[PROF-12841] Fix profiler not identifying executables with gems they belong to

Steepfile

@@ -224,6 +224,7 @@ target :datadog do
   # References `RubyVM::YJIT`, which does not have type information.
   ignore 'lib/datadog/core/environment/yjit.rb'
 
+  library 'bundler'
   library 'pathname'
   library 'cgi'
   library 'logger', 'monitor'

lib/datadog/profiling/collectors/code_provenance.rb

@@ -22,14 +22,15 @@ def initialize(
           @libraries_by_path = {}
           @seen_files = Set.new
           @seen_libraries = Set.new
+          @executable_paths = [Gem.bindir, (Bundler.bin_path.to_s if defined?(Bundler))].uniq.compact.freeze
 
           record_library(
             Library.new(
               kind: "standard library",
               name: "stdlib",
               version: RUBY_VERSION,
               path: standard_library_path,
-              extra_path: ruby_native_filename,
+              extra_paths: [ruby_native_filename],
             )
           )
         end
@@ -51,7 +52,8 @@ def generate_json
           :libraries_by_name,
           :libraries_by_path,
           :seen_files,
-          :seen_libraries
+          :seen_libraries,
+          :executable_paths
 
         def record_library(library)
           libraries_by_name[library.name] = library
@@ -79,13 +81,20 @@ def record_loaded_specs(loaded_specs)
           loaded_specs.each do |spec|
             next if libraries_by_name.key?(spec.name)
 
+            extra_paths = [(spec.extension_dir if spec.extensions.any?)]
+            spec.executables&.each do |executable|
+              executable_paths.each do |path|
+                extra_paths << File.join(path, executable)
+              end
+            end
+
             record_library(
               Library.new(
                 kind: "library",
                 name: spec.name,
                 version: spec.version,
                 path: spec.gem_dir,
-                extra_path: (spec.extension_dir if spec.extensions.any?),
+                extra_paths: extra_paths,
               )
             )
             recorded_library = true
@@ -118,12 +127,12 @@ def record_loaded_files(loaded_files)
         class Library
           attr_reader :kind, :name, :version
 
-          def initialize(kind:, name:, version:, path:, extra_path: nil)
-            extra_path = nil if extra_path&.empty?
+          def initialize(kind:, name:, version:, path:, extra_paths:)
+            extra_paths = Array(extra_paths).compact.reject(&:empty?).map { |p| p.dup.freeze }
             @kind = kind.freeze
             @name = name.dup.freeze
             @version = version.to_s.dup.freeze
-            @paths = [path.dup.freeze, extra_path.dup.freeze].compact.freeze
+            @paths = [path.dup.freeze, *extra_paths].freeze
             freeze
           end
 

sig/datadog/profiling/collectors/code_provenance.rbs

@@ -12,6 +12,7 @@ module Datadog
         attr_reader libraries_by_path: Hash[String, Library]
         attr_reader seen_files: Set[String]
         attr_reader seen_libraries: Set[Library]
+        attr_reader executable_paths: Array[String]
 
         def record_library: (Library) -> void
         def sort_libraries_by_longest_path_first: () -> void
@@ -24,7 +25,7 @@ module Datadog
           @version: String
           @paths: Array[String]
 
-          def initialize: (kind: String, name: String, version: String, path: String, ?extra_path: String?) -> void
+          def initialize: (kind: String, name: String, version: String, path: String, extra_paths: Array[String?]) -> void
 
           def kind: -> String
           def name: -> String

spec/datadog/profiling/collectors/code_provenance_spec.rb

@@ -32,11 +32,16 @@
           kind: "library",
           name: "datadog",
           version: Datadog::VERSION::STRING,
-          paths: contain_exactly(start_with("/"), include("extensions").and(include(RUBY_PLATFORM))),
+          paths: contain_exactly(
+            start_with("/"),
+            include("extensions").and(include(RUBY_PLATFORM)),
+            "#{Gem.bindir}/ddprofrb",
+            "#{Bundler.bin_path}/ddprofrb",
+          ),
         },
         {
           kind: "library",
-          name: "rspec-core",
+          name: "rspec",
           version: start_with("3."), # This will one day need to be bumped for RSpec 4
           paths: contain_exactly(start_with("/")),
         },
@@ -59,6 +64,20 @@
       )
     end
 
+    it "includes the executables for gems with executables" do
+      refresh
+
+      expect(generate_result.find { |it| it[:name] == "rspec-core" }.fetch(:paths)).to contain_exactly(
+        Gem.loaded_specs.fetch("rspec-core").gem_dir,
+        "#{Gem.bindir}/rspec",
+        "#{Bundler.bin_path}/rspec",
+      )
+
+      # Sanity checks
+      expect(Gem.bindir).to start_with("/")
+      expect(Bundler.bin_path.to_s).to start_with("/")
+    end
+
     it "records the correct path for datadog" do
       refresh
 
@@ -78,13 +97,15 @@
             version: "not_loaded_version",
             gem_dir: "/not_loaded/",
             extensions: [],
+            executables: [],
           ),
           instance_double(
             Gem::Specification,
             name: "is_loaded",
             version: "is_loaded_version",
             gem_dir: "/is_loaded/",
             extensions: [],
+            executables: [],
           )
         ],
       )
@@ -131,13 +152,15 @@
               version: "1.2.3",
               gem_dir: "/dd-trace-rb",
               extensions: [],
+              executables: [],
             ),
             instance_double(
               Gem::Specification,
               name: "byebug",
               version: "4.5.6",
               gem_dir: "/dd-trace-rb/vendor/bundle/ruby/2.7.0/gems/byebug-11.1.3",
               extensions: [],
+              executables: [],
             )
           ],
         )
@@ -288,6 +311,7 @@
           kind: "library",
           version: "1.2.3",
           path: "/example/path/to/datadog/gem",
+          extra_paths: ["/example/path/to/datadog/extensions"],
         )
 
         serialized_without_to_json = YAML.dump(instance)
@@ -298,7 +322,7 @@
           "name" => "datadog",
           "kind" => "library",
           "version" => "1.2.3",
-          "paths" => ["/example/path/to/datadog/gem"],
+          "paths" => ["/example/path/to/datadog/gem", "/example/path/to/datadog/extensions"],
         )
       end
     end

vendor/rbs/bundler/0/bundler.rbs

@@ -0,0 +1,3 @@
+module Bundler
+  def self.bin_path: () -> Pathname
+end

vendor/rbs/gem/0/gem.rbs

@@ -16,4 +16,5 @@ class Gem::BasicSpecification
   def gem_dir: () -> String
   def extension_dir: () -> String
   def extensions: () -> Array[String]
+  def executables: () -> Array[String]?
 end