All of lore.kernel.org
 help / color / mirror / Atom feed
From: Linus Torvalds <torvalds@linux-foundation.org>
To: "Michael S. Tsirkin" <mst@redhat.com>
Cc: Davidlohr Bueso <dave@stgolabs.net>,
	Peter Zijlstra <peterz@infradead.org>,
	Ingo Molnar <mingo@kernel.org>,
	Thomas Gleixner <tglx@linutronix.de>,
	"Paul E. McKenney" <paulmck@linux.vnet.ibm.com>,
	Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
	"the arch/x86 maintainers" <x86@kernel.org>,
	Davidlohr Bueso <dbueso@suse.de>,
	"H. Peter Anvin" <hpa@zytor.com>,
	virtualization <virtualization@lists.linux-foundation.org>
Subject: Re: [PATCH 3/4] x86,asm: Re-work smp_store_mb()
Date: Tue, 12 Jan 2016 10:04:13 -0800	[thread overview]
Message-ID: <CA+55aFxarbsXdebbcrB=590qtagh178hhxr5ST+qr7XnjZkgpg@mail.gmail.com> (raw)
In-Reply-To: <20160112193027-mutt-send-email-mst@redhat.com>

On Tue, Jan 12, 2016 at 9:45 AM, Michael S. Tsirkin <mst@redhat.com> wrote:
>
> By the way, the comment in barrier.h says:
>
> /*
>  * Some non-Intel clones support out of order store. wmb() ceases to be
>  * a nop for these.
>  */
>
> and while the 1st sentence may well be true, if you have
> an SMP system with out of order stores, making wmb
> not a nop will not help.
>
> Additionally as you point out, wmb is not a nop even
> for regular intel CPUs because of these weird use-cases.
>
> Drop this comment?

We should drop it, yes. We dropped support for CONFIG_X86_OOSTORE
almost two years ago. See commit 09df7c4c8097 ("x86: Remove
CONFIG_X86_OOSTORE") and it was questionable for a long time even
before that (perhaps ever).

So the comment is stale.

We *do* still use the non-nop rmb/wmb for IO barriers, but even that
is generally questionable. See our "copy_user_64.S" for an actual use
of "movnt" followed by sfence. There's a couple of other cases too. So
that's all correct, but the point is that when we use "movnt" we don't
actually use "wmb()", we are doing assembly, and the assembly should
just use sfence directly.

So it's actually very questionable to ever make even the IO
wmb()/rmb() functions use lfence/sfence. They should never really need
it.

But at the same time, I _really_ don't think we care enough. I'd
rather leave those non-smp barrier cases alone as historial unless
somebody can point to a case where they care about the performance.

We also do have the whole PPRO_FENCE thing, which we can hopefully get
rid of at some point too.

                 Linus

WARNING: multiple messages have this Message-ID (diff)
From: Linus Torvalds <torvalds@linux-foundation.org>
To: "Michael S. Tsirkin" <mst@redhat.com>
Cc: Davidlohr Bueso <dave@stgolabs.net>,
	Davidlohr Bueso <dbueso@suse.de>,
	Peter Zijlstra <peterz@infradead.org>,
	the arch/x86 maintainers <x86@kernel.org>,
	Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
	virtualization <virtualization@lists.linux-foundation.org>,
	"H. Peter Anvin" <hpa@zytor.com>,
	Thomas Gleixner <tglx@linutronix.de>,
	"Paul E. McKenney" <paulmck@linux.vnet.ibm.com>,
	Ingo Molnar <mingo@kernel.org>
Subject: Re: [PATCH 3/4] x86,asm: Re-work smp_store_mb()
Date: Tue, 12 Jan 2016 10:04:13 -0800	[thread overview]
Message-ID: <CA+55aFxarbsXdebbcrB=590qtagh178hhxr5ST+qr7XnjZkgpg@mail.gmail.com> (raw)
In-Reply-To: <20160112193027-mutt-send-email-mst@redhat.com>

On Tue, Jan 12, 2016 at 9:45 AM, Michael S. Tsirkin <mst@redhat.com> wrote:
>
> By the way, the comment in barrier.h says:
>
> /*
>  * Some non-Intel clones support out of order store. wmb() ceases to be
>  * a nop for these.
>  */
>
> and while the 1st sentence may well be true, if you have
> an SMP system with out of order stores, making wmb
> not a nop will not help.
>
> Additionally as you point out, wmb is not a nop even
> for regular intel CPUs because of these weird use-cases.
>
> Drop this comment?

We should drop it, yes. We dropped support for CONFIG_X86_OOSTORE
almost two years ago. See commit 09df7c4c8097 ("x86: Remove
CONFIG_X86_OOSTORE") and it was questionable for a long time even
before that (perhaps ever).

So the comment is stale.

We *do* still use the non-nop rmb/wmb for IO barriers, but even that
is generally questionable. See our "copy_user_64.S" for an actual use
of "movnt" followed by sfence. There's a couple of other cases too. So
that's all correct, but the point is that when we use "movnt" we don't
actually use "wmb()", we are doing assembly, and the assembly should
just use sfence directly.

