Auto merge of rust-lang#132173 - veluca93:abi_checks, r=<try>

Emit warning when calling/declaring functions with unavailable vectors.

On some architectures, vector types may have a different ABI depending on whether the relevant target features are enabled. (The ABI when the feature is disabled is often not specified, but LLVM implements some de-facto ABI.)

As discussed in rust-lang/lang-team#235, this turns out to very easily lead to unsound code.

This commit makes it a post-monomorphization future-incompat warning to declare or call functions using those vector types in a context in which the corresponding target features are disabled, if using an ABI for which the difference is relevant. This ensures that these functions are always called with a consistent ABI.

See the [nomination comment](rust-lang#127731 (comment)) for more discussion.

Part of rust-lang#116558

r? RalfJung

compiler/rustc_lint_defs/src/builtin.rs

@@ -16,6 +16,7 @@ declare_lint_pass! {
     /// that are used by other parts of the compiler.
     HardwiredLints => [
         // tidy-alphabetical-start
+        ABI_UNSUPPORTED_VECTOR_TYPES,
         ABSOLUTE_PATHS_NOT_STARTING_WITH_CRATE,
         AMBIGUOUS_ASSOCIATED_ITEMS,
         AMBIGUOUS_GLOB_IMPORTS,
@@ -5031,3 +5032,69 @@ declare_lint! {
     };
     crate_level_only
 }
+
+declare_lint! {
+    /// The `abi_unsupported_vector_types` lint detects function definitions and calls
+    /// whose ABI depends on enabling certain target features, but those features are not enabled.
+    ///
+    /// ### Example
+    ///
+    /// ```rust,ignore (fails on non-x86_64)
+    /// extern "C" fn missing_target_feature(_: std::arch::x86_64::__m256) {
+    ///   todo!()
+    /// }
+    ///
+    /// #[target_feature(enable = "avx")]
+    /// unsafe extern "C" fn with_target_feature(_: std::arch::x86_64::__m256) {
+    ///   todo!()
+    /// }
+    ///
+    /// fn main() {
+    ///   let v = unsafe { std::mem::zeroed() };
+    ///   unsafe { with_target_feature(v); }
+    /// }
+    /// ```
+    ///
+    /// ```text
+    /// warning: ABI error: this function call uses a avx vector type, which is not enabled in the caller
+    ///  --> lint_example.rs:18:12
+    ///   |
+    ///   |   unsafe { with_target_feature(v); }
+    ///   |            ^^^^^^^^^^^^^^^^^^^^^^ function called here
+    ///   |
+    ///   = warning: this was previously accepted by the compiler but is being phased out; it will become a hard error in a future release!
+    ///   = note: for more information, see issue #116558 <https://github.com/rust-lang/rust/issues/116558>
+    ///   = help: consider enabling it globally (-C target-feature=+avx) or locally (#[target_feature(enable="avx")])
+    ///   = note: `#[warn(abi_unsupported_vector_types)]` on by default
+    ///
+    ///
+    /// warning: ABI error: this function definition uses a avx vector type, which is not enabled
+    ///  --> lint_example.rs:3:1
+    ///   |
+    ///   | pub extern "C" fn with_target_feature(_: std::arch::x86_64::__m256) {
+    ///   | ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ function defined here
+    ///   |
+    ///   = warning: this was previously accepted by the compiler but is being phased out; it will become a hard error in a future release!
+    ///   = note: for more information, see issue #116558 <https://github.com/rust-lang/rust/issues/116558>
+    ///   = help: consider enabling it globally (-C target-feature=+avx) or locally (#[target_feature(enable="avx")])
+    /// ```
+    ///
+    ///
+    ///
+    /// ### Explanation
+    ///
+    /// The C ABI for `__m256` requires the value to be passed in an AVX register,
+    /// which is only possible when the `avx` target feature is enabled.
+    /// Therefore, `missing_target_feature` cannot be compiled without that target feature.
+    /// A similar (but complementary) message is triggered when `with_target_feature` is called
+    /// by a function that does not enable the `avx` target feature.
+    ///
+    /// Note that this lint is very similar to the `-Wpsabi` warning in `gcc`/`clang`.
+    pub ABI_UNSUPPORTED_VECTOR_TYPES,
+    Warn,
+    "this function call or definition uses a vector type which is not enabled",
+    @future_incompatible = FutureIncompatibleInfo {
+        reason: FutureIncompatibilityReason::FutureReleaseErrorDontReportInDeps,
+        reference: "issue #116558 <https://github.com/rust-lang/rust/issues/116558>",
+    };
+}

