linux-arch.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: "Michael S. Tsirkin" <mst@redhat.com>
To: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
Cc: linux-kernel@vger.kernel.org,
	Peter Zijlstra <peterz@infradead.org>,
	Arnd Bergmann <arnd@arndb.de>,
	linux-arch@vger.kernel.org,
	Andrew Cooper <andrew.cooper3@citrix.com>,
	virtualization@lists.linux-foundation.org,
	Jonathan Corbet <corbet@lwn.net>,
	linux-doc@vger.kernel.org
Subject: Re: [PATCH 01/34] Documentation/memory-barriers.txt: document __smb_mb()
Date: Mon, 4 Jan 2016 13:27:36 +0200	[thread overview]
Message-ID: <20160104132634-mutt-send-email-mst@redhat.com> (raw)
In-Reply-To: <alpine.DEB.2.02.1601041104290.19710@kaball.uk.xensource.com>

On Mon, Jan 04, 2016 at 11:08:19AM +0000, Stefano Stabellini wrote:
> On Wed, 30 Dec 2015, Michael S. Tsirkin wrote:
> > Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
> > ---
> >  Documentation/memory-barriers.txt | 33 ++++++++++++++++++++++++++++-----
> >  1 file changed, 28 insertions(+), 5 deletions(-)
> > 
> > diff --git a/Documentation/memory-barriers.txt b/Documentation/memory-barriers.txt
> > index aef9487..a20f7ef 100644
> > --- a/Documentation/memory-barriers.txt
> > +++ b/Documentation/memory-barriers.txt
> > @@ -1655,17 +1655,18 @@ macro is a good place to start looking.
> >  SMP memory barriers are reduced to compiler barriers on uniprocessor compiled
> >  systems because it is assumed that a CPU will appear to be self-consistent,
> >  and will order overlapping accesses correctly with respect to itself.
> > +However, see the subsection on "Virtual Machine Guests" below.
> >  
> >  [!] Note that SMP memory barriers _must_ be used to control the ordering of
> >  references to shared memory on SMP systems, though the use of locking instead
> >  is sufficient.
> >  
> >  Mandatory barriers should not be used to control SMP effects, since mandatory
> > -barriers unnecessarily impose overhead on UP systems. They may, however, be
> > -used to control MMIO effects on accesses through relaxed memory I/O windows.
> > -These are required even on non-SMP systems as they affect the order in which
> > -memory operations appear to a device by prohibiting both the compiler and the
> > -CPU from reordering them.
> > +barriers impose unnecessary overhead on both SMP and UP systems. They may,
> > +however, be used to control MMIO effects on accesses through relaxed memory I/O
> > +windows.  These barriers are required even on non-SMP systems as they affect
> > +the order in which memory operations appear to a device by prohibiting both the
> > +compiler and the CPU from reordering them.
> >  
> >  
> >  There are some more advanced barrier functions:
> > @@ -2948,6 +2949,28 @@ The Alpha defines the Linux kernel's memory barrier model.
> >  
> >  See the subsection on "Cache Coherency" above.
> >  
> > +VIRTUAL MACHINE GUESTS
> > +-------------------
> > +
> > +Guests running within virtual machines might be affected by
> > +SMP effects even if the guest itself is compiled within
> 
>                                                     ^ without

Right - this is fixed in v2.
Could you review that one please?

> > +SMP support.
> > +
> > +This is an artifact of interfacing with an SMP host while
> > +running an UP kernel.
> > +
> > +Using mandatory barriers for this use-case would be possible
> > +but is often suboptimal.
> > +
> > +To handle this case optimally, low-level __smp_mb() etc macros are available.
> > +These have the same effect as smp_mb() etc when SMP is enabled, but generate
> > +identical code for SMP and non-SMP systems. For example, virtual machine guests
> > +should use __smp_mb() rather than smp_mb() when synchronizing against a
> > +(possibly SMP) host.
> > +
> > +These are equivalent to smp_mb() etc counterparts in all other respects,
> > +in particular, they do not control MMIO effects: to control
> > +MMIO effects, use mandatory barriers.
> >  
> >  ============
> >  EXAMPLE USES
> > -- 
> > MST
> > 

  reply	other threads:[~2016-01-04 11:27 UTC|newest]

