Auto merge of rust-lang#118534 - RalfJung:extern-type-size-of-val, r=WaffleLapkin

codegen: panic when trying to compute size/align of extern type

The alignment is also computed when accessing a field of extern type at non-zero offset, so we also panic in that case.

Previously `size_of_val` worked because the code path there assumed that "thin pointer" means "sized". But that's not true any more with extern types. The returned size and align are just blatantly wrong, so it seems better to panic than returning wrong results. We use a non-unwinding panic since code probably does not expect size_of_val to panic.

Author: bors

compiler/rustc_codegen_cranelift/build_system/tests.rs

@@ -75,11 +75,6 @@ const BASE_SYSROOT_SUITE: &[TestCase] = &[
         "example/arbitrary_self_types_pointers_and_wrappers.rs",
         &[],
     ),
-    TestCase::build_bin_and_run(
-        "aot.issue_91827_extern_types",
-        "example/issue-91827-extern-types.rs",
-        &[],
-    ),
     TestCase::build_lib("build.alloc_system", "example/alloc_system.rs", "lib"),
     TestCase::build_bin_and_run("aot.alloc_example", "example/alloc_example.rs", &[]),
     TestCase::jit_bin("jit.std_example", "example/std_example.rs", ""),

compiler/rustc_codegen_cranelift/example/issue-91827-extern-types.rs

@@ -51,10 +51,10 @@ pub mod codegen_attrs;
 pub mod common;
 pub mod debuginfo;
 pub mod errors;
-pub mod glue;
 pub mod meth;
 pub mod mir;
 pub mod mono_item;
+pub mod size_of_val;
 pub mod target_features;
 pub mod traits;
 

compiler/rustc_codegen_ssa/src/lib.rs

@@ -4,8 +4,8 @@ use super::FunctionCx;
 use crate::common::IntPredicate;
 use crate::errors;
 use crate::errors::InvalidMonomorphization;
-use crate::glue;
 use crate::meth;
+use crate::size_of_val;
 use crate::traits::*;
 use crate::MemFlags;
 
@@ -88,21 +88,17 @@ impl<'a, 'tcx, Bx: BuilderMethods<'a, 'tcx>> FunctionCx<'a, 'tcx, Bx> {
             sym::va_end => bx.va_end(args[0].immediate()),
             sym::size_of_val => {
                 let tp_ty = fn_args.type_at(0);
-                if let OperandValue::Pair(_, meta) = args[0].val {
-                    let (llsize, _) = glue::size_and_align_of_dst(bx, tp_ty, Some(meta));
-                    llsize
-                } else {
-                    bx.const_usize(bx.layout_of(tp_ty).size.bytes())
-                }
+                let meta =
+                    if let OperandValue::Pair(_, meta) = args[0].val { Some(meta) } else { None };
+                let (llsize, _) = size_of_val::size_and_align_of_dst(bx, tp_ty, meta);
+                llsize
             }
             sym::min_align_of_val => {
                 let tp_ty = fn_args.type_at(0);
-                if let OperandValue::Pair(_, meta) = args[0].val {
-                    let (_, llalign) = glue::size_and_align_of_dst(bx, tp_ty, Some(meta));
-                    llalign
-                } else {
-                    bx.const_usize(bx.layout_of(tp_ty).align.abi.bytes())
-                }
+                let meta =
+                    if let OperandValue::Pair(_, meta) = args[0].val { Some(meta) } else { None };
+                let (_, llalign) = size_of_val::size_and_align_of_dst(bx, tp_ty, meta);
+                llalign
             }
             sym::vtable_size | sym::vtable_align => {
                 let vtable = args[0].immediate();

compiler/rustc_codegen_ssa/src/mir/intrinsic.rs

@@ -2,7 +2,7 @@ use super::place::PlaceRef;
 use super::{FunctionCx, LocalRef};
 
 use crate::base;
-use crate::glue;
+use crate::size_of_val;
 use crate::traits::*;
 use crate::MemFlags;
 
@@ -466,13 +466,13 @@ impl<'a, 'tcx, V: CodegenObject> OperandValue<V> {
             .ty;
 
         let OperandValue::Ref(llptr, Some(llextra), _) = self else {
-            bug!("store_unsized called with a sized value")
+            bug!("store_unsized called with a sized value (or with an extern type)")
         };
 
