/*
* Sigh. Inside a local zone, we don't have access to /etc/zfs/zpool.cache,
* and we don't want to allow the local zone to see all the pools anyway.
* So we have to invent the ZFS_IOC_CONFIG ioctl to grab the configuration
* information for all pool visible within the zone.
*/
nvlist_t *
spa_all_configs(uint64_t *generation)
{
nvlist_t *pools;
spa_t *spa;
if (*generation == spa_config_generation)
return (NULL);
VERIFY(nvlist_alloc(&pools, NV_UNIQUE_NAME, KM_SLEEP) == 0);
spa = NULL;
mutex_enter(&spa_namespace_lock);
while ((spa = spa_next(spa)) != NULL) {
if (INGLOBALZONE(curproc) ||
zone_dataset_visible(spa_name(spa), NULL)) {
mutex_enter(&spa->spa_config_cache_lock);
VERIFY(nvlist_add_nvlist(pools, spa_name(spa),
spa->spa_config) == 0);
mutex_exit(&spa->spa_config_cache_lock);
}
}
mutex_exit(&spa_namespace_lock);
*generation = spa_config_generation;
return (pools);
}
/*
* Sigh. Inside a local zone, we don't have access to /etc/zfs/zpool.cache,
* and we don't want to allow the local zone to see all the pools anyway.
* So we have to invent the ZFS_IOC_CONFIG ioctl to grab the configuration
* information for all pool visible within the zone.
*/
nvlist_t *
spa_all_configs(uint64_t *generation)
{
nvlist_t *pools;
spa_t *spa = NULL;
if (*generation == spa_config_generation)
return (NULL);
pools = fnvlist_alloc();
mutex_enter(&spa_namespace_lock);
while ((spa = spa_next(spa)) != NULL) {
if (INGLOBALZONE(curproc) ||
zone_dataset_visible(spa_name(spa), NULL)) {
mutex_enter(&spa->spa_props_lock);
fnvlist_add_nvlist(pools, spa_name(spa),
spa->spa_config);
mutex_exit(&spa->spa_props_lock);
}
}
*generation = spa_config_generation;
mutex_exit(&spa_namespace_lock);
return (pools);
}
/*
* Create minors for specified pool, if pool is NULL create minors
* for all available pools.
*/
int
zvol_create_minors(const char *pool)
{
spa_t *spa = NULL;
int error = 0;
if (zvol_inhibit_dev)
return (0);
mutex_enter(&zvol_state_lock);
if (pool) {
error = dmu_objset_find_spa(NULL, pool, zvol_create_minors_cb,
NULL, DS_FIND_CHILDREN | DS_FIND_SNAPSHOTS);
} else {
mutex_enter(&spa_namespace_lock);
while ((spa = spa_next(spa)) != NULL) {
error = dmu_objset_find_spa(NULL,
spa_name(spa), zvol_create_minors_cb, NULL,
DS_FIND_CHILDREN | DS_FIND_SNAPSHOTS);
if (error)
break;
}
mutex_exit(&spa_namespace_lock);
}
mutex_exit(&zvol_state_lock);
return error;
}
/*
* Callback for device termination events, ie, disks removed.
*/
bool IOkit_disk_removed_callback(void* target,
void* refCon,
IOService* newService,
IONotifier* notifier)
{
OSObject *prop = 0;
OSString* bsdnameosstr = 0;
prop = newService->getProperty(kIOBSDNameKey, gIOServicePlane,
kIORegistryIterateRecursively);
if (prop) {
spa_t *spa = NULL;
bsdnameosstr = OSDynamicCast(OSString, prop);
printf("ZFS: Device removal detected: '%s'\n",
bsdnameosstr->getCStringNoCopy());
for (spa = spa_next(NULL);
spa != NULL; spa = spa_next(spa)) {
vdev_t *vd;
dprintf("ZFS: Scanning pool '%s'\n", spa_name(spa));
vd = vdev_lookup_by_path(spa->spa_root_vdev,
bsdnameosstr->getCStringNoCopy());
if (vd && vd->vdev_path) {
printf("ZFS: Device '%s' removal requested\n",
vd->vdev_path);
(void) thread_create(NULL, 0, vdev_close_thread,
vd, 0, &p0,
TS_RUN, minclsyspri);
vd->vdev_remove_wanted = B_TRUE;
spa_async_request(spa, SPA_ASYNC_REMOVE);
break;
}
} // for all spa
} // if has BSDname
return true;
}
void
mmp_signal_all_threads(void)
{
spa_t *spa = NULL;
mutex_enter(&spa_namespace_lock);
while ((spa = spa_next(spa))) {
if (spa->spa_state == POOL_STATE_ACTIVE)
mmp_signal_thread(spa);
}
mutex_exit(&spa_namespace_lock);
}
/*
* Synchronize all pools to disk. This must be called with the namespace lock
* held.
