rustc_target: callconv: powerpc64: Use llvm_abiname rather than target_abi for ABI determination by Gelbpunkt · Pull Request #153035 · rust-lang/rust

Gelbpunkt · 2026-02-24T06:44:52Z

Currently on PowerPC64 targets, llvm_abiname and target_abi will be the same unless we're on AIX. Since llvm_abiname is what we pass on to LLVM, it is preferable to use the value of that to determine the calling convention rather than target_abi.

All PowerPC64 targets set both llvm_abiname and target_abi to the respective ELF ABIs, with the exception of AIX. This is a non-functional change.

Noticed this in the follow-up discussion from #150468 and also requested by @RalfJung here.

r? @RalfJung

rustbot · 2026-02-24T06:44:57Z

RalfJung is not on the review rotation at the moment.
They may take a while to respond.

…t_abi for ABI determination Currently on PowerPC64 targets, llvm_abiname and target_abi will be the same unless we're on AIX. Since llvm_abiname is what we pass on to LLVM, it is preferable to use the value of that to determine the calling convention rather than target_abi. All PowerPC64 targets set both llvm_abiname and target_abi to the respective ELF ABIs, with the exception of AIX. This is a non-functional change.

RalfJung · 2026-02-24T07:59:48Z

LGTM, thanks! r=me when CI is green.

@bors delegate+

rust-bors · 2026-02-24T07:59:52Z

✌️ @Gelbpunkt, you can now approve this pull request!

If @RalfJung told you to "r=me" after making some further change, then please make that change and post @bors r=RalfJung.

Gelbpunkt · 2026-02-25T10:14:55Z

CI is green without changes, so:

@bors r=RalfJung

rust-bors · 2026-02-25T10:14:58Z

📌 Commit d501f96 has been approved by RalfJung

It is now in the queue for this repository.

@RalfJung

…ame, r=RalfJung rustc_target: callconv: powerpc64: Use llvm_abiname rather than target_abi for ABI determination Currently on PowerPC64 targets, `llvm_abiname` and `target_abi` will be the same unless we're on AIX. Since `llvm_abiname` is what we pass on to LLVM, it is preferable to use the value of that to determine the calling convention rather than `target_abi`. All PowerPC64 targets set both `llvm_abiname` and `target_abi` to the respective ELF ABIs, with the exception of AIX. This is a non-functional change. Noticed this in the follow-up discussion from rust-lang#150468 and also requested by @RalfJung [here](rust-lang#150468 (comment)). r? @RalfJung

@RalfJung

…ame, r=RalfJung rustc_target: callconv: powerpc64: Use llvm_abiname rather than target_abi for ABI determination Currently on PowerPC64 targets, `llvm_abiname` and `target_abi` will be the same unless we're on AIX. Since `llvm_abiname` is what we pass on to LLVM, it is preferable to use the value of that to determine the calling convention rather than `target_abi`. All PowerPC64 targets set both `llvm_abiname` and `target_abi` to the respective ELF ABIs, with the exception of AIX. This is a non-functional change. Noticed this in the follow-up discussion from rust-lang#150468 and also requested by @RalfJung [here](rust-lang#150468 (comment)). r? @RalfJung

@RalfJung

…ame, r=RalfJung rustc_target: callconv: powerpc64: Use llvm_abiname rather than target_abi for ABI determination Currently on PowerPC64 targets, `llvm_abiname` and `target_abi` will be the same unless we're on AIX. Since `llvm_abiname` is what we pass on to LLVM, it is preferable to use the value of that to determine the calling convention rather than `target_abi`. All PowerPC64 targets set both `llvm_abiname` and `target_abi` to the respective ELF ABIs, with the exception of AIX. This is a non-functional change. Noticed this in the follow-up discussion from rust-lang#150468 and also requested by @RalfJung [here](rust-lang#150468 (comment)). r? @RalfJung

