/*
* If this dataset has a non-empty intent log, replay it and destroy it.
*/
void
zil_replay(objset_t *os, void *arg, zil_replay_func_t *replay_func[TX_MAX_TYPE])
{
zilog_t *zilog = dmu_objset_zil(os);
const zil_header_t *zh = zilog->zl_header;
zil_replay_arg_t zr;
if ((zh->zh_flags & ZIL_REPLAY_NEEDED) == 0) {
zil_destroy(zilog, B_TRUE);
return;
}
zr.zr_os = os;
zr.zr_replay = replay_func;
zr.zr_arg = arg;
zr.zr_byteswap = BP_SHOULD_BYTESWAP(&zh->zh_log);
/* XXX: Changed to use vmem_alloc instead of kmem_alloc for
* large allocation size (I think this is safe here).
*/
zr.zr_lrbuf = vmem_alloc(2 * SPA_MAXBLOCKSIZE, KM_SLEEP);
/*
* Wait for in-progress removes to sync before starting replay.
*/
txg_wait_synced(zilog->zl_dmu_pool, 0);
zilog->zl_replay = B_TRUE;
zilog->zl_replay_time = lbolt;
ASSERT(zilog->zl_replay_blks == 0);
(void) zil_parse(zilog, zil_incr_blks, zil_replay_log_record, &zr,
zh->zh_claim_txg);
vmem_free(zr.zr_lrbuf, 2 * SPA_MAXBLOCKSIZE);
zil_destroy(zilog, B_FALSE);
txg_wait_synced(zilog->zl_dmu_pool, zilog->zl_destroy_txg);
zilog->zl_replay = B_FALSE;
}
/*
* If this dataset has a non-empty intent log, replay it and destroy it.
*/
void
zil_replay(objset_t *os, void *arg, zil_replay_func_t *replay_func[TX_MAX_TYPE])
{
zilog_t *zilog = dmu_objset_zil(os);
const zil_header_t *zh = zilog->zl_header;
zil_replay_arg_t zr;
if ((zh->zh_flags & ZIL_REPLAY_NEEDED) == 0) {
zil_destroy(zilog, B_TRUE);
return;
}
//printf("ZFS: Replaying ZIL on %s...\n", os->os->os_spa->spa_name);
zr.zr_os = os;
zr.zr_replay = replay_func;
zr.zr_arg = arg;
zr.zr_byteswap = BP_SHOULD_BYTESWAP(&zh->zh_log);
zr.zr_lrbuf = kmem_alloc(2 * SPA_MAXBLOCKSIZE, KM_SLEEP);
/*
* Wait for in-progress removes to sync before starting replay.
*/
txg_wait_synced(zilog->zl_dmu_pool, 0);
zilog->zl_replay = B_TRUE;
zilog->zl_replay_time = LBOLT;
ASSERT(zilog->zl_replay_blks == 0);
(void) zil_parse(zilog, zil_incr_blks, zil_replay_log_record, &zr,
zh->zh_claim_txg);
kmem_free(zr.zr_lrbuf, 2 * SPA_MAXBLOCKSIZE);
zil_destroy(zilog, B_FALSE);
txg_wait_synced(zilog->zl_dmu_pool, zilog->zl_destroy_txg);
zilog->zl_replay = B_FALSE;
//printf("ZFS: Replay of ZIL on %s finished.\n", os->os->os_spa->spa_name);
}
int
zio_checksum_error(zio_t *zio, zio_bad_cksum_t *info)
{
blkptr_t *bp = zio->io_bp;
uint_t checksum = (bp == NULL ? zio->io_prop.zp_checksum :
(BP_IS_GANG(bp) ? ZIO_CHECKSUM_GANG_HEADER : BP_GET_CHECKSUM(bp)));
int byteswap;
int error;
uint64_t size = (bp == NULL ? zio->io_size :
(BP_IS_GANG(bp) ? SPA_GANGBLOCKSIZE : BP_GET_PSIZE(bp)));
uint64_t offset = zio->io_offset;
void *data = zio->io_data;
zio_checksum_info_t *ci = &zio_checksum_table[checksum];
zio_cksum_t actual_cksum, expected_cksum, verifier;
if (checksum >= ZIO_CHECKSUM_FUNCTIONS || ci->ci_func[0] == NULL)
return (EINVAL);
if (ci->ci_eck) {
zio_eck_t *eck;
if (checksum == ZIO_CHECKSUM_ZILOG2) {
zil_chain_t *zilc = data;
uint64_t nused;
eck = &zilc->zc_eck;
