Could boot performance be improved? Caching default gems perhaps?

[casperisfine]

Sorry if this was discussed elsewhere, I tried to search for previous discussions but given how generic the subject is I couldn't find anything.

Context

As many people know, Rubygems initialization is the biggest part of MRI's boot:

$ time ruby -e '1'

real	0m0.076s
user	0m0.051s
sys	0m0.021s
$ time ruby --disable-gems -e '1'

real	0m0.029s
user	0m0.011s
sys	0m0.015s

So (on my machine) "raw" MRI initialization is 29ms, and rubygems adds 47ms to it.

When profiling, we can see that a large part of that comes from loading default gems:

==================================
  Mode: cpu(10)
  Samples: 545 (52.98% miss rate)
  GC: 25 (4.59%)
==================================
     TOTAL    (pct)     SAMPLES    (pct)     FRAME
       237  (43.5%)         190  (34.9%)     Gem::Specification.load
       419  (76.9%)          40   (7.3%)     <require:/opt/rubies/2.8.0-dev/lib/ruby/2.8.0/rubygems.rb>
        66  (12.1%)          34   (6.2%)     <require:/opt/rubies/2.8.0-dev/lib/ruby/2.8.0/did_you_mean.rb>
        70  (12.8%)          33   (6.1%)     <require:/opt/rubies/2.8.0-dev/lib/ruby/2.8.0/rubygems/specification.rb>
        30   (5.5%)          30   (5.5%)     Gem::BasicSpecification#internal_init
        22   (4.0%)          22   (4.0%)     (sweeping)
        93  (17.1%)          21   (3.9%)     <module:Gem>
       106  (19.4%)          16   (2.9%)     Kernel#require
        15   (2.8%)          14   (2.6%)     Gem::Specification#files
        10   (1.8%)          10   (1.8%)     Gem::PathSupport#expand
        11   (2.0%)           9   (1.7%)     <require:/opt/rubies/2.8.0-dev/lib/ruby/2.8.0/did_you_mean/spell_checker.rb>
        10   (1.8%)           9   (1.7%)     <require:/opt/rubies/2.8.0-dev/lib/ruby/2.8.0/rubygems/user_interaction.rb>
        11   (2.0%)           9   (1.7%)     <require:/opt/rubies/2.8.0-dev/lib/ruby/2.8.0/did_you_mean/spell_checkers/name_error_checkers.rb>
         8   (1.5%)           8   (1.5%)     Dir.glob
        14   (2.6%)           7   (1.3%)     Gem::Specification#add_dependency_with_type
        17   (3.1%)           7   (1.3%)     <require:/opt/rubies/2.8.0-dev/lib/ruby/2.8.0/rubygems/specification_policy.rb>
        23   (4.2%)           5   (0.9%)     Gem.register_default_spec
        10   (1.8%)           4   (0.7%)     <class:Specification>
         5   (0.9%)           4   (0.7%)     <require:/opt/rubies/2.8.0-dev/lib/ruby/2.8.0/rubygems/core_ext/kernel_require.rb>
         5   (0.9%)           4   (0.7%)     <require:/opt/rubies/2.8.0-dev/lib/ruby/2.8.0/did_you_mean/spell_checkers/require_path_checker.rb>
         5   (0.9%)           4   (0.7%)     <require:/opt/rubies/2.8.0-dev/lib/ruby/2.8.0/rubygems/dependency.rb>

If I comment out Gem::Specification.load_defaults, the overhead is down to 26ms:

$ time ruby -e '1'

real	0m0.058s
user	0m0.034s
sys	0m0.020s
$ time ruby --disable-gems  -e '1'

real	0m0.032s
user	0m0.012s
sys	0m0.016s

And most of that is in each_spec([Gem.default_specifications_dir]).

Ideas

This makes me wonder what kind of assumptions we could make about default gems. Can we assume they are immutable? If so couldn't we cache all these specs in a single marshal dump file rather than evaluate all the default gems (55 on 2.8.0-dev) ?

I tried a very crude hack on my machine:

  def self.load_defaults
    cached_specs.each do |spec|
      # #load returns nil if the spec is bad, so we just ignore
      # it at this stage
      Gem.register_default_spec(spec)
    end
  end

  def self.cached_specs
    Marshal.load(File.read('/tmp/default_specs.marshal'))
  rescue SystemCallError
    specs = []
    each_spec([Gem.default_specifications_dir]) { |s| specs << s }
    File.write('/tmp/default_specs.marshal', Marshal.dump(specs))
    specs
  end

With the following result:

$ time ruby -e '1'

real	0m0.060s
user	0m0.036s
sys	0m0.020s

So 16ms faster, which is about 34% of rubygems load time.

