2007-07-26 11:41:02 -06:00
|
|
|
/*P:100 This is the Launcher code, a simple program which lays out the
|
|
|
|
* "physical" memory for the new Guest by mapping the kernel image and the
|
|
|
|
* virtual devices, then reads repeatedly from /dev/lguest to run the Guest.
|
2007-10-21 19:03:26 -06:00
|
|
|
:*/
|
2007-07-19 02:49:29 -06:00
|
|
|
#define _LARGEFILE64_SOURCE
|
|
|
|
#define _GNU_SOURCE
|
|
|
|
#include <stdio.h>
|
|
|
|
#include <string.h>
|
|
|
|
#include <unistd.h>
|
|
|
|
#include <err.h>
|
|
|
|
#include <stdint.h>
|
|
|
|
#include <stdlib.h>
|
|
|
|
#include <elf.h>
|
|
|
|
#include <sys/mman.h>
|
2007-08-28 15:35:59 -06:00
|
|
|
#include <sys/param.h>
|
2007-07-19 02:49:29 -06:00
|
|
|
#include <sys/types.h>
|
|
|
|
#include <sys/stat.h>
|
|
|
|
#include <sys/wait.h>
|
|
|
|
#include <fcntl.h>
|
|
|
|
#include <stdbool.h>
|
|
|
|
#include <errno.h>
|
|
|
|
#include <ctype.h>
|
|
|
|
#include <sys/socket.h>
|
|
|
|
#include <sys/ioctl.h>
|
|
|
|
#include <sys/time.h>
|
|
|
|
#include <time.h>
|
|
|
|
#include <netinet/in.h>
|
|
|
|
#include <net/if.h>
|
|
|
|
#include <linux/sockios.h>
|
|
|
|
#include <linux/if_tun.h>
|
|
|
|
#include <sys/uio.h>
|
|
|
|
#include <termios.h>
|
|
|
|
#include <getopt.h>
|
|
|
|
#include <zlib.h>
|
2007-10-21 19:24:22 -06:00
|
|
|
#include <assert.h>
|
|
|
|
#include <sched.h>
|
2007-10-21 18:56:24 -06:00
|
|
|
#include "linux/lguest_launcher.h"
|
2007-10-21 19:24:22 -06:00
|
|
|
#include "linux/virtio_config.h"
|
|
|
|
#include "linux/virtio_net.h"
|
|
|
|
#include "linux/virtio_blk.h"
|
|
|
|
#include "linux/virtio_console.h"
|
|
|
|
#include "linux/virtio_ring.h"
|
2007-10-21 19:29:57 -06:00
|
|
|
#include "asm-x86/bootparam.h"
|
2007-10-24 22:09:25 -06:00
|
|
|
/*L:110 We can ignore the 38 include files we need for this program, but I do
|
|
|
|
* want to draw attention to the use of kernel-style types.
|
|
|
|
*
|
|
|
|
* As Linus said, "C is a Spartan language, and so should your naming be." I
|
|
|
|
* like these abbreviations, so we define them here. Note that u64 is always
|
|
|
|
* unsigned long long, which works on all Linux systems: this means that we can
|
|
|
|
* use %llu in printf for any u64. */
|
|
|
|
typedef unsigned long long u64;
|
|
|
|
typedef uint32_t u32;
|
|
|
|
typedef uint16_t u16;
|
|
|
|
typedef uint8_t u8;
|
2007-07-26 11:41:03 -06:00
|
|
|
/*:*/
|
2007-07-19 02:49:29 -06:00
|
|
|
|
|
|
|
#define PAGE_PRESENT 0x7 /* Present, RW, Execute */
|
|
|
|
#define NET_PEERNUM 1
|
|
|
|
#define BRIDGE_PFX "bridge:"
|
|
|
|
#ifndef SIOCBRADDIF
|
|
|
|
#define SIOCBRADDIF 0x89a2 /* add interface to bridge */
|
|
|
|
#endif
|
2007-10-21 19:03:26 -06:00
|
|
|
/* We can have up to 256 pages for devices. */
|
|
|
|
#define DEVICE_PAGES 256
|
2007-11-11 19:39:18 -07:00
|
|
|
/* This will occupy 2 pages: it must be a power of 2. */
|
|
|
|
#define VIRTQUEUE_NUM 128
|
2007-07-19 02:49:29 -06:00
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/*L:120 verbose is both a global flag and a macro. The C preprocessor allows
|
|
|
|
* this, and although I wouldn't recommend it, it works quite nicely here. */
|
2007-07-19 02:49:29 -06:00
|
|
|
static bool verbose;
|
|
|
|
#define verbose(args...) \
|
|
|
|
do { if (verbose) printf(args); } while(0)
|
2007-07-26 11:41:03 -06:00
|
|
|
/*:*/
|
|
|
|
|
|
|
|
/* The pipe to send commands to the waker process */
|
2007-07-19 02:49:29 -06:00
|
|
|
static int waker_fd;
|
2007-10-21 19:03:26 -06:00
|
|
|
/* The pointer to the start of guest memory. */
|
|
|
|
static void *guest_base;
|
|
|
|
/* The maximum guest physical address allowed, and maximum possible. */
|
|
|
|
static unsigned long guest_limit, guest_max;
|
2007-07-19 02:49:29 -06:00
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/* This is our list of devices. */
|
2007-07-19 02:49:29 -06:00
|
|
|
struct device_list
|
|
|
|
{
|
2007-07-26 11:41:03 -06:00
|
|
|
/* Summary information about the devices in our list: ready to pass to
|
|
|
|
* select() to ask which need servicing.*/
|
2007-07-19 02:49:29 -06:00
|
|
|
fd_set infds;
|
|
|
|
int max_infd;
|
|
|
|
|
2007-10-21 19:24:22 -06:00
|
|
|
/* Counter to assign interrupt numbers. */
|
|
|
|
unsigned int next_irq;
|
|
|
|
|
|
|
|
/* Counter to print out convenient device numbers. */
|
|
|
|
unsigned int device_num;
|
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/* The descriptor page for the devices. */
|
2007-10-21 19:24:22 -06:00
|
|
|
u8 *descpage;
|
|
|
|
|
|
|
|
/* The tail of the last descriptor. */
|
|
|
|
unsigned int desc_used;
|
2007-07-26 11:41:03 -06:00
|
|
|
|
|
|
|
/* A single linked list of devices. */
|
2007-07-19 02:49:29 -06:00
|
|
|
struct device *dev;
|
2007-07-26 11:41:03 -06:00
|
|
|
/* ... And an end pointer so we can easily append new devices */
|
2007-07-19 02:49:29 -06:00
|
|
|
struct device **lastdev;
|
|
|
|
};
|
|
|
|
|
2007-10-21 19:24:22 -06:00
|
|
|
/* The list of Guest devices, based on command line arguments. */
|
|
|
|
static struct device_list devices;
|
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/* The device structure describes a single device. */
|
2007-07-19 02:49:29 -06:00
|
|
|
struct device
|
|
|
|
{
|
2007-07-26 11:41:03 -06:00
|
|
|
/* The linked-list pointer. */
|
2007-07-19 02:49:29 -06:00
|
|
|
struct device *next;
|
2007-10-21 19:24:22 -06:00
|
|
|
|
|
|
|
/* The this device's descriptor, as mapped into the Guest. */
|
2007-07-19 02:49:29 -06:00
|
|
|
struct lguest_device_desc *desc;
|
2007-10-21 19:24:22 -06:00
|
|
|
|
|
|
|
/* The name of this device, for --verbose. */
|
|
|
|
const char *name;
|
2007-07-19 02:49:29 -06:00
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/* If handle_input is set, it wants to be called when this file
|
|
|
|
* descriptor is ready. */
|
2007-07-19 02:49:29 -06:00
|
|
|
int fd;
|
|
|
|
bool (*handle_input)(int fd, struct device *me);
|
|
|
|
|
2007-10-21 19:24:22 -06:00
|
|
|
/* Any queues attached to this device */
|
|
|
|
struct virtqueue *vq;
|
2007-07-19 02:49:29 -06:00
|
|
|
|
|
|
|
/* Device-specific data. */
|
|
|
|
void *priv;
|
|
|
|
};
|
|
|
|
|
2007-10-21 19:24:22 -06:00
|
|
|
/* The virtqueue structure describes a queue attached to a device. */
|
|
|
|
struct virtqueue
|
|
|
|
{
|
|
|
|
struct virtqueue *next;
|
|
|
|
|
|
|
|
/* Which device owns me. */
|
|
|
|
struct device *dev;
|
|
|
|
|
|
|
|
/* The configuration for this queue. */
|
|
|
|
struct lguest_vqconfig config;
|
|
|
|
|
|
|
|
/* The actual ring of buffers. */
|
|
|
|
struct vring vring;
|
|
|
|
|
|
|
|
/* Last available index we saw. */
|
|
|
|
u16 last_avail_idx;
|
|
|
|
|
|
|
|
/* The routine to call when the Guest pings us. */
|
|
|
|
void (*handle_output)(int fd, struct virtqueue *me);
|
|
|
|
};
|
|
|
|
|
|
|
|
/* Since guest is UP and we don't run at the same time, we don't need barriers.
|
|
|
|
* But I include them in the code in case others copy it. */
|
|
|
|
#define wmb()
|
|
|
|
|
|
|
|
/* Convert an iovec element to the given type.
|
|
|
|
*
|
|
|
|
* This is a fairly ugly trick: we need to know the size of the type and
|
|
|
|
* alignment requirement to check the pointer is kosher. It's also nice to
|
|
|
|
* have the name of the type in case we report failure.
|
|
|
|
*
|
|
|
|
* Typing those three things all the time is cumbersome and error prone, so we
|
|
|
|
* have a macro which sets them all up and passes to the real function. */
|
|
|
|
#define convert(iov, type) \
|
|
|
|
((type *)_convert((iov), sizeof(type), __alignof__(type), #type))
|
|
|
|
|
|
|
|
static void *_convert(struct iovec *iov, size_t size, size_t align,
|
|
|
|
const char *name)
|
|
|
|
{
|
|
|
|
if (iov->iov_len != size)
|
|
|
|
errx(1, "Bad iovec size %zu for %s", iov->iov_len, name);
|
|
|
|
if ((unsigned long)iov->iov_base % align != 0)
|
|
|
|
errx(1, "Bad alignment %p for %s", iov->iov_base, name);
|
|
|
|
return iov->iov_base;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* The virtio configuration space is defined to be little-endian. x86 is
|
|
|
|
* little-endian too, but it's nice to be explicit so we have these helpers. */
|
|
|
|
#define cpu_to_le16(v16) (v16)
|
|
|
|
#define cpu_to_le32(v32) (v32)
|
|
|
|
#define cpu_to_le64(v64) (v64)
|
|
|
|
#define le16_to_cpu(v16) (v16)
|
|
|
|
#define le32_to_cpu(v32) (v32)
|
|
|
|
#define le64_to_cpu(v32) (v64)
|
|
|
|
|
2007-10-21 19:03:26 -06:00
|
|
|
/*L:100 The Launcher code itself takes us out into userspace, that scary place
|
|
|
|
* where pointers run wild and free! Unfortunately, like most userspace
|
|
|
|
* programs, it's quite boring (which is why everyone likes to hack on the
|
|
|
|
* kernel!). Perhaps if you make up an Lguest Drinking Game at this point, it
|
|
|
|
* will get you through this section. Or, maybe not.
|
|
|
|
*
|
|
|
|
* The Launcher sets up a big chunk of memory to be the Guest's "physical"
|
|
|
|
* memory and stores it in "guest_base". In other words, Guest physical ==
|
|
|
|
* Launcher virtual with an offset.
|
|
|
|
*
|
|
|
|
* This can be tough to get your head around, but usually it just means that we
|
|
|
|
* use these trivial conversion functions when the Guest gives us it's
|
|
|
|
* "physical" addresses: */
|
|
|
|
static void *from_guest_phys(unsigned long addr)
|
|
|
|
{
|
|
|
|
return guest_base + addr;
|
|
|
|
}
|
|
|
|
|
|
|
|
static unsigned long to_guest_phys(const void *addr)
|
|
|
|
{
|
|
|
|
return (addr - guest_base);
|
|
|
|
}
|
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/*L:130
|
|
|
|
* Loading the Kernel.
|
|
|
|
*
|
|
|
|
* We start with couple of simple helper routines. open_or_die() avoids
|
|
|
|
* error-checking code cluttering the callers: */
|
2007-07-19 02:49:29 -06:00
|
|
|
static int open_or_die(const char *name, int flags)
|
|
|
|
{
|
|
|
|
int fd = open(name, flags);
|
|
|
|
if (fd < 0)
|
|
|
|
err(1, "Failed to open %s", name);
|
|
|
|
return fd;
|
|
|
|
}
|
|
|
|
|
2007-10-21 19:03:26 -06:00
|
|
|
/* map_zeroed_pages() takes a number of pages. */
|
|
|
|
static void *map_zeroed_pages(unsigned int num)
|
2007-07-19 02:49:29 -06:00
|
|
|
{
|
2007-10-21 19:03:26 -06:00
|
|
|
int fd = open_or_die("/dev/zero", O_RDONLY);
|
|
|
|
void *addr;
|
2007-07-19 02:49:29 -06:00
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/* We use a private mapping (ie. if we write to the page, it will be
|
2007-10-21 19:03:26 -06:00
|
|
|
* copied). */
|
|
|
|
addr = mmap(NULL, getpagesize() * num,
|
|
|
|
PROT_READ|PROT_WRITE|PROT_EXEC, MAP_PRIVATE, fd, 0);
|
|
|
|
if (addr == MAP_FAILED)
|
|
|
|
err(1, "Mmaping %u pages of /dev/zero", num);
|
|
|
|
|
|
|
|
return addr;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Get some more pages for a device. */
|
|
|
|
static void *get_pages(unsigned int num)
|
|
|
|
{
|
|
|
|
void *addr = from_guest_phys(guest_limit);
|
|
|
|
|
|
|
|
guest_limit += num * getpagesize();
|
|
|
|
if (guest_limit > guest_max)
|
|
|
|
errx(1, "Not enough memory for devices");
|
|
|
|
return addr;
|
2007-07-19 02:49:29 -06:00
|
|
|
}
|
|
|
|
|
2007-08-28 15:35:59 -06:00
|
|
|
/* This routine is used to load the kernel or initrd. It tries mmap, but if
|
|
|
|
* that fails (Plan 9's kernel file isn't nicely aligned on page boundaries),
|
|
|
|
* it falls back to reading the memory in. */
|
|
|
|
static void map_at(int fd, void *addr, unsigned long offset, unsigned long len)
|
|
|
|
{
|
|
|
|
ssize_t r;
|
|
|
|
|
|
|
|
/* We map writable even though for some segments are marked read-only.