@RalfJung

…ame, r=RalfJung rustc_target: callconv: powerpc64: Use llvm_abiname rather than target_abi for ABI determination Currently on PowerPC64 targets, `llvm_abiname` and `target_abi` will be the same unless we're on AIX. Since `llvm_abiname` is what we pass on to LLVM, it is preferable to use the value of that to determine the calling convention rather than `target_abi`. All PowerPC64 targets set both `llvm_abiname` and `target_abi` to the respective ELF ABIs, with the exception of AIX. This is a non-functional change. Noticed this in the follow-up discussion from rust-lang#150468 and also requested by @RalfJung [here](rust-lang#150468 (comment)). r? @RalfJung

…uwer Rollup of 13 pull requests Successful merges: - #153079 (Revert "Move aarch64-apple dist builder to dynamic llvm linking") - #148146 (CI: use alternative disks if available) - #149937 (spliit out `linker-info` from `linker-messages`) - #151771 (Fix: On wasm targets, call `panic_in_cleanup` if panic occurs in cleanup) - #153035 (rustc_target: callconv: powerpc64: Use llvm_abiname rather than target_abi for ABI determination) - #153075 (mGCA: Lower negated literals directly and reject non-integer negations) - #153078 (Remove `QuerySystemFns`) - #149978 (deprecate `Eq::assert_receiver_is_total_eq` and emit FCW on manual impls) - #153029 (Rename `rustc::pass_by_value` lint as `rustc::disallowed_pass_by_ref`.) - #153063 (`is_ty_must_use`: do not require a `span` argument) - #153071 (Update books) - #153092 (Remove redundant self usages) - #153094 (Simplify `AppendOnlyVec` iterators) Failed merges: - #153091 (Migration of `LintDiagnostic` - part 4)

Rollup of 13 pull requests Successful merges: - #148146 (CI: use alternative disks if available) - #151771 (Fix: On wasm targets, call `panic_in_cleanup` if panic occurs in cleanup) - #153035 (rustc_target: callconv: powerpc64: Use llvm_abiname rather than target_abi for ABI determination) - #153075 (mGCA: Lower negated literals directly and reject non-integer negations) - #153078 (Remove `QuerySystemFns`) - #153089 (interpret: avoid dummy spans in the stacktrace) - #153111 (Refactor url_parts to return is_absolute instead of out param) - #149978 (deprecate `Eq::assert_receiver_is_total_eq` and emit FCW on manual impls) - #153029 (Rename `rustc::pass_by_value` lint as `rustc::disallowed_pass_by_ref`.) - #153063 (`is_ty_must_use`: do not require a `span` argument) - #153071 (Update books) - #153092 (Remove redundant self usages) - #153094 (Simplify `AppendOnlyVec` iterators) Failed merges: - #153091 (Migration of `LintDiagnostic` - part 4)

@RalfJung

Rollup merge of #153035 - Gelbpunkt:ppc64-callconv-llvm-abiname, r=RalfJung rustc_target: callconv: powerpc64: Use llvm_abiname rather than target_abi for ABI determination Currently on PowerPC64 targets, `llvm_abiname` and `target_abi` will be the same unless we're on AIX. Since `llvm_abiname` is what we pass on to LLVM, it is preferable to use the value of that to determine the calling convention rather than `target_abi`. All PowerPC64 targets set both `llvm_abiname` and `target_abi` to the respective ELF ABIs, with the exception of AIX. This is a non-functional change. Noticed this in the follow-up discussion from #150468 and also requested by @RalfJung [here](#150468 (comment)). r? @RalfJung

Rollup of 13 pull requests Successful merges: - rust-lang/rust#148146 (CI: use alternative disks if available) - rust-lang/rust#151771 (Fix: On wasm targets, call `panic_in_cleanup` if panic occurs in cleanup) - rust-lang/rust#153035 (rustc_target: callconv: powerpc64: Use llvm_abiname rather than target_abi for ABI determination) - rust-lang/rust#153075 (mGCA: Lower negated literals directly and reject non-integer negations) - rust-lang/rust#153078 (Remove `QuerySystemFns`) - rust-lang/rust#153089 (interpret: avoid dummy spans in the stacktrace) - rust-lang/rust#153111 (Refactor url_parts to return is_absolute instead of out param) - rust-lang/rust#149978 (deprecate `Eq::assert_receiver_is_total_eq` and emit FCW on manual impls) - rust-lang/rust#153029 (Rename `rustc::pass_by_value` lint as `rustc::disallowed_pass_by_ref`.) - rust-lang/rust#153063 (`is_ty_must_use`: do not require a `span` argument) - rust-lang/rust#153071 (Update books) - rust-lang/rust#153092 (Remove redundant self usages) - rust-lang/rust#153094 (Simplify `AppendOnlyVec` iterators) Failed merges: - rust-lang/rust#153091 (Migration of `LintDiagnostic` - part 4)

