static void viva_power_init(struct power_module *module)
{
struct viva_power_module *viva =
(struct viva_power_module *) module;
int tmp;
char freq_buf[MAX_FREQ_NUMBER*10];
tmp = sysfs_read(CPUFREQ_CPU0 "scaling_available_frequencies",
freq_buf, sizeof(freq_buf));
if (tmp <= 0) {
return;
}
freq_num = str_to_tokens(freq_buf, freq_list, MAX_FREQ_NUMBER);
if (!freq_num) {
return;
}
max_freq = freq_list[freq_num - 1];
tmp = (NOM_FREQ_INDEX > freq_num) ? freq_num : NOM_FREQ_INDEX;
nom_freq = freq_list[tmp - 1];
sysfs_write(CPUFREQ_INTERACTIVE "timer_rate", "20000");
sysfs_write(CPUFREQ_INTERACTIVE "min_sample_time","60000");
sysfs_write(CPUFREQ_INTERACTIVE "hispeed_freq", nom_freq);
sysfs_write(CPUFREQ_INTERACTIVE "go_hispeed_load", "60");
sysfs_write(CPUFREQ_INTERACTIVE "above_hispeed_delay", "100000");
ALOGI("Initialized successfully");
viva->inited = 1;
}
int sysfs_gpio_getvalue(uint8_t pin)
{
char buf[MAX_BUF], value[4];
snprintf(buf, sizeof(buf), SYSFS_GPIO_DIR "/gpio%d/", pin);
sysfs_read(buf, "value", value);
return atoi(value);
}
/**
* Read the value from an ADC.
*
* It is recommended to use adc_sample() in preference to this as the ADC values can sometimes be erroneous.
*
* @see adc_sample()
*
* @param int adc the ADC channel to read
*
* @return int
*/
int adc_get_value(int adc)
{
char ctrlFile[1024];
char adcVal[16];
int bytesRead;
switch (adc)
{
case 0:
snprintf(ctrlFile, sizeof(ctrlFile), "%s%s", ADC_DIR_PREFIX, ADC_0_DIR);
break;
case 1:
snprintf(ctrlFile, sizeof(ctrlFile), "%s%s", ADC_DIR_PREFIX, ADC_1_DIR);
break;
case 2:
snprintf(ctrlFile, sizeof(ctrlFile), "%s%s", ADC_DIR_PREFIX, ADC_2_DIR);
break;
case 3:
snprintf(ctrlFile, sizeof(ctrlFile), "%s%s", ADC_DIR_PREFIX, ADC_3_DIR);
break;
case 4:
snprintf(ctrlFile, sizeof(ctrlFile), "%s%s", ADC_DIR_PREFIX, ADC_4_DIR);
break;
case 5:
snprintf(ctrlFile, sizeof(ctrlFile), "%s%s", ADC_DIR_PREFIX, ADC_5_DIR);
break;
case 6:
snprintf(ctrlFile, sizeof(ctrlFile), "%s%s", ADC_DIR_PREFIX, ADC_6_DIR);
break;
case 7:
snprintf(ctrlFile, sizeof(ctrlFile), "%s%s", ADC_DIR_PREFIX, ADC_7_DIR);
break;
default:
Logger::getInstance()->error("adc::adc_get_value: unknown ADC");
return -1;
}
memset(adcVal, 0, sizeof(adcVal));
bytesRead = sysfs_read(ctrlFile, adcVal, sizeof(adcVal));
if (bytesRead < 0)
{
Logger::getInstance()->error("adc::adc_get_value: failed to read ADC sysfs file");
return -1;
}
if (bytesRead >= sizeof(adcVal))
{
Logger::getInstance()->error("adc::adc_get_value: overflow");
return -1;
}
else
{
adcVal[bytesRead+1] = '\0';
}
return atoi(adcVal);
}
static ssize_t dven_read(struct file *filp, char __user *buf,
size_t count, loff_t *pos)
{
char src[] = "0x15b3";
return sysfs_read(buf, count, pos, src);
}
static ssize_t vsd_read(struct file *filp, char __user *buf,
size_t count, loff_t *pos)
{
char *src = "puma20_A1-10.2.3.0";
return sysfs_read(buf, count, pos, src);
}
void unblock_monitor(char *container, const int unfreeze)
{
struct mdstat_ent *ent, *e;
struct mdinfo *sra = NULL;
int to_ping = 0;
ent = mdstat_read(0, 0);
if (!ent) {
fprintf(stderr, Name
": failed to read /proc/mdstat while unblocking container\n");
return;
}
/* unfreeze container contents */
for (e = ent; e; e = e->next) {
if (!is_container_member(e, container))
continue;
sysfs_free(sra);
sra = sysfs_read(-1, e->devnum, GET_VERSION|GET_LEVEL);
if (sra->array.level > 0)
to_ping++;
if (unblock_subarray(sra, unfreeze))
fprintf(stderr, Name ": Failed to unfreeze %s\n", e->dev);
}
if (to_ping)
ping_monitor(container);
sysfs_free(sra);
free_mdstat(ent);
}
static ssize_t dname_read(struct file *filp, char __user *buf,
size_t count, loff_t *pos)
{
struct ib_uverbs_device *uvp;
uvp = (struct ib_uverbs_device *)get_fs_info(filp);
return sysfs_read(buf, count, pos, uvp->ib_dev->name);
}
static ssize_t ib_api_ver_read(struct file *filp, char __user *buf,
size_t count, loff_t *pos)
{
char src[4] = { 0, 0, 0, 0};
src[0] = '0' + IB_USER_VERBS_ABI_VERSION;
return sysfs_read(buf, count, pos, src);
}
static ssize_t mlx4_mgm_read(struct file *filp, char __user *buf,
size_t count, loff_t *pos)
{
#if CONFIG_MLX4_DEFAULT_MGM_LOG_ENTRY_SIZE == -1
char src[4] = { '-', '1', 0, 0 };
#else
char src[4] = { '1', '0', 0, 0 };
#endif
return sysfs_read(buf, count, pos, src);
}
static ssize_t dver_read(struct file *filp, char __user *buf,
size_t count, loff_t *pos)
{
struct ib_uverbs_device *uvp;
char src[4] = { 0, 0, 0, 0};
uvp = (struct ib_uverbs_device *)get_fs_info(filp);
src[0] = '0' + uvp->ib_dev->uverbs_abi_ver;
return sysfs_read(buf, count, pos, src);
}
void power_set_interactive(struct power_module *, int on) {
if (on) {
if (hispeed_freq[0]) {
sysfs_write(HISPEED_FREQ_PATH, hispeed_freq);
}
} else {
if (sysfs_read(HISPEED_FREQ_PATH, hispeed_freq) < 0) {
hispeed_freq[0] = '\0';
}
sysfs_write(HISPEED_FREQ_PATH, HISPEED_FREQ_OFF);
}
}
int get_scaling_governor_check_cores(char governor[], int size,int core_num)
{
if (sysfs_read(scaling_gov_path[core_num], governor,
size) == -1) {
// Can't obtain the scaling governor. Return.
return -1;
}
// Strip newline at the end.
int len = strlen(governor);
len--;
while (len >= 0 && (governor[len] == '\n' || governor[len] == '\r'))
governor[len--] = '\0';
return 0;
}
static int get_scaling_governor() {
if (sysfs_read(SCALING_GOVERNOR_PATH, governor,
sizeof(governor)) == -1) {
return -1;
} else {
// Strip newline at the end.
int len = strlen(governor);
len--;
while (len >= 0 && (governor[len] == '\n' || governor[len] == '\r'))
governor[len--] = '\0';
}
return 0;
}
static int get_scaling_governor(char governor[], int size) {
if (sysfs_read(SCALING_GOVERNOR_PATH, governor,
size) == -1) {
// Can't obtain the scaling governor. Return.
return -1;
} else {
// Strip newline at the end.
int len = strlen(governor);
len--;
while (len >= 0 && (governor[len] == '\n' || governor[len] == '\r'))
governor[len--] = '\0';
}
return 0;
}
static ssize_t cpu_read(struct file *filp, char __user *buf,
size_t count, loff_t *pos)
{
char cpu_info_str[128];
long freq = __proc_global_info.tsc_freq, idx;
strncpy(cpu_info_str, "cpu MHz\t\t: ", 16);
idx = strlen(cpu_info_str);
stradd(cpu_info_str + idx, freq / 1000000, 4);
idx += 4;
strncpy(cpu_info_str + idx, ".", 1);
idx++;
stradd(cpu_info_str + idx, freq % 1000000, 3);
idx += 3;
cpu_info_str[idx] = 0;
return sysfs_read(buf, count, pos, cpu_info_str);
}
void RebuildMap(void)
{
struct mdstat_ent *mdstat = mdstat_read(0, 0);
struct mdstat_ent *md;
struct map_ent *map = NULL;
int require_homehost;
char sys_hostname[256];
char *homehost = conf_get_homehost(&require_homehost);
if (homehost == NULL || strcmp(homehost, "<system>")==0) {
if (gethostname(sys_hostname, sizeof(sys_hostname)) == 0) {
sys_hostname[sizeof(sys_hostname)-1] = 0;
homehost = sys_hostname;
}
}
for (md = mdstat ; md ; md = md->next) {
struct mdinfo *sra = sysfs_read(-1, md->devnm, GET_DEVS);
struct mdinfo *sd;
if (!sra)
continue;
for (sd = sra->devs ; sd ; sd = sd->next) {
char namebuf[100];
char dn[30];
int dfd;
int ok;
int devid;
struct supertype *st;
char *subarray = NULL;
char *path;
struct mdinfo *info;
sprintf(dn, "%d:%d", sd->disk.major, sd->disk.minor);
dfd = dev_open(dn, O_RDONLY);
if (dfd < 0)
continue;
st = guess_super(dfd);
if ( st == NULL)
ok = -1;
else {
subarray = get_member_info(md);
ok = st->ss->load_super(st, dfd, NULL);
}
close(dfd);
if (ok != 0)
continue;
if (subarray)
info = st->ss->container_content(st, subarray);
else {
info = xmalloc(sizeof(*info));
st->ss->getinfo_super(st, info, NULL);
}
if (!info)
continue;
devid = devnm2devid(md->devnm);
path = map_dev(major(devid), minor(devid), 0);
if (path == NULL ||
strncmp(path, "/dev/md/", 8) != 0) {
/* We would really like a name that provides
* an MD_DEVNAME for udev.
* The name needs to be unique both in /dev/md/
* and in this mapfile.
* It needs to match what -I or -As would come
* up with.
* That means:
* Check if array is in mdadm.conf
* - if so use that.
* determine trustworthy from homehost etc
* find a unique name based on metadata name.
*
*/
struct mddev_ident *match = conf_match(st, info,
NULL, 0,
NULL);
struct stat stb;
if (match && match->devname && match->devname[0] == '/') {
path = match->devname;
if (path[0] != '/') {
strcpy(namebuf, "/dev/md/");
strcat(namebuf, path);
path = namebuf;
}
} else {
int unum = 0;
char *sep = "_";
const char *name;
int conflict = 1;
if ((homehost == NULL ||
st->ss->match_home(st, homehost) != 1) &&
st->ss->match_home(st, "any") != 1 &&
(require_homehost
|| ! conf_name_is_free(info->name)))
/* require a numeric suffix */
unum = 0;
else
/* allow name to be used as-is if no conflict */
unum = -1;
name = info->name;
if (!*name) {
name = st->ss->name;
if (!isdigit(name[strlen(name)-1]) &&
unum == -1) {
unum = 0;
sep = "";
}
}
if (strchr(name, ':')) {
/* Probably a uniquifying
* hostname prefix. Allow
* without a suffix, and strip
* hostname if it is us.
