Lustre 2.9.0 Changelog

Server support for kernels: 


 * RHEL 7.3 (3.10.0-514.el7)

Client support for unpatched kernels: 


 * RHEL 7.3 (3.10.0-514.el7)
 * SLES12 SP1 (3.12.62-60.64.8)

Recommended e2fsprogs version: 


 * v1.42.13.wc5-7

Interoperability Support: 


 * Clients & Servers: Latest 2.8.X

Issues fixed between 2.8.0 and 2.9.0
New Features 


 * LU-3289: IU Shared Secret Key authentication and encryption
 * LU-3291: IU UID/GID Mapping Feature
 * LU-4931: New feature of giving server/storage side advice of accessing file
 * LU-7335: store default filesystem layout directly on root inode
 * LU-7846: Impose subdirectory mount via nodemap
 * LU-7990: 16MB bulk IO support
 * LU-8289: add ll_decode_linkea tool

Improvements 


 * LU-1118: Push file_update_time into .page_mkwrite
 * LU-1595: force wrap at 70 columns during git commit-msg editing
 * LU-3434: improve git commit hooks for code style checking
 * LU-3888: lfs getstripe should print FID
 * LU-3953: lustre build system improvements
 * LU-4865: osd-zfs: increase object block size dynamically as object grows
 * LU-5505: Fix some problems with current read ahead algorithm
 * LU-5560: SELinux support on the client side
 * LU-5725: Add lu_object cache site_stats for OFD
 * LU-5783: o2iblnd: investigate new memory registration mechanisms
 * LU-5953: lustre[-dkms] needs to automatically account for OFED
 * LU-6391: Option for client not to stop pinging when server suppress_pings enabled
 * LU-6429: add ability to use an external journals for all targets.
 * LU-6527: Journal commit callback optimization
 * LU-6910: Configurable values for OST reserved size
 * LU-7060: Add MGS NID substitution to ldev
 * LU-7134: Ensure ZFS hostid protection if servicenode/failover options given to mkfs.lustre
 * LU-7156: Provide size of changelogs
 * LU-7225: change ladvise wire protocol for lockahead and future usage
 * LU-7227: Add a check to t-f that dmesg doesn't have critical messages
 * LU-7334: Refactor lprocfs helpers
 * LU-7346: Reintroduce tests from SLOW list back to review process
 * LU-7355: osd_trans_start is too noisy on tiny systems
 * LU-7521: LDLM_DEBUG should be used without \n
 * LU-7655: To drop write data on OST to help performance benchmark
 * LU-7689: limit lu_site hash table size on clients
 * LU-7704: LOOP_CTL_GET_FREE should be checked against target kernel
 * LU-7712: migration is too noisy in case of error
 * LU-7733: ptlrpc_lprocfs_svc_req_history_show (Seq_printf) time resolution should be higher than 1 second.
 * LU-7795: least_qunit should be tunnable
 * LU-7835: Improved gerrit/checkpatch.pl integration
 * LU-7896: lu_object_limit is called too frequently
 * LU-7898: remove unnecessary declarations from osd-zfs
 * LU-7900: lu_object_assign_fid doesn't need to lookup FID
 * LU-7901: osd-zfs doesn't need to allocate bulks for SA updates
 * LU-7902: osd_attr_set can be turned into a noop in some cases
 * LU-7904: osd_fid_lookup should honor LOC_F_NEW
 * LU-7905: osd-zfs should pin OI objects
 * LU-7906: osd-zfs should not set blocksize on MDT
 * LU-7908: fake object precreation for benchmarking purpose
 * LU-7918: remove unnecessary declarations in MDD and TGT
 * LU-7931: Allow per-subtest setup/cleanup logic
 * LU-7968: "build on master taking too long
 * LU-8055: Further ldev improvements
 * LU-8056: Support for linux 4.5 kernels
 * LU-8058: Remove #if 0 sections older than 4 years
 * LU-8068: Large ZFS Dnode support
 * LU-8102: Correlate ptlrpc request with AST error
 * LU-8157: Functionnal testing for layout swaps
 * LU-8159: cache xattr in ldiskfs OSD
 * LU-8176: Add file_dentry and file_inode wrappers
 * LU-8251: sanityn test_40* - test_47* : Delete only test files
 * LU-8339: Lock improvement for kiblnd_check_sends
 * LU-8549: optimize restore_and_check_size in sanity-hsm
 * LU-8599: Restore lshowmount utility
 * LU-8619: Lustre to build against ZFS 0.7.0
 * LU-8648: Remove all reference to Sun Inc in our source code.
 * LU-8655: customize run_mdtest
 * LU-8775: ZFS OST has an extra 128 byte for each IO

