/*
* mdadm - manage Linux "md" devices aka RAID arrays.
*
- * Copyright (C) 2001-2006 Neil Brown <neilb@suse.de>
+ * Copyright (C) 2001-2009 Neil Brown <neilb@suse.de>
*
*
* This program is free software; you can redistribute it and/or modify
* Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
*
* Author: Neil Brown
- * Email: <neilb@cse.unsw.edu.au>
- * Paper: Neil Brown
- * School of Computer Science and Engineering
- * The University of New South Wales
- * Sydney, 2052
- * Australia
+ * Email: <neilb@suse.de>
*/
#include "mdadm.h"
#include "md_p.h"
+#include <sys/socket.h>
#include <sys/utsname.h>
+#include <sys/wait.h>
+#include <sys/un.h>
#include <ctype.h>
+#include <dirent.h>
+#include <signal.h>
+
+int __offroot;
/*
* following taken from linux/blkpg.h because they aren't
char volname[BLKPG_VOLNAMELTH]; /* volume label */
};
+#include "part.h"
+
+/* Force a compilation error if condition is true */
+#define BUILD_BUG_ON(condition) ((void)BUILD_BUG_ON_ZERO(condition))
+
+/* Force a compilation error if condition is true, but also produce a
+ result (of value 0 and type size_t), so the expression can be used
+ e.g. in a structure initializer (or where-ever else comma expressions
+ 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
{
struct utsname name;
char *cp;
- int a,b,c;
+ int a = 0, b = 0,c = 0;
if (uname(&name) <0)
return -1;
cp = name.release;
a = strtoul(cp, &cp, 10);
- if (*cp != '.') return -1;
- b = strtoul(cp+1, &cp, 10);
- if (*cp != '.') return -1;
- c = strtoul(cp+1, NULL, 10);
+ if (*cp == '.')
+ b = strtoul(cp+1, &cp, 10);
+ if (*cp == '.')
+ c = strtoul(cp+1, &cp, 10);
+
+ return (a*1000000)+(b*1000)+c;
+}
+#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
+ * followed by 'K', 'M', or 'G'.
+ * Without a suffix, K is assumed.
+ * Number returned is in sectors (half-K)
+ */
+ char *c;
+ long long s = strtoll(size, &c, 10);
+ if (s > 0) {
+ switch (*c) {
+ case 'K':
+ c++;
+ default:
+ s *= 2;
+ break;
+ case 'M':
+ c++;
+ s *= 1024 * 2;
+ break;
+ case 'G':
+ c++;
+ s *= 1024 * 1024 * 2;
+ break;
+ }
+ }
+ if (*c)
+ s = 0;
+ return s;
+}
+
+int parse_layout_10(char *layout)
+{
+ int copies, rv;
+ char *cp;
+ /* Parse the layout string for raid10 */
+ /* 'f', 'o' or 'n' followed by a number <= raid_disks */
+ if ((layout[0] != 'n' && layout[0] != 'f' && layout[0] != 'o') ||
+ (copies = strtoul(layout+1, &cp, 10)) < 1 ||
+ copies > 200 ||
+ *cp)
+ return -1;
+ if (layout[0] == 'n')
+ rv = 256 + copies;
+ else if (layout[0] == 'o')
+ rv = 0x10000 + (copies<<8) + 1;
+ else
+ rv = 1 + (copies<<8);
+ return rv;
+}
+
+int parse_layout_faulty(char *layout)
+{
+ /* Parse the layout string for 'faulty' */
+ int ln = strcspn(layout, "0123456789");
+ char *m = strdup(layout);
+ int mode;
+ m[ln] = 0;
+ mode = map_name(faultylayout, m);
+ if (mode == UnSet)
+ return -1;
+
+ return mode | (atoi(layout+ln)<< ModeShift);
+}
+#endif
+
void remove_partitions(int fd)
{
/* remove partitions from this block devices.
#endif
}
+int test_partition(int fd)
+{
+ /* Check if fd is a whole-disk or a partition.
+ * BLKPG will return EINVAL on a partition, and BLKPG_DEL_PARTITION
+ * will return ENXIO on an invalid partition number.
