Rollup merge of #156210 - BorrowSanitizer:codegen-emit-retag-2, r=saethlin

Emit retags in codegen to support BorrowSanitizer (part 2)

Tracking issue: #154760
[Zulip Thread](https://rust-lang.zulipchat.com/#narrow/channel/131828-t-compiler/topic/Staging.20for.20emitting.20retags.20in.20codegen/with/593004012)

This is one of several PRs that will add experimental support for emitting retags as function calls in codegen. Each PR will be a minimal, improved slice of the changes in #155965.

This PR adds a new unstable flag `-Zcodegen-emit-retag`, which will enable experimental retag calls in generated code. This flag is a nop for now, but the relevant methods have been added to codegen_ssa, and they are called wherever retags are necessary. Subsequent PRs will complete this implementation.

This does not depend on #156208.

r? @RalfJung

Author: jhpratt

compiler/rustc_codegen_ssa/src/mir/block.rs

@@ -264,6 +264,12 @@ impl<'a, 'tcx> TerminatorCodegenHelper<'tcx> {
                     bx.lifetime_end(tmp, size);
                 }
                 fx.store_return(bx, ret_dest, &fn_abi.ret, invokeret);
+
+                // If the return value was retagged as it was stored,
+                // then we might be in a different basic block now.
+                // Update the cached block for `target` to point to this new
+                // block, where codegen will continue.
+                fx.cached_llbbs[target] = CachedLlbb::Some(bx.llbb());
             }
             MergingSucc::False
         } else {
@@ -2186,19 +2192,27 @@ impl<'a, 'tcx, Bx: BuilderMethods<'a, 'tcx>> FunctionCx<'a, 'tcx, Bx> {
         llval: Bx::Value,
     ) {
         use self::ReturnDest::*;
-
+        let retags_enabled = bx.tcx().sess.opts.unstable_opts.codegen_emit_retag.is_some();
         match dest {
             Nothing => (),
-            Store(dst) => bx.store_arg(ret_abi, llval, dst),
+            Store(dst) => {
+                bx.store_arg(ret_abi, llval, dst);
+                if retags_enabled {
+                    self.codegen_retag_place(bx, dst, false);
+                }
+            }
             IndirectOperand(tmp, index) => {
-                let op = bx.load_operand(tmp);
+                let mut op = bx.load_operand(tmp);
                 tmp.storage_dead(bx);
+                if retags_enabled {
+                    op = self.codegen_retag_operand(bx, op, false);
+                }
                 self.overwrite_local(index, LocalRef::Operand(op));
                 self.debug_introduce_local(bx, index);
             }
             DirectOperand(index) => {
                 // If there is a cast, we have to store and reload.
-                let op = if let PassMode::Cast { .. } = ret_abi.mode {
+                let mut op = if let PassMode::Cast { .. } = ret_abi.mode {
                     let tmp = PlaceRef::alloca(bx, ret_abi.layout);
                     tmp.storage_live(bx);
                     bx.store_arg(ret_abi, llval, tmp);
@@ -2208,6 +2222,9 @@ impl<'a, 'tcx, Bx: BuilderMethods<'a, 'tcx>> FunctionCx<'a, 'tcx, Bx> {
                 } else {
                     OperandRef::from_immediate_or_packed_pair(bx, llval, ret_abi.layout)
                 };
+                if retags_enabled {
+                    op = self.codegen_retag_operand(bx, op, false);
+                }
                 self.overwrite_local(index, LocalRef::Operand(op));
                 self.debug_introduce_local(bx, index);
             }

compiler/rustc_codegen_ssa/src/mir/mod.rs

@@ -24,6 +24,7 @@ mod locals;
 pub mod naked_asm;
 pub mod operand;
 pub mod place;
+mod retag;
 mod rvalue;
 mod statement;
 
@@ -425,7 +426,7 @@ fn arg_local_refs<'a, 'tcx, Bx: BuilderMethods<'a, 'tcx>>(
         return vec![];
     }
 
