From fd313b306ec1637859430780e25f21a21cc9bdc8 Mon Sep 17 00:00:00 2001 From: =?utf8?q?Fabian=20Gr=C3=BCnbichler?= Date: Mon, 30 Apr 2018 13:46:29 +0200 Subject: [PATCH] no-change update to 0.7.8 to reduce version number confusion --- spl-patches/0014-Tag-spl-0.7.8.patch | 63 ++ spl-patches/series | 1 + ...-zap_add-failures-in-mixed-case-mode.patch | 553 ++++++++++++++++++ ...se-a-fixed-tag-for-spa_config-locks.patch} | 0 ...ib-initramfs-add-missing-conf.d-zfs.patch} | 0 ...-receive-fails-with-dataset-is-busy.patch} | 0 ...remove_child-panic-due-to-already-d.patch} | 0 ...-Clarify-zinject-8-explanation-of-e.patch} | 0 ...onfig-issues-frame-size-and-headers.patch} | 0 ...rect-count_uberblocks-in-mmp.kshlib.patch} | 0 ...d-SMART-attributes-for-SSD-and-NVMe.patch} | 0 ...-to-fail-when-called-within-systemd.patch} | 0 ...16-compat-use-correct-_dec_and_test.patch} | 0 ...-MMP-writes-while-pool-is-suspended.patch} | 0 ...on-and-error-in-mmp_history-entries.patch} | 0 ...ix-free-memory-calculation-on-v3.14.patch} | 0 ...Add-scrub-after-resilver-zed-script.patch} | 2 +- ...self-test-results-to-zpool-status-c.patch} | 0 ...list-example-to-match-actual-format.patch} | 0 ...-typos.patch => 0045-Fix-some-typos.patch} | 0 ...ent-zil_itx_needcopy_bytes-properly.patch} | 0 ...-checksum-IO-delay-ratelimit-values.patch} | 0 ...nux-4.16-compat-get_disk_and_module.patch} | 0 ...long-config-lock-acquisition-in-mmp.patch} | 0 ...paces-into-account-in-policy-checks.patch} | 2 +- ...se-tests-don-t-pass-with-real-disks.patch} | 0 ...zfs-kmod-builds-when-using-rpm-4.14.patch} | 0 ... 0053-Handle-zio_resume-and-mmp-off.patch} | 0 ...MMP-write-frequency-for-large-pools.patch} | 0 ...s-0.7.7.patch => 0055-Tag-zfs-0.7.7.patch} | 0 ...ert-Handle-zap_add-failures-in-mixed.patch | 551 +++++++++++++++++ zfs-patches/0057-Tag-zfs-0.7.8.patch | 63 ++ zfs-patches/series | 57 +- 33 files changed, 1263 insertions(+), 29 deletions(-) create mode 100644 spl-patches/0014-Tag-spl-0.7.8.patch create mode 100644 zfs-patches/0028-Handle-zap_add-failures-in-mixed-case-mode.patch rename zfs-patches/{0028-mmp-should-use-a-fixed-tag-for-spa_config-locks.patch => 0029-mmp-should-use-a-fixed-tag-for-spa_config-locks.patch} (100%) rename zfs-patches/{0029-contrib-initramfs-add-missing-conf.d-zfs.patch => 0030-contrib-initramfs-add-missing-conf.d-zfs.patch} (100%) rename zfs-patches/{0030-zfs-receive-fails-with-dataset-is-busy.patch => 0031-zfs-receive-fails-with-dataset-is-busy.patch} (100%) rename zfs-patches/{0031-OpenZFS-8857-zio_remove_child-panic-due-to-already-d.patch => 0032-OpenZFS-8857-zio_remove_child-panic-due-to-already-d.patch} (100%) rename zfs-patches/{0032-Clarify-zinject-8-explanation-of-e.patch => 0033-Clarify-zinject-8-explanation-of-e.patch} (100%) rename zfs-patches/{0033-Fix-config-issues-frame-size-and-headers.patch => 0034-Fix-config-issues-frame-size-and-headers.patch} (100%) rename zfs-patches/{0034-Correct-count_uberblocks-in-mmp.kshlib.patch => 0035-Correct-count_uberblocks-in-mmp.kshlib.patch} (100%) rename zfs-patches/{0035-Add-SMART-attributes-for-SSD-and-NVMe.patch => 0036-Add-SMART-attributes-for-SSD-and-NVMe.patch} (100%) rename zfs-patches/{0036-Allow-modprobe-to-fail-when-called-within-systemd.patch => 0037-Allow-modprobe-to-fail-when-called-within-systemd.patch} (100%) rename zfs-patches/{0037-Linux-4.16-compat-use-correct-_dec_and_test.patch => 0038-Linux-4.16-compat-use-correct-_dec_and_test.patch} (100%) rename zfs-patches/{0038-Do-not-initiate-MMP-writes-while-pool-is-suspended.patch => 0039-Do-not-initiate-MMP-writes-while-pool-is-suspended.patch} (100%) rename zfs-patches/{0039-Report-duration-and-error-in-mmp_history-entries.patch => 0040-Report-duration-and-error-in-mmp_history-entries.patch} (100%) rename zfs-patches/{0040-Fix-free-memory-calculation-on-v3.14.patch => 0041-Fix-free-memory-calculation-on-v3.14.patch} (100%) rename zfs-patches/{0041-Add-scrub-after-resilver-zed-script.patch => 0042-Add-scrub-after-resilver-zed-script.patch} (99%) rename zfs-patches/{0042-Add-SMART-self-test-results-to-zpool-status-c.patch => 0043-Add-SMART-self-test-results-to-zpool-status-c.patch} (100%) rename zfs-patches/{0043-Fix-zpool-8-list-example-to-match-actual-format.patch => 0044-Fix-zpool-8-list-example-to-match-actual-format.patch} (100%) rename zfs-patches/{0044-Fix-some-typos.patch => 0045-Fix-some-typos.patch} (100%) rename zfs-patches/{0045-Increment-zil_itx_needcopy_bytes-properly.patch => 0046-Increment-zil_itx_needcopy_bytes-properly.patch} (100%) rename zfs-patches/{0046-Change-checksum-IO-delay-ratelimit-values.patch => 0047-Change-checksum-IO-delay-ratelimit-values.patch} (100%) rename zfs-patches/{0047-Linux-4.16-compat-get_disk_and_module.patch => 0048-Linux-4.16-compat-get_disk_and_module.patch} (100%) rename zfs-patches/{0048-Detect-long-config-lock-acquisition-in-mmp.patch => 0049-Detect-long-config-lock-acquisition-in-mmp.patch} (100%) rename zfs-patches/{0049-Take-user-namespaces-into-account-in-policy-checks.patch => 0050-Take-user-namespaces-into-account-in-policy-checks.patch} (99%) rename zfs-patches/{0050-zdb-and-inuse-tests-don-t-pass-with-real-disks.patch => 0051-zdb-and-inuse-tests-don-t-pass-with-real-disks.patch} (100%) rename zfs-patches/{0051-Fix-zfs-kmod-builds-when-using-rpm-4.14.patch => 0052-Fix-zfs-kmod-builds-when-using-rpm-4.14.patch} (100%) rename zfs-patches/{0052-Handle-zio_resume-and-mmp-off.patch => 0053-Handle-zio_resume-and-mmp-off.patch} (100%) rename zfs-patches/{0053-Fix-MMP-write-frequency-for-large-pools.patch => 0054-Fix-MMP-write-frequency-for-large-pools.patch} (100%) rename zfs-patches/{0054-Tag-zfs-0.7.7.patch => 0055-Tag-zfs-0.7.7.patch} (100%) create mode 100644 zfs-patches/0056-Revert-Handle-zap_add-failures-in-mixed.patch create mode 100644 zfs-patches/0057-Tag-zfs-0.7.8.patch diff --git a/spl-patches/0014-Tag-spl-0.7.8.patch b/spl-patches/0014-Tag-spl-0.7.8.patch new file mode 100644 index 0000000..85a281a --- /dev/null +++ b/spl-patches/0014-Tag-spl-0.7.8.patch @@ -0,0 +1,63 @@ +From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001 +From: Tony Hutter +Date: Mon, 9 Apr 2018 14:31:47 -0700 +Subject: [PATCH] Tag spl-0.7.8 +MIME-Version: 1.0 +Content-Type: text/plain; charset=UTF-8 +Content-Transfer-Encoding: 8bit + +META file and changelog updated. + +Signed-off-by: Tony Hutter +(cherry picked from commit 557af2687fd70191120beb406f644392de39f747) +Signed-off-by: Fabian Grünbichler +--- + rpm/generic/spl-kmod.spec.in | 3 +++ + rpm/generic/spl.spec.in | 3 +++ + META | 2 +- + 3 files changed, 7 insertions(+), 1 deletion(-) + +diff --git a/rpm/generic/spl-kmod.spec.in b/rpm/generic/spl-kmod.spec.in +index 6af292d..a604e1f 100644 +--- a/rpm/generic/spl-kmod.spec.in ++++ b/rpm/generic/spl-kmod.spec.in +@@ -167,6 +167,9 @@ chmod u+x ${RPM_BUILD_ROOT}%{kmodinstdir_prefix}/*/extra/*/*/* + rm -rf $RPM_BUILD_ROOT + + %changelog ++* Mon Apr 09 2018 Tony Hutter - 0.7.8-1 ++- Released 0.7.8-1, detailed release notes are available at: ++- https://github.com/zfsonlinux/zfs/releases/tag/zfs-0.7.8 + * Wed Mar 14 2018 Tony Hutter - 0.7.7-1 + - Released 0.7.7-1, detailed