Security fixes:
CVE-2017-9233 -- External entity infinite loop DoS
Details: https://libexpat.github.io/doc/cve-2017-9233/
Commit c4bf96bb51dd2a1b0e185374362ee136fe2c9d7f
[MOX-002] CVE-2016-9063 -- Detect integer overflow; commit
d4f735b88d9932bd5039df2335eefdd0723dbe20
(Fixed version of existing downstream patches!)
(SF.net) #539 Fix regression from fix to CVE-2016-0718 cutting off
longer tag names; commits
* 896b6c1fd3b842f377d1b62135dccf0a579cf65d
* af507cef2c93cb8d40062a0abe43a4f4e9158fb2
#16 * 0dbbf43fdb20f593ddf4fa1ff67288000dd4a7fd
#25 More integer overflow detection (function poolGrow); commits
* 810b74e4703dcfdd8f404e3cb177d44684775143
* 44178553f3539ce69d34abee77a05e879a7982ac
[MOX-002] Detect overflow from len=INT_MAX call to XML_Parse; commits
* 4be2cb5afcc018d996f34bbbce6374b7befad47f
* 7e5b71b748491b6e459e5c9a1d090820f94544d8
[MOX-005] #30 Use high quality entropy for hash initialization:
* arc4random_buf on BSD, systems with libbsd
(when configured with --with-libbsd), CloudABI
* RtlGenRandom on Windows XP / Server 2003 and later
* getrandom on Linux 3.17+
In a way, that's still part of CVE-2016-5300.
https://github.com/libexpat/libexpat/pull/30/commits
[MOX-005] For the low quality entropy extraction fallback code,
the parser instance address can no longer leak, commit
04ad658bd3079dd15cb60fc67087900f0ff4b083
[MOX-003] Prevent use of uninitialised variable; commit
[MOX-004] a4dc944f37b664a3ca7199c624a98ee37babdb4b
Add missing parameter validation to public API functions
and dedicated error code XML_ERROR_INVALID_ARGUMENT:
[MOX-006] * NULL checks; commits
* d37f74b2b7149a3a95a680c4c4cd2a451a51d60a (merge/many)
* 9ed727064b675b7180c98cb3d4f75efba6966681
* 6a747c837c50114dfa413994e07c0ba477be4534
* Negative length (XML_Parse); commit
[MOX-002] 70db8d2538a10f4c022655d6895e4c3e78692e7f
[MOX-001] #35 Change hash algorithm to William Ahern's version of SipHash
to go further with fixing CVE-2012-0876.
https://github.com/libexpat/libexpat/pull/39/commits
Bug fixes:
#32 Fix sharing of hash salt across parsers;
relevant where XML_ExternalEntityParserCreate is called
prior to XML_Parse, in particular (e.g. FBReader)
#28 xmlwf: Auto-disable use of memory-mapping (and parsing
as a single chunk) for files larger than ~1 GB (2^30 bytes)
rather than failing with error "out of memory"
#3 Fix double free after malloc failure in DTD code; commit
7ae9c3d3af433cd4defe95234eae7dc8ed15637f
#17 Fix memory leak on parser error for unbound XML attribute
prefix with new namespaces defined in the same tag;
found by Google's OSS-Fuzz; commits
* 16f87daae5a16132e479e4f71862128c7a915c73
* b47dbc9745932c160893d433220e462bd605f8cd
xmlwf on Windows: Add missing calls to CloseHandle
New features:
#30 Introduced environment switch EXPAT_ENTROPY_DEBUG=1
for runtime debugging of entropy extraction
Other changes:
Increase code coverage
#33 Reject use of XML_UNICODE_WCHAR_T with sizeof(wchar_t) != 2;
XML_UNICODE_WCHAR_T was never meant to be used outside
of Windows; 4-byte wchar_t is common on Linux
(SF.net) #538 Start using -fno-strict-aliasing
(SF.net) #540 Support compilation against cloudlibc of CloudABI
Allow MinGW cross-compilation
(SF.net) #534 CMake: Introduce option "BUILD_doc" (enabled by default)
to bypass compilation of the xmlwf.1 man page
(SF.net) pr2 CMake: Introduce option "INSTALL" (enabled by default)
to bypass installation of expat files
CMake: Fix ninja support
Autotools: Add parameters --enable-xml-context [COUNT]
and --disable-xml-context; default of context of 1024
bytes enabled unchanged
#14 Drop AmigaOS 4.x code and includes
#14 Drop ancient build systems:
* Borland C++ Builder
* OpenVMS
* Open Watcom
* Visual Studio 6.0
* Pre-X Mac OS (MPW Makefile)
If you happen to rely on some of these, please get in
touch for joining with maintenance.
