|
1dafb427
|
2018-09-03T15:29:50
|
|
Don't include SAX.h from globals.h
SAX.h contains a legacy interface with several unprefixed symbols
like `reference`, causing severe namespace pollution. The globals.h
header doesn't need any of these symbols, so remove the #include.
|
|
bfec41b3
|
2018-06-23T12:14:28
|
|
Support xmlTextReaderNextSibling w/o preparsed doc
This implements missing support for readers that are not based on a
preparsed document in xmlTextReaderNextSibling.
|
|
|
d2ef114c
|
2018-06-23T13:08:46
|
|
Fix xmlTextReaderNext with preparsed document
This fixes the traversal of parent nodes using xmlTextReaderNext()
when the reader is based on a preparsed document (created using
xmlReaderWalker(doc)).
Without this fix the parser will abort even though there are parent
nodes it should traverse to, if it is not currently on an element or
attribute node. This is incorrect, since it can be for example on a
text node when it needs to enter backtracking.
|
|
c65c9e8e
|
2018-08-31T11:42:14
|
|
Really declare dllexport/dllimport for Cygwin
Cygwin does not define _WIN32, but still requires dllexport/dllimport
tags for when applications use the --disable-auto-import linker flag,
probably set by the gl_WOE32_DLL autoconf macro in woe32-dll.m4 file.
|
|
|
33d7ca53
|
2018-08-24T10:10:02
|
|
Merge branch 'patch-2' into 'master'
Change dir to $THEDIR after ACLOCAL_PATH check
See merge request GNOME/libxml2!6
|
|
19f5efd6
|
2018-08-23T20:42:23
|
|
Change dir to $THEDIR after ACLOCAL_PATH check
autoreconf creates aclocal.m4 in $srcdir
|
|
|
b6ab8b44
|
2018-08-19T18:59:10
|
|
Improve error message if pkg.m4 couldn't be found
Check whether aclocal.m4 contains the pkg.m4 macro package after running
autoreconf and print a more helpful error message.
Thanks to 林博仁 (Buo-ren Lin) for the initial patch.
Closes: #21
|
|
173ea684
|
2018-08-18T14:22:57
|
|
Don't instruct user to run make when autogen.sh failed
Currently the autogen script will always tell user to run make even
when the configure script failed, this patch changes the behavior to
only show this message when the configure script terminates without
error.
|
|
|
2240fbf5
|
2018-07-30T13:14:11
|
|
Fix infinite loop in LZMA decompression
Check the liblzma error code more thoroughly to avoid infinite loops.
Closes: https://gitlab.gnome.org/GNOME/libxml2/issues/13
Closes: https://bugzilla.gnome.org/show_bug.cgi?id=794914
This is CVE-2018-9251 and CVE-2018-14567.
Thanks to Dongliang Mu and Simon Wörner for the reports.
|
|
|
a4363749
|
2018-07-30T12:54:38
|
|
Fix nullptr deref with XPath logic ops
If the XPath stack is corrupted, for example by a misbehaving extension
function, the "and" and "or" XPath operators could dereference NULL
pointers. Check that the XPath stack isn't empty and optimize the
logic operators slightly.
Closes: https://gitlab.gnome.org/GNOME/libxml2/issues/5
Also see
https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=901817
https://bugzilla.redhat.com/show_bug.cgi?id=1595985
This is CVE-2018-14404.
Thanks to Guy Inbar for the report.
|
|
|
b7c50b8d
|
2018-04-17T12:07:08
|
|
Remove stray character from comment
Fixes bug #795316:
https://bugzilla.gnome.org/show_bug.cgi?id=795316
|
|
d2293cdb
|
2018-01-30T15:04:11
|
|
Remove a misleading line from xmlCharEncOutput
Closes: https://bugzilla.gnome.org/show_bug.cgi?id=793028
It seams this line was accidentally copied over from xmlCharEncOutFunc.
