Enable Inter-VM Communication Based on Ivshmem

You can use inter-VM communication based on the ivshmem dm-land solution or hv-land solution, according to the usage scenario needs. (See ACRN Shared Memory Based Inter-VM Communication for a high-level description of these solutions.) While both solutions can be used at the same time, VMs using different solutions cannot communicate with each other.

Enable Ivshmem Support

The ivshmem solution is disabled by default in ACRN. You can enable it using the ACRN configuration toolset with these steps:

  • Enable ivshmem via ACRN configuration tool GUI.

    • Set hv.FEATURES.IVSHMEM.IVSHMEM_ENABLED to y

    • Edit hv.FEATURES.IVSHMEM.IVSHMEM_REGION to specify the shared memory name, size and communication VMs. The IVSHMEM_REGION format is shm_name,shm_size,VM IDs:

      • shm_name - Specify a shared memory name. The name needs to start with the hv:/ prefix for hv-land, or dm:/ for dm-land. For example, hv:/shm_region_0 for hv-land and dm:/shm_region_0 for dm-land.

      • shm_size - Specify a shared memory size. The unit is megabyte. The size ranges from 2 megabytes to 512 megabytes and must be a power of 2 megabytes. For example, to set up a shared memory of 2 megabytes, use 2 instead of shm_size.

      • VM IDs - Specify the VM IDs to use the same shared memory communication and separate it with :. For example, the communication between VM0 and VM2, it can be written as 0:2

  • Build with the XML configuration, refer to Build ACRN From Source.

Ivshmem DM-Land Usage

Follow Enable Ivshmem Support and add below line as an acrn-dm boot parameter:

-s slot,ivshmem,shm_name,shm_size

where

  • -s slot - Specify the virtual PCI slot number

  • ivshmem - Virtual PCI device emulating the Shared Memory

  • shm_name - Specify a shared memory name. This shm_name must be listed in hv.FEATURES.IVSHMEM.IVSHMEM_REGION in Enable Ivshmem Support section and needs to start with dm:/ prefix.

  • shm_size - Shared memory size of selected shm_name.

There are two ways to insert above boot parameter for acrn-dm

  • Manually edit launch script file, in this case, user shall ensure that both shm_name and shm_size match with that are defined via configuration tool GUI.

  • Use the command following below format to create a launch script, when IVSHMEM is enabled

    and hv.FEATURES.IVSHMEM.IVSHMEM_REGION is properly configured via configuration tool GUI.

    python3 misc/config_tools/launch_config/launch_cfg_gen.py \
    --board <path_to_your_boardxml> \
    --scenario <path_to_your_scenarioxml> \
    --launch <path_to_your_launched_script_xml>  \
    --uosid <desired_single_vmid_or_0_for_all_vmids>
    

Note

This device can be used with real-time VM (RTVM) as well.

Ivshmem HV-Land Usage

Follow Enable Ivshmem Support to setup HV-Land Ivshmem support.

Ivshmem Notification Mechanism

Notification (doorbell) of ivshmem device allows VMs with ivshmem devices enabled to notify (interrupt) each other following this flow:

Notification Sender (VM):

VM triggers the notification to target VM by writing target Peer ID (Equals to VM ID of target VM) and vector index to doorbell register of ivshmem device, the layout of doorbell register is described in ACRN Shared Memory Based Inter-VM Communication.

Hypervisor:

When doorbell register is programmed, hypervisor will search the target VM by target Peer ID and inject MSI interrupt to the target VM.

Notification Receiver (VM):

VM receives MSI interrupt and forward it to related application.

ACRN supports up to 8 (MSI-X) interrupt vectors for ivshmem device. Guest VMs shall implement their own mechanism to forward MSI interrupts to applications.

Note

Notification is supported only for HV-land ivshmem devices. (Future support may include notification for DM-land ivshmem devices.)

Inter-VM Communication Examples

DM-Land Example

This example uses dm-land inter-VM communication between two Linux-based post-launched VMs (VM1 and VM2).

Note

An ivshmem Windows driver exists and can be found here.