|
|
|
|
* The kernel really wants to be writable: it patches its own
|
|
|
|
* instructions.
|
|
|
|
*
|
|
|
|
* MAP_PRIVATE means that the page won't be copied until a write is
|
|
|
|
* done to it. This allows us to share untouched memory between
|
|
|
|
* Guests. */
|
|
|
|
if (mmap(addr, len, PROT_READ|PROT_WRITE|PROT_EXEC,
|
|
|
|
MAP_FIXED|MAP_PRIVATE, fd, offset) != MAP_FAILED)
|
|
|
|
return;
|
|
|
|
|
|
|
|
/* pread does a seek and a read in one shot: saves a few lines. */
|
|
|
|
r = pread(fd, addr, len, offset);
|
|
|
|
if (r != len)
|
|
|
|
err(1, "Reading offset %lu len %lu gave %zi", offset, len, r);
|
|
|
|
}
|
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/* This routine takes an open vmlinux image, which is in ELF, and maps it into
|
|
|
|
* the Guest memory. ELF = Embedded Linking Format, which is the format used
|
|
|
|
* by all modern binaries on Linux including the kernel.
|
|
|
|
*
|
|
|
|
* The ELF headers give *two* addresses: a physical address, and a virtual
|
2007-10-21 19:03:36 -06:00
|
|
|
* address. We use the physical address; the Guest will map itself to the
|
|
|
|
* virtual address.
|
2007-07-26 11:41:03 -06:00
|
|
|
*
|
|
|
|
* We return the starting address. */
|
2007-10-21 19:03:36 -06:00
|
|
|
static unsigned long map_elf(int elf_fd, const Elf32_Ehdr *ehdr)
|
2007-07-19 02:49:29 -06:00
|
|
|
{
|
|
|
|
Elf32_Phdr phdr[ehdr->e_phnum];
|
|
|
|
unsigned int i;
|
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/* Sanity checks on the main ELF header: an x86 executable with a
|
|
|
|
* reasonable number of correctly-sized program headers. */
|
2007-07-19 02:49:29 -06:00
|
|
|
if (ehdr->e_type != ET_EXEC
|
|
|
|
|| ehdr->e_machine != EM_386
|
|
|
|
|| ehdr->e_phentsize != sizeof(Elf32_Phdr)
|
|
|
|
|| ehdr->e_phnum < 1 || ehdr->e_phnum > 65536U/sizeof(Elf32_Phdr))
|
|
|
|
errx(1, "Malformed elf header");
|
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/* An ELF executable contains an ELF header and a number of "program"
|
|
|
|
* headers which indicate which parts ("segments") of the program to
|
|
|
|
* load where. */
|
|
|
|
|
|
|
|
/* We read in all the program headers at once: */
|
2007-07-19 02:49:29 -06:00
|
|
|
if (lseek(elf_fd, ehdr->e_phoff, SEEK_SET) < 0)
|
|
|
|
err(1, "Seeking to program headers");
|
|
|
|
if (read(elf_fd, phdr, sizeof(phdr)) != sizeof(phdr))
|
|
|
|
err(1, "Reading program headers");
|
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/* Try all the headers: there are usually only three. A read-only one,
|
|
|
|
* a read-write one, and a "note" section which isn't loadable. */
|
2007-07-19 02:49:29 -06:00
|
|
|
for (i = 0; i < ehdr->e_phnum; i++) {
|
2007-07-26 11:41:03 -06:00
|
|
|
/* If this isn't a loadable segment, we ignore it */
|
2007-07-19 02:49:29 -06:00
|
|
|
if (phdr[i].p_type != PT_LOAD)
|
|
|
|
continue;
|
|
|
|
|
|
|
|
verbose("Section %i: size %i addr %p\n",
|
|
|
|
i, phdr[i].p_memsz, (void *)phdr[i].p_paddr);
|
|
|
|
|
2007-08-28 15:35:59 -06:00
|
|
|
/* We map this section of the file at its physical address. */
|
2007-10-21 19:03:26 -06:00
|
|
|
map_at(elf_fd, from_guest_phys(phdr[i].p_paddr),
|
2007-08-28 15:35:59 -06:00
|
|
|
phdr[i].p_offset, phdr[i].p_filesz);
|
2007-07-19 02:49:29 -06:00
|
|
|
}
|
|
|
|
|
2007-10-21 19:29:44 -06:00
|
|
|
/* The entry point is given in the ELF header. */
|
|
|
|
return ehdr->e_entry;
|
2007-07-19 02:49:29 -06:00
|
|
|
}
|
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/*L:150 A bzImage, unlike an ELF file, is not meant to be loaded. You're
|
2007-10-21 19:29:56 -06:00
|
|
|
* supposed to jump into it and it will unpack itself. We used to have to
|
|
|
|
* perform some hairy magic because the unpacking code scared me.
|
2007-07-26 11:41:03 -06:00
|
|
|
*
|
2007-10-21 19:29:56 -06:00
|
|
|
* Fortunately, Jeremy Fitzhardinge convinced me it wasn't that hard and wrote
|
|
|
|
* a small patch to jump over the tricky bits in the Guest, so now we just read
|
|
|
|
* the funky header so we know where in the file to load, and away we go! */
|
2007-10-21 19:03:36 -06:00
|
|
|
static unsigned long load_bzimage(int fd)
|
2007-07-19 02:49:29 -06:00
|
|
|
{
|
2007-10-21 19:29:57 -06:00
|
|
|
struct boot_params boot;
|
2007-10-21 19:29:56 -06:00
|
|
|
int r;
|
|
|
|
/* Modern bzImages get loaded at 1M. */
|
|
|
|
void *p = from_guest_phys(0x100000);
|
|
|
|
|
|
|
|
/* Go back to the start of the file and read the header. It should be
|
|
|
|
* a Linux boot header (see Documentation/i386/boot.txt) */
|
|
|
|
lseek(fd, 0, SEEK_SET);
|
2007-10-21 19:29:57 -06:00
|
|
|
read(fd, &boot, sizeof(boot));
|
2007-10-21 19:29:56 -06:00
|
|
|
|
2007-10-21 19:29:57 -06:00
|
|
|
/* Inside the setup_hdr, we expect the magic "HdrS" */
|
|
|
|
if (memcmp(&boot.hdr.header, "HdrS", 4) != 0)
|
2007-10-21 19:29:56 -06:00
|
|
|
errx(1, "This doesn't look like a bzImage to me");
|
|
|
|
|
2007-10-21 19:29:57 -06:00
|
|
|
/* Skip over the extra sectors of the header. */
|
|
|
|
lseek(fd, (boot.hdr.setup_sects+1) * 512, SEEK_SET);
|
2007-10-21 19:29:56 -06:00
|
|
|
|
|
|
|
/* Now read everything into memory. in nice big chunks. */
|
|
|
|
while ((r = read(fd, p, 65536)) > 0)
|
|
|
|
p += r;
|
|
|
|
|
2007-10-21 19:29:57 -06:00
|
|
|
/* Finally, code32_start tells us where to enter the kernel. */
|
|
|
|
return boot.hdr.code32_start;
|
2007-07-19 02:49:29 -06:00
|
|
|
}
|
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/*L:140 Loading the kernel is easy when it's a "vmlinux", but most kernels
|
2007-10-24 23:02:50 -06:00
|
|
|
* come wrapped up in the self-decompressing "bzImage" format. With a little
|
|
|
|
* work, we can load those, too. */
|
2007-10-21 19:03:36 -06:00
|
|
|
static unsigned long load_kernel(int fd)
|
2007-07-19 02:49:29 -06:00
|
|
|
{
|
|
|
|
Elf32_Ehdr hdr;
|
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/* Read in the first few bytes. */
|
2007-07-19 02:49:29 -06:00
|
|
|
if (read(fd, &hdr, sizeof(hdr)) != sizeof(hdr))
|
|
|
|
err(1, "Reading kernel");
|
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/* If it's an ELF file, it starts with "\177ELF" */
|
2007-07-19 02:49:29 -06:00
|
|
|
if (memcmp(hdr.e_ident, ELFMAG, SELFMAG) == 0)
|
2007-10-21 19:03:36 -06:00
|
|
|
return map_elf(fd, &hdr);
|
2007-07-19 02:49:29 -06:00
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/* Otherwise we assume it's a bzImage, and try to unpack it */
|
2007-10-21 19:03:36 -06:00
|
|
|
return load_bzimage(fd);
|
2007-07-19 02:49:29 -06:00
|
|
|
}
|
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/* This is a trivial little helper to align pages. Andi Kleen hated it because
|
|
|
|
* it calls getpagesize() twice: "it's dumb code."
|
|
|
|
*
|
|
|
|
* Kernel guys get really het up about optimization, even when it's not
|
|
|
|
* necessary. I leave this code as a reaction against that. */
|
2007-07-19 02:49:29 -06:00
|
|
|
static inline unsigned long page_align(unsigned long addr)
|
|
|
|
{
|
2007-07-26 11:41:03 -06:00
|
|
|
/* Add upwards and truncate downwards. */
|
2007-07-19 02:49:29 -06:00
|
|
|
return ((addr + getpagesize()-1) & ~(getpagesize()-1));
|
|
|
|
}
|
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/*L:180 An "initial ram disk" is a disk image loaded into memory along with
|
|
|
|
* the kernel which the kernel can use to boot from without needing any
|
|
|
|
* drivers. Most distributions now use this as standard: the initrd contains
|
|
|
|
* the code to load the appropriate driver modules for the current machine.
|
|
|
|
*
|
|
|
|
* Importantly, James Morris works for RedHat, and Fedora uses initrds for its
|
|
|
|
* kernels. He sent me this (and tells me when I break it). */
|
2007-07-19 02:49:29 -06:00
|
|
|
static unsigned long load_initrd(const char *name, unsigned long mem)
|
|
|
|
{
|
|
|
|
int ifd;
|
|
|
|
struct stat st;
|
|
|
|
unsigned long len;
|
|
|
|
|
|
|
|
ifd = open_or_die(name, O_RDONLY);
|
2007-07-26 11:41:03 -06:00
|
|
|
/* fstat() is needed to get the file size. */
|
2007-07-19 02:49:29 -06:00
|
|
|
if (fstat(ifd, &st) < 0)
|
|
|
|
err(1, "fstat() on initrd '%s'", name);
|
|
|
|
|
2007-08-28 15:35:59 -06:00
|
|
|
/* We map the initrd at the top of memory, but mmap wants it to be
|
|
|
|
* page-aligned, so we round the size up for that. */
|
2007-07-19 02:49:29 -06:00
|
|
|
len = page_align(st.st_size);
|
2007-10-21 19:03:26 -06:00
|
|
|
map_at(ifd, from_guest_phys(mem - len), 0, st.st_size);
|
2007-07-26 11:41:03 -06:00
|
|
|
/* Once a file is mapped, you can close the file descriptor. It's a
|
|
|
|
* little odd, but quite useful. */
|
2007-07-19 02:49:29 -06:00
|
|
|
close(ifd);
|
2007-08-28 15:35:59 -06:00
|
|
|
verbose("mapped initrd %s size=%lu @ %p\n", name, len, (void*)mem-len);
|
2007-07-26 11:41:03 -06:00
|
|
|
|
|
|
|
/* We return the initrd size. */
|
2007-07-19 02:49:29 -06:00
|
|
|
return len;
|
|
|
|
}
|
|
|
|
|
2007-10-21 19:03:36 -06:00
|
|
|
/* Once we know how much memory we have, we can construct simple linear page
|
|
|
|
* tables which set virtual == physical which will get the Guest far enough
|
2007-10-21 19:03:26 -06:00
|
|
|
* into the boot to create its own.
