[MODERATED] Re: [PATCH SSBv11 0/3] seccomp 1

[Andi Kleen <ak@linux.intel.com>]

> The other problem with 'site isolation' is that it doesn't necessarily solve or 
> even mitigate the problem: if for example malicious Javascript is injected from an 
> ad network, supposedly safely sandboxed, but it can still anomalously read site 

If the ad network injects JS on your site it can already read everything
of that site in the JS context. So there's no threat on the JS level data.

But I believe normally ads are running in a different site context anyways,
because they are served from the adservers, not the site's server.

> local data via leaky speculation then that's still a dangerous violation of 
> sandboxing constraints: it could read pointers to defeat ASLR, 

That's true, but then it would still be jailed in the seccomp syscall sandbox.
Also if there's an attack where the pointers help it's likely already
exploitable with standard spraying etc. techniques.

> it could read local keys or other data it's not supposed to read.

Everything sensitive (and especially keys) is supposed to be in other processes.

> Once a browser specifically knows that it has fully mitigated against an attack it 
> can turn off any default mitigation early in its init sequence via the prctl, when 
> it still has full OS access and no seccomp isolation. All child tasks should 
> inherit that.

Ok, so the browser has to then essentially work around that Linux bogosity.
Would be better to not have it in the first place.

-Andi