ACK: [SRU][N:linux-aws][PULL] Backport patches to support NVIDIA GB200
Massimiliano Pellizzer
massimiliano.pellizzer at canonical.com
Mon Mar 24 09:37:17 UTC 2025
On Fri, 21 Mar 2025 at 15:38, Philip Cox <philip.cox at canonical.com> wrote:
>
>
> On 2025-03-17 3:57 p.m., Magali Lemes wrote:
> > BugLink: https://bugs.launchpad.net/bugs/2101185
> >
> > [Impact]
> > AWS requested the following patchsets to support GB200 to be backported
> > to 6.8 kernels and newer:
> > #1: Update SMMUv3 to the modern iommu API (part 2b/3)
> > (https://lore.kernel.org/linux-iommu/0-v9-5cd718286059+79186-smmuv3_newapi_p2b_jgg@nvidia.com/)
> >
> > #2: Add Tegra241 (Grace) CMDQV Support (part 1/2)
> > (https://lore.kernel.org/linux-iommu/cover.1724970714.git.nicolinc@nvidia.com/)
> >
> > #3: iommu/tegra241-cmdqv: Fix alignment failure at max_n_shift
> > (https://lore.kernel.org/all/20241111030226.1940737-1-nicolinc@nvidia.com/)
> >
> >
> > [Fix]
> > Backporting patchset #1 required backporting the patchsets it was
> > built on top of:
> > - All patches from "Update SMMUv3 to the modern iommu API (part
> > 1/3)"
> > (https://lore.kernel.org/all/0-v6-96275f25c39d+2d4-smmuv3_newapi_p1_jgg@nvidia.com/),
> > except for "iommu/arm-smmu-v3: Hold arm_smmu_asid_lock during all of
> > attach_dev", which is already in the tree. And
> > * "iommu/arm-smmu-v3: Add cpu_to_le64() around STRTAB_STE_0_V", as
> > a fix to "iommu/arm-smmu-v3: Make STE programming independent of the
> > callers".
> > - Parts of "Update SMMUv3 to the modern iommu API (part 2/3)"
> > (https://lore.kernel.org/all/0-v6-228e7adf25eb+4155-smmuv3_newapi_p2_jgg@nvidia.com/).
> > - All patches from "Make the SMMUv3 CD logic match the new STE
> > design" (part 2a/3)
> > (https://lore.kernel.org/all/0-v9-5040dc602008+177d7-smmuv3_newapi_p2_jgg@nvidia.com/).
> > And
> > * "iommu: Introduce iommu_group_mutex_assert()", as pre-req for
> > "iommu/arm-smmu-v3: Move the CD generation for SVA into a function".
> > * "iommu/arm-smmu-v3: Fix access for STE.SHCFG", as pre-req for
> > "iommu/arm-smmu-v3: Add unit tests for arm_smmu_write_entry".
> > * "iommu/arm-smmu-v3: Avoid uninitialized asid in case of error",
> > as a fix for "iommu/arm-smmu-v3: Build the whole CD in
> > arm_smmu_make_s1_cd()".
> > Other pre-reqs for patchset #1 included:
> > - "iommu: Pass domain to remove_dev_pasid() op", as pre-req for
> > "iommu/arm-smmu-v3: Put the SVA mmu notifier in the smmu_domain".
> > - A backport of "iommu: Add ops->domain_alloc_sva()", as pre-req for
> > "iommu/arm-smmu-v3: Convert to domain_alloc_sva()".
> >
> > Patchset #2 was almost all clean cherry-picks except for
> > "iommu/arm-smmu-v3: Add acpi_smmu_iort_probe_model for impl".
> > Additionally, these were needed:
> > * "iommu/arm-smmu-v3: Make the kunit into a module" and
> > "iommu/arm-smmu-v3: Use *-y instead of *-objs in Makefile" to clean
> > cherry-pick "iommu/arm-smmu-v3: Add in-kernel support for NVIDIA
> > Tegra241 (Grace) CMDQV".
> > * "iommu/arm-smmu-v3: add missing MODULE_DESCRIPTION() macro", as a
> > fix for "iommu/arm-smmu-v3: Make the kunit into a module".
> > * Other `Fixes` to commit 918eb5c856f6 ("iommu/arm-smmu-v3: Add
> > in-kernel support for NVIDIA Tegra241 (Grace) CMDQV"), which were
> > clean cherry-picks.
> >
> > Patchset #3's only patch and one fix to it were clean cherry-picks.
> >
> > [Test Case]
> > Compile and boot tested.
> > Tested by AWS.
