#include <sys/un.h>
#include <sys/resource.h>
#include <sys/vfs.h>
+#include <sys/mman.h>
#include <linux/magic.h>
+#include <poll.h>
#include <ctype.h>
#include <dirent.h>
#include <signal.h>
+#include <dlfcn.h>
+
/*
* following taken from linux/blkpg.h because they aren't
aren't permitted). */
#define BUILD_BUG_ON_ZERO(e) (sizeof(struct { int:-!!(e); }))
-/*
- * 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])
+static int is_dlm_hooks_ready = 0;
+
+int dlm_funs_ready(void)
{
- 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;
+ return is_dlm_hooks_ready ? 1 : 0;
+}
+
+static struct dlm_hooks *dlm_hooks = NULL;
+struct dlm_lock_resource *dlm_lock_res = NULL;
+static int ast_called = 0;
+
+struct dlm_lock_resource {
+ dlm_lshandle_t *ls;
+ struct dlm_lksb lksb;
+};
- if (hit<32) {
- uuid[hit/8] <<= 4;
- uuid[hit/8] += n;
+/* Using poll(2) to wait for and dispatch ASTs */
+static int poll_for_ast(dlm_lshandle_t ls)
+{
+ struct pollfd pfd;
+
+ pfd.fd = dlm_hooks->ls_get_fd(ls);
+ pfd.events = POLLIN;
+
+ while (!ast_called)
+ {
+ if (poll(&pfd, 1, 0) < 0)
+ {
+ perror("poll");
+ return -1;
}
- hit++;
+ dlm_hooks->dispatch(dlm_hooks->ls_get_fd(ls));
}
- if (hit == 32)
- return 1;
+ ast_called = 0;
+
return 0;
}
-/*
- * Get the md version number.
- * We use the RAID_VERSION ioctl if it is supported
- * If not, but we have a block device with major '9', we assume
- * 0.36.0
- *
- * Return version number as 24 but number - assume version parts
- * always < 255
- */
+static void dlm_ast(void *arg)
+{
+ ast_called = 1;
+}
-int md_get_version(int fd)
+static char *cluster_name = NULL;
+/* Create the lockspace, take bitmapXXX locks on all the bitmaps. */
+int cluster_get_dlmlock(void)
{
- struct stat stb;
- mdu_version_t vers;
+ int ret = -1;
+ char str[64];
+ int flags = LKF_NOQUEUE;
+ int retry_count = 0;
- if (fstat(fd, &stb)<0)
+ if (!dlm_funs_ready()) {
+ pr_err("Something wrong with dlm library\n");
return -1;
- if ((S_IFMT&stb.st_mode) != S_IFBLK)
+ }
+
+ ret = get_cluster_name(&cluster_name);
+ if (ret) {
+ pr_err("The md can't get cluster name\n");
return -1;
+ }
- if (ioctl(fd, RAID_VERSION, &vers) == 0)
- return (vers.major*10000) + (vers.minor*100) + vers.patchlevel;
- if (errno == EACCES)
+ dlm_lock_res = xmalloc(sizeof(struct dlm_lock_resource));
+ dlm_lock_res->ls = dlm_hooks->open_lockspace(cluster_name);
+ if (!dlm_lock_res->ls) {
+ dlm_lock_res->ls = dlm_hooks->create_lockspace(cluster_name, O_RDWR);
+ if (!dlm_lock_res->ls) {
+ pr_err("%s failed to create lockspace\n", cluster_name);
+ return -ENOMEM;
+ }
+ } else {
+ pr_err("open existed %s lockspace\n", cluster_name);
+ }
+
+ snprintf(str, 64, "bitmap%s", cluster_name);
+retry:
+ ret = dlm_hooks->ls_lock(dlm_lock_res->ls, LKM_PWMODE,
+ &dlm_lock_res->lksb, flags, str, strlen(str),
+ 0, dlm_ast, dlm_lock_res, NULL, NULL);
+ if (ret) {
+ 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);
+ retry_count++;
+ goto retry;
+ }
+ dlm_hooks->release_lockspace(cluster_name, dlm_lock_res->ls, 1);
+ return ret;
+ }
+
+ /* Wait for it to complete */
+ poll_for_ast(dlm_lock_res->ls);
+
+ if (dlm_lock_res->lksb.sb_status) {
+ pr_err("failed to lock cluster\n");
return -1;
- if (major(stb.st_rdev) == MD_MAJOR)
- return (3600);
- return -1;
+ }
+ return 1;
+}
+
+int cluster_release_dlmlock(void)
+{
+ int ret = -1;
+
+ if (!cluster_name)
+ goto out;
+
+ if (!dlm_lock_res->lksb.sb_lkid)
+ goto out;
+
+ ret = dlm_hooks->ls_unlock_wait(dlm_lock_res->ls,
+ dlm_lock_res->lksb.sb_lkid, 0,
+ &dlm_lock_res->lksb);
+ if (ret) {
+ pr_err("error %d happened when unlock\n", errno);
+ /* XXX make sure the lock is unlocked eventually */
+ goto out;
+ }
+
+ /* Wait for it to complete */
+ poll_for_ast(dlm_lock_res->ls);
+
+ errno = dlm_lock_res->lksb.sb_status;
+ if (errno != EUNLOCK) {