release notes are available at: + - https://github.com/zfsonlinux/zfs/releases/tag/zfs-0.7.7 +diff --git a/rpm/generic/spl.spec.in b/rpm/generic/spl.spec.in +index b693bdc..e7a718a 100644 +--- a/rpm/generic/spl.spec.in ++++ b/rpm/generic/spl.spec.in +@@ -34,6 +34,9 @@ make install DESTDIR=%{?buildroot} + %{_mandir}/man5/* + + %changelog ++* Mon Apr 09 2018 Tony Hutter - 0.7.8-1 ++- Released 0.7.8-1, detailed release notes are available at: ++- https://github.com/zfsonlinux/zfs/releases/tag/zfs-0.7.8 + * Wed Mar 14 2018 Tony Hutter - 0.7.7-1 + - Released 0.7.7-1, detailed release notes are available at: + - https://github.com/zfsonlinux/zfs/releases/tag/zfs-0.7.7 +diff --git a/META b/META +index 64a07f1..0a5132f 100644 +--- a/META ++++ b/META +@@ -1,7 +1,7 @@ + Meta: 1 + Name: spl + Branch: 1.0 +-Version: 0.7.7 ++Version: 0.7.8 + Release: 1 + Release-Tags: relext + License: GPL +-- +2.14.2 + diff --git a/spl-patches/series b/spl-patches/series index 88fb8e2..84344e9 100644 --- a/spl-patches/series +++ b/spl-patches/series @@ -11,3 +11,4 @@ 0011-Fix-more-cstyle-warnings.patch 0012-Fix-spl-kmod-builds-when-using-rpm-4.14.patch 0013-Tag-spl-0.7.7.patch +0014-Tag-spl-0.7.8.patch diff --git a/zfs-patches/0028-Handle-zap_add-failures-in-mixed-case-mode.patch b/zfs-patches/0028-Handle-zap_add-failures-in-mixed-case-mode.patch new file mode 100644 index 0000000..94d0ca1 --- /dev/null +++ b/zfs-patches/0028-Handle-zap_add-failures-in-mixed-case-mode.patch @@ -0,0 +1,553 @@ +From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001 +From: sanjeevbagewadi +Date: Fri, 9 Feb 2018 23:45:53 +0530 +Subject: [PATCH] Handle zap_add() failures in mixed case mode +MIME-Version: 1.0 +Content-Type: text/plain; charset=UTF-8 +Content-Transfer-Encoding: 8bit + +With "casesensitivity=mixed", zap_add() could fail when the number of +files/directories with the same name (varying in case) exceed the +capacity of the leaf node of a Fatzap. This results in a ASSERT() +failure as zfs_link_create() does not expect zap_add() to fail. The fix +is to handle these failures and rollback the transactions. + +Reviewed by: Matt Ahrens +Reviewed-by: Chunwei Chen +Reviewed-by: Brian Behlendorf +Signed-off-by: Sanjeev Bagewadi +Closes #7011 +Closes #7054 +(cherry picked from commit b3da003ebfad673bb4ada35f87a18a1ef175e95d) +Signed-off-by: Fabian Grünbichler +--- + .../tests/functional/casenorm/Makefile.am | 1 + + include/sys/zap_leaf.h | 15 ++- + module/zfs/zap.c | 25 +++- + module/zfs/zap_leaf.c | 2 +- + module/zfs/zap_micro.c | 38 +++++- + module/zfs/zfs_dir.c | 29 ++++- + module/zfs/zfs_vnops.c | 73 ++++++++--- + tests/runfiles/linux.run | 2 +- + .../functional/casenorm/mixed_create_failure.ksh | 136 +++++++++++++++++++++ + 9 files changed, 289 insertions(+), 32 deletions(-) + create mode 100755 tests/zfs-tests/tests/functional/casenorm/mixed_create_failure.ksh + +diff --git a/tests/zfs-tests/tests/functional/casenorm/Makefile.am b/tests/zfs-tests/tests/functional/casenorm/Makefile.am +index 00a19c7ff..00cb59074 100644 +--- a/tests/zfs-tests/tests/functional/casenorm/Makefile.am ++++ b/tests/zfs-tests/tests/functional/casenorm/Makefile.am +@@ -9,6 +9,7 @@ dist_pkgdata_SCRIPTS = \ + insensitive_formd_lookup.ksh \ + insensitive_none_delete.ksh \ + insensitive_none_lookup.ksh \ ++ mixed_create_failure.ksh \ + mixed_formd_delete.ksh \ + mixed_formd_lookup_ci.ksh \ + mixed_formd_lookup.ksh \ +diff --git a/include/sys/zap_leaf.h b/include/sys/zap_leaf.h +index e784c5963..a3da1036a 100644 +--- a/include/sys/zap_leaf.h ++++ b/include/sys/zap_leaf.h +@@ -46,10 +46,15 @@ struct zap_stats; + * block size (1<l_bs) - hash entry size (2) * number of hash + * entries - header space (2*chunksize) + */ +-#define ZAP_LEAF_NUMCHUNKS(l) \ +- (((1<<(l)->l_bs) - 2*ZAP_LEAF_HASH_NUMENTRIES(l)) / \ ++#define ZAP_LEAF_NUMCHUNKS_BS(bs) \ ++ (((1<<(bs)) - 2*ZAP_LEAF_HASH_NUMENTRIES_BS(bs)) / \ + ZAP_LEAF_CHUNKSIZE - 2) + ++#define ZAP_LEAF_NUMCHUNKS(l) (ZAP_LEAF_NUMCHUNKS_BS(((l)->l_bs))) ++ ++#define ZAP_LEAF_NUMCHUNKS_DEF \ ++ (ZAP_LEAF_NUMCHUNKS_BS(fzap_default_block_shift)) ++ + /* + * The amount of space within the chunk available for the array is: + * chunk size - space for type (1) - space for next pointer (2) +@@ -74,8 +79,10 @@ struct zap_stats; + * which is less than block size / CHUNKSIZE (24) / minimum number of + * chunks per entry (3). + */ +-#define ZAP_LEAF_HASH_SHIFT(l) ((l)->l_bs - 5) +-#define ZAP_LEAF_HASH_NUMENTRIES(l) (1 << ZAP_LEAF_HASH_SHIFT(l)) ++#define ZAP_LEAF_HASH_SHIFT_BS(bs) ((bs) - 5) ++#define ZAP_LEAF_HASH_NUMENTRIES_BS(bs) (1 << ZAP_LEAF_HASH_SHIFT_BS(bs)) ++#define ZAP_LEAF_HASH_SHIFT(l) (ZAP_LEAF_HASH_SHIFT_BS(((l)->l_bs))) ++#define ZAP_LEAF_HASH_NUMENTRIES(l) (ZAP_LEAF_HASH_NUMENTRIES_BS(((l)->l_bs))) + + /* + * The chunks start immediately after the hash table. The end of the +diff --git a/module/zfs/zap.c b/module/zfs/zap.c +index ee9962bff..9843d8c50 100644 +--- a/module/zfs/zap.c ++++ b/module/zfs/zap.c +@@ -819,15 +819,19 @@ fzap_lookup(zap_name_t *zn, + return (err); + } + ++#define MAX_EXPAND_RETRIES 2 ++ + int + fzap_add_cd(zap_name_t *zn, + uint64_t integer_size, uint64_t num_integers, + const void *val, uint32_t cd, void *tag, dmu_tx_t *tx) + { + zap_leaf_t *l; ++ zap_leaf_t *prev_l = NULL; + int err; + zap_entry_handle_t zeh; + zap_t *zap = zn->zn_zap; ++ int expand_retries = 0; + + ASSERT(RW_LOCK_HELD(&zap->zap_rwlock)); + ASSERT(!zap->zap_ismicro); +@@ -851,10 +855,29 @@ retry: + if (err == 0) { + zap_increment_num_entries(zap, 1, tx); + } else if (err == EAGAIN) { ++ /* ++ * If the last two expansions did not help, there is no point ++ * trying to expand again ++ */ ++ if (expand_retries > MAX_EXPAND_RETRIES && prev_l == l) { ++ err = SET_ERROR(ENOSPC); ++ goto out; ++ } ++ + err = zap_expand_leaf(zn, l, tag, tx, &l); + zap = zn->zn_zap; /* zap_expand_leaf() may change zap */ +- if (err == 0) ++ if (err == 0) { ++ prev_l = l; ++ expand_retries++; + goto retry; ++ } else if (err == ENOSPC) { ++ /* ++ * If we failed to expand the leaf, then bailout ++ * as there is no point trying ++ * zap_put_leaf_maybe_grow_ptrtbl(). ++ */ ++ return (err); ++ } + } + + out: +diff --git a/module/zfs/zap_leaf.c b/module/zfs/zap_leaf.c +index c342695c7..526e46606 100644 +--- a/module/zfs/zap_leaf.c ++++ b/module/zfs/zap_leaf.c +@@ -53,7 +53,7 @@ static uint16_t *zap_leaf_rehash_entry(zap_leaf_t *l, uint16_t entry); + ((h) >> \ + (64 - ZAP_LEAF_HASH_SHIFT(l) - zap_leaf_phys(l)->l_hdr.lh_prefix_len))) + +-#define LEAF_HASH_ENTPTR(l, h) (&zap_leaf_phys(l)->l_hash[LEAF_HASH(l, h)]) ++#define LEAF_HASH_ENTPTR(l, h) (&zap_leaf_phys(l)->l_hash[LEAF_HASH(l, h)]) + + extern inline zap_leaf_phys_t *zap_leaf_phys(zap_leaf_t *l); + +diff --git a/module/zfs/zap_micro.c b/module/zfs/zap_micro.c +index 3ebf995c6..34bef3e63 100644 +--- a/module/zfs/zap_micro.c ++++ b/module/zfs/zap_micro.c +@@ -363,6 +363,41 @@ mze_find_unused_cd(zap_t *zap, uint64_t hash) + return (cd); + } + ++/* ++ * Each mzap entry requires at max : 4 chunks ++ * 3 chunks for names + 1 chunk for value. ++ */ ++#define MZAP_ENT_CHUNKS (1 + ZAP_LEAF_ARRAY_NCHUNKS(MZAP_NAME_LEN) + \ ++ ZAP_LEAF_ARRAY_NCHUNKS(sizeof (uint64_t))) ++ ++/* ++ * Check if the current entry keeps the colliding entries under the fatzap leaf ++ * size. ++ */ ++static boolean_t ++mze_canfit_fzap_leaf(zap_name_t *zn, uint64_t hash) ++{ ++ zap_t *zap = zn->zn_zap; ++ mzap_ent_t mze_tofind; ++ mzap_ent_t *mze; ++ avl_index_t idx; ++ avl_tree_t *avl = &zap->zap_m.zap_avl; ++ uint32_t mzap_ents = 0; ++ ++ mze_tofind.mze_hash = hash; ++ mze_tofind.mze_cd = 0; ++ ++ for (mze = avl_find(avl, &mze_tofind, &idx); ++ mze && mze->mze_hash == hash; mze = AVL_NEXT(avl, mze)) { ++ mzap_ents++; ++ } ++ ++ /* Include the new entry being added */ ++ mzap_ents++; ++ ++ return (ZAP_LEAF_NUMCHUNKS_DEF > (mzap_ents * MZAP_ENT_CHUNKS)); ++} ++ + static void + mze_remove(zap_t *zap, mzap_ent_t *mze) + { +@@ -1191,7 +1226,8 @@ zap_add_impl(zap_t *zap, const char *key, + err = fzap_add(zn, integer_size, num_integers, val, tag, tx); + zap = zn->zn_zap; /* fzap_add() may change zap */ + } else if (integer_size != 8 || num_integers != 1 || +- strlen(key) >= MZAP_NAME_LEN) { ++ strlen(key) >= MZAP_NAME_LEN || ++ !mze_canfit_fzap_leaf(zn, zn->zn_hash)) { + err = mzap_upgrade(&zn->zn_zap, tag, tx, 0); + if (err == 0) { + err = fzap_add(zn, integer_size, num_integers, val, +diff --git a/module/zfs/zfs_dir.c b/module/zfs/zfs_dir.c +index 9a8bbccd9..6398a1d15 100644 +--- a/module/zfs/zfs_dir.c ++++ b/module/zfs/zfs_dir.c +@@ -742,7 +742,11 @@ zfs_dirent(znode_t *zp, uint64_t mode) + } + + /* +- * Link zp into dl. Can only fail if zp has been unlinked. ++ * Link zp into dl. Can fail in the following cases : ++ * - if zp has been unlinked. ++ * - if the number of entries with the same hash (aka. colliding entries) ++ * exceed the capacity of a leaf-block of fatzap and splitting of the ++ * leaf-block does not help. + */ + int + zfs_link_create(zfs_dirlock_t *dl, znode_t *zp, dmu_tx_t *tx, int flag) +@@ -776,6 +780,24 @@ zfs_link_create(zfs_dirlock_t *dl, znode_t *zp, dmu_tx_t *tx, int flag) + NULL, &links, sizeof (links)); + } + } ++ ++ value = zfs_dirent(zp, zp->z_mode); ++ error = zap_add(ZTOZSB(zp)->z_os, dzp->z_id, dl->dl_name, 8, 1, ++ &value, tx); ++ ++ /* ++ * zap_add could fail to add the entry if it exceeds the capacity of the ++ * leaf-block and zap_leaf_split() failed to help. ++ * The caller of this routine is responsible for failing the transaction ++ * which will rollback the SA updates done above. ++ */ ++ if (error != 0) { ++ if (!(flag & ZRENAMING) && !(flag & ZNEW)) ++ drop_nlink(ZTOI(zp)); ++ mutex_exit(&zp->z_lock); ++ return (error); ++ } ++ + SA_ADD_BULK_ATTR(bulk, count, SA_ZPL_PARENT(zfsvfs), NULL, + &dzp->z_id, sizeof (dzp->z_id)); + SA_ADD_BULK_ATTR(bulk, count, SA_ZPL_FLAGS(zfsvfs), NULL, +@@ -813,11 +835,6 @@ zfs_link_create(zfs_dirlock_t *dl, znode_t *zp, dmu_tx_t *tx, int flag) + ASSERT(error == 0); + mutex_exit(&dzp->z_lock); + +- value = zfs_dirent(zp, zp->z_mode); +- error = zap_add(ZTOZSB(zp)->z_os, dzp->z_id, dl->dl_name, +- 8, 1, &value, tx); +- ASSERT(error == 0); +- + return (0); + } + +diff --git a/module/zfs/zfs_vnops.c b/module/zfs/zfs_vnops.c +index 6f6ce79db..8a7ad702c 100644 +--- a/module/zfs/zfs_vnops.c ++++ b/module/zfs/zfs_vnops.c +@@ -1443,10 +1443,22 @@ top: + } + zfs_mknode(dzp, vap, tx, cr, 0, &zp, &acl_ids); + ++ error = zfs_link_create(dl, zp, tx, ZNEW); ++ if (error != 0) { ++ /* ++ * Since, we failed to add the directory entry for it, ++ * delete the newly created dnode. ++ */ ++ zfs_znode_delete(zp, tx); ++ remove_inode_hash(ZTOI(zp)); ++ zfs_acl_ids_free(&acl_ids); ++ dmu_tx_commit(tx); ++ goto out; ++ } ++ + if (fuid_dirtied) + zfs_fuid_sync(zfsvfs, tx); + +- (void) zfs_link_create(dl, zp, tx, ZNEW); + txtype = zfs_log_create_txtype(Z_FILE, vsecp, vap); + if (flag & FIGNORECASE) + txtype |= TX_CI; +@@ -2037,13 +2049,18 @@ top: + */ + zfs_mknode(dzp, vap, tx, cr, 0, &zp, &acl_ids); + +- if (fuid_dirtied) +- zfs_fuid_sync(zfsvfs, tx); +- + /* + * Now put new name in parent dir. + */ +- (void) zfs_link_create(dl, zp, tx, ZNEW); ++ error = zfs_link_create(dl, zp, tx, ZNEW); ++ if (error != 0) { ++ zfs_znode_delete(zp, tx); ++ remove_inode_hash(ZTOI(zp)); ++ goto out; ++ } ++ ++ if (fuid_dirtied) ++ zfs_fuid_sync(zfsvfs, tx); + + *ipp = ZTOI(zp); + +@@ -2053,6 +2070,7 @@ top: + zfs_log_create(zilog, tx, txtype, dzp, zp, dirname, vsecp, + acl_ids.z_fuidp, vap); + ++out: + zfs_acl_ids_free(&acl_ids); + + dmu_tx_commit(tx); +@@ -2062,10 +2080,14 @@ top: + if (zfsvfs->z_os->os_sync == ZFS_SYNC_ALWAYS) + zil_commit(zilog, 0); + +- zfs_inode_update(dzp); +- zfs_inode_update(zp); ++ if (error != 0) { ++ iput(ZTOI(zp)); ++ } else { ++ zfs_inode_update(dzp); ++ zfs_inode_update(zp); ++ } + ZFS_EXIT(zfsvfs); +- return (0); ++ return (error); + } + + /* +@@ -3683,6 +3705,13 @@ top: + VERIFY3U(zfs_link_destroy(tdl, szp, tx, + ZRENAMING, NULL), ==, 0); + } ++ } else { ++ /* ++ * If we had removed the existing target, subsequent ++ * call to zfs_link_create() to add back the same entry ++ * but, the new dnode (szp) should not fail. ++ */ ++ ASSERT(tzp == NULL); + } + } + +@@ -3853,14 +3882,18 @@ top: + /* + * Insert the new object into the directory. + */ +- (void) zfs_link_create(dl, zp, tx, ZNEW); +- +- if (flags & FIGNORECASE) +- txtype |= TX_CI; +- zfs_log_symlink(zilog, tx, txtype, dzp, zp, name, link); ++ error = zfs_link_create(dl, zp, tx, ZNEW); ++ if (error != 0) { ++ zfs_znode_delete(zp, tx); ++ remove_inode_hash(ZTOI(zp)); ++ } else { ++ if (flags & FIGNORECASE) ++ txtype |= TX_CI; ++ zfs_log_symlink(zilog, tx, txtype, dzp, zp, name, link); + +- zfs_inode_update(dzp); +- zfs_inode_update(zp); ++ zfs_inode_update(dzp); ++ zfs_inode_update(zp); ++ } + + zfs_acl_ids_free(&acl_ids); + +@@ -3868,10 +3901,14 @@ top: + + zfs_dirent_unlock(dl); + +- *ipp = ZTOI(zp); ++ if (error == 0) { ++ *ipp = ZTOI(zp); + +- if (zfsvfs->z_os->os_sync == ZFS_SYNC_ALWAYS) +- zil_commit(zilog, 0); ++ if (zfsvfs->z_os->os_sync == ZFS_SYNC_ALWAYS) ++ zil_commit(zilog, 0); ++ } else { ++ iput(ZTOI(zp)); ++ } + + ZFS_EXIT(zfsvfs); + return (error); +diff --git a/tests/runfiles/linux.run b/tests/runfiles/linux.run +index ea2dbb282..8be3e1c62 100644 +--- a/tests/runfiles/linux.run ++++ b/tests/runfiles/linux.run +@@ -55,7 +55,7 @@ tags = ['functional', 'cachefile'] + # 'mixed_none_lookup', 'mixed_none_lookup_ci', 'mixed_none_delete', + # 'mixed_formd_lookup', 'mixed_formd_lookup_ci', 'mixed_formd_delete'] + [tests/functional/casenorm] +-tests = ['case_all_values', 'norm_all_values'] ++tests = ['case_all_values', 'norm_all_values', 'mixed_create_failure'] + tags = ['functional', 'casenorm'] + + [tests/functional/chattr] +diff --git a/tests/zfs-tests/tests/functional/casenorm/mixed_create_failure.ksh b/tests/zfs-tests/tests/functional/casenorm/mixed_create_failure.ksh +new file mode 100755 +index 000000000..51b5bb3f6 +--- /dev/null ++++ b/tests/zfs-tests/tests/functional/casenorm/mixed_create_failure.ksh +@@ -0,0 +1,136 @@ ++#!