Claudio André

I have a feeling that the `-cl-std=CL1.2` option is too invasive as it stands now. We can easily add a john.conf attribute to prevent it (or, on the contrary, enable...

It's simple as we can see below, but what will happen to the "array arguments" in the new driver version when we disable it, and what effect will the OpenCL...

> OK with you? Since bleeding is no longer a place for experimentation, that's okay. In fact, this even helps with a possible policy change to allow frequent (source only)...

> I should try dropping that -cl-std=CL1.2 I think reverting and testing this (and using a header that is compatible with OpenCL 3.0) is sensible and/or will be necessary sooner...

Or there is something wrong in the source code and no tests with `pcapngbh.block_type` == 2 were run Or There is another possibility: - if `pcapngbh.block_type == 6` is read...

Thanks for reporting. I'm afraid I don't understand your complaint here. What I understood is: - you say the `7z2john` command doesn't work; - I just tested using a Ubuntu...

No. The interesting thing is that the error message is very clear and direct! So I tried to debug it myself, but I couldn't [1]. [1] It's a non-OpenMP build,...

They are already enabled. This is the log obtained this Monday. ``` echo '------------------------- UBSAN fuzzing --------------------------' ../run/john --test=0 Testing: descrypt, traditional crypt(3) [DES 256/256 AVX2]... PASS [...] Testing: andOTP...

> to use my `ofs` instead of `2 * key_len - late_skip` silences UbSan. In fact, the below also silences UbSan: So ubsan is not good at analyzing and doing...