+ */
+ struct blkpg_ioctl_arg a;
+ struct blkpg_partition p;
+ a.op = BLKPG_DEL_PARTITION;
+ a.data = (void*)&p;
+ a.datalen = sizeof(p);
+ a.flags = 0;
+ memset(a.data, 0, a.datalen);
+ p.pno = 1<<30;
+ if (ioctl(fd, BLKPG, &a) == 0)
+ /* Very unlikely, but not a partition */
+ return 0;
+ if (errno == ENXIO)
+ /* not a partition */
+ return 0;
+
+ 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)
{
} while (first != 0);
return 1;
- case -4:
+ case LEVEL_MULTIPATH:
return avail_disks>= 1;
- case -1:
+ case LEVEL_LINEAR:
case 0:
return avail_disks == raid_disks;
case 1:
}
}
+int enough_fd(int fd)
+{
+ struct mdu_array_info_s array;
+ struct mdu_disk_info_s disk;
+ int avail_disks = 0;
+ int i, rv;
+ 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 < 1024 && 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_disks++;
+ 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, avail_disks);
+ 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) {
memcpy(a, b, 16);
}
+char *__fname_from_uuid(int id[4], int swap, char *buf, char sep)
+{
+ int i, j;
+ char uuid[16];
+ char *c = buf;
+ strcpy(c, "UUID-");
+ c += strlen(c);
+ copy_uuid(uuid, id, swap);
+ for (i = 0; i < 4; i++) {
+ if (i)
+ *c++ = sep;
+ for (j = 3; j >= 0; j--) {
+ sprintf(c,"%02x", (unsigned char) uuid[j+4*i]);
+ c+= 2;
+ }
+ }
+ return buf;
+
+}
+
+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);
+}
+
#ifndef MDASSEMBLE
int check_ext2(int fd, char *name)
{
struct supertype *st = guess_super(fd);
if (!st) return 0;
+ st->ignore_hw_compat = 1;
st->ss->load_super(st, fd, name);
/* 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);
}
#endif /* MDASSEMBLE */
-char *map_num(mapping_t *map, int num)
-{
- while (map->name) {
- if (map->num == num)
- return map->name;
- map++;
- }
- return NULL;
-}
-
-int map_name(mapping_t *map, char *name)
-{
- while (map->name) {
- if (strcmp(map->name, name)==0)
- return map->num;
- map++;
- }
- return UnSet;
-}
-
-
int is_standard(char *dev, int *nump)
{
/* tests if dev is a "standard" md dev name.
* i.e if the last component is "/dNN" or "/mdNN",
* where NN is a string of digits
+ * Returns 1 if a partitionable standard,
+ * -1 if non-partitonable,
+ * 0 if not a standard name.
*/
char *d = strrchr(dev, '/');
int type=0;
if (strncmp(d, "/d",2)==0)
d += 2, type=1; /* /dev/md/dN{pM} */
else if (strncmp(d, "/md_d", 5)==0)
- d += 5, type=1; /* /dev/md_dNpM */
+ d += 5, type=1; /* /dev/md_dN{pM} */
else if (strncmp(d, "/md", 3)==0)
d += 3, type=-1; /* /dev/mdN */
else if (d-dev > 3 && strncmp(d-2, "md/", 3)==0)
return type;
}
-
-/*
- * convert a major/minor pair for a block device into a name in /dev, if possible.
- * On the first call, walk /dev collecting name.
- * Put them in a simple linked listfor now.
- */
-struct devmap {
- int major, minor;
- char *name;
- struct devmap *next;
-} *devlist = NULL;
-int devlist_ready = 0;
-
-int add_dev(const char *name, const struct stat *stb, int flag, struct FTW *s)
-{
- struct stat st;
- if (S_ISLNK(stb->st_mode)) {
- stat(name, &st);
- stb = &st;
- }
-
- if ((stb->st_mode&S_IFMT)== S_IFBLK) {
- char *n = strdup(name);
- struct devmap *dm = malloc(sizeof(*dm));
- if (strncmp(n, "/dev/./", 7)==0)
- strcpy(n+4, name+6);
- if (dm) {
- dm->major = major(stb->st_rdev);
- dm->minor = minor(stb->st_rdev);
- dm->name = n;
- dm->next = devlist;
- devlist = dm;
- }
- }
- return 0;
-}
-
-#ifndef HAVE_NFTW
-#ifdef HAVE_FTW
-int add_dev_1(const char *name, const struct stat *stb, int flag)
-{
- return add_dev(name, stb, flag, NULL);
-}
-int nftw(const char *path, int (*han)(const char *name, const struct stat *stb, int flag, struct FTW *s), int nopenfd, int flags)
-{
- return ftw(path, add_dev_1, nopenfd);
-}
-#else
-int nftw(const char *path, int (*han)(const char *name, const struct stat *stb, int flag, struct FTW *s), int nopenfd, int flags)
-{
- return 0;
-}
-#endif /* HAVE_FTW */
-#endif /* HAVE_NFTW */
-
-/*
- * Find a block device with the right major/minor number.
- * If we find multiple names, choose the shortest.
- * If we find a non-standard name, it is probably there
- * deliberately so prefer it over a standard name.
- * This applies only to names for MD devices.
- */
-char *map_dev(int major, int minor, int create)
-{
- struct devmap *p;
- char *std = NULL, *nonstd=NULL;
- int did_check = 0;
-
- if (major == 0 && minor == 0)
- return NULL;
-
- retry:
- if (!devlist_ready) {
- char *dev = "/dev";
- struct stat stb;
- while(devlist) {
- struct devmap *d = devlist;
- devlist = d->next;
- free(d->name);
- free(d);
- }
- if (lstat(dev, &stb)==0 &&
- S_ISLNK(stb.st_mode))
- dev = "/dev/.";
- nftw(dev, add_dev, 10, FTW_PHYS);
- devlist_ready=1;
- did_check = 1;
- }
-
- for (p=devlist; p; p=p->next)
- if (p->major == major &&
- p->minor == minor) {
- if (is_standard(p->name, NULL)) {
- if (std == NULL ||
- strlen(p->name) < strlen(std))
- std = p->name;
- } else {
- if (nonstd == NULL ||
- strlen(p->name) < strlen(nonstd))
- nonstd = p->name;
- }
- }
- if (!std && !nonstd && !did_check) {
- devlist_ready = 0;
- goto retry;
- }
- if (create && !std && !nonstd) {
- static char buf[30];
- snprintf(buf, sizeof(buf), "%d:%d", major, minor);
- nonstd = buf;
- }
-
- return nonstd ? nonstd : std;
-}
-
unsigned long calc_csum(void *super, int bytes)
{
unsigned long long newcsum = 0;
* We allow upto 2048Megabytes before converting to
* gigabytes, as that shows more precision and isn't
* too large a number.
- * Terrabytes are not yet handled.
+ * Terabytes are not yet handled.
*/
if (bytes < 5000*1024)
);
return buf;
}
-#endif
-#if !defined(MDASSEMBLE) || defined(MDASSEMBLE) && defined(MDASSEMBLE_AUTO)
-int get_mdp_major(void)
-{
-static int mdp_major = -1;
- FILE *fl;
- char *w;
- int have_block = 0;
- int have_devices = 0;
- int last_num = -1;
-
- if (mdp_major != -1)
- return mdp_major;
- fl = fopen("/proc/devices", "r");
- if (!fl)
- return -1;
- while ((w = conf_word(fl, 1))) {
- if (have_block && strcmp(w, "devices:")==0)
- have_devices = 1;
- have_block = (strcmp(w, "Block")==0);
- if (isdigit(w[0]))
- last_num = atoi(w);
- if (have_devices && strcmp(w, "mdp")==0)
- mdp_major = last_num;
- free(w);
+void print_r10_layout(int layout)
+{
+ int near = layout & 255;
+ int far = (layout >> 8) & 255;
+ int offset = (layout&0x10000);
+ char *sep = "";
+
+ if (near != 1) {
+ printf("%s near=%d", sep, near);
+ sep = ",";
}
- fclose(fl);
- return mdp_major;
+ if (far != 1)
+ printf("%s %s=%d", sep, offset?"offset":"far", far);
+ 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)
+{
+ devsize &= ~(unsigned long long)((chunksize>>9)-1);
+ return get_data_disks(level, layout, raid_disks) * devsize;
+}
+
+int get_data_disks(int level, int layout, int raid_disks)
+{
+ int data_disks = 0;
+ switch (level) {
+ case 0: data_disks = raid_disks; break;
+ case 1: data_disks = 1; break;
+ case 4:
+ case 5: data_disks = raid_disks - 1; break;
+ case 6: data_disks = raid_disks - 2; break;
+ case 10: data_disks = raid_disks / (layout & 255) / ((layout>>8)&255);
+ break;
+ }
+ return data_disks;
+}
+#if !defined(MDASSEMBLE) || defined(MDASSEMBLE) && defined(MDASSEMBLE_AUTO)