In xmlCharEncOutput output is a pointer so incrementing it by ret can
point it where it wasn't supposed to be pointing. Luckily the current
implementation doesn't dereference the pointer after advancing it.
Signed-off-by: Daniel Veillard <veillard@redhat.com>
|
|
35e83488
|
2018-04-18T15:58:42
|
|
HTML noscript should not close p
For https://bugzilla.gnome.org/show_bug.cgi?id=795343
- HTMLparser.c: noscript should not close <p> but it should close <script>
|
|
|
e61c8217
|
2018-04-16T20:46:13
|
|
Run Travis ASan tests with "sudo: required"
See https://github.com/travis-ci/travis-ci/issues/9033
|
|
|
fa33bf31
|
2017-05-25T00:45:10
|
|
Improve restoring of context size and position
Restore context size and position where it is modified, not in
seemingly random places.
|
|
|
665df41d
|
2018-04-16T19:37:34
|
|
Simplify and harden nodeset filtering
If a nodeset to be filtered is empty, it can be returned without popping
it from the stack.
Make sure to restore the context node in all error paths and never set
it to NULL.
Save and restore the context node in RANGETO operations.
|
|
|
029d0e96
|
2017-05-25T01:28:27
|
|
Avoid unnecessary backups of the context node
|
|
|
938835e7
|
2017-05-25T01:21:57
|
|
Don't change context node in xmlXPathRoot
|
|
|
e22a83b1
|
2017-05-25T01:18:36
|
|
Stop using XPATH_OP_RESET
It only sets the context node to NULL which doesn't seem useful and can
even cause bugs like bug #795299:
https://bugzilla.gnome.org/show_bug.cgi?id=795299
|
|
|
ebe12882
|
2018-04-16T18:18:11
|
|
Fix inconsistency in xmlXPathIsInf
We don't use HUGE_VAL for INFINITY after the most recent fix.
|
|
|
7a1bd7f6
|
2018-03-17T00:03:24
|
|
Revert "Change calls to xmlCharEncInput to set flush false"
This reverts commit 6e6ae5daa6cd9640c9a83c1070896273e9b30d14 which
broke decoding of larger documents with ICU.
See https://bugs.chromium.org/p/chromium/issues/detail?id=820163
|
|
|
7abec671
|
2018-03-15T19:33:52
|
|
NaN and Inf fixes for pre-C99 compilers
On some pre-C99 compilers, the NAN and INFINITY macros don't expand to
constant expressions.
Some MSVC versions complain about floating point division by zero in
constants.
Thanks to Fabrice Manfroi for the report.
|
|
|
18890f47
|
2018-03-05T17:09:43
|
|
Release of libxml2-2.9.8
* configure.ac doc/* libxml2.syms testapi.c: updated for the new release
|
|
d910e99c
|
2017-11-06T16:20:25
|
|
python: remove single use of _PyVerify_fd
_PyVerify_fd disappeared from python with:
https://hg.python.org/cpython/rev/e88e2049b793
https://bugzilla.gnome.org/show_bug.cgi?id=776815
|
|
|
707ad080
|
2018-01-23T16:37:54
|
|
Fix xmlParserEntityCheck
A previous commit removed the check for XML_ERR_ENTITY_LOOP which is
required to abort early in case of excessive entity recursion.
|
|
|
ab362ab0
|
2018-01-22T15:40:05
|
|
Halt parser in case of encoding error
Should fix crbug.com/793715, although I wasn't able to reproduce the
issue.
|
|
|
60dded12
|
2018-01-22T15:04:58
|
|
Clear entity content in case of errors
This only affects recovery mode and avoids integer overflow in
xmlStringGetNodeList and possibly other nasty surprises.