-    let args = mir
+    let mut args = mir
         .args_iter()
         .enumerate()
         .map(|(arg_index, local)| {
@@ -562,6 +563,32 @@ fn arg_local_refs<'a, 'tcx, Bx: BuilderMethods<'a, 'tcx>>(
             }
         })
         .collect::<Vec<_>>();
+    if bx.tcx().sess.opts.unstable_opts.codegen_emit_retag.is_some() {
+        args = args
+            .iter()
+            .map(|arg| match arg {
+                &LocalRef::Place(place_ref) => {
+                    fx.codegen_retag_place(bx, place_ref, true);
+                    LocalRef::Place(place_ref)
+                }
+                &LocalRef::UnsizedPlace(place_ref) => {
+                    let operand = bx.load_operand(place_ref);
+                    let retagged = fx.codegen_retag_operand(bx, operand, true);
+                    assert!(matches!(retagged.val, OperandValue::Pair(_, _)));
+                    retagged.val.store(bx, place_ref);
+                    LocalRef::UnsizedPlace(place_ref)
+                }
+                &LocalRef::Operand(operand_ref) => {
+                    let retagged = fx.codegen_retag_operand(bx, operand_ref, true);
+                    LocalRef::Operand(retagged)
+                }
+                LocalRef::PendingOperand => LocalRef::PendingOperand,
+            })
+            .collect::<Vec<_>>();
+        // If we branched during retagging, then we need to update the
+        // start block to the new location.
+        fx.cached_llbbs[mir::START_BLOCK] = CachedLlbb::Some(bx.llbb());
+    }
 
     if fx.instance.def.requires_caller_location(bx.tcx()) {
         let mir_args = if let Some(num_untupled) = num_untupled {

compiler/rustc_codegen_ssa/src/mir/retag.rs

@@ -0,0 +1,34 @@
+//! Experimental support for emitting retags as function calls in generated code.
+
+use rustc_middle::mir::{Rvalue, WithRetag};
+
+use crate::mir::FunctionCx;
+use crate::mir::operand::OperandRef;
+use crate::mir::place::PlaceRef;
+use crate::traits::BuilderMethods;
+
+pub(crate) fn rvalue_needs_retag(rvalue: &Rvalue<'_>) -> bool {
+    // `Ref` has its own internal retagging
+    !matches!(rvalue, Rvalue::Ref(..)) && !matches!(rvalue, Rvalue::Use(.., WithRetag::No))
+}
+
+impl<'a, 'tcx, Bx: BuilderMethods<'a, 'tcx>> FunctionCx<'a, 'tcx, Bx> {
+    /// Retags the pointers within an [`OperandRef`].
+    pub(crate) fn codegen_retag_operand(
+        &mut self,
+        _bx: &mut Bx,
+        operand: OperandRef<'tcx, Bx::Value>,
+        _is_fn_entry: bool,
+    ) -> OperandRef<'tcx, Bx::Value> {
+        operand
+    }
+
+    /// Retags the pointers within a [`PlaceRef`].
+    pub(crate) fn codegen_retag_place(
+        &mut self,
+        _bx: &mut Bx,
+        _place_ref: PlaceRef<'tcx, Bx::Value>,
+        _is_fn_entry: bool,
+    ) {
+    }
+}

compiler/rustc_codegen_ssa/src/mir/rvalue.rs

@@ -522,7 +522,12 @@ impl<'a, 'tcx, Bx: BuilderMethods<'a, 'tcx>> FunctionCx<'a, 'tcx, Bx> {
                 let mk_ref = move |tcx: TyCtxt<'tcx>, ty: Ty<'tcx>| {
                     Ty::new_ref(tcx, tcx.lifetimes.re_erased, ty, bk.to_mutbl_lossy())
                 };
-                self.codegen_place_to_pointer(bx, place, mk_ref)
+                let op = self.codegen_place_to_pointer(bx, place, mk_ref);
+                if self.cx.tcx().sess.opts.unstable_opts.codegen_emit_retag.is_some() {
+                    self.codegen_retag_operand(bx, op, false)
+                } else {
+                    op
+                }
             }
 
             // Note: Exclusive reborrowing is always equal to a memcpy, as the types do not change.

compiler/rustc_codegen_ssa/src/mir/statement.rs

@@ -3,6 +3,7 @@ use rustc_middle::{bug, span_bug, ty};
 use tracing::instrument;
 
 use super::{FunctionCx, LocalRef};
+use crate::mir::retag;
 use crate::traits::*;
 
