/*
* mdadm - manage Linux "md" devices aka RAID arrays.
*
- * Copyright (C) 2001-2009 Neil Brown <neilb@suse.de>
+ * Copyright (C) 2001-2012 Neil Brown <neilb@suse.de>
*
*
* This program is free software; you can redistribute it and/or modify
#include <dirent.h>
#include <signal.h>
+int __offroot;
+
/*
* following taken from linux/blkpg.h because they aren't
* anywhere else and it isn't safe to #include linux/ * stuff.
int hit = 0; /* number of Hex digIT */
int i;
char c;
- for (i=0; i<4; i++) uuid[i]=0;
+ for (i = 0; i < 4; i++)
+ uuid[i] = 0;
- while ((c= *str++)) {
+ while ((c = *str++) != 0) {
int n;
- if (c>='0' && c<='9')
+ if (c >= '0' && c <= '9')
n = c-'0';
- else if (c>='a' && c <= 'f')
+ else if (c >= 'a' && c <= 'f')
n = 10 + c - 'a';
- else if (c>='A' && c <= 'F')
+ else if (c >= 'A' && c <= 'F')
n = 10 + c - 'A';
else if (strchr(":. -", c))
continue;
return 0;
}
-
/*
* Get the md version number.
* We use the RAID_VERSION ioctl if it is supported
{
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;
}
return (a*1000000)+(b*1000)+c;
}
-long long parse_size(char *size)
+unsigned 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)
+ * 0 returned on error.
*/
char *c;
long long s = strtoll(size, &c, 10);
s *= 1024 * 1024 * 2;
break;
}
- }
+ } else
+ s = 0;
if (*c)
s = 0;
return s;
{
/* Parse the layout string for 'faulty' */
int ln = strcspn(layout, "0123456789");
- char *m = strdup(layout);
+ char *m = xstrdup(layout);
int mode;
m[ln] = 0;
mode = map_name(faultylayout, m);
return mode | (atoi(layout+ln)<< ModeShift);
}
+
+long parse_num(char *num)
+{
+ /* Either return a valid number, or -1 */
+ char *c;
+ long rv = strtol(num, &c, 10);
+ if (rv < 0 || *c || !num[0])
+ return -1;
+ else
+ return rv;
+}
#endif
void remove_partitions(int fd)
a.datalen = sizeof(p);
a.flags = 0;
memset(a.data, 0, a.datalen);
- for (p.pno=0; p.pno < 16; p.pno++)
+ for (p.pno = 0; p.pno < 16; p.pno++)
ioctl(fd, BLKPG, &a);
#endif
}
return rv;
}
-int enough(int level, int raid_disks, int layout, int clean,
- char *avail, int avail_disks)
+int enough(int level, int raid_disks, int layout, int clean, char *avail)
{
int copies, first;
+ int i;
+ int avail_disks = 0;
+
+ for (i = 0; i < raid_disks; i++)
+ avail_disks += !!avail[i];
+
switch (level) {
case 10:
/* This is the tricky one - we need to check
* which actual disks are present.
*/
copies = (layout&255)* ((layout>>8) & 255);
- first=0;
+ first = 0;
do {
/* there must be one of the 'copies' form 'first' */
int n = copies;
- int cnt=0;
+ int cnt = 0;
while (n--) {
if (avail[first])
cnt++;
struct mdu_array_info_s array;
struct mdu_disk_info_s disk;
int avail_disks = 0;
- int i;
+ 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<array.raid_disks + array.nr_disks; i++) {
+ 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)
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);
+ 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)
unsigned char *ac = (unsigned char *)a;
unsigned char *bc = (unsigned char *)b;
int i;
- for (i=0; i<16; i+= 4) {
+ 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] ||
return 0;
}
}
+
void copy_uuid(void *a, int b[4], int swapuuid)
{
if (swapuuid) {
unsigned char *ac = (unsigned char *)a;
unsigned char *bc = (unsigned char *)b;
int i;
- for (i=0; i<16; i+= 4) {
+ 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];
mtime = sb[44]|(sb[45]|(sb[46]|sb[47]<<8)<<8)<<8;
bsize = sb[24]|(sb[25]|(sb[26]|sb[27]<<8)<<8)<<8;
size = sb[4]|(sb[5]|(sb[6]|sb[7]<<8)<<8)<<8;
- fprintf(stderr, Name ": %s appears to contain an ext2fs file system\n",
+ pr_err("%s appears to contain an ext2fs file system\n",
name);
fprintf(stderr," size=%dK mtime=%s",
size*(1<<bsize), ctime(&mtime));
return 0;
if (read(fd, sb, 1024) != 1024)
return 0;
- if (strncmp((char*)sb+52, "ReIsErFs",8)!=0 &&
- strncmp((char*)sb+52, "ReIsEr2Fs",9)!=0)
+ if (strncmp((char*)sb+52, "ReIsErFs",8) != 0 &&
+ strncmp((char*)sb+52, "ReIsEr2Fs",9) != 0)
return 0;
- fprintf(stderr, Name ": %s appears to contain a reiserfs file system\n",name);
+ pr_err("%s appears to contain a reiserfs file system\n",name);
size = sb[0]|(sb[1]|(sb[2]|sb[3]<<8)<<8)<<8;
fprintf(stderr, " size = %luK\n", size*4);
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",
+ 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);
{
char *add = "";
int i;
- for (i=0; i<5; i++) {
+ for (i = 0; i < 5; i++) {
char buf[100];
fprintf(stderr, "%s%s", mesg, add);
fflush(stderr);
return 0;
add = "(y/n) ";
}
- fprintf(stderr, Name ": assuming 'no'\n");
+ pr_err("assuming 'no'\n");
return 0;
}
#endif /* MDASSEMBLE */
* 0 if not a standard name.
