// Copyright 2017 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.

mod fdt;

const E820_RAM: u32 = 1;
const SETUP_DTB: u32 = 2;
const X86_64_FDT_MAX_SIZE: u64 = 0x200000;

#[allow(dead_code)]
#[allow(non_upper_case_globals)]
#[allow(non_camel_case_types)]
#[allow(non_snake_case)]
mod bootparam;

// boot_params is just a series of ints, it is safe to initialize it.
unsafe impl data_model::DataInit for bootparam::boot_params {}

#[allow(dead_code)]
#[allow(non_upper_case_globals)]
mod msr_index;

#[allow(dead_code)]
#[allow(non_upper_case_globals)]
#[allow(non_camel_case_types)]
#[allow(clippy::all)]
mod mpspec;
// These mpspec types are only data, reading them from data is a safe initialization.
unsafe impl data_model::DataInit for mpspec::mpc_bus {}
unsafe impl data_model::DataInit for mpspec::mpc_cpu {}
unsafe impl data_model::DataInit for mpspec::mpc_intsrc {}
unsafe impl data_model::DataInit for mpspec::mpc_ioapic {}
unsafe impl data_model::DataInit for mpspec::mpc_table {}
unsafe impl data_model::DataInit for mpspec::mpc_lintsrc {}
unsafe impl data_model::DataInit for mpspec::mpf_intel {}

mod acpi;
mod bzimage;
mod cpuid;
mod gdt;
mod interrupts;
mod mptable;
mod regs;
mod smbios;

use std::collections::BTreeMap;
use std::error::Error as StdError;
use std::ffi::{CStr, CString};
use std::fmt::{self, Display};
use std::fs::File;
use std::io::{self, Seek};
use std::mem;
use std::sync::Arc;

use crate::bootparam::boot_params;
use acpi_tables::aml::Aml;
use acpi_tables::sdt::SDT;
use arch::{
    get_serial_cmdline, GetSerialCmdlineError, RunnableLinuxVm, SerialHardware, SerialParameters,
    VmComponents, VmImage,
};
use devices::split_irqchip_common::GsiRelay;
use devices::{
    Ioapic, PciAddress, PciConfigIo, PciDevice, PciInterruptPin, Pic, IOAPIC_BASE_ADDRESS,
    IOAPIC_MEM_LENGTH_BYTES,
};
use io_jail::Minijail;
use kvm::*;
use msg_socket::MsgOnSocket;
use remain::sorted;
use resources::SystemAllocator;
use sync::Mutex;
use sys_util::{Clock, EventFd, GuestAddress, GuestMemory, GuestMemoryError};
use vm_control::VmIrqRequestSocket;

#[sorted]
#[derive(Debug)]
pub enum Error {
    AllocateIOResouce(resources::Error),
    AllocateIrq,
    CloneEventFd(sys_util::Error),
    Cmdline(kernel_cmdline::Error),
    ConfigureSystem,
    CreateDevices(Box<dyn StdError>),
    CreateEventFd(sys_util::Error),
    CreateFdt(arch::fdt::Error),
    CreateIoapicDevice(sys_util::Error),
    CreateIrqChip(sys_util::Error),
    CreateKvm(sys_util::Error),
    CreatePciRoot(arch::DeviceRegistrationError),
    CreatePit(sys_util::Error),
    CreatePitDevice(devices::PitError),
    CreateSerialDevices(arch::DeviceRegistrationError),
    CreateSocket(io::Error),
    CreateVcpu(sys_util::Error),
    CreateVm(sys_util::Error),
    E820Configuration,
    EnableSplitIrqchip(sys_util::Error),
    GetSerialCmdline(GetSerialCmdlineError),
    KernelOffsetPastEnd,
    LoadBios(io::Error),
    LoadBzImage(bzimage::Error),
    LoadCmdline(kernel_loader::Error),
    LoadInitrd(arch::LoadImageError),
    LoadKernel(kernel_loader::Error),
    Pstore(arch::pstore::Error),
    RegisterIrqfd(sys_util::Error),
    RegisterVsock(arch::DeviceRegistrationError),
    SetLint(interrupts::Error),
    SetTssAddr(sys_util::Error),
    SetupCpuid(cpuid::Error),
    SetupFpu(regs::Error),
    SetupGuestMemory(GuestMemoryError),
    SetupMptable(mptable::Error),
    SetupMsrs(regs::Error),
    SetupRegs(regs::Error),
    SetupSmbios(smbios::Error),
    SetupSregs(regs::Error),
    ZeroPagePastRamEnd,
    ZeroPageSetup,
}

impl Display for Error {
    #[remain::check]
    fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
        use self::Error::*;

