#include <poll.h>
#include <ctype.h>
#include <dirent.h>
-#include <signal.h>
#include <dlfcn.h>
pr_err("error %d when get PW mode on lock %s\n", errno, str);
/* let's try several times if EAGAIN happened */
if (dlm_lock_res->lksb.sb_status == EAGAIN && retry_count < 10) {
- sleep(10);
+ sleep_for(10, 0, true);
retry_count++;
goto retry;
}
* GET_ARRAY_INFO doesn't provide access to the proper state
* information, so fallback to a basic check for raid_disks != 0
*/
- ret = ioctl(fd, GET_ARRAY_INFO, &array);
+ ret = md_get_array_info(fd, &array);
}
return !ret;
return ioctl(fd, GET_DISK_INFO, disk);
}
-/*
- * Parse a 128 bit uuid in 4 integers
- * format is 32 hexx nibbles with options :.<space> separator
- * If not exactly 32 hex digits are found, return 0
- * else return 1
- */
-int parse_uuid(char *str, int uuid[4])
-{
- int hit = 0; /* number of Hex digIT */
- int i;
- char c;
- for (i = 0; i < 4; i++)
- uuid[i] = 0;
-
- while ((c = *str++) != 0) {
- int n;
- if (c >= '0' && c <= '9')
- n = c-'0';
- else if (c >= 'a' && c <= 'f')
- n = 10 + c - 'a';
- else if (c >= 'A' && c <= 'F')
- n = 10 + c - 'A';
- else if (strchr(":. -", c))
- continue;
- else return 0;
-
- if (hit<32) {
- uuid[hit/8] <<= 4;
- uuid[hit/8] += n;
- }
- hit++;
- }
- if (hit == 32)
- return 1;
- return 0;
-}
-
int get_linux_version()
{
struct utsname name;
unsigned long long parse_size(char *size)
{
/* parse 'size' which should be a number optionally
- * followed by 'K', 'M', or 'G'.
+ * followed by 'K', 'M'. 'G' or 'T'.
* Without a suffix, K is assumed.
* Number returned is in sectors (half-K)
* INVALID_SECTORS returned on error.
c++;
s *= 1024 * 1024 * 2;
break;
+ case 'T':
+ c++;
+ s *= 1024 * 1024 * 1024 * 2LL;
+ break;
case 's': /* sectors */
c++;
break;
int parse_layout_faulty(char *layout)
{
+ int ln, mode;
+ char *m;
+
+ if (!layout)
+ return -1;
+
/* Parse the layout string for 'faulty' */
- int ln = strcspn(layout, "0123456789");
- char *m = xstrdup(layout);
- int mode;
+ ln = strcspn(layout, "0123456789");
+ m = xstrdup(layout);
m[ln] = 0;
mode = map_name(faultylayout, m);
+ free(m);
+
if (mode == UnSet)
return -1;
return mode | (atoi(layout+ln)<< ModeShift);
}
-long parse_num(char *num)
-{
- /* Either return a valid number, or -1 */
- char *c;
- long rv = strtol(num, &c, 10);
- if (rv < 0 || *c || !num[0])
- return -1;
- else
- return rv;
-}
-
int parse_cluster_confirm_arg(char *input, char **devname, int *slot)
{
char *dev;
}
}
-const int uuid_zero[4] = { 0, 0, 0, 0 };
-
-int same_uuid(int a[4], int b[4], int swapuuid)
-{
- if (swapuuid) {
- /* parse uuids are hostendian.
- * uuid's from some superblocks are big-ending
- * if there is a difference, we need to swap..