See bug 783052 and
https://bugs.chromium.org/p/oss-fuzz/issues/detail?id=3874
https://bugs.chromium.org/p/chromium/issues/detail?id=796804
|
|
6e6ae5da
|
2018-01-02T21:47:35
|
|
Change calls to xmlCharEncInput to set flush false when not final call. Having flush incorrectly set to true causes errors for ICU.
|
|
|
1637d612
|
2018-01-08T19:19:42
|
|
Disable pointer-overflow UBSan checks under Travis
clang 5 added UBSan checks for pointer overflow which are triggered by
an API test.
|
|
|
132af1a0
|
2018-01-08T18:48:01
|
|
Fix buffer over-read in xmlParseNCNameComplex
Calling GROW can halt the parser if the buffer grows too large. This
will set the buffer to an empty string. Return immediately in this case,
otherwise the "current" pointer is advanced leading to a buffer over-read.
Found with OSS-Fuzz. See
https://oss-fuzz.com/testcase?key=6683819592646656
https://bugs.chromium.org/p/oss-fuzz/issues/detail?id=5031
|
|
|
ad88b54f
|
2017-12-08T09:42:31
|
|
Improve handling of context input_id
For https://bugzilla.gnome.org/show_bug.cgi?id=772726
This was used in xmlsec to detect issues with accessing external entities
and prevent them, but was unreliable, based on a patch from Aleksey Sanin
* parser.c: make sure input_id is incremented when creating sub-entities
for parsing or when parsing out of context
|
|
|
8ac759a6
|
2017-11-27T18:23:58
|
|
Build more test executables on Windows/MSVC
Build all test executables required by 'make checktests'.
|
|
|
706926fd
|
2017-11-27T18:17:38
|
|
Fix ICU library filenames on Windows/MSVC
Fixes bug 790119.
|
|
d4049198
|
2017-11-27T15:08:00
|
|
Add resource file to Windows DLL
Add a resource file containing verison information for the Windows DLL.
This is only used by the MSVC build for now.
|
|
|
790c230f
|
2017-11-12T19:19:53
|
|
Run Travis tests with -Werror
|
|
|
45efd0b0
|
2017-11-13T22:05:22
|
|
Build with "-Wall -Wextra"
Remove warning options that are enabled with "-Wall -Wextra".
Disable -Warray-bounds warning that produces a (seemingly) false
positive in xpath.c.
|
|
|
2677fbf4
|
2017-11-27T14:20:31
|
|
Fix -Wtautological-pointer-compare warnings
Skip tautological pointer comparisons on pthread systems where we don't
use the weak symbols hack.
|
|
|
ff628d46
|
2017-11-13T18:35:51
|
|
Stop including ansidecl.h
This seems to be an undocumented, internal GCC header added a long time
ago. I don't know why it was included, but I think it can be safely
removed.
|
|
|
bf3b4563
|
2017-11-13T18:33:23
|
|
Remove unused AC_CHECKs
|
|
|
cb5541c9
|
2017-11-13T17:08:38
|
|
Fix libz and liblzma detection
If libz or liblzma are detected with pkg-config, AC_CHECK_HEADERS must
not be run because the correct CPPFLAGS aren't set. It is actually not
required have separate checks for LIBXML_ZLIB_ENABLED and HAVE_ZLIB_H.
Only check for LIBXML_ZLIB_ENABLED and remove HAVE_ZLIB_H macro.
Fixes bug 764657, bug 787041.
|
|
|
ddbb075b
|
2017-11-27T14:30:19
|
|
Fix xmlXPathIsNaN broken by recent commit
|
|
|
c1002101
|
2017-11-12T20:06:21
|
|
Update information about contributing
The contents of the HACKING file were hopelessly outdated. Remove the
file and start with a CONTRIBUTING document.
|
|
|
b2189571
|
2017-11-13T21:23:17
|
|
Fix -Wenum-compare warnings
Also fixes a serious bug in xmlXPtrNewRangeNodePoint.
|
|
|
359e7504
|
2017-11-13T21:13:46
|
|
Fix -Wmisleading-indentation warnings
|
|
|
8cb8e317
|
2017-11-12T18:22:23
|
|
Don't touch CFLAGS in configure.ac
CFLAGS shouldn't be touched by configure.ac. The variable is supplied by
the user and must come after other flags, so flags like warning options
can be overridden.
