replace more std::unordered_* types by faster boost hash maps

src/libexpr-c/nix_api_expr.cc

@@ -16,7 +16,7 @@
 #include "nix_api_util_internal.h"
 
 #if NIX_USE_BOEHMGC
-#  include <mutex>
+#  include <boost/unordered/concurrent_flat_map.hpp>
 #endif
 
 /**
@@ -207,28 +207,20 @@ void nix_state_free(EvalState * state)
 }
 
 #if NIX_USE_BOEHMGC
-std::unordered_map<
+boost::concurrent_flat_map<
     const void *,
     unsigned int,
     std::hash<const void *>,
     std::equal_to<const void *>,
     traceable_allocator<std::pair<const void * const, unsigned int>>>
-    nix_refcounts;
-
-std::mutex nix_refcount_lock;
+    nix_refcounts{};
 
 nix_err nix_gc_incref(nix_c_context * context, const void * p)
 {
     if (context)
         context->last_err_code = NIX_OK;
     try {
-        std::scoped_lock lock(nix_refcount_lock);
-        auto f = nix_refcounts.find(p);
-        if (f != nix_refcounts.end()) {
-            f->second++;
-        } else {
-            nix_refcounts[p] = 1;
-        }
+        nix_refcounts.insert_or_visit({p, 1}, [](auto & kv) { kv.second++; });
     }
     NIXC_CATCH_ERRS
 }
@@ -239,12 +231,12 @@ nix_err nix_gc_decref(nix_c_context * context, const void * p)
     if (context)
         context->last_err_code = NIX_OK;
     try {
-        std::scoped_lock lock(nix_refcount_lock);
-        auto f = nix_refcounts.find(p);
-        if (f != nix_refcounts.end()) {
-            if (--f->second == 0)
-                nix_refcounts.erase(f);
-        } else
+        bool fail = true;
+        nix_refcounts.erase_if(p, [&](auto & kv) {
+            fail = false;
+            return !--kv.second;
+        });
+        if (fail)
             throw std::runtime_error("nix_gc_decref: object was not referenced");
     }
     NIXC_CATCH_ERRS

src/libexpr/primops.cc

@@ -4571,11 +4571,13 @@ struct RegexCache
         std::regex regex;
         /* No std::regex constructor overload from std::string_view, but can be constructed
            from a pointer + size or an iterator range. */
-        cache.try_emplace_and_cvisit(re,
-            /*s=*/re.data(), /*count=*/re.size(), std::regex::extended,
+        cache.try_emplace_and_cvisit(
+            re,
+            /*s=*/re.data(),
+            /*count=*/re.size(),
+            std::regex::extended,
             [&regex](const auto & kv) { regex = kv.second; },
-            [&regex](const auto & kv) { regex = kv.second; }
-        );
+            [&regex](const auto & kv) { regex = kv.second; });
         return regex;
     }
 };

src/libexpr/print.cc

@@ -1,5 +1,4 @@
 #include <limits>
-#include <unordered_set>
 #include <sstream>
 
 #include "nix/expr/print.hh"
@@ -10,6 +9,8 @@
 #include "nix/util/english.hh"
 #include "nix/expr/eval.hh"
 
+#include <boost/unordered/unordered_flat_set.hpp>
+
 namespace nix {
 
 void printElided(
@@ -81,7 +82,7 @@ std::ostream & printLiteralBool(std::ostream & str, bool boolean)
 // For example `or' doesn't need to be quoted.
 bool isReservedKeyword(const std::string_view str)
 {
-    static const std::unordered_set<std::string_view> reservedKeywords = {
+    static const boost::unordered_flat_set<std::string_view> reservedKeywords = {
         "if", "then", "else", "assert", "with", "let", "in", "rec", "inherit"};
     return reservedKeywords.contains(str);
 }

src/libfetchers/filtering-source-accessor.cc

@@ -1,5 +1,7 @@
 #include "nix/fetchers/filtering-source-accessor.hh"
 