- */
- unsigned char *ac = (unsigned char *)a;
- unsigned char *bc = (unsigned char *)b;
- int i;
- for (i = 0; i < 16; i += 4) {
- if (ac[i+0] != bc[i+3] ||
- ac[i+1] != bc[i+2] ||
- ac[i+2] != bc[i+1] ||
- ac[i+3] != bc[i+0])
- return 0;
- }
- return 1;
- } else {
- if (a[0]==b[0] &&
- a[1]==b[1] &&
- a[2]==b[2] &&
- a[3]==b[3])
- return 1;
- return 0;
- }
-}
-
-void copy_uuid(void *a, int b[4], int swapuuid)
-{
- if (swapuuid) {
- /* parse uuids are hostendian.
- * uuid's from some superblocks are big-ending
- * if there is a difference, we need to swap..
- */
- unsigned char *ac = (unsigned char *)a;
- unsigned char *bc = (unsigned char *)b;
- int i;
- for (i = 0; i < 16; i += 4) {
- ac[i+0] = bc[i+3];
- ac[i+1] = bc[i+2];
- ac[i+2] = bc[i+1];
- ac[i+3] = bc[i+0];
- }
- } else
- memcpy(a, b, 16);
-}
-
char *__fname_from_uuid(int id[4], int swap, char *buf, char sep)
{
int i, j;
{
static char buf[47];
- /* We convert bytes to either centi-M{ega,ibi}bytes or
- * centi-G{igi,ibi}bytes, with appropriate rounding,
- * and then print 1/100th of those as a decimal.
+ /* We convert bytes to either centi-M{ega,ibi}bytes,
+ * centi-G{igi,ibi}bytes or centi-T{era,ebi}bytes
+ * with appropriate rounding, and then print
+ * 1/100th of those as a decimal.
* We allow upto 2048Megabytes before converting to
- * gigabytes, as that shows more precision and isn't
+ * gigabytes and 2048Gigabytes before converting to
+ * terabytes, as that shows more precision and isn't
* too large a number.
- * Terabytes are not yet handled.
*/
if (bytes < 5000*1024)
long cMB = (bytes / ( 1000000LL / 200LL ) +1) /2;
snprintf(buf, sizeof(buf), " (%ld.%02ld MiB %ld.%02ld MB)",
cMiB/100, cMiB % 100, cMB/100, cMB % 100);
- } else {
+ } else if (bytes < 2*1024LL*1024LL*1024LL*1024LL) {
long cGiB = (bytes * 200LL / (1LL<<30) +1) / 2;
long cGB = (bytes / (1000000000LL/200LL ) +1) /2;
snprintf(buf, sizeof(buf), " (%ld.%02ld GiB %ld.%02ld GB)",
cGiB/100, cGiB % 100, cGB/100, cGB % 100);
+ } else {
+ long cTiB = (bytes * 200LL / (1LL<<40) + 1) / 2;
+ long cTB = (bytes / (1000000000000LL / 200LL) + 1) / 2;
+ snprintf(buf, sizeof(buf), " (%ld.%02ld TiB %ld.%02ld TB)",
+ cTiB/100, cTiB % 100, cTB/100, cTB % 100);
}
return buf;
}
{
static char buf[30];
- /* We convert bytes to either centi-M{ega,ibi}bytes or
- * centi-G{igi,ibi}bytes, with appropriate rounding,
- * and then print 1/100th of those as a decimal.
+ /* We convert bytes to either centi-M{ega,ibi}bytes,
+ * centi-G{igi,ibi}bytes or centi-T{era,ebi}bytes
+ * with appropriate rounding, and then print
+ * 1/100th of those as a decimal.
* We allow upto 2048Megabytes before converting to
- * gigabytes, as that shows more precision and isn't
+ * gigabytes and 2048Gigabytes before converting to
+ * terabytes, as that shows more precision and isn't
* too large a number.
- * Terabytes are not yet handled.
*
* If prefix == IEC, we mean prefixes like kibi,mebi,gibi etc.
