webrick: prevent response splitting and header injection
Original patch by tenderlove (with minor style adjustments).
* lib/webrick/httpresponse.rb (send_header): call check_header
(check_header): raise on embedded CRLF in header value
* test/webrick/test_httpresponse.rb
(test_prevent_response_splitting_headers): new test
* (test_prevent_response_splitting_cookie_headers): ditto
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_2@63022 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
webrick: use IO.copy_stream for multipart response
Use the new Proc response body feature to generate a multipart
range response dynamically. We use a flat array to minimize
object overhead as much as possible; as many ranges may fit
into an HTTP request header.
* lib/webrick/httpservlet/filehandler.rb (multipart_body): new method
(make_partial_content): use multipart_body
------------------------------------------------------------------------
r62960 | normal | 2018-03-28 17:06:23 +0900 (水, 28 3 2018) | 13 lines
webrick/httprequest: limit request headers size
We use the same 112 KB limit started (AFAIK) by Mongrel, Thin,
and Puma to prevent malicious users from using up all the memory
with a single request. This also limits the damage done by
excessive ranges in multipart Range: requests.
Due to the way we rely on IO#gets and the desire to keep
the code simple, the actual maximum header may be 4093 bytes
larger than 112 KB, but we're splitting hairs at that point.
* lib/webrick/httprequest.rb: define MAX_HEADER_LENGTH
(read_header): raise when headers exceed max length
------------------------------------------------------------------------
r62961 | normal | 2018-03-28 17:06:28 +0900 (水, 28 3 2018) | 9 lines
webrick/httpservlet/cgihandler: reduce memory use
WEBrick::HTTPRequest#body can be passed a block to process the
body in chunks. Use this feature to avoid building a giant
string in memory.
* lib/webrick/httpservlet/cgihandler.rb (do_GET):
avoid reading entire request body into memory
(do_POST is aliased to do_GET, so it handles bodies)
------------------------------------------------------------------------
r62962 | normal | 2018-03-28 17:06:34 +0900 (水, 28 3 2018) | 7 lines
webrick/httprequest: raise correct exception
"BadRequest" alone does not resolve correctly, it is in the
HTTPStatus namespace.
* lib/webrick/httprequest.rb (read_chunked): use correct exception
* test/webrick/test_httpserver.rb (test_eof_in_chunk): new test
------------------------------------------------------------------------
r62963 | normal | 2018-03-28 17:06:39 +0900 (水, 28 3 2018) | 9 lines
webrick/httprequest: use InputBufferSize for chunked requests
While WEBrick::HTTPRequest#body provides a Proc interface
for streaming large request bodies, clients must not force
the server to use an excessively large chunk size.
* lib/webrick/httprequest.rb (read_chunk_size): limit each
read and block.call to :InputBufferSize in config.
* test/webrick/test_httpserver.rb (test_big_chunks): new test
------------------------------------------------------------------------
r62964 | normal | 2018-03-28 17:06:44 +0900 (水, 28 3 2018) | 9 lines
webrick: add test for Digest auth-int
No changes to the actual code, this is a new test for
a feature for which no tests existed. I don't understand
the Digest authentication code well at all, but this is
necessary for the subsequent change.
* test/webrick/test_httpauth.rb (test_digest_auth_int): new test
(credentials_for_request): support bodies with POST
------------------------------------------------------------------------
r62965 | normal | 2018-03-28 17:06:49 +0900 (水, 28 3 2018) | 18 lines
webrick/httpauth/digestauth: stream req.body
WARNING! WARNING! WARNING! LIKELY BROKEN CHANGE
Pass a proc to WEBrick::HTTPRequest#body to avoid reading a
potentially large request body into memory during
authentication.
WARNING! this will break apps completely which want to do
something with the body besides calculating the MD5 digest
of it.
Also, keep in mind that probably nobody uses "auth-int".
Servers such as Apache, lighttpd, nginx don't seem to
support it; nor does curl when using POST/PUT bodies;
and we didn't have tests for it until now...