/bin/ksh -p ++# ++# ++# This file and its contents are supplied under the terms of the ++# Common Development and Distribution License ("CDDL"), version 1.0. ++# You may only use this file in accordance with the terms of version ++# 1.0 of the CDDL. ++# ++# A full copy of the text of the CDDL should have accompanied this ++# source. A copy of the CDDL is also available via the Internet at ++# http://www.illumos.org/license/CDDL. ++# ++# ++# Copyright 2018 Nutanix Inc. All rights reserved. ++# ++ ++. $STF_SUITE/tests/functional/casenorm/casenorm.kshlib ++ ++# DESCRIPTION: ++# For the filesystem with casesensitivity=mixed, normalization=none, ++# when multiple files with the same name (differing only in case) are created, ++# the number of files is limited to what can fit in a fatzap leaf-block. ++# And beyond that, it fails with ENOSPC. ++# ++# Ensure that the create/rename operations fail gracefully and not trigger an ++# ASSERT. ++# ++# STRATEGY: ++# Repeat the below steps for objects: files, directories, symlinks and hardlinks ++# 1. Create objects with same name but varying in case. ++# E.g. 'abcdefghijklmnop', 'Abcdefghijklmnop', 'ABcdefghijklmnop' etc. ++# The create should fail with ENOSPC. ++# 2. Create an object with name 'tmp_obj' and try to rename it to name that we ++# failed to add in step 1 above. ++# This should fail as well. ++ ++verify_runnable "global" ++ ++function cleanup ++{ ++ destroy_testfs ++} ++ ++log_onexit cleanup ++log_assert "With mixed mode: ensure create fails with ENOSPC beyond a certain limit" ++ ++create_testfs "-o casesensitivity=mixed -o normalization=none" ++ ++# Different object types ++obj_type=('file' 'dir' 'symlink' 'hardlink') ++ ++# Commands to create different object types ++typeset -A ops ++ops['file']='touch' ++ops['dir']='mkdir' ++ops['symlink']='ln -s' ++ops['hardlink']='ln' ++ ++# This function tests the following for a give object type : ++# - Create multiple objects with the same name (varying only in case). ++# Ensure that it eventually fails once the leaf-block limit is exceeded. ++# - Create another object with a different name. And attempt rename it to the ++# name (for which the create had failed in the previous step). ++# This should fail as well. ++# Args : ++# $1 - object type (file/dir/symlink/hardlink) ++# $2 - test directory ++# ++function test_ops ++{ ++ typeset obj_type=$1 ++ typeset testdir=$2 ++ ++ target_obj='target-file' ++ ++ op="${ops[$obj_type]}" ++ ++ log_note "The op : $op" ++ log_note "testdir=$testdir obj_type=$obj_type" ++ ++ test_path="$testdir/$obj_type" ++ mkdir $test_path ++ log_note "Created test dir $test_path" ++ ++ if [[ $obj_type = "symlink" || $obj_type = "hardlink" ]]; then ++ touch $test_path/$target_obj ++ log_note "Created target: $test_path/$target_obj" ++ op="$op $test_path/$target_obj" ++ fi ++ ++ log_note "op : $op" ++ names='{a,A}{b,B}{c,C}{d,D}{e,E}{f,F}{g,G}{h,H}{i,I}{j,J}{k,K}{l,L}' ++ for name in $names; do ++ cmd="$op $test_path/$name" ++ out=$($cmd 2>&1) ++ ret=$? ++ log_note "cmd: $cmd ret: $ret out=$out" ++ if (($ret != 0)); then ++ if [[ $out = *@(No space left on device)* ]]; then ++ save_name="$test_path/$name" ++ break; ++ else ++ log_err "$cmd failed with unexpected error : $out" ++ fi ++ fi ++ done ++ ++ log_note 'Test rename \"sample_name\" rename' ++ TMP_OBJ="$test_path/tmp_obj" ++ cmd="$op $TMP_OBJ" ++ out=$($cmd 2>&1) ++ ret=$? ++ if (($ret != 0)); then ++ log_err "cmd:$cmd failed out:$out" ++ fi ++ ++ # Now, try to rename the tmp_obj to the name which we failed to add earlier. ++ # This should fail as well. ++ out=$(mv $TMP_OBJ $save_name 2>&1) ++ ret=$? ++ if (($ret != 0)); then ++ if [[ $out = *@(No space left on device)* ]]; then ++ log_note "$cmd failed as expected : $out" ++ else ++ log_err "$cmd failed with : $out" ++ fi ++ fi ++} ++ ++for obj_type in ${obj_type[*]}; ++do ++ log_note "Testing create of $obj_type" ++ test_ops $obj_type $TESTDIR ++done ++ ++log_pass "Mixed mode FS: Ops on large number of colliding names fail gracefully" +-- +2.14.2 + diff --git a/zfs-patches/0028-mmp-should-use-a-fixed-tag-for-spa_config-locks.patch b/zfs-patches/0029-mmp-should-use-a-fixed-tag-for-spa_config-locks.patch similarity index 100% rename from zfs-patches/0028-mmp-should-use-a-fixed-tag-for-spa_config-locks.patch rename to zfs-patches/0029-mmp-should-use-a-fixed-tag-for-spa_config-locks.patch diff --git a/zfs-patches/0029-contrib-initramfs-add-missing-conf.d-zfs.patch b/zfs-patches/0030-contrib-initramfs-add-missing-conf.d-zfs.patch similarity index 100% rename from zfs-patches/0029-contrib-initramfs-add-missing-conf.d-zfs.patch rename to zfs-patches/0030-contrib-initramfs-add-missing-conf.d-zfs.patch diff --git a/zfs-patches/0030-zfs-receive-fails-with-dataset-is-busy.patch b/zfs-patches/0031-zfs-receive-fails-with-dataset-is-busy.patch similarity index 100% rename from zfs-patches/0030-zfs-receive-fails-with-dataset-is-busy.patch rename to zfs-patches/0031-zfs-receive-fails-with-dataset-is-busy.patch diff --git a/zfs-patches/0031-OpenZFS-8857-zio_remove_child-panic-due-to-already-d.patch b/zfs-patches/0032-OpenZFS-8857-zio_remove_child-panic-due-to-already-d.patch similarity index 100% rename from zfs-patches/0031-OpenZFS-8857-zio_remove_child-panic-due-to-already-d.patch rename to zfs-patches/0032-OpenZFS-8857-zio_remove_child-panic-due-to-already-d.patch diff --git a/zfs-patches/0032-Clarify-zinject-8-explanation-of-e.patch b/zfs-patches/0033-Clarify-zinject-8-explanation-of-e.patch similarity index 100% rename from zfs-patches/0032-Clarify-zinject-8-explanation-of-e.patch rename to zfs-patches/0033-Clarify-zinject-8-explanation-of-e.patch diff --git a/zfs-patches/0033-Fix-config-issues-frame-size-and-headers.patch b/zfs-patches/0034-Fix-config-issues-frame-size-and-headers.patch similarity index 100% rename from zfs-patches/0033-Fix-config-issues-frame-size-and-headers.patch rename to zfs-patches/0034-Fix-config-issues-frame-size-and-headers.patch diff --git a/zfs-patches/0034-Correct-count_uberblocks-in-mmp.kshlib.patch b/zfs-patches/0035-Correct-count_uberblocks-in-mmp.kshlib.patch similarity index 100% rename from zfs-patches/0034-Correct-count_uberblocks-in-mmp.kshlib.patch rename to zfs-patches/0035-Correct-count_uberblocks-in-mmp.kshlib.patch diff --git a/zfs-patches/0035-Add-SMART-attributes-for-SSD-and-NVMe.patch b/zfs-patches/0036-Add-SMART-attributes-for-SSD-and-NVMe.patch similarity index 100% rename from zfs-patches/0035-Add-SMART-attributes-for-SSD-and-NVMe.patch rename to zfs-patches/0036-Add-SMART-attributes-for-SSD-and-NVMe.patch diff --git a/zfs-patches/0036-Allow-modprobe-to-fail-when-called-within-systemd.patch b/zfs-patches/0037-Allow-modprobe-to-fail-when-called-within-systemd.patch similarity index 100% rename from zfs-patches/0036-Allow-modprobe-to-fail-when-called-within-systemd.patch rename to zfs-patches/0037-Allow-modprobe-to-fail-when-called-within-systemd.patch diff --git a/zfs-patches/0037-Linux-4.16-compat-use-correct-_dec_and_test.patch b/zfs-patches/0038-Linux-4.16-compat-use-correct-_dec_and_test.patch similarity index 100% rename from zfs-patches/0037-Linux-4.16-compat-use-correct-_dec_and_test.patch rename to zfs-patches/0038-Linux-4.16-compat-use-correct-_dec_and_test.patch diff --git a/zfs-patches/0038-Do-not-initiate-MMP-writes-while-pool-is-suspended.patch