If this happens the imsg may no longer be usable as there may be queued
messages, but this is a) already the case with the code now, and b)
would be the case if recvmsg() fails anyway, so we can document that -1
from imsg_read() invalidates the struct imsgbuf.
discussed with and ok eric

freeing the msgbuf.
While here also remove an unnecessary while loop.
ok eric pyr

Make the imsg protocol network-safe.
it might be network safe, but half the imsg based daemons on my firewalls
dont run anymore.

Currently the receiver fetches an imsg via imsg_get() and if he expects
an fd, he then calls imsg_get_fd() to fetch the next fd queued on the
imsgbuf from which the imsg came.
This changes hides the fd queueing mechanism to the API user.  When closing
an imsg with an fd, the message is flagged so that the receiving end knows
it must dequeue the fd in imsg_get() and return it with the imsg structure.
This way there is no (less) possible screw up from imsg_get_fd() not being
called directly after imsg_get() by the user. The retreived imsg is
self-contained.
ok pyr@, "I like that" henning@

ok pyr@

add a flag field, use u_int32_t for pid_t and extend type to 32 bits
for padding.
ok pyr@

behavior change.
ok eric@

signal "connection closed" upstream.
spotted by Valentin Kozamernik <tin@komna.com>

From: Alexander von Gernler <grunk@pestilenz.org>

ssize_t
From: Alexander von Gernler <grunk@pestilenz.org>

there are some bytes left (less than an imsg header, or less than the
imsg header len field says) we copy it to the very beginning of the buffer.
use memmove instead of memcpy since it is not guaranteed that there's no
overlap. while memcpy on OpenBSD is safe, it might not elsewhere, and
we want our code to be correct anyways.
funny enough theo and I talked at length about that last week in dublin,
and I said I believe I had no memcpys with the chance of overlap in ntpd/bgpd -
well, here is one, and Alexander von Gernler <grunk@pestilenz.org> pointed
me to it.

enum imsg_type and not an int

-don't buf_free() on imsg_add() errors, it already does that for us
-use imsg_close() instead of buf_close(), does error handling already

does the error checking for us.

-kill the _pid flavors of imsg_create and imsg_compose, and just add pid as
argument to those
-use imsg_create in imsg_compose instead of duplicating code
-check for datalen overflow

it is just capable of answering (s)ntp4 requests with the local time
for now.
imsg/buffer and logging framework from bgpd, ntp protocol hackery
with Alexander Guy