mbox series

[0/3] vfio/type1: batch page pinning

Message ID 20210203204756.125734-1-daniel.m.jordan@oracle.com (mailing list archive)
Headers show
Series vfio/type1: batch page pinning | expand


Daniel Jordan Feb. 3, 2021, 8:47 p.m. UTC

The VFIO type1 driver is calling pin_user_pages_remote() once per 4k page, so
let's do it once per 512 4k pages to bring VFIO in line with other drivers such
as IB and vDPA.

qemu guests with at least 2G memory start about 8% faster on a Xeon server,
with more detailed results in the last changelog.

Thanks to Matthew, who first suggested the idea to me.


Test Cases

 1) qemu passthrough with IOMMU-capable PCI device

 2) standalone program to hit
        vfio_pin_map_dma() -> vfio_pin_pages_remote()

 3) standalone program to hit
        vfio_iommu_replay() -> vfio_pin_pages_remote()

Each was run...

 - with varying sizes
 - with/without disable_hugepages=1
 - with/without LOCKED_VM exceeded

I didn't test vfio_pin_page_external() because there was no readily available
hardware, but the changes there are pretty minimal.

Series based on v5.11-rc6.

Daniel Jordan (3):
  vfio/type1: change success value of vaddr_get_pfn()
  vfio/type1: prepare for batched pinning with struct vfio_batch
  vfio/type1: batch page pinning

 drivers/vfio/vfio_iommu_type1.c | 213 +++++++++++++++++++++++---------
 1 file changed, 154 insertions(+), 59 deletions(-)

base-commit: 1048ba83fb1c00cd24172e23e8263972f6b5d9ac