From: Eric Auger <eric.auger@redhat.com>
To: Jean-Philippe Brucker <jean-philippe@linaro.org>,
Nicolin Chen <nicolinc@nvidia.com>
Cc: jgg@nvidia.com, robin.murphy@arm.com, will@kernel.org,
kevin.tian@intel.com, baolu.lu@linux.intel.com, joro@8bytes.org,
shameerali.kolothum.thodi@huawei.com,
linux-arm-kernel@lists.infradead.org, iommu@lists.linux.dev,
linux-kernel@vger.kernel.org, yi.l.liu@intel.com
Subject: Re: [PATCH v1 02/14] iommufd: Add nesting related data structures for ARM SMMUv3
Date: Fri, 10 Mar 2023 12:33:12 +0100 [thread overview]
Message-ID: <5288c0e9-b806-370d-e7de-8d69d5b8e902@redhat.com> (raw)
In-Reply-To: <20230309134217.GA1673607@myrica>
Hi,
On 3/9/23 14:42, Jean-Philippe Brucker wrote:
> Hi Nicolin,
>
> On Thu, Mar 09, 2023 at 02:53:38AM -0800, Nicolin Chen wrote:
>> Add the following data structures for corresponding ioctls:
>> iommu_hwpt_arm_smmuv3 => IOMMUFD_CMD_HWPT_ALLOC
>> iommu_hwpt_invalidate_arm_smmuv3 => IOMMUFD_CMD_HWPT_INVALIDATE
>>
>> Also, add IOMMU_HW_INFO_TYPE_ARM_SMMUV3 and IOMMU_PGTBL_TYPE_ARM_SMMUV3_S1
>> to the header and corresponding type/size arrays.
>>
>> Signed-off-by: Nicolin Chen <nicolinc@nvidia.com>
>> +/**
>> + * struct iommu_hwpt_arm_smmuv3 - ARM SMMUv3 specific page table data
>> + *
>> + * @flags: page table entry attributes
>> + * @s2vmid: Virtual machine identifier
>> + * @s1ctxptr: Stage-1 context descriptor pointer
>> + * @s1cdmax: Number of CDs pointed to by s1ContextPtr
>> + * @s1fmt: Stage-1 Format
>> + * @s1dss: Default substream
>> + */
>> +struct iommu_hwpt_arm_smmuv3 {
>> +#define IOMMU_SMMUV3_FLAG_S2 (1 << 0) /* if unset, stage-1 */
> I don't understand the purpose of this flag, since the structure only
> provides stage-1 configuration fields
>
>> +#define IOMMU_SMMUV3_FLAG_VMID (1 << 1) /* vmid override */
> Doesn't this break isolation? The VMID space is global for the SMMU, so
> the guest could access cached mappings of another device
>
>> + __u64 flags;
>> + __u32 s2vmid;
>> + __u32 __reserved;
>> + __u64 s1ctxptr;
>> + __u64 s1cdmax;
>> + __u64 s1fmt;
>> + __u64 s1dss;
>> +};
>> +
>
>> +/**
>> + * struct iommu_hwpt_invalidate_arm_smmuv3 - ARM SMMUv3 cahce invalidation info
>> + * @flags: boolean attributes of cache invalidation command
>> + * @opcode: opcode of cache invalidation command
>> + * @ssid: SubStream ID
>> + * @granule_size: page/block size of the mapping in bytes
>> + * @range: IOVA range to invalidate
>> + */
>> +struct iommu_hwpt_invalidate_arm_smmuv3 {
>> +#define IOMMU_SMMUV3_CMDQ_TLBI_VA_LEAF (1 << 0)
>> + __u64 flags;
>> + __u8 opcode;
>> + __u8 padding[3];
>> + __u32 asid;
>> + __u32 ssid;
>> + __u32 granule_size;
>> + struct iommu_iova_range range;
>> +};
> Although we can keep the alloc and hardware info separate for each IOMMU
> architecture, we should try to come up with common invalidation methods.
>
> It matters because things like vSVA, or just efficient dynamic mappings,
> will require optimal invalidation latency. A paravirtual interface like
> vhost-iommu can help with that, as the host kernel will handle guest
> invalidations directly instead of doing a round-trip to host userspace
> (and we'll likely want the same path for PRI.)
>
> Supporting HW page tables for a common PV IOMMU does require some
> architecture-specific knowledge, but invalidation messages contain roughly
> the same information on all architectures. The PV IOMMU won't include
> command opcodes for each possible architecture if one generic command does
> the same job.