Bugs 


 * LU-28: Ability for client to mount subdirectories of a Lustre filesystem
 * LU-522: "sanity.sh test_27x failed with ""OST0 was degraded but new created file still use it"""
 * LU-815: "BUG: unable to handle kernel NULL pointer dereference"" in lprocfs_rd_import"
 * LU-937: CLIENTONLY testing no longer works
 * LU-1482: attribute not permitted on Lustre versions 2.x
 * LU-1882: conf-sanity test_45: umount2: Device or resource busy
 * LU-2547: "test: recovery-small test_24a
 * LU-2766: lov_object.c:635:lov_layout_change) ASSERTION( atomic_read(&lov->lo_active_ios) == 0 ) failed
 * LU-3782: Divizion by zero in ost-pools 18
 * LU-4039: "Failure on test suite replay-single test_90: wrong stripe: f0
 * LU-4182: sanity-scrub test_15: FAIL: (10) Failed to start OI scrub on mds1
 * LU-4185: Incorrect permission handling when creating existing directories at ICHEC
 * LU-4257: parallel dds are slower than serial dds
 * LU-4433: mds-survey test_1: FAIL: OST lustre-MDT0001 not setup
 * LU-4444: conf-sanity test_69: ZFS took too long to create 100k files
 * LU-4474: deadlock of ldiskfs_quota_off
 * LU-4739: Code inconsistencies when referring to config_llog_instance
 * LU-4781: lustre-rsync-test test_2b: Replication of operation failed(-17)
 * LU-4825: lfs migrate not freeing space on OST
 * LU-5050: cpu partitioning oddities
 * LU-5282: (mdc_request.c:791:mdc_free_open) ASSERTION( mod->mod_open_req->rq_replay == 0 ) failed
 * LU-5473: Test failure sanity test_51b: test_51b failed: fnum: No space left on device
 * LU-5564: open handle leak after setuid
 * LU-5614: use %kernel_module_package for weak-updates
 * LU-5704: lfs getstripe hangs on fifo files
 * LU-5814: encapsulate lov_stripe_md (LSM) to LOV layer
 * LU-5954: IOC_MDC_GETFILEINFO returns the wrong ino
 * LU-6083: IB with Ubuntu 14.04 client
 * LU-6119: Add testing for Lustre exported via Samba
 * LU-6123: conf-sanity test_72: FAIL: mount client failed
 * LU-6163: 64-bit divides should use do_div
 * LU-6284: FLD read is not swabbed correctly
 * LU-6285: Assert fails in staging client module crashes kernel if CPUMASK_OFFSTACK set
 * LU-6303: comparison between signed and unsigned integer expressions
 * LU-6304: crash on umount in cleanup_resource
 * LU-6441: "OST problems following router node crash
 * LU-6447: mdt_identity_upcall calls sleeping function under rwlock
 * LU-6512: ll_dir_ioctl has some pretty strange code now.
 * LU-6518: Unhandled possible allocation failure in lustre_start_mgc
 * LU-6523: Unchecked usage of potential ERR_PTR result in lmv_hsm_req_count and lmv_hsm_req_build
 * LU-6601: deactivated OSTs do not appear to be inactive according to 'lctl dl'
 * LU-6622: lnet-selftest test_smoke: @@@@@@ FAIL: lst Error found
 * LU-6635: sanity-lfsck test_18e:FAIL: (8) .lustre/lost+found/MDT0000/ should not be empty
 * LU-6638: sanity-hsm test_37:request on 0x200000401:0x2d:0x0 is not SUCCEED on mds1
 * LU-6688: sanityn 77a and 77b fail to set the NRS policy
 * LU-6696: "ASSERTION( rc == 0 || rc == LLOG_PROC_BREAK ) failed: 0 changes
 * LU-6722: sanity-lfsck test_1a: FAIL: (3) Fail to start LFSCK for namespace!
 * LU-6754: wiretest / wirecheck don't match up
 * LU-6756: sanity test_225b: test failed to respond and timed out
 * LU-6808: Interop 2.5.3<->master sanity test_224c: Bulk IO write error
 * LU-6820: osd_handler.c:2280:osd_declare_object_destroy) ASSERTION( inode ) failed:
 * LU-6844: replay-single test 70b failure: 'rundbench load on * failed!'
 * LU-6861: sanity-scrub test 4a and 4b fail: Auto trigger full scrub unexpectedly
 * LU-6888: osp_update_request_create should check return value
 * LU-6937: sanity-hsm test_106 test_403: error opening lhsmtool_posix: No such file or directory
 * LU-6966: filesystem name with dash causes test-framework.sh warnings
 * LU-6971: "Remove obsolete remote client and ""lfs cp|ls|getacl|setacl"" commands"
 * LU-7002: mkfs.lustre mishandles SELinux context of temporary directory
 * LU-7003: append_context_for_mount must quote the value of the context option
 * LU-7024: 2.5.3<->2.7.58 interop: sanity-sec failed with lctl nodemap_modify
 * LU-7027: Suspicios dereference in class_config_llog_handler
 * LU-7038: obdfilter-survey test_3a: (lu_object.c:1224:lu_device_fini) ASSERTION( atomic_read(&d->ld_ref) == 0 ) failed: Refcount is 3
 * LU-7044: Interop 2.5.3<->master sanityn test_77e: failed to set TBF policy
 * LU-7055: osd_ost_fini calls sleeping function under rwlock
 * LU-7061: BUG: unable to handle kernel NULL pointer dereference at 0000000000000004 IP: osd_scrub_refresh_mapping+0x39d/0x410
 * LU-7064: llog_cancel_rec does not detect errors from llog_declare_destroy
 * LU-7084: sanityn test_77c: cfs_hash_find_or_add) ASSERTION( hlist_unhashed(hnode) ) failed
 * LU-7099: Crash in kiblnd_pool_alloc_node
 * LU-7101: Lnet: Support per NI map-on-demand
 * LU-7110: sanity-hsm.sh can erase all files on filesystem
 * LU-7111: misleading error message in mgs_write_log_param
 * LU-7117: replay-single test_70d: timeout and mkdir/rmdir stopped
 * LU-7121: Extra items shown in the stack trace in t-f
 * LU-7124: MLX5: Limit hit in cap.max_send_wr
