Whamcloud - gitweb
dzogin [Wed, 2 Sep 2009 14:30:33 +0000 (14:30 +0000)]
Branch b1_8
b=20146
i=andrew.perepechko
i=adilger
Modified Files:
Tag: b1_8
lustre/ChangeLog lustre/mds/mds_reint.c
Description: Do not update ctime for the deleted inode.
Details : Check in mds_reint_unlink() before calling fsfilt_setattr().
brian [Wed, 2 Sep 2009 13:53:46 +0000 (13:53 +0000)]
b=20595
i=yangsheng
Build support for OFED 1.4.2.
w.li [Wed, 2 Sep 2009 13:07:24 +0000 (13:07 +0000)]
Branch b1_8
b=20347
i=tappro
i=yong.fan
In llite's SETFLAGS ioctl, the local i_flags should be updated even when
"lsm == NULL".
green [Tue, 1 Sep 2009 00:01:05 +0000 (00:01 +0000)]
b=18534
r=adilger,shadow
Reduce number of RPCs for mkdir/mknod by 1.
yangsheng [Sat, 29 Aug 2009 11:28:40 +0000 (11:28 +0000)]
Branch b1_8
b=19205
i=adilger, nathan
Rename the sanity tests 200z[a-c] to 201[a-c].
dzogin [Wed, 26 Aug 2009 19:41:55 +0000 (19:41 +0000)]
Branch b1_8
b=14951
i=grev
Modified Files:
lustre/ChangeLog lustre/tests/conf-sanity.sh
Description: Improved cleanup procedure for conf-sanity test_46a.
grev [Tue, 25 Aug 2009 16:55:06 +0000 (16:55 +0000)]
b=18956
i=Jian.Yu
add write_disjoint to acc-sm
grev [Tue, 25 Aug 2009 16:13:12 +0000 (16:13 +0000)]
b=18909
i=Adilger
t-f fix for mgs different from MDSDEV
grev [Mon, 24 Aug 2009 11:27:32 +0000 (11:27 +0000)]
b=20371
i=Oleg.Drokin
i=Alexey.Lyashkov
test_151 fix for single node setup
zhanghc [Mon, 24 Aug 2009 05:43:52 +0000 (05:43 +0000)]
b=19791
obdclass/llog_lvfs.c: llog_lvfs_create
l_dentry_open will call dput if there is an error, then
there should be no similar call(l_dput) if error returned
i=adilger@sun.com
i=alexey.lyashkov@sun.com
zhanghc [Mon, 24 Aug 2009 03:19:15 +0000 (03:19 +0000)]
b=20183
fix a race in lov_update_create_set between lov_create and
osc_interpret_create/handle_async_create/cb_create_update
i=alexey.lyashkov@sun.com
i=andrew.perepechko@sun.com
bobijam [Mon, 24 Aug 2009 02:18:43 +0000 (02:18 +0000)]
Branch b1_8
b=17197
o=tom.wang (wangdi)
i=adilger
i=zhenyu.xu (bobijam)
* fix the discard pages counting
* reset the ra window during cache miss
* stride detect fixes
* comments for read-ahead algorithm
zhanghc [Mon, 24 Aug 2009 01:39:50 +0000 (01:39 +0000)]
b=20183
replace multiop with multiop_bg_pause in test_53b in replay-single.sh
in order to make it sure that the operation is paused on return
i=grev@sun.com
dzogin [Sat, 22 Aug 2009 16:07:33 +0000 (16:07 +0000)]
Branch b1_8
b=19566
i=oleg.drokin
i=nathan.rutman
Modified Files:
Tag: b1_8
lustre/ChangeLog lustre/obdclass/obd_mount.c
Description: Prevent inconsistences between linux and lustre mount structures.
Details : Wait indefinitely in server_wait_finished() until mnt_count drops.
Make the sleep interruptible.
johann [Fri, 21 Aug 2009 15:42:44 +0000 (15:42 +0000)]
Branch b1_8
b=20321
i=bzzz
i=dmitry
Take i_alloc_sem in filter_destroy() when truncating the object & fix inode mutex leak
w.li [Fri, 21 Aug 2009 01:29:07 +0000 (01:29 +0000)]
Branch b1_8
b=19743
i=tappro
i=nathan.rutman,robert.read (non-test part)
Attachment 25260: lfs path2fid. Add a "path2fid <path>" command to
lfs(1). Ported from HEAD (Bug 17187 and 19856).