*/
void
spa_config_sync(void)
{
spa_t *spa = NULL;
nvlist_t *config;
size_t buflen;
char *buf;
vnode_t *vp;
int oflags = FWRITE | FTRUNC | FCREAT | FOFFMAX;
char pathname[128];
char pathname2[128];
ASSERT(MUTEX_HELD(&spa_namespace_lock));
VERIFY(nvlist_alloc(&config, NV_UNIQUE_NAME, KM_SLEEP) == 0);
/*
* Add all known pools to the configuration list, ignoring those with
* alternate root paths.
*/
spa = NULL;
while ((spa = spa_next(spa)) != NULL) {
mutex_enter(&spa->spa_config_cache_lock);
if (spa->spa_config && spa->spa_name && spa->spa_root == NULL)
VERIFY(nvlist_add_nvlist(config, spa->spa_name,
spa->spa_config) == 0);
mutex_exit(&spa->spa_config_cache_lock);
}
/*
* Pack the configuration into a buffer.
*/
VERIFY(nvlist_size(config, &buflen, NV_ENCODE_XDR) == 0);
buf = kmem_alloc(buflen, KM_SLEEP);
VERIFY(nvlist_pack(config, &buf, &buflen, NV_ENCODE_XDR,
KM_SLEEP) == 0);
/*
* Write the configuration to disk. We need to do the traditional
* 'write to temporary file, sync, move over original' to make sure we
* always have a consistent view of the data.
*/
(void) snprintf(pathname, sizeof (pathname), "%s/%s", spa_config_dir,
ZPOOL_CACHE_TMP);
if (vn_open(pathname, UIO_SYSSPACE, oflags, 0644, &vp, CRCREAT, 0) != 0)
goto out;
if (vn_rdwr(UIO_WRITE, vp, buf, buflen, 0, UIO_SYSSPACE,
0, RLIM64_INFINITY, kcred, NULL) == 0 &&
VOP_FSYNC(vp, FSYNC, kcred) == 0) {
(void) snprintf(pathname2, sizeof (pathname2), "%s/%s",
spa_config_dir, ZPOOL_CACHE_FILE);
(void) vn_rename(pathname, pathname2, UIO_SYSSPACE);
}
(void) VOP_CLOSE(vp, oflags, 1, 0, kcred);
VN_RELE(vp);
out:
(void) vn_remove(pathname, UIO_SYSSPACE, RMFILE);
spa_config_generation++;
kmem_free(buf, buflen);
nvlist_free(config);
}
/*
* Synchronize pool configuration to disk. This must be called with the
* namespace lock held. Synchronizing the pool cache is typically done after
* the configuration has been synced to the MOS. This exposes a window where
* the MOS config will have been updated but the cache file has not. If
* the system were to crash at that instant then the cached config may not
* contain the correct information to open the pool and an explicity import
* would be required.
*/
void
spa_config_sync(spa_t *target, boolean_t removing, boolean_t postsysevent)
{
spa_config_dirent_t *dp, *tdp;
nvlist_t *nvl;
boolean_t ccw_failure;
int error;
ASSERT(MUTEX_HELD(&spa_namespace_lock));
if (rootdir == NULL || !(spa_mode_global & FWRITE))
return;
/*
* Iterate over all cachefiles for the pool, past or present. When the
* cachefile is changed, the new one is pushed onto this list, allowing
* us to update previous cachefiles that no longer contain this pool.
*/
ccw_failure = B_FALSE;
for (dp = list_head(&target->spa_config_list); dp != NULL;
dp = list_next(&target->spa_config_list, dp)) {
spa_t *spa = NULL;
if (dp->scd_path == NULL)
continue;
/*
* Iterate over all pools, adding any matching pools to 'nvl'.
*/
nvl = NULL;
while ((spa = spa_next(spa)) != NULL) {
nvlist_t *nvroot = NULL;
/*
* Skip over our own pool if we're about to remove
* ourselves from the spa namespace or any pool that
* is readonly. Since we cannot guarantee that a
* readonly pool would successfully import upon reboot,
* we don't allow them to be written to the cache file.
