Rollup of 12 pull requests

compiler/rustc/Cargo.toml

@@ -27,6 +27,7 @@ features = ['unprefixed_malloc_on_supported_platforms']
 
 [features]
 # tidy-alphabetical-start
+check_only = ['rustc_driver_impl/check_only']
 jemalloc = ['dep:tikv-jemalloc-sys']
 llvm = ['rustc_driver_impl/llvm']
 max_level_info = ['rustc_driver_impl/max_level_info']

compiler/rustc_codegen_gcc/src/debuginfo.rs

@@ -36,10 +36,6 @@ impl<'a, 'gcc, 'tcx> DebugInfoBuilderMethods for Builder<'a, 'gcc, 'tcx> {
         _variable_alloca.set_location(_dbg_loc);
     }
 
-    fn insert_reference_to_gdb_debug_scripts_section_global(&mut self) {
-        // TODO(antoyo): insert reference to gdb debug scripts section global.
-    }
-
     /// FIXME(tempdragon): Currently, this function is not yet implemented. It seems that the
     /// debug name and the mangled name should both be included in the LValues.
     /// Besides, a function to get the rvalue type(m_is_lvalue) should also be included.
@@ -254,7 +250,8 @@ impl<'gcc, 'tcx> DebugInfoCodegenMethods<'tcx> for CodegenCx<'gcc, 'tcx> {
         // TODO(antoyo): implement.
     }
 
-    fn debuginfo_finalize(&self) {
+    fn debuginfo_finalize(&mut self) {
+        // TODO: emit section `.debug_gdb_scripts`.
         self.context.set_debug_info(true)
     }
 

compiler/rustc_codegen_llvm/Cargo.toml

@@ -43,3 +43,6 @@ serde_json = "1"
 smallvec = { version = "1.8.1", features = ["union", "may_dangle"] }
 tracing = "0.1"
 # tidy-alphabetical-end
+
+[features]
+check_only = ["rustc_llvm/check_only"]

compiler/rustc_codegen_llvm/src/base.rs

@@ -109,11 +109,16 @@ pub(crate) fn compile_codegen_unit(
             }
 
             // Finalize code coverage by injecting the coverage map. Note, the coverage map will
-            // also be added to the `llvm.compiler.used` variable, created next.
+            // also be added to the `llvm.compiler.used` variable, created below.
             if cx.sess().instrument_coverage() {
                 cx.coverageinfo_finalize();
             }
 
+            // Finalize debuginfo. This adds to `llvm.used`, created below.
+            if cx.sess().opts.debuginfo != DebugInfo::None {
+                cx.debuginfo_finalize();
+            }
+
             // Create the llvm.used and llvm.compiler.used variables.
             if !cx.used_statics.is_empty() {
                 cx.create_used_variable_impl(c"llvm.used", &cx.used_statics);
@@ -130,11 +135,6 @@ pub(crate) fn compile_codegen_unit(
                     llvm::LLVMDeleteGlobal(old_g);
                 }
             }
-
-            // Finalize debuginfo
-            if cx.sess().opts.debuginfo != DebugInfo::None {
-                cx.debuginfo_finalize();
-            }
         }
 
         ModuleCodegen::new_regular(cgu_name.to_string(), llvm_module)

compiler/rustc_codegen_llvm/src/debuginfo/gdb.rs

@@ -1,5 +1,7 @@
 // .debug_gdb_scripts binary section.
 
+use std::ffi::CString;
+
 use rustc_ast::attr;
 use rustc_codegen_ssa::base::collect_debugger_visualizers_transitive;
 use rustc_codegen_ssa::traits::*;
@@ -9,31 +11,21 @@ use rustc_middle::middle::debugger_visualizer::DebuggerVisualizerType;
 use rustc_session::config::{CrateType, DebugInfo};
 use rustc_span::sym;
 
-use crate::builder::Builder;
 use crate::common::CodegenCx;
 use crate::llvm;
 use crate::value::Value;
 
