Summary:
Preparing statements allocates memory and is generally best done once and for all.
We need to re-prepare statements when opening a new DB connexion (after each fork).
Reviewed By: mbouaziz
Differential Revision: D5824157
fbshipit-source-id: 4d239ac
master
Jules Villard7 years agocommitted byFacebook Github Bot
(* Higher level of "synchronous" are only useful to guarantee that the db will not be corrupted if the machine crashes for some reason before the data has been actually written to disk. We do not need this kind of guarantee for infer results as one can always rerun infer if interrupted. *)
(* Higher level of "synchronous" are only useful to guarantee that the db will not be corrupted if the machine crashes for some reason before the data has been actually written to disk. We do not need this kind of guarantee for infer results as one can always rerun infer if interrupted. *)
(** Return a reversed list of results obtained by repeatedly stepping through [stmt] and saving only column 0 of each returned row (all that's been needed so far). *)
(** Return a reversed list of results obtained by repeatedly stepping through [stmt] and saving only column 0 of each returned row (all that's been needed so far). *)