JonathanBrouwer · 2026-02-26T21:31:25Z

@rust-timer build a52ed29

rust-timer · 2026-02-27T00:03:07Z

Finished benchmarking commit (a52ed29): comparison URL.

Overall result: no relevant changes - no action needed

Benchmarking this pull request means it may be perf-sensitive – we'll automatically label it not fit for rolling up. You can override this, but we strongly advise not to, due to possible changes in compiler perf.

@bors rollup=never
@rustbot label: -S-waiting-on-perf -perf-regression

Instruction count

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

Max RSS (memory usage)

Results (primary -3.9%, secondary 6.1%)

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)	6.1%	[4.5%, 7.7%]	2
Improvements ✅ (primary)	-3.9%	[-3.9%, -3.9%]	1
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-3.9%	[-3.9%, -3.9%]	1

Cycles

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

Binary size

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

Bootstrap: 492.161s -> 481.091s (-2.25%)
Artifact size: 395.78 MiB -> 397.79 MiB (0.51%)

Rollup of 13 pull requests Successful merges: - rust-lang/rust#148146 (CI: use alternative disks if available) - rust-lang/rust#151771 (Fix: On wasm targets, call `panic_in_cleanup` if panic occurs in cleanup) - rust-lang/rust#153035 (rustc_target: callconv: powerpc64: Use llvm_abiname rather than target_abi for ABI determination) - rust-lang/rust#153075 (mGCA: Lower negated literals directly and reject non-integer negations) - rust-lang/rust#153078 (Remove `QuerySystemFns`) - rust-lang/rust#153089 (interpret: avoid dummy spans in the stacktrace) - rust-lang/rust#153111 (Refactor url_parts to return is_absolute instead of out param) - rust-lang/rust#149978 (deprecate `Eq::assert_receiver_is_total_eq` and emit FCW on manual impls) - rust-lang/rust#153029 (Rename `rustc::pass_by_value` lint as `rustc::disallowed_pass_by_ref`.) - rust-lang/rust#153063 (`is_ty_must_use`: do not require a `span` argument) - rust-lang/rust#153071 (Update books) - rust-lang/rust#153092 (Remove redundant self usages) - rust-lang/rust#153094 (Simplify `AppendOnlyVec` iterators) Failed merges: - rust-lang/rust#153091 (Migration of `LintDiagnostic` - part 4)

@Darksonn

prefer actual ABI-controling fields over target.abi when making ABI decisions We don't actually check that `abi` is consistent with the fields that control the ABI, e.g. one could set `llvm_abiname` to "ilp32e" on a riscv target without setting a matching `abi`. So, if we need to make actual decisions, better to use the source of truth we forward to LLVM than the informational string we forward to the user. This is a breaking change for aarch64 JSON target specs: setting `abi` to "softfloat" is no longer enough; one has to also set `rustc_abi` to "softfloat". That is consistent with riscv and arm32, but it's still surprising. Cc @Darksonn in case this affects the Linux kernel. Also see rust-lang#153035 which does something similar for PowerPC, and [Zulip](https://rust-lang.zulipchat.com/#narrow/channel/131828-t-compiler/topic/De-spaghettifying.20ABI.20controls/with/575095372). Happy to delay this PR if someone has a better idea. Cc @folkertdev @workingjubilee