#10 Move from WIN32 to _WIN32
#13 Fix "make run-xmltest" order instability
Address compile warnings
Bump version info from 7:2:6 to 7:3:6
Add AUTHORS file
Infrastructure:
#1 Migrate from SourceForge to GitHub (except downloads):
https://github.com/libexpat/#1 Re-create http://libexpat.org/ project website
Start utilizing Travis CI
Special thanks to:
Andy Wang
Don Lewis
Ed Schouten
Karl Waclawek
Pascal Cuoq
Rhodri James
Sergei Nikulov
Tobias Taschner
Viktor Szakats
and
Core Infrastructure Initiative
Mozilla Foundation (MOSS Track 3: Secure Open Source)
Radically Open Security
Originally, MKCRYPTO was introduced because the United States
classified cryptography as a munition and restricted its export. The
export controls were substantially relaxed fifteen years ago, and are
essentially irrelevant for software with published source code.
In the intervening time, nobody bothered to remove the option after
its motivation -- the US export restriction -- was eliminated. I'm
not aware of any other operating system that has a similar option; I
expect it is mainly out of apathy for churn that we still have it.
Today, cryptography is an essential part of modern computing -- you
can't use the internet responsibly without cryptography.
The position of the TNF board of directors is that TNF makes no
representation that MKCRYPTO=no satisfies any country's cryptography
regulations.
My personal position is that the availability of cryptography is a
basic human right; that any local laws restricting it to a privileged
few are fundamentally immoral; and that it is wrong for developers to
spend effort crippling cryptography to work around such laws.
As proposed on tech-crypto, tech-security, and tech-userlevel to no
objections:
https://mail-index.netbsd.org/tech-crypto/2017/05/06/msg000719.htmlhttps://mail-index.netbsd.org/tech-security/2017/05/06/msg000928.htmlhttps://mail-index.netbsd.org/tech-userlevel/2017/05/06/msg010547.html
P.S. Reviewing all the uses of MKCRYPTO in src revealed a lot of
*bad* crypto that was conditional on it, e.g. DES in telnet... That
should probably be removed too, but on the grounds that it is bad,
not on the grounds that it is (nominally) crypto.
Release 2.2.0 Tue June 21 2016
Security fixes:
#537 CVE-2016-0718 -- Fix crash on malformed input
CVE-2016-4472 -- Improve insufficient fix to CVE-2015-1283 /
CVE-2015-2716 introduced with Expat 2.1.1
#499 CVE-2016-5300 -- Use more entropy for hash initialization
than the original fix to CVE-2012-0876
#519 CVE-2012-6702 -- Resolve troublesome internal call to srand
that was introduced with Expat 2.1.0
when addressing CVE-2012-0876 (issue #496)
Bug fixes:
Fix uninitialized reads of size 1
(e.g. in little2_updatePosition)
Fix detection of UTF-8 character boundaries
Other changes:
#532 Fix compilation for Visual Studio 2010 (keyword "C99")
Autotools: Resolve use of "$<" to better support bmake
Autotools: Add QA script "qa.sh" (and make target "qa")
Autotools: Respect CXXFLAGS if given
Autotools: Fix "make run-xmltest"
Autotools: Have "make run-xmltest" check for expected output
p90 CMake: Fix static build (BUILD_shared=OFF) on Windows
#536 CMake: Add soversion, support -DNO_SONAME=yes to bypass
#323 CMake: Add suffix "d" to differentiate debug from release
CMake: Define WIN32 with CMake on Windows
Annotate memory allocators for GCC
Address all currently known compile warnings
Make sure that API symbols remain visible despite
-fvisibility=hidden
Remove executable flag from source files
Resolve COMPILED_FROM_DSP in favor of WIN32
introducing since release of software to be recognised. This should hopefully
allow the builds to progress a littles further on systems such as the POWER8
which features a little endian 64-bit PowerPC CPU identified as ppc64le.
christos