From: Joerg Roedel <joro@8bytes.org>
To: Sai Praneeth Prakhya <sai.praneeth.prakhya@intel.com>
Cc: Ashok Raj <ashok.raj@intel.com>,
Will Deacon <will.deacon@arm.com>,
iommu@lists.linux-foundation.org,
Robin Murphy <robin.murphy@arm.com>,
Christoph Hellwig <hch@lst.de>
Subject: Re: [PATCH V2 3/5] iommu: Add support to change default domain of an iommu_group
Date: Mon, 2 Mar 2020 16:08:34 +0100 [thread overview]
Message-ID: <20200302150833.GA6540@8bytes.org> (raw)
In-Reply-To: <5aa5ef20ff81f706aafa9a6af68cef98fe60ad0f.1581619464.git.sai.praneeth.prakhya@intel.com>
Hello Sai, Baolu,
On Sun, Feb 16, 2020 at 01:57:26PM -0800, Sai Praneeth Prakhya wrote:
> Hence it will be helpful if there is some way to change the default
> domain of a B:D.F dynamically. Since, linux iommu subsystem prefers to
> deal at iommu_group level instead of B:D.F level, it might be helpful
> if there is some way to change the default domain of a *iommu_group*
> dynamically. Hence, add such support.
The question is how this plays together with the per-device private
domains in the Intel VT-d driver. I recently debugged an issue there and
I think there are more. The overall code for this seems to be pretty
fragile, so I had the idea to make the private default domains more
general.
IOMMU default domains don't necessarily need to stick to the iommu-group
granularity, because the default domain is used by in-kernel drivers
only, and the kernel trusts itself.
So my idea was to make the private-domain concept of the VT-d driver
more generic and move it to the iommu core code. With that we can
configure real per-device default domain types and don't have the race
condition with driver probing when changing the default domain of
multiple devices. We have to limit the ability to change default domain
types to devices with no PCI aliases, but that should not be a problem
for the intended use-case.
What do you think?
Regards,
Joerg
_______________________________________________
iommu mailing list
iommu@lists.linux-foundation.org
https://lists.linuxfoundation.org/mailman/listinfo/iommu
next prev parent reply other threads:[~2020-03-02 15:08 UTC|newest]
Thread overview: 28+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-02-16 21:57 [PATCH V2 0/5] iommu: Add support to change default domain of a group Sai Praneeth Prakhya
2020-02-16 21:57 ` [PATCH V2 1/5] iommu: Add dev_def_domain_type() call back function to iommu_ops Sai Praneeth Prakhya
2020-02-22 23:37 ` Lu Baolu
2020-02-22 23:39 ` Prakhya, Sai Praneeth
2020-02-16 21:57 ` [PATCH V2 2/5] iommu/vt-d: Rename device_def_domain_type() to intel_iommu_dev_def_domain_type() Sai Praneeth Prakhya
2020-02-22 23:42 ` Lu Baolu
2020-02-22 23:59 ` Prakhya, Sai Praneeth
2020-02-23 1:50 ` Lu Baolu
2020-02-24 3:23 ` Prakhya, Sai Praneeth
2020-02-16 21:57 ` [PATCH V2 3/5] iommu: Add support to change default domain of an iommu_group Sai Praneeth Prakhya
2020-02-23 1:20 ` Lu Baolu
2020-02-24 3:20 ` Prakhya, Sai Praneeth
2020-02-24 5:46 ` Lu Baolu
2020-02-24 7:03 ` Prakhya, Sai Praneeth
2020-02-24 7:39 ` Lu Baolu
2020-02-24 7:57 ` Prakhya, Sai Praneeth
2020-02-24 8:12 ` Lu Baolu
2020-02-24 8:39 ` Lu Baolu
2020-02-24 8:44 ` Prakhya, Sai Praneeth
2020-03-02 15:08 ` Joerg Roedel [this message]
2020-03-03 6:47 ` Lu Baolu
2020-03-03 13:13 ` Joerg Roedel
2020-03-04 12:17 ` Lu Baolu
2020-02-16 21:57 ` [PATCH V2 4/5] iommu: Take lock before reading iommu_group default domain type Sai Praneeth Prakhya
2020-02-16 21:57 ` [PATCH V2 5/5] iommu: Document usage of "/sys/kernel/iommu_groups/<grp_id>/type" file Sai Praneeth Prakhya
2020-02-23 1:38 ` Lu Baolu
2020-02-24 2:18 ` Prakhya, Sai Praneeth
2020-02-22 23:40 ` [PATCH V2 0/5] iommu: Add support to change default domain of a group Prakhya, Sai Praneeth
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=20200302150833.GA6540@8bytes.org \
--to=joro@8bytes.org \
--cc=ashok.raj@intel.com \
--cc=hch@lst.de \
--cc=iommu@lists.linux-foundation.org \
--cc=robin.murphy@arm.com \
--cc=sai.praneeth.prakhya@intel.com \
--cc=will.deacon@arm.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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox