5.3Pci虚拟化

5.3.1 虚拟pci控制器

(1) 初始化

pc_init1 ==>

i440fx_init(&i440fx_state,&piix3_devfn, &isa_bus, gsi,.....); ==> i440fx_common_init

gsi为全局中断qemu_irq.

isa_bus为pc isa_bus设备对象;

下面是该函数的流程分析:

a) dev =qdev_create(NULL, "i440FX-pcihost");

s = PCI_HOST_BRIDGE(dev);

s->address_space = address_space_mem;

b = pci_bus_new(dev, NULL,pci_address_space,

1. address_space_io,0);

创建 虚拟pcihost, 然后创建pcibus , 该pci bus的parent为i440fx-pcihost.

b) d =pci_create_simple(b, 0, device_name);

*pi440fx_state = DO_UPCAST(PCII440FXState,dev, d);

f = *pi440fx_state;

f->system_memory = address_space_mem;

在pci bus上创建i440fx,并设置其memory

c) 在pci bus上创建pci dev piix3

piix3 = DO_UPCAST(PIIX3State, dev,

pci_create_simple_multifunction(b,-1, true, "PIIX3"));

pci bus的终端map_irq函数为pci_slot_get_pirq

set_irq为piix3_set_irq, 数据为piix3

pci_bus_irqs(b, piix3_set_irq,pci_slot_get_pirq, piix3, PIIX_NUM_PIRQS);

bus->route_intx_pin_to_irq = piix3_route_intx_pin_to_irq

pci_bus_set_route_irq_fn(b,piix3_route_intx_pin_to_irq);

piix3的创建

static const TypeInfo piix3_info = {

.name = "PIIX3",

.parent = TYPE_PCI_DEVICE,

.instance_size = sizeof(PIIX3State),

.class_init = piix3_class_init,

};

piix3_class_init ==>

k->init = piix3_initfn;

k->config_write = piix3_write_config;

d) piix3的init 为piix3_initfn==> isa_bus_new(&d->dev.qdev, pci_address_space_io(dev));

piix3中包含了一个isa_bus

piix3->pic = pic; //为gsi

*isa_bus = DO_UPCAST(ISABus, qbus, qdev_get_child_bus(&piix3->dev.qdev,"isa.0"));

e) i440fx_update_memory_mappings(f);

为i440fx 分配内存

(2) i440fx-pcihost 与 piix3

static const TypeInfoi440fx_pcihost_info = {

.name = "i440FX-pcihost",

.parent = TYPE_PCI_HOST_BRIDGE,

.instance_size = sizeof(I440FXState),

.class_init = i440fx_pcihost_class_init,

};

static inti440fx_pcihost_initfn(SysBusDevice *dev)

{

memory_region_init_io(&s->conf_mem,&pci_host_conf_le_ops, s,

"pci-conf-idx", 4);

sysbus_add_io(dev, 0xcf8,&s->conf_mem);

sysbus_init_ioports(&s->busdev,0xcf8, 4);//port 操作地址

memory_region_init_io(&s->data_mem,&pci_host_data_le_ops, s,

"pci-conf-data", 4);

sysbus_add_io(dev, 0xcfc, &s->data_mem);////port 操作地址

sysbus_init_ioports(&s->busdev,0xcfc, 4);

}

0xcf8, 0xcfc为pc机pci访问的端口,下面看下该接口的虚拟化

const MemoryRegionOpspci_host_conf_le_ops = {

.read = pci_host_config_read,

.write = pci_host_config_write,

.endianness = DEVICE_LITTLE_ENDIAN,

};

static uint64_tpci_host_data_read(void *opaque,

target_phys_addr_t addr, unsigned len)

{

......

val = pci_data_read(s->bus,s->config_reg | (addr & 3), len);

return val;

}

s->config_reg由0xcf8的端口虚拟化操作控制.

uint32_t pci_data_read(PCIBus*s, uint32_t addr, int len)

{

PCIDevice *pci_dev =pci_dev_find_by_addr(s, addr);

uint32_t config_addr = addr &(PCI_CONFIG_SPACE_SIZE - 1);

uint32_t val;

if (!pci_dev) {

return ~0x0;

}

val = pci_host_config_read_common(pci_dev,config_addr,

PCI_CONFIG_SPACE_SIZE, len);

return val;

}

pci_dev_find_by_addr 根据pci设备地址得到pci设备

static inline PCIDevice*pci_dev_find_by_addr(PCIBus *bus, uint32_t addr)

{

uint8_t bus_num = addr >> 16;

uint8_t devfn = addr >> 8;

return pci_find_device(bus, bus_num,devfn);

}

PCIDevice*pci_find_device(PCIBus *bus, int bus_num, uint8_t devfn)

{

bus = pci_find_bus_nr(bus, bus_num);

if (!bus)

return NULL;

return bus->devices[devfn];

}

pci_create_xxx创建pci dev时要指定devfn.和bus.

pci_host_config_read_common最终会调用pci_dev->config_read

因此i440FX-pcihost的职责就是负责向pci_dev转发config的rw.

piix3的write_config为

static void piix3_write_config(PCIDevice*dev,

uint32_taddress, uint32_t val, int len)

