linux-arm-kernel.lists.infradead.org archive mirror
 help / color / mirror / Atom feed
From: will.deacon@arm.com (Will Deacon)
To: linux-arm-kernel@lists.infradead.org
Subject: [PATCH 3/6] irqchip: gic: use writel instead of dsb + writel_relaxed
Date: Fri, 7 Feb 2014 11:23:38 +0000	[thread overview]
Message-ID: <20140207112337.GC5976@mudshark.cambridge.arm.com> (raw)
In-Reply-To: <20140206152048.GD32520@arm.com>

On Thu, Feb 06, 2014 at 03:20:48PM +0000, Catalin Marinas wrote:
> On Thu, Feb 06, 2014 at 01:26:44PM +0000, Will Deacon wrote:
> > On Thu, Feb 06, 2014 at 12:23:40PM +0000, Catalin Marinas wrote:
> > > On Thu, Feb 06, 2014 at 12:13:50PM +0000, Will Deacon wrote:
> > > > Ok, so if we assume that a dsb(ishst) is sufficient because the CPU we're
> > > > talking to is either (a) coherent in the inner-shareable domain or (b)
> > > > incoherent, and we flushed everything to PoC, then why wouldn't a dmb(ishst)
> > > > work?
> > > 
> > > Because you want to guarantee the ordering between a store to Normal
> > > Cacheable memory vs store to Device for the IPI (see the mailbox example
> > > in the Barrier Litmus section ;)). The second is just a slave access, DMB
> > > guarantees observability from the master access perspective.
> > 
> > Ok, my reasoning is as follows:
> > 
> >   - CPU0 tries to message CPU1. It writes to a location in normal memory,
> >     then writes to the GICD to send the SGI
> > 
> >   - We need to ensure that CPU1 observes the write to normal memory before
> >     the write to GICD reaches the distributor. This is *not* about end-point
> >     ordering (the usual non-coherent DMA example).
> > 
> >   - A dmb ishst ensures that the two writes are observed in order by CPU1
> >     (and, in fact, the inner-shareable domain containing CPU0).
> 
> The last bullet point is not correct. DMB would only guarantee that the
> two writes (memory and GICD) are observed by CPU1 if CPU1 actually read
> the GICD (observability is defined for master accesses).

No, that's not how observability is defined, unfortunately.

Rather than attempt to solve this via email (your examples below are already
getting hard to follow :), how about we sit down with $drink_of_choice and
post back here with our conclusions?

In the meantime, I'll use dsb(ishst) because I think we now agree that
works. The question is whether it can be relaxed further to a dmb.

Will

  reply	other threads:[~2014-02-07 11:23 UTC|newest]

Thread overview: 29+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-02-06 11:30 [PATCH 1/6] arm64: barriers: allow dsb macro to take option parameter Will Deacon
2014-02-06 11:30 ` [PATCH 2/6] arm64: barriers: make use of barrier options with explicit barriers Will Deacon
2014-02-06 11:41   ` Catalin Marinas
2014-02-06 11:45     ` Will Deacon
2014-02-06 11:49       ` Catalin Marinas
2014-02-06 11:52         ` Will Deacon
2014-02-06 11:30 ` [PATCH 3/6] irqchip: gic: use writel instead of dsb + writel_relaxed Will Deacon
2014-02-06 11:39   ` Marc Zyngier
2014-02-06 11:45   ` Catalin Marinas
2014-02-06 11:51     ` Will Deacon
2014-02-06 11:54       ` Catalin Marinas
2014-02-06 11:57         ` Will Deacon
2014-02-06 12:00           ` Catalin Marinas
2014-02-06 12:13             ` Will Deacon
2014-02-06 12:23               ` Catalin Marinas
2014-02-06 13:26                 ` Will Deacon
2014-02-06 15:20                   ` Catalin Marinas
2014-02-07 11:23                     ` Will Deacon [this message]
2014-02-07 12:57                       ` Catalin Marinas
2014-02-14 16:30                       ` Will Deacon
2014-02-14 16:48                         ` Catalin Marinas
2014-02-14 17:18                           ` Rob Herring
2014-02-06 11:30 ` [PATCH 4/6] iommu/arm-smmu: provide option to dsb macro when publishing tables Will Deacon
2014-02-06 11:51   ` Catalin Marinas
2014-02-06 11:30 ` [PATCH 5/6] arm64: barriers: wire up new barrier options Will Deacon
2014-02-06 11:55   ` Catalin Marinas
2014-02-06 11:30 ` [PATCH 6/6] arm64: barriers: use barrier() instead of smp_mb() when !SMP Will Deacon
2014-02-06 11:56   ` Catalin Marinas
2014-02-06 11:39 ` [PATCH 1/6] arm64: barriers: allow dsb macro to take option parameter Catalin Marinas

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=20140207112337.GC5976@mudshark.cambridge.arm.com \
    --to=will.deacon@arm.com \
    --cc=linux-arm-kernel@lists.infradead.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).