Attachment 25261: lctl getobjversion. Add a new lctl(8) command,
"getobjversion <fid>", to get the version of an object. Currently, only
versions of MDT objects can be retrieved.
Attachment 24412: replay-vbr rmultiop. Make the command sequence an
argument, as required by the new replay-vbr tests. Fix rmultiop_stop's
return code which would always be zero.
Attachment 25094: replay-vbr tests. Move rmultiop_{start,stop} to top.
Add a zconf_mount_clients to balance the former zconf_umount_clients.
Add new tests 0* and 2b (2 renamed to 2a).
dzogin [Thu, 20 Aug 2009 22:34:48 +0000 (22:34 +0000)]
b=20146
i=alex.zhuravlev
i=johann
Description: Increase of the size of the LDLM resource hash.
Details : Bump up RES_HASH_BITS=12.
Modified Files:
Tag: b1_8
lustre/ChangeLog lustre/include/lustre_dlm.h
anserper [Thu, 20 Aug 2009 18:51:07 +0000 (18:51 +0000)]
b=16890
i=Johann Lombardi
i=ZhiYong Tian
do not reset grace times if setquota is given only hardlimit params
shadow [Thu, 20 Aug 2009 15:57:27 +0000 (15:57 +0000)]
correctly send lsm on open replay
Branch b1_8
b=19934
i=tappro
i=panda
zhanghc [Thu, 20 Aug 2009 15:53:51 +0000 (15:53 +0000)]
b=19069
pass (re)connect flag to lprocfs_exp_setup to optimize
the implmentation of nid stats in export (obd_export->exp_nid_stats)
i=andrew.perepechko@sun.com
i=hongchao.zhang@sun.com
yangsheng [Thu, 20 Aug 2009 13:07:44 +0000 (13:07 +0000)]
Branch b1_8
b=16504
i=shadow, green, brian, cliff
Update kernel for RHEL4.7 2.6.9-78.0.24.EL.
yangsheng [Thu, 20 Aug 2009 05:19:50 +0000 (05:19 +0000)]
Branch b1_8
b=19808
i=green, bobijam
Add flock config check.
brian [Wed, 19 Aug 2009 21:21:54 +0000 (21:21 +0000)]
b=20451
i=yangsheng
i=adilger
Need to also make sure the kernel supports grab_cache_page_write_begin()
when enabling the .write_{begin,end} struct address_space_operations
members since write_begin() uses grab_cache_page_write_begin().
brian [Wed, 19 Aug 2009 21:14:47 +0000 (21:14 +0000)]
b=20479
i=yangsheng
i=adilger
i=shadow
Ubuntu's 2.6.24-19-generic (and perhaps other kernels) remove_suid() takes
a struct path * argument.
brian [Wed, 19 Aug 2009 17:30:08 +0000 (17:30 +0000)]
b=20478
i=wangyb
i=adilger
Fix bug where version_tag.pl failed when there was no kernel source
available.
cvs2svn [Wed, 19 Aug 2009 16:32:22 +0000 (16:32 +0000)]
This commit was manufactured by cvs2svn to create branch 'b1_8'.
brian [Wed, 19 Aug 2009 16:32:21 +0000 (16:32 +0000)]
b=19721
i=adilger
Add a debian/ dir to allow building on Debian/Ubuntu systems. The process
is basically (from the top-level) lustre dir:
$ dpkg-buildpackage
$ sudo m-a build ../lustre-source_1.8.1-1_all.deb
The above needs to be captured in a "make debs" Makefile target so that it
works like the "make rpms" target, including building lustre packages as
well as a binary kernel modules package. This will be the next step in this
enhancement.
All of this has really only been tested to build the patchless client with
Ubuntu's 2.6.24-19-generic kernel at this point. However the bits necessary
to build a patched server kernel are included, even if they need to be tested
and perhaps tweaked for supported kernels. At least theoretically, it should
work for newer kernels even.
anserper [Wed, 19 Aug 2009 15:48:44 +0000 (15:48 +0000)]
b=16267
i=Vitaly Fertman
liblustre man pages
adilger [Tue, 18 Aug 2009 20:35:11 +0000 (20:35 +0000)]
Branch b1_8
Allow automake 1.11 for building (tested on FC10 client).
