Auto merge of #82780 - cjgillot:dep-stream, r=michaelwoerister

Stream the dep-graph to a file instead of storing it in-memory.

This is a reimplementation of #60035.

Instead of storing the dep-graph in-memory, the nodes are encoded as they come
into the a temporary file as they come. At the end of a successful the compilation,
this file is renamed to be the persistent dep-graph, to be decoded during the next
compilation session.

This two-files scheme avoids overwriting the dep-graph on unsuccessful or crashing compilations.

The structure of the file is modified to be the sequence of `(DepNode, Fingerprint, EdgesVec)`.
The deserialization is responsible for going to the more compressed representation.
The `node_count` and `edge_count` are stored in the last 16 bytes of the file,
in order to accurately reserve capacity for the vectors.

At the end of the compilation, the encoder is flushed and dropped.
The graph is not usable after this point: any creation of a node will ICE.

I had to retrofit the debugging options, which is not really pretty.

compiler/rustc_incremental/src/assert_dep_graph.rs

@@ -40,8 +40,9 @@ use rustc_graphviz as dot;
 use rustc_hir as hir;
 use rustc_hir::def_id::DefId;
 use rustc_hir::intravisit::{self, NestedVisitorMap, Visitor};
-use rustc_middle::dep_graph::debug::{DepNodeFilter, EdgeFilter};
-use rustc_middle::dep_graph::{DepGraphQuery, DepKind, DepNode, DepNodeExt};
+use rustc_middle::dep_graph::{
+    DepGraphQuery, DepKind, DepNode, DepNodeExt, DepNodeFilter, EdgeFilter,
+};
 use rustc_middle::hir::map::Map;
 use rustc_middle::ty::TyCtxt;
 use rustc_span::symbol::{sym, Symbol};
@@ -54,7 +55,7 @@ use std::io::{BufWriter, Write};
 pub fn assert_dep_graph(tcx: TyCtxt<'_>) {
     tcx.dep_graph.with_ignore(|| {
         if tcx.sess.opts.debugging_opts.dump_dep_graph {
-            dump_graph(tcx);
+            tcx.dep_graph.with_query(dump_graph);
         }
 
         if !tcx.sess.opts.debugging_opts.query_dep_graph {
@@ -200,29 +201,29 @@ fn check_paths<'tcx>(tcx: TyCtxt<'tcx>, if_this_changed: &Sources, then_this_wou
         }
         return;
     }
-    let query = tcx.dep_graph.query();
-    for &(_, source_def_id, ref source_dep_node) in if_this_changed {
-        let dependents = query.transitive_predecessors(source_dep_node);
-        for &(target_span, ref target_pass, _, ref target_dep_node) in then_this_would_need {
-            if !dependents.contains(&target_dep_node) {
-                tcx.sess.span_err(
-                    target_span,
-                    &format!(
-                        "no path from `{}` to `{}`",
-                        tcx.def_path_str(source_def_id),
-                        target_pass
-                    ),
-                );
-            } else {
-                tcx.sess.span_err(target_span, "OK");
+    tcx.dep_graph.with_query(|query| {
+        for &(_, source_def_id, ref source_dep_node) in if_this_changed {
+            let dependents = query.transitive_predecessors(source_dep_node);
+            for &(target_span, ref target_pass, _, ref target_dep_node) in then_this_would_need {
+                if !dependents.contains(&target_dep_node) {
+                    tcx.sess.span_err(
+                        target_span,
+                        &format!(
+                            "no path from `{}` to `{}`",
+                            tcx.def_path_str(source_def_id),
+                            target_pass
+                        ),
+                    );
+                } else {
+                    tcx.sess.span_err(target_span, "OK");
+                }
             }
         }
-    }
+    });
 }
 
