On Tue, Oct 26, 2021 at 1:38 PM Eric Dumazet <eric.dumazet@xxxxxxxxx> wrote:
>
> From: Eric Dumazet <edumazet@xxxxxxxxxx>
>
> __bpf_prog_run() can run from non IRQ contexts, meaning
> it could be re entered if interrupted.
>
> This calls for the irq safe variant of u64_stats_update_{begin|end},
> or risk a deadlock.
>
> This patch is a nop on 64bit arches, fortunately.
u64_stats_update_begin_irqsave is a nop. Good!
We just sent the last bpf tree PR for this cycle.
We'll probably take it into bpf-next after CI has a chance to run it.