b/zfs-patches/0039-Do-not-initiate-MMP-writes-while-pool-is-suspended.patch similarity index 100% rename from zfs-patches/0038-Do-not-initiate-MMP-writes-while-pool-is-suspended.patch rename to zfs-patches/0039-Do-not-initiate-MMP-writes-while-pool-is-suspended.patch diff --git a/zfs-patches/0039-Report-duration-and-error-in-mmp_history-entries.patch b/zfs-patches/0040-Report-duration-and-error-in-mmp_history-entries.patch similarity index 100% rename from zfs-patches/0039-Report-duration-and-error-in-mmp_history-entries.patch rename to zfs-patches/0040-Report-duration-and-error-in-mmp_history-entries.patch diff --git a/zfs-patches/0040-Fix-free-memory-calculation-on-v3.14.patch b/zfs-patches/0041-Fix-free-memory-calculation-on-v3.14.patch similarity index 100% rename from zfs-patches/0040-Fix-free-memory-calculation-on-v3.14.patch rename to zfs-patches/0041-Fix-free-memory-calculation-on-v3.14.patch diff --git a/zfs-patches/0041-Add-scrub-after-resilver-zed-script.patch b/zfs-patches/0042-Add-scrub-after-resilver-zed-script.patch similarity index 99% rename from zfs-patches/0041-Add-scrub-after-resilver-zed-script.patch rename to zfs-patches/0042-Add-scrub-after-resilver-zed-script.patch index 7c6002c..ee8b608 100644 --- a/zfs-patches/0041-Add-scrub-after-resilver-zed-script.patch +++ b/zfs-patches/0042-Add-scrub-after-resilver-zed-script.patch @@ -261,7 +261,7 @@ index a1dd33704..8b0e476d5 100644 ## # The syslog priority (e.g., specified as a "facility.level" pair). diff --git a/tests/runfiles/linux.run b/tests/runfiles/linux.run -index ea2dbb282..e4a11a0fa 100644 +index 8be3e1c62..89c923db1 100644 --- a/tests/runfiles/linux.run +++ b/tests/runfiles/linux.run @@ -421,7 +421,7 @@ tests = ['exec_001_pos', 'exec_002_neg'] diff --git a/zfs-patches/0042-Add-SMART-self-test-results-to-zpool-status-c.patch b/zfs-patches/0043-Add-SMART-self-test-results-to-zpool-status-c.patch similarity index 100% rename from zfs-patches/0042-Add-SMART-self-test-results-to-zpool-status-c.patch rename to zfs-patches/0043-Add-SMART-self-test-results-to-zpool-status-c.patch diff --git a/zfs-patches/0043-Fix-zpool-8-list-example-to-match-actual-format.patch b/zfs-patches/0044-Fix-zpool-8-list-example-to-match-actual-format.patch similarity index 100% rename from zfs-patches/0043-Fix-zpool-8-list-example-to-match-actual-format.patch rename to zfs-patches/0044-Fix-zpool-8-list-example-to-match-actual-format.patch diff --git a/zfs-patches/0044-Fix-some-typos.patch b/zfs-patches/0045-Fix-some-typos.patch similarity index 100% rename from zfs-patches/0044-Fix-some-typos.patch rename to zfs-patches/0045-Fix-some-typos.patch diff --git a/zfs-patches/0045-Increment-zil_itx_needcopy_bytes-properly.patch b/zfs-patches/0046-Increment-zil_itx_needcopy_bytes-properly.patch similarity index 100% rename from zfs-patches/0045-Increment-zil_itx_needcopy_bytes-properly.patch rename to zfs-patches/0046-Increment-zil_itx_needcopy_bytes-properly.patch diff --git a/zfs-patches/0046-Change-checksum-IO-delay-ratelimit-values.patch b/zfs-patches/0047-Change-checksum-IO-delay-ratelimit-values.patch similarity index 100% rename from zfs-patches/0046-Change-checksum-IO-delay-ratelimit-values.patch rename to zfs-patches/0047-Change-checksum-IO-delay-ratelimit-values.patch diff --git a/zfs-patches/0047-Linux-4.16-compat-get_disk_and_module.patch b/zfs-patches/0048-Linux-4.16-compat-get_disk_and_module.patch similarity index 100% rename from zfs-patches/0047-Linux-4.16-compat-get_disk_and_module.patch rename to zfs-patches/0048-Linux-4.16-compat-get_disk_and_module.patch diff --git a/zfs-patches/0048-Detect-long-config-lock-acquisition-in-mmp.patch b/zfs-patches/0049-Detect-long-config-lock-acquisition-in-mmp.patch similarity index 100% rename from zfs-patches/0048-Detect-long-config-lock-acquisition-in-mmp.patch rename to zfs-patches/0049-Detect-long-config-lock-acquisition-in-mmp.patch diff --git a/zfs-patches/0049-Take-user-namespaces-into-account-in-policy-checks.patch b/zfs-patches/0050-Take-user-namespaces-into-account-in-policy-checks.patch similarity index 99% rename from zfs-patches/0049-Take-user-namespaces-into-account-in-policy-checks.patch rename to zfs-patches/0050-Take-user-namespaces-into-account-in-policy-checks.patch index c0349e8..604a33d 100644 --- a/zfs-patches/0049-Take-user-namespaces-into-account-in-policy-checks.patch +++ b/zfs-patches/0050-Take-user-namespaces-into-account-in-policy-checks.patch @@ -528,7 +528,7 @@ index 419ed1a2c..910d4ff25 100644 AS_IF([test "$LINUX_OBJ" != "$LINUX"], [ KERNELMAKE_PARAMS="$KERNELMAKE_PARAMS O=$LINUX_OBJ" diff --git a/tests/runfiles/linux.run b/tests/runfiles/linux.run -index e4a11a0fa..4b2694202 100644 +index 89c923db1..25ae3fe5e 100644 --- a/tests/runfiles/linux.run +++ b/tests/runfiles/linux.run @@ -648,6 +648,10 @@ tags = ['functional', 'truncate'] diff --git a/zfs-patches/0050-zdb-and-inuse-tests-don-t-pass-with-real-disks.patch b/zfs-patches/0051-zdb-and-inuse-tests-don-t-pass-with-real-disks.patch similarity index 100% rename from zfs-patches/0050-zdb-and-inuse-tests-don-t-pass-with-real-disks.patch rename to zfs-patches/0051-zdb-and-inuse-tests-don-t-pass-with-real-disks.patch diff --git a/zfs-patches/0051-Fix-zfs-kmod-builds-when-using-rpm-4.14.patch b/zfs-patches/0052-Fix-zfs-kmod-builds-when-using-rpm-4.14.patch similarity index 100% rename from zfs-patches/0051-Fix-zfs-kmod-builds-when-using-rpm-4.14.patch rename to zfs-patches/0052-Fix-zfs-kmod-builds-when-using-rpm-4.14.patch diff --git a/zfs-patches/0052-Handle-zio_resume-and-mmp-off.patch b/zfs-patches/0053-Handle-zio_resume-and-mmp-off.patch similarity index 100% rename from zfs-patches/0052-Handle-zio_resume-and-mmp-off.patch rename to zfs-patches/0053-Handle-zio_resume-and-mmp-off.patch diff --git a/zfs-patches/0053-Fix-MMP-write-frequency-for-large-pools.patch b/zfs-patches/0054-Fix-MMP-write-frequency-for-large-pools.patch similarity index 100% rename from zfs-patches/0053-Fix-MMP-write-frequency-for-large-pools.patch rename to zfs-patches/0054-Fix-MMP-write-frequency-for-large-pools.patch diff --git a/zfs-patches/0054-Tag-zfs-0.7.7.patch b/zfs-patches/0055-Tag-zfs-0.7.7.patch similarity index 100% rename from zfs-patches/0054-Tag-zfs-0.7.7.patch rename to zfs-patches/0055-Tag-zfs-0.7.7.patch diff --git a/zfs-patches/0056-Revert-Handle-zap_add-failures-in-mixed.patch b/zfs-patches/0056-Revert-Handle-zap_add-failures-in-mixed.patch new file mode 100644 index 0000000..77f3ecb --- /dev/null +++ b/zfs-patches/0056-Revert-Handle-zap_add-failures-in-mixed.patch @@ -0,0 +1,551 @@ +From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001 +From: Tony Hutter +Date: Mon, 9 Apr 2018 14:24:46 -0700 +Subject: [PATCH] Revert "Handle zap_add() failures in mixed ... " +MIME-Version: 1.0 +Content-Type: text/plain; charset=UTF-8 +Content-Transfer-Encoding: 8bit + +This reverts commit cc63068e95ee725cce03b1b7ce50179825a6cda5. + +Under certain circumstances this change can result in an ENOSPC +error when adding new files to a directory. See #7401 for full +details. + +Reviewed-by: Brian Behlendorf +Signed-off-by: Tony Hutter +Issue #7401 +Closes #7416 +(cherry picked from commit 9a2e90c9fc469d377c14eb863952261f9ec12d2c) +Signed-off-by: Fabian Grünbichler +--- + .../tests/functional/casenorm/Makefile.am | 1 - + include/sys/zap_leaf.h | 15 +-- + module/zfs/zap.c | 25 +--- + module/zfs/zap_leaf.c | 2 +- + module/zfs/zap_micro.c | 38 +----- + module/zfs/zfs_dir.c | 29 +---- + module/zfs/zfs_vnops.c | 73 +++-------- + tests/runfiles/linux.run | 2 +- + .../functional/casenorm/mixed_create_failure.ksh | 136 --------------------- + 9 files changed, 32 insertions(+), 289 deletions(-) + delete mode 100755 tests/zfs-tests/tests/functional/casenorm/mixed_create_failure.ksh + +diff --git a/tests/zfs-tests/tests/functional/casenorm/Makefile.am b/tests/zfs-tests/tests/functional/casenorm/Makefile.am +index 00cb59074..00a19c7ff 100644 +--- a/tests/zfs-tests/tests/functional/casenorm/Makefile.am ++++ b/tests/zfs-tests/tests/functional/casenorm/Makefile.am +@@ -9,7 +9,6 @@ dist_pkgdata_SCRIPTS = \ + insensitive_formd_lookup.ksh \ + insensitive_none_delete.ksh \ + insensitive_none_lookup.ksh \ +- mixed_create_failure.ksh \ + mixed_formd_delete.ksh \ + mixed_formd_lookup_ci.ksh \ + mixed_formd_lookup.ksh \ +diff --git a/include/sys/zap_leaf.h b/include/sys/zap_leaf.h +index a3da1036a..e784c5963 100644 +--- a/include/sys/zap_leaf.h ++++ b/include/sys/zap_leaf.h +@@ -46,15 +46,10 @@ struct zap_stats; + * block size (1<l_bs) - hash entry size (2) * number of hash + * entries - header space (2*chunksize) + */ +-#define ZAP_LEAF_NUMCHUNKS_BS(bs) \ +- (((1<<(bs)) - 2*ZAP_LEAF_HASH_NUMENTRIES_BS(bs)) / \ ++#define ZAP_LEAF_NUMCHUNKS(l) \ ++ (((1<<(l)->l_bs) - 2*ZAP_LEAF_HASH_NUMENTRIES(l)) / \ + ZAP_LEAF_CHUNKSIZE - 2) + +-#define ZAP_LEAF_NUMCHUNKS(l) (ZAP_LEAF_NUMCHUNKS_BS(((l)->l_bs))) +- +-#define ZAP_LEAF_NUMCHUNKS_DEF \ +- (ZAP_LEAF_NUMCHUNKS_BS(fzap_default_block_shift)) +- + /* + * The amount of space within the chunk available for the array is: + * chunk size - space for type (1) - space for next pointer (2) +@@ -79,10 +74,8 @@ struct zap_stats; + * which is less than block size / CHUNKSIZE (24) / minimum number of + * chunks per entry (3). + */ +-#define ZAP_LEAF_HASH_SHIFT_BS(bs) ((bs) - 5) +-#define ZAP_LEAF_HASH_NUMENTRIES_BS(bs) (1 << ZAP_LEAF_HASH_SHIFT_BS(bs)) +-#define ZAP_LEAF_HASH_SHIFT(l) (ZAP_LEAF_HASH_SHIFT_BS(((l)->l_bs))) +-#define ZAP_LEAF_HASH_NUMENTRIES(l) (ZAP_LEAF_HASH_NUMENTRIES_BS(((l)->l_bs))) ++#define ZAP_LEAF_HASH_SHIFT(l) ((l)->l_bs - 5) ++#define ZAP_LEAF_HASH_NUMENTRIES(l) (1 << ZAP_LEAF_HASH_SHIFT(l)) + + /* + * The chunks start immediately after the hash table. The end of the +diff --git a/module/zfs/zap.c b/module/zfs/zap.c +index 9843d8c50..ee9962bff 100644 +--- a/module/zfs/zap.c ++++ b/module/zfs/zap.c +@@ -819,19 +819,15 @@ fzap_lookup(zap_name_t *zn, + return (err); + } + +-#define MAX_EXPAND_RETRIES 2 +- + int + fzap_add_cd(zap_name_t *zn, + uint64_t integer_size, uint64_t num_integers, + const void *val, uint32_t cd, void *tag, dmu_tx_t *tx) + { + zap_leaf_t *l; +- zap_leaf_t *prev_l = NULL; + int err; + zap_entry_handle_t zeh; + zap_t *zap = zn->zn_zap; +- int expand_retries = 0; + + ASSERT(RW_LOCK_HELD(&zap->zap_rwlock)); + ASSERT(!zap->zap_ismicro); +@@ -855,29 +851,10 @@ retry: + if (err == 0) { + zap_increment_num_entries(zap, 1, tx); + } else if (err == EAGAIN) { +- /* +- * If the last two expansions did not help, there is no point +- * trying to expand again +- */ +- if (expand_retries > MAX_EXPAND_RETRIES && prev_l == l) { +- err = SET_ERROR(ENOSPC); +- goto out; +- } +- + err = zap_expand_leaf(zn, l, tag, tx, &l); + zap = zn->zn_zap; /* zap_expand_leaf() may change zap */ +- if (err == 0) { +- prev_l = l; +- expand_retries++; ++ if (err == 0) + goto retry; +- } else if (err == ENOSPC) { +- /* +- * If we failed to expand the leaf, then bailout +- * as there is no point trying +- * zap_put_leaf_maybe_grow_ptrtbl(). +- */ +- return (err); +- } + } + + out: +diff --git a/module/zfs/zap_leaf.c b/module/zfs/zap_leaf.c +index 526e46606..c342695c7 100644 +--- a/module/zfs/zap_leaf.c ++++ b/module/zfs/zap_leaf.c +@@ -53,7 +53,7 @@ static uint16_t *zap_leaf_rehash_entry(zap_leaf_t *l, uint16_t entry); + ((h) >> \ + (64 - ZAP_LEAF_HASH_SHIFT(l) - zap_leaf_phys(l)->l_hdr.lh_prefix_len))) + +-#define LEAF_HASH_ENTPTR(l, h) (&zap_leaf_phys(l)->l_hash[LEAF_HASH(l, h)]) ++#define LEAF_HASH_ENTPTR(l, h) (&zap_leaf_phys(l)->l_hash[LEAF_HASH(l, h)]) + + extern inline zap_leaf_phys_t *zap_leaf_phys(zap_leaf_t *l); + +diff --git a/module/zfs/zap_micro.c b/module/zfs/zap_micro.c +index 34bef3e63..3ebf995c6 100644 +--- a/module/zfs/zap_micro.c ++++ b/module/zfs/zap_micro.c +@@ -363,41 +363,6 @@ mze_find_unused_cd(zap_t *zap, uint64_t hash) + return (cd); + } + +-/* +- * Each mzap entry requires at max : 4 chunks +- * 3 chunks for names + 1 chunk for value. +- */ +-#define MZAP_ENT_CHUNKS (1 + ZAP_LEAF_ARRAY_NCHUNKS(MZAP_NAME_LEN) + \ +- ZAP_LEAF_ARRAY_NCHUNKS(sizeof (uint64_t))) +- +-/* +- * Check if the current entry keeps the colliding entries under the fatzap leaf +- * size. +- */ +-static boolean_t +-mze_canfit_fzap_leaf(zap_name_t *zn, uint64_t hash) +-{ +- zap_t *zap = zn->zn_zap; +- mzap_ent_t mze_tofind; +- mzap_ent_t *mze; +- avl_index_t idx; +- avl_tree_t *avl = &zap->zap_m.zap_avl; +- uint32_t mzap_ents = 0; +- +- mze_tofind.mze_hash = hash; +- mze_tofind.mze_cd = 0; +- +- for (mze = avl_find(avl, &mze_tofind, &idx); +- mze && mze->mze_hash == hash; mze = AVL_NEXT(avl, mze)) { +- mzap_ents++; +- } +- +- /* Include the new entry being added */ +- mzap_ents++; +- +- return (ZAP_LEAF_NUMCHUNKS_DEF > (mzap_ents * MZAP_ENT_CHUNKS)); +-} +- + static void + mze_remove(zap_t *zap, mzap_ent_t *mze) + { +@@ -1226,8 +1191,7 @@ zap_add_impl(zap_t *zap, const char *key, + err = fzap_add(zn, integer_size, num_integers, val, tag, tx); + zap = zn->zn_zap; /* fzap_add() may change zap */ + } else if (integer_size != 8 || num_integers != 1 || +- strlen(key) >= MZAP_NAME_LEN || +- !mze_canfit_fzap_leaf(zn, zn->zn_hash)) { ++ strlen(key) >= MZAP_NAME_LEN) { + err = mzap_upgrade(&zn->zn_zap, tag, tx, 0); + if (err == 0) { + err = fzap_add(zn, integer_size, num_integers, val, +diff --git a/module/zfs/zfs_dir.c b/module/zfs/zfs_dir.c +index 6398a1d15..9a8bbccd9 100644 +--- a/module/zfs/zfs_dir.c ++++ b/module/zfs/zfs_dir.c +@@ -742,11 +742,7 @@ zfs_dirent(znode_t *zp, uint64_t mode) + } + + /* +- * Link zp into dl. Can fail in the following cases : +- * - if zp has been unlinked. +- * - if the number of entries with the same hash (aka. colliding entries) +- * exceed the capacity of a leaf-block of fatzap and splitting of the +- * leaf-block does not help. ++ * Link zp into dl. Can only fail if zp has been unlinked. + */ + int + zfs_link_create(zfs_dirlock_t *dl, znode_t *zp, dmu_tx_t *tx, int flag) +@@ -780,24 +776,6 @@ zfs_link_create(zfs_dirlock_t *dl, znode_t *zp, dmu_tx_t *tx, int flag) + NULL, &links, sizeof (links)); + } + } +- +- value = zfs_dirent(zp, zp->z_mode); +- error = zap_add(ZTOZSB(zp)->z_os, dzp->z_id, dl->dl_name, 8, 1, +- &value, tx); +- +- /* +- * zap_add could fail to add the entry if it exceeds the capacity of the +- * leaf-block and zap_leaf_split() failed to help. +- * The caller of this routine is responsible for failing the transaction +- * which will rollback the SA updates done above. +- */ +- if (error != 0) { +- if (!