*/
if (homehost && unum == -1 &&
strncmp(name, homehost,
strlen(homehost)) == 0 &&
name[strlen(homehost)] == ':')
name += strlen(homehost)+1;
unum = -1;
}
while (conflict) {
if (unum >= 0)
sprintf(namebuf, "/dev/md/%s%s%d",
name, sep, unum);
else
sprintf(namebuf, "/dev/md/%s",
name);
unum++;
if (lstat(namebuf, &stb) != 0 &&
(map == NULL ||
!map_by_name(&map, namebuf+8)))
conflict = 0;
}
path = namebuf;
}
}
map_add(&map, md->devnm,
info->text_version,
info->uuid, path);
st->ss->free_super(st);
free(info);
break;
}
sysfs_free(sra);
}
/* Only trigger a change if we wrote a new map file */
if (map_write(map))
for (md = mdstat ; md ; md = md->next) {
struct mdinfo *sra = sysfs_read(-1, md->devnm,
GET_VERSION);
if (sra)
sysfs_uevent(sra, "change");
sysfs_free(sra);
}
map_free(map);
free_mdstat(mdstat);
}
int Manage_subdevs(char *devname, int fd,
struct mddev_dev *devlist, int verbose, int test,
char *update)
{
/* do something to each dev.
* devmode can be
* 'a' - add the device
* try HOT_ADD_DISK
* If that fails EINVAL, try ADD_NEW_DISK
* 'r' - remove the device HOT_REMOVE_DISK
* device can be 'faulty' or 'detached' in which case all
* matching devices are removed.
* 'f' - set the device faulty SET_DISK_FAULTY
* device can be 'detached' in which case any device that
* is inaccessible will be marked faulty.
* For 'f' and 'r', the device can also be a kernel-internal
* name such as 'sdb'.
*/
struct mddev_dev *add_devlist = NULL;
mdu_array_info_t array;
mdu_disk_info_t disc;
unsigned long long array_size;
struct mddev_dev *dv, *next = NULL;
struct stat stb;
int j, jnext = 0;
int tfd = -1;
struct supertype *st, *tst;
char *subarray = NULL;
int duuid[4];
int ouuid[4];
int lfd = -1;
int sysfd = -1;
int count = 0; /* number of actions taken */
if (ioctl(fd, GET_ARRAY_INFO, &array)) {
fprintf(stderr, Name ": cannot get array info for %s\n",
devname);
return 1;
}
/* array.size is only 32 bit and may be truncated.
* So read from sysfs if possible, and record number of sectors
*/
array_size = get_component_size(fd);
if (array_size <= 0)
array_size = array.size * 2;
tst = super_by_fd(fd, &subarray);
if (!tst) {
fprintf(stderr, Name ": unsupport array - version %d.%d\n",
array.major_version, array.minor_version);
return 1;
}
stb.st_rdev = 0;
for (dv = devlist, j=0 ; dv; dv = next, j = jnext) {
unsigned long long ldsize;
char dvname[20];
char *dnprintable = dv->devname;
char *add_dev = dv->devname;
int err;
int re_add_failed = 0;
next = dv->next;
jnext = 0;
if (strcmp(dv->devname, "failed")==0 ||
strcmp(dv->devname, "faulty")==0) {
int remaining_disks = array.nr_disks;
if (dv->disposition != 'r') {
fprintf(stderr, Name ": %s only meaningful "
"with -r, not -%c\n",
dv->devname, dv->disposition);
return 1;
}
for (; j < 1024 && remaining_disks > 0; j++) {
unsigned dev;
disc.number = j;
if (ioctl(fd, GET_DISK_INFO, &disc))
continue;
if (disc.major == 0 && disc.minor == 0)
continue;
remaining_disks --;
if ((disc.state & 1) == 0) /* faulty */
continue;
dev = makedev(disc.major, disc.minor);
if (stb.st_rdev == dev)
/* already did that one */
continue;
stb.st_rdev = dev;
next = dv;
/* same slot again next time - things might
* have reshuffled */
jnext = j;
sprintf(dvname,"%d:%d", disc.major, disc.minor);
dnprintable = dvname;
break;
}
if (next != dv)
continue;
} else if (strcmp(dv->devname, "detached") == 0) {
int remaining_disks = array.nr_disks;
if (dv->disposition != 'r' && dv->disposition != 'f') {
fprintf(stderr, Name ": %s only meaningful "
"with -r of -f, not -%c\n",
dv->devname, dv->disposition);
return 1;
}
for (; j < 1024 && remaining_disks > 0; j++) {
int sfd;
unsigned dev;
disc.number = j;
if (ioctl(fd, GET_DISK_INFO, &disc))
continue;
if (disc.major == 0 && disc.minor == 0)
continue;
remaining_disks --;
sprintf(dvname,"%d:%d", disc.major, disc.minor);
sfd = dev_open(dvname, O_RDONLY);
if (sfd >= 0) {
close(sfd);
continue;
}
if (dv->disposition == 'f' &&
(disc.state & 1) == 1) /* already faulty */
continue;
if (errno != ENXIO)
continue;
dev = makedev(disc.major, disc.minor);
if (stb.st_rdev == dev)
/* already did that one */
continue;
stb.st_rdev = dev;
next = dv;
/* same slot again next time - things might
* have reshuffled */
jnext = j;
dnprintable = dvname;
break;
}
if (next != dv)
continue;
} else if (strcmp(dv->devname, "missing") == 0) {
if (dv->disposition != 'a' || dv->re_add == 0) {
fprintf(stderr, Name ": 'missing' only meaningful "
"with --re-add\n");
return 1;
}
if (add_devlist == NULL)
add_devlist = conf_get_devs();
if (add_devlist == NULL) {
fprintf(stderr, Name ": no devices to scan for missing members.");
continue;
}
add_dev = add_devlist->devname;
add_devlist = add_devlist->next;
if (add_devlist != NULL)
next = dv;
if (stat(add_dev, &stb) < 0)
continue;
} else if (strchr(dv->devname, '/') == NULL &&
strchr(dv->devname, ':') == NULL &&
strlen(dv->devname) < 50) {
/* Assume this is a kernel-internal name like 'sda1' */
int found = 0;
char dname[55];
if (dv->disposition != 'r' && dv->disposition != 'f') {
fprintf(stderr, Name ": %s only meaningful "
"with -r or -f, not -%c\n",
dv->devname, dv->disposition);
return 1;
}
sprintf(dname, "dev-%s", dv->devname);
sysfd = sysfs_open(fd2devnum(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);
found = 1;
}
close(sysfd);
sysfd = -1;
}
if (!found) {
sysfd = sysfs_open(fd2devnum(fd), dname, "state");
if (sysfd < 0) {
fprintf(stderr, Name ": %s does not appear "
"to be a component of %s\n",
dv->devname, devname);
return 1;
}
}
} else {
j = 0;
tfd = dev_open(dv->devname, O_RDONLY);
if (tfd < 0 && dv->disposition == 'r' &&
lstat(dv->devname, &stb) == 0)
/* Be happy, the lstat worked, that is
* enough for --remove
*/
;
else {
if (tfd < 0 || fstat(tfd, &stb) != 0) {
fprintf(stderr, Name ": cannot find %s: %s\n",
dv->devname, strerror(errno));
if (tfd >= 0)
close(tfd);
return 1;
}
close(tfd);
tfd = -1;
}
if ((stb.st_mode & S_IFMT) != S_IFBLK) {
fprintf(stderr, Name ": %s is not a "
"block device.\n",
dv->devname);
return 1;
}
}
switch(dv->disposition){
default:
fprintf(stderr, Name ": internal error - devmode[%s]=%d\n",
dv->devname, dv->disposition);
return 1;
case 'a':
/* add the device */
if (subarray) {
fprintf(stderr, Name ": Cannot add disks to a"
" \'member\' array, perform this"
" operation on the parent container\n");
return 1;
}
/* Make sure it isn't in use (in 2.6 or later) */
tfd = dev_open(add_dev, O_RDONLY|O_EXCL|O_DIRECT);
if (tfd < 0 && add_dev != dv->devname)
continue;
if (tfd < 0) {
fprintf(stderr, Name ": Cannot open %s: %s\n",
dv->devname, strerror(errno));
return 1;
}
st = dup_super(tst);
if (array.not_persistent==0)
st->ss->load_super(st, tfd, NULL);
if (add_dev == dv->devname) {
if (!get_dev_size(tfd, dv->devname, &ldsize)) {
close(tfd);
return 1;
}
} else if (!get_dev_size(tfd, NULL, &ldsize)) {
close(tfd);
tfd = -1;
continue;
}
if (!tst->ss->external &&
array.major_version == 0 &&
md_get_version(fd)%100 < 2) {
close(tfd);
tfd = -1;
if (ioctl(fd, HOT_ADD_DISK,
(unsigned long)stb.st_rdev)==0) {
if (verbose >= 0)
fprintf(stderr, Name ": hot added %s\n",
add_dev);
continue;
}
fprintf(stderr, Name ": hot add failed for %s: %s\n",
add_dev, strerror(errno));
return 1;
}
if (array.not_persistent == 0 || tst->ss->external) {
/* need to find a sample superblock to copy, and
* a spare slot to use.
* For 'external' array (well, container based),
* We can just load the metadata for the array.
*/
if (tst->sb)
/* already loaded */;
else if (tst->ss->external) {
tst->ss->load_container(tst, fd, NULL);
} else for (j = 0; j < tst->max_devs; j++) {
char *dev;
int dfd;
disc.number = j;
if (ioctl(fd, GET_DISK_INFO, &disc))
continue;
if (disc.major==0 && disc.minor==0)
continue;
if ((disc.state & 4)==0) continue; /* sync */
/* Looks like a good device to try */
dev = map_dev(disc.major, disc.minor, 1);
if (!dev) continue;
dfd = dev_open(dev, O_RDONLY);
if (dfd < 0) continue;
if (tst->ss->load_super(tst, dfd,
NULL)) {
close(dfd);
continue;
}
close(dfd);
break;
}
/* FIXME this is a bad test to be using */
if (!tst->sb) {
close(tfd);
fprintf(stderr, Name ": cannot load array metadata from %s\n", devname);
return 1;
}
/* Make sure device is large enough */
if (tst->ss->avail_size(tst, ldsize/512) <
array_size) {
close(tfd);
tfd = -1;
if (add_dev != dv->devname)
continue;
fprintf(stderr, Name ": %s not large enough to join array\n",
dv->devname);
return 1;
}
/* Possibly this device was recently part of the array
* and was temporarily removed, and is now being re-added.
* If so, we can simply re-add it.
*/
tst->ss->uuid_from_super(tst, duuid);
if (st->sb) {
struct mdinfo mdi;
st->ss->getinfo_super(st, &mdi, NULL);
st->ss->uuid_from_super(st, ouuid);
if ((mdi.disk.state & (1<<MD_DISK_ACTIVE)) &&
!(mdi.disk.state & (1<<MD_DISK_FAULTY)) &&
memcmp(duuid, ouuid, sizeof(ouuid))==0) {
/* look like it is worth a try. Need to
* make sure kernel will accept it though.