char *get_md_name(int dev)
{
/* find /dev/md%d or /dev/md/%d or make a device /dev/.tmp.md%d */
if (strncmp(name, "/dev/.tmp.md", 12)==0)
unlink(name);
}
+
+int find_free_devnum(int use_partitions)
+{
+ int devnum;
+ for (devnum = 127; devnum != 128;
+ devnum = devnum ? devnum-1 : (1<<20)-1) {
+ char *dn;
+ int _devnum;
+
+ _devnum = use_partitions ? (-1-devnum) : devnum;
+ if (mddev_busy(_devnum))
+ continue;
+ /* make sure it is new to /dev too, at least as a
+ * non-standard */
+ dn = map_dev(dev2major(_devnum), dev2minor(_devnum), 0);
+ if (dn && ! is_standard(dn, NULL))
+ continue;
+ break;
+ }
+ if (devnum == 128)
+ return NoMdDev;
+ return use_partitions ? (-1-devnum) : devnum;
+}
#endif /* !defined(MDASSEMBLE) || defined(MDASSEMBLE) && defined(MDASSEMBLE_AUTO) */
int dev_open(char *dev, int flags)
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", major, minor);
- if (mknod(devname, S_IFBLK|0600, makedev(major, minor))==0) {
- fd = open(devname, flags);
- 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);
return fd;
}
-struct superswitch *superlist[] = { &super0, &super1, NULL };
+int open_dev_flags(int devnum, int flags)
+{
+ char buf[20];
+
+ sprintf(buf, "%d:%d", dev2major(devnum), dev2minor(devnum));
+ return dev_open(buf, flags);
+}
+
+int open_dev(int devnum)
+{
+ return open_dev_flags(devnum, O_RDONLY);
+}
+
+int open_dev_excl(int devnum)
+{
+ char buf[20];
+ int i;
+ int flags = O_RDWR;
+
+ sprintf(buf, "%d:%d", dev2major(devnum), dev2minor(devnum));
+ for (i=0 ; i<25 ; i++) {
+ int fd = dev_open(buf, flags|O_EXCL);
+ if (fd >= 0)
+ return fd;
+ if (errno == EACCES && flags == O_RDWR) {
+ flags = O_RDONLY;
+ continue;
+ }
+ if (errno != EBUSY)
+ return fd;
+ usleep(200000);
+ }
+ return -1;
+}
+
+int same_dev(char *one, char *two)
+{
+ struct stat st1, st2;
+ if (stat(one, &st1) != 0)
+ return 0;
+ if (stat(two, &st2) != 0)
+ return 0;
+ if ((st1.st_mode & S_IFMT) != S_IFBLK)
+ return 0;
+ if ((st2.st_mode & S_IFMT) != S_IFBLK)
+ return 0;
+ return st1.st_rdev == st2.st_rdev;
+}
+
+void wait_for(char *dev, int fd)
+{
+ int i;
+ struct stat stb_want;
+
+ if (fstat(fd, &stb_want) != 0 ||
+ (stb_want.st_mode & S_IFMT) != S_IFBLK)
+ return;
-struct supertype *super_by_fd(int fd)
+ for (i=0 ; i<25 ; i++) {
+ struct stat stb;
+ if (stat(dev, &stb) == 0 &&
+ (stb.st_mode & S_IFMT) == S_IFBLK &&
+ (stb.st_rdev == stb_want.st_rdev))
+ return;
+ usleep(200000);
+ }
+ if (i == 25)
+ dprintf("%s: timeout waiting for %s\n", __func__, dev);
+}
+
+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, char **subarrayp)
{
mdu_array_info_t array;
int vers;
char *verstr;
char version[20];
int i;
+ char *subarray = NULL;
+ int container = NoMdDev;
sra = sysfs_read(fd, 0, GET_VERSION);
sprintf(version, "%d.%d", vers, minor);
verstr = version;
}
+ if (minor == -2 && is_subarray(verstr)) {
+ char *dev = verstr+1;
+
+ subarray = strchr(dev, '/');
+ if (subarray)
+ *subarray++ = '\0';
+ subarray = strdup(subarray);
+ container = devname2devnum(dev);
+ if (sra)
+ sysfs_free(sra);
+ sra = sysfs_read(-1, container, GET_VERSION);
+ if (sra && sra->text_version[0])
+ verstr = sra->text_version;