* If prefix == JEDEC, we mean prefixes like kilo,mega,giga etc.
long cMiB = (bytes * 200LL / (1LL<<20) +1) /2;
snprintf(buf, sizeof(buf), "%ld.%02ldMiB",
cMiB/100, cMiB % 100);
- } else {
+ } else if (bytes < 2*1024LL*1024LL*1024LL*1024LL) {
long cGiB = (bytes * 200LL / (1LL<<30) +1) /2;
snprintf(buf, sizeof(buf), "%ld.%02ldGiB",
cGiB/100, cGiB % 100);
+ } else {
+ long cTiB = (bytes * 200LL / (1LL<<40) + 1) / 2;
+ snprintf(buf, sizeof(buf), "%ld.%02ldTiB",
+ cTiB/100, cTiB % 100);
}
}
else if (prefix == JEDEC) {
long cMB = (bytes / ( 1000000LL / 200LL ) +1) /2;
snprintf(buf, sizeof(buf), "%ld.%02ldMB",
cMB/100, cMB % 100);
- } else {
+ } else if (bytes < 2*1024LL*1024LL*1024LL*1024LL) {
long cGB = (bytes / (1000000000LL/200LL ) +1) /2;
snprintf(buf, sizeof(buf), "%ld.%02ldGB",
cGB/100, cGB % 100);
+ } else {
+ long cTB = (bytes / (1000000000000LL / 200LL) + 1) / 2;
+ snprintf(buf, sizeof(buf), "%ld.%02ldTB",
+ cTB/100, cTB % 100);
}
}
else
/* First look in /sys/block/$DEVNM/dev for %d:%d
* If that fails, try parsing out a number
*/
- char path[100];
+ char path[PATH_MAX];
char *ep;
int fd;
int mjr,mnr;
- sprintf(path, "/sys/block/%s/dev", devnm);
+ snprintf(path, sizeof(path), "/sys/block/%s/dev", devnm);
fd = open(path, O_RDONLY);
if (fd >= 0) {
char buf[20];
return 0;
}
-char *get_md_name(char *devnm)
+/**
+ * is_devname_numbered() - helper for numbered devname verification.
+ * @devname: path or name to check.
+ * @pref: expected devname prefix.
+ * @pref_len: prefix len.
+ */
+static bool is_devname_numbered(const char *devname, const char *pref, const int pref_len)
{
- /* find /dev/md%d or /dev/md/%d or make a device /dev/.tmp.md%d */
- /* if dev < 0, want /dev/md/d%d or find mdp in /proc/devices ... */
+ int val;
- static char devname[50];
- struct stat stb;
- dev_t rdev = devnm2devid(devnm);
- char *dn;
+ assert(devname && pref);
- if (rdev == 0)
- return 0;
- if (strncmp(devnm, "md_", 3) == 0) {
- snprintf(devname, sizeof(devname), "/dev/md/%s",
- devnm + 3);
- if (stat(devname, &stb) == 0 &&
- (S_IFMT&stb.st_mode) == S_IFBLK && (stb.st_rdev == rdev))
- return devname;
- }
- snprintf(devname, sizeof(devname), "/dev/%s", devnm);
- if (stat(devname, &stb) == 0 && (S_IFMT&stb.st_mode) == S_IFBLK &&
- (stb.st_rdev == rdev))
- return devname;
+ if (strncmp(devname, pref, pref_len) != 0)
+ return false;
- snprintf(devname, sizeof(devname), "/dev/md/%s", devnm+2);
- if (stat(devname, &stb) == 0 && (S_IFMT&stb.st_mode) == S_IFBLK &&
- (stb.st_rdev == rdev))
- return devname;
+ if (parse_num(&val, devname + pref_len) != 0)
+ return false;
+
+ if (val > 127)
+ return false;
+
+ return true;
+}
+
+/**
+ * is_devname_md_numbered() - check if &devname is numbered MD device (md).
+ * @devname: path or name to check.
+ */
+bool is_devname_md_numbered(const char *devname)
+{
+ return is_devname_numbered(devname, DEV_NUM_PREF, DEV_NUM_PREF_LEN);
+}
+
+/**
+ * is_devname_md_d_numbered() - check if &devname is secondary numbered MD device (md_d).