>
> Ideally I'd like something like this for vhost-iommu:
>
> * slow path through userspace for complex requests like attach-table and
> probe, where the VMM can decode arch-specific information and translate
> them to iommufd and vhost-iommu ioctls to update the configuration.
>
> * fast path within the kernel for performance-critical requests like
> invalidate, page request and response. It would be absurd for the
> vhost-iommu driver to translate generic invalidation requests from the
> guest into arch-specific commands with special opcodes, when the next
> step is calling the IOMMU driver which does that for free.
>
> During previous discussions we came up with generic invalidations that
> could fit both Arm and x86 [1][2]. The only difference was the ASID
> (called archid/id in those proposals) which VT-d didn't need. Could we try
> to build on that?
I do agree with Jean. We spent a lot of efforts all together to define
this generic invalidation API and if there is compelling reason that
prevents from using it, we should try to reuse it.
Thanks
Eric
>
> [1] https://elixir.bootlin.com/linux/v5.17/source/include/uapi/linux/iommu.h#L161
> [2] https://lists.oasis-open.org/archives/virtio-dev/202102/msg00014.html
>
> Thanks,
> Jean
>
next prev parent reply other threads:[~2023-03-10 11:34 UTC|newest]
Thread overview: 165+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-03-09 10:53 [PATCH v1 00/14] Add Nested Translation Support for SMMUv3 Nicolin Chen
2023-03-09 10:53 ` [PATCH v1 01/14] iommu: Add iommu_get_unmanaged_domain helper Nicolin Chen
2023-03-09 12:51 ` Robin Murphy
2023-03-09 14:19 ` Jason Gunthorpe
2023-03-09 19:04 ` Robin Murphy
2023-03-10 0:23 ` Jason Gunthorpe
2023-03-10 8:41 ` Eric Auger
2023-03-10 15:55 ` Jason Gunthorpe
2023-03-16 1:21 ` Nicolin Chen
2023-03-16 18:42 ` Robin Murphy
2023-03-16 20:01 ` Nicolin Chen
2023-03-20 12:51 ` Jason Gunthorpe
2023-03-10 10:14 ` Eric Auger
2023-03-10 15:33 ` Jason Gunthorpe
2023-03-10 15:44 ` Shameerali Kolothum Thodi
2023-03-10 15:56 ` Jason Gunthorpe
2023-03-10 16:07 ` Shameerali Kolothum Thodi
2023-03-10 16:21 ` Jason Gunthorpe
2023-03-10 16:30 ` Shameerali Kolothum Thodi
2023-03-10 17:03 ` Jason Gunthorpe
2023-03-22 16:07 ` Eric Auger
2023-03-22 17:02 ` Jason Gunthorpe
2023-03-22 17:41 ` Eric Auger
2023-03-22 18:07 ` Jason Gunthorpe
2023-03-16 19:51 ` Nicolin Chen
2023-03-16 19:56 ` Shameerali Kolothum Thodi
2023-03-22 15:44 ` Eric Auger
2023-03-09 10:53 ` [PATCH v1 02/14] iommufd: Add nesting related data structures for ARM SMMUv3 Nicolin Chen
2023-03-09 13:42 ` Jean-Philippe Brucker
2023-03-09 14:48 ` Jason Gunthorpe
2023-03-09 18:26 ` Jean-Philippe Brucker
2023-03-09 21:01 ` Jason Gunthorpe
2023-03-10 12:16 ` Jean-Philippe Brucker
2023-03-10 14:52 ` Robin Murphy
2023-03-10 15:25 ` Jason Gunthorpe
2023-03-10 15:57 ` Robin Murphy
