Auto merge of #74967 - Aaron1011:feature/incr-def-path-table, r=pnkfelix

Implement lazy decoding of DefPathTable during incremental compilation

PR https://github.com/rust-lang/rust/pull/75813 implemented lazy decoding of the `DefPathTable` from crate metadata. However, it requires decoding the entire `DefPathTable` when incremental compilation is active, so that we can map a decoded `DefPathHash` to a `DefId` from an arbitrary crate.

This PR adds support for lazy decoding of dependency `DefPathTable`s when incremental compilation si active.

When we load the incremental cache and dep
graph, we need the ability to map a `DefPathHash` to a `DefId` in the
current compilation session (if the corresponding definition still
exists).

This is accomplished by storing the old `DefId` (that is, the `DefId`
from the previous compilation session) for each `DefPathHash` we need to
remap. Since a `DefPathHash` includes the owning crate, the old crate is
guaranteed to be the right one (if the definition still exists). We then
use the old `DefIndex` as an initial guess, which we validate by
comparing the expected and actual `DefPathHash`es. In most cases,
foreign crates will be completely unchanged, which means that we our
guess will be correct. If our guess is wrong, we fall back to decoding
the entire `DefPathTable` for the foreign crate. This still represents
an improvement over the status quo, since we can skip decoding the
entire `DefPathTable` for other crates (where all of our guesses were
correct).

11 files changed, 289 insertions, 40 deletions

diff --git a/compiler/rustc_incremental/src/persist/load.rs b/compiler/rustc_incremental/src/persist/load.rs
index 201f381f889..35428dc8d84 100644
--- a/compiler/rustc_incremental/src/persist/load.rs
+++ b/compiler/rustc_incremental/src/persist/load.rs
@@ -1,6 +1,7 @@
 //! Code to save/load the dep-graph from files.
 
 use rustc_data_structures::fx::FxHashMap;
+use rustc_hir::definitions::Definitions;
 use rustc_middle::dep_graph::{PreviousDepGraph, SerializedDepGraph, WorkProduct, WorkProductId};
 use rustc_middle::ty::query::OnDiskCache;
 use rustc_middle::ty::TyCtxt;
@@ -204,7 +205,10 @@ pub fn load_dep_graph(sess: &Session) -> DepGraphFuture {
 /// If we are not in incremental compilation mode, returns `None`.
 /// Otherwise, tries to load the query result cache from disk,
 /// creating an empty cache if it could not be loaded.
-pub fn load_query_result_cache(sess: &Session) -> Option<OnDiskCache<'_>> {
+pub fn load_query_result_cache<'a>(
+    sess: &'a Session,
+    definitions: &Definitions,
+) -> Option<OnDiskCache<'a>> {
     if sess.opts.incremental.is_none() {
         return None;
     }
@@ -217,7 +221,7 @@ pub fn load_query_result_cache(sess: &Session) -> Option<OnDiskCache<'_>> {
         sess.is_nightly_build(),
     ) {
         LoadResult::Ok { data: (bytes, start_pos) } => {
-            Some(OnDiskCache::new(sess, bytes, start_pos))
+            Some(OnDiskCache::new(sess, bytes, start_pos, definitions))
         }
         _ => Some(OnDiskCache::new_empty(sess.source_map())),
     }
diff --git a/compiler/rustc_interface/src/passes.rs b/compiler/rustc_interface/src/passes.rs
index 1f820024f77..c13d26c79d7 100644
--- a/compiler/rustc_interface/src/passes.rs
+++ b/compiler/rustc_interface/src/passes.rs
@@ -747,7 +747,7 @@ pub fn create_global_ctxt<'tcx>(
         Definitions::new(crate_name, sess.local_crate_disambiguator()),
     ));
 
-    let query_result_on_disk_cache = rustc_incremental::load_query_result_cache(sess);
+    let query_result_on_disk_cache = rustc_incremental::load_query_result_cache(sess, defs);
 