*/
/* re-add doesn't work for version-1 superblocks
* before 2.6.18 :-(
*/
if (array.major_version == 1 &&
get_linux_version() <= 2006018)
goto skip_re_add;
disc.number = mdi.disk.number;
if (ioctl(fd, GET_DISK_INFO, &disc) != 0
|| disc.major != 0 || disc.minor != 0
|| !enough_fd(fd))
goto skip_re_add;
disc.major = major(stb.st_rdev);
disc.minor = minor(stb.st_rdev);
disc.number = mdi.disk.number;
disc.raid_disk = mdi.disk.raid_disk;
disc.state = mdi.disk.state;
if (dv->writemostly == 1)
disc.state |= 1 << MD_DISK_WRITEMOSTLY;
if (dv->writemostly == 2)
disc.state &= ~(1 << MD_DISK_WRITEMOSTLY);
remove_partitions(tfd);
close(tfd);
tfd = -1;
if (update) {
int rv = -1;
tfd = dev_open(dv->devname, O_RDWR);
if (tfd >= 0)
rv = st->ss->update_super(
st, NULL, update,
devname, verbose, 0, NULL);
if (rv == 0)
rv = st->ss->store_super(st, tfd);
close(tfd);
tfd = -1;
if (rv != 0) {
fprintf(stderr, Name ": failed to update"
" superblock during re-add\n");
return 1;
}
}
/* don't even try if disk is marked as faulty */
errno = 0;
if (ioctl(fd, ADD_NEW_DISK, &disc) == 0) {
if (verbose >= 0)
fprintf(stderr, Name ": re-added %s\n", add_dev);
count++;
continue;
}
if (errno == ENOMEM || errno == EROFS) {
fprintf(stderr, Name ": add new device failed for %s: %s\n",
add_dev, strerror(errno));
if (add_dev != dv->devname)
continue;
return 1;
}
skip_re_add:
re_add_failed = 1;
}
st->ss->free_super(st);
}
if (add_dev != dv->devname) {
if (verbose > 0)
fprintf(stderr, Name
": --re-add for %s to %s is not possible\n",
add_dev, devname);
if (tfd >= 0) {
close(tfd);
tfd = -1;
}
continue;
}
if (dv->re_add) {
if (tfd >= 0)
close(tfd);
fprintf(stderr, Name
": --re-add for %s to %s is not possible\n",
dv->devname, devname);
return 1;
}
if (re_add_failed) {
fprintf(stderr, Name ": %s reports being an active member for %s, but a --re-add fails.\n",
dv->devname, devname);
fprintf(stderr, Name ": not performing --add as that would convert %s in to a spare.\n",
dv->devname);
fprintf(stderr, Name ": To make this a spare, use \"mdadm --zero-superblock %s\" first.\n",
dv->devname);
if (tfd >= 0)
close(tfd);
return 1;
}
} else {
/* non-persistent. Must ensure that new drive
* is at least array.size big.
*/
if (ldsize/512 < array_size) {
fprintf(stderr, Name ": %s not large enough to join array\n",
dv->devname);
if (tfd >= 0)
close(tfd);
return 1;
}
}
/* committed to really trying this device now*/
if (tfd >= 0) {
remove_partitions(tfd);
close(tfd);
tfd = -1;
}
/* in 2.6.17 and earlier, version-1 superblocks won't
* use the number we write, but will choose a free number.
* we must choose the same free number, which requires
* starting at 'raid_disks' and counting up
*/
for (j = array.raid_disks; j< tst->max_devs; j++) {
disc.number = j;
if (ioctl(fd, GET_DISK_INFO, &disc))
break;
if (disc.major==0 && disc.minor==0)
break;
if (disc.state & 8) /* removed */
break;
}
disc.major = major(stb.st_rdev);
disc.minor = minor(stb.st_rdev);
disc.number =j;
disc.state = 0;
if (array.not_persistent==0) {
int dfd;
if (dv->writemostly == 1)
disc.state |= 1 << MD_DISK_WRITEMOSTLY;
dfd = dev_open(dv->devname, O_RDWR | O_EXCL|O_DIRECT);
if (tst->ss->add_to_super(tst, &disc, dfd,
dv->devname)) {
close(dfd);
return 1;
}
if (tst->ss->write_init_super(tst)) {
close(dfd);
return 1;
}
} else if (dv->re_add) {
/* this had better be raid1.
* As we are "--re-add"ing we must find a spare slot
* to fill.
*/
char *used = malloc(array.raid_disks);
memset(used, 0, array.raid_disks);
for (j=0; j< tst->max_devs; j++) {
mdu_disk_info_t disc2;
disc2.number = j;
if (ioctl(fd, GET_DISK_INFO, &disc2))
continue;
if (disc2.major==0 && disc2.minor==0)
continue;
if (disc2.state & 8) /* removed */
continue;
if (disc2.raid_disk < 0)
continue;
if (disc2.raid_disk > array.raid_disks)
continue;
used[disc2.raid_disk] = 1;
}
for (j=0 ; j<array.raid_disks; j++)
if (!used[j]) {
disc.raid_disk = j;
disc.state |= (1<<MD_DISK_SYNC);
break;
}
free(used);
}
if (dv->writemostly == 1)
disc.state |= (1 << MD_DISK_WRITEMOSTLY);
if (tst->ss->external) {
/* add a disk
* to an external metadata container */
struct mdinfo new_mdi;
struct mdinfo *sra;
int container_fd;
int devnum = fd2devnum(fd);
int dfd;
container_fd = open_dev_excl(devnum);
if (container_fd < 0) {
fprintf(stderr, Name ": add failed for %s:"
" could not get exclusive access to container\n",
dv->devname);
tst->ss->free_super(tst);
return 1;
}
dfd = dev_open(dv->devname, O_RDWR | O_EXCL|O_DIRECT);
if (mdmon_running(tst->container_dev))
tst->update_tail = &tst->updates;
if (tst->ss->add_to_super(tst, &disc, dfd,
dv->devname)) {
close(dfd);
close(container_fd);
return 1;
}
if (tst->update_tail)
flush_metadata_updates(tst);
else
tst->ss->sync_metadata(tst);
sra = sysfs_read(container_fd, -1, 0);
if (!sra) {
fprintf(stderr, Name ": add failed for %s: sysfs_read failed\n",
dv->devname);
close(container_fd);
tst->ss->free_super(tst);
return 1;
}
sra->array.level = LEVEL_CONTAINER;
/* Need to set data_offset and component_size */
tst->ss->getinfo_super(tst, &new_mdi, NULL);
new_mdi.disk.major = disc.major;
new_mdi.disk.minor = disc.minor;
new_mdi.recovery_start = 0;
/* Make sure fds are closed as they are O_EXCL which
* would block add_disk */
tst->ss->free_super(tst);
if (sysfs_add_disk(sra, &new_mdi, 0) != 0) {
fprintf(stderr, Name ": add new device to external metadata"
" failed for %s\n", dv->devname);
close(container_fd);
sysfs_free(sra);
return 1;
}
ping_monitor_by_id(devnum);
sysfs_free(sra);
close(container_fd);
} else {
tst->ss->free_super(tst);
if (ioctl(fd, ADD_NEW_DISK, &disc)) {
fprintf(stderr, Name ": add new device failed for %s as %d: %s\n",
dv->devname, j, strerror(errno));
return 1;
}
}
if (verbose >= 0)
fprintf(stderr, Name ": added %s\n", dv->devname);
break;
case 'r':
/* hot remove */
if (subarray) {
fprintf(stderr, Name ": Cannot remove disks from a"
" \'member\' array, perform this"
" operation on the parent container\n");
if (sysfd >= 0)
close(sysfd);
return 1;
}
if (tst->ss->external) {
/* To remove a device from a container, we must
* check that it isn't in use in an array.
* This involves looking in the 'holders'
* directory - there must be just one entry,
* the container.
* To ensure that it doesn't get used as a
* hold spare while we are checking, we
* get an O_EXCL open on the container
*/
int dnum = fd2devnum(fd);
lfd = open_dev_excl(dnum);
if (lfd < 0) {
fprintf(stderr, Name
": Cannot get exclusive access "
" to container - odd\n");
if (sysfd >= 0)
close(sysfd);
return 1;
}
/* in the detached case it is not possible to
* check if we are the unique holder, so just
* rely on the 'detached' checks
*/
if (strcmp(dv->devname, "detached") == 0 ||
sysfd >= 0 ||
sysfs_unique_holder(dnum, stb.st_rdev))
/* pass */;
else {
fprintf(stderr, Name
": %s is %s, cannot remove.\n",
dnprintable,
errno == EEXIST ? "still in use":
"not a member");
close(lfd);
return 1;
}
}
/* FIXME check that it is a current member */
if (sysfd >= 0) {
/* device has been removed and we don't know
* the major:minor number
*/
int n = write(sysfd, "remove", 6);
if (n != 6)
err = -1;
else
err = 0;
close(sysfd);
sysfd = -1;
} else {
err = ioctl(fd, HOT_REMOVE_DISK, (unsigned long)stb.st_rdev);
if (err && errno == ENODEV) {
/* Old kernels rejected this if no personality
* registered */
struct mdinfo *sra = sysfs_read(fd, 0, GET_DEVS);
struct mdinfo *dv = NULL;
if (sra)
dv = sra->devs;
for ( ; dv ; dv=dv->next)
if (dv->disk.major == (int)major(stb.st_rdev) &&
dv->disk.minor == (int)minor(stb.st_rdev))
break;
if (dv)
err = sysfs_set_str(sra, dv,
"state", "remove");
else
err = -1;
if (sra)
sysfs_free(sra);
}
}
if (err) {
fprintf(stderr, Name ": hot remove failed "
"for %s: %s\n", dnprintable,
strerror(errno));
if (lfd >= 0)
close(lfd);
return 1;
}
if (tst->ss->external) {
/*
* Before dropping our exclusive open we make an
* attempt at preventing mdmon from seeing an
* 'add' event before reconciling this 'remove'
* event.