 impl<'a, 'tcx, Bx: BuilderMethods<'a, 'tcx>> FunctionCx<'a, 'tcx, Bx> {
@@ -12,9 +13,17 @@ impl<'a, 'tcx, Bx: BuilderMethods<'a, 'tcx>> FunctionCx<'a, 'tcx, Bx> {
         self.set_debug_loc(bx, statement.source_info);
         match statement.kind {
             mir::StatementKind::Assign((ref place, ref rvalue)) => {
+                let needs_retag = bx.tcx().sess.opts.unstable_opts.codegen_emit_retag.is_some()
+                    && retag::rvalue_needs_retag(rvalue);
+
                 if let Some(index) = place.as_local() {
                     match self.locals[index] {
-                        LocalRef::Place(cg_dest) => self.codegen_rvalue(bx, cg_dest, rvalue),
+                        LocalRef::Place(cg_dest) => {
+                            self.codegen_rvalue(bx, cg_dest, rvalue);
+                            if needs_retag {
+                                self.codegen_retag_place(bx, cg_dest, false);
+                            }
+                        }
                         LocalRef::UnsizedPlace(cg_indirect_dest) => {
                             let ty = cg_indirect_dest.layout.ty;
                             span_bug!(
@@ -24,7 +33,10 @@ impl<'a, 'tcx, Bx: BuilderMethods<'a, 'tcx>> FunctionCx<'a, 'tcx, Bx> {
                             );
                         }
                         LocalRef::PendingOperand => {
-                            let operand = self.codegen_rvalue_operand(bx, rvalue);
+                            let mut operand = self.codegen_rvalue_operand(bx, rvalue);
+                            if needs_retag {
+                                operand = self.codegen_retag_operand(bx, operand, false);
+                            }
                             self.overwrite_local(index, LocalRef::Operand(operand));
                             self.debug_introduce_local(bx, index);
                         }
@@ -39,12 +51,19 @@ impl<'a, 'tcx, Bx: BuilderMethods<'a, 'tcx>> FunctionCx<'a, 'tcx, Bx> {
 
                             // If the type is zero-sized, it's already been set here,
                             // but we still need to make sure we codegen the operand
-                            self.codegen_rvalue_operand(bx, rvalue);
+                            // and emit a retag.
+                            let operand = self.codegen_rvalue_operand(bx, rvalue);
+                            if needs_retag {
+                                self.codegen_retag_operand(bx, operand, false);
+                            }
                         }
                     }
                 } else {
                     let cg_dest = self.codegen_place(bx, place.as_ref());
                     self.codegen_rvalue(bx, cg_dest, rvalue);
+                    if needs_retag {
+                        self.codegen_retag_place(bx, cg_dest, false);
+                    }
                 }
             }
             mir::StatementKind::SetDiscriminant { ref place, variant_index } => {

compiler/rustc_interface/src/tests.rs

@@ -10,9 +10,9 @@ use rustc_errors::ColorConfig;
 use rustc_errors::emitter::HumanReadableErrorType;
 use rustc_hir::attrs::{CollapseMacroDebuginfo, NativeLibKind};
 use rustc_session::config::{
-    AnnotateMoves, AutoDiff, BranchProtection, CFGuard, Cfg, CoverageLevel, CoverageOptions,
-    DebugInfo, DumpMonoStatsFormat, ErrorOutputType, ExternEntry, ExternLocation, Externs,
-    FmtDebug, FunctionReturn, IncrementalStateAssertion, InliningThreshold, Input,
+    AnnotateMoves, AutoDiff, BranchProtection, CFGuard, Cfg, CodegenRetagOptions, CoverageLevel,
+    CoverageOptions, DebugInfo, DumpMonoStatsFormat, ErrorOutputType, ExternEntry, ExternLocation,
+    Externs, FmtDebug, FunctionReturn, IncrementalStateAssertion, InliningThreshold, Input,
     InstrumentCoverage, InstrumentXRay, LinkSelfContained, LinkerPluginLto, LocationDetail, LtoCli,
     MirIncludeSpans, NextSolverConfig, Offload, Options, OutFileName, OutputType, OutputTypes,
     PAuthKey, PacRet, Passes, PatchableFunctionEntry, Polonius, ProcMacroExecutionStrategy, Strip,
@@ -772,6 +772,7 @@ fn test_unstable_options_tracking_hash() {
         })
     );
     tracked!(codegen_backend, Some("abc".to_string()));
+    tracked!(codegen_emit_retag, Some(CodegenRetagOptions::default()));
     tracked!(
         coverage_options,
         CoverageOptions {

compiler/rustc_session/src/config.rs

@@ -200,6 +200,15 @@ pub enum Offload {
     Test,
 }
 
+/// The different settings that the `-Z codegen-emit-retag` flag can have.
+#[derive(Copy, Clone, Debug, Default, PartialEq, Hash, Encodable, Decodable)]
+pub struct CodegenRetagOptions {
+    /// Track interior mutable data on the level of references, instead of on the byte level.
+    pub no_precise_im: bool,
+    /// Track `UnsafePinned` data on the level of references, instead of on the byte level.
+    pub no_precise_pin: bool,
+}
+
 /// The different settings that the `-Z autodiff` flag can have.
 #[derive(Clone, PartialEq, Hash, Debug, Encodable, Decodable)]
 pub enum AutoDiff {
@@ -3045,12 +3054,13 @@ pub(crate) mod dep_tracking {
     };
 
