[PATCH] Zlib::GzipWriter#write segfaults when given large (2^23 byte) strings

0001-Example-for-compression-of-large-strings.-1178.patch

From 28fa610732b8621bbac1019be00622ac77caee24 Mon Sep 17 00:00:00 2001
From: David Yip <[email protected]>
Date: Fri, 30 Sep 2011 21:34:24 -0500
Subject: [PATCH 1/2] Example for compression of large strings.  #1178.

The expected output was generated by the following Ruby program run in
Ruby 1.8.7-p352:

    require 'stringio'
    require 'zlib'

    io = StringIO.new
    input = '.' * (2 ** 23)

    Zlib::GzipWriter.wrap(io) do |gzio|
      gzio.write(input)
    end

    print io.string

It was run as

    ruby gen.rb | hexdump -C

which resulted in the following output:

    00000000  1f 8b 08 00 b9 a8 8a 4e  00 03 ec c1 81 00 00 00  |.......N........|
    00000010  00 80 20 a3 fd f1 16 a9  0a 00 00 00 00 00 00 00  |.. .............|
    00000020  00 00 00 00 00 00 00 00  00 00 00 00 00 00 00 00  |................|
    *
    00001010  00 00 00 00 00 00 00 80  db 83 03 02 00 00 00 00  |................|
    00001020  21 ff 5f 37 24 00 00 00  00 00 00 00 00 00 00 00  |!._7$...........|
    00001030  00 00 00 00 00 00 00 00  00 00 00 00 00 00 00 00  |................|
    *
    00001fe0  00 00 00 00 00 00 9c 05  c2 9f 5a 5a 00 00 80 00  |..........ZZ....|
    00001ff0

This output was then transcribed into a concatenation of Ruby strings.
---
 spec/ruby/library/zlib/gzipwriter/write_spec.rb |   27 +++++++++++++++++++---
 1 files changed, 23 insertions(+), 4 deletions(-)

diff --git a/spec/ruby/library/zlib/gzipwriter/write_spec.rb b/spec/ruby/library/zlib/gzipwriter/write_spec.rb
index 21c6f4e..0e6fe2c 100644
--- a/spec/ruby/library/zlib/gzipwriter/write_spec.rb
+++ b/spec/ruby/library/zlib/gzipwriter/write_spec.rb
@@ -6,19 +6,38 @@ require 'zlib'
 describe "GzipWriter#write" do
 
   before :each do
-    @data = '12345abcde'
-    @zip = "\037\213\b\000,\334\321G\000\00334261MLJNI\005\000\235\005\000$\n\000\000\000"
     @io = StringIO.new ""
   end
 
   it "writes some compressed data" do
+    input = '12345abcde'
+    output = "\037\213\b\000,\334\321G\000\00334261MLJNI\005\000\235\005\000$\n\000\000\000"
+
     Zlib::GzipWriter.wrap @io do |gzio|
-      gzio.write @data
+      gzio.write input
     end
 
     # skip gzip header for now
-    @io.string[10..-1].should == @zip[10..-1]
+    @io.string[10..-1].should == output[10..-1]
   end
 
+  it "handles inputs of 2^23 bytes" do
+    input = '.' * (2 ** 23)
+    output = "\x1F\x8B\x08\x00\x5D\xA3\x8A\x4E" + "\x00\x03\xEC\xC1\x81\x00\x00\x00" +
+             "\x00\x80\x20\xA3\xFD\xF1\x16\xA9" + "\x0A\x00\x00\x00\x00\x00\x00\x00" +
+             "\x00\x00\x00\x00\x00\x00\x00\x00" + "\x00\x00\x00\x00\x00\x00\x00\x00" +
+             ("\x00" * 0xFE0) +
+             "\x00\x00\x00\x00\x00\x00\x00\x80" + "\xDB\x83\x03\x02\x00\x00\x00\x00" +
+             "\x21\xFF\x5F\x37\x24\x00\x00\x00" + "\x00\x00\x00\x00\x00\x00\x00\x00" +
+             "\x00\x00\x00\x00\x00\x00\x00\x00" + "\x00\x00\x00\x00\x00\x00\x00\x00" +
+             ("\x00" * 0xFA0) +
+             "\x00\x00\x00\x00\x00\x00\x9C\x05" + "\xC2\x9F\x5A\x5A\x00\x00\x80\x00"
+
+    Zlib::GzipWriter.wrap @io do |gzio|
+      gzio.write input
+    end
+
+    @io.string[10..-1].should == output[10..-1]
+  end
 end
 
-- 
1.7.6.3

0002-Calculate-a-checksum-CHUNK-bytes-at-a-time.-1178.patch

From f0fd2c4c2175ea8769e5d79c20a632825dd6a73a Mon Sep 17 00:00:00 2001
From: David Yip <[email protected]>
Date: Tue, 4 Oct 2011 01:07:43 -0500
Subject: [PATCH 2/2] Calculate a checksum CHUNK bytes at a time.  #1178.

---
 lib/zlib.rb.ffi |   12 +++++++++---
 1 files changed, 9 insertions(+), 3 deletions(-)

diff --git a/lib/zlib.rb.ffi b/lib/zlib.rb.ffi
index 867816e..73ac7d6 100644
--- a/lib/zlib.rb.ffi
+++ b/lib/zlib.rb.ffi
@@ -1,3 +1,5 @@
+require 'stringio'
+
 ##
 # The Zlib module contains several classes for compressing and decompressing
 # streams, and for working with "gzip" files.
@@ -1305,11 +1307,15 @@ module Zlib
       data = String data
 
       if data.length > 0 or sync? then
-        @crc = Zlib.crc32_c @crc, data, data.length
+        io = StringIO.new(data)
+
+        while chunk = io.read(CHUNK)
+          @crc = Zlib.crc32_c @crc, chunk, chunk.length
 
-        flush = sync? ? Zlib::SYNC_FLUSH : Zlib::NO_FLUSH
+          flush = sync? ? Zlib::SYNC_FLUSH : Zlib::NO_FLUSH
 
-        @zstream.run data, flush
+          @zstream.run chunk, flush
+        end
       end
 
       write_raw
-- 
1.7.6.3