*/
char *d = strrchr(dev, '/');
- int type=0;
+ int type = 0;
int num;
if (!d)
return 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_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)
- d += 1, type=-1; /* /dev/md/N */
+ 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_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)
+ d += 1, type = -1; /* /dev/md/N */
else
return 0;
if (!*d)
unsigned int csum;
unsigned int *superc = (unsigned int*) super;
- for(i=0; i<bytes/4; i++)
- newcsum+= superc[i];
+ for(i = 0; i < bytes/4; i++)
+ newcsum += superc[i];
csum = (newcsum& 0xffffffff) + (newcsum>>32);
#ifdef __alpha__
/* The in-kernel checksum calculation is always 16bit on
* the alpha, though it is 32 bit on i386...
- * I wonder what it is elsewhere... (it uses and API in
+ * I wonder what it is elsewhere... (it uses an API in
* a way that it shouldn't).
*/
csum = (csum & 0xffff) + (csum >> 16);
* 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)
- buf[0]=0;
+ buf[0] = 0;
else if (bytes < 2*1024LL*1024LL*1024LL) {
long cMiB = (bytes / ( (1LL<<20) / 200LL ) +1) /2;
long cMB = (bytes / ( 1000000LL / 200LL ) +1) /2;
unsigned long long calc_array_size(int level, int raid_disks, int layout,
int chunksize, unsigned long long devsize)
+{
+ if (level == 1)
+ return 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 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 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;
}
- devsize &= ~(unsigned long long)((chunksize>>9)-1);
- return data_disks * devsize;
+
+ return data_disks;
}
#if !defined(MDASSEMBLE) || defined(MDASSEMBLE) && defined(MDASSEMBLE_AUTO)
void put_md_name(char *name)
{
- if (strncmp(name, "/dev/.tmp.md", 12)==0)
+ if (strncmp(name, "/dev/.tmp.md", 12) == 0)
unlink(name);
}
devnum = devnum ? devnum-1 : (1<<20)-1) {
char *dn;
int _devnum;
+ char nbuf[50];
_devnum = use_partitions ? (-1-devnum) : devnum;
if (mddev_busy(_devnum))
continue;
+ sprintf(nbuf, "%s%d", use_partitions?"mdp":"md", devnum);
+ if (!conf_name_is_free(nbuf))
+ continue;
/* make sure it is new to /dev too, at least as a
* non-standard */
dn = map_dev(dev2major(_devnum), dev2minor(_devnum), 0);
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) {
+ 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) {
+ if (mknod(devname, S_IFBLK|0600, makedev(major, minor)) == 0) {
fd = open(devname, flags);
unlink(devname);
}
int flags = O_RDWR;
sprintf(buf, "%d:%d", dev2major(devnum), dev2minor(devnum));
- 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 &&
char *dev = verstr+1;
subarray = strchr(dev, '/');
- if (subarray)
+ if (subarray) {
*subarray++ = '\0';
- subarray = strdup(subarray);
+ subarray = xstrdup(subarray);
+ }
container = devname2devnum(dev);
if (sra)
sysfs_free(sra);
if (!orig)
return orig;
- st = malloc(sizeof(*st));
- if (!st)
- return st;
- memset(st, 0, sizeof(*st));
+ st = xcalloc(1, sizeof(*st));
st->ss = orig->ss;
st->max_devs = orig->max_devs;
st->minor_version = orig->minor_version;
int bestsuper = -1;
int i;
- st = malloc(sizeof(*st));
- memset(st, 0, sizeof(*st));
+ st = xcalloc(1, sizeof(*st));
st->container_dev = NoMdDev;
- 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)
- fprintf(stderr, Name ": Cannot get size of %s: %s\b",
+ pr_err("Cannot get size of %s: %s\b",
dname, strerror(errno));
return 0;
}
static int get_gpt_last_partition_end(int fd, unsigned long long *endofpart)
{
struct GPT gpt;
- unsigned char buf[512];
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);
- /* read GPT header */
+ /* skip protective MBR */
lseek(fd, 512, SEEK_SET);
+ /* read GPT header */
if (read(fd, &gpt, 512) != 512)
return 0;