     let codegen_backend = compiler.codegen_backend();
     let mut local_providers = *DEFAULT_QUERY_PROVIDERS;
diff --git a/compiler/rustc_metadata/src/rmeta/decoder.rs b/compiler/rustc_metadata/src/rmeta/decoder.rs
index e82449f69fd..f53a4027806 100644
--- a/compiler/rustc_metadata/src/rmeta/decoder.rs
+++ b/compiler/rustc_metadata/src/rmeta/decoder.rs
@@ -77,6 +77,10 @@ crate struct CrateMetadata {
     raw_proc_macros: Option<&'static [ProcMacro]>,
     /// Source maps for code from the crate.
     source_map_import_info: OnceCell<Vec<ImportedSourceFile>>,
+    /// For every definition in this crate, maps its `DefPathHash` to its
+    /// `DefIndex`. See `raw_def_id_to_def_id` for more details about how
+    /// this is used.
+    def_path_hash_map: OnceCell<FxHashMap<DefPathHash, DefIndex>>,
     /// Used for decoding interpret::AllocIds in a cached & thread-safe manner.
     alloc_decoding_state: AllocDecodingState,
     /// The `DepNodeIndex` of the `DepNode` representing this upstream crate.
@@ -1519,6 +1523,53 @@ impl<'a, 'tcx> CrateMetadataRef<'a> {
             .or_insert_with(|| self.root.tables.def_keys.get(self, index).unwrap().decode(self))
     }
 
+    /// Finds the corresponding `DefId` for the provided `DefPathHash`, if it exists.
+    /// This is used by incremental compilation to map a serialized `DefPathHash` to
+    /// its `DefId` in the current session.
+    /// Normally, only one 'main' crate will change between incremental compilation sessions:
+    /// all dependencies will be completely unchanged. In this case, we can avoid
+    /// decoding every `DefPathHash` in the crate, since the `DefIndex` from the previous
+    /// session will still be valid. If our 'guess' is wrong (the `DefIndex` no longer exists,
+    /// or has a different `DefPathHash`, then we need to decode all `DefPathHashes` to determine
+    /// the correct mapping).
+    fn def_path_hash_to_def_id(
+        &self,
+        krate: CrateNum,
+        index_guess: u32,
+        hash: DefPathHash,
+    ) -> Option<DefId> {
+        let def_index_guess = DefIndex::from_u32(index_guess);
+        let old_hash = self
+            .root
+            .tables
+            .def_path_hashes
+            .get(self, def_index_guess)
+            .map(|lazy| lazy.decode(self));
+
+        // Fast path: the definition and its index is unchanged from the
+        // previous compilation session. There is no need to decode anything
+        // else
+        if old_hash == Some(hash) {
+            return Some(DefId { krate, index: def_index_guess });
+        }
+
+        // Slow path: We need to find out the new `DefIndex` of the provided
+        // `DefPathHash`, if its still exists. This requires decoding every `DefPathHash`
+        // stored in this crate.
+        let map = self.cdata.def_path_hash_map.get_or_init(|| {
+            let end_id = self.root.tables.def_path_hashes.size() as u32;
+            let mut map = FxHashMap::with_capacity_and_hasher(end_id as usize, Default::default());
+            for i in 0..end_id {
+                let def_index = DefIndex::from_u32(i);
+                let hash =
+                    self.root.tables.def_path_hashes.get(self, def_index).unwrap().decode(self);
+                map.insert(hash, def_index);
+            }
+            map
+        });
+        map.get(&hash).map(|index| DefId { krate, index: *index })
+    }
+
     // Returns the path leading to the thing with this `id`.
     fn def_path(&self, id: DefIndex) -> DefPath {
         debug!("def_path(cnum={:?}, id={:?})", self.cnum, id);
@@ -1797,6 +1848,7 @@ impl CrateMetadata {
             trait_impls,
             raw_proc_macros,
             source_map_import_info: OnceCell::new(),
+            def_path_hash_map: Default::default(),
             alloc_decoding_state,
             dep_node_index: AtomicCell::new(DepNodeIndex::INVALID),
             cnum,
diff --git a/compiler/rustc_metadata/src/rmeta/decoder/cstore_impl.rs b/compiler/rustc_metadata/src/rmeta/decoder/cstore_impl.rs
index 85dc60d7eed..2ffd239b2f0 100644
--- a/compiler/rustc_metadata/src/rmeta/decoder/cstore_impl.rs
+++ b/compiler/rustc_metadata/src/rmeta/decoder/cstore_impl.rs
@@ -506,6 +506,16 @@ impl CrateStore for CStore {
         self.get_crate_data(cnum).num_def_ids()
     }
 
