Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
MIME-Version: 1.0
References: <20181211222326.14581-1-bp@alien8.de> <20181211222326.14581-5-bp@alien8.de>
 <CALCETrVArfAAbQKJ-g_5pOy6dMVgx-FC_mjdC9xzL1jU9T07jw@mail.gmail.com>
 <59aad362-4a5b-dd8b-642f-0dc3f83cf7ee@amd.com> <20181211233901.GV27375@zn.tnic>
 <CALCETrXvZ4avdUh2bTLVTTz3Jxogo3zK5akWTsJ=rkJpf1XiDg@mail.gmail.com>
 <20181212100814.GB6653@zn.tnic> <CALCETrWSG5825HrxwEJGqbSoRz=A_2O-GSBD+K_q9YPHu2TB6A@mail.gmail.com>
 <20181212184459.GE6653@zn.tnic> <d7d04973d6d54a969c9324ab3240b401@AcuMS.aculab.com>
In-Reply-To: <d7d04973d6d54a969c9324ab3240b401@AcuMS.aculab.com>
From:   Andy Lutomirski <luto@kernel.org>
Date:   Sat, 15 Dec 2018 10:53:32 -0800
Message-ID: <CALCETrW6QgD4SLo7S+b+MgX4X4w1wAv1oJoi8px_q4+XGydyOQ@mail.gmail.com>
Subject: Re: [RFC PATCH 4/4] x86/TSC: Use RDTSCP
To:     David Laight <David.Laight@aculab.com>
Cc:     Borislav Petkov <bp@alien8.de>, Andy Lutomirski <luto@kernel.org>,
        Tom Lendacky <Thomas.Lendacky@amd.com>,
        LKML <linux-kernel@vger.kernel.org>, X86 ML <x86@kernel.org>,
        "H. Peter Anvin" <hpa@zytor.com>,
        Josh Poimboeuf <jpoimboe@redhat.com>,
        Peter Zijlstra <peterz@infradead.org>,
        John Stultz <john.stultz@linaro.org>
Content-Type: text/plain; charset="UTF-8"
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Fri, Dec 14, 2018 at 5:39 AM David Laight <David.Laight@aculab.com> wrote:
>
> From:  Borislav Petkov
> > Sent: 12 December 2018 18:45
> ...
> > > The property I want for RDTSC ordering is much weaker: I want it to be
> > > ordered like a load.  Imagine that, instead of an on-chip TSC, the TSC
> > > is literally a location in main memory that gets incremented by an
> > > extra dedicated CPU every nanosecond or so.  I want users of RDTSC to
> > > work as if they were reading such a location in memory using an
> > > ordinary load.  I believe this gives the real desired property that it
> > > should be impossible to observe the TSC going backwards.  This is a
> > > much weaker form of serialization.
> >
> > Well, in that case you need something new.
> >
> > Because, the moment you have a RDTSC in flight and a second RDTSC comes
> > in and that second RDTSC must *not* bypass the first one and execute
> > earlier due to OoO, you need to impose some ordering. And that's pretty
> > much uarch-dependent, I'd say.
> >
> > And I guess on AMD the way to do that is to stop dispatch until the
> > first RDTSC retires.
> >
> > Can it be done faster? Sure. And I'm pretty sure there's a lot of pesky
> > little hw details we're not even hearing of, which get in the way.
>
> ISTR one of the problems with RDTSC serialising is that it is used
> for micro-benchmarks.

If you're benchmarking with that level of detail, you're probably
doing RDTSC directly instead of using the vDSO.  Or, even better,
RDPMC.