Subject: Re: fork1(), uvm_fork() and cpu_fork()
To: Alan Barrett <apb@cequrux.com>
From: Greywolf <greywolf@starwolf.com>
List: tech-kern
Date: 12/01/2002 09:23:35
On Sun, 1 Dec 2002, Alan Barrett wrote:

# No, that can't be the reason, because the code is quite willing to
# display that stuff to unprivileged users if it's in memory.  There's a
# macro that tests whether either the forceuread flag is set or a bit in
# some structure says that data is in memory; sorry I can't easily find
# the exact macro or variable names right now, but grep for forceuread to
# find it.
#
# I expect that the original reason was more like limiting the amount of
# paging activity done by ps on behalf of unprivileged users.

which is silly given that ps is not installed setuid on NetBSD,
which means that anyone can go and get the source and recompile
it without that check.

I don't see that restricting it to the super-user is at all useful --
either leave out the check for uid==0 or leave out the option altogether,
IM(NS)HO.

# --apb (Alan Barrett)


				--*greywolf;
--
NetBSD: u_long uptime;