{

pci_default_write_config(dev, address, val,len);

if (ranges_overlap(address, len,PIIX_PIRQC, 4)) {

PIIX3State *piix3 = DO_UPCAST(PIIX3State,dev, dev);

int pic_irq;

pci_bus_fire_intx_routing_notifier(piix3->dev.bus);

piix3_update_irq_levels(piix3);

for (pic_irq = 0; pic_irq <PIIX_NUM_PIC_IRQS; pic_irq++) {

piix3_set_irq_pic(piix3, pic_irq);

}

}

}

piix_set_irq_pic会调用qemu_set_irq发射中断。piix3->pic为gsi

5.3.2 pci device 虚拟化

(1) pci device 初始化

pci device的parent为pci_type.其初始化函数如下：

static TypeInfopci_device_type_info = {

.name = TYPE_PCI_DEVICE,

.parent = TYPE_DEVICE,

.instance_size = sizeof(PCIDevice),

.abstract = true,

.class_size = sizeof(PCIDeviceClass),

.class_init = pci_device_class_init,

};

init 函数 pci_qdev_init ==》

a) do_pci_register_device

b) 为pci设备添加rom

do_pci_register_device的职责是为

a) 根据config_size分配conf用于config寄存器虚拟化

b) bus->dma_context_fn用于pci全硬件虚拟化(iommu)，第7章分析

c) pci_device的标准寄存器赋初值，如vendor_id

d) pcidevice中断管理pci_dev->irq =qemu_allocate_irqs(pci_set_irq, pci_dev, PCI_NUM_PINS);对于piix3其irq处理也为pci_set_irq,因此上一节的ide中断最终也由pci_set_irq处理

(2) pci 地址分配

guest_os运行后会为pci device分配baraddress,该地址为物理地址。该过程由config的虚拟化操作实现，就是对pci->conf.

config的虚拟化default函数为pci_default_write_config(对于多数设备直接使用该函数，程序员也可以自定义config_write).

voidpci_default_write_config(PCIDevice *d, uint32_t addr, uint32_t val, int l)

{

int i, was_irq_disabled =pci_irq_disabled(d);

for (i = 0; i < l; val >>= 8, ++i){

uint8_t wmask = d->wmask[addr + i];

uint8_t w1cmask = d->w1cmask[addr +i];

assert(!(wmask & w1cmask));

d->config[addr + i] =(d->config[addr + i] & ~wmask) | (val & wmask);

d->config[addr + i] &= ~(val& w1cmask); /* W1C: Write 1 to Clear */

}

if (ranges_overlap(addr, l,PCI_BASE_ADDRESS_0, 24) ||

ranges_overlap(addr, l,PCI_ROM_ADDRESS, 4) ||

ranges_overlap(addr, l, PCI_ROM_ADDRESS1,4) ||

range_covers_byte(addr, l,PCI_COMMAND))

pci_update_mappings(d);

if (range_covers_byte(addr, l,PCI_COMMAND))

pci_update_irq_disabled(d,was_irq_disabled);

msi_write_config(d, addr, val, l);

msix_write_config(d, addr, val, l);

}

当bar地址被更新后，pci_update_mappings就被调用

该函数根据bar地址来删除或建立内存区

new_addr = pci_bar_address(d,i, r->type, r->size);

/* now do the real mapping */

if (r->addr != PCI_BAR_UNMAPPED) {

memory_region_del_subregion(r->address_space,r->memory);

}

r->addr = new_addr;

if (r->addr != PCI_BAR_UNMAPPED) {

memory_region_add_subregion_overlap(r->address_space,

r->addr, r->memory, 1);

5.3.3 pci中断注入

(1) 基本中断处理

static void pci_set_irq(void*opaque, int irq_num, int level)

{

PCIDevice *pci_dev = opaque;

int change;

change = level - pci_irq_state(pci_dev,irq_num);

if (!change)

return;

pci_set_irq_state(pci_dev, irq_num, level);

pci_update_irq_status(pci_dev);

if (pci_irq_disabled(pci_dev))

return;

pci_change_irq_level(pci_dev, irq_num,change);

}

pci_change_irq_level ==> bus->set_irq

bus->irq在初始化时被设为了piix3_set_irq

static voidpiix3_set_irq_pic(PIIX3State *piix3, int pic_irq)

{

qemu_set_irq(piix3->pic[pic_irq],

!!(piix3->pic_levels &

(((1ULL <<PIIX_NUM_PIRQS) - 1) <<

(pic_irq *PIIX_NUM_PIRQS))));

}

所以pci中断最终于gsi关联起来了。

(2) msi中断

x86处理器使用FSB Interrupt Message总线事务转发MSI/MSI-X中断请求。使用这种方法的优点是向CPU内核提交中断请求的同时，提交PCIe设备使用的中断向量，从而CPU不需要使用中断响应周期从寄存器中获得中断向量。msi_init 用于pci_dev初始化时初始化msi能力.

msi_notify 用于发射中断时设置msi的gpa

void msi_notify(PCIDevice*dev, unsigned int vector)

{

.......

if (msi64bit) {

address = pci_get_quad(dev->config +msi_address_lo_off(dev));

} else {

address = pci_get_long(dev->config +msi_address_lo_off(dev));

}

/* upper bit 31:16 is zero */

data = pci_get_word(dev->config +msi_data_off(dev, msi64bit));

if (nr_vectors > 1) {

data &= ~(nr_vectors - 1);

data |= vector;

}

stl_le_phys(address, data); //将数据写入gpa address, 注入msi中断

}