+    // See `CrateMetadataRef::def_path_hash_to_def_id` for more details
+    fn def_path_hash_to_def_id(
+        &self,
+        cnum: CrateNum,
+        index_guess: u32,
+        hash: DefPathHash,
+    ) -> Option<DefId> {
+        self.get_crate_data(cnum).def_path_hash_to_def_id(cnum, index_guess, hash)
+    }
+
     fn crates_untracked(&self) -> Vec<CrateNum> {
         let mut result = vec![];
         self.iter_crate_data(|cnum, _| result.push(cnum));
diff --git a/compiler/rustc_middle/src/dep_graph/dep_node.rs b/compiler/rustc_middle/src/dep_graph/dep_node.rs
index 38bc3b46b0f..d954c8ab5fb 100644
--- a/compiler/rustc_middle/src/dep_graph/dep_node.rs
+++ b/compiler/rustc_middle/src/dep_graph/dep_node.rs
@@ -252,8 +252,7 @@ macro_rules! define_dep_nodes {
             /// has been removed.
             fn extract_def_id(&self, tcx: TyCtxt<'tcx>) -> Option<DefId> {
                 if self.kind.can_reconstruct_query_key() {
-                    let def_path_hash = DefPathHash(self.hash.into());
-                    tcx.def_path_hash_to_def_id.as_ref()?.get(&def_path_hash).cloned()
+                    tcx.queries.on_disk_cache.as_ref()?.def_path_hash_to_def_id(tcx, DefPathHash(self.hash.into()))
                 } else {
                     None
                 }
@@ -320,7 +319,17 @@ impl<'tcx> DepNodeParams<TyCtxt<'tcx>> for DefId {
     }
 
     fn to_fingerprint(&self, tcx: TyCtxt<'tcx>) -> Fingerprint {
-        tcx.def_path_hash(*self).0
+        let hash = tcx.def_path_hash(*self);
+        // If this is a foreign `DefId`, store its current value
+        // in the incremental cache. When we decode the cache,
+        // we will use the old DefIndex as an initial guess for
+        // a lookup into the crate metadata.
+        if !self.is_local() {
+            if let Some(cache) = &tcx.queries.on_disk_cache {
+                cache.store_foreign_def_id_hash(*self, hash);
+            }
+        }
+        hash.0
     }
 
     fn to_debug_str(&self, tcx: TyCtxt<'tcx>) -> String {
@@ -359,7 +368,7 @@ impl<'tcx> DepNodeParams<TyCtxt<'tcx>> for CrateNum {
 
     fn to_fingerprint(&self, tcx: TyCtxt<'tcx>) -> Fingerprint {
         let def_id = DefId { krate: *self, index: CRATE_DEF_INDEX };
-        tcx.def_path_hash(def_id).0
+        def_id.to_fingerprint(tcx)
     }
 
     fn to_debug_str(&self, tcx: TyCtxt<'tcx>) -> String {
diff --git a/compiler/rustc_middle/src/dep_graph/mod.rs b/compiler/rustc_middle/src/dep_graph/mod.rs
index a85b66b65da..a94f6d25fc7 100644
--- a/compiler/rustc_middle/src/dep_graph/mod.rs
+++ b/compiler/rustc_middle/src/dep_graph/mod.rs
@@ -5,7 +5,7 @@ use rustc_data_structures::profiling::SelfProfilerRef;
 use rustc_data_structures::sync::Lock;
 use rustc_data_structures::thin_vec::ThinVec;
 use rustc_errors::Diagnostic;
-use rustc_hir::def_id::LocalDefId;
+use rustc_hir::def_id::{DefPathHash, LocalDefId};
 
 mod dep_node;
 
@@ -91,6 +91,12 @@ impl<'tcx> DepContext for TyCtxt<'tcx> {
     type DepKind = DepKind;
     type StableHashingContext = StableHashingContext<'tcx>;
 
+    fn register_reused_dep_path_hash(&self, hash: DefPathHash) {
+        if let Some(cache) = self.queries.on_disk_cache.as_ref() {
+            cache.register_reused_dep_path_hash(hash)
+        }
+    }
+
     fn create_stable_hashing_context(&self) -> Self::StableHashingContext {
         TyCtxt::create_stable_hashing_context(*self)
     }
diff --git a/compiler/rustc_middle/src/middle/cstore.rs b/compiler/rustc_middle/src/middle/cstore.rs
index ae9e4d364d3..bd7121ca1d7 100644
--- a/compiler/rustc_middle/src/middle/cstore.rs
+++ b/compiler/rustc_middle/src/middle/cstore.rs
@@ -191,6 +191,12 @@ pub trait CrateStore {
     fn def_path_hash(&self, def: DefId) -> DefPathHash;
     fn all_def_path_hashes_and_def_ids(&self, cnum: CrateNum) -> Vec<(DefPathHash, DefId)>;
     fn num_def_ids(&self, cnum: CrateNum) -> usize;
+    fn def_path_hash_to_def_id(
+        &self,
+        cnum: CrateNum,
+        index_guess: u32,
+        hash: DefPathHash,
+    ) -> Option<DefId>;
 
