From: Nicolin Chen <nicolinc@nvidia.com>
To: Pranjal Shrivastava <praan@google.com>
Cc: <jgg@nvidia.com>, <kevin.tian@intel.com>, <corbet@lwn.net>,
<will@kernel.org>, <bagasdotme@gmail.com>, <robin.murphy@arm.com>,
<joro@8bytes.org>, <thierry.reding@gmail.com>,
<vdumpa@nvidia.com>, <jonathanh@nvidia.com>, <shuah@kernel.org>,
<jsnitsel@redhat.com>, <nathan@kernel.org>,
<peterz@infradead.org>, <yi.l.liu@intel.com>,
<mshavit@google.com>, <zhangzekun11@huawei.com>,
<iommu@lists.linux.dev>, <linux-doc@vger.kernel.org>,
<linux-kernel@vger.kernel.org>,
<linux-arm-kernel@lists.infradead.org>,
<linux-tegra@vger.kernel.org>, <linux-kselftest@vger.kernel.org>,
<patches@lists.linux.dev>, <mochs@nvidia.com>,
<alok.a.tiwari@oracle.com>, <vasant.hegde@amd.com>
Subject: Re: [PATCH v2 21/22] iommu/tegra241-cmdqv: Add user-space use support
Date: Wed, 30 Apr 2025 15:39:15 -0700 [thread overview]
Message-ID: <aBKmk6PNFreeyfLh@Asurada-Nvidia> (raw)
In-Reply-To: <aBKdMaFLPFJYegIS@google.com>
On Wed, Apr 30, 2025 at 09:59:13PM +0000, Pranjal Shrivastava wrote:
> On Fri, Apr 25, 2025 at 10:58:16PM -0700, Nicolin Chen wrote:
> > The CMDQV HW supports a user-space use for virtualization cases. It allows
> > the VM to issue guest-level TLBI or ATC_INV commands directly to the queue
> > and executes them without a VMEXIT, as HW will replace the VMID field in a
> > TLBI command and the SID field in an ATC_INV command with the preset VMID
> > and SID.
> >
> > This is built upon the vIOMMU infrastructure by allowing VMM to allocate a
> > VINTF (as a vIOMMU object) and assign VCMDQs (vCMDQ objects) to the VINTF.
> >
> > So firstly, replace the standard vSMMU model with the VINTF implementation
> > but reuse the standard cache_invalidate op (for unsupported commands) and
> > the standard alloc_domain_nested op (for standard nested STE).
> >
> > Each VINTF has two 64KB MMIO pages (128B per logical vCMDQ):
> > - Page0 (directly accessed by guest) has all the control and status bits.
> > - Page1 (trapped by VMM) has guest-owned queue memory location/size info.
> >
> > VMM should trap the emulated VINTF0's page1 of the guest VM for the guest-
> > level VCMDQ location/size info and forward that to the kernel to translate
> > to a physical memory location to program the VCMDQ HW during an allocation
> > call. Then, it should mmap the assigned VINTF's page0 to the VINTF0 page0
> > of the guest VM. This allows the guest OS to read and write the guest-own
> > VINTF's page0 for direct control of the VCMDQ HW.
> >
> > For ATC invalidation commands that hold an SID, it requires all devices to
> > register their virtual SIDs to the SID_MATCH registers and their physical
> > SIDs to the pairing SID_REPLACE registers, so that HW can use those as a
> > lookup table to replace those virtual SIDs with the correct physical SIDs.
> > Thus, implement the driver-allocated vDEVICE op with a tegra241_vintf_sid
> > structure to allocate SID_REPLACE and to program the SIDs accordingly.
> >
> > This enables the HW accelerated feature for NVIDIA Grace CPU. Compared to
> > the standard SMMUv3 operating in the nested translation mode trapping CMDQ
> > for TLBI and ATC_INV commands, this gives a huge performance improvement:
> > 70% to 90% reductions of invalidation time were measured by various DMA
> > unmap tests running in a guest OS.
