login.c bug?

[Cristian Ionescu-Idbohrn]

The statement [1] in functions: read_or_build_utent and write_utent
is (IMHO) nonsensical.

Function read_or_build_utent: the utent structure is 1st zeroed (using
memset) and then zeroed [1] again. Does it need to?

Function write_utent: sets [1] again (and is also doing more than
documented a few lines above).

I would have expected something like:

	if (time(&utent.ut_time) == -1) {
		<error handling>
	}

instead of plain [1], to make it usable. Current behaviour was
introduced in r16075 (see diff -r16074:16075).

Thoughts?


Cheers,
Cristian

[1] utent.ut_time = time(NULL);