Magisk/native/jni/core/magiskinit.c

490 lines
12 KiB
C
Raw Normal View History

/* magiskinit.c - Pre-init Magisk support
*
* This code has to be compiled statically to work properly.
*
2017-11-10 00:54:54 +08:00
* To unify Magisk support for both legacy "normal" devices and new skip_initramfs devices,
* magisk binary compilation is split into two parts - first part only compiles "magisk".
* The python build script will load the magisk main binary and compress with lzma2, dumping
* the results into "dump.h". The "magisk" binary is embedded into this binary, and will
* get extracted to the overlay folder along with init.magisk.rc.
*
* This tool does all pre-init operations to setup a Magisk environment, which pathces rootfs
* on the fly, providing fundamental support such as init, init.rc, and sepolicy patching.
*
* Magiskinit is also responsible for constructing a proper rootfs on skip_initramfs devices.
* On skip_initramfs devices, it will parse kernel cmdline, mount sysfs, parse through
* uevent files to make the system (or vendor if available) block device node, then copy
2017-11-10 00:54:54 +08:00
* rootfs files from system.
*
* This tool will be replaced with the real init to continue the boot process, but hardlinks are
* preserved as it also provides CLI for sepolicy patching (magiskpolicy)
*/
2017-12-05 21:05:20 +08:00
#define _GNU_SOURCE
#include <stdio.h>
#include <stdlib.h>
#include <unistd.h>
#include <string.h>
#include <dirent.h>
#include <fcntl.h>
#include <libgen.h>
#include <sys/stat.h>
#include <sys/types.h>
#include <sys/mount.h>
#include <sys/mman.h>
#include <sys/sendfile.h>
#include <sys/sysmacros.h>
2017-11-10 00:54:54 +08:00
#include <lzma.h>
#include <cil/cil.h>
2017-11-28 03:42:48 +08:00
#include "dump.h"
#include "magiskrc.h"
#include "utils.h"
#include "magiskpolicy.h"
2017-11-28 03:42:48 +08:00
#include "daemon.h"
2017-12-06 12:51:16 +08:00
#include "cpio.h"
2017-11-28 03:42:48 +08:00
#include "magisk.h"
#ifdef MAGISK_DEBUG
#define VLOG(fmt, ...) printf(fmt, __VA_ARGS__)
#else
2017-11-06 22:33:41 +08:00
#define VLOG(fmt, ...)
#endif
2017-11-06 22:33:41 +08:00
extern policydb_t *policydb;
2017-11-28 04:43:46 +08:00
int (*init_applet_main[]) (int, char *[]) = { magiskpolicy_main, magiskpolicy_main, NULL };
static char RAND_SOCKET_NAME[sizeof(SOCKET_NAME)];
static int SOCKET_OFF = -1;
2017-11-28 04:43:46 +08:00
struct cmdline {
char skip_initramfs;
char slot[3];
};
struct device {
dev_t major;
dev_t minor;
char devname[32];
char partname[32];
char path[64];
};
static void parse_cmdline(struct cmdline *cmd) {
2017-11-23 18:38:12 +08:00
// cleanup
memset(cmd, 0, sizeof(&cmd));
2017-11-23 18:38:12 +08:00
char cmdline[4096];
mkdir("/proc", 0555);
xmount("proc", "/proc", "proc", 0, NULL);
int fd = open("/proc/cmdline", O_RDONLY | O_CLOEXEC);
2017-11-23 18:38:12 +08:00
cmdline[read(fd, cmdline, sizeof(cmdline))] = '\0';
close(fd);
umount("/proc");
2017-12-07 01:30:48 +08:00
for (char *tok = strtok(cmdline, " "); tok; tok = strtok(NULL, " ")) {
if (strncmp(tok, "androidboot.slot_suffix", 23) == 0) {
sscanf(tok, "androidboot.slot_suffix=%s", cmd->slot);
2017-11-06 22:33:41 +08:00
} else if (strncmp(tok, "androidboot.slot", 16) == 0) {
cmd->slot[0] = '_';
sscanf(tok, "androidboot.slot=%c", cmd->slot + 1);
} else if (strcmp(tok, "skip_initramfs") == 0) {
cmd->skip_initramfs = 1;
}
}
VLOG("cmdline: skip_initramfs[%d] slot[%s]\n", cmd->skip_initramfs, cmd->slot);
}
