Linux IOMMU Development
 help / color / mirror / Atom feed
From: Baolu Lu <baolu.lu@linux.intel.com>
To: Yi Liu <yi.l.liu@intel.com>,
	joro@8bytes.org, jgg@nvidia.com, kevin.tian@intel.com
Cc: baolu.lu@linux.intel.com, eric.auger@redhat.com,
	nicolinc@nvidia.com, chao.p.peng@linux.intel.com,
	iommu@lists.linux.dev, vasant.hegde@amd.com, will@kernel.org
Subject: Re: [PATCH v6 09/14] iommu/vt-d: Add IOMMU_HWPT_ALLOC_PASID support
Date: Wed, 25 Dec 2024 15:13:04 +0800	[thread overview]
Message-ID: <5c5180cb-9bbe-4fa7-b109-5dad2ca9516a@linux.intel.com> (raw)
In-Reply-To: <c10755cb-6966-479b-b68c-9bc022c894d9@intel.com>

On 2024/12/25 12:30, Yi Liu wrote:
> On 2024/12/25 09:02, Baolu Lu wrote:
>> On 12/24/24 19:35, Yi Liu wrote:
>>>> Another related consideration is the support for page faults in nested
>>>> domains once PASID is available in user space. Would it be 
>>>> reasonable to
>>>> support page faults for nested domains?
>>>
>>> yeah, it's good to discuss it.
>>>
>>>> If so, perhaps it's time to open intel_iommu_domain_alloc_nested() to
>>>> support IOMMU_HWPT_FAULT_ID_VALID when PRI is enabled on the device?
>>>
>>> IMHO, the PRQ support for nested domains requires some more facility. 
>>> PRQ
>>> can happen at either stage-1 or stage-2, iommu driver may need to tell
>>> it and forward to the correct domain (nested domain or parent 
>>> domain). Or
>>> the stage-2 is always pinned just as the VFIO/iommufd does. Hence, 
>>> any PRQ
>>> happens under nested translation should be due to stage-1.
>>
>> The parent domain is currently always pinned and does not yet support
>> page faults. Therefore, when a page fault occurs within a nested domain,
>> it apparently should be routed to user space...
>>
>> If we decide to support page faults on the stage-2 domain in the future,
>> we will need to figure out the correct destination of each page fault
>> and route it accordingly, either to the parent domain or the user space
>> nested domain. Hardware assistance would be beneficial, otherwise the
>> software may need to traverse the parent domain, which is not
>> performance friendly.
> 
> this is my question. Is it still true the parent domain is always pinned
> after the below series? If yes, then it's fine to enable IOPF for nested
> domain.
> 
> https://lore.kernel.org/linux-iommu/20241015-jag-iopfv8-v4-0- 
> b696ca89ba29@kernel.org/

Then, perhaps we could enforce this in iommufd for a short-term purpose?

When allocating a hwpt in iommufd, we should enforce that the flags
IOMMU_HWPT_ALLOC_NEST_PARENT and IOMMU_HWPT_FAULT_ID_VALID cannot be set
at the same time.

---
baolu

  reply	other threads:[~2024-12-25  7:13 UTC|newest]