+ * @devname: path or name to check.
+ */
+bool is_devname_md_d_numbered(const char *devname)
+{
+ static const char d_dev[] = DEV_NUM_PREF "_d";
+
+ return is_devname_numbered(devname, d_dev, sizeof(d_dev) - 1);
+}
- dn = map_dev(major(rdev), minor(rdev), 0);
- if (dn)
- return dn;
- snprintf(devname, sizeof(devname), "/dev/.tmp.%s", devnm);
- if (mknod(devname, S_IFBLK | 0600, rdev) == -1)
- if (errno != EEXIST)
- return NULL;
+/**
+ * get_md_name() - Get main dev node of the md device.
+ * @devnm: Md device name or path.
+ *
+ * Function checks if the full name was passed and returns md name
+ * if it is the MD device.
+ *
+ * Return: Main dev node of the md device or NULL if not found.
+ */
+char *get_md_name(char *devnm)
+{
+ static char devname[NAME_MAX];
+ struct stat stb;
+
+ if (strncmp(devnm, "/dev/", 5) == 0)
+ snprintf(devname, sizeof(devname), "%s", devnm);
+ else
+ snprintf(devname, sizeof(devname), "/dev/%s", devnm);
- if (stat(devname, &stb) == 0 && (S_IFMT&stb.st_mode) == S_IFBLK &&
- (stb.st_rdev == rdev))
+ if (!is_mddev(devname))
+ return NULL;
+ if (stat(devname, &stb) == 0 && (S_IFMT&stb.st_mode) == S_IFBLK)
return devname;
- unlink(devname);
+
return NULL;
}
*e == 0);
}
+/**
+ * is_bit_set() - get bit value by index.
+ * @val: value.
+ * @index: index of the bit (LSB numbering).
+ *
+ * Return: bit value.
+ */
+bool is_bit_set(int *val, unsigned char index)
+{
+ if ((*val) & (1 << index))
+ return true;
+ return false;
+}
+
int dev_open(char *dev, int flags)
{
/* like 'open', but if 'dev' matches %d:%d, create a temp
int i;
int flags = O_RDWR;
dev_t devid = devnm2devid(devnm);
- long delay = 1000;
+ unsigned int delay = 1; // miliseconds
sprintf(buf, "%d:%d", major(devid), minor(devid));
for (i = 0; i < 25; i++) {
}
if (errno != EBUSY)
return fd;
- usleep(delay);
- if (delay < 200000)
+ sleep_for(0, MSEC_TO_NSEC(delay), true);
+ if (delay < 200)
delay *= 2;
}
return -1;
{
int i;
struct stat stb_want;
- long delay = 1000;
+ unsigned int delay = 1; // miliseconds
if (fstat(fd, &stb_want) != 0 ||
(stb_want.st_mode & S_IFMT) != S_IFBLK)
(stb.st_mode & S_IFMT) == S_IFBLK &&
(stb.st_rdev == stb_want.st_rdev))
return;
- usleep(delay);
- if (delay < 200000)
+ sleep_for(0, MSEC_TO_NSEC(delay), true);
+ if (delay < 200)
delay *= 2;
}
if (i == 25)
int i;
char *subarray = NULL;
char container[32] = "";
+ char *devnm = NULL;
+
+ devnm = fd2devnm(fd);
+ if (!devnm)
+ return NULL;
sra = sysfs_read(fd, NULL, GET_VERSION);
if (subarrayp)
*subarrayp = subarray;
strcpy(st->container_devnm, container);
- strcpy(st->devnm, fd2devnm(fd));
+ strncpy(st->devnm, devnm, MD_NAME_MAX - 1);
} else
free(subarray);
/* 'fd' is a block device. Find out if it is in use
* by a container, and return an open fd on that container.
