diff --git a/compiler/rustc_middle/src/lib.rs b/compiler/rustc_middle/src/lib.rs index 69c2e099080c9..9798130e69ef0 100644 --- a/compiler/rustc_middle/src/lib.rs +++ b/compiler/rustc_middle/src/lib.rs @@ -80,6 +80,7 @@ pub mod metadata; pub mod middle; pub mod mir; pub mod mono; +pub mod ptrauth; pub mod queries; pub mod query; pub mod thir; diff --git a/compiler/rustc_middle/src/ptrauth/discriminator.rs b/compiler/rustc_middle/src/ptrauth/discriminator.rs new file mode 100644 index 0000000000000..76d7fe9b1f0ea --- /dev/null +++ b/compiler/rustc_middle/src/ptrauth/discriminator.rs @@ -0,0 +1,493 @@ +/*! +Function pointer type discrimination for pointer authentication. + +This module implements Rust's equivalent of Clang's function pointer type +discriminator computation used in pointer authentication. + +Compatibility with Clang is a primary goal. The discriminator produced for a +given external "C" function type must match the value computed by Clang so that +function pointers can be exchanged safely between Rust and C code while +preserving pointer authentication semantics. + +The implementation mirrors Clang's behavior in +`ASTContext::encodeTypeForFunctionPointerAuth`, ensuring that identical +C-compatible function types produce identical discriminators. See: +. + +## Overview + +The computation is structured into three conceptual stages: + +### 1. Type normalization and lowering + Rust types are converted into a language-independent representation + (`ClangDiscTy`) that mirrors the type categories used by Clang when computing + function pointer discriminators. This includes canonicalization such as + treating all pointer-like types uniformly and mapping Rust constructs onto + their closest C equivalents. + One notable exception is C `_Complex`. Rust has no corresponding native type, + so there is no canonical Rust representation to map onto Clang's `_Complex` + type category. Rather than infer one (for example, by treating `(f32, f32)` + or `(f64, f64)` as complex numbers), this implementation leaves such + representation choices to users and does not provide dedicated `_Complex` + encoding. + +### 2. Type encoding + The lowered representation is serialized into a byte stream using rules + intended to match Clang's implementation in: + `encodeTypeForFunctionPointerAuth`. The resulting encoding describes the + function signature in a target-independent form suitable for hashing. + +### 3. Discriminator hashing + The encoded byte stream is hashed using LLVM's stable SipHash-2-4 based + discriminator algorithm. The implementation here is a direct translation + of LLVM/Clang's logic and must remain bit-for-bit compatible. See: + . + Defined in `llvm_siphash.rs`. + +## Module structure + +- High-level API + - `FnPtrDiscriminatorSource` + - `compute_fn_ptr_type_discriminator_for` + - `clone_discriminated_ptrauth_schema_for` + +- Low-level API + - `FnPtrTypeDiscriminatorInput` + - `compute_fn_ptr_type_discriminator` + +- Signature extraction + - `extract_fn_ptr_type` + +- Clang-compatible type model + - `ClangDiscTy` + - `canonicalize_c_type` + - `to_clang_disc_ty` + +- Encoding + - `PtrauthEncoder` + - `encode_ty` + +## Compatibility requirements + +Any changes to the encoding or hashing