Skip to content

c_variadic: provide our own va_arg implementation for more targets #150094

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

Merged
rust-bors merged 1 commit into from
Jan 8, 2026
+83 −23
Merged

c_variadic: provide our own va_arg implementation for more targets #150094

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
106 changes: 83 additions & 23 deletions compiler/rustc_codegen_llvm/src/va_arg.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,12 +1,13 @@
use rustc_abi::{Align, BackendRepr, Endian, HasDataLayout, Primitive, Size, TyAndLayout};
use rustc_abi::{Align, BackendRepr, Endian, HasDataLayout, Primitive, Size};
use rustc_codegen_ssa::MemFlags;
use rustc_codegen_ssa::common::IntPredicate;
use rustc_codegen_ssa::mir::operand::OperandRef;
use rustc_codegen_ssa::traits::{
BaseTypeCodegenMethods, BuilderMethods, ConstCodegenMethods, LayoutTypeCodegenMethods,
};
use rustc_middle::bug;
use rustc_middle::ty::Ty;
use rustc_middle::ty::layout::{HasTyCtxt, LayoutOf};
use rustc_middle::ty::layout::{HasTyCtxt, LayoutOf, TyAndLayout};
use rustc_target::spec::{Abi, Arch, Env};

use crate::builder::Builder;
Expand Down Expand Up @@ -82,6 +83,7 @@ enum PassMode {
enum SlotSize {
Bytes8 = 8,
Bytes4 = 4,
Bytes1 = 1,
}

enum AllowHigherAlign {
Expand Down Expand Up @@ -728,7 +730,7 @@ fn emit_x86_64_sysv64_va_arg<'ll, 'tcx>(
fn copy_to_temporary_if_more_aligned<'ll, 'tcx>(
bx: &mut Builder<'_, 'll, 'tcx>,
reg_addr: &'ll Value,
layout: TyAndLayout<'tcx, Ty<'tcx>>,
layout: TyAndLayout<'tcx>,
src_align: Align,
) -> &'ll Value {
if layout.layout.align.abi > src_align {
Expand All @@ -751,7 +753,7 @@ fn copy_to_temporary_if_more_aligned<'ll, 'tcx>(
fn x86_64_sysv64_va_arg_from_memory<'ll, 'tcx>(
bx: &mut Builder<'_, 'll, 'tcx>,
va_list_addr: &'ll Value,
layout: TyAndLayout<'tcx, Ty<'tcx>>,
layout: TyAndLayout<'tcx>,
) -> &'ll Value {
let dl = bx.cx.data_layout();
let ptr_align_abi = dl.data_layout().pointer_align().abi;
Expand Down Expand Up @@ -1003,15 +1005,17 @@ fn emit_xtensa_va_arg<'ll, 'tcx>(
return bx.load(layout.llvm_type(bx), value_ptr, layout.align.abi);
}

/// Determine the va_arg implementation to use. The LLVM va_arg instruction
/// is lacking in some instances, so we should only use it as a fallback.
pub(super) fn emit_va_arg<'ll, 'tcx>(
bx: &mut Builder<'_, 'll, 'tcx>,
addr: OperandRef<'tcx, &'ll Value>,
target_ty: Ty<'tcx>,
) -> &'ll Value {
// Determine the va_arg implementation to use. The LLVM va_arg instruction
// is lacking in some instances, so we should only use it as a fallback.
let target = &bx.cx.tcx.sess.target;
let layout = bx.cx.layout_of(target_ty);
let target_ty_size = layout.layout.size().bytes();

let target = &bx.cx.tcx.sess.target;
match target.arch {
Arch::X86 => emit_ptr_va_arg(
bx,
Expand Down Expand Up @@ -1069,23 +1073,79 @@ pub(super) fn emit_va_arg<'ll, 'tcx>(
AllowHigherAlign::Yes,
ForceRightAdjust::No,
),
Arch::LoongArch32 => emit_ptr_va_arg(
bx,
addr,
target_ty,
if target_ty_size > 2 * 4 { PassMode::Indirect } else { PassMode::Direct },
SlotSize::Bytes4,
AllowHigherAlign::Yes,
ForceRightAdjust::No,
),
Arch::LoongArch64 => emit_ptr_va_arg(
bx,
addr,
target_ty,
if target_ty_size > 2 * 8 { PassMode::Indirect } else { PassMode::Direct },
SlotSize::Bytes8,
AllowHigherAlign::Yes,
ForceRightAdjust::No,
),
Arch::AmdGpu => emit_ptr_va_arg(
bx,
addr,
target_ty,
PassMode::Direct,
SlotSize::Bytes4,
AllowHigherAlign::No,
ForceRightAdjust::No,
),
Comment on lines +1094 to +1102
Copy link
Contributor Author

