c/mirror_zfs
1
0
Fork 0
mirror of https://git.proxmox.com/git/mirror_zfs.git synced 2024-12-25 02:49:32 +03:00
Code Issues Packages Projects Releases Wiki Activity

Illumos 5911 - ZFS "hangs" while deleting file

Browse Source 
5911 ZFS "hangs" while deleting file
Reviewed by: Bayard Bell <buffer.g.overflow@gmail.com>
Reviewed by: Alek Pinchuk <alek@nexenta.com>
Reviewed by: Simon Klinkert <simon.klinkert@gmail.com>
Reviewed by: Dan McDonald <danmcd@omniti.com>
Approved by: Richard Lowe <richlowe@richlowe.net>

References:
  https://www.illumos.org/issues/5911
  https://github.com/illumos/illumos-gate/commit/46e1baa

Porting notes:

Resolved ISO C90 forbids mixed declarations and code wanting in
the dnode_free_range() function.

Ported-by: kernelOfTruth kerneloftruth@gmail.com
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes #3554
This commit is contained in:
Matthew Ahrens 2015-07-02 18:23:20 +02:00 committed by Brian Behlendorf
parent 5e8cd5d17f
commit 4bda3bd0e7
5 changed files with 91 additions and 34 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
include/sys/dbuf.h
View File

@ -20,7 +20,7 @@
*/
/*
* Copyright (c) 2005, 2010, Oracle and/or its affiliates. All rights reserved.
* Copyright (c) 2012, 2014 by Delphix. All rights reserved.
* Copyright (c) 2012, 2015 by Delphix. All rights reserved.
* Copyright (c) 2013 by Saso Kiselkov. All rights reserved.
* Copyright (c) 2014 Spectra Logic Corporation, All rights reserved.
*/
@ -287,7 +287,7 @@ void dbuf_clear(dmu_buf_impl_t *db);
void dbuf_evict(dmu_buf_impl_t *db);
void dbuf_unoverride(dbuf_dirty_record_t *dr);
void dbuf_sync_list(list_t *list, dmu_tx_t *tx);
void dbuf_sync_list(list_t *list, int level, dmu_tx_t *tx);
void dbuf_release_bp(dmu_buf_impl_t *db);
void dbuf_free_range(struct dnode *dn, uint64_t start, uint64_t end,

