#include <sys/arc.h>
#include <sys/ddt.h>
#include <sys/zfeature.h>
+#include <zfs_comutil.h>
#undef ZFS_MAXNAMELEN
#include <libzfs.h>
uint64_t *zopt_object = NULL;
int zopt_objects = 0;
libzfs_handle_t *g_zfs;
+uint64_t max_inflight = 200;
/*
* These libumem hooks provide a reasonable set of defaults for the allocator's
{
(void) fprintf(stderr,
"Usage: %s [-CumdibcsDvhLXFPA] [-t txg] [-e [-p path...]] "
- "poolname [object...]\n"
- " %s [-divPA] [-e -p path...] dataset [object...]\n"
- " %s -m [-LXFPA] [-t txg] [-e [-p path...]] "
+ "[-U config] [-M inflight I/Os] poolname [object...]\n"
+ " %s [-divPA] [-e -p path...] [-U config] dataset "
+ "[object...]\n"
+ " %s -m [-LXFPA] [-t txg] [-e [-p path...]] [-U config] "
"poolname [vdev [metaslab...]]\n"
" %s -R [-A] [-e [-p path...]] poolname "
"vdev:offset:size[:flags]\n"
- " %s -S [-PA] [-e [-p path...]] poolname\n"
+ " %s -S [-PA] [-e [-p path...]] [-U config] poolname\n"
" %s -l [-uA] device\n"
" %s -C [-A] [-U config]\n\n",
cmdname, cmdname, cmdname, cmdname, cmdname, cmdname, cmdname);
(void) fprintf(stderr, " -P print numbers in parseable form\n");
(void) fprintf(stderr, " -t <txg> -- highest txg to use when "
"searching for uberblocks\n");
+ (void) fprintf(stderr, " -M <number of inflight I/Os> -- "
+ "specify the maximum number of checksumming I/Os "
+ "[default is 200]\n");
(void) fprintf(stderr, "Specify an option more than once (e.g. -bb) "
"to make only that option verbose\n");
(void) fprintf(stderr, "Default is to dump everything non-verbosely\n");
nvlist_free(nv);
}
+/* ARGSUSED */
+static void
+dump_history_offsets(objset_t *os, uint64_t object, void *data, size_t size)
+{
+ spa_history_phys_t *shp = data;
+
+ if (shp == NULL)
+ return;
+
+ (void) printf("\t\tpool_create_len = %llu\n",
+ (u_longlong_t)shp->sh_pool_create_len);
+ (void) printf("\t\tphys_max_off = %llu\n",
+ (u_longlong_t)shp->sh_phys_max_off);
+ (void) printf("\t\tbof = %llu\n",
+ (u_longlong_t)shp->sh_bof);
+ (void) printf("\t\teof = %llu\n",
+ (u_longlong_t)shp->sh_eof);
+ (void) printf("\t\trecords_lost = %llu\n",
+ (u_longlong_t)shp->sh_records_lost);
+}
+
static void
zdb_nicenum(uint64_t num, char *buf)
{
dump_metaslab_stats(metaslab_t *msp)
{
char maxbuf[32];
- space_map_t *sm = &msp->ms_map;
+ space_map_t *sm = msp->ms_map;
avl_tree_t *t = sm->sm_pp_root;
int free_pct = sm->sm_space * 100 / sm->sm_size;
{
vdev_t *vd = msp->ms_group->mg_vd;
spa_t *spa = vd->vdev_spa;
- space_map_t *sm = &msp->ms_map;
+ space_map_t *sm = msp->ms_map;
space_map_obj_t *smo = &msp->ms_smo;
char freebuf[32];
for (i = 0; i < num; i++) {
uint64_t time, txg, ievent;
char *cmd, *intstr;
+ boolean_t printed = B_FALSE;
if (nvlist_lookup_uint64(events[i], ZPOOL_HIST_TIME,
&time) != 0)
- continue;
+ goto next;
if (nvlist_lookup_string(events[i], ZPOOL_HIST_CMD,
&cmd) != 0) {
if (nvlist_lookup_uint64(events[i],
ZPOOL_HIST_INT_EVENT, &ievent) != 0)
- continue;
+ goto next;
verify(nvlist_lookup_uint64(events[i],
ZPOOL_HIST_TXG, &txg) == 0);
verify(nvlist_lookup_string(events[i],
ZPOOL_HIST_INT_STR, &intstr) == 0);
- if (ievent >= LOG_END)
- continue;
+ if (ievent >= ZFS_NUM_LEGACY_HISTORY_EVENTS)
+ goto next;
(void) snprintf(internalstr,
sizeof (internalstr),