*/
char *name = devnum2devname(fd2devnum(fd));
if (!name) {
fprintf(stderr, Name ": unable to get container name\n");
return 1;
}
ping_manager(name);
free(name);
}
if (lfd >= 0)
close(lfd);
count++;
if (verbose >= 0)
fprintf(stderr, Name ": hot removed %s from %s\n",
dnprintable, devname);
break;
case 'f': /* set faulty */
/* FIXME check current member */
if ((sysfd >= 0 && write(sysfd, "faulty", 6) != 6) ||
(sysfd < 0 && ioctl(fd, SET_DISK_FAULTY,
(unsigned long) stb.st_rdev))) {
fprintf(stderr, Name ": set device faulty failed for %s: %s\n",
dnprintable, strerror(errno));
if (sysfd >= 0)
close(sysfd);
return 1;
}
if (sysfd >= 0)
close(sysfd);
sysfd = -1;
count++;
if (verbose >= 0)
fprintf(stderr, Name ": set %s faulty in %s\n",
dnprintable, devname);
break;
}
}
if (test && count == 0)
return 2;
return 0;
}
/* Turn file (or device node) into class name */
static int affinity_file(struct bitmask *mask, char *cls, const char *file)
{
struct stat st;
DIR *dir;
int n;
unsigned maj = 0, min = 0;
dev_t d;
struct dirent de, *dep;
cls = "block";
char fn[sizeof("/sys/class/") + strlen(cls)];
if (stat(file, &st) < 0) {
numa_warn(W_blockdev1, "Cannot stat file %s", file);
return -1;
}
d = st.st_dev;
if (S_ISCHR(st.st_mode)) {
/* Better choice than misc? Most likely misc will not work
anyways unless the kernel is fixed. */
cls = "misc";
d = st.st_rdev;
} else if (S_ISBLK(st.st_mode))
d = st.st_rdev;
sprintf(fn, "/sys/class/%s", cls);
dir = opendir(fn);
if (!dir) {
numa_warn(W_blockdev2, "Cannot enumerate %s devices in sysfs",
cls);
return -1;
}
while (readdir_r(dir, &de, &dep) == 0 && dep) {
char *name = dep->d_name;
if (*name == '.')
continue;
char *dev;
char fn2[sizeof("/sys/class/block//dev") + strlen(name)];
n = -1;
if (sprintf(fn2, "/sys/class/block/%s/dev", name) < 0)
break;
dev = sysfs_read(fn2);
if (dev) {
n = sscanf(dev, "%u:%u", &maj, &min);
free(dev);
}
if (n != 2) {
numa_warn(W_blockdev3, "Cannot parse sysfs device %s",
name);
continue;
}
if (major(d) != maj || minor(d) != min)
continue;
closedir(dir);
return affinity_class(mask, "block", name);
}
closedir(dir);
numa_warn(W_blockdev5, "Cannot find block device %x:%x in sysfs for `%s'",
maj, min, file);
return -1;
}
int main(int argc, char *argv[])
{
/* md_device start length */
int *fds = NULL;
char *buf = NULL;
char **disk_name = NULL;
unsigned long long *offsets = NULL;
int raid_disks = 0;
int active_disks;
int chunk_size = 0;
int layout = -1;
int level = 6;
int repair = 0;
int failed_disk1, failed_disk2;
unsigned long long start, length;
int i;
int mdfd;
struct mdinfo *info = NULL, *comp = NULL;
char *err = NULL;
int exit_err = 0;
int close_flag = 0;
char *prg = strrchr(argv[0], '/');
if (prg == NULL)
prg = argv[0];
else
prg++;
if (argc < 4) {
fprintf(stderr, "Usage: %s md_device start_stripe length_stripes [autorepair]\n", prg);
fprintf(stderr, " or: %s md_device repair stripe failed_slot_1 failed_slot_2\n", prg);
exit_err = 1;
goto exitHere;
}
mdfd = open(argv[1], O_RDONLY);
if(mdfd < 0) {
perror(argv[1]);
fprintf(stderr, "%s: cannot open %s\n", prg, argv[1]);
exit_err = 2;
goto exitHere;
}
info = sysfs_read(mdfd, -1,
GET_LEVEL|
GET_LAYOUT|
GET_DISKS|
GET_DEGRADED |
GET_COMPONENT|
GET_CHUNK|
GET_DEVS|
GET_OFFSET|
GET_SIZE);
if(info == NULL) {
fprintf(stderr, "%s: Error reading sysfs information of %s\n", prg, argv[1]);
exit_err = 9;
goto exitHere;
}
if(info->array.level != level) {
fprintf(stderr, "%s: %s not a RAID-6\n", prg, argv[1]);
exit_err = 3;
goto exitHere;
}
if(info->array.failed_disks > 0) {
fprintf(stderr, "%s: %s degraded array\n", prg, argv[1]);
exit_err = 8;
goto exitHere;
}
printf("layout: %d\n", info->array.layout);
printf("disks: %d\n", info->array.raid_disks);
printf("component size: %llu\n", info->component_size * 512);
printf("total stripes: %llu\n", (info->component_size * 512) / info->array.chunk_size);
printf("chunk size: %d\n", info->array.chunk_size);
printf("\n");
comp = info->devs;
for(i = 0, active_disks = 0; active_disks < info->array.raid_disks; i++) {
printf("disk: %d - offset: %llu - size: %llu - name: %s - slot: %d\n",
i, comp->data_offset * 512, comp->component_size * 512,
map_dev(comp->disk.major, comp->disk.minor, 0),
comp->disk.raid_disk);
if(comp->disk.raid_disk >= 0)
active_disks++;
comp = comp->next;
}
printf("\n");
close(mdfd);
raid_disks = info->array.raid_disks;
chunk_size = info->array.chunk_size;
layout = info->array.layout;
if (strcmp(argv[2], "repair")==0) {
if (argc < 6) {
fprintf(stderr, "For repair mode, call %s md_device repair stripe failed_slot_1 failed_slot_2\n", prg);
exit_err = 1;
goto exitHere;
}
repair = 1;
start = getnum(argv[3], &err);
length = 1;
failed_disk1 = getnum(argv[4], &err);
failed_disk2 = getnum(argv[5], &err);
if(failed_disk1 >= info->array.raid_disks) {
fprintf(stderr, "%s: failed_slot_1 index is higher than number of devices in raid\n", prg);
exit_err = 4;
goto exitHere;
}
if(failed_disk2 >= info->array.raid_disks) {
fprintf(stderr, "%s: failed_slot_2 index is higher than number of devices in raid\n", prg);
exit_err = 4;
goto exitHere;
}
if(failed_disk1 == failed_disk2) {
fprintf(stderr, "%s: failed_slot_1 and failed_slot_2 are the same\n", prg);
exit_err = 4;
goto exitHere;
}
}
else {
start = getnum(argv[2], &err);
length = getnum(argv[3], &err);
if (argc >= 5 && strcmp(argv[4], "autorepair")==0)
repair = 2;
}
if (err) {
fprintf(stderr, "%s: Bad number: %s\n", prg, err);
exit_err = 4;
goto exitHere;
}
if(start > ((info->component_size * 512) / chunk_size)) {
start = (info->component_size * 512) / chunk_size;
fprintf(stderr, "%s: start beyond disks size\n", prg);
}
if((length == 0) ||
((length + start) > ((info->component_size * 512) / chunk_size))) {
length = (info->component_size * 512) / chunk_size - start;
}
disk_name = xmalloc(raid_disks * sizeof(*disk_name));
fds = xmalloc(raid_disks * sizeof(*fds));
offsets = xcalloc(raid_disks, sizeof(*offsets));
buf = xmalloc(raid_disks * chunk_size);
for(i=0; i<raid_disks; i++) {
fds[i] = -1;
}
close_flag = 1;
comp = info->devs;
for (i=0, active_disks=0; active_disks<raid_disks; i++) {
int disk_slot = comp->disk.raid_disk;
if(disk_slot >= 0) {
disk_name[disk_slot] = map_dev(comp->disk.major, comp->disk.minor, 0);
offsets[disk_slot] = comp->data_offset * 512;
fds[disk_slot] = open(disk_name[disk_slot], O_RDWR);
if (fds[disk_slot] < 0) {
perror(disk_name[disk_slot]);
fprintf(stderr,"%s: cannot open %s\n", prg, disk_name[disk_slot]);
exit_err = 6;
goto exitHere;
}
active_disks++;
}
comp = comp->next;
}
int rv = check_stripes(info, fds, offsets,
raid_disks, chunk_size, level, layout,
start, length, disk_name, repair, failed_disk1, failed_disk2);
if (rv != 0) {
fprintf(stderr,
"%s: check_stripes returned %d\n", prg, rv);
exit_err = 7;
goto exitHere;
}
exitHere:
if (close_flag)
for(i = 0; i < raid_disks; i++)
close(fds[i]);
free(disk_name);
free(fds);
free(offsets);
free(buf);
exit(exit_err);
}
int Monitor(mddev_dev_t devlist,
char *mailaddr, char *alert_cmd,
int period, int daemonise, int scan, int oneshot,
int dosyslog, int test, char* pidfile, int increments)
{
/*
* Every few seconds, scan every md device looking for changes
* When a change is found, log it, possibly run the alert command,
* and possibly send Email
*
* For each array, we record:
* Update time
* active/working/failed/spare drives
* State of each device.
* %rebuilt if rebuilding
*
* If the update time changes, check out all the data again
* It is possible that we cannot get the state of each device
* due to bugs in the md kernel module.
* We also read /proc/mdstat to get rebuild percent,
* and to get state on all active devices incase of kernel bug.
