Whamcloud - gitweb
fs/lustre-release.git
14 years agob=19312
grev [Fri, 17 Jul 2009 15:59:34 +0000 (15:59 +0000)]
b=19312
i=Brian
i=Oleg
port parallel_grouplock.c from cmd3

14 years agob=20179
grev [Fri, 17 Jul 2009 15:33:20 +0000 (15:33 +0000)]
b=20179
i=Robert.Read
process TESTSUITE log to fail if error

14 years agoBranch b1_8
yangsheng [Fri, 17 Jul 2009 13:30:27 +0000 (13:30 +0000)]
Branch b1_8
b=19615

i=shadow, johann

Disable server build for rhel4 & sles9.

14 years agob=19415
grev [Thu, 16 Jul 2009 15:37:43 +0000 (15:37 +0000)]
b=19415
i=Robert.Read
fixes related to new load var

14 years agob=18801
anserper [Thu, 16 Jul 2009 14:27:01 +0000 (14:27 +0000)]
b=18801

the lost part of the previous commit

14 years agob=19215
grev [Thu, 16 Jul 2009 10:40:59 +0000 (10:40 +0000)]
b=19215
i=Robert.Read
verbose do_nodes
do_node print HOSTNAME for no_dsh

14 years agoBranch b1_8
tianzy [Thu, 16 Jul 2009 09:55:52 +0000 (09:55 +0000)]
Branch b1_8
fix a typo for LQUOTA_QAQ_CEATE_LQS
b=20031

14 years agoadd time.h include to avoid warning
anserper [Thu, 16 Jul 2009 09:44:20 +0000 (09:44 +0000)]
add time.h include to avoid warning

14 years ago b=18801
anserper [Thu, 16 Jul 2009 09:03:28 +0000 (09:03 +0000)]
  b=18801
  i=Oleg Drokin
  i=Alexander Zarochentsev

  atime, mtime and additional POSIX compliance stuff for direct I/O and lockless I/O

14 years agob=19872
girish [Thu, 16 Jul 2009 07:35:57 +0000 (07:35 +0000)]
b=19872
i=adilger
i=dmitry.zogin

fix test 130e to write 64k blocks with 128k stripes so FIEMAP can work properly across mixed platform clusters

14 years agoBranch b1_8
tianzy [Thu, 16 Jul 2009 06:40:12 +0000 (06:40 +0000)]
Branch b1_8
Fix group quota "too many blocks" OSS crashes.

In filter_commitrw_write():
1. lquota_chkquota();
2. obdo_from_inode();
3. lquota_pending_commit();

2. has changed uid and gid in this case, then caused
lquota_chkquota() and lquota_pending_commit() to use
different uid and gid. That is the root of the bug.
b=18793
i=yong.fan
i=panda

14 years agob=19870
zhanghc [Thu, 16 Jul 2009 05:22:21 +0000 (05:22 +0000)]
b=19870

fix a bug of processing export->exp_rpc_count during recovery

i=tappro@sun.com
i=alexey.lyashkov@sun.com

14 years agob=18539
zhanghc [Thu, 16 Jul 2009 04:46:30 +0000 (04:46 +0000)]
b=18539

don't allocate new files on degraded OSTs

i=adilger@sun.com
i=hongchao.zhang@sun.com

14 years agorevert attachment 24367 due compilation problems.
shadow [Thu, 16 Jul 2009 04:46:01 +0000 (04:46 +0000)]
revert attachment 24367 due compilation problems.
Branch b1_8

b=18801

14 years agoBranch b1_8
yujian [Thu, 16 Jul 2009 02:01:58 +0000 (02:01 +0000)]
Branch b1_8
b=20047
i=robert.read
i=yong.fan

reset debug level if lustre was already mounted

14 years agoBranch b1_8
yujian [Thu, 16 Jul 2009 01:47:12 +0000 (01:47 +0000)]
Branch b1_8
b=19432
i=robert.read
i=grev

1) move MPIRUN and MPI_USER from ncli.sh to local.sh
2) remove MPIBIN
3) redirect the stderr of "which mpirun" to /dev/null

14 years agoBranch b1_8
bobijam [Thu, 16 Jul 2009 01:32:15 +0000 (01:32 +0000)]
Branch b1_8
b=17682
i=alexey.lyashkov (shadow)
i=sheng.yang

DLM debug message improvement.