* lib/webrick/httpauth/digestauth.rb (_authenticate): stream req.body
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_2@63021 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
webrick: support Proc objects as body responses
* lib/webrick/httpresponse.rb (send_body): call send_body_proc
(send_body_proc): new method
(class ChunkedWrapper): new class
* test/webrick/test_httpresponse.rb (test_send_body_proc): new test
(test_send_body_proc_chunked): ditto
[Feature #855]
webrick: favor .write over << method
This will make the next change to use IO.copy_stream
easier-to-read. When we can drop Ruby 2.4 support in a few
years, this will allow us to use writev(2) with multiple
arguments for headers and chunked responses.
* lib/webrick/cgi.rb (write): new wrapper method
lib/webrick/httpresponse.rb: (send_header): use socket.write
(send_body_io): ditto
(send_body_string): ditto
(send_body_proc): ditto
(_write_data): ditto
(ChunkedWrapper#write): ditto
(_send_file): ditto
------------------------------------------------------------------------
r62954 | normal | 2018-03-28 17:05:52 +0900 (水, 28 3 2018) | 14 lines
webrick/httpresponse: IO.copy_stream for regular files
Remove the redundant _send_file method since its functionality
is unnecessary with IO.copy_stream. IO.copy_stream also allows
the use of sendfile under some OSes to speed up copies to
non-TLS sockets.
Testing with "curl >/dev/null" and "ruby -run -e httpd" to
read a 1G file over Linux loopback reveals a reduction from
around ~0.770 to ~0.490 seconds on the client side.
* lib/webrick/httpresponse.rb (send_body_io): use IO.copy_stream
(_send_file): remove
[Feature #14237]
------------------------------------------------------------------------
r62955 | normal | 2018-03-28 17:05:57 +0900 (水, 28 3 2018) | 10 lines
webrick: use IO.copy_stream for single range response
This is also compatible with range responses generated
by Rack::File (tested with rack 2.0.3).
* lib/webrick/httpresponse.rb (send_body_io): use Content-Range
* lib/webrick/httpservlet/filehandler.rb (make_partial_content):
use File object for the single range case
* test/webrick/test_filehandler.rb (get_res_body): use send_body
to test result
------------------------------------------------------------------------
r62956 | normal | 2018-03-28 17:06:02 +0900 (水, 28 3 2018) | 7 lines
test/webrick/test_filehandler.rb: stricter multipart range test
We need to ensure we generate compatibile output in
the face of future changes
* test/webrick/test_filehandler.rb (test_make_partial_content):
check response body
------------------------------------------------------------------------
r62957 | normal | 2018-03-28 17:06:08 +0900 (水, 28 3 2018) | 8 lines
webrick: quiet warning for multi-part ranges
Content-Length is ignored by WEBrick::HTTPResponse even if we
calculate it, so instead we chunk responses to HTTP/1.1 clients
and terminate HTTP/1.0 connections.
* lib/webrick/httpservlet/filehandler.rb (make_partial_content):
quiet warning
------------------------------------------------------------------------
r62958 | normal | 2018-03-28 17:06:13 +0900 (水, 28 3 2018) | 7 lines
webrick/httpresponse: make ChunkedWrapper copy_stream-compatible
The .write method needs to return the number of bytes written
to avoid confusing IO.copy_stream.
* lib/webrick/httpresponse.rb (ChunkedWrapper#write): return bytes written
(ChunkedWrapper#<<): return self
------------------------------------------------------------------------
r62959 | normal | 2018-03-28 17:06:18 +0900 (水, 28 3 2018) | 9 lines
webrick: use IO.copy_stream for multipart response
Use the new Proc response body feature to generate a multipart
range response dynamically. We use a flat array to minimize
object overhead as much as possible; as many ranges may fit
into an HTTP request header.