> >
> > [Where problems could occur]
> > These patches introduce significant refactoring and features to the
> > SMMUv3 driver. Additionally, support to Tegra241's CMDQ-Virtualization
> > is being added.
> >
> > [Other info]
> > SF #00404773
> >
> > ------------------------------------------------------------------------
> >
> > The following changes since commit
> > 21b44fd410f8f0e8330c93553ebaba64ab9dd36e:
> >
> > UBUNTU: Ubuntu-aws-6.8.0-1025.27 (2025-02-19 11:40:06 -0500)
> >
> > are available in the Git repository at:
> >
> >
> > git://git.launchpad.net/~magalilemes/ubuntu/+source/linux-aws/+git/noble
> > gb200
> >
> > for you to fetch changes up to 45b2986d7b60d6d562aac2e844632dfbc70d8115:
> >
> > iommu/tegra241-cmdqv: Read SMMU IDR1.CMDQS instead of hardcoding
> > (2025-03-17 13:57:50 -0300)
> >
> > ----------------------------------------------------------------
> > Andy Shevchenko (1):
> > iommu/arm-smmu-v3: Use *-y instead of *-objs in Makefile
> >
> > Dan Carpenter (1):
> > iommu/tegra241-cmdqv: Fix ioremap() error handling in probe()
> >
> > Jason Gunthorpe (44):
> > iommu/arm-smmu-v3: Make STE programming independent of the callers
> > iommu/arm-smmu-v3: Consolidate the STE generation for abort/bypass
> > iommu/arm-smmu-v3: Move the STE generation for S1 and S2 domains
> > into functions
> > iommu/arm-smmu-v3: Build the whole STE in
> > arm_smmu_make_s2_domain_ste()
> > iommu/arm-smmu-v3: Compute the STE only once for each master
> > iommu/arm-smmu-v3: Do not change the STE twice during
> > arm_smmu_attach_dev()
> > iommu/arm-smmu-v3: Put writing the context descriptor in the
> > right order
> > iommu/arm-smmu-v3: Pass smmu_domain to arm_enable/disable_ats()
> > iommu/arm-smmu-v3: Remove arm_smmu_master->domain
> > iommu/arm-smmu-v3: Check that the RID domain is S1 in SVA
> > iommu/arm-smmu-v3: Add a global static IDENTITY domain
> > iommu/arm-smmu-v3: Add a global static BLOCKED domain
> > iommu/arm-smmu-v3: Use the identity/blocked domain during release
> > iommu/arm-smmu-v3: Pass arm_smmu_domain and arm_smmu_device to
> > finalize
> > iommu/arm-smmu-v3: Convert to domain_alloc_paging()
> > iommu/arm-smmu-v3: Add cpu_to_le64() around STRTAB_STE_0_V
> > iommu/arm-smmu-v3: Do not ATC invalidate the entire domain
> > iommu/arm-smmu-v3: Add a type for the CD entry
> > iommu/arm-smmu-v3: Add an ops indirection to the STE code
> > iommu/arm-smmu-v3: Make CD programming use arm_smmu_write_entry()
> > iommu/arm-smmu-v3: Move the CD generation for S1 domains into a
> > function
> > iommu/arm-smmu-v3: Consolidate clearing a CD table entry
> > iommu/arm-smmu-v3: Make arm_smmu_alloc_cd_ptr()
> > iommu/arm-smmu-v3: Allocate the CD table entry in advance
> > iommu/arm-smmu-v3: Move the CD generation for SVA into a function
> > iommu/arm-smmu-v3: Build the whole CD in arm_smmu_make_s1_cd()
> > iommu/arm-smmu-v3: Add unit tests for arm_smmu_write_entry
> > iommu: Add ops->domain_alloc_sva()
> > iommu/arm-smmu-v3: Convert to domain_alloc_sva()
> > iommu/arm-smmu-v3: Start building a generic PASID layer
> > iommu/arm-smmu-v3: Make smmu_domain->devices into an allocated list
> > iommu/arm-smmu-v3: Make changing domains be hitless for ATS
> > iommu/arm-smmu-v3: Add ssid to struct arm_smmu_master_domain
> > iommu/arm-smmu-v3: Do not use master->sva_enable to restrict
> > attaches
> > iommu/arm-smmu-v3: Thread SSID through the arm_smmu_attach_*()
> > interface
> > iommu/arm-smmu-v3: Make SVA allocate a normal arm_smmu_domain
> > iommu/arm-smmu-v3: Keep track of arm_smmu_master_domain for SVA
