Auto merge of rust-lang#80463 - tgnottingham:incr_comp_serial_mem_usage, r=oli-obk

Serialize incr comp structures to file via fixed-size buffer

Reduce a large memory spike that happens during serialization by writing
the incr comp structures to file by way of a fixed-size buffer, rather
than an unbounded vector.

Effort was made to keep the instruction count close to that of the
previous implementation. However, buffered writing to a file inherently
has more overhead than writing to a vector, because each write may
result in a handleable error. To reduce this overhead, arrangements are
made so that each LEB128-encoded integer can be written to the buffer
with only one capacity and error check. Higher-level optimizations in
which entire composite structures can be written with one capacity and
error check are possible, but would require much more work.

The performance is mostly on par with the previous implementation, with
small to moderate instruction count regressions. The memory reduction is
significant, however, so it seems like a worth-while trade-off.

Author: bors

compiler/rustc_data_structures/src/fingerprint.rs

@@ -1,6 +1,6 @@
 use crate::stable_hasher;
 use rustc_serialize::{
-    opaque::{self, EncodeResult},
+    opaque::{self, EncodeResult, FileEncodeResult},
     Decodable, Encodable,
 };
 use std::hash::{Hash, Hasher};
@@ -53,13 +53,6 @@ impl Fingerprint {
         format!("{:x}{:x}", self.0, self.1)
     }
 
-    pub fn encode_opaque(&self, encoder: &mut opaque::Encoder) -> EncodeResult {
-        let bytes: [u8; 16] = unsafe { mem::transmute([self.0.to_le(), self.1.to_le()]) };
-
-        encoder.emit_raw_bytes(&bytes);
-        Ok(())
-    }
-
     pub fn decode_opaque(decoder: &mut opaque::Decoder<'_>) -> Result<Fingerprint, String> {
         let mut bytes: [MaybeUninit<u8>; 16] = MaybeUninit::uninit_array();
 
@@ -142,7 +135,16 @@ impl<E: rustc_serialize::Encoder> FingerprintEncoder for E {
 
 impl FingerprintEncoder for opaque::Encoder {
     fn encode_fingerprint(&mut self, f: &Fingerprint) -> EncodeResult {
-        f.encode_opaque(self)
+        let bytes: [u8; 16] = unsafe { mem::transmute([f.0.to_le(), f.1.to_le()]) };
+        self.emit_raw_bytes(&bytes);
+        Ok(())
+    }
+}
+
+impl FingerprintEncoder for opaque::FileEncoder {
+    fn encode_fingerprint(&mut self, f: &Fingerprint) -> FileEncodeResult {
+        let bytes: [u8; 16] = unsafe { mem::transmute([f.0.to_le(), f.1.to_le()]) };
+        self.emit_raw_bytes(&bytes)
     }
 }
 
@@ -198,7 +200,7 @@ impl<E: rustc_serialize::Encoder> Encodable<E> for PackedFingerprint {
 impl<D: rustc_serialize::Decoder> Decodable<D> for PackedFingerprint {
     #[inline]
     fn decode(d: &mut D) -> Result<Self, D::Error> {
-        Fingerprint::decode(d).map(|f| PackedFingerprint(f))
+        Fingerprint::decode(d).map(PackedFingerprint)
     }
 }
 

compiler/rustc_incremental/src/persist/file_format.rs

@@ -14,7 +14,7 @@ use std::fs;
 use std::io::{self, Read};
 use std::path::Path;
 
-use rustc_serialize::opaque::Encoder;
+use rustc_serialize::opaque::{FileEncodeResult, FileEncoder};
 
 /// The first few bytes of files generated by incremental compilation.
 const FILE_MAGIC: &[u8] = b"RSIC";
@@ -27,15 +27,17 @@ const HEADER_FORMAT_VERSION: u16 = 0;
 /// the Git commit hash.
 const RUSTC_VERSION: Option<&str> = option_env!("CFG_VERSION");
 
-pub fn write_file_header(stream: &mut Encoder, nightly_build: bool) {
-    stream.emit_raw_bytes(FILE_MAGIC);
-    stream
-        .emit_raw_bytes(&[(HEADER_FORMAT_VERSION >> 0) as u8, (HEADER_FORMAT_VERSION >> 8) as u8]);
+pub fn write_file_header(stream: &mut FileEncoder, nightly_build: bool) -> FileEncodeResult {
+    stream.emit_raw_bytes(FILE_MAGIC)?;
+    stream.emit_raw_bytes(&[
+        (HEADER_FORMAT_VERSION >> 0) as u8,
+        (HEADER_FORMAT_VERSION >> 8) as u8,
+    ])?;
 
     let rustc_version = rustc_version(nightly_build);
     assert_eq!(rustc_version.len(), (rustc_version.len() as u8) as usize);
-    stream.emit_raw_bytes(&[rustc_version.len() as u8]);
-    stream.emit_raw_bytes(rustc_version.as_bytes());
+    stream.emit_raw_bytes(&[rustc_version.len() as u8])?;
+    stream.emit_raw_bytes(rustc_version.as_bytes())
 }
 