+#include <boost/unordered/unordered_flat_set.hpp>
+
 namespace nix {
 
 std::optional<std::filesystem::path> FilteringSourceAccessor::getPhysicalPath(const CanonPath & path)
@@ -57,12 +59,12 @@ void FilteringSourceAccessor::checkAccess(const CanonPath & path)
 struct AllowListSourceAccessorImpl : AllowListSourceAccessor
 {
     std::set<CanonPath> allowedPrefixes;
-    std::unordered_set<CanonPath> allowedPaths;
+    boost::unordered_flat_set<CanonPath, std::hash<CanonPath>> allowedPaths;
 
     AllowListSourceAccessorImpl(
         ref<SourceAccessor> next,
         std::set<CanonPath> && allowedPrefixes,
-        std::unordered_set<CanonPath> && allowedPaths,
+        boost::unordered_flat_set<CanonPath, std::hash<CanonPath>> && allowedPaths,
         MakeNotAllowedError && makeNotAllowedError)
         : AllowListSourceAccessor(SourcePath(next), std::move(makeNotAllowedError))
         , allowedPrefixes(std::move(allowedPrefixes))
@@ -84,7 +86,7 @@ struct AllowListSourceAccessorImpl : AllowListSourceAccessor
 ref<AllowListSourceAccessor> AllowListSourceAccessor::create(
     ref<SourceAccessor> next,
     std::set<CanonPath> && allowedPrefixes,
-    std::unordered_set<CanonPath> && allowedPaths,
+    boost::unordered_flat_set<CanonPath, std::hash<CanonPath>> && allowedPaths,
     MakeNotAllowedError && makeNotAllowedError)
 {
     return make_ref<AllowListSourceAccessorImpl>(

src/libfetchers/git-utils.cc

@@ -30,8 +30,9 @@
 #include <git2/sys/mempack.h>
 #include <git2/tree.h>
 
+#include <boost/unordered/unordered_flat_map.hpp>
+#include <boost/unordered/unordered_flat_set.hpp>
 #include <iostream>
-#include <unordered_set>
 #include <queue>
 #include <regex>
 #include <span>
@@ -315,7 +316,7 @@ struct GitRepoImpl : GitRepo, std::enable_shared_from_this<GitRepoImpl>
 
     uint64_t getRevCount(const Hash & rev) override
     {
-        std::unordered_set<git_oid> done;
+        boost::unordered_flat_set<git_oid, std::hash<git_oid>> done;
         std::queue<Commit> todo;
 
         todo.push(peelObject<Commit>(lookupObject(*this, hashToOID(rev)).get(), GIT_OBJECT_COMMIT));
@@ -569,7 +570,7 @@ struct GitRepoImpl : GitRepo, std::enable_shared_from_this<GitRepoImpl>
     void verifyCommit(const Hash & rev, const std::vector<fetchers::PublicKey> & publicKeys) override
     {
         // Map of SSH key types to their internal OpenSSH representations
-        static const std::unordered_map<std::string_view, std::string_view> keyTypeMap = {
+        static const boost::unordered_flat_map<std::string_view, std::string_view> keyTypeMap = {
             {"ssh-dsa", "ssh-dsa"},
             {"ssh-ecdsa", "ssh-ecdsa"},
             {"ssh-ecdsa-sk", "sk-ecdsa-sha2-nistp256@openssh.com"},
@@ -816,7 +817,7 @@ struct GitSourceAccessor : SourceAccessor
         return toHash(*git_tree_entry_id(entry));
     }
 
-    std::unordered_map<CanonPath, TreeEntry> lookupCache;
+    boost::unordered_flat_map<CanonPath, TreeEntry, std::hash<CanonPath>> lookupCache;
 
