Some asserts on oni_read_frame should be replaced by normal sanity checks about liboni HOT 3 OPEN

Of course I agree that it would be nice to have these checks. Are we worried about performance at all? This is the most performance critical call in the library by a wide margin. I guess a simple if statement is not going to be an issue in the grand scheme of things (e.g. in comparison to what needs to be done to data after it is collected).

from liboni.

Performance is why I discarded checking that the reported frame size equals to the expected frame size for that device, as that would mean looking through the hashtable, which is expensive.

But just checking against two constants (>0 && <= max_frame_size ) should be negligible in comparison with the mallocs that we are already doing there.

from liboni.

Yeah, OK. Lets go ahead and add it then.

from liboni.