if (md_get_version(fd) < 9000) {
if (ioctl(fd, STOP_MD, 0) == 0)
return 0;
- pr_err("stopping device %s "
- "failed: %s\n",
+ pr_err("stopping device %s failed: %s\n",
devname, strerror(errno));
return 1;
}
- /* If this is an mdmon managed array, just write 'inactive'
- * to the array state and let mdmon clear up.
- */
strcpy(devnm, fd2devnm(fd));
/* Get EXCL access first. If this fails, then attempting
* to stop is probably a bad idea.
container[0] = 0;
close(fd);
count = 5;
- while (((fd = ((devnm[0] == '/')
+ while (((fd = ((devname[0] == '/')
?open(devname, O_RDONLY|O_EXCL)
:open_dev_flags(devnm, O_RDONLY|O_EXCL))) < 0
|| strcmp(fd2devnm(fd), devnm) != 0)
&& container[0]
&& mdmon_running(container)
&& count) {
+ /* Can't open, so something might be wrong. However it
+ * is a container, so we might be racing with mdmon, so
+ * retry for a bit.
+ */
if (fd >= 0)
close(fd);
flush_mdmon(container);
if (fd >= 0)
close(fd);
if (verbose >= 0)
- pr_err("Cannot get exclusive access to %s:"
- "Perhaps a running "
- "process, mounted filesystem "
- "or active volume group?\n",
+ pr_err("Cannot get exclusive access to %s:Perhaps a running process, mounted filesystem or active volume group?\n",
devname);
return 1;
}
+ /* If this is an mdmon managed array, just write 'inactive'
+ * to the array state and let mdmon clear up.
+ */
if (mdi &&
mdi->array.level > 0 &&
is_subarray(mdi->text_version)) {
/* This is mdmon managed. */
close(fd);
- /* As we have an O_EXCL open, any use of the device
+ /* As we had an O_EXCL open, any use of the device
* which blocks STOP_ARRAY is probably a transient use,
* so it is reasonable to retry for a while - 5 seconds.
*/
fd = open_dev_excl(devnm);
if (fd < 0) {
if (verbose >= 0)
- pr_err("failed to completely stop %s"
- ": Device is busy\n",
+ pr_err("failed to completely stop %s: Device is busy\n",
devname);
rv = 1;
goto out;
metadata_container_matches(m->metadata_version+9,
devnm)) {
if (verbose >= 0)
- pr_err("Cannot stop container %s: "
- "member %s still active\n",
- devname, m->dev);
+ pr_err("Cannot stop container %s: member %s still active\n",
+ devname, m->devnm);
free_mdstat(mds);
rv = 1;
goto out;
sysfs_attribute_available(mdi, NULL, "reshape_direction") &&
sysfs_get_str(mdi, NULL, "sync_action", buf, 20) > 0 &&
strcmp(buf, "reshape\n") == 0 &&
- sysfs_get_two(mdi, NULL, "raid_disks", &rd1, &rd2) == 2 &&
- sysfs_set_str(mdi, NULL, "sync_action", "frozen") == 0) {
- /* Array is frozen */
+ sysfs_get_two(mdi, NULL, "raid_disks", &rd1, &rd2) == 2) {
unsigned long long position, curr;
unsigned long long chunk1, chunk2;
unsigned long long rddiv, chunkdiv;
int delay;
int scfd;
+ delay = 40;
+ while (rd1 > rd2 && delay > 0 &&
+ sysfs_get_ll(mdi, NULL, "sync_max", &old_sync_max) == 0) {
+ /* must be in the critical section - wait a bit */
+ delay -= 1;
+ usleep(100000);
+ }
+
+ if (sysfs_set_str(mdi, NULL, "sync_action", "frozen") != 0)
+ goto done;
+ /* Array is frozen */
+
rd1 -= mdi->array.level == 6 ? 2 : 1;
rd2 -= mdi->array.level == 6 ? 2 : 1;
sysfs_get_str(mdi, NULL, "reshape_direction", buf, sizeof(buf));
if (strncmp(buf, "back", 4) == 0)
backwards = 1;
- sysfs_get_ll(mdi, NULL, "reshape_position", &position);
+ if (sysfs_get_ll(mdi, NULL, "reshape_position", &position) != 0) {
+ /* reshape must have finished now */
+ sysfs_set_str(mdi, NULL, "sync_action", "idle");
+ goto done;
+ }
sysfs_get_two(mdi, NULL, "chunk_size", &chunk1, &chunk2);
chunk1 /= 512;
chunk2 /= 512;
size &= ~(chunk1-1);
size &= ~(chunk2-1);
/* rd1 must be smaller */
+ /* Reshape may have progressed further backwards than
+ * recorded, so target even further back (hence "-1")
+ */
position = (position / sectors - 1) * sectors;
+ /* rd1 is always the conversion factor between 'sync'
+ * position and 'reshape' position.
