From: Niklas Schnelle <schnelle@linux.ibm.com>
To: Jason Gunthorpe <jgg@nvidia.com>, Robin Murphy <robin.murphy@arm.com>
Cc: Baolu Lu <baolu.lu@linux.intel.com>,
Matthew Rosato <mjrosato@linux.ibm.com>,
Gerd Bayer <gbayer@linux.ibm.com>,
iommu@lists.linux.dev, Joerg Roedel <joro@8bytes.org>,
Will Deacon <will@kernel.org>,
Wenjia Zhang <wenjia@linux.ibm.com>,
Pierre Morel <pmorel@linux.ibm.com>,
linux-s390@vger.kernel.org, borntraeger@linux.ibm.com,
hca@linux.ibm.com, gor@linux.ibm.com,
gerald.schaefer@linux.ibm.com, agordeev@linux.ibm.com,
svens@linux.ibm.com, linux-kernel@vger.kernel.org,
Julian Ruess <julianr@linux.ibm.com>
Subject: Re: [PATCH v2 4/7] iommu: Let iommu.strict override ops->def_domain_type
Date: Mon, 05 Dec 2022 16:34:08 +0100 [thread overview]
Message-ID: <4b34be4433856d9a3a5bf13dad7a77c1ba93db13.camel@linux.ibm.com> (raw)
In-Reply-To: <Y4Zm53o1ovdIAqr/@nvidia.com>
On Tue, 2022-11-29 at 16:09 -0400, Jason Gunthorpe wrote:
> On Tue, Nov 29, 2022 at 06:41:22PM +0000, Robin Murphy wrote:
> > On 2022-11-29 17:33, Jason Gunthorpe wrote:
> > > On Mon, Nov 28, 2022 at 09:01:43PM +0000, Robin Murphy wrote:
> > >
> > > > I'm hardly an advocate for trying to save users from themselves, but I
> > > > honestly can't see any justifiable reason for not having sysfs respect
> > > > iommu_get_def_domain_type().
> > >
> > > We really need to rename this value if it is not actually just an
> > > advisory "default" but a functional requirement ..
> >
> > It represents a required default domain type. As in, the type for the
> > device's default domain. Not the default type for a domain. It's the
> > iommu_def_domain_type variable that holds the *default* default domain type
> > ;)
>
> I find the name "default domain" incredibly confusing at this point in
> time.
Yes it definitely confused me as evidenced by this patch.
>
> I would like to call that the "dma-api domain" - its primary purpose
> is to be the domain that the DMA API uses to operate the IOMMU, there
> is little "default" about it. This meshes better with our apis talking
> about ownership and so forth.
>
> So, if the op was called
> get_dma_api_domain_type()
>
> It is pretty clear that it is the exact type of domain that should be
> created to support the DMA API, which is what I think you have been
> describing it is supposed to do?
>
> And with Lu's series we have the set_platform_dma() (Lu perhaps you
> should call this set_platform_dma_api() to re-enforce it is about the
> DMA API, not some nebulous DMA thing)
>
> Which is basically the other way to configure the DMA API for
> operation.
>
> And encapsulating more of the logic to setup and manage the DMA API's
> domain into dma-iommu.c would also be helpful to understanding.
>
> > Which reminds me I should finish that patch undoing my terrible
> > ops->default_domain_ops idea, not least because they are misleadingly
> > unrelated to default domains...
>
> :)
>
> > > It is close to being clear, once we get the last touches of dma-iommu
> > > stuff out of the drivers it should be quite clear
> >
> > Cool, some upheaval of .domain_alloc is next on my hitlist anyway, so that
> > might be a good excuse to upheave it a bit more and streamline the type
> > stuff along the way.
>
> Yes, I think so. I want to tidy things a bit so adding this "user
> space" domain concept is a little nicer
>
> Jason
Ok, so it sounds to me like there is going to be quite a bit of change
in this area. Thus I'm a little unsure however how to proceed here. I
think especially with the proposed multiple queue types in this series
it makes sense for an IOMMU driver to express its preference of a
particular domain type if it supports multiple which clearly isn't what
iommu_get_def_domain_type() is currently supposed to do.
Looking at the current code I don't see a trivial way to integrate this
especially with a lot of reworks going on.
At the moment the preference for IOMMU_DOMAIN_DMA_FQ over
IOMMU_DOMAIN_DMA during initial boot is hard coded in
iommu_subsys_init() before we even know what IOMMU driver will be used.
so can't really access its ops there. On the other hand this decision
may still be rolled back if iommu_dma_init_fq() fails in
iommu_dma_init_domain() so maybe it would make sense to assume a DMA
domain type of IOMMU_DOMAIN_DMA until that point and only then prefer
IOMMU_DOMAIN_DMA_FQ or something else if the driver has a preference.