|
2007-07-26 11:41:03 -06:00
|
|
|
*
|
|
|
|
* We lay them out of the way, just below the initrd (which is why we need to
|
|
|
|
* know its size). */
|
2007-07-19 02:49:29 -06:00
|
|
|
static unsigned long setup_pagetables(unsigned long mem,
|
2007-10-21 19:03:36 -06:00
|
|
|
unsigned long initrd_size)
|
2007-07-19 02:49:29 -06:00
|
|
|
{
|
2007-10-21 19:03:31 -06:00
|
|
|
unsigned long *pgdir, *linear;
|
2007-07-19 02:49:29 -06:00
|
|
|
unsigned int mapped_pages, i, linear_pages;
|
2007-10-21 19:03:31 -06:00
|
|
|
unsigned int ptes_per_page = getpagesize()/sizeof(void *);
|
2007-07-19 02:49:29 -06:00
|
|
|
|
2007-10-21 19:03:36 -06:00
|
|
|
mapped_pages = mem/getpagesize();
|
2007-07-19 02:49:29 -06:00
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/* Each PTE page can map ptes_per_page pages: how many do we need? */
|
2007-07-19 02:49:29 -06:00
|
|
|
linear_pages = (mapped_pages + ptes_per_page-1)/ptes_per_page;
|
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/* We put the toplevel page directory page at the top of memory. */
|
2007-10-21 19:03:26 -06:00
|
|
|
pgdir = from_guest_phys(mem) - initrd_size - getpagesize();
|
2007-07-26 11:41:03 -06:00
|
|
|
|
|
|
|
/* Now we use the next linear_pages pages as pte pages */
|
2007-07-19 02:49:29 -06:00
|
|
|
linear = (void *)pgdir - linear_pages*getpagesize();
|
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/* Linear mapping is easy: put every page's address into the mapping in
|
|
|
|
* order. PAGE_PRESENT contains the flags Present, Writable and
|
|
|
|
* Executable. */
|
2007-07-19 02:49:29 -06:00
|
|
|
for (i = 0; i < mapped_pages; i++)
|
|
|
|
linear[i] = ((i * getpagesize()) | PAGE_PRESENT);
|
|
|
|
|
2007-10-21 19:03:36 -06:00
|
|
|
/* The top level points to the linear page table pages above. */
|
2007-07-19 02:49:29 -06:00
|
|
|
for (i = 0; i < mapped_pages; i += ptes_per_page) {
|
2007-10-21 19:03:36 -06:00
|
|
|
pgdir[i/ptes_per_page]
|
2007-10-21 19:03:31 -06:00
|
|
|
= ((to_guest_phys(linear) + i*sizeof(void *))
|
2007-10-21 19:03:26 -06:00
|
|
|
| PAGE_PRESENT);
|
2007-07-19 02:49:29 -06:00
|
|
|
}
|
|
|
|
|
2007-10-21 19:03:26 -06:00
|
|
|
verbose("Linear mapping of %u pages in %u pte pages at %#lx\n",
|
|
|
|
mapped_pages, linear_pages, to_guest_phys(linear));
|
2007-07-19 02:49:29 -06:00
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/* We return the top level (guest-physical) address: the kernel needs
|
|
|
|
* to know where it is. */
|
2007-10-21 19:03:26 -06:00
|
|
|
return to_guest_phys(pgdir);
|
2007-07-19 02:49:29 -06:00
|
|
|
}
|
2007-10-24 23:02:50 -06:00
|
|
|
/*:*/
|
2007-07-19 02:49:29 -06:00
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/* Simple routine to roll all the commandline arguments together with spaces
|
|
|
|
* between them. */
|
2007-07-19 02:49:29 -06:00
|
|
|
static void concat(char *dst, char *args[])
|
|
|
|
{
|
|
|
|
unsigned int i, len = 0;
|
|
|
|
|
|
|
|
for (i = 0; args[i]; i++) {
|
|
|
|
strcpy(dst+len, args[i]);
|
|
|
|
strcat(dst+len, " ");
|
|
|
|
len += strlen(args[i]) + 1;
|
|
|
|
}
|
|
|
|
/* In case it's empty. */
|
|
|
|
dst[len] = '\0';
|
|
|
|
}
|
|
|
|
|
2007-10-24 23:02:50 -06:00
|
|
|
/*L:185 This is where we actually tell the kernel to initialize the Guest. We
|
|
|
|
* saw the arguments it expects when we looked at initialize() in lguest_user.c:
|
|
|
|
* the base of Guest "physical" memory, the top physical page to allow, the
|
2007-10-21 19:03:36 -06:00
|
|
|
* top level pagetable and the entry point for the Guest. */
|
|
|
|
static int tell_kernel(unsigned long pgdir, unsigned long start)
|
2007-07-19 02:49:29 -06:00
|
|
|
{
|
2007-10-21 19:03:31 -06:00
|
|
|
unsigned long args[] = { LHREQ_INITIALIZE,
|
|
|
|
(unsigned long)guest_base,
|
2007-10-21 19:03:36 -06:00
|
|
|
guest_limit / getpagesize(), pgdir, start };
|
2007-07-19 02:49:29 -06:00
|
|
|
int fd;
|
|
|
|
|
2007-10-21 19:03:26 -06:00
|
|
|
verbose("Guest: %p - %p (%#lx)\n",
|
|
|
|
guest_base, guest_base + guest_limit, guest_limit);
|
2007-07-19 02:49:29 -06:00
|
|
|
fd = open_or_die("/dev/lguest", O_RDWR);
|
|
|
|
if (write(fd, args, sizeof(args)) < 0)
|
|
|
|
err(1, "Writing to /dev/lguest");
|
2007-07-26 11:41:03 -06:00
|
|
|
|
|
|
|
/* We return the /dev/lguest file descriptor to control this Guest */
|
2007-07-19 02:49:29 -06:00
|
|
|
return fd;
|
|
|
|
}
|
2007-07-26 11:41:03 -06:00
|
|
|
/*:*/
|
2007-07-19 02:49:29 -06:00
|
|
|
|
2007-10-21 19:24:22 -06:00
|
|
|
static void add_device_fd(int fd)
|
2007-07-19 02:49:29 -06:00
|
|
|
{
|
2007-10-21 19:24:22 -06:00
|
|
|
FD_SET(fd, &devices.infds);
|
|
|
|
if (fd > devices.max_infd)
|
|
|
|
devices.max_infd = fd;
|
2007-07-19 02:49:29 -06:00
|
|
|
}
|
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/*L:200
|
|
|
|
* The Waker.
|
|
|
|
*
|
2007-10-24 23:02:50 -06:00
|
|
|
* With console, block and network devices, we can have lots of input which we
|
|
|
|
* need to process. We could try to tell the kernel what file descriptors to
|
|
|
|
* watch, but handing a file descriptor mask through to the kernel is fairly
|
|
|
|
* icky.
|
2007-07-26 11:41:03 -06:00
|
|
|
*
|
|
|
|
* Instead, we fork off a process which watches the file descriptors and writes
|
2007-10-24 23:02:50 -06:00
|
|
|
* the LHREQ_BREAK command to the /dev/lguest file descriptor to tell the Host
|
|
|
|
* stop running the Guest. This causes the Launcher to return from the
|
2007-07-26 11:41:03 -06:00
|
|
|
* /dev/lguest read with -EAGAIN, where it will write to /dev/lguest to reset
|
|
|
|
* the LHREQ_BREAK and wake us up again.
|
|
|
|
*
|
|
|
|
* This, of course, is merely a different *kind* of icky.
|
|
|
|
*/
|
2007-10-21 19:24:22 -06:00
|
|
|
static void wake_parent(int pipefd, int lguest_fd)
|
2007-07-19 02:49:29 -06:00
|
|
|
{
|
2007-07-26 11:41:03 -06:00
|
|
|
/* Add the pipe from the Launcher to the fdset in the device_list, so
|
|
|
|
* we watch it, too. */
|
2007-10-21 19:24:22 -06:00
|
|
|
add_device_fd(pipefd);
|
2007-07-19 02:49:29 -06:00
|
|
|
|
|
|
|
for (;;) {
|
2007-10-21 19:24:22 -06:00
|
|
|
fd_set rfds = devices.infds;
|
2007-10-21 19:03:31 -06:00
|
|
|
unsigned long args[] = { LHREQ_BREAK, 1 };
|
2007-07-19 02:49:29 -06:00
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/* Wait until input is ready from one of the devices. */
|
2007-10-21 19:24:22 -06:00
|
|
|
select(devices.max_infd+1, &rfds, NULL, NULL, NULL);
|
2007-07-26 11:41:03 -06:00
|
|
|
/* Is it a message from the Launcher? */
|
2007-07-19 02:49:29 -06:00
|
|
|
if (FD_ISSET(pipefd, &rfds)) {
|
2007-10-21 19:24:23 -06:00
|
|
|
int fd;
|
2007-07-26 11:41:03 -06:00
|
|
|
/* If read() returns 0, it means the Launcher has
|
|
|
|
* exited. We silently follow. */
|
2007-10-21 19:24:23 -06:00
|
|
|
if (read(pipefd, &fd, sizeof(fd)) == 0)
|
2007-07-19 02:49:29 -06:00
|
|
|
exit(0);
|
2007-10-21 19:24:23 -06:00
|
|
|
/* Otherwise it's telling us to change what file
|
2007-10-24 23:02:50 -06:00
|
|
|
* descriptors we're to listen to. Positive means
|
|
|
|
* listen to a new one, negative means stop
|
|
|
|
* listening. */
|
2007-10-21 19:24:23 -06:00
|
|
|
if (fd >= 0)
|
|
|
|
FD_SET(fd, &devices.infds);
|
|
|
|
else
|
|
|
|
FD_CLR(-fd - 1, &devices.infds);
|
2007-07-26 11:41:03 -06:00
|
|
|
} else /* Send LHREQ_BREAK command. */
|
2007-07-19 02:49:29 -06:00
|
|
|
write(lguest_fd, args, sizeof(args));
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/* This routine just sets up a pipe to the Waker process. */
|
2007-10-21 19:24:22 -06:00
|
|
|
static int setup_waker(int lguest_fd)
|
2007-07-19 02:49:29 -06:00
|
|
|
{
|
|
|
|
int pipefd[2], child;
|
|
|
|
|
2007-10-24 23:02:50 -06:00
|
|
|
/* We create a pipe to talk to the Waker, and also so it knows when the
|
2007-07-26 11:41:03 -06:00
|
|
|
* Launcher dies (and closes pipe). */
|
2007-07-19 02:49:29 -06:00
|
|
|
pipe(pipefd);
|
|
|
|
child = fork();
|
|
|
|
if (child == -1)
|
|
|
|
err(1, "forking");
|
|
|
|
|
|
|
|
if (child == 0) {
|
2007-10-24 23:02:50 -06:00
|
|
|
/* We are the Waker: close the "writing" end of our copy of the
|
|
|
|
* pipe and start waiting for input. */
|
2007-07-19 02:49:29 -06:00
|
|
|
close(pipefd[1]);
|
2007-10-21 19:24:22 -06:00
|
|
|
wake_parent(pipefd[0], lguest_fd);
|
2007-07-19 02:49:29 -06:00
|
|
|
}
|
2007-07-26 11:41:03 -06:00
|
|
|
/* Close the reading end of our copy of the pipe. */
|
2007-07-19 02:49:29 -06:00
|
|
|
close(pipefd[0]);
|
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/* Here is the fd used to talk to the waker. */
|
2007-07-19 02:49:29 -06:00
|
|
|
return pipefd[1];
|
|
|
|
}
|
|
|
|
|
2007-10-24 23:02:50 -06:00
|
|
|
/*
|
2007-07-26 11:41:03 -06:00
|
|
|
* Device Handling.
|
|
|
|
*
|
2007-10-24 23:02:50 -06:00
|
|
|
* When the Guest gives us a buffer, it sends an array of addresses and sizes.
|
2007-07-26 11:41:03 -06:00
|
|
|
* We need to make sure it's not trying to reach into the Launcher itself, so
|
2007-10-24 23:02:50 -06:00
|
|
|
* we have a convenient routine which checks it and exits with an error message
|
2007-07-26 11:41:03 -06:00
|
|
|
* if something funny is going on:
|
|
|
|
*/
|
2007-07-19 02:49:29 -06:00
|
|
|
static void *_check_pointer(unsigned long addr, unsigned int size,
|
|
|
|
unsigned int line)
|
|
|
|
{
|
2007-07-26 11:41:03 -06:00
|
|
|
/* We have to separately check addr and addr+size, because size could
|
|
|
|
* be huge and addr + size might wrap around. */
|
2007-10-21 19:03:26 -06:00
|
|
|
if (addr >= guest_limit || addr + size >= guest_limit)
|
2007-10-21 19:24:22 -06:00
|
|
|
errx(1, "%s:%i: Invalid address %#lx", __FILE__, line, addr);
|
2007-07-26 11:41:03 -06:00
|
|
|
/* We return a pointer for the caller's convenience, now we know it's
|
|
|
|
* safe to use. */
|
2007-10-21 19:03:26 -06:00
|
|
|
return from_guest_phys(addr);
|
2007-07-19 02:49:29 -06:00
|
|
|
}
|
2007-07-26 11:41:03 -06:00
|
|
|
/* A macro which transparently hands the line number to the real function. */
|
2007-07-19 02:49:29 -06:00
|
|
|
#define check_pointer(addr,size) _check_pointer(addr, size, __LINE__)
|
|
|
|
|
2007-10-24 23:02:50 -06:00
|
|
|
/* Each buffer in the virtqueues is actually a chain of descriptors. This
|
|
|
|
* function returns the next descriptor in the chain, or vq->vring.num if we're
|
|
|
|
* at the end. */
|
2007-10-21 19:24:22 -06:00
|
|
|
static unsigned next_desc(struct virtqueue *vq, unsigned int i)
|
|
|
|
{
|
|
|
|
unsigned int next;
|
|
|
|
|
|
|
|
/* If this descriptor says it doesn't chain, we're done. */
|
|
|
|
if (!(vq->vring.desc[i].flags & VRING_DESC_F_NEXT))
|
|
|
|
return vq->vring.num;
|
|
|
|
|
|
|
|
/* Check they're not leading us off end of descriptors. */
|
|
|
|
next = vq->vring.desc[i].next;
|
|
|
|
/* Make sure compiler knows to grab that: we don't want it changing! */
|
|
|
|
wmb();
|
|
|
|
|
|
|
|
if (next >= vq->vring.num)
|
|
|
|
errx(1, "Desc next is %u", next);
|
|
|
|
|
|
|
|
return next;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* This looks in the virtqueue and for the first available buffer, and converts
|
|
|
|
* it to an iovec for convenient access. Since descriptors consist of some
|
|
|
|
* number of output then some number of input descriptors, it's actually two
|
|
|
|
* iovecs, but we pack them into one and note how many of each there were.