        #[sorted]
        match self {
            AllocateIOResouce(e) => write!(f, "error allocating IO resource: {}", e),
            AllocateIrq => write!(f, "error allocating a single irq"),
            CloneEventFd(e) => write!(f, "unable to clone an EventFd: {}", e),
            Cmdline(e) => write!(f, "the given kernel command line was invalid: {}", e),
            ConfigureSystem => write!(f, "error configuring the system"),
            CreateDevices(e) => write!(f, "error creating devices: {}", e),
            CreateEventFd(e) => write!(f, "unable to make an EventFd: {}", e),
            CreateFdt(e) => write!(f, "failed to create fdt: {}", e),
            CreateIoapicDevice(e) => write!(f, "failed to create IOAPIC device: {}", e),
            CreateIrqChip(e) => write!(f, "failed to create irq chip: {}", e),
            CreateKvm(e) => write!(f, "failed to open /dev/kvm: {}", e),
            CreatePciRoot(e) => write!(f, "failed to create a PCI root hub: {}", e),
            CreatePit(e) => write!(f, "unable to create PIT: {}", e),
            CreatePitDevice(e) => write!(f, "unable to make PIT device: {}", e),
            CreateSerialDevices(e) => write!(f, "unable to create serial devices: {}", e),
            CreateSocket(e) => write!(f, "failed to create socket: {}", e),
            CreateVcpu(e) => write!(f, "failed to create VCPU: {}", e),
            CreateVm(e) => write!(f, "failed to create VM: {}", e),
            E820Configuration => write!(f, "invalid e820 setup params"),
            EnableSplitIrqchip(e) => write!(f, "failed to enable split irqchip: {}", e),
            GetSerialCmdline(e) => write!(f, "failed to get serial cmdline: {}", e),
            KernelOffsetPastEnd => write!(f, "the kernel extends past the end of RAM"),
            LoadBios(e) => write!(f, "error loading bios: {}", e),
            LoadBzImage(e) => write!(f, "error loading kernel bzImage: {}", e),
            LoadCmdline(e) => write!(f, "error loading command line: {}", e),
            LoadInitrd(e) => write!(f, "error loading initrd: {}", e),
            LoadKernel(e) => write!(f, "error loading Kernel: {}", e),
            Pstore(e) => write!(f, "failed to allocate pstore region: {}", e),
            RegisterIrqfd(e) => write!(f, "error registering an IrqFd: {}", e),
            RegisterVsock(e) => write!(f, "error registering virtual socket device: {}", e),
            SetLint(e) => write!(f, "failed to set interrupts: {}", e),
            SetTssAddr(e) => write!(f, "failed to set tss addr: {}", e),
            SetupCpuid(e) => write!(f, "failed to set up cpuid: {}", e),
            SetupFpu(e) => write!(f, "failed to set up FPU: {}", e),
            SetupGuestMemory(e) => write!(f, "failed to set up guest memory: {}", e),
            SetupMptable(e) => write!(f, "failed to set up mptable: {}", e),
            SetupMsrs(e) => write!(f, "failed to set up MSRs: {}", e),
            SetupRegs(e) => write!(f, "failed to set up registers: {}", e),
            SetupSmbios(e) => write!(f, "failed to set up SMBIOS: {}", e),
            SetupSregs(e) => write!(f, "failed to set up sregs: {}", e),
            ZeroPagePastRamEnd => write!(f, "the zero page extends past the end of guest_mem"),
            ZeroPageSetup => write!(f, "error writing the zero page of guest memory"),
        }
    }
}

pub type Result<T> = std::result::Result<T, Error>;

impl std::error::Error for Error {}

pub struct X8664arch;

const BOOT_STACK_POINTER: u64 = 0x8000;
// Make sure it align to 256MB for MTRR convenient
const MEM_32BIT_GAP_SIZE: u64 = 768 << 20;
const FIRST_ADDR_PAST_32BITS: u64 = 1 << 32;
const END_ADDR_BEFORE_32BITS: u64 = FIRST_ADDR_PAST_32BITS - MEM_32BIT_GAP_SIZE;
const MMIO_SIZE: u64 = MEM_32BIT_GAP_SIZE - 0x8000000;
const KERNEL_64BIT_ENTRY_OFFSET: u64 = 0x200;
const ZERO_PAGE_OFFSET: u64 = 0x7000;
/// The x86 reset vector for i386+ and x86_64 puts the processor into an "unreal mode" where it
/// can access the last 1 MB of the 32-bit address space in 16-bit mode, and starts the instruction
/// pointer at the effective physical address 0xFFFFFFF0.
const BIOS_LEN: usize = 1 << 20;
const BIOS_START: u64 = FIRST_ADDR_PAST_32BITS - (BIOS_LEN as u64);

const KERNEL_START_OFFSET: u64 = 0x200000;
const CMDLINE_OFFSET: u64 = 0x20000;
const CMDLINE_MAX_SIZE: u64 = KERNEL_START_OFFSET - CMDLINE_OFFSET;
const X86_64_SERIAL_1_3_IRQ: u32 = 4;
const X86_64_SERIAL_2_4_IRQ: u32 = 3;
// X86_64_SCI_IRQ is used to fill the ACPI FACP table.
// The sci_irq number is better to be a legacy
// IRQ number which is less than 16(actually most of the
// platforms have fixed IRQ number 9). So we can
// reserve the IRQ number 5 for SCI and let the
// the other devices starts from next.
pub const X86_64_SCI_IRQ: u32 = 5;
// So the IRQ_BASE start from SCI_IRQ + 1
pub const X86_64_IRQ_BASE: u32 = X86_64_SCI_IRQ + 1;
const ACPI_HI_RSDP_WINDOW_BASE: u64 = 0x000E0000;