Questions

So the main question here is what would it take to move this from a crude hack to an acceptable patch.

• Can we assume default gems are immutable?
• If not could we do a mtime comparison? on the directory or all individual specs?
• Is there an existing location where it would be acceptable to store that cache?

@deivid-rodriguez any thoughts on this?

[deivid-rodriguez]

So funny, I'm working on that exact path 🤣. I have this, which I believe is the same idea:

commit 0441af11bcf6d1418f52f0c0dd2c87c60d694e8b
Author: David Rodríguez <deivid.rodriguez@riseup.net>
Date:   Wed Jul 8 09:09:31 2020 +0200

    Load a marshaled version of the default specs map if available

diff --git a/lib/rubygems.rb b/lib/rubygems.rb
index 8bc0ef4b74..12936671db 100644
--- a/lib/rubygems.rb
+++ b/lib/rubygems.rb
@@ -1242,8 +1242,14 @@ def register_default_spec(spec)
     # Loads the default specifications. It should be called only once.
 
     def load_default_specs
-      Gem::Specification.each_spec([default_specifications_dir]) do |spec|
-        register_default_spec(spec)
+      if File.exist?(marshaled_default_spec_file)
+        @path_to_default_spec_map = Marshal.load read_binary(marshaled_default_spec_file)
+      else
+        Gem::Specification.each_spec([default_specifications_dir]) do |spec|
+          register_default_spec(spec)
+        end
+
+        File.open(marshaled_default_spec_file, "w") {|io| Marshal.dump @path_to_default_spec_map, io }
       end
     end
 
@@ -1319,6 +1325,9 @@ def default_gem_load_paths
       @default_gem_load_paths ||= $LOAD_PATH[load_path_insert_index..-1]
     end
 
+    def marshaled_default_spec_file
+      File.join(dir, "default_specs.#{marshal_version}")
+    end
   end
 
   ##

commit 1ee905c8eea7efa2257f1acdde3a27074c5ac3e5
Author: David Rodríguez <deivid.rodriguez@riseup.net>
Date:   Tue Jul 7 20:11:30 2020 +0200

    Move `Gem::Specification.load_defaults` to `Gem.load_default_specs`
    
    And deprecate the former. Makes sense I believe for upcoming work.

diff --git a/lib/rubygems.rb b/lib/rubygems.rb
index 1a70ee9d0a..8bc0ef4b74 100644
--- a/lib/rubygems.rb
+++ b/lib/rubygems.rb
@@ -1238,6 +1238,15 @@ def register_default_spec(spec)
       end
     end
 
+    ##
+    # Loads the default specifications. It should be called only once.
+
+    def load_default_specs
+      Gem::Specification.each_spec([default_specifications_dir]) do |spec|
+        register_default_spec(spec)
+      end
+    end
+
     ##
     # Find a Gem::Specification of default gem from +path+
 
@@ -1358,7 +1367,7 @@ def default_gem_load_paths
 
 ##
 # Loads the default specs.
-Gem::Specification.load_defaults
+Gem.load_default_specs
 
 require 'rubygems/core_ext/kernel_gem'
 require 'rubygems/core_ext/kernel_require'
diff --git a/lib/rubygems/specification.rb b/lib/rubygems/specification.rb
index e6ecc7649a..3c83494b8e 100644
--- a/lib/rubygems/specification.rb
+++ b/lib/rubygems/specification.rb
@@ -845,15 +845,14 @@ def self._resort!(specs) # :nodoc:
     end
   end
 
-  ##
-  # Loads the default specifications. It should be called only once.
+  def self.load_defaults # :nodoc:
+    Gem.load_default_specs
+  end
 
-  def self.load_defaults
-    each_spec([Gem.default_specifications_dir]) do |spec|
-      # #load returns nil if the spec is bad, so we just ignore
-      # it at this stage
-      Gem.register_default_spec(spec)
-    end
+  class << self
+    extend Gem::Deprecate
+
+    rubygems_deprecate :load_defaults, "Gem.load_default_specs"
   end

[casperisfine]

Oh! That's awesome. I'll ignore that area then, and focus on other ideas. Should I close this issue?

[deivid-rodriguez]

I'll definitely tag you on the PR, which will close this, so we call leave this opened 👍.

I believe we can assume default specs are immutable, yeah.

Regarding the cache location, we could use Gem.spec_cache_dir, and put ruby version specific files in there like 2.7.1/default_specs.4.8, or something like that. I think that could make things align quite nicely.

[eregon]

Refs: https://bugs.ruby-lang.org/issues/18568 and #4199