*
* Events are:
* Fail
* An active device had Faulty set or Active/Sync removed
* FailSpare
* A spare device had Faulty set
* SpareActive
* An active device had a reverse transition
* RebuildStarted
* percent went from -1 to +ve
* RebuildNN
* percent went from below to not-below NN%
* DeviceDisappeared
* Couldn't access a device which was previously visible
*
* if we detect an array with active<raid and spare==0
* we look at other arrays that have same spare-group
* If we find one with active==raid and spare>0,
* and if we can get_disk_info and find a name
* Then we hot-remove and hot-add to the other array
*
* If devlist is NULL, then we can monitor everything because --scan
* was given. We get an initial list from config file and add anything
* that appears in /proc/mdstat
*/
struct state {
char *devname;
int devnum; /* to sync with mdstat info */
long utime;
int err;
char *spare_group;
int active, working, failed, spare, raid;
int expected_spares;
int devstate[MaxDisks];
unsigned devid[MaxDisks];
int percent;
struct state *next;
} *statelist = NULL;
int finished = 0;
struct mdstat_ent *mdstat = NULL;
char *mailfrom = NULL;
if (!mailaddr) {
mailaddr = conf_get_mailaddr();
if (mailaddr && ! scan)
fprintf(stderr, Name ": Monitor using email address \"%s\" from config file\n",
mailaddr);
}
mailfrom = conf_get_mailfrom();
if (!alert_cmd) {
alert_cmd = conf_get_program();
if (alert_cmd && ! scan)
fprintf(stderr, Name ": Monitor using program \"%s\" from config file\n",
alert_cmd);
}
if (scan && !mailaddr && !alert_cmd) {
fprintf(stderr, Name ": No mail address or alert command - not monitoring.\n");
return 1;
}
if (daemonise) {
int pid = fork();
if (pid > 0) {
if (!pidfile)
printf("%d\n", pid);
else {
FILE *pid_file;
pid_file=fopen(pidfile, "w");
if (!pid_file)
perror("cannot create pid file");
else {
fprintf(pid_file,"%d\n", pid);
fclose(pid_file);
}
}
return 0;
}
if (pid < 0) {
perror("daemonise");
return 1;
}
close(0);
open("/dev/null", O_RDWR);
dup2(0,1);
dup2(0,2);
setsid();
}
if (devlist == NULL) {
mddev_ident_t mdlist = conf_get_ident(NULL);
for (; mdlist; mdlist=mdlist->next) {
struct state *st;
if (mdlist->devname == NULL)
continue;
if (strcasecmp(mdlist->devname, "<ignore>") == 0)
continue;
st = malloc(sizeof *st);
if (st == NULL)
continue;
if (mdlist->devname[0] == '/')
st->devname = strdup(mdlist->devname);
else {
st->devname = malloc(8+strlen(mdlist->devname)+1);
strcpy(strcpy(st->devname, "/dev/md/"),
mdlist->devname);
}
st->utime = 0;
st->next = statelist;
st->err = 0;
st->devnum = INT_MAX;
st->percent = -2;
st->expected_spares = mdlist->spare_disks;
if (mdlist->spare_group)
st->spare_group = strdup(mdlist->spare_group);
else
st->spare_group = NULL;
statelist = st;
}
} else {
mddev_dev_t dv;
for (dv=devlist ; dv; dv=dv->next) {
mddev_ident_t mdlist = conf_get_ident(dv->devname);
struct state *st = malloc(sizeof *st);
if (st == NULL)
continue;
st->devname = strdup(dv->devname);
st->utime = 0;
st->next = statelist;
st->err = 0;
st->devnum = INT_MAX;
st->percent = -2;
st->expected_spares = -1;
st->spare_group = NULL;
if (mdlist) {
st->expected_spares = mdlist->spare_disks;
if (mdlist->spare_group)
st->spare_group = strdup(mdlist->spare_group);
}
statelist = st;
}
}
while (! finished) {
int new_found = 0;
struct state *st;
if (mdstat)
free_mdstat(mdstat);
mdstat = mdstat_read(oneshot?0:1, 0);
for (st=statelist; st; st=st->next) {
struct { int state, major, minor; } info[MaxDisks];
mdu_array_info_t array;
struct mdstat_ent *mse = NULL, *mse2;
char *dev = st->devname;
int fd;
int i;
if (test)
alert("TestMessage", dev, NULL, mailaddr, mailfrom, alert_cmd, dosyslog);
fd = open(dev, O_RDONLY);
if (fd < 0) {
if (!st->err)
alert("DeviceDisappeared", dev, NULL,
mailaddr, mailfrom, alert_cmd, dosyslog);
/* fprintf(stderr, Name ": cannot open %s: %s\n",
dev, strerror(errno));
*/ st->err=1;
continue;
}
fcntl(fd, F_SETFD, FD_CLOEXEC);
if (ioctl(fd, GET_ARRAY_INFO, &array)<0) {
if (!st->err)
alert("DeviceDisappeared", dev, NULL,
mailaddr, mailfrom, alert_cmd, dosyslog);
/* fprintf(stderr, Name ": cannot get array info for %s: %s\n",
dev, strerror(errno));
*/ st->err=1;
close(fd);
continue;
}
/* It's much easier to list what array levels can't
* have a device disappear than all of them that can
*/
if (array.level == 0 || array.level == -1) {
if (!st->err)
alert("DeviceDisappeared", dev, "Wrong-Level",
mailaddr, mailfrom, alert_cmd, dosyslog);
st->err = 1;
close(fd);
continue;
}
if (st->devnum == INT_MAX) {
struct stat stb;
if (fstat(fd, &stb) == 0 &&
(S_IFMT&stb.st_mode)==S_IFBLK) {
if (major(stb.st_rdev) == MD_MAJOR)
st->devnum = minor(stb.st_rdev);
else
st->devnum = -1- (minor(stb.st_rdev)>>6);
}
}
for (mse2 = mdstat ; mse2 ; mse2=mse2->next)
if (mse2->devnum == st->devnum) {
mse2->devnum = INT_MAX; /* flag it as "used" */
mse = mse2;
}
if (array.utime == 0)
/* external arrays don't update utime */
array.utime = time(0);
if (st->utime == array.utime &&
st->failed == array.failed_disks &&
st->working == array.working_disks &&
st->spare == array.spare_disks &&
(mse == NULL || (
mse->percent == st->percent
))) {
close(fd);
st->err = 0;
continue;
}
if (st->utime == 0 && /* new array */
mse && /* is in /proc/mdstat */
mse->pattern && strchr(mse->pattern, '_') /* degraded */
)
alert("DegradedArray", dev, NULL, mailaddr, mailfrom, alert_cmd, dosyslog);
if (st->utime == 0 && /* new array */
st->expected_spares > 0 &&
array.spare_disks < st->expected_spares)
alert("SparesMissing", dev, NULL, mailaddr, mailfrom, alert_cmd, dosyslog);
if (mse &&
st->percent == -1 &&
mse->percent >= 0)
alert("RebuildStarted", dev, NULL, mailaddr, mailfrom, alert_cmd, dosyslog);
if (mse &&
st->percent >= 0 &&
mse->percent >= 0 &&
(mse->percent / increments) > (st->percent / increments)) {
char percentalert[15]; // "RebuildNN" (10 chars) or "RebuildStarted" (15 chars)
if((mse->percent / increments) == 0)
snprintf(percentalert, sizeof(percentalert), "RebuildStarted");
else
snprintf(percentalert, sizeof(percentalert), "Rebuild%02d", mse->percent);
alert(percentalert,
dev, NULL, mailaddr, mailfrom, alert_cmd, dosyslog);
}
if (mse &&
mse->percent == -1 &&
st->percent >= 0) {
/* Rebuild/sync/whatever just finished.
* If there is a number in /mismatch_cnt,
* we should report that.
*/
struct mdinfo *sra =
sysfs_read(-1, st->devnum, GET_MISMATCH);
if (sra && sra->mismatch_cnt > 0) {
char cnt[40];
sprintf(cnt, " mismatches found: %d", sra->mismatch_cnt);
alert("RebuildFinished", dev, cnt, mailaddr, mailfrom, alert_cmd, dosyslog);
} else
alert("RebuildFinished", dev, NULL, mailaddr, mailfrom, alert_cmd, dosyslog);
if (sra)
free(sra);
}
if (mse)
st->percent = mse->percent;
for (i=0; i<MaxDisks && i <= array.raid_disks + array.nr_disks;
i++) {
mdu_disk_info_t disc;
disc.number = i;
if (ioctl(fd, GET_DISK_INFO, &disc) >= 0) {
info[i].state = disc.state;
info[i].major = disc.major;
info[i].minor = disc.minor;
} else
info[i].major = info[i].minor = 0;
}
close(fd);
for (i=0; i<MaxDisks; i++) {
mdu_disk_info_t disc = {0,0,0,0,0};
int newstate=0;
int change;
char *dv = NULL;
disc.number = i;
if (i > array.raid_disks + array.nr_disks) {
newstate = 0;
disc.major = disc.minor = 0;
} else if (info[i].major || info[i].minor) {
newstate = info[i].state;
dv = map_dev(info[i].major, info[i].minor, 1);
disc.state = newstate;
disc.major = info[i].major;
disc.minor = info[i].minor;
} else if (mse && mse->pattern && i < (int)strlen(mse->pattern)) {
switch(mse->pattern[i]) {
case 'U': newstate = 6 /* ACTIVE/SYNC */; break;
case '_': newstate = 0; break;
}
disc.major = disc.minor = 0;
}
if (dv == NULL && st->devid[i])
dv = map_dev(major(st->devid[i]),
minor(st->devid[i]), 1);
change = newstate ^ st->devstate[i];
if (st->utime && change && !st->err) {
if (i < array.raid_disks &&
(((newstate&change)&(1<<MD_DISK_FAULTY)) ||
((st->devstate[i]&change)&(1<<MD_DISK_ACTIVE)) ||
((st->devstate[i]&change)&(1<<MD_DISK_SYNC)))
)
alert("Fail", dev, dv, mailaddr, mailfrom, alert_cmd, dosyslog);
else if (i >= array.raid_disks &&
(disc.major || disc.minor) &&
st->devid[i] == makedev(disc.major, disc.minor) &&
((newstate&change)&(1<<MD_DISK_FAULTY))
)
alert("FailSpare", dev, dv, mailaddr, mailfrom, alert_cmd, dosyslog);
else if (i < array.raid_disks &&
! (newstate & (1<<MD_DISK_REMOVED)) &&
(((st->devstate[i]&change)&(1<<MD_DISK_FAULTY)) ||
((newstate&change)&(1<<MD_DISK_ACTIVE)) ||
((newstate&change)&(1<<MD_DISK_SYNC)))
)
alert("SpareActive", dev, dv, mailaddr, mailfrom, alert_cmd, dosyslog);
}
st->devstate[i] = newstate;
st->devid[i] = makedev(disc.major, disc.minor);
}
st->active = array.active_disks;
st->working = array.working_disks;
st->spare = array.spare_disks;
st->failed = array.failed_disks;
st->utime = array.utime;
st->raid = array.raid_disks;
st->err = 0;
}
/* now check if there are any new devices found in mdstat */
if (scan) {
struct mdstat_ent *mse;
for (mse=mdstat; mse; mse=mse->next)
if (mse->devnum != INT_MAX &&
mse->level &&
(strcmp(mse->level, "raid0")!=0 &&
strcmp(mse->level, "linear")!=0)
) {
struct state *st = malloc(sizeof *st);
mdu_array_info_t array;
int fd;
if (st == NULL)
continue;
st->devname = strdup(get_md_name(mse->devnum));
if ((fd = open(st->devname, O_RDONLY)) < 0 ||
ioctl(fd, GET_ARRAY_INFO, &array)< 0) {
/* no such array */
if (fd >=0) close(fd);
put_md_name(st->devname);
free(st->devname);
free(st);
continue;
}
close(fd);
st->utime = 0;
st->next = statelist;
st->err = 1;
st->devnum = mse->devnum;
st->percent = -2;
st->spare_group = NULL;
st->expected_spares = -1;
statelist = st;
if (test)
alert("TestMessage", st->devname, NULL, mailaddr, mailfrom, alert_cmd, dosyslog);
alert("NewArray", st->devname, NULL, mailaddr, mailfrom, alert_cmd, dosyslog);
new_found = 1;
}
}
/* If an array has active < raid && spare == 0 && spare_group != NULL
* Look for another array with spare > 0 and active == raid and same spare_group
* if found, choose a device and hotremove/hotadd
*/
for (st = statelist; st; st=st->next)
if (st->active < st->raid &&
st->spare == 0 &&
st->spare_group != NULL) {
struct state *st2;
for (st2=statelist ; st2 ; st2=st2->next)
if (st2 != st &&
st2->spare > 0 &&
st2->active == st2->raid &&
st2->spare_group != NULL &&
strcmp(st->spare_group, st2->spare_group) == 0) {
/* try to remove and add */
int fd1 = open(st->devname, O_RDONLY);
int fd2 = open(st2->devname, O_RDONLY);
int dev = -1;
int d;
if (fd1 < 0 || fd2 < 0) {
if (fd1>=0) close(fd1);
if (fd2>=0) close(fd2);
continue;
}
for (d=st2->raid; d < MaxDisks; d++) {
if (st2->devid[d] > 0 &&
st2->devstate[d] == 0) {
dev = st2->devid[d];
break;
}
}
if (dev > 0) {
struct mddev_dev_s devlist;
char devname[20];
devlist.next = NULL;
devlist.used = 0;
devlist.re_add = 0;
devlist.writemostly = 0;
devlist.devname = devname;
sprintf(devname, "%d:%d", major(dev), minor(dev));
devlist.disposition = 'r';
if (Manage_subdevs(st2->devname, fd2, &devlist, -1, 0) == 0) {
devlist.disposition = 'a';
if (Manage_subdevs(st->devname, fd1, &devlist, -1, 0) == 0) {
alert("MoveSpare", st->devname, st2->devname, mailaddr, mailfrom, alert_cmd, dosyslog);
close(fd1);
close(fd2);
break;
}
else Manage_subdevs(st2->devname, fd2, &devlist, -1, 0);
}
}
close(fd1);
close(fd2);
}
}
if (!new_found) {
if (oneshot)
break;
else
mdstat_wait(period);
}
test = 0;
}
if (pidfile)
unlink(pidfile);
return 0;
}
uint64_t ad_read(int idx) {
char tmp[256];
snprintf(tmp, 256, "/sys/bus/iio/devices/iio:device0/in_voltage%d_raw",idx);
return (uint64_t) sysfs_read(tmp);
}
int WaitClean(char *dev, int sock, int verbose)
{
int fd;
struct mdinfo *mdi;
int rv = 1;
char devnm[32];
fd = open(dev, O_RDONLY);
if (fd < 0) {
if (verbose)
pr_err("Couldn't open %s: %s\n", dev, strerror(errno));
return 1;
}
strcpy(devnm, fd2devnm(fd));
mdi = sysfs_read(fd, devnm, GET_VERSION|GET_LEVEL|GET_SAFEMODE);
if (!mdi) {
if (verbose)
pr_err("Failed to read sysfs attributes for %s\n", dev);
close(fd);
return 0;
}
switch(mdi->array.level) {
case LEVEL_LINEAR:
case LEVEL_MULTIPATH:
case 0:
/* safemode delay is irrelevant for these levels */
rv = 0;
}
/* for internal metadata the kernel handles the final clean
* transition, containers can never be dirty
*/
if (!is_subarray(mdi->text_version))
rv = 0;
/* safemode disabled ? */
if (mdi->safe_mode_delay == 0)
rv = 0;
if (rv) {
int state_fd = sysfs_open(fd2devnm(fd), NULL, "array_state");
char buf[20];
int delay = 5000;
/* minimize the safe_mode_delay and prepare to wait up to 5s
* for writes to quiesce
*/
sysfs_set_safemode(mdi, 1);
/* wait for array_state to be clean */
while (1) {
rv = read(state_fd, buf, sizeof(buf));
if (rv < 0)
break;
if (sysfs_match_word(buf, clean_states) <= 4)
break;
rv = sysfs_wait(state_fd, &delay);
if (rv < 0 && errno != EINTR)
break;
lseek(state_fd, 0, SEEK_SET);
}
if (rv < 0)
rv = 1;
else if (fping_monitor(sock) == 0 ||
ping_monitor(mdi->text_version) == 0) {
/* we need to ping to close the window between array
* state transitioning to clean and the metadata being
* marked clean
*/
rv = 0;
} else
rv = 1;
if (rv && verbose)
pr_err("Error waiting for %s to be clean\n",
dev);
/* restore the original safe_mode_delay */
sysfs_set_safemode(mdi, mdi->safe_mode_delay);
close(state_fd);
}
sysfs_free(mdi);
close(fd);
return rv;
}
static int check_array(struct state *st, struct mdstat_ent *mdstat,
int test, struct alert_info *ainfo,
int increments, char *prefer)
{
/* Update the state 'st' to reflect any changes shown in mdstat,
* or found by directly examining the array, and return
* '1' if the array is degraded, or '0' if it is optimal (or dead).
