* You may not use this file except in compliance with the License.
*
* You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
- * or http://www.opensolaris.org/os/licensing.
+ * or https://opensource.org/licenses/CDDL-1.0.
* See the License for the specific language governing permissions
* and limitations under the License.
*
/*
* Copyright (c) 2009, 2010, Oracle and/or its affiliates. All rights reserved.
* Copyright (c) 2012, 2016 by Delphix. All rights reserved.
+ * Copyright (c) 2022 by Pawel Jakub Dawidek
*/
#include <sys/zfs_context.h>
#include <sys/arc.h>
#include <sys/dsl_pool.h>
#include <sys/zio_checksum.h>
-#include <sys/zio_compress.h>
#include <sys/dsl_scan.h>
#include <sys/abd.h>
*/
int zfs_dedup_prefetch = 0;
-static const ddt_ops_t *ddt_ops[DDT_TYPES] = {
+static const ddt_ops_t *const ddt_ops[DDT_TYPES] = {
&ddt_zap_ops,
};
-static const char *ddt_class_name[DDT_CLASSES] = {
+static const char *const ddt_class_name[DDT_CLASSES] = {
"ditto",
"duplicate",
"unique",
ddt_object_name(ddt, type, class, name);
- ASSERT(*objectp == 0);
- VERIFY(ddt_ops[type]->ddt_op_create(os, objectp, tx, prehash) == 0);
- ASSERT(*objectp != 0);
+ ASSERT3U(*objectp, ==, 0);
+ VERIFY0(ddt_ops[type]->ddt_op_create(os, objectp, tx, prehash));
+ ASSERT3U(*objectp, !=, 0);
- VERIFY(zap_add(os, DMU_POOL_DIRECTORY_OBJECT, name,
- sizeof (uint64_t), 1, objectp, tx) == 0);
+ VERIFY0(zap_add(os, DMU_POOL_DIRECTORY_OBJECT, name,
+ sizeof (uint64_t), 1, objectp, tx));
- VERIFY(zap_add(os, spa->spa_ddt_stat_object, name,
+ VERIFY0(zap_add(os, spa->spa_ddt_stat_object, name,
sizeof (uint64_t), sizeof (ddt_histogram_t) / sizeof (uint64_t),
- &ddt->ddt_histogram[type][class], tx) == 0);
+ &ddt->ddt_histogram[type][class], tx));
}
static void
ddt_object_name(ddt, type, class, name);
- ASSERT(*objectp != 0);
+ ASSERT3U(*objectp, !=, 0);
ASSERT(ddt_histogram_empty(&ddt->ddt_histogram[type][class]));
- VERIFY(ddt_object_count(ddt, type, class, &count) == 0 && count == 0);
- VERIFY(zap_remove(os, DMU_POOL_DIRECTORY_OBJECT, name, tx) == 0);
- VERIFY(zap_remove(os, spa->spa_ddt_stat_object, name, tx) == 0);
- VERIFY(ddt_ops[type]->ddt_op_destroy(os, *objectp, tx) == 0);
- bzero(&ddt->ddt_object_stats[type][class], sizeof (ddt_object_t));
+ VERIFY0(ddt_object_count(ddt, type, class, &count));
+ VERIFY0(count);
+ VERIFY0(zap_remove(os, DMU_POOL_DIRECTORY_OBJECT, name, tx));
+ VERIFY0(zap_remove(os, spa->spa_ddt_stat_object, name, tx));
+ VERIFY0(ddt_ops[type]->ddt_op_destroy(os, *objectp, tx));
+ memset(&ddt->ddt_object_stats[type][class], 0, sizeof (ddt_object_t));
*objectp = 0;
}
ddt_object_name(ddt, type, class, name);
- VERIFY(zap_update(ddt->ddt_os, ddt->ddt_spa->spa_ddt_stat_object, name,
+ VERIFY0(zap_update(ddt->ddt_os, ddt->ddt_spa->spa_ddt_stat_object, name,
sizeof (uint64_t), sizeof (ddt_histogram_t) / sizeof (uint64_t),
- &ddt->ddt_histogram[type][class], tx) == 0);
+ &ddt->ddt_histogram[type][class], tx));
/*
* Cache DDT statistics; this is the only time they'll change.