fn configure_system(
    guest_mem: &GuestMemory,
    _mem_size: u64,
    kernel_addr: GuestAddress,
    cmdline_addr: GuestAddress,
    cmdline_size: usize,
    num_cpus: u8,
    pci_irqs: Vec<(PciAddress, u32, PciInterruptPin)>,
    setup_data: Option<GuestAddress>,
    initrd: Option<(GuestAddress, usize)>,
    mut params: boot_params,
    acpi_dev_resource: acpi::ACPIDevResource,
) -> Result<()> {
    const EBDA_START: u64 = 0x0009fc00;
    const KERNEL_BOOT_FLAG_MAGIC: u16 = 0xaa55;
    const KERNEL_HDR_MAGIC: u32 = 0x53726448;
    const KERNEL_LOADER_OTHER: u8 = 0xff;
    const KERNEL_MIN_ALIGNMENT_BYTES: u32 = 0x1000000; // Must be non-zero.
    let first_addr_past_32bits = GuestAddress(FIRST_ADDR_PAST_32BITS);
    let end_32bit_gap_start = GuestAddress(END_ADDR_BEFORE_32BITS);

    // Note that this puts the mptable at 0x0 in guest physical memory.
    mptable::setup_mptable(guest_mem, num_cpus, pci_irqs).map_err(Error::SetupMptable)?;

    smbios::setup_smbios(guest_mem).map_err(Error::SetupSmbios)?;

    params.hdr.type_of_loader = KERNEL_LOADER_OTHER;
    params.hdr.boot_flag = KERNEL_BOOT_FLAG_MAGIC;
    params.hdr.header = KERNEL_HDR_MAGIC;
    params.hdr.cmd_line_ptr = cmdline_addr.offset() as u32;
    params.hdr.cmdline_size = cmdline_size as u32;
    params.hdr.kernel_alignment = KERNEL_MIN_ALIGNMENT_BYTES;
    if let Some(setup_data) = setup_data {
        params.hdr.setup_data = setup_data.offset();
    }
    if let Some((initrd_addr, initrd_size)) = initrd {
        params.hdr.ramdisk_image = initrd_addr.offset() as u32;
        params.hdr.ramdisk_size = initrd_size as u32;
    }

    add_e820_entry(&mut params, 0, EBDA_START, E820_RAM)?;

    let mem_end = guest_mem.end_addr();
    if mem_end < end_32bit_gap_start {
        add_e820_entry(
            &mut params,
            kernel_addr.offset() as u64,
            mem_end.offset_from(kernel_addr) as u64,
            E820_RAM,
        )?;
    } else {
        add_e820_entry(
            &mut params,
            kernel_addr.offset() as u64,
            end_32bit_gap_start.offset_from(kernel_addr) as u64,
            E820_RAM,
        )?;
        if mem_end > first_addr_past_32bits {
            add_e820_entry(
                &mut params,
                first_addr_past_32bits.offset() as u64,
                mem_end.offset_from(first_addr_past_32bits) as u64,
                E820_RAM,
            )?;
        }
    }

    let zero_page_addr = GuestAddress(ZERO_PAGE_OFFSET);
    guest_mem
        .checked_offset(zero_page_addr, mem::size_of::<boot_params>() as u64)
        .ok_or(Error::ZeroPagePastRamEnd)?;
    guest_mem
        .write_obj_at_addr(params, zero_page_addr)
        .map_err(|_| Error::ZeroPageSetup)?;

    if let Some(rsdp_addr) =
        acpi::create_acpi_tables(guest_mem, num_cpus, X86_64_SCI_IRQ, acpi_dev_resource)
    {
        params.acpi_rsdp_addr = rsdp_addr.0;
    }

    Ok(())
}

/// Add an e820 region to the e820 map.
/// Returns Ok(()) if successful, or an error if there is no space left in the map.
fn add_e820_entry(params: &mut boot_params, addr: u64, size: u64, mem_type: u32) -> Result<()> {
    if params.e820_entries >= params.e820_table.len() as u8 {
        return Err(Error::E820Configuration);
    }

    params.e820_table[params.e820_entries as usize].addr = addr;
    params.e820_table[params.e820_entries as usize].size = size;
    params.e820_table[params.e820_entries as usize].type_ = mem_type;
    params.e820_entries += 1;

    Ok(())
}

#[derive(Clone, Copy, Debug, MsgOnSocket)]
pub struct MemoryParams {
    /// Physical memory size in bytes for the VM.
    pub size: u64,
    pub has_bios: bool,
}

impl MemoryParams {
    fn new(components: &VmComponents) -> Self {
        let has_bios = match components.vm_image {
            VmImage::Bios(_) => true,
            _ => false,
        };

        MemoryParams {
            size: components.memory_size,
            has_bios,
        }
    }
}