 /// Reads the contents of a file with a file header as defined in this module.

compiler/rustc_incremental/src/persist/save.rs

@@ -2,7 +2,7 @@ use rustc_data_structures::fx::FxHashMap;
 use rustc_data_structures::sync::join;
 use rustc_middle::dep_graph::{DepGraph, DepKind, WorkProduct, WorkProductId};
 use rustc_middle::ty::TyCtxt;
-use rustc_serialize::opaque::Encoder;
+use rustc_serialize::opaque::{FileEncodeResult, FileEncoder};
 use rustc_serialize::Encodable as RustcEncodable;
 use rustc_session::Session;
 use std::fs;
@@ -33,12 +33,12 @@ pub fn save_dep_graph(tcx: TyCtxt<'_>) {
         join(
             move || {
                 sess.time("incr_comp_persist_result_cache", || {
-                    save_in(sess, query_cache_path, |e| encode_query_cache(tcx, e));
+                    save_in(sess, query_cache_path, "query cache", |e| encode_query_cache(tcx, e));
                 });
             },
             || {
                 sess.time("incr_comp_persist_dep_graph", || {
-                    save_in(sess, dep_graph_path, |e| {
+                    save_in(sess, dep_graph_path, "dependency graph", |e| {
                         sess.time("incr_comp_encode_dep_graph", || encode_dep_graph(tcx, e))
                     });
                 });
@@ -65,7 +65,7 @@ pub fn save_work_product_index(
     debug!("save_work_product_index()");
     dep_graph.assert_ignored();
     let path = work_products_path(sess);
-    save_in(sess, path, |e| encode_work_product_index(&new_work_products, e));
+    save_in(sess, path, "work product index", |e| encode_work_product_index(&new_work_products, e));
 
     // We also need to clean out old work-products, as not all of them are
     // deleted during invalidation. Some object files don't change their
@@ -92,13 +92,13 @@ pub fn save_work_product_index(
     });
 }
 
-fn save_in<F>(sess: &Session, path_buf: PathBuf, encode: F)
+fn save_in<F>(sess: &Session, path_buf: PathBuf, name: &str, encode: F)
 where
-    F: FnOnce(&mut Encoder),
+    F: FnOnce(&mut FileEncoder) -> FileEncodeResult,
 {
     debug!("save: storing data in {}", path_buf.display());
 
-    // delete the old dep-graph, if any
+    // Delete the old file, if any.
     // Note: It's important that we actually delete the old file and not just
     // truncate and overwrite it, since it might be a shared hard-link, the
     // underlying data of which we don't want to modify
@@ -109,34 +109,44 @@ where
         Err(err) if err.kind() == io::ErrorKind::NotFound => (),
         Err(err) => {
             sess.err(&format!(
-                "unable to delete old dep-graph at `{}`: {}",
+                "unable to delete old {} at `{}`: {}",
+                name,
                 path_buf.display(),
                 err
             ));
             return;
         }
     }
 
-    // generate the data in a memory buffer
-    let mut encoder = Encoder::new(Vec::new());
-    file_format::write_file_header(&mut encoder, sess.is_nightly_build());
-    encode(&mut encoder);
-
-    // write the data out
-    let data = encoder.into_inner();
-    match fs::write(&path_buf, data) {
-        Ok(_) => {
-            debug!("save: data written to disk successfully");
-        }
+    let mut encoder = match FileEncoder::new(&path_buf) {
+        Ok(encoder) => encoder,
         Err(err) => {
-            sess.err(&format!("failed to write dep-graph to `{}`: {}", path_buf.display(), err));
+            sess.err(&format!("failed to create {} at `{}`: {}", name, path_buf.display(), err));
+            return;
         }
+    };
+
+    if let Err(err) = file_format::write_file_header(&mut encoder, sess.is_nightly_build()) {
+        sess.err(&format!("failed to write {} header to `{}`: {}", name, path_buf.display(), err));
+        return;
+    }
+
+    if let Err(err) = encode(&mut encoder) {
+        sess.err(&format!("failed to write {} to `{}`: {}", name, path_buf.display(), err));
+        return;
     }
+
+    if let Err(err) = encoder.flush() {
+        sess.err(&format!("failed to flush {} to `{}`: {}", name, path_buf.display(), err));
+        return;
+    }
+
+    debug!("save: data written to disk successfully");
 }
 