+ pr_err("error %d happened in ast when unlock lockspace\n",
+ errno);
+ /* XXX make sure the lockspace is unlocked eventually */
+ goto out;
+ }
+
+ ret = dlm_hooks->release_lockspace(cluster_name, dlm_lock_res->ls, 1);
+ if (ret) {
+ pr_err("error %d happened when release lockspace\n", errno);
+ /* XXX make sure the lockspace is released eventually */
+ goto out;
+ }
+ free(dlm_lock_res);
+
+out:
+ return ret;
+}
+
+int md_array_valid(int fd)
+{
+ struct mdinfo *sra;
+ int ret;
+
+ sra = sysfs_read(fd, NULL, GET_ARRAY_STATE);
+ if (sra) {
+ if (sra->array_state != ARRAY_UNKNOWN_STATE)
+ ret = 0;
+ else
+ ret = -ENODEV;
+
+ free(sra);
+ } else {
+ /*
+ * 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, RAID_VERSION);
+ }
+
+ return !ret;
+}
+
+int md_array_active(int fd)
+{
+ struct mdinfo *sra;
+ struct mdu_array_info_s array;
+ int ret = 0;
+
+ sra = sysfs_read(fd, NULL, GET_ARRAY_STATE);
+ if (sra) {
+ if (!md_array_is_active(sra))
+ ret = -ENODEV;
+
+ free(sra);
+ } else {
+ /*
+ * 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);
+ }
+
+ return !ret;
+}
+
+int md_array_is_active(struct mdinfo *info)
+{
+ return (info->array_state != ARRAY_CLEAR &&
+ info->array_state != ARRAY_INACTIVE &&
+ info->array_state != ARRAY_UNKNOWN_STATE);
+}
+
+/*
+ * Get array info from the kernel. Longer term we want to deprecate the
+ * ioctl and get it from sysfs.
+ */
+int md_get_array_info(int fd, struct mdu_array_info_s *array)
+{
+ return ioctl(fd, GET_ARRAY_INFO, array);
+}
+
+/*
+ * Set array info
+ */
+int md_set_array_info(int fd, struct mdu_array_info_s *array)
+{
+ return ioctl(fd, SET_ARRAY_INFO, array);
+}
+
+/*
+ * Get disk info from the kernel.
+ */
+int md_get_disk_info(int fd, struct mdu_disk_info_s *disk)
+{
+ return ioctl(fd, GET_DISK_INFO, disk);
}
int get_linux_version()
return (a*1000000)+(b*1000)+c;
}
-#ifndef MDASSEMBLE
int mdadm_version(char *version)
{
int a, b, c;
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;
return s;
}
+int is_near_layout_10(int layout)
+{
+ int fc, fo;
+
+ fc = (layout >> 8) & 255;
+ fo = layout & (1 << 16);
+ if (fc > 1 || fo > 0)
+ return 0;
+ return 1;
+}
+
int parse_layout_10(char *layout)
{
int copies, rv;
else
return rv;
}
-#endif
+
+int parse_cluster_confirm_arg(char *input, char **devname, int *slot)
+{
+ char *dev;
+ *slot = strtoul(input, &dev, 10);
+ if (dev == input || dev[0] != ':')
+ return -1;
+ *devname = dev+1;
+ return 0;
+}
void remove_partitions(int fd)
{
case 1:
return avail_disks >= 1;
case 4:
+ if (avail_disks == raid_disks - 1 &&
+ !avail[raid_disks - 1])
+ /* If just the parity device is missing, then we
+ * have enough, even if not clean
+ */
+ return 1;
+ /* FALL THROUGH */
case 5:
if (clean)
return avail_disks >= raid_disks-1;
}
}
-int enough_fd(int fd)
-{
- struct mdu_array_info_s array;
- struct mdu_disk_info_s disk;
- int i, rv;
- char *avail;
-
- if (ioctl(fd, GET_ARRAY_INFO, &array) != 0 ||
- array.raid_disks <= 0)
- return 0;
- avail = xcalloc(array.raid_disks, 1);
- for (i = 0; i < MAX_DISKS && array.nr_disks > 0; i++) {
- disk.number = i;
- if (ioctl(fd, GET_DISK_INFO, &disk) != 0)
- continue;
- if (disk.major == 0 && disk.minor == 0)
- continue;
- array.nr_disks--;
-
- if (! (disk.state & (1<<MD_DISK_SYNC)))
- continue;
- if (disk.raid_disk < 0 || disk.raid_disk >= array.raid_disks)
- continue;
- avail[disk.raid_disk] = 1;
- }
- /* This is used on an active array, so assume it is clean */
- rv = enough(array.level, array.raid_disks, array.layout,
- 1, avail);
- free(avail);
- return rv;
-}
-
-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;
}
-char *fname_from_uuid(struct supertype *st, struct mdinfo *info, char *buf, char sep)
+char *fname_from_uuid(struct supertype *st, struct mdinfo *info,
+ char *buf, char sep)
{
// dirty hack to work around an issue with super1 superblocks...