34
module/zfs/dbuf.c
View File

@ -21,7 +21,7 @@
/*
* Copyright (c) 2005, 2010, Oracle and/or its affiliates. All rights reserved.
* Copyright 2011 Nexenta Systems, Inc. All rights reserved.
* Copyright (c) 2012, 2014 by Delphix. All rights reserved.
* Copyright (c) 2012, 2015 by Delphix. All rights reserved.
* Copyright (c) 2013 by Saso Kiselkov. All rights reserved.
* Copyright (c) 2014 Spectra Logic Corporation, All rights reserved.
*/
@ -1455,6 +1455,16 @@ dbuf_undirty(dmu_buf_impl_t *db, dmu_tx_t *tx)
dbuf_dirty_record_t *dr, **drp;
ASSERT(txg != 0);
/*
* Due to our use of dn_nlevels below, this can only be called
* in open context, unless we are operating on the MOS.
* From syncing context, dn_nlevels may be different from the
* dn_nlevels used when dbuf was dirtied.
*/
ASSERT(db->db_objset ==
dmu_objset_pool(db->db_objset)->dp_meta_objset ||
txg != spa_syncing_txg(dmu_objset_spa(db->db_objset)));
ASSERT(db->db_blkid != DMU_BONUS_BLKID);
ASSERT0(db->db_level);
ASSERT(MUTEX_HELD(&db->db_mtx));
@ -1477,11 +1487,8 @@ dbuf_undirty(dmu_buf_impl_t *db, dmu_tx_t *tx)
ASSERT(db->db.db_size != 0);
/*
* Any space we accounted for in dp_dirty_* will be cleaned up by
* dsl_pool_sync(). This is relatively rare so the discrepancy
* is not a big deal.
*/
dsl_pool_undirty_space(dmu_objset_pool(dn->dn_objset),
dr->dr_accounted, txg);
*drp = dr->dr_next;
@ -1496,7 +1503,7 @@ dbuf_undirty(dmu_buf_impl_t *db, dmu_tx_t *tx)
list_remove(&dr->dr_parent->dt.di.dr_children, dr);
mutex_exit(&dr->dr_parent->dt.di.dr_mtx);
} else if (db->db_blkid == DMU_SPILL_BLKID ||
db->db_level+1 == dn->dn_nlevels) {
db->db_level + 1 == dn->dn_nlevels) {
ASSERT(db->db_blkptr == NULL || db->db_parent == dn->dn_dbuf);
mutex_enter(&dn->dn_mtx);
list_remove(&dn->dn_dirty_records[txg & TXG_MASK], dr);
@ -1513,11 +1520,6 @@ dbuf_undirty(dmu_buf_impl_t *db, dmu_tx_t *tx)
VERIFY(arc_buf_remove_ref(dr->dt.dl.dr_data, db));
}
if (db->db_level != 0) {
mutex_destroy(&dr->dt.di.dr_mtx);
list_destroy(&dr->dt.di.dr_children);
}
kmem_free(dr, sizeof (dbuf_dirty_record_t));
ASSERT(db->db_dirtycnt > 0);
@ -2603,7 +2605,7 @@ dbuf_sync_indirect(dbuf_dirty_record_t *dr, dmu_tx_t *tx)
zio = dr->dr_zio;
mutex_enter(&dr->dt.di.dr_mtx);
dbuf_sync_list(&dr->dt.di.dr_children, tx);
dbuf_sync_list(&dr->dt.di.dr_children, db->db_level - 1, tx);
ASSERT(list_head(&dr->dt.di.dr_children) == NULL);
mutex_exit(&dr->dt.di.dr_mtx);
zio_nowait(zio);
@ -2754,7 +2756,7 @@ dbuf_sync_leaf(dbuf_dirty_record_t *dr, dmu_tx_t *tx)
}
void
dbuf_sync_list(list_t *list, dmu_tx_t *tx)
dbuf_sync_list(list_t *list, int level, dmu_tx_t *tx)
{
dbuf_dirty_record_t *dr;
@ -2771,6 +2773,10 @@ dbuf_sync_list(list_t *list, dmu_tx_t *tx)
DMU_META_DNODE_OBJECT);
break;
}
if (dr->dr_dbuf->db_blkid != DMU_BONUS_BLKID &&
dr->dr_dbuf->db_blkid != DMU_SPILL_BLKID) {
VERIFY3U(dr->dr_dbuf->db_level, ==, level);
}
list_remove(list, dr);
if (dr->dr_dbuf->db_level > 0)
dbuf_sync_indirect(dr, tx);

4
module/zfs/dmu_tx.c
View File

@ -21,7 +21,7 @@
/*
* Copyright (c) 2005, 2010, Oracle and/or its affiliates. All rights reserved.
* Copyright 2011 Nexenta Systems, Inc. All rights reserved.
* Copyright (c) 2013 by Delphix. All rights reserved.
* Copyright (c) 2012, 2015 by Delphix. All rights reserved.
*/
#include <sys/dmu.h>
@ -679,7 +679,7 @@ dmu_tx_hold_free(dmu_tx_t *tx, uint64_t object, uint64_t off, uint64_t len)
uint64_t ibyte = i << shift;
err = dnode_next_offset(dn, 0, &ibyte, 2, 1, 0);
i = ibyte >> shift;
if (err == ESRCH)
if (err == ESRCH || i > end)
break;
if (err) {
tx->tx_err = err;