> >
>
> The write-up is super helpful to understand how the HW works from a high
> level. Thanks for explaining this well! :)
>
> I'm curious to know the DMA unmap tests that were run for perf?
tools/testing/selftests/dma/dma_map_benchmark.c
> > /**
> > * struct iommu_hw_info_arm_smmuv3 - ARM SMMUv3 hardware information
> > * (IOMMU_HW_INFO_TYPE_ARM_SMMUV3)
> > *
> > - * @flags: Must be set to 0
> > - * @impl: Must be 0
> > + * @flags: Combination of enum iommu_hw_info_arm_smmuv3_flags
> > + * @impl: Implementation-defined bits when the following flags are set:
> > + * - IOMMU_HW_INFO_ARM_SMMUV3_HAS_TEGRA241_CMDQV
> > + * Bits[15:12] - Log2 of the total number of SID replacements
> > + * Bits[07:04] - Log2 of the total number of vCMDQs per vIOMMU
> > + * Bits[03:00] - Version number for the CMDQ-V HW
>
> Nit: It seems that we deliberately chose not to reveal `NUM_VINTF_LOG2`
> to the user-space. If so, maybe we shall mark those bitfields as unused
> or reserved for clarity? Bits[11:08] - Reserved / Unused (even 31:16).
I think it should have been there, but kernel should just report 0.
Bits[11:08] - Log2 of the total number of virtual interface
> > * @idr: Implemented features for ARM SMMU Non-secure programming interface
> > * @iidr: Information about the implementation and implementer of ARM SMMU,
> > * and architecture version supported
> > @@ -952,10 +965,28 @@ struct iommu_fault_alloc {
> > * enum iommu_viommu_type - Virtual IOMMU Type
> > * @IOMMU_VIOMMU_TYPE_DEFAULT: Reserved for future use
> > * @IOMMU_VIOMMU_TYPE_ARM_SMMUV3: ARM SMMUv3 driver specific type
> > + * @IOMMU_VIOMMU_TYPE_TEGRA241_CMDQV: NVIDIA Tegra241 CMDQV Extension for SMMUv3
> > */
> > enum iommu_viommu_type {
> > IOMMU_VIOMMU_TYPE_DEFAULT = 0,
> > IOMMU_VIOMMU_TYPE_ARM_SMMUV3 = 1,
> > + IOMMU_VIOMMU_TYPE_TEGRA241_CMDQV = 2,
> > +};
>
> This is a little confusing.. I understand that we need a new viommu type
> to copy the new struct iommu_viommu_tegra241_cmdqv b/w the user & kernel
>
> But, in a previous patch (Add vsmmu_alloc impl op), we add a check to
> fallback to the standard type SMMUv3, if the impl_ops->vsmmu_alloc
> returns -EOPNOTSUPP:
>
> if (master->smmu->impl_ops && master->smmu->impl_ops->vsmmu_alloc)
> vsmmu = master->smmu->impl_ops->vsmmu_alloc(
> master->smmu, s2_parent, ictx, viommu_type, user_data);
> if (PTR_ERR(vsmmu) == -EOPNOTSUPP) {
> if (viommu_type != IOMMU_VIOMMU_TYPE_ARM_SMMUV3)
> return ERR_PTR(-EOPNOTSUPP);
> /* Fallback to standard SMMUv3 type if viommu_type matches */
> vsmmu = iommufd_viommu_alloc(ictx, struct arm_vsmmu, core,
> &arm_vsmmu_ops);
>
> Now, if we'll ALWAYS try to allocate an impl-specified vsmmu first, even
> when the viommu_type == IOMMU_VIOMMU_TYPE_ARM_SMMUV3, we are anyways
> going to return back from the impl_ops->vsmmu_alloc with -EOPNOTSUPP.