(void) localtime_r(&tsec, &t);
(void) strftime(tbuf, sizeof (tbuf), "%F.%T", &t);
(void) printf("%s %s\n", tbuf, cmd);
+ printed = B_TRUE;
+
+next:
+ if (dump_opt['h'] > 1) {
+ if (!printed)
+ (void) printf("unrecognized record:\n");
+ dump_nvlist(events[i], 2);
+ }
}
}
arc_buf_t *buf;
uint64_t fill = 0;
- err = arc_read_nolock(NULL, spa, bp, arc_getbuf_func, &buf,
+ err = arc_read(NULL, spa, bp, arc_getbuf_func, &buf,
ZIO_PRIORITY_ASYNC_READ, ZIO_FLAG_CANFAIL, &flags, zb);
if (err)
return (err);
}
static void
-dump_bpobj(bpobj_t *bpo, char *name)
+dump_bpobj(bpobj_t *bpo, char *name, int indent)
{
char bytes[32];
char comp[32];
char uncomp[32];
+ uint64_t i;
if (dump_opt['d'] < 3)
return;
zdb_nicenum(bpo->bpo_phys->bpo_bytes, bytes);
- if (bpo->bpo_havesubobj) {
+ if (bpo->bpo_havesubobj && bpo->bpo_phys->bpo_subobjs != 0) {
zdb_nicenum(bpo->bpo_phys->bpo_comp, comp);
zdb_nicenum(bpo->bpo_phys->bpo_uncomp, uncomp);
- (void) printf("\n %s: %llu local blkptrs, %llu subobjs, "
- "%s (%s/%s comp)\n",
- name, (u_longlong_t)bpo->bpo_phys->bpo_num_blkptrs,
+ (void) printf(" %*s: object %llu, %llu local blkptrs, "
+ "%llu subobjs, %s (%s/%s comp)\n",
+ indent * 8, name,
+ (u_longlong_t)bpo->bpo_object,
+ (u_longlong_t)bpo->bpo_phys->bpo_num_blkptrs,
(u_longlong_t)bpo->bpo_phys->bpo_num_subobjs,
bytes, comp, uncomp);
+
+ for (i = 0; i < bpo->bpo_phys->bpo_num_subobjs; i++) {
+ uint64_t subobj;
+ bpobj_t subbpo;
+ int error;
+ VERIFY0(dmu_read(bpo->bpo_os,
+ bpo->bpo_phys->bpo_subobjs,
+ i * sizeof (subobj), sizeof (subobj), &subobj, 0));
+ error = bpobj_open(&subbpo, bpo->bpo_os, subobj);
+ if (error != 0) {
+ (void) printf("ERROR %u while trying to open "
+ "subobj id %llu\n",
+ error, (u_longlong_t)subobj);
+ continue;
+ }
+ dump_bpobj(&subbpo, "subobj", indent + 1);
+ }
} else {
- (void) printf("\n %s: %llu blkptrs, %s\n",
- name, (u_longlong_t)bpo->bpo_phys->bpo_num_blkptrs, bytes);
+ (void) printf(" %*s: object %llu, %llu blkptrs, %s\n",
+ indent * 8, name,
+ (u_longlong_t)bpo->bpo_object,
+ (u_longlong_t)bpo->bpo_phys->bpo_num_blkptrs,
+ bytes);
}
if (dump_opt['d'] < 5)
return;
- (void) printf("\n");
- (void) bpobj_iterate_nofree(bpo, dump_bpobj_cb, NULL, NULL);
+ if (indent == 0) {
+ (void) bpobj_iterate_nofree(bpo, dump_bpobj_cb, NULL, NULL);
+ (void) printf("\n");
+ }
}
static void
dump_deadlist(dsl_deadlist_t *dl)
{
dsl_deadlist_entry_t *dle;
+ uint64_t unused;
char bytes[32];
char comp[32];
char uncomp[32];
(void) printf("\n");
+ /* force the tree to be loaded */
+ dsl_deadlist_space_range(dl, 0, UINT64_MAX, &unused, &unused, &unused);
+
for (dle = avl_first(&dl->dl_tree); dle;
dle = AVL_NEXT(&dl->dl_tree, dle)) {
- (void) printf(" mintxg %llu -> obj %llu\n",
- (longlong_t)dle->dle_mintxg,
- (longlong_t)dle->dle_bpobj.bpo_object);
+ if (dump_opt['d'] >= 5) {
+ char buf[128];
+ (void) snprintf(buf, sizeof (buf),
+ "mintxg %llu -> obj %llu",
+ (longlong_t)dle->dle_mintxg,
+ (longlong_t)dle->dle_bpobj.bpo_object);
+
+ dump_bpobj(&dle->dle_bpobj, buf, 0);
+ } else {
+ (void) printf("mintxg %llu -> obj %llu\n",
+ (longlong_t)dle->dle_mintxg,
+ (longlong_t)dle->dle_bpobj.bpo_object);
- if (dump_opt['d'] >= 5)
- dump_bpobj(&dle->dle_bpobj, "");
+ }
}
}
* print uid or gid information.