Rename CFLAGS to EXTRA_CFLAGS and add the value to AM_CFLAGS. This also
makes it possible to override flags for each Automake program or library.
|
|
|
00859f91
|
2017-11-12T17:38:08
|
|
Fix callback signature in testapi.c
|
|
|
13acadbb
|
2017-11-12T17:28:12
|
|
Ignore function pointer cast warnings
Use GCC pragmas as these warnings are enabled by -Wpedantic and there's
no way to disable them selectively.
|
|
|
8813f397
|
2017-09-21T00:11:26
|
|
Simplify XPath NaN, inf and -0 handling
Use C99 macros NAN, INFINITY, isnan, isinf. If they're not available:
- Assume that (0.0 / 0.0) generates a NaN and !(x == x) tests for NaN.
- Use C89's HUGE_VAL for INFINITY.
Remove manual handling of NaN, infinity and negative zero in functions
xmlXPathValueFlipSign and xmlXPathDivValues.
Remove xmlXPathGetSign. All the tests for negative zero can be replaced
with a test for negative or positive zero.
Simplify xmlXPathRoundFunction.
Remove Trio dependency.
This should work on IEEE 754 compliant implementations even if the C99
macros aren't available, but will likely break some ancient platforms.
If problems arise, my plan is to port the relevant trionan.c solution
to xpath.c. Note that non-compliant implementations are impossible
to fully support, anyway, since XPath requires IEEE 754.
|
|
|
86182390
|
2017-11-13T22:17:53
|
|
Revert "Compile testapi with -Wno-unused-function"
This reverts commit f404c4f500a0ec18b04463581d5d1b3aae7f7730.
The change broke non-GCC compilers.
|
|
|
772c0648
|
2017-11-09T17:56:31
|
|
Fix unused parameter warning without ICU
|
|
|
86615e43
|
2017-11-09T17:47:47
|
|
Fix IO callback signatures
|
|
|
e5f33e56
|
2017-11-09T17:29:22
|
|
Fix misc callback signatures
Fix parameter and return types for xmlC14NIsVisibleCallback and
xmlRegExecCallbacks.
|
|
|
4dd6d7a5
|
2017-11-09T17:28:00
|
|
Fix list callback signatures
Make sure that all parameters and return values of list callback
functions exactly match the callback function type. This is required
to pass clang's Control Flow Integrity checks and to allow compilation
to asm.js with Emscripten.
Also change the `user` parameter type from `const void *` to `void *`.
|
|
|
e03f0a19
|
2017-11-09T16:42:47
|
|
Fix hash callback signatures
Make sure that all parameters and return values of hash callback
functions exactly match the callback function type. This is required
to pass clang's Control Flow Integrity checks and to allow compilation
to asm.js with Emscripten.
Fixes bug 784861.
|
|
|
71462a60
|
2017-11-09T13:42:33
|
|
Introduce xmlPosixStrdup and update xmlMemStrdup
Introduce xmlPosixStrdup, an internal strdup implementation matching the
POSIX strdup type signature, and update xmlMemStrdup to use it.
Thanks to Vlad Tsyrklevich for the initial patch.
|
|
28f52fe8
|
2017-08-10T15:08:48
|
|
Refactor name and type signature for xmlNop
Update xmlNop's name to xmlInputReadCallbackNop and its type signature
to match xmlInputReadCallback.
Fixes bug 786134.
|
|
|
72182550
|
2017-11-04T15:38:58
|
|
Add test for ICU flush and pivot buffer
|
|
|
0b19f236
|
2017-10-25T18:11:12
|
|
Fixed ICU to set flush correctly and provide pivot buffer.
