This RFC series aims at enabling KVM platform device passthrough. It implements a VFIO platform device which is bound to be dynamically instantiated using -device option.
The VFIO platform device uses an host VFIO platform driver which must be bound to the assigned device prior to the QEMU system start. - the guest can directly access the device register space - assigned device IRQs are transparently routed to the guest by QEMU/KVM (2 methods currently are supported) - iommu is transparently programmed to prevent the device from accessing physical pages outside of the guest address space The patch series was fully reworked between v3 and v4 to ease the review of PCI modifications. Dynamic instantiation from command line was cleaned up thanks to Alex Graf "Dynamic sysbus device allocation support" patch series and its porting onto machvirt. the patch relies on the following QEMU patch series: - Alex Graf's "Dynamic sysbus device allocation support" http://lists.gnu.org/archive/html/qemu-ppc/2014-07/msg00047.html - "machvirt dynamic sysbus device instantiation" Port Alex mechanics from e500 to virt. Propose to implement device tree generation in devices instead of machine file The patch series is made of the following patch files: 1-6) Modifications to PCI code to prepare for VFIO platform device: 7) split of PCI specific code and generic code (move) 8) EXEC_FLAG setting 9) creation of the VFIO platform device, without irqfd support (MMIO direct access and IRQ assignment). 10-11) addition of irqfd/virqfd support 12) capability to dynamically instantiate the device 13) example derived VFIO device: calxeda xgmac v3->v4 changes (Eric Auger, Alvise Rigo) - rebase on last VFIO PCI code (v2.1.0-rc0) - full git history rework to ease PCI code change review - mv include files in hw/vfio - DPRINTF reformatting temporarily moved out - support of VFIO virq (removal of resamplefd handler on user-side) - integration with sysbus dynamic instantiation framwork - removal of unrealize and cleanup routines until it is better understood what is really needed - Support of VFIO for Amba devices should be handled in an inherited device to specialize the device tree generation (clock handle currently missing in framework however) - "Always use eventfd as notifying mechanism" temporarily moved out - static instantiation is not mainstream (although it remains possible) note if static instantiation is used, irqfd must be setup in machine file when virtual IRQ is known - create the GSI routing table on qemu side v2->v3 changes (Alvise Rigo, Eric Auger): - Following Alex W recommandations, further efforts to factorize the code between PCI:introduction of VFIODevice and VFIORegion as base classes - unique reset handler for platform and PCI - cleanup following Kim's comments - multiple IRQ support mechanics should be in place although not tested - Better handling of MMIO multiple regions - New features and fixes by Alvise (multiple compat string, exec flag, force eventfd usage, amba device tree support) - irqfd support v1->v2 changes (Kim Phillips, Eric Auger): - IRQ initial support (legacy mode where eventfds are handled on user side) - hacked dynamic instantiation v1 (Kim Phillips): - initial split between PCI and platform - MMIO support only - static instantiation This patch has the following kernel side dependencies: - [RFC Patch v6 0/20] VFIO support for platform devices https://www.mail-archive.com/kvm@vger.kernel.org/msg103247.html - [Patch] ARM: KVM: Handle IPA unmapping on memory region deletion https://patches.linaro.org/27691/ - [PATCH v3] ARM: KVM: add irqfd and irq routing support https://patches.linaro.org/32261/ - [PATCH] ARM: KVM: Enable the KVM-VFIO device https://lists.cs.columbia.edu/pipermail/kvmarm/2014-March/008629.html - [PATCH v2] ARM: KVM: user_mem_abort: support stage 2 MMIO page mapping http://www.spinics.net/lists/kvm/msg105083.html The patch series was tested on Calxeda Midway (ARMv7) where one xgmac is assigned to KVM host while the second one is assigned to the guest. Unfortunately a single IRQ is exercised. Next steps: - use of "ARM: Forwarding physical interrupts to a guest VM" - unbind/migration/reset problematics Here are the instructions to test on a Calxeda Midway: https://wiki.linaro.org/LEG/Engineering/Virtualization/Platform_Device_Passthrough_on_Midway git://git.linaro.org/people/eric.auger/linux.git (branch irqfd_integ_v3) git://git.linaro.org/people/eric.auger/qemu.git (branch vfio_integ_v4) Best Regards Eric Alvise Rigo (1): hw/vfio/common: Add EXEC_FLAG to VFIO DMA mappings Eric Auger (11): hw/vfio/pci: Rename VFIODevice into VFIOPCIDevice hw/vfio/pci: Remove unneeded include files hw/vfio/pci: introduce VFIODevice hw/vfio/pci: Introduce VFIORegion hw/vfio/pci: split vfio_get_device hw/vfio: create common module hw/vfio/platform: add vfio-platform support hw/intc/arm_gic_kvm: enable irqfd and set routing table hw/vfio/platform: Add irqfd support hw/vfio/platform: add default dt generation for vfio device hw/vfio: add an example calxeda_xgmac Kim Phillips (1): vfio: move hw/misc/vfio.c to hw/vfio/pci.c Move vfio.h into include/hw/vfio LICENSE | 2 +- MAINTAINERS | 2 +- hw/Makefile.objs | 1 + hw/intc/arm_gic_kvm.c | 11 + hw/misc/Makefile.objs | 1 - hw/ppc/spapr_pci_vfio.c | 2 +- hw/vfio/Makefile.objs | 6 + hw/vfio/calxeda_xgmac.c | 165 +++++ hw/vfio/common.c | 1003 +++++++++++++++++++++++++ hw/{misc/vfio.c => vfio/pci.c} | 1514 +++++++------------------------------- hw/vfio/platform.c | 766 +++++++++++++++++++ include/hw/vfio/vfio-common.h | 149 ++++ include/hw/vfio/vfio-platform.h | 75 ++ include/hw/{misc => vfio}/vfio.h | 0 linux-headers/linux/vfio.h | 2 + 15 files changed, 2465 insertions(+), 1234 deletions(-) create mode 100644 hw/vfio/Makefile.objs create mode 100644 hw/vfio/calxeda_xgmac.c create mode 100644 hw/vfio/common.c rename hw/{misc/vfio.c => vfio/pci.c} (71%) create mode 100644 hw/vfio/platform.c create mode 100644 include/hw/vfio/vfio-common.h create mode 100644 include/hw/vfio/vfio-platform.h rename include/hw/{misc => vfio}/vfio.h (100%) -- 1.8.3.2