@Darksonn

prefer actual ABI-controling fields over target.abi when making ABI decisions We don't actually check that `abi` is consistent with the fields that control the ABI, e.g. one could set `llvm_abiname` to "ilp32e" on a riscv target without setting a matching `abi`. So, if we need to make actual decisions, better to use the source of truth we forward to LLVM than the informational string we forward to the user. This is a breaking change for aarch64 JSON target specs: setting `abi` to "softfloat" is no longer enough; one has to also set `rustc_abi` to "softfloat". That is consistent with riscv and arm32, but it's still surprising. Cc @Darksonn in case this affects the Linux kernel. Also see rust-lang#153035 which does something similar for PowerPC, and [Zulip](https://rust-lang.zulipchat.com/#narrow/channel/131828-t-compiler/topic/De-spaghettifying.20ABI.20controls/with/575095372). Happy to delay this PR if someone has a better idea. Cc @folkertdev @workingjubilee

@Darksonn

prefer actual ABI-controling fields over target.abi when making ABI decisions We don't actually check that `abi` is consistent with the fields that control the ABI, e.g. one could set `llvm_abiname` to "ilp32e" on a riscv target without setting a matching `abi`. So, if we need to make actual decisions, better to use the source of truth we forward to LLVM than the informational string we forward to the user. This is a breaking change for aarch64 JSON target specs: setting `abi` to "softfloat" is no longer enough; one has to also set `rustc_abi` to "softfloat". That is consistent with riscv and arm32, but it's still surprising. Cc @Darksonn in case this affects the Linux kernel. Also see rust-lang#153035 which does something similar for PowerPC, and [Zulip](https://rust-lang.zulipchat.com/#narrow/channel/131828-t-compiler/topic/De-spaghettifying.20ABI.20controls/with/575095372). Happy to delay this PR if someone has a better idea. Cc @folkertdev @workingjubilee

@Darksonn

prefer actual ABI-controling fields over target.abi when making ABI decisions We don't actually check that `abi` is consistent with the fields that control the ABI, e.g. one could set `llvm_abiname` to "ilp32e" on a riscv target without setting a matching `abi`. So, if we need to make actual decisions, better to use the source of truth we forward to LLVM than the informational string we forward to the user. This is a breaking change for aarch64 JSON target specs: setting `abi` to "softfloat" is no longer enough; one has to also set `rustc_abi` to "softfloat". That is consistent with riscv and arm32, but it's still surprising. Cc @Darksonn in case this affects the Linux kernel. Also see rust-lang#153035 which does something similar for PowerPC, and [Zulip](https://rust-lang.zulipchat.com/#narrow/channel/131828-t-compiler/topic/De-spaghettifying.20ABI.20controls/with/575095372). Happy to delay this PR if someone has a better idea. Cc @folkertdev @workingjubilee

@Darksonn

Rollup merge of #152941 - RalfJung:abi-control, r=mati865 prefer actual ABI-controling fields over target.abi when making ABI decisions We don't actually check that `abi` is consistent with the fields that control the ABI, e.g. one could set `llvm_abiname` to "ilp32e" on a riscv target without setting a matching `abi`. So, if we need to make actual decisions, better to use the source of truth we forward to LLVM than the informational string we forward to the user. This is a breaking change for aarch64 JSON target specs: setting `abi` to "softfloat" is no longer enough; one has to also set `rustc_abi` to "softfloat". That is consistent with riscv and arm32, but it's still surprising. Cc @Darksonn in case this affects the Linux kernel. Also see #153035 which does something similar for PowerPC, and [Zulip](https://rust-lang.zulipchat.com/#narrow/channel/131828-t-compiler/topic/De-spaghettifying.20ABI.20controls/with/575095372). Happy to delay this PR if someone has a better idea. Cc @folkertdev @workingjubilee

