Auto merge of rust-lang#76260 - xd009642:rfc/2867, r=jonas-schievink

Implementation of RFC2867

rust-lang#74727

So I've started work on this, I think my next steps are to make use of the `instruction_set` value in the llvm codegen but this is the point where I begin to get a bit lost. I'm looking at the code but it would be nice to have some guidance on what I've currently done and what I'm doing next 😄

compiler/rustc_attr/src/builtin.rs

@@ -75,6 +75,12 @@ pub enum InlineAttr {
     Never,
 }
 
+#[derive(Clone, Encodable, Decodable)]
+pub enum InstructionSetAttr {
+    ArmA32,
+    ArmT32,
+}
+
 #[derive(Clone, Encodable, Decodable)]
 pub enum OptimizeAttr {
     None,

compiler/rustc_codegen_llvm/src/attributes.rs

@@ -18,7 +18,7 @@ use crate::attributes;
 use crate::llvm::AttributePlace::Function;
 use crate::llvm::{self, Attribute};
 use crate::llvm_util;
-pub use rustc_attr::{InlineAttr, OptimizeAttr};
+pub use rustc_attr::{InlineAttr, InstructionSetAttr, OptimizeAttr};
 
 use crate::context::CodegenCx;
 use crate::value::Value;
@@ -310,6 +310,10 @@ pub fn from_fn_attrs(cx: &CodegenCx<'ll, 'tcx>, llfn: &'ll Value, instance: ty::
             let feature = &f.as_str();
             format!("+{}", llvm_util::to_llvm_feature(cx.tcx.sess, feature))
         }))
+        .chain(codegen_fn_attrs.instruction_set.iter().map(|x| match x {
+            InstructionSetAttr::ArmA32 => "-thumb-mode".to_string(),
+            InstructionSetAttr::ArmT32 => "+thumb-mode".to_string(),
+        }))
         .collect::<Vec<String>>()
         .join(",");
 

compiler/rustc_error_codes/src/error_codes.rs

@@ -460,6 +460,8 @@ E0774: include_str!("./error_codes/E0774.md"),
 E0775: include_str!("./error_codes/E0775.md"),
 E0776: include_str!("./error_codes/E0776.md"),
 E0777: include_str!("./error_codes/E0777.md"),
+E0778: include_str!("./error_codes/E0778.md"),
+E0779: include_str!("./error_codes/E0779.md"),
 ;
 //  E0006, // merged with E0005
 //  E0008, // cannot bind by-move into a pattern guard

compiler/rustc_error_codes/src/error_codes/E0778.md

@@ -0,0 +1,35 @@
+The `instruction_set` attribute was malformed.
+
+Erroneous code example:
+
+```compile_fail,E0778
+#![feature(isa_attribute)]
+
+#[instruction_set()] // error: expected one argument
+pub fn something() {}
+fn main() {}
+```
+
+The parenthesized `instruction_set` attribute requires the parameter to be
+specified:
+
+```
+#![feature(isa_attribute)]
+
+#[cfg_attr(target_arch="arm", instruction_set(arm::a32))]
+fn something() {}
+```
+
+or:
+
+```
+#![feature(isa_attribute)]
+
+#[cfg_attr(target_arch="arm", instruction_set(arm::t32))]
+fn something() {}
+```
+
+For more information see the [`instruction_set` attribute][isa-attribute]
+section of the Reference.
+
+[isa-attribute]: https://doc.rust-lang.org/reference/attributes/codegen.html

compiler/rustc_error_codes/src/error_codes/E0779.md

@@ -0,0 +1,32 @@
+An unknown argument was given to the `instruction_set` attribute.
+
+Erroneous code example:
+
+```compile_fail,E0779
+#![feature(isa_attribute)]
+
+#[instruction_set(intel::x64)] // error: invalid argument
+pub fn something() {}
+fn main() {}
+```
+
+The `instruction_set` attribute only supports two arguments currently:
+
+ * arm::a32
+ * arm::t32
+
+All other arguments given to the `instruction_set` attribute will return this
+error. Example:
+
+```
+#![feature(isa_attribute)]
+
+#[cfg_attr(target_arch="arm", instruction_set(arm::a32))] // ok!
+pub fn something() {}
+fn main() {}
+```
+
+For more information see the [`instruction_set` attribute][isa-attribute]
+section of the Reference.
+
+[isa-attribute]: https://doc.rust-lang.org/reference/attributes/codegen.html

compiler/rustc_feature/src/active.rs

@@ -599,6 +599,9 @@ declare_features! (
     /// Allows argument and return position `impl Trait` in a `const fn`.
     (active, const_impl_trait, "1.48.0", Some(77463), None),
 
+    /// Allows `#[instruction_set(_)]` attribute
+    (active, isa_attribute, "1.48.0", Some(74727), None),
+
     // -------------------------------------------------------------------------
     // feature-group-end: actual feature gates
     // -------------------------------------------------------------------------

compiler/rustc_feature/src/builtin_attrs.rs

@@ -336,6 +336,8 @@ pub const BUILTIN_ATTRIBUTES: &[BuiltinAttribute] = &[
         optimize, AssumedUsed, template!(List: "size|speed"), optimize_attribute,
         experimental!(optimize),
     ),
+    // RFC 2867
+    gated!(instruction_set, AssumedUsed, template!(List: "set"), isa_attribute, experimental!(instruction_set)),
 
     gated!(ffi_returns_twice, AssumedUsed, template!(Word), experimental!(ffi_returns_twice)),
     gated!(ffi_pure, AssumedUsed, template!(Word), experimental!(ffi_pure)),

compiler/rustc_middle/src/ich/impls_hir.rs

@@ -221,6 +221,12 @@ impl<'hir> HashStable<StableHashingContext<'hir>> for attr::InlineAttr {
     }
 }
 
+impl<'hir> HashStable<StableHashingContext<'hir>> for attr::InstructionSetAttr {
+    fn hash_stable(&self, hcx: &mut StableHashingContext<'hir>, hasher: &mut StableHasher) {
+        mem::discriminant(self).hash_stable(hcx, hasher);
+    }
+}
+
 impl<'hir> HashStable<StableHashingContext<'hir>> for attr::OptimizeAttr {
     fn hash_stable(&self, hcx: &mut StableHashingContext<'hir>, hasher: &mut StableHasher) {
         mem::discriminant(self).hash_stable(hcx, hasher);

compiler/rustc_middle/src/middle/codegen_fn_attrs.rs

@@ -1,5 +1,5 @@
 use crate::mir::mono::Linkage;
-use rustc_attr::{InlineAttr, OptimizeAttr};
+use rustc_attr::{InlineAttr, InstructionSetAttr, OptimizeAttr};
 use rustc_session::config::SanitizerSet;
 use rustc_span::symbol::Symbol;
 
@@ -34,6 +34,10 @@ pub struct CodegenFnAttrs {
     /// The `#[no_sanitize(...)]` attribute. Indicates sanitizers for which
     /// instrumentation should be disabled inside the annotated function.
     pub no_sanitize: SanitizerSet,
+    /// The `#[instruction_set(set)]` attribute. Indicates if the generated code should
+    /// be generated against a specific instruction set. Only usable on architectures which allow
+    /// switching between multiple instruction sets.
+    pub instruction_set: Option<InstructionSetAttr>,
 }
 
 bitflags! {
@@ -98,6 +102,7 @@ impl CodegenFnAttrs {
             linkage: None,
             link_section: None,
             no_sanitize: SanitizerSet::empty(),
+            instruction_set: None,
         }
     }
 