     /* Recursively look up 'path' relative to the root. */
     git_tree_entry * lookup(State & state, const CanonPath & path)
@@ -1253,7 +1254,7 @@ GitRepoImpl::getAccessor(const WorkdirInfo & wd, bool exportIgnore, MakeNotAllow
                                            makeFSSourceAccessor(path),
                                            std::set<CanonPath>{wd.files},
                                            // Always allow access to the root, but not its children.
-                                           std::unordered_set<CanonPath>{CanonPath::root},
+                                           boost::unordered_flat_set<CanonPath, std::hash<CanonPath>>{CanonPath::root},
                                            std::move(makeNotAllowedError))
                                            .cast<SourceAccessor>();
     if (exportIgnore)

src/libfetchers/include/nix/fetchers/filtering-source-accessor.hh

@@ -2,7 +2,7 @@
 
 #include "nix/util/source-path.hh"
 
-#include <unordered_set>
+#include <boost/unordered/unordered_flat_set_fwd.hpp>
 
 namespace nix {
 
@@ -72,7 +72,7 @@ struct AllowListSourceAccessor : public FilteringSourceAccessor
     static ref<AllowListSourceAccessor> create(
         ref<SourceAccessor> next,
         std::set<CanonPath> && allowedPrefixes,
-        std::unordered_set<CanonPath> && allowedPaths,
+        boost::unordered_flat_set<CanonPath, std::hash<CanonPath>> && allowedPaths,
         MakeNotAllowedError && makeNotAllowedError);
 
     using FilteringSourceAccessor::FilteringSourceAccessor;

src/libflake/lockfile.cc

@@ -1,5 +1,3 @@
-#include <unordered_set>
-
 #include "nix/fetchers/fetch-settings.hh"
 #include "nix/flake/settings.hh"
 #include "nix/flake/lockfile.hh"
@@ -9,6 +7,7 @@
 #include <algorithm>
 #include <iomanip>
 
+#include <boost/unordered/unordered_flat_set.hpp>
 #include <iterator>
 #include <nlohmann/json.hpp>
 
@@ -162,7 +161,7 @@ std::pair<nlohmann::json, LockFile::KeyMap> LockFile::toJSON() const
 {
     nlohmann::json nodes;
     KeyMap nodeKeys;
-    std::unordered_set<std::string> keys;
+    boost::unordered_flat_set<std::string> keys;
 
     std::function<std::string(const std::string & key, ref<const Node> node)> dumpNode;
 

src/libstore/derivations.cc

@@ -11,6 +11,7 @@
 #include "nix/util/json-utils.hh"
 
 #include <boost/container/small_vector.hpp>
+#include <boost/unordered/concurrent_flat_map.hpp>
 #include <nlohmann/json.hpp>
 
 namespace nix {
@@ -834,7 +835,7 @@ DerivationType BasicDerivation::type() const
     throw Error("can't mix derivation output types");
 }
 
-Sync<DrvHashes> drvHashes;
+DrvHashes drvHashes;
 
 /* pathDerivationModulo and hashDerivationModulo are mutually recursive
  */
@@ -844,16 +845,13 @@ Sync<DrvHashes> drvHashes;
  */
 static const DrvHash pathDerivationModulo(Store & store, const StorePath & drvPath)
 {
-    {
-        auto hashes = drvHashes.lock();
-        auto h = hashes->find(drvPath);
-        if (h != hashes->end()) {
-            return h->second;
-        }
+    std::optional<DrvHash> hash;
+    if (drvHashes.cvisit(drvPath, [&hash](const auto & kv) { hash.emplace(kv.second); })) {
+        return *hash;
     }
     auto h = hashDerivationModulo(store, store.readInvalidDerivation(drvPath), false);
     // Cache it
-    drvHashes.lock()->insert_or_assign(drvPath, h);
+    drvHashes.insert_or_assign(drvPath, h);
     return h;
 }
 

src/libstore/gc.cc

@@ -1,6 +1,7 @@
 #include "nix/store/derivations.hh"
 #include "nix/store/globals.hh"
 #include "nix/store/local-store.hh"
+#include "nix/store/path.hh"
 #include "nix/util/finally.hh"
 #include "nix/util/unix-domain-socket.hh"
 #include "nix/util/signals.hh"
@@ -13,14 +14,10 @@
 #  include "nix/util/processes.hh"
 #endif
 