14 years agoBranch b1_8
w.li [Thu, 16 Jul 2009 01:03:30 +0000 (01:03 +0000)]
Branch b1_8
b=19999
i=grev
i=johann

Remove PID reading from multiop_bg_pause.  Turn to $! for multiop PID in
runmultiop_bg_pause.  The result looks more consistent with HEAD.

14 years agoBranch b1_8
yangsheng [Wed, 15 Jul 2009 15:27:53 +0000 (15:27 +0000)]
Branch b1_8
b=19205

i=adilger, Jacques-Charles

'lfs df' for pool support.

14 years agoBranch b1_8
fanyong [Wed, 15 Jul 2009 15:04:51 +0000 (15:04 +0000)]
Branch b1_8
b=20002
i=tianzy
i=robert.read

Check "auto_quota_on()" return value before "build_lqs()" called.

14 years agob=19150
grev [Wed, 15 Jul 2009 13:35:12 +0000 (13:35 +0000)]
b=19150
i=Robert.Read
cleanup: work in TESTDIR instead of DIR

14 years agob=19544
anserper [Wed, 15 Jul 2009 13:28:47 +0000 (13:28 +0000)]
b=19544
i=Robert Read
i=Dmitry Zogin

move to static on-stack allocations

14 years agob=19529
dzogin [Wed, 15 Jul 2009 13:21:11 +0000 (13:21 +0000)]
b=19529
i=andrew.perepechko
i=adilger
Description: Avoid deadlock for local client writes
Details    : Use new OBD_BRW_MEMALLOC flag to notify OST about writes in the
     memory freeing context. This allows OST threads to set the
     PF_MEMALLOC flag on task structures in order to allocate memory
     from reserved pools and complete IO.
     Use GFP_HIGHUSER for OST allocations for non-local client writes,
     so that the OST threads generate memory pressure and allow
     inactive pages to be reclaimed.

14 years agoBranch b1_8
w.li [Wed, 15 Jul 2009 13:19:41 +0000 (13:19 +0000)]
Branch b1_8
b=19980
i=alexey.lyashkov
i=andrew.perepechko

Subtract evicted clients from "completed_clients" in recovery_status.

14 years agob=18801
anserper [Wed, 15 Jul 2009 12:47:33 +0000 (12:47 +0000)]
b=18801
i=Oleg Drokin
i=Alexander Zarochentsev

atime, mtime and addition POSIX compliance stuff for direct I/O and lockless I/o

14 years agob=16267
anserper [Wed, 15 Jul 2009 12:41:15 +0000 (12:41 +0000)]
b=16267
added the man page to the manpage list

14 years agob=16267
anserper [Wed, 15 Jul 2009 12:27:00 +0000 (12:27 +0000)]
b=16267
i=ZhiYong Tian
a man page for llapi_file_open and llapi_file_create

14 years agofile llapi_file_open.3 was added on branch b1_8 on 2009-07-15 12:27:00 +0000
anserper [Wed, 15 Jul 2009 12:23:39 +0000 (12:23 +0000)]
file llapi_file_open.3 was added on branch b1_8 on 2009-07-15 12:27:00 +0000

14 years agob=20074
grev [Wed, 15 Jul 2009 11:47:33 +0000 (11:47 +0000)]
b=20074
i=Alexey.Lyashkov
i=Robert.Read
process tar's rc correctly

14 years agofirst part of new dcache handling (bug 16654)
shadow [Tue, 14 Jul 2009 06:23:12 +0000 (06:23 +0000)]
first part of new dcache handling (bug 16654)
make DCACHE_LUSTRE_INVALID is always defined.

