Hi,

I have recently locked down a bunch of electron apps using AppArmor and
I noticed something that doesn't yet make sense in my mind:

All electron-apps I'm using do split into multiple executables, for one
the named executable which I call to start it and for two the app.asar,
which seems to be the electron executable, which is in turn started by
the "named executable".

I locked down the named executable and added

  /path/to/app.asar rix,

to the profile and I would expect that this app.asar is then confined
just like the executable the profile is made for (and which is
originally called). According to htop, the app.asar is indeed a
subprocess of the named executable.

However, it doesn't seem to be so (at least with regards to the
filesystem access). To achieve this, I have to add an additional
profile for /path/to/app.asar, and then modify the line above to

  /path/to/app.asar rpx,

which achieves the desired containment effect.

In practice this is what I'm doing anyways mostly, as the app.asar
usually works with a tighter harness, but according to the
documentation, shouldn't "ix" also have such an effect instead of the
subprocess falling out of confinement?

Or in other words: where is my mental model of AppArmor still incorrect?

(I do have other execution flags in the profiles in question, but all of
them are ix.)


Thanks in advance,
Jonas

-- 
AppArmor mailing list
[email protected]
Modify settings or unsubscribe at: 
https://lists.ubuntu.com/mailman/listinfo/apparmor

Reply via email to