Auto merge of rust-lang#116892 - ojeda:rethunk, r=wesleywiser

Add `-Zfunction-return={keep,thunk-extern}` option

This is intended to be used for Linux kernel RETHUNK builds.

With this commit (optionally backported to Rust 1.73.0), plus a patched Linux kernel to pass the flag, I get a RETHUNK build with Rust enabled that is `objtool`-warning-free and is able to boot in QEMU and load a sample Rust kernel module.

Issue: rust-lang#116853.

Author: bors

compiler/rustc_codegen_llvm/src/attributes.rs

@@ -4,7 +4,7 @@ use rustc_codegen_ssa::traits::*;
 use rustc_hir::def_id::DefId;
 use rustc_middle::middle::codegen_fn_attrs::CodegenFnAttrFlags;
 use rustc_middle::ty::{self, TyCtxt};
-use rustc_session::config::OptLevel;
+use rustc_session::config::{FunctionReturn, OptLevel};
 use rustc_span::symbol::sym;
 use rustc_target::spec::abi::Abi;
 use rustc_target::spec::{FramePointer, SanitizerSet, StackProbeType, StackProtector};
@@ -118,6 +118,15 @@ pub fn frame_pointer_type_attr<'ll>(cx: &CodegenCx<'ll, '_>) -> Option<&'ll Attr
     Some(llvm::CreateAttrStringValue(cx.llcx, "frame-pointer", attr_value))
 }
 
+fn function_return_attr<'ll>(cx: &CodegenCx<'ll, '_>) -> Option<&'ll Attribute> {
+    let function_return_attr = match cx.sess().opts.unstable_opts.function_return {
+        FunctionReturn::Keep => return None,
+        FunctionReturn::ThunkExtern => AttributeKind::FnRetThunkExtern,
+    };
+
+    Some(function_return_attr.create_attr(cx.llcx))
+}
+
 /// Tell LLVM what instrument function to insert.
 #[inline]
 fn instrument_function_attr<'ll>(cx: &CodegenCx<'ll, '_>) -> SmallVec<[&'ll Attribute; 4]> {
@@ -331,8 +340,9 @@ pub fn from_fn_attrs<'ll, 'tcx>(
         to_add.push(llvm::CreateAttrString(cx.llcx, "use-sample-profile"));
     }
 
-    // FIXME: none of these three functions interact with source level attributes.
+    // FIXME: none of these functions interact with source level attributes.
     to_add.extend(frame_pointer_type_attr(cx));
+    to_add.extend(function_return_attr(cx));
     to_add.extend(instrument_function_attr(cx));
     to_add.extend(nojumptables_attr(cx));
     to_add.extend(probestack_attr(cx));

compiler/rustc_codegen_llvm/src/llvm/ffi.rs

@@ -200,6 +200,7 @@ pub enum AttributeKind {
     AllocatedPointer = 38,
     AllocAlign = 39,
     SanitizeSafeStack = 40,
+    FnRetThunkExtern = 41,
 }
 
 /// LLVMIntPredicate

compiler/rustc_interface/src/tests.rs

@@ -5,10 +5,11 @@ use rustc_errors::{emitter::HumanReadableErrorType, registry, ColorConfig};
 use rustc_session::config::{
     build_configuration, build_session_options, rustc_optgroups, BranchProtection, CFGuard, Cfg,
     DebugInfo, DumpMonoStatsFormat, ErrorOutputType, ExternEntry, ExternLocation, Externs,
-    InliningThreshold, Input, InstrumentCoverage, InstrumentXRay, LinkSelfContained,
-    LinkerPluginLto, LocationDetail, LtoCli, MirSpanview, OomStrategy, Options, OutFileName,
-    OutputType, OutputTypes, PAuthKey, PacRet, Passes, Polonius, ProcMacroExecutionStrategy, Strip,
-    SwitchWithOptPath, SymbolManglingVersion, TraitSolver, WasiExecModel,
+    FunctionReturn, InliningThreshold, Input, InstrumentCoverage, InstrumentXRay,
+    LinkSelfContained, LinkerPluginLto, LocationDetail, LtoCli, MirSpanview, OomStrategy, Options,
+    OutFileName, OutputType, OutputTypes, PAuthKey, PacRet, Passes, Polonius,
+    ProcMacroExecutionStrategy, Strip, SwitchWithOptPath, SymbolManglingVersion, TraitSolver,
+    WasiExecModel,
 };
 use rustc_session::lint::Level;
 use rustc_session::search_paths::SearchPath;
@@ -758,6 +759,7 @@ fn test_unstable_options_tracking_hash() {
     tracked!(flatten_format_args, false);
     tracked!(force_unstable_if_unmarked, true);
     tracked!(fuel, Some(("abc".to_string(), 99)));
+    tracked!(function_return, FunctionReturn::ThunkExtern);
     tracked!(function_sections, Some(false));
     tracked!(human_readable_cgu_names, true);
     tracked!(incremental_ignore_spans, true);