Thread overview: 90+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-12-30 13:24 [PATCH 00/34] arch: barrier cleanup + __smp_XXX barriers for virt Michael S. Tsirkin
2015-12-30 12:58 ` [PATCH 00/34] arch: barrier cleanup + __smp_xxx " Michael S. Tsirkin
2015-12-30 13:24 ` [PATCH 01/34] Documentation/memory-barriers.txt: document __smb_mb() Michael S. Tsirkin
2016-01-04 11:08   ` Stefano Stabellini
2016-01-04 11:27     ` Michael S. Tsirkin [this message]
2015-12-30 13:24 ` [PATCH 02/34] asm-generic: guard smp_store_release/load_acquire Michael S. Tsirkin
2015-12-30 13:24 ` [PATCH 03/34] ia64: rename nop->iosapic_nop Michael S. Tsirkin
2015-12-30 13:24   ` Michael S. Tsirkin
2015-12-30 16:39   ` Luck, Tony
2015-12-30 16:39     ` Luck, Tony
2015-12-30 13:24 ` [PATCH 04/34] ia64: reuse asm-generic/barrier.h Michael S. Tsirkin
2015-12-30 13:24   ` Michael S. Tsirkin
2015-12-30 16:41   ` Luck, Tony
2015-12-30 16:41     ` Luck, Tony
2015-12-30 13:24 ` [PATCH 05/34] powerpc: " Michael S. Tsirkin
2015-12-30 13:24   ` Michael S. Tsirkin
2015-12-30 13:24 ` [PATCH 06/34] s390: " Michael S. Tsirkin
2015-12-30 13:24   ` Michael S. Tsirkin
2015-12-30 13:24 ` [PATCH 07/34] sparc: " Michael S. Tsirkin
2015-12-30 13:24   ` Michael S. Tsirkin
2015-12-30 17:56   ` David Miller
2015-12-30 19:55     ` Michael S. Tsirkin
2015-12-30 19:55       ` Michael S. Tsirkin
2015-12-30 20:47       ` David Miller
2015-12-30 20:47         ` David Miller
2015-12-30 13:24 ` [PATCH 08/34] asm-generic: smp_store_mb should use smp_mb Michael S. Tsirkin
2015-12-30 13:24   ` Michael S. Tsirkin
2015-12-30 13:44   ` Arnd Bergmann
2015-12-30 20:30     ` Michael S. Tsirkin
2015-12-30 22:43       ` Arnd Bergmann
2015-12-30 13:24 ` [PATCH 09/34] arm: reuse asm-generic/barrier.h Michael S. Tsirkin
2015-12-30 13:24   ` Michael S. Tsirkin
2015-12-30 13:24 ` [PATCH 10/34] arm64: " Michael S. Tsirkin
2015-12-30 13:24   ` Michael S. Tsirkin
2015-12-30 13:25 ` [PATCH 11/34] metag: " Michael S. Tsirkin
2015-12-30 13:25   ` Michael S. Tsirkin
2015-12-30 13:25 ` [PATCH 12/34] mips: " Michael S. Tsirkin
2015-12-30 13:25   ` Michael S. Tsirkin
2015-12-30 13:25 ` [PATCH 13/34] x86/um: " Michael S. Tsirkin
2015-12-30 13:25   ` Michael S. Tsirkin
2015-12-30 13:25 ` [PATCH 14/34] x86: " Michael S. Tsirkin
2015-12-30 13:25   ` Michael S. Tsirkin
2015-12-30 13:25 ` [PATCH 15/34] asm-generic: add __smp_XXX wrappers Michael S. Tsirkin
2015-12-30 13:25   ` [PATCH 15/34] asm-generic: add __smp_xxx wrappers Michael S. Tsirkin
2015-12-30 13:25 ` [PATCH 16/34] powerpc: define __smp_XXX Michael S. Tsirkin
2015-12-30 13:25   ` [PATCH 16/34] powerpc: define __smp_xxx Michael S. Tsirkin
2015-12-30 13:25 ` [PATCH 17/34] arm64: define __smp_XXX Michael S. Tsirkin
2015-12-30 13:25   ` [PATCH 17/34] arm64: define __smp_xxx Michael S. Tsirkin
2015-12-30 13:25 ` [PATCH 18/34] arm: define __smp_XXX Michael S. Tsirkin
2015-12-30 13:25   ` [PATCH 18/34] arm: define __smp_xxx Michael S. Tsirkin
2015-12-30 13:25 ` [PATCH 19/34] blackfin: define __smp_XXX Michael S. Tsirkin
2015-12-30 13:25   ` [PATCH 19/34] blackfin: define __smp_xxx Michael S. Tsirkin
2015-12-30 13:25 ` [PATCH 20/34] ia64: define __smp_XXX Michael S. Tsirkin