compiler/rustc_middle/src/query/keys.rs

@@ -591,3 +591,11 @@ impl<'tcx> Key for (ValidityRequirement, ty::ParamEnvAnd<'tcx, Ty<'tcx>>) {
         }
     }
 }
+
+impl<'tcx> Key for (Ty<'tcx>, DefId) {
+    type Cache<V> = DefaultCache<Self, V>;
+
+    fn default_span(&self, tcx: TyCtxt<'_>) -> Span {
+        self.1.default_span(tcx)
+    }
+}

compiler/rustc_middle/src/query/mod.rs

@@ -2299,6 +2299,19 @@ rustc_queries! {
         desc { "whether the item should be made inlinable across crates" }
         separate_provide_extern
     }
+
+    /// Returns the list of missing features for a call to a callee of a certain type in a certain
+    /// caller (identified by DefId).
+    query call_site_abi_missing_features(key: (Ty<'tcx>, DefId)) -> &'tcx Vec<String> {
+        arena_cache
+        desc { "missing features for ABI compatibility at call sites" }
+        cache_on_disk_if { true }
+    }
+
+    query check_instance_abi(key: ty::Instance<'tcx>) {
+        desc { "check ABI mismatch for instances" }
+        cache_on_disk_if { true }
+    }
 }
 
 rustc_query_append! { define_callbacks! }

compiler/rustc_monomorphize/messages.ftl

@@ -1,3 +1,12 @@
+monomorphize_abi_error_disabled_vector_type_call =
+  ABI error: this function call uses a vector type that requires the `{$required_feature}` target feature, which is not enabled in the caller
+  .label = function called here
+  .help = consider enabling it globally (`-C target-feature=+{$required_feature}`) or locally (`#[target_feature(enable="{$required_feature}")]`)
+monomorphize_abi_error_disabled_vector_type_def =
+  ABI error: this function definition uses a vector type that requires the `{$required_feature}` target feature, which is not enabled
+  .label = function defined here
+  .help = consider enabling it globally (`-C target-feature=+{$required_feature}`) or locally (`#[target_feature(enable="{$required_feature}")]`)
+
 monomorphize_couldnt_dump_mono_stats =
     unexpected error occurred while dumping monomorphization stats: {$error}
 

compiler/rustc_monomorphize/src/collector.rs

@@ -205,6 +205,7 @@
 //! this is not implemented however: a mono item will be produced
 //! regardless of whether it is actually needed or not.
 
+mod abi_check;
 mod move_check;
 
 use std::path::PathBuf;
@@ -223,11 +224,12 @@ use rustc_middle::mir::visit::Visitor as MirVisitor;
 use rustc_middle::mir::{self, Location, MentionedItem, traversal};
 use rustc_middle::query::TyCtxtAt;
 use rustc_middle::ty::adjustment::{CustomCoerceUnsized, PointerCoercion};
+use rustc_middle::ty::inherent::*;
 use rustc_middle::ty::layout::ValidityRequirement;
 use rustc_middle::ty::print::{shrunk_instance_name, with_no_trimmed_paths};
 use rustc_middle::ty::{
-    self, AssocKind, GenericArgs, GenericParamDefKind, Instance, InstanceKind, Ty, TyCtxt,
-    TypeFoldable, TypeVisitableExt, VtblEntry,
+    self, AssocKind, GenericArgs, GenericParamDefKind, Instance, InstanceKind, ParamEnv, Ty,
+    TyCtxt, TypeFoldable, TypeVisitableExt, VtblEntry,
 };
 use rustc_middle::util::Providers;
 use rustc_middle::{bug, span_bug};
@@ -766,6 +768,7 @@ impl<'a, 'tcx> MirVisitor<'tcx> for MirUsedCollector<'a, 'tcx> {
                 self.used_mentioned_items.insert(MentionedItem::Fn(callee_ty));
                 let callee_ty = self.monomorphize(callee_ty);
                 self.check_fn_args_move_size(callee_ty, args, *fn_span, location);
+                abi_check::check_call_site_abi(tcx, callee_ty, *fn_span, self.body.source.instance);
                 visit_fn_use(self.tcx, callee_ty, true, source, &mut self.used_items)
             }
             mir::TerminatorKind::Drop { ref place, .. } => {
@@ -1207,6 +1210,19 @@ fn collect_items_of_instance<'tcx>(
     mentioned_items: &mut MonoItems<'tcx>,
     mode: CollectionMode,
 ) {
+    // Check the instance for feature-dependent ABI if we cannot determine that the function has
+    // "Rust" ABI, which is known not to be feature-dependent.
+    // Note that the check is still correct on Rust ABI functions, but somewhat expensive. Hence,
+    // checking for "Rust" ABI is just an optimization.
+    let instance_ty = instance.ty(tcx, ParamEnv::reveal_all());
+    // We only check the signature if the instance is actually function-like, as otherwise this
+    // causes an ICE.
+    if !instance_ty.is_fn() || !instance_ty.fn_sig(tcx).abi().is_rust() {
+        // The query returns nothing, but is declared #[must_use] nonetheless. This silences the
+        // warning.
+        let _ = tcx.check_instance_abi(instance);
+    }
+
     let body = tcx.instance_mir(instance.def);
     // Naively, in "used" collection mode, all functions get added to *both* `used_items` and
     // `mentioned_items`. Mentioned items processing will then notice that they have already been
@@ -1623,4 +1639,5 @@ pub(crate) fn collect_crate_mono_items<'tcx>(
 