*/
struct { int state, major, minor; } info[MAX_DISKS];
mdu_array_info_t array;
struct mdstat_ent *mse = NULL, *mse2;
char *dev = st->devname;
int fd = -1;
int i;
int remaining_disks;
int last_disk;
int new_array = 0;
if (test)
alert("TestMessage", dev, NULL, ainfo);
if (st->devnm[0])
fd = open("/sys/block", O_RDONLY|O_DIRECTORY);
if (fd >= 0) {
/* Don't open the device unless it is present and
* active in sysfs.
*/
char buf[10];
close(fd);
fd = sysfs_open(st->devnm, NULL, "array_state");
if (fd < 0 ||
read(fd, buf, 10) < 5 ||
strncmp(buf,"clear",5) == 0 ||
strncmp(buf,"inact",5) == 0) {
if (fd >= 0)
close(fd);
fd = sysfs_open(st->devnm, NULL, "level");
if (fd < 0 || read(fd, buf, 10) != 0) {
if (fd >= 0)
close(fd);
if (!st->err)
alert("DeviceDisappeared", dev, NULL, ainfo);
st->err++;
return 0;
}
}
close(fd);
}
fd = open(dev, O_RDONLY);
if (fd < 0) {
if (!st->err)
alert("DeviceDisappeared", dev, NULL, ainfo);
st->err++;
return 0;
}
fcntl(fd, F_SETFD, FD_CLOEXEC);
if (ioctl(fd, GET_ARRAY_INFO, &array)<0) {
if (!st->err)
alert("DeviceDisappeared", dev, NULL, ainfo);
st->err++;
close(fd);
return 0;
}
/* It's much easier to list what array levels can't
* have a device disappear than all of them that can
*/
if (array.level == 0 || array.level == -1) {
if (!st->err && !st->from_config)
alert("DeviceDisappeared", dev, " Wrong-Level", ainfo);
st->err++;
close(fd);
return 0;
}
if (st->devnm[0] == 0)
strcpy(st->devnm, fd2devnm(fd));
for (mse2 = mdstat ; mse2 ; mse2=mse2->next)
if (strcmp(mse2->devnm, st->devnm) == 0) {
mse2->devnm[0] = 0; /* flag it as "used" */
mse = mse2;
}
if (!mse) {
/* duplicated array in statelist
* or re-created after reading mdstat*/
st->err++;
close(fd);
return 0;
}
/* this array is in /proc/mdstat */
if (array.utime == 0)
/* external arrays don't update utime, so
* just make sure it is always different. */
array.utime = st->utime + 1;;
if (st->err) {
/* New array appeared where previously had an error */
st->err = 0;
st->percent = RESYNC_NONE;
new_array = 1;
alert("NewArray", st->devname, NULL, ainfo);
}
if (st->utime == array.utime &&
st->failed == array.failed_disks &&
st->working == array.working_disks &&
st->spare == array.spare_disks &&
(mse == NULL || (
mse->percent == st->percent
))) {
close(fd);
if ((st->active < st->raid) && st->spare == 0)
return 1;
else
return 0;
}
if (st->utime == 0 && /* new array */
mse->pattern && strchr(mse->pattern, '_') /* degraded */
)
alert("DegradedArray", dev, NULL, ainfo);
if (st->utime == 0 && /* new array */
st->expected_spares > 0 &&
array.spare_disks < st->expected_spares)
alert("SparesMissing", dev, NULL, ainfo);
if (st->percent < 0 && st->percent != RESYNC_UNKNOWN &&
mse->percent >= 0)
alert("RebuildStarted", dev, NULL, ainfo);
if (st->percent >= 0 &&
mse->percent >= 0 &&
(mse->percent / increments) > (st->percent / increments)) {
char percentalert[15]; // "RebuildNN" (10 chars) or "RebuildStarted" (15 chars)
if((mse->percent / increments) == 0)
snprintf(percentalert, sizeof(percentalert), "RebuildStarted");
else
snprintf(percentalert, sizeof(percentalert), "Rebuild%02d", mse->percent);
alert(percentalert, dev, NULL, ainfo);
}
if (mse->percent == RESYNC_NONE &&
st->percent >= 0) {
/* Rebuild/sync/whatever just finished.
* If there is a number in /mismatch_cnt,
* we should report that.
*/
struct mdinfo *sra =
sysfs_read(-1, st->devnm, GET_MISMATCH);
if (sra && sra->mismatch_cnt > 0) {
char cnt[80];
snprintf(cnt, sizeof(cnt),
" mismatches found: %d (on raid level %d)",
sra->mismatch_cnt, array.level);
alert("RebuildFinished", dev, cnt, ainfo);
} else
alert("RebuildFinished", dev, NULL, ainfo);
if (sra)
free(sra);
}
st->percent = mse->percent;
remaining_disks = array.nr_disks;
for (i=0; i<MAX_DISKS && remaining_disks > 0;
i++) {
mdu_disk_info_t disc;
disc.number = i;
if (ioctl(fd, GET_DISK_INFO, &disc) >= 0) {
info[i].state = disc.state;
info[i].major = disc.major;
info[i].minor = disc.minor;
if (disc.major || disc.minor)
remaining_disks --;
} else
info[i].major = info[i].minor = 0;
}
last_disk = i;
if (mse->metadata_version &&
strncmp(mse->metadata_version, "external:", 9) == 0 &&
is_subarray(mse->metadata_version+9)) {
char *sl;
strcpy(st->parent_devnm,
mse->metadata_version+10);
sl = strchr(st->parent_devnm, '/');
if (sl)
*sl = 0;
} else
st->parent_devnm[0] = 0;
if (st->metadata == NULL &&
st->parent_devnm[0] == 0)
st->metadata = super_by_fd(fd, NULL);
close(fd);
for (i=0; i<MAX_DISKS; i++) {
mdu_disk_info_t disc = {0,0,0,0,0};
int newstate=0;
int change;
char *dv = NULL;
disc.number = i;
if (i < last_disk &&
(info[i].major || info[i].minor)) {
newstate = info[i].state;
dv = map_dev_preferred(
info[i].major, info[i].minor, 1,
prefer);
disc.state = newstate;
disc.major = info[i].major;
disc.minor = info[i].minor;
} else
newstate = (1 << MD_DISK_REMOVED);
if (dv == NULL && st->devid[i])
dv = map_dev_preferred(
major(st->devid[i]),
minor(st->devid[i]), 1, prefer);
change = newstate ^ st->devstate[i];
if (st->utime && change && !st->err && !new_array) {
if ((st->devstate[i]&change)&(1<<MD_DISK_SYNC))
alert("Fail", dev, dv, ainfo);
else if ((newstate & (1<<MD_DISK_FAULTY)) &&
(disc.major || disc.minor) &&
st->devid[i] == makedev(disc.major, disc.minor))
alert("FailSpare", dev, dv, ainfo);
else if ((newstate&change)&(1<<MD_DISK_SYNC))
alert("SpareActive", dev, dv, ainfo);
}
st->devstate[i] = newstate;
st->devid[i] = makedev(disc.major, disc.minor);
}
st->active = array.active_disks;
st->working = array.working_disks;
st->spare = array.spare_disks;
st->failed = array.failed_disks;
st->utime = array.utime;
st->raid = array.raid_disks;
st->err = 0;
if ((st->active < st->raid) && st->spare == 0)
return 1;
return 0;
}
JsVarInt sysfs_read_int(const char *path) {
char buf[20];
sysfs_read(path, buf, sizeof(buf));
return stringToIntWithRadix(buf, 10, 0);
}
static int mdmon(char *devnm, int must_fork, int takeover)
{
int mdfd;
struct mdinfo *mdi, *di;
struct supertype *container;
sigset_t set;
struct sigaction act;
int pfd[2];
int status;
int ignore;
pid_t victim = -1;
int victim_sock = -1;
dprintf("starting mdmon for %s\n", devnm);
mdfd = open_dev(devnm);
if (mdfd < 0) {
pr_err("%s: %s\n", devnm, strerror(errno));
return 1;
}
if (md_get_version(mdfd) < 0) {
pr_err("%s: Not an md device\n", devnm);
return 1;
}
/* Fork, and have the child tell us when they are ready */
if (must_fork) {
if (pipe(pfd) != 0) {
pr_err("failed to create pipe\n");
return 1;
}
switch(fork()) {
case -1:
pr_err("failed to fork: %s\n", strerror(errno));
return 1;
case 0: /* child */
close(pfd[0]);
break;
default: /* parent */
close(pfd[1]);
if (read(pfd[0], &status, sizeof(status)) != sizeof(status)) {
wait(&status);
status = WEXITSTATUS(status);
}
close(pfd[0]);
return status;
}
} else
pfd[0] = pfd[1] = -1;
container = xcalloc(1, sizeof(*container));
strcpy(container->devnm, devnm);
container->arrays = NULL;
container->sock = -1;
mdi = sysfs_read(mdfd, container->devnm, GET_VERSION|GET_LEVEL|GET_DEVS);
if (!mdi) {
pr_err("failed to load sysfs info for %s\n", container->devnm);
exit(3);
}
if (mdi->array.level != UnSet) {
pr_err("%s is not a container - cannot monitor\n", devnm);
exit(3);
}
if (mdi->array.major_version != -1 ||
mdi->array.minor_version != -2) {
pr_err("%s does not use external metadata - cannot monitor\n",
devnm);
exit(3);
}
container->ss = version_to_superswitch(mdi->text_version);
if (container->ss == NULL) {
pr_err("%s uses unsupported metadata: %s\n",
devnm, mdi->text_version);
exit(3);
}
container->devs = NULL;
for (di = mdi->devs; di; di = di->next) {
struct mdinfo *cd = xmalloc(sizeof(*cd));
*cd = *di;
cd->next = container->devs;
container->devs = cd;
}
sysfs_free(mdi);
/* SIGUSR is sent between parent and child. So both block it
* and enable it only with pselect.