compiler/rustc_span/src/symbol.rs

@@ -212,6 +212,7 @@ symbols! {
         _d,
         _e,
         _task_context,
+        a32,
         aarch64_target_feature,
         abi,
         abi_amdgpu_kernel,
@@ -256,6 +257,7 @@ symbols! {
         arbitrary_enum_discriminant,
         arbitrary_self_types,
         arith_offset,
+        arm,
         arm_target_feature,
         array,
         arrays,
@@ -592,11 +594,13 @@ symbols! {
         inlateout,
         inline,
         inout,
+        instruction_set,
         intel,
         into_iter,
         into_result,
         intrinsics,
         irrefutable_let_patterns,
+        isa_attribute,
         isize,
         issue,
         issue_5723_bootstrap,
@@ -1065,6 +1069,7 @@ symbols! {
         sym,
         sync,
         sync_trait,
+        t32,
         target_arch,
         target_endian,
         target_env,

compiler/rustc_target/src/spec/armv4t_unknown_linux_gnueabi.rs

@@ -20,6 +20,7 @@ pub fn target() -> TargetResult {
             max_atomic_width: Some(32),
             unsupported_abis: super::arm_base::unsupported_abis(),
             target_mcount: "\u{1}__gnu_mcount_nc".to_string(),
+            has_thumb_interworking: true,
             ..base
         },
     })

compiler/rustc_target/src/spec/armv5te_unknown_linux_gnueabi.rs

@@ -20,6 +20,7 @@ pub fn target() -> TargetResult {
             max_atomic_width: Some(32),
             unsupported_abis: super::arm_base::unsupported_abis(),
             target_mcount: "\u{1}__gnu_mcount_nc".to_string(),
+            has_thumb_interworking: true,
             ..base
         },
     })

compiler/rustc_target/src/spec/armv5te_unknown_linux_musleabi.rs

@@ -23,6 +23,7 @@ pub fn target() -> TargetResult {
             max_atomic_width: Some(32),
             unsupported_abis: super::arm_base::unsupported_abis(),
             target_mcount: "\u{1}mcount".to_string(),
+            has_thumb_interworking: true,
             ..base
         },
     })

compiler/rustc_target/src/spec/mod.rs

@@ -994,6 +994,10 @@ pub struct TargetOptions {
     /// used to locate unwinding information is passed
     /// (only has effect if the linker is `ld`-like).
     pub eh_frame_header: bool,
+
+    /// Is true if the target is an ARM architecture using thumb v1 which allows for
+    /// thumb and arm interworking.
+    pub has_thumb_interworking: bool,
 }
 
 impl Default for TargetOptions {
@@ -1086,6 +1090,7 @@ impl Default for TargetOptions {
             llvm_args: vec![],
             use_ctors_section: false,
             eh_frame_header: true,
+            has_thumb_interworking: false,
         }
     }
 }
@@ -1479,6 +1484,7 @@ impl Target {
         key!(llvm_args, list);
         key!(use_ctors_section, bool);
         key!(eh_frame_header, bool);
+        key!(has_thumb_interworking, bool);
 
         // NB: The old name is deprecated, but support for it is retained for
         // compatibility.
@@ -1717,6 +1723,7 @@ impl ToJson for Target {
         target_option_val!(llvm_args);
         target_option_val!(use_ctors_section);
         target_option_val!(eh_frame_header);
+        target_option_val!(has_thumb_interworking);
 
         if default.unsupported_abis != self.options.unsupported_abis {
             d.insert(

compiler/rustc_target/src/spec/thumbv4t_none_eabi.rs

@@ -55,6 +55,7 @@ pub fn target() -> TargetResult {
 
             // don't have atomic compare-and-swap
             atomic_cas: false,
+            has_thumb_interworking: true,
 
