Auto merge of #155793 - JonathanBrouwer:rollup-q75rqML, r=JonathanBrouwer

Rollup of 4 pull requests

Successful merges:

 - #146181 (Add intrinsic for launch-sized workgroup memory on GPUs)
 - #155065 (Error on invalid macho section specifier)
 - #155676 ( Reject implementing const Drop for types that are not const `Destruct` already)
 - #155783 (Do not suggest internal cfg trace attributes)

Author: bors

compiler/rustc_abi/src/lib.rs

@@ -1753,6 +1753,9 @@ pub struct AddressSpace(pub u32);
 impl AddressSpace {
     /// LLVM's `0` address space.
     pub const ZERO: Self = AddressSpace(0);
+    /// The address space for workgroup memory on nvptx and amdgpu.
+    /// See e.g. the `gpu_launch_sized_workgroup_mem` intrinsic for details.
+    pub const GPU_WORKGROUP: Self = AddressSpace(3);
 }
 
 /// How many scalable vectors are in a `BackendRepr::ScalableVector`?

compiler/rustc_attr_parsing/src/attributes/link_attrs.rs

@@ -16,8 +16,9 @@ use crate::attributes::cfg::parse_cfg_entry;
 use crate::session_diagnostics::{
     AsNeededCompatibility, BundleNeedsStatic, EmptyLinkName, ExportSymbolsNeedsStatic,
     ImportNameTypeRaw, ImportNameTypeX86, IncompatibleWasmLink, InvalidLinkModifier,
-    LinkFrameworkApple, LinkOrdinalOutOfRange, LinkRequiresName, MultipleModifiers,
-    NullOnLinkSection, RawDylibNoNul, RawDylibOnlyWindows, WholeArchiveNeedsStatic,
+    InvalidMachoSection, InvalidMachoSectionReason, LinkFrameworkApple, LinkOrdinalOutOfRange,
+    LinkRequiresName, MultipleModifiers, NullOnLinkSection, RawDylibNoNul, RawDylibOnlyWindows,
+    WholeArchiveNeedsStatic,
 };
 
 pub(crate) struct LinkNameParser;
@@ -462,6 +463,29 @@ impl LinkParser {
 
 pub(crate) struct LinkSectionParser;
 
+fn check_link_section_macho(name: Symbol) -> Result<(), InvalidMachoSectionReason> {
+    let mut parts = name.as_str().split(',').map(|s| s.trim());
+
+    // The segment can be empty.
+    let _segment = parts.next();
+
+    // But the section is required.
+    let section = match parts.next() {
+        None | Some("") => return Err(InvalidMachoSectionReason::MissingSection),
+        Some(section) => section,
+    };
+
+    if section.len() > 16 {
+        return Err(InvalidMachoSectionReason::SectionTooLong { section: section.to_string() });
+    }
+
+    // LLVM also checks the other components of the section specifier, but that logic is hard to
+    // keep in sync. We skip it here for now, assuming that if you got that far you'll be able
+    // to interpret the LLVM errors.
+
+    Ok(())
+}
+
 impl<S: Stage> SingleAttributeParser<S> for LinkSectionParser {
     const PATH: &[Symbol] = &[sym::link_section];
     const ON_DUPLICATE: OnDuplicate<S> = OnDuplicate::WarnButFutureError;
@@ -495,6 +519,18 @@ impl<S: Stage> SingleAttributeParser<S> for LinkSectionParser {
             return None;
         }
 
+        // We (currently) only validate macho section specifiers.
+        match cx.sess.target.binary_format {
+            BinaryFormat::MachO => match check_link_section_macho(name) {
+                Ok(()) => {}
+                Err(reason) => {
+                    cx.emit_err(InvalidMachoSection { name_span: nv.value_span, reason });
+                    return None;
+                }
+            },
+            BinaryFormat::Coff | BinaryFormat::Elf | BinaryFormat::Wasm | BinaryFormat::Xcoff => {}
+        }
+
         Some(LinkSection { name, span: cx.attr_span })
     }
 }

