* update bundled libs
* apply ICU ubsan patch from libreoffice
This fixes a few things that are not real problems but ubsan doesn't
like them, e.g., ofz#3926 .
* ofz#3670 fix heap-buffer overflow in ICU
* Get the shared library to build for unrar
* Fuzz by writing temp file and calling CmdExtract::DoExtract()
* Incorporate review feedback
* Incorporate review feedback
* Refs #925 -- Added build for spidermonkey
* review feedback
* Switch to the git mirror
git has --depth, which lets us speed up clones. we can switch to the hg upstream once @durin42 implements shallow clones for hg :-)
* Clone this as m-c
* gstreamer: build glib ourselves
Allows having full backtraces of leaks amongst other things
* gstreamer: Don't build orc
it's still too problematic
* gstreamer: Add new typefind fuzzing target
Covers just the typefind element, which has a lot of parsing code within
* projects: Add GStreamer
This is an initial fuzzer which goes over ogg/theora/vorbis files
using the discoverer process
* gstreamer/build.sh: Cleanup file
* gstreamer/Dockerfile: Update copyright date
* gstreamer: Update project.yaml
Use the security mailing list as the primary contact
Remove explicit sanitizer listing
* gstreamer: Simplify base fuzzer
Removed almost all outputting
I am the original author of the code this is taken for, relicensing
an ultra-simplified version of my original code to Apache.
* gstreamer: Cleanup of build file and dockerfile
* gstreamer: Code minimization and avoid leaks
Data provided by the fuzzer shouldn't be freed (but the wrapping
GstBuffer should).
Avoid logging by default
* gstreamer: Download corpus in Dockerfile
And extract in build.sh
* gstreamer: Move code to repository and more cleanups
Remove custom LDFLAGS (not needed)
Use fuzzing target code from upstream repository