cancelled workflows:

https://github.com/rust-lang/rust/actions/runs/27461008959

The next pull request likely to be tested is #157852.

Rollup of 10 pull requests Successful merges: - #157714 (Improve polymorphization of raw pointer formatting) - #157653 (remove AliasTerm::def_id()) - #157773 (Remove AnonConstKind::GCA and reject generic anon consts) - #157846 (Avoid ICE in unconstrained generic parameter suggestion for enums) - #157594 (note which target a missing target_feature belongs to) - #157740 (rustdoc: correct doctest span for trailing semicolon after item) - #157782 (Added `PhantomPinned` diagnostic item and prevented dead field warning on `PhantomPinned`) - #157802 (Ensure that optimize attributes on closures are inherited by the shim.) - #157812 (rustc_public: implement `ty::Alias` conversion) - #157833 (Update wasm-component-ld to 0.5.25)

rust-bors · 2026-06-13T08:10:04Z

⌛ Testing commit c043f23 with merge c180c64...

Workflow: https://github.com/rust-lang/rust/actions/runs/27461267747

Improve polymorphization of raw pointer formatting This reduces the amount of LLVM IR rustc generates for basically every use of `fmt::Pointer` by several hundred lines per pointee type. Most of the gain is from side-stepping a code size issue with the current implementation of `Formatter::with_field`. But even if those were fixed, it would still be an improvement: * The "is it a thin pointer?" check can't be eliminated during codegen, so rustc codegens *both* branches for every pointer type and LLVM has to clean it up later. * For wide pointers, the use of `DebugStruct` isn't a *lot* of code, but it's still a bit of extra code per wide pointer type that can easily be shared across all wide pointer types.

JonathanBrouwer · 2026-06-13T08:14:07Z

@bors yield
Yielding to enclosing rollup

rust-bors · 2026-06-13T08:14:11Z

Auto build was cancelled. Cancelled workflows:

https://github.com/rust-lang/rust/actions/runs/27461267747

The next pull request likely to be tested is #157852.

Rollup of 10 pull requests Successful merges: - #157714 (Improve polymorphization of raw pointer formatting) - #157653 (remove AliasTerm::def_id()) - #157773 (Remove AnonConstKind::GCA and reject generic anon consts) - #157846 (Avoid ICE in unconstrained generic parameter suggestion for enums) - #157594 (note which target a missing target_feature belongs to) - #157740 (rustdoc: correct doctest span for trailing semicolon after item) - #157782 (Added `PhantomPinned` diagnostic item and prevented dead field warning on `PhantomPinned`) - #157802 (Ensure that optimize attributes on closures are inherited by the shim.) - #157812 (rustc_public: implement `ty::Alias` conversion) - #157833 (Update wasm-component-ld to 0.5.25)

Rollup merge of #157714 - hanna-kruppe:fmt-pointer, r=LawnGnome Improve polymorphization of raw pointer formatting This reduces the amount of LLVM IR rustc generates for basically every use of `fmt::Pointer` by several hundred lines per pointee type. Most of the gain is from side-stepping a code size issue with the current implementation of `Formatter::with_field`. But even if those were fixed, it would still be an improvement: * The "is it a thin pointer?" check can't be eliminated during codegen, so rustc codegens *both* branches for every pointer type and LLVM has to clean it up later. * For wide pointers, the use of `DebugStruct` isn't a *lot* of code, but it's still a bit of extra code per wide pointer type that can easily be shared across all wide pointer types.

rustbot assigned LawnGnome Jun 10, 2026

rustbot added S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. T-libs Relevant to the library team, which will review and decide on the PR/issue. labels Jun 10, 2026

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 Jun 10, 2026

hanna-kruppe changed the title ~~improve polymorphization of raw pointer formatting~~ Improve polymorphization of raw pointer formatting Jun 10, 2026

This comment has been minimized.

Sign in to view

rust-bors Bot pushed a commit that referenced this pull request Jun 12, 2026

Auto merge of #157714 - hanna-kruppe:fmt-pointer, r=<try>

8a88eb6

Improve polymorphization of raw pointer formatting

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 Jun 12, 2026

rust-bors Bot 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 Jun 12, 2026

This comment has been minimized.

Sign in to view

rust-bors Bot 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 Jun 13, 2026

rust-bors Bot added the S-waiting-on-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion. label Jun 13, 2026

rust-bors Bot removed the S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. label Jun 13, 2026

jhpratt mentioned this pull request Jun 13, 2026

Rollup of 10 pull requests #157852

Merged

rust-bors Bot merged commit 7b303c6 into rust-lang:main Jun 13, 2026
13 of 14 checks passed

rustbot added this to the 1.98.0 milestone Jun 13, 2026

Uh oh!

Conversation

hanna-kruppe commented Jun 10, 2026 • edited by rustbot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

rustbot commented Jun 10, 2026

Uh oh!

rustbot commented Jun 10, 2026

Uh oh!

Kobzol commented Jun 10, 2026

Uh oh!

This comment has been minimized.

rust-bors Bot commented Jun 10, 2026

Uh oh!

Kobzol commented Jun 12, 2026

Uh oh!

This comment has been minimized.

rust-bors Bot commented Jun 12, 2026

Uh oh!

This comment has been minimized.

rust-timer commented Jun 12, 2026

Overall result: ✅ improvements - no action needed

Instruction count

Max RSS (memory usage)

Cycles

Binary size

Uh oh!

LawnGnome commented Jun 12, 2026

Uh oh!

rust-bors Bot commented Jun 12, 2026

Uh oh!

This comment has been minimized.

rust-bors Bot commented Jun 13, 2026

Uh oh!

rust-log-analyzer commented Jun 13, 2026

Uh oh!

JonathanBrouwer commented Jun 13, 2026

Uh oh!

JonathanBrouwer commented Jun 13, 2026

Uh oh!

rust-bors Bot commented Jun 13, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

jhpratt commented Jun 13, 2026

Uh oh!

rust-bors Bot commented Jun 13, 2026

Uh oh!

rust-bors Bot commented Jun 13, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

JonathanBrouwer commented Jun 13, 2026

Uh oh!

rust-bors Bot commented Jun 13, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

8 participants

hanna-kruppe commented Jun 10, 2026 •

edited by rustbot

Loading

rust-bors Bot commented Jun 13, 2026 •

edited

Loading

rust-bors Bot commented Jun 13, 2026 •

edited

Loading