/// Returns a Vec of the valid memory addresses.
/// These should be used to configure the GuestMemory structure for the platform.
/// For x86_64 all addresses are valid from the start of the kernel except a
/// carve out at the end of 32bit address space.
pub fn arch_memory_regions(params: MemoryParams) -> Vec<(GuestAddress, u64)> {
    let mem_end = GuestAddress(params.size);
    let first_addr_past_32bits = GuestAddress(FIRST_ADDR_PAST_32BITS);
    let end_32bit_gap_start = GuestAddress(END_ADDR_BEFORE_32BITS);

    let mut regions = Vec::new();
    if mem_end < end_32bit_gap_start {
        regions.push((GuestAddress(0), params.size));
        if params.has_bios {
            regions.push((GuestAddress(BIOS_START), BIOS_LEN as u64));
        }
    } else {
        regions.push((GuestAddress(0), end_32bit_gap_start.offset()));
        if params.has_bios {
            regions.push((GuestAddress(BIOS_START), BIOS_LEN as u64));
        }
        regions.push((
            first_addr_past_32bits,
            mem_end.offset_from(end_32bit_gap_start),
        ));
    }

    regions
}

impl arch::LinuxArch for X8664arch {
    type Error = Error;
    type MemoryParams = MemoryParams;

    fn build_vm<F, E>(
        mut components: VmComponents,
        split_irqchip: bool,
        ioapic_device_socket: VmIrqRequestSocket,
        serial_parameters: &BTreeMap<(SerialHardware, u8), SerialParameters>,
        serial_jail: Option<Minijail>,
        create_devices: F,
    ) -> Result<RunnableLinuxVm>
    where
        F: FnOnce(
            &GuestMemory,
            &mut Vm,
            &mut SystemAllocator,
            &EventFd,
        ) -> std::result::Result<Vec<(Box<dyn PciDevice>, Option<Minijail>)>, E>,
        E: StdError + 'static,
    {
        let mem_params = MemoryParams::new(&components);
        let mem = Self::setup_memory(mem_params)?;
        let mut resources = Self::get_resource_allocator(&mem, components.wayland_dmabuf);

        let kvm = Kvm::new().map_err(Error::CreateKvm)?;
        let mut vm = Self::create_vm(&kvm, split_irqchip, mem.clone())?;
        let vcpu_count = components.vcpu_count;
        let mut vcpus = Vec::with_capacity(vcpu_count as usize);
        for cpu_id in 0..vcpu_count {
            let vcpu = Vcpu::new(cpu_id as libc::c_ulong, &kvm, &vm).map_err(Error::CreateVcpu)?;
            if let VmImage::Kernel(_) = components.vm_image {
                Self::configure_vcpu(
                    vm.get_memory(),
                    &kvm,
                    &vm,
                    &vcpu,
                    cpu_id as u64,
                    vcpu_count as u64,
                )?;
            }
            vcpus.push(vcpu);
        }

        let vcpu_affinity = components.vcpu_affinity;

        let irq_chip = Self::create_irq_chip(&vm)?;

        let mut mmio_bus = devices::Bus::new();

        let exit_evt = EventFd::new().map_err(Error::CreateEventFd)?;

        let (split_irqchip, mut gsi_relay) = if split_irqchip {
            let gsi_relay = GsiRelay::new();
            let pic = Arc::new(Mutex::new(Pic::new()));
            let ioapic = Arc::new(Mutex::new(
                Ioapic::new(&mut vm, ioapic_device_socket).map_err(Error::CreateIoapicDevice)?,
            ));
            mmio_bus
                .insert(
                    ioapic.clone(),
                    IOAPIC_BASE_ADDRESS,
                    IOAPIC_MEM_LENGTH_BYTES,
                    false,
                )
                .unwrap();
            (Some((pic, ioapic)), Some(gsi_relay))
        } else {
            (None, None)
        };
        let pci_devices = create_devices(&mem, &mut vm, &mut resources, &exit_evt)
            .map_err(|e| Error::CreateDevices(Box::new(e)))?;
        let (pci, pci_irqs, pid_debug_label_map) = arch::generate_pci_root(
            pci_devices,
            &mut gsi_relay,
            &mut mmio_bus,
            &mut resources,
            &mut vm,
        )
        .map_err(Error::CreatePciRoot)?;
        let pci_bus = Arc::new(Mutex::new(PciConfigIo::new(pci)));

        // Event used to notify crosvm that guest OS is trying to suspend.
        let suspend_evt = EventFd::new().map_err(Error::CreateEventFd)?;

        let mut io_bus = Self::setup_io_bus(
            &mut vm,
            &mut gsi_relay,
            exit_evt.try_clone().map_err(Error::CloneEventFd)?,
            Some(pci_bus.clone()),
            components.memory_size,
        )?;

        Self::setup_serial_devices(
            &mut vm,
            &mut io_bus,
            &mut gsi_relay,
            serial_parameters,
            serial_jail,
        )?;

        let acpi_dev_resource = Self::setup_acpi_devices(
            &mut io_bus,
            &mut resources,
            suspend_evt.try_clone().map_err(Error::CloneEventFd)?,
            components.acpi_sdts,
        )?;

        let ramoops_region = match components.pstore {
            Some(pstore) => Some(
                arch::pstore::create_memory_region(&mut vm, &mut resources, &pstore)
                    .map_err(Error::Pstore)?,
            ),
            None => None,
        };

        let gsi_relay = if let Some((pic, ioapic)) = &split_irqchip {
            io_bus.insert(pic.clone(), 0x20, 0x2, true).unwrap();
            io_bus.insert(pic.clone(), 0xa0, 0x2, true).unwrap();
            io_bus.insert(pic.clone(), 0x4d0, 0x2, true).unwrap();

            let mut irq_num = resources.allocate_irq().unwrap();
            while irq_num < kvm::NUM_IOAPIC_PINS as u32 {
                irq_num = resources.allocate_irq().unwrap();
            }