-fn encode_dep_graph(tcx: TyCtxt<'_>, encoder: &mut Encoder) {
+fn encode_dep_graph(tcx: TyCtxt<'_>, encoder: &mut FileEncoder) -> FileEncodeResult {
     // First encode the commandline arguments hash
-    tcx.sess.opts.dep_tracking_hash().encode(encoder).unwrap();
+    tcx.sess.opts.dep_tracking_hash().encode(encoder)?;
 
     // Encode the graph data.
     let serialized_graph =
@@ -214,15 +224,13 @@ fn encode_dep_graph(tcx: TyCtxt<'_>, encoder: &mut Encoder) {
         println!("[incremental]");
     }
 
-    tcx.sess.time("incr_comp_encode_serialized_dep_graph", || {
-        serialized_graph.encode(encoder).unwrap();
-    });
+    tcx.sess.time("incr_comp_encode_serialized_dep_graph", || serialized_graph.encode(encoder))
 }
 
 fn encode_work_product_index(
     work_products: &FxHashMap<WorkProductId, WorkProduct>,
-    encoder: &mut Encoder,
-) {
+    encoder: &mut FileEncoder,
+) -> FileEncodeResult {
     let serialized_products: Vec<_> = work_products
         .iter()
         .map(|(id, work_product)| SerializedWorkProduct {
@@ -231,11 +239,9 @@ fn encode_work_product_index(
         })
         .collect();
 
-    serialized_products.encode(encoder).unwrap();
+    serialized_products.encode(encoder)
 }
 
-fn encode_query_cache(tcx: TyCtxt<'_>, encoder: &mut Encoder) {
-    tcx.sess.time("incr_comp_serialize_result_cache", || {
-        tcx.serialize_query_result_cache(encoder).unwrap();
-    })
+fn encode_query_cache(tcx: TyCtxt<'_>, encoder: &mut FileEncoder) -> FileEncodeResult {
+    tcx.sess.time("incr_comp_serialize_result_cache", || tcx.serialize_query_result_cache(encoder))
 }

compiler/rustc_metadata/src/rmeta/encoder.rs

@@ -308,7 +308,7 @@ impl<'a, 'tcx> Encodable<EncodeContext<'a, 'tcx>> for Span {
 
 impl<'a, 'tcx> FingerprintEncoder for EncodeContext<'a, 'tcx> {
     fn encode_fingerprint(&mut self, f: &Fingerprint) -> Result<(), Self::Error> {
-        f.encode_opaque(&mut self.opaque)
+        self.opaque.encode_fingerprint(f)
     }
 }
 

compiler/rustc_middle/src/ty/codec.rs

@@ -50,22 +50,6 @@ impl<'tcx, E: TyEncoder<'tcx>> EncodableWithShorthand<'tcx, E> for ty::Predicate
     }
 }
 
-pub trait OpaqueEncoder: Encoder {
-    fn opaque(&mut self) -> &mut rustc_serialize::opaque::Encoder;
-    fn encoder_position(&self) -> usize;
-}
-
-impl OpaqueEncoder for rustc_serialize::opaque::Encoder {
-    #[inline]
-    fn opaque(&mut self) -> &mut rustc_serialize::opaque::Encoder {
-        self
-    }
-    #[inline]
-    fn encoder_position(&self) -> usize {
-        self.position()
-    }
-}
-
 pub trait TyEncoder<'tcx>: Encoder {
     const CLEAR_CROSS_CRATE: bool;
 

compiler/rustc_middle/src/ty/context.rs

@@ -47,6 +47,7 @@ use rustc_hir::{
 };
 use rustc_index::vec::{Idx, IndexVec};
 use rustc_macros::HashStable;
+use rustc_serialize::opaque::{FileEncodeResult, FileEncoder};
 use rustc_session::config::{BorrowckMode, CrateType, OutputFilenames};
 use rustc_session::lint::{Level, Lint};
 use rustc_session::Session;
@@ -1336,10 +1337,7 @@ impl<'tcx> TyCtxt<'tcx> {
         }
     }
 
-    pub fn serialize_query_result_cache<E>(self, encoder: &mut E) -> Result<(), E::Error>
-    where
-        E: ty::codec::OpaqueEncoder,
-    {
+    pub fn serialize_query_result_cache(self, encoder: &mut FileEncoder) -> FileEncodeResult {
         self.queries.on_disk_cache.as_ref().map(|c| c.serialize(self, encoder)).unwrap_or(Ok(()))
     }