Merge pull request #14040 from NixOS/import-thunk

 Ensure that files are parsed/evaluated only once (2nd attempt)

Author: roberth

src/libexpr/eval.cc

@@ -38,6 +38,7 @@
 
 #include <nlohmann/json.hpp>
 #include <boost/container/small_vector.hpp>
+#include <boost/unordered/concurrent_flat_map.hpp>
 
 #include "nix/util/strings-inline.hh"
 
@@ -264,6 +265,9 @@ EvalState::EvalState(
     , debugRepl(nullptr)
     , debugStop(false)
     , trylevel(0)
+    , srcToStore(make_ref<decltype(srcToStore)::element_type>())
+    , importResolutionCache(make_ref<decltype(importResolutionCache)::element_type>())
+    , fileEvalCache(make_ref<decltype(fileEvalCache)::element_type>())
     , regexCache(makeRegexCache())
 #if NIX_USE_BOEHMGC
     , valueAllocCache(std::allocate_shared<void *>(traceable_allocator<void *>(), nullptr))
@@ -1022,63 +1026,90 @@ Value * ExprPath::maybeThunk(EvalState & state, Env & env)
     return &v;
 }
 
-void EvalState::evalFile(const SourcePath & path, Value & v, bool mustBeTrivial)
+/**
+ * A helper `Expr` class to lets us parse and evaluate Nix expressions
+ * from a thunk, ensuring that every file is parsed/evaluated only
+ * once (via the thunk stored in `EvalState::fileEvalCache`).
+ */
+struct ExprParseFile : Expr, gc
 {
-    FileEvalCache::iterator i;
-    if ((i = fileEvalCache.find(path)) != fileEvalCache.end()) {
-        v = i->second;
-        return;
-    }
+    // FIXME: make this a reference (see below).
+    SourcePath path;
+    bool mustBeTrivial;
 
-    auto resolvedPath = resolveExprPath(path);
-    if ((i = fileEvalCache.find(resolvedPath)) != fileEvalCache.end()) {
-        v = i->second;
-        return;
+    ExprParseFile(SourcePath & path, bool mustBeTrivial)
+        : path(path)
+        , mustBeTrivial(mustBeTrivial)
+    {
     }
 
-    printTalkative("evaluating file '%1%'", resolvedPath);
-    Expr * e = nullptr;
+    void eval(EvalState & state, Env & env, Value & v) override
+    {
+        printTalkative("evaluating file '%s'", path);
+
+        auto e = state.parseExprFromFile(path);
 
-    auto j = fileParseCache.find(resolvedPath);
-    if (j != fileParseCache.end())
-        e = j->second;
+        try {
+            auto dts =
+                state.debugRepl
+                    ? makeDebugTraceStacker(
+                          state, *e, state.baseEnv, e->getPos(), "while evaluating the file '%s':", path.to_string())
+                    : nullptr;
+
+            // Enforce that 'flake.nix' is a direct attrset, not a
+            // computation.
+            if (mustBeTrivial && !(dynamic_cast<ExprAttrs *>(e)))
+                state.error<EvalError>("file '%s' must be an attribute set", path).debugThrow();
+
+            state.eval(e, v);
+        } catch (Error & e) {
+            state.addErrorTrace(e, "while evaluating the file '%s':", path.to_string());
+            throw;
+        }
+    }
+};
 
-    if (!e)
-        e = parseExprFromFile(resolvedPath);
+void EvalState::evalFile(const SourcePath & path, Value & v, bool mustBeTrivial)
+{
+    auto resolvedPath = getConcurrent(*importResolutionCache, path);
 
-    fileParseCache.emplace(resolvedPath, e);
+    if (!resolvedPath) {
+        resolvedPath = resolveExprPath(path);
+        importResolutionCache->emplace(path, *resolvedPath);
+    }
 