-fn dump_graph(tcx: TyCtxt<'_>) {
+fn dump_graph(query: &DepGraphQuery) {
     let path: String = env::var("RUST_DEP_GRAPH").unwrap_or_else(|_| "dep_graph".to_string());
-    let query = tcx.dep_graph.query();
 
     let nodes = match env::var("RUST_DEP_GRAPH_FILTER") {
         Ok(string) => {

compiler/rustc_incremental/src/lib.rs

@@ -14,7 +14,7 @@ mod assert_dep_graph;
 pub mod assert_module_sources;
 mod persist;
 
-pub use assert_dep_graph::assert_dep_graph;
+use assert_dep_graph::assert_dep_graph;
 pub use persist::copy_cgu_workproduct_to_incr_comp_cache_dir;
 pub use persist::delete_workproduct_files;
 pub use persist::finalize_session_directory;
@@ -26,4 +26,4 @@ pub use persist::prepare_session_directory;
 pub use persist::save_dep_graph;
 pub use persist::save_work_product_index;
 pub use persist::LoadResult;
-pub use persist::{load_dep_graph, DepGraphFuture};
+pub use persist::{build_dep_graph, load_dep_graph, DepGraphFuture};

compiler/rustc_incremental/src/persist/dirty_clean.rs

@@ -14,7 +14,6 @@
 //! the required condition is not met.
 
 use rustc_ast::{self as ast, Attribute, NestedMetaItem};
-use rustc_data_structures::fingerprint::Fingerprint;
 use rustc_data_structures::fx::FxHashSet;
 use rustc_hir as hir;
 use rustc_hir::def_id::{DefId, LocalDefId};
@@ -381,39 +380,18 @@ impl DirtyCleanVisitor<'tcx> {
     fn assert_dirty(&self, item_span: Span, dep_node: DepNode) {
         debug!("assert_dirty({:?})", dep_node);
 
-        let current_fingerprint = self.get_fingerprint(&dep_node);
-        let prev_fingerprint = self.tcx.dep_graph.prev_fingerprint_of(&dep_node);
-
-        if current_fingerprint == prev_fingerprint {
+        if self.tcx.dep_graph.is_green(&dep_node) {
             let dep_node_str = self.dep_node_str(&dep_node);
             self.tcx
                 .sess
                 .span_err(item_span, &format!("`{}` should be dirty but is not", dep_node_str));
         }
     }
 
-    fn get_fingerprint(&self, dep_node: &DepNode) -> Option<Fingerprint> {
-        if self.tcx.dep_graph.dep_node_exists(dep_node) {
-            let dep_node_index = self.tcx.dep_graph.dep_node_index_of(dep_node);
-            Some(self.tcx.dep_graph.fingerprint_of(dep_node_index))
-        } else {
-            None
-        }
-    }
-
     fn assert_clean(&self, item_span: Span, dep_node: DepNode) {
         debug!("assert_clean({:?})", dep_node);
 
-        let current_fingerprint = self.get_fingerprint(&dep_node);
-        let prev_fingerprint = self.tcx.dep_graph.prev_fingerprint_of(&dep_node);
-
-        // if the node wasn't previously evaluated and now is (or vice versa),
-        // then the node isn't actually clean or dirty.
-        if (current_fingerprint == None) ^ (prev_fingerprint == None) {
-            return;
-        }
-
-        if current_fingerprint != prev_fingerprint {
+        if self.tcx.dep_graph.is_red(&dep_node) {
             let dep_node_str = self.dep_node_str(&dep_node);
             self.tcx
                 .sess

compiler/rustc_incremental/src/persist/fs.rs

@@ -122,6 +122,7 @@ mod tests;
 
 const LOCK_FILE_EXT: &str = ".lock";
 const DEP_GRAPH_FILENAME: &str = "dep-graph.bin";
+const STAGING_DEP_GRAPH_FILENAME: &str = "dep-graph.part.bin";
 const WORK_PRODUCTS_FILENAME: &str = "work-products.bin";
 const QUERY_CACHE_FILENAME: &str = "query-cache.bin";
 
