#include <dirent.h>
#include <ctype.h>
-int load_sys(char *path, char *buf)
+#define MAX_SYSFS_PATH_LEN 120
+
+int load_sys(char *path, char *buf, int len)
{
int fd = open(path, O_RDONLY);
int n;
if (fd < 0)
return -1;
- n = read(fd, buf, 1024);
+ n = read(fd, buf, len);
close(fd);
- if (n <0 || n >= 1024)
+ if (n <0 || n >= len)
return -1;
buf[n] = 0;
if (n && buf[n-1] == '\n')
while (sra->devs) {
struct mdinfo *d = sra->devs;
sra->devs = d->next;
+ free(d->bb.entries);
free(d);
}
+ free(sra->bb.entries);
free(sra);
sra = sra2;
}
}
-int sysfs_open(int devnum, char *devname, char *attr)
+int sysfs_open(char *devnm, char *devname, char *attr)
{
- char fname[50];
+ char fname[MAX_SYSFS_PATH_LEN];
int fd;
- char *mdname = devnum2devname(devnum);
-
- if (!mdname)
- return -1;
- sprintf(fname, "/sys/block/%s/md/", mdname);
+ snprintf(fname, MAX_SYSFS_PATH_LEN, "/sys/block/%s/md/", devnm);
if (devname) {
- strcat(fname, devname);
- strcat(fname, "/");
+ strncat(fname, devname, MAX_SYSFS_PATH_LEN - strlen(fname));
+ strncat(fname, "/", MAX_SYSFS_PATH_LEN - strlen(fname));
}
- strcat(fname, attr);
+ strncat(fname, attr, MAX_SYSFS_PATH_LEN - strlen(fname));
fd = open(fname, O_RDWR);
if (fd < 0 && errno == EACCES)
fd = open(fname, O_RDONLY);
- free(mdname);
return fd;
}
-void sysfs_init(struct mdinfo *mdi, int fd, int devnum)
+void sysfs_init_dev(struct mdinfo *mdi, dev_t devid)
{
- mdi->sys_name[0] = 0;
- if (fd >= 0) {
- mdu_version_t vers;
- if (ioctl(fd, RAID_VERSION, &vers) != 0)
- return;
- devnum = fd2devnum(fd);
- }
- if (devnum == NoMdDev)
- return;
- fmt_devname(mdi->sys_name, devnum);
+ snprintf(mdi->sys_name,
+ sizeof(mdi->sys_name), "dev-%s", devid2kname(devid));
}
+int sysfs_init(struct mdinfo *mdi, int fd, char *devnm)
+{
+ struct stat stb;
+ char fname[MAX_SYSFS_PATH_LEN];
+ int retval = -ENODEV;
+
+ mdi->sys_name[0] = 0;
+ if (fd >= 0)
+ devnm = fd2devnm(fd);
+
+ if (devnm == NULL)
+ goto out;
-struct mdinfo *sysfs_read(int fd, int devnum, unsigned long options)
+ snprintf(fname, MAX_SYSFS_PATH_LEN, "/sys/block/%s/md", devnm);
+
+ if (stat(fname, &stb))
+ goto out;
+ if (!S_ISDIR(stb.st_mode))
+ goto out;
+ strcpy(mdi->sys_name, devnm);
+
+ retval = 0;
+out:
+ return retval;
+}
+
+struct mdinfo *sysfs_read(int fd, char *devnm, unsigned long options)
{
char fname[PATH_MAX];
char buf[PATH_MAX];
char *base;
char *dbase;
struct mdinfo *sra;
- struct mdinfo *dev;
+ struct mdinfo *dev, **devp;
DIR *dir = NULL;
struct dirent *de;
sra = xcalloc(1, sizeof(*sra));
- sysfs_init(sra, fd, devnum);
- if (sra->sys_name[0] == 0) {
+ if (sysfs_init(sra, fd, devnm)) {
free(sra);
return NULL;
}
sra->devs = NULL;
if (options & GET_VERSION) {
strcpy(base, "metadata_version");
- if (load_sys(fname, buf))
+ if (load_sys(fname, buf, sizeof(buf)))
goto abort;
if (strncmp(buf, "none", 4) == 0) {
sra->array.major_version =
}
if (options & GET_LEVEL) {
strcpy(base, "level");
- if (load_sys(fname, buf))