By always setting flush=TRUE when doing multiple reads, ICU
will not correctly handle truncated utf8 chars across read
boundaries.
The fix is to set flush=TRUE only on final read, and to
provide a pivot buffer which is maintained by libxml
between calls to ucnv_convertEx.
|
|
|
4b413597
|
2017-10-31T17:17:16
|
|
Skip EBCDIC tests if EBCDIC isn't supported
Fixes bug 603432.
|
|
|
f404c4f5
|
2017-10-31T17:14:43
|
|
Compile testapi with -Wno-unused-function
|
|
|
bc5a5d65
|
2017-11-02T21:26:55
|
|
Release of libxnl2-2.9.7
* configure.ac doc/news.html : update for release
* doc/news.html doc/xmlcatalog.1 doc/xmlcatalog_man.html: regenerated
|
|
|
09f51ecb
|
2017-10-24T14:29:26
|
|
Add Makefile rules to rebuild HTML man pages
This should make sure that the online version of the man pages for
xmllint and xmlcatalog stay up to date. Also rebuild both HTML
documents.
|
|
ca01f813
|
2017-10-24T13:51:34
|
|
xmlcatalog: refresh man page wrt. quering system catalog easily
Also normalize "lookup" spelling.
|
|
|
6b780f65
|
2017-10-19T13:27:29
|
|
xmlcatalog: restore ability to query system catalog easily
I've noticed that easy way of locating a DocBook XSLT in a configure
script of another project doesn't work anymore. It is using something
like: xmlcatalog "" ${DOCBOOK_XSL_URI}/${DOCBOOK_XSL_PATH}. The script
is then forced to a plain, suboptimal search using find utility.
Indeed, I retrospectively realize that the check was working just
by mere accident given that the window this presumably side-effect
was applicable had not lasted long, some 8 months between
054c716ea1bf001544127a4ab4f4346d1b9947e7
introducing xmlInitializeCatalog in the LIBXML_TEST_VERSION-rooted
call chain, and
f65128f38289d77ff322d63aef2858cc0a819c34
reverting that again.
So while one can state /etc/xml/catalog constant explicitly, in
some use cases (such as the mentioned one -- that's why I wanted
to omit stating the full path in the first place), this is rather
an implementation detail, perhaps subject to change from that POV.
Therefore I propose to restore that behaviour in the targeted
manner without global disruptions this time around.
As a side-effect, this fixes indenting of the affected part.
|
|
|
5e986e3b
|
2017-10-21T15:09:33
|
|
Fix mixed decls and code in timsort.h
|
|
|
cfa912dd
|
2017-10-21T15:00:47
|
|
Fix deprecated Travis compiler flag
|
|
|
cf820462
|
2017-10-21T14:43:00
|
|
Rework handling of return values in thread tests
Pass a struct to each thread that contains the filename and the return
value. This avoids casting a DWORD to a pointer on Windows.
|
|
f05af837
|
2017-10-21T14:09:16
|
|
Add declaration for DllMain
Fixes a compiler warning.
|
|
|
882a165a
|
2017-10-21T14:04:20
|
|
Fix preprocessor conditional in threads.h
Make sure that the preprocessor conditions and types for xmlDllMain
match exactly in threads.h and threads.c.
|
|
|
154824ee
|
2017-10-21T14:07:38
|
|
Fix unused variable warnings in testrecurse
|
|
|
a4864c2f
|
2017-10-21T14:01:10
|
|
Fix pointer comparison warnings on 64-bit Windows
|
|
|
8bc6c878
|
2017-10-21T13:57:44
|
|
Fix macro redefinition warning
|
|
|
d2c329a9
|
2017-10-21T13:49:31
|
|
Fix -Wimplicit-fallthrough warnings
Add "falls through" comments to quench implicit-fallthrough warnings
which are enabled by -Wextra under GCC 7.