Rollup of 13 pull requests Successful merges: - rust-lang/rust#148146 (CI: use alternative disks if available) - rust-lang/rust#151771 (Fix: On wasm targets, call `panic_in_cleanup` if panic occurs in cleanup) - rust-lang/rust#153035 (rustc_target: callconv: powerpc64: Use llvm_abiname rather than target_abi for ABI determination) - rust-lang/rust#153075 (mGCA: Lower negated literals directly and reject non-integer negations) - rust-lang/rust#153078 (Remove `QuerySystemFns`) - rust-lang/rust#153089 (interpret: avoid dummy spans in the stacktrace) - rust-lang/rust#153111 (Refactor url_parts to return is_absolute instead of out param) - rust-lang/rust#149978 (deprecate `Eq::assert_receiver_is_total_eq` and emit FCW on manual impls) - rust-lang/rust#153029 (Rename `rustc::pass_by_value` lint as `rustc::disallowed_pass_by_ref`.) - rust-lang/rust#153063 (`is_ty_must_use`: do not require a `span` argument) - rust-lang/rust#153071 (Update books) - rust-lang/rust#153092 (Remove redundant self usages) - rust-lang/rust#153094 (Simplify `AppendOnlyVec` iterators) Failed merges: - rust-lang/rust#153091 (Migration of `LintDiagnostic` - part 4)

rustbot assigned RalfJung Feb 24, 2026

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 Feb 24, 2026

This comment has been minimized.
Sign in to view

Gelbpunkt force-pushed the ppc64-callconv-llvm-abiname branch from a3237e2 to d501f96Compare February 24, 2026 06:56

RalfJung mentioned this pull request Feb 24, 2026
prefer actual ABI-controling fields over target.abi when making ABI decisions #152941
Merged

rust-borsBot 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 Feb 25, 2026

JonathanBrouwer mentioned this pull request Feb 25, 2026
Rollup of 10 pull requests #153084
Closed

JonathanBrouwer mentioned this pull request Feb 25, 2026
Rollup of 9 pull requests #153086
Closed

JonathanBrouwer mentioned this pull request Feb 25, 2026
Rollup of 10 pull requests #153088
Closed

JonathanBrouwer mentioned this pull request Feb 25, 2026
Rollup of 13 pull requests #153098
Closed

jhpratt mentioned this pull request Feb 26, 2026
Rollup of 13 pull requests #153116
Merged

rust-borsBot merged commit d108355 into rust-lang:mainFeb 26, 2026
11 checks passed

rustbot added this to the 1.95.0 milestone Feb 26, 2026

This comment has been minimized.
Sign in to view

This was referenced Mar 14, 2026
target specs: stricter checks for LLVM ABI values, and correlate that with cfg(target_abi) #153769
Merged
Rename target.abi to target.cfg_abi and enum-ify llvm_abiname #153857
Merged

Navigation Menu

Uh oh!

Uh oh!

rustc_target: callconv: powerpc64: Use llvm_abiname rather than target_abi for ABI determination - #153035

rustc_target: callconv: powerpc64: Use llvm_abiname rather than target_abi for ABI determination#153035
rust-bors[bot] merged 1 commit into
rust-lang:mainfrom
Gelbpunkt:ppc64-callconv-llvm-abiname

Gelbpunkt commented Feb 24, 2026•
edited by rustbot
Loading

Uh oh!

rustbot commented Feb 24, 2026

Uh oh!

This comment has been minimized.

RalfJung commented Feb 24, 2026

Uh oh!

rust-borsBot commented Feb 24, 2026

Uh oh!

Gelbpunkt commented Feb 25, 2026

Uh oh!

rust-borsBot commented Feb 25, 2026

Uh oh!

Uh oh!

JonathanBrouwer commented Feb 26, 2026

Uh oh!

This comment has been minimized.

rust-timer commented Feb 27, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

6 participants

Uh oh!

Uh oh!

Conversation

Gelbpunkt commented Feb 24, 2026• edited by rustbot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

rustbot commented Feb 24, 2026

Uh oh!

This comment has been minimized.

RalfJung commented Feb 24, 2026

Uh oh!

rust-borsBot commented Feb 24, 2026

Uh oh!

Gelbpunkt commented Feb 25, 2026

Uh oh!

rust-borsBot commented Feb 25, 2026

Uh oh!

Uh oh!

JonathanBrouwer commented Feb 26, 2026

Uh oh!

This comment has been minimized.

rust-timer commented Feb 27, 2026

Overall result: no relevant changes - no action needed

Instruction count

Max RSS (memory usage)

Cycles

Binary size

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

6 participants

Gelbpunkt commented Feb 24, 2026•
edited by rustbot
Loading