 pub(crate) fn provide(providers: &mut Providers) {
     providers.hooks.should_codegen_locally = should_codegen_locally;
+    abi_check::provide(providers);
 }

compiler/rustc_monomorphize/src/collector/abi_check.rs

@@ -0,0 +1,134 @@
+//! This module ensures that if a function's ABI requires a particular target feature,
+//! that target feature is enabled both on the callee and all callers.
+use rustc_hir::CRATE_HIR_ID;
+use rustc_middle::query::Providers;
+use rustc_middle::ty::inherent::*;
+use rustc_middle::ty::{self, Instance, InstanceKind, ParamEnv, Ty, TyCtxt};
+use rustc_session::lint::builtin::ABI_UNSUPPORTED_VECTOR_TYPES;
+use rustc_span::def_id::DefId;
+use rustc_span::{DUMMY_SP, Span, Symbol};
+use rustc_target::abi::call::{FnAbi, PassMode};
+use rustc_target::abi::{Abi as PassAbi, RegKind};
+
+use crate::errors::{AbiErrorDisabledVectorTypeCall, AbiErrorDisabledVectorTypeDef};
+
+fn uses_vector_registers(mode: &PassMode, abi: &PassAbi) -> bool {
+    match mode {
+        PassMode::Ignore | PassMode::Indirect { .. } => false,
+        PassMode::Cast { pad_i32: _, cast } => {
+            cast.prefix.iter().any(|r| r.is_some_and(|x| x.kind == RegKind::Vector))
+                || cast.rest.unit.kind == RegKind::Vector
+        }
+        PassMode::Direct(..) | PassMode::Pair(..) => matches!(abi, PassAbi::Vector { .. }),
+    }
+}
+
+fn do_check_abi<'tcx>(
+    tcx: TyCtxt<'tcx>,
+    abi: &FnAbi<'tcx, Ty<'tcx>>,
+    target_feature_def: DefId,
+    mut emit_err: impl FnMut(&'static str),
+) {
+    let Some(feature_def) = tcx.sess.target.features_for_correct_vector_abi() else {
+        return;
+    };
+    let codegen_attrs = tcx.codegen_fn_attrs(target_feature_def);
+    for arg_abi in abi.args.iter().chain(std::iter::once(&abi.ret)) {
+        let size = arg_abi.layout.size;
+        if uses_vector_registers(&arg_abi.mode, &arg_abi.layout.abi) {
+            // Find the first feature that provides at least this vector size.
+            let feature = match feature_def.iter().find(|(bits, _)| size.bits() <= *bits) {
+                Some((_, feature)) => feature,
+                None => {
+                    emit_err("<no available feature for this size>");
+                    continue;
+                }
+            };
+            let feature_sym = Symbol::intern(feature);
+            if !tcx.sess.unstable_target_features.contains(&feature_sym)
+                && !codegen_attrs.target_features.iter().any(|x| x.name == feature_sym)
+            {
+                emit_err(feature);
+            }
+        }
+    }
+}
+
+/// Checks that the ABI of a given instance of a function does not contain vector-passed arguments
+/// or return values for which the corresponding target feature is not enabled.
+fn check_instance_abi<'tcx>(tcx: TyCtxt<'tcx>, instance: Instance<'tcx>) {
+    let param_env = ParamEnv::reveal_all();
+    let Ok(abi) = tcx.fn_abi_of_instance(param_env.and((instance, ty::List::empty()))) else {
+        // An error will be reported during codegen if we cannot determine the ABI of this
+        // function.
+        return;
+    };
+    do_check_abi(tcx, abi, instance.def_id(), |required_feature| {
+        let span = tcx.def_span(instance.def_id());
+        tcx.emit_node_span_lint(
+            ABI_UNSUPPORTED_VECTOR_TYPES,
+            CRATE_HIR_ID,
+            span,
+            AbiErrorDisabledVectorTypeDef { span, required_feature },
+        );
+    })
+}
+
+fn call_site_abi_missing_features<'tcx>(
+    tcx: TyCtxt<'tcx>,
+    (callee, caller): (Ty<'tcx>, DefId),
+) -> Vec<String> {
+    let mut answer = vec![];
+    let param_env = ParamEnv::reveal_all();
+    let callee_abi = match *callee.kind() {
+        ty::FnPtr(..) => {
+            tcx.fn_abi_of_fn_ptr(param_env.and((callee.fn_sig(tcx), ty::List::empty())))
+        }
+        ty::FnDef(def_id, args) => {
+            // Intrinsics are handled separately by the compiler.
+            if tcx.intrinsic(def_id).is_some() {
+                return vec![];
+            }
+            let instance = ty::Instance::expect_resolve(tcx, param_env, def_id, args, DUMMY_SP);
+            tcx.fn_abi_of_instance(param_env.and((instance, ty::List::empty())))
+        }
+        _ => {
+            panic!("Invalid function call");
+        }
+    };
+
+    let Ok(callee_abi) = callee_abi else {
+        // ABI failed to compute; this will not get through codegen.
+        return vec![];
+    };
+    do_check_abi(tcx, callee_abi, caller, |required_feature| {
+        answer.push(required_feature.to_string());
+    });
+    answer
+}
+
+/// Checks that a call expression does not try to pass a vector-passed argument which requires a
+/// target feature that the caller does not have, as doing so causes UB because of ABI mismatch.
+pub(super) fn check_call_site_abi<'tcx>(
+    tcx: TyCtxt<'tcx>,
+    callee: Ty<'tcx>,
+    span: Span,
+    caller: InstanceKind<'tcx>,
+) {
+    if callee.fn_sig(tcx).abi().is_rust() {
+        // "Rust" ABI never passes arguments in vector registers.
+        return;
+    }
+    for required_feature in tcx.call_site_abi_missing_features((callee, caller.def_id())) {
+        tcx.emit_node_span_lint(
+            ABI_UNSUPPORTED_VECTOR_TYPES,
+            CRATE_HIR_ID,
+            span,
+            AbiErrorDisabledVectorTypeCall { span, required_feature },
+        );
+    }
+}
+
+pub(super) fn provide(providers: &mut Providers) {
+    *providers = Providers { check_instance_abi, call_site_abi_missing_features, ..*providers }
+}

