[PATCH 0/2] Reduce IOTLB flush when pass-through dGPU devices

From: Suravee Suthikulpanit
Date: Fri Nov 17 2017 - 16:12:06 EST


From: Suravee Suthikulpanit <suravee.suthikulpanit@xxxxxxx>

Currently, when pass-through dGPU to a guest VM, there are thousands
of IOTLB flush commands sent from IOMMU to end-point-device. This cause
performance issue when launching new VMs, and could cause IOTLB invalidate
time-out issue on certain dGPUs.

This can be avoided by adopting the new fast IOTLB flush APIs.

Cc: Alex Williamson <alex.williamson@xxxxxxxxxx>
Cc: Joerg Roedel <jroedel@xxxxxxx>

Suravee Suthikulpanit (2):
vfio/type1: Adopt fast IOTLB flush interface when unmap IOVAs
iommu/amd: Add support for fast IOTLB flushing

drivers/iommu/amd_iommu.c | 77 ++++++++++++++++++++++++++++++++++++++++-
drivers/iommu/amd_iommu_init.c | 2 --
drivers/iommu/amd_iommu_types.h | 2 ++
drivers/vfio/vfio_iommu_type1.c | 12 +++++--
4 files changed, 87 insertions(+), 6 deletions(-)

--
1.8.3.1