if (eck->zec_magic == ZEC_MAGIC)
nused = zilc->zc_nused;
else if (eck->zec_magic == BSWAP_64(ZEC_MAGIC))
nused = BSWAP_64(zilc->zc_nused);
else
return (ECKSUM);
if (nused > size)
return (ECKSUM);
size = P2ROUNDUP_TYPED(nused, ZIL_MIN_BLKSZ, uint64_t);
} else {
eck = (zio_eck_t *)((char *)data + size) - 1;
}
if (checksum == ZIO_CHECKSUM_GANG_HEADER)
zio_checksum_gang_verifier(&verifier, bp);
else if (checksum == ZIO_CHECKSUM_LABEL)
zio_checksum_label_verifier(&verifier, offset);
else
verifier = bp->blk_cksum;
byteswap = (eck->zec_magic == BSWAP_64(ZEC_MAGIC));
if (byteswap)
byteswap_uint64_array(&verifier, sizeof (zio_cksum_t));
expected_cksum = eck->zec_cksum;
eck->zec_cksum = verifier;
ci->ci_func[byteswap](data, size, &actual_cksum);
eck->zec_cksum = expected_cksum;
if (byteswap)
byteswap_uint64_array(&expected_cksum,
sizeof (zio_cksum_t));
} else {
ASSERT(!BP_IS_GANG(bp));
byteswap = BP_SHOULD_BYTESWAP(bp);
expected_cksum = bp->blk_cksum;
ci->ci_func[byteswap](data, size, &actual_cksum);
}
info->zbc_expected = expected_cksum;
info->zbc_actual = actual_cksum;
info->zbc_checksum_name = ci->ci_name;
info->zbc_byteswapped = byteswap;
info->zbc_injected = 0;
info->zbc_has_cksum = 1;
if (!ZIO_CHECKSUM_EQUAL(actual_cksum, expected_cksum)) {
return (ECKSUM);
}
if (zio_injection_enabled && !zio->io_error &&
(error = zio_handle_fault_injection(zio, ECKSUM)) != 0) {
info->zbc_injected = 1;
return (error);
}
return (0);
}
int
zio_checksum_error_impl(spa_t *spa, const blkptr_t *bp,
enum zio_checksum checksum, abd_t *abd, uint64_t size, uint64_t offset,
zio_bad_cksum_t *info)
{
zio_checksum_info_t *ci = &zio_checksum_table[checksum];
zio_cksum_t actual_cksum, expected_cksum;
zio_eck_t eck;
int byteswap;
if (checksum >= ZIO_CHECKSUM_FUNCTIONS || ci->ci_func[0] == NULL)
return (SET_ERROR(EINVAL));
zio_checksum_template_init(checksum, spa);
if (ci->ci_flags & ZCHECKSUM_FLAG_EMBEDDED) {
zio_cksum_t verifier;
size_t eck_offset;
if (checksum == ZIO_CHECKSUM_ZILOG2) {
zil_chain_t zilc;
uint64_t nused;
abd_copy_to_buf(&zilc, abd, sizeof (zil_chain_t));
eck = zilc.zc_eck;
eck_offset = offsetof(zil_chain_t, zc_eck) +
offsetof(zio_eck_t, zec_cksum);
if (eck.zec_magic == ZEC_MAGIC) {
nused = zilc.zc_nused;
} else if (eck.zec_magic == BSWAP_64(ZEC_MAGIC)) {
nused = BSWAP_64(zilc.zc_nused);
} else {
return (SET_ERROR(ECKSUM));
}
if (nused > size) {
return (SET_ERROR(ECKSUM));
}
size = P2ROUNDUP_TYPED(nused, ZIL_MIN_BLKSZ, uint64_t);
} else {
eck_offset = size - sizeof (zio_eck_t);
abd_copy_to_buf_off(&eck, abd, eck_offset,
sizeof (zio_eck_t));
eck_offset += offsetof(zio_eck_t, zec_cksum);
}
if (checksum == ZIO_CHECKSUM_GANG_HEADER)
zio_checksum_gang_verifier(&verifier, bp);
else if (checksum == ZIO_CHECKSUM_LABEL)
zio_checksum_label_verifier(&verifier, offset);
else
verifier = bp->blk_cksum;
byteswap = (eck.zec_magic == BSWAP_64(ZEC_MAGIC));
if (byteswap)
byteswap_uint64_array(&verifier, sizeof (zio_cksum_t));