+ if (load_sys(fname, buf, sizeof(buf)))
goto abort;
sra->array.level = map_name(pers, buf);
}
if (options & GET_LAYOUT) {
strcpy(base, "layout");
- if (load_sys(fname, buf))
+ if (load_sys(fname, buf, sizeof(buf)))
goto abort;
sra->array.layout = strtoul(buf, NULL, 0);
}
- if (options & GET_DISKS) {
+ if (options & (GET_DISKS|GET_STATE)) {
strcpy(base, "raid_disks");
- if (load_sys(fname, buf))
+ if (load_sys(fname, buf, sizeof(buf)))
goto abort;
sra->array.raid_disks = strtoul(buf, NULL, 0);
}
- if (options & GET_DEGRADED) {
- strcpy(base, "degraded");
- if (load_sys(fname, buf))
- goto abort;
- sra->array.failed_disks = strtoul(buf, NULL, 0);
- }
if (options & GET_COMPONENT) {
strcpy(base, "component_size");
- if (load_sys(fname, buf))
+ if (load_sys(fname, buf, sizeof(buf)))
goto abort;
sra->component_size = strtoull(buf, NULL, 0);
/* sysfs reports "K", but we want sectors */
}
if (options & GET_CHUNK) {
strcpy(base, "chunk_size");
- if (load_sys(fname, buf))
+ if (load_sys(fname, buf, sizeof(buf)))
goto abort;
sra->array.chunk_size = strtoul(buf, NULL, 0);
}
if (options & GET_CACHE) {
strcpy(base, "stripe_cache_size");
- if (load_sys(fname, buf))
+ if (load_sys(fname, buf, sizeof(buf)))
/* Probably level doesn't support it */
sra->cache_size = 0;
else
}
if (options & GET_MISMATCH) {
strcpy(base, "mismatch_cnt");
- if (load_sys(fname, buf))
+ if (load_sys(fname, buf, sizeof(buf)))
goto abort;
sra->mismatch_cnt = strtoul(buf, NULL, 0);
}
size_t len;
strcpy(base, "safe_mode_delay");
- if (load_sys(fname, buf))
+ if (load_sys(fname, buf, sizeof(buf)))
goto abort;
/* remove a period, and count digits after it */
}
if (options & GET_BITMAP_LOCATION) {
strcpy(base, "bitmap/location");
- if (load_sys(fname, buf))
+ if (load_sys(fname, buf, sizeof(buf)))
goto abort;
if (strncmp(buf, "file", 4) == 0)
sra->bitmap_offset = 1;
goto abort;
}
+ if (options & GET_ARRAY_STATE) {
+ strcpy(base, "array_state");
+ if (load_sys(fname, buf, sizeof(buf)))
+ goto abort;
+ sra->array_state = map_name(sysfs_array_states, buf);
+ }
+
+ if (options & GET_CONSISTENCY_POLICY) {
+ strcpy(base, "consistency_policy");
+ if (load_sys(fname, buf, sizeof(buf)))
+ sra->consistency_policy = CONSISTENCY_POLICY_UNKNOWN;
+ else
+ sra->consistency_policy = map_name(consistency_policies,
+ buf);
+ }
+
if (! (options & GET_DEVS))
return sra;
if (!dir)
goto abort;
sra->array.spare_disks = 0;
+ sra->array.active_disks = 0;
+ sra->array.failed_disks = 0;
+ sra->array.working_disks = 0;
+ devp = &sra->devs;
+ sra->devs = NULL;
while ((de = readdir(dir)) != NULL) {
char *ep;
if (de->d_ino == 0 ||
dbase = base + strlen(base);
*dbase++ = '/';
- dev = xmalloc(sizeof(*dev));
+ dev = xcalloc(1, sizeof(*dev));
/* Always get slot, major, minor */
strcpy(dbase, "slot");
- if (load_sys(fname, buf)) {
+ if (load_sys(fname, buf, sizeof(buf))) {
/* hmm... unable to read 'slot' maybe the device
* is going away?