+#include <boost/unordered/unordered_flat_map.hpp>
+#include <boost/unordered/unordered_flat_set.hpp>
 #include <boost/regex.hpp>
-
-#include <functional>
 #include <queue>
-#include <algorithm>
-#include <random>
-
-#include <climits>
 #include <errno.h>
 #include <fcntl.h>
 #include <sys/stat.h>
@@ -314,7 +311,7 @@ Roots LocalStore::findRoots(bool censor)
 /**
  * Key is a mere string because cannot has path with macOS's libc++
  */
-typedef std::unordered_map<std::string, std::unordered_set<std::string>> UncheckedRoots;
+typedef boost::unordered_flat_map<std::string, boost::unordered_flat_set<std::string>> UncheckedRoots;
 
 static void readProcLink(const std::filesystem::path & file, UncheckedRoots & roots)
 {
@@ -463,13 +460,13 @@ void LocalStore::collectGarbage(const GCOptions & options, GCResults & results)
     bool gcKeepOutputs = settings.gcKeepOutputs;
     bool gcKeepDerivations = settings.gcKeepDerivations;
 
-    std::unordered_set<StorePath> roots, dead, alive;
+    boost::unordered_flat_set<StorePath, std::hash<StorePath>> roots, dead, alive;
 
     struct Shared
     {
         // The temp roots only store the hash part to make it easier to
         // ignore suffixes like '.lock', '.chroot' and '.check'.
-        std::unordered_set<std::string> tempRoots;
+        boost::unordered_flat_set<std::string> tempRoots;
 
         // Hash part of the store path currently being deleted, if
         // any.
@@ -672,7 +669,7 @@ void LocalStore::collectGarbage(const GCOptions & options, GCResults & results)
         }
     };
 
-    std::unordered_map<StorePath, StorePathSet> referrersCache;
+    boost::unordered_flat_map<StorePath, StorePathSet, std::hash<StorePath>> referrersCache;
 
     /* Helper function that visits all paths reachable from `start`
        via the referrers edges and optionally derivers and derivation

src/libstore/include/nix/store/derivations.hh

@@ -11,7 +11,7 @@
 #include "nix/util/sync.hh"
 #include "nix/util/variant-wrapper.hh"
 
-#include <map>
+#include <boost/unordered/concurrent_flat_map_fwd.hpp>
 #include <variant>
 
 namespace nix {
@@ -507,13 +507,23 @@ DrvHash hashDerivationModulo(Store & store, const Derivation & drv, bool maskOut
  */
 std::map<std::string, Hash> staticOutputHashes(Store & store, const Derivation & drv);
 
+struct DrvHashFct
+{
+    using is_avalanching = std::true_type;
+
+    std::size_t operator()(const StorePath & path) const noexcept
+    {
+        return std::hash<std::string_view>{}(path.to_string());
+    }
+};
+
 /**
  * Memoisation of hashDerivationModulo().
  */
-typedef std::map<StorePath, DrvHash> DrvHashes;
+typedef boost::concurrent_flat_map<StorePath, DrvHash, DrvHashFct> DrvHashes;
 
 // FIXME: global, though at least thread-safe.
-extern Sync<DrvHashes> drvHashes;
+extern DrvHashes drvHashes;
 
 struct Source;
 struct Sink;

src/libstore/include/nix/store/gc-store.hh

@@ -1,13 +1,13 @@
 #pragma once
 ///@file
 
-#include <unordered_set>
-
 #include "nix/store/store-api.hh"
+#include <boost/unordered/unordered_flat_map.hpp>
+#include <boost/unordered/unordered_flat_set.hpp>
 
 namespace nix {
 
-typedef std::unordered_map<StorePath, std::unordered_set<std::string>> Roots;
+typedef boost::unordered_flat_map<StorePath, boost::unordered_flat_set<std::string>, std::hash<StorePath>> Roots;
 
 struct GCOptions
 {

src/libstore/include/nix/store/local-store.hh

@@ -11,7 +11,7 @@
 #include <chrono>
 #include <future>
 #include <string>
-#include <unordered_set>
+#include <boost/unordered/unordered_flat_set.hpp>
 
 namespace nix {
 
@@ -442,7 +442,7 @@ private:
 
     std::pair<std::filesystem::path, AutoCloseFD> createTempDirInStore();
 