@@ -134,6 +135,9 @@ const INT_ENCODE_BASE: usize = base_n::CASE_INSENSITIVE;
 pub fn dep_graph_path(sess: &Session) -> PathBuf {
     in_incr_comp_dir_sess(sess, DEP_GRAPH_FILENAME)
 }
+pub fn staging_dep_graph_path(sess: &Session) -> PathBuf {
+    in_incr_comp_dir_sess(sess, STAGING_DEP_GRAPH_FILENAME)
+}
 pub fn dep_graph_path_from(incr_comp_session_dir: &Path) -> PathBuf {
     in_incr_comp_dir(incr_comp_session_dir, DEP_GRAPH_FILENAME)
 }

compiler/rustc_incremental/src/persist/load.rs

@@ -5,7 +5,7 @@ use rustc_hir::definitions::Definitions;
 use rustc_middle::dep_graph::{PreviousDepGraph, SerializedDepGraph, WorkProduct, WorkProductId};
 use rustc_middle::ty::query::OnDiskCache;
 use rustc_serialize::opaque::Decoder;
-use rustc_serialize::Decodable as RustcDecodable;
+use rustc_serialize::Decodable;
 use rustc_session::Session;
 use std::path::Path;
 
@@ -120,7 +120,7 @@ pub fn load_dep_graph(sess: &Session) -> DepGraphFuture {
             // Decode the list of work_products
             let mut work_product_decoder = Decoder::new(&work_products_data[..], start_pos);
             let work_products: Vec<SerializedWorkProduct> =
-                RustcDecodable::decode(&mut work_product_decoder).unwrap_or_else(|e| {
+                Decodable::decode(&mut work_product_decoder).unwrap_or_else(|e| {
                     let msg = format!(
                         "Error decoding `work-products` from incremental \
                                     compilation session directory: {}",

compiler/rustc_incremental/src/persist/mod.rs

@@ -18,6 +18,7 @@ pub use fs::prepare_session_directory;
 pub use load::load_query_result_cache;
 pub use load::LoadResult;
 pub use load::{load_dep_graph, DepGraphFuture};
+pub use save::build_dep_graph;
 pub use save::save_dep_graph;
 pub use save::save_work_product_index;
 pub use work_product::copy_cgu_workproduct_to_incr_comp_cache_dir;

compiler/rustc_incremental/src/persist/save.rs

@@ -1,6 +1,6 @@
 use rustc_data_structures::fx::FxHashMap;
 use rustc_data_structures::sync::join;
-use rustc_middle::dep_graph::{DepGraph, WorkProduct, WorkProductId};
+use rustc_middle::dep_graph::{DepGraph, PreviousDepGraph, WorkProduct, WorkProductId};
 use rustc_middle::ty::TyCtxt;
 use rustc_serialize::opaque::{FileEncodeResult, FileEncoder};
 use rustc_serialize::Encodable as RustcEncodable;
@@ -15,6 +15,9 @@ use super::file_format;
 use super::fs::*;
 use super::work_product;
 
+/// Save and dump the DepGraph.
+///
+/// No query must be invoked after this function.
 pub fn save_dep_graph(tcx: TyCtxt<'_>) {
     debug!("save_dep_graph()");
     tcx.dep_graph.with_ignore(|| {
@@ -29,23 +32,41 @@ pub fn save_dep_graph(tcx: TyCtxt<'_>) {
 
         let query_cache_path = query_cache_path(sess);
         let dep_graph_path = dep_graph_path(sess);
+        let staging_dep_graph_path = staging_dep_graph_path(sess);
+
+        sess.time("assert_dep_graph", || crate::assert_dep_graph(tcx));
+        sess.time("check_dirty_clean", || dirty_clean::check_dirty_clean_annotations(tcx));
+
+        if sess.opts.debugging_opts.incremental_info {
+            tcx.dep_graph.print_incremental_info()
+        }
 