* For normal POSIX id just the id is printed in decimal format.
* For CIFS files with FUID the fuid is printed in hex followed by
- * the doman-rid string.
+ * the domain-rid string.
*/
static void
print_idstr(uint64_t id, const char *id_type)
print_idstr(gid, "gid");
}
+static void
+dump_znode_sa_xattr(sa_handle_t *hdl)
+{
+ nvlist_t *sa_xattr;
+ nvpair_t *elem = NULL;
+ int sa_xattr_size = 0;
+ int sa_xattr_entries = 0;
+ int error;
+ char *sa_xattr_packed;
+
+ error = sa_size(hdl, sa_attr_table[ZPL_DXATTR], &sa_xattr_size);
+ if (error || sa_xattr_size == 0)
+ return;
+
+ sa_xattr_packed = malloc(sa_xattr_size);
+ if (sa_xattr_packed == NULL)
+ return;
+
+ error = sa_lookup(hdl, sa_attr_table[ZPL_DXATTR],
+ sa_xattr_packed, sa_xattr_size);
+ if (error) {
+ free(sa_xattr_packed);
+ return;
+ }
+
+ error = nvlist_unpack(sa_xattr_packed, sa_xattr_size, &sa_xattr, 0);
+ if (error) {
+ free(sa_xattr_packed);
+ return;
+ }
+
+ while ((elem = nvlist_next_nvpair(sa_xattr, elem)) != NULL)
+ sa_xattr_entries++;
+
+ (void) printf("\tSA xattrs: %d bytes, %d entries\n\n",
+ sa_xattr_size, sa_xattr_entries);
+ while ((elem = nvlist_next_nvpair(sa_xattr, elem)) != NULL) {
+ uchar_t *value;
+ uint_t cnt, idx;
+
+ (void) printf("\t\t%s = ", nvpair_name(elem));
+ nvpair_value_byte_array(elem, &value, &cnt);
+ for (idx = 0; idx < cnt; ++idx) {
+ if (isprint(value[idx]))
+ (void) putchar(value[idx]);
+ else
+ (void) printf("\\%3.3o", value[idx]);
+ }
+ (void) putchar('\n');
+ }
+
+ nvlist_free(sa_xattr);
+ free(sa_xattr_packed);
+}
+
/*ARGSUSED*/
static void
dump_znode(objset_t *os, uint64_t object, void *data, size_t size)
if (sa_lookup(hdl, sa_attr_table[ZPL_RDEV], &rdev,
sizeof (uint64_t)) == 0)
(void) printf("\trdev 0x%016llx\n", (u_longlong_t)rdev);
+ dump_znode_sa_xattr(hdl);
sa_handle_destroy(hdl);
}
dump_zap, /* other ZAP */
dump_zap, /* persistent error log */
dump_uint8, /* SPA history */
- dump_uint64, /* SPA history offsets */
+ dump_history_offsets, /* SPA history offsets */
dump_zap, /* Pool properties */
dump_zap, /* DSL permissions */
dump_acl, /* ZFS ACL */
int print_header = 1;
int i, error;
+ dsl_pool_config_enter(dmu_objset_pool(os), FTAG);
dmu_objset_fast_stat(os, &dds);
+ dsl_pool_config_exit(dmu_objset_pool(os), FTAG);
if (dds.dds_type < DMU_OST_NUMTYPES)
type = objset_types[dds.dds_type];
bp, NULL, NULL, ZIO_FLAG_CANFAIL)), ==, 0);
}
-/* ARGSUSED */
+static void
+zdb_blkptr_done(zio_t *zio)
+{
+ spa_t *spa = zio->io_spa;
+ blkptr_t *bp = zio->io_bp;
+ int ioerr = zio->io_error;
+ zdb_cb_t *zcb = zio->io_private;
+ zbookmark_t *zb = &zio->io_bookmark;
+
+ zio_data_buf_free(zio->io_data, zio->io_size);
+
+ mutex_enter(&spa->spa_scrub_lock);
+ spa->spa_scrub_inflight--;
+ cv_broadcast(&spa->spa_scrub_io_cv);
+