+ * We read 1 "new" stripe worth of data from where-ever,
+ * and when write out that full stripe.
+ */
sync_max = size - position/rd1;
} else {
+ /* Reshape will very likely be beyond position, and it may
+ * be too late to stop at '+1', so aim for '+2'
+ */
position = (position / sectors + 2) * sectors;
sync_max = position/rd1;
}
delay = 3000;
scfd = sysfs_open(mdi->sys_name, NULL, "sync_completed");
while (scfd >= 0 && delay > 0 && old_sync_max > 0) {
+ unsigned long long max_completed;
sysfs_get_ll(mdi, NULL, "reshape_position", &curr);
sysfs_fd_get_str(scfd, buf, sizeof(buf));
if (strncmp(buf, "none", 4) == 0) {
break;
}
- if (sysfs_fd_get_ll(scfd, &completed) == 0 &&
+ if (sysfs_fd_get_two(scfd, &completed,
+ &max_completed) == 2 &&
+ /* 'completed' sometimes reads as max-uulong */
+ completed < max_completed &&
(completed > sync_max ||
(completed == sync_max && curr != position))) {
while (completed > sync_max) {
close(scfd);
}
+done:
/* As we have an O_EXCL open, any use of the device
* which blocks STOP_ARRAY is probably a transient use,
pr_err("failed to stop array %s: %s\n",
devname, strerror(errno));
if (errno == EBUSY)
- cont_err("Perhaps a running "
- "process, mounted filesystem "
- "or active volume group?\n");
+ cont_err("Perhaps a running process, mounted filesystem or active volume group?\n");
}
rv = 1;
goto out;
int rv = -1;
tfd = dev_open(dv->devname, O_RDWR);
if (tfd < 0) {
- pr_err("failed to open %s for"
- " superblock update during re-add\n", dv->devname);
+ pr_err("failed to open %s for superblock update during re-add\n", dv->devname);
return -1;
}
rv = dev_st->ss->store_super(dev_st, tfd);
close(tfd);
if (rv != 0) {
- pr_err("failed to update"
- " superblock during re-add\n");
+ pr_err("failed to update superblock during re-add\n");
return -1;
}
}
int Manage_add(int fd, int tfd, struct mddev_dev *dv,
struct supertype *tst, mdu_array_info_t *array,
int force, int verbose, char *devname,
- char *update, unsigned long rdev, unsigned long long array_size)
+ char *update, unsigned long rdev, unsigned long long array_size,
+ int raid_slot)
{
unsigned long long ldsize;
struct supertype *dev_st = NULL;
if (tst->ss == &super0 && ldsize > 4ULL*1024*1024*1024*1024) {
/* More than 4TB is wasted on v0.90 */
if (!force) {
- pr_err("%s is larger than %s can "
- "effectively use.\n"
- " Add --force is you "
- "really want to add this device.\n",
+ pr_err("%s is larger than %s can effectively use.\n"
+ " Add --force is you really want to add this device.\n",
dv->devname, devname);
return -1;
}
- pr_err("%s is larger than %s can "
- "effectively use.\n"
- " Adding anyway as --force "
- "was given.\n",
+ pr_err("%s is larger than %s can effectively use.\n"
+ " Adding anyway as --force was given.\n",
dv->devname, devname);
}
if (!tst->ss->external &&
}
/* Make sure device is large enough */
- if (tst->ss->avail_size(tst, ldsize/512, INVALID_SECTORS) <
+ if (tst->sb &&
+ tst->ss->avail_size(tst, ldsize/512, INVALID_SECTORS) <
array_size) {
if (dv->disposition == 'M')
return 0;
int d;
int found = 0;
- for (d = 0; d < MAX_DISKS && found < array->active_disks; d++) {
+ for (d = 0; d < MAX_DISKS && found < array->nr_disks; d++) {
disc.number = d;
if (ioctl(fd, GET_DISK_INFO, &disc))
continue;
if (disc.major == 0 && disc.minor == 0)
continue;
+ found++;
if (!(disc.state & (1<<MD_DISK_SYNC)))
continue;
avail[disc.raid_disk] = 1;
