char volname[BLKPG_VOLNAMELTH]; /* volume label */
};
-/* partition table structures so we can check metadata position
- * against the end of the last partition.
- * Only handle MBR ant GPT partition tables.
- */
-struct MBR_part_record {
- __u8 bootable;
- __u8 first_head;
- __u8 first_sector;
- __u8 first_cyl;
- __u8 part_type;
- __u8 last_head;
- __u8 last_sector;
- __u8 last_cyl;
- __u32 first_sect_lba;
- __u32 blocks_num;
-};
-
-struct MBR {
- __u8 pad[446];
- struct MBR_part_record parts[4];
- __u16 magic;
-} __attribute__((packed));
-
-struct GPT_part_entry {
- unsigned char type_guid[16];
- unsigned char partition_guid[16];
- __u64 starting_lba;
- __u64 ending_lba;
- unsigned char attr_bits[8];
- unsigned char name[72];
-} __attribute__((packed));
-
-struct GPT {
- __u64 magic;
- __u32 revision;
- __u32 header_size;
- __u32 crc;
- __u32 pad1;
- __u64 current_lba;
- __u64 backup_lba;
- __u64 first_lba;
- __u64 last_lba;
- __u8 guid[16];
- __u64 part_start;
- __u32 part_cnt;
- __u32 part_size;
- __u32 part_crc;
- __u8 pad2[420];
-} __attribute__((packed));
+#include "part.h"
/* Force a compilation error if condition is true */
#define BUILD_BUG_ON(condition) ((void)BUILD_BUG_ON_ZERO(condition))
aren't permitted). */
#define BUILD_BUG_ON_ZERO(e) (sizeof(struct { int:-!!(e); }))
-
-/* MBR/GPT magic numbers */
-#define MBR_SIGNATURE_MAGIC __cpu_to_le16(0xAA55)
-#define GPT_SIGNATURE_MAGIC __cpu_to_le64(0x5452415020494645ULL)
-
-#define MBR_PARTITIONS 4
-#define MBR_GPT_PARTITION_TYPE 0xEE
-
/*
* Parse a 128 bit uuid in 4 integers
* format is 32 hexx nibbles with options :.<space> separator
}
#ifndef MDASSEMBLE
+int mdadm_version(char *version)
+{
+ int a, b, c;
+ char *cp;
+
+ if (!version)
+ version = Version;
+
+ cp = strchr(version, '-');
+ if (!cp || *(cp+1) != ' ' || *(cp+2) != 'v')
+ return -1;
+ cp += 3;
+ a = strtoul(cp, &cp, 10);
+ if (*cp != '.')
+ return -1;
+ b = strtoul(cp+1, &cp, 10);
+ if (*cp == '.')
+ c = strtoul(cp+1, &cp, 10);
+ else
+ c = 0;
+ if (*cp != ' ' && *cp != '-')
+ return -1;
+ return (a*1000000)+(b*1000)+c;
+}
+
long long parse_size(char *size)
{
/* parse 'size' which should be a number optionally
return 1;
}
+int test_partition_from_id(dev_t id)
+{
+ char buf[20];
+ int fd, rv;
+
+ sprintf(buf, "%d:%d", major(id), minor(id));
+ fd = dev_open(buf, O_RDONLY);
+ if (fd < 0)
+ return -1;
+ rv = test_partition(fd);
+ close(fd);
+ return rv;
+}
int enough(int level, int raid_disks, int layout, int clean,
char *avail, int avail_disks)
}
}
-const int uuid_match_any[4] = { ~0, ~0, ~0, ~0 };
-int same_uuid(int a[4], int b[4], int swapuuid)
+int enough_fd(int fd)
{
- if (memcmp(a, uuid_match_any, sizeof(int[4])) == 0 ||
- memcmp(b, uuid_match_any, sizeof(int[4])) == 0)
- return 1;
+ struct mdu_array_info_s array;
+ struct mdu_disk_info_s disk;
+ int avail_disks = 0;
+ int i;
+ char *avail;
+
+ if (ioctl(fd, GET_ARRAY_INFO, &array) != 0 ||
+ array.raid_disks <= 0)
+ return 0;
+ avail = calloc(array.raid_disks, 1);
+ for (i=0; i<array.raid_disks + array.nr_disks; i++) {
+ disk.number = i;
+ if (ioctl(fd, GET_DISK_INFO, &disk) != 0)
+ continue;
+ if (! (disk.state & (1<<MD_DISK_SYNC)))
+ continue;
+ if (disk.raid_disk < 0 || disk.raid_disk >= array.raid_disks)
+ continue;
+ avail_disks++;
+ avail[disk.raid_disk] = 1;
+ }
+ /* This is used on an active array, so assume it is clean */
+ return enough(array.level, array.raid_disks, array.layout,
