Auto merge of #144019 - RalfJung:miri-sync, r=RalfJung

Miri subtree update

r? `@ghost`

Author: bors

src/tools/miri/README.md

@@ -342,9 +342,9 @@ environment variable. We first document the most relevant and most commonly used
   is enabled (the default), this is also used to emulate system entropy. The default seed is 0. You
   can increase test coverage by running Miri multiple times with different seeds.
 * `-Zmiri-strict-provenance` enables [strict
-  provenance](https://github.com/rust-lang/rust/issues/95228) checking in Miri. This means that
-  casting an integer to a pointer will stop execution because the provenance of the pointer
-  cannot be determined.
+  provenance](https://doc.rust-lang.org/nightly/std/ptr/index.html#strict-provenance) checking in
+  Miri. This means that casting an integer to a pointer will stop execution because the provenance
+  of the pointer cannot be determined.
 * `-Zmiri-symbolic-alignment-check` makes the alignment check more strict.  By default, alignment is
   checked by casting the pointer to an integer, and making sure that is a multiple of the alignment.
   This can lead to cases where a program passes the alignment check by pure chance, because things

src/tools/miri/miri-script/Cargo.toml

@@ -7,6 +7,7 @@ repository = "https://github.com/rust-lang/miri"
 version = "0.1.0"
 default-run = "miri-script"
 edition = "2024"
+rust-version = "1.85"
 
 [workspace]
 # We make this a workspace root so that cargo does not go looking in ../Cargo.toml for the workspace root.

src/tools/miri/miri-script/src/commands.rs

@@ -702,7 +702,6 @@ impl Command {
         let mut early_flags = Vec::<OsString>::new();
 
         // In `dep` mode, the target is already passed via `MIRI_TEST_TARGET`
-        #[expect(clippy::collapsible_if)] // we need to wait until this is stable
         if !dep {
             if let Some(target) = &target {
                 early_flags.push("--target".into());
@@ -735,7 +734,6 @@ impl Command {
         // Add Miri flags
         let mut cmd = cmd.args(&miri_flags).args(&early_flags).args(&flags);
         // For `--dep` we also need to set the target in the env var.
-        #[expect(clippy::collapsible_if)] // we need to wait until this is stable
         if dep {
             if let Some(target) = &target {
                 cmd = cmd.env("MIRI_TEST_TARGET", target);

src/tools/miri/rust-version

@@ -1 +1 @@
-733b47ea4b1b86216f14ef56e49440c33933f230
+7f2065a4bae1faed5bab928c670964eafbf43b55

src/tools/miri/src/alloc/isolated_alloc.rs

@@ -302,31 +302,28 @@ impl IsolatedAlloc {
         }
     }
 
-    /// Returns a vector of page addresses managed by the allocator.
-    pub fn pages(&self) -> Vec<usize> {
-        let mut pages: Vec<usize> =
-            self.page_ptrs.iter().map(|p| p.expose_provenance().get()).collect();
-        for (ptr, size) in self.huge_ptrs.iter() {
-            for i in 0..size / self.page_size {
-                pages.push(ptr.expose_provenance().get().strict_add(i * self.page_size));
-            }
-        }
-        pages
+    /// Returns a list of page addresses managed by the allocator.
+    pub fn pages(&self) -> impl Iterator<Item = usize> {
+        let pages = self.page_ptrs.iter().map(|p| p.expose_provenance().get());
+        pages.chain(self.huge_ptrs.iter().flat_map(|(ptr, size)| {
+            (0..size / self.page_size)
+                .map(|i| ptr.expose_provenance().get().strict_add(i * self.page_size))
+        }))
     }
 
     /// Protects all owned memory as `PROT_NONE`, preventing accesses.
     ///
     /// SAFETY: Accessing memory after this point will result in a segfault
     /// unless it is first unprotected.
-    pub unsafe fn prepare_ffi(&mut self) -> Result<(), nix::errno::Errno> {
+    pub unsafe fn start_ffi(&mut self) -> Result<(), nix::errno::Errno> {
         let prot = mman::ProtFlags::PROT_NONE;
         unsafe { self.mprotect(prot) }
     }
 