logic should be validated against Clang's +discriminator computation. Divergence from Clang will result in incompatible +pointer authentication values across language boundaries. + +This implementation intentionally approximates Clang's behavior for extern "C" +function types only. It does NOT attempt to model full type system rules. +*/ + +use rustc_abi::ExternAbi; +use rustc_middle::ty::{self, Instance, Ty, TyCtxt, Unnormalized}; +use rustc_session::PointerAuthSchema; +use rustc_span::sym; + +use crate::ptrauth::llvm_siphash::llvm_pointer_auth_stable_siphash; + +/// Types that can serve as a source for function pointer type discrimination. +/// +/// This trait abstracts over the different compiler representations from which +/// a function signature can be obtained. Implementations construct the +/// canonical `FnPtrTypeDiscriminatorInput` consumed by the discriminator +/// computation. +/// +/// This is intended primarily for ergonomic use at call sites, allowing code +/// to compute discriminators directly from an `Instance`, `Ty`, or `FnSig` +/// without manually constructing the intermediate representation. +pub trait FnPtrDiscriminatorSource<'tcx>: Sized { + fn discriminator_input(self, tcx: TyCtxt<'tcx>) -> Option>; +} + +/// Enables discriminator computation directly from Rust function types. +/// +/// Accepts both: +/// - `FnPtr`: actual function pointer types +/// - `FnDef`: function items +/// +/// FnDef is only accepted for convenience; the discriminator is still computed +/// from the instantiated function signature. +impl<'tcx> FnPtrDiscriminatorSource<'tcx> for Ty<'tcx> { + fn discriminator_input(self, tcx: TyCtxt<'tcx>) -> Option> { + let ty = extract_fn_ptr_type(tcx, self)?; + + match ty.kind() { + ty::FnPtr(sig, header) => { + let sig = sig.skip_binder(); + Some(FnPtrTypeDiscriminatorInput::from_sig_tys(sig, header)) + } + + ty::FnDef(def_id, args) => { + let sig = tcx.fn_sig(*def_id).instantiate(tcx, args.skip_binder()).skip_binder(); + + Some(FnPtrTypeDiscriminatorInput::from_sig(sig)) + } + + _ => None, + } + } +} +/// Enables discriminator computation directly from monomorphized function +/// instances. +/// +/// The instance's signature is instantiated using its generic arguments and +/// normalized before constructing the canonical discriminator input. +impl<'tcx> FnPtrDiscriminatorSource<'tcx> for Instance<'tcx> { + fn discriminator_input(self, tcx: TyCtxt<'tcx>) -> Option> { + let sig = tcx + .instantiate_and_normalize_erasing_regions( + self.args, + ty::TypingEnv::fully_monomorphized(), + tcx.fn_sig(self.def_id()), + ) + .skip_binder(); + + Some(FnPtrTypeDiscriminatorInput::from_sig(sig)) + } +} +/// Enables discriminator computation directly from instantiated function +/// signatures. +/// +/// The signature is assumed to already be instantiated and normalized. +impl<'tcx> FnPtrDiscriminatorSource<'tcx> for ty::FnSig<'tcx> { + fn discriminator_input(self, _: TyCtxt<'tcx>) -> Option> { + Some(FnPtrTypeDiscriminatorInput::from_sig(self)) + } +} + +/// Computes the function pointer type discriminator directly from a supported +/// source. +/// +/// This is a convenience wrapper around +/// `FnPtrDiscriminatorSource::discriminator_input` and +/// `compute_fn_ptr_type_discriminator`. +/// +/// Returns `None` if the supplied source does not represent a function pointer +/// type (for example, a non-function `Ty`). +pub fn compute_fn_ptr_type_discriminator_for<'tcx, S>(tcx: TyCtxt<'tcx>, source: S) -> Option +where + S: FnPtrDiscriminatorSource<'tcx>, +{ + let input = source.discriminator_input(tcx)?; + Some(compute_fn_ptr_type_discriminator(tcx, &input)) +} + +/// Clones a pointer authentication schema and updates its constant +/// discriminator. +/// +/// If `schema` is `Some`, the function computes a function pointer type +/// discriminator from `source` and stores it in the cloned schema's +/// `constant_discriminator` field. +/// +/// If no discriminator can be computed (for example, because `source` does not +/// represent a function pointer type), the schema is returned unchanged. +/// +/// This is intended as a convenience helper for code generation sites that need +/// to attach function pointer type discrimination to a generic schema before +/// calling `get_fn_addr`. +pub fn clone_discriminated_ptrauth_schema_for<'tcx, S>( + tcx: TyCtxt<'tcx>, + mut schema: Option, + source: S, +) -> Option +where + S: FnPtrDiscriminatorSource<'tcx>, +{ + if let Some(ref mut s) = schema { + if let Some(disc) = compute_fn_ptr_type_discriminator_for(tcx, source) { + s.constant_discriminator = disc; + } + } + + schema +} + +/// Canonical representation of a function signature used for pointer +/// authentication discriminator generation. +#[derive(Debug)] +pub struct FnPtrTypeDiscriminatorInput<'tcx> { + inputs: &'tcx [Ty<'tcx>], + output: Ty<'tcx>, + abi: ExternAbi, + c_variadic: bool, +} + +impl<'tcx> FnPtrTypeDiscriminatorInput<'tcx> { + fn from_sig(sig: ty::FnSig<'tcx>) -> Self { + FnPtrTypeDiscriminatorInput { + inputs: sig.inputs(), + output: sig.output(), + abi: sig.abi(), + c_variadic: sig.c_variadic(), + } + } + + fn from_sig_tys(sig: ty::FnSigTys>, header: &ty::FnHeader>) -> Self { + FnPtrTypeDiscriminatorInput { + inputs: sig.inputs(), + output: sig.output(), + abi: header.abi(), + c_variadic: header.c_variadic(), + } + } +} + +/// Unwraps optional function pointers and normalizes the type. +/// +/// Only `Option` is supported for nullability modeling, matching C ABI +/// null pointer conventions. +fn extract_fn_ptr_type<'tcx>(tcx: TyCtxt<'tcx>, mut ty: Ty<'tcx>) -> Option> { + ty = tcx.normalize_erasing_regions(ty::TypingEnv::fully_monomorphized(), Unnormalized::new(ty)); + + loop { + match ty.kind() { + ty::Adt(def, args) if tcx.is_diagnostic_item(sym::Option, def.did()) => { + ty = args.type_at(0); + continue; + } + + ty::FnPtr(..) | ty::FnDef(..) => { + return Some(ty); + } + + _ => return None, + } + } +} + +/// Computes the Clang-compatible function pointer type discriminator. +/// +/// This is the low-level discriminator computation routine operating on an +/// already constructed `FnPtrTypeDiscriminatorInput`. +fn compute_fn_ptr_type_discriminator<'tcx>( + tcx: TyCtxt<'tcx>, + input: &FnPtrTypeDiscriminatorInput<'tcx>, +) -> u16 { + if !matches!