         // Allocate an appropriate region on the stack, and copy the value into it. Since alloca
         // doesn't support dynamic alignment, we allocate an extra align - 1 bytes, and align the
         // pointer manually.
-        let (size, align) = glue::size_and_align_of_dst(bx, unsized_ty, Some(llextra));
+        let (size, align) = size_of_val::size_and_align_of_dst(bx, unsized_ty, Some(llextra));
         let one = bx.const_usize(1);
         let align_minus_1 = bx.sub(align, one);
         let size_extra = bx.add(size, align_minus_1);

compiler/rustc_codegen_ssa/src/mir/operand.rs

@@ -2,7 +2,7 @@ use super::operand::OperandValue;
 use super::{FunctionCx, LocalRef};
 
 use crate::common::IntPredicate;
-use crate::glue;
+use crate::size_of_val;
 use crate::traits::*;
 
 use rustc_middle::mir;
@@ -99,6 +99,8 @@ impl<'a, 'tcx, V: CodegenObject> PlaceRef<'tcx, V> {
         let offset = self.layout.fields.offset(ix);
         let effective_field_align = self.align.restrict_for_offset(offset);
 
+        // `simple` is called when we don't need to adjust the offset to
+        // the dynamic alignment of the field.
         let mut simple = || {
             let llval = match self.layout.abi {
                 _ if offset.bytes() == 0 => {
@@ -141,28 +143,21 @@ impl<'a, 'tcx, V: CodegenObject> PlaceRef<'tcx, V> {
         };
 
         // Simple cases, which don't need DST adjustment:
-        //   * no metadata available - just log the case
-        //   * known alignment - sized types, `[T]`, `str` or a foreign type
+        //   * known alignment - sized types, `[T]`, `str`
+        //   * offset 0 -- rounding up to alignment cannot change the offset
         // Note that looking at `field.align` is incorrect since that is not necessarily equal
         // to the dynamic alignment of the type.
         match field.ty.kind() {
-            _ if self.llextra.is_none() => {
-                debug!(
-                    "unsized field `{}`, of `{:?}` has no metadata for adjustment",
-                    ix, self.llval
-                );
-                return simple();
-            }
             _ if field.is_sized() => return simple(),
-            ty::Slice(..) | ty::Str | ty::Foreign(..) => return simple(),
+            ty::Slice(..) | ty::Str => return simple(),
+            _ if offset.bytes() == 0 => return simple(),
             _ => {}
         }
 
         // We need to get the pointer manually now.
         // We do this by casting to a `*i8`, then offsetting it by the appropriate amount.
         // We do this instead of, say, simply adjusting the pointer from the result of a GEP
-        // because the field may have an arbitrary alignment in the LLVM representation
-        // anyway.
+        // because the field may have an arbitrary alignment in the LLVM representation.
         //
         // To demonstrate:
         //
@@ -179,7 +174,7 @@ impl<'a, 'tcx, V: CodegenObject> PlaceRef<'tcx, V> {
         let unaligned_offset = bx.cx().const_usize(offset.bytes());
 
         // Get the alignment of the field
-        let (_, mut unsized_align) = glue::size_and_align_of_dst(bx, field.ty, meta);
+        let (_, mut unsized_align) = size_of_val::size_and_align_of_dst(bx, field.ty, meta);
 
         // For packed types, we need to cap alignment.
         if let ty::Adt(def, _) = self.layout.ty.kind()

compiler/rustc_codegen_ssa/src/mir/place.rs

@@ -1,10 +1,11 @@
-//!
-//
-// Code relating to drop glue.
+//! Computing the size and alignment of a value.
 
+use crate::common;
 use crate::common::IntPredicate;
 use crate::meth;
 use crate::traits::*;
+use rustc_hir::LangItem;
+use rustc_middle::ty::print::{with_no_trimmed_paths, with_no_visible_paths};
 use rustc_middle::ty::{self, Ty};
 use rustc_target::abi::WrappingRange;
 