     // "queries" used in resolve that aren't tracked for incremental compilation
     fn crate_name_untracked(&self, cnum: CrateNum) -> Symbol;
diff --git a/compiler/rustc_middle/src/ty/context.rs b/compiler/rustc_middle/src/ty/context.rs
index 744aa39b96c..f078bbacfe9 100644
--- a/compiler/rustc_middle/src/ty/context.rs
+++ b/compiler/rustc_middle/src/ty/context.rs
@@ -1,7 +1,7 @@
 //! Type context book-keeping.
 
 use crate::arena::Arena;
-use crate::dep_graph::{self, DepConstructor, DepGraph};
+use crate::dep_graph::{self, DepGraph, DepKind, DepNode, DepNodeExt};
 use crate::hir::exports::ExportMap;
 use crate::ich::{NodeIdHashingMode, StableHashingContext};
 use crate::infer::canonical::{Canonical, CanonicalVarInfo, CanonicalVarInfos};
@@ -34,12 +34,12 @@ use rustc_data_structures::stable_hasher::{
 };
 use rustc_data_structures::steal::Steal;
 use rustc_data_structures::sync::{self, Lock, Lrc, WorkerLocal};
-use rustc_data_structures::unhash::UnhashMap;
 use rustc_errors::ErrorReported;
 use rustc_hir as hir;
 use rustc_hir::def::{DefKind, Res};
-use rustc_hir::def_id::{CrateNum, DefId, DefIdMap, LocalDefId, LOCAL_CRATE};
-use rustc_hir::definitions::{DefPathHash, Definitions};
+use rustc_hir::def_id::{CrateNum, DefId, DefIdMap, LocalDefId};
+use rustc_hir::def_id::{CRATE_DEF_INDEX, LOCAL_CRATE};
+use rustc_hir::definitions::Definitions;
 use rustc_hir::intravisit::Visitor;
 use rustc_hir::lang_items::LangItem;
 use rustc_hir::{
@@ -945,10 +945,6 @@ pub struct GlobalCtxt<'tcx> {
     pub(crate) untracked_crate: &'tcx hir::Crate<'tcx>,
     pub(crate) definitions: &'tcx Definitions,
 
-    /// A map from `DefPathHash` -> `DefId`. Includes `DefId`s from the local crate
-    /// as well as all upstream crates. Only populated in incremental mode.
-    pub def_path_hash_to_def_id: Option<UnhashMap<DefPathHash, DefId>>,
-
     pub queries: query::Queries<'tcx>,
 
     maybe_unused_trait_imports: FxHashSet<LocalDefId>,
@@ -1113,21 +1109,6 @@ impl<'tcx> TyCtxt<'tcx> {
         let mut providers = IndexVec::from_elem_n(extern_providers, max_cnum + 1);
         providers[LOCAL_CRATE] = local_providers;
 