*/
- char path[256];
+ char path[288];
char *e;
DIR *dir;
struct dirent *de;
while ((ret = ioctl(mdfd, HOT_REMOVE_DISK, dev)) == -1 &&
errno == EBUSY &&
cnt-- > 0)
- usleep(10000);
+ sleep_for(0, MSEC_TO_NSEC(10), true);
return ret;
}
while ((ret = write(statefd, "remove", 6)) == -1 &&
errno == EBUSY &&
cnt-- > 0)
- usleep(10000);
+ sleep_for(0, MSEC_TO_NSEC(10), true);
return ret == 6 ? 0 : -1;
}
int start_mdmon(char *devnm)
{
- int i, skipped;
+ int i;
int len;
pid_t pid;
int status;
+ char *prefix = in_initrd() ? "initrd-" : "";
char pathbuf[1024];
char *paths[4] = {
pathbuf,
if (check_env("MDADM_NO_MDMON"))
return 0;
+ if (continue_via_systemd(devnm, MDMON_SERVICE, prefix))
+ return 0;
+ /* That failed, try running mdmon directly */
len = readlink("/proc/self/exe", pathbuf, sizeof(pathbuf)-1);
if (len > 0) {
char *sl;
} else
pathbuf[0] = '\0';
- /* First try to run systemctl */
- if (!check_env("MDADM_NO_SYSTEMCTL"))
- switch(fork()) {
- case 0:
- /* FIXME yuk. CLOSE_EXEC?? */
- skipped = 0;
- for (i = 3; skipped < 20; i++)
- if (close(i) < 0)
- skipped++;
- else
- skipped = 0;
-
- /* Don't want to see error messages from
- * systemctl. If the service doesn't exist,
- * we start mdmon ourselves.
- */
- close(2);
- open("/dev/null", O_WRONLY);
- snprintf(pathbuf, sizeof(pathbuf), "mdmon@%s.service",
- devnm);
- status = execl("/usr/bin/systemctl", "systemctl",
- "start",
- pathbuf, NULL);
- status = execl("/bin/systemctl", "systemctl", "start",
- pathbuf, NULL);
- exit(1);
- case -1: pr_err("cannot run mdmon. Array remains readonly\n");
- return -1;
- default: /* parent - good */
- pid = wait(&status);
- if (pid >= 0 && status == 0)
- return 0;
- }
-
- /* That failed, try running mdmon directly */
switch(fork()) {
case 0:
- /* FIXME yuk. CLOSE_EXEC?? */
- skipped = 0;
- for (i = 3; skipped < 20; i++)
- if (close(i) < 0)
- skipped++;
- else
- skipped = 0;
-
+ manage_fork_fds(1);
for (i = 0; paths[i]; i++)
if (paths[i][0]) {
execl(paths[i], paths[i],
setrlimit(RLIMIT_NOFILE, &lim);
}
+/* Close all opened descriptors if needed and redirect
+ * streams to /dev/null.
+ * For debug purposed, leave STDOUT and STDERR untouched
+ * Returns:
+ * 1- if any error occurred
+ * 0- otherwise
+ */
+void manage_fork_fds(int close_all)
+{
+ DIR *dir;
+ struct dirent *dirent;
+
+ close(0);
+ open("/dev/null", O_RDWR);
+
+#ifndef DEBUG
+ dup2(0, 1);
+ dup2(0, 2);
+#endif
+
+ if (close_all == 0)
+ return;
+
+ dir = opendir("/proc/self/fd");
+ if (!dir) {
+ pr_err("Cannot open /proc/self/fd directory.\n");
+ return;
+ }
+ for (dirent = readdir(dir); dirent; dirent = readdir(dir)) {
+ int fd = -1;
+
+ if ((strcmp(dirent->d_name, ".") == 0) ||
+ (strcmp(dirent->d_name, "..")) == 0)
+ continue;
+
+ fd = strtol(dirent->d_name, NULL, 10);
+ if (fd > 2)
+ close(fd);
+ }
+}
+
+/* In a systemd/udev world, it is best to get systemd to
+ * run daemon rather than running in the background.
