[Cryptech Tech] Coverity Scan

[Fredrik Thulin]

On tisdag 9 oktober 2018 kl. 14:01:55 CEST Joachim Strömbergson wrote:
...
> Start fuzzing at all is really what I think we should aim for at this
> point. I've used AFL before and will look at using it for cryptech. The
> big benifit of OSS-Fuzz as I see it is the significant amount of
> non-artificial intelligence doing analysis of findings.

Great initiative! I did AFL fuzzing of the CLI and HSM (device side of the RPC 
code) some years ago. I used the attached program to expose the RPC code to 
AFL - you might want to use it as a starting point.

It seems I put it in libhal/utils/ and hacked the GNUMakefile to build it, and 
started AFL like this:

  afl-fuzz -i ../in/ -o ../out/ -M fuzzer1 -m none -- ./utils/afl-run @@

I'm attaching what I used to generate the corpus (the '../in/' directory) too. 
This was quite some time ago, so I guess some RPCs might have changed. I also 
think I compiled libhal with an in-memory keystore which we had back in the 
day, don't know if that is still available.

/Fredrik
-------------- next part --------------
A non-text attachment was scrubbed...
Name: afl-run.c
Type: text/x-csrc
Size: 1064 bytes
Desc: not available
URL: <https://lists.cryptech.is/archives/tech/attachments/20181009/8d04b3c9/attachment.c>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: testdata.py
Type: text/x-python
Size: 2224 bytes
Desc: not available
URL: <https://lists.cryptech.is/archives/tech/attachments/20181009/8d04b3c9/attachment.py>