             ..super::thumb_base::opts()
         },

compiler/rustc_typeck/src/collect.rs

@@ -21,8 +21,8 @@ use crate::constrained_generic_params as cgp;
 use crate::errors;
 use crate::middle::resolve_lifetime as rl;
 use rustc_ast as ast;
-use rustc_ast::MetaItemKind;
-use rustc_attr::{list_contains_name, InlineAttr, OptimizeAttr};
+use rustc_ast::{MetaItemKind, NestedMetaItem};
+use rustc_attr::{list_contains_name, InlineAttr, InstructionSetAttr, OptimizeAttr};
 use rustc_data_structures::captures::Captures;
 use rustc_data_structures::fx::{FxHashMap, FxHashSet, FxIndexSet};
 use rustc_errors::{struct_span_err, Applicability};
@@ -2647,6 +2647,75 @@ fn codegen_fn_attrs(tcx: TyCtxt<'_>, id: DefId) -> CodegenFnAttrs {
                     }
                 }
             }
+        } else if tcx.sess.check_name(attr, sym::instruction_set) {
+            codegen_fn_attrs.instruction_set = match attr.meta().map(|i| i.kind) {
+                Some(MetaItemKind::List(ref items)) => match items.as_slice() {
+                    [NestedMetaItem::MetaItem(set)] => {
+                        let segments =
+                            set.path.segments.iter().map(|x| x.ident.name).collect::<Vec<_>>();
+                        match segments.as_slice() {
+                            [sym::arm, sym::a32] | [sym::arm, sym::t32] => {
+                                if !tcx.sess.target.target.options.has_thumb_interworking {
+                                    struct_span_err!(
+                                        tcx.sess.diagnostic(),
+                                        attr.span,
+                                        E0779,
+                                        "target does not support `#[instruction_set]`"
+                                    )
+                                    .emit();
+                                    None
+                                } else if segments[1] == sym::a32 {
+                                    Some(InstructionSetAttr::ArmA32)
+                                } else if segments[1] == sym::t32 {
+                                    Some(InstructionSetAttr::ArmT32)
+                                } else {
+                                    unreachable!()
+                                }
+                            }
+                            _ => {
+                                struct_span_err!(
+                                    tcx.sess.diagnostic(),
+                                    attr.span,
+                                    E0779,
+                                    "invalid instruction set specified",
+                                )
+                                .emit();
+                                None
+                            }
+                        }
+                    }
+                    [] => {
+                        struct_span_err!(
+                            tcx.sess.diagnostic(),
+                            attr.span,
+                            E0778,
+                            "`#[instruction_set]` requires an argument"
+                        )
+                        .emit();
+                        None
+                    }
+                    _ => {
+                        struct_span_err!(
+                            tcx.sess.diagnostic(),
+                            attr.span,
+                            E0779,
+                            "cannot specify more than one instruction set"
+                        )
+                        .emit();
+                        None
+                    }
+                },
+                _ => {
+                    struct_span_err!(
+                        tcx.sess.diagnostic(),
+                        attr.span,
+                        E0778,
+                        "must specify an instruction set"
+                    )
+                    .emit();
+                    None
+                }
+            };
         }
     }
 

src/test/ui/error-codes/E0778.rs

@@ -0,0 +1,8 @@
+#![feature(isa_attribute)]
+
+#[instruction_set()] //~ ERROR
+fn no_isa_defined() {
+}
+
+fn main() {
+}

src/test/ui/error-codes/E0778.stderr

@@ -0,0 +1,9 @@
+error[E0778]: `#[instruction_set]` requires an argument
+  --> $DIR/E0778.rs:3:1
+   |
+LL | #[instruction_set()]
+   | ^^^^^^^^^^^^^^^^^^^^
+
+error: aborting due to previous error
+
+For more information about this error, try `rustc --explain E0778`.

src/test/ui/error-codes/E0779.rs

@@ -0,0 +1,6 @@
+#![feature(isa_attribute)]
+
+#[instruction_set(arm::magic)] //~ ERROR
+fn main() {
+
+}

src/test/ui/error-codes/E0779.stderr

@@ -0,0 +1,9 @@
+error[E0779]: invalid instruction set specified
+  --> $DIR/E0779.rs:3:1
+   |
+LL | #[instruction_set(arm::magic)]
+   | ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+
+error: aborting due to previous error
+
+For more information about this error, try `rustc --explain E0779`.