*/
free(dev);
goto abort;
}
-
+
}
strcpy(dev->sys_name, de->d_name);
dev->disk.raid_disk = strtoul(buf, &ep, 10);
if (*ep) dev->disk.raid_disk = -1;
+ sra->array.nr_disks++;
strcpy(dbase, "block/dev");
- if (load_sys(fname, buf)) {
+ if (load_sys(fname, buf, sizeof(buf))) {
/* assume this is a stale reference to a hot
* removed device
*/
- free(dev);
- continue;
+ if (!(options & GET_DEVS_ALL)) {
+ free(dev);
+ continue;
+ }
+ } else {
+ sscanf(buf, "%d:%d", &dev->disk.major, &dev->disk.minor);
}
- sscanf(buf, "%d:%d", &dev->disk.major, &dev->disk.minor);
- /* special case check for block devices that can go 'offline' */
- strcpy(dbase, "block/device/state");
- if (load_sys(fname, buf) == 0 &&
- strncmp(buf, "offline", 7) == 0) {
- free(dev);
- continue;
+ if (!(options & GET_DEVS_ALL)) {
+ /* special case check for block devices that can go 'offline' */
+ strcpy(dbase, "block/device/state");
+ if (load_sys(fname, buf, sizeof(buf)) == 0 &&
+ strncmp(buf, "offline", 7) == 0) {
+ free(dev);
+ continue;
+ }
}
/* finally add this disk to the array */
- dev->next = sra->devs;
- sra->devs = dev;
+ *devp = dev;
+ devp = & dev->next;
+ dev->next = NULL;
if (options & GET_OFFSET) {
strcpy(dbase, "offset");
- if (load_sys(fname, buf))
+ if (load_sys(fname, buf, sizeof(buf)))
goto abort;
dev->data_offset = strtoull(buf, NULL, 0);
strcpy(dbase, "new_offset");
- if (load_sys(fname, buf) == 0)
+ if (load_sys(fname, buf, sizeof(buf)) == 0)
dev->new_data_offset = strtoull(buf, NULL, 0);
else
dev->new_data_offset = dev->data_offset;
}
if (options & GET_SIZE) {
strcpy(dbase, "size");
- if (load_sys(fname, buf))
+ if (load_sys(fname, buf, sizeof(buf)))
goto abort;
dev->component_size = strtoull(buf, NULL, 0) * 2;
}
if (options & GET_STATE) {
dev->disk.state = 0;
strcpy(dbase, "state");
- if (load_sys(fname, buf))
+ if (load_sys(fname, buf, sizeof(buf)))
goto abort;
- if (strstr(buf, "in_sync"))
- dev->disk.state |= (1<<MD_DISK_SYNC);
if (strstr(buf, "faulty"))
dev->disk.state |= (1<<MD_DISK_FAULTY);
- if (dev->disk.state == 0)
- sra->array.spare_disks++;
+ else {
+ sra->array.working_disks++;
+ if (strstr(buf, "in_sync")) {
+ dev->disk.state |= (1<<MD_DISK_SYNC);
+ sra->array.active_disks++;
+ }
+ if (dev->disk.state == 0)
+ sra->array.spare_disks++;
+ }
}
if (options & GET_ERROR) {
strcpy(buf, "errors");
- if (load_sys(fname, buf))
+ if (load_sys(fname, buf, sizeof(buf)))
goto abort;
dev->errors = strtoul(buf, NULL, 0);
}
}
+
+ if ((options & GET_STATE) && sra->array.raid_disks)
+ sra->array.failed_disks = sra->array.raid_disks -
+ sra->array.active_disks - sra->array.spare_disks;
+
closedir(dir);
return sra;
* This returns in units of sectors.