-        let def_path_hash_to_def_id = if s.opts.build_dep_graph() {
-            let capacity = definitions.def_path_table().num_def_ids()
-                + crates.iter().map(|cnum| cstore.num_def_ids(*cnum)).sum::<usize>();
-            let mut map = UnhashMap::with_capacity_and_hasher(capacity, Default::default());
-
-            map.extend(definitions.def_path_table().all_def_path_hashes_and_def_ids(LOCAL_CRATE));
-            for cnum in &crates {
-                map.extend(cstore.all_def_path_hashes_and_def_ids(*cnum).into_iter());
-            }
-
-            Some(map)
-        } else {
-            None
-        };
-
         let mut trait_map: FxHashMap<_, FxHashMap<_, _>> = FxHashMap::default();
         for (hir_id, v) in krate.trait_map.iter() {
             let map = trait_map.entry(hir_id.owner).or_default();
@@ -1155,7 +1136,6 @@ impl<'tcx> TyCtxt<'tcx> {
             extern_prelude: resolutions.extern_prelude,
             untracked_crate: krate,
             definitions,
-            def_path_hash_to_def_id,
             queries: query::Queries::new(providers, extern_providers, on_disk_query_result_cache),
             ty_rcache: Default::default(),
             pred_rcache: Default::default(),
@@ -1329,7 +1309,8 @@ impl<'tcx> TyCtxt<'tcx> {
         // We cannot use the query versions of crates() and crate_hash(), since
         // those would need the DepNodes that we are allocating here.
         for cnum in self.cstore.crates_untracked() {
-            let dep_node = DepConstructor::CrateMetadata(self, cnum);
+            let def_path_hash = self.def_path_hash(DefId { krate: cnum, index: CRATE_DEF_INDEX });
+            let dep_node = DepNode::from_def_path_hash(def_path_hash, DepKind::CrateMetadata);
             let crate_hash = self.cstore.crate_hash_untracked(cnum);
             self.dep_graph.with_task(
                 dep_node,
diff --git a/compiler/rustc_middle/src/ty/query/on_disk_cache.rs b/compiler/rustc_middle/src/ty/query/on_disk_cache.rs
index 173e9a31928..898cc24992b 100644
--- a/compiler/rustc_middle/src/ty/query/on_disk_cache.rs
+++ b/compiler/rustc_middle/src/ty/query/on_disk_cache.rs
@@ -11,6 +11,7 @@ use rustc_data_structures::thin_vec::ThinVec;
 use rustc_errors::Diagnostic;
 use rustc_hir::def_id::{CrateNum, DefId, DefIndex, LocalDefId, LOCAL_CRATE};
 use rustc_hir::definitions::DefPathHash;
+use rustc_hir::definitions::Definitions;
 use rustc_index::vec::{Idx, IndexVec};
 use rustc_serialize::{opaque, Decodable, Decoder, Encodable, Encoder};
 use rustc_session::{CrateDisambiguator, Session};
@@ -21,6 +22,8 @@ use rustc_span::hygiene::{
 use rustc_span::source_map::{SourceMap, StableSourceFileId};
 use rustc_span::CachingSourceMapView;
 use rustc_span::{BytePos, ExpnData, SourceFile, Span, DUMMY_SP};
+use std::collections::hash_map::Entry;
+use std::iter::FromIterator;
 use std::mem;
 
 const TAG_FILE_FOOTER: u128 = 0xC0FFEE_C0FFEE_C0FFEE_C0FFEE_C0FFEE;
@@ -80,6 +83,31 @@ pub struct OnDiskCache<'sess> {
     expn_data: FxHashMap<u32, AbsoluteBytePos>,
     // Additional information used when decoding hygiene data.
     hygiene_context: HygieneDecodeContext,
+    // Maps `DefPathHash`es to their `RawDefId`s from the *previous*
+    // compilation session. This is used as an initial 'guess' when
+    // we try to map a `DefPathHash` to its `DefId` in the current compilation
+    // session.
+    foreign_def_path_hashes: FxHashMap<DefPathHash, RawDefId>,
+
+    // The *next* compilation sessison's `foreign_def_path_hashes` - at
+    // the end of our current compilation session, this will get written
+    // out to the `foreign_def_path_hashes` field of the `Footer`, which
+    // will become `foreign_def_path_hashes` of the next compilation session.
+    // This stores any `DefPathHash` that we may need to map to a `DefId`
+    // during the next compilation session.
+    latest_foreign_def_path_hashes: Lock<FxHashMap<DefPathHash, RawDefId>>,
+
+    // Maps `DefPathHashes` to their corresponding `LocalDefId`s for all
+    // local items in the current compilation session. This is only populated
+    // when we are in incremental mode and have loaded a pre-existing cache
+    // from disk, since this map is only used when deserializing a `DefPathHash`
+    // from the incremental cache.
+    local_def_path_hash_to_def_id: FxHashMap<DefPathHash, LocalDefId>,
+    // Caches all lookups of `DefPathHashes`, both for local and foreign
+    // definitions. A definition from the previous compilation session
+    // may no longer exist in the current compilation session, so
+    // we use `Option<DefId>` so that we can cache a lookup failure.
+    def_path_hash_to_def_id_cache: Lock<FxHashMap<DefPathHash, Option<DefId>>>,
 }
 
 // This type is used only for serialization and deserialization.
@@ -95,6 +123,7 @@ struct Footer {
     syntax_contexts: FxHashMap<u32, AbsoluteBytePos>,
     // See `OnDiskCache.expn_data`
     expn_data: FxHashMap<u32, AbsoluteBytePos>,
+    foreign_def_path_hashes: FxHashMap<DefPathHash, RawDefId>,
 }
 
 type EncodedQueryResultIndex = Vec<(SerializedDepNodeIndex, AbsoluteBytePos)>;
@@ -118,9 +147,36 @@ impl AbsoluteBytePos {
     }
 }
 