Branch b1_8
b=16417
i=green
i=johann

14 years agob=20094
grev [Thu, 9 Jul 2009 08:44:23 +0000 (08:44 +0000)]
b=20094
i=Alexander.Zarochentsev (Zam)
test_121: ignore useless dd output

14 years agob=19889
nathan [Wed, 8 Jul 2009 19:51:03 +0000 (19:51 +0000)]
b=19889
i=rread
i=manoj
return error codes from pool_cmd (and others)

14 years agolock ordering violation between &cli->cl_sem and _lprocfs_lock
shadow [Wed, 8 Jul 2009 03:45:44 +0000 (03:45 +0000)]
lock ordering violation between &cli->cl_sem and _lprocfs_lock

Branch b1_8
b=18380
i=jay
i=deen

14 years agob=20055
brian [Tue, 7 Jul 2009 19:26:09 +0000 (19:26 +0000)]
b=20055
i=green
i=zhanghc
o=johann

Fix for connectathon, NFSv4: ASSERTION(last_discon == NULL) failed

14 years agob=16774
zhanghc [Tue, 7 Jul 2009 16:09:41 +0000 (16:09 +0000)]
b=16774

cache the extent lock gotten before doing I/O in "ll_file_aio_read"
and "ll_file_aio_write", then "ll_read_page", "ll_read_ahead_page"
and "ll_writepage" will save much time spent on "ldlm_lock_match"
(in cache_add_extent) to find the appropriate extent lock

1, using current->journal_info to store the locks(struct ll_thread_data)
2, replace o_reget_short_lock, o_release_short_lock with o_get_lock in obd_ops

i=johann@sun.com
i=oleg.drokin@sun.com

14 years agob=19737
grev [Mon, 6 Jul 2009 13:35:23 +0000 (13:35 +0000)]
b=19737
i=Robert.Read
acc-sm, t-f NFSCLIENT changes

14 years agob=20054
girish [Sat, 4 Jul 2009 13:21:36 +0000 (13:21 +0000)]
b=20054
o=rread
i=girish
i=eric.mei

print out stripe data. regression caused by 18442

14 years agob=16890
anserper [Fri, 3 Jul 2009 22:40:50 +0000 (22:40 +0000)]
b=16890
i=Elena

SHOW_GROUP_INFO call consisting of two commands was sometimes interpreted as a long command with

14 years agoBranch b1_8
zam [Fri, 3 Jul 2009 15:04:45 +0000 (15:04 +0000)]
Branch b1_8
b=19963
i=johann
i=andrew.perepechko

remove temporary file upon successful completion of replay-single test 13;
check whether it is deletable and can it cause an EEXIST failure in
replay-dual test 13 (see bug 19963).

14 years agoBranch b1_8
zam [Fri, 3 Jul 2009 14:55:46 +0000 (14:55 +0000)]
Branch b1_8
b=19963
i=johann

mds_lookup: misplaced EXIT debug marker.

14 years agob=18442
girish [Fri, 3 Jul 2009 05:39:34 +0000 (05:39 +0000)]
b=18442
i=girish
o=rread

fix test 200d to check for new output pattern

14 years agob=19980
anserper [Thu, 2 Jul 2009 21:59:21 +0000 (21:59 +0000)]
b=19980
i=Andrew Perepechko
i=Alexey Lyashkov
o=Mike Pershin

fix for recovery messages

14 years agoBranch b1_8
yangsheng [Thu, 2 Jul 2009 15:36:01 +0000 (15:36 +0000)]
Branch b1_8
b=19969

Fixed for rpm install dependencies failed.

14 years agoBranch b1_8
yangsheng [Thu, 2 Jul 2009 13:41:04 +0000 (13:41 +0000)]
Branch b1_8

b=19846

Fork the patch get rid of the fuzz.

14 years agofile sd_iostats-2.6-sles10.patch was added on branch b1_8 on 2009-07-02 13:41:03...
yangsheng [Thu, 2 Jul 2009 13:07:36 +0000 (13:07 +0000)]
file sd_iostats-2.6-sles10.patch was added on branch b1_8 on 2009-07-02 13:41:03 +0000

14 years agoBranch b1_8
yangsheng [Thu, 2 Jul 2009 03:48:40 +0000 (03:48 +0000)]
Branch b1_8
b=19846

i=shadow, johann

Update kernel to 2.6.18-128.1.14.el5.

14 years agoBranch b1_8
bobijam [Thu, 2 Jul 2009 01:12:04 +0000 (01:12 +0000)]
Branch b1_8
b=18478
i=yibin.wang

Improve sanity test_180.

14 years agoBranch b1_8
johann [Wed, 1 Jul 2009 12:01:16 +0000 (12:01 +0000)]
Branch b1_8
b=19559
i=shadow (patch from Oleg)
i=panda

Fix NFS export: properly calculate fh size & reintroduce ll_get_parent.