compiler/rustc_llvm/llvm-wrapper/LLVMWrapper.h

@@ -89,6 +89,7 @@ enum LLVMRustAttribute {
   AllocatedPointer = 38,
   AllocAlign = 39,
   SanitizeSafeStack = 40,
+  FnRetThunkExtern = 41,
 };
 
 typedef struct OpaqueRustString *RustStringRef;

compiler/rustc_llvm/llvm-wrapper/RustWrapper.cpp

@@ -278,6 +278,8 @@ static Attribute::AttrKind fromRust(LLVMRustAttribute Kind) {
     return Attribute::AllocAlign;
   case SanitizeSafeStack:
     return Attribute::SafeStack;
+  case FnRetThunkExtern:
+    return Attribute::FnRetThunkExtern;
   }
   report_fatal_error("bad AttributeKind");
 }

compiler/rustc_session/messages.ftl

@@ -26,6 +26,10 @@ session_file_is_not_writeable = output file {$file} is not writeable -- check it
 
 session_file_write_fail = failed to write `{$path}` due to error `{$err}`
 
+session_function_return_requires_x86_or_x86_64 = `-Zfunction-return` (except `keep`) is only supported on x86 and x86_64
+
+session_function_return_thunk_extern_requires_non_large_code_model = `-Zfunction-return=thunk-extern` is only supported on non-large code models
+
 session_hexadecimal_float_literal_not_supported = hexadecimal float literal is not supported
 
 session_incompatible_linker_flavor = linker flavor `{$flavor}` is incompatible with the current target

compiler/rustc_session/src/config.rs

@@ -3164,9 +3164,9 @@ impl PpMode {
 pub(crate) mod dep_tracking {
     use super::{
         BranchProtection, CFGuard, CFProtection, CrateType, DebugInfo, DebugInfoCompression,
-        ErrorOutputType, InliningThreshold, InstrumentCoverage, InstrumentXRay, LinkerPluginLto,
-        LocationDetail, LtoCli, OomStrategy, OptLevel, OutFileName, OutputType, OutputTypes,
-        Polonius, RemapPathScopeComponents, ResolveDocLinks, SourceFileHashAlgorithm,
+        ErrorOutputType, FunctionReturn, InliningThreshold, InstrumentCoverage, InstrumentXRay,
+        LinkerPluginLto, LocationDetail, LtoCli, OomStrategy, OptLevel, OutFileName, OutputType,
+        OutputTypes, Polonius, RemapPathScopeComponents, ResolveDocLinks, SourceFileHashAlgorithm,
         SplitDwarfKind, SwitchWithOptPath, SymbolManglingVersion, TraitSolver, TrimmedDefPaths,
     };
     use crate::lint;
@@ -3273,6 +3273,7 @@ pub(crate) mod dep_tracking {
         TraitSolver,
         Polonius,
         InliningThreshold,
+        FunctionReturn,
     );
 
     impl<T1, T2> DepTrackingHash for (T1, T2)
@@ -3451,3 +3452,14 @@ impl Default for InliningThreshold {
         Self::Sometimes(100)
     }
 }
+
+/// The different settings that the `-Zfunction-return` flag can have.
+#[derive(Clone, Copy, PartialEq, Hash, Debug, Default)]
+pub enum FunctionReturn {
+    /// Keep the function return unmodified.
+    #[default]
+    Keep,
+
+    /// Replace returns with jumps to thunk, without emitting the thunk.
+    ThunkExtern,
+}