// super1 superblocks need swapuuid set in order for assembly to
// work, but can't have it set if we want this printout to match
// all the other uuid printouts in super1.c, so we force swapuuid
// to 1 to make our printout match the rest of super1
- return __fname_from_uuid(info->uuid, (st->ss == &super1) ? 1 : st->ss->swapuuid, buf, sep);
+#if __BYTE_ORDER == BIG_ENDIAN
+ return __fname_from_uuid(info->uuid, 1, buf, sep);
+#else
+ return __fname_from_uuid(info->uuid, (st->ss == &super1) ? 1 :
+ st->ss->swapuuid, buf, sep);
+#endif
}
-#ifndef MDASSEMBLE
int check_ext2(int fd, char *name)
{
/*
if (!st)
return 0;
- st->ss->load_super(st, fd, name);
- /* Looks like a raid array .. */
- pr_err("%s appears to be part of a raid array:\n",
- name);
- st->ss->getinfo_super(st, &info, NULL);
- st->ss->free_super(st);
- crtime = info.array.ctime;
- level = map_num(pers, info.array.level);
- if (!level) level = "-unknown-";
- cont_err("level=%s devices=%d ctime=%s",
- level, info.array.raid_disks, ctime(&crtime));
+ if (st->ss->add_to_super != NULL) {
+ st->ss->load_super(st, fd, name);
+ /* Looks like a raid array .. */
+ pr_err("%s appears to be part of a raid array:\n", name);
+ st->ss->getinfo_super(st, &info, NULL);
+ st->ss->free_super(st);
+ crtime = info.array.ctime;
+ level = map_num(pers, info.array.level);
+ if (!level)
+ level = "-unknown-";
+ cont_err("level=%s devices=%d ctime=%s",
+ level, info.array.raid_disks, ctime(&crtime));
+ } else {
+ /* Looks like GPT or MBR */
+ pr_err("partition table exists on %s\n", name);
+ }
+ return 1;
+}
+
+int fstat_is_blkdev(int fd, char *devname, dev_t *rdev)
+{
+ struct stat stb;
+
+ if (fstat(fd, &stb) != 0) {
+ pr_err("fstat failed for %s: %s\n", devname, strerror(errno));
+ return 0;
+ }
+ if ((S_IFMT & stb.st_mode) != S_IFBLK) {
+ pr_err("%s is not a block device.\n", devname);
+ return 0;
+ }
+ if (rdev)
+ *rdev = stb.st_rdev;
+ return 1;
+}
+
+int stat_is_blkdev(char *devname, dev_t *rdev)
+{
+ struct stat stb;
+
+ if (stat(devname, &stb) != 0) {
+ pr_err("stat failed for %s: %s\n", devname, strerror(errno));
+ return 0;
+ }
+ if ((S_IFMT & stb.st_mode) != S_IFBLK) {
+ pr_err("%s is not a block device.\n", devname);
+ return 0;
+ }
+ if (rdev)
+ *rdev = stb.st_rdev;
return 1;
}
pr_err("assuming 'no'\n");
return 0;
}
-#endif /* MDASSEMBLE */
int is_standard(char *dev, int *nump)
{
return csum;
}
-#ifndef MDASSEMBLE
char *human_size(long long bytes)
{
- static char buf[30];
+ 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 cMiB = (bytes * 200LL / (1LL<<20) + 1) / 2;
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 {
+ cMiB/100, cMiB % 100, cMB/100, cMB % 100);
+ } 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);
+ 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.
