bjoernager/rust
bjoernager/rust
1
Fork 0
Code Issues Pull requests Activity

move ABI sanity check from LLVM codegen backend to ABI computation logic

Browse source
This commit is contained in:
Ralf Jung 2023-10-29 12:24:32 +01:00
parent 2429818b20
commit 405e4204d0
3 changed files with 79 additions and 43 deletions
Download patch file Download diff file Expand all files Collapse all files

43
compiler/rustc_codegen_llvm/src/abi.rs
View file

@ -348,50 +348,18 @@ impl<'ll, 'tcx> FnAbiLlvmExt<'ll, 'tcx> for FnAbi<'tcx, Ty<'tcx>> {
PassMode::Direct(_) => {
// ABI-compatible Rust types have the same `layout.abi` (up to validity ranges),
// and for Scalar ABIs the LLVM type is fully determined by `layout.abi`,
// guarnateeing that we generate ABI-compatible LLVM IR. Things get tricky for
// aggregates...
if matches!(arg.layout.abi, abi::Abi::Aggregate { .. }) {
assert!(
arg.layout.is_sized(),
"`PassMode::Direct` for unsized type: {}",
arg.layout.ty
);
// This really shouldn't happen, since `immediate_llvm_type` will use
// `layout.fields` to turn this Rust type into an LLVM type. This means all
// sorts of Rust type details leak into the ABI. However wasm sadly *does*
// currently use this mode so we have to allow it -- but we absolutely
// shouldn't let any more targets do that.
// (Also see <https://github.com/rust-lang/rust/issues/115666>.)
//
// The unstable abi `PtxKernel` also uses Direct for now.
// It needs to switch to something else before stabilization can happen.
// (See issue: https://github.com/rust-lang/rust/issues/117271)
assert!(
matches!(&*cx.tcx.sess.target.arch, "wasm32" | "wasm64")
|| self.conv == Conv::PtxKernel,
"`PassMode::Direct` for aggregates only allowed on wasm and `extern \"ptx-kernel\"` fns\nProblematic type: {:#?}",
arg.layout,
);
}
// guaranteeing that we generate ABI-compatible LLVM IR.
arg.layout.immediate_llvm_type(cx)
}
PassMode::Pair(..) => {
// ABI-compatible Rust types have the same `layout.abi` (up to validity ranges),
// so for ScalarPair we can easily be sure that we are generating ABI-compatible
// LLVM IR.
assert!(
matches!(arg.layout.abi, abi::Abi::ScalarPair(..)),
"PassMode::Pair for type {}",
arg.layout.ty
);
llargument_tys.push(arg.layout.scalar_pair_element_llvm_type(cx, 0, true));
llargument_tys.push(arg.layout.scalar_pair_element_llvm_type(cx, 1, true));
continue;
}
PassMode::Indirect { attrs: _, meta_attrs: Some(_), on_stack } => {
// `Indirect` with metadata is only for unsized types, and doesn't work with
// on-stack passing.
assert!(arg.layout.is_unsized() && !on_stack);
PassMode::Indirect { attrs: _, meta_attrs: Some(_), on_stack: _ } => {
// Construct the type of a (wide) pointer to `ty`, and pass its two fields.
// Any two ABI-compatible unsized types have the same metadata type and
// moreover the same metadata value leads to the same dynamic size and
@ -402,13 +370,8 @@ impl<'ll, 'tcx> FnAbiLlvmExt<'ll, 'tcx> for FnAbi<'tcx, Ty<'tcx>> {
llargument_tys.push(ptr_layout.scalar_pair_element_llvm_type(cx, 1, true));
continue;
}
PassMode::Indirect { attrs: _, meta_attrs: None, on_stack: _ } => {
assert!(arg.layout.is_sized());
cx.type_ptr()
}
PassMode::Indirect { attrs: _, meta_attrs: None, on_stack: _ } => cx.type_ptr(),
PassMode::Cast { cast, pad_i32 } => {
// `Cast` means "transmute to `CastType`"; that only makes sense for sized types.
assert!(arg.layout.is_sized());
// add padding
if *pad_i32 {
llargument_tys.push(Reg::i32().llvm_type(cx));