     /// Deprotects all owned memory by setting it to RW. Erroring here is very
     /// likely unrecoverable, so it may panic if applying those permissions
     /// fails.
-    pub fn unprep_ffi(&mut self) {
+    pub fn end_ffi(&mut self) {
         let prot = mman::ProtFlags::PROT_READ | mman::ProtFlags::PROT_WRITE;
         unsafe {
             self.mprotect(prot).unwrap();

src/tools/miri/src/bin/miri.rs

@@ -233,8 +233,6 @@ impl rustc_driver::Callbacks for MiriCompilerCalls {
         } else {
             let return_code = miri::eval_entry(tcx, entry_def_id, entry_type, &config, None)
                 .unwrap_or_else(|| {
-                    //#[cfg(target_os = "linux")]
-                    //miri::native_lib::register_retcode_sv(rustc_driver::EXIT_FAILURE);
                     tcx.dcx().abort_if_errors();
                     rustc_driver::EXIT_FAILURE
                 });
@@ -337,6 +335,9 @@ impl rustc_driver::Callbacks for MiriBeRustCompilerCalls {
 fn exit(exit_code: i32) -> ! {
     // Drop the tracing guard before exiting, so tracing calls are flushed correctly.
     deinit_loggers();
+    // Make sure the supervisor knows about the code code.
+    #[cfg(target_os = "linux")]
+    miri::native_lib::register_retcode_sv(exit_code);
     std::process::exit(exit_code);
 }
 
@@ -355,6 +356,11 @@ fn run_compiler_and_exit(
     args: &[String],
     callbacks: &mut (dyn rustc_driver::Callbacks + Send),
 ) -> ! {
+    // Install the ctrlc handler that sets `rustc_const_eval::CTRL_C_RECEIVED`, even if
+    // MIRI_BE_RUSTC is set. We do this late so that when `native_lib::init_sv` is called,
+    // there are no other threads.
+    rustc_driver::install_ctrlc_handler();
+
     // Invoke compiler, catch any unwinding panics and handle return code.
     let exit_code =
         rustc_driver::catch_with_exit_code(move || rustc_driver::run_compiler(args, callbacks));
@@ -439,10 +445,6 @@ fn main() {
     let args = rustc_driver::catch_fatal_errors(|| rustc_driver::args::raw_args(&early_dcx))
         .unwrap_or_else(|_| std::process::exit(rustc_driver::EXIT_FAILURE));
 
-    // Install the ctrlc handler that sets `rustc_const_eval::CTRL_C_RECEIVED`, even if
-    // MIRI_BE_RUSTC is set.
-    rustc_driver::install_ctrlc_handler();
-
     // If the environment asks us to actually be rustc, then do that.
     if let Some(crate_kind) = env::var_os("MIRI_BE_RUSTC") {
         // Earliest rustc setup.
@@ -750,15 +752,15 @@ fn main() {
 
     debug!("rustc arguments: {:?}", rustc_args);
     debug!("crate arguments: {:?}", miri_config.args);
-    #[cfg(target_os = "linux")]
     if !miri_config.native_lib.is_empty() && miri_config.native_lib_enable_tracing {
-        // FIXME: This should display a diagnostic / warning on error
-        // SAFETY: If any other threads exist at this point (namely for the ctrlc
-        // handler), they will not interact with anything on the main rustc/Miri
-        // thread in an async-signal-unsafe way such as by accessing shared
-        // semaphores, etc.; the handler only calls `sleep()` and `exit()`, which
-        // are async-signal-safe, as is accessing atomics
-        //let _ = unsafe { miri::native_lib::init_sv() };
+        // SAFETY: No other threads are running
+        #[cfg(target_os = "linux")]
+        if unsafe { miri::native_lib::init_sv() }.is_err() {
+            eprintln!(
+                "warning: The native-lib tracer could not be started. Is this an x86 Linux system, and does Miri have permissions to ptrace?\n\
+                Falling back to non-tracing native-lib mode."
+            );
+        }
     }
     run_compiler_and_exit(
         &rustc_args,