*/
sigemptyset(&set);
sigaddset(&set, SIGUSR1);
sigaddset(&set, SIGTERM);
sigprocmask(SIG_BLOCK, &set, NULL);
act.sa_handler = wake_me;
act.sa_flags = 0;
sigaction(SIGUSR1, &act, NULL);
act.sa_handler = term;
sigaction(SIGTERM, &act, NULL);
act.sa_handler = SIG_IGN;
sigaction(SIGPIPE, &act, NULL);
victim = mdmon_pid(container->devnm);
if (victim >= 0)
victim_sock = connect_monitor(container->devnm);
ignore = chdir("/");
if (!takeover && victim > 0 && victim_sock >= 0) {
if (fping_monitor(victim_sock) == 0) {
pr_err("%s already managed\n", container->devnm);
exit(3);
}
close(victim_sock);
victim_sock = -1;
}
if (container->ss->load_container(container, mdfd, devnm)) {
pr_err("Cannot load metadata for %s\n", devnm);
exit(3);
}
close(mdfd);
/* Ok, this is close enough. We can say goodbye to our parent now.
*/
if (victim > 0)
remove_pidfile(devnm);
if (make_pidfile(devnm) < 0) {
exit(3);
}
container->sock = make_control_sock(devnm);
status = 0;
if (pfd[1] >= 0) {
if (write(pfd[1], &status, sizeof(status)) < 0)
pr_err("failed to notify our parent: %d\n",
getppid());
close(pfd[1]);
}
mlockall(MCL_CURRENT | MCL_FUTURE);
if (clone_monitor(container) < 0) {
pr_err("failed to start monitor process: %s\n",
strerror(errno));
exit(2);
}
if (victim > 0) {
try_kill_monitor(victim, container->devnm, victim_sock);
if (victim_sock >= 0)
close(victim_sock);
}
setsid();
close(0);
open("/dev/null", O_RDWR);
close(1);
ignore = dup(0);
#ifndef DEBUG
close(2);
ignore = dup(0);
#endif
/* This silliness is to stop the compiler complaining
* that we ignore 'ignore'
*/
if (ignore)
ignore++;
do_manager(container);
exit(0);
}
int Manage_ro(char *devname, int fd, int readonly)
{
/* switch to readonly or rw
*
* requires >= 0.90.0
* first check that array is runing
* use RESTART_ARRAY_RW or STOP_ARRAY_RO
*
*/
mdu_array_info_t array;
#ifndef MDASSEMBLE
struct mdinfo *mdi;
#endif
if (md_get_version(fd) < 9000) {
fprintf(stderr, Name ": need md driver version 0.90.0 or later\n");
return 1;
}
#ifndef MDASSEMBLE
/* If this is an externally-manage array, we need to modify the
* metadata_version so that mdmon doesn't undo our change.
*/
mdi = sysfs_read(fd, -1, GET_LEVEL|GET_VERSION);
if (mdi &&
mdi->array.major_version == -1 &&
is_subarray(mdi->text_version)) {
char vers[64];
strcpy(vers, "external:");
strcat(vers, mdi->text_version);
if (readonly > 0) {
int rv;
/* We set readonly ourselves. */
vers[9] = '-';
sysfs_set_str(mdi, NULL, "metadata_version", vers);
close(fd);
rv = sysfs_set_str(mdi, NULL, "array_state", "readonly");
if (rv < 0) {
fprintf(stderr, Name ": failed to set readonly for %s: %s\n",
devname, strerror(errno));
vers[9] = mdi->text_version[0];
sysfs_set_str(mdi, NULL, "metadata_version", vers);
return 1;
}
} else {
char *cp;
/* We cannot set read/write - must signal mdmon */
vers[9] = '/';
sysfs_set_str(mdi, NULL, "metadata_version", vers);
cp = strchr(vers+10, '/');
if (*cp)
*cp = 0;
ping_monitor(vers+10);
if (mdi->array.level <= 0)
sysfs_set_str(mdi, NULL, "array_state", "active");
}
return 0;
}
#endif
if (ioctl(fd, GET_ARRAY_INFO, &array)) {
fprintf(stderr, Name ": %s does not appear to be active.\n",
devname);
return 1;
}
if (readonly>0) {
if (ioctl(fd, STOP_ARRAY_RO, NULL)) {
fprintf(stderr, Name ": failed to set readonly for %s: %s\n",
devname, strerror(errno));
return 1;
}
} else if (readonly < 0) {
if (ioctl(fd, RESTART_ARRAY_RW, NULL)) {
fprintf(stderr, Name ": failed to set writable for %s: %s\n",
devname, strerror(errno));
return 1;
}
}
return 0;
}
/**
* block_monitor - prevent mdmon spare assignment
* @container - container to block
* @freeze - flag to additionally freeze sync_action
*
* This is used by the reshape code to freeze the container, and the
* auto-rebuild implementation to atomically move spares.
* In both cases we need to stop mdmon from assigning spares to replace
* failed devices as we might have other plans for the spare.
* For the reshape case we also need to 'freeze' sync_action so that
* no recovery happens until we have fully prepared for the reshape.
*
* We tell mdmon that the array is frozen by marking the 'metadata' name
* with a leading '-'. The previously told mdmon "Don't make this array
* read/write, leave it readonly". Now it means a more general "Don't
* reconfigure this array at all".
* As older versions of mdmon (which might run from initrd) don't understand
* this, we first check that the running mdmon is new enough.
*/
int block_monitor(char *container, const int freeze)
{
int devnum = devname2devnum(container);
struct mdstat_ent *ent, *e, *e2;
struct mdinfo *sra = NULL;
char *version = NULL;
char buf[64];
int rv = 0;
if (!mdmon_running(devnum)) {
/* if mdmon is not active we assume that any instance that is
* later started will match the current mdadm version, if this
* assumption is violated we may inadvertantly rebuild an array
* that was meant for reshape, or start rebuild on a spare that
* was to be moved to another container
*/
/* pass */;
} else {
int ver;
version = ping_monitor_version(container);
ver = version ? mdadm_version(version) : -1;
free(version);
if (ver < 3002000) {
fprintf(stderr, Name
": mdmon instance for %s cannot be disabled\n",
container);
return -1;
}
}
ent = mdstat_read(0, 0);
if (!ent) {
fprintf(stderr, Name
": failed to read /proc/mdstat while disabling mdmon\n");
return -1;
}
/* freeze container contents */
for (e = ent; e; e = e->next) {
if (!is_container_member(e, container))
continue;
sysfs_free(sra);
sra = sysfs_read(-1, e->devnum, GET_VERSION);
if (!sra) {
fprintf(stderr, Name
": failed to read sysfs for subarray%s\n",
to_subarray(e, container));
break;
}
/* can't reshape an array that we can't monitor */
if (sra->text_version[0] == '-')
break;
if (freeze && sysfs_freeze_array(sra) < 1)
break;
/* flag this array to not be modified by mdmon (close race with
* takeover in reshape case and spare reassignment in the
* auto-rebuild case)
*/
if (block_subarray(sra))
break;
ping_monitor(container);
/* check that we did not race with recovery */
if ((freeze &&
!sysfs_attribute_available(sra, NULL, "sync_action")) ||
(freeze &&
sysfs_attribute_available(sra, NULL, "sync_action") &&
sysfs_get_str(sra, NULL, "sync_action", buf, 20) > 0 &&
strcmp(buf, "frozen\n") == 0))
/* pass */;
else {
unblock_subarray(sra, 0);
break;
}
/* Double check against races - there should be no spares
* or part-spares
*/
sysfs_free(sra);
sra = sysfs_read(-1, e->devnum, GET_DEVS | GET_STATE);
if (sra && sra->array.spare_disks > 0) {
unblock_subarray(sra, freeze);
break;
}
}
if (e) {
fprintf(stderr, Name ": failed to freeze subarray%s\n",
to_subarray(e, container));
/* thaw the partially frozen container */
for (e2 = ent; e2 && e2 != e; e2 = e2->next) {
if (!is_container_member(e2, container))
continue;
sysfs_free(sra);
sra = sysfs_read(-1, e2->devnum, GET_VERSION);
if (unblock_subarray(sra, freeze))
fprintf(stderr, Name ": Failed to unfreeze %s\n", e2->dev);
}
ping_monitor(container); /* cleared frozen */
rv = -1;
}
sysfs_free(sra);
free_mdstat(ent);
return rv;
}
void set_interactive(struct power_module *module, int on)
{
char governor[80];
char tmp_str[NODE_MAX];
struct video_encode_metadata_t video_encode_metadata;
int rc;
if (set_interactive_override(module, on) == HINT_HANDLED) {
return;
}
ALOGI("Got set_interactive hint");
if (get_scaling_governor(governor, sizeof(governor)) == -1) {
ALOGE("Can't obtain scaling governor.");
return;
}
if (!on) {
/* Display off. */
if ((strncmp(governor, ONDEMAND_GOVERNOR, strlen(ONDEMAND_GOVERNOR)) == 0) &&
(strlen(governor) == strlen(ONDEMAND_GOVERNOR))) {
int resource_values[] = {DISPLAY_OFF, MS_500, THREAD_MIGRATION_SYNC_OFF};
if (!display_hint_sent) {
perform_hint_action(DISPLAY_STATE_HINT_ID,
resource_values, sizeof(resource_values)/sizeof(resource_values[0]));
display_hint_sent = 1;
}
} else if ((strncmp(governor, INTERACTIVE_GOVERNOR, strlen(INTERACTIVE_GOVERNOR)) == 0) &&
(strlen(governor) == strlen(INTERACTIVE_GOVERNOR))) {
int resource_values[] = {TR_MS_500, THREAD_MIGRATION_SYNC_OFF};
if (!display_hint_sent) {
perform_hint_action(DISPLAY_STATE_HINT_ID,