That's not necessarily true. An impl_ops->vsmmu_alloc can support
IOMMU_VIOMMU_TYPE_ARM_SMMUV3 potentially, e.g. an impl could just
toggle a few special bits in a register and return a valid vsmmu
pointer.
It doesn't work like this with VCMDQ as it supports its own type,
but for the long run I think we should pass in the standard type
to impl_ops->vsmmu_alloc too.
> Then we'll again check if the retval was -EOPNOTSUPP and re-check the
> viommu_type requested.. which seems a little counter intuitive.
It's just prioritizing the impl_ops->vsmmu_alloc. Similar to the
probe, if VCMDQ is missing or encountering some initialization
problem, give it a chance to fallback to the standard SMMU.
> > + /*
> > + * @length must be a power of 2, in range of
> > + * [ 32, 1 ^ (idr[1].CMDQS + CMDQ_ENT_SZ_SHIFT) ]
> > + */
>
> Nit: 2 ^ (idr[1].CMDQS + CMDQ_ENT_SZ_SHIFT) to match the comment in uapi
Alok pointed it out too. Fixed.
> > + vcmdq = iommufd_vcmdq_alloc(viommu, struct tegra241_vcmdq, core);
> > + if (!vcmdq)
> > + return ERR_PTR(-ENOMEM);
> > +
> > + /*
> > + * HW requires to unmap LVCMDQs in descending order, so destroy() must
> > + * follow this rule. Set a dependency on its previous LVCMDQ so iommufd
> > + * core will help enforce it.
> > + */
> > + if (prev) {
> > + ret = iommufd_vcmdq_depend(vcmdq, prev, core);
> > + if (ret)
> > + goto free_vcmdq;
> > + }
> > + vcmdq->prev = prev;
> > +
> > + ret = tegra241_vintf_init_lvcmdq(vintf, index, vcmdq);
> > + if (ret)
> > + goto free_vcmdq;
> > +
> > + dev_dbg(cmdqv->dev, "%sallocated\n",
> > + lvcmdq_error_header(vcmdq, header, 64));
> > +
> > + tegra241_vcmdq_map_lvcmdq(vcmdq);
> > +
> > + vcmdq->cmdq.q.q_base = q_base & VCMDQ_ADDR;
> > + vcmdq->cmdq.q.q_base |= log2size;
> > +
> > + ret = tegra241_vcmdq_hw_init_user(vcmdq);
> > + if (ret)
> > + goto free_vcmdq;
> > + vintf->lvcmdqs[index] = vcmdq;
> > +
> > + return &vcmdq->core;
> > +free_vcmdq:
> > + iommufd_struct_destroy(viommu->ictx, vcmdq, core);
> > + return ERR_PTR(ret);
>
> Are we missing an undepend here?
Right. The iommufd_struct_destroy doesn't invoke obj->ops.abort().
The whole revert flow is wonky, missing all the unmap/deinit steps.
> > +static void tegra241_vintf_destroy_vdevice(struct iommufd_vdevice *vdev)
> > +{
> > + struct tegra241_vintf_sid *vsid =
> > + container_of(vdev, struct tegra241_vintf_sid, core);
> > + struct tegra241_vintf *vintf = vsid->vintf;
> > +
> > + writel_relaxed(0, REG_VINTF(vintf, SID_REPLACE(vsid->idx)));
> > + writel_relaxed(0, REG_VINTF(vintf, SID_MATCH(vsid->idx)));
>
> Just a thought: Should these be writel to avoid races?
> Although I believe all user-queues would be free-d by this point?
Yea. They should be. I will change them.
