From: Nick Piggin <nickpiggin@yahoo.com.au>
To: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Keith Packard <keithp@keithp.com>, Dave Airlie <airlied@linux.ie>,
Andrew Morton <akpm@linux-foundation.org>,
Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
dri-devel@lists.sf.net
Subject: Re: [git pull] drm patches for 2.6.27-rc1
Date: Sun, 19 Oct 2008 14:14:34 +1100 [thread overview]
Message-ID: <200810191414.34558.nickpiggin@yahoo.com.au> (raw)
In-Reply-To: <alpine.LFD.2.00.0810181220510.3438@nehalem.linux-foundation.org>
On Sunday 19 October 2008 06:31, Linus Torvalds wrote:
> On Sat, 18 Oct 2008, Keith Packard wrote:
> > The basic plan is to have four new functions (yes, I'm making up names
> > here):
> >
> > struct io_mapping *io_reserve_pci_resource(struct pci_dev *dev,
> > int bar,
> > int prot);
> > void io_mapping_free(struct io_mapping *mapping);
> >
> > void *io_map_atomic(struct io_mapping *mapping, unsigned long pfn);
> > void io_unmap_atomic(struct io_mapping *mapping, unsigned long pfn);
>
> The important thing is that mappings need to be per-CPU, so the above may
> work, but only if it's designed so that "io_reserve_pci_resource()" will
> actually reserve space for 'nr_possible_cpu' page mappings, and then the
> "io_[un]map_atomic()" functions do per-CPU mappings.
>
> Anything else is a disaster, because anything else implies TLB shootdown.
TLB shootdown need only be implied if the behaviour required is to
unmap the virtual address *and* cause any other CPU that subsequently
touches it to fault.
For kva, that would be a bug anyway (use after free). The only thing
it implies is that a TLB shootdown happens at some point before the
address get reused.
Still, it's always going to be faster than a global mapping, if done
properly. I was thinking about doing a vmap_atomic thing generically
in the vmap layer... why exactly do we need the FIXMAP stuff for it?
prev parent reply other threads:[~2008-10-19 3:14 UTC|newest]
Thread overview: 75+ messages / expand[flat|nested] mbox.gz Atom feed top
2008-10-17 21:29 [git pull] drm patches for 2.6.27-rc1 Dave Airlie
2008-10-17 22:43 ` Linus Torvalds
2008-10-18 2:10 ` Eric Anholt
2008-10-18 2:47 ` Linus Torvalds
2008-10-18 3:49 ` Keith Packard
2008-10-18 6:44 ` Corbin Simpson
2008-10-18 7:49 ` Eric Anholt
2008-10-19 17:52 ` Peter Zijlstra
2008-10-20 4:17 ` Steven J Newbury
2008-10-20 16:31 ` Linus Torvalds
2008-10-20 20:04 ` Jesse Barnes
2008-10-18 9:11 ` Dave Airlie
2008-10-18 1:37 ` Nick Piggin
2008-10-18 19:11 ` Keith Packard
2008-10-18 19:31 ` Linus Torvalds
2008-10-18 20:07 ` Thomas Hellström
2008-10-18 20:20 ` Keith Packard
2008-10-18 20:37 ` Ingo Molnar
2008-10-18 21:51 ` Keith Packard
2008-10-18 22:32 ` Ingo Molnar
2008-10-18 22:47 ` Jon Smirl
2008-10-18 22:53 ` Linus Torvalds
2008-10-19 0:38 ` Keith Packard
2008-10-19 1:06 ` Arjan van de Ven
2008-10-19 1:15 ` Keith Packard
2008-10-20 10:01 ` Ingo Molnar
2008-10-19 4:14 ` Keith Packard
2008-10-19 6:41 ` Keith Packard
2008-10-19 17:53 ` io resources and cached mappings (was: [git pull] drm patches for 2.6.27-rc1) Ingo Molnar
2008-10-19 18:00 ` Arjan van de Ven
2008-10-19 19:07 ` Eric Anholt
2008-10-20 11:55 ` Ingo Molnar
2008-10-20 12:20 ` Ingo Molnar
2008-10-19 21:04 ` Keith Packard
2008-10-20 11:58 ` Ingo Molnar
2008-10-20 15:49 ` Keith Packard
2008-10-22 9:36 ` Ingo Molnar
2008-10-23 7:14 ` Keith Packard
2008-10-23 7:14 ` [PATCH] Add io-mapping functions to dynamically map large device apertures Keith Packard
2008-10-23 7:14 ` [PATCH] [drm/i915] Use io-mapping interfaces instead of a variety of mapping kludges Keith Packard
2008-10-24 4:49 ` [PATCH] Add io-mapping functions to dynamically map large device apertures Randy Dunlap
2008-10-24 6:26 ` Keith Packard
2008-10-23 8:05 ` io resources and cached mappings (was: [git pull] drm patches for 2.6.27-rc1) Ingo Molnar
2008-10-23 15:39 ` Keith Packard
2008-11-03 7:00 ` Dave Airlie
2008-11-03 10:48 ` Ingo Molnar
2008-11-03 16:36 ` Linus Torvalds
2008-11-03 16:53 ` Ingo Molnar
2008-11-03 17:29 ` [git pull] IO mappings, #2 Ingo Molnar
2008-11-04 22:36 ` Jonathan Corbet
2008-11-05 9:01 ` Ingo Molnar
2008-10-23 20:22 ` Adding kmap_atomic_prot_pfn (was: [git pull] drm patches for 2.6.27-rc1) Keith Packard
2008-10-23 20:38 ` Andrew Morton
2008-10-23 21:03 ` Keith Packard
2008-10-23 21:24 ` Linus Torvalds
2008-10-24 1:50 ` Keith Packard
2008-10-24 2:48 ` Linus Torvalds
2008-10-24 3:24 ` Benjamin Herrenschmidt
2008-10-24 5:37 ` Keith Packard
2008-10-24 14:53 ` Linus Torvalds
2008-10-24 15:45 ` Keith Packard
2008-10-24 4:29 ` Keith Packard
2008-10-24 6:22 ` Keith Packard
2008-10-24 7:33 ` Adding kmap_atomic_prot_pfn Thomas Hellström
2008-10-24 8:38 ` Ingo Molnar
2008-10-24 9:19 ` Thomas Hellström
2008-10-24 9:32 ` Ingo Molnar
2008-10-24 11:04 ` Thomas Hellström
2008-10-24 15:48 ` Keith Packard
2008-10-24 10:18 ` Thomas Hellström
2008-10-24 9:14 ` Adding kmap_atomic_prot_pfn (was: [git pull] drm patches for 2.6.27-rc1) Ingo Molnar
2008-10-24 3:21 ` Benjamin Herrenschmidt
2008-10-20 10:10 ` io resources and cached mappings " Ingo Molnar
2008-10-19 4:28 ` [git pull] drm patches for 2.6.27-rc1 Yinghai Lu
2008-10-19 3:14 ` Nick Piggin [this message]
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=200810191414.34558.nickpiggin@yahoo.com.au \
--to=nickpiggin@yahoo.com.au \
--cc=airlied@linux.ie \
--cc=akpm@linux-foundation.org \
--cc=dri-devel@lists.sf.net \
--cc=keithp@keithp.com \
--cc=linux-kernel@vger.kernel.org \
--cc=torvalds@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).