-/// Inserts a side-effect free instruction sequence that makes sure that the
-/// .debug_gdb_scripts global is referenced, so it isn't removed by the linker.
-pub(crate) fn insert_reference_to_gdb_debug_scripts_section_global(bx: &mut Builder<'_, '_, '_>) {
-    if needs_gdb_debug_scripts_section(bx) {
-        let gdb_debug_scripts_section = get_or_insert_gdb_debug_scripts_section_global(bx);
-        // Load just the first byte as that's all that's necessary to force
-        // LLVM to keep around the reference to the global.
-        let volatile_load_instruction = bx.volatile_load(bx.type_i8(), gdb_debug_scripts_section);
-        unsafe {
-            llvm::LLVMSetAlignment(volatile_load_instruction, 1);
-        }
-    }
-}
-
 /// Allocates the global variable responsible for the .debug_gdb_scripts binary
 /// section.
 pub(crate) fn get_or_insert_gdb_debug_scripts_section_global<'ll>(
-    cx: &CodegenCx<'ll, '_>,
+    cx: &mut CodegenCx<'ll, '_>,
 ) -> &'ll Value {
-    let c_section_var_name = c"__rustc_debug_gdb_scripts_section__";
+    let c_section_var_name = CString::new(format!(
+        "__rustc_debug_gdb_scripts_section_{}_{:08x}",
+        cx.tcx.crate_name(LOCAL_CRATE),
+        cx.tcx.stable_crate_id(LOCAL_CRATE),
+    ))
+    .unwrap();
     let section_var_name = c_section_var_name.to_str().unwrap();
 
     let section_var = unsafe { llvm::LLVMGetNamedGlobal(cx.llmod, c_section_var_name.as_ptr()) };
@@ -80,6 +72,8 @@ pub(crate) fn get_or_insert_gdb_debug_scripts_section_global<'ll>(
             // This should make sure that the whole section is not larger than
             // the string it contains. Otherwise we get a warning from GDB.
             llvm::LLVMSetAlignment(section_var, 1);
+            // Make sure that the linker doesn't optimize the global away.
+            cx.add_used_global(section_var);
             section_var
         }
     })

compiler/rustc_codegen_llvm/src/debuginfo/mod.rs

@@ -30,7 +30,7 @@ use tracing::debug;
 
 use self::metadata::{UNKNOWN_COLUMN_NUMBER, UNKNOWN_LINE_NUMBER, file_metadata, type_di_node};
 use self::namespace::mangled_name_of_instance;
-use self::utils::{DIB, create_DIArray, is_node_local_to_unit};
+use self::utils::{DIB, create_DIArray, debug_context, is_node_local_to_unit};
 use crate::builder::Builder;
 use crate::common::{AsCCharPtr, CodegenCx};
 use crate::llvm;
@@ -131,20 +131,22 @@ impl<'ll, 'tcx> CodegenUnitDebugContext<'ll, 'tcx> {
 }
 
 /// Creates any deferred debug metadata nodes
-pub(crate) fn finalize(cx: &CodegenCx<'_, '_>) {
-    if let Some(dbg_cx) = &cx.dbg_cx {
-        debug!("finalize");
-
-        if gdb::needs_gdb_debug_scripts_section(cx) {
-            // Add a .debug_gdb_scripts section to this compile-unit. This will
-            // cause GDB to try and load the gdb_load_rust_pretty_printers.py file,
-            // which activates the Rust pretty printers for binary this section is
-            // contained in.
-            gdb::get_or_insert_gdb_debug_scripts_section_global(cx);
-        }
+pub(crate) fn finalize(cx: &mut CodegenCx<'_, '_>) {
+    if cx.dbg_cx.is_none() {
+        return;
+    }
+
+    debug!("finalize");
 
-        dbg_cx.finalize(cx.sess());
+    if gdb::needs_gdb_debug_scripts_section(cx) {
+        // Add a .debug_gdb_scripts section to this compile-unit. This will
+        // cause GDB to try and load the gdb_load_rust_pretty_printers.py file,
+        // which activates the Rust pretty printers for binary this section is
+        // contained in.
+        gdb::get_or_insert_gdb_debug_scripts_section_global(cx);
     }
+
+    debug_context(cx).finalize(cx.sess());
 }
 
 impl<'ll> Builder<'_, 'll, '_> {
@@ -215,10 +217,6 @@ impl<'ll> DebugInfoBuilderMethods for Builder<'_, 'll, '_> {
         }
     }
 