-    try {
-        auto dts = debugRepl ? makeDebugTraceStacker(
-                                   *this,
-                                   *e,
-                                   this->baseEnv,
-                                   e->getPos(),
-                                   "while evaluating the file '%1%':",
-                                   resolvedPath.to_string())
-                             : nullptr;
-
-        // Enforce that 'flake.nix' is a direct attrset, not a
-        // computation.
-        if (mustBeTrivial && !(dynamic_cast<ExprAttrs *>(e)))
-            error<EvalError>("file '%s' must be an attribute set", path).debugThrow();
-        eval(e, v);
-    } catch (Error & e) {
-        addErrorTrace(e, "while evaluating the file '%1%':", resolvedPath.to_string());
-        throw;
+    if (auto v2 = getConcurrent(*fileEvalCache, *resolvedPath)) {
+        forceValue(**v2, noPos);
+        v = **v2;
+        return;
     }
 
-    fileEvalCache.emplace(resolvedPath, v);
-    if (path != resolvedPath)
-        fileEvalCache.emplace(path, v);
+    Value * vExpr;
+    // FIXME: put ExprParseFile on the stack instead of the heap once
+    // https://github.com/NixOS/nix/pull/13930 is merged. That will ensure
+    // the post-condition that `expr` is unreachable after
+    // `forceValue()` returns.
+    auto expr = new ExprParseFile{*resolvedPath, mustBeTrivial};
+
+    fileEvalCache->try_emplace_and_cvisit(
+        *resolvedPath,
+        nullptr,
+        [&](auto & i) {
+            vExpr = allocValue();
+            vExpr->mkThunk(&baseEnv, expr);
+            i.second = vExpr;
+        },
+        [&](auto & i) { vExpr = i.second; });
+
+    forceValue(*vExpr, noPos);
+
+    v = *vExpr;
 }
 
 void EvalState::resetFileCache()
 {
-    fileEvalCache.clear();
-    fileEvalCache.rehash(0);
-    fileParseCache.clear();
-    fileParseCache.rehash(0);
+    importResolutionCache->clear();
+    fileEvalCache->clear();
     inputCache->clear();
 }
 
@@ -2397,24 +2428,26 @@ StorePath EvalState::copyPathToStore(NixStringContext & context, const SourcePat
     if (nix::isDerivation(path.path.abs()))
         error<EvalError>("file names are not allowed to end in '%1%'", drvExtension).debugThrow();
 
-    std::optional<StorePath> dstPath;
-    if (!srcToStore.cvisit(path, [&dstPath](const auto & kv) { dstPath.emplace(kv.second); })) {
-        dstPath.emplace(fetchToStore(
+    auto dstPathCached = getConcurrent(*srcToStore, path);
+
+    auto dstPath = dstPathCached ? *dstPathCached : [&]() {
+        auto dstPath = fetchToStore(
             fetchSettings,
             *store,
             path.resolveSymlinks(SymlinkResolution::Ancestors),
             settings.readOnlyMode ? FetchMode::DryRun : FetchMode::Copy,
             path.baseName(),
             ContentAddressMethod::Raw::NixArchive,
             nullptr,
-            repair));
-        allowPath(*dstPath);
-        srcToStore.try_emplace(path, *dstPath);
-        printMsg(lvlChatty, "copied source '%1%' -> '%2%'", path, store->printStorePath(*dstPath));
-    }
+            repair);
+        allowPath(dstPath);
+        srcToStore->try_emplace(path, dstPath);
+        printMsg(lvlChatty, "copied source '%1%' -> '%2%'", path, store->printStorePath(dstPath));
+        return dstPath;
+    }();
 
-    context.insert(NixStringContextElem::Opaque{.path = *dstPath});
-    return *dstPath;
+    context.insert(NixStringContextElem::Opaque{.path = dstPath});
+    return dstPath;
 }
 
 SourcePath EvalState::coerceToPath(const PosIdx pos, Value & v, NixStringContext & context, std::string_view errorCtx)

src/libexpr/include/nix/expr/eval.hh

@@ -20,8 +20,9 @@
 // For `NIX_USE_BOEHMGC`, and if that's set, `GC_THREADS`
 #include "nix/expr/config.hh"
 