14 years agob=18442
girish [Wed, 1 Jul 2009 06:51:40 +0000 (06:51 +0000)]
b=18442
i=adilger
i=rread
o=james simmons

integrate "ll_getstripe_info" into "lfs getstripe"

14 years agoBranch b1_8
johann [Tue, 30 Jun 2009 21:04:16 +0000 (21:04 +0000)]
Branch b1_8
b=20011
i=shadow
i=panda

ll_shrink_cache() can sleep while holding the ll_sb_lock.
Convert ll_sb_lock to a read/write semaphore to fix the problem.

14 years agoBranch b1_8
yangsheng [Tue, 30 Jun 2009 15:25:50 +0000 (15:25 +0000)]
Branch b1_8
b=18970

Update kernel to 2.6.5-7.317.

14 years agoBranch b1_8
yangsheng [Tue, 30 Jun 2009 15:20:33 +0000 (15:20 +0000)]
Branch b1_8
b=19848

Update kernel to SLES11 2.6.27.23-0.1.1.

14 years agoBranch b1_8
tianzy [Tue, 30 Jun 2009 06:24:40 +0000 (06:24 +0000)]
Branch b1_8
fix quota spurious message
b=19898
i=tianzy
i=panda

14 years agoBranch b1_8
vitaly [Mon, 29 Jun 2009 17:10:00 +0000 (17:10 +0000)]
Branch b1_8
b=19207
i=zam
i=panda

memleak fix for zero ll_get_user_pages case

14 years agoBranch b1_8
bobijam [Mon, 29 Jun 2009 10:39:12 +0000 (10:39 +0000)]
Branch b1_8
b=17197

roll back this patch, it causes lbugs.

14 years agoBranch b1_8
bobijam [Mon, 29 Jun 2009 07:09:43 +0000 (07:09 +0000)]
Branch b1_8
b=18478

obdecho cannot attach remote ost device.

14 years agob=17569
girish [Mon, 29 Jun 2009 04:34:57 +0000 (04:34 +0000)]
b=17569
i=adilger
i=johann

add check for >8TB filesystems. Don't allo mounting them until force_over_8tb mount option is specified

14 years agoBranch b1_8
bobijam [Mon, 29 Jun 2009 01:36:02 +0000 (01:36 +0000)]
Branch b1_8
b=18478
i=zhen.liang
i=johann

Fix obdecho to make it work.

14 years agowith AT enabled, we can have connect request timeout ~ reconnect
shadow [Sat, 27 Jun 2009 04:59:44 +0000 (04:59 +0000)]
with AT enabled, we can have connect request timeout ~ reconnect
timeout and test can't see real disconnect

Branch b1_8
b=19280
i=rread

14 years agob=18948
anserper [Fri, 26 Jun 2009 08:52:28 +0000 (08:52 +0000)]
b=18948

reverting patch because of the memory leak

14 years agoBranch b1_8
zam [Thu, 25 Jun 2009 17:07:19 +0000 (17:07 +0000)]
Branch b1_8
b=19630
i=alexander.zarochentsev
i=alexey.lyashkov

fail truncate syscall if one ost is inactive.

14 years agob=18021
grev [Thu, 25 Jun 2009 17:05:19 +0000 (17:05 +0000)]
b=18021
i=Adilger
i=Nathan
i=Robert
t-f CLIENTONLY does not depend on *_HOST

14 years agob=18948
anserper [Thu, 25 Jun 2009 07:06:46 +0000 (07:06 +0000)]
b=18948
i=Nathan Rutman
i=Andreas Dilger

Tunable recovery timeouts (v1)

14 years agoBranch b1_8
bobijam [Thu, 25 Jun 2009 01:23:33 +0000 (01:23 +0000)]
Branch b1_8
b=18702
i=nathan.rutman
i=johann

Server failover cleanup needs destroy llog.

14 years agob=18144
nathan [Wed, 24 Jun 2009 16:31:31 +0000 (16:31 +0000)]
b=18144
i=ericm
i=wangybin
hold the fsdb flag while assigning new OST indicies to prevent racing

14 years agoBranch b1_8
yangsheng [Wed, 24 Jun 2009 13:51:25 +0000 (13:51 +0000)]
Branch b1_8
b=19869

i=shadow, johann, tianzy

Fixes for Werror flag enabled.

