From: Roland Dreier <rdreier@cisco.com>
To: Jason Gunthorpe <jgunthorpe@obsidianresearch.com>
Cc: Matthew Wilcox <matthew@wil.cx>,
linux-ia64@vger.kernel.org, Jeff Garzik <jeff@garzik.org>,
linux-kernel@vger.kernel.org, openib-general@openib.org,
linux-pci@atrey.karlin.mff.cuni.cz
Subject: Re: [openib-general] Ordering between PCI config space writes and MMIO reads?
Date: Wed, 25 Oct 2006 14:04:56 +0000 [thread overview]
Message-ID: <adairi8biev.fsf@cisco.com> (raw)
In-Reply-To: <20061024225935.GK4054@obsidianresearch.com> (Jason Gunthorpe's message of "Tue, 24 Oct 2006 16:59:35 -0600")
> I'm not sure that can work either. The PCI-X spec is very clear, you
> must wait for a non-posted completion if you care about order. Doing a
> config read in the driver as a surrogate flush is not good enough in
> the general case. Like you say, a pci bridge is free to reorder all
> in flight non-posted operations.
No, hang on. Nothing can reorder a dependent read to start after a
write that it depends on, can it? So a config read of PCI_COMMAND
can't start until the completion of a config write of the same
register, right?
- R.
WARNING: multiple messages have this Message-ID (diff)
From: Roland Dreier <rdreier@cisco.com>
To: Jason Gunthorpe <jgunthorpe@obsidianresearch.com>
Cc: Matthew Wilcox <matthew@wil.cx>,
linux-ia64@vger.kernel.org, Jeff Garzik <jeff@garzik.org>,
linux-kernel@vger.kernel.org, openib-general@openib.org,
linux-pci@atrey.karlin.mff.cuni.cz
Subject: Re: [openib-general] Ordering between PCI config space writes and MMIO reads?
Date: Wed, 25 Oct 2006 07:04:56 -0700 [thread overview]
Message-ID: <adairi8biev.fsf@cisco.com> (raw)
In-Reply-To: <20061024225935.GK4054@obsidianresearch.com> (Jason Gunthorpe's message of "Tue, 24 Oct 2006 16:59:35 -0600")
> I'm not sure that can work either. The PCI-X spec is very clear, you
> must wait for a non-posted completion if you care about order. Doing a
> config read in the driver as a surrogate flush is not good enough in
> the general case. Like you say, a pci bridge is free to reorder all
> in flight non-posted operations.
No, hang on. Nothing can reorder a dependent read to start after a
write that it depends on, can it? So a config read of PCI_COMMAND
can't start until the completion of a config write of the same
register, right?
- R.
next prev parent reply other threads:[~2006-10-25 14:04 UTC|newest]
Thread overview: 76+ messages / expand[flat|nested] mbox.gz Atom feed top
2006-10-24 19:13 Ordering between PCI config space writes and MMIO reads? Roland Dreier
2006-10-24 19:13 ` Roland Dreier
2006-10-24 19:22 ` Jeff Garzik
2006-10-24 19:22 ` Jeff Garzik
2006-10-24 21:47 ` Matthew Wilcox
2006-10-24 21:47 ` Matthew Wilcox
2006-10-24 21:51 ` Roland Dreier
2006-10-24 21:51 ` Roland Dreier
2006-10-24 22:12 ` John Partridge
2006-10-24 22:12 ` John Partridge
2006-10-24 22:36 ` Matthew Wilcox
2006-10-24 22:36 ` Matthew Wilcox
2006-10-24 22:43 ` David Miller
2006-10-24 22:43 ` David Miller
2006-10-25 14:15 ` Roland Dreier
2006-10-25 14:15 ` Roland Dreier
2006-10-31 19:02 ` Roland Dreier
2006-10-31 19:02 ` Roland Dreier
2006-10-31 19:53 ` Michael S. Tsirkin
2006-10-31 19:53 ` Michael S. Tsirkin
2006-10-31 19:53 ` Roland Dreier
2006-10-31 19:53 ` Roland Dreier
2006-10-31 19:58 ` Matthew Wilcox
2006-10-31 19:58 ` Matthew Wilcox
2006-10-31 20:28 ` Michael S. Tsirkin
2006-10-31 20:28 ` Michael S. Tsirkin
2006-10-31 20:34 ` Richard B. Johnson
2006-10-31 20:34 ` Richard B. Johnson
2006-10-31 20:47 ` Matthew Wilcox
2006-10-31 20:47 ` Matthew Wilcox
2006-10-31 22:30 ` Roland Dreier
2006-10-31 22:30 ` Roland Dreier
2006-11-01 16:27 ` John Partridge
2006-11-01 16:27 ` John Partridge
2006-11-01 16:46 ` Matthew Wilcox
2006-11-01 16:46 ` Matthew Wilcox
2006-11-01 17:08 ` John Partridge
2006-11-01 17:08 ` John Partridge
2006-11-01 17:14 ` Matthew Wilcox
2006-11-01 17:14 ` Matthew Wilcox
2006-11-01 23:04 ` David Miller
2006-11-01 23:04 ` David Miller
2006-11-02 1:08 ` John Partridge
2006-11-02 1:08 ` John Partridge
2006-10-31 20:50 ` Michael S. Tsirkin
2006-10-31 20:50 ` Michael S. Tsirkin
2006-10-24 22:59 ` [openib-general] " Jason Gunthorpe
2006-10-24 22:59 ` Jason Gunthorpe
2006-10-25 14:04 ` Roland Dreier [this message]
2006-10-25 14:04 ` Roland Dreier
2006-10-24 23:09 ` Michael S. Tsirkin
2006-10-24 23:09 ` Michael S. Tsirkin
2006-10-24 23:27 ` Jack Steiner
2006-10-24 23:27 ` Jack Steiner
2006-10-25 14:05 ` Roland Dreier
2006-10-25 14:05 ` Roland Dreier
2006-11-02 3:05 ` Jeremy Higdon
2006-11-02 3:05 ` Jeremy Higdon
2006-10-24 21:01 ` [openib-general] " JWM
2006-10-24 21:01 ` JWM
2006-10-24 21:24 ` Alan Cox
2006-10-24 21:24 ` Alan Cox
2006-10-24 21:29 ` Roland Dreier
2006-10-24 21:29 ` Roland Dreier
2006-10-24 21:37 ` Jeff Garzik
2006-10-24 21:37 ` Jeff Garzik
2006-10-25 6:30 ` Grant Grundler
2006-10-25 6:30 ` Grant Grundler
2006-10-25 14:11 ` Roland Dreier
2006-10-25 14:11 ` Roland Dreier
2006-10-25 14:18 ` Matthew Wilcox
2006-10-25 14:18 ` Matthew Wilcox
2006-10-25 17:15 ` [openib-general] " Jason Gunthorpe
2006-10-25 17:15 ` Jason Gunthorpe
2006-10-25 18:22 ` Michael S. Tsirkin
2006-10-25 18:22 ` Michael S. Tsirkin
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=adairi8biev.fsf@cisco.com \
--to=rdreier@cisco.com \
--cc=jeff@garzik.org \
--cc=jgunthorpe@obsidianresearch.com \
--cc=linux-ia64@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-pci@atrey.karlin.mff.cuni.cz \
--cc=matthew@wil.cx \
--cc=openib-general@openib.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.