Thanks
Nicolin
next prev parent reply other threads:[~2025-04-30 22:39 UTC|newest]
Thread overview: 146+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-04-26 5:57 [PATCH v2 00/22] iommufd: Add vIOMMU infrastructure (Part-4 vCMDQ) Nicolin Chen
2025-04-26 5:57 ` [PATCH v2 01/22] iommufd/viommu: Add driver-allocated vDEVICE support Nicolin Chen
2025-04-27 6:23 ` Baolu Lu
2025-04-28 0:41 ` Tian, Kevin
2025-04-28 18:08 ` Nicolin Chen
2025-04-26 5:57 ` [PATCH v2 02/22] iommu: Pass in a driver-level user data structure to viommu_alloc op Nicolin Chen
2025-04-27 6:31 ` Baolu Lu
2025-04-28 17:19 ` Nicolin Chen
2025-04-28 17:28 ` Pranjal Shrivastava
2025-04-26 5:57 ` [PATCH v2 03/22] iommufd/viommu: Allow driver-specific user data for a vIOMMU object Nicolin Chen
2025-04-27 6:36 ` Baolu Lu
2025-04-28 17:52 ` Pranjal Shrivastava
2025-04-30 14:58 ` ALOK TIWARI
2025-04-26 5:57 ` [PATCH v2 04/22] iommu: Add iommu_copy_struct_to_user helper Nicolin Chen
2025-04-27 6:39 ` Baolu Lu
2025-04-28 17:50 ` Pranjal Shrivastava
2025-04-28 18:21 ` Nicolin Chen
2025-04-29 8:31 ` Pranjal Shrivastava
2025-04-26 5:58 ` [PATCH v2 05/22] iommufd: Add iommufd_struct_destroy to revert iommufd_viommu_alloc Nicolin Chen
2025-04-27 6:55 ` Baolu Lu
2025-04-28 17:24 ` Nicolin Chen
2025-04-26 5:58 ` [PATCH v2 06/22] iommufd/selftest: Support user_data in mock_viommu_alloc Nicolin Chen
2025-04-28 18:56 ` Pranjal Shrivastava
2025-04-26 5:58 ` [PATCH v2 07/22] iommufd/selftest: Add covearge for viommu data Nicolin Chen
2025-04-28 19:02 ` Pranjal Shrivastava
2025-04-26 5:58 ` [PATCH v2 08/22] iommufd: Abstract iopt_pin_pages and iopt_unpin_pages helpers Nicolin Chen
2025-04-27 7:22 ` Baolu Lu
2025-04-28 17:41 ` Nicolin Chen
2025-05-05 15:01 ` Jason Gunthorpe
2025-05-05 15:44 ` Nicolin Chen
2025-05-05 15:55 ` Jason Gunthorpe
2025-05-05 16:03 ` Nicolin Chen
2025-05-05 16:05 ` Jason Gunthorpe
2025-05-05 16:19 ` Nicolin Chen
2025-05-05 16:56 ` Jason Gunthorpe
2025-04-28 20:14 ` Pranjal Shrivastava
2025-04-28 22:12 ` Nicolin Chen
2025-04-28 23:34 ` Nicolin Chen
2025-04-29 18:03 ` Pranjal Shrivastava
2025-05-06 9:36 ` Tian, Kevin
2025-05-06 19:17 ` Nicolin Chen
2025-05-07 7:22 ` Tian, Kevin
2025-05-07 7:36 ` Nicolin Chen
2025-05-07 7:51 ` Tian, Kevin
2025-04-26 5:58 ` [PATCH v2 09/22] iommufd/viommu: Introduce IOMMUFD_OBJ_VCMDQ and its related struct Nicolin Chen
2025-04-28 1:09 ` Baolu Lu
2025-04-28 18:10 ` Nicolin Chen
2025-05-05 15:02 ` Jason Gunthorpe
2025-05-05 15:45 ` Nicolin Chen
2025-04-28 21:01 ` Pranjal Shrivastava