if (bytes < 2*1024LL*1024LL*1024LL) {
long cMiB = (bytes * 200LL / (1LL<<20) +1) /2;
snprintf(buf, sizeof(buf), "%ld.%02ldMiB",
- cMiB/100 , cMiB % 100);
- } else {
+ cMiB/100, cMiB % 100);
+ } 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);
+ 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) {
if (bytes < 2*1024LL*1024LL*1024LL) {
long cMB = (bytes / ( 1000000LL / 200LL ) +1) /2;
snprintf(buf, sizeof(buf), "%ld.%02ldMB",
- cMB/100, cMB % 100);
- } else {
+ cMB/100, cMB % 100);
+ } 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);
+ cGB/100, cGB % 100);
+ } else {
+ long cTB = (bytes / (1000000000000LL / 200LL) + 1) / 2;
+ snprintf(buf, sizeof(buf), "%ld.%02ldTB",
+ cTB/100, cTB % 100);
}
}
else
if (near*far == 1)
printf("NO REDUNDANCY");
}
-#endif
unsigned long long calc_array_size(int level, int raid_disks, int layout,
int chunksize, unsigned long long devsize)
return data_disks;
}
-int devnm2devid(char *devnm)
+dev_t devnm2devid(char *devnm)
{
/* First look in /sys/block/$DEVNM/dev for %d:%d
* If that fails, try parsing out a number
return 0;
}
-#if !defined(MDASSEMBLE) || defined(MDASSEMBLE) && defined(MDASSEMBLE_AUTO)
char *get_md_name(char *devnm)
{
/* find /dev/md%d or /dev/md/%d or make a device /dev/.tmp.md%d */
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))
+ 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))
+ if (stat(devname, &stb) == 0 && (S_IFMT&stb.st_mode) == S_IFBLK &&
+ (stb.st_rdev == rdev))
return devname;
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))
+ if (stat(devname, &stb) == 0 && (S_IFMT&stb.st_mode) == S_IFBLK &&
+ (stb.st_rdev == rdev))
return devname;
dn = map_dev(major(rdev), minor(rdev), 0);
if (errno != EEXIST)
return NULL;
- if (stat(devname, &stb) == 0
- && (S_IFMT&stb.st_mode) == S_IFBLK
- && (stb.st_rdev == rdev))
+ if (stat(devname, &stb) == 0 && (S_IFMT&stb.st_mode) == S_IFBLK &&
+ (stb.st_rdev == rdev))
return devname;
unlink(devname);
return NULL;
if (strncmp(name, "/dev/.tmp.md", 12) == 0)
unlink(name);
}
-#endif /* !defined(MDASSEMBLE) || defined(MDASSEMBLE) && defined(MDASSEMBLE_AUTO) */
int get_maj_min(char *dev, int *major, int *minor)
{
int major;
int minor;
- if (!dev) return -1;
+ if (!dev)
+ return -1;
flags |= O_DIRECT;
if (get_maj_min(dev, &major, &minor)) {
}
if (fd < 0) {
/* Try /tmp as /dev appear to be read-only */
- snprintf(devname, sizeof(devname), "/tmp/.tmp.md.%d:%d:%d",
+ snprintf(devname, sizeof(devname),
+ "/tmp/.tmp.md.%d:%d:%d",
(int)getpid(), major, minor);
- if (mknod(devname, S_IFBLK|0600, makedev(major, minor)) == 0) {
+ if (mknod(devname, S_IFBLK|0600,
+ makedev(major, minor)) == 0) {
fd = open(devname, flags);
unlink(devname);
}
int open_dev_flags(char *devnm, int flags)
{
- int devid;
+ dev_t devid;
char buf[20];
devid = devnm2devid(devnm);
char buf[20];
int i;
int flags = O_RDWR;
- int devid = devnm2devid(devnm);
+ dev_t devid = devnm2devid(devnm);