*/
struct stat stb;
- char fname[50];
+ char fname[MAX_SYSFS_PATH_LEN];
int n;
- if (fstat(fd, &stb)) return 0;
- if (major(stb.st_rdev) != (unsigned)get_mdp_major())
- sprintf(fname, "/sys/block/md%d/md/component_size",
- (int)minor(stb.st_rdev));
- else
- sprintf(fname, "/sys/block/md_d%d/md/component_size",
- (int)minor(stb.st_rdev)>>MdpMinorShift);
+ if (fstat(fd, &stb))
+ return 0;
+ snprintf(fname, MAX_SYSFS_PATH_LEN,
+ "/sys/block/%s/md/component_size", stat2devnm(&stb));
fd = open(fname, O_RDONLY);
if (fd < 0)
return 0;
int sysfs_set_str(struct mdinfo *sra, struct mdinfo *dev,
char *name, char *val)
{
- char fname[50];
+ char fname[MAX_SYSFS_PATH_LEN];
unsigned int n;
int fd;
- sprintf(fname, "/sys/block/%s/md/%s/%s",
+ snprintf(fname, MAX_SYSFS_PATH_LEN, "/sys/block/%s/md/%s/%s",
sra->sys_name, dev?dev->sys_name:"", name);
fd = open(fname, O_WRONLY);
if (fd < 0)
n = write(fd, val, strlen(val));
close(fd);
if (n != strlen(val)) {
- dprintf(Name ": failed to write '%s' to '%s' (%s)\n",
+ dprintf("failed to write '%s' to '%s' (%s)\n",
val, fname, strerror(errno));
return -1;
}
int sysfs_uevent(struct mdinfo *sra, char *event)
{
- char fname[50];
+ char fname[MAX_SYSFS_PATH_LEN];
int n;
int fd;
- sprintf(fname, "/sys/block/%s/uevent",
+ snprintf(fname, MAX_SYSFS_PATH_LEN, "/sys/block/%s/uevent",
sra->sys_name);
fd = open(fname, O_WRONLY);
if (fd < 0)
n = write(fd, event, strlen(event));
close(fd);
if (n != (int)strlen(event)) {
- dprintf(Name ": failed to write '%s' to '%s' (%s)\n",
+ dprintf("failed to write '%s' to '%s' (%s)\n",
event, fname, strerror(errno));
return -1;
}
return 0;
-}
+}
int sysfs_attribute_available(struct mdinfo *sra, struct mdinfo *dev, char *name)
{
- char fname[50];
+ char fname[MAX_SYSFS_PATH_LEN];
struct stat st;
- sprintf(fname, "/sys/block/%s/md/%s/%s",
+ snprintf(fname, MAX_SYSFS_PATH_LEN, "/sys/block/%s/md/%s/%s",
sra->sys_name, dev?dev->sys_name:"", name);
return stat(fname, &st) == 0;
int sysfs_get_fd(struct mdinfo *sra, struct mdinfo *dev,
char *name)
{
- char fname[50];
+ char fname[MAX_SYSFS_PATH_LEN];
int fd;
- sprintf(fname, "/sys/block/%s/md/%s/%s",
+ snprintf(fname, MAX_SYSFS_PATH_LEN, "/sys/block/%s/md/%s/%s",
sra->sys_name, dev?dev->sys_name:"", name);
fd = open(fname, O_RDWR);
if (fd < 0)
lseek(fd, 0, 0);
n = read(fd, buf, sizeof(buf));
- if (n <= 0)
+ if (n <= 0 || n == sizeof(buf))
return -2;
buf[n] = 0;
*val = strtoull(buf, &ep, 0);
return n;
}
+int sysfs_fd_get_two(int fd, unsigned long long *v1, unsigned long long *v2)
+{
+ /* two numbers in this sysfs file, either
+ * NNN (NNN)
+ * or
+ * NNN / NNN
+ */
+ char buf[80];
+ int n;
+ char *ep, *ep2;
+
+ lseek(fd, 0, 0);
+ n = read(fd, buf, sizeof(buf));
+ if (n <= 0 || n == sizeof(buf))