-#include <boost/unordered/concurrent_flat_map.hpp>
 #include <boost/unordered/unordered_flat_map.hpp>
+#include <boost/unordered/concurrent_flat_map_fwd.hpp>
+
 #include <map>
 #include <optional>
 #include <functional>
@@ -403,37 +404,30 @@ private:
 
     /* Cache for calls to addToStore(); maps source paths to the store
        paths. */
-    boost::concurrent_flat_map<SourcePath, StorePath, std::hash<SourcePath>> srcToStore;
+    ref<boost::concurrent_flat_map<SourcePath, StorePath>> srcToStore;
 
     /**
-     * A cache from path names to parse trees.
+     * A cache that maps paths to "resolved" paths for importing Nix
+     * expressions, i.e. `/foo` to `/foo/default.nix`.
      */
-    typedef boost::unordered_flat_map<
-        SourcePath,
-        Expr *,
-        std::hash<SourcePath>,
-        std::equal_to<SourcePath>,
-        traceable_allocator<std::pair<const SourcePath, Expr *>>>
-        FileParseCache;
-    FileParseCache fileParseCache;
+    ref<boost::concurrent_flat_map<SourcePath, SourcePath>> importResolutionCache;
 
     /**
-     * A cache from path names to values.
+     * A cache from resolved paths to values.
      */
-    typedef boost::unordered_flat_map<
+    ref<boost::concurrent_flat_map<
         SourcePath,
-        Value,
+        Value *,
         std::hash<SourcePath>,
         std::equal_to<SourcePath>,
-        traceable_allocator<std::pair<const SourcePath, Value>>>
-        FileEvalCache;
-    FileEvalCache fileEvalCache;
+        traceable_allocator<std::pair<const SourcePath, Value *>>>>
+        fileEvalCache;
 
     /**
      * Associate source positions of certain AST nodes with their preceding doc comment, if they have one.
      * Grouped by file.
      */
-    boost::unordered_flat_map<SourcePath, DocCommentMap, std::hash<SourcePath>> positionToDocComment;
+    boost::unordered_flat_map<SourcePath, DocCommentMap> positionToDocComment;
 
     LookupPath lookupPath;
 

src/libfetchers/filtering-source-accessor.cc

@@ -59,12 +59,12 @@ void FilteringSourceAccessor::checkAccess(const CanonPath & path)
 struct AllowListSourceAccessorImpl : AllowListSourceAccessor
 {
     std::set<CanonPath> allowedPrefixes;
-    boost::unordered_flat_set<CanonPath, std::hash<CanonPath>> allowedPaths;
+    boost::unordered_flat_set<CanonPath> allowedPaths;
 
     AllowListSourceAccessorImpl(
         ref<SourceAccessor> next,
         std::set<CanonPath> && allowedPrefixes,
-        boost::unordered_flat_set<CanonPath, std::hash<CanonPath>> && allowedPaths,
+        boost::unordered_flat_set<CanonPath> && allowedPaths,
         MakeNotAllowedError && makeNotAllowedError)
         : AllowListSourceAccessor(SourcePath(next), std::move(makeNotAllowedError))
         , allowedPrefixes(std::move(allowedPrefixes))
@@ -86,7 +86,7 @@ struct AllowListSourceAccessorImpl : AllowListSourceAccessor
 ref<AllowListSourceAccessor> AllowListSourceAccessor::create(
     ref<SourceAccessor> next,
     std::set<CanonPath> && allowedPrefixes,
-    boost::unordered_flat_set<CanonPath, std::hash<CanonPath>> && allowedPaths,
+    boost::unordered_flat_set<CanonPath> && allowedPaths,
     MakeNotAllowedError && makeNotAllowedError)
 {
     return make_ref<AllowListSourceAccessorImpl>(

src/libfetchers/git-utils.cc

@@ -817,7 +817,7 @@ struct GitSourceAccessor : SourceAccessor
         return toHash(*git_tree_entry_id(entry));
     }
 
-    boost::unordered_flat_map<CanonPath, TreeEntry, std::hash<CanonPath>> lookupCache;
+    boost::unordered_flat_map<CanonPath, TreeEntry> lookupCache;
 
