From: Baolu Lu <baolu.lu@linux.intel.com>
To: "Tian, Kevin" <kevin.tian@intel.com>,
"iommu@lists.linux.dev" <iommu@lists.linux.dev>
Cc: baolu.lu@linux.intel.com, Joerg Roedel <joro@8bytes.org>,
Will Deacon <will@kernel.org>,
Robin Murphy <robin.murphy@arm.com>,
"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH 1/1] iommu/vt-d: Enable PASID during iommu device probe
Date: Thu, 15 Sep 2022 15:21:18 +0800 [thread overview]
Message-ID: <78da3b7e-c930-f433-8c80-b1ac53f9ab58@linux.intel.com> (raw)
In-Reply-To: <BN9PR11MB527688C87CEA57CACD5324608C499@BN9PR11MB5276.namprd11.prod.outlook.com>
On 2022/9/15 11:22, Tian, Kevin wrote:
>> From: Baolu Lu <baolu.lu@linux.intel.com>
>> Sent: Tuesday, September 13, 2022 5:25 PM
>>
>> On 2022/9/13 16:01, Tian, Kevin wrote:
>>>> From: Lu Baolu <baolu.lu@linux.intel.com>
>>>> Sent: Monday, September 12, 2022 10:48 AM
>>>>
>>>> @@ -1401,7 +1403,6 @@ static void iommu_enable_dev_iotlb(struct
>>>> device_domain_info *info)
>>>
>>> This is not the right name now as dev_iotlb is only related to ATS.
>>
>> Yes. This name is confusing. Perhaps we can split it into some specific
>> helpers,
>>
>> - intel_iommu_enable_pci_ats()
>> - intel_iommu_enabel_pci_pri()
>> - intel_iommu_enable_pci_pasid()
>> ?
>
> Probably intel_iommu_enable_pci_caps()
It's better.
>
>>
>>>
>>>> info->pfsid = pci_dev_id(pf_pdev);
>>>> }
>>>>
>>>> -#ifdef CONFIG_INTEL_IOMMU_SVM
>>>> /* The PCIe spec, in its wisdom, declares that the behaviour of
>>>> the device if you enable PASID support after ATS support is
>>>> undefined. So always enable PASID support on devices which
>>>> @@ -1414,7 +1415,7 @@ static void iommu_enable_dev_iotlb(struct
>>>> device_domain_info *info)
>>>> (info->pasid_enabled ? pci_prg_resp_pasid_required(pdev) : 1)
>>>> &&
>>>> !pci_reset_pri(pdev) && !pci_enable_pri(pdev, PRQ_DEPTH))
>>>> info->pri_enabled = 1;
>>>> -#endif
>>>> +
>>>> if (info->ats_supported && pci_ats_page_aligned(pdev) &&
>>>> !pci_enable_ats(pdev, VTD_PAGE_SHIFT)) {
>>>> info->ats_enabled = 1;
>>>
>>> iommu_enable_dev_iotlb() is currently called both when the device is
>> probed
>>> and when sva is enabled (which is actually useless). From this angle the
>> commit
>>> msg is inaccurate.
>>
>> The logic is a bit tricky. iommu_support_dev_iotlb() only returns a
>> devinfo pointer when ATS is supported on the device. So, you are right
>> if device supports both ATS and PASID; otherwise PASID will not be
>> enabled.
>
> Yes, that is what the first part of this patch fixes.
>
> But my point is about the message that previously PASID was enabled
> only when FEAT_SVA is enabled and now the patch moves it to the
> probe time.
>
> My point is that even in old way iommu_enable_dev_iotlb() was called
> twice: one at probe time and the other at FEAT_SVA. If ATS exists
> then PASID is enabled at probe time already. If ATS doesn't exist then
> PASID is always disabled.
>
> So this patch is really to decouple PASID enabling from ATS and remove
> the unnecessary/duplicated call of iommu_enable_dev_iotlb() in
> intel_iommu_enable_sva().
Yes. Exactly. I will rephrase the commit message and send a v2.
Best regards,
baolu
prev parent reply other threads:[~2022-09-15 7:21 UTC|newest]
Thread overview: 15+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-09-12 2:48 [PATCH 1/1] iommu/vt-d: Enable PASID during iommu device probe Lu Baolu
2022-09-13 3:13 ` Tian, Kevin
2022-09-13 6:01 ` Baolu Lu
2022-09-13 7:42 ` Tian, Kevin
2022-09-13 7:46 ` Ethan Zhao
2022-09-13 9:30 ` Baolu Lu
[not found] ` <e26efaee-d84a-3b60-8400-90d8e49a9b25@linux.intel.com>
2022-09-15 3:00 ` Baolu Lu
2022-09-16 2:40 ` Ethan Zhao
2022-09-16 3:05 ` Baolu Lu
2022-09-16 3:37 ` Ethan Zhao
[not found] ` <a6ac5953-7372-9894-58d4-d1ee2905d4dd@linux.intel.com>
2022-09-16 4:01 ` Baolu Lu
2022-09-13 8:01 ` Tian, Kevin
2022-09-13 9:25 ` Baolu Lu
2022-09-15 3:22 ` Tian, Kevin
2022-09-15 7:21 ` Baolu Lu [this message]
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=78da3b7e-c930-f433-8c80-b1ac53f9ab58@linux.intel.com \
--to=baolu.lu@linux.intel.com \
--cc=iommu@lists.linux.dev \
--cc=joro@8bytes.org \
--cc=kevin.tian@intel.com \
--cc=linux-kernel@vger.kernel.org \
--cc=robin.murphy@arm.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).