On Thu, Jan 13, 2022 at 02:01:24AM +0300, Sergey Shinderuk wrote: > Gave it a thorough read. Looks good, except for errstr not set in a couple > of places (see the diff attached).
Thanks for the review. The comments about pg_hmac_ctx->data were wrong from the beginning, coming, I guess, from one of the earlier patch versions where this was discussed. So I have applied that independently.
I have also spent a good amount of time on that to close the loop and make sure that no code paths are missing an error context, adjusted a couple of comments to explain more the role of *errstr in all the SCRAM routines, and finally applied it on HEAD. -- Michael