(flag & ZRENAMING) && !(flag & ZNEW)) +- drop_nlink(ZTOI(zp)); +- mutex_exit(&zp->z_lock); +- return (error); +- } +- + SA_ADD_BULK_ATTR(bulk, count, SA_ZPL_PARENT(zfsvfs), NULL, + &dzp->z_id, sizeof (dzp->z_id)); + SA_ADD_BULK_ATTR(bulk, count, SA_ZPL_FLAGS(zfsvfs), NULL, +@@ -835,6 +813,11 @@ zfs_link_create(zfs_dirlock_t *dl, znode_t *zp, dmu_tx_t *tx, int flag) + ASSERT(error == 0); + mutex_exit(&dzp->z_lock); + ++ value = zfs_dirent(zp, zp->z_mode); ++ error = zap_add(ZTOZSB(zp)->z_os, dzp->z_id, dl->dl_name, ++ 8, 1, &value, tx); ++ ASSERT(error == 0); ++ + return (0); + } + +diff --git a/module/zfs/zfs_vnops.c b/module/zfs/zfs_vnops.c +index 8a7ad702c..6f6ce79db 100644 +--- a/module/zfs/zfs_vnops.c ++++ b/module/zfs/zfs_vnops.c +@@ -1443,22 +1443,10 @@ top: + } + zfs_mknode(dzp, vap, tx, cr, 0, &zp, &acl_ids); + +- error = zfs_link_create(dl, zp, tx, ZNEW); +- if (error != 0) { +- /* +- * Since, we failed to add the directory entry for it, +- * delete the newly created dnode. +- */ +- zfs_znode_delete(zp, tx); +- remove_inode_hash(ZTOI(zp)); +- zfs_acl_ids_free(&acl_ids); +- dmu_tx_commit(tx); +- goto out; +- } +- + if (fuid_dirtied) + zfs_fuid_sync(zfsvfs, tx); + ++ (void) zfs_link_create(dl, zp, tx, ZNEW); + txtype = zfs_log_create_txtype(Z_FILE, vsecp, vap); + if (flag & FIGNORECASE) + txtype |= TX_CI; +@@ -2049,18 +2037,13 @@ top: + */ + zfs_mknode(dzp, vap, tx, cr, 0, &zp, &acl_ids); + ++ if (fuid_dirtied) ++ zfs_fuid_sync(zfsvfs, tx); ++ + /* + * Now put new name in parent dir. + */ +- error = zfs_link_create(dl, zp, tx, ZNEW); +- if (error != 0) { +- zfs_znode_delete(zp, tx); +- remove_inode_hash(ZTOI(zp)); +- goto out; +- } +- +- if (fuid_dirtied) +- zfs_fuid_sync(zfsvfs, tx); ++ (void) zfs_link_create(dl, zp, tx, ZNEW); + + *ipp = ZTOI(zp); + +@@ -2070,7 +2053,6 @@ top: + zfs_log_create(zilog, tx, txtype, dzp, zp, dirname, vsecp, + acl_ids.z_fuidp, vap); + +-out: + zfs_acl_ids_free(&acl_ids); + + dmu_tx_commit(tx); +@@ -2080,14 +2062,10 @@ out: + if (zfsvfs->z_os->os_sync == ZFS_SYNC_ALWAYS) + zil_commit(zilog, 0); + +- if (error != 0) { +- iput(ZTOI(zp)); +- } else { +- zfs_inode_update(dzp); +- zfs_inode_update(zp); +- } ++ zfs_inode_update(dzp); ++ zfs_inode_update(zp); + ZFS_EXIT(zfsvfs); +- return (error); ++ return (0); + } + + /* +@@ -3705,13 +3683,6 @@ top: + VERIFY3U(zfs_link_destroy(tdl, szp, tx, + ZRENAMING, NULL), ==, 0); + } +- } else { +- /* +- * If we had removed the existing target, subsequent +- * call to zfs_link_create() to add back the same entry +- * but, the new dnode (szp) should not fail. +- */ +- ASSERT(tzp == NULL); + } + } + +@@ -3882,18 +3853,14 @@ top: + /* + * Insert the new object into the directory. + */ +- error = zfs_link_create(dl, zp, tx, ZNEW); +- if (error != 0) { +- zfs_znode_delete(zp, tx); +- remove_inode_hash(ZTOI(zp)); +- } else { +- if (flags & FIGNORECASE) +- txtype |= TX_CI; +- zfs_log_symlink(zilog, tx, txtype, dzp, zp, name, link); ++ (void) zfs_link_create(dl, zp, tx, ZNEW); + +- zfs_inode_update(dzp); +- zfs_inode_update(zp); +- } ++ if (flags & FIGNORECASE) ++ txtype |= TX_CI; ++ zfs_log_symlink(zilog, tx, txtype, dzp, zp, name, link); ++ ++ zfs_inode_update(dzp); ++ zfs_inode_update(zp); + + zfs_acl_ids_free(&acl_ids); + +@@ -3901,14 +3868,10 @@ top: + + zfs_dirent_unlock(dl); + +- if (error == 0) { +- *ipp = ZTOI(zp); ++ *ipp = ZTOI(zp); + +- if (zfsvfs->z_os->os_sync == ZFS_SYNC_ALWAYS) +- zil_commit(zilog, 0); +- } else { +- iput(ZTOI(zp)); +- } ++ if (zfsvfs->z_os->os_sync == ZFS_SYNC_ALWAYS) ++ zil_commit(zilog, 0); + + ZFS_EXIT(zfsvfs); + return (error); +diff --git a/tests/runfiles/linux.run b/tests/runfiles/linux.run +index 25ae3fe5e..4b2694202 100644 +--- a/tests/runfiles/linux.run ++++ b/tests/runfiles/linux.run +@@ -55,7 +55,7 @@ tags = ['functional', 'cachefile'] + # 'mixed_none_lookup', 'mixed_none_lookup_ci', 'mixed_none_delete', + # 'mixed_formd_lookup', 'mixed_formd_lookup_ci', 'mixed_formd_delete'] + [tests/functional/casenorm] +-tests = ['case_all_values', 'norm_all_values', 'mixed_create_failure'] ++tests = ['case_all_values', 'norm_all_values'] + tags = ['functional', 'casenorm'] + + [tests/functional/chattr] +diff --git a/tests/zfs-tests/tests/functional/casenorm/mixed_create_failure.ksh b/tests/zfs-tests/tests/functional/casenorm/mixed_create_failure.ksh +deleted file mode 100755 +index 51b5bb3f6..000000000 +--- a/tests/zfs-tests/tests/functional/casenorm/mixed_create_failure.ksh ++++ /dev/null +@@ -1,136 +0,0 @@ +-#!/bin/ksh -p +-# +-# +-# This file and its contents are supplied under the terms of the +-# Common Development and Distribution License ("CDDL"), version 1.0. +-# You may only use this file in accordance with the terms of version +-# 1.0 of the CDDL. +-# +-# A full copy of the text of the CDDL should have accompanied this +-# source. A copy of the CDDL is also available via the Internet at +-# http://www.illumos.org/license/CDDL. +-# +-# +-# Copyright 2018 Nutanix Inc. All rights reserved. +-# +- +-. $STF_SUITE/tests/functional/casenorm/casenorm.kshlib +- +-# DESCRIPTION: +-# For the filesystem with casesensitivity=mixed, normalization=none, +-# when multiple files with the same name (differing only in case) are created, +-# the number of files is limited to what can fit in a fatzap leaf-block. +-# And beyond that, it fails with ENOSPC. +-# +-# Ensure that the create/rename operations fail gracefully and not trigger an +-# ASSERT. +-# +-# STRATEGY: +-# Repeat the below steps for objects: files, directories, symlinks and hardlinks +-# 1. Create objects with same name but varying in case. +-# E.g. 'abcdefghijklmnop', 'Abcdefghijklmnop', 'ABcdefghijklmnop' etc. +-# The create should fail with ENOSPC. +-# 2. Create an object with name 'tmp_obj' and try to rename it to name that we +-# failed to add in step 1 above. +-# This should fail as well. +- +-verify_runnable "global" +- +-function cleanup +-{ +- destroy_testfs +-} +- +-log_onexit cleanup +-log_assert "With mixed mode: ensure create fails with ENOSPC beyond a certain limit" +- +-create_testfs "-o casesensitivity=mixed -o normalization=none" +- +-# Different object types +-obj_type=('file' 'dir' 'symlink' 'hardlink') +- +-# Commands to create different object types +-typeset -A ops +-ops['file']='touch' +-ops['dir']='mkdir' +-ops['symlink']='ln -s' +-ops['hardlink']='ln' +- +-# This function tests the following for a give object type : +-# - Create multiple objects with the same name (varying only in case). +-# Ensure that it eventually fails once the leaf-block limit is exceeded. +-# - Create another object with a different name. And attempt rename it to the +-# name (for which the create had failed in the previous step). +-# This should fail as well. +-# Args : +-# $1 - object type (file/dir/symlink/hardlink) +-# $2 - test directory +-# +-function test_ops +-{ +- typeset obj_type=$1 +- typeset