|
|
|
|
*
|
|
|
|
* This function returns the descriptor number found, or vq->vring.num (which
|
|
|
|
* is never a valid descriptor number) if none was found. */
|
|
|
|
static unsigned get_vq_desc(struct virtqueue *vq,
|
|
|
|
struct iovec iov[],
|
|
|
|
unsigned int *out_num, unsigned int *in_num)
|
|
|
|
{
|
|
|
|
unsigned int i, head;
|
|
|
|
|
|
|
|
/* Check it isn't doing very strange things with descriptor numbers. */
|
|
|
|
if ((u16)(vq->vring.avail->idx - vq->last_avail_idx) > vq->vring.num)
|
|
|
|
errx(1, "Guest moved used index from %u to %u",
|
|
|
|
vq->last_avail_idx, vq->vring.avail->idx);
|
|
|
|
|
|
|
|
/* If there's nothing new since last we looked, return invalid. */
|
|
|
|
if (vq->vring.avail->idx == vq->last_avail_idx)
|
|
|
|
return vq->vring.num;
|
|
|
|
|
|
|
|
/* Grab the next descriptor number they're advertising, and increment
|
|
|
|
* the index we've seen. */
|
|
|
|
head = vq->vring.avail->ring[vq->last_avail_idx++ % vq->vring.num];
|
|
|
|
|
|
|
|
/* If their number is silly, that's a fatal mistake. */
|
|
|
|
if (head >= vq->vring.num)
|
|
|
|
errx(1, "Guest says index %u is available", head);
|
|
|
|
|
|
|
|
/* When we start there are none of either input nor output. */
|
|
|
|
*out_num = *in_num = 0;
|
|
|
|
|
|
|
|
i = head;
|
|
|
|
do {
|
|
|
|
/* Grab the first descriptor, and check it's OK. */
|
|
|
|
iov[*out_num + *in_num].iov_len = vq->vring.desc[i].len;
|
|
|
|
iov[*out_num + *in_num].iov_base
|
|
|
|
= check_pointer(vq->vring.desc[i].addr,
|
|
|
|
vq->vring.desc[i].len);
|
|
|
|
/* If this is an input descriptor, increment that count. */
|
|
|
|
if (vq->vring.desc[i].flags & VRING_DESC_F_WRITE)
|
|
|
|
(*in_num)++;
|
|
|
|
else {
|
|
|
|
/* If it's an output descriptor, they're all supposed
|
|
|
|
* to come before any input descriptors. */
|
|
|
|
if (*in_num)
|
|
|
|
errx(1, "Descriptor has out after in");
|
|
|
|
(*out_num)++;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* If we've got too many, that implies a descriptor loop. */
|
|
|
|
if (*out_num + *in_num > vq->vring.num)
|
|
|
|
errx(1, "Looped descriptor");
|
|
|
|
} while ((i = next_desc(vq, i)) != vq->vring.num);
|
2007-07-26 11:41:03 -06:00
|
|
|
|
2007-10-21 19:24:22 -06:00
|
|
|
return head;
|
2007-07-19 02:49:29 -06:00
|
|
|
}
|
|
|
|
|
2007-10-24 23:02:50 -06:00
|
|
|
/* After we've used one of their buffers, we tell them about it. We'll then
|
2007-10-21 19:24:22 -06:00
|
|
|
* want to send them an interrupt, using trigger_irq(). */
|
|
|
|
static void add_used(struct virtqueue *vq, unsigned int head, int len)
|
2007-07-19 02:49:29 -06:00
|
|
|
{
|
2007-10-21 19:24:22 -06:00
|
|
|
struct vring_used_elem *used;
|
|
|
|
|
2007-10-24 23:02:50 -06:00
|
|
|
/* The virtqueue contains a ring of used buffers. Get a pointer to the
|
|
|
|
* next entry in that used ring. */
|
2007-10-21 19:24:22 -06:00
|
|
|
used = &vq->vring.used->ring[vq->vring.used->idx % vq->vring.num];
|
|
|
|
used->id = head;
|
|
|
|
used->len = len;
|
|
|
|
/* Make sure buffer is written before we update index. */
|
|
|
|
wmb();
|
|
|
|
vq->vring.used->idx++;
|
2007-07-19 02:49:29 -06:00
|
|
|
}
|
|
|
|
|
2007-10-21 19:24:22 -06:00
|
|
|
/* This actually sends the interrupt for this virtqueue */
|
|
|
|
static void trigger_irq(int fd, struct virtqueue *vq)
|
2007-07-19 02:49:29 -06:00
|
|
|
{
|
2007-10-21 19:24:22 -06:00
|
|
|
unsigned long buf[] = { LHREQ_IRQ, vq->config.irq };
|
|
|
|
|
2007-10-24 23:02:50 -06:00
|
|
|
/* If they don't want an interrupt, don't send one. */
|
2007-10-21 19:24:22 -06:00
|
|
|
if (vq->vring.avail->flags & VRING_AVAIL_F_NO_INTERRUPT)
|
|
|
|
return;
|
|
|
|
|
|
|
|
/* Send the Guest an interrupt tell them we used something up. */
|
2007-07-19 02:49:29 -06:00
|
|
|
if (write(fd, buf, sizeof(buf)) != 0)
|
2007-10-21 19:24:22 -06:00
|
|
|
err(1, "Triggering irq %i", vq->config.irq);
|
2007-07-19 02:49:29 -06:00
|
|
|
}
|
|
|
|
|
2007-10-21 19:24:22 -06:00
|
|
|
/* And here's the combo meal deal. Supersize me! */
|
|
|
|
static void add_used_and_trigger(int fd, struct virtqueue *vq,
|
|
|
|
unsigned int head, int len)
|
2007-07-19 02:49:29 -06:00
|
|
|
{
|
2007-10-21 19:24:22 -06:00
|
|
|
add_used(vq, head, len);
|
|
|
|
trigger_irq(fd, vq);
|
2007-07-19 02:49:29 -06:00
|
|
|
}
|
|
|
|
|
2007-10-24 23:02:50 -06:00
|
|
|
/*
|
|
|
|
* The Console
|
|
|
|
*
|
|
|
|
* Here is the input terminal setting we save, and the routine to restore them
|
|
|
|
* on exit so the user gets their terminal back. */
|
2007-07-19 02:49:29 -06:00
|
|
|
static struct termios orig_term;
|
|
|
|
static void restore_term(void)
|
|
|
|
{
|
|
|
|
tcsetattr(STDIN_FILENO, TCSANOW, &orig_term);
|
|
|
|
}
|
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/* We associate some data with the console for our exit hack. */
|
2007-07-19 02:49:29 -06:00
|
|
|
struct console_abort
|
|
|
|
{
|
2007-07-26 11:41:03 -06:00
|
|
|
/* How many times have they hit ^C? */
|
2007-07-19 02:49:29 -06:00
|
|
|
int count;
|
2007-07-26 11:41:03 -06:00
|
|
|
/* When did they start? */
|
2007-07-19 02:49:29 -06:00
|
|
|
struct timeval start;
|
|
|
|
};
|
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/* This is the routine which handles console input (ie. stdin). */
|
2007-07-19 02:49:29 -06:00
|
|
|
static bool handle_console_input(int fd, struct device *dev)
|
|
|
|
{
|
|
|
|
int len;
|
2007-10-21 19:24:22 -06:00
|
|
|
unsigned int head, in_num, out_num;
|
|
|
|
struct iovec iov[dev->vq->vring.num];
|
2007-07-19 02:49:29 -06:00
|
|
|
struct console_abort *abort = dev->priv;
|
|
|
|
|
2007-10-21 19:24:22 -06:00
|
|
|
/* First we need a console buffer from the Guests's input virtqueue. */
|
|
|
|
head = get_vq_desc(dev->vq, iov, &out_num, &in_num);
|
2007-10-21 19:24:23 -06:00
|
|
|
|
|
|
|
/* If they're not ready for input, stop listening to this file
|
|
|
|
* descriptor. We'll start again once they add an input buffer. */
|
|
|
|
if (head == dev->vq->vring.num)
|
|
|
|
return false;
|
|
|
|
|
|
|
|
if (out_num)
|
2007-10-21 19:24:22 -06:00
|
|
|
errx(1, "Output buffers in console in queue?");
|
2007-07-19 02:49:29 -06:00
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/* This is why we convert to iovecs: the readv() call uses them, and so
|
|
|
|
* it reads straight into the Guest's buffer. */
|
2007-10-21 19:24:22 -06:00
|
|
|
len = readv(dev->fd, iov, in_num);
|
2007-07-19 02:49:29 -06:00
|
|
|
if (len <= 0) {
|
2007-07-26 11:41:03 -06:00
|
|
|
/* This implies that the console is closed, is /dev/null, or
|
2007-10-21 19:24:22 -06:00
|
|
|
* something went terribly wrong. */
|
2007-07-19 02:49:29 -06:00
|
|
|
warnx("Failed to get console input, ignoring console.");
|
2007-10-21 19:24:23 -06:00
|
|
|
/* Put the input terminal back. */
|
2007-10-21 19:24:22 -06:00
|
|
|
restore_term();
|
2007-10-21 19:24:23 -06:00
|
|
|
/* Remove callback from input vq, so it doesn't restart us. */
|
|
|
|
dev->vq->handle_output = NULL;
|
|
|
|
/* Stop listening to this fd: don't call us again. */
|
2007-10-21 19:24:22 -06:00
|
|
|
return false;
|
2007-07-19 02:49:29 -06:00
|
|
|
}
|
|
|
|
|
2007-10-21 19:24:23 -06:00
|
|
|
/* Tell the Guest about the new input. */
|
|
|
|
add_used_and_trigger(fd, dev->vq, head, len);
|
2007-07-19 02:49:29 -06:00
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/* Three ^C within one second? Exit.
|
|
|
|
*
|
|
|
|
* This is such a hack, but works surprisingly well. Each ^C has to be
|
|
|
|
* in a buffer by itself, so they can't be too fast. But we check that
|
|
|
|
* we get three within about a second, so they can't be too slow. */
|
2007-07-19 02:49:29 -06:00
|
|
|
if (len == 1 && ((char *)iov[0].iov_base)[0] == 3) {
|
|
|
|
if (!abort->count++)
|
|
|
|
gettimeofday(&abort->start, NULL);
|
|
|
|
else if (abort->count == 3) {
|
|
|
|
struct timeval now;
|
|
|
|
gettimeofday(&now, NULL);
|
|
|
|
if (now.tv_sec <= abort->start.tv_sec+1) {
|
2007-10-21 19:03:31 -06:00
|
|
|
unsigned long args[] = { LHREQ_BREAK, 0 };
|
2007-07-26 11:41:03 -06:00
|
|
|
/* Close the fd so Waker will know it has to
|
|
|
|
* exit. */
|
2007-07-19 02:49:29 -06:00
|
|
|
close(waker_fd);
|
2007-07-26 11:41:03 -06:00
|
|
|
/* Just in case waker is blocked in BREAK, send
|
|
|
|
* unbreak now. */
|
2007-07-19 02:49:29 -06:00
|
|
|
write(fd, args, sizeof(args));
|
|
|
|
exit(2);
|
|
|
|
}
|
|
|
|
abort->count = 0;
|
|
|
|
}
|
|
|
|
} else
|
2007-07-26 11:41:03 -06:00
|
|
|
/* Any other key resets the abort counter. */
|
2007-07-19 02:49:29 -06:00
|
|
|
abort->count = 0;
|
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/* Everything went OK! */
|
2007-07-19 02:49:29 -06:00
|
|
|
return true;
|
|
|
|
}
|
|
|
|
|
2007-10-21 19:24:22 -06:00
|
|
|
/* Handling output for console is simple: we just get all the output buffers
|
|
|
|
* and write them to stdout. */
|
|
|
|
static void handle_console_output(int fd, struct virtqueue *vq)
|
2007-07-19 02:49:29 -06:00
|
|
|
{
|
2007-10-21 19:24:22 -06:00
|
|
|
unsigned int head, out, in;
|
|
|
|
int len;
|
|
|
|
struct iovec iov[vq->vring.num];
|
|
|
|
|
|
|
|
/* Keep getting output buffers from the Guest until we run out. */
|
|
|
|
while ((head = get_vq_desc(vq, iov, &out, &in)) != vq->vring.num) {
|
|
|
|
if (in)
|
|
|
|
errx(1, "Input buffers in output queue?");
|
|
|
|
len = writev(STDOUT_FILENO, iov, out);
|
|
|
|
add_used_and_trigger(fd, vq, head, len);
|
|
|
|
}
|
2007-07-19 02:49:29 -06:00
|
|
|
}
|
|
|
|
|
2007-10-24 23:02:50 -06:00
|
|
|
/*
|
|
|
|
* The Network
|
|
|
|
*
|
|
|
|
* Handling output for network is also simple: we get all the output buffers
|
2007-10-21 19:24:22 -06:00
|
|
|
* and write them (ignoring the first element) to this device's file descriptor
|
|
|
|
* (stdout). */
|
|
|
|
static void handle_net_output(int fd, struct virtqueue *vq)
|
2007-07-19 02:49:29 -06:00
|
|
|
{
|
2007-10-21 19:24:22 -06:00
|
|
|
unsigned int head, out, in;
|
|
|
|
int len;
|
|
|
|
struct iovec iov[vq->vring.num];
|
|
|
|
|
|
|
|
/* Keep getting output buffers from the Guest until we run out. */
|
|
|
|
while ((head = get_vq_desc(vq, iov, &out, &in)) != vq->vring.num) {
|
|
|
|
if (in)
|
|
|
|
errx(1, "Input buffers in output queue?");
|
2007-10-24 23:02:50 -06:00
|
|
|
/* Check header, but otherwise ignore it (we told the Guest we
|
|
|
|
* supported no features, so it shouldn't have anything
|
|
|
|
* interesting). */
|
2007-10-21 19:24:22 -06:00
|
|
|
(void)convert(&iov[0], struct virtio_net_hdr);
|
|
|
|
len = writev(vq->dev->fd, iov+1, out-1);
|
|
|
|
add_used_and_trigger(fd, vq, head, len);
|
|
|
|
}
|
2007-07-19 02:49:29 -06:00
|
|
|
}
|
|
|
|
|
2007-10-21 19:24:22 -06:00
|
|
|
/* This is where we handle a packet coming in from the tun device to our
|
|
|
|
* Guest. */
|
2007-07-19 02:49:29 -06:00
|
|
|
static bool handle_tun_input(int fd, struct device *dev)
