> -----Original Message----- > From: Nathan Chen <nath...@nvidia.com> > Sent: Wednesday, July 2, 2025 2:02 AM > To: qemu-devel@nongnu.org > Cc: berra...@redhat.com; ddut...@redhat.com; eric.au...@redhat.com; > gustavo.rom...@linaro.org; imamm...@redhat.com; j...@nvidia.com; > jiangkunkun <jiangkun...@huawei.com>; Jonathan Cameron > <jonathan.came...@huawei.com>; Linuxarm <linux...@huawei.com>; > mo...@nvidia.com; nath...@nvidia.com; nicol...@nvidia.com; > peter.mayd...@linaro.org; qemu-...@nongnu.org; Shameerali Kolothum > Thodi <shameerali.kolothum.th...@huawei.com>; smost...@google.com; > Wangzhou (B) <wangzh...@hisilicon.com>; zhangfei....@linaro.org > Subject: Re: [PATCH v5 00/11] hw/arm/virt: Add support for user creatable > SMMUv3 device > > > To address this, patch #6 in the series introduces a new helper > > function pci_setup_iommu_per_bus(), which explicitly sets the > > iommu_per_bus field in the PCIBus structure. This allows > > pci_device_get_iommu_bus_devfn() to retrieve IOMMU ops based > > on the specific bus. > > > > This patch series introduces support for a user-creatable SMMUv3 > device > > (-device arm-smmuv3) in QEMU. > > Tested-by: Nathan Chen <nath...@nvidia.com> > > I re-ran the test from v3 [0] and am able to create 16 SMMUv3 devices in > a qemu VM with emulated devices properly associated with the guest > SMMUs > in guest sysfs - verified with some guest SMMUs having two or three > emulated NICs assigned to them while other guest SMMUs have a minimum > of > one assigned. > > Removing SMMUv3 devices from the VM config described above, I do not > observe the problematic behavior where devices behind PXBs without > SMMUs > erroneously use the address space from pcie.0's SMMU. I removed SMMUv3 > devices from PXBs with one, two, and three emulated NICs assigned to > them. Below are the guest topology and qemu command used where > SMMUv3 > devices are excluded from the original test:
Cool. Appreciate testing that specifically. Thanks, Shameer