+ 1,
+ avail, avail_disks);
+}
+
+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
/* Looks like a raid array .. */
fprintf(stderr, Name ": %s appears to be part of a raid array:\n",
name);
- st->ss->getinfo_super(st, &info);
+ st->ss->getinfo_super(st, &info, NULL);
st->ss->free_super(st);
crtime = info.array.ctime;
level = map_num(pers, info.array.level);
int minor;
if (!dev) return -1;
+ flags |= O_DIRECT;
major = strtoul(dev, &e, 0);
if (e > dev && *e == ':' && e[1] &&
(minor = strtoul(e+1, &e, 0)) >= 0 &&
*e == 0) {
- snprintf(devname, sizeof(devname), "/dev/.tmp.md.%d:%d:%d",
- (int)getpid(), major, minor);
- if (mknod(devname, S_IFBLK|0600, makedev(major, minor))==0) {
- fd = open(devname, flags|O_DIRECT);
- unlink(devname);
+ char *path = map_dev(major, minor, 0);
+ if (path)
+ fd = open(path, flags);
+ if (fd < 0) {
+ snprintf(devname, sizeof(devname), "/dev/.tmp.md.%d:%d:%d",
+ (int)getpid(), major, minor);
+ if (mknod(devname, S_IFBLK|0600, makedev(major, minor))==0) {
+ fd = open(devname, flags);
+ unlink(devname);
+ }
+ }
+ if (fd < 0) {
+ snprintf(devname, sizeof(devname), "/tmp/.tmp.md.%d:%d:%d",
+ (int)getpid(), major, minor);
+ if (mknod(devname, S_IFBLK|0600, makedev(major, minor))==0) {
+ fd = open(devname, flags);
+ unlink(devname);
+ }
}
} else
- fd = open(dev, flags|O_DIRECT);
+ fd = open(dev, flags);
return fd;
}
dprintf("%s: timeout waiting for %s\n", __func__, dev);
}
-struct superswitch *superlist[] = { &super0, &super1, &super_ddf, &super_imsm, NULL };
+struct superswitch *superlist[] =
+{
+ &super0, &super1,
+ &super_ddf, &super_imsm,
+ &mbr, &gpt,
+ NULL };
#if !defined(MDASSEMBLE) || defined(MDASSEMBLE) && defined(MDASSEMBLE_AUTO)
-struct supertype *super_by_fd(int fd)
+struct supertype *super_by_fd(int fd, char **subarrayp)
{
mdu_array_info_t array;
int vers;
char version[20];
int i;
char *subarray = NULL;
+ int container = NoMdDev;
sra = sysfs_read(fd, 0, GET_VERSION);
}
if (minor == -2 && is_subarray(verstr)) {
char *dev = verstr+1;
+
subarray = strchr(dev, '/');
- int devnum;
if (subarray)
*subarray++ = '\0';
- devnum = devname2devnum(dev);
subarray = strdup(subarray);
+ container = devname2devnum(dev);
if (sra)
sysfs_free(sra);
- sra = sysfs_read(-1, devnum, GET_VERSION);
+ sra = sysfs_read(-1, container, GET_VERSION);
if (sra && sra->text_version[0])
verstr = sra->text_version;
else
sysfs_free(sra);
if (st) {
st->sb = NULL;
- if (subarray) {
- strncpy(st->subarray, subarray, 32);
- st->subarray[31] = 0;
- free(subarray);
- } else
- st->subarray[0] = 0;
- }
+ if (subarrayp)
+ *subarrayp = subarray;
+ st->container_dev = container;
+ st->devnum = fd2devnum(fd);
+ } else
+ free(subarray);
+
return st;
}
#endif /* !defined(MDASSEMBLE) || defined(MDASSEMBLE) && defined(MDASSEMBLE_AUTO) */
+int dev_size_from_id(dev_t id, unsigned long long *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_size(fd, NULL, size)) {
+ close(fd);
+ return 1;
+ }
+ close(fd);
+ return 0;
+}
struct supertype *dup_super(struct supertype *orig)
{
st->ss = orig->ss;
st->max_devs = orig->max_devs;
st->minor_version = orig->minor_version;
- strcpy(st->subarray, orig->subarray);
st->sb = NULL;
st->info = NULL;
return st;
}
-struct supertype *guess_super(int fd)
+struct supertype *guess_super_type(int fd, enum guess_types guess_type)
{
/* try each load_super to find the best match,
* and return the best superswitch
