linux-kernel - Re: [PATCH v2 1/1] PCI: Add translated request only flag for pci_enable

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <3b2c162f-7ba2-5e84-2389-a8686c772467@linux.intel.com>
Date:   Fri, 13 Jan 2023 10:36:37 +0800
From:   Baolu Lu <baolu.lu@...ux.intel.com>
To:     "Tian, Kevin" <kevin.tian@...el.com>,
        Bjorn Helgaas <bhelgaas@...gle.com>,
        "Rodel, Jorg" <jroedel@...e.de>
Cc:     baolu.lu@...ux.intel.com,
        Christian König <christian.koenig@....com>,
        Jason Gunthorpe <jgg@...dia.com>,
        Vasant Hegde <vasant.hegde@....com>,
        "Zhu, Tony" <tony.zhu@...el.com>,
        "linux-pci@...r.kernel.org" <linux-pci@...r.kernel.org>,
        "iommu@...ts.linux.dev" <iommu@...ts.linux.dev>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
        Matt Fagnani <matt.fagnani@...l.net>
Subject: Re: [PATCH v2 1/1] PCI: Add translated request only flag for
 pci_enable_pasid()

On 2023/1/13 10:24, Tian, Kevin wrote:
>> From: Lu Baolu <baolu.lu@...ux.intel.com>
>> Sent: Friday, January 13, 2023 9:44 AM
>>
>> One alternative is ATS which lets the device resolve the PASID+addr pair
>> before a memory request is made into a routeable TLB address through the
>> TA. Those resolved addresses are then cached on the device instead of
>> in the IOMMU TLB and the device always sets the translated bit for PASID.
>> One example of those devices are AMD graphic devices that always have ACS
>> or ATS enabled together with PASID.
> 
> this should be made clear that ATS alone doesn't imply that translated
> bit is always set. It's just an optimization so the device may cache
> translation for selective requests. Only combining with PRI to
> support SVA has such implication.
> 
>> + * @flags: device-specific flags
>> + *   - PCI_PASID_XLATED_REQ_ONLY: The PCI device only issues PASID
>> + *                                memory requests of translated type.
> 
> this reads like the device even doesn't issue non-PASID requests.
> 
> It is clearer to be "The PCI device always use translated type for all
> PASID memory requests".
> 
> Otherwise looks good to me:
> 
> Reviewed-by: Kevin Tian <kevin.tian@...el.com>

Both make sense to me. It's updated. Thanks a lot!

--
Best regards,
baolu