3.0.2: rlm_sql_null duplicating its statements

[Alan DeKok]

Phil Mayers wrote:
> Any write() needs to be wrapped in a loop to ensure the entire buffer
> was handled.

  Yes.  I'll see if I can put a wrapper function into src/lib/

> IIRC there are some weak guarantees in which a write() will tend to be
> atomic - something related to PIPE_BUF even though it's a plain file? -
> which is why rlm_linelog doesn't suffer interleaving problems (much?).

  Small writes are less likely to be broken up.  It's luck, mostly.

  Alan DeKok.