src/tools/miri/src/borrow_tracker/mod.rs

@@ -260,6 +260,7 @@ impl GlobalStateInner {
         kind: MemoryKind,
         machine: &MiriMachine<'_>,
     ) -> AllocState {
+        let _span = enter_trace_span!(borrow_tracker::new_allocation, ?id, ?alloc_size, ?kind);
         match self.borrow_tracker_method {
             BorrowTrackerMethod::StackedBorrows =>
                 AllocState::StackedBorrows(Box::new(RefCell::new(Stacks::new_allocation(
@@ -280,6 +281,7 @@ pub trait EvalContextExt<'tcx>: crate::MiriInterpCxExt<'tcx> {
         kind: RetagKind,
         val: &ImmTy<'tcx>,
     ) -> InterpResult<'tcx, ImmTy<'tcx>> {
+        let _span = enter_trace_span!(borrow_tracker::retag_ptr_value, ?kind, ?val.layout);
         let this = self.eval_context_mut();
         let method = this.machine.borrow_tracker.as_ref().unwrap().borrow().borrow_tracker_method;
         match method {
@@ -293,6 +295,7 @@ pub trait EvalContextExt<'tcx>: crate::MiriInterpCxExt<'tcx> {
         kind: RetagKind,
         place: &PlaceTy<'tcx>,
     ) -> InterpResult<'tcx> {
+        let _span = enter_trace_span!(borrow_tracker::retag_place_contents, ?kind, ?place);
         let this = self.eval_context_mut();
         let method = this.machine.borrow_tracker.as_ref().unwrap().borrow().borrow_tracker_method;
         match method {
@@ -302,6 +305,7 @@ pub trait EvalContextExt<'tcx>: crate::MiriInterpCxExt<'tcx> {
     }
 
     fn protect_place(&mut self, place: &MPlaceTy<'tcx>) -> InterpResult<'tcx, MPlaceTy<'tcx>> {
+        let _span = enter_trace_span!(borrow_tracker::protect_place, ?place);
         let this = self.eval_context_mut();
         let method = this.machine.borrow_tracker.as_ref().unwrap().borrow().borrow_tracker_method;
         match method {
@@ -311,6 +315,8 @@ pub trait EvalContextExt<'tcx>: crate::MiriInterpCxExt<'tcx> {
     }
 
     fn expose_tag(&self, alloc_id: AllocId, tag: BorTag) -> InterpResult<'tcx> {
+        let _span =
+            enter_trace_span!(borrow_tracker::expose_tag, alloc_id = alloc_id.0, tag = tag.0);
         let this = self.eval_context_ref();
         let method = this.machine.borrow_tracker.as_ref().unwrap().borrow().borrow_tracker_method;
         match method {
@@ -354,6 +360,7 @@ pub trait EvalContextExt<'tcx>: crate::MiriInterpCxExt<'tcx> {
         &self,
         frame: &Frame<'tcx, Provenance, FrameExtra<'tcx>>,
     ) -> InterpResult<'tcx> {
+        let _span = enter_trace_span!(borrow_tracker::on_stack_pop);
         let this = self.eval_context_ref();
         let borrow_tracker = this.machine.borrow_tracker.as_ref().unwrap();
         // The body of this loop needs `borrow_tracker` immutably
@@ -431,6 +438,7 @@ impl AllocState {
         range: AllocRange,
         machine: &MiriMachine<'tcx>,
     ) -> InterpResult<'tcx> {
+        let _span = enter_trace_span!(borrow_tracker::before_memory_read, alloc_id = alloc_id.0);
         match self {
             AllocState::StackedBorrows(sb) =>
                 sb.borrow_mut().before_memory_read(alloc_id, prov_extra, range, machine),
@@ -452,6 +460,7 @@ impl AllocState {
         range: AllocRange,
         machine: &MiriMachine<'tcx>,
     ) -> InterpResult<'tcx> {
+        let _span = enter_trace_span!(borrow_tracker::before_memory_write, alloc_id = alloc_id.0);
         match self {
             AllocState::StackedBorrows(sb) =>
                 sb.get_mut().before_memory_write(alloc_id, prov_extra, range, machine),
@@ -473,6 +482,8 @@ impl AllocState {
         size: Size,
         machine: &MiriMachine<'tcx>,
     ) -> InterpResult<'tcx> {
+        let _span =
+            enter_trace_span!(borrow_tracker::before_memory_deallocation, alloc_id = alloc_id.0);
         match self {
             AllocState::StackedBorrows(sb) =>
                 sb.get_mut().before_memory_deallocation(alloc_id, prov_extra, size, machine),
@@ -482,6 +493,7 @@ impl AllocState {
     }
 