*/
- VERIFY(ddt_object_info(ddt, type, class, &doi) == 0);
- VERIFY(ddt_object_count(ddt, type, class, &count) == 0);
+ VERIFY0(ddt_object_info(ddt, type, class, &doi));
+ VERIFY0(ddt_object_count(ddt, type, class, &count));
ddo->ddo_count = count;
ddo->ddo_dspace = doi.doi_physical_blocks_512 << 9;
void
ddt_bp_fill(const ddt_phys_t *ddp, blkptr_t *bp, uint64_t txg)
{
- ASSERT(txg != 0);
+ ASSERT3U(txg, !=, 0);
for (int d = 0; d < SPA_DVAS_PER_BP; d++)
bp->blk_dva[d] = ddp->ddp_dva[d];
void
ddt_phys_fill(ddt_phys_t *ddp, const blkptr_t *bp)
{
- ASSERT(ddp->ddp_phys_birth == 0);
+ ASSERT0(ddp->ddp_phys_birth);
for (int d = 0; d < SPA_DVAS_PER_BP; d++)
ddp->ddp_dva[d] = bp->blk_dva[d];
void
ddt_phys_clear(ddt_phys_t *ddp)
{
- bzero(ddp, sizeof (*ddp));
+ memset(ddp, 0, sizeof (*ddp));
}
void
ddt_phys_decref(ddt_phys_t *ddp)
{
if (ddp) {
- ASSERT(ddp->ddp_refcnt > 0);
+ ASSERT3U(ddp->ddp_refcnt, >, 0);
ddp->ddp_refcnt--;
}
}
uint64_t lsize = DDK_GET_LSIZE(ddk);
uint64_t psize = DDK_GET_PSIZE(ddk);
- bzero(dds, sizeof (*dds));
+ memset(dds, 0, sizeof (*dds));
for (int p = 0; p < DDT_PHYS_TYPES; p++, ddp++) {
uint64_t dsize = 0;
ddt_stat_generate(ddt, dde, &dds);
bucket = highbit64(dds.dds_ref_blocks) - 1;
- ASSERT(bucket >= 0);
+ ASSERT3U(bucket, >=, 0);
ddh = &ddt->ddt_histogram[dde->dde_type][dde->dde_class];
void
ddt_histogram_stat(ddt_stat_t *dds, const ddt_histogram_t *ddh)
{
- bzero(dds, sizeof (*dds));
+ memset(dds, 0, sizeof (*dds));
for (int h = 0; h < 64; h++)
ddt_stat_add(dds, &ddh->ddh_stat[h], 0);
if (spa->spa_dedup_dspace != ~0ULL)
return (spa->spa_dedup_dspace);
- bzero(&dds_total, sizeof (ddt_stat_t));
+ memset(&dds_total, 0, sizeof (ddt_stat_t));
/* Calculate and cache the stats */
ddt_get_dedup_stats(spa, &dds_total);
return (dds_total.dds_ref_dsize * 100 / dds_total.dds_dsize);
}
-size_t
-ddt_compress(void *src, uchar_t *dst, size_t s_len, size_t d_len)
-{
- uchar_t *version = dst++;
- int cpfunc = ZIO_COMPRESS_ZLE;
- zio_compress_info_t *ci = &zio_compress_table[cpfunc];
- size_t c_len;
-
- ASSERT(d_len >= s_len + 1); /* no compression plus version byte */
-
- c_len = ci->ci_compress(src, dst, s_len, d_len - 1, ci->ci_level);
-
- if (c_len == s_len) {
- cpfunc = ZIO_COMPRESS_OFF;
- bcopy(src, dst, s_len);
- }
-
- *version = cpfunc;
- if (ZFS_HOST_BYTEORDER)
- *version |= DDT_COMPRESS_BYTEORDER_MASK;
-
- return (c_len + 1);
-}
-
-void
-ddt_decompress(uchar_t *src, void *dst, size_t s_len, size_t d_len)
-{
- uchar_t version = *src++;
- int cpfunc = version & DDT_COMPRESS_FUNCTION_MASK;
- zio_compress_info_t *ci = &zio_compress_table[cpfunc];
-
- if (ci->ci_decompress != NULL)
- (void) ci->ci_decompress(src, dst, s_len, d_len, ci->ci_level);
- else
- bcopy(src, dst, d_len);
-
- if (((version & DDT_COMPRESS_BYTEORDER_MASK) != 0) !=
- (ZFS_HOST_BYTEORDER != 0))
- byteswap_uint64_array(dst, d_len);
-}
-
ddt_t *
ddt_select(spa_t *spa, const blkptr_t *bp)
{