+ else
+ verstr = "-no-metadata-";
+ }
+
for (i = 0; st == NULL && superlist[i] ; i++)
st = superlist[i]->match_metadata_desc(verstr);
if (sra)
sysfs_free(sra);
- st->sb = NULL;
+ if (st) {
+ st->sb = NULL;
+ 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) */
-struct supertype *dup_super(struct supertype *st)
+int dev_size_from_id(dev_t id, unsigned long long *size)
{
- struct supertype *stnew = NULL;
- char *verstr = NULL;
- char version[20];
- int i;
+ char buf[20];
+ int fd;
- if (!st)
- return st;
-
- if (st->minor_version == -1)
- sprintf(version, "%d", st->ss->major);
- else
- sprintf(version, "%d.%d", st->ss->major, st->minor_version);
- verstr = version;
+ 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;
+}
- for (i = 0; stnew == NULL && superlist[i] ; i++)
- stnew = superlist[i]->match_metadata_desc(verstr);
+struct supertype *dup_super(struct supertype *orig)
+{
+ struct supertype *st;
- stnew->sb = NULL;
- return stnew;
+ if (!orig)
+ return orig;
+ st = malloc(sizeof(*st));
+ if (!st)
+ return st;
+ memset(st, 0, sizeof(*st));
+ st->ss = orig->ss;
+ st->max_devs = orig->max_devs;
+ st->minor_version = orig->minor_version;
+ 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];
- st->ss = NULL;
+ 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));
+ st->ignore_hw_compat = 1;
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;
}
if (bestsuper != -1) {
int rv;
- st->ss = NULL;
+ memset(st, 0, sizeof(*st));
+ st->ignore_hw_compat = 1;
rv = superlist[bestsuper]->load_super(st, fd, NULL);
if (rv == 0) {
superlist[bestsuper]->free_super(st);
+ st->ignore_hw_compat = 0;
return st;
}
}
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
+ * -1 for unknown partition type
+ * 0 for other errors
+ */
+static int get_gpt_last_partition_end(int fd, unsigned long long *endofpart)
+{
+ struct GPT gpt;
+ unsigned char empty_gpt_entry[16]= {0};
+ struct GPT_part_entry *part;
+ char buf[512];
+ unsigned long long curr_part_end;
+ unsigned all_partitions, entry_size;
+ unsigned part_nr;
+
+ *endofpart = 0;
+
+ BUILD_BUG_ON(sizeof(gpt) != 512);
+ /* skip protective MBR */
+ lseek(fd, 512, SEEK_SET);
+ /* read GPT header */
+ if (read(fd, &gpt, 512) != 512)
+ return 0;
+
+ /* get the number of partition entries and the entry size */
+ all_partitions = __le32_to_cpu(gpt.part_cnt);
+ entry_size = __le32_to_cpu(gpt.part_size);
+
+ /* Check GPT signature*/
+ if (gpt.magic != GPT_SIGNATURE_MAGIC)
+ return -1;
+
+ /* sanity checks */
+ if (all_partitions > 1024 ||
+ entry_size > sizeof(buf))
+ return -1;
+
+ part = (struct GPT_part_entry *)buf;
+
+ for (part_nr=0; part_nr < all_partitions; part_nr++) {
+ /* read partition entry */
+ if (read(fd, buf, entry_size) != (ssize_t)entry_size)
+ return 0;
+
+ /* is this valid partition? */
+ if (memcmp(part->type_guid, empty_gpt_entry, 16) != 0) {
+ /* check the last lba for the current partition */
+ curr_part_end = __le64_to_cpu(part->ending_lba);
+ if (curr_part_end > *endofpart)
+ *endofpart = curr_part_end;
+ }
+
+ }
+ return 1;
+}
+
+/* Sets endofpart parameter to the last block used by the last partition on the device.