*/
if ((spa == target && removing) ||
(spa_state(spa) == POOL_STATE_ACTIVE &&
!spa_writeable(spa)))
continue;
mutex_enter(&spa->spa_props_lock);
tdp = list_head(&spa->spa_config_list);
if (spa->spa_config == NULL ||
tdp->scd_path == NULL ||
strcmp(tdp->scd_path, dp->scd_path) != 0) {
mutex_exit(&spa->spa_props_lock);
continue;
}
if (nvl == NULL)
VERIFY(nvlist_alloc(&nvl, NV_UNIQUE_NAME,
KM_SLEEP) == 0);
VERIFY(nvlist_add_nvlist(nvl, spa->spa_name,
spa->spa_config) == 0);
mutex_exit(&spa->spa_props_lock);
if (nvlist_lookup_nvlist(nvl, spa->spa_name, &nvroot) == 0)
spa_config_clean(nvroot);
}
error = spa_config_write(dp, nvl);
if (error != 0)
ccw_failure = B_TRUE;
nvlist_free(nvl);
}
if (ccw_failure) {
/*
* Keep trying so that configuration data is
* written if/when any temporary filesystem
* resource issues are resolved.
*/
if (target->spa_ccw_fail_time == 0) {
zfs_ereport_post(FM_EREPORT_ZFS_CONFIG_CACHE_WRITE,
target, NULL, NULL, 0, 0);
}
target->spa_ccw_fail_time = gethrtime();
spa_async_request(target, SPA_ASYNC_CONFIG_UPDATE);
} else {
/*
* Do not rate limit future attempts to update
* the config cache.
*/
target->spa_ccw_fail_time = 0;
}
/*
* Remove any config entries older than the current one.
*/
dp = list_head(&target->spa_config_list);
while ((tdp = list_next(&target->spa_config_list, dp)) != NULL) {
list_remove(&target->spa_config_list, tdp);
if (tdp->scd_path != NULL)
spa_strfree(tdp->scd_path);
kmem_free(tdp, sizeof (spa_config_dirent_t));
}
spa_config_generation++;
if (postsysevent)
spa_event_notify(target, NULL, ESC_ZFS_CONFIG_SYNC);
}
/*
* Synchronize pool configuration to disk. This must be called with the
* namespace lock held. Synchronizing the pool cache is typically done after
* the configuration has been synced to the MOS. This exposes a window where
* the MOS config will have been updated but the cache file has not. If
* the system were to crash at that instant then the cached config may not
* contain the correct information to open the pool and an explicity import
* would be required.
*/
void
spa_config_sync(spa_t *target, boolean_t removing, boolean_t postsysevent)
{
spa_config_dirent_t *dp, *tdp;
nvlist_t *nvl;
char *pool_name;
ASSERT(MUTEX_HELD(&spa_namespace_lock));
if (rootdir == NULL || !(spa_mode_global & FWRITE))
return;
/*
* Iterate over all cachefiles for the pool, past or present. When the
* cachefile is changed, the new one is pushed onto this list, allowing
* us to update previous cachefiles that no longer contain this pool.
*/
for (dp = list_head(&target->spa_config_list); dp != NULL;
dp = list_next(&target->spa_config_list, dp)) {
spa_t *spa = NULL;
if (dp->scd_path == NULL)
continue;
/*
* Iterate over all pools, adding any matching pools to 'nvl'.
*/
nvl = NULL;
while ((spa = spa_next(spa)) != NULL) {
/*
* Skip over our own pool if we're about to remove
* ourselves from the spa namespace or any pool that
* is readonly. Since we cannot guarantee that a
* readonly pool would successfully import upon reboot,
* we don't allow them to be written to the cache file.
*/
if ((spa == target && removing) ||
!spa_writeable(spa))
continue;
mutex_enter(&spa->spa_props_lock);
tdp = list_head(&spa->spa_config_list);
if (spa->spa_config == NULL ||
tdp->scd_path == NULL ||
strcmp(tdp->scd_path, dp->scd_path) != 0) {
mutex_exit(&spa->spa_props_lock);
continue;
}
if (nvl == NULL)
VERIFY(nvlist_alloc(&nvl, NV_UNIQUE_NAME,
KM_SLEEP) == 0);
if (spa->spa_import_flags & ZFS_IMPORT_TEMP_NAME) {
VERIFY0(nvlist_lookup_string(spa->spa_config,
ZPOOL_CONFIG_POOL_NAME, &pool_name));
} else
pool_name = spa_name(spa);
VERIFY(nvlist_add_nvlist(nvl, pool_name,
spa->spa_config) == 0);
mutex_exit(&spa->spa_props_lock);
}
spa_config_write(dp, nvl);
nvlist_free(nvl);
}
/*
* Remove any config entries older than the current one.
*/
dp = list_head(&target->spa_config_list);
while ((tdp = list_next(&target->spa_config_list, dp)) != NULL) {
list_remove(&target->spa_config_list, tdp);
if (tdp->scd_path != NULL)
spa_strfree(tdp->scd_path);
kmem_free(tdp, sizeof (spa_config_dirent_t));
}
spa_config_generation++;
if (postsysevent)
spa_event_notify(target, NULL, FM_EREPORT_ZFS_CONFIG_SYNC);
}
/*
* Synchronize pool configuration to disk. This must be called with the
* namespace lock held.