2025-04-26 5:58 ` [PATCH v2 10/22] iommufd/viommmu: Add IOMMUFD_CMD_VCMDQ_ALLOC ioctl Nicolin Chen
2025-04-28 1:32 ` Baolu Lu
2025-04-28 18:58 ` Nicolin Chen
2025-04-29 6:11 ` Baolu Lu
2025-04-28 12:12 ` Vasant Hegde
2025-04-28 20:02 ` Nicolin Chen
2025-04-29 5:34 ` Vasant Hegde
2025-04-29 6:45 ` Nicolin Chen
2025-04-29 10:22 ` Vasant Hegde
2025-04-29 17:14 ` Nicolin Chen
2025-04-30 4:22 ` Vasant Hegde
2025-04-30 8:01 ` Nicolin Chen
2025-04-30 10:21 ` Vasant Hegde
2025-05-06 9:25 ` Tian, Kevin
2025-05-06 20:12 ` Nicolin Chen
2025-05-07 7:25 ` Tian, Kevin
2025-05-07 7:37 ` Nicolin Chen
2025-05-07 12:33 ` Jason Gunthorpe
2025-05-07 20:51 ` Nicolin Chen
2025-04-28 21:34 ` Pranjal Shrivastava
2025-04-28 22:44 ` Nicolin Chen
2025-04-29 8:28 ` Pranjal Shrivastava
2025-04-29 18:10 ` Pranjal Shrivastava
2025-04-29 18:15 ` Nicolin Chen
2025-04-29 18:57 ` Pranjal Shrivastava
2025-04-26 5:58 ` [PATCH v2 11/22] iommufd: Add for-driver helpers iommufd_vcmdq_depend/undepend() Nicolin Chen
2025-04-28 2:22 ` Baolu Lu
2025-04-28 18:17 ` Nicolin Chen
2025-04-29 12:40 ` Pranjal Shrivastava
2025-04-29 17:10 ` Nicolin Chen
2025-04-29 17:59 ` Pranjal Shrivastava
2025-04-29 18:07 ` Nicolin Chen
2025-04-29 18:44 ` Pranjal Shrivastava
2025-04-26 5:58 ` [PATCH v2 12/22] iommufd/selftest: Add coverage for IOMMUFD_CMD_VCMDQ_ALLOC Nicolin Chen
2025-04-26 5:58 ` [PATCH v2 13/22] iommufd: Add mmap interface Nicolin Chen
2025-04-28 2:50 ` Baolu Lu
2025-04-28 18:54 ` Nicolin Chen
2025-05-05 16:50 ` Jason Gunthorpe
2025-05-05 17:21 ` Nicolin Chen
2025-05-05 17:28 ` Jason Gunthorpe
2025-05-05 20:07 ` Nicolin Chen
2025-05-06 9:22 ` Tian, Kevin
2025-05-06 12:55 ` Jason Gunthorpe
2025-05-06 12:54 ` Jason Gunthorpe
2025-05-06 20:54 ` Nicolin Chen
2025-05-07 12:36 ` Jason Gunthorpe
2025-05-07 20:49 ` Nicolin Chen
2025-04-29 20:24 ` Pranjal Shrivastava
2025-04-29 20:34 ` Pranjal Shrivastava
2025-04-29 20:39 ` Nicolin Chen
2025-04-29 20:55 ` Pranjal Shrivastava
2025-04-29 21:05 ` Nicolin Chen
2025-04-29 21:35 ` Pranjal Shrivastava
2025-04-29 21:46 ` Nicolin Chen
2025-04-29 21:57 ` Pranjal Shrivastava
2025-05-05 16:55 ` Jason Gunthorpe
2025-05-05 17:27 ` Nicolin Chen
2025-05-05 17:31 ` Jason Gunthorpe
2025-05-05 19:50 ` Nicolin Chen
2025-05-06 12:52 ` Jason Gunthorpe
2025-05-06 19:30 ` Nicolin Chen
2025-05-07 12:39 ` Jason Gunthorpe
2025-05-07 21:09 ` Nicolin Chen
2025-05-07 22:08 ` Jason Gunthorpe
2025-05-08 3:49 ` Nicolin Chen
2025-05-08 9:15 ` Tian, Kevin
2025-05-08 12:12 ` Jason Gunthorpe
2025-05-08 17:14 ` Nicolin Chen
2025-05-05 18:47 ` Pranjal Shrivastava