2015-12-30 13:25   ` [PATCH 20/34] ia64: define __smp_xxx Michael S. Tsirkin
2015-12-30 16:43   ` [PATCH 20/34] ia64: define __smp_XXX Luck, Tony
2015-12-30 13:25 ` [PATCH 21/34] metag: " Michael S. Tsirkin
2015-12-30 13:25   ` [PATCH 21/34] metag: define __smp_xxx Michael S. Tsirkin
2015-12-30 13:25 ` [PATCH 22/34] mips: define __smp_XXX Michael S. Tsirkin
2015-12-30 13:25   ` [PATCH 22/34] mips: define __smp_xxx Michael S. Tsirkin
2015-12-30 13:25 ` [PATCH 23/34] s390: define __smp_XXX Michael S. Tsirkin
2015-12-30 13:25   ` [PATCH 23/34] s390: define __smp_xxx Michael S. Tsirkin
2015-12-30 13:25 ` [PATCH 24/34] sh: define __smp_XXX, fix smp_store_mb for !SMP Michael S. Tsirkin
2015-12-30 13:25   ` [PATCH 24/34] sh: define __smp_xxx, " Michael S. Tsirkin
2015-12-30 13:26 ` [PATCH 25/34] sparc: define __smp_XXX Michael S. Tsirkin
2015-12-30 13:26   ` [PATCH 25/34] sparc: define __smp_xxx Michael S. Tsirkin
2015-12-30 13:26 ` [PATCH 26/34] tile: define __smp_XXX Michael S. Tsirkin
2015-12-30 13:26   ` [PATCH 26/34] tile: define __smp_xxx Michael S. Tsirkin
2015-12-30 13:26 ` [PATCH 27/34] xtensa: define __smp_XXX Michael S. Tsirkin
2015-12-30 13:26   ` [PATCH 27/34] xtensa: define __smp_xxx Michael S. Tsirkin
2015-12-30 13:26 ` [PATCH 28/34] x86: define __smp_XXX Michael S. Tsirkin
2015-12-30 13:26   ` [PATCH 28/34] x86: define __smp_xxX Michael S. Tsirkin
2015-12-30 13:26 ` [PATCH 29/34] Revert "virtio_ring: Update weak barriers to use dma_wmb/rmb" Michael S. Tsirkin
2015-12-30 13:26   ` Michael S. Tsirkin
2015-12-30 13:26 ` [PATCH 30/34] virtio_ring: update weak barriers to use __smp_XXX Michael S. Tsirkin
2015-12-30 13:26 ` [PATCH 31/34] xenbus: use __smp_XXX barriers Michael S. Tsirkin
2015-12-30 13:26   ` [PATCH 31/34] xenbus: use __smp_xxx barriers Michael S. Tsirkin
2016-01-04 11:58   ` [PATCH 31/34] xenbus: use __smp_XXX barriers Stefano Stabellini
2015-12-30 13:26 ` [PATCH 32/34] xen/io: " Michael S. Tsirkin
2015-12-30 13:26   ` [PATCH 32/34] xen/io: use __smp_xxx barriers Michael S. Tsirkin
2016-01-04 11:59   ` [PATCH 32/34] xen/io: use __smp_XXX barriers Stefano Stabellini
2015-12-30 13:26 ` [PATCH 33/34] virtio: use __smp_load_acquire/__smp_store_release Michael S. Tsirkin
2015-12-30 13:26 ` [PATCH 34/34] virtio_ring: use __smp_store_mb Michael S. Tsirkin
2015-12-30 13:49 ` [PATCH 00/34] arch: barrier cleanup + __smp_XXX barriers for virt Arnd Bergmann
2015-12-30 21:45   ` [PATCH 00/34] arch: barrier cleanup + __smp_xxx " Michael S. Tsirkin
2015-12-30 22:45     ` Arnd Bergmann
2015-12-30 20:46 ` David Miller
2015-12-30 20:46   ` David Miller
2015-12-30 21:36   ` Michael S. Tsirkin
2015-12-30 21:59     ` David Miller
2015-12-30 21:59       ` David Miller

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=20160104132634-mutt-send-email-mst@redhat.com \
    --to=mst@redhat.com \
    --cc=andrew.cooper3@citrix.com \
    --cc=arnd@arndb.de \
    --cc=corbet@lwn.net \
    --cc=linux-arch@vger.kernel.org \
    --cc=linux-doc@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=peterz@infradead.org \
    --cc=stefano.stabellini@eu.citrix.com \
    --cc=virtualization@lists.linux-foundation.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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).