@@ -14,7 +15,7 @@ pub fn size_and_align_of_dst<'a, 'tcx, Bx: BuilderMethods<'a, 'tcx>>(
     info: Option<Bx::Value>,
 ) -> (Bx::Value, Bx::Value) {
     let layout = bx.layout_of(t);
-    debug!("size_and_align_of_dst(ty={}, info={:?}): layout: {:?}", t, info, layout);
+    trace!("size_and_align_of_dst(ty={}, info={:?}): layout: {:?}", t, info, layout);
     if layout.is_sized() {
         let size = bx.const_usize(layout.size.bytes());
         let align = bx.const_usize(layout.align.abi.bytes());
@@ -51,7 +52,31 @@ pub fn size_and_align_of_dst<'a, 'tcx, Bx: BuilderMethods<'a, 'tcx>>(
                 bx.const_usize(unit.align.abi.bytes()),
             )
         }
-        _ => {
+        ty::Foreign(_) => {
+            // `extern` type. We cannot compute the size, so panic.
+            let msg_str = with_no_visible_paths!({
+                with_no_trimmed_paths!({
+                    format!("attempted to compute the size or alignment of extern type `{t}`")
+                })
+            });
+            let msg = bx.const_str(&msg_str);
+
+            // Obtain the panic entry point.
+            let (fn_abi, llfn) = common::build_langcall(bx, None, LangItem::PanicNounwind);
+
+            // Generate the call.
+            // Cannot use `do_call` since we don't have a MIR terminator so we can't create a `TerminationCodegenHelper`.
+            // (But we are in good company, this code is duplicated plenty of times.)
+            let fn_ty = bx.fn_decl_backend_type(fn_abi);
+
+            bx.call(fn_ty, /* fn_attrs */ None, Some(fn_abi), llfn, &[msg.0, msg.1], None);
+
+            // This function does not return so we can now return whatever we want.
+            let size = bx.const_usize(layout.size.bytes());
+            let align = bx.const_usize(layout.align.abi.bytes());
+            (size, align)
+        }
+        ty::Adt(..) | ty::Tuple(..) => {
             // First get the size of all statically known fields.
             // Don't use size_of because it also rounds up to alignment, which we
             // want to avoid, as the unsized field's alignment could be smaller.
@@ -122,5 +147,6 @@ pub fn size_and_align_of_dst<'a, 'tcx, Bx: BuilderMethods<'a, 'tcx>>(
 
             (size, align)
         }
+        _ => bug!("size_and_align_of_dst: {t} not supported"),
     }
 }

compiler/rustc_codegen_ssa/src/glue.rs renamed to compiler/rustc_codegen_ssa/src/size_of_val.rs

@@ -174,12 +174,15 @@ where
                     };
                     (base_meta, offset.align_to(align))
                 }
-                None => {
-                    // For unsized types with an extern type tail we perform no adjustments.
-                    // NOTE: keep this in sync with `PlaceRef::project_field` in the codegen backend.
-                    assert!(matches!(base_meta, MemPlaceMeta::None));
+                None if offset == Size::ZERO => {
+                    // If the offset is 0, then rounding it up to alignment wouldn't change anything,
+                    // so we can do this even for types where we cannot determine the alignment.
                     (base_meta, offset)
                 }
+                None => {
+                    // We don't know the alignment of this field, so we cannot adjust.
+                    throw_unsup_format!("`extern type` does not have a known offset")
+                }
             }
         } else {
             // base_meta could be present; we might be accessing a sized field of an unsized

compiler/rustc_const_eval/src/interpret/projection.rs

@@ -0,0 +1,29 @@
+#![feature(extern_types)]
+
+extern "C" {
+    type Opaque;
+}
+
+struct Newtype(Opaque);
+
+struct S {
+    i: i32,
+    j: i32,
+    a: Newtype,
+}
+
+fn main() {
+    let buf = [0i32; 4];
+
+    let x: &Newtype = unsafe { &*(&buf as *const _ as *const Newtype) };
+    // Projecting to the newtype works, because it is always at offset 0.
+    let _field = &x.0;
+
+    let x: &S = unsafe { &*(&buf as *const _ as *const S) };
+    // Accessing sized fields is perfectly fine, even at non-zero offsets.
+    let _field = &x.i;
+    let _field = &x.j;
+    // This needs to compute the field offset, but we don't know the type's alignment,
+    // so this panics.
+    let _field = &x.a; //~ERROR: does not have a known offset
+}

src/tools/miri/tests/fail/extern-type-field-offset.rs

@@ -0,0 +1,14 @@
+error: unsupported operation: `extern type` does not have a known offset
+  --> $DIR/extern-type-field-offset.rs:LL:CC
+   |
+LL |     let _field = &x.a;
+   |                  ^^^^ `extern type` does not have a known offset
+   |
+   = help: this is likely not a bug in the program; it indicates that the program performed an operation that the interpreter does not support
+   = note: BACKTRACE:
+   = note: inside `main` at $DIR/extern-type-field-offset.rs:LL:CC
+
+note: some details are omitted, run with `MIRIFLAGS=-Zmiri-backtrace=full` for a verbose backtrace
+
+error: aborting due to 1 previous error
+