 * LU-7130: sanity.sh test 162 does not properly handle some DIR values
 * LU-7144: Interop 2.7.0<->master- sanity-scrub test_14: (6) Some entry under /lost+found should be repaired
 * LU-7149: Test framework doesn't cleanup so test always passed when executed consecutively
 * LU-7151: sanity test_205:No jobstats for id.205.dd.3 found
 * LU-7152: sanity-hsm test_24a:FAIL: restore changed ctime from 1442224769 to 1442224774
 * LU-7160: Saved by change_nids .bak files on MGS should never be processed
 * LU-7185: restore flags on ptlrpc_connect_import failure to prevent LBUG
 * LU-7198: vvp_io.c:701:vvp_io_fault_start) binary changed while waiting for the page fault lock
 * LU-7204: replay-single test_85b: syntax error : unary operator expected
 * LU-7206: LBUG osp_sync.c:1541:osp_sync_id_traction_fini) ASSERTION( list_empty(&tr->otr_wakeup_list) ) failed:
 * LU-7220: Resource acquired to fd may be lost
 * LU-7234: recovery-small 105 test failed recovery-small test_105: @@@@@@ FAIL: MGS IR state must be partial
 * LU-7256: sanity-lfsck TIMEOUT on umount /mnt/mds4
 * LU-7258: Bit field 'oi_lockless' has one bit and is signed
 * LU-7259: Bit field 'ot_remove_agents' has one bit and is signed
 * LU-7301: change test-framework to exclude the cleanup between the tests suites
 * LU-7302: sanity-lfsck test_23b hangs on START_NAMESPACE
 * LU-7310: sanityn test_39a fails with 'mtime is not updated on write: 1444890978 <= 1413354977 <= 1444890979'
 * LU-7311: Linux Kernel 3.18 LTS support
 * LU-7317: error function not called upon error
 * LU-7323: "ldiskfs: Unknown symbol mb_cache_create (err 0)"" on RHEL7"
 * LU-7352: conf-sanity test_78: no space left on device
 * LU-7402: double free of request in ll_atomic_open and ll_lookup_nd
 * LU-7403: lookup_intent handling is obscure
 * LU-7408: multislot RPC support didn't declare write for reply_data object
 * LU-7418: sanity test_29 failed with 'CURRENT: 7 > 6'
 * LU-7422: incorrect ENOENT handling in mdt_intent_reint causes panic on client side
 * LU-7428: "conf-sanity test_84
 * LU-7429: sanity-lfsck test_23c: @@@@@@ FAIL: (8) unexpected size
 * LU-7433: xattr locks are lost on mdt
 * LU-7434: lost bulk leads to a hang
 * LU-7435: ost-pools defect: suite is not functional on config MGS/MDS not combined
 * LU-7442: conf-sanity test_41c: @@@@@@ FAIL: unexpected concurent MDT mounts rc=17 rc2=0
 * LU-7445: sanity test_51ba defect: test depends on test_51b and always skipped if run separately
 * LU-7469: lustre-rsync-test test_1a: FAIL: Dir create not replicated : Error writing to target path /tmp/target
 * LU-7472: ior tests are to be improved to operate with smaller blockSize
 * LU-7483: sanity test_103a: run_acl_subtest cp failed
 * LU-7486: Inconsistent state information in health_check
 * LU-7501: "inconsistencies between ""lfs find""
 * LU-7524: sanity-hsm_404 test failed: >LustreError: 7253:0:(client.c:771:__ptlrpc_request_alloc) ASSERTION( (unsigned long)imp > 0x1000 ) failed: (null)
 * LU-7528: sanity test_160b test_160c tests defect: changelog user is not deregistered on test failure
 * LU-7536: racer_1 test: mds crash : BUG: unable to handle kernel NULL pointer dereference at (null) Oops: 0010 [#1] SMP
 * LU-7542: possible recursive locking detected on cli->cl_sem
 * LU-7558: niobuf.c:721:ptl_send_rpc LASSERT(AT_OFF || imp_state != LUSTRE_IMP_FULL || imp_msghdr_flags & MSGHDR_AT_SUPPORT ...)
 * LU-7593: umount vs tgt_last_rcvd_update deadlock
 * LU-7608: kernel upgrade [SLES12 SP1 3.12.51-60.25]
 * LU-7613: "racer crash on lustre nfs mount
 * LU-7619: USE_LU_REF/lu_ref feature broken after some REFASSERTs have been added without lu_ref::lf_guard protection
 * LU-7623: Clean up lustre user/kernel pointer types.
 * LU-7626: ldlm_resource_complain) echotgt-echo_srv_UUID: namespace resource [0x1:0x0:0x0].0 (ffff88005e7ade00) refcount nonzero (1) after lock cleanup; forcing cleanup.
 * LU-7627: "Lustre sanity tests 130b
 * LU-7646: Infinite CON RACE Condition after rebooting LNet router
 * LU-7656: replay-single_70c test failed tar: Exiting with failure status due to previous errors
 * LU-7660: FS default striping settings only honored on MDT 0
 * LU-7664: set_conf_param_and_check defect : it should wait update on facet_active_host instead of facet_host
 * LU-7667: conf-sanity test_52 fails due to $TMP/files already existing
 * LU-7669: lmv_connect leaks export if lmv_connect_mdc fails
 * LU-7674: remove unused placement parameter from lmv
 * LU-7702: ASSERTION( lock->l_export == opd->opd_exp ) failed
 * LU-7707: general protection fault in qsd_reint_main
 * LU-7718: lfs getstripe does not work on bind mount
 * LU-7727: open with FMODE_EXEC fails with ETXTBSY after a failed FMODE_WRITE open attempt on a read only client
 * LU-7728: soft lockup in osp_precreate_reserve
 * LU-7732: "check_for_next_transno) lustre-MDT0000: waking for gap in transno
 * LU-7736: lustre_rmmod does not remove all the Lustre modules
 * LU-7747: "sanity test_56w: dataversion changed during copy
 * LU-7756: oss_num_threads max value is sometimes too low to feed disk controllers
 * LU-7759: umount hanging in modern distros when OST is unavailable