> > iommu/arm-smmu-v3: Put the SVA mmu notifier in the smmu_domain
> > iommu/arm-smmu-v3: Allow IDENTITY/BLOCKED to be set while PASID
> > is used
> > iommu/arm-smmu-v3: Test the STE S1DSS functionality
> > iommu/arm-smmu-v3: Allow a PASID to be set when RID is
> > IDENTITY/BLOCKED
> > iommu/arm-smmu-v3: Allow setting a S1 domain to a PASID
> > iommu/arm-smmu-v3: Make the kunit into a module
> > iommu/arm-smmu-v3: Add struct arm_smmu_impl_ops
> >
> > Jeff Johnson (1):
> > iommu/arm-smmu-v3: add missing MODULE_DESCRIPTION() macro
> >
> > Luis Claudio R. Goncalves (1):
> > iommu/tegra241-cmdqv: do not use smp_processor_id in preemptible
> > context
> >
> > Magali Lemes (1):
> > UBUNTU: [Config] updateconfigs to enable CONFIG_TEGRA241_CMDQV
> >
> > Mostafa Saleh (2):
> > iommu/arm-smmu-v3: Fix access for STE.SHCFG
> > iommu/arm-smmu-v3: Avoid uninitialized asid in case of error
> >
> > Nate Watterson (1):
> > iommu/arm-smmu-v3: Add in-kernel support for NVIDIA Tegra241
> > (Grace) CMDQV
> >
> > Nicolin Chen (14):
> > iommu/arm-smmu-v3: Issue a batch of commands to the same cmdq
> > iommu/arm-smmu-v3: Pass in cmdq pointer to
> > arm_smmu_cmdq_build_sync_cmd
> > iommu/arm-smmu-v3: Pass in cmdq pointer to arm_smmu_cmdq_init
> > iommu/arm-smmu-v3: Make symbols public for CONFIG_TEGRA241_CMDQV
> > iommu/arm-smmu-v3: Add ARM_SMMU_OPT_TEGRA241_CMDQV
> > iommu/arm-smmu-v3: Add acpi_smmu_iort_probe_model for impl
> > iommu/arm-smmu-v3: Start a new batch if new command is not
> > supported
> > iommu/tegra241-cmdqv: Limit CMDs for VCMDQs of a guest owned VINTF
> > iommu/tegra241-cmdqv: Fix -Wformat-truncation warnings in
> > lvcmdq_error_header
> > iommu/tegra241-cmdqv: Drop static at local variable
> > iommu/tegra241-cmdqv: Do not allocate vcmdq until
> > dma_set_mask_and_coherent
> > iommu/tegra241-cmdqv: Staticize cmdqv_debugfs_dir
> > iommu/tegra241-cmdqv: Fix alignment failure at max_n_shift
> > iommu/tegra241-cmdqv: Read SMMU IDR1.CMDQS instead of hardcoding
> >
> > Vasant Hegde (1):
> > iommu: Introduce iommu_group_mutex_assert()
> >
> > Yi Liu (1):
> > iommu: Pass domain to remove_dev_pasid() op
> >
>
>
>
> Acked-by: Philip Cox <philip.cox at canonical.com>
>
>
>
>
>
>
> > MAINTAINERS | 1 +
> > debian.aws/config/annotations | 1 +
> > drivers/iommu/Kconfig | 24 +-
> > drivers/iommu/arm/arm-smmu-v3/Makefile | 8 +-
> > drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3-sva.c | 526
> > +++++++------------
> > drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3-test.c | 571
> > ++++++++++++++++++++
> > drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c | 1666
> > ++++++++++++++++++++++++++++++++++++++++++-----------------
> > drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.h | 151 ++++--
> > drivers/iommu/arm/arm-smmu-v3/tegra241-cmdqv.c | 912
> > ++++++++++++++++++++++++++++++++
> > drivers/iommu/intel/iommu.c | 11 +-
> > drivers/iommu/iommu-sva.c | 4 +-
> > drivers/iommu/iommu.c | 40 +-
> > include/linux/iommu.h | 14 +-
> > 13 files changed, 3054 insertions(+), 875 deletions(-)
> > create mode 100644 drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3-test.c
> > create mode 100644 drivers/iommu/arm/arm-smmu-v3/tegra241-cmdqv.c
> >
>
> --
> kernel-team mailing list
> kernel-team at lists.ubuntu.com
> https://lists.ubuntu.com/mailman/listinfo/kernel-team
Acked-by: Massimiliano Pellizzer <massimiliano.pellizzer at canonical.com>
--
Massimiliano Pellizzer
More information about the kernel-team
mailing list