+ * Returns:
+ * 1- if systemd service has been started
+ * 0- otherwise
+ */
+int continue_via_systemd(char *devnm, char *service_name, char *prefix)
+{
+ int pid, status;
+ char pathbuf[1024];
+
+ /* Simply return that service cannot be started */
+ if (check_env("MDADM_NO_SYSTEMCTL"))
+ return 0;
+ switch (fork()) {
+ case 0:
+ manage_fork_fds(1);
+ snprintf(pathbuf, sizeof(pathbuf),
+ "%s@%s%s.service", service_name, prefix ?: "", devnm);
+ status = execl("/usr/bin/systemctl", "systemctl", "restart",
+ pathbuf, NULL);
+ status = execl("/bin/systemctl", "systemctl", "restart",
+ pathbuf, NULL);
+ exit(1);
+ case -1: /* Just do it ourselves. */
+ break;
+ default: /* parent - good */
+ pid = wait(&status);
+ if (pid >= 0 && status == 0)
+ return 1;
+ }
+ return 0;
+}
+
int in_initrd(void)
{
- /* This is based on similar function in systemd. */
- struct statfs s;
- /* statfs.f_type is signed long on s390x and MIPS, causing all
- sorts of sign extension problems with RAMFS_MAGIC being
- defined as 0x858458f6 */
- return statfs("/", &s) >= 0 &&
- ((unsigned long)s.f_type == TMPFS_MAGIC ||
- ((unsigned long)s.f_type & 0xFFFFFFFFUL) ==
- ((unsigned long)RAMFS_MAGIC & 0xFFFFFFFFUL));
+ return access("/etc/initrd-release", F_OK) >= 0;
}
void reopen_mddev(int mdfd)
close(fd_zero);
return ret;
}
+
+/**
+ * sleep_for() - Sleeps for specified time.
+ * @sec: Seconds to sleep for.
+ * @nsec: Nanoseconds to sleep for, has to be less than one second.
+ * @wake_after_interrupt: If set, wake up if interrupted.
+ *
+ * Function immediately returns if error different than EINTR occurs.
+ */
+void sleep_for(unsigned int sec, long nsec, bool wake_after_interrupt)
+{
+ struct timespec delay = {.tv_sec = sec, .tv_nsec = nsec};
+
+ assert(nsec < MSEC_TO_NSEC(1000));
+
+ do {
+ errno = 0;
+ nanosleep(&delay, &delay);
+ if (errno != 0 && errno != EINTR) {
+ pr_err("Error sleeping for %us %ldns: %s\n", sec, nsec, strerror(errno));
+ return;
+ }
+ } while (!wake_after_interrupt && errno == EINTR);
+}
+
+/* is_directory() - Checks if directory provided by path is indeed a regular directory.
+ * @path: directory path to be checked
+ *
+ * Doesn't accept symlinks.
+ *
+ * Return: true if is a directory, false if not
+ */
+bool is_directory(const char *path)
+{
+ struct stat st;
+
+ if (lstat(path, &st) != 0) {
+ pr_err("%s: %s\n", strerror(errno), path);
+ return false;
+ }
+
+ if (!S_ISDIR(st.st_mode))
+ return false;
+
+ return true;
+}
+
+/*
+ * is_file() - Checks if file provided by path is indeed a regular file.
+ * @path: file path to be checked
+ *
+ * Doesn't accept symlinks.
+ *
+ * Return: true if is a file, false if not
+ */
+bool is_file(const char *path)
+{
+ struct stat st;
+
+ if (lstat(path, &st) != 0) {
+ pr_err("%s: %s\n", strerror(errno), path);
+ return false;
+ }
+
+ if (!S_ISREG(st.st_mode))
+ return false;
+
+ return true;
+}
projects / thirdparty / mdadm.git / blobdiff