Commit Briefs
make the rtable command actually update the rtable of the nsh process (rtable)
The assumption that nsh always runs in rdomain 0 is invalid. Stop assuming that cli_rtable is set to zero by default, and switch the nsh process rdomain accordingly as needed.
on startup, initialize cli_rtable to the current rdomain
Fixes behaviour when nsh starts in a non-default rdomain. We used to assume to start out in rdomain 0, which is not necessarily the case.
remove references to dhclient(8); obsolete since OpenBSD 7.2
Should fix github issue #163 as a side effect
Merge pull request #161 from smytht/logger
add logger command for entering messages into the ..
Update extensive-nsh-openbsd-integrate.sh move motd on import
move /etc/motd on import so the sym link works then
add logger command for entering messages into the ..
system log or /var/log/messages
Update README.md remove chmod commands and use sh ...
.. with shell script instead based on feedback from STSP
Version Bump for release 1.2.2 (tags/v1.2.2)
minor corrections to nsh.8 before release
mandoc -Tlint
add a 'make static' target which builds statically linked binaries
The goal is to make it easier to build the -static flavour of nsh in the OpenBSD ports tree.
su is an alias of enable for convenience
su should not be displayed in help / context help in nsh su will work but not be suggested to user during tab completion
Update extensive-nsh-openbsd-integrate.sh
remove hostname.if and mygate on Integration script run per comment above those lines and add commands to backup the hostname and hosts file configuration
fix adding blackhole and reject routes; a gateway is required
Fixes an invalid route message error during nsh -i at boot time: % cmd: route (line 49) margv [route 224.0.0.0/4 127.0.0.1 iface nompath reject] % RTM_ADD: Add Route: len 128, table 0, pid: 0, seq 1, errno 0, flags:<UP,REJECT,STATIC> % locks: inits: % sockaddrs: <DST,NETMASK> 224.0.0.0 240.0.0.0 % ip_route: rtmsg: Invalid argument test + ok Tom
silence ifconfig errors related to carp demotion at boot time
This happens when nsh is started by /etc/rc.d/nsh during boot. The /etc/rc script starts package scripts with carp-interlock already enabled, and both ifconfig and nsh would attempt to reset the carp demotion counter. Generally, this is harmless, apart from an error message displayed by ifconfig which might mislead users into assuming there was a problem: starting package daemons: nsh. ifconfig: carp: SIOCSIFGATTR: Invalid argument To avoid this issue, nsh can check whether the demotion counter has already been set to the desired level and only change the counter if changing it is indeed required. test + ok Tom