From: Teddy Astie <teddy.astie@vates.tech>
To: Jason Gunthorpe <jgg@ziepe.ca>
Cc: xen-devel@lists.xenproject.org, iommu@lists.linux.dev,
"Juergen Gross" <jgross@suse.com>,
"Stefano Stabellini" <sstabellini@kernel.org>,
"Oleksandr Tyshchenko" <oleksandr_tyshchenko@epam.com>,
"Joerg Roedel" <joro@8bytes.org>, "Will Deacon" <will@kernel.org>,
"Robin Murphy" <robin.murphy@arm.com>,
"Marek Marczykowski-Górecki" <marmarek@invisiblethingslab.com>
Subject: Re: [RFC PATCH] iommu/xen: Add Xen PV-IOMMU driver
Date: Fri, 21 Jun 2024 15:09:08 +0000 [thread overview]
Message-ID: <750967b7-252f-4523-872f-64b79358c97c@vates.tech> (raw)
In-Reply-To: <20240619163000.GK791043@ziepe.ca>
Hello Jason,
Le 19/06/2024 à 18:30, Jason Gunthorpe a écrit :
> On Thu, Jun 13, 2024 at 01:50:22PM +0000, Teddy Astie wrote:
>
>> +struct iommu_domain *xen_iommu_domain_alloc(unsigned type)
>> +{
>> + struct xen_iommu_domain *domain;
>> + u16 ctx_no;
>> + int ret;
>> +
>> + if (type & IOMMU_DOMAIN_IDENTITY) {
>> + /* use default domain */
>> + ctx_no = 0;
>
> Please use the new ops, domain_alloc_paging and the static identity domain.
Yes, in the v2, I will use this newer interface.
I have a question on this new interface : is it valid to not have a
identity domain (and "default domain" being blocking); well in the
current implementation it doesn't really matter, but at some point, we
may want to allow not having it (thus making this driver mandatory).
>
>> +static struct iommu_group *xen_iommu_device_group(struct device *dev)
>> +{
>> + if (!dev_is_pci(dev))
>> + return ERR_PTR(-ENODEV);
>> +
>
> device_group is only called after probe_device, since you already
> exclude !pci during probe there is no need for this wrapper, just set
> the op directly to pci_device_group.
>
>> + if (!dev_is_pci(dev))
>> + return;
>
> No op is ever called on a non-probed device, remove all these checks.
>
>
> A paging domain should be the only domain ops that have a populated
> map so this should be made impossible by construction
Makes sense, will remove these redundant checks in v2.
>
>> +static void xen_iommu_release_device(struct device *dev)
>> +{
>> + int ret;
>> + struct pci_dev *pdev;
>> + struct pv_iommu_op op = {
>> + .subop_id = IOMMUOP_reattach_device,
>> + .flags = 0,
>> + .ctx_no = 0 /* reattach device back to default context */
>> + };
>
> Consider if you can use release_domain for this, I think this is
> probably a BLOCKED domain behavior.
The goal is to put back all devices where they were at the beginning
(the default "context"), which is what release_domain looks like it is
doing. Will use it for v2.
>
> Jason
Teddy
Teddy Astie | Vates XCP-ng Intern
XCP-ng & Xen Orchestra - Vates solutions
web: https://vates.tech
next prev parent reply other threads:[~2024-06-21 15:09 UTC|newest]
Thread overview: 10+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-06-13 13:50 [RFC PATCH] iommu/xen: Add Xen PV-IOMMU driver Teddy Astie
2024-06-13 14:32 ` Jan Beulich
2024-06-17 13:36 ` Teddy Astie
2024-06-17 13:36 ` Teddy Astie
2024-06-17 13:43 ` Jan Beulich
2024-06-19 16:30 ` Jason Gunthorpe
2024-06-21 15:09 ` Teddy Astie [this message]
2024-06-23 3:21 ` Baolu Lu
2024-06-24 11:09 ` Robin Murphy
2024-06-25 1:18 ` Baolu 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=750967b7-252f-4523-872f-64b79358c97c@vates.tech \
--to=teddy.astie@vates.tech \
--cc=iommu@lists.linux.dev \
--cc=jgg@ziepe.ca \
--cc=jgross@suse.com \
--cc=joro@8bytes.org \
--cc=marmarek@invisiblethingslab.com \
--cc=oleksandr_tyshchenko@epam.com \
--cc=robin.murphy@arm.com \
--cc=sstabellini@kernel.org \
--cc=will@kernel.org \
--cc=xen-devel@lists.xenproject.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.