compiler/rustc_attr_parsing/src/session_diagnostics.rs

@@ -1137,3 +1137,22 @@ pub(crate) struct UnstableAttrForAlreadyStableFeature {
     #[label("the stability attribute annotates this item")]
     pub item_span: Span,
 }
+
+#[derive(Diagnostic)]
+#[diag("invalid Mach-O section specifier")]
+pub(crate) struct InvalidMachoSection {
+    #[primary_span]
+    #[label("not a valid Mach-O section specifier")]
+    pub name_span: Span,
+    #[subdiagnostic]
+    pub reason: InvalidMachoSectionReason,
+}
+
+#[derive(Subdiagnostic)]
+pub(crate) enum InvalidMachoSectionReason {
+    #[note("a Mach-O section specifier requires a segment and a section, separated by a comma")]
+    #[help("an example of a valid Mach-O section specifier is `__TEXT,__cstring`")]
+    MissingSection,
+    #[note("section name `{$section}` is longer than 16 bytes")]
+    SectionTooLong { section: String },
+}

compiler/rustc_codegen_llvm/src/declare.rs

@@ -14,6 +14,7 @@
 use std::borrow::Borrow;
 
 use itertools::Itertools;
+use rustc_abi::AddressSpace;
 use rustc_codegen_ssa::traits::{MiscCodegenMethods, TypeMembershipCodegenMethods};
 use rustc_data_structures::fx::FxIndexSet;
 use rustc_middle::ty::{Instance, Ty};
@@ -104,6 +105,28 @@ impl<'ll, CX: Borrow<SCx<'ll>>> GenericCx<'ll, CX> {
             )
         }
     }
+
+    /// Declare a global value in a specific address space.
+    ///
+    /// If there’s a value with the same name already declared, the function will
+    /// return its Value instead.
+    pub(crate) fn declare_global_in_addrspace(
+        &self,
+        name: &str,
+        ty: &'ll Type,
+        addr_space: AddressSpace,
+    ) -> &'ll Value {
+        debug!("declare_global(name={name:?}, addrspace={addr_space:?})");
+        unsafe {
+            llvm::LLVMRustGetOrInsertGlobalInAddrspace(
+                (**self).borrow().llmod,
+                name.as_c_char_ptr(),
+                name.len(),
+                ty,
+                addr_space.0,
+            )
+        }
+    }
 }
 
 impl<'ll, 'tcx> CodegenCx<'ll, 'tcx> {

compiler/rustc_codegen_llvm/src/intrinsic.rs

@@ -3,8 +3,8 @@ use std::ffi::c_uint;
 use std::{assert_matches, iter, ptr};
 
 use rustc_abi::{
-    Align, BackendRepr, Float, HasDataLayout, Integer, NumScalableVectors, Primitive, Size,
-    WrappingRange,
+    AddressSpace, Align, BackendRepr, Float, HasDataLayout, Integer, NumScalableVectors, Primitive,
+    Size, WrappingRange,
 };
 use rustc_codegen_ssa::base::{compare_simd_types, wants_msvc_seh, wants_wasm_eh};
 use rustc_codegen_ssa::common::{IntPredicate, TypeKind};
@@ -178,6 +178,7 @@ impl<'ll, 'tcx> IntrinsicCallBuilderMethods<'tcx> for Builder<'_, 'll, 'tcx> {
         span: Span,
     ) -> Result<(), ty::Instance<'tcx>> {
         let tcx = self.tcx;
+        let llvm_version = crate::llvm_util::get_version();
 
         let name = tcx.item_name(instance.def_id());
         let fn_args = instance.args;
@@ -194,7 +195,7 @@ impl<'ll, 'tcx> IntrinsicCallBuilderMethods<'tcx> for Builder<'_, 'll, 'tcx> {
             | sym::maximum_number_nsz_f64
             | sym::maximum_number_nsz_f128
                 // Need at least LLVM 22 for `min/maximumnum` to not crash LLVM.
-                if crate::llvm_util::get_version() >= (22, 0, 0) =>
+                if llvm_version >= (22, 0, 0) =>
             {
                 let intrinsic_name = if name.as_str().starts_with("min") {
                     "llvm.minimumnum"
@@ -420,7 +421,7 @@ impl<'ll, 'tcx> IntrinsicCallBuilderMethods<'tcx> for Builder<'_, 'll, 'tcx> {
             }
 
             // FIXME move into the branch below when LLVM 22 is the lowest version we support.
-            sym::carryless_mul if crate::llvm_util::get_version() >= (22, 0, 0) => {
+            sym::carryless_mul if llvm_version >= (22, 0, 0) => {
                 let ty = args[0].layout.ty;
                 if !ty.is_integral() {
                     tcx.dcx().emit_err(InvalidMonomorphization::BasicIntegerType {
@@ -620,6 +621,46 @@ impl<'ll, 'tcx> IntrinsicCallBuilderMethods<'tcx> for Builder<'_, 'll, 'tcx> {
                 return Ok(());
             }
 
+            sym::gpu_launch_sized_workgroup_mem => {
+                // Generate an anonymous global per call, with these properties:
+                // 1. The global is in the address space for workgroup memory
+                // 2. It is an `external` global
+                // 3. It is correctly aligned for the pointee `T`
+                // All instances of extern addrspace(gpu_workgroup) globals are merged in the LLVM backend.
+                // The name is irrelevant.
+                // See https://docs.nvidia.com/cuda/cuda-c-programming-guide/#shared
+                let name = if llvm_version < (23, 0, 0) && tcx.sess.target.arch == Arch::Nvptx64 {
+                    // The auto-assigned name for extern shared globals in the nvptx backend does
+                    // not compile in ptxas. Workaround this issue by assigning a name.
+                    // Fixed in LLVM 23.
+                    "gpu_launch_sized_workgroup_mem"
+                } else {
+                    ""
+                };
+                let global = self.declare_global_in_addrspace(
+                    name,
+                    self.type_array(self.type_i8(), 0),
+                    AddressSpace::GPU_WORKGROUP,
+                );
+                let ty::RawPtr(inner_ty, _) = result.layout.ty.kind() else { unreachable!() };
+                // The alignment of the global is used to specify the *minimum* alignment that
+                // must be obeyed by the GPU runtime.
+                // When multiple of these global variables are used by a kernel, the maximum alignment is taken.
+                // See https://github.com/llvm/llvm-project/blob/a271d07488a85ce677674bbe8101b10efff58c95/llvm/lib/Target/AMDGPU/AMDGPULowerModuleLDSPass.cpp#L821
+                let alignment = self.align_of(*inner_ty).bytes() as u32;
+                unsafe {
+                    // FIXME Workaround the above issue by taking maximum alignment if the global existed
+                    if tcx.sess.target.arch == Arch::Nvptx64 {
+                        if alignment > llvm::LLVMGetAlignment(global) {
+                            llvm::LLVMSetAlignment(global, alignment);
+                        }
+                    } else {
+                        llvm::LLVMSetAlignment(global, alignment);
+                    }
+                }
+                self.cx().const_pointercast(global, self.type_ptr())
+            }
+
             sym::amdgpu_dispatch_ptr => {
                 let val = self.call_intrinsic("llvm.amdgcn.dispatch.ptr", &[], &[]);
                 // Relying on `LLVMBuildPointerCast` to produce an addrspacecast

compiler/rustc_codegen_llvm/src/llvm/ffi.rs

@@ -2003,6 +2003,13 @@ unsafe extern "C" {
         NameLen: size_t,
         T: &'a Type,
     ) -> &'a Value;
+    pub(crate) fn LLVMRustGetOrInsertGlobalInAddrspace<'a>(
+        M: &'a Module,
+        Name: *const c_char,
+        NameLen: size_t,
+        T: &'a Type,
+        AddressSpace: c_uint,
+    ) -> &'a Value;
     pub(crate) fn LLVMRustGetNamedValue(
         M: &Module,
         Name: *const c_char,

compiler/rustc_codegen_ssa/src/mir/intrinsic.rs

@@ -111,6 +111,7 @@ impl<'a, 'tcx, Bx: BuilderMethods<'a, 'tcx>> FunctionCx<'a, 'tcx, Bx> {
                 sym::abort
                 | sym::unreachable
                 | sym::cold_path
+                | sym::gpu_launch_sized_workgroup_mem
                 | sym::breakpoint
                 | sym::amdgpu_dispatch_ptr
                 | sym::assert_zero_valid

compiler/rustc_hir/src/attrs/data_structures.rs

@@ -1098,8 +1098,8 @@ pub enum AttributeKind {
 
     /// Represents [`#[link_section]`](https://doc.rust-lang.org/reference/abi.html#the-link_section-attribute)
     LinkSection {
-        name: Symbol,
         span: Span,
+        name: Symbol,
     },
 
     /// Represents `#[linkage]`.

compiler/rustc_hir_analysis/src/check/always_applicable.rs

@@ -11,7 +11,7 @@ use rustc_infer::infer::{RegionResolutionError, TyCtxtInferExt};
 use rustc_infer::traits::{ObligationCause, ObligationCauseCode};
 use rustc_middle::span_bug;
 use rustc_middle::ty::util::CheckRegions;
-use rustc_middle::ty::{self, GenericArgsRef, Ty, TyCtxt, TypingMode};
+use rustc_middle::ty::{self, GenericArgsRef, Ty, TyCtxt, TypeVisitableExt, TypingMode};
 use rustc_trait_selection::regions::InferCtxtRegionExt;
 use rustc_trait_selection::traits::{self, ObligationCtxt};
 