*/
struct superswitch *ss;
struct supertype *st;
- unsigned long besttime = 0;
+ time_t besttime = 0;
int bestsuper = -1;
int i;
st = malloc(sizeof(*st));
+ memset(st, 0, sizeof(*st));
+ st->container_dev = NoMdDev;
+
for (i=0 ; superlist[i]; i++) {
int rv;
ss = superlist[i];
+ if (guess_type == guess_array && ss->add_to_super == NULL)
+ continue;
+ if (guess_type == guess_partitions && ss->add_to_super != NULL)
+ continue;
memset(st, 0, sizeof(*st));
rv = ss->load_super(st, fd, NULL);
if (rv == 0) {
struct mdinfo info;
- st->ss->getinfo_super(st, &info);
+ st->ss->getinfo_super(st, &info, NULL);
if (bestsuper == -1 ||
besttime < info.array.ctime) {
bestsuper = i;
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)
+{
+ unsigned long long size;
+ if (md_get_version(fd) < 0)
+ return 0;
+ if (get_dev_size(fd, NULL, &size) == 0)
+ return 1;
+ if (size == 0)
+ return 1;
+ return 0;
+}
/* Sets endofpart parameter to the last block used by the last GPT partition on the device.
* Returns: 1 if successful
struct GPT_part_entry *part;
unsigned long long curr_part_end;
unsigned all_partitions, entry_size;
- int part_nr;
+ unsigned part_nr;
*endofpart = 0;
struct MBR boot_sect;
struct MBR_part_record *part;
unsigned long long curr_part_end;
- int part_nr;
+ unsigned part_nr;
int retval = 0;
*endofpart = 0;
continue;
n = read(dfd, buf, sizeof(buf));
close(dfd);
- if (n <= 0 || n >= sizeof(buf))
+ if (n <= 0 || (unsigned)n >= sizeof(buf))
continue;
buf[n] = 0;
if (sscanf(buf, "%d:%d", &major, &minor) != 2)
struct mdstat_ent *mdstat = mdstat_read(0, 0);
struct mdstat_ent *ent;
- for (ent = mdstat; ent; ent = ent->next) {
- if (is_container_member(ent, container)) {
- char *inst = &ent->metadata_version[10+strlen(container)+1];
-
- if (strcmp(inst, subarray) == 0)
+ for (ent = mdstat; ent; ent = ent->next)
+ if (is_container_member(ent, container))
+ if (strcmp(to_subarray(ent, container), subarray) == 0)
break;
- }
- }
free_mdstat(mdstat);
/* open_subarray - opens a subarray in a container
* @dev: container device name
- * @st: supertype with only ->subarray set
+ * @st: empty supertype
* @quiet: block reporting errors flag
*
* On success returns an fd to a container and fills in *st
*/
-int open_subarray(char *dev, struct supertype *st, int quiet)
+int open_subarray(char *dev, char *subarray, struct supertype *st, int quiet)
{
struct mdinfo *mdi;
+ struct mdinfo *info;
int fd, err = 1;
fd = open(dev, O_RDWR|O_EXCL);
goto free_sysfs;
}
- if (st->ss->load_super(st, fd, NULL)) {
+ if (!st->ss->load_container) {
if (!quiet)
- fprintf(stderr, Name ": Failed to find subarray-%s in %s\n",
- st->subarray, dev);
+ fprintf(stderr, Name ": %s is not a container\n", dev);
goto free_name;
}
- if (!st->loaded_container) {
+ if (st->ss->load_container(st, fd, NULL)) {
if (!quiet)
- fprintf(stderr, Name ": %s is not a container\n", dev);
+ fprintf(stderr, Name ": Failed to load metadata for %s\n",
+ dev);
+ goto free_name;
+ }
+
+ info = st->ss->container_content(st, subarray);
+ if (!info) {
+ if (!quiet)
+ fprintf(stderr, Name ": Failed to find subarray-%s in %s\n",
+ subarray, dev);
goto free_super;
}
+ free(info);
err = 0;
return rv;
}
+int remove_disk(int mdfd, struct supertype *st,
+ 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 set_array_info(int mdfd, struct supertype *st, struct mdinfo *info)
{
/* Initialise kernel's knowledge of array.