-    fn insert_reference_to_gdb_debug_scripts_section_global(&mut self) {
-        gdb::insert_reference_to_gdb_debug_scripts_section_global(self)
-    }
-
     fn set_var_name(&mut self, value: &'ll Value, name: &str) {
         // Avoid wasting time if LLVM value names aren't even enabled.
         if self.sess().fewer_names() {
@@ -614,7 +612,7 @@ impl<'ll, 'tcx> DebugInfoCodegenMethods<'tcx> for CodegenCx<'ll, 'tcx> {
         metadata::extend_scope_to_file(self, scope_metadata, file)
     }
 
-    fn debuginfo_finalize(&self) {
+    fn debuginfo_finalize(&mut self) {
         finalize(self)
     }
 

compiler/rustc_codegen_ssa/src/base.rs

@@ -528,8 +528,6 @@ pub fn maybe_create_entry_wrapper<'a, 'tcx, Bx: BuilderMethods<'a, 'tcx>>(
         let llbb = Bx::append_block(cx, llfn, "top");
         let mut bx = Bx::build(cx, llbb);
 
-        bx.insert_reference_to_gdb_debug_scripts_section_global();
-
         let isize_ty = cx.type_isize();
         let ptr_ty = cx.type_ptr();
         let (arg_argc, arg_argv) = get_argc_argv(&mut bx);

compiler/rustc_codegen_ssa/src/traits/debuginfo.rs

@@ -50,7 +50,7 @@ pub trait DebugInfoCodegenMethods<'tcx>: BackendTypes {
         scope_metadata: Self::DIScope,
         file: &SourceFile,
     ) -> Self::DIScope;
-    fn debuginfo_finalize(&self);
+    fn debuginfo_finalize(&mut self);
 
     // FIXME(eddyb) find a common convention for all of the debuginfo-related
     // names (choose between `dbg`, `debug`, `debuginfo`, `debug_info` etc.).
@@ -81,6 +81,5 @@ pub trait DebugInfoBuilderMethods: BackendTypes {
     );
     fn set_dbg_loc(&mut self, dbg_loc: Self::DILocation);
     fn clear_dbg_loc(&mut self);
-    fn insert_reference_to_gdb_debug_scripts_section_global(&mut self);
     fn set_var_name(&mut self, value: Self::Value, name: &str);
 }

compiler/rustc_const_eval/src/interpret/memory.rs

@@ -1498,7 +1498,7 @@ impl<'tcx, M: Machine<'tcx>> InterpCx<'tcx, M> {
             dest_alloc
                 .write_uninit(&tcx, dest_range)
                 .map_err(|e| e.to_interp_error(dest_alloc_id))?;
-            // We can forget about the provenance, this is all not initialized anyway.
+            // `write_uninit` also resets the provenance, so we are done.
             return interp_ok(());
         }
 

compiler/rustc_driver_impl/Cargo.toml

@@ -72,6 +72,7 @@ ctrlc = "3.4.4"
 
 [features]
 # tidy-alphabetical-start
+check_only = ['rustc_interface/check_only']
 llvm = ['rustc_interface/llvm']
 max_level_info = ['rustc_log/max_level_info']
 rustc_randomized_layouts = [

compiler/rustc_interface/Cargo.toml

@@ -56,5 +56,6 @@ rustc_abi = { path = "../rustc_abi" }
 
 [features]
 # tidy-alphabetical-start
+check_only = ['rustc_codegen_llvm?/check_only']
 llvm = ['dep:rustc_codegen_llvm']
 # tidy-alphabetical-end

compiler/rustc_llvm/Cargo.toml

@@ -14,3 +14,7 @@ libc = "0.2.73"
 # pinned `cc` in `rustc_codegen_ssa` if you update `cc` here.
 cc = "=1.2.16"
 # tidy-alphabetical-end
+
+[features]
+# Used by ./x.py check --compile-time-deps to skip building C++ code
+check_only = []

compiler/rustc_llvm/build.rs

@@ -106,6 +106,10 @@ fn output(cmd: &mut Command) -> String {
 }
 
 fn main() {
+    if cfg!(feature = "check_only") {
+        return;
+    }
+
     for component in REQUIRED_COMPONENTS.iter().chain(OPTIONAL_COMPONENTS.iter()) {
         println!("cargo:rustc-check-cfg=cfg(llvm_component,values(\"{component}\"))");
     }

compiler/rustc_middle/src/mir/interpret/allocation.rs

@@ -101,6 +101,8 @@ pub struct Allocation<Prov: Provenance = CtfeProvenance, Extra = (), Bytes = Box
     /// at the given offset.
     provenance: ProvenanceMap<Prov>,
     /// Denotes which part of this allocation is initialized.
+    ///
+    /// Invariant: the uninitialized parts have no provenance.
     init_mask: InitMask,
     /// The alignment of the allocation to detect unaligned reads.
     /// (`Align` guarantees that this is a power of two.)
@@ -796,24 +798,19 @@ impl<Prov: Provenance, Extra, Bytes: AllocBytes> Allocation<Prov, Extra, Bytes>
         Ok(())
     }
 
