Cimple cannot actually find these without also causing false positives,
but I found them with cimple before removing the code causing false
positives again.
These were found by the new stronger type check in cimple. The one
bugfix is in `crypto_sha512_cmp`, which used to think `crypto_verify_32`
returns bool while actually it's -1/0/1.
Use of `strcpy` in these particular cases was safe, but it's hard to
tell and also useless. `strcpy` would effectively need to do another
`strlen` which we already did.
Also removed sprintf, which was also safe in this case but it's easier to
be "obviously safe", especially for static analysers.
These have all dependencies built with asan/tsan, so they can do deeper
sanity checks such as checking what opus is doing with our buffers.
The asan check currently fails for me locally, so these are not yet set
to be required for PRs to be merged.
This seems to be the sweet spot for the current tokstyle implementation.
4 cores gives the same speedup, 5 also, 6 makes it slower, and 2 also
makes it slower.
These display some idea, but the tests are not implemented correctly. We
will need to implement the idea correctly later, but for now we can't use
these.
This change does not include the addition of VP9. We do that in a
separate pull request.
Changes:
* fix the video bug (video frames larger than 65KBytes) by sending full
frame length in alternate header field
* improve video frame reconstruction logic with slots
* configure video encoder and decoder to be multihtreaded
* set error resilience flags on video codec
* change encoder and decoder softdeadline
