about summary refs log tree commit homepage
path: root/lib/unicorn/configurator.rb
diff options
context:
space:
mode:
authorJeremy Evans <code@jeremyevans.net>2017-03-08 10:19:02 -0800
committerEric Wong <e@80x24.org>2017-03-10 20:27:41 +0000
commitea1a4360d66a833d75fbd887388d8cd4fe4ae299 (patch)
treec2f631143f2acb5fcfd092097ee6db42ad55f762 /lib/unicorn/configurator.rb
parent59e39cbc68c49623949b3e5c2d7e113e96e90a27 (diff)
downloadunicorn-ea1a4360d66a833d75fbd887388d8cd4fe4ae299.tar.gz
The worker_exec configuration option makes all worker processes
exec after forking.  This initializes the worker processes with
separate memory layouts, defeating address space discovery
attacks on operating systems supporting address space layout
randomization, such as Linux, MacOS X, NetBSD, OpenBSD, and
Solaris.

Support for execing workers is very similar to support for reexecing
the master process.  The main difference is the worker's to_i and
master pipes also need to be inherited after worker exec just as the
listening sockets need to be inherited after reexec.

Because execing working is similar to reexecing the master, this
extracts a couple of methods from reexec (listener_sockets and
close_sockets_on_exec), so they can be reused in worker_spawn.
Diffstat (limited to 'lib/unicorn/configurator.rb')
-rw-r--r--lib/unicorn/configurator.rb10
1 files changed, 10 insertions, 0 deletions
diff --git a/lib/unicorn/configurator.rb b/lib/unicorn/configurator.rb
index 7ed5ffa..f69f220 100644
--- a/lib/unicorn/configurator.rb
+++ b/lib/unicorn/configurator.rb
@@ -53,6 +53,7 @@ class Unicorn::Configurator
         server.logger.info("worker=#{worker.nr} ready")
       },
     :pid => nil,
+    :worker_exec => false,
     :preload_app => false,
     :check_client_connection => false,
     :rewindable_input => true, # for Rack 2.x: (Rack::VERSION[0] <= 1),
@@ -239,6 +240,15 @@ class Unicorn::Configurator
     set[:timeout] = seconds > max ? max : seconds
   end
 
+  # Whether to exec in each worker process after forking.  This changes the
+  # memory layout of each worker process, which is a security feature designed
+  # to defeat possible address space discovery attacks.  Note that using
+  # worker_exec only makes sense if you are not preloading the application,
+  # and will result in higher memory usage.
+  def worker_exec(bool)
+    set_bool(:worker_exec, bool)
+  end
+
   # sets the current number of worker_processes to +nr+.  Each worker
   # process will serve exactly one client at a time.  You can
   # increment or decrement this value at runtime by sending SIGTTIN