-    /// Initialize all previously uninitialized bytes in the entire allocation, and set
-    /// provenance of everything to `Wildcard`. Before calling this, make sure all
-    /// provenance in this allocation is exposed!
-    pub fn prepare_for_native_access(&mut self) {
-        let full_range = AllocRange { start: Size::ZERO, size: Size::from_bytes(self.len()) };
-        // Overwrite uninitialized bytes with 0, to ensure we don't leak whatever their value happens to be.
-        for chunk in self.init_mask.range_as_init_chunks(full_range) {
-            if !chunk.is_init() {
-                let uninit_bytes = &mut self.bytes
-                    [chunk.range().start.bytes_usize()..chunk.range().end.bytes_usize()];
-                uninit_bytes.fill(0);
-            }
-        }
-        // Mark everything as initialized now.
-        self.mark_init(full_range, true);
-
-        // Set provenance of all bytes to wildcard.
-        self.provenance.write_wildcards(self.len());
+    /// Mark all bytes in the given range as initialised and reset the provenance
+    /// to wildcards. This entirely breaks the normal mechanisms for tracking
+    /// initialisation and is only provided for Miri operating in native-lib
+    /// mode. UB will be missed if the underlying bytes were not actually written to.
+    ///
+    /// If `range` is `None`, defaults to performing this on the whole allocation.
+    pub fn process_native_write(&mut self, cx: &impl HasDataLayout, range: Option<AllocRange>) {
+        let range = range.unwrap_or_else(|| AllocRange {
+            start: Size::ZERO,
+            size: Size::from_bytes(self.len()),
+        });
+        self.mark_init(range, true);
+        self.provenance.write_wildcards(cx, range);
     }
 
     /// Remove all provenance in the given memory range.

compiler/rustc_middle/src/mir/interpret/allocation/provenance_map.rs

@@ -212,21 +212,37 @@ impl<Prov: Provenance> ProvenanceMap<Prov> {
         Ok(())
     }
 
-    /// Overwrites all provenance in the allocation with wildcard provenance.
+    /// Overwrites all provenance in the given range with wildcard provenance.
+    /// Pointers partially overwritten will have their provenances preserved
+    /// bytewise on their remaining bytes.
     ///
     /// Provided for usage in Miri and panics otherwise.
-    pub fn write_wildcards(&mut self, alloc_size: usize) {
+    pub fn write_wildcards(&mut self, cx: &impl HasDataLayout, range: AllocRange) {
         assert!(
             Prov::OFFSET_IS_ADDR,
             "writing wildcard provenance is not supported when `OFFSET_IS_ADDR` is false"
         );
         let wildcard = Prov::WILDCARD.unwrap();
 
-        // Remove all pointer provenances, then write wildcards into the whole byte range.
-        self.ptrs.clear();
-        let last = Size::from_bytes(alloc_size);
         let bytes = self.bytes.get_or_insert_with(Box::default);
-        for offset in Size::ZERO..last {
+
+        // Remove pointer provenances that overlap with the range, then readd the edge ones bytewise.
+        let ptr_range = Self::adjusted_range_ptrs(range, cx);
+        let ptrs = self.ptrs.range(ptr_range.clone());
+        if let Some((offset, prov)) = ptrs.first() {
+            for byte_ofs in *offset..range.start {
+                bytes.insert(byte_ofs, *prov);
+            }
+        }
+        if let Some((offset, prov)) = ptrs.last() {
+            for byte_ofs in range.end()..*offset + cx.data_layout().pointer_size() {
+                bytes.insert(byte_ofs, *prov);
+            }
+        }
+        self.ptrs.remove_range(ptr_range);
+
+        // Overwrite bytewise provenance.
+        for offset in range.start..range.end() {
             bytes.insert(offset, wildcard);
         }
     }