testdir=$2 +- +- target_obj='target-file' +- +- op="${ops[$obj_type]}" +- +- log_note "The op : $op" +- log_note "testdir=$testdir obj_type=$obj_type" +- +- test_path="$testdir/$obj_type" +- mkdir $test_path +- log_note "Created test dir $test_path" +- +- if [[ $obj_type = "symlink" || $obj_type = "hardlink" ]]; then +- touch $test_path/$target_obj +- log_note "Created target: $test_path/$target_obj" +- op="$op $test_path/$target_obj" +- fi +- +- log_note "op : $op" +- names='{a,A}{b,B}{c,C}{d,D}{e,E}{f,F}{g,G}{h,H}{i,I}{j,J}{k,K}{l,L}' +- for name in $names; do +- cmd="$op $test_path/$name" +- out=$($cmd 2>&1) +- ret=$? +- log_note "cmd: $cmd ret: $ret out=$out" +- if (($ret != 0)); then +- if [[ $out = *@(No space left on device)* ]]; then +- save_name="$test_path/$name" +- break; +- else +- log_err "$cmd failed with unexpected error : $out" +- fi +- fi +- done +- +- log_note 'Test rename \"sample_name\" rename' +- TMP_OBJ="$test_path/tmp_obj" +- cmd="$op $TMP_OBJ" +- out=$($cmd 2>&1) +- ret=$? +- if (($ret != 0)); then +- log_err "cmd:$cmd failed out:$out" +- fi +- +- # Now, try to rename the tmp_obj to the name which we failed to add earlier. +- # This should fail as well. +- out=$(mv $TMP_OBJ $save_name 2>&1) +- ret=$? +- if (($ret != 0)); then +- if [[ $out = *@(No space left on device)* ]]; then +- log_note "$cmd failed as expected : $out" +- else +- log_err "$cmd failed with : $out" +- fi +- fi +-} +- +-for obj_type in ${obj_type[*]}; +-do +- log_note "Testing create of $obj_type" +- test_ops $obj_type $TESTDIR +-done +- +-log_pass "Mixed mode FS: Ops on large number of colliding names fail gracefully" +-- +2.14.2 + diff --git a/zfs-patches/0057-Tag-zfs-0.7.8.patch b/zfs-patches/0057-Tag-zfs-0.7.8.patch new file mode 100644 index 0000000..e923c1b --- /dev/null +++ b/zfs-patches/0057-Tag-zfs-0.7.8.patch @@ -0,0 +1,63 @@ +From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001 +From: Tony Hutter +Date: Mon, 9 Apr 2018 14:31:57 -0700 +Subject: [PATCH] Tag zfs-0.7.8 +MIME-Version: 1.0 +Content-Type: text/plain; charset=UTF-8 +Content-Transfer-Encoding: 8bit + +META file and changelog updated. + +Signed-off-by: Tony Hutter +(cherry picked from commit bd69ae3b537ac4e796c53b0c3d6efdc92d652da6) +Signed-off-by: Fabian Grünbichler +--- + rpm/generic/zfs-kmod.spec.in | 3 +++ + rpm/generic/zfs.spec.in | 3 +++ + META | 2 +- + 3 files changed, 7 insertions(+), 1 deletion(-) + +diff --git a/rpm/generic/zfs-kmod.spec.in b/rpm/generic/zfs-kmod.spec.in +index bf52fa2c4..9db8e370a 100644 +--- a/rpm/generic/zfs-kmod.spec.in ++++ b/rpm/generic/zfs-kmod.spec.in +@@ -191,6 +191,9 @@ chmod u+x ${RPM_BUILD_ROOT}%{kmodinstdir_prefix}/*/extra/*/*/* + rm -rf $RPM_BUILD_ROOT + + %changelog ++* Mon Apr 09 2018 Tony Hutter - 0.7.8-1 ++- Released 0.7.8-1, detailed release notes are available at: ++- https://github.com/zfsonlinux/zfs/releases/tag/zfs-0.7.8 + * Wed Mar 14 2018 Tony Hutter - 0.7.7-1 + - Released 0.7.7-1, detailed release notes are available at: + - https://github.com/zfsonlinux/zfs/releases/tag/zfs-0.7.7 +diff --git a/rpm/generic/zfs.spec.in b/rpm/generic/zfs.spec.in +index 7479e987e..769bdab10 100644 +--- a/rpm/generic/zfs.spec.in ++++ b/rpm/generic/zfs.spec.in +@@ -352,6 +352,9 @@ systemctl --system daemon-reload >/dev/null || true + %endif + + %changelog ++* Mon Apr 09 2018 Tony Hutter - 0.7.8-1 ++- Released 0.7.8-1, detailed release notes are available at: ++- https://github.com/zfsonlinux/zfs/releases/tag/zfs-0.7.8 + * Wed Mar 14 2018 Tony Hutter - 0.7.7-1 + - Released 0.7.7-1, detailed release notes are available at: + - https://github.com/zfsonlinux/zfs/releases/tag/zfs-0.7.7 +diff --git a/META b/META +index eefafb10d..615008960 100644 +--- a/META ++++ b/META +@@ -1,7 +1,7 @@ + Meta: 1 + Name: zfs + Branch: 1.0 +-Version: 0.7.7 ++Version: 0.7.8 + Release: 1 + Release-Tags: relext + License: CDDL +-- +2.14.2 + diff --git a/zfs-patches/series b/zfs-patches/series index e248e85..5ae2941 100644 --- a/zfs-patches/series +++ b/zfs-patches/series @@ -25,30 +25,33 @@ 0025-Fix-zdb-R-decompression.patch 0026-Fix-zdb-E-segfault.patch 0027-Fix-zdb-ed-on-objset-for-exported-pool.patch -0028-mmp-should-use-a-fixed-tag-for-spa_config-locks.patch -0029-contrib-initramfs-add-missing-conf.d-zfs.patch -0030-zfs-receive-fails-with-dataset-is-busy.patch -0031-OpenZFS-8857-zio_remove_child-panic-due-to-already-d.patch -0032-Clarify-zinject-8-explanation-of-e.patch -0033-Fix-config-issues-frame-size-and-headers.patch -0034-Correct-count_uberblocks-in-mmp.kshlib.patch -0035-Add-SMART-attributes-for-SSD-and-NVMe.patch -0036-Allow-modprobe-to-fail-when-called-within-systemd.patch -0037-Linux-4.16-compat-use-correct-_dec_and_test.patch -0038-Do-not-initiate-MMP-writes-while-pool-is-suspended.patch -0039-Report-duration-and-error-in-mmp_history-entries.patch -0040-Fix-free-memory-calculation-on-v3.14.patch -0041-Add-scrub-after-resilver-zed-script.patch -0042-Add-SMART-self-test-results-to-zpool-status-c.patch -0043-Fix-zpool-8-list-example-to-match-actual-format.patch -0044-Fix-some-typos.patch -0045-Increment-zil_itx_needcopy_bytes-properly.patch -0046-Change-checksum-IO-delay-ratelimit-values.patch -0047-Linux-4.16-compat-get_disk_and_module.patch -0048-Detect-long-config-lock-acquisition-in-mmp.patch -0049-Take-user-namespaces-into-account-in-policy-checks.patch -0050-zdb-and-inuse-tests-don-t-pass-with-real-disks.patch -0051-Fix-zfs-kmod-builds-when-using-rpm-4.14.patch -0052-Handle-zio_resume-and-mmp-off.patch -0053-Fix-MMP-write-frequency-for-large-pools.patch -0054-Tag-zfs-0.7.7.patch +0028-Handle-zap_add-failures-in-mixed-case-mode.patch +0029-mmp-should-use-a-fixed-tag-for-spa_config-locks.patch +0030-contrib-initramfs-add-missing-conf.d-zfs.patch +0031-zfs-receive-fails-with-dataset-is-busy.patch +0032-OpenZFS-8857-zio_remove_child-panic-due-to-already-d.patch +0033-Clarify-zinject-8-explanation-of-e.patch +0034-Fix-config-issues-frame-size-and-headers.patch +0035-Correct-count_uberblocks-in-mmp.kshlib.patch +0036-Add-SMART-attributes-for-SSD-and-NVMe.patch +0037-Allow-modprobe-to-fail-when-called-within-systemd.patch +0038-Linux-4.16-compat-use-correct-_dec_and_test.patch +0039-Do-not-initiate-MMP-writes-while-pool-is-suspended.patch +0040-Report-duration-and-error-in-mmp_history-entries.patch +0041-Fix-free-memory-calculation-on-v3.14.patch +0042-Add-scrub-after-resilver-zed-script.patch +0043-Add-SMART-self-test-results-to-zpool-status-c.patch +0044-Fix-zpool-8-list-example-to-match-actual-format.patch +0045-Fix-some-typos.patch +0046-Increment-zil_itx_needcopy_bytes-properly.patch +0047-Change-checksum-IO-delay-ratelimit-values.patch +0048-Linux-4.16-compat-get_disk_and_module.patch +0049-Detect-long-config-lock-acquisition-in-mmp.patch +0050-Take-user-namespaces-into-account-in-policy-checks.patch +0051-zdb-and-inuse-tests-don-t-pass-with-real-disks.patch +0052-Fix-zfs-kmod-builds-when-using-rpm-4.14.patch +0053-Handle-zio_resume-and-mmp-off.patch +0054-Fix-MMP-write-frequency-for-large-pools.patch +0055-Tag-zfs-0.7.7.patch +0056-Revert-Handle-zap_add-failures-in-mixed.patch +0057-Tag-zfs-0.7.8.patch -- 2.39.2