* lib/webrick/httpservlet/filehandler.rb (multipart_body): new method
(make_partial_content): use multipart_body
get rid of test error/failure on Windows introduced at r62955
* lib/webrick/httpresponse.rb (send_body_io): use seek if NotImplementedError
is raised in IO.copy_stream with offset.
* lib/webrick/httpservlet/filehandler.rb (multipart_body): ditto.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_2@63020 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
load.c: backtrace of circular require
* load.c (load_lock): print backtrace of circular require via
`Warning.warn` [ruby-core:80850] [Bug #13505]
Send the backtrace of the circular require warning as a single String to Warning.warn
* load.c: send as a single string.
* error.c: expose the string formatted by rb_warning as rb_warning_string().
* test/ruby/test_exception.rb: update tests.
[ruby-core:80850] [Bug #13505]
fix regexp literal warning.
* test/rubygems/test_gem_server.rb: eliminate duplicated character class warning.
[Bug #14481]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_2@62441 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
Fix a command injection vulnerability in Net::FTP.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_2@61246 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
lib/webrick/log.rb: sanitize any type of logs
It had failed to sanitize some type of exception messages. Reported and
patched by Yusuke Endoh (mame) at https://hackerone.com/reports/223363
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_2@59902 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* configure.in: improve ICC (Intel C Compiler) support.
* configure.in (CXX): The name of icc's c++ compiler is `icpc`.
* configure.in (warnings): Add `-diag-disable=2259` to suppress
noisy warnings: "non-pointer conversion from "..." to "..." may
lose significant bits".
* configure.in (optflags): Add `-fp-model precise` like -fno-fast-math.
* lib/mkmf.rb: icc supports -Werror=division-by-zero
and -Werror=deprecated-declarations, but doesn't support
-Wdivision-by-zero and -Wdeprecated-declarations.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_2@58143 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
lib/ostruct.rb: [DOC] revise docs for OpenStruct
* update paragraph on implementation:
define_singleton_method is used, not define_method
* add call-seq with return values for each_pair
* adopt description of dig from Array and Hash
* fix description of the hash method
* :nodoc: initialize_copy, respond_to_missing?
* other small improvements, e.g. use the term `attribute' in the docs
(instead of `member'), which is clearer for users of the class
* improve code examples: e.g. use more consistent style (always use
double quotes, drop `p' and `puts', ...), update inspect output,
use example data that is not prone to change (like population)
* add more code examples
* fix some small errors and grammar
[ruby-core:79265] [Bug #13159]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_2@58124 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
nodoc OptParse
* lib/optparse.rb: [DOC] nodoc OptParse, introduced with r46126,
to avoid leaking of its documentation (OptionParser's docs) into
the class documentation of Object. [ruby-core:79909] [Bug #13281]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_2@58122 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
fileutils.rb: do not make root
* lib/fileutils.rb (FileUtils#mkdir_p): no need to make root
directory which should be exist and cannot be made with mkdir
recent Cygwin can make a directory contains a colon.
[Bug #13214]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_2@58120 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
lib/ostruct.rb: Fix returned value of each_pair.
From a patch by Marcus Stollsteimer. [Fixes#13169]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_2@58100 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
ruby-lex.rb: fix for label
* lib/irb/ruby-lex.rb (identify_identifier): treat identifier just
followed by a colon as a lable. this is not a precise solution
but enough for the time being. [ruby-core:78526] [Bug #13012]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_2@58086 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* lib/forwardable.rb (def_instance_delegator): adjust backtrace of
method body by tail call optimization. adjusting the delegated
target is still done by deleting backtrace.
* lib/forwardable.rb (def_single_delegator): ditto.
* lib/forwardable.rb (Forwardable._delegator_method): extract
method generator and deal with non-module objects.
[ruby-dev:49656] [Bug #12478]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_2@58085 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* lib/net/http.rb (transport_request): other than HTTPContinue
in 1xx (HTTPInformation) also needs to continue. [Bug #12890]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_2@56787 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* lib/logger.rb (Logger::Period#next_rotate_time): fix monthly log
rotate when DST is applied during a month of 31 days.