- found++;
}
array_failed = !enough(array->level, array->raid_disks,
array->layout, 1, avail);
}
disc.major = major(rdev);
disc.minor = minor(rdev);
- disc.number =j;
+ if (raid_slot < 0)
+ disc.number = j;
+ else
+ disc.number = raid_slot;
disc.state = 0;
if (array->not_persistent==0) {
int dfd;
}
free(used);
}
+
+ if (array->state & (1 << MD_SB_CLUSTERED)) {
+ if (dv->disposition == 'c')
+ disc.state |= (1 << MD_DISK_CANDIDATE);
+ else
+ disc.state |= (1 << MD_DISK_CLUSTER_ADD);
+ }
+
if (dv->writemostly == 1)
disc.state |= (1 << MD_DISK_WRITEMOSTLY);
if (tst->ss->external) {
container_fd = open_dev_excl(devnm);
if (container_fd < 0) {
- pr_err("add failed for %s:"
- " could not get exclusive access to container\n",
+ pr_err("add failed for %s: could not get exclusive access to container\n",
dv->devname);
tst->ss->free_super(tst);
return -1;
* would block add_disk */
tst->ss->free_super(tst);
if (sysfs_add_disk(sra, &new_mdi, 0) != 0) {
- pr_err("add new device to external metadata"
- " failed for %s\n", dv->devname);
+ pr_err("add new device to external metadata failed for %s\n", dv->devname);
close(container_fd);
sysfs_free(sra);
return -1;
strcpy(devnm, fd2devnm(fd));
lfd = open_dev_excl(devnm);
if (lfd < 0) {
- pr_err("Cannot get exclusive access "
- " to container - odd\n");
+ pr_err("Cannot get exclusive access to container - odd\n");
return -1;
}
/* We may not be able to check on holders in
}
}
if (err) {
- pr_err("hot remove failed "
- "for %s: %s\n", dv->devname,
+ pr_err("hot remove failed for %s: %s\n", dv->devname,
strerror(errno));
if (lfd >= 0)
close(lfd);
* variant on 'A'
* 'F' - Another variant of 'A', where the device was faulty
* so must be removed from the array first.
+ * 'c' - confirm the device as found (for clustered environments)
*
* For 'f' and 'r', the device can also be a kernel-internal
* name such as 'sdb'.
mdu_array_info_t array;
unsigned long long array_size;
struct mddev_dev *dv;
- struct stat stb;
int tfd = -1;
struct supertype *tst;
char *subarray = NULL;
struct mdinfo info;
int frozen = 0;
int busy = 0;
+ int raid_slot = -1;
if (ioctl(fd, GET_ARRAY_INFO, &array)) {
pr_err("Cannot get array info for %s\n",
goto abort;
}
- stb.st_rdev = 0;
for (dv = devlist; dv; dv = dv->next) {
+ unsigned long rdev = 0; /* device to add/remove etc */
int rv;
+ int mj,mn;
+
+ raid_slot = -1;
+ if (dv->disposition == 'c') {
+ rv = parse_cluster_confirm_arg(dv->devname,
+ &dv->devname,
+ &raid_slot);
+ if (rv) {
+ pr_err("Could not get the devname of cluster\n");
+ goto abort;
+ }
+ }
if (strcmp(dv->devname, "failed") == 0 ||
strcmp(dv->devname, "faulty") == 0) {
if (dv->disposition != 'A'
&& dv->disposition != 'r') {
- pr_err("%s only meaningful "
- "with -r or --re-add, not -%c\n",
+ pr_err("%s only meaningful with -r or --re-add, not -%c\n",
dv->devname, dv->disposition);
goto abort;
}
}
if (strcmp(dv->devname, "detached") == 0) {
if (dv->disposition != 'r' && dv->disposition != 'f') {
- pr_err("%s only meaningful "
- "with -r of -f, not -%c\n",
+ pr_err("%s only meaningful with -r of -f, not -%c\n",
dv->devname, dv->disposition);
goto abort;
}
if (strcmp(dv->devname, "missing") == 0) {
struct mddev_dev *add_devlist = NULL;
struct mddev_dev **dp;
+ if (dv->disposition == 'c') {
+ rv = ioctl(fd, CLUSTERED_DISK_NACK, NULL);
+ break;
+ }
+
if (dv->disposition != 'A') {
- pr_err("'missing' only meaningful "