b=19808
zhanghc [Tue, 18 Aug 2009 11:54:41 +0000 (11:54 +0000)]
b=20183
fix a race in lov_update_create_set between lov_create and
osc_interpret_create/handle_async_create/cb_create_update
i=alexey.lyashkov@sun.com
i=andrew.perepechko@sun.com
tappro [Tue, 18 Aug 2009 08:23:13 +0000 (08:23 +0000)]
Branch HEAD
b=15390
i=fanyong
i=pravin
Port permission sync functionality from 1.8 to 2.0.
adilger [Tue, 18 Aug 2009 01:06:43 +0000 (01:06 +0000)]
Branch HEAD
Fix whitespace, remove horrific nested indenting.
adilger [Tue, 18 Aug 2009 00:37:33 +0000 (00:37 +0000)]
Branch HEAD
cscope will index directory names, and "find -prune" only prunes entries
within the specified sub-directory. Exclude such pathnames entirely.
b=19690
anserper [Mon, 17 Aug 2009 19:22:43 +0000 (19:22 +0000)]
ChangeLog
anserper [Mon, 17 Aug 2009 19:20:54 +0000 (19:20 +0000)]
ChangeLog
anserper [Mon, 17 Aug 2009 18:43:23 +0000 (18:43 +0000)]
b=20008
i=Oleg Drokin
i=Alex Tomas
pin the last page before starting transaction in setattr to avoid GFP_FS allocation under transaction
anserper [Mon, 17 Aug 2009 17:43:14 +0000 (17:43 +0000)]
revert incidental commit
anserper [Mon, 17 Aug 2009 17:38:36 +0000 (17:38 +0000)]
b=20008
i=Oleg Drokin
i=Alex Tomas
pin the last page before starting transaction in setattr to avoid GFP_FS allocation under transaction
tappro [Mon, 17 Aug 2009 06:20:35 +0000 (06:20 +0000)]
Branch HEAD
b=19844
i=shadow
i=deen
rc == 1 is not error, make rc checking like in 1.8
zhanghc [Fri, 14 Aug 2009 15:31:00 +0000 (15:31 +0000)]
b=20126
replace multiop with multiop_bg_pause in test_53b in replay-single.sh
in order to make it sure that the operation is paused on return.
i=grev@sun.com
manoj [Thu, 13 Aug 2009 19:42:06 +0000 (19:42 +0000)]
b=19780
Removing unwanted debug messages added during testing.
stinson1947 [Thu, 13 Aug 2009 19:27:08 +0000 (19:27 +0000)]
Updated for 2.0 build 24.
lisaweek [Thu, 13 Aug 2009 17:25:40 +0000 (17:25 +0000)]
b=17167
Creating b_hd_nn branch for libcfs name normaliation effort (libcfs, lnet and lustre modules are branched)
manoj [Thu, 13 Aug 2009 17:15:17 +0000 (17:15 +0000)]
b=19780
r=nathan.rutman
r=grev
r=edg
Tests for the OST Pools feature
liangzhen [Thu, 13 Aug 2009 16:06:05 +0000 (16:06 +0000)]
Branch HEAD
entry for MD perf tests
fanyong [Thu, 13 Aug 2009 01:27:03 +0000 (01:27 +0000)]
Branch HEAD
b=20133
i=pravin.shelar
i=rahul.deshmukh
Insert ".." item just at the second slot of the whole directory.
pravins [Wed, 12 Aug 2009 19:10:59 +0000 (19:10 +0000)]
b=20339
i=alexey.lyashkov
i=ed
fix Cray build failure.
grev [Wed, 12 Aug 2009 18:20:33 +0000 (18:20 +0000)]
b=20371
i=Oleg.Drokin
i=Alexey.Lyashkov
test_151 fix for single node setup
grev [Wed, 12 Aug 2009 18:14:09 +0000 (18:14 +0000)]
b=17951
o=Ed Giesen
skip conf-sanity test_46a for too few OSTs
isaac [Wed, 12 Aug 2009 17:53:54 +0000 (17:53 +0000)]
b=19782,i=isaac:
- fixed a typo in LST sample test script.