         join(
             move || {
                 sess.time("incr_comp_persist_result_cache", || {
                     save_in(sess, query_cache_path, "query cache", |e| encode_query_cache(tcx, e));
                 });
             },
-            || {
+            move || {
                 sess.time("incr_comp_persist_dep_graph", || {
-                    save_in(sess, dep_graph_path, "dependency graph", |e| {
-                        sess.time("incr_comp_encode_dep_graph", || encode_dep_graph(tcx, e))
-                    });
+                    if let Err(err) = tcx.dep_graph.encode(&tcx.sess.prof) {
+                        sess.err(&format!(
+                            "failed to write dependency graph to `{}`: {}",
+                            staging_dep_graph_path.display(),
+                            err
+                        ));
+                    }
+                    if let Err(err) = fs::rename(&staging_dep_graph_path, &dep_graph_path) {
+                        sess.err(&format!(
+                            "failed to move dependency graph from `{}` to `{}`: {}",
+                            staging_dep_graph_path.display(),
+                            dep_graph_path.display(),
+                            err
+                        ));
+                    }
                 });
             },
         );
-
-        dirty_clean::check_dirty_clean_annotations(tcx);
     })
 }
 
@@ -92,7 +113,7 @@ pub fn save_work_product_index(
     });
 }
 
-fn save_in<F>(sess: &Session, path_buf: PathBuf, name: &str, encode: F)
+pub(crate) fn save_in<F>(sess: &Session, path_buf: PathBuf, name: &str, encode: F)
 where
     F: FnOnce(&mut FileEncoder) -> FileEncodeResult,
 {
@@ -144,21 +165,6 @@ where
     debug!("save: data written to disk successfully");
 }
 
-fn encode_dep_graph(tcx: TyCtxt<'_>, encoder: &mut FileEncoder) -> FileEncodeResult {
-    // First encode the commandline arguments hash
-    tcx.sess.opts.dep_tracking_hash().encode(encoder)?;
-
-    if tcx.sess.opts.debugging_opts.incremental_info {
-        tcx.dep_graph.print_incremental_info();
-    }
-
-    // There is a tiny window between printing the incremental info above and encoding the dep
-    // graph below in which the dep graph could change, thus making the printed incremental info
-    // slightly out of date. If this matters to you, please feel free to submit a patch. :)
-
-    tcx.sess.time("incr_comp_encode_serialized_dep_graph", || tcx.dep_graph.encode(encoder))
-}
-
 fn encode_work_product_index(
     work_products: &FxHashMap<WorkProductId, WorkProduct>,
     encoder: &mut FileEncoder,
@@ -177,3 +183,56 @@ fn encode_work_product_index(
 fn encode_query_cache(tcx: TyCtxt<'_>, encoder: &mut FileEncoder) -> FileEncodeResult {
     tcx.sess.time("incr_comp_serialize_result_cache", || tcx.serialize_query_result_cache(encoder))
 }
+
+pub fn build_dep_graph(
+    sess: &Session,
+    prev_graph: PreviousDepGraph,
+    prev_work_products: FxHashMap<WorkProductId, WorkProduct>,
+) -> Option<DepGraph> {
+    if sess.opts.incremental.is_none() {
+        // No incremental compilation.
+        return None;
+    }
+
+    // Stream the dep-graph to an alternate file, to avoid overwriting anything in case of errors.
+    let path_buf = staging_dep_graph_path(sess);
+
+    let mut encoder = match FileEncoder::new(&path_buf) {
+        Ok(encoder) => encoder,
+        Err(err) => {
+            sess.err(&format!(
+                "failed to create dependency graph at `{}`: {}",
+                path_buf.display(),
+                err
+            ));
+            return None;
+        }
+    };
+
+    if let Err(err) = file_format::write_file_header(&mut encoder, sess.is_nightly_build()) {
+        sess.err(&format!(
+            "failed to write dependency graph header to `{}`: {}",
+            path_buf.display(),
+            err
+        ));
+        return None;
+    }
+
+    // First encode the commandline arguments hash
+    if let Err(err) = sess.opts.dep_tracking_hash().encode(&mut encoder) {
+        sess.err(&format!(
+            "failed to write dependency graph hash `{}`: {}",
+            path_buf.display(),
+            err
+        ));
+        return None;
+    }
+
+    Some(DepGraph::new(
+        prev_graph,
+        prev_work_products,
+        encoder,
+        sess.opts.debugging_opts.query_dep_graph,
+        sess.opts.debugging_opts.incremental_info,
+    ))
+}