  1. Add a new virtual PCI device for both VMs: the device type is ivshmem, shared memory name is dm:/test, and shared memory size is 2MB. Both VMs must have the same shared memory name and size:

    • VM1 Launch Script Sample

      acrn-dm -A -m $mem_size -s 0:0,hostbridge \
       -s 2,pci-gvt -G "$2" \
       -s 5,virtio-console,@stdio:stdio_port \
       -s 6,virtio-hyper_dmabuf \
       -s 3,virtio-blk,/home/acrn/uos1.img \
       -s 4,virtio-net,tap0 \
       -s 6,ivshmem,dm:/test,2 \
       -s 7,virtio-rnd \
       --ovmf /usr/share/acrn/bios/OVMF.fd \
       $vm_name
      
    • VM2 Launch Script Sample

      acrn-dm -A -m $mem_size -s 0:0,hostbridge \
       -s 2,pci-gvt -G "$2" \
       -s 3,virtio-blk,/home/acrn/uos2.img \
       -s 4,virtio-net,tap0 \
       -s 5,ivshmem,dm:/test,2 \
       --ovmf /usr/share/acrn/bios/OVMF.fd \
       $vm_name
      
  2. Boot two VMs and use lspci | grep "shared memory" to verify that the virtual device is ready for each VM.

    • For VM1, it shows 00:06.0 RAM memory: Red Hat, Inc. Inter-VM shared memory (rev 01)

    • For VM2, it shows 00:05.0 RAM memory: Red Hat, Inc. Inter-VM shared memory (rev 01)

  3. As recorded in the PCI ID Repository, the ivshmem device vendor ID is 1af4 (Red Hat) and device ID is 1110 (Inter-VM shared memory). Use these commands to probe the device:

    $ sudo modprobe uio
    $ sudo modprobe uio_pci_generic
    $ sudo echo "1af4 1110" > /sys/bus/pci/drivers/uio_pci_generic/new_id
    

Note

These commands are applicable to Linux-based guests with CONFIG_UIO and CONFIG_UIO_PCI_GENERIC enabled.

  1. Finally, a user application can get the shared memory base address from the ivshmem device BAR resource (/sys/class/uio/uioX/device/resource2) and the shared memory size from the ivshmem device config resource (/sys/class/uio/uioX/device/config).

    The X in uioX above, is a number that can be retrieved using the ls command:

    • For VM1 use ls -lh /sys/bus/pci/devices/0000:00:06.0/uio

    • For VM2 use ls -lh /sys/bus/pci/devices/0000:00:05.0/uio

HV-Land Example

This example uses hv-land inter-VM communication between two Linux-based VMs (VM0 is a pre-launched VM and VM2 is a post-launched VM).

  1. Make a copy of the predefined hybrid_rt scenario on whl-ipc-i5 (available at acrn-hypervisor/misc/config_tools/data/whl-ipc-i5/hybrid_rt.xml) and configure shared memory for the communication between VM0 and VM2. The shared memory name is hv:/shm_region_0, and shared memory size is 2M bytes. The resulting scenario XML should look like this:

    <IVSHMEM desc="IVSHMEM configuration">
           <IVSHMEM_ENABLED>y</IVSHMEM_ENABLED>
           <IVSHMEM_REGION>hv:/shm_region_0, 2, 0:2</IVSHMEM_REGION>
    </IVSHMEM>
    
  2. Build ACRN based on the XML configuration for hybrid_rt scenario on whl-ipc-i5 board:

    make BOARD=whl-ipc-i5 SCENARIO=<path/to/edited/scenario.xml> TARGET_DIR=xxx
    
  3. Add a new virtual PCI device for VM2 (post-launched VM): the device type is ivshmem, shared memory name is hv:/shm_region_0, and shared memory size is 2MB.

    • VM2 Launch Script Sample

      acrn-dm -A -m $mem_size -s 0:0,hostbridge \
       -s 2,pci-gvt -G "$2" \
       -s 3,virtio-blk,/home/acrn/uos2.img \
       -s 4,virtio-net,tap0 \
       -s 5,ivshmem,hv:/shm_region_0,2 \
       --ovmf /usr/share/acrn/bios/OVMF.fd \
       $vm_name
      
  4. Continue following the dm-land steps 2-4 and the ivshmem device BDF may be different depending on the configuration.