resource_values, sizeof(resource_values)/sizeof(resource_values[0]));
display_hint_sent = 1;
}
} else if ((strncmp(governor, MSMDCVS_GOVERNOR, strlen(MSMDCVS_GOVERNOR)) == 0) &&
(strlen(governor) == strlen(MSMDCVS_GOVERNOR))) {
if (saved_interactive_mode == 1){
/* Display turned off. */
if (sysfs_read(DCVS_CPU0_SLACK_MAX_NODE, tmp_str, NODE_MAX - 1)) {
if (!slack_node_rw_failed) {
ALOGE("Failed to read from %s", DCVS_CPU0_SLACK_MAX_NODE);
}
rc = 1;
} else {
saved_dcvs_cpu0_slack_max = atoi(tmp_str);
}
if (sysfs_read(DCVS_CPU0_SLACK_MIN_NODE, tmp_str, NODE_MAX - 1)) {
if (!slack_node_rw_failed) {
ALOGE("Failed to read from %s", DCVS_CPU0_SLACK_MIN_NODE);
}
rc = 1;
} else {
saved_dcvs_cpu0_slack_min = atoi(tmp_str);
}
if (sysfs_read(MPDECISION_SLACK_MAX_NODE, tmp_str, NODE_MAX - 1)) {
if (!slack_node_rw_failed) {
ALOGE("Failed to read from %s", MPDECISION_SLACK_MAX_NODE);
}
rc = 1;
} else {
saved_mpdecision_slack_max = atoi(tmp_str);
}
if (sysfs_read(MPDECISION_SLACK_MIN_NODE, tmp_str, NODE_MAX - 1)) {
if(!slack_node_rw_failed) {
ALOGE("Failed to read from %s", MPDECISION_SLACK_MIN_NODE);
}
rc = 1;
} else {
saved_mpdecision_slack_min = atoi(tmp_str);
}
/* Write new values. */
if (saved_dcvs_cpu0_slack_max != -1) {
snprintf(tmp_str, NODE_MAX, "%d", 10 * saved_dcvs_cpu0_slack_max);
if (sysfs_write(DCVS_CPU0_SLACK_MAX_NODE, tmp_str) != 0) {
if (!slack_node_rw_failed) {
ALOGE("Failed to write to %s", DCVS_CPU0_SLACK_MAX_NODE);
}
rc = 1;
}
}
if (saved_dcvs_cpu0_slack_min != -1) {
snprintf(tmp_str, NODE_MAX, "%d", 10 * saved_dcvs_cpu0_slack_min);
if (sysfs_write(DCVS_CPU0_SLACK_MIN_NODE, tmp_str) != 0) {
if(!slack_node_rw_failed) {
ALOGE("Failed to write to %s", DCVS_CPU0_SLACK_MIN_NODE);
}
rc = 1;
}
}
if (saved_mpdecision_slack_max != -1) {
snprintf(tmp_str, NODE_MAX, "%d", 10 * saved_mpdecision_slack_max);
if (sysfs_write(MPDECISION_SLACK_MAX_NODE, tmp_str) != 0) {
if(!slack_node_rw_failed) {
ALOGE("Failed to write to %s", MPDECISION_SLACK_MAX_NODE);
}
rc = 1;
}
}
if (saved_mpdecision_slack_min != -1) {
snprintf(tmp_str, NODE_MAX, "%d", 10 * saved_mpdecision_slack_min);
if (sysfs_write(MPDECISION_SLACK_MIN_NODE, tmp_str) != 0) {
if(!slack_node_rw_failed) {
ALOGE("Failed to write to %s", MPDECISION_SLACK_MIN_NODE);
}
rc = 1;
}
}
}
slack_node_rw_failed = rc;
}
} else {
/* Display on. */
if ((strncmp(governor, ONDEMAND_GOVERNOR, strlen(ONDEMAND_GOVERNOR)) == 0) &&
(strlen(governor) == strlen(ONDEMAND_GOVERNOR))) {
undo_hint_action(DISPLAY_STATE_HINT_ID);
display_hint_sent = 0;
} else if ((strncmp(governor, INTERACTIVE_GOVERNOR, strlen(INTERACTIVE_GOVERNOR)) == 0) &&
(strlen(governor) == strlen(INTERACTIVE_GOVERNOR))) {
undo_hint_action(DISPLAY_STATE_HINT_ID);
display_hint_sent = 0;
} else if ((strncmp(governor, MSMDCVS_GOVERNOR, strlen(MSMDCVS_GOVERNOR)) == 0) &&
(strlen(governor) == strlen(MSMDCVS_GOVERNOR))) {
if (saved_interactive_mode == -1 || saved_interactive_mode == 0) {
/* Display turned on. Restore if possible. */
if (saved_dcvs_cpu0_slack_max != -1) {
snprintf(tmp_str, NODE_MAX, "%d", saved_dcvs_cpu0_slack_max);
if (sysfs_write(DCVS_CPU0_SLACK_MAX_NODE, tmp_str) != 0) {
if (!slack_node_rw_failed) {
ALOGE("Failed to write to %s", DCVS_CPU0_SLACK_MAX_NODE);
}
rc = 1;
}
}
if (saved_dcvs_cpu0_slack_min != -1) {
snprintf(tmp_str, NODE_MAX, "%d", saved_dcvs_cpu0_slack_min);
if (sysfs_write(DCVS_CPU0_SLACK_MIN_NODE, tmp_str) != 0) {
if (!slack_node_rw_failed) {
ALOGE("Failed to write to %s", DCVS_CPU0_SLACK_MIN_NODE);
}
rc = 1;
}
}
if (saved_mpdecision_slack_max != -1) {
snprintf(tmp_str, NODE_MAX, "%d", saved_mpdecision_slack_max);
if (sysfs_write(MPDECISION_SLACK_MAX_NODE, tmp_str) != 0) {
if (!slack_node_rw_failed) {
ALOGE("Failed to write to %s", MPDECISION_SLACK_MAX_NODE);
}
rc = 1;
}
}
if (saved_mpdecision_slack_min != -1) {
snprintf(tmp_str, NODE_MAX, "%d", saved_mpdecision_slack_min);
if (sysfs_write(MPDECISION_SLACK_MIN_NODE, tmp_str) != 0) {
if (!slack_node_rw_failed) {
ALOGE("Failed to write to %s", MPDECISION_SLACK_MIN_NODE);
}
rc = 1;
}
}
}
slack_node_rw_failed = rc;
}
}
saved_interactive_mode = !!on;
}
int Manage_runstop(char *devname, int fd, int runstop, int quiet)
{
/* Run or stop the array. array must already be configured
* required >= 0.90.0
* Only print failure messages if quiet == 0;
* quiet > 0 means really be quiet
* quiet < 0 means we will try again if it fails.
*/
mdu_param_t param; /* unused */
if (runstop == -1 && md_get_version(fd) < 9000) {
if (ioctl(fd, STOP_MD, 0)) {
if (quiet == 0) fprintf(stderr,
Name ": stopping device %s "
"failed: %s\n",
devname, strerror(errno));
return 1;
}
}
if (md_get_version(fd) < 9000) {
fprintf(stderr, Name ": need md driver version 0.90.0 or later\n");
return 1;
}
/*
if (ioctl(fd, GET_ARRAY_INFO, &array)) {
fprintf(stderr, Name ": %s does not appear to be active.\n",
devname);
return 1;
}
*/
if (runstop>0) {
if (ioctl(fd, RUN_ARRAY, ¶m)) {
fprintf(stderr, Name ": failed to run array %s: %s\n",
devname, strerror(errno));
return 1;
}
if (quiet <= 0)
fprintf(stderr, Name ": started %s\n", devname);
} else if (runstop < 0){
struct map_ent *map = NULL;
struct stat stb;
struct mdinfo *mdi;
int devnum;
int err;
int count;
/* If this is an mdmon managed array, just write 'inactive'
* to the array state and let mdmon clear up.
*/
devnum = fd2devnum(fd);
/* Get EXCL access first. If this fails, then attempting
* to stop is probably a bad idea.
*/
close(fd);
fd = open(devname, O_RDONLY|O_EXCL);
if (fd < 0 || fd2devnum(fd) != devnum) {
if (fd >= 0)
close(fd);
fprintf(stderr,
Name ": Cannot get exclusive access to %s:"
"Perhaps a running "
"process, mounted filesystem "
"or active volume group?\n",
devname);
return 1;
}
mdi = sysfs_read(fd, -1, GET_LEVEL|GET_VERSION);
if (mdi &&
mdi->array.level > 0 &&
is_subarray(mdi->text_version)) {
int err;
/* This is mdmon managed. */
close(fd);
count = 25;
while (count &&
(err = sysfs_set_str(mdi, NULL,
"array_state",
"inactive")) < 0
&& errno == EBUSY) {
usleep(200000);
count--;
}
if (err && !quiet) {
fprintf(stderr, Name
": failed to stop array %s: %s\n",
devname, strerror(errno));
return 1;
}
/* Give monitor a chance to act */
ping_monitor(mdi->text_version);
fd = open_dev_excl(devnum);
if (fd < 0) {
fprintf(stderr, Name
": failed to completely stop %s"
": Device is busy\n",
devname);
return 1;
}
} else if (mdi &&
mdi->array.major_version == -1 &&
mdi->array.minor_version == -2 &&
!is_subarray(mdi->text_version)) {
struct mdstat_ent *mds, *m;
/* container, possibly mdmon-managed.
* Make sure mdmon isn't opening it, which
* would interfere with the 'stop'
*/
ping_monitor(mdi->sys_name);
/* now check that there are no existing arrays
* which are members of this array
*/
mds = mdstat_read(0, 0);
for (m=mds; m; m=m->next)
if (m->metadata_version &&
strncmp(m->metadata_version, "external:", 9)==0 &&
is_subarray(m->metadata_version+9) &&
devname2devnum(m->metadata_version+10) == devnum) {
if (!quiet)
fprintf(stderr, Name
": Cannot stop container %s: "
"member %s still active\n",
devname, m->dev);
free_mdstat(mds);
if (mdi)
sysfs_free(mdi);
return 1;
}
}
/* As we have 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.
*/
count = 25; err = 0;
while (count && fd >= 0
&& (err = ioctl(fd, STOP_ARRAY, NULL)) < 0
&& errno == EBUSY) {
usleep(200000);
count --;
}
if (fd >= 0 && err) {
if (quiet == 0) {
fprintf(stderr, Name
": failed to stop array %s: %s\n",
devname, strerror(errno));
if (errno == EBUSY)
fprintf(stderr, "Perhaps a running "
"process, mounted filesystem "
"or active volume group?\n");
}
if (mdi)
sysfs_free(mdi);
return 1;
}
/* prior to 2.6.28, KOBJ_CHANGE was not sent when an md array
* was stopped, so We'll do it here just to be sure. Drop any
* partitions as well...
*/
if (fd >= 0)
ioctl(fd, BLKRRPART, 0);
if (mdi)
sysfs_uevent(mdi, "change");
if (devnum != NoMdDev &&
(stat("/dev/.udev", &stb) != 0 ||
check_env("MDADM_NO_UDEV"))) {
struct map_ent *mp = map_by_devnum(&map, devnum);
remove_devices(devnum, mp ? mp->path : NULL);
}
if (quiet <= 0)
fprintf(stderr, Name ": stopped %s\n", devname);
map_lock(&map);
map_remove(&map, devnum);
map_unlock(&map);
}
return 0;
}