 * LU-7768: Do not keep retrieving the FLD entries if MDT does not support
 * LU-7772: catalogs shouldn't destroy plain llogs twice
 * LU-7776: lustre-single lnet-selftest test failed
 * LU-7778: mount of MDT(==MGS) failed after MDS restart
 * LU-7781: kernel update [RHEL7.2 3.10.0-327.10.1.el7]
 * LU-7782: sanity-scrub test_2: NULL pointer dereference at 0x10 in lu_context_key_get on mds2
 * LU-7786: "improve racer cleanup
 * LU-7796: """lctl set_param jobid_var"" should return error about missing value"
 * LU-7799: mdt.*.hsm.actions skips some records
 * LU-7800: Panic during recovery of soak-test.
 * LU-7803: sanity test 78 failures in interop
 * LU-7805: ll_orphan_dentry_list is unused and should be removed
 * LU-7809: general protection fault: 0000 during failback of MDS disk resources
 * LU-7813: default pool not inherited when specifying any layout parameter
 * LU-7815: pinger thread evict the export between MDTs.
 * LU-7818: kernel update [SLES12 SP1 3.12.53-60.30]
 * LU-7828: Traversal directory is very slow
 * LU-7829: Glimpse lock shouldn't match not granted locks
 * LU-7830: evict during truncate
 * LU-7836: MDSes crashed with oom-killer
 * LU-7841: Lutsre2.8 fails to build on SLES11 SP4 due to unavailability of rst2man/python-docutils
 * LU-7848: Recovery process on MDS stalled
 * LU-7850: gnilnd Fix niov calculation with offset kiov
 * LU-7851: ZFS rolling upgrade: cannot mount MDS: mdt: lustre-MDT0000 unknown param som=disabled
 * LU-7859: User info should not be pulled directly from /etc/passwd in lustre/tests
 * LU-7860: LustreError: 19445:0:(ldlm_lock.c:2273:ldlm_lock_cancel) ASSERTION( !(((( lock))->l_flags & (1ULL << 53)) != 0) ) failed
 * LU-7861: MDS Contention during unlinks due to llog spinlock
 * LU-7883: sanity test_60a invokes oom-killer in subtest 10b and times out
 * LU-7887: lbuild not building lnetctl
 * LU-7888: kernel: INFO: task mount.lustre:22219 blocked for more than 120 seconds
 * LU-7890: EFBIG on 2TB singly stripped file with ZFS
 * LU-7893: osd-zfs: calls dmu_objset_disown with NULL objset_t *
 * LU-7903: recovery-small test_23: hang on umount
 * LU-7907: kernel update [RHEL6.7 2.6.32-573.22.1.el6]
 * LU-7914: cl_page_list_own is unused and could be removed
 * LU-7919: "Buffer overflow in mount_lustre: parse_ldd
 * LU-7920: hsm coordinator request_count and max_requests not used consistently
 * LU-7922: "ROOT dir created at mkfs time is using a high #d inode
 * LU-7925: ll_d_iput can clear i_nlink for an inode in use
 * LU-7927: Deadlock between ll_setattr and ll_file_write->ll_fsync
 * LU-7928: Wrong params for mdd_declare_rename at mdd_rename
 * LU-7936: Resource acquired to 'read_fd' at line 500 may be lost here.
 * LU-7937: Result of malloc that may return NULL will be dereferenced
 * LU-7939: ldiskfs_find_entry returns ERR_PTR but checked for NULL
 * LU-7949: Suspicious dereference of pointer before NULL check in osd_declare_qid
 * LU-7950: Suspicious dereference of pointer before NULL check in lwp_device_fini
 * LU-7955: Add ability to set gnilnd bte_get/put_dlvr_mode
 * LU-7957: kernel update [SLES11 SP4 3.0.101-71]
 * LU-7961: Problem with kernel location detection on CentOS 6
 * LU-7962: Support build against weak-module ZFS install
 * LU-7965: mkfs.lustre sometimes fails with a segmentation fault
 * LU-7967: kernel update [RHEL7.2 3.10.0-327.13.1.el7]
 * LU-7972: $REFORMAT in t-f is not used correctly
 * LU-7973: Lustre client crash in __d_lookup - BUG: unable to handle kernel paging request
 * LU-7975: """(lod_object.c:700:lod_ah_init) ASSERTION( lc->ldo_stripenr == 0 )"" LBUG/Assert on MDS"
 * LU-7977: Change MB/s to MiB/s in lnet_selftest reporting
 * LU-7979: dereferencing ERR PTR in kiblnd_alloc_freg_pool
 * LU-7981: double read of lli_trunc_sem in ll_page_mkwrite and vvp_io_fault_start leads to deadlock
 * LU-7986: "HSM: Duplicated cookies in ""actions"""
 * LU-7992: HSM: mutual deadlock when killing a copytool
 * LU-7993: sanity test_116a: division by 0
 * LU-7997: RCU stalls waiting for lu_sites_guard mutex in lu_cache_shrink_count
 * LU-7998: Resource acquired to fd may be lost in mount_utils.c
 * LU-7999: Object 'uuids' was freed at line 1895 after being freed by calling 'realloc' at line 1892
 * LU-8001: Null pointer dereference in nm_member_reclassify_nodemap
 * LU-8003: ll_ddelete has obsolete reference to find_cbdata
 * LU-8005: LBUG on osc_req_attr_set
 * LU-8009: Lustre_rsync does not sync files on MDT1
 * LU-8010: lfs hsm command hangs up after lfs hsm_cancel
 * LU-8014: potential memory dereference in kuc_ispayload
 * LU-8015: Replication of operation failed(-22): 2885 HLINK (3)
 * LU-8017: "All Nodes report NOT HEALTHY
 * LU-8018: lov_init_raid0 ASSERTION( subdev != NULL ) failed: not init ost 0
 * LU-8019: Openlock breakage
 * LU-8021: interop: 2.1.x server <-> clients version > 2.3: t-f debugsave debugrestore defect
 * LU-8022: LNet: BUG: unable to handle kernel NULL pointer dereference
 * LU-8023: build problem on master due to recent change
 * LU-8024: kernel update [SLES12 SP1 3.12.57-60.35]
 * LU-8025: "ll_direct_io code introduced in LU-6260 is based on 2.4/2.5
 * LU-8027: Ensure OSC OBD has been set up in onu_upcall
 * LU-8030: hsm: inserting duplicate requests
 * LU-8035: finish renaming MDS_GETSTATUS to MDS_GET_ROOT
 * LU-8036: lustreapi.h should not define ARRAY_SIZE
 * LU-8037: subtree mounts should reject '..' as a path component
 * LU-8038: lfs migrate fails with --block option
 * LU-8040: changelog_mask modifications from default do not persist across MDT restart
 * LU-8041: Close should update atime increasing only
 * LU-8044: class_process_config no device for: lustre-MDT0021-mdtlov
 * LU-8048: LBUG during cfs_hash_destroy
 * LU-8049: Do not use LUSTRE_MSG_MAGIC to check local endianness
 * LU-8051: "sanity tests 300[a-i] failed
 * LU-8052: test-framework change breaks test
 * LU-8057: o2iblnd driver is causing memory corruption due to improper handling of scatter list.
 * LU-8059: wrong ext4 patch in LU-7132
 * LU-8064: Minor fix in param_display
 * LU-8065: lctl changelog_{de}register help message is unclear
 * LU-8071: lvcreate --snapshot of MDT hangs in ldiskfs_journal_start_sb
 * LU-8078: plot-obdfilter doesn't parse summaries correctly.
 * LU-8079: Remove unneeded wrapper llog_cat_init_and_process
 * LU-8080: Replace calls to signal with sigaction in multithreaded environments
 * LU-8081: Lustre 2.8.0
 * LU-8083: namespace LFSCK needs to handle symbol file properly
 * LU-8084: namespace LFSCK needs to handle linkea record length properly
 * LU-8085: Inode Iteration should increase the cursor to skip unused inodes
 * LU-8089: MGT/MDT mount fails on secondary HA node
 * LU-8091: server_lsi2mti should return error when fail to get nid
 * LU-8093: kernel update [RHEL6.7 2.6.32-573.26.1.el6]
 * LU-8094: floods of log messages seen in syslog
 * LU-8097: Do not assert for missing lock for resent
 * LU-8098: Add lfs mdts to the man page
 * LU-8100: Missing MDTs in /proc/fs/lustre/lmv/lustre-clilmv-.../target_obd
 * LU-8106: kiblnd_pool_alloc_node crashed because of NULL pointer dereference
 * LU-8117: lustre-ppc fails to build on ppc64 el7
 * LU-8124: MDT zpool capacity consumed at greater rate than inode allocation
 * LU-8125: "ASSERTION( strncmp(infos[pol_idx].pi_arg
 * LU-8126: new kernel [RHEL6.8 2.6.32-642.el6]
 * LU-8127: Build failure after master new tag 2.8.53
 * LU-8129: sanity test_102n: before and after xattrs of '/mnt/lustre/f102n.sanity.1' differ
 * LU-8132: tests: llmount.sh does not use fs-specific paths for the ost and mdt it mounts
 * LU-8134: kernel update [RHEL7.2 3.10.0-327.18.2.el7]
 * LU-8135: sanity test_101g fails with 'not all RPCs are 16 MiB BRW rpcs'
 * LU-8136: sanity-hsm test_9 fails with 'request on 0x200000405:0x4:0x0 is not SUCCEED on mds1'
 * LU-8138: newwiretest build failures
 * LU-8141: refresh acl test fix
 * LU-8147: sanity test_208: zfs softlockup in mount error path
 * LU-8150: Track creates in MDS stats
 * LU-8155: sanity-hsm test_16: Transfer is too fast 19 < 20
 * LU-8163: hsm: user_request_mask can't be unset in test 24c
 * LU-8165: (tgt_lastrcvd.c:656:tgt_client_del) lustre-OST0000: client 4294967295: bit already clear in bitmap!!
 * LU-8171: stack overrun in hai_dump_data_field
 * LU-8174: Taking a lease for the 2nd time fails with -EBUSY
 * LU-8175: conflicting PW & PR extent locks on a client
 * LU-8178: recovery_duration value becomes negative
 * LU-8179: libcfs/util/ioctl.h unconditionally redefines __user
 * LU-8182: Suspicious code in mdt_obd_connect
 * LU-8186: Typo in ll_rw_extents_stats_pp_seq_show
 * LU-8189: osc_match_base prototype differs from declaration
 * LU-8193: request mbits isn't set properly for EINPROGRESS resend
 * LU-8196: class_handle_init alloates 1.5M
 * LU-8197: early reply causes replay request deadline decrease
 * LU-8199: NULL pointer dereference in tgt_free_reply_data
 * LU-8203: sanity-hsm test_60 failed as we do not wait for copytool operation to complete
 * LU-8209: glimpse lock request does not engage ELC to drop unneeded locks
 * LU-8210: OST Read Cache does not work in Centos7 platform
 * LU-8216: Quota updates are not properly journaled
 * LU-8218: lfsck not able to recover files lost from MDT
 * LU-8220: "parallel-scale-cifs test iozone:fsync: No space left on device
 * LU-8226: t-f check_catastrophe defect
 * LU-8227: kernel update [RHEL6.8 2.6.32-642.1.1.el6]
 * LU-8232: Dead code in obdclass
 * LU-8239: llanalyze logs parser is broken
 * LU-8241: lustre builds with MOFED 3.3 fail
 * LU-8246: Leaks on ldlm granted locks counter on MDS leading to canceling loop
 * LU-8248: sanity test_248: fast read was not 4 times faster
 * LU-8249: Potential deadlock in lnet
 * LU-8250: MDT recovery stalled on secondary node
 * LU-8257: Update error if /etc/mtab is a symlink
 * LU-8258: Unsafe userspace address access from proc handler in nodemap code
 * LU-8259: ldiskfs osd proc interface lacks __user qualifiers
 * LU-8270: List corruption with Shared Secret Key enabled
 * LU-8277: missing commas in checkpatch.pl
 * LU-8279: sanity-scrub test_4b: @@@@@@ FAIL: Error in dmesg detected
 * LU-8284: i_size updates from BRW writes are not atomic
 * LU-8285: Autotest local.sh incorrectly turns off LNet logging
 * LU-8287: "sanity-sec test_16: mgs and c0 idmap mismatch
 * LU-8297: mount with missing key leads to leaked MGC device and LBUG
 * LU-8298: enabling gss causes possibly misleading warnings from tgt_init_sec_level
 * LU-8300: unable create .lustre/lost+found/MDTxxxx if lost oi.16.xxx
 * LU-8301: namespace LFSCK cannot handle ROOT fid properly
 * LU-8303: Melanox / O2ib lnd cause a OOM on OST node
 * LU-8305: sanity-sec test_27: @@@@@@ FAIL: fileset not cleared on nodemap c0
 * LU-8311: Target does not mount with the new mgsnode parameter format in case of multirail configuration
 * LU-8316: BUG: unable to handle kernel NULL pointer dereference at tgt_free_reply_data+0x97/0x330
 * LU-8319: ASSERTION( inode == parent_dir ) failed:
 * LU-8320: :(llog_osd.c:338:llog_osd_write_rec) ASSERTION( llh ) failed:
 * LU-8321: kernel update [RHEL7.2 3.10.0-327.22.2.el7]
 * LU-8323: lustre-initialization-1: fails for older servers and master clients
 * LU-8327: conf_sanity test_61: soft lockup
 * LU-8328: conf-sanity test_87 test failed to respond and timed out
 * LU-8331: kernel update [SLES11 SP4 3.0.101-77]
 * LU-8337: kernel update [SLES12 SP1 3.12.59-60.41]
 * LU-8340: sanity-sec test_25: /usr/bin/lfs setquota -u quota_usr -b 13761540 -B 14449617 -i 916168 -I 961976 /mnt/lustre FAILED!
 * LU-8343: build failures on SLES
 * LU-8347: granting conflicting locks
 * LU-8348: Don't send an last_commited after journal abort to ability a replay uncommitted data
 * LU-8349: MDS crash ldlm_flock.c:849:ldlm_export_flock_put) ASSERTION( flock->blocking_export != ((void *)0) ) failed
 * LU-8352: Mellanox/OFED build breakage due to file_inode macro.
 * LU-8353: mdt unlink should lock parent before lookup child
 * LU-8354: soft lockup in ldlm_plain_compat_queue
 * LU-8356: umount hang/stonith on mds failback
 * LU-8357: sanity-sec LBUG on MDS umount with ASSERTION( exp->u.eu_target_data.ted_nodemap == nodemap )
 * LU-8360: obd_set_info_async error suppressed
 * LU-8361: lctl lfsck_start --all does not start lfsck on all devices
 * LU-8368: Use kgnilnd_vzalloc for copy buffer allocation
 * LU-8370: ASSERTION( lur->lur_hdr.lrh_len <= ctxt->loc_chunk_size )
 * LU-8371: Optimize open of known negative dentry
 * LU-8372: Suspicious dereference of pointer 'vma->vm_file' in ll_vm_{open|close}
 * LU-8377: lustre fail to build on el7
 * LU-8378: unused but set variables
 * LU-8383: Wrong tests package dependencies
 * LU-8386: Lustre master does not build with zfs master because ofåÊ'ZFS_MAXNAMELEN' undeclared
 * LU-8387: has_warning check in sanity.sh:test_129 always return true
 * LU-8389: kernel update [RHEL6.8 2.6.32-642.3.1.el6]
 * LU-8394: parallel-scale-nfsv3: FAIL: cleanup_nfs failed
 * LU-8399: MDT hung at lu_object_find_at during umount
 * LU-8401: modprobe: ERROR: could not insert 'lnet': Unknown symbol in module
 * LU-8407: Recovery timer hangs at zero on DNE MDTs
 * LU-8408: mgc_request.c:141:config_log_put) ASSERTION( atomic_read(&cld->cld_refcount) > 0 )
 * LU-8410: fiemap vs walk race
 * LU-8413: sanity test_101f fails with 'misses too much pages!'
 * LU-8415: fix t-f to customise MPIRUN (for mpiexec.hydra)
 * LU-8418: node fails to kdump after lbug crash
 * LU-8426: "rpmbuild build fails with option ""--without lustre_utils"""
 * LU-8427: Need clear file lustre-tests.files before add to it
 * LU-8429: Add option for gnilnd to not reconnect after connection timeout
 * LU-8436: sanity-sec test_14: FAIL: nodemap_add failed with 1
 * LU-8443: Can't create Lustre FS with enabled meta_bg option
 * LU-8446: metadata-updates: FAIL: wrong timestamps
 * LU-8450: replay-single test 70c: mount MDS hung
 * LU-8454: non-root user is able to change stripe size for mount point
 * LU-8460: (osc_cache.c:658:osc_extent_find) ASSERTION( (max_pages & ~chunk_mask) == 0 ) failed: LBUG
 * LU-8468: kernel update [RHEL7.2 3.10.0-327.28.2.el7]
 * LU-8471: missing EXPORT_SYMBOL for several lu_ref_* functions
 * LU-8472: sanity-scrub test_5 times out
 * LU-8473: conf-sanity test_41a with separate MGS stuck on starting client and timed out
 * LU-8474: conf-sanity test_55 with separate MGS stuck on starting client and timed out
 * LU-8479: Reserve 0x220x for OBD_FAIL_* macros
 * LU-8480: Server syslog: ofd_grant.c:183:ofd_grant_sanity_check) ofd_obd_disconnect: tot_granted 69347328 != fo_tot_granted 102901760
 * LU-8492: ptlrpc: Correctly calculate hrp->hrp_nthrs
 * LU-8493: Do not set stale flag for new created OSP object
 * LU-8495: kernel update [SLES11 SP4 3.0.101-80]
 * LU-8495: kernel update [SLES11 SP4 3.0.101-80]
 * LU-8498: configuration from log 'nodemap' failed (-22)
 * LU-8499: wireshark plugins don't build any more
 * LU-8500: lustre-2.7.2 hits MGS is waiting for obd_unlinked_exports more than 256 seconds. The obd refcount = 5. Is it stuck?
 * LU-8501: DLC: Routing setting using DLC failed the first time
 * LU-8507: DLC: Enable peer_credit management
 * LU-8508: kernel:LustreError: 3842:0:(lu_object.c:1243:lu_device_fini) ASSERTION( atomic_read(&d->ld_ref) == 0 ) failed: Refcount is 1
 * LU-8509: drop_caches hangs in cl_inode_fini
 * LU-8510: ASSERTION( dt->do_ops->do_invalidate ) failed