|
|
|
|
{
|
2007-10-21 19:24:22 -06:00
|
|
|
unsigned int head, in_num, out_num;
|
2007-07-19 02:49:29 -06:00
|
|
|
int len;
|
2007-10-21 19:24:22 -06:00
|
|
|
struct iovec iov[dev->vq->vring.num];
|
|
|
|
struct virtio_net_hdr *hdr;
|
2007-07-19 02:49:29 -06:00
|
|
|
|
2007-10-21 19:24:22 -06:00
|
|
|
/* First we need a network buffer from the Guests's recv virtqueue. */
|
|
|
|
head = get_vq_desc(dev->vq, iov, &out_num, &in_num);
|
|
|
|
if (head == dev->vq->vring.num) {
|
2007-07-26 11:41:03 -06:00
|
|
|
/* Now, it's expected that if we try to send a packet too
|
2007-10-21 19:24:22 -06:00
|
|
|
* early, the Guest won't be ready yet. Wait until the device
|
|
|
|
* status says it's ready. */
|
|
|
|
/* FIXME: Actually want DRIVER_ACTIVE here. */
|
|
|
|
if (dev->desc->status & VIRTIO_CONFIG_S_DRIVER_OK)
|
2007-07-19 02:49:29 -06:00
|
|
|
warn("network: no dma buffer!");
|
2007-10-21 19:24:23 -06:00
|
|
|
/* We'll turn this back on if input buffers are registered. */
|
|
|
|
return false;
|
2007-10-21 19:24:22 -06:00
|
|
|
} else if (out_num)
|
|
|
|
errx(1, "Output buffers in network recv queue?");
|
|
|
|
|
|
|
|
/* First element is the header: we set it to 0 (no features). */
|
|
|
|
hdr = convert(&iov[0], struct virtio_net_hdr);
|
|
|
|
hdr->flags = 0;
|
|
|
|
hdr->gso_type = VIRTIO_NET_HDR_GSO_NONE;
|
2007-07-19 02:49:29 -06:00
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/* Read the packet from the device directly into the Guest's buffer. */
|
2007-10-21 19:24:22 -06:00
|
|
|
len = readv(dev->fd, iov+1, in_num-1);
|
2007-07-19 02:49:29 -06:00
|
|
|
if (len <= 0)
|
|
|
|
err(1, "reading network");
|
2007-07-26 11:41:03 -06:00
|
|
|
|
2007-10-21 19:24:23 -06:00
|
|
|
/* Tell the Guest about the new packet. */
|
|
|
|
add_used_and_trigger(fd, dev->vq, head, sizeof(*hdr) + len);
|
2007-10-21 19:24:22 -06:00
|
|
|
|
2007-07-19 02:49:29 -06:00
|
|
|
verbose("tun input packet len %i [%02x %02x] (%s)\n", len,
|
2007-10-21 19:24:22 -06:00
|
|
|
((u8 *)iov[1].iov_base)[0], ((u8 *)iov[1].iov_base)[1],
|
|
|
|
head != dev->vq->vring.num ? "sent" : "discarded");
|
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/* All good. */
|
2007-07-19 02:49:29 -06:00
|
|
|
return true;
|
|
|
|
}
|
|
|
|
|
2007-10-24 23:02:50 -06:00
|
|
|
/*L:215 This is the callback attached to the network and console input
|
|
|
|
* virtqueues: it ensures we try again, in case we stopped console or net
|
2007-10-21 19:24:23 -06:00
|
|
|
* delivery because Guest didn't have any buffers. */
|
|
|
|
static void enable_fd(int fd, struct virtqueue *vq)
|
|
|
|
{
|
|
|
|
add_device_fd(vq->dev->fd);
|
|
|
|
/* Tell waker to listen to it again */
|
|
|
|
write(waker_fd, &vq->dev->fd, sizeof(vq->dev->fd));
|
|
|
|
}
|
|
|
|
|
2007-10-21 19:24:22 -06:00
|
|
|
/* This is the generic routine we call when the Guest uses LHCALL_NOTIFY. */
|
|
|
|
static void handle_output(int fd, unsigned long addr)
|
2007-07-19 02:49:29 -06:00
|
|
|
{
|
|
|
|
struct device *i;
|
2007-10-21 19:24:22 -06:00
|
|
|
struct virtqueue *vq;
|
|
|
|
|
|
|
|
/* Check each virtqueue. */
|
|
|
|
for (i = devices.dev; i; i = i->next) {
|
|
|
|
for (vq = i->vq; vq; vq = vq->next) {
|
|
|
|
if (vq->config.pfn == addr/getpagesize()
|
|
|
|
&& vq->handle_output) {
|
|
|
|
verbose("Output to %s\n", vq->dev->name);
|
|
|
|
vq->handle_output(fd, vq);
|
|
|
|
return;
|
|
|
|
}
|
2007-07-19 02:49:29 -06:00
|
|
|
}
|
|
|
|
}
|
2007-07-26 11:41:03 -06:00
|
|
|
|
2007-10-21 19:24:22 -06:00
|
|
|
/* Early console write is done using notify on a nul-terminated string
|
|
|
|
* in Guest memory. */
|
|
|
|
if (addr >= guest_limit)
|
|
|
|
errx(1, "Bad NOTIFY %#lx", addr);
|
|
|
|
|
|
|
|
write(STDOUT_FILENO, from_guest_phys(addr),
|
|
|
|
strnlen(from_guest_phys(addr), guest_limit - addr));
|
2007-07-19 02:49:29 -06:00
|
|
|
}
|
|
|
|
|
2007-10-24 23:02:50 -06:00
|
|
|
/* This is called when the Waker wakes us up: check for incoming file
|
2007-07-26 11:41:03 -06:00
|
|
|
* descriptors. */
|
2007-10-21 19:24:22 -06:00
|
|
|
static void handle_input(int fd)
|
2007-07-19 02:49:29 -06:00
|
|
|
{
|
2007-07-26 11:41:03 -06:00
|
|
|
/* select() wants a zeroed timeval to mean "don't wait". */
|
2007-07-19 02:49:29 -06:00
|
|
|
struct timeval poll = { .tv_sec = 0, .tv_usec = 0 };
|
|
|
|
|
|
|
|
for (;;) {
|
|
|
|
struct device *i;
|
2007-10-21 19:24:22 -06:00
|
|
|
fd_set fds = devices.infds;
|
2007-07-19 02:49:29 -06:00
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/* If nothing is ready, we're done. */
|
2007-10-21 19:24:22 -06:00
|
|
|
if (select(devices.max_infd+1, &fds, NULL, NULL, &poll) == 0)
|
2007-07-19 02:49:29 -06:00
|
|
|
break;
|
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/* Otherwise, call the device(s) which have readable
|
|
|
|
* file descriptors and a method of handling them. */
|
2007-10-21 19:24:22 -06:00
|
|
|
for (i = devices.dev; i; i = i->next) {
|
2007-07-19 02:49:29 -06:00
|
|
|
if (i->handle_input && FD_ISSET(i->fd, &fds)) {
|
2007-10-21 19:24:23 -06:00
|
|
|
int dev_fd;
|
|
|
|
if (i->handle_input(fd, i))
|
|
|
|
continue;
|
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/* If handle_input() returns false, it means we
|
2007-10-21 19:24:23 -06:00
|
|
|
* should no longer service it. Networking and
|
|
|
|
* console do this when there's no input
|
|
|
|
* buffers to deliver into. Console also uses
|
|
|
|
* it when it discovers that stdin is
|
|
|
|
* closed. */
|
|
|
|
FD_CLR(i->fd, &devices.infds);
|
|
|
|
/* Tell waker to ignore it too, by sending a
|
|
|
|
* negative fd number (-1, since 0 is a valid
|
|
|
|
* FD number). */
|
|
|
|
dev_fd = -i->fd - 1;
|
|
|
|
write(waker_fd, &dev_fd, sizeof(dev_fd));
|
2007-07-19 02:49:29 -06:00
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/*L:190
|
|
|
|
* Device Setup
|
|
|
|
*
|
|
|
|
* All devices need a descriptor so the Guest knows it exists, and a "struct
|
|
|
|
* device" so the Launcher can keep track of it. We have common helper
|
|
|
|
* routines to allocate them.
|
|
|
|
*
|
|
|
|
* This routine allocates a new "struct lguest_device_desc" from descriptor
|
2007-10-21 19:24:22 -06:00
|
|
|
* table just above the Guest's normal memory. It returns a pointer to that
|
|
|
|
* descriptor. */
|
|
|
|
static struct lguest_device_desc *new_dev_desc(u16 type)
|
2007-07-19 02:49:29 -06:00
|
|
|
{
|
2007-10-21 19:24:22 -06:00
|
|
|
struct lguest_device_desc *d;
|
2007-07-19 02:49:29 -06:00
|
|
|
|
2007-10-21 19:24:22 -06:00
|
|
|
/* We only have one page for all the descriptors. */
|
|
|
|
if (devices.desc_used + sizeof(*d) > getpagesize())
|
|
|
|
errx(1, "Too many devices");
|
|
|
|
|
|
|
|
/* We don't need to set config_len or status: page is 0 already. */
|
|
|
|
d = (void *)devices.descpage + devices.desc_used;
|
|
|
|
d->type = type;
|
|
|
|
devices.desc_used += sizeof(*d);
|
|
|
|
|
|
|
|
return d;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Each device descriptor is followed by some configuration information.
|
2007-10-24 23:02:50 -06:00
|
|
|
* Each configuration field looks like: u8 type, u8 len, [... len bytes...].
|
2007-10-21 19:24:22 -06:00
|
|
|
*
|
|
|
|
* This routine adds a new field to an existing device's descriptor. It only
|
|
|
|
* works for the last device, but that's OK because that's how we use it. */
|
|
|
|
static void add_desc_field(struct device *dev, u8 type, u8 len, const void *c)
|
|
|
|
{
|
|
|
|
/* This is the last descriptor, right? */
|
|
|
|
assert(devices.descpage + devices.desc_used
|
|
|
|
== (u8 *)(dev->desc + 1) + dev->desc->config_len);
|
|
|
|
|
|
|
|
/* We only have one page of device descriptions. */
|
|
|
|
if (devices.desc_used + 2 + len > getpagesize())
|
|
|
|
errx(1, "Too many devices");
|
|
|
|
|
|
|
|
/* Copy in the new config header: type then length. */
|
|
|
|
devices.descpage[devices.desc_used++] = type;
|
|
|
|
devices.descpage[devices.desc_used++] = len;
|
|
|
|
memcpy(devices.descpage + devices.desc_used, c, len);
|
|
|
|
devices.desc_used += len;
|
|
|
|
|
|
|
|
/* Update the device descriptor length: two byte head then data. */
|
|
|
|
dev->desc->config_len += 2 + len;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* This routine adds a virtqueue to a device. We specify how many descriptors
|
|
|
|
* the virtqueue is to have. */
|
|
|
|
static void add_virtqueue(struct device *dev, unsigned int num_descs,
|
|
|
|
void (*handle_output)(int fd, struct virtqueue *me))
|
|
|
|
{
|
|
|
|
unsigned int pages;
|
|
|
|
struct virtqueue **i, *vq = malloc(sizeof(*vq));
|
|
|
|
void *p;
|
|
|
|
|
|
|
|
/* First we need some pages for this virtqueue. */
|
2007-11-11 19:39:18 -07:00
|
|
|
pages = (vring_size(num_descs, getpagesize()) + getpagesize() - 1)
|
|
|
|
/ getpagesize();
|
2007-10-21 19:24:22 -06:00
|
|
|
p = get_pages(pages);
|
|
|
|
|
|
|
|
/* Initialize the configuration. */
|
|
|
|
vq->config.num = num_descs;
|
|
|
|
vq->config.irq = devices.next_irq++;
|
|
|
|
vq->config.pfn = to_guest_phys(p) / getpagesize();
|
|
|
|
|
|
|
|
/* Initialize the vring. */
|
2007-11-11 19:39:18 -07:00
|
|
|
vring_init(&vq->vring, num_descs, p, getpagesize());
|
2007-10-21 19:24:22 -06:00
|
|
|
|
|
|
|
/* Add the configuration information to this device's descriptor. */
|
|
|
|
add_desc_field(dev, VIRTIO_CONFIG_F_VIRTQUEUE,
|
|
|
|
sizeof(vq->config), &vq->config);
|
|
|
|
|
|
|
|
/* Add to tail of list, so dev->vq is first vq, dev->vq->next is
|
|
|
|
* second. */
|
|
|
|
for (i = &dev->vq; *i; i = &(*i)->next);
|
|
|
|
*i = vq;
|
|
|
|
|
|
|
|
/* Link virtqueue back to device. */
|
|
|
|
vq->dev = dev;
|
|
|
|
|
2007-10-24 23:02:50 -06:00
|
|
|
/* Set the routine to call when the Guest does something to this
|
|
|
|
* virtqueue. */
|
2007-10-21 19:24:22 -06:00
|
|
|
vq->handle_output = handle_output;
|
2007-10-24 23:02:50 -06:00
|
|
|
|
|
|
|
/* Set the "Don't Notify Me" flag if we don't have a handler */
|
2007-10-21 19:24:22 -06:00
|
|
|
if (!handle_output)
|
|
|
|
vq->vring.used->flags = VRING_USED_F_NO_NOTIFY;
|
2007-07-19 02:49:29 -06:00
|
|
|
}
|
|
|
|
|
2007-10-21 19:24:22 -06:00
|
|
|
/* This routine does all the creation and setup of a new device, including
|
2007-10-24 23:02:50 -06:00
|
|
|
* calling new_dev_desc() to allocate the descriptor and device memory. */
|
2007-10-21 19:24:22 -06:00
|
|
|
static struct device *new_device(const char *name, u16 type, int fd,
|
|
|
|
bool (*handle_input)(int, struct device *))
|
2007-07-19 02:49:29 -06:00
|
|
|
{
|
|
|
|
struct device *dev = malloc(sizeof(*dev));
|
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/* Append to device list. Prepending to a single-linked list is
|
|
|
|
* easier, but the user expects the devices to be arranged on the bus
|
|
|
|
* in command-line order. The first network device on the command line