14 years agob=19927
brian [Wed, 24 Jun 2009 13:28:04 +0000 (13:28 +0000)]
b=19927
i=yangsheng
i=wangyb

If the $LUSTRE_VERS environment variable is set, don't bother doing the
legwork to determine the version, and just exit early with it.

14 years agoBranch b1_8
yangsheng [Wed, 24 Jun 2009 13:27:28 +0000 (13:27 +0000)]
Branch b1_8

b=19553

i=johann, shadow

Workaround for SLES10 + OFED-1.4.1.

14 years agob=19870
zhanghc [Wed, 24 Jun 2009 02:18:48 +0000 (02:18 +0000)]
b=19870

fix a bug in "mds_reint_rename" which casuse
one lock not to be canceled

i=johann@sun.com
i=tappro@sun.com

14 years agoBranch b1_8
bobijam [Wed, 24 Jun 2009 01:27:06 +0000 (01:27 +0000)]
Branch b1_8
b=17197
o=tom.wang (wangdi)
i=adilger
i=johann

Fix for detecting read-ahead problem.

14 years agoBranch b1_8
johann [Tue, 23 Jun 2009 23:37:51 +0000 (23:37 +0000)]
Branch b1_8
b=19933
i=grev

add error messages in test 103.

14 years agoBranch b1_8
johann [Tue, 23 Jun 2009 14:39:03 +0000 (14:39 +0000)]
Branch b1_8
b=19797
i=grev

Fix return code of test 84.

14 years agoBranch b1_8
zam [Mon, 22 Jun 2009 16:13:31 +0000 (16:13 +0000)]
Branch b1_8
b=19321
i=johann

add the jbd2 fix (wrong jiffie rounding) to the rhel5 series files.

14 years agodon't lose wakeup for replay_single::test_72a
shadow [Sat, 20 Jun 2009 11:59:51 +0000 (11:59 +0000)]
don't lose wakeup for replay_single::test_72a

Branch b1_8
b=19581
i=tappro
i=johann

14 years ago Branch b1_8
dzogin [Fri, 19 Jun 2009 22:18:08 +0000 (22:18 +0000)]
 Branch b1_8
 b=18624
 i=dmitry.zogin
 i=zhenyu.xu

14 years agoBranch b1_8
tappro [Fri, 19 Jun 2009 21:45:50 +0000 (21:45 +0000)]
Branch b1_8
b=19528
i=rread
i=johann

14 years agob=18775
anserper [Fri, 19 Jun 2009 13:23:37 +0000 (13:23 +0000)]
b=18775
i=Andreas Dilger
i=Robert Read
o=Emoly Liu

 cleanup get/set_param patch

14 years agob=19913
grev [Fri, 19 Jun 2009 13:22:09 +0000 (13:22 +0000)]
b=19913
i=Johann
skip test_151 for remote ost && no_dsh

14 years agoBranch b1_8
johann [Fri, 19 Jun 2009 12:21:30 +0000 (12:21 +0000)]
Branch b1_8
b=19913
i=grev
i=bobijam

Add support remote ost in test 151.

14 years agob=19875
girish [Fri, 19 Jun 2009 11:24:47 +0000 (11:24 +0000)]
b=19875
i=johann
i=adilger

use i_fs_version in sles11

14 years agoBranch b1_8
yangsheng [Thu, 18 Jun 2009 09:07:44 +0000 (09:07 +0000)]
Branch b1_8
b=19813

i=johann, yangsheng

Complete error handling.
Author: bobijam

14 years agoMove read/write CATALOG into osc layer.
shadow [Thu, 18 Jun 2009 05:39:39 +0000 (05:39 +0000)]
Move read/write CATALOG into osc layer.