            // This will never fail because gsi_relay is Some iff split_irqchip is Some.
            let gsi_relay = Arc::new(gsi_relay.unwrap());
            pic.lock().register_relay(gsi_relay.clone());
            ioapic.lock().register_relay(gsi_relay.clone());
            Some(gsi_relay)
        } else {
            None
        };

        match components.vm_image {
            VmImage::Bios(ref mut bios) => Self::load_bios(&mem, bios)?,
            VmImage::Kernel(ref mut kernel_image) => {
                let mut cmdline = Self::get_base_linux_cmdline();

                get_serial_cmdline(&mut cmdline, serial_parameters, "io")
                    .map_err(Error::GetSerialCmdline)?;

                for param in components.extra_kernel_params {
                    cmdline.insert_str(&param).map_err(Error::Cmdline)?;
                }

                // It seems that default record_size is only 4096 byte even if crosvm allocates
                // more memory. It means that one crash can only 4096 byte.
                // Set record_size and console_size to 1/4 of allocated memory size.
                // This configulation is same as the host.
                if let Some(ramoops_region) = ramoops_region {
                    let ramoops_opts = [
                        ("mem_address", ramoops_region.address),
                        ("mem_size", ramoops_region.size as u64),
                        ("console_size", (ramoops_region.size / 4) as u64),
                        ("record_size", (ramoops_region.size / 4) as u64),
                        ("dump_oops", 1_u64),
                    ];
                    for (name, val) in &ramoops_opts {
                        cmdline
                            .insert_str(format!("ramoops.{}={:#x}", name, val))
                            .map_err(Error::Cmdline)?;
                    }
                }

                // separate out load_kernel from other setup to get a specific error for
                // kernel loading
                let (params, kernel_end) = Self::load_kernel(&mem, kernel_image)?;

                Self::setup_system_memory(
                    &mem,
                    components.memory_size,
                    vcpu_count,
                    &CString::new(cmdline).unwrap(),
                    components.initrd_image,
                    pci_irqs,
                    components.android_fstab,
                    kernel_end,
                    params,
                    acpi_dev_resource,
                )?;
            }
        }
        Ok(RunnableLinuxVm {
            vm,
            kvm,
            resources,
            exit_evt,
            vcpus,
            vcpu_affinity,
            irq_chip,
            split_irqchip,
            gsi_relay,
            io_bus,
            mmio_bus,
            pid_debug_label_map,
            suspend_evt,
        })
    }

    fn setup_memory(params: MemoryParams) -> Result<GuestMemory> {
        let arch_mem_regions = arch_memory_regions(params);
        let mem = GuestMemory::new(&arch_mem_regions).map_err(Error::SetupGuestMemory)?;
        Ok(mem)
    }
}

impl X8664arch {
    /// Loads the bios from an open file.
    ///
    /// # Arguments
    ///
    /// * `mem` - The memory to be used by the guest.
    /// * `bios_image` - the File object for the specified bios
    fn load_bios(mem: &GuestMemory, bios_image: &mut File) -> Result<()> {
        let bios_image_length = bios_image
            .seek(io::SeekFrom::End(0))
            .map_err(Error::LoadBios)?;
        if bios_image_length != BIOS_LEN as u64 {
            return Err(Error::LoadBios(io::Error::new(
                io::ErrorKind::InvalidData,
                format!(
                    "bios was {} bytes, expected {}",
                    bios_image_length, BIOS_LEN
                ),
            )));
        }
        bios_image
            .seek(io::SeekFrom::Start(0))
            .map_err(Error::LoadBios)?;
        mem.read_to_memory(GuestAddress(BIOS_START), bios_image, BIOS_LEN)
            .map_err(Error::SetupGuestMemory)?;
        Ok(())
    }

    /// Loads the kernel from an open file.
    ///
    /// # Arguments
    ///
    /// * `mem` - The memory to be used by the guest.
    /// * `kernel_image` - the File object for the specified kernel.
    fn load_kernel(mem: &GuestMemory, kernel_image: &mut File) -> Result<(boot_params, u64)> {
        let elf_result =
            kernel_loader::load_kernel(mem, GuestAddress(KERNEL_START_OFFSET), kernel_image);
        if elf_result == Err(kernel_loader::Error::InvalidElfMagicNumber) {
            bzimage::load_bzimage(mem, GuestAddress(KERNEL_START_OFFSET), kernel_image)
                .map_err(Error::LoadBzImage)
        } else {
            let kernel_end = elf_result.map_err(Error::LoadKernel)?;
            Ok((Default::default(), kernel_end))
        }
    }

    /// Configures the system memory space should be called once per vm before
    /// starting vcpu threads.
    ///
    /// # Arguments
    ///
    /// * `mem` - The memory to be used by the guest.
    /// * `vcpu_count` - Number of virtual CPUs the guest will have.
    /// * `cmdline` - the kernel commandline
    /// * `initrd_file` - an initial ramdisk image
    fn setup_system_memory(
        mem: &GuestMemory,
        mem_size: u64,
        vcpu_count: u32,
        cmdline: &CStr,
        initrd_file: Option<File>,
        pci_irqs: Vec<(PciAddress, u32, PciInterruptPin)>,
        android_fstab: Option<File>,
        kernel_end: u64,
        params: boot_params,
        acpi_dev_resource: acpi::ACPIDevResource,
    ) -> Result<()> {
        kernel_loader::load_cmdline(mem, GuestAddress(CMDLINE_OFFSET), cmdline)
            .map_err(Error::LoadCmdline)?;

