about summary refs log tree commit homepage
diff options
context:
space:
mode:
authorEric Wong <normalperson@yhbt.net>2010-02-12 23:59:55 -0800
committerEric Wong <normalperson@yhbt.net>2010-02-13 00:34:50 -0800
commitafdc6fce5f7e34a5c07f24204625cd466e93e5ac (patch)
tree1dfb1c5d5e0dccae3480609147b6e7dfb1226b48
parent615cede23579906b486c7a18bdeb22d04864a6ea (diff)
downloadunicorn-afdc6fce5f7e34a5c07f24204625cd466e93e5ac.tar.gz
First off, this memory leak DOES NOT affect Unicorn itself.
Unicorn allocates the HttpParser once and always reuses it
in every sequential request.

This leak affects applications which repeatedly allocate a new
HTTP parser.  Thus this bug affects _all_ deployments of
Rainbows! and Zbatery.  These servers allocate a new parser for
every client connection.

I misread the Data_Make_Struct/Data_Wrap_Struct documentation
and ended up passing NULL as the "free" argument instead of -1,
causing the memory to never be freed.

From README.EXT in the MRI source which I misread:
> The free argument is the function to free the pointer
> allocation.  If this is -1, the pointer will be just freed.
> The functions mark and free will be called from garbage
> collector.
-rw-r--r--ext/unicorn_http/unicorn_http.rl2
-rw-r--r--test/unit/test_http_parser.rb18
2 files changed, 18 insertions, 2 deletions
diff --git a/ext/unicorn_http/unicorn_http.rl b/ext/unicorn_http/unicorn_http.rl
index 6232e2c..a95224c 100644
--- a/ext/unicorn_http/unicorn_http.rl
+++ b/ext/unicorn_http/unicorn_http.rl
@@ -439,7 +439,7 @@ static void hp_mark(void *ptr)
 static VALUE HttpParser_alloc(VALUE klass)
 {
   struct http_parser *hp;
-  return Data_Make_Struct(klass, struct http_parser, hp_mark, NULL, hp);
+  return Data_Make_Struct(klass, struct http_parser, hp_mark, -1, hp);
 }
 
 
diff --git a/test/unit/test_http_parser.rb b/test/unit/test_http_parser.rb
index 0443b46..5b0ca9f 100644
--- a/test/unit/test_http_parser.rb
+++ b/test/unit/test_http_parser.rb
@@ -527,5 +527,21 @@ class HttpParserTest < Test::Unit::TestCase
     end
 
   end
-end
 
+  # so we don't  care about the portability of this test
+  # if it doesn't leak on Linux, it won't leak anywhere else
+  # unless your C compiler or platform is otherwise broken
+  LINUX_PROC_PID_STATUS = "/proc/self/status"
+  def test_memory_leak
+    match_rss = /^VmRSS:\s+(\d+)/
+    if File.read(LINUX_PROC_PID_STATUS) =~ match_rss
+      before = $1.to_i
+      1000000.times { Unicorn::HttpParser.new }
+      File.read(LINUX_PROC_PID_STATUS) =~ match_rss
+      after = $1.to_i
+      diff = after - before
+      assert(diff < 10000, "memory grew more than 10M: #{diff}")
+    end
+  end if RUBY_PLATFORM =~ /linux/ && test(?r, LINUX_PROC_PID_STATUS)
+
+end