ddt_entry_t *dde;
dde = kmem_cache_alloc(ddt_entry_cache, KM_SLEEP);
- bzero(dde, sizeof (ddt_entry_t));
+ memset(dde, 0, sizeof (ddt_entry_t));
cv_init(&dde->dde_cv, NULL, CV_DEFAULT, NULL);
dde->dde_key = *ddk;
ASSERT(!dde->dde_loading);
for (int p = 0; p < DDT_PHYS_TYPES; p++)
- ASSERT(dde->dde_lead_zio[p] == NULL);
+ ASSERT3P(dde->dde_lead_zio[p], ==, NULL);
if (dde->dde_repair_abd != NULL)
abd_free(dde->dde_repair_abd);
ddt_enter(ddt);
- ASSERT(dde->dde_loaded == B_FALSE);
- ASSERT(dde->dde_loading == B_TRUE);
+ ASSERT(!dde->dde_loaded);
+ ASSERT(dde->dde_loading);
dde->dde_type = type; /* will be DDT_TYPES if no entry found */
dde->dde_class = class; /* will be DDT_CLASSES if no entry found */
ddt_t *ddt;
ddt = kmem_cache_alloc(ddt_cache, KM_SLEEP);
- bzero(ddt, sizeof (ddt_t));
+ memset(ddt, 0, sizeof (ddt_t));
mutex_init(&ddt->ddt_lock, NULL, MUTEX_DEFAULT, NULL);
avl_create(&ddt->ddt_tree, ddt_entry_compare,
static void
ddt_table_free(ddt_t *ddt)
{
- ASSERT(avl_numnodes(&ddt->ddt_tree) == 0);
- ASSERT(avl_numnodes(&ddt->ddt_repair_tree) == 0);
+ ASSERT0(avl_numnodes(&ddt->ddt_tree));
+ ASSERT0(avl_numnodes(&ddt->ddt_repair_tree));
avl_destroy(&ddt->ddt_tree);
avl_destroy(&ddt->ddt_repair_tree);
mutex_destroy(&ddt->ddt_lock);
/*
* Seed the cached histograms.
*/
- bcopy(ddt->ddt_histogram, &ddt->ddt_histogram_cache,
+ memcpy(&ddt->ddt_histogram_cache, ddt->ddt_histogram,
sizeof (ddt->ddt_histogram));
spa->spa_dedup_dspace = ~0ULL;
}
}
}
- bzero(dde->dde_phys, sizeof (dde->dde_phys));
+ memset(dde->dde_phys, 0, sizeof (dde->dde_phys));
return (dde);
}
for (int p = 0; p < DDT_PHYS_TYPES; p++, ddp++, rddp++) {
if (ddp->ddp_phys_birth == 0 ||
ddp->ddp_phys_birth != rddp->ddp_phys_birth ||
- bcmp(ddp->ddp_dva, rddp->ddp_dva, sizeof (ddp->ddp_dva)))
+ memcmp(ddp->ddp_dva, rddp->ddp_dva, sizeof (ddp->ddp_dva)))
continue;
ddt_bp_create(ddt->ddt_checksum, ddk, ddp, &blk);
zio_nowait(zio_rewrite(zio, zio->io_spa, 0, &blk,
ASSERT(!dde->dde_loading);
for (int p = 0; p < DDT_PHYS_TYPES; p++, ddp++) {
- ASSERT(dde->dde_lead_zio[p] == NULL);
+ ASSERT3P(dde->dde_lead_zio[p], ==, NULL);
if (ddp->ddp_phys_birth == 0) {
- ASSERT(ddp->ddp_refcnt == 0);
+ ASSERT0(ddp->ddp_refcnt);
continue;
}
if (p == DDT_PHYS_DITTO) {
if (otype != DDT_TYPES &&
(otype != ntype || oclass != nclass || total_refcnt == 0)) {
- VERIFY(ddt_object_remove(ddt, otype, oclass, dde, tx) == 0);
- ASSERT(ddt_object_lookup(ddt, otype, oclass, dde) == ENOENT);
+ VERIFY0(ddt_object_remove(ddt, otype, oclass, dde, tx));
+ ASSERT3U(
+ ddt_object_lookup(ddt, otype, oclass, dde), ==, ENOENT);
}
if (total_refcnt != 0) {
ddt_stat_update(ddt, dde, 0);
if (!ddt_object_exists(ddt, ntype, nclass))
ddt_object_create(ddt, ntype, nclass, tx);
- VERIFY(ddt_object_update(ddt, ntype, nclass, dde, tx) == 0);
+ VERIFY0(ddt_object_update(ddt, ntype, nclass, dde, tx));
/*
* If the class changes, the order that we scan this bp