        // Track the first free address after the kernel - this is where extra
        // data like the device tree blob and initrd will be loaded.
        let mut free_addr = kernel_end;

        let setup_data = if let Some(android_fstab) = android_fstab {
            let free_addr_aligned = (((free_addr + 64 - 1) / 64) * 64) + 64;
            let dtb_start = GuestAddress(free_addr_aligned);
            let dtb_size = fdt::create_fdt(
                X86_64_FDT_MAX_SIZE as usize,
                mem,
                dtb_start.offset(),
                android_fstab,
            )
            .map_err(Error::CreateFdt)?;
            free_addr = dtb_start.offset() + dtb_size as u64;
            Some(dtb_start)
        } else {
            None
        };

        let initrd = match initrd_file {
            Some(mut initrd_file) => {
                let mut initrd_addr_max = u64::from(params.hdr.initrd_addr_max);
                // Default initrd_addr_max for old kernels (see Documentation/x86/boot.txt).
                if initrd_addr_max == 0 {
                    initrd_addr_max = 0x37FFFFFF;
                }

                let mem_max = mem.end_addr().offset() - 1;
                if initrd_addr_max > mem_max {
                    initrd_addr_max = mem_max;
                }

                let (initrd_start, initrd_size) = arch::load_image_high(
                    mem,
                    &mut initrd_file,
                    GuestAddress(free_addr),
                    GuestAddress(initrd_addr_max),
                    sys_util::pagesize() as u64,
                )
                .map_err(Error::LoadInitrd)?;
                Some((initrd_start, initrd_size))
            }
            None => None,
        };

        configure_system(
            mem,
            mem_size,
            GuestAddress(KERNEL_START_OFFSET),
            GuestAddress(CMDLINE_OFFSET),
            cmdline.to_bytes().len() + 1,
            vcpu_count as u8,
            pci_irqs,
            setup_data,
            initrd,
            params,
            acpi_dev_resource,
        )?;
        Ok(())
    }

    /// Creates a new VM object and initializes architecture specific devices
    ///
    /// # Arguments
    ///
    /// * `kvm` - The opened /dev/kvm object.
    /// * `split_irqchip` - Whether to use a split IRQ chip.
    /// * `mem` - The memory to be used by the guest.
    fn create_vm(kvm: &Kvm, split_irqchip: bool, mem: GuestMemory) -> Result<Vm> {
        let mut vm = Vm::new(&kvm, mem).map_err(Error::CreateVm)?;
        let tss_addr = GuestAddress(0xfffbd000);
        vm.set_tss_addr(tss_addr).map_err(Error::SetTssAddr)?;
        if !split_irqchip {
            vm.create_pit().map_err(Error::CreatePit)?;
            vm.create_irq_chip().map_err(Error::CreateIrqChip)?;
        } else {
            vm.enable_split_irqchip()
                .map_err(Error::EnableSplitIrqchip)?;
            for i in 0..kvm::NUM_IOAPIC_PINS {
                // Add dummy MSI routes to replace the default IRQChip routes.
                let route = IrqRoute {
                    gsi: i as u32,
                    source: IrqSource::Msi {
                        address: 0,
                        data: 0,
                    },
                };
                // Safe to ignore errors because errors are caused by the default routes and dummy
                // MSI routes will always be registered.
                let _ = vm.add_irq_route_entry(route);
            }
        }
        Ok(vm)
    }

    /// The creates the interrupt controller device and optionally returns the fd for it.
    /// Some architectures may not have a separate descriptor for the interrupt
    /// controller, so they would return None even on success.
    ///
    /// # Arguments
    ///
    /// * `vm` - the vm object
    fn create_irq_chip(_vm: &kvm::Vm) -> Result<Option<File>> {
        // Unfortunately X86 and ARM have to do this in completely different order
        // X86 needs to create the irq chip before creating cpus and
        // ARM needs to do it afterwards.
        Ok(None)
    }

    /// This returns the start address of high mmio
    ///
    /// # Arguments
    ///
    /// * mem: The memory to be used by the guest
    fn get_high_mmio_base(mem: &GuestMemory) -> u64 {
        // Put device memory at a 2MB boundary after physical memory or 4gb, whichever is greater.
        const MB: u64 = 1 << 20;
        const GB: u64 = 1 << 30;
        let ram_end_round_2mb = (mem.end_addr().offset() + 2 * MB - 1) / (2 * MB) * (2 * MB);
        std::cmp::max(ram_end_round_2mb, 4 * GB)
    }

    /// This returns a minimal kernel command for this architecture
    fn get_base_linux_cmdline() -> kernel_cmdline::Cmdline {
        let mut cmdline = kernel_cmdline::Cmdline::new(CMDLINE_MAX_SIZE as usize);
        cmdline.insert_str("pci=noacpi reboot=k panic=-1").unwrap();

        cmdline
    }

    /// Returns a system resource allocator.
    fn get_resource_allocator(mem: &GuestMemory, gpu_allocation: bool) -> SystemAllocator {
        let high_mmio_start = Self::get_high_mmio_base(mem);
        SystemAllocator::builder()
            .add_io_addresses(0xc000, 0x10000)
            .add_low_mmio_addresses(END_ADDR_BEFORE_32BITS, MMIO_SIZE)
            .add_high_mmio_addresses(high_mmio_start, u64::max_value() - high_mmio_start)
            .create_allocator(X86_64_IRQ_BASE, gpu_allocation)
            .unwrap()
    }

