Mark .cargo/git and .cargo/registry as cache dirs

crates/cargo-util/src/paths.rs

@@ -632,7 +632,7 @@ pub fn create_dir_all_excluded_from_backups_atomic(p: impl AsRef<Path>) -> Resul
     let parent = path.parent().unwrap();
     let base = path.file_name().unwrap();
     create_dir_all(parent)?;
-    // We do this in two steps (first create a temporary directory and exlucde
+    // We do this in two steps (first create a temporary directory and exclude
     // it from backups, then rename it to the desired name. If we created the
     // directory directly where it should be and then excluded it from backups
     // we would risk a situation where cargo is interrupted right after the directory
@@ -660,6 +660,14 @@ pub fn create_dir_all_excluded_from_backups_atomic(p: impl AsRef<Path>) -> Resul
     Ok(())
 }
 
+/// Mark an existing directory as excluded from backups and indexing.
+pub fn exclude_from_backups_and_indexing(p: impl AsRef<Path>) -> Result<()> {
+    let path = p.as_ref();
+    exclude_from_backups(path);
+    exclude_from_content_indexing(path);
+    Ok(())
+}
+
 /// Marks the directory as excluded from archives/backups.
 ///
 /// This is recommended to prevent derived/temporary files from bloating backups. There are two

src/cargo/sources/git/source.rs

@@ -7,6 +7,7 @@ use crate::util::errors::CargoResult;
 use crate::util::hex::short_hash;
 use crate::util::Config;
 use anyhow::Context;
+use cargo_util::paths::exclude_from_backups_and_indexing;
 use log::trace;
 use std::fmt::{self, Debug, Formatter};
 use std::task::Poll;
@@ -122,8 +123,20 @@ impl<'cfg> Source for GitSource<'cfg> {
             return Ok(());
         }
 
-        let git_path = self.config.git_path();
-        let git_path = self.config.assert_package_cache_locked(&git_path);
+        let git_fs = self.config.git_path();
+        git_fs.create_dir()?;
+        let git_path = self.config.assert_package_cache_locked(&git_fs);
+
+        // Before getting a checkout, make sure that `<cargo_home>/git` is
+        // marked as excluded from indexing and backups. Older versions of Cargo
+        // didn't do this, so we do it here regardless of whether `<cargo_home>`
+        // exists.
+        //
+        // This does not use `create_dir_all_excluded_from_backups_atomic` for
+        // the same reason: we want to exclude it even if the directory already
+        // exists.
+        exclude_from_backups_and_indexing(&git_path)?;
+
         let db_path = git_path.join("db").join(&self.ident);
 
         let db = self.remote.db_at(&db_path).ok();

src/cargo/sources/registry/mod.rs

@@ -167,6 +167,7 @@ use std::path::{Path, PathBuf};
 use std::task::Poll;
 
 use anyhow::Context as _;
+use cargo_util::paths::exclude_from_backups_and_indexing;
 use flate2::read::GzDecoder;
 use log::debug;
 use semver::Version;
@@ -552,6 +553,7 @@ impl<'cfg> RegistrySource<'cfg> {
         } else {
             Box::new(remote::RemoteRegistry::new(source_id, config, &name)) as Box<_>
         };
+
         Ok(RegistrySource::new(
             source_id,
             config,
@@ -812,6 +814,18 @@ impl<'cfg> Source for RegistrySource<'cfg> {
     }
 
     fn block_until_ready(&mut self) -> CargoResult<()> {
+        // Before starting to work on the registry, make sure that
+        // `<cargo_home>/registry` is marked as excluded from indexing and
+        // backups. Older versions of Cargo didn't do this, so we do it here
+        // regardless of whether `<cargo_home>` exists.
+        //
+        // This does not use `create_dir_all_excluded_from_backups_atomic` for
+        // the same reason: we want to exclude it even if the directory already
+        // exists.
+        let registry_base = self.config.registry_base_path();
+        registry_base.create_dir()?;
+        exclude_from_backups_and_indexing(&registry_base.into_path_unlocked())?;
+
         self.ops.block_until_ready()
     }
 }

src/cargo/util/config/mod.rs

@@ -315,19 +315,24 @@ impl Config {
         self.home_path.join("git")
     }
 
+    /// Gets the Cargo base directory for all registry information (`<cargo_home>/registry`).
+    pub fn registry_base_path(&self) -> Filesystem {
+        self.home_path.join("registry")
+    }
+
     /// Gets the Cargo registry index directory (`<cargo_home>/registry/index`).
     pub fn registry_index_path(&self) -> Filesystem {
-        self.home_path.join("registry").join("index")
+        self.registry_base_path().join("index")
     }
 
     /// Gets the Cargo registry cache directory (`<cargo_home>/registry/path`).
     pub fn registry_cache_path(&self) -> Filesystem {
-        self.home_path.join("registry").join("cache")
+        self.registry_base_path().join("cache")
     }
 
     /// Gets the Cargo registry source directory (`<cargo_home>/registry/src`).
     pub fn registry_source_path(&self) -> Filesystem {
-        self.home_path.join("registry").join("src")
+        self.registry_base_path().join("src")
     }
 
     /// Gets the default Cargo registry.

tests/testsuite/git.rs

@@ -2247,6 +2247,8 @@ fn add_a_git_dep() {
 
     p.cargo("build").run();
 
+    assert!(paths::home().join(".cargo/git/CACHEDIR.TAG").is_file());
+
     p.change_file(
         "a/Cargo.toml",
         &format!(
@@ -2580,6 +2582,7 @@ fn use_the_cli() {
 ";
 
     project.cargo("build -v").with_stderr(stderr).run();
+    assert!(paths::home().join(".cargo/git/CACHEDIR.TAG").is_file());
 }
 
 #[cargo_test]

tests/testsuite/registry.rs

@@ -93,6 +93,8 @@ fn simple(cargo: fn(&Project, &str) -> Execs) {
 
     cargo(&p, "clean").run();
 
+    assert!(paths::home().join(".cargo/registry/CACHEDIR.TAG").is_file());
+
     // Don't download a second time
     cargo(&p, "build")
         .with_stderr(
@@ -150,6 +152,8 @@ fn deps(cargo: fn(&Project, &str) -> Execs) {
 ",
         )
         .run();
+
+    assert!(paths::home().join(".cargo/registry/CACHEDIR.TAG").is_file());
 }
 
 #[cargo_test]
@@ -1231,6 +1235,13 @@ fn updating_a_dep(cargo: fn(&Project, &str) -> Execs) {
 ",
         )
         .run();
+    assert!(paths::home().join(".cargo/registry/CACHEDIR.TAG").is_file());
+
+    // Now delete the CACHEDIR.TAG file: this is the situation we'll be in after
+    // upgrading from a version of Cargo that doesn't mark this directory, to one that
+    // does. It should be recreated.
+    fs::remove_file(paths::home().join(".cargo/registry/CACHEDIR.TAG"))
+        .expect("remove CACHEDIR.TAG");
 
     p.change_file(
         "a/Cargo.toml",
@@ -1260,6 +1271,11 @@ fn updating_a_dep(cargo: fn(&Project, &str) -> Execs) {
 ",
         )
         .run();
+
+    assert!(
+        paths::home().join(".cargo/registry/CACHEDIR.TAG").is_file(),
+        "CACHEDIR.TAG recreated in existing registry"
+    );
 }
 
 #[cargo_test]