- "with --re-add\n");
+ pr_err("'missing' only meaningful with --re-add\n");
goto abort;
}
add_devlist = conf_get_devs();
int found = 0;
char dname[55];
if (dv->disposition != 'r' && dv->disposition != 'f') {
- pr_err("%s only meaningful "
- "with -r or -f, not -%c\n",
+ pr_err("%s only meaningful with -r or -f, not -%c\n",
dv->devname, dv->disposition);
goto abort;
}
sysfd = sysfs_open(fd2devnm(fd), dname, "block/dev");
if (sysfd >= 0) {
char dn[20];
- int mj,mn;
if (sysfs_fd_get_str(sysfd, dn, 20) > 0 &&
sscanf(dn, "%d:%d", &mj,&mn) == 2) {
- stb.st_rdev = makedev(mj,mn);
+ rdev = makedev(mj,mn);
found = 1;
}
close(sysfd);
if (!found) {
sysfd = sysfs_open(fd2devnm(fd), dname, "state");
if (sysfd < 0) {
- pr_err("%s does not appear "
- "to be a component of %s\n",
+ pr_err("%s does not appear to be a component of %s\n",
dv->devname, devname);
goto abort;
}
}
+ } else if ((dv->disposition == 'r' || dv->disposition == 'f')
+ && get_maj_min(dv->devname, &mj, &mn)) {
+ /* for 'fail' and 'remove', the device might
+ * not exist.
+ */
+ rdev = makedev(mj, mn);
} else {
+ struct stat stb;
tfd = dev_open(dv->devname, O_RDONLY);
if (tfd >= 0)
fstat(tfd, &stb);
goto abort;
}
}
+ rdev = stb.st_rdev;
}
switch(dv->disposition){
default:
case 'A':
case 'M': /* --re-add missing */
case 'F': /* --re-add faulty */
+ case 'c': /* --cluster-confirm */
/* add the device */
if (subarray) {
- pr_err("Cannot add disks to a"
- " \'member\' array, perform this"
- " operation on the parent container\n");
+ pr_err("Cannot add disks to a \'member\' array, perform this operation on the parent container\n");
goto abort;
}
if (dv->disposition == 'F')
/* Need to remove first */
- ioctl(fd, HOT_REMOVE_DISK,
- (unsigned long)stb.st_rdev);
+ ioctl(fd, HOT_REMOVE_DISK, rdev);
/* Make sure it isn't in use (in 2.6 or later) */
tfd = dev_open(dv->devname, O_RDONLY|O_EXCL);
if (tfd >= 0) {
}
rv = Manage_add(fd, tfd, dv, tst, &array,
force, verbose, devname, update,
- stb.st_rdev, array_size);
+ rdev, array_size, raid_slot);
close(tfd);
tfd = -1;
if (rv < 0)
case 'r':
/* hot remove */
if (subarray) {
- pr_err("Cannot remove disks from a"
- " \'member\' array, perform this"
- " operation on the parent container\n");
+ pr_err("Cannot remove disks from a \'member\' array, perform this operation on the parent container\n");
rv = -1;
} else
rv = Manage_remove(tst, fd, dv, sysfd,
- stb.st_rdev, verbose,
+ rdev, verbose,
devname);
if (sysfd >= 0)
close(sysfd);
/* FIXME check current member */
if ((sysfd >= 0 && write(sysfd, "faulty", 6) != 6) ||
(sysfd < 0 && ioctl(fd, SET_DISK_FAULTY,
- (unsigned long) stb.st_rdev))) {
+ rdev))) {
if (errno == EBUSY)
busy = 1;
pr_err("set device faulty failed for %s: %s\n",
break;
case 'R': /* Mark as replaceable */
if (subarray) {
- pr_err("Cannot replace disks in a"
- " \'member\' array, perform this"
- " operation on the parent container\n");
+ pr_err("Cannot replace disks in a \'member\' array, perform this operation on the parent container\n");
rv = -1;
} else {
if (!frozen) {
frozen = -1;
}
rv = Manage_replace(tst, fd, dv,
- stb.st_rdev, verbose,
+ rdev, verbose,
devname);
}
if (rv < 0)
goto abort;
case 'w': /* --with device which was matched */
rv = Manage_with(tst, fd, dv,
- stb.st_rdev, verbose, devname);
+ rdev, verbose, devname);
if (rv < 0)
goto abort;
break;
projects / thirdparty / mdadm.git / blobdiff