|
2007-10-24 23:02:50 -06:00
|
|
|
* is eth0, the first block device /dev/vda, etc. */
|
2007-10-21 19:24:22 -06:00
|
|
|
*devices.lastdev = dev;
|
2007-07-19 02:49:29 -06:00
|
|
|
dev->next = NULL;
|
2007-10-21 19:24:22 -06:00
|
|
|
devices.lastdev = &dev->next;
|
2007-07-19 02:49:29 -06:00
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/* Now we populate the fields one at a time. */
|
2007-07-19 02:49:29 -06:00
|
|
|
dev->fd = fd;
|
2007-07-26 11:41:03 -06:00
|
|
|
/* If we have an input handler for this file descriptor, then we add it
|
|
|
|
* to the device_list's fdset and maxfd. */
|
2007-07-19 02:49:29 -06:00
|
|
|
if (handle_input)
|
2007-10-21 19:24:22 -06:00
|
|
|
add_device_fd(dev->fd);
|
|
|
|
dev->desc = new_dev_desc(type);
|
2007-07-19 02:49:29 -06:00
|
|
|
dev->handle_input = handle_input;
|
2007-10-21 19:24:22 -06:00
|
|
|
dev->name = name;
|
2007-07-19 02:49:29 -06:00
|
|
|
return dev;
|
|
|
|
}
|
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/* Our first setup routine is the console. It's a fairly simple device, but
|
|
|
|
* UNIX tty handling makes it uglier than it could be. */
|
2007-10-21 19:24:22 -06:00
|
|
|
static void setup_console(void)
|
2007-07-19 02:49:29 -06:00
|
|
|
{
|
|
|
|
struct device *dev;
|
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/* If we can save the initial standard input settings... */
|
2007-07-19 02:49:29 -06:00
|
|
|
if (tcgetattr(STDIN_FILENO, &orig_term) == 0) {
|
|
|
|
struct termios term = orig_term;
|
2007-07-26 11:41:03 -06:00
|
|
|
/* Then we turn off echo, line buffering and ^C etc. We want a
|
|
|
|
* raw input stream to the Guest. */
|
2007-07-19 02:49:29 -06:00
|
|
|
term.c_lflag &= ~(ISIG|ICANON|ECHO);
|
|
|
|
tcsetattr(STDIN_FILENO, TCSANOW, &term);
|
2007-07-26 11:41:03 -06:00
|
|
|
/* If we exit gracefully, the original settings will be
|
|
|
|
* restored so the user can see what they're typing. */
|
2007-07-19 02:49:29 -06:00
|
|
|
atexit(restore_term);
|
|
|
|
}
|
|
|
|
|
2007-10-21 19:24:22 -06:00
|
|
|
dev = new_device("console", VIRTIO_ID_CONSOLE,
|
|
|
|
STDIN_FILENO, handle_console_input);
|
2007-07-26 11:41:03 -06:00
|
|
|
/* We store the console state in dev->priv, and initialize it. */
|
2007-07-19 02:49:29 -06:00
|
|
|
dev->priv = malloc(sizeof(struct console_abort));
|
|
|
|
((struct console_abort *)dev->priv)->count = 0;
|
|
|
|
|
2007-10-21 19:24:23 -06:00
|
|
|
/* The console needs two virtqueues: the input then the output. When
|
|
|
|
* they put something the input queue, we make sure we're listening to
|
|
|
|
* stdin. When they put something in the output queue, we write it to
|
2007-10-24 23:02:50 -06:00
|
|
|
* stdout. */
|
2007-10-21 19:24:23 -06:00
|
|
|
add_virtqueue(dev, VIRTQUEUE_NUM, enable_fd);
|
2007-10-21 19:24:22 -06:00
|
|
|
add_virtqueue(dev, VIRTQUEUE_NUM, handle_console_output);
|
|
|
|
|
|
|
|
verbose("device %u: console\n", devices.device_num++);
|
2007-07-19 02:49:29 -06:00
|
|
|
}
|
2007-10-21 19:24:22 -06:00
|
|
|
/*:*/
|
2007-07-19 02:49:29 -06:00
|
|
|
|
2007-10-21 19:24:22 -06:00
|
|
|
/*M:010 Inter-guest networking is an interesting area. Simplest is to have a
|
|
|
|
* --sharenet=<name> option which opens or creates a named pipe. This can be
|
|
|
|
* used to send packets to another guest in a 1:1 manner.
|
2007-07-26 11:41:03 -06:00
|
|
|
*
|
2007-10-21 19:24:22 -06:00
|
|
|
* More sopisticated is to use one of the tools developed for project like UML
|
|
|
|
* to do networking.
|
2007-07-26 11:41:03 -06:00
|
|
|
*
|
2007-10-21 19:24:22 -06:00
|
|
|
* Faster is to do virtio bonding in kernel. Doing this 1:1 would be
|
|
|
|
* completely generic ("here's my vring, attach to your vring") and would work
|
|
|
|
* for any traffic. Of course, namespace and permissions issues need to be
|
|
|
|
* dealt with. A more sophisticated "multi-channel" virtio_net.c could hide
|
|
|
|
* multiple inter-guest channels behind one interface, although it would
|
|
|
|
* require some manner of hotplugging new virtio channels.
|
|
|
|
*
|
|
|
|
* Finally, we could implement a virtio network switch in the kernel. :*/
|
2007-07-19 02:49:29 -06:00
|
|
|
|
|
|
|
static u32 str2ip(const char *ipaddr)
|
|
|
|
{
|
|
|
|
unsigned int byte[4];
|
|
|
|
|
|
|
|
sscanf(ipaddr, "%u.%u.%u.%u", &byte[0], &byte[1], &byte[2], &byte[3]);
|
|
|
|
return (byte[0] << 24) | (byte[1] << 16) | (byte[2] << 8) | byte[3];
|
|
|
|
}
|
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/* This code is "adapted" from libbridge: it attaches the Host end of the
|
|
|
|
* network device to the bridge device specified by the command line.
|
|
|
|
*
|
|
|
|
* This is yet another James Morris contribution (I'm an IP-level guy, so I
|
|
|
|
* dislike bridging), and I just try not to break it. */
|
2007-07-19 02:49:29 -06:00
|
|
|
static void add_to_bridge(int fd, const char *if_name, const char *br_name)
|
|
|
|
{
|
|
|
|
int ifidx;
|
|
|
|
struct ifreq ifr;
|
|
|
|
|
|
|
|
if (!*br_name)
|
|
|
|
errx(1, "must specify bridge name");
|
|
|
|
|
|
|
|
ifidx = if_nametoindex(if_name);
|
|
|
|
if (!ifidx)
|
|
|
|
errx(1, "interface %s does not exist!", if_name);
|
|
|
|
|
|
|
|
strncpy(ifr.ifr_name, br_name, IFNAMSIZ);
|
|
|
|
ifr.ifr_ifindex = ifidx;
|
|
|
|
if (ioctl(fd, SIOCBRADDIF, &ifr) < 0)
|
|
|
|
err(1, "can't add %s to bridge %s", if_name, br_name);
|
|
|
|
}
|
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/* This sets up the Host end of the network device with an IP address, brings
|
|
|
|
* it up so packets will flow, the copies the MAC address into the hwaddr
|
2007-10-21 19:24:22 -06:00
|
|
|
* pointer. */
|
2007-07-19 02:49:29 -06:00
|
|
|
static void configure_device(int fd, const char *devname, u32 ipaddr,
|
|
|
|
unsigned char hwaddr[6])
|
|
|
|
{
|
|
|
|
struct ifreq ifr;
|
|
|
|
struct sockaddr_in *sin = (struct sockaddr_in *)&ifr.ifr_addr;
|
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/* Don't read these incantations. Just cut & paste them like I did! */
|
2007-07-19 02:49:29 -06:00
|
|
|
memset(&ifr, 0, sizeof(ifr));
|
|
|
|
strcpy(ifr.ifr_name, devname);
|
|
|
|
sin->sin_family = AF_INET;
|
|
|
|
sin->sin_addr.s_addr = htonl(ipaddr);
|
|
|
|
if (ioctl(fd, SIOCSIFADDR, &ifr) != 0)
|
|
|
|
err(1, "Setting %s interface address", devname);
|
|
|
|
ifr.ifr_flags = IFF_UP;
|
|
|
|
if (ioctl(fd, SIOCSIFFLAGS, &ifr) != 0)
|
|
|
|
err(1, "Bringing interface %s up", devname);
|
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/* SIOC stands for Socket I/O Control. G means Get (vs S for Set
|
|
|
|
* above). IF means Interface, and HWADDR is hardware address.
|
|
|
|
* Simple! */
|
2007-07-19 02:49:29 -06:00
|
|
|
if (ioctl(fd, SIOCGIFHWADDR, &ifr) != 0)
|
|
|
|
err(1, "getting hw address for %s", devname);
|
|
|
|
memcpy(hwaddr, ifr.ifr_hwaddr.sa_data, 6);
|
|
|
|
}
|
|
|
|
|
2007-10-21 19:24:22 -06:00
|
|
|
/*L:195 Our network is a Host<->Guest network. This can either use bridging or
|
|
|
|
* routing, but the principle is the same: it uses the "tun" device to inject
|
|
|
|
* packets into the Host as if they came in from a normal network card. We
|
|
|
|
* just shunt packets between the Guest and the tun device. */
|
|
|
|
static void setup_tun_net(const char *arg)
|
2007-07-19 02:49:29 -06:00
|
|
|
{
|
|
|
|
struct device *dev;
|
|
|
|
struct ifreq ifr;
|
|
|
|
int netfd, ipfd;
|
|
|
|
u32 ip;
|
|
|
|
const char *br_name = NULL;
|
2007-10-21 19:24:22 -06:00
|
|
|
u8 hwaddr[6];
|
2007-07-19 02:49:29 -06:00
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/* We open the /dev/net/tun device and tell it we want a tap device. A
|
|
|
|
* tap device is like a tun device, only somehow different. To tell
|
|
|
|
* the truth, I completely blundered my way through this code, but it
|
|
|
|
* works now! */
|
2007-07-19 02:49:29 -06:00
|
|
|
netfd = open_or_die("/dev/net/tun", O_RDWR);
|
|
|
|
memset(&ifr, 0, sizeof(ifr));
|
|
|
|
ifr.ifr_flags = IFF_TAP | IFF_NO_PI;
|
|
|
|
strcpy(ifr.ifr_name, "tap%d");
|
|
|
|
if (ioctl(netfd, TUNSETIFF, &ifr) != 0)
|
|
|
|
err(1, "configuring /dev/net/tun");
|
2007-07-26 11:41:03 -06:00
|
|
|
/* We don't need checksums calculated for packets coming in this
|
|
|
|
* device: trust us! */
|
2007-07-19 02:49:29 -06:00
|
|
|
ioctl(netfd, TUNSETNOCSUM, 1);
|
|
|
|
|
2007-10-21 19:24:22 -06:00
|
|
|
/* First we create a new network device. */
|
|
|
|
dev = new_device("net", VIRTIO_ID_NET, netfd, handle_tun_input);
|
2007-07-26 11:41:03 -06:00
|
|
|
|
2007-10-21 19:24:23 -06:00
|
|
|
/* Network devices need a receive and a send queue, just like
|
|
|
|
* console. */
|
|
|
|
add_virtqueue(dev, VIRTQUEUE_NUM, enable_fd);
|
2007-10-21 19:24:22 -06:00
|
|
|
add_virtqueue(dev, VIRTQUEUE_NUM, handle_net_output);
|
2007-07-19 02:49:29 -06:00
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/* We need a socket to perform the magic network ioctls to bring up the
|
|
|
|
* tap interface, connect to the bridge etc. Any socket will do! */
|
2007-07-19 02:49:29 -06:00
|
|
|
ipfd = socket(PF_INET, SOCK_DGRAM, IPPROTO_IP);
|
|
|
|
if (ipfd < 0)
|
|
|
|
err(1, "opening IP socket");
|
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/* If the command line was --tunnet=bridge:<name> do bridging. */
|
2007-07-19 02:49:29 -06:00
|
|
|
if (!strncmp(BRIDGE_PFX, arg, strlen(BRIDGE_PFX))) {
|
|
|
|
ip = INADDR_ANY;
|
|
|
|
br_name = arg + strlen(BRIDGE_PFX);
|
|
|
|
add_to_bridge(ipfd, ifr.ifr_name, br_name);
|
2007-07-26 11:41:03 -06:00
|
|
|
} else /* It is an IP address to set up the device with */
|
2007-07-19 02:49:29 -06:00
|
|
|
ip = str2ip(arg);
|
|
|
|
|
2007-10-21 19:24:22 -06:00
|
|
|
/* Set up the tun device, and get the mac address for the interface. */
|
|
|
|
configure_device(ipfd, ifr.ifr_name, ip, hwaddr);
|
2007-07-19 02:49:29 -06:00
|
|
|
|
2007-10-21 19:24:22 -06:00
|
|
|
/* Tell Guest what MAC address to use. */
|
|
|
|
add_desc_field(dev, VIRTIO_CONFIG_NET_MAC_F, sizeof(hwaddr), hwaddr);
|
2007-07-19 02:49:29 -06:00
|
|
|
|
2007-10-21 19:24:22 -06:00
|
|
|
/* We don't seed the socket any more; setup is done. */
|
2007-07-19 02:49:29 -06:00
|
|
|
close(ipfd);
|
|
|
|
|
2007-10-21 19:24:22 -06:00
|
|
|
verbose("device %u: tun net %u.%u.%u.%u\n",
|
|
|
|
devices.device_num++,
|
|
|
|
(u8)(ip>>24),(u8)(ip>>16),(u8)(ip>>8),(u8)ip);
|
2007-07-19 02:49:29 -06:00
|
|
|
if (br_name)
|
|
|
|
verbose("attached to bridge: %s\n", br_name);
|
|
|
|
}
|
2007-10-21 19:24:22 -06:00
|
|
|
|
2007-10-24 23:02:50 -06:00
|
|
|
/* Our block (disk) device should be really simple: the Guest asks for a block
|
|
|
|
* number and we read or write that position in the file. Unfortunately, that
|
|
|
|
* was amazingly slow: the Guest waits until the read is finished before
|
|
|
|
* running anything else, even if it could have been doing useful work.
