Re: Running JITed and interpreted programs simultaneously

From: Andrii Nakryiko <andrii.nakryiko@gmail.com>
To: Juraj Vijtiuk <juraj.vijtiuk@sartura.hr>
Cc: bpf <bpf@vger.kernel.org>, Luka Perkov <luka.perkov@sartura.hr>,
	David Marcinkovic <david.marcinkovic@sartura.hr>
Subject: Re: Running JITed and interpreted programs simultaneously
Date: Tue, 13 Oct 2020 15:05:02 -0700	[thread overview]
Message-ID: <CAEf4BzazaFZQHLcNARGWn4TTJJTQPdBVbskg+bJGp-dds-t1xw@mail.gmail.com> (raw)
In-Reply-To: <CAOjtDRXzkwG84UCUVw0J_WmRt585OhOSjuWbdenYFNFinsSG0Q@mail.gmail.com>

On Fri, Oct 9, 2020 at 12:58 PM Juraj Vijtiuk <juraj.vijtiuk@sartura.hr> wrote:
>
> It would be great to hear if anyone has any thoughts on running a set
> of BPF programs JITed while other programs are run by the interpreter.
>
> Something like that would be useful on 32-bit architectures, as the
> JIT compiler there doesn't support some instructions, primarily
> instructions that work with 64-bit data. As far as I can tell, it is
> unlikely that support will be coming soon as it is a general issue for
> all 32-bit architectures. Atomic operations like BPF_XADD look
> especially problematic regarding support on 32 bit platforms. From
> what I managed to see such a conclusion appeared in a few patches
> where support for 32-bit JITs was added, for example [0].
> That results in some programs being runnable with BPF JIT enabled, and
> some failing during load time, but running successfully without JIT on
> 32-bit platforms.
>
> The only way to run some programs with JIT and some without, that
> seems possible right now, is to manually change
> /proc/sys/net/core/bpf_jit_enable every time a program is loaded.
> Although I've managed to do that and it seems to be working, it seems
> pretty hacky and looks like it could cause race conditions if multiple
> programs were loaded, especially by independent loaders.

I agree, the global file is not flexible enough and can cause problems
in production environment.

I don't see any reason why we shouldn't allow to decide interpreted vs
jitted mode per program during BPF_PROG_LOAD.

See kernel/bpf/core.c, bpf_prog's jit_requested field determines
whether a program is going to be jitted or not. It should be trivial
to allow overriding that during BPF_PROG_LOAD command.

We can probably also generalize this to allow to "force-jit" or
"force-interpret" by users, which would fail if kernel didn't support
requested mode.

>
> At first glance it seems that if something like this was to be added
> to a loader, it would have to either somehow be aware of other BPF
> programs being loaded or possibly implement some sort of locking
> mechanism which also seems hacky. From what I understand, doing it in
> the kernel looks even less promising as bpf_jit_enable is a system
> wide setting, and I imagine that changing it to work on a per program
> basis would pretty much require a rework of the current design, so
> that looks even less promising.
>
> It looks like the best option right now is to just run everything in
> interpreted mode, but I want to make sure that I am not missing
> something. If someone has tried doing something similar, it would be
> great to know about that.
>
> Thanks,
> Juraj Vijtiuk
>
> [0] https://lore.kernel.org/netdev/20200305050207.4159-3-luke.r.nels@gmail.com/