+/// Represents a potentially invalid `DefId`. This is used during incremental
+/// compilation to represent a `DefId` from the *previous* compilation session,
+/// which may no longer be valid. This is used to help map a `DefPathHash`
+/// to a `DefId` in the current compilation session.
+#[derive(Encodable, Decodable, Copy, Clone, Debug)]
+crate struct RawDefId {
+    // We deliberately do not use `CrateNum` and `DefIndex`
+    // here, since a crate/index from the previous compilation
+    // session may no longer exist.
+    pub krate: u32,
+    pub index: u32,
+}
+
+fn make_local_def_path_hash_map(definitions: &Definitions) -> FxHashMap<DefPathHash, LocalDefId> {
+    FxHashMap::from_iter(
+        definitions
+            .def_path_table()
+            .all_def_path_hashes_and_def_ids(LOCAL_CRATE)
+            .map(|(hash, def_id)| (hash, def_id.as_local().unwrap())),
+    )
+}
+
 impl<'sess> OnDiskCache<'sess> {
     /// Creates a new `OnDiskCache` instance from the serialized data in `data`.
-    pub fn new(sess: &'sess Session, data: Vec<u8>, start_pos: usize) -> Self {
+    pub fn new(
+        sess: &'sess Session,
+        data: Vec<u8>,
+        start_pos: usize,
+        definitions: &Definitions,
+    ) -> Self {
         debug_assert!(sess.opts.incremental.is_some());
 
         // Wrap in a scope so we can borrow `data`.
@@ -155,6 +211,10 @@ impl<'sess> OnDiskCache<'sess> {
             syntax_contexts: footer.syntax_contexts,
             expn_data: footer.expn_data,
             hygiene_context: Default::default(),
+            foreign_def_path_hashes: footer.foreign_def_path_hashes,
+            latest_foreign_def_path_hashes: Default::default(),
+            local_def_path_hash_to_def_id: make_local_def_path_hash_map(definitions),
+            def_path_hash_to_def_id_cache: Default::default(),
         }
     }
 
@@ -173,6 +233,10 @@ impl<'sess> OnDiskCache<'sess> {
             syntax_contexts: FxHashMap::default(),
             expn_data: FxHashMap::default(),
             hygiene_context: Default::default(),
+            foreign_def_path_hashes: Default::default(),
+            latest_foreign_def_path_hashes: Default::default(),
+            local_def_path_hash_to_def_id: Default::default(),
+            def_path_hash_to_def_id_cache: Default::default(),
         }
     }
 
@@ -200,6 +264,15 @@ impl<'sess> OnDiskCache<'sess> {
                 (file_to_file_index, file_index_to_stable_id)
             };
 
+            // Load everything into memory so we can write it out to the on-disk
+            // cache. The vast majority of cacheable query results should already
+            // be in memory, so this should be a cheap operation.
+            // Do this *before* we clone 'latest_foreign_def_path_hashes', since
+            // loading existing queries may cause us to create new DepNodes, which
+            // may in turn end up invoking `store_foreign_def_id_hash`
+            tcx.dep_graph.exec_cache_promotions(tcx);
+
+            let latest_foreign_def_path_hashes = self.latest_foreign_def_path_hashes.lock().clone();
             let hygiene_encode_context = HygieneEncodeContext::default();
 
             let mut encoder = CacheEncoder {
@@ -211,13 +284,9 @@ impl<'sess> OnDiskCache<'sess> {
                 source_map: CachingSourceMapView::new(tcx.sess.source_map()),
                 file_to_file_index,
                 hygiene_context: &hygiene_encode_context,
+                latest_foreign_def_path_hashes,
             };
 
-            // Load everything into memory so we can write it out to the on-disk
-            // cache. The vast majority of cacheable query results should already
-            // be in memory, so this should be a cheap operation.
-            tcx.dep_graph.exec_cache_promotions(tcx);
-
             // Encode query results.
             let mut query_result_index = EncodedQueryResultIndex::new();
 
@@ -312,6 +381,9 @@ impl<'sess> OnDiskCache<'sess> {
                 },
             )?;
 
+            let foreign_def_path_hashes =
+                std::mem::take(&mut encoder.latest_foreign_def_path_hashes);
+
             // `Encode the file footer.
             let footer_pos = encoder.position() as u64;
             encoder.encode_tagged(
@@ -324,6 +396,7 @@ impl<'sess> OnDiskCache<'sess> {
                     interpret_alloc_index,
                     syntax_contexts,
                     expn_data: expn_ids,
+                    foreign_def_path_hashes,
                 },
             )?;
 
@@ -374,6 +447,31 @@ impl<'sess> OnDiskCache<'sess> {
         debug_assert!(prev.is_none());
     }
 