+ * Returns: 1 if successful
+ * -1 for unknown partition type
+ * 0 for other errors
+ */
+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;
+ int retval = 0;
+
+ *endofpart = 0;
+
+ BUILD_BUG_ON(sizeof(boot_sect) != 512);
+ /* read MBR */
+ lseek(fd, 0, 0);
+ if (read(fd, &boot_sect, 512) != 512)
+ goto abort;
+
+ /* check MBP signature */
+ 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++) {
+ /* check for GPT type */
+ if (part->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);
+ if (curr_part_end > *endofpart)
+ *endofpart = curr_part_end;
+
+ part++;
+ }
+ } else {
+ /* Unknown partition table */
+ retval = -1;
+ }
+ abort:
+ return retval;
+}
+
+int check_partitions(int fd, char *dname, unsigned long long freesize,
+ unsigned long long size)
+{
+ /*
+ * Check where the last partition ends
+ */
+ unsigned long long endofpart;
+ int ret;
+
+ if ((ret = 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 */
+ fprintf(stderr,
+ Name ": partition table exists on %s but will be lost or\n"
+ " meaningless after creating array\n",
+ dname);
+ return 1;
+ } else if (endofpart > freesize) {
+ /* last partition overlaps metadata */
+ fprintf(stderr,
+ Name ": metadata will over-write last partition on %s.\n",
+ dname);
+ return 1;
+ } else if (size && endofpart > size) {
+ /* partitions will be truncated in new device */
+ fprintf(stderr,
+ Name ": array size is too small to cover all partitions on %s.\n",
+ dname);
+ return 1;
+ }
+ }
+ return 0;
+}
+
void get_one_disk(int mdfd, mdu_array_info_t *ainf, mdu_disk_info_t *disk)
{
int d;
+
ioctl(mdfd, GET_ARRAY_INFO, ainf);
- for (d = 0 ; d < ainf->raid_disks + ainf->nr_disks ; d++)
- if (ioctl(mdfd, GET_DISK_INFO, disk) == 0)
+ for (d = 0 ; d < 1024 ; d++) {
+ if (ioctl(mdfd, GET_DISK_INFO, disk) == 0 &&
+ (disk->major || disk->minor))
return;
+ }
+}
+
+int open_container(int fd)
+{
+ /* '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 *e;
+ DIR *dir;
+ struct dirent *de;
+ int dfd, n;
+ char buf[200];
+ int major, minor;
+ struct stat st;
+
+ if (fstat(fd, &st) != 0)
+ return -1;
+ sprintf(path, "/sys/dev/block/%d:%d/holders",
+ (int)major(st.st_rdev), (int)minor(st.st_rdev));
+ e = path + strlen(path);
+
+ dir = opendir(path);
+ if (!dir)
+ return -1;
+ while ((de = readdir(dir))) {
+ if (de->d_ino == 0)
+ continue;
+ if (de->d_name[0] == '.')
+ continue;
+ sprintf(e, "/%s/dev", de->d_name);
+ dfd = open(path, O_RDONLY);
+ if (dfd < 0)
+ continue;
+ n = read(dfd, buf, sizeof(buf));
+ close(dfd);
+ if (n <= 0 || (unsigned)n >= sizeof(buf))
+ continue;
+ buf[n] = 0;
+ if (sscanf(buf, "%d:%d", &major, &minor) != 2)
+ continue;
+ sprintf(buf, "%d:%d", major, minor);
+ dfd = dev_open(buf, O_RDONLY);
+ if (dfd >= 0) {
+ closedir(dir);
+ return dfd;
+ }
+ }
+ closedir(dir);
+ return -1;
+}
+
+struct superswitch *version_to_superswitch(char *vers)
+{
+ int i;
+
+ for (i = 0; superlist[i]; i++) {
+ struct superswitch *ss = superlist[i];
+
+ if (strcmp(vers, ss->name) == 0)
+ return ss;
+ }
+
+ return NULL;
+}
+
+int is_container_member(struct mdstat_ent *mdstat, char *container)
+{
+ if (mdstat->metadata_version == NULL ||
+ strncmp(mdstat->metadata_version, "external:", 9) != 0 ||
+ !is_subarray(mdstat->metadata_version+9) ||
