Closed maelp closed 3 years ago
Ah, I actually see from FFI that the ffi.new
keeps ownership of allocated data, so there shouldn't be a memory leak
I think this is going to need a major bump. No problem with that though.
I updated the version to 5.0.0
I merged this code in the other PR
I updated the code using something similar to what is done for CNAME responses, but looking at the code it seems that there might be some memory leaks in many of those handlers: each time we do a
ffi.new
and there is a parsing error, it seems we do not free the allocated memory