[Fix GH-1458]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_2@56504 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* lib/uri/generic.rb (def check_password): don't include bad password
in URI exception output
* test/uri/test_generic.rb (def test_set_component): test for behavior
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_2@56307 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* lib/uri/mailto.rb (initialize): RFC3986_Parser#split sets opaque
only if the URI has path-rootless, not path-empty.
[ruby-core:76055] [Bug #12498]
patched by Chris Heisterkamp <cheister@squareup.com>
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_2@55941 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* lib/net/http/generic_rquest.rb (write_header): A Request-Line must
not contain CR or LF.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_2@55874 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* lib/net/ftp.rb (putline): raise an ArgumentError when
CR or LF is included in a line.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_2@55873 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* lib/optparse.rb (OptionParser::Completion.candidate): get rid of
nil as key names. [ruby-core:75773] [Bug #12438]
* lib/optparse.rb (OptionParser#make_switch): char class option
cannot be NoArgument, default to RequiredArgument.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_2@55364 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* lib/rubygems/test_case.rb: Fix test on Windows for inconsistent temp path.
https://github.com/rubygems/rubygems/pull/1554
[Bug #12193][ruby-core:74431]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_2@54692 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* lib/forwardable.rb (def_instance_delegator) fix delegating to
'args' and 'block', clashing with local variables in generated
methods. [ruby-core:72579] [Bug #11916]
* lib/forwardable.rb (def_single_delegator): ditto.
* lib/forwardable.rb: Convert given accessors to String.
r53381 changed to accept only Symbol or String for accessors, but
there are several rubygems that pass classes (e.g. Array,
Hash, ...) as accessors. Prior r53381, it was accepted because Class#to_s
returns its class name. After r53381 given accessors are checked
with define_method, but it accepts only Symbol or String, otherwise
raises TypeError.
def_delegator Foo, :some_method
This change is to revert unwanted incompatibility. But this behavior
may change in the future.
This change is to revert unexpected incompatibility. But this behavior
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_2@54673 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* lib/uri/mailto.rb: raising URI::InvalidComponentError instead
of failing with undefined method `split' for nil:NilClass for
mailto: URIs without opaque part. [Bug #10738]
* test/uri/testuri.rb: Test for above
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_2@54341 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* net/ftp.rb: add NullSocket#closed? to fix closing not opened
connection. [Fix GH-1232]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_2@54330 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* lib/logger.rb: Remove block from Logger.add as it's not needed
patch provided by Daniel Lobato Garcia [fix GH-1240] [Bug #12054]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_2@54072 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* lib/xmlrpc/client.rb: Support SSL options in async methods of
XMLRPC::Client.
[Bug #11489]
Reported by Aleksandar Kostadinov. Thanks!!!
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_2@54069 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* lib/rubygems.rb: bump version to 2.4.5.1. this version fixed
CVE-2015-3900.
* lib/rubygems/remote_fetcher.rb: ditto.
* test/rubygems/test_gem_remote_fetcher.rb: added testcase for CVE-2015-3900
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_2@51619 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* lib/net/ftp.rb (makeport): close the TCPServer
when sending the port fails.
* test/net/ftp/test_ftp.rb: test for above.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_2@51556 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* lib/timeout.rb (ExitException): removed internal exception class
and use Timeout::Error instead, as using throw/catch to isolate
each timeouts now. [ruby-dev:49179] [Bug #11344]
* lib/timeout.rb (ExitException): leave Timeout::ExitException as an
alias of Timeout::Error for backward compatibility in stable branch.
[ruby-dev:49179] [Bug #11344]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_2@51523 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* lib/net/http/response.rb (inflater): CONTENT_ENCODING can be upper
case. [ruby-core:69670] [Bug #11285] patched by Andy Chu
* test/net/http/test_httpresponse.rb
(HTTPResponseTest#test_read_body_content_encoding_deflate_uppercase):
fix a failure without zlib.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_2@51522 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
usa