(input.abi, ExternAbi::C { .. } | ExternAbi::System { .. }) { + return 0; + } + + let mut enc = PtrauthEncoder::new(); + enc.push(b'F'); + + encode_ty(&mut enc, tcx, input.output); + + for &arg in input.inputs { + encode_ty(&mut enc, tcx, arg); + } + + if input.c_variadic { + enc.push(b'z'); + } + + enc.push(b'E'); + + let hash = enc.finish(); + + hash.into() +} + +// Clang disc type. +#[derive(Debug)] +enum ClangDiscTy<'tcx> { + Int, + Float(&'tcx ty::FloatTy), + Bool, + Char, + + // Pointer-like types in the C ABI sense. + // This includes: + // - raw pointers (`*const T`, `*mut T`) + // - Rust references (`&T`, `&mut T`) + // - function pointers + // All collapse to a single Clang-compatible 'P' node. + Pointer, + + Array { elem: Ty<'tcx> }, + + // FIXME(jchlands) Decide if to support Complex types in future. Clang has + // dedicated node for this `Type::Complex`, Rust does not. So we could match + // against a Tuple(FP_TYPE, FP_TYPE). + // Complex(Ty<'tcx>), + Vector { bytes: u64 }, + + EnumLikeInt, + AdtName(String), + Opaque, + Void, +} + +// Canonicalize Option-wrapped pointer types used to model C nullable pointers. +// +// Rust and Clang should compute identical discriminators for equivalent C APIs. +// Clang does not distinguish nullable from non-nullable pointer types when +// computing function pointer authentication discriminators, so +// `Option` and `Option<*mut T>` are encoded identically to their +// underlying pointer types. +// +// Although `Option<*mut T>` is not considered FFI-safe by Rust and triggers the +// `improper_ctypes`/`improper_ctypes_definitions` lints, this is a warning +// rather than a hard error. Canonicalizing it here preserves Clang-compatible +// discriminator computation. +// +// Please see the following tests for sample use cases: +// pauth-fn-ptr-type-discrimination-option-callback.rs, +// pauth-fn-ptr-type-discrimination-option-return.rs and pauth-fn-ptr-type-discrimination-option.rs +fn canonicalize_c_type<'tcx>(tcx: TyCtxt<'tcx>, ty: Ty<'tcx>) -> Ty<'tcx> { + if let ty::Adt(def, args) = ty.kind() + && tcx.is_diagnostic_item(sym::Option, def.did()) + { + let inner = args.type_at(0); + + match inner.kind() { + ty::FnPtr(..) | ty::RawPtr(..) => return inner, + _ => {} + } + } + + ty +} + +/// Lowers a Rust type into a Clang-compatible discriminator type. +/// +/// This is not a full semantic translation of Rust types. It is a lossy mapping +/// that intentionally matches Clang's function pointer authentication encoding +/// rules. +/// This is not a full semantic translation of Rust types. It is a lossy mapping +/// that intentionally matches Clang's function pointer authentication encoding +/// rules where Rust has a direct language-level equivalent. +/// +/// In particular C `_Complex`, without a canonical Rust equivalent, is not +/// recognized. This avoids introducing heuristics for user-defined +/// representations that may vary across codebases. +/// +/// Important invariants: +/// - All pointer-like types (Rust refs, raw pointers, fn pointers) collapse to +/// `Pointer`. +/// - Struct/union types are encoded using name only, not layout. +/// - Enums are treated as integers. +/// - SIMD types are encoded only by total byte size (no lane semantics). +/// - No attempt is made to recognize user-defined representations of C +/// `_Complex` types. +/// This must remain in sync with Clang's `encodeTypeForFunctionPointerAuth`. +fn