79
module/zfs/dnode.c
View File

@ -20,7 +20,7 @@
*/
/*
* Copyright (c) 2005, 2010, Oracle and/or its affiliates. All rights reserved.
* Copyright (c) 2012, 2014 by Delphix. All rights reserved.
* Copyright (c) 2012, 2015 by Delphix. All rights reserved.
* Copyright (c) 2014 Spectra Logic Corporation, All rights reserved.
*/
@ -1517,6 +1517,16 @@ out:
rw_downgrade(&dn->dn_struct_rwlock);
}
static void
dnode_dirty_l1(dnode_t *dn, uint64_t l1blkid, dmu_tx_t *tx)
{
dmu_buf_impl_t *db = dbuf_hold_level(dn, 1, l1blkid, FTAG);
if (db != NULL) {
dmu_buf_will_dirty(&db->db, tx);
dbuf_rele(db, FTAG);
}
}
void
dnode_free_range(dnode_t *dn, uint64_t off, uint64_t len, dmu_tx_t *tx)
{
@ -1637,27 +1647,68 @@ dnode_free_range(dnode_t *dn, uint64_t off, uint64_t len, dmu_tx_t *tx)
nblks += 1;
/*
* Dirty the first and last indirect blocks, as they (and/or their
* parents) will need to be written out if they were only
* partially freed. Interior indirect blocks will be themselves freed,
* by free_children(), so they need not be dirtied. Note that these
* interior blocks have already been prefetched by dmu_tx_hold_free().
* Dirty all the indirect blocks in this range. Note that only
* the first and last indirect blocks can actually be written
* (if they were partially freed) -- they must be dirtied, even if
* they do not exist on disk yet. The interior blocks will
* be freed by free_children(), so they will not actually be written.
* Even though these interior blocks will not be written, we
* dirty them for two reasons:
*
* - It ensures that the indirect blocks remain in memory until
* syncing context. (They have already been prefetched by
* dmu_tx_hold_free(), so we don't have to worry about reading
* them serially here.)
*
* - The dirty space accounting will put pressure on the txg sync
* mechanism to begin syncing, and to delay transactions if there
* is a large amount of freeing. Even though these indirect
* blocks will not be written, we could need to write the same
* amount of space if we copy the freed BPs into deadlists.
*/
if (dn->dn_nlevels > 1) {
uint64_t first, last;
uint64_t first, last, i, ibyte;
int shift, err;
first = blkid >> epbs;
if ((db = dbuf_hold_level(dn, 1, first, FTAG))) {
dmu_buf_will_dirty(&db->db, tx);
dbuf_rele(db, FTAG);
}
dnode_dirty_l1(dn, first, tx);
if (trunc)
last = dn->dn_maxblkid >> epbs;
else
last = (blkid + nblks - 1) >> epbs;
if (last > first && (db = dbuf_hold_level(dn, 1, last, FTAG))) {
dmu_buf_will_dirty(&db->db, tx);
dbuf_rele(db, FTAG);
if (last != first)
dnode_dirty_l1(dn, last, tx);
shift = dn->dn_datablkshift + dn->dn_indblkshift -
SPA_BLKPTRSHIFT;
for (i = first + 1; i < last; i++) {
/*
* Set i to the blockid of the next non-hole
* level-1 indirect block at or after i. Note
* that dnode_next_offset() operates in terms of
* level-0-equivalent bytes.
*/
ibyte = i << shift;
err = dnode_next_offset(dn, DNODE_FIND_HAVELOCK,
&ibyte, 2, 1, 0);
i = ibyte >> shift;
if (i >= last)
break;
/*
* Normally we should not see an error, either
* from dnode_next_offset() or dbuf_hold_level()
* (except for ESRCH from dnode_next_offset).
* If there is an i/o error, then when we read
* this block in syncing context, it will use
* ZIO_FLAG_MUSTSUCCEED, and thus hang/panic according
* to the "failmode" property. dnode_next_offset()
* doesn't have a flag to indicate MUSTSUCCEED.
*/
if (err != 0)
break;
dnode_dirty_l1(dn, i, tx);
}
}

4
module/zfs/dnode_sync.c
View File

@ -21,7 +21,7 @@
/*
* Copyright (c) 2005, 2010, Oracle and/or its affiliates. All rights reserved.
* Copyright (c) 2012, 2014 by Delphix. All rights reserved.
* Copyright (c) 2012, 2015 by Delphix. All rights reserved.
* Copyright (c) 2014 Spectra Logic Corporation, All rights reserved.
*/
@ -718,7 +718,7 @@ dnode_sync(dnode_t *dn, dmu_tx_t *tx)
mutex_exit(&dn->dn_mtx);
}
dbuf_sync_list(list, tx);
dbuf_sync_list(list, dn->dn_phys->dn_nlevels - 1, tx);
if (!DMU_OBJECT_IS_SPECIAL(dn->dn_object)) {
ASSERT3P(list_head(list), ==, NULL);