return recovery_start;
}
-char *devnum2devname(int num)
+void fmt_devname(char *name, int num)
{
- char name[100];
if (num >= 0)
sprintf(name, "md%d", num);
else
sprintf(name, "md_d%d", -1-num);
+}
+
+char *devnum2devname(int num)
+{
+ char name[100];
+ fmt_devname(name,num);
return strdup(name);
}
if ((S_IFMT & st->st_mode) == S_IFBLK) {
if (major(st->st_rdev) == MD_MAJOR)
return minor(st->st_rdev);
- else if (major(st->st_rdev) == get_mdp_major())
+ else if (major(st->st_rdev) == (unsigned)get_mdp_major())
return -1- (minor(st->st_rdev)>>MdpMinorShift);
/* must be an extended-minor partition. Look at the
link[n] = 0;
cp = strrchr(link, '/');
if (cp) *cp = 0;
- cp = strchr(link, '/');
+ cp = strrchr(link, '/');
if (cp && strncmp(cp, "/md", 3) == 0)
return devname2devnum(cp+1);
}
return NoMdDev;
}
-char *pid_dir = VAR_RUN;
-
int mdmon_pid(int devnum)
{
char path[100];
int n;
char *devname = devnum2devname(devnum);
- sprintf(path, "%s/%s.pid", pid_dir, devname);
+ sprintf(path, "%s/%s.pid", MDMON_DIR, devname);
free(devname);
fd = open(path, O_RDONLY | O_NOATIME, 0);
mu->buf = buf;
mu->len = len;
mu->space = NULL;
+ mu->space_list = NULL;
mu->next = NULL;
*st->update_tail = mu;
st->update_tail = &mu->next;
unsigned int __invalid_size_argument_for_IOC = 0;
#endif
+int experimental(void)
+{
+ if (check_env("MDADM_EXPERIMENTAL"))
+ return 1;
+ else {
+ fprintf(stderr, Name ": To use this feature MDADM_EXPERIMENTAL enviroment variable has to 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 domainlist *domlist,
+ char *spare_group,
+ const char *metadata, int get_one)
+{
+ struct mdinfo *d, **dp, *disks = NULL;
+
+ /* get list of all disks in container */
+ if (st->ss->getinfo_super_disks)
+ disks = st->ss->getinfo_super_disks(st);
+
+ if (!disks)
+ return disks;
+ /* find spare devices on the list */
+ dp = &disks->devs;
+ disks->array.spare_disks = 0;
+ while (*dp) {
+ int found = 0;
+ d = *dp;
+ if (d->disk.state == 0) {
+ /* check if size is acceptable */
+ unsigned long long dev_size;
+ dev_t dev = makedev(d->disk.major,d->disk.minor);
+
+ if (!min_size ||
+ (dev_size_from_id(dev, &dev_size) &&
+ dev_size >= min_size))
+ found = 1;
+ /* check if domain matches */
+ if (found && domlist) {
+ struct dev_policy *pol = devnum_policy(dev);
+ if (spare_group)
+ pol_add(&pol, pol_domain,
+ spare_group, NULL);
+ if (domain_test(domlist, pol, metadata) != 1)
+ found = 0;
+ dev_policy_free(pol);
+ }
+ }
+ if (found) {
+ dp = &d->next;
+ disks->array.spare_disks++;
+ if (get_one) {
+ sysfs_free(*dp);
+ d->next = NULL;
+ }
+ } else {
+ *dp = d->next;
+ d->next = NULL;
+ sysfs_free(d);
+ }
+ }
+ return disks;
+}