/* sanity checks */
if (all_partitions > 1024 ||
- entry_size > 512)
+ entry_size > sizeof(buf))
return -1;
- /* read first GPT partition entries */
- if (read(fd, buf, 512) != 512)
- return 0;
+ part = (struct GPT_part_entry *)buf;
- 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;
- for (part_nr=0; part_nr < all_partitions; part_nr++) {
/* is this valid partition? */
if (memcmp(part->type_guid, empty_gpt_entry, 16) != 0) {
/* check the last lba for the current partition */
*endofpart = curr_part_end;
}
- part = (struct GPT_part_entry*)((unsigned char*)part + entry_size);
-
- if ((unsigned char *)part >= buf + 512) {
- if (read(fd, buf, 512) != 512)
- return 0;
- part = (struct GPT_part_entry*)buf;
- }
}
return 1;
}
/* found the correct signature */
part = boot_sect.parts;
- for (part_nr=0; part_nr < MBR_PARTITIONS; part_nr++) {
+ 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);
/* 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);
+ pr_err("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);
+ pr_err("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);
+ pr_err("array size is too small to cover all partitions on %s.\n",
+ dname);
return 1;
}
}
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 < MAX_DISKS ; d++) {
+ if (ioctl(mdfd, GET_DISK_INFO, disk) == 0 &&
+ (disk->major || disk->minor))
return;
+ }
}
int open_container(int fd)
fd = open(dev, O_RDWR|O_EXCL);
if (fd < 0) {
if (!quiet)
- fprintf(stderr, Name ": Couldn't open %s, aborting\n",
+ pr_err("Couldn't open %s, aborting\n",
dev);
- return 2;
+ return -1;
}
st->devnum = fd2devnum(fd);
if (st->devnum == NoMdDev) {
if (!quiet)
- fprintf(stderr,
- Name ": Failed to determine device number for %s\n",
- dev);
+ pr_err("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",
+ pr_err("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);
+ pr_err("%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);
+ pr_err("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");
+ pr_err("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);
+ pr_err("%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",
+ pr_err("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",
+ pr_err("Failed to find subarray-%s in %s\n",
subarray, dev);
goto free_super;
}
if (sd2 == info)
break;
if (sd2 == NULL) {
- sd2 = malloc(sizeof(*sd2));
+ sd2 = xmalloc(sizeof(*sd2));
*sd2 = *info;
sd2->next = sra->devs;
sra->devs = sd2;
int start_mdmon(int devnum)
{
- int i;
+ int i, skipped;
int len;
- pid_t pid;
+ pid_t pid;
int status;
char pathbuf[1024];
char *paths[4] = {
if (check_env("MDADM_NO_MDMON"))
return 0;
- len = readlink("/proc/self/exe", pathbuf, sizeof(pathbuf));
+ len = readlink("/proc/self/exe", pathbuf, sizeof(pathbuf)-1);
if (len > 0) {
char *sl;
pathbuf[len] = 0;
switch(fork()) {
case 0:
/* FIXME yuk. CLOSE_EXEC?? */
- for (i=3; i < 100; i++)
- close(i);
- for (i=0; paths[i]; i++)
- if (paths[i][0])
- execl(paths[i], "mdmon",
- devnum2devname(devnum),
- NULL);
+ 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. "
+ case -1: pr_err("cannot run mdmon. "
"Array remains readonly\n");
return -1;
default: /* parent - good */
void append_metadata_update(struct supertype *st, void *buf, int len)
{
- struct metadata_update *mu = malloc(sizeof(*mu));
+ struct metadata_update *mu = xmalloc(sizeof(*mu));
mu->buf = buf;
mu->len = len;
if (check_env("MDADM_EXPERIMENTAL"))
return 1;
else {
- fprintf(stderr, Name ": To use this feature MDADM_EXPERIMENTAL enviroment variable has to defined.\n");
+ pr_err("To use this feature MDADM_EXPERIMENTAL"
+ " environment variable has to be defined.\n");
return 0;
}
}
projects / thirdparty / mdadm.git / blobdiff