long delay = 1000;
sprintf(buf, "%d:%d", major(devid), minor(devid));
- for (i = 0 ; i < 25 ; i++) {
+ for (i = 0; i < 25; i++) {
int fd = dev_open(buf, flags|O_EXCL);
if (fd >= 0)
return fd;
(stb_want.st_mode & S_IFMT) != S_IFBLK)
return;
- for (i = 0 ; i < 25 ; i++) {
+ for (i = 0; i < 25; i++) {
struct stat stb;
if (stat(dev, &stb) == 0 &&
(stb.st_mode & S_IFMT) == S_IFBLK &&
delay *= 2;
}
if (i == 25)
- dprintf("timeout waiting for %s\n", dev);
+ pr_err("timeout waiting for %s\n", dev);
}
struct superswitch *superlist[] =
&super0, &super1,
&super_ddf, &super_imsm,
&mbr, &gpt,
- NULL };
-
-#if !defined(MDASSEMBLE) || defined(MDASSEMBLE) && defined(MDASSEMBLE_AUTO)
+ NULL
+};
struct supertype *super_by_fd(int fd, char **subarrayp)
{
minor = sra->array.minor_version;
verstr = sra->text_version;
} else {
- if (ioctl(fd, GET_ARRAY_INFO, &array))
+ if (md_get_array_info(fd, &array))
array.major_version = array.minor_version = 0;
vers = array.major_version;
minor = array.minor_version;
subarray = xstrdup(subarray);
}
strcpy(container, dev);
- if (sra)
- sysfs_free(sra);
+ sysfs_free(sra);
sra = sysfs_read(-1, container, GET_VERSION);
if (sra && sra->text_version[0])
verstr = sra->text_version;
verstr = "-no-metadata-";
}
- for (i = 0; st == NULL && superlist[i] ; i++)
+ for (i = 0; st == NULL && superlist[i]; i++)
st = superlist[i]->match_metadata_desc(verstr);
- if (sra)
- sysfs_free(sra);
+ sysfs_free(sra);
if (st) {
st->sb = NULL;
if (subarrayp)
return st;
}
-#endif /* !defined(MDASSEMBLE) || defined(MDASSEMBLE) && defined(MDASSEMBLE_AUTO) */
int dev_size_from_id(dev_t id, unsigned long long *size)
{
return 0;
}
+int dev_sector_size_from_id(dev_t id, unsigned int *size)
+{
+ char buf[20];
+ int fd;
+
+ sprintf(buf, "%d:%d", major(id), minor(id));
+ fd = dev_open(buf, O_RDONLY);
+ if (fd < 0)
+ return 0;
+ if (get_dev_sector_size(fd, NULL, size)) {
+ close(fd);
+ return 1;
+ }
+ close(fd);
+ return 0;
+}
+
struct supertype *dup_super(struct supertype *orig)
{
struct supertype *st;
*/
struct superswitch *ss;
struct supertype *st;
- time_t besttime = 0;
+ unsigned int besttime = 0;
int bestsuper = -1;
int i;
st = xcalloc(1, sizeof(*st));
st->container_devnm[0] = 0;
- for (i = 0 ; superlist[i]; i++) {
+ for (i = 0; superlist[i]; i++) {
int rv;
ss = superlist[i];
if (guess_type == guess_array && ss->add_to_super == NULL)
ldsize <<= 9;
} else {
if (dname)
- pr_err("Cannot get size of %s: %s\b",
+ pr_err("Cannot get size of %s: %s\n",
dname, strerror(errno));
return 0;
}
return 1;
}
+/* Return sector size of device in bytes */
+int get_dev_sector_size(int fd, char *dname, unsigned int *sectsizep)
+{
+ unsigned int sectsize;
+
+ if (ioctl(fd, BLKSSZGET, §size) != 0) {
+ if (dname)
+ pr_err("Cannot get sector size of %s: %s\n",
+ dname, strerror(errno));
+ return 0;
+ }
+
+ *sectsizep = sectsize;
+ return 1;
+}
+
/* Return true if this can only be a container, not a member device.