compiler/rustc_session/src/errors.rs

@@ -436,3 +436,11 @@ pub struct IncompatibleLinkerFlavor {
     pub flavor: &'static str,
     pub compatible_list: String,
 }
+
+#[derive(Diagnostic)]
+#[diag(session_function_return_requires_x86_or_x86_64)]
+pub(crate) struct FunctionReturnRequiresX86OrX8664;
+
+#[derive(Diagnostic)]
+#[diag(session_function_return_thunk_extern_requires_non_large_code_model)]
+pub(crate) struct FunctionReturnThunkExternRequiresNonLargeCodeModel;

compiler/rustc_session/src/options.rs

@@ -430,6 +430,7 @@ mod desc {
     pub const parse_inlining_threshold: &str =
         "either a boolean (`yes`, `no`, `on`, `off`, etc), or a non-negative number";
     pub const parse_llvm_module_flag: &str = "<key>:<type>:<value>:<behavior>. Type must currently be `u32`. Behavior should be one of (`error`, `warning`, `require`, `override`, `append`, `appendunique`, `max`, `min`)";
+    pub const parse_function_return: &str = "`keep` or `thunk-extern`";
 }
 
 mod parse {
@@ -1359,6 +1360,15 @@ mod parse {
         slot.push((key.to_string(), value, behavior));
         true
     }
+
+    pub(crate) fn parse_function_return(slot: &mut FunctionReturn, v: Option<&str>) -> bool {
+        match v {
+            Some("keep") => *slot = FunctionReturn::Keep,
+            Some("thunk-extern") => *slot = FunctionReturn::ThunkExtern,
+            _ => return false,
+        }
+        true
+    }
 }
 
 options! {
@@ -1603,6 +1613,8 @@ options! {
         "force all crates to be `rustc_private` unstable (default: no)"),
     fuel: Option<(String, u64)> = (None, parse_optimization_fuel, [TRACKED],
         "set the optimization fuel quota for a crate"),
+    function_return: FunctionReturn = (FunctionReturn::default(), parse_function_return, [TRACKED],
+        "replace returns with jumps to `__x86_return_thunk` (default: `keep`)"),
     function_sections: Option<bool> = (None, parse_opt_bool, [TRACKED],
         "whether each function should go in its own section"),
     future_incompat_test: bool = (false, parse_bool, [UNTRACKED],

compiler/rustc_session/src/session.rs

@@ -1,7 +1,7 @@
 use crate::code_stats::CodeStats;
 pub use crate::code_stats::{DataTypeKind, FieldInfo, FieldKind, SizeKind, VariantInfo};
 use crate::config::{
-    self, CrateType, InstrumentCoverage, OptLevel, OutFileName, OutputType,
+    self, CrateType, FunctionReturn, InstrumentCoverage, OptLevel, OutFileName, OutputType,
     RemapPathScopeComponents, SwitchWithOptPath,
 };
 use crate::config::{ErrorOutputType, Input};
@@ -1678,6 +1678,28 @@ fn validate_commandline_args_with_session_available(sess: &Session) {
             sess.emit_err(errors::IncompatibleLinkerFlavor { flavor, compatible_list });
         }
     }
+
+    if sess.opts.unstable_opts.function_return != FunctionReturn::default() {
+        if sess.target.arch != "x86" && sess.target.arch != "x86_64" {
+            sess.emit_err(errors::FunctionReturnRequiresX86OrX8664);
+        }
+    }
+
+    // The code model check applies to `thunk` and `thunk-extern`, but not `thunk-inline`, so it is
+    // kept as a `match` to force a change if new ones are added, even if we currently only support
+    // `thunk-extern` like Clang.
+    match sess.opts.unstable_opts.function_return {
+        FunctionReturn::Keep => (),
+        FunctionReturn::ThunkExtern => {
+            // FIXME: In principle, the inherited base LLVM target code model could be large,
+            // but this only checks whether we were passed one explicitly (like Clang does).
+            if let Some(code_model) = sess.code_model()
+                && code_model == CodeModel::Large
+            {
+                sess.emit_err(errors::FunctionReturnThunkExternRequiresNonLargeCodeModel);
+            }
+        }
+    }
 }
 
 /// Holds data on the current incremental compilation session, if there is one.