    /// Sets up the IO bus for this platform
    ///
    /// # Arguments
    ///
    /// * - `vm` the vm object
    /// * - `gsi_relay`: only valid for split IRQ chip (i.e. userspace PIT/PIC/IOAPIC)
    /// * - `exit_evt` - the event fd object which should receive exit events
    /// * - `mem_size` - the size in bytes of physical ram for the guest
    fn setup_io_bus(
        _vm: &mut Vm,
        gsi_relay: &mut Option<GsiRelay>,
        exit_evt: EventFd,
        pci: Option<Arc<Mutex<devices::PciConfigIo>>>,
        mem_size: u64,
    ) -> Result<devices::Bus> {
        struct NoDevice;
        impl devices::BusDevice for NoDevice {
            fn debug_label(&self) -> String {
                "no device".to_owned()
            }
        }

        let mut io_bus = devices::Bus::new();

        let mem_gap_start = END_ADDR_BEFORE_32BITS;
        let mem_below_4g = std::cmp::min(mem_gap_start, mem_size);
        let mem_above_4g = mem_size.saturating_sub(FIRST_ADDR_PAST_32BITS);

        io_bus
            .insert(
                Arc::new(Mutex::new(devices::Cmos::new(mem_below_4g, mem_above_4g))),
                0x70,
                0x2,
                false,
            )
            .unwrap();

        let nul_device = Arc::new(Mutex::new(NoDevice));
        let i8042 = Arc::new(Mutex::new(devices::I8042Device::new(
            exit_evt.try_clone().map_err(Error::CloneEventFd)?,
        )));

        if let Some(gsi_relay) = gsi_relay {
            let pit_evt = EventFd::new().map_err(Error::CreateEventFd)?;
            let pit = Arc::new(Mutex::new(
                devices::Pit::new(
                    pit_evt.try_clone().map_err(Error::CloneEventFd)?,
                    Arc::new(Mutex::new(Clock::new())),
                )
                .map_err(Error::CreatePitDevice)?,
            ));
            io_bus.insert(pit.clone(), 0x040, 0x8, true).unwrap();
            io_bus.insert(pit.clone(), 0x061, 0x1, true).unwrap();
            io_bus.insert(i8042, 0x062, 0x3, true).unwrap();
            gsi_relay.register_irqfd(pit_evt, 0);
        } else {
            io_bus
                .insert(nul_device.clone(), 0x040, 0x8, false)
                .unwrap(); // ignore pit
            io_bus.insert(i8042, 0x061, 0x4, true).unwrap();
        }

        io_bus
            .insert(nul_device.clone(), 0x0ed, 0x1, false)
            .unwrap(); // most likely this one does nothing
        io_bus
            .insert(nul_device.clone(), 0x0f0, 0x2, false)
            .unwrap(); // ignore fpu

        if let Some(pci_root) = pci {
            io_bus.insert(pci_root, 0xcf8, 0x8, false).unwrap();
        } else {
            // ignore pci.
            io_bus
                .insert(nul_device.clone(), 0xcf8, 0x8, false)
                .unwrap();
        }

        Ok(io_bus)
    }

    /// Sets up the acpi devices for this platform and
    /// return the resources which is used to set the ACPI tables.
    ///
    /// # Arguments
    ///
    /// * - `io_bus` the I/O bus to add the devices to
    /// * - `resources` the SystemAllocator to allocate IO and MMIO for acpi
    ///                devices.
    /// * - `suspend_evt` - the event fd object which used to suspend the vm
    fn setup_acpi_devices(
        io_bus: &mut devices::Bus,
        resources: &mut SystemAllocator,
        suspend_evt: EventFd,
        sdts: Vec<SDT>,
    ) -> Result<acpi::ACPIDevResource> {
        // The AML data for the acpi devices
        let mut amls = Vec::new();

        let pm_alloc = resources.get_anon_alloc();
        let pm_iobase = match resources.io_allocator() {
            Some(io) => io
                .allocate_with_align(
                    devices::acpi::ACPIPM_RESOURCE_LEN as u64,
                    pm_alloc,
                    "ACPIPM".to_string(),
                    devices::acpi::ACPIPM_RESOURCE_LEN as u64,
                )
                .map_err(Error::AllocateIOResouce)?,
            None => 0x600,
        };

        let pmresource = devices::ACPIPMResource::new(suspend_evt);
        Aml::to_aml_bytes(&pmresource, &mut amls);
        let pm = Arc::new(Mutex::new(pmresource));
        io_bus
            .insert(
                pm.clone(),
                pm_iobase as u64,
                devices::acpi::ACPIPM_RESOURCE_LEN as u64,
                false,
            )
            .unwrap();
        io_bus.notify_on_resume(pm);

        Ok(acpi::ACPIDevResource {
            amls,
            pm_iobase,
            sdts,
        })
    }