static void parse_device(struct device *dev, char *uevent) {
2017-11-06 22:33:41 +08:00
dev->partname[0] = '\0';
char *tok;
tok = strtok(uevent, "\n");
while (tok != NULL) {
if (strncmp(tok, "MAJOR", 5) == 0) {
sscanf(tok, "MAJOR=%ld", (long*) &dev->major);
} else if (strncmp(tok, "MINOR", 5) == 0) {
sscanf(tok, "MINOR=%ld", (long*) &dev->minor);
} else if (strncmp(tok, "DEVNAME", 7) == 0) {
sscanf(tok, "DEVNAME=%s", dev->devname);
} else if (strncmp(tok, "PARTNAME", 8) == 0) {
sscanf(tok, "PARTNAME=%s", dev->partname);
}
tok = strtok(NULL, "\n");
}
2017-11-06 22:33:41 +08:00
VLOG("%s [%s] (%u, %u)\n", dev->devname, dev->partname, (unsigned) dev->major, (unsigned) dev->minor);
}
static int setup_block(struct device *dev, const char *partname) {
char buffer[1024], path[128];
struct dirent *entry;
DIR *dir = opendir("/sys/dev/block");
if (dir == NULL)
return 1;
int found = 0;
while ((entry = readdir(dir))) {
if (strcmp(entry->d_name, ".") == 0 || strcmp(entry->d_name, "..") == 0)
continue;
snprintf(path, sizeof(path), "/sys/dev/block/%s/uevent", entry->d_name);
int fd = open(path, O_RDONLY | O_CLOEXEC);
ssize_t size = read(fd, buffer, sizeof(buffer));
buffer[size] = '\0';
close(fd);
parse_device(dev, buffer);
if (strcmp(dev->partname, partname) == 0) {
snprintf(dev->path, sizeof(dev->path), "/dev/block/%s", dev->devname);
found = 1;
break;
}
}
closedir(dir);
if (!found)
return 1;
mkdir("/dev", 0755);
mkdir("/dev/block", 0755);
mknod(dev->path, S_IFBLK | 0600, makedev(dev->major, dev->minor));
return 0;
}
static void patch_ramdisk() {
void *addr;
size_t size;
mmap_rw("/init", &addr, &size);
for (int i = 0; i < size; ++i) {
2017-12-01 03:33:25 +08:00
if (memcmp(addr + i, SPLIT_PLAT_CIL, sizeof(SPLIT_PLAT_CIL) - 1) == 0) {
memcpy(addr + i + sizeof(SPLIT_PLAT_CIL) - 4, "xxx", 3);
break;
}
}
munmap(addr, size);
2017-12-07 01:30:48 +08:00
full_read("/init.rc", &addr, &size);
patch_init_rc(&addr, &size);
int fd = creat("/init.rc", 0750);
write(fd, addr, size);
close(fd);
2017-12-07 01:30:48 +08:00
free(addr);
}
static int strend(const char *s1, const char *s2) {
2018-02-02 03:22:38 +08:00
size_t l1 = strlen(s1);
size_t l2 = strlen(s2);
return strcmp(s1 + l1 - l2, s2);
}
static int compile_cil() {
DIR *dir;
struct dirent *entry;
char path[128];
struct cil_db *db = NULL;
sepol_policydb_t *pdb = NULL;
void *addr;
size_t size;
cil_db_init(&db);
cil_set_mls(db, 1);
cil_set_multiple_decls(db, 1);
cil_set_disable_neverallow(db, 1);
cil_set_target_platform(db, SEPOL_TARGET_SELINUX);
cil_set_policy_version(db, POLICYDB_VERSION_XPERMS_IOCTL);
cil_set_attrs_expand_generated(db, 0);
// plat
2017-12-01 03:33:25 +08:00
mmap_ro(SPLIT_PLAT_CIL, &addr, &size);
VLOG("cil_add[%s]\n", SPLIT_PLAT_CIL);
cil_add_file(db, SPLIT_PLAT_CIL, addr, size);
munmap(addr, size);
// mapping
char plat[10];
2017-12-01 03:33:25 +08:00
int fd = open(SPLIT_NONPLAT_VER, O_RDONLY | O_CLOEXEC);
plat[read(fd, plat, sizeof(plat)) - 1] = '\0';
snprintf(path, sizeof(path), SPLIT_PLAT_MAPPING, plat);
mmap_ro(path, &addr, &size);
VLOG("cil_add[%s]\n", path);
cil_add_file(db, path, addr, size);
munmap(addr, size);
close(fd);
2017-12-01 03:33:25 +08:00
// nonplat
dir = opendir(NONPLAT_POLICY_DIR);
while ((entry = readdir(dir))) {
if (strcmp(entry->d_name, ".") == 0 || strcmp(entry->d_name, "..") == 0)
continue;
if (strend(entry->d_name, ".cil") == 0) {
2017-12-01 03:33:25 +08:00
snprintf(path, sizeof(path), NONPLAT_POLICY_DIR "%s", entry->d_name);
mmap_ro(path, &addr, &size);
