Aleksander Alekseev <aleksander(at)timescale(dot)com> writes:
> This leaves us one patch to deal with.
> [ v4-0001-Pass-Size-size_t-as-a-2nd-argument-of-snprintf.patch ]
I looked at this and am inclined to reject it. None of these
places realistically need to deal with strings longer than
MAXPATHLEN or so, let alone multiple gigabytes. So it's just
code churn, creating backpatch hazards (admittedly not big ones)
for no real gain.
regards, tom lane