 * LU-8513: kernel update [RHEL7.2 3.10.0-327.28.3.el7]
 * LU-8514: tgt_main.c:121:tgt_cancel_slc_locks) ASSERTION( lock->l_client_cookie != 0 ) failed
 * LU-8518: kernel update [SLES12 SP1 3.12.62-60.62]
 * LU-8519: build is broken on sles12
 * LU-8523: sanity test_311: objs not destroyed after unlink
 * LU-8524: sanity-selinux: FAIL: Accepting only targeted policy
 * LU-8527: Lustre 2.8 server crashed on bring up during large scale test shot
 * LU-8529: AC_MSG_RESULT unhelpful when ldiskfs patch series not identified
 * LU-8534: RHEL7.3 ldiskfs patch set
 * LU-8535: RR policy should not allocate on same ost
 * LU-8537: kernel update [RHEL6.8 2.6.32-642.4.2.el6]
 * LU-8540: ib_map_mr_sg with five arguments
 * LU-8544: recovery-double-scale test_pairwise_fail: start client on trevis-54vm5 failed
 * LU-8547: Interop b2_8<->master: ost-pools test_24: Pool name not inherited in /mnt/lustre/d24.ost-pools/dir3/f24.ost-pools0(test_85b)
 * LU-8551: Use mds1 rather than mds to operate on MDT0000
 * LU-8559: sanity test_56xa: cannot swap layouts: Operation not permitted
 * LU-8561: Redundant check in osd_mount
 * LU-8565: "sanity test 255a fails with <89>Û÷Speedup with willread is less than X%
 * LU-8573: IOR: niobuf.c:319:ptlrpc_register_bulk) ASSERTION( desc->bd_md_count == 0 ) failed
 * LU-8574: The logic of verifying FID-in-dirent may handle the dangling or corrupted name entry improperly
 * LU-8579: "oti_dev is never set
 * LU-8580: general protection fault: osd_xattr_get+0x32c/0x5b0 [osd_ldiskfs]
 * LU-8581: Kernel Panic - osd_oxc_lookup+0x38/0x70 [osd_ldiskfs]
 * LU-8590: Fix issues with SK privacy and integrity mode
 * LU-8592: MDS crashed with ASSERTION( atomic_read(&o->lo_header->loh_ref) > 0 )
 * LU-8600: sanity test_399:fake write is slower
 * LU-8608: "Rolling upgrade between 2.8.x and master failed: Upon upgrading OSS
 * LU-8612: Sync-on-Lock-Cancel locks should be handled per MDT
 * LU-8615: kernel update [RHEL7.2 3.10.0-327.36.1.el7]
 * LU-8625: replay-single test_87 defect: test_87 is to be renamed to test_87a
 * LU-8629: Error exit path doing wrong cleanup in cl_env_percpu_init
 * LU-8630: Install of lustre-dkms fails
 * LU-8631: lfs quota should return useful error on non-lustre file systems
 * LU-8632: Address of function 'page_count' is used instead of a local variable
 * LU-8634: 2.8.0 MDS (layout.c:2025:__req_capsule_get) @@@ Wrong buffer for field `quota_body' (3 of 1) in format `LDLM_INTENT_QUOTA': 0 vs. 112 (server)
 * LU-8636: lbuild: sharing lbuild cache is not working
 * LU-8637: cont-sanity test_71c failed: class_export_put+0x18/0x310 [obdclass]
 * LU-8645: Update imp_known_replied_xid for resend-replay
 * LU-8647: lfsck_namespace_double_scan) ASSERTION( list_empty(&lad->lad_req_list) ) failed
 * LU-8654: obd_connect_data handling must account for clients that send obd_connect_data_v1
 * LU-8658: Suppress ptlrpc error message for LDLM_ENQUEUE with -EAGAIN
 * LU-8660: Don't ignore server_make_name return value
 * LU-8667: Validate sequence on OSP/OFD
 * LU-8669: kernel update [RHEL6.8 2.6.32-642.6.1.el6]
 * LU-8680: replay-single test_20b: BUG: soft lockup - osc_makes_rpc
 * LU-8681: osd: ignore ENODATA during unlink agent parent
 * LU-8682: sles11sp4 client panic in lov_read_and_clear_async_rc
 * LU-8683: readahead: update ras window correctly
 * LU-8685: Fix JBD2 issue in EL7 Kernels
 * LU-8692: kernel update [RHEL7.2 3.10.0-327.36.2.el7]
 * LU-8694: ZFS format fails when /etc/hostid is set or unset.
 * LU-8695: tgt_txn_start_cb should declare write at unknown offset
 * LU-8697: remove IS_ERR(master_inode) check.
 * LU-8700: dkms fails to build lustre on Power8 due to llite_loop missing
 * LU-8705: fix lnet-selftest to be not skipped on DNE
 * LU-8707: sles11sp3/4 stop working for lbuild
 * LU-8723: "llapi_obd_statfs tests errno in non error cases
 * LU-8731: lfs df exits with status 0 on failures
 * LU-8733: gnilnd: Remove read capability of cksum_test proc file
 * LU-8738: sanity test_255b: FAIL: Ladvise willread should use more memory than 76800 KiB
 * LU-8748: set block size for zfs echo object
 * LU-8749: ldiskfs OSD doesn't inherit S_ISGID correctly on remote directory
 * LU-8751: kernel update [RHEL7.2 3.10.0-327.36.3.el7]
 * LU-8755: kernel update [SLES12 SP1 3.12.62-60.64.8]
 * LU-8763: LustreError: 18461:0:(update_records.c:72:update_records_dump) after lctl abort_recovery
 * LU-8774: copy_from_user should not use MAX_STRING_SIZE
 * LU-8778: osd_index_declare_ea_delete doesn't declare enough credits
 * LU-8788: modify create_pool to use as wrapper
 * LU-8791: Suspicious deadlock on OST stack
 * LU-8795: The user cannot access lustre even if they successfully authenticate by kinit
 * LU-8796: kernel upgrade [RHEL7.3 3.10.0-514.el7]
 * LU-8805: Failover: recovery-mds-scale test_failover_mds: test_failover_mds returned 4
 * LU-8816: test-framework needs to reload SPL module after set hostid
 * LU-8824: sanity-sec test_9: ASSERTION( config->nmc_default_nodemap )
 * LU-8829: 'default' Secure RPC rule display issue
 * LU-8838: kernel update [RHEL6.8 2.6.32-642.11.1.el6]
 * LU-8861: llapi_ladvise.3 man page correction