*/
void
spa_config_sync(spa_t *target, boolean_t removing, boolean_t postsysevent)
{
spa_config_dirent_t *dp, *tdp;
nvlist_t *nvl;
ASSERT(MUTEX_HELD(&spa_namespace_lock));
if (rootdir == NULL || !(spa_mode_global & FWRITE))
return;
/*
* Iterate over all cachefiles for the pool, past or present. When the
* cachefile is changed, the new one is pushed onto this list, allowing
* us to update previous cachefiles that no longer contain this pool.
*/
for (dp = list_head(&target->spa_config_list); dp != NULL;
dp = list_next(&target->spa_config_list, dp)) {
spa_t *spa = NULL;
if (dp->scd_path == NULL)
continue;
/*
* Iterate over all pools, adding any matching pools to 'nvl'.
*/
nvl = NULL;
while ((spa = spa_next(spa)) != NULL) {
if (spa == target && removing)
continue;
mutex_enter(&spa->spa_props_lock);
tdp = list_head(&spa->spa_config_list);
if (spa->spa_config == NULL ||
tdp->scd_path == NULL ||
strcmp(tdp->scd_path, dp->scd_path) != 0) {
mutex_exit(&spa->spa_props_lock);
continue;
}
if (nvl == NULL)
VERIFY(nvlist_alloc(&nvl, NV_UNIQUE_NAME,
KM_SLEEP) == 0);
VERIFY(nvlist_add_nvlist(nvl, spa->spa_name,
spa->spa_config) == 0);
mutex_exit(&spa->spa_props_lock);
}
spa_config_write(dp, nvl);
nvlist_free(nvl);
}
/*
* Remove any config entries older than the current one.
*/
dp = list_head(&target->spa_config_list);
while ((tdp = list_next(&target->spa_config_list, dp)) != NULL) {
list_remove(&target->spa_config_list, tdp);
if (tdp->scd_path != NULL)
spa_strfree(tdp->scd_path);
kmem_free(tdp, sizeof (spa_config_dirent_t));
}
spa_config_generation++;
if (postsysevent)
spa_event_notify(target, NULL, FM_EREPORT_ZFS_CONFIG_SYNC);
}
/*
* Synchronize pool configuration to disk. This must be called with the
* namespace lock held.
*/
void
spa_config_sync(spa_t *target, boolean_t removing, boolean_t postsysevent)
{
spa_t *spa = NULL;
spa_config_dirent_t *dp, *tdp;
nvlist_t *nvl;
ASSERT(MUTEX_HELD(&spa_namespace_lock));
/*
* Iterate over all cachefiles for the pool, past or present. When the
* cachefile is changed, the new one is pushed onto this list, allowing
* us to update previous cachefiles that no longer contain this pool.
*/
for (dp = list_head(&target->spa_config_list); dp != NULL;
dp = list_next(&target->spa_config_list, dp)) {
spa = NULL;
if (dp->scd_path == NULL)
continue;
/*
* Iterate over all pools, adding any matching pools to 'nvl'.
*/
nvl = NULL;
while ((spa = spa_next(spa)) != NULL) {
if (spa->spa_config == NULL || spa->spa_name == NULL)
continue;
if (spa == target && removing)
continue;
#ifdef __APPLE__
/* OS X - Omit disk based pools */
if (vdev_contains_disks(spa->spa_root_vdev))
continue;
#endif
tdp = list_head(&spa->spa_config_list);
ASSERT(tdp != NULL);
if (tdp->scd_path == NULL ||
strcmp(tdp->scd_path, dp->scd_path) != 0)
continue;
if (nvl == NULL)
VERIFY(nvlist_alloc(&nvl, NV_UNIQUE_NAME,
KM_SLEEP) == 0);
VERIFY(nvlist_add_nvlist(nvl, spa->spa_name,
spa->spa_config) == 0);
}
spa_config_write(dp, nvl);
nvlist_free(nvl);
}
/*
* Remove any config entries older than the current one.
*/
dp = list_head(&target->spa_config_list);
while ((tdp = list_next(&target->spa_config_list, dp)) != NULL) {
list_remove(&target->spa_config_list, tdp);
if (tdp->scd_path != NULL)
spa_strfree(tdp->scd_path);
kmem_free(tdp, sizeof (spa_config_dirent_t));
}
spa_config_generation++;
if (postsysevent)
spa_event_notify(target, NULL, ESC_ZFS_CONFIG_SYNC);
}