2017-12-01 03:33:25 +08:00
VLOG("cil_add[%s]\n", path);
cil_add_file(db, path, addr, size);
munmap(addr, size);
}
}
closedir(dir);
cil_compile(db);
cil_build_policydb(db, &pdb);
cil_db_destroy(&db);
policydb = &pdb->p;
return 0;
}
static int verify_precompiled() {
DIR *dir;
struct dirent *entry;
int fd;
char sys_sha[70], ven_sha[70];
2017-12-01 03:33:25 +08:00
// init the strings with different value
sys_sha[0] = 0;
ven_sha[0] = 1;
dir = opendir(NONPLAT_POLICY_DIR);
while ((entry = readdir(dir))) {
if (strcmp(entry->d_name, ".") == 0 || strcmp(entry->d_name, "..") == 0)
continue;
if (strend(entry->d_name, ".sha256") == 0) {
fd = openat(dirfd(dir), entry->d_name, O_RDONLY | O_CLOEXEC);
2017-12-01 03:33:25 +08:00
ven_sha[read(fd, ven_sha, sizeof(ven_sha)) - 1] = '\0';
close(fd);
break;
}
}
closedir(dir);
2017-12-01 03:33:25 +08:00
dir = opendir(PLAT_POLICY_DIR);
while ((entry = readdir(dir))) {
if (strcmp(entry->d_name, ".") == 0 || strcmp(entry->d_name, "..") == 0)
continue;
if (strend(entry->d_name, ".sha256") == 0) {
fd = openat(dirfd(dir), entry->d_name, O_RDONLY | O_CLOEXEC);
2017-12-01 03:33:25 +08:00
sys_sha[read(fd, sys_sha, sizeof(sys_sha)) - 1] = '\0';
close(fd);
break;
}
}
closedir(dir);
2017-12-01 03:33:25 +08:00
VLOG("sys_sha[%s]\nven_sha[%s]\n", sys_sha, ven_sha);
return strcmp(sys_sha, ven_sha) == 0;
}
static int patch_sepolicy() {
2017-12-01 03:33:25 +08:00
if (access("/sepolicy", R_OK) == 0)
load_policydb("/sepolicy");
2017-12-01 03:33:25 +08:00
else if (access(SPLIT_PRECOMPILE, R_OK) == 0 && verify_precompiled())
load_policydb(SPLIT_PRECOMPILE);
else if (access(SPLIT_PLAT_CIL, R_OK) == 0)
compile_cil();
else
return 1;
2017-11-30 20:57:40 +08:00
sepol_magisk_rules();
dump_policydb("/sepolicy");
2018-02-02 04:17:13 +08:00
// Remove the stupid debug sepolicy and use our own
if (access("/sepolicy_debug", F_OK) == 0) {
unlink("/sepolicy_debug");
link("/sepolicy", "/sepolicy_debug");
}
return 0;
}
2017-11-10 00:54:54 +08:00
#define BUFSIZE (1 << 20)
static int unxz(const void *buf, size_t size, int fd) {
lzma_stream strm = LZMA_STREAM_INIT;
if (lzma_auto_decoder(&strm, UINT64_MAX, 0) != LZMA_OK)
return 1;
2018-02-02 03:22:38 +08:00
lzma_ret ret;
2017-11-10 00:54:54 +08:00
void *out = malloc(BUFSIZE);
strm.next_in = buf;
strm.avail_in = size;
do {
strm.next_out = out;
strm.avail_out = BUFSIZE;
ret = lzma_code(&strm, LZMA_RUN);
write(fd, out, BUFSIZE - strm.avail_out);
} while (strm.avail_out == 0 && ret == LZMA_OK);
free(out);
lzma_end(&strm);
if (ret != LZMA_OK && ret != LZMA_STREAM_END)
return 1;
return 0;
}
static int dump_magisk(const char *path, mode_t mode) {
unlink(path);
int fd = creat(path, mode);
int ret = unxz(magisk_dump, sizeof(magisk_dump), fd);
close(fd);
return ret;
}
static int dump_magiskrc(const char *path, mode_t mode) {
int fd = creat(path, mode);
write(fd, magiskrc, sizeof(magiskrc));
close(fd);
return 0;
}
static void patch_socket_name(const char *path) {
void *buf;
size_t size;
mmap_rw(path, &buf, &size);
if (SOCKET_OFF < 0) {
for (int i = 0; i < size; ++i) {
2018-04-15 03:18:18 +08:00
if (memcmp(buf + i, SOCKET_NAME, sizeof(SOCKET_NAME)) == 0) {
SOCKET_OFF = i;
break;
}
}
}
gen_rand_str(RAND_SOCKET_NAME, sizeof(SOCKET_NAME));
memcpy(buf + SOCKET_OFF, RAND_SOCKET_NAME, sizeof(SOCKET_NAME));
munmap(buf, size);
}
int main(int argc, char *argv[]) {
2017-11-23 18:38:12 +08:00
umask(0);
2017-11-28 04:43:46 +08:00
for (int i = 0; init_applet[i]; ++i) {
