I've backpatched Thomas's fixes for the potential buffer overruns in
the datetime code to the REL7_2_STABLE branch. The required changes
are pretty minimal, and the code passes the regression tests.
However, I haven't yet seen a test-case that demonstrates the buffer
overrun in 7.2 code (I originally found the problem when playing with
datetime on 7.3-dev with integer datetimes) -- I've tested it a little
bit and it seems to work, but this patch probably requires pretty
widespread testing.
Cheers,
Neil
--
Neil Conway <neilc@samurai.com> || PGP Key ID: DB3C29FC