|
2007-10-21 19:24:22 -06:00
|
|
|
*
|
2007-10-24 23:02:50 -06:00
|
|
|
* We could use async I/O, except it's reputed to suck so hard that characters
|
|
|
|
* actually go missing from your code when you try to use it.
|
2007-10-21 19:24:22 -06:00
|
|
|
*
|
|
|
|
* So we farm the I/O out to thread, and communicate with it via a pipe. */
|
|
|
|
|
2007-10-24 23:02:50 -06:00
|
|
|
/* This hangs off device->priv. */
|
2007-10-21 19:24:22 -06:00
|
|
|
struct vblk_info
|
|
|
|
{
|
|
|
|
/* The size of the file. */
|
|
|
|
off64_t len;
|
|
|
|
|
|
|
|
/* The file descriptor for the file. */
|
|
|
|
int fd;
|
|
|
|
|
|
|
|
/* IO thread listens on this file descriptor [0]. */
|
|
|
|
int workpipe[2];
|
|
|
|
|
|
|
|
/* IO thread writes to this file descriptor to mark it done, then
|
|
|
|
* Launcher triggers interrupt to Guest. */
|
|
|
|
int done_fd;
|
|
|
|
};
|
2007-10-24 23:02:50 -06:00
|
|
|
/*:*/
|
2007-10-21 19:24:22 -06:00
|
|
|
|
2007-10-24 23:02:50 -06:00
|
|
|
/*L:210
|
|
|
|
* The Disk
|
|
|
|
*
|
|
|
|
* Remember that the block device is handled by a separate I/O thread. We head
|
|
|
|
* straight into the core of that thread here:
|
|
|
|
*/
|
2007-10-21 19:24:22 -06:00
|
|
|
static bool service_io(struct device *dev)
|
|
|
|
{
|
|
|
|
struct vblk_info *vblk = dev->priv;
|
|
|
|
unsigned int head, out_num, in_num, wlen;
|
|
|
|
int ret;
|
|
|
|
struct virtio_blk_inhdr *in;
|
|
|
|
struct virtio_blk_outhdr *out;
|
|
|
|
struct iovec iov[dev->vq->vring.num];
|
|
|
|
off64_t off;
|
|
|
|
|
2007-10-24 23:02:50 -06:00
|
|
|
/* See if there's a request waiting. If not, nothing to do. */
|
2007-10-21 19:24:22 -06:00
|
|
|
head = get_vq_desc(dev->vq, iov, &out_num, &in_num);
|
|
|
|
if (head == dev->vq->vring.num)
|
|
|
|
return false;
|
|
|
|
|
2007-10-24 23:02:50 -06:00
|
|
|
/* Every block request should contain at least one output buffer
|
|
|
|
* (detailing the location on disk and the type of request) and one
|
|
|
|
* input buffer (to hold the result). */
|
2007-10-21 19:24:22 -06:00
|
|
|
if (out_num == 0 || in_num == 0)
|
|
|
|
errx(1, "Bad virtblk cmd %u out=%u in=%u",
|
|
|
|
head, out_num, in_num);
|
|
|
|
|
|
|
|
out = convert(&iov[0], struct virtio_blk_outhdr);
|
|
|
|
in = convert(&iov[out_num+in_num-1], struct virtio_blk_inhdr);
|
|
|
|
off = out->sector * 512;
|
|
|
|
|
2007-10-24 23:02:50 -06:00
|
|
|
/* The block device implements "barriers", where the Guest indicates
|
|
|
|
* that it wants all previous writes to occur before this write. We
|
|
|
|
* don't have a way of asking our kernel to do a barrier, so we just
|
|
|
|
* synchronize all the data in the file. Pretty poor, no? */
|
2007-10-21 19:24:22 -06:00
|
|
|
if (out->type & VIRTIO_BLK_T_BARRIER)
|
|
|
|
fdatasync(vblk->fd);
|
|
|
|
|
2007-10-24 23:02:50 -06:00
|
|
|
/* In general the virtio block driver is allowed to try SCSI commands.
|
|
|
|
* It'd be nice if we supported eject, for example, but we don't. */
|
2007-10-21 19:24:22 -06:00
|
|
|
if (out->type & VIRTIO_BLK_T_SCSI_CMD) {
|
|
|
|
fprintf(stderr, "Scsi commands unsupported\n");
|
|
|
|
in->status = VIRTIO_BLK_S_UNSUPP;
|
2007-11-08 20:13:44 -07:00
|
|
|
wlen = sizeof(*in);
|
2007-10-21 19:24:22 -06:00
|
|
|
} else if (out->type & VIRTIO_BLK_T_OUT) {
|
|
|
|
/* Write */
|
|
|
|
|
|
|
|
/* Move to the right location in the block file. This can fail
|
|
|
|
* if they try to write past end. */
|
|
|
|
if (lseek64(vblk->fd, off, SEEK_SET) != off)
|
|
|
|
err(1, "Bad seek to sector %llu", out->sector);
|
|
|
|
|
|
|
|
ret = writev(vblk->fd, iov+1, out_num-1);
|
|
|
|
verbose("WRITE to sector %llu: %i\n", out->sector, ret);
|
|
|
|
|
|
|
|
/* Grr... Now we know how long the descriptor they sent was, we
|
|
|
|
* make sure they didn't try to write over the end of the block
|
|
|
|
* file (possibly extending it). */
|
|
|
|
if (ret > 0 && off + ret > vblk->len) {
|
|
|
|
/* Trim it back to the correct length */
|
|
|
|
ftruncate64(vblk->fd, vblk->len);
|
|
|
|
/* Die, bad Guest, die. */
|
|
|
|
errx(1, "Write past end %llu+%u", off, ret);
|
|
|
|
}
|
2007-11-08 20:13:44 -07:00
|
|
|
wlen = sizeof(*in);
|
2007-10-21 19:24:22 -06:00
|
|
|
in->status = (ret >= 0 ? VIRTIO_BLK_S_OK : VIRTIO_BLK_S_IOERR);
|
|
|
|
} else {
|
|
|
|
/* Read */
|
|
|
|
|
|
|
|
/* Move to the right location in the block file. This can fail
|
|
|
|
* if they try to read past end. */
|
|
|
|
if (lseek64(vblk->fd, off, SEEK_SET) != off)
|
|
|
|
err(1, "Bad seek to sector %llu", out->sector);
|
|
|
|
|
|
|
|
ret = readv(vblk->fd, iov+1, in_num-1);
|
|
|
|
verbose("READ from sector %llu: %i\n", out->sector, ret);
|
|
|
|
if (ret >= 0) {
|
2007-11-08 20:13:44 -07:00
|
|
|
wlen = sizeof(*in) + ret;
|
2007-10-21 19:24:22 -06:00
|
|
|
in->status = VIRTIO_BLK_S_OK;
|
|
|
|
} else {
|
2007-11-08 20:13:44 -07:00
|
|
|
wlen = sizeof(*in);
|
2007-10-21 19:24:22 -06:00
|
|
|
in->status = VIRTIO_BLK_S_IOERR;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/* We can't trigger an IRQ, because we're not the Launcher. It does
|
|
|
|
* that when we tell it we're done. */
|
|
|
|
add_used(dev->vq, head, wlen);
|
|
|
|
return true;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* This is the thread which actually services the I/O. */
|
|
|
|
static int io_thread(void *_dev)
|
|
|
|
{
|
|
|
|
struct device *dev = _dev;
|
|
|
|
struct vblk_info *vblk = dev->priv;
|
|
|
|
char c;
|
|
|
|
|
|
|
|
/* Close other side of workpipe so we get 0 read when main dies. */
|
|
|
|
close(vblk->workpipe[1]);
|
|
|
|
/* Close the other side of the done_fd pipe. */
|
|
|
|
close(dev->fd);
|
|
|
|
|
|
|
|
/* When this read fails, it means Launcher died, so we follow. */
|
|
|
|
while (read(vblk->workpipe[0], &c, 1) == 1) {
|
2007-10-24 23:02:50 -06:00
|
|
|
/* We acknowledge each request immediately to reduce latency,
|
2007-10-21 19:24:22 -06:00
|
|
|
* rather than waiting until we've done them all. I haven't
|
|
|
|
* measured to see if it makes any difference. */
|
|
|
|
while (service_io(dev))
|
|
|
|
write(vblk->done_fd, &c, 1);
|
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2007-10-24 23:02:50 -06:00
|
|
|
/* Now we've seen the I/O thread, we return to the Launcher to see what happens
|
|
|
|
* when the thread tells us it's completed some I/O. */
|
2007-10-21 19:24:22 -06:00
|
|
|
static bool handle_io_finish(int fd, struct device *dev)
|
|
|
|
{
|
|
|
|
char c;
|
|
|
|
|
2007-10-24 23:02:50 -06:00
|
|
|
/* If the I/O thread died, presumably it printed the error, so we
|
|
|
|
* simply exit. */
|
2007-10-21 19:24:22 -06:00
|
|
|
if (read(dev->fd, &c, 1) != 1)
|
|
|
|
exit(1);
|
|
|
|
|
|
|
|
/* It did some work, so trigger the irq. */
|
|
|
|
trigger_irq(fd, dev->vq);
|
|
|
|
return true;
|
|
|
|
}
|
|
|
|
|
2007-10-24 23:02:50 -06:00
|
|
|
/* When the Guest submits some I/O, we just need to wake the I/O thread. */
|
2007-10-21 19:24:22 -06:00
|
|
|
static void handle_virtblk_output(int fd, struct virtqueue *vq)
|
|
|
|
{
|
|
|
|
struct vblk_info *vblk = vq->dev->priv;
|
|
|
|
char c = 0;
|
|
|
|
|
|
|
|
/* Wake up I/O thread and tell it to go to work! */
|
|
|
|
if (write(vblk->workpipe[1], &c, 1) != 1)
|
|
|
|
/* Presumably it indicated why it died. */
|
|
|
|
exit(1);
|
|
|
|
}
|
|
|
|
|
2007-10-24 23:02:50 -06:00
|
|
|
/*L:198 This actually sets up a virtual block device. */
|
2007-10-21 19:24:22 -06:00
|
|
|
static void setup_block_file(const char *filename)
|
|
|
|
{
|
|
|
|
int p[2];
|
|
|
|
struct device *dev;
|
|
|
|
struct vblk_info *vblk;
|
|
|
|
void *stack;
|
|
|
|
u64 cap;
|
|
|
|
unsigned int val;
|
|
|
|
|
|
|
|
/* This is the pipe the I/O thread will use to tell us I/O is done. */
|
|
|
|
pipe(p);
|
|
|
|
|
|
|
|
/* The device responds to return from I/O thread. */
|
|
|
|
dev = new_device("block", VIRTIO_ID_BLOCK, p[0], handle_io_finish);
|
|
|
|
|
2007-10-24 23:02:50 -06:00
|
|
|
/* The device has one virtqueue, where the Guest places requests. */
|
2007-10-21 19:24:22 -06:00
|
|
|
add_virtqueue(dev, VIRTQUEUE_NUM, handle_virtblk_output);
|
|
|
|
|
|
|
|
/* Allocate the room for our own bookkeeping */
|
|
|
|
vblk = dev->priv = malloc(sizeof(*vblk));
|
|
|
|
|
|
|
|
/* First we open the file and store the length. */
|
|
|
|
vblk->fd = open_or_die(filename, O_RDWR|O_LARGEFILE);
|
|
|
|
vblk->len = lseek64(vblk->fd, 0, SEEK_END);
|
|
|
|
|
|
|
|
/* Tell Guest how many sectors this device has. */
|
|
|
|
cap = cpu_to_le64(vblk->len / 512);
|
|
|
|
add_desc_field(dev, VIRTIO_CONFIG_BLK_F_CAPACITY, sizeof(cap), &cap);
|
|
|
|
|
|
|
|
/* Tell Guest not to put in too many descriptors at once: two are used
|
|
|
|
* for the in and out elements. */
|
|
|
|
val = cpu_to_le32(VIRTQUEUE_NUM - 2);
|
|
|
|
add_desc_field(dev, VIRTIO_CONFIG_BLK_F_SEG_MAX, sizeof(val), &val);
|
|
|
|
|
|
|
|
/* The I/O thread writes to this end of the pipe when done. */
|
|
|
|
vblk->done_fd = p[1];
|
|
|
|
|
2007-10-24 23:02:50 -06:00
|
|
|
/* This is the second pipe, which is how we tell the I/O thread about
|
|
|
|
* more work. */
|
2007-10-21 19:24:22 -06:00
|
|
|
pipe(vblk->workpipe);
|
|
|
|
|
|
|
|
/* Create stack for thread and run it */
|
|
|
|
stack = malloc(32768);
|
|
|
|
if (clone(io_thread, stack + 32768, CLONE_VM, dev) == -1)
|
|
|
|
err(1, "Creating clone");
|
|
|
|
|
|
|
|
/* We don't need to keep the I/O thread's end of the pipes open. */
|
|
|
|
close(vblk->done_fd);
|
|
|
|
close(vblk->workpipe[0]);
|
|
|
|
|
|
|
|
verbose("device %u: virtblock %llu sectors\n",
|
|
|
|
devices.device_num, cap);
|
|
|
|
}
|
2007-07-26 11:41:03 -06:00
|
|
|
/* That's the end of device setup. */
|
2007-07-19 02:49:29 -06:00
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/*L:220 Finally we reach the core of the Launcher, which runs the Guest, serves
|
|
|
|
* its input and output, and finally, lays it to rest. */
|
2007-10-21 19:24:22 -06:00
|
|
|
static void __attribute__((noreturn)) run_guest(int lguest_fd)
|
2007-07-19 02:49:29 -06:00
|
|
|
{
|
|
|
|
for (;;) {
|
2007-10-21 19:03:31 -06:00
|
|
|
unsigned long args[] = { LHREQ_BREAK, 0 };
|
2007-10-21 19:24:22 -06:00
|
|
|
unsigned long notify_addr;
|
2007-07-19 02:49:29 -06:00
|
|
|
int readval;
|
|
|
|
|
|
|
|
/* We read from the /dev/lguest device to run the Guest. */
|
2007-10-21 19:24:22 -06:00
|
|
|
readval = read(lguest_fd, ¬ify_addr, sizeof(notify_addr));
|
2007-07-19 02:49:29 -06:00
|
|
|
|
2007-10-21 19:24:22 -06:00
|
|
|
/* One unsigned long means the Guest did HCALL_NOTIFY */
|
|
|
|
if (readval == sizeof(notify_addr)) {
|
|
|
|
verbose("Notify on address %#lx\n", notify_addr);
|
|
|
|
handle_output(lguest_fd, notify_addr);
|
2007-07-19 02:49:29 -06:00
|
|
|
continue;
|
2007-07-26 11:41:03 -06:00
|
|
|
/* ENOENT means the Guest died. Reading tells us why. */
|
2007-07-19 02:49:29 -06:00
|
|
|
} else if (errno == ENOENT) {
|
|
|
|
char reason[1024] = { 0 };
|
|
|
|
read(lguest_fd, reason, sizeof(reason)-1);
|
|
|
|
errx(1, "%s", reason);
|
2007-10-24 23:02:50 -06:00
|
|
|
/* EAGAIN means the Waker wanted us to look at some input.