Thread overview: 71+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-12-19 13:27 [PATCH v6 00/14] iommufd support pasid attach/replace Yi Liu
2024-12-19 13:27 ` [PATCH v6 01/14] iommu: Introduce a replace API for device pasid Yi Liu
2024-12-20  2:47   ` Baolu Lu
2025-01-09  7:08   ` Tian, Kevin
2025-01-09  7:20   ` Tian, Kevin
2025-01-09 14:43     ` Jason Gunthorpe
2025-01-10  2:31       ` Baolu Lu
2025-01-10  7:21         ` Tian, Kevin
2025-01-16 10:00           ` Yi Liu
2025-01-13 20:21   ` Jason Gunthorpe
2025-01-14  8:10     ` Tian, Kevin
2025-01-14 13:45       ` Jason Gunthorpe
2025-01-15  4:43         ` Tian, Kevin
2025-01-15 14:43           ` Jason Gunthorpe
2025-01-16  5:48             ` Tian, Kevin
2025-01-17 10:32               ` Yi Liu
2024-12-19 13:27 ` [PATCH v6 02/14] iommufd: Refactor __fault_domain_replace_dev() to be a wrapper of iommu_replace_group_handle() Yi Liu
2024-12-19 13:27 ` [PATCH v6 03/14] iommufd: Move the iommufd_handle helpers to device.c Yi Liu
2024-12-20  3:31   ` Baolu Lu
2024-12-20  6:34     ` Yi Liu
2024-12-19 13:27 ` [PATCH v6 04/14] iommufd: Always pass iommu_attach_handle to iommu core Yi Liu
2024-12-20  4:35   ` Nicolin Chen
2024-12-20  6:40     ` Yi Liu
2024-12-20  6:58       ` Nicolin Chen
2025-01-09  7:44   ` Tian, Kevin
2025-01-17 12:33     ` Yi Liu
2025-01-17 19:03       ` Nicolin Chen
2024-12-19 13:27 ` [PATCH v6 05/14] iommufd: Pass pasid through the device attach/replace path Yi Liu
2025-01-09  7:53   ` Tian, Kevin
2025-01-09 14:51     ` Jason Gunthorpe
2025-01-10  7:22       ` Tian, Kevin
2024-12-19 13:27 ` [PATCH v6 06/14] iommufd: Mark PASID-compatible domain Yi Liu
2025-01-09  7:56   ` Tian, Kevin
2025-01-09 14:54   ` Jason Gunthorpe
2025-01-17 10:50     ` Yi Liu
2024-12-19 13:27 ` [PATCH v6 07/14] iommufd: Support pasid attach/replace Yi Liu
2025-01-09  8:25   ` Tian, Kevin
2024-12-19 13:27 ` [PATCH v6 08/14] iommufd: Enforce PASID-compatible domain for RID Yi Liu
2025-01-09  8:31   ` Tian, Kevin
2024-12-19 13:27 ` [PATCH v6 09/14] iommu/vt-d: Add IOMMU_HWPT_ALLOC_PASID support Yi Liu
2024-12-23  2:51   ` Baolu Lu
2024-12-24 11:35     ` Yi Liu
2024-12-25  1:02       ` Baolu Lu
2024-12-25  4:30         ` Yi Liu
2024-12-25  7:13           ` Baolu Lu [this message]
2025-02-12  7:47             ` Yi Liu
2025-02-12 12:59               ` Jason Gunthorpe
2025-02-13  9:34                 ` Yi Liu
2025-02-13 12:56                   ` Jason Gunthorpe
2025-02-14  3:24                     ` Yi Liu
2025-02-12 13:00               ` Robin Murphy
2025-02-12 13:08                 ` Jason Gunthorpe
2025-02-13 10:10                 ` Yi Liu
2025-02-13 10:24                   ` Robin Murphy
2025-02-13 12:53                     ` Yi Liu
2025-02-19  8:02                     ` Tian, Kevin
2025-02-19 12:50                       ` Yi Liu
2025-02-20  6:57                         ` Tian, Kevin
2025-01-09 15:27     ` Jason Gunthorpe
2025-01-10  2:41       ` Baolu Lu
2025-01-10  7:34         ` Tian, Kevin
2025-01-17 10:57           ` Yi Liu
2025-01-10  7:38   ` Tian, Kevin
2025-01-14  8:13     ` Tian, Kevin
2025-01-13 20:31   ` Jason Gunthorpe
2025-01-14  8:19     ` Tian, Kevin
2024-12-19 13:27 ` [PATCH v6 10/14] iommufd: Allow allocating PASID-compatible domain Yi Liu
2024-12-19 13:27 ` [PATCH v6 11/14] iommufd/selftest: Add set_dev_pasid in mock iommu Yi Liu
2024-12-19 13:27 ` [PATCH v6 12/14] iommufd/selftest: Add a helper to get test device Yi Liu
2024-12-19 13:27 ` [PATCH v6 13/14] iommufd/selftest: Add test ops to test pasid attach/detach Yi Liu
2024-12-19 13:27 ` [PATCH v6 14/14] iommufd/selftest: Add coverage for iommufd " Yi Liu

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=5c5180cb-9bbe-4fa7-b109-5dad2ca9516a@linux.intel.com \
    --to=baolu.lu@linux.intel.com \
    --cc=chao.p.peng@linux.intel.com \
    --cc=eric.auger@redhat.com \
    --cc=iommu@lists.linux.dev \
    --cc=jgg@nvidia.com \
    --cc=joro@8bytes.org \
    --cc=kevin.tian@intel.com \
    --cc=nicolinc@nvidia.com \
    --cc=vasant.hegde@amd.com \
    --cc=will@kernel.org \
    --cc=yi.l.liu@intel.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