Branch b1_8
b=18800
i=tappro
i=wangdi

14 years agob=18801
anserper [Wed, 17 Jun 2009 17:44:13 +0000 (17:44 +0000)]
b=18801

unused variable warning

14 years agob=18801
anserper [Wed, 17 Jun 2009 17:12:40 +0000 (17:12 +0000)]
b=18801
i=Dmitry Zogin
i=Alexander Zarochentsev

remove lockless I/O code which is duplicated by direct I/O code

14 years agob=19778
anserper [Wed, 17 Jun 2009 16:44:02 +0000 (16:44 +0000)]
b=19778
i=Johann Lombardi
i=ZhiYong Tian

reduce qonoff lock contention for getquota

14 years agob=16424
anserper [Wed, 17 Jun 2009 12:48:30 +0000 (12:48 +0000)]
b=16424
i=Sheng Yang
i=Yibin Wang
o=Brian Murrell

fix for sles' -obj dirs

14 years agob=19259
zhanghc [Wed, 17 Jun 2009 07:36:28 +0000 (07:36 +0000)]
b=19259

add a NULL proc in osc_destroy to allow
the blocking thread to process LUSTRE_FATAL_SIGS

i=adilger@sun.com
i=johann@sun.com

14 years agob=19706
zhanghc [Tue, 16 Jun 2009 11:15:45 +0000 (11:15 +0000)]
b=19706

fix a deadlock bug introduced from kernel 2.6.23

i=adilger@sun.com
i=johann@sun.com

14 years agoBranch b1_8
yangsheng [Tue, 16 Jun 2009 09:00:19 +0000 (09:00 +0000)]
Branch b1_8
b=19868

i=johann, shadow

Move up the THREAD_SIZE to comply the Werror.

14 years agoBranch b1_8
bobijam [Tue, 16 Jun 2009 05:10:48 +0000 (05:10 +0000)]
Branch b1_8
b=17213

Minor change.

14 years agob=19852
girish [Tue, 16 Jun 2009 02:44:06 +0000 (02:44 +0000)]
b=19852
update changelog to use new e2fsprogs version

14 years agoBranch b1_8
bobijam [Mon, 15 Jun 2009 01:42:31 +0000 (01:42 +0000)]
Branch b1_8
b=17213
i=andrew.perepechko (panda)
i=johann

avoid getting client uuid from invalid client_obd.

14 years agoBranch b1_8
zam [Fri, 12 Jun 2009 05:33:40 +0000 (05:33 +0000)]
Branch b1_8
b=19321
i=johann
i=alex.zhuravlev

jbd and jbd2 fixes: remove jiffie rounding from jbd commit timer setup code.
The rounding can cause the timer to be fired before the corresponding
transaction gets expired. That transaction can stay not committed forever if
no new transaction is created or expicit sync/umount happens.

14 years agofile jbd2-commit-timer-no-jiffies-rounding.diff was added on branch b1_8 on 2009...
zam [Fri, 12 Jun 2009 05:29:46 +0000 (05:29 +0000)]
file jbd2-commit-timer-no-jiffies-rounding.diff was added on branch b1_8 on 2009-06-12 05:33:40 +0000

14 years agofile jbd-commit-timer-no-jiffies-rounding.diff was added on branch b1_8 on 2009-06...
zam [Fri, 12 Jun 2009 05:29:45 +0000 (05:29 +0000)]
file jbd-commit-timer-no-jiffies-rounding.diff was added on branch b1_8 on 2009-06-12 05:33:40 +0000

14 years agob=19793
grev [Thu, 11 Jun 2009 14:31:15 +0000 (14:31 +0000)]
b=19793
o=Nathan
i=Johann
minor script issue in test_17

14 years agoBranch b1_8
johann [Thu, 11 Jun 2009 14:13:14 +0000 (14:13 +0000)]
Branch b1_8
b=19820
i=bobijam

stop os2 at the end of test 50f.

14 years agoBranch b1_8
yangsheng [Thu, 11 Jun 2009 12:15:40 +0000 (12:15 +0000)]
Branch b1_8
b=18668

Recommit the kernel config file.

14 years agoBranch b1_8
adilger [Wed, 10 Jun 2009 21:38:39 +0000 (21:38 +0000)]
Branch b1_8
Ignore generated files.

14 years agob=16424
brian [Wed, 10 Jun 2009 21:09:58 +0000 (21:09 +0000)]
b=16424
i=rread
i=yangsheng

Allow builds from tarball (i.e. those without CVS subdirs) to properly
populate the Lustre version field.

This is done by saving the date that would have been found in the CVS files
into a status file when buiding the tarball and then using that when building
from tarball.