     /* Recursively look up 'path' relative to the root. */
     git_tree_entry * lookup(State & state, const CanonPath & path)
@@ -1254,7 +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.
-                                           boost::unordered_flat_set<CanonPath, std::hash<CanonPath>>{CanonPath::root},
+                                           boost::unordered_flat_set<CanonPath>{CanonPath::root},
                                            std::move(makeNotAllowedError))
                                            .cast<SourceAccessor>();
     if (exportIgnore)

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

@@ -72,7 +72,7 @@ struct AllowListSourceAccessor : public FilteringSourceAccessor
     static ref<AllowListSourceAccessor> create(
         ref<SourceAccessor> next,
         std::set<CanonPath> && allowedPrefixes,
-        boost::unordered_flat_set<CanonPath, std::hash<CanonPath>> && allowedPaths,
+        boost::unordered_flat_set<CanonPath> && allowedPaths,
         MakeNotAllowedError && makeNotAllowedError);
 
     using FilteringSourceAccessor::FilteringSourceAccessor;

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

@@ -8,6 +8,8 @@
 #include <set>
 #include <vector>
 
+#include <boost/container_hash/hash.hpp>
+
 namespace nix {
 
 /**
@@ -258,20 +260,26 @@ public:
      */
     std::string makeRelative(const CanonPath & path) const;
 
-    friend struct std::hash<CanonPath>;
+    friend std::size_t hash_value(const CanonPath &);
 };
 
 std::ostream & operator<<(std::ostream & stream, const CanonPath & path);
 
+inline std::size_t hash_value(const CanonPath & path)
+{
+    boost::hash<std::string_view> hasher;
+    return hasher(path.path);
+}
+
 } // namespace nix
 
 template<>
 struct std::hash<nix::CanonPath>
 {
     using is_avalanching = std::true_type;
 
-    std::size_t operator()(const nix::CanonPath & s) const noexcept
+    std::size_t operator()(const nix::CanonPath & path) const noexcept
     {
-        return std::hash<std::string>{}(s.path);
+        return nix::hash_value(path);
     }
 };

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

@@ -119,15 +119,23 @@ struct SourcePath
 
 std::ostream & operator<<(std::ostream & str, const SourcePath & path);
 
+inline std::size_t hash_value(const SourcePath & path)
+{
+    std::size_t hash = 0;
+    boost::hash_combine(hash, path.accessor->number);
+    boost::hash_combine(hash, path.path);
+    return hash;
+}
+
 } // namespace nix
 
 template<>
 struct std::hash<nix::SourcePath>
 {
+    using is_avalanching = std::true_type;
+
     std::size_t operator()(const nix::SourcePath & s) const noexcept
     {
-        std::size_t hash = 0;
-        hash_combine(hash, s.accessor->number, s.path);
-        return hash;
+        return nix::hash_value(s);
     }
 };

src/libutil/include/nix/util/util.hh

@@ -220,6 +220,17 @@ typename T::mapped_type * get(T & map, const K & key)
 template<class T, typename K>
 typename T::mapped_type * get(T && map, const K & key) = delete;
 
+/**
+ * Look up a value in a `boost::concurrent_flat_map`.
+ */
+template<class T>
+std::optional<typename T::mapped_type> getConcurrent(const T & map, const typename T::key_type & key)
+{
+    std::optional<typename T::mapped_type> res;
+    map.cvisit(key, [&](auto & x) { res = x.second; });
+    return res;
+}
+
 /**
  * Get a value for the specified key from an associate container, or a default value if the key isn't present.
  */

src/libutil/posix-source-accessor.cc

@@ -95,9 +95,7 @@ std::optional<struct stat> PosixSourceAccessor::cachedLstat(const CanonPath & pa
     // former is not hashable on libc++.
     Path absPath = makeAbsPath(path).string();
 
-    std::optional<Cache::mapped_type> res;
-    cache.cvisit(absPath, [&](auto & x) { res.emplace(x.second); });
-    if (res)
+    if (auto res = getConcurrent(cache, absPath))
         return *res;
 
     auto st = nix::maybeLstat(absPath.c_str());