+ return -2;
+ buf[n] = 0;
+ *v1 = strtoull(buf, &ep, 0);
+ if (ep == buf || (*ep != 0 && *ep != '\n' && *ep != ' '))
+ return -1;
+ while (*ep == ' ' || *ep == '/' || *ep == '(')
+ ep++;
+ *v2 = strtoull(ep, &ep2, 0);
+ if (ep2 == ep || (*ep2 != 0 && *ep2 != '\n' && *ep2 != ' ' && *ep2 != ')')) {
+ *v2 = *v1;
+ return 1;
+ }
+ return 2;
+}
+
+int sysfs_get_two(struct mdinfo *sra, struct mdinfo *dev,
+ char *name, unsigned long long *v1, unsigned long long *v2)
+{
+ int n;
+ int fd;
+
+ fd = sysfs_get_fd(sra, dev, name);
+ if (fd < 0)
+ return -1;
+ n = sysfs_fd_get_two(fd, v1, v2);
+ close(fd);
+ return n;
+}
+
int sysfs_fd_get_str(int fd, char *val, int size)
{
int n;
lseek(fd, 0, 0);
n = read(fd, val, size);
- if (n <= 0)
+ if (n <= 0 || n == size)
return -1;
val[n] = 0;
return n;
if ((vers % 100) < 2 ||
sysfs_set_str(info, NULL, "metadata_version",
ver) < 0) {
- pr_err("This kernel does not "
- "support external metadata.\n");
+ pr_err("This kernel does not support external metadata.\n");
return 1;
}
}
rc = sysfs_set_num(info, NULL, "array_size",
info->custom_array_size/2);
if (rc && errno == ENOENT) {
- pr_err("This kernel does not "
- "have the md/array_size attribute, "
- "the array may be larger than expected\n");
+ pr_err("This kernel does not have the md/array_size attribute, the array may be larger than expected\n");
rc = 0;
}
rv |= rc;
* once the reshape completes.
*/
}
+
+ if (info->consistency_policy == CONSISTENCY_POLICY_PPL) {
+ if (sysfs_set_str(info, NULL, "consistency_policy",
+ map_num(consistency_policies,
+ info->consistency_policy))) {
+ pr_err("This kernel does not support PPL. Falling back to consistency-policy=resync.\n");
+ info->consistency_policy = CONSISTENCY_POLICY_RESYNC;
+ }
+ }
+
return rv;
}
char nm[PATH_MAX];
char *dname;
int rv;
+ int i;
sprintf(dv, "%d:%d", sd->disk.major, sd->disk.minor);
rv = sysfs_set_str(sra, NULL, "new_dev", dv);
return rv;
memset(nm, 0, sizeof(nm));
- sprintf(dv, "/sys/dev/block/%d:%d", sd->disk.major, sd->disk.minor);
- rv = readlink(dv, nm, sizeof(nm)-1);
- if (rv <= 0)
- return -1;
- nm[rv] = '\0';
- dname = strrchr(nm, '/');
- if (dname) dname++;
+ dname = devid2kname(makedev(sd->disk.major, sd->disk.minor));
strcpy(sd->sys_name, "dev-");
strcpy(sd->sys_name+4, dname);
rv = sysfs_set_num(sra, sd, "offset", sd->data_offset);
rv |= sysfs_set_num(sra, sd, "size", (sd->component_size+1) / 2);
if (sra->array.level != LEVEL_CONTAINER) {
+ if (sra->consistency_policy == CONSISTENCY_POLICY_PPL) {
+ rv |= sysfs_set_num(sra, sd, "ppl_sector", sd->ppl_sector);
+ rv |= sysfs_set_num(sra, sd, "ppl_size", sd->ppl_size);
+ }
if (sd->recovery_start == MaxSector)
/* This can correctly fail if array isn't started,
* yet, so just ignore status for now.