+    fn get_raw_def_id(&self, hash: &DefPathHash) -> Option<RawDefId> {
+        self.foreign_def_path_hashes.get(hash).copied()
+    }
+
+    fn try_remap_cnum(&self, tcx: TyCtxt<'_>, cnum: u32) -> Option<CrateNum> {
+        let cnum_map =
+            self.cnum_map.get_or_init(|| Self::compute_cnum_map(tcx, &self.prev_cnums[..]));
+
+        cnum_map[CrateNum::from_u32(cnum)]
+    }
+
+    pub(crate) fn store_foreign_def_id_hash(&self, def_id: DefId, hash: DefPathHash) {
+        // We may overwrite an existing entry, but it will have the same value,
+        // so it's fine
+        self.latest_foreign_def_path_hashes
+            .lock()
+            .insert(hash, RawDefId { krate: def_id.krate.as_u32(), index: def_id.index.as_u32() });
+    }
+
+    pub fn register_reused_dep_path_hash(&self, hash: DefPathHash) {
+        if let Some(old_id) = self.foreign_def_path_hashes.get(&hash) {
+            self.latest_foreign_def_path_hashes.lock().insert(hash, *old_id);
+        }
+    }
+
     /// Returns the cached query result if there is something in the cache for
     /// the given `SerializedDepNodeIndex`; otherwise returns `None`.
     crate fn try_load_query_result<'tcx, T>(
@@ -481,6 +579,53 @@ impl<'sess> OnDiskCache<'sess> {
             map
         })
     }
+
+    /// Converts a `DefPathHash` to its corresponding `DefId` in the current compilation
+    /// session, if it still exists. This is used during incremental compilation to
+    /// turn a deserialized `DefPathHash` into its current `DefId`.
+    pub(crate) fn def_path_hash_to_def_id(
+        &self,
+        tcx: TyCtxt<'tcx>,
+        hash: DefPathHash,
+    ) -> Option<DefId> {
+        let mut cache = self.def_path_hash_to_def_id_cache.lock();
+        match cache.entry(hash) {
+            Entry::Occupied(e) => *e.get(),
+            Entry::Vacant(e) => {
+                // Check if the `DefPathHash` corresponds to a definition in the current
+                // crate
+                if let Some(def_id) = self.local_def_path_hash_to_def_id.get(&hash).cloned() {
+                    let def_id = def_id.to_def_id();
+                    e.insert(Some(def_id));
+                    return Some(def_id);
+                }
+                // This `raw_def_id` represents the `DefId` of this `DefPathHash` in
+                // the *previous* compliation session. The `DefPathHash` includes the
+                // owning crate, so if the corresponding definition still exists in the
+                // current compilation session, the crate is guaranteed to be the same
+                // (otherwise, we would compute a different `DefPathHash`).
+                let raw_def_id = self.get_raw_def_id(&hash)?;
+                // If the owning crate no longer exists, the corresponding definition definitely
+                // no longer exists.
+                let krate = self.try_remap_cnum(tcx, raw_def_id.krate)?;
+                // If our `DefPathHash` corresponded to a definition in the local crate,
+                // we should have either found it in `local_def_path_hash_to_def_id`, or
+                // never attempted to load it in the first place. Any query result or `DepNode`
+                // that references a local `DefId` should depend on some HIR-related `DepNode`.
+                // If a local definition is removed/modified such that its old `DefPathHash`
+                // no longer has a corresponding definition, that HIR-related `DepNode` should
+                // end up red. This should prevent us from ever calling
+                // `tcx.def_path_hash_to_def_id`, since we'll end up recomputing any
+                // queries involved.
+                debug_assert_ne!(krate, LOCAL_CRATE);
+                // Try to find a definition in the current session, using the previous `DefIndex`
+                // as an initial guess.
+                let opt_def_id = tcx.cstore.def_path_hash_to_def_id(krate, raw_def_id.index, hash);
+                e.insert(opt_def_id);
+                opt_def_id
+            }
+        }
+    }
 }
 
 //- DECODING -------------------------------------------------------------------
@@ -713,7 +858,17 @@ impl<'a, 'tcx> Decodable<CacheDecoder<'a, 'tcx>> for DefId {
         let def_path_hash = DefPathHash::decode(d)?;
 