* i.e. is and md device and size is zero
*/
int must_be_container(int fd)
{
+ struct mdinfo *mdi;
unsigned long long size;
- if (md_get_version(fd) < 0)
+
+ mdi = sysfs_read(fd, NULL, GET_VERSION);
+ if (!mdi)
return 0;
+ sysfs_free(mdi);
+
if (get_dev_size(fd, NULL, &size) == 0)
return 1;
if (size == 0)
unsigned long long curr_part_end;
unsigned all_partitions, entry_size;
unsigned part_nr;
+ unsigned int sector_size = 0;
*endofpart = 0;
BUILD_BUG_ON(sizeof(gpt) != 512);
/* skip protective MBR */
- lseek(fd, 512, SEEK_SET);
+ if (!get_dev_sector_size(fd, NULL, §or_size))
+ return 0;
+ lseek(fd, sector_size, SEEK_SET);
/* read GPT header */
if (read(fd, &gpt, 512) != 512)
return 0;
part = (struct GPT_part_entry *)buf;
+ /* set offset to third block (GPT entries) */
+ lseek(fd, sector_size*2, SEEK_SET);
for (part_nr = 0; part_nr < all_partitions; part_nr++) {
/* read partition entry */
if (read(fd, buf, entry_size) != (ssize_t)entry_size)
static int get_last_partition_end(int fd, unsigned long long *endofpart)
{
struct MBR boot_sect;
- struct MBR_part_record *part;
unsigned long long curr_part_end;
unsigned part_nr;
+ unsigned int sector_size;
int retval = 0;
*endofpart = 0;
if (boot_sect.magic == MBR_SIGNATURE_MAGIC) {
retval = 1;
/* found the correct signature */
- part = boot_sect.parts;
for (part_nr = 0; part_nr < MBR_PARTITIONS; part_nr++) {
+ /*
+ * Have to make every access through boot_sect rather
+ * than using a pointer to the partition table (or an
+ * entry), since the entries are not properly aligned.
+ */
+
/* check for GPT type */
- if (part->part_type == MBR_GPT_PARTITION_TYPE) {
+ if (boot_sect.parts[part_nr].part_type ==
+ MBR_GPT_PARTITION_TYPE) {
retval = get_gpt_last_partition_end(fd, endofpart);
break;
}
/* check the last used lba for the current partition */
- curr_part_end = __le32_to_cpu(part->first_sect_lba) +
- __le32_to_cpu(part->blocks_num);
+ curr_part_end =
+ __le32_to_cpu(boot_sect.parts[part_nr].first_sect_lba) +
+ __le32_to_cpu(boot_sect.parts[part_nr].blocks_num);
if (curr_part_end > *endofpart)
*endofpart = curr_part_end;
-
- part++;
}
} else {
/* Unknown partition table */
retval = -1;
}
+ /* calculate number of 512-byte blocks */
+ if (get_dev_sector_size(fd, NULL, §or_size))
+ *endofpart *= (sector_size / 512);
abort:
return retval;
}
* Check where the last partition ends
*/
unsigned long long endofpart;
- int ret;
- if ((ret = get_last_partition_end(fd, &endofpart)) > 0) {
+ if (get_last_partition_end(fd, &endofpart) > 0) {
/* There appears to be a partition table here */
if (freesize == 0) {
/* partitions will not be visible in new device */
{
/* Add a device to an array, in one of 2 ways. */
int rv;
-#ifndef MDASSEMBLE
+
if (st->ss->external) {
if (info->disk.state & (1<<MD_DISK_SYNC))
info->recovery_start = MaxSector;
}
}
} else
-#endif
rv = ioctl(mdfd, ADD_NEW_DISK, &info->disk);
return rv;
}
struct mdinfo *sra, struct mdinfo *info)
{
int rv;
+
/* Remove the disk given by 'info' from the array */
-#ifndef MDASSEMBLE
if (st->ss->external)
rv = sysfs_set_str(sra, info, "slot", "none");
else
-#endif
rv = ioctl(mdfd, HOT_REMOVE_DISK, makedev(info->disk.major,
info->disk.minor));
return rv;
}
+int hot_remove_disk(int mdfd, unsigned long dev, int force)
+{
+ int cnt = force ? 500 : 5;
+ int ret;
+
+ /* HOT_REMOVE_DISK can fail with EBUSY if there are
+ * outstanding IO requests to the device.
+ * In this case, it can be helpful to wait a little while,
+ * up to 5 seconds if 'force' is set, or 50 msec if not.
+ */
+ while ((ret = ioctl(mdfd, HOT_REMOVE_DISK, dev)) == -1 &&
+ errno == EBUSY &&
+ cnt-- > 0)
+ usleep(10000);
+
+ return ret;
+}
+
+int sys_hot_remove_disk(int statefd, int force)
+{
+ int cnt = force ? 500 : 5;
+ int ret;
+
+ while ((ret = write(statefd, "remove", 6)) == -1 &&
+ errno == EBUSY &&
+ cnt-- > 0)
+ usleep(10000);
+ return ret == 6 ? 0 : -1;
+}
+
int set_array_info(int mdfd, struct supertype *st, struct mdinfo *info)
{
/* Initialise kernel's knowledge of array.
