@HulaHoop in T544#16070

How are we doing on RAM use? Is it any more or less efficient than socat after you cut down the number of spawned instances?

Done.

(superseded by systemd socket activation)

Had to be done for Whonix 14 to stop wasting RAM by too many socat instances.

Done and closed.

JasonJAyalaP (Jason J. Ayala P.):

JasonJAyalaP added a comment.

> Try to remove as much as possible.

I pasted the wrong code, I meant to say that I'm assuming we still need:

  for i in /etc/controlportfilt.d/*.conf /rw/config/controlportfilt.d/*.conf; do
     if [ -f "$i" ]; then
        . "$i"
     fi
  done

Try to remove as much as possible.

I pasted the wrong code, I meant to say that I'm assuming we still need:

for i in /etc/controlportfilt.d/*.conf /rw/config/controlportfilt.d/*.conf; do
   if [ -f "$i" ]; then
      . "$i"
   fi
done

Did you try new identity? Does it work?

In T657#13429, @JasonJAyalaP wrote:

I commented out the three lines and rebooted. Both stable and alpha connect fine.

I commented out the three lines and rebooted. Both stable and alpha connect fine.

systemd feature request:
make ListenStream= port number available as ${PORT} for ExecStart
https://github.com/systemd/systemd/issues/5403

As per https://lists.freedesktop.org/archives/systemd-devel/2017-February/038261.html the package would not be that cleanly implemented. Disadvantage: lots of systemd unit files. One per port. (Also one per unix domain socket file.)

Good idea.

• https://github.com/Whonix/anon-ws-disable-stacked-tor/commit/24421e2dadbcec6911ecb41bc8f7fce5b23a018f
• https://github.com/Whonix/anon-ws-disable-stacked-tor/commit/678d81fd2d1a87517c54874792b6d18935507878

anon-ws-disable-stacked-tor stable upgrade package was released:
https://forums.whonix.org/t/tor-browser-6-5a4-connectivity-broken-blocked-by-apparmor-profile-since-tbb-changed-to-sockssocket/3154/10?u=patrick

The variable names ended up being:

I kept this as review, since this all has to be tested again when a new build has been created.

Got an answer. Will be probably TOR_SOCKS_SOCKET.

In latest torbutton version there already is environment variable TOR_CONTROL_SOCKET.

https://github.com/Whonix/anon-ws-disable-stacked-tor/commit/63a0d3ccd15646cf11d7c37a2a7fbde7e4b0bb61

Since rinetd has been replaced with socat in T464, this code can be removed.

https://github.com/Whonix/anon-ws-disable-stacked-tor/commit/57673fa1923536319853dcc93a148703dbb3eabc
https://github.com/Whonix/whonix-legacy/commit/7d2ae02c19b907926c63ce6613310740e70e53dc

https://github.com/Whonix/anon-ws-disable-stacked-tor/commit/2ead73c2d15ba39836ec14f122a14f5047bb768a

https://github.com/Whonix/anon-ws-disable-stacked-tor

https://forums.whonix.org/t/retroshare-tickets/2364/2?u=patrick makes me wonder, if this whole thing should be configurable using a .d configuration folder.

https://forums.whonix.org/t/retroshare-tickets/2364/2?u=patrick makes me wonder, if this whole thing should be configurable using a .d configuration folder.

Ricochet is moving to UNIX sockets too:

marmarek (Marek Marczykowski-Górecki):

How is Tor started in TBB?

How is Tor started in TBB? If it is possible to disable it (I hope so), there should be no race condition with it. When user start TBB, the socket should be already there, and wont be overwritten because Tor is disabled. Right?

Other crazy approaches:

As no crypto is used in Whonix (and generally nothing more than simple
TCP connection), I think it doesn't matter.

A very ugly crypto bug/backdoor was dug up in socat. Maybe an aternative is better?

iptables cannot redirect local port traffic to remote ports, otherwise
these tools would not be needed.

What about iptables redirect? Isn't that enough, or some problems with it?