if (strcmp(basename(argv[0]), init_applet[i]) == 0)
return (*init_applet_main[i])(argc, argv);
}
2017-11-23 18:38:12 +08:00
if (argc > 1 && strcmp(argv[1], "-x") == 0) {
2017-11-10 00:54:54 +08:00
if (strcmp(argv[2], "magisk") == 0)
return dump_magisk(argv[3], 0755);
else if (strcmp(argv[2], "magiskrc") == 0)
return dump_magiskrc(argv[3], 0755);
}
2017-12-05 21:05:20 +08:00
// Prevent file descriptor confusion
mknod("/null", S_IFCHR | 0666, makedev(1, 3));
int null = open("/null", O_RDWR | O_CLOEXEC);
unlink("/null");
dup3(null, STDIN_FILENO, O_CLOEXEC);
dup3(null, STDOUT_FILENO, O_CLOEXEC);
dup3(null, STDERR_FILENO, O_CLOEXEC);
2017-12-16 04:40:54 +08:00
if (null > STDERR_FILENO)
close(null);
2017-12-05 21:05:20 +08:00
2018-04-21 20:16:59 +08:00
// Backup self
link("/init", "/init.bak");
2017-11-10 00:54:54 +08:00
struct cmdline cmd;
parse_cmdline(&cmd);
/* ***********
* Initialize
* ***********/
2017-11-06 22:33:41 +08:00
int root = open("/", O_RDONLY | O_CLOEXEC);
if (cmd.skip_initramfs) {
// Clear rootfs
2018-04-21 20:16:59 +08:00
excl_list = (char *[]) { "overlay", ".backup", "init.bak", NULL };
frm_rf(root);
} else if (access("/ramdisk.cpio.xz", R_OK) == 0) {
// High compression mode
void *addr;
size_t size;
mmap_ro("/ramdisk.cpio.xz", &addr, &size);
int fd = creat("/ramdisk.cpio", 0);
unxz(addr, size, fd);
munmap(addr, size);
close(fd);
struct vector v;
vec_init(&v);
parse_cpio(&v, "/ramdisk.cpio");
excl_list = (char *[]) { "overlay", ".backup", NULL };
frm_rf(root);
chdir("/");
cpio_extract_all(&v);
cpio_vec_destroy(&v);
} else {
// Revert original init binary
unlink("/init");
link("/.backup/init", "/init");
}
/* ************
* Early Mount
* ************/
// If skip_initramfs or using split policies, we need early mount
if (cmd.skip_initramfs || access("/sepolicy", R_OK) != 0) {
char partname[32];
struct device dev;
// Mount sysfs
mkdir("/sys", 0755);
xmount("sysfs", "/sys", "sysfs", 0, NULL);
// Mount system
snprintf(partname, sizeof(partname), "system%s", cmd.slot);
setup_block(&dev, partname);
if (cmd.skip_initramfs) {
mkdir("/system_root", 0755);
xmount(dev.path, "/system_root", "ext4", MS_RDONLY, NULL);
int system_root = open("/system_root", O_RDONLY | O_CLOEXEC);
// Clone rootfs except /system
excl_list = (char *[]) { "system", NULL };
clone_dir(system_root, root);
close(system_root);
mkdir("/system", 0755);
xmount("/system_root/system", "/system", NULL, MS_BIND, NULL);
} else {
xmount(dev.path, "/system", "ext4", MS_RDONLY, NULL);
}
// Mount vendor
snprintf(partname, sizeof(partname), "vendor%s", cmd.slot);
if (setup_block(&dev, partname) == 0)
xmount(dev.path, "/vendor", "ext4", MS_RDONLY, NULL);
}
/* *************
* Patch rootfs
* *************/
// Only patch rootfs if not intended to run in recovery
2017-12-05 21:05:20 +08:00
if (access("/etc/recovery.fstab", F_OK) != 0) {
2018-04-21 20:16:59 +08:00
int fd;
// Handle ramdisk overlays
fd = open("/overlay", O_RDONLY | O_CLOEXEC);
if (fd >= 0) {
mv_dir(fd, root);
close(fd);
rmdir("/overlay");
}
patch_ramdisk();
patch_sepolicy();
2018-04-21 20:16:59 +08:00
// Dump binaries
dump_magiskrc("/init.magisk.rc", 0750);
dump_magisk("/sbin/magisk", 0755);
patch_socket_name("/sbin/magisk");
rename("/init.bak", "/sbin/magiskinit");
2017-11-28 03:42:48 +08:00
}
// Clean up
close(root);
if (!cmd.skip_initramfs)
umount("/system");
2017-12-05 21:05:20 +08:00
umount("/vendor");
2017-11-23 18:38:12 +08:00
// Finally, give control back!
execv("/init", argv);
}