expected_cksum = eck.zec_cksum;
abd_copy_from_buf_off(abd, &verifier, eck_offset,
sizeof (zio_cksum_t));
ci->ci_func[byteswap](abd, size,
spa->spa_cksum_tmpls[checksum], &actual_cksum);
abd_copy_from_buf_off(abd, &expected_cksum, eck_offset,
sizeof (zio_cksum_t));
if (byteswap) {
byteswap_uint64_array(&expected_cksum,
sizeof (zio_cksum_t));
}
} else {
byteswap = BP_SHOULD_BYTESWAP(bp);
expected_cksum = bp->blk_cksum;
ci->ci_func[byteswap](abd, size,
spa->spa_cksum_tmpls[checksum], &actual_cksum);
}
/*
* MAC checksums are a special case since half of this checksum will
* actually be the encryption MAC. This will be verified by the
* decryption process, so we just check the truncated checksum now.
* Objset blocks use embedded MACs so we don't truncate the checksum
* for them.
*/
if (bp != NULL && BP_USES_CRYPT(bp) &&
BP_GET_TYPE(bp) != DMU_OT_OBJSET) {
if (!(ci->ci_flags & ZCHECKSUM_FLAG_DEDUP)) {
actual_cksum.zc_word[0] ^= actual_cksum.zc_word[2];
actual_cksum.zc_word[1] ^= actual_cksum.zc_word[3];
}
actual_cksum.zc_word[2] = 0;
actual_cksum.zc_word[3] = 0;
expected_cksum.zc_word[2] = 0;
expected_cksum.zc_word[3] = 0;
}
if (info != NULL) {
info->zbc_expected = expected_cksum;
info->zbc_actual = actual_cksum;
info->zbc_checksum_name = ci->ci_name;
info->zbc_byteswapped = byteswap;
info->zbc_injected = 0;
info->zbc_has_cksum = 1;
}
if (!ZIO_CHECKSUM_EQUAL(actual_cksum, expected_cksum))
return (SET_ERROR(ECKSUM));
return (0);
}
int
zio_checksum_error_impl(spa_t *spa, blkptr_t *bp, enum zio_checksum checksum,
abd_t *abd, uint64_t size, uint64_t offset, zio_bad_cksum_t *info)
{
zio_checksum_info_t *ci = &zio_checksum_table[checksum];
int byteswap;
zio_cksum_t actual_cksum, expected_cksum;
if (checksum >= ZIO_CHECKSUM_FUNCTIONS || ci->ci_func[0] == NULL)
return (SET_ERROR(EINVAL));
zio_checksum_template_init(checksum, spa);
if (ci->ci_flags & ZCHECKSUM_FLAG_EMBEDDED) {
zio_eck_t *eck;
zio_cksum_t verifier;
size_t eck_offset;
uint64_t data_size = size;
void *data = abd_borrow_buf_copy(abd, data_size);
if (checksum == ZIO_CHECKSUM_ZILOG2) {
zil_chain_t *zilc = data;
uint64_t nused;
eck = &zilc->zc_eck;
if (eck->zec_magic == ZEC_MAGIC) {
nused = zilc->zc_nused;
} else if (eck->zec_magic == BSWAP_64(ZEC_MAGIC)) {
nused = BSWAP_64(zilc->zc_nused);
} else {
abd_return_buf(abd, data, data_size);
return (SET_ERROR(ECKSUM));
}
if (nused > data_size) {
abd_return_buf(abd, data, data_size);
return (SET_ERROR(ECKSUM));
}
size = P2ROUNDUP_TYPED(nused, ZIL_MIN_BLKSZ, uint64_t);
} else {
eck = (zio_eck_t *)((char *)data + data_size) - 1;
}
if (checksum == ZIO_CHECKSUM_GANG_HEADER)
zio_checksum_gang_verifier(&verifier, bp);
else if (checksum == ZIO_CHECKSUM_LABEL)
zio_checksum_label_verifier(&verifier, offset);
else
verifier = bp->blk_cksum;
byteswap = (eck->zec_magic == BSWAP_64(ZEC_MAGIC));
if (byteswap)
byteswap_uint64_array(&verifier, sizeof (zio_cksum_t));
eck_offset = (size_t)(&eck->zec_cksum) - (size_t)data;