+ strncmp(mdstat->metadata_version+10, container, strlen(container)) != 0 ||
+ mdstat->metadata_version[10+strlen(container)] != '/')
+ return 0;
+
+ return 1;
+}
+
+int is_subarray_active(char *subarray, char *container)
+{
+ 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))
+ if (strcmp(to_subarray(ent, container), subarray) == 0)
+ break;
+
+ free_mdstat(mdstat);
+
+ return ent != NULL;
+}
+
+/* open_subarray - opens a subarray in a container
+ * @dev: container device name
+ * @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, char *subarray, struct supertype *st, int quiet)
+{
+ struct mdinfo *mdi;
+ struct mdinfo *info;
+ int fd, err = 1;
+
+ fd = open(dev, O_RDWR|O_EXCL);
+ if (fd < 0) {
+ if (!quiet)
+ fprintf(stderr, Name ": Couldn't open %s, aborting\n",
+ dev);
+ return -1;
+ }
+
+ st->devnum = fd2devnum(fd);
+ if (st->devnum == NoMdDev) {
+ if (!quiet)
+ fprintf(stderr,
+ Name ": Failed to determine device number for %s\n",
+ dev);
+ goto close_fd;
+ }
+
+ mdi = sysfs_read(fd, st->devnum, GET_VERSION|GET_LEVEL);
+ if (!mdi) {
+ if (!quiet)
+ fprintf(stderr, Name ": Failed to read sysfs for %s\n",
+ dev);
+ goto close_fd;
+ }
+
+ if (mdi->array.level != UnSet) {
+ if (!quiet)
+ fprintf(stderr, Name ": %s is not a container\n", dev);
+ goto free_sysfs;
+ }
+
+ st->ss = version_to_superswitch(mdi->text_version);
+ if (!st->ss) {
+ if (!quiet)
+ fprintf(stderr,
+ Name ": Operation not supported for %s metadata\n",
+ mdi->text_version);
+ goto free_sysfs;
+ }
+
+ st->devname = devnum2devname(st->devnum);
+ if (!st->devname) {
+ if (!quiet)
+ fprintf(stderr, Name ": Failed to allocate device name\n");
+ goto free_sysfs;
+ }
+
+ if (!st->ss->load_container) {
+ if (!quiet)
+ fprintf(stderr, Name ": %s is not a container\n", dev);
+ goto free_name;
+ }
+
+ if (st->ss->load_container(st, fd, NULL)) {
+ if (!quiet)
+ 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;
+
+ free_super:
+ if (err)
+ st->ss->free_super(st);
+ free_name:
+ if (err)
+ free(st->devname);
+ free_sysfs:
+ sysfs_free(mdi);
+ close_fd:
+ if (err)
+ close(fd);
+
+ if (err)
+ return -1;
+ else
+ return fd;
+}
+
+int add_disk(int mdfd, struct supertype *st,
+ struct mdinfo *sra, struct mdinfo *info)
+{
+ /* 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
+ info->recovery_start = 0;
+ rv = sysfs_add_disk(sra, info, 0);
+ if (! rv) {
+ struct mdinfo *sd2;
+ for (sd2 = sra->devs; sd2; sd2=sd2->next)
+ if (sd2 == info)
+ break;
+ if (sd2 == NULL) {
+ sd2 = malloc(sizeof(*sd2));
+ *sd2 = *info;
+ sd2->next = sra->devs;
+ sra->devs = sd2;
+ }
+ }
+ } else
+#endif
+ rv = ioctl(mdfd, ADD_NEW_DISK, &info->disk);
+ return rv;
}
-static int dev2major(int d)
+int remove_disk(int mdfd, struct supertype *st,
+ struct mdinfo *sra, struct mdinfo *info)
{
- if (d >= 0)
- return MD_MAJOR;
+ 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
- return get_mdp_major();
+#endif
+ rv = ioctl(mdfd, HOT_REMOVE_DISK, makedev(info->disk.major,
+ info->disk.minor));
+ return rv;
}
-static int dev2minor(int d)
+int set_array_info(int mdfd, struct supertype *st, struct mdinfo *info)
{
- if (d >= 0)
- return d;
- return (-1-d) << MdpMinorShift;
+ /* Initialise kernel's knowledge of array.