+ if (ioerr && !(zio->io_flags & ZIO_FLAG_SPECULATIVE)) {
+ char blkbuf[BP_SPRINTF_LEN];
+
+ zcb->zcb_haderrors = 1;
+ zcb->zcb_errors[ioerr]++;
+
+ if (dump_opt['b'] >= 2)
+ sprintf_blkptr(blkbuf, bp);
+ else
+ blkbuf[0] = '\0';
+
+ (void) printf("zdb_blkptr_cb: "
+ "Got error %d reading "
+ "<%llu, %llu, %lld, %llx> %s -- skipping\n",
+ ioerr,
+ (u_longlong_t)zb->zb_objset,
+ (u_longlong_t)zb->zb_object,
+ (u_longlong_t)zb->zb_level,
+ (u_longlong_t)zb->zb_blkid,
+ blkbuf);
+ }
+ mutex_exit(&spa->spa_scrub_lock);
+}
+
static int
-zdb_blkptr_cb(spa_t *spa, zilog_t *zilog, const blkptr_t *bp, arc_buf_t *pbuf,
+zdb_blkptr_cb(spa_t *spa, zilog_t *zilog, const blkptr_t *bp,
const zbookmark_t *zb, const dnode_phys_t *dnp, void *arg)
{
zdb_cb_t *zcb = arg;
is_metadata = (BP_GET_LEVEL(bp) != 0 || DMU_OT_IS_METADATA(type));
if (dump_opt['c'] > 1 || (dump_opt['c'] && is_metadata)) {
- int ioerr;
size_t size = BP_GET_PSIZE(bp);
- void *data = malloc(size);
+ void *data = zio_data_buf_alloc(size);
int flags = ZIO_FLAG_CANFAIL | ZIO_FLAG_SCRUB | ZIO_FLAG_RAW;
/* If it's an intent log block, failure is expected. */
if (zb->zb_level == ZB_ZIL_LEVEL)
flags |= ZIO_FLAG_SPECULATIVE;
- ioerr = zio_wait(zio_read(NULL, spa, bp, data, size,
- NULL, NULL, ZIO_PRIORITY_ASYNC_READ, flags, zb));
-
- free(data);
-
- if (ioerr && !(flags & ZIO_FLAG_SPECULATIVE)) {
- zcb->zcb_haderrors = 1;
- zcb->zcb_errors[ioerr]++;
+ mutex_enter(&spa->spa_scrub_lock);
+ while (spa->spa_scrub_inflight > max_inflight)
+ cv_wait(&spa->spa_scrub_io_cv, &spa->spa_scrub_lock);
+ spa->spa_scrub_inflight++;
+ mutex_exit(&spa->spa_scrub_lock);
- if (dump_opt['b'] >= 2)
- sprintf_blkptr(blkbuf, bp);
- else
- blkbuf[0] = '\0';
-
- (void) printf("zdb_blkptr_cb: "
- "Got error %d reading "
- "<%llu, %llu, %lld, %llx> %s -- skipping\n",
- ioerr,
- (u_longlong_t)zb->zb_objset,
- (u_longlong_t)zb->zb_object,
- (u_longlong_t)zb->zb_level,
- (u_longlong_t)zb->zb_blkid,
- blkbuf);
- }
+ zio_nowait(zio_read(NULL, spa, bp, data, size,
+ zdb_blkptr_done, zcb, ZIO_PRIORITY_ASYNC_READ, flags, zb));
}
zcb->zcb_readfails = 0;
for (m = 0; m < vd->vdev_ms_count; m++) {
metaslab_t *msp = vd->vdev_ms[m];
mutex_enter(&msp->ms_lock);
- space_map_unload(&msp->ms_map);
- VERIFY(space_map_load(&msp->ms_map,
+ space_map_unload(msp->ms_map);
+ VERIFY(space_map_load(msp->ms_map,
&zdb_space_map_ops, SM_ALLOC, &msp->ms_smo,
spa->spa_meta_objset) == 0);
- msp->ms_map.sm_ppd = vd;
+ msp->ms_map->sm_ppd = vd;
mutex_exit(&msp->ms_lock);
}
}
for (m = 0; m < vd->vdev_ms_count; m++) {
metaslab_t *msp = vd->vdev_ms[m];
mutex_enter(&msp->ms_lock);
- space_map_unload(&msp->ms_map);
+ space_map_unload(msp->ms_map);
mutex_exit(&msp->ms_lock);
}
}
* it's not part of any space map) is a double allocation,
* reference to a freed block, or an unclaimed log block.