2025-04-26 5:58 ` [PATCH v2 14/22] iommufd/selftest: Add coverage for the new " Nicolin Chen
2025-04-26 5:58 ` [PATCH v2 15/22] Documentation: userspace-api: iommufd: Update vCMDQ Nicolin Chen
2025-04-28 14:31 ` Bagas Sanjaya
2025-04-28 19:00 ` Nicolin Chen
2025-04-26 5:58 ` [PATCH v2 16/22] iommu/arm-smmu-v3-iommufd: Add vsmmu_alloc impl op Nicolin Chen
2025-04-29 21:36 ` Pranjal Shrivastava
2025-04-26 5:58 ` [PATCH v2 17/22] iommu/arm-smmu-v3-iommufd: Support implementation-defined hw_info Nicolin Chen
2025-04-29 21:44 ` Pranjal Shrivastava
2025-04-26 5:58 ` [PATCH v2 18/22] iommu/tegra241-cmdqv: Use request_threaded_irq Nicolin Chen
2025-04-29 21:47 ` Pranjal Shrivastava
2025-04-26 5:58 ` [PATCH v2 19/22] iommu/tegra241-cmdqv: Simplify deinit flow in tegra241_cmdqv_remove_vintf() Nicolin Chen
2025-04-29 22:05 ` Pranjal Shrivastava
2025-04-26 5:58 ` [PATCH v2 20/22] iommu/tegra241-cmdqv: Do not statically map LVCMDQs Nicolin Chen
2025-04-29 20:43 ` ALOK TIWARI
2025-04-29 22:32 ` Pranjal Shrivastava
2025-04-29 22:37 ` Nicolin Chen
2025-04-26 5:58 ` [PATCH v2 21/22] iommu/tegra241-cmdqv: Add user-space use support Nicolin Chen
2025-04-29 19:47 ` ALOK TIWARI
2025-04-29 21:12 ` Nicolin Chen
2025-04-30 21:59 ` Pranjal Shrivastava
2025-04-30 22:39 ` Nicolin Chen [this message]
2025-05-01 0:54 ` Nicolin Chen
2025-05-01 21:46 ` Pranjal Shrivastava
2025-05-01 21:45 ` Pranjal Shrivastava
2025-04-26 5:58 ` [PATCH v2 22/22] iommu/tegra241-cmdqv: Add IOMMU_VEVENTQ_TYPE_TEGRA241_CMDQV support Nicolin Chen
2025-04-30 15:07 ` ALOK TIWARI
2025-04-30 22:03 ` Pranjal Shrivastava
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=aBKmk6PNFreeyfLh@Asurada-Nvidia \
--to=nicolinc@nvidia.com \
--cc=alok.a.tiwari@oracle.com \
--cc=bagasdotme@gmail.com \
--cc=corbet@lwn.net \
--cc=iommu@lists.linux.dev \
--cc=jgg@nvidia.com \
--cc=jonathanh@nvidia.com \
--cc=joro@8bytes.org \
--cc=jsnitsel@redhat.com \
--cc=kevin.tian@intel.com \
--cc=linux-arm-kernel@lists.infradead.org \
--cc=linux-doc@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-kselftest@vger.kernel.org \
--cc=linux-tegra@vger.kernel.org \
--cc=mochs@nvidia.com \
--cc=mshavit@google.com \
--cc=nathan@kernel.org \
--cc=patches@lists.linux.dev \
--cc=peterz@infradead.org \
--cc=praan@google.com \
--cc=robin.murphy@arm.com \
--cc=shuah@kernel.org \
--cc=thierry.reding@gmail.com \
--cc=vasant.hegde@amd.com \
--cc=vdumpa@nvidia.com \
--cc=will@kernel.org \
--cc=yi.l.liu@intel.com \
--cc=zhangzekun11@huawei.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).