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;
Date:   Tue, 20 Mar 2018 11:54:27 +0100
From:   Ingo Molnar <mingo@kernel.org>
To:     Thomas Gleixner <tglx@linutronix.de>
Cc:     David Laight <David.Laight@ACULAB.COM>,
        'Rahul Lakkireddy' <rahul.lakkireddy@chelsio.com>,
        "x86@kernel.org" <x86@kernel.org>,
        "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
        "netdev@vger.kernel.org" <netdev@vger.kernel.org>,
        "mingo@redhat.com" <mingo@redhat.com>,
        "hpa@zytor.com" <hpa@zytor.com>,
        "davem@davemloft.net" <davem@davemloft.net>,
        "akpm@linux-foundation.org" <akpm@linux-foundation.org>,
        "torvalds@linux-foundation.org" <torvalds@linux-foundation.org>,
        "ganeshgr@chelsio.com" <ganeshgr@chelsio.com>,
        "nirranjan@chelsio.com" <nirranjan@chelsio.com>,
        "indranil@chelsio.com" <indranil@chelsio.com>,
        Andy Lutomirski <luto@kernel.org>,
        Peter Zijlstra <a.p.zijlstra@chello.nl>,
        Fenghua Yu <fenghua.yu@intel.com>,
        Eric Biggers <ebiggers3@gmail.com>
Subject: Re: [RFC PATCH 0/3] kernel: add support for 256-bit IO access
Message-ID: <20180320105427.bm4od7cpessbraag@gmail.com>
References: <cover.1521469118.git.rahul.lakkireddy@chelsio.com>
 <7f0ddb3678814c7bab180714437795e0@AcuMS.aculab.com>
 <alpine.DEB.2.21.1803191557400.2010@nanos.tec.linutronix.de>
 <7f8d811e79284a78a763f4852984eb3f@AcuMS.aculab.com>
 <alpine.DEB.2.21.1803191625080.2010@nanos.tec.linutronix.de>
 <20180320082651.jmxvvii2xvmpyr2s@gmail.com>
 <alpine.DEB.2.21.1803200933320.6506@nanos.tec.linutronix.de>
 <20180320090802.qw4tqjmhy6yfd6sf@gmail.com>
 <alpine.DEB.2.21.1803201039460.6506@nanos.tec.linutronix.de>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <alpine.DEB.2.21.1803201039460.6506@nanos.tec.linutronix.de>
User-Agent: NeoMutt/20170609 (1.8.3)
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk


* Thomas Gleixner <tglx@linutronix.de> wrote:

> On Tue, 20 Mar 2018, Ingo Molnar wrote:
> > * Thomas Gleixner <tglx@linutronix.de> wrote:
> > 
> > > > So I do think we could do more in this area to improve driver performance, if the 
> > > > code is correct and if there's actual benchmarks that are showing real benefits.
> > > 
> > > If it's about hotpath performance I'm all for it, but the use case here is
> > > a debug facility...
> > > 
> > > And if we go down that road then we want a AVX based memcpy()
> > > implementation which is runtime conditional on the feature bit(s) and
> > > length dependent. Just slapping a readqq() at it and use it in a loop does
> > > not make any sense.
> > 
> > Yeah, so generic memcpy() replacement is only feasible I think if the most 
> > optimistic implementation is actually correct:
> > 
> >  - if no preempt disable()/enable() is required
> > 
> >  - if direct access to the AVX[2] registers does not disturb legacy FPU state in 
> >    any fashion
> > 
> >  - if direct access to the AVX[2] registers cannot raise weird exceptions or have
> >    weird behavior if the FPU control word is modified to non-standard values by 
> >    untrusted user-space
> > 
> > If we have to touch the FPU tag or control words then it's probably only good for 
> > a specialized API.
> 
> I did not mean to have a general memcpy replacement. Rather something like
> magic_memcpy() which falls back to memcpy when AVX is not usable or the
> length does not justify the AVX stuff at all.

OK, fair enough.

Note that a generic version might still be worth trying out, if and only if it's 
safe to access those vector registers directly: modern x86 CPUs will do their 
non-constant memcpy()s via the common memcpy_erms() function - which could in 
theory be an easy common point to be (cpufeatures-) patched to an AVX2 variant, if 
size (and alignment, perhaps) is a multiple of 32 bytes or so.

Assuming it's correct with arbitrary user-space FPU state and if it results in any 
measurable speedups, which might not be the case: ERMS is supposed to be very 
fast.

So even if it's possible (which it might not be), it could end up being slower 
than the ERMS version.

Thanks,

	Ingo