Add c_enum_min_bits to target spec

Author: Manishearth

compiler/rustc_middle/src/ty/layout.rs

@@ -112,9 +112,6 @@ impl IntegerExt for Integer {
         let unsigned_fit = Integer::fit_unsigned(cmp::max(min as u128, max as u128));
         let signed_fit = cmp::max(Integer::fit_signed(min), Integer::fit_signed(max));
 
-        let mut min_from_extern = None;
-        let min_default = I8;
-
         if let Some(ity) = repr.int {
             let discr = Integer::from_attr(&tcx, ity);
             let fit = if ity.is_signed() { signed_fit } else { unsigned_fit };
@@ -128,19 +125,14 @@ impl IntegerExt for Integer {
             return (discr, ity.is_signed());
         }
 
-        if repr.c() {
-            match &tcx.sess.target.arch[..] {
-                "hexagon" => min_from_extern = Some(I8),
-                // WARNING: the ARM EABI has two variants; the one corresponding
-                // to `at_least == I32` appears to be used on Linux and NetBSD,
-                // but some systems may use the variant corresponding to no
-                // lower bound. However, we don't run on those yet...?
-                "arm" => min_from_extern = Some(I32),
-                _ => min_from_extern = Some(I32),
-            }
-        }
-
-        let at_least = min_from_extern.unwrap_or(min_default);
+        let at_least = if repr.c() {
+            // This is usually I32, however it can be different on some platforms,
+            // notably hexagon and arm-none/thumb-none
+            tcx.data_layout().c_enum_min_size
+        } else {
+            // repr(Rust) enums try to be as small as possible
+            I8
+        };
 
         // If there are no negative values, we can use the unsigned fit.
         if min >= 0 {

compiler/rustc_target/src/abi/mod.rs

@@ -36,6 +36,9 @@ pub struct TargetDataLayout {
     pub vector_align: Vec<(Size, AbiAndPrefAlign)>,
 
     pub instruction_address_space: AddressSpace,
+
+    /// Minimum size of #[repr(C)] enums (default I32 bits)
+    pub c_enum_min_size: Integer,
 }
 
 impl Default for TargetDataLayout {
@@ -60,6 +63,7 @@ impl Default for TargetDataLayout {
                 (Size::from_bits(128), AbiAndPrefAlign::new(align(128))),
             ],
             instruction_address_space: AddressSpace::DATA,
+            c_enum_min_size: Integer::I32,
         }
     }
 }
@@ -173,6 +177,8 @@ impl TargetDataLayout {
             ));
         }
 
+        dl.c_enum_min_size = Integer::from_size(Size::from_bits(target.c_enum_min_bits))?;
+
         Ok(dl)
     }
 
@@ -610,6 +616,17 @@ impl Integer {
         }
         I8
     }
+
+    fn from_size(size: Size) -> Result<Self, String> {
+        match size.bits() {
+            8 => Ok(Integer::I8),
+            16 => Ok(Integer::I16),
+            32 => Ok(Integer::I32),
+            64 => Ok(Integer::I64),
+            128 => Ok(Integer::I128),
+            _ => Err(format!("rust does not support integers with {} bits", size.bits())),
+        }
+    }
 }
 
 /// Fundamental unit of memory access and layout.

compiler/rustc_target/src/spec/hexagon_unknown_linux_musl.rs

@@ -13,6 +13,8 @@ pub fn target() -> Target {
     base.dynamic_linking = true;
     base.executables = true;
 
+    base.c_enum_min_bits = 8;
+
     Target {
         llvm_target: "hexagon-unknown-linux-musl".to_string(),
         pointer_width: 32,

compiler/rustc_target/src/spec/mod.rs

@@ -1335,6 +1335,9 @@ pub struct TargetOptions {
 
     /// If present it's a default value to use for adjusting the C ABI.
     pub default_adjusted_cabi: Option<Abi>,
+
+    /// Minimum number of bits in #[repr(C)] enum. Defaults to 32.
+    pub c_enum_min_bits: u64,
 }
 
 impl Default for TargetOptions {
@@ -1439,6 +1442,7 @@ impl Default for TargetOptions {
             split_debuginfo: SplitDebuginfo::Off,
             supported_sanitizers: SanitizerSet::empty(),
             default_adjusted_cabi: None,
+            c_enum_min_bits: 32,
         }
     }
 }
@@ -1603,6 +1607,12 @@ impl Target {
                     base.$key_name = s;
                 }
             } );
+            ($key_name:ident, u64) => ( {
+                let name = (stringify!($key_name)).replace("_", "-");
+                if let Some(s) = obj.remove_key(&name).and_then(|j| Json::as_u64(&j)) {
+                    base.$key_name = s;
+                }
+            } );
             ($key_name:ident, Option<u32>) => ( {
                 let name = (stringify!($key_name)).replace("_", "-");
                 if let Some(s) = obj.remove_key(&name).and_then(|j| Json::as_u64(&j)) {
@@ -2016,6 +2026,7 @@ impl Target {
         key!(split_debuginfo, SplitDebuginfo)?;
         key!(supported_sanitizers, SanitizerSet)?;
         key!(default_adjusted_cabi, Option<Abi>)?;
+        key!(c_enum_min_bits, u64);
 
         if base.is_builtin {
             // This can cause unfortunate ICEs later down the line.
@@ -2254,6 +2265,7 @@ impl ToJson for Target {
         target_option_val!(has_thumb_interworking);
         target_option_val!(split_debuginfo);
         target_option_val!(supported_sanitizers);
+        target_option_val!(c_enum_min_bits);
 
         if let Some(abi) = self.default_adjusted_cabi {
             d.insert("default-adjusted-cabi".to_string(), Abi::name(abi).to_json());

src/test/ui/layout/thumb-enum.rs

@@ -0,0 +1,34 @@
+// compile-flags: --target thumbv8m.main-none-eabihf
+// needs-llvm-components: arm
+//
+// Verify that thumb targets implement the repr(C) for enums correctly.
+//
+// See #87917
+#![feature(never_type, rustc_attrs, no_core, lang_items)]
+#![crate_type = "lib"]
+#![no_core]
+
+#[lang="sized"]
+trait Sized {}
+
+#[rustc_layout(debug)]
+#[repr(C)]
+enum A { Apple } //~ ERROR: layout_of
+
+#[rustc_layout(debug)]
+#[repr(C)]
+enum B { Banana = 255, } //~ ERROR: layout_of
+
+#[rustc_layout(debug)]
+#[repr(C)]
+enum C { Chaenomeles = 256, } //~ ERROR: layout_of
+
+#[rustc_layout(debug)]
+#[repr(C)]
+enum P { Peach = 0x1000_0000isize, } //~ ERROR: layout_of
+
+const TANGERINE: usize = 0x8100_0000; // hack to get negative numbers without negation operator!
+
+#[rustc_layout(debug)]
+#[repr(C)]
+enum T { Tangerine = TANGERINE as isize } //~ ERROR: layout_of