compiler/rustc_interface/src/passes.rs

@@ -1021,9 +1021,6 @@ pub fn start_codegen<'tcx>(
         rustc_symbol_mangling::test::report_symbol_names(tcx);
     }
 
-    tcx.sess.time("assert_dep_graph", || rustc_incremental::assert_dep_graph(tcx));
-    tcx.sess.time("serialize_dep_graph", || rustc_incremental::save_dep_graph(tcx));
-
     info!("Post-codegen\n{:?}", tcx.debug_stats());
 
     if tcx.sess.opts.output_types.contains_key(&OutputType::Mir) {

compiler/rustc_interface/src/queries.rs

@@ -207,7 +207,13 @@ impl<'tcx> Queries<'tcx> {
                                 })
                                 .open(self.session())
                         });
-                    DepGraph::new(prev_graph, prev_work_products)
+
+                    rustc_incremental::build_dep_graph(
+                        self.session(),
+                        prev_graph,
+                        prev_work_products,
+                    )
+                    .unwrap_or_else(DepGraph::new_disabled)
                 }
             })
         })
@@ -435,6 +441,9 @@ impl Compiler {
             if self.session().opts.debugging_opts.query_stats {
                 gcx.enter(rustc_query_impl::print_stats);
             }
+
+            self.session()
+                .time("serialize_dep_graph", || gcx.enter(rustc_incremental::save_dep_graph));
         }
 
         _timer = Some(self.session().timer("free_global_ctxt"));

compiler/rustc_middle/src/dep_graph/mod.rs

@@ -8,8 +8,8 @@ use rustc_session::Session;
 mod dep_node;
 
 pub use rustc_query_system::dep_graph::{
-    debug, hash_result, DepContext, DepNodeColor, DepNodeIndex, SerializedDepNodeIndex,
-    WorkProduct, WorkProductId,
+    debug::DepNodeFilter, hash_result, DepContext, DepNodeColor, DepNodeIndex,
+    SerializedDepNodeIndex, WorkProduct, WorkProductId,
 };
 
 crate use dep_node::make_compile_codegen_unit;
@@ -20,6 +20,7 @@ pub type TaskDeps = rustc_query_system::dep_graph::TaskDeps<DepKind>;
 pub type DepGraphQuery = rustc_query_system::dep_graph::DepGraphQuery<DepKind>;
 pub type PreviousDepGraph = rustc_query_system::dep_graph::PreviousDepGraph<DepKind>;
 pub type SerializedDepGraph = rustc_query_system::dep_graph::SerializedDepGraph<DepKind>;
+pub type EdgeFilter = rustc_query_system::dep_graph::debug::EdgeFilter<DepKind>;
 
 impl rustc_query_system::dep_graph::DepKind for DepKind {
     const NULL: Self = DepKind::Null;

compiler/rustc_query_impl/src/plumbing.rs

@@ -477,10 +477,7 @@ macro_rules! define_queries {
                         return
                     }
 
-                    debug_assert!(tcx.dep_graph
-                                     .node_color(dep_node)
-                                     .map(|c| c.is_green())
-                                     .unwrap_or(false));
+                    debug_assert!(tcx.dep_graph.is_green(dep_node));
 
                     let key = recover(*tcx, dep_node).unwrap_or_else(|| panic!("Failed to recover key for {:?} with hash {}", dep_node, dep_node.hash));
                     if queries::$name::cache_on_disk(tcx, &key, None) {