     pub fn remove_unreachable_tags(&self, tags: &FxHashSet<BorTag>) {
+        let _span = enter_trace_span!(borrow_tracker::remove_unreachable_tags);
         match self {
             AllocState::StackedBorrows(sb) => sb.borrow_mut().remove_unreachable_tags(tags),
             AllocState::TreeBorrows(tb) => tb.borrow_mut().remove_unreachable_tags(tags),
@@ -496,6 +508,11 @@ impl AllocState {
         tag: BorTag,
         alloc_id: AllocId, // diagnostics
     ) -> InterpResult<'tcx> {
+        let _span = enter_trace_span!(
+            borrow_tracker::release_protector,
+            alloc_id = alloc_id.0,
+            tag = tag.0
+        );
         match self {
             AllocState::StackedBorrows(_sb) => interp_ok(()),
             AllocState::TreeBorrows(tb) =>
@@ -506,6 +523,7 @@ impl AllocState {
 
 impl VisitProvenance for AllocState {
     fn visit_provenance(&self, visit: &mut VisitWith<'_>) {
+        let _span = enter_trace_span!(borrow_tracker::visit_provenance);
         match self {
             AllocState::StackedBorrows(sb) => sb.visit_provenance(visit),
             AllocState::TreeBorrows(tb) => tb.visit_provenance(visit),

src/tools/miri/src/borrow_tracker/stacked_borrows/mod.rs

@@ -30,7 +30,7 @@ pub type AllocState = Stacks;
 #[derive(Clone, Debug)]
 pub struct Stacks {
     // Even reading memory can have effects on the stack, so we need a `RefCell` here.
-    stacks: RangeMap<Stack>,
+    stacks: DedupRangeMap<Stack>,
     /// Stores past operations on this allocation
     history: AllocHistory,
     /// The set of tags that have been exposed inside this allocation.
@@ -468,7 +468,7 @@ impl<'tcx> Stacks {
         let stack = Stack::new(item);
 
         Stacks {
-            stacks: RangeMap::new(size, stack),
+            stacks: DedupRangeMap::new(size, stack),
             history: AllocHistory::new(id, item, machine),
             exposed_tags: FxHashSet::default(),
         }

src/tools/miri/src/borrow_tracker/tree_borrows/mod.rs

@@ -314,7 +314,7 @@ trait EvalContextPrivExt<'tcx>: crate::MiriInterpCxExt<'tcx> {
         let span = this.machine.current_span();
 
         // Store initial permissions and their corresponding range.
-        let mut perms_map: RangeMap<LocationState> = RangeMap::new(
+        let mut perms_map: DedupRangeMap<LocationState> = DedupRangeMap::new(
             ptr_size,
             LocationState::new_accessed(Permission::new_disabled(), IdempotentForeignAccess::None), // this will be overwritten
         );

src/tools/miri/src/borrow_tracker/tree_borrows/tree.rs

@@ -247,7 +247,7 @@ pub struct Tree {
     /// `unwrap` any `perm.get(key)`.
     ///
     /// We do uphold the fact that `keys(perms)` is a subset of `keys(nodes)`
-    pub(super) rperms: RangeMap<UniValMap<LocationState>>,
+    pub(super) rperms: DedupRangeMap<UniValMap<LocationState>>,
     /// The index of the root node.
     pub(super) root: UniIndex,
 }
@@ -609,7 +609,7 @@ impl Tree {
                     IdempotentForeignAccess::None,
                 ),
             );
-            RangeMap::new(size, perms)
+            DedupRangeMap::new(size, perms)
         };
         Self { root: root_idx, nodes, rperms, tag_mapping }
     }
@@ -631,7 +631,7 @@ impl<'tcx> Tree {
         base_offset: Size,
         parent_tag: BorTag,
         new_tag: BorTag,
-        initial_perms: RangeMap<LocationState>,
+        initial_perms: DedupRangeMap<LocationState>,
         default_perm: Permission,
         protected: bool,
         span: Span,