if (resume)
sysfs_set_num(sra, sd, "recovery_start", sd->recovery_start);
}
+ if (sd->bb.supported) {
+ if (sysfs_set_str(sra, sd, "state", "external_bbl")) {
+ /*
+ * backward compatibility - if kernel doesn't support
+ * bad blocks for external metadata, let it continue
+ * as long as there are none known so far
+ */
+ if (sd->bb.count) {
+ pr_err("The kernel has no support for bad blocks in external metadata\n");
+ return -1;
+ }
+ }
+
+ for (i = 0; i < sd->bb.count; i++) {
+ char s[30];
+ const struct md_bb_entry *entry = &sd->bb.entries[i];
+
+ snprintf(s, sizeof(s) - 1, "%llu %d\n", entry->sector,
+ entry->length);
+ rv |= sysfs_set_str(sra, sd, "bad_blocks", s);
+ }
+ }
return rv;
}
struct stat st;
char path[256];
char sg_path[256];
- char sg_major_minor[8];
+ char sg_major_minor[10];
char *c;
DIR *dir;
struct dirent *de;
rv = read(fd, sg_major_minor, sizeof(sg_major_minor));
close(fd);
- if (rv < 0)
+ if (rv < 0 || rv == sizeof(sg_major_minor))
return -1;
else
sg_major_minor[rv - 1] = '\0';
return 0;
}
-
-int sysfs_unique_holder(int devnum, long rdev)
+int sysfs_unique_holder(char *devnm, long rdev)
{
- /* Check that devnum is a holder of rdev,
+ /* Check that devnm is a holder of rdev,
* and is the only holder.
* we should be locked against races by
- * an O_EXCL on devnum
+ * an O_EXCL on devnm
* Return values:
* 0 - not unique, not even a holder
* 1 - unique, this is the only holder.
return -1;
l = strlen(dirname);
while ((de = readdir(dir)) != NULL) {
- char buf[10];
+ char buf[100];
+ char *sl;
int n;
- int mj, mn;
- char c;
- int fd;
if (de->d_ino == 0)
continue;
continue;
strcpy(dirname+l, "/");
strcat(dirname+l, de->d_name);
- strcat(dirname+l, "/dev");
- fd = open(dirname, O_RDONLY);
- if (fd < 0) {
- /* Probably a race, just ignore this */
- continue;
- }
- n = read(fd, buf, sizeof(buf)-1);
- close(fd);
- if (n < 0)
+ n = readlink(dirname, buf, sizeof(buf)-1);
+ if (n <= 0)
continue;
buf[n] = 0;
- if (sscanf(buf, "%d:%d%c", &mj, &mn, &c) != 3 ||
- c != '\n')
+ sl = strrchr(buf, '/');
+ if (!sl)
continue;
- if (mj != MD_MAJOR)
- mn = -1-(mn>>6);
+ sl++;
- if (devnum == mn)
+ if (strcmp(devnm, sl) == 0)
ret |= 1;
else
ret |= 2;
if (strcmp(buf, "frozen\n") == 0)
/* Already frozen */
return 0;
- if (strcmp(buf, "idle\n") != 0)
+ if (strcmp(buf, "idle\n") != 0 && strcmp(buf, "recover\n") != 0)
return -1;
if (sysfs_set_str(sra, NULL, "sync_action", "frozen") < 0)
return 0;
return 1;
}
+
+int sysfs_wait(int fd, int *msec)
+{
+ /* Wait up to '*msec' for fd to have an exception condition.
+ * if msec == NULL, wait indefinitely.
+ */
+ fd_set fds;
+ int n;
+ FD_ZERO(&fds);
+ FD_SET(fd, &fds);
+ if (msec == NULL)
+ n = select(fd+1, NULL, NULL, &fds, NULL);
+ else if (*msec < 0)
+ n = 0;
+ else {
+ struct timeval start, end, tv;
+ gettimeofday(&start, NULL);
+ if (*msec < 1000) {
+ tv.tv_sec = 0;
+ tv.tv_usec = (*msec)*1000;
+ } else {
+ tv.tv_sec = (*msec)/1000;
+ tv.tv_usec = 0;
+ }
+ n = select(fd+1, NULL, NULL, &fds, &tv);
+ gettimeofday(&end, NULL);
+ end.tv_sec -= start.tv_sec;
+ *msec -= (end.tv_sec * 1000 + end.tv_usec/1000
+ - start.tv_usec/1000) + 1;
+ }
+ return n;
+}