Re: [PATCH v4 54/57] x86/mm: convert arch_within_stack_frames() to use the new unwinder

[Josh Poimboeuf]

On Mon, Aug 22, 2016 at 05:59:18PM -0700, Kees Cook wrote:
> On Mon, Aug 22, 2016 at 4:33 PM, Josh Poimboeuf <jpoimboe@xxxxxxxxxx> wrote:
> > On Mon, Aug 22, 2016 at 03:27:19PM -0500, Josh Poimboeuf wrote:
> >> On Fri, Aug 19, 2016 at 04:55:22PM -0500, Josh Poimboeuf wrote:
> >> > On Fri, Aug 19, 2016 at 11:27:18AM -0700, Kees Cook wrote:
> >> > > On Thu, Aug 18, 2016 at 6:06 AM, Josh Poimboeuf <jpoimboe@xxxxxxxxxx> wrote:
> >> > > > Convert arch_within_stack_frames() to use the new unwinder.
> >> > > >
> >> > > > This also changes some existing behavior:
> >> > > >
> >> > > > - Skip checking of pt_regs frames.
> >> > > > - Warn if it can't reach the grandparent's stack frame.
> >> > > > - Warn if it doesn't unwind to the end of the stack.
> >> > > >
> >> > > > Signed-off-by: Josh Poimboeuf <jpoimboe@xxxxxxxxxx>
> >> > >
> >> > > All the stuff touching usercopy looks good to me. One question,
> >> > > though, in looking through the unwinder. It seems like it's much more
> >> > > complex than just the frame-hopping that the old
> >> > > arch_within_stack_frames() did, but I'm curious to hear what you think
> >> > > about its performance. We'll be calling this with every usercopy that
> >> > > touches the stack, so I'd like to be able to estimate the performance
> >> > > impact of this replacement...
> >> >
> >> > Yeah, good point.  I'll take some measurements from before and after and
> >> > get back to you.
> >>
> >> I took some before/after measurements by enclosing the affected
> >> functions with ktime calls to get the total time spent in each function,
> >> and did a "find /usr >/dev/null" to trigger a bunch of user copies.
> >>
> >>       copy_to/from_user       check_object_size       arch_within_stack_frames
> >> before: 13ms                  6.8ms                   0.61ms
> >> after:        17ms                    11ms                    4.6ms
> >>
> >> The unwinder port made arch_within_stack_frames() *much* (8x) slower
> >> than its current simple implementation, and added about 30% (4ms) to the
> >> total copy_to/from_user() run time.
> >>
> >> Note that hardened usercopy itself is already quite slow: it made user
> >> copies about 52% slower.  With the unwinder port, that worsened to ~65%.
> >
> > FWIW, I think I messed up my math summary here.  Hardened usercopy was
> > roughly 110% slower than normal usercopy (i.e., it took more than twice
> > as long) with 52% of the usercopy time being consumed by
> > check_object_size().
> 
> And this is comparing usercopy to hardened usercopy, which isn't
> expected to be super fast, it's just a cheap expense in comparison to
> the rest of the work being done for a given syscall.
> 
> > With the unwinder, that worsened to 180% slower -- with 65% of the
> > usercopy time being consumed by check_object_size().
> 
> That's quite a bit more than just a simple frame walk. You mentioned a
> few benefits to using the unwinder, but I'm trying to make sure the
> cases it covers can actually happen during a usercopy?

Yeah, I really don't know.  And given Linus's objections, I think I'll
drop it, and the other usercopy patches.

Though I think "move arch_within_stack_frames() to usercopy.c" is still
a nice cleanup if you want to pick that one up.

-- 
Josh