-    typedef std::unordered_set<ino_t> InodeHash;
+    typedef boost::unordered_flat_set<ino_t> InodeHash;
 
     InodeHash loadInodeHash();
     Strings readDirectoryIgnoringInodes(const Path & path, const InodeHash & inodeHash);

src/libstore/misc.cc

@@ -1,5 +1,3 @@
-#include <unordered_set>
-
 #include "nix/store/derivations.hh"
 #include "nix/store/parsed-derivations.hh"
 #include "nix/store/derivation-options.hh"
@@ -13,6 +11,8 @@
 #include "nix/store/filetransfer.hh"
 #include "nix/util/strings.hh"
 
+#include <boost/unordered/unordered_flat_set.hpp>
+
 namespace nix {
 
 void Store::computeFSClosure(
@@ -106,7 +106,7 @@ MissingPaths Store::queryMissing(const std::vector<DerivedPath> & targets)
 
     struct State
     {
-        std::unordered_set<std::string> done;
+        boost::unordered_flat_set<std::string> done;
         MissingPaths res;
     };
 

src/libutil/include/nix/util/canon-path.hh

@@ -258,7 +258,7 @@ public:
      */
     std::string makeRelative(const CanonPath & path) const;
 
-    friend class std::hash<CanonPath>;
+    friend struct std::hash<CanonPath>;
 };
 
 std::ostream & operator<<(std::ostream & stream, const CanonPath & path);
@@ -268,6 +268,8 @@ std::ostream & operator<<(std::ostream & stream, const CanonPath & path);
 template<>
 struct std::hash<nix::CanonPath>
 {
+    using is_avalanching = std::true_type;
+
     std::size_t operator()(const nix::CanonPath & s) const noexcept
     {
         return std::hash<std::string>{}(s.path);

src/libutil/linux/cgroup.cc

@@ -4,10 +4,10 @@
 #include "nix/util/file-system.hh"
 #include "nix/util/finally.hh"
 
+#include <boost/unordered/unordered_flat_set.hpp>
 #include <chrono>
 #include <cmath>
 #include <regex>
-#include <unordered_set>
 #include <thread>
 
 #include <dirent.h>
@@ -76,7 +76,7 @@ static CgroupStats destroyCgroup(const std::filesystem::path & cgroup, bool retu
 
     int round = 1;
 
-    std::unordered_set<pid_t> pidsShown;
+    boost::unordered_flat_set<pid_t> pidsShown;
 
     while (true) {
         auto pids = tokenizeString<std::vector<std::string>>(readFile(procsFile));

src/libutil/posix-source-accessor.cc

@@ -104,7 +104,7 @@ std::optional<struct stat> PosixSourceAccessor::cachedLstat(const CanonPath & pa
 
     if (cache.size() >= 16384)
         cache.clear();
-    cache.emplace(absPath, st);
+    cache.emplace(std::move(absPath), st);
 
     return st;
 }

src/nix/env.cc

@@ -71,7 +71,7 @@ struct CmdShell : InstallablesCommand, MixEnvironment
         auto outPaths =
             Installable::toStorePaths(getEvalStore(), store, Realise::Outputs, OperateOn::Output, installables);
 
-        std::unordered_set<StorePath> done;
+        boost::unordered_flat_set<StorePath, std::hash<StorePath>> done;
         std::queue<StorePath> todo;
         for (auto & path : outPaths)
             todo.push(path);