2023-03-10 16:03 ` Jason Gunthorpe
2023-03-17 10:10 ` Tian, Kevin
2023-03-17 10:04 ` Tian, Kevin
2023-03-10 4:50 ` Nicolin Chen
2023-03-10 12:54 ` Jean-Philippe Brucker
2023-03-10 14:00 ` Jason Gunthorpe
2023-03-10 16:06 ` Jason Gunthorpe
2023-03-16 0:59 ` Nicolin Chen
2023-03-09 15:26 ` Shameerali Kolothum Thodi
2023-03-09 15:40 ` Jason Gunthorpe
2023-03-09 15:51 ` Shameerali Kolothum Thodi
2023-03-09 15:59 ` Jason Gunthorpe
2023-03-09 16:07 ` Shameerali Kolothum Thodi
2023-03-10 5:26 ` Nicolin Chen
2023-03-10 5:36 ` Nicolin Chen
2023-03-10 12:55 ` Jason Gunthorpe
2023-03-10 5:18 ` Nicolin Chen
2023-03-10 5:04 ` Nicolin Chen
2023-03-10 11:33 ` Eric Auger [this message]
2023-03-10 12:51 ` Jason Gunthorpe
2023-03-17 10:17 ` Tian, Kevin
2023-03-09 10:53 ` [PATCH v1 03/14] iommufd/device: Setup MSI on kernel-managed domains Nicolin Chen
2023-03-10 16:45 ` Eric Auger
2023-03-11 0:17 ` Nicolin Chen
2023-03-09 10:53 ` [PATCH v1 04/14] iommu/arm-smmu-v3: Add arm_smmu_hw_info Nicolin Chen
2023-03-09 13:03 ` Robin Murphy
2023-03-10 1:17 ` Nicolin Chen
2023-03-10 15:28 ` Robin Murphy
2023-03-16 0:13 ` Nicolin Chen
2023-03-16 15:19 ` Robin Murphy
2023-03-16 20:06 ` Nicolin Chen
2023-04-12 7:47 ` Nicolin Chen
2023-03-09 10:53 ` [PATCH v1 05/14] iommu/arm-smmu-v3: Remove ARM_SMMU_DOMAIN_NESTED Nicolin Chen
2023-03-10 16:39 ` Eric Auger
2023-03-10 17:05 ` Jason Gunthorpe
2023-03-11 0:24 ` Nicolin Chen
2023-03-11 0:23 ` Nicolin Chen
2023-03-09 10:53 ` [PATCH v1 06/14] iommu/arm-smmu-v3: Unset corresponding STE fields when s2_cfg is NULL Nicolin Chen
2023-03-09 13:13 ` Robin Murphy
2023-03-09 18:24 ` Shameerali Kolothum Thodi
2023-03-10 1:54 ` Nicolin Chen
2023-03-09 10:53 ` [PATCH v1 07/14] iommu/arm-smmu-v3: Add STRTAB_STE_0_CFG_NESTED for 2-stage translation Nicolin Chen
2023-03-09 10:53 ` [PATCH v1 08/14] iommu/arm-smmu-v3: Prepare for nested domain support Nicolin Chen
2023-03-10 20:39 ` Robin Murphy
2023-03-11 12:40 ` Nicolin Chen
2023-03-09 10:53 ` [PATCH v1 09/14] iommu/arm-smmu-v3: Implement arm_smmu_get_unmanaged_domain Nicolin Chen
2023-03-09 10:53 ` [PATCH v1 10/14] iommu/arm-smmu-v3: Pass in user_cfg to arm_smmu_domain_finalise Nicolin Chen
2023-03-09 10:53 ` [PATCH v1 11/14] iommu/arm-smmu-v3: Add arm_smmu_domain_alloc_user Nicolin Chen
2023-03-24 15:28 ` Eric Auger
2023-03-24 17:40 ` Nicolin Chen
2023-03-24 17:50 ` Jason Gunthorpe
2023-03-24 18:00 ` Nicolin Chen
2023-03-24 15:33 ` Eric Auger
2023-03-24 17:43 ` Nicolin Chen
2023-03-09 10:53 ` [PATCH v1 12/14] iommu/arm-smmu-v3: Support IOMMU_DOMAIN_NESTED type of allocations Nicolin Chen
2023-03-09 13:20 ` Robin Murphy
2023-03-09 14:28 ` Robin Murphy
2023-03-10 1:34 ` Nicolin Chen
2023-03-24 15:44 ` Eric Auger
2023-03-24 16:30 ` Jason Gunthorpe
2023-03-24 17:50 ` Nicolin Chen
2023-03-24 17:51 ` Jason Gunthorpe
2023-03-24 17:55 ` Nicolin Chen