if (avl_numnodes(&ddt->ddt_tree) == 0)
return;
- ASSERT(spa->spa_uberblock.ub_version >= SPA_VERSION_DEDUP);
+ ASSERT3U(spa->spa_uberblock.ub_version, >=, SPA_VERSION_DEDUP);
if (spa->spa_ddt_stat_object == 0) {
spa->spa_ddt_stat_object = zap_create_link(ddt->ddt_os,
for (enum ddt_class class = 0; class < DDT_CLASSES; class++) {
if (ddt_object_exists(ddt, type, class)) {
ddt_object_sync(ddt, type, class, tx);
- VERIFY(ddt_object_count(ddt, type, class,
- &add) == 0);
+ VERIFY0(ddt_object_count(ddt, type, class,
+ &add));
count += add;
}
}
}
}
- bcopy(ddt->ddt_histogram, &ddt->ddt_histogram_cache,
+ memcpy(&ddt->ddt_histogram_cache, ddt->ddt_histogram,
sizeof (ddt->ddt_histogram));
spa->spa_dedup_dspace = ~0ULL;
}
dmu_tx_t *tx;
zio_t *rio;
- ASSERT(spa_syncing_txg(spa) == txg);
+ ASSERT3U(spa_syncing_txg(spa), ==, txg);
tx = dmu_tx_create_assigned(spa->spa_dsl_pool, txg);
return (SET_ERROR(ENOENT));
}
+/*
+ * This function is used by Block Cloning (brt.c) to increase reference
+ * counter for the DDT entry if the block is already in DDT.
+ *
+ * Return false if the block, despite having the D bit set, is not present
+ * in the DDT. Currently this is not possible but might be in the future.
+ * See the comment below.
+ */
+boolean_t
+ddt_addref(spa_t *spa, const blkptr_t *bp)
+{
+ ddt_t *ddt;
+ ddt_entry_t *dde;
+ boolean_t result;
+
+ spa_config_enter(spa, SCL_ZIO, FTAG, RW_READER);
+ ddt = ddt_select(spa, bp);
+ ddt_enter(ddt);
+
+ dde = ddt_lookup(ddt, bp, B_TRUE);
+ ASSERT3P(dde, !=, NULL);
+
+ if (dde->dde_type < DDT_TYPES) {
+ ddt_phys_t *ddp;
+
+ ASSERT3S(dde->dde_class, <, DDT_CLASSES);
+
+ ddp = &dde->dde_phys[BP_GET_NDVAS(bp)];
+
+ /*
+ * This entry already existed (dde_type is real), so it must
+ * have refcnt >0 at the start of this txg. We are called from
+ * brt_pending_apply(), before frees are issued, so the refcnt
+ * can't be lowered yet. Therefore, it must be >0. We assert
+ * this because if the order of BRT and DDT interactions were
+ * ever to change and the refcnt was ever zero here, then
+ * likely further action is required to fill out the DDT entry,
+ * and this is a place that is likely to be missed in testing.
+ */
+ ASSERT3U(ddp->ddp_refcnt, >, 0);
+
+ ddt_phys_addref(ddp);
+ result = B_TRUE;
+ } else {
+ /*
+ * At the time of implementating this if the block has the
+ * DEDUP flag set it must exist in the DEDUP table, but
+ * there are many advocates that want ability to remove
+ * entries from DDT with refcnt=1. If this will happen,
+ * we may have a block with the DEDUP set, but which doesn't
+ * have a corresponding entry in the DDT. Be ready.
+ */
+ ASSERT3S(dde->dde_class, ==, DDT_CLASSES);
+ ddt_remove(ddt, dde);
+ result = B_FALSE;
+ }
+
+ ddt_exit(ddt);
+ spa_config_exit(spa, SCL_ZIO, FTAG);
+
+ return (result);
+}
+
ZFS_MODULE_PARAM(zfs_dedup, zfs_dedup_, prefetch, INT, ZMOD_RW,
"Enable prefetching dedup-ed blks");
projects / mirror_zfs.git / blobdiff