[PATCH v7 2/3] iommu/arm-smmu-v3: Introduce CFGI/TLBI-repeat workaround infrastructure

Nicolin Chen nicolinc at nvidia.com
Sun Jul 19 13:37:20 PDT 2026


On Sat, Jul 18, 2026 at 01:45:48PM -0300, Jason Gunthorpe wrote:
> On Fri, Jul 17, 2026 at 02:16:53PM -0700, Nicolin Chen wrote:
> > Given the use cases on Tegra264, instead of patching the iommufd
> > path as this patch does, perhaps we should simply spit a WARN in
> > arm_vsmmu_init():
> 
> Nope, in this case you need to make the errata discoverable through
> the viommu info so the VMM can validate the right DT is used, or apply
> the WAR itself before submitting commands to the kernel.

OK.

> Some kind of new flag in the viommu struct I guess

struct iommu_hw_info_arm_smmuv3 has an unused "flags" field:

/**
 * enum iommu_hw_info_smmuv3_flags - Flags for ARM SMMUv3 hw_info
 * @IOMMU_HW_INFO_SMMUV3_ERRATA_REPEAT_TLBI_CFGI: If set, user space must issue
 *                                                TLBI/CFGI+SYNC commands twice
 *                                                due to a hardware erratum:
 *                                                <link-to-erratum>
 */
enum iommu_hw_info_smmuv3_flags {
	IOMMU_HW_INFO_SMMUV3_ERRATA_REPEAT_TLBI_CFGI = 1 << 0,
};

Thanks
Nicolin



More information about the linux-arm-kernel mailing list