Maybe it would also make sense not to mix the type of flush queue used
with the domain and maybe the queue type could be passed in via
iommu_setup_dma_ops() -> iommu_dma_init_domain() though I do remember
that there is also a planned rework for that. Any suggestions?
Thanks,
Niklas
next prev parent reply other threads:[~2022-12-05 15:34 UTC|newest]
Thread overview: 39+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-11-16 17:16 [PATCH v2 0/7] iommu/dma: s390 DMA API conversion and optimized IOTLB flushing Niklas Schnelle
2022-11-16 17:16 ` [PATCH v2 1/7] s390/ism: Set DMA coherent mask Niklas Schnelle
2022-11-16 17:16 ` [PATCH v2 2/7] s390/pci: prepare is_passed_through() for dma-iommu Niklas Schnelle
2022-11-16 17:16 ` [PATCH v2 3/7] s390/pci: Use dma-iommu layer Niklas Schnelle
2022-11-28 18:03 ` Robin Murphy
2022-12-19 15:17 ` Niklas Schnelle
2022-11-16 17:16 ` [PATCH v2 4/7] iommu: Let iommu.strict override ops->def_domain_type Niklas Schnelle
2022-11-17 1:55 ` Baolu Lu
2022-11-28 11:10 ` Niklas Schnelle
2022-11-28 13:00 ` Baolu Lu
2022-11-28 13:29 ` Jason Gunthorpe
2022-11-28 15:54 ` Niklas Schnelle
2022-11-28 16:35 ` Jason Gunthorpe
2022-11-28 21:01 ` Robin Murphy
2022-11-29 17:33 ` Jason Gunthorpe
2022-11-29 18:41 ` Robin Murphy
2022-11-29 20:09 ` Jason Gunthorpe
2022-11-30 1:28 ` Baolu Lu
2022-12-05 15:34 ` Niklas Schnelle [this message]
2022-12-06 23:09 ` Jason Gunthorpe
2022-12-07 13:18 ` Baolu Lu
2022-12-07 13:23 ` Jason Gunthorpe
2022-12-07 14:18 ` Robin Murphy
2022-12-07 14:30 ` Jason Gunthorpe
2022-11-28 16:56 ` Robin Murphy
2022-11-16 17:16 ` [PATCH v2 5/7] iommu/dma: Allow a single FQ in addition to per-CPU FQs Niklas Schnelle
2022-11-16 17:16 ` [PATCH v2 6/7] iommu/dma: Enable variable queue size and use larger single queue Niklas Schnelle
2022-11-16 17:16 ` [PATCH v2 7/7] iommu/s390: flush queued IOVAs on RPCIT out of resource indication Niklas Schnelle
2022-11-28 14:52 ` Robin Murphy
2022-11-29 12:00 ` Niklas Schnelle
2022-11-29 12:53 ` Robin Murphy
2022-11-29 14:40 ` Niklas Schnelle
2022-12-02 14:29 ` Niklas Schnelle
2022-12-02 14:42 ` Jason Gunthorpe
2022-12-02 15:12 ` Niklas Schnelle
2022-12-02 15:24 ` Jason Gunthorpe
2022-12-05 18:24 ` Robin Murphy
2022-12-06 10:13 ` Niklas Schnelle
2022-11-29 13:51 ` Matthew Rosato
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=4b34be4433856d9a3a5bf13dad7a77c1ba93db13.camel@linux.ibm.com \
--to=schnelle@linux.ibm.com \
--cc=agordeev@linux.ibm.com \
--cc=baolu.lu@linux.intel.com \
--cc=borntraeger@linux.ibm.com \
--cc=gbayer@linux.ibm.com \
--cc=gerald.schaefer@linux.ibm.com \
--cc=gor@linux.ibm.com \
--cc=hca@linux.ibm.com \
--cc=iommu@lists.linux.dev \
--cc=jgg@nvidia.com \
--cc=joro@8bytes.org \
--cc=julianr@linux.ibm.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-s390@vger.kernel.org \
--cc=mjrosato@linux.ibm.com \
--cc=pmorel@linux.ibm.com \
--cc=robin.murphy@arm.com \
--cc=svens@linux.ibm.com \
--cc=wenjia@linux.ibm.com \
--cc=will@kernel.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).