    /// Sets up the serial devices for this platform. Returns the serial port number and serial
    /// device to be used for stdout
    ///
    /// # Arguments
    ///
    /// * - `vm` the vm object
    /// * - `io_bus` the I/O bus to add the devices to
    /// * - `gsi_relay`: only valid for split IRQ chip (i.e. userspace PIT/PIC/IOAPIC)
    /// * - `serial_parmaters` - definitions for how the serial devices should be configured
    fn setup_serial_devices(
        vm: &mut Vm,
        io_bus: &mut devices::Bus,
        gsi_relay: &mut Option<GsiRelay>,
        serial_parameters: &BTreeMap<(SerialHardware, u8), SerialParameters>,
        serial_jail: Option<Minijail>,
    ) -> Result<()> {
        let com_evt_1_3 = EventFd::new().map_err(Error::CreateEventFd)?;
        let com_evt_2_4 = EventFd::new().map_err(Error::CreateEventFd)?;

        arch::add_serial_devices(
            io_bus,
            &com_evt_1_3,
            &com_evt_2_4,
            &serial_parameters,
            serial_jail,
        )
        .map_err(Error::CreateSerialDevices)?;

        if let Some(gsi_relay) = gsi_relay {
            gsi_relay.register_irqfd(com_evt_1_3, X86_64_SERIAL_1_3_IRQ as usize);
            gsi_relay.register_irqfd(com_evt_2_4, X86_64_SERIAL_2_4_IRQ as usize);
        } else {
            vm.register_irqfd(&com_evt_1_3, X86_64_SERIAL_1_3_IRQ)
                .map_err(Error::RegisterIrqfd)?;
            vm.register_irqfd(&com_evt_2_4, X86_64_SERIAL_2_4_IRQ)
                .map_err(Error::RegisterIrqfd)?;
        }

        Ok(())
    }

    /// Configures the vcpu and should be called once per vcpu from the vcpu's thread.
    ///
    /// # Arguments
    ///
    /// * `guest_mem` - The memory to be used by the guest.
    /// * `kernel_load_offset` - Offset in bytes from `guest_mem` at which the
    ///                          kernel starts.
    /// * `kvm` - The /dev/kvm object that created vcpu.
    /// * `vm` - The VM object associated with this VCPU.
    /// * `vcpu` - The VCPU object to configure.
    /// * `cpu_id` - The id of the given `vcpu`.
    /// * `num_cpus` - Number of virtual CPUs the guest will have.
    fn configure_vcpu(
        guest_mem: &GuestMemory,
        kvm: &Kvm,
        _vm: &Vm,
        vcpu: &Vcpu,
        cpu_id: u64,
        num_cpus: u64,
    ) -> Result<()> {
        let kernel_load_addr = GuestAddress(KERNEL_START_OFFSET);
        cpuid::setup_cpuid(kvm, vcpu, cpu_id, num_cpus).map_err(Error::SetupCpuid)?;
        regs::setup_msrs(vcpu, END_ADDR_BEFORE_32BITS).map_err(Error::SetupMsrs)?;
        let kernel_end = guest_mem
            .checked_offset(kernel_load_addr, KERNEL_64BIT_ENTRY_OFFSET)
            .ok_or(Error::KernelOffsetPastEnd)?;
        regs::setup_regs(
            vcpu,
            (kernel_end).offset() as u64,
            BOOT_STACK_POINTER as u64,
            ZERO_PAGE_OFFSET as u64,
        )
        .map_err(Error::SetupRegs)?;
        regs::setup_fpu(vcpu).map_err(Error::SetupFpu)?;
        regs::setup_sregs(guest_mem, vcpu).map_err(Error::SetupSregs)?;
        interrupts::set_lint(vcpu).map_err(Error::SetLint)?;
        Ok(())
    }
}
#[cfg(test)]
mod tests {
    use super::*;

    #[test]
    fn regions_lt_4gb_nobios() {
        let regions = arch_memory_regions(MemoryParams {
            size: 1u64 << 29,
            has_bios: false,
        });
        assert_eq!(1, regions.len());
        assert_eq!(GuestAddress(0), regions[0].0);
        assert_eq!(1u64 << 29, regions[0].1);
    }

    #[test]
    fn regions_gt_4gb_nobios() {
        let regions = arch_memory_regions(MemoryParams {
            size: (1u64 << 32) + 0x8000,
            has_bios: false,
        });
        assert_eq!(2, regions.len());
        assert_eq!(GuestAddress(0), regions[0].0);
        assert_eq!(GuestAddress(1u64 << 32), regions[1].0);
    }

    #[test]
    fn regions_lt_4gb_bios() {
        let regions = arch_memory_regions(MemoryParams {
            size: 1u64 << 29,
            has_bios: true,
        });
        assert_eq!(2, regions.len());
        assert_eq!(GuestAddress(0), regions[0].0);
        assert_eq!(1u64 << 29, regions[0].1);
        assert_eq!(GuestAddress(BIOS_START), regions[1].0);
        assert_eq!(BIOS_LEN as u64, regions[1].1);
    }

    #[test]
    fn regions_gt_4gb_bios() {
        let regions = arch_memory_regions(MemoryParams {
            size: (1u64 << 32) + 0x8000,
            has_bios: true,
        });
        assert_eq!(3, regions.len());
        assert_eq!(GuestAddress(0), regions[0].0);
        assert_eq!(GuestAddress(BIOS_START), regions[1].0);
        assert_eq!(BIOS_LEN as u64, regions[1].1);
        assert_eq!(GuestAddress(1u64 << 32), regions[2].0);
    }
}