Unveil /tftpboot only if needed

Unless -t is used, this directory is not accessed in any way.

OK millert

No need to declare optind, optarg or opterr; unistd.h does this for us.
From Jan Stary. OK deraadt@

repair missing paths on unveil failure

getifaddrs() can return entries where ifa_addr is NULL. Check for this
before accessing anything in ifa_addr.
ok claudio@

When system calls indicate an error they return -1, not some arbitrary
value < 0. errno is only updated in this case. Change all (most?)
callers of syscalls to follow this better, and let's see if

When system calls indicate an error they return -1, not some arbitrary
value < 0. errno is only updated in this case. Change all (most?)
callers of syscalls to follow this better, and let's see if this strictness
helps us in the future.

show more ...

Unveil is used at tail of initialization to allow "r" of /tftpboot
(that decides whether rarpd should reply), and /etc/ethers "r" for
debug reporting.

Just over 15 years and 35 revisions ago, itojun removed the last use of
SIOCGIFNETMASK in #ifdef'ed out code. Adjust two comments accordingly.

Use <fcntl.h> instead of <sys/file.h> for open() and friends.
Delete a bunch of unnecessary #includes and sort to match style(9)
while doing the above cleanup.

ok deraadt@ krw@

Switch base tools from /dev/bpf0 to /dev/bpf. Now that /dev/bpf has been
around for two releases, it should be safe to do so.

ok bluhm deraadt sthen tb yasuoka

Oops. one "error(NONFATAL,..." got flipped to "error(...)" instead
of "warning(...)".

Spotted by & ok procter@

Split error() into error() and warn() so that error() can be marked
__dead and thus let gcc sleep at night.

ok tb@ beck@

Nuke more unused variables.

ok millert@

Replace the /dev/bpf* open loop with a plain open("/dev/bpf0", ...).
ok deraadt jca

hoist the SOCK_RAW setup to the top before pledge
issues discovered by sebastia
comments from semarie

pledge "stdio rpath dns" right at the start of the servicing loop.
Commiting to gather reports.

Consistency in error messages.

No functional change.

remove util.h include and -lutil after pidfile removal

ok deraadt

stop using pidfile()
ok benno jung jca

use daemon(), jca had the same diff in his tree

stdlib.h is in scope; do not cast malloc/calloc/realloc*
ok millert krw

typo

Routing sockets are here to stay, so delete the wrapper that would use the
defunct ioctl(SIOCSARP)

ok millert@

Replace <sys/param.h> with <limits.h> and other less dirty headers where
possible. Annotate <sys/param.h> lines with their current reasons. Switch
to PATH_MAX, NGROUPS_MAX, HOST_NAME_MAX+1, LOGIN_N

Replace <sys/param.h> with <limits.h> and other less dirty headers where
possible. Annotate <sys/param.h> lines with their current reasons. Switch
to PATH_MAX, NGROUPS_MAX, HOST_NAME_MAX+1, LOGIN_NAME_MAX, etc. Change
MIN() and MAX() to local definitions of MINIMUM() and MAXIMUM() where
sensible to avoid pulling in the pollution. These are the files confirmed
through binary verification.
ok guenther, millert, doug (helped with the verification protocol)

show more ...

Replace setpgrp(0, getpid()) with setpgid(0, 0). OK deraadt@ tedu@

convert select() to poll(). Bit more complicated here. Some interesting
discussion with millert about POLLHUP, which is not needed here because
this is a bpf description, but indicates we need to c

convert select() to poll(). Bit more complicated here. Some interesting
discussion with millert about POLLHUP, which is not needed here because
this is a bpf description, but indicates we need to consider that elsewhere.
ok millert

show more ...