@folkertdev folkertdev Dec 17, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

https://github.com/llvm/llvm-project/blob/c386d6d3bfa400e94ac9cadb85a0b1a72ed54b2b/clang/lib/CodeGen/Targets/AMDGPU.cpp#L136

Copy link
Member

@workingjubilee workingjubilee Dec 30, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

just making sure @Flakebi sees this, since we had one platform weirdness and the idea of varargs on a GPU at least feels inherently fishy, but maybe it's like... fine actually? feels like a bad idea performance wise, but

Copy link
Contributor

@Flakebi Flakebi Dec 30, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Comparing the Rust and clang implementation, this looks fine (thanks for checking).
I wanted to say I never tried varargs on GPUs but I did use printf, so… (GPUs inline everything anyway, so as long as it’s inlined before instruction selection it’s not a problem performance wise)

Arch::Nvptx64 => emit_ptr_va_arg(
bx,
addr,
target_ty,
PassMode::Direct,
SlotSize::Bytes1,
AllowHigherAlign::Yes,
ForceRightAdjust::No,
),
Comment on lines +1103 to +1111
Copy link
Contributor Author

@folkertdev folkertdev Dec 17, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

https://github.com/llvm/llvm-project/blob/c386d6d3bfa400e94ac9cadb85a0b1a72ed54b2b/clang/lib/CodeGen/Targets/NVPTX.cpp#L231

Copy link
Member

@workingjubilee workingjubilee Dec 30, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

cc @kjetilkjeka for similar reasons: just letting you know about this if you have a comment.

Arch::Wasm32 => emit_ptr_va_arg(
bx,
addr,
target_ty,
if layout.is_aggregate() || layout.is_zst() || layout.is_1zst() {
PassMode::Indirect
} else {
PassMode::Direct
},
SlotSize::Bytes4,
AllowHigherAlign::Yes,
ForceRightAdjust::No,
),
Arch::Wasm64 => bug!("c-variadic functions are not fully implemented for wasm64"),
Arch::CSky => emit_ptr_va_arg(
bx,
addr,
target_ty,
PassMode::Direct,
SlotSize::Bytes4,
AllowHigherAlign::Yes,
ForceRightAdjust::No,
),
Comment on lines +1126 to +1134
Copy link
Contributor Author

@folkertdev folkertdev Dec 17, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

https://github.com/llvm/llvm-project/blob/c386d6d3bfa400e94ac9cadb85a0b1a72ed54b2b/clang/lib/CodeGen/Targets/CSKY.cpp#L70

Copy link
Member

@workingjubilee workingjubilee Dec 30, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

cc @Dirreke if there's anything of particular interest here.

// Windows x86_64
Arch::X86_64 if target.is_like_windows => {
let target_ty_size = bx.cx.size_of(target_ty).bytes();
emit_ptr_va_arg(
bx,
addr,
target_ty,
if target_ty_size > 8 || !target_ty_size.is_power_of_two() {
PassMode::Indirect
} else {
PassMode::Direct
},
SlotSize::Bytes8,
AllowHigherAlign::No,
ForceRightAdjust::No,
)
}
Arch::X86_64 if target.is_like_windows => emit_ptr_va_arg(
bx,
addr,
target_ty,
if target_ty_size > 8 || !target_ty_size.is_power_of_two() {
PassMode::Indirect
} else {
PassMode::Direct
},
SlotSize::Bytes8,
AllowHigherAlign::No,
ForceRightAdjust::No,
),
// This includes `target.is_like_darwin`, which on x86_64 targets is like sysv64.
Arch::X86_64 => emit_x86_64_sysv64_va_arg(bx, addr, target_ty),
Arch::Xtensa => emit_xtensa_va_arg(bx, addr, target_ty),
Expand Down
Loading