On (Wed) 15 Jul 2015 [23:44:52], Nils Carlson wrote:
> On Mon, 13 Jul 2015, Nils Carlson wrote:
>
> >On Mon, 13 Jul 2015, Amit Shah wrote:
>
> <snip>
>
> >>Also, returning TRUE there isn't right - if the connection ends, we
> >>should return FALSE.
> >
> >I agree that this seems reasonable. I will change it and re-test.
> >
>
> I had a closer look, and it seems always returning true is intentional here,
> the called function, tcp_chr_disconnect(chr), handles the deregistration
> from handlers. If we were to return FALSE we would be duplicating work and
> possibly breaking things.
Not sure how.
Anyway, can you please start a new thread, with the author and
reviewers of the patch CC'ed, so they can chime in as well?
Amit
