Re: [RFC PATCH v3 for 4.15 08/24] Provide cpu_opv system call

From: Andi Kleen
Date: Thu Nov 16 2017 - 19:14:35 EST

My preference would be just to drop this new super ugly system call.

It's also not just the ugliness, but the very large attack surface
that worries me here.

As far as I know it is only needed to support single stepping, correct?

We already have other code that cannot be single stepped, most
prominently the ring 3 vdso time functions that rely on seq locks.

The code that actually cannot be single stepped is very small
and only a few lines.

As far as I know this wasn't ever a significant problem for anybody, and
there is always a simple workaround (set a temporary break point
after it and continue)

Same should apply to the new rseq regions. They should be all
tiny, and we should just accept that they cannot be single-stepped,
but only skipped.

Then this whole mess would disappear.