compiler/rustc_monomorphize/src/errors.rs

@@ -92,3 +92,21 @@ pub(crate) struct StartNotFound;
 pub(crate) struct UnknownCguCollectionMode<'a> {
     pub mode: &'a str,
 }
+
+#[derive(LintDiagnostic)]
+#[diag(monomorphize_abi_error_disabled_vector_type_def)]
+#[help]
+pub(crate) struct AbiErrorDisabledVectorTypeDef<'a> {
+    #[label]
+    pub span: Span,
+    pub required_feature: &'a str,
+}
+
+#[derive(LintDiagnostic)]
+#[diag(monomorphize_abi_error_disabled_vector_type_call)]
+#[help]
+pub(crate) struct AbiErrorDisabledVectorTypeCall<'a> {
+    #[label]
+    pub span: Span,
+    pub required_feature: &'a str,
+}

compiler/rustc_target/src/target_features.rs

@@ -524,6 +524,13 @@ pub fn all_known_features() -> impl Iterator<Item = (&'static str, Stability)> {
         .map(|(f, s, _)| (f, s))
 }
 
+// These arrays represent the least-constraining feature that is required for vector types up to a
+// certain size to have their "proper" ABI on each architecture.
+// Note that they must be kept sorted by vector size.
+const X86_FEATURES_FOR_CORRECT_VECTOR_ABI: &'static [(u64, &'static str)] =
+    &[(128, "sse"), (256, "avx"), (512, "avx512f")];
+const AARCH64_FEATURES_FOR_CORRECT_VECTOR_ABI: &'static [(u64, &'static str)] = &[(128, "neon")];
+
 impl super::spec::Target {
     pub fn supported_target_features(
         &self,
@@ -545,6 +552,16 @@ impl super::spec::Target {
         }
     }
 
+    // Returns None if we do not support ABI checks on the given target yet.
+    pub fn features_for_correct_vector_abi(&self) -> Option<&'static [(u64, &'static str)]> {
+        match &*self.arch {
+            "x86" | "x86_64" => Some(X86_FEATURES_FOR_CORRECT_VECTOR_ABI),
+            "aarch64" => Some(AARCH64_FEATURES_FOR_CORRECT_VECTOR_ABI),
+            // FIXME: add support for non-tier1 architectures
+            _ => None,
+        }
+    }
+
     pub fn tied_target_features(&self) -> &'static [&'static [&'static str]] {
         match &*self.arch {
             "aarch64" | "arm64ec" => AARCH64_TIED_FEATURES,