+ * This varies between externally managed arrays
+ * and older kernels
+ */
+ int vers = md_get_version(mdfd);
+ 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 rv;
}
-int find_free_devnum(int use_partitions)
+unsigned long long min_recovery_start(struct mdinfo *array)
{
- int devnum;
- for (devnum = 127; devnum != 128;
- devnum = devnum ? devnum-1 : (1<<22)-1) {
- char *dn;
- int _devnum;
+ /* find the minimum recovery_start in an array for metadata
+ * formats that only record per-array recovery progress instead
+ * of per-device
+ */
+ unsigned long long recovery_start = MaxSector;
+ struct mdinfo *d;
- _devnum = use_partitions ? (-1-devnum) : devnum;
- if (mddev_busy(_devnum))
- continue;
- /* make sure it is new to /dev too, at least as a
- * non-standard */
- dn = map_dev(dev2major(_devnum), dev2minor(_devnum), 0);
- if (dn && ! is_standard(dn, NULL))
- continue;
- break;
+ for (d = array->devs; d; d = d->next)
+ recovery_start = min(recovery_start, d->recovery_start);
+
+ return recovery_start;
+}
+
+int mdmon_pid(int devnum)
+{
+ char path[100];
+ char pid[10];
+ int fd;
+ int n;
+ char *devname = devnum2devname(devnum);
+
+ sprintf(path, "%s/%s.pid", MDMON_DIR, devname);
+ free(devname);
+
+ fd = open(path, O_RDONLY | O_NOATIME, 0);
+
+ if (fd < 0)
+ return -1;
+ n = read(fd, pid, 9);
+ close(fd);
+ if (n <= 0)
+ return -1;
+ return atoi(pid);
+}
+
+int mdmon_running(int devnum)
+{
+ int pid = mdmon_pid(devnum);
+ if (pid <= 0)
+ return 0;
+ if (kill(pid, 0) == 0)
+ return 1;
+ return 0;
+}
+
+int start_mdmon(int devnum)
+{
+ int i, skipped;
+ int len;
+ pid_t pid;
+ int status;
+ char pathbuf[1024];
+ char *paths[4] = {
+ pathbuf,
+ "/sbin/mdmon",
+ "mdmon",
+ NULL
+ };
+
+ if (check_env("MDADM_NO_MDMON"))
+ return 0;
+
+ len = readlink("/proc/self/exe", pathbuf, sizeof(pathbuf)-1);
+ if (len > 0) {
+ char *sl;
+ pathbuf[len] = 0;
+ sl = strrchr(pathbuf, '/');
+ if (sl)
+ sl++;
+ else
+ sl = pathbuf;
+ strcpy(sl, "mdmon");
+ } else
+ pathbuf[0] = '\0';
+
+ switch(fork()) {
+ case 0:
+ /* FIXME yuk. CLOSE_EXEC?? */
+ skipped = 0;
+ for (i=3; skipped < 20; i++)
+ if (close(i) < 0)
+ skipped++;
+ else
+ skipped = 0;
+
+ for (i=0; paths[i]; i++)
+ if (paths[i][0]) {
+ if (__offroot) {
+ execl(paths[i], "mdmon", "--offroot",
+ devnum2devname(devnum),
+ NULL);
+ } else {
+ execl(paths[i], "mdmon",
+ devnum2devname(devnum),
+ NULL);
+ }
+ }
+ exit(1);
+ case -1: fprintf(stderr, Name ": cannot run mdmon. "
+ "Array remains readonly\n");
+ return -1;
+ default: /* parent - good */
+ pid = wait(&status);
+ if (pid < 0 || status != 0)
+ return -1;
}
- if (devnum == 128)
- return NoMdDev;
- return use_partitions ? (-1-devnum) : devnum;
+ return 0;
}
+int check_env(char *name)
+{
+ char *val = getenv(name);
+
+ if (val && atoi(val) == 1)
+ return 1;
+
+ return 0;
+}
+
+__u32 random32(void)
+{
+ __u32 rv;
+ int rfd = open("/dev/urandom", O_RDONLY);
+ if (rfd < 0 || read(rfd, &rv, 4) != 4)
+ rv = random();
+ if (rfd >= 0)
+ close(rfd);
+ return rv;
+}
+
+#ifndef MDASSEMBLE
+int flush_metadata_updates(struct supertype *st)
+{
+ int sfd;
+ if (!st->updates) {
+ st->update_tail = NULL;
+ return -1;
+ }
+
+ sfd = connect_monitor(devnum2devname(st->container_dev));
+ if (sfd < 0)
+ return -1;
+
+ while (st->updates) {
+ struct metadata_update *mu = st->updates;
+ st->updates = mu->next;
+
+ send_message(sfd, mu, 0);
+ wait_reply(sfd, 0);
+ free(mu->buf);
+ free(mu);
+ }
+ ack(sfd, 0);
+ wait_reply(sfd, 0);
+ close(sfd);
+ st->update_tail = NULL;
+ return 0;
+}
+
+void append_metadata_update(struct supertype *st, void *buf, int len)
+{
+
+ struct metadata_update *mu = malloc(sizeof(*mu));
+
+ 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;
+}
+#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 {
+ fprintf(stderr, Name ": 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 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;
+}