     use super::{
-        AnnotateMoves, AutoDiff, BranchProtection, CFGuard, CFProtection, CoverageOptions,
-        CrateType, DebugInfo, DebugInfoCompression, ErrorOutputType, FmtDebug, FunctionReturn,
-        InliningThreshold, InstrumentCoverage, InstrumentXRay, LinkerPluginLto, LocationDetail,
-        LtoCli, MirStripDebugInfo, NextSolverConfig, Offload, OptLevel, OutFileName, OutputType,
-        OutputTypes, PatchableFunctionEntry, Polonius, ResolveDocLinks, SourceFileHashAlgorithm,
-        SplitDwarfKind, SwitchWithOptPath, SymbolManglingVersion, WasiExecModel,
+        AnnotateMoves, AutoDiff, BranchProtection, CFGuard, CFProtection, CodegenRetagOptions,
+        CoverageOptions, CrateType, DebugInfo, DebugInfoCompression, ErrorOutputType, FmtDebug,
+        FunctionReturn, InliningThreshold, InstrumentCoverage, InstrumentXRay, LinkerPluginLto,
+        LocationDetail, LtoCli, MirStripDebugInfo, NextSolverConfig, Offload, OptLevel,
+        OutFileName, OutputType, OutputTypes, PatchableFunctionEntry, Polonius, ResolveDocLinks,
+        SourceFileHashAlgorithm, SplitDwarfKind, SwitchWithOptPath, SymbolManglingVersion,
+        WasiExecModel,
     };
     use crate::lint;
     use crate::utils::NativeLib;
@@ -3154,6 +3164,7 @@ pub(crate) mod dep_tracking {
         InliningThreshold,
         FunctionReturn,
         Align,
+        CodegenRetagOptions
     );
 
     impl<T1, T2> DepTrackingHash for (T1, T2)

compiler/rustc_session/src/options.rs

@@ -783,6 +783,8 @@ mod desc {
     pub(crate) const parse_dump_mono_stats: &str = "`markdown` (default) or `json`";
     pub(crate) const parse_instrument_coverage: &str = parse_bool;
     pub(crate) const parse_coverage_options: &str = "`block` | `branch` | `condition`";
+    pub(crate) const parse_codegen_retag_options: &str =
+        "either no value or a comma-separated list of settings: `no-precise-im`, `no-precise-pin`";
     pub(crate) const parse_instrument_xray: &str = "either a boolean (`yes`, `no`, `on`, `off`, etc), or a comma separated list of settings: `always` or `never` (mutually exclusive), `ignore-loops`, `instruction-threshold=N`, `skip-entry`, `skip-exit`";
     pub(crate) const parse_unpretty: &str = "`string` or `string=string`";
     pub(crate) const parse_treat_err_as_bug: &str = "either no value or a non-negative number";
@@ -1523,6 +1525,29 @@ pub mod parse {
         true
     }
 
+    pub(crate) fn parse_codegen_retag_options(
+        slot: &mut Option<CodegenRetagOptions>,
+        v: Option<&str>,
+    ) -> bool {
+        let mut no_precise_im = false;
+        let mut no_precise_pin = false;
+        if let Some(opt_list) = v.map(|s| s.split(',')) {
+            for opt in opt_list {
+                match opt {
+                    "no-precise-im" => {
+                        no_precise_im = true;
+                    }
+                    "no-precise-pin" => {
+                        no_precise_pin = true;
+                    }
+                    _ => return false,
+                }
+            }
+        }
+        *slot = Some(CodegenRetagOptions { no_precise_im, no_precise_pin });
+        true
+    }
+
     pub(crate) fn parse_coverage_options(slot: &mut CoverageOptions, v: Option<&str>) -> bool {
         let Some(v) = v else { return true };
 
@@ -2242,6 +2267,8 @@ options! {
         "hash algorithm of source files used to check freshness in cargo (`blake3` or `sha256`)"),
     codegen_backend: Option<String> = (None, parse_opt_string, [TRACKED],
         "the backend to use"),
+    codegen_emit_retag: Option<CodegenRetagOptions> = (None, parse_codegen_retag_options, [TRACKED],
+        "emit retag function calls in generated code"),
     codegen_source_order: bool = (false, parse_bool, [UNTRACKED],
         "emit mono items in the order of spans in source files (default: no)"),
     contract_checks: Option<bool> = (None, parse_opt_bool, [TRACKED],

compiler/rustc_session/src/session.rs

@@ -553,6 +553,8 @@ impl Session {
         // HWAddressSanitizer and KernelHWAddressSanitizer will use lifetimes to detect use after
         // scope bugs in the future.
         || self.sanitizers().intersects(SanitizerSet::ADDRESS | SanitizerSet::KERNELADDRESS | SanitizerSet::MEMORY | SanitizerSet::HWADDRESS | SanitizerSet::KERNELHWADDRESS)
+        // Lifetimes are necessary for retagging semantics.
+        || self.opts.unstable_opts.codegen_emit_retag.is_some()
     }
 
     pub fn diagnostic_width(&self) -> usize {