So it's actually very questionable to ever make even the IO
wmb()/rmb() functions use lfence/sfence. They should never really need
it.

But at the same time, I _really_ don't think we care enough. I'd
rather leave those non-smp barrier cases alone as historial unless
somebody can point to a case where they care about the performance.

We also do have the whole PPRO_FENCE thing, which we can hopefully get
rid of at some point too.

                 Linus

  reply	other threads:[~2016-01-12 18:04 UTC|newest]

Thread overview: 56+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-10-27 19:53 [PATCH -tip 0/4] A few updates around smp_store_mb() Davidlohr Bueso
2015-10-27 19:53 ` [PATCH 1/4] arch,cmpxchg: Remove tas() definitions Davidlohr Bueso
2015-12-04 12:01   ` [tip:locking/core] locking/cmpxchg, arch: " tip-bot for Davidlohr Bueso
2015-10-27 19:53 ` [PATCH 2/4] arch,barrier: Use smp barriers in smp_store_release() Davidlohr Bueso
2015-10-27 20:03   ` Davidlohr Bueso
2015-12-04 12:01   ` [tip:locking/core] lcoking/barriers, arch: " tip-bot for Davidlohr Bueso
2015-10-27 19:53 ` [PATCH 3/4] x86,asm: Re-work smp_store_mb() Davidlohr Bueso
2015-10-27 21:33   ` Linus Torvalds
2015-10-27 22:01     ` Davidlohr Bueso
2015-10-27 22:37     ` Peter Zijlstra
2015-10-28 19:49       ` Davidlohr Bueso
2015-11-02 20:15       ` Davidlohr Bueso
2015-11-03  0:06         ` Linus Torvalds
2015-11-03  1:36           ` Davidlohr Bueso
2016-01-12 13:57           ` Michael S. Tsirkin
2016-01-12 13:57           ` Michael S. Tsirkin
2016-01-12 17:20             ` Linus Torvalds
2016-01-12 17:20               ` Linus Torvalds
2016-01-12 17:45               ` Michael S. Tsirkin
2016-01-12 17:45                 ` Michael S. Tsirkin
2016-01-12 18:04                 ` Linus Torvalds [this message]
2016-01-12 18:04                   ` Linus Torvalds
2016-01-12 20:30               ` Andy Lutomirski
2016-01-12 20:54                 ` Linus Torvalds
2016-01-12 20:54                   ` Linus Torvalds
2016-01-12 20:59                   ` Andy Lutomirski
2016-01-12 20:59                     ` Andy Lutomirski
2016-01-12 21:37                     ` Linus Torvalds
2016-01-12 21:37                       ` Linus Torvalds
2016-01-12 22:14                       ` Michael S. Tsirkin
2016-01-12 22:14                       ` Michael S. Tsirkin
2016-01-13 16:20                       ` Michael S. Tsirkin
2016-01-13 16:20                         ` Michael S. Tsirkin
2016-01-12 22:21                     ` Michael S. Tsirkin
2016-01-12 22:21                       ` Michael S. Tsirkin
2016-01-12 22:55                       ` H. Peter Anvin
2016-01-12 22:55                         ` H. Peter Anvin
2016-01-12 23:24                         ` Linus Torvalds
2016-01-12 23:24                         ` Linus Torvalds
2016-01-13 16:17                           ` Borislav Petkov
2016-01-13 16:17                             ` Borislav Petkov
2016-01-13 16:25                             ` Michael S. Tsirkin
2016-01-13 16:25                               ` Michael S. Tsirkin
2016-01-13 16:33                               ` Borislav Petkov
2016-01-13 16:33                                 ` Borislav Petkov
2016-01-13 16:42                                 ` Michael S. Tsirkin
2016-01-13 16:42                                   ` Michael S. Tsirkin
2016-01-13 16:53                                   ` Borislav Petkov
2016-01-13 16:53                                     ` Borislav Petkov
2016-01-13 17:00                                     ` Michael S. Tsirkin
2016-01-13 17:00                                     ` Michael S. Tsirkin
2016-01-13 18:38                                   ` Linus Torvalds
2016-01-13 18:38                                     ` Linus Torvalds
2015-10-27 19:53 ` [PATCH 4/4] doc,smp: Remove ambiguous statement in smp_store_mb() Davidlohr Bueso
2015-12-04 12:01   ` [tip:locking/core] locking/barriers, arch: Remove ambiguous statement in the smp_store_mb() documentation tip-bot for Davidlohr Bueso
2015-10-27 23:27 ` [PATCH 1/4] arch,cmpxchg: Remove tas() definitions David Howells

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to='CA+55aFxarbsXdebbcrB=590qtagh178hhxr5ST+qr7XnjZkgpg@mail.gmail.com' \
    --to=torvalds@linux-foundation.org \
    --cc=dave@stgolabs.net \
    --cc=dbueso@suse.de \
    --cc=hpa@zytor.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mingo@kernel.org \
    --cc=mst@redhat.com \
    --cc=paulmck@linux.vnet.ibm.com \
    --cc=peterz@infradead.org \
    --cc=tglx@linutronix.de \
    --cc=virtualization@lists.linux-foundation.org \
    --cc=x86@kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body. 
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.