|
|
|
bec3c17f
|
2017-10-12T15:15:58
|
|
Upgrade timsort.h to latest revision
Upgrade timsort.h to revision 05fd77bfec049ce8b7c408c4d3dd2d51ee061a15
from https://github.com/swenson/sort
Removed all code unrelated to Timsort and made minor adjustments for
cross-platform compatibility.
|
|
|
25499fc2
|
2017-10-09T02:39:16
|
|
Increase warning level to /W3 under MSVC
With the exception of timsort.h and warnings 4244 and 4267, the MSVC
build is now warnings-free with /W3.
|
|
|
6472dfe3
|
2017-10-09T16:50:57
|
|
Fix a couple of warnings in dict.c and threads.c
Only visible on Windows.
|
|
|
8575c117
|
2017-10-09T16:41:58
|
|
Default to native threads on MinGW-w64
|
|
|
56723974
|
2017-10-09T15:35:32
|
|
Simplify Windows IO functions
Remove "native" non-Unicode functions which were only needed for
pre-NT systems like Windows 95/98.
Don't redefine `stat` but use `struct _stat` and `_stat()` instead.
|
|
|
d1505ef5
|
2017-10-09T02:10:11
|
|
Update .gitignore for Windows
|
|
|
82e0394d
|
2017-10-09T02:05:41
|
|
Fix runtest on Windows
Unbuffer the output and fix some compiler warnings.
|
|
|
ecbdfa9e
|
2017-10-09T02:01:00
|
|
Fix unused variable warnings in nanohttp.c
Some variables are only used if certain macros are set. Move the
declarations around.
|
|
|
e3890546
|
2017-10-09T00:20:01
|
|
Fix the Windows header mess
Don't include windows.h and wsockcompat.h from config.h but only when
needed.
Don't define _WINSOCKAPI_ manually. This was apparently done to stop
windows.h from including winsock.h which is a problem if winsock2.h
wasn't included first. But on MinGW, this causes compiler warnings.
Define WIN32_LEAN_AND_MEAN instead which has the same effect.
Always use the compiler-defined _WIN32 macro instead of WIN32.
|
|
|
bee8f1e4
|
2017-10-09T00:19:06
|
|
Don't include winsock2.h in xmllint.c
I'm not sure why xmllint.c did include winsock2.h at all. Stop
including the header as both MinGW and MSVC builds don't seem to
require it.
|
|
|
45b0ebdc
|
2017-10-09T00:08:10
|
|
socklen_t is always int on Windows
Define XML_SOCKLEN_T as `int` unconditionally in wsockcompat.h. Fixes
compiler warnings and removes some duplicated code.
|
|
|
5b2324b6
|
2017-10-09T00:05:04
|
|
Don't redefine socket error codes on Windows
Translate WSA error codes instead of blindly redefining POSIX error
codes on Windows. The redefinitions caused warnings in modern Windows
environments, but simply ifdef'ing out the redefinitions seems like
the wrong approach.
|
|
|
d422b954
|
2017-10-09T13:37:42
|
|
Fix pointer/int cast warnings on 64-bit Windows
On 64-bit Windows, `long` is 32 bits wide and can't hold a pointer.
Switch to ptrdiff_t instead which should be the same size as a pointer
on every somewhat sane platform without requiring C99 types like
intptr_t.
Fixes bug 788312.
Thanks to J. Peter Mugaas for the report and initial patch.
|
|
|
41c0a13f
|
2017-10-09T13:32:20
|
|
Fix Windows compiler warnings in xmlCanonicPath
The code handling Windows paths assigned some char/xmlChar pointers
without explicit casts. Also remove an unused variable.
|
|
|
5af594d8
|
2017-10-07T14:54:45
|
|
Fix comparison of nodesets to strings
Fix two bugs in xmlXPathNodeValHash which could lead to errors when
comparing nodesets to strings:
- Only use contents of text nodes to compute the hash for element nodes.
Comments, PIs, and other node types don't affect the string-value and
must be ignored.
