// Copyright 2018 The Chromium OS Authors. All rights reserved. // Use of this source code is governed by a BSD-style license that can be // found in the LICENSE file. pub mod fdt; extern crate byteorder; extern crate devices; extern crate io_jail; extern crate kernel_cmdline; extern crate kvm; extern crate libc; extern crate resources; extern crate sync; extern crate sys_util; use std::collections::BTreeMap; use std::fmt; use std::fs::File; use std::io::{Read, Seek, SeekFrom}; use std::os::unix::io::AsRawFd; use std::result; use std::sync::Arc; use devices::virtio::VirtioDevice; use devices::{ Bus, BusDevice, BusError, PciDevice, PciDeviceError, PciInterruptPin, PciRoot, ProxyDevice, Serial, }; use io_jail::Minijail; use kvm::{IoeventAddress, Kvm, Vcpu, Vm}; use resources::SystemAllocator; use sync::Mutex; use sys_util::{syslog, EventFd, GuestAddress, GuestMemory, GuestMemoryError}; pub type Result = result::Result>; /// Holds the pieces needed to build a VM. Passed to `build_vm` in the `LinuxArch` trait below to /// create a `RunnableLinuxVm`. pub struct VmComponents { pub memory_mb: u64, pub vcpu_count: u32, pub kernel_image: File, pub android_fstab: Option, pub extra_kernel_params: Vec, pub wayland_dmabuf: bool, } /// Holds the elements needed to run a Linux VM. Created by `build_vm`. pub struct RunnableLinuxVm { pub vm: Vm, pub kvm: Kvm, pub resources: SystemAllocator, pub stdio_serial: Arc>, pub exit_evt: EventFd, pub vcpus: Vec, pub irq_chip: Option, pub io_bus: Bus, pub mmio_bus: Bus, pub pid_debug_label_map: BTreeMap, } /// The device and optional jail. pub struct VirtioDeviceStub { pub dev: Box, pub jail: Option, } /// Trait which is implemented for each Linux Architecture in order to /// set up the memory, cpus, and system devices and to boot the kernel. pub trait LinuxArch { /// Takes `VmComponents` and generates a `RunnableLinuxVm`. /// /// # Arguments /// /// * `components` - Parts to use to build the VM. /// * `virtio_devs` - Function to generate a list of virtio devices. fn build_vm(components: VmComponents, virtio_devs: F) -> Result where F: FnOnce( &GuestMemory, &EventFd, ) -> Result, Option)>>; } /// Errors for device manager. #[derive(Debug)] pub enum DeviceRegistrationError { /// Could not allocate IO space for the device. AllocateIoAddrs(PciDeviceError), /// Could not allocate an IRQ number. AllocateIrq, /// Could not create the mmio device to wrap a VirtioDevice. CreateMmioDevice(sys_util::Error), /// Could not create an event fd. EventFdCreate(sys_util::Error), /// Could not add a device to the mmio bus. MmioInsert(BusError), /// Failed to register ioevent with VM. RegisterIoevent(sys_util::Error), /// Failed to register irq eventfd with VM. RegisterIrqfd(sys_util::Error), /// Failed to initialize proxy device for jailed device. ProxyDeviceCreation(devices::ProxyError), /// Appending to kernel command line failed. Cmdline(kernel_cmdline::Error), /// No more IRQs are available. IrqsExhausted, /// No more MMIO space available. AddrsExhausted, } impl fmt::Display for DeviceRegistrationError { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { match self { DeviceRegistrationError::AllocateIoAddrs(e) => { write!(f, "Allocating IO addresses: {:?}", e) } DeviceRegistrationError::AllocateIrq => write!(f, "Allocating IRQ number"), DeviceRegistrationError::CreateMmioDevice(e) => { write!(f, "failed to create mmio device: {:?}", e) } DeviceRegistrationError::Cmdline(e) => { write!(f, "unable to add device to kernel command line: {}", e) } DeviceRegistrationError::EventFdCreate(e) => { write!(f, "failed to create eventfd: {:?}", e) } DeviceRegistrationError::MmioInsert(e) => { write!(f, "failed to add to mmio bus: {:?}", e) } DeviceRegistrationError::RegisterIoevent(e) => { write!(f, "failed to register ioevent to VM: {:?}", e) } DeviceRegistrationError::RegisterIrqfd(e) => { write!(f, "failed to register irq eventfd to VM: {:?