2023-03-09 10:53 ` [PATCH v1 13/14] iommu/arm-smmu-v3: Add CMDQ_OP_TLBI_NH_VAA and CMDQ_OP_TLBI_NH_ALL Nicolin Chen
2023-03-09 13:44 ` Robin Murphy
2023-03-10 1:19 ` Nicolin Chen
2023-03-09 10:53 ` [PATCH v1 14/14] iommu/arm-smmu-v3: Add arm_smmu_cache_invalidate_user Nicolin Chen
2023-03-09 14:49 ` Robin Murphy
2023-03-09 15:31 ` Jason Gunthorpe
2023-03-10 4:20 ` Nicolin Chen
2023-03-10 16:19 ` Jason Gunthorpe
2023-03-11 11:56 ` Nicolin Chen
2023-03-11 12:53 ` Nicolin Chen
2023-03-20 13:03 ` Jason Gunthorpe
2023-03-20 15:56 ` Nicolin Chen
2023-03-20 16:04 ` Jason Gunthorpe
2023-03-20 16:59 ` Nicolin Chen
2023-03-20 18:45 ` Jason Gunthorpe
2023-03-20 21:22 ` Nicolin Chen
2023-03-20 22:19 ` Jason Gunthorpe
2023-03-22 20:57 ` Nicolin Chen
2023-03-23 12:17 ` Jason Gunthorpe
2023-03-17 9:41 ` Tian, Kevin
2023-03-17 14:24 ` Nicolin Chen
2023-03-20 12:59 ` Jason Gunthorpe
2023-03-20 16:12 ` Nicolin Chen
2023-03-20 18:00 ` Jason Gunthorpe
2023-03-21 8:34 ` Tian, Kevin
2023-03-21 11:48 ` Jason Gunthorpe
2023-03-22 6:42 ` Nicolin Chen
2023-03-22 12:43 ` Jason Gunthorpe
2023-03-22 17:11 ` Nicolin Chen
2023-03-22 17:28 ` Jason Gunthorpe
2023-03-22 19:21 ` Nicolin Chen
2023-03-22 19:41 ` Jason Gunthorpe
2023-03-22 20:43 ` Nicolin Chen
2023-03-23 12:16 ` Jason Gunthorpe
2023-03-23 18:13 ` Nicolin Chen
2023-03-23 18:27 ` Jason Gunthorpe
2023-03-24 9:02 ` Tian, Kevin
2023-03-24 14:57 ` Jason Gunthorpe
2023-03-24 17:35 ` Nicolin Chen
2023-03-28 3:03 ` Tian, Kevin
2023-03-24 8:47 ` Tian, Kevin
2023-03-24 14:44 ` Jason Gunthorpe
2023-03-28 2:48 ` Tian, Kevin
2023-03-28 12:26 ` Jason Gunthorpe
2023-03-31 8:09 ` Tian, Kevin
2023-03-17 9:24 ` Tian, Kevin
2023-03-10 3:51 ` Nicolin Chen
2023-03-10 17:53 ` Robin Murphy
2023-03-10 18:49 ` Jason Gunthorpe
2023-03-11 12:38 ` Nicolin Chen
2023-03-13 13:07 ` Robin Murphy
2023-03-16 0:01 ` Nicolin Chen
2023-03-16 14:58 ` Robin Murphy
2023-03-16 21:09 ` Nicolin Chen
2023-03-20 1:32 ` Nicolin Chen
2023-03-20 13:11 ` Jason Gunthorpe
2023-03-20 15:28 ` Nicolin Chen
2023-03-20 16:01 ` Jason Gunthorpe
2023-03-20 16:35 ` Nicolin Chen
2023-03-20 18:07 ` Jason Gunthorpe
2023-03-20 20:46 ` Nicolin Chen
2023-03-20 22:14 ` Jason Gunthorpe
2023-03-22 5:14 ` Nicolin Chen
2023-03-24 8:55 ` Tian, Kevin
2023-03-17 9:47 ` Tian, Kevin
2023-03-17 14:16 ` Nicolin Chen
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=5288c0e9-b806-370d-e7de-8d69d5b8e902@redhat.com \
--to=eric.auger@redhat.com \
--cc=baolu.lu@linux.intel.com \
--cc=iommu@lists.linux.dev \
--cc=jean-philippe@linaro.org \
--cc=jgg@nvidia.com \
--cc=joro@8bytes.org \
--cc=kevin.tian@intel.com \
--cc=linux-arm-kernel@lists.infradead.org \
--cc=linux-kernel@vger.kernel.org \
--cc=nicolinc@nvidia.com \
--cc=robin.murphy@arm.com \
--cc=shameerali.kolothum.thodi@huawei.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;
as well as URLs for NNTP newsgroup(s).