* This varies between externally managed arrays
* and older kernels
*/
- int vers = md_get_version(mdfd);
+ mdu_array_info_t inf;
int rv;
-#ifndef MDASSEMBLE
if (st->ss->external)
- rv = sysfs_set_array(info, vers);
- else
-#endif
- if ((vers % 100) >= 1) { /* can use different versions */
- mdu_array_info_t inf;
- memset(&inf, 0, sizeof(inf));
- inf.major_version = info->array.major_version;
- inf.minor_version = info->array.minor_version;
- rv = ioctl(mdfd, SET_ARRAY_INFO, &inf);
- } else
- rv = ioctl(mdfd, SET_ARRAY_INFO, NULL);
+ return sysfs_set_array(info, 9003);
+
+ memset(&inf, 0, sizeof(inf));
+ inf.major_version = info->array.major_version;
+ inf.minor_version = info->array.minor_version;
+ rv = md_set_array_info(mdfd, &inf);
+
return rv;
}
int start_mdmon(char *devnm)
{
- int i, skipped;
+ int i;
int len;
pid_t pid;
int status;
if (check_env("MDADM_NO_MDMON"))
return 0;
+ if (continue_via_systemd(devnm, MDMON_SERVICE))
+ 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],
return rv;
}
-#ifndef MDASSEMBLE
+void random_uuid(__u8 *buf)
+{
+ int fd, i, len;
+ __u32 r[4];
+
+ fd = open("/dev/urandom", O_RDONLY);
+ if (fd < 0)
+ goto use_random;
+ len = read(fd, buf, 16);
+ close(fd);
+ if (len != 16)
+ goto use_random;
+
+ return;
+
+use_random:
+ for (i = 0; i < 4; i++)
+ r[i] = random();
+ memcpy(buf, r, 16);
+}
+
int flush_metadata_updates(struct supertype *st)
{
int sfd;
*st->update_tail = mu;
st->update_tail = &mu->next;
}
-#endif /* MDASSEMBLE */
#ifdef __TINYC__
/* tinyc doesn't optimize this check in ioctl.h out ... */
unsigned int __invalid_size_argument_for_IOC = 0;
#endif
-int experimental(void)
-{
- if (check_env("MDADM_EXPERIMENTAL"))
- return 1;
- else {
- pr_err("To use this feature MDADM_EXPERIMENTAL environment variable has to be defined.\n");
- return 0;
- }
-}
-
/* Pick all spares matching given criteria from a container
* if min_size == 0 do not check size
* if domlist == NULL do not check domains
* if spare_group given add it to domains of each spare
* metadata allows to test domains using metadata of destination array */
struct mdinfo *container_choose_spares(struct supertype *st,
- unsigned long long min_size,
+ struct spare_criteria *criteria,
struct domainlist *domlist,
char *spare_group,
const char *metadata, int get_one)
if (d->disk.state == 0) {
/* check if size is acceptable */
unsigned long long dev_size;
+ unsigned int dev_sector_size;
+ int size_valid = 0;
+ int sector_size_valid = 0;
+
dev_t dev = makedev(d->disk.major,d->disk.minor);
- if (!min_size ||
+ if (!criteria->min_size ||
(dev_size_from_id(dev, &dev_size) &&
- dev_size >= min_size))
- found = 1;
+ dev_size >= criteria->min_size))
+ size_valid = 1;
+
+ if (!criteria->sector_size ||
+ (dev_sector_size_from_id(dev, &dev_sector_size) &&
+ criteria->sector_size == dev_sector_size))
+ sector_size_valid = 1;
+
+ found = size_valid && sector_size_valid;
+
/* check if domain matches */
if (found && domlist) {
struct dev_policy *pol = devid_policy(dev);
{
unsigned int fds = 20 + devices;
struct rlimit lim;
- if (getrlimit(RLIMIT_NOFILE, &lim) != 0
- || lim.rlim_cur >= fds)
+ if (getrlimit(RLIMIT_NOFILE, &lim) != 0 || lim.rlim_cur >= fds)
return;
if (lim.rlim_max < fds)
lim.rlim_max = fds;
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)
