From: Jeremy Fitzhardinge <jeremy@goop.org>
To: Avi Kivity <avi@qumranet.com>
Cc: Ingo Molnar <mingo@elte.hu>,
Xen Devel <xen-devel@lists.xensource.com>,
Andi Kleen <andi@firstfloor.org>,
linux-kernel@vger.kernel.org, "H. Peter Anvin" <hpa@zytor.com>
Subject: Re: [Xen-devel] Re: [PATCH 1 of 7] x86: add _PAGE_IOMAP pte flag for IO mappings
Date: Wed, 10 Sep 2008 09:38:36 -0700 [thread overview]
Message-ID: <48C7F80C.6000807@goop.org> (raw)
In-Reply-To: <48C7998D.1030408@qumranet.com>
Avi Kivity wrote:
> Actually it's more of a "no struct page" flag, which implies no
> refcounting.
Hm, is that actually true enough to define it? Could we rename it
something like _PAGE_NOSTRUCTPAGE or something a bit more specific than
"special"?
> And not having a struct page should correspond well to a pte not
> requiring pfn->mfn conversion and being an I/O page.
But _PAGE_SPECIAL is only set in a few places. It's not set in ioremap
mappings and so on. Should it be?
There's also the hiccup that it gets set in a pte with pte_mkspecial() -
but at that point its too late because you've already constructed the
pte and done the pfn->mfn conversion. _PAGE_IOMAP can only be set when
you initially construct the pte out of a frame number and a pgprot.
J
WARNING: multiple messages have this Message-ID (diff)
From: Jeremy Fitzhardinge <jeremy@goop.org>
To: Avi Kivity <avi@qumranet.com>
Cc: Ingo Molnar <mingo@elte.hu>,
Xen Devel <xen-devel@lists.xensource.com>,
Andi Kleen <andi@firstfloor.org>,
linux-kernel@vger.kernel.org, "H. Peter Anvin" <hpa@zytor.com>
Subject: Re: Re: [PATCH 1 of 7] x86: add _PAGE_IOMAP pte flag for IO mappings
Date: Wed, 10 Sep 2008 09:38:36 -0700 [thread overview]
Message-ID: <48C7F80C.6000807@goop.org> (raw)
In-Reply-To: <48C7998D.1030408@qumranet.com>
Avi Kivity wrote:
> Actually it's more of a "no struct page" flag, which implies no
> refcounting.
Hm, is that actually true enough to define it? Could we rename it
something like _PAGE_NOSTRUCTPAGE or something a bit more specific than
"special"?
> And not having a struct page should correspond well to a pte not
> requiring pfn->mfn conversion and being an I/O page.
But _PAGE_SPECIAL is only set in a few places. It's not set in ioremap
mappings and so on. Should it be?
There's also the hiccup that it gets set in a pte with pte_mkspecial() -
but at that point its too late because you've already constructed the
pte and done the pfn->mfn conversion. _PAGE_IOMAP can only be set when
you initially construct the pte out of a frame number and a pgprot.
J
next prev parent reply other threads:[~2008-09-10 16:38 UTC|newest]
Thread overview: 53+ messages / expand[flat|nested] mbox.gz Atom feed top
2008-09-07 22:21 [PATCH 0 of 7] x86: lay groundwork for Xen domain 0 support Jeremy Fitzhardinge
2008-09-07 22:21 ` [PATCH 1 of 7] x86: add _PAGE_IOMAP pte flag for IO mappings Jeremy Fitzhardinge
2008-09-09 13:32 ` Avi Kivity
2008-09-09 13:32 ` Avi Kivity
2008-09-09 14:47 ` Jeremy Fitzhardinge
2008-09-09 14:47 ` Jeremy Fitzhardinge
2008-09-09 14:56 ` Avi Kivity
2008-09-09 14:56 ` Avi Kivity
2008-09-09 15:29 ` [Xen-devel] " Keir Fraser
2008-09-09 15:29 ` Keir Fraser
2008-09-09 15:48 ` [Xen-devel] " Jeremy Fitzhardinge
2008-09-09 15:48 ` Jeremy Fitzhardinge
2008-09-09 16:05 ` [Xen-devel] " Keir Fraser
2008-09-09 16:05 ` Keir Fraser
2008-09-10 9:55 ` [Xen-devel] " Avi Kivity
2008-09-10 9:55 ` Avi Kivity
2008-09-10 16:38 ` Jeremy Fitzhardinge [this message]
2008-09-10 16:38 ` Jeremy Fitzhardinge
2008-09-10 16:55 ` [Xen-devel] " Nick Piggin
2008-09-10 17:27 ` Jeremy Fitzhardinge
2008-09-07 22:21 ` [PATCH 2 of 7] x86: remove duplicate early_ioremap declarations Jeremy Fitzhardinge
2008-09-07 22:21 ` Jeremy Fitzhardinge
2008-09-07 22:21 ` [PATCH 3 of 7] x86: add early_memremap() Jeremy Fitzhardinge
2008-09-07 22:21 ` Jeremy Fitzhardinge
2008-09-07 22:21 ` [PATCH 4 of 7] x86: use early_memremap() in setup.c Jeremy Fitzhardinge
2008-09-07 22:21 ` [PATCH 5 of 7] x86-64: don't check for map replacement Jeremy Fitzhardinge
2008-09-07 22:21 ` [PATCH 6 of 7] x86: use early_ioremap in __acpi_map_table Jeremy Fitzhardinge
2008-09-07 23:44 ` Andi Kleen
2008-09-08 0:03 ` Jeremy Fitzhardinge
2008-09-08 0:03 ` Jeremy Fitzhardinge
2008-09-08 14:26 ` Ingo Molnar
2008-09-08 14:26 ` Ingo Molnar
2008-09-08 16:29 ` Jeremy Fitzhardinge
2008-09-08 16:29 ` Jeremy Fitzhardinge
2008-09-08 19:41 ` Jeremy Fitzhardinge
2008-09-08 19:41 ` Jeremy Fitzhardinge
2008-09-10 11:55 ` Ingo Molnar
2008-09-10 16:49 ` Jeremy Fitzhardinge
2008-09-11 7:33 ` Ingo Molnar
2008-09-11 18:36 ` Jeremy Fitzhardinge
2008-09-11 18:56 ` Ingo Molnar
2008-09-11 20:34 ` Yinghai Lu
2008-09-11 21:07 ` Jeremy Fitzhardinge
2008-09-12 9:49 ` Ingo Molnar
2008-09-12 17:31 ` Yinghai Lu
[not found] ` <20080911125748.GA14698@elte.hu>
2008-09-11 21:33 ` [PATCH] acpi: remove final __acpi_map_table mapping before setting acpi_gbl_permanent_mmap Jeremy Fitzhardinge
2008-09-07 22:21 ` [PATCH 7 of 7] x86: always explicitly map acpi memory Jeremy Fitzhardinge
2008-09-07 23:35 ` [Xen-devel] " Yinghai Lu
2008-09-07 23:35 ` Yinghai Lu
2008-09-08 0:02 ` [Xen-devel] " Jeremy Fitzhardinge
2008-09-08 0:02 ` Jeremy Fitzhardinge
2008-09-08 0:14 ` [Xen-devel] " Yinghai Lu
2008-09-08 0:14 ` Yinghai Lu
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=48C7F80C.6000807@goop.org \
--to=jeremy@goop.org \
--cc=andi@firstfloor.org \
--cc=avi@qumranet.com \
--cc=hpa@zytor.com \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@elte.hu \
--cc=xen-devel@lists.xensource.com \
/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.