}", e) } DeviceRegistrationError::ProxyDeviceCreation(e) => { write!(f, "failed to create proxy device: {}", e) } DeviceRegistrationError::IrqsExhausted => write!(f, "no more IRQs are available"), DeviceRegistrationError::AddrsExhausted => write!(f, "no more addresses are available"), } } } /// Creates a root PCI device for use by this Vm. pub fn generate_pci_root( devices: Vec<(Box, Option)>, mmio_bus: &mut Bus, resources: &mut SystemAllocator, vm: &mut Vm, ) -> std::result::Result< (PciRoot, Vec<(u32, PciInterruptPin)>, BTreeMap), DeviceRegistrationError, > { let mut root = PciRoot::new(); let mut pci_irqs = Vec::new(); let mut pid_labels = BTreeMap::new(); for (dev_idx, (mut device, jail)) in devices.into_iter().enumerate() { let mut keep_fds = device.keep_fds(); syslog::push_fds(&mut keep_fds); let irqfd = EventFd::new().map_err(DeviceRegistrationError::EventFdCreate)?; let irq_resample_fd = EventFd::new().map_err(DeviceRegistrationError::EventFdCreate)?; let irq_num = resources .allocate_irq() .ok_or(DeviceRegistrationError::AllocateIrq)? as u32; let pci_irq_pin = match dev_idx % 4 { 0 => PciInterruptPin::IntA, 1 => PciInterruptPin::IntB, 2 => PciInterruptPin::IntC, 3 => PciInterruptPin::IntD, _ => panic!(""), // Obviously not possible, but the compiler is not smart enough. }; vm.register_irqfd_resample(&irqfd, &irq_resample_fd, irq_num) .map_err(DeviceRegistrationError::RegisterIrqfd)?; keep_fds.push(irqfd.as_raw_fd()); keep_fds.push(irq_resample_fd.as_raw_fd()); device.assign_irq(irqfd, irq_resample_fd, irq_num, pci_irq_pin); pci_irqs.push((dev_idx as u32, pci_irq_pin)); let ranges = device .allocate_io_bars(resources) .map_err(DeviceRegistrationError::AllocateIoAddrs)?; for (event, addr, datamatch) in device.ioeventfds() { let io_addr = IoeventAddress::Mmio(addr); vm.register_ioevent(&event, io_addr, datamatch) .map_err(DeviceRegistrationError::RegisterIoevent)?; keep_fds.push(event.as_raw_fd()); } let arced_dev: Arc> = if let Some(jail) = jail { let proxy = ProxyDevice::new(device, &jail, keep_fds) .map_err(DeviceRegistrationError::ProxyDeviceCreation)?; pid_labels.insert(proxy.pid() as u32, proxy.debug_label()); Arc::new(Mutex::new(proxy)) } else { Arc::new(Mutex::new(device)) }; root.add_device(arced_dev.clone()); for range in &ranges { mmio_bus .insert(arced_dev.clone(), range.0, range.1, true) .map_err(DeviceRegistrationError::MmioInsert)?; } } Ok((root, pci_irqs, pid_labels)) } /// Errors for image loading. #[derive(Debug)] pub enum LoadImageError { Seek(std::io::Error), ImageSizeTooLarge(u64), ReadToMemory(GuestMemoryError), } impl fmt::Display for LoadImageError { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { match self { LoadImageError::Seek(e) => write!(f, "Seek failed: {:?}", e), LoadImageError::ImageSizeTooLarge(size) => { write!(f, "Image size too large: {:?}", size) } LoadImageError::ReadToMemory(e) => { write!(f, "Reading image into memory failed: {:?}", e) } } } } /// Load an image from a file into guest memory. /// /// # Arguments /// /// * `guest_mem` - The memory to be used by the guest. /// * `guest_addr` - The starting address to load the image in the guest memory. /// * `max_size` - The amount of space in bytes available in the guest memory for the image. /// * `image` - The file containing the image to be loaded. /// /// The size in bytes of the loaded image is returned. pub fn load_image( guest_mem: &GuestMemory, image: &mut F, guest_addr: GuestAddress, max_size: u64, ) -> std::result::Result where F: Read + Seek, { let size = image.seek(SeekFrom::End(0)).map_err(LoadImageError::Seek)?; if size > usize::max_value() as u64 || size > max_size { return Err(LoadImageError::ImageSizeTooLarge(size)); } // This is safe due to the bounds check above. let size = size as usize; image .seek(SeekFrom::Start(0)) .map_err(LoadImageError::Seek)?; guest_mem .read_to_memory(guest_addr, image, size) .map_err(LoadImageError::ReadToMemory)?; Ok(size) }