|
2007-07-26 11:41:03 -06:00
|
|
|
* Anything else means a bug or incompatible change. */
|
2007-07-19 02:49:29 -06:00
|
|
|
} else if (errno != EAGAIN)
|
|
|
|
err(1, "Running guest failed");
|
2007-07-26 11:41:03 -06:00
|
|
|
|
2007-10-24 23:02:50 -06:00
|
|
|
/* Service input, then unset the BREAK to release the Waker. */
|
2007-10-21 19:24:22 -06:00
|
|
|
handle_input(lguest_fd);
|
2007-07-19 02:49:29 -06:00
|
|
|
if (write(lguest_fd, args, sizeof(args)) < 0)
|
|
|
|
err(1, "Resetting break");
|
|
|
|
}
|
|
|
|
}
|
2007-07-26 11:41:03 -06:00
|
|
|
/*
|
2007-10-24 23:02:50 -06:00
|
|
|
* This is the end of the Launcher. The good news: we are over halfway
|
|
|
|
* through! The bad news: the most fiendish part of the code still lies ahead
|
|
|
|
* of us.
|
2007-07-26 11:41:03 -06:00
|
|
|
*
|
2007-10-24 23:02:50 -06:00
|
|
|
* Are you ready? Take a deep breath and join me in the core of the Host, in
|
|
|
|
* "make Host".
|
|
|
|
:*/
|
2007-07-19 02:49:29 -06:00
|
|
|
|
|
|
|
static struct option opts[] = {
|
|
|
|
{ "verbose", 0, NULL, 'v' },
|
|
|
|
{ "tunnet", 1, NULL, 't' },
|
|
|
|
{ "block", 1, NULL, 'b' },
|
|
|
|
{ "initrd", 1, NULL, 'i' },
|
|
|
|
{ NULL },
|
|
|
|
};
|
|
|
|
static void usage(void)
|
|
|
|
{
|
|
|
|
errx(1, "Usage: lguest [--verbose] "
|
2007-10-21 19:24:22 -06:00
|
|
|
"[--tunnet=(<ipaddr>|bridge:<bridgename>)\n"
|
2007-07-19 02:49:29 -06:00
|
|
|
"|--block=<filename>|--initrd=<filename>]...\n"
|
|
|
|
"<mem-in-mb> vmlinux [args...]");
|
|
|
|
}
|
|
|
|
|
2007-10-21 19:03:26 -06:00
|
|
|
/*L:105 The main routine is where the real work begins: */
|
2007-07-19 02:49:29 -06:00
|
|
|
int main(int argc, char *argv[])
|
|
|
|
{
|
2007-10-21 19:03:36 -06:00
|
|
|
/* Memory, top-level pagetable, code startpoint and size of the
|
|
|
|
* (optional) initrd. */
|
|
|
|
unsigned long mem = 0, pgdir, start, initrd_size = 0;
|
2007-10-24 23:02:50 -06:00
|
|
|
/* Two temporaries and the /dev/lguest file descriptor. */
|
2007-07-23 19:43:56 -06:00
|
|
|
int i, c, lguest_fd;
|
2007-10-21 19:03:26 -06:00
|
|
|
/* The boot information for the Guest. */
|
2007-10-21 19:29:57 -06:00
|
|
|
struct boot_params *boot;
|
2007-07-26 11:41:03 -06:00
|
|
|
/* If they specify an initrd file to load. */
|
2007-07-19 02:49:29 -06:00
|
|
|
const char *initrd_name = NULL;
|
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/* First we initialize the device list. Since console and network
|
|
|
|
* device receive input from a file descriptor, we keep an fdset
|
|
|
|
* (infds) and the maximum fd number (max_infd) with the head of the
|
|
|
|
* list. We also keep a pointer to the last device, for easy appending
|
2007-10-21 19:24:22 -06:00
|
|
|
* to the list. Finally, we keep the next interrupt number to hand out
|
|
|
|
* (1: remember that 0 is used by the timer). */
|
|
|
|
FD_ZERO(&devices.infds);
|
|
|
|
devices.max_infd = -1;
|
|
|
|
devices.lastdev = &devices.dev;
|
|
|
|
devices.next_irq = 1;
|
2007-07-19 02:49:29 -06:00
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/* We need to know how much memory so we can set up the device
|
|
|
|
* descriptor and memory pages for the devices as we parse the command
|
|
|
|
* line. So we quickly look through the arguments to find the amount
|
|
|
|
* of memory now. */
|
2007-07-23 19:43:56 -06:00
|
|
|
for (i = 1; i < argc; i++) {
|
|
|
|
if (argv[i][0] != '-') {
|
2007-10-21 19:03:26 -06:00
|
|
|
mem = atoi(argv[i]) * 1024 * 1024;
|
|
|
|
/* We start by mapping anonymous pages over all of
|
|
|
|
* guest-physical memory range. This fills it with 0,
|
|
|
|
* and ensures that the Guest won't be killed when it
|
|
|
|
* tries to access it. */
|
|
|
|
guest_base = map_zeroed_pages(mem / getpagesize()
|
|
|
|
+ DEVICE_PAGES);
|
|
|
|
guest_limit = mem;
|
|
|
|
guest_max = mem + DEVICE_PAGES*getpagesize();
|
2007-10-21 19:24:22 -06:00
|
|
|
devices.descpage = get_pages(1);
|
2007-07-23 19:43:56 -06:00
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
2007-07-26 11:41:03 -06:00
|
|
|
|
|
|
|
/* The options are fairly straight-forward */
|
2007-07-19 02:49:29 -06:00
|
|
|
while ((c = getopt_long(argc, argv, "v", opts, NULL)) != EOF) {
|
|
|
|
switch (c) {
|
|
|
|
case 'v':
|
|
|
|
verbose = true;
|
|
|
|
break;
|
|
|
|
case 't':
|
2007-10-21 19:24:22 -06:00
|
|
|
setup_tun_net(optarg);
|
2007-07-19 02:49:29 -06:00
|
|
|
break;
|
|
|
|
case 'b':
|
2007-10-21 19:24:22 -06:00
|
|
|
setup_block_file(optarg);
|
2007-07-19 02:49:29 -06:00
|
|
|
break;
|
|
|
|
case 'i':
|
|
|
|
initrd_name = optarg;
|
|
|
|
break;
|
|
|
|
default:
|
|
|
|
warnx("Unknown argument %s", argv[optind]);
|
|
|
|
usage();
|
|
|
|
}
|
|
|
|
}
|
2007-07-26 11:41:03 -06:00
|
|
|
/* After the other arguments we expect memory and kernel image name,
|
|
|
|
* followed by command line arguments for the kernel. */
|
2007-07-19 02:49:29 -06:00
|
|
|
if (optind + 2 > argc)
|
|
|
|
usage();
|
|
|
|
|
2007-10-21 19:03:26 -06:00
|
|
|
verbose("Guest base is at %p\n", guest_base);
|
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/* We always have a console device */
|
2007-10-21 19:24:22 -06:00
|
|
|
setup_console();
|
2007-07-19 02:49:29 -06:00
|
|
|
|
|
|
|
/* Now we load the kernel */
|
2007-10-21 19:03:36 -06:00
|
|
|
start = load_kernel(open_or_die(argv[optind+1], O_RDONLY));
|
2007-07-19 02:49:29 -06:00
|
|
|
|
2007-10-21 19:03:26 -06:00
|
|
|
/* Boot information is stashed at physical address 0 */
|
|
|
|
boot = from_guest_phys(0);
|
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/* Map the initrd image if requested (at top of physical memory) */
|
2007-07-19 02:49:29 -06:00
|
|
|
if (initrd_name) {
|
|
|
|
initrd_size = load_initrd(initrd_name, mem);
|
2007-07-26 11:41:03 -06:00
|
|
|
/* These are the location in the Linux boot header where the
|
|
|
|
* start and size of the initrd are expected to be found. */
|
2007-10-21 19:29:57 -06:00
|
|
|
boot->hdr.ramdisk_image = mem - initrd_size;
|
|
|
|
boot->hdr.ramdisk_size = initrd_size;
|
2007-07-26 11:41:03 -06:00
|
|
|
/* The bootloader type 0xFF means "unknown"; that's OK. */
|
2007-10-21 19:29:57 -06:00
|
|
|
boot->hdr.type_of_loader = 0xFF;
|
2007-07-19 02:49:29 -06:00
|
|
|
}
|
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/* Set up the initial linear pagetables, starting below the initrd. */
|
2007-10-21 19:03:36 -06:00
|
|
|
pgdir = setup_pagetables(mem, initrd_size);
|
2007-07-19 02:49:29 -06:00
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/* The Linux boot header contains an "E820" memory map: ours is a
|
|
|
|
* simple, single region. */
|
2007-10-21 19:29:57 -06:00
|
|
|
boot->e820_entries = 1;
|
|
|
|
boot->e820_map[0] = ((struct e820entry) { 0, mem, E820_RAM });
|
2007-07-26 11:41:03 -06:00
|
|
|
/* The boot header contains a command line pointer: we put the command
|
2007-10-21 19:29:57 -06:00
|
|
|
* line after the boot header. */
|
|
|
|
boot->hdr.cmd_line_ptr = to_guest_phys(boot + 1);
|
2007-10-24 23:02:50 -06:00
|
|
|
/* We use a simple helper to copy the arguments separated by spaces. */
|
2007-10-21 19:29:57 -06:00
|
|
|
concat((char *)(boot + 1), argv+optind+2);
|
2007-07-26 11:41:03 -06:00
|
|
|
|
2007-10-21 19:29:44 -06:00
|
|
|
/* Boot protocol version: 2.07 supports the fields for lguest. */
|
2007-10-21 19:29:57 -06:00
|
|
|
boot->hdr.version = 0x207;
|
2007-10-21 19:29:44 -06:00
|
|
|
|
|
|
|
/* The hardware_subarch value of "1" tells the Guest it's an lguest. */
|
2007-10-21 19:29:57 -06:00
|
|
|
boot->hdr.hardware_subarch = 1;
|
2007-10-21 19:29:44 -06:00
|
|
|
|
2007-10-21 19:29:57 -06:00
|
|
|
/* Tell the entry path not to try to reload segment registers. */
|
|
|
|
boot->hdr.loadflags |= KEEP_SEGMENTS;
|
2007-07-19 02:49:29 -06:00
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/* We tell the kernel to initialize the Guest: this returns the open
|
|
|
|
* /dev/lguest file descriptor. */
|
2007-10-21 19:03:36 -06:00
|
|
|
lguest_fd = tell_kernel(pgdir, start);
|
2007-07-26 11:41:03 -06:00
|
|
|
|
|
|
|
/* We fork off a child process, which wakes the Launcher whenever one
|
|
|
|
* of the input file descriptors needs attention. Otherwise we would
|
|
|
|
* run the Guest until it tries to output something. */
|
2007-10-21 19:24:22 -06:00
|
|
|
waker_fd = setup_waker(lguest_fd);
|
2007-07-19 02:49:29 -06:00
|
|
|
|
2007-07-26 11:41:03 -06:00
|
|
|
/* Finally, run the Guest. This doesn't return. */
|
2007-10-21 19:24:22 -06:00
|
|
|
run_guest(lguest_fd);
|
2007-07-19 02:49:29 -06:00
|
|
|
}
|
2007-07-26 11:41:05 -06:00
|
|
|
/*:*/
|
|
|
|
|
|
|
|
/*M:999
|
|
|
|
* Mastery is done: you now know everything I do.
|
|
|
|
*
|
|
|
|
* But surely you have seen code, features and bugs in your wanderings which
|
|
|
|
* you now yearn to attack? That is the real game, and I look forward to you
|
|
|
|
* patching and forking lguest into the Your-Name-Here-visor.
|
|
|
|
*
|
|
|
|
* Farewell, and good coding!
|
|
|
|
* Rusty Russell.
|
|
|
|
*/
|