Re: [PATCH v2 4/4] x86/static_call: Add inline static call implementation for x86-64

From: Linus Torvalds
Date: Thu Nov 29 2018 - 18:04:44 EST


On Thu, Nov 29, 2018 at 12:25 PM Josh Poimboeuf <jpoimboe@xxxxxxxxxx> wrote:
>
> On Thu, Nov 29, 2018 at 11:27:00AM -0800, Andy Lutomirski wrote:
> >
> > I propose a different solution:
> >
> > As in this patch set, we have a direct and an indirect version. The
> > indirect version remains exactly the same as in this patch set. The
> > direct version just only does the patching when all seems well: the
> > call instruction needs to be 0xe8, and we only do it when the thing
> > doesn't cross a cache line. Does that work? In the rare case where
> > the compiler generates something other than 0xe8 or crosses a cache
> > line, then the thing just remains as a call to the out of line jmp
> > trampoline. Does that seem reasonable? It's a very minor change to
> > the patch set.
>
> Maybe that would be ok. If my math is right, we would use the
> out-of-line version almost 5% of the time due to cache misalignment of
> the address.

Note that I don't think cache-line alignment is necessarily sufficient.

The I$ fetch from the cacheline can happen in smaller chunks, because
the bus between the I$ and the instruction decode isn't a full
cacheline (well, it is _now_ in modern big cores, but it hasn't always
been).

So even if the cacheline is updated atomically, I could imagine seeing
a partial fetch from the I$ (old values) and then a second partial
fetch (new values).

It would be interesting to know what the exact fetch rules are.

Linus