to_clang_disc_ty<'tcx>(tcx: TyCtxt<'tcx>, ty: Ty<'tcx>) -> ClangDiscTy<'tcx> { + let ty = canonicalize_c_type(tcx, ty); + match ty.kind() { + // C void / Rust () + ty::Tuple(list) if list.is_empty() => ClangDiscTy::Void, + + // scalars + ty::Bool => ClangDiscTy::Bool, + ty::Char => ClangDiscTy::Char, + + ty::Int(_) | ty::Uint(_) => ClangDiscTy::Int, + ty::Float(f) => ClangDiscTy::Float(f), + + // everything pointer-like collapses + ty::RawPtr(..) | ty::Ref(..) | ty::FnPtr(..) | ty::Dynamic(..) | ty::Slice(_) | ty::Str => { + ClangDiscTy::Pointer + } + + // arrays ignore size + ty::Array(elem, _) => ClangDiscTy::Array { elem: *elem }, + + // enums to integer collapse + ty::Adt(def, _) if def.is_enum() => ClangDiscTy::EnumLikeInt, + // simd vectors + ty::Adt(def, args) if def.repr().simd() => { + // Clang encodes SIMD vectors by their total size + let input = ty::PseudoCanonicalInput { + typing_env: ty::TypingEnv::fully_monomorphized(), + value: ty, + }; + + let Ok(layout) = tcx.layout_of(input) else { + tcx.dcx().delayed_bug("could not compute SIMD layout"); + return ClangDiscTy::Opaque; + }; + + let bytes = layout.size.bytes(); + + ClangDiscTy::Vector { bytes } + } + // structs/unions to name-based identity + ty::Adt(def, _) => { + let name = tcx.item_name(def.did()).to_string(); + ClangDiscTy::AdtName(name) + } + + ty::Foreign(_) => ClangDiscTy::Opaque, + + _ => ClangDiscTy::Opaque, + } +} + +// Encoder +struct PtrauthEncoder { + buf: Vec, +} + +impl PtrauthEncoder { + fn new() -> Self { + Self { buf: Vec::new() } + } + + fn push(&mut self, b: u8) { + self.buf.push(b); + } + + fn push_str(&mut self, s: &str) { + self.buf.extend_from_slice(s.as_bytes()); + } + + fn finish(&self) -> u16 { + llvm_pointer_auth_stable_siphash(&self.buf) + } +} + +/// Encodes a ClangDiscTy into the discriminator byte stream. +/// +/// This format is intended to be bit-for-bit compatible with Clang's +/// `encodeTypeForFunctionPointerAuth`. +fn encode_ty<'tcx>(enc: &mut PtrauthEncoder, tcx: TyCtxt<'tcx>, ty: Ty<'tcx>) { + let cty = to_clang_disc_ty(tcx, ty); + + match cty { + // scalars + ClangDiscTy::Bool | ClangDiscTy::Char | ClangDiscTy::Int => enc.push(b'i'), + + ClangDiscTy::Float(f) => match f.bit_width() { + 16 => enc.push_str("Dh"), + 32 => enc.push(b'f'), + 64 => enc.push(b'd'), + 128 => enc.push(b'g'), + _ => enc.push(b'?'), + }, + + ClangDiscTy::Void => enc.push(b'v'), + + // pointer boundary (NO RECURSION) + ClangDiscTy::Pointer => enc.push(b'P'), + + // arrays ignore size + ClangDiscTy::Array { elem } => { + enc.push(b'A'); + encode_ty(enc, tcx, elem); + } + + // enums collapse + ClangDiscTy::EnumLikeInt => enc.push(b'i'), + + // ADT identity + ClangDiscTy::AdtName(name) => { + enc.push_str(&name.len().to_string()); + enc.push_str(&name); + } + + ClangDiscTy::Opaque => enc.push(b'?'), + + ClangDiscTy::Vector { bytes } => { + enc.push_str("Dv"); + enc.push_str(&bytes.to_string()); + } + } +} diff --git a/compiler/rustc_middle/src/ptrauth/llvm_siphash.rs b/compiler/rustc_middle/src/ptrauth/llvm_siphash.rs new file mode 100644 index 0000000000000..68aeac11a0429 --- /dev/null +++ b/compiler/rustc_middle/src/ptrauth/llvm_siphash.rs @@ -0,0 +1,142 @@ +// LLVM SipHash-2-4 +pub fn llvm_pointer_auth_stable_siphash(data: &[u8]) -> u16 { + let raw = llvm_siphash_2_4_64(data); + + ((raw % 0xFFFF) + 1) as u16 +} + +const SIPHASH_KEY: [u8; 16] = [ + 0xb5, 0xd4, 0xc9, 0xeb, 0x79, 0x10, 0x4a, 0x79, 0x6f, 0xec, 0x8b, 0x1b, 0x42, 0x87, 0x81, 0xd4, +]; + +#[inline(always)] +fn rotl(x: u64, b: u32) -> u64 { + (x << b) | (x >> (64 - b)) +} + +#[inline(always)] +fn sipround(v0: &mut u64, v1: &mut u64, v2: &mut u64, v3: &mut u64) { + *v0 = v0.wrapping_add(*v1); + *v1 = rotl(*v1, 13); + *v1 ^= *v0; + *v0 = rotl(*v0, 32); + + *v2 = v2.wrapping_add(*v3); + *v3 = rotl(*v3, 16); + *v3 ^= *v2; + + *v0 = v0.wrapping_add(*v3); + *v3 = rotl(*v3, 21); + *v3 ^= *v0; + + *v2 = v2.wrapping_add(*v1); + *v1 = rotl(*v1, 17); + *v1 ^= *v2; + *v2 = rotl(*v2, 32); +} + +fn u64_from_le(bytes: &[u8]) -> u64 { + u64::from_le_bytes(bytes.try_into().unwrap()) +} + +fn load_u64_partial(bytes: &[u8]) -> u64 { + let mut b = 0u64; + + match bytes.len() { + 7 => { + b |= (bytes[6] as u64) << 48; + b |= (bytes[5] as u64) << 40; + b |= (bytes[4] as u64) << 32; + b |= (bytes[3] as u64) << 24; + b |= (bytes[2] as u64) << 16; + b |= (bytes[1] as u64) << 8; + b |= bytes[0] as u64; + } + 6 => { + b |= (bytes[5] as u64) << 40; + b |= (bytes[4] as u64) << 32; + b |= (bytes[3] as u64) << 24; + b |= (bytes[2] as u64) << 16; + b |= (bytes[1] as u64) << 8; + b |= bytes[0] as u64; + } + 5 => { + b |= (bytes[4] as u64) << 32; + b |= (bytes[3] as u64) << 24; + b |= (bytes[2] as u64) << 16; + b |= (bytes[1] as u64) << 8; + b |= bytes[0] as u64; + } + 4 => { + b |= (bytes[3] as u64) << 24; + b |= (bytes[2] as u64) << 16; + b |= (bytes[1] as u64) << 8; + b |= bytes[0] as u64; + } + 3 => { + b |= (bytes[2] as u64) << 16; + b |= (bytes[1] as u64) << 8; + b |= bytes[0] as u64; + } + 2 => { + b |= (bytes[1] as u64) << 8; + b |= bytes[0] as u64; + } + 1 => { + b |= bytes[0] as u64; + } + _ => {} + } + + b +} + +// LLVM siphash<2,4> 64-bit output +fn llvm_siphash_2_4_64(data: &[u8]) -> u64 { + let k0 = u64_from_le(&SIPHASH_KEY[0..8]); + let k1 = u64_from_le(&SIPHASH_KEY[8..16]); + + let mut v0: u64 = 0x736f6d6570736575 ^ k0; + let mut v1: u64 = 0x646f72616e646f6d ^ k1; + let mut v2: u64 = 0x6c7967656e657261 ^ k0; + let mut v3: u64 = 0x7465646279746573 ^ k1; + + let mut b: u64 = (data.len() as u64) << 56; + let mut i = 0; + + // compression + while i + 8 <= data.len() { + let m = u64_from_le(&data[i..i + 8]); + i += 8; + + v3 ^= m; + + for _ in 0..2 { + sipround(&mut v0, &mut v1, &mut v2, &mut v3); + } + + v0 ^= m; + } + + // tail + let tail = &data[i..]; + + b |= load_u64_partial(tail); + + v3 ^= b; + + for _ in 0..2 { + sipround(&mut v0, &mut v1, &mut v2, &mut v3); + } + + v0 ^= b; + + // finalization + v2 ^= 0xff; + + for _ in 0..4 { + sipround(&mut v0, &mut v1, &mut v2, &mut v3); + } + + v0 ^ v1 ^ v2 ^ v3 +} diff --git a/compiler/rustc_middle/src/ptrauth/mod.rs b/compiler/rustc_middle/src/ptrauth/mod.rs new file mode 100644 index 0000000000000..ec874c3ef9015 --- /dev/null +++ b/compiler/rustc_middle/src/ptrauth/mod.rs @@ -0,0 +1,7 @@ +pub mod discriminator; +pub mod llvm_siphash; + +pub use discriminator::{ + FnPtrDiscriminatorSource, FnPtrTypeDiscriminatorInput, clone_discriminated_ptrauth_schema_for, + compute_fn_ptr_type_discriminator_for, +};