- Reset `string` to NULL for node types other than text.
Reported by Aleksei on the mailing list:
https://mail.gnome.org/archives/xml/2017-September/msg00016.html
|
|
|
0c913916
|
2017-10-07T14:40:34
|
|
Remove generated file python/setup.py from version control
|
|
|
03a238f9
|
2017-09-29T13:00:50
|
|
Use __linux__ macro in generated code
|
|
|
4b4d3d85
|
2017-10-06T09:00:53
|
|
Release of libxml2-2.9.6
* configure.ac doc/xml.html doc/news.html: updated for release
|
|
|
0f3b843b
|
2017-06-01T23:12:19
|
|
Fix XPath stack frame logic
Move the calls to xmlXPathSetFrame and xmlXPathPopFrame around in
xmlXPathCompOpEvalPositionalPredicate to make sure that the context
object on the stack is actually protected. Otherwise, memory corruption
can occur when calling sloppily coded XPath extension functions.
Fixes bug 783160.
|
|
|
3157cf4e
|
2017-09-20T16:13:29
|
|
Report undefined XPath variable error message
Commit c851970 removed a redundant error message if XPath evaluation
failed. This uncovered a case where an undefined XPath variable error
wasn't reported correctly.
Thanks to Petr Pisar for the report.
Fixes bug 787941.
|
|
|
b90d8989
|
2017-09-19T15:45:35
|
|
Fix regression with librsvg
Instead of using xmlCreateIOParserCtxt, librsvg pushes its own
xmlParserInput on top of a memory push parser. This incorrect use of
the API confuses several parser checks and, since 2.9.5, completely
breaks documents with internal subsets. Work around the problem with
internal subsets.
Thanks to Petr Sumbera for the report:
https://mail.gnome.org/archives/xml/2017-September/msg00011.html
Also see
https://bugzilla.gnome.org/show_bug.cgi?id=787895
|
|
|
2cdaaab1
|
2017-09-14T21:30:51
|
|
Change preprocessor OS tests to __linux__
"linux" without leading underscores is deprecated and less reliable.
|
|
|
abbda93c
|
2017-09-11T01:14:16
|
|
Handle more invalid entity values in recovery mode
In attribute content, don't emit entity references if there are
problems with the entity value. Otherwise some illegal entity values
like
<!ENTITY a '&#x123456789;'>
would later cause problems like integer overflow.
Make xmlStringLenDecodeEntities return NULL on more error conditions
including invalid char refs and errors from recursive calls. Remove
some fragile error checks based on lastError that shouldn't be
needed now. Clear the entity content in xmlParseAttValueComplex if
an error was found.
Found by OSS-Fuzz. Should fix bug 783052.
Also see https://bugs.chromium.org/p/oss-fuzz/issues/detail?id=3343
|
|
|
c709f3f2
|
2017-09-07T19:52:39
|
|
Fix structured validation errors
Also pass 'str2'. Fixes bug 777473.
|
|
|
07e227ed
|
2017-09-07T18:55:46
|
|
Fix memory leak in LZMA decompressor
|
|
|
e2a9122b
|
2017-09-07T18:36:01
|
|
Set memory limit for LZMA decompression
Otherwise malicious LZMA compressed files could consume large amounts
of memory when decompressed.
According to the xz man page, files compressed with `xz -9` currently
require 65 MB to decompress, so set the limit to 100 MB.
Should fix bug 786696.
|
|
|
0fcab658
|
2017-09-07T18:25:11
|
|
Handle illegal entity values in recovery mode
Make xmlParseEntityValue always return NULL on error. Otherwise some
illegal entity values like
<!ENTITY e '&%#4294967298;'>
would later cause problems like integer overflow.
Found by OSS-Fuzz. Should fix bug 783052.
Also see
https://bugs.chromium.org/p/oss-fuzz/issues/detail?id=592
https://bugs.chromium.org/p/oss-fuzz/issues/detail?id=2732
|