@@ -65,6 +65,8 @@ pub(crate) fn check_drop_impl(
                 adt_to_impl_args,
             )?;
 
+            ensure_all_fields_are_const_destruct(tcx, drop_impl_did, adt_def.did())?;
+
             ensure_impl_predicates_are_implied_by_item_defn(
                 tcx,
                 drop_impl_did,
@@ -173,6 +175,64 @@ fn ensure_impl_params_and_item_params_correspond<'tcx>(
     Err(err.emit())
 }
 
+fn ensure_all_fields_are_const_destruct<'tcx>(
+    tcx: TyCtxt<'tcx>,
+    impl_def_id: LocalDefId,
+    adt_def_id: DefId,
+) -> Result<(), ErrorGuaranteed> {
+    if !tcx.is_conditionally_const(impl_def_id) {
+        return Ok(());
+    }
+    let infcx = tcx.infer_ctxt().build(TypingMode::non_body_analysis());
+    let ocx = ObligationCtxt::new_with_diagnostics(&infcx);
+
+    let impl_span = tcx.def_span(impl_def_id.to_def_id());
+    let env =
+        ty::EarlyBinder::bind(tcx.param_env(impl_def_id)).instantiate_identity().skip_norm_wip();
+    let args = ty::GenericArgs::identity_for_item(tcx, impl_def_id);
+    let destruct_trait = tcx.lang_items().destruct_trait().unwrap();
+    for field in tcx.adt_def(adt_def_id).all_fields() {
+        let field_ty = field.ty(tcx, args);
+        let cause = traits::ObligationCause::new(
+            tcx.def_span(field.did),
+            impl_def_id,
+            ObligationCauseCode::Misc,
+        );
+        ocx.register_obligation(traits::Obligation::new(
+            tcx,
+            cause,
+            env,
+            ty::ClauseKind::HostEffect(ty::HostEffectPredicate {
+                trait_ref: ty::TraitRef::new(tcx, destruct_trait, [field_ty]),
+                constness: ty::BoundConstness::Maybe,
+            }),
+        ));
+    }
+    ocx.evaluate_obligations_error_on_ambiguity()
+        .into_iter()
+        .map(|error| {
+            let ty::ClauseKind::HostEffect(eff) =
+                error.root_obligation.predicate.expect_clause().kind().no_bound_vars().unwrap()
+            else {
+                unreachable!()
+            };
+            let field_ty = eff.trait_ref.self_ty();
+            let diag = struct_span_code_err!(
+                tcx.dcx(),
+                error.root_obligation.cause.span,
+                E0367,
+                "`{field_ty}` does not implement `[const] Destruct`",
+            )
+            .with_span_note(impl_span, "required for this `Drop` impl");
+            if field_ty.has_param() {
+                // FIXME: suggest adding `[const] Destruct` by teaching
+                // `suggest_restricting_param_bound` about const traits.
+            }
+            Err(diag.emit())
+        })
+        .collect()
+}
+
 /// Confirms that all predicates defined on the `Drop` impl (`drop_impl_def_id`) are able to be
 /// proven from within `adt_def_id`'s environment. I.e. all the predicates on the impl are
 /// implied by the ADT being well formed.

compiler/rustc_hir_analysis/src/check/intrinsic.rs

@@ -130,6 +130,7 @@ fn intrinsic_operation_unsafety(tcx: TyCtxt<'_>, intrinsic_id: LocalDefId) -> hi
         | sym::forget
         | sym::frem_algebraic
         | sym::fsub_algebraic
+        | sym::gpu_launch_sized_workgroup_mem
         | sym::is_val_statically_known
         | sym::log2f16
         | sym::log2f32
@@ -297,6 +298,7 @@ pub(crate) fn check_intrinsic_type(
         sym::field_offset => (1, 0, vec![], tcx.types.usize),
         sym::rustc_peek => (1, 0, vec![param(0)], param(0)),
         sym::caller_location => (0, 0, vec![], tcx.caller_location_ty()),
+        sym::gpu_launch_sized_workgroup_mem => (1, 0, vec![], Ty::new_mut_ptr(tcx, param(0))),
         sym::assert_inhabited | sym::assert_zero_valid | sym::assert_mem_uninitialized_valid => {
             (1, 0, vec![], tcx.types.unit)
         }