         // Using the `DefPathHash`, we can lookup the new `DefId`.
-        Ok(d.tcx().def_path_hash_to_def_id.as_ref().unwrap()[&def_path_hash])
+        // Subtle: We only encode a `DefId` as part of a query result.
+        // If we get to this point, then all of the query inputs were green,
+        // which means that the definition with this hash is guaranteed to
+        // still exist in the current compilation session.
+        Ok(d.tcx()
+            .queries
+            .on_disk_cache
+            .as_ref()
+            .unwrap()
+            .def_path_hash_to_def_id(d.tcx(), def_path_hash)
+            .unwrap())
     }
 }
 
@@ -773,6 +928,7 @@ struct CacheEncoder<'a, 'tcx, E: OpaqueEncoder> {
     source_map: CachingSourceMapView<'tcx>,
     file_to_file_index: FxHashMap<*const SourceFile, SourceFileIndex>,
     hygiene_context: &'a HygieneEncodeContext,
+    latest_foreign_def_path_hashes: FxHashMap<DefPathHash, RawDefId>,
 }
 
 impl<'a, 'tcx, E> CacheEncoder<'a, 'tcx, E>
@@ -895,6 +1051,15 @@ where
 {
     fn encode(&self, s: &mut CacheEncoder<'a, 'tcx, E>) -> Result<(), E::Error> {
         let def_path_hash = s.tcx.def_path_hash(*self);
+        // Store additional information when we encode a foreign `DefId`,
+        // so that we can map its `DefPathHash` back to a `DefId` in the next
+        // compilation session.
+        if !self.is_local() {
+            s.latest_foreign_def_path_hashes.insert(
+                def_path_hash,
+                RawDefId { krate: self.krate.as_u32(), index: self.index.as_u32() },
+            );
+        }
         def_path_hash.encode(s)
     }
 }
diff --git a/compiler/rustc_query_system/src/dep_graph/graph.rs b/compiler/rustc_query_system/src/dep_graph/graph.rs
index 617ec84ae71..ac37b296b53 100644
--- a/compiler/rustc_query_system/src/dep_graph/graph.rs
+++ b/compiler/rustc_query_system/src/dep_graph/graph.rs
@@ -7,6 +7,7 @@ use rustc_data_structures::sync::{AtomicU32, AtomicU64, Lock, Lrc, Ordering};
 use rustc_data_structures::unlikely;
 use rustc_errors::Diagnostic;
 use rustc_index::vec::{Idx, IndexVec};
+use rustc_span::def_id::DefPathHash;
 
 use parking_lot::{Condvar, Mutex};
 use smallvec::{smallvec, SmallVec};
@@ -699,6 +700,18 @@ impl<K: DepKind> DepGraph<K> {
             data.current.intern_node(*dep_node, current_deps, fingerprint)
         };
 
+        // We have just loaded a deserialized `DepNode` from the previous
+        // compilation session into the current one. If this was a foreign `DefId`,
+        // then we stored additional information in the incr comp cache when we
+        // initially created its fingerprint (see `DepNodeParams::to_fingerprint`)
+        // We won't be calling `to_fingerprint` again for this `DepNode` (we no longer
+        // have the original value), so we need to copy over this additional information
+        // from the old incremental cache into the new cache that we serialize
+        // and the end of this compilation session.
+        if dep_node.kind.can_reconstruct_query_key() {
+            tcx.register_reused_dep_path_hash(DefPathHash(dep_node.hash.into()));
+        }
+
         // ... emitting any stored diagnostic ...
 
         // FIXME: Store the fact that a node has diagnostics in a bit in the dep graph somewhere
diff --git a/compiler/rustc_query_system/src/dep_graph/mod.rs b/compiler/rustc_query_system/src/dep_graph/mod.rs
index e8d02692f37..3b4b62ad6be 100644
--- a/compiler/rustc_query_system/src/dep_graph/mod.rs
+++ b/compiler/rustc_query_system/src/dep_graph/mod.rs
@@ -15,6 +15,7 @@ use rustc_data_structures::profiling::SelfProfilerRef;
 use rustc_data_structures::sync::Lock;
 use rustc_data_structures::thin_vec::ThinVec;
 use rustc_errors::Diagnostic;
+use rustc_span::def_id::DefPathHash;
 
 use std::fmt;
 use std::hash::Hash;
@@ -32,6 +33,8 @@ pub trait DepContext: Copy {
     /// Try to force a dep node to execute and see if it's green.
     fn try_force_from_dep_node(&self, dep_node: &DepNode<Self::DepKind>) -> bool;
 
+    fn register_reused_dep_path_hash(&self, hash: DefPathHash);
+
     /// Return whether the current session is tainted by errors.
     fn has_errors_or_delayed_span_bugs(&self) -> bool;