expected_cksum = eck->zec_cksum;
eck->zec_cksum = verifier;
abd_return_buf_copy(abd, data, data_size);
ci->ci_func[byteswap](abd, size,
spa->spa_cksum_tmpls[checksum], &actual_cksum);
abd_copy_from_buf_off(abd, &expected_cksum,
eck_offset, sizeof (zio_cksum_t));
if (byteswap) {
byteswap_uint64_array(&expected_cksum,
sizeof (zio_cksum_t));
}
} else {
byteswap = BP_SHOULD_BYTESWAP(bp);
expected_cksum = bp->blk_cksum;
ci->ci_func[byteswap](abd, size,
spa->spa_cksum_tmpls[checksum], &actual_cksum);
}
if (info != NULL) {
info->zbc_expected = expected_cksum;
info->zbc_actual = actual_cksum;
info->zbc_checksum_name = ci->ci_name;
info->zbc_byteswapped = byteswap;
info->zbc_injected = 0;
info->zbc_has_cksum = 1;
}
if (!ZIO_CHECKSUM_EQUAL(actual_cksum, expected_cksum))
return (SET_ERROR(ECKSUM));
return (0);
}
int
zio_checksum_error(zio_t *zio, zio_bad_cksum_t *info)
{
blkptr_t *bp = zio->io_bp;
uint_t checksum = (bp == NULL ? zio->io_prop.zp_checksum :
(BP_IS_GANG(bp) ? ZIO_CHECKSUM_GANG_HEADER : BP_GET_CHECKSUM(bp)));
int byteswap;
int error;
uint64_t size = (bp == NULL ? zio->io_size :
(BP_IS_GANG(bp) ? SPA_GANGBLOCKSIZE : BP_GET_PSIZE(bp)));
uint64_t offset = zio->io_offset;
void *data = zio->io_data;
zio_checksum_info_t *ci = &zio_checksum_table[checksum];
zio_cksum_t actual_cksum, expected_cksum, verifier;
if (checksum >= ZIO_CHECKSUM_FUNCTIONS || ci->ci_func[0] == NULL)
return (SET_ERROR(EINVAL));
if (ci->ci_eck) {
zio_eck_t *eck;
if (checksum == ZIO_CHECKSUM_ZILOG2) {
zil_chain_t *zilc = data;
uint64_t nused;
eck = &zilc->zc_eck;
if (eck->zec_magic == ZEC_MAGIC)
nused = zilc->zc_nused;
else if (eck->zec_magic == BSWAP_64(ZEC_MAGIC))
nused = BSWAP_64(zilc->zc_nused);
else
return (SET_ERROR(ECKSUM));
if (nused > size)
return (SET_ERROR(ECKSUM));
size = P2ROUNDUP_TYPED(nused, ZIL_MIN_BLKSZ, uint64_t);
} else {
eck = (zio_eck_t *)((char *)data + size) - 1;
}
if (checksum == ZIO_CHECKSUM_GANG_HEADER)
zio_checksum_gang_verifier(&verifier, bp);
else if (checksum == ZIO_CHECKSUM_LABEL)
zio_checksum_label_verifier(&verifier, offset);
else
verifier = bp->blk_cksum;
byteswap = (eck->zec_magic == BSWAP_64(ZEC_MAGIC));
if (byteswap)
byteswap_uint64_array(&verifier, sizeof (zio_cksum_t));
expected_cksum = eck->zec_cksum;
eck->zec_cksum = verifier;
ci->ci_func[byteswap](data, size, &actual_cksum);
eck->zec_cksum = expected_cksum;
if (byteswap)
byteswap_uint64_array(&expected_cksum,
sizeof (zio_cksum_t));
} else {
ASSERT(!BP_IS_GANG(bp));
byteswap = BP_SHOULD_BYTESWAP(bp);
expected_cksum = bp->blk_cksum;
ci->ci_func[byteswap](data, size, &actual_cksum);
}
info->zbc_expected = expected_cksum;
info->zbc_actual = actual_cksum;
info->zbc_checksum_name = ci->ci_name;
info->zbc_byteswapped = byteswap;
info->zbc_injected = 0;
info->zbc_has_cksum = 1;
/*
* Special case for truncated checksums with crypto MAC
* This may not be the best place to deal with this but it is here now.