+{
+ 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.service", service_name, 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. */
if (fd >= 0 && fd != mdfd)
dup2(fd, mdfd);
}
+
+static struct cmap_hooks *cmap_hooks = NULL;
+static int is_cmap_hooks_ready = 0;
+
+void set_cmap_hooks(void)
+{
+ cmap_hooks = xmalloc(sizeof(struct cmap_hooks));
+ cmap_hooks->cmap_handle = dlopen("libcmap.so.4", RTLD_NOW | RTLD_LOCAL);
+ if (!cmap_hooks->cmap_handle)
+ return;
+
+ cmap_hooks->initialize =
+ dlsym(cmap_hooks->cmap_handle, "cmap_initialize");
+ cmap_hooks->get_string =
+ dlsym(cmap_hooks->cmap_handle, "cmap_get_string");
+ cmap_hooks->finalize = dlsym(cmap_hooks->cmap_handle, "cmap_finalize");
+
+ if (!cmap_hooks->initialize || !cmap_hooks->get_string ||
+ !cmap_hooks->finalize)
+ dlclose(cmap_hooks->cmap_handle);
+ else
+ is_cmap_hooks_ready = 1;
+}
+
+int get_cluster_name(char **cluster_name)
+{
+ int rv = -1;
+ cmap_handle_t handle;
+
+ if (!is_cmap_hooks_ready)
+ return rv;
+
+ rv = cmap_hooks->initialize(&handle);
+ if (rv != CS_OK)
+ goto out;
+
+ rv = cmap_hooks->get_string(handle, "totem.cluster_name", cluster_name);
+ if (rv != CS_OK) {
+ free(*cluster_name);
+ rv = -1;
+ goto name_err;
+ }
+
+ rv = 0;
+name_err:
+ cmap_hooks->finalize(handle);
+out:
+ return rv;
+}
+
+void set_dlm_hooks(void)
+{
+ dlm_hooks = xmalloc(sizeof(struct dlm_hooks));
+ dlm_hooks->dlm_handle = dlopen("libdlm_lt.so.3", RTLD_NOW | RTLD_LOCAL);
+ if (!dlm_hooks->dlm_handle)
+ return;
+
+ dlm_hooks->open_lockspace =
+ dlsym(dlm_hooks->dlm_handle, "dlm_open_lockspace");
+ dlm_hooks->create_lockspace =
+ dlsym(dlm_hooks->dlm_handle, "dlm_create_lockspace");
+ dlm_hooks->release_lockspace =
+ dlsym(dlm_hooks->dlm_handle, "dlm_release_lockspace");
+ dlm_hooks->ls_lock = dlsym(dlm_hooks->dlm_handle, "dlm_ls_lock");
+ dlm_hooks->ls_unlock_wait =
+ dlsym(dlm_hooks->dlm_handle, "dlm_ls_unlock_wait");
+ dlm_hooks->ls_get_fd = dlsym(dlm_hooks->dlm_handle, "dlm_ls_get_fd");
+ dlm_hooks->dispatch = dlsym(dlm_hooks->dlm_handle, "dlm_dispatch");
+
+ if (!dlm_hooks->open_lockspace || !dlm_hooks->create_lockspace ||
+ !dlm_hooks->ls_lock || !dlm_hooks->ls_unlock_wait ||
+ !dlm_hooks->release_lockspace || !dlm_hooks->ls_get_fd ||
+ !dlm_hooks->dispatch)
+ dlclose(dlm_hooks->dlm_handle);
+ else
+ is_dlm_hooks_ready = 1;
+}
+
+void set_hooks(void)
+{
+ set_dlm_hooks();
+ set_cmap_hooks();
+}
+
+int zero_disk_range(int fd, unsigned long long sector, size_t count)
+{
+ int ret = 0;
+ int fd_zero;
+ void *addr = NULL;
+ size_t written = 0;
+ size_t len = count * 512;
+ ssize_t n;
+
+ fd_zero = open("/dev/zero", O_RDONLY);
+ if (fd_zero < 0) {
+ pr_err("Cannot open /dev/zero\n");
+ return -1;
+ }
+
+ if (lseek64(fd, sector * 512, SEEK_SET) < 0) {
+ ret = -errno;
+ pr_err("Failed to seek offset for zeroing\n");
+ goto out;
+ }
+
+ addr = mmap(NULL, len, PROT_READ, MAP_PRIVATE, fd_zero, 0);
+
+ if (addr == MAP_FAILED) {
+ ret = -errno;
+ pr_err("Mapping /dev/zero failed\n");
+ goto out;
+ }
+
+ do {
+ n = write(fd, addr + written, len - written);
+ if (n < 0) {
+ if (errno == EINTR)
+ continue;
+ ret = -errno;
+ pr_err("Zeroing disk range failed\n");
+ break;
+ }
+ written += n;
+ } while (written != len);
+
+ munmap(addr, len);
+
+out:
+ close(fd_zero);
+ return ret;
+}