*/
- bzero(&zcb, sizeof(zdb_cb_t));
+ bzero(&zcb, sizeof (zdb_cb_t));
zdb_leak_init(spa, &zcb);
/*
*/
(void) bpobj_iterate_nofree(&spa->spa_deferred_bpobj,
count_block_cb, &zcb, NULL);
- (void) bpobj_iterate_nofree(&spa->spa_dsl_pool->dp_free_bpobj,
- count_block_cb, &zcb, NULL);
+ if (spa_version(spa) >= SPA_VERSION_DEADLISTS) {
+ (void) bpobj_iterate_nofree(&spa->spa_dsl_pool->dp_free_bpobj,
+ count_block_cb, &zcb, NULL);
+ }
if (spa_feature_is_active(spa,
&spa_feature_table[SPA_FEATURE_ASYNC_DESTROY])) {
VERIFY3U(0, ==, bptree_iterate(spa->spa_meta_objset,
zcb.zcb_haderrors |= traverse_pool(spa, 0, flags, zdb_blkptr_cb, &zcb);
+ /*
+ * If we've traversed the data blocks then we need to wait for those
+ * I/Os to complete. We leverage "The Godfather" zio to wait on
+ * all async I/Os to complete.
+ */
+ if (dump_opt['c']) {
+ (void) zio_wait(spa->spa_async_zio_root);
+ spa->spa_async_zio_root = zio_root(spa, NULL, NULL,
+ ZIO_FLAG_CANFAIL | ZIO_FLAG_SPECULATIVE |
+ ZIO_FLAG_GODFATHER);
+ }
+
if (zcb.zcb_haderrors) {
(void) printf("\nError counts:\n\n");
(void) printf("\t%5s %s\n", "errno", "count");
/* ARGSUSED */
static int
zdb_ddt_add_cb(spa_t *spa, zilog_t *zilog, const blkptr_t *bp,
- arc_buf_t *pbuf, const zbookmark_t *zb, const dnode_phys_t *dnp, void *arg)
+ const zbookmark_t *zb, const dnode_phys_t *dnp, void *arg)
{
avl_tree_t *t = arg;
avl_index_t where;
if (dump_opt['d'] || dump_opt['i']) {
dump_dir(dp->dp_meta_objset);
if (dump_opt['d'] >= 3) {
- dump_bpobj(&spa->spa_deferred_bpobj, "Deferred frees");
+ dump_bpobj(&spa->spa_deferred_bpobj,
+ "Deferred frees", 0);
if (spa_version(spa) >= SPA_VERSION_DEADLISTS) {
dump_bpobj(&spa->spa_dsl_pool->dp_free_bpobj,
- "Pool snapshot frees");
+ "Pool snapshot frees", 0);
}
if (spa_feature_is_active(spa,
psize = size;
lsize = size;
- pbuf = umem_alloc(SPA_MAXBLOCKSIZE, UMEM_NOFAIL);
+ pbuf = umem_alloc_aligned(SPA_MAXBLOCKSIZE, 512, UMEM_NOFAIL);
lbuf = umem_alloc(SPA_MAXBLOCKSIZE, UMEM_NOFAIL);
BP_ZERO(bp);
nvlist_t *policy = NULL;
uint64_t max_txg = UINT64_MAX;
int rewind = ZPOOL_NEVER_REWIND;
+ char *spa_config_path_env;
(void) setrlimit(RLIMIT_NOFILE, &rl);
(void) enable_extended_FILE_stdio(-1, -1);
dprintf_setup(&argc, argv);
- while ((c = getopt(argc, argv, "bcdhilmsuCDRSAFLXevp:t:U:P")) != -1) {
+ /*
+ * If there is an environment variable SPA_CONFIG_PATH it overrides
+ * default spa_config_path setting. If -U flag is specified it will
+ * override this environment variable settings once again.
+ */
+ spa_config_path_env = getenv("SPA_CONFIG_PATH");
+ if (spa_config_path_env != NULL)
+ spa_config_path = spa_config_path_env;
+
+ while ((c = getopt(argc, argv, "bcdhilmM:suCDRSAFLXevp:t:U:P")) != -1) {
switch (c) {
case 'b':
case 'c':
case 'v':
verbose++;
break;
+ case 'M':
+ max_inflight = strtoull(optarg, NULL, 0);
+ if (max_inflight == 0) {
+ (void) fprintf(stderr, "maximum number "
+ "of inflight I/Os must be greater "
+ "than 0\n");
+ usage();
+ }
+ break;
case 'p':
if (searchdirs == NULL) {
searchdirs = umem_alloc(sizeof (char *),
projects / mirror_zfs.git / blobdiff