Technical Tasks


 * LU-3815: "Fix sanity-hsm test 34: ""Remove file during restore"""
 * LU-4048: Fix 'control flow' issues
 * LU-5092: "nodemap: transfer idmaps between MGS
 * LU-5494: mdd: Null pointer may be passed to function that may dereference it
 * LU-6862: Improve dist tarball inclusion of contrib subdirectory
 * LU-7518: Remove non-standard cross build stuff from standard spec file
 * LU-7642: Allow lustre source build without git working directory
 * LU-7643: Remove kernel version string from Lustre release field
 * LU-7645: Stop controlling the RPM Release field from Lustre's build system
 * LU-7648: split lctl lfsck sub-commands to new man pages
 * LU-7699: Overhaul lustre's versioning
 * LU-7839: nodemap: add a deny access flag to nodemap to deny access to non-mapped users
 * LU-8072: Restore module debuginfo
 * LU-8073: Eliminate lustre-source rpm package
 * LU-8113: Remove obsolete meta packages from spec file
 * LU-8116: Cleanup of various configure output messages
 * LU-8369: nodemap: ignore 0@lo for nodemap classification purposes
 * LU-7863: ZFS dmu_prefetch API change.
 * LU-7870: Update ZFS/SPL version to 0.6.5.6
 * LU-8146: Update ZFS/SPL version to 0.6.5.7
 * LU-8713: Load zfs modules during libzfs_init
 * LU-8006: Specify ordering of TBF policy rules