*
* Words 0 and 1 and 32 bits of word 2 of the checksum are the
* first 160 bytes of SHA256 hash.
* The rest of words 2 and all of word 3 are the crypto MAC so
* ignore those because we can't check them until we do the decryption
* later, nor could we do them if the key wasn't present
*/
if (ci->ci_trunc) {
if (!(0 == (
(actual_cksum.zc_word[0] - expected_cksum.zc_word[0]) |
(actual_cksum.zc_word[1] - expected_cksum.zc_word[1]) |
(BF64_GET(actual_cksum.zc_word[2], 0, 32) -
BF64_GET(expected_cksum.zc_word[2], 0, 32))))) {
return (ECKSUM);
}
} else if (!ZIO_CHECKSUM_EQUAL(actual_cksum, expected_cksum)) {
return (SET_ERROR(ECKSUM));
}
if (zio_injection_enabled && !zio->io_error &&
(error = zio_handle_fault_injection(zio, ECKSUM)) != 0) {
info->zbc_injected = 1;
return (error);
}
return (0);
}
/*
* Create an on-disk intent log.
*/
static void
zil_create(zilog_t *zilog)
{
const zil_header_t *zh = zilog->zl_header;
lwb_t *lwb;
uint64_t txg = 0;
dmu_tx_t *tx = NULL;
blkptr_t blk;
int error = 0;
/*
* Wait for any previous destroy to complete.
*/
txg_wait_synced(zilog->zl_dmu_pool, zilog->zl_destroy_txg);
ASSERT(zh->zh_claim_txg == 0);
ASSERT(zh->zh_replay_seq == 0);
blk = zh->zh_log;
/*
* If we don't already have an initial log block or we have one
* but it's the wrong endianness then allocate one.
*/
if (BP_IS_HOLE(&blk) || BP_SHOULD_BYTESWAP(&blk)) {
tx = dmu_tx_create(zilog->zl_os);
(void) dmu_tx_assign(tx, TXG_WAIT);
dsl_dataset_dirty(dmu_objset_ds(zilog->zl_os), tx);
txg = dmu_tx_get_txg(tx);
if (!BP_IS_HOLE(&blk)) {
zio_free_blk(zilog->zl_spa, &blk, txg);
BP_ZERO(&blk);
}
error = zio_alloc_blk(zilog->zl_spa, ZIL_MIN_BLKSZ, &blk,
NULL, txg);
if (error == 0)
zil_init_log_chain(zilog, &blk);
}
/*
* Allocate a log write buffer (lwb) for the first log block.
*/
if (error == 0) {
lwb = kmem_cache_alloc(zil_lwb_cache, KM_SLEEP);
lwb->lwb_zilog = zilog;
lwb->lwb_blk = blk;
lwb->lwb_nused = 0;
lwb->lwb_sz = BP_GET_LSIZE(&lwb->lwb_blk);
lwb->lwb_buf = zio_buf_alloc(lwb->lwb_sz);
lwb->lwb_max_txg = txg;
lwb->lwb_zio = NULL;
mutex_enter(&zilog->zl_lock);
list_insert_tail(&zilog->zl_lwb_list, lwb);
mutex_exit(&zilog->zl_lock);
}
/*
* If we just allocated the first log block, commit our transaction
* and wait for zil_sync() to stuff the block poiner into zh_log.
* (zh is part of the MOS, so we cannot modify it in open context.)
*/
if (tx != NULL) {
dmu_tx_commit(tx);
txg_wait_synced(zilog->zl_dmu_pool, txg);
}
ASSERT(bcmp(&blk, &zh->zh_log, sizeof (blk)) == 0);
}