anserper [Wed, 12 Aug 2009 17:36:53 +0000 (17:36 +0000)]
b=16267
i=Vitaly Fertman
liblustre library API man pages
tianzy [Wed, 12 Aug 2009 08:42:03 +0000 (08:42 +0000)]
Branch HEAD
make test_11 of sanity_quota.sh invalid
b=19835
i=johann
grev [Tue, 11 Aug 2009 20:07:31 +0000 (20:07 +0000)]
b=18909
i=Adilger
t-f fix for mgs different from MDSDEV1
vitaly [Tue, 11 Aug 2009 13:02:15 +0000 (13:02 +0000)]
Branch HEAD
b=19674
reverting the patch which fails the sles10 build
wangdi [Mon, 10 Aug 2009 18:16:01 +0000 (18:16 +0000)]
Branch: HEAD
Fix for basetest to avoid skipping test like test_300ca when
runing sanity with ONLY=300
b=20385
o=adilger (patch from Andreas)
i=wangdi
vitaly [Mon, 10 Aug 2009 10:53:34 +0000 (10:53 +0000)]
Branch HEAD
b=17545
reverting the patch
shadow [Mon, 10 Aug 2009 09:32:29 +0000 (09:32 +0000)]
correctly handle get_lock vs lock cancel race
Branch b1_8
b=20162
i=gree
i=panda
grev [Mon, 10 Aug 2009 07:54:47 +0000 (07:54 +0000)]
b=19418
i=Brian
lost part of patch: add metadata-updates.sh to rpm
bobijam [Mon, 10 Aug 2009 02:27:34 +0000 (02:27 +0000)]
Branch b1_8
b=19640
i=johann
i=adilger
Avoid calling vfs_link upon a file with link count of 0 in mds_orphan_add_link().
bobijam [Mon, 10 Aug 2009 02:27:28 +0000 (02:27 +0000)]
Branch b1_8
b=19640
i=johann
i=adilger
Avoid calling vfs_link upon a file with link count of 0 in mds_orphan_add_link().
dzogin [Sun, 9 Aug 2009 17:16:57 +0000 (17:16 +0000)]
b=20200
i=andrew.perepechko
i=tianzy
Description: Enhancement for lfs(1) command to use numeric uid/gid.
dzogin [Sun, 9 Aug 2009 17:01:39 +0000 (17:01 +0000)]
b=18674
i=alexey.lyashkov
i=dmitry.zogin
Modified Files:
Tag: b1_8
lustre/ChangeLog lustre/ost/ost_handler.c
Description: Do not start bulk transfer, if client tried to reconnect.
Details : Do not start bulk transfer, if client tried to reconnect,
but failed due to active rpcs. Check for exp_abort_active_req
has been implemented.
wangdi [Sat, 8 Aug 2009 13:27:30 +0000 (13:27 +0000)]
Branch b1_8
Fix for basetest to avoid skipping test like test_300ca when
runing sanity with ONLY=300
b=20385
o=adilger (patch from Andreas)
i=wangdi
fanyong [Sat, 8 Aug 2009 12:40:59 +0000 (12:40 +0000)]
Branch b1_8
b=18902
i=alexey.lyashkov
i=vladimir.saveliev
1) do not reset statahead if conflict was found when opendir
2) do not release "sai" if someone else want to use it for race condition
3) patch from bug 20139
4) increase timeout value for racer killing threads to process client evicted case
There are race conditions for statahead:
1) just before statahead thread is triggered, the first item of the dentry many moved to other
directory, which will cause the statahead thread to be triggered for different directory, and then
cause the original "ls -l" thread which is waiting for the expected directory can not be waken up.
2) when some items are in statahead, such items maybe move to other directory, then the interpret
call maybe find different dentry under such changed directory, and left the original original "ls
-l" thread which is waiting for the expected directory can not be waken up.
3) lli_fid should be set just after vfs inode created on client-side, to prevent the same inode was
created multiple times by statahead thread and other stat thread.
zhanghc [Thu, 6 Aug 2009 11:21:19 +0000 (11:21 +0000)]
b=19406
fix a bug in search_fsname which doesn't find
the correct mountpoint if there are sevearl Lustre
filesystems mounted within each other
i=adilger@sun.com
yangsheng [Thu, 6 Aug 2009 07:09:22 +0000 (07:09 +0000)]
Branch HEAD
b=19674
i=johann, adilger
Initialize variable to get rid of warning.
deshmukh [Thu, 6 Aug 2009 05:59:42 +0000 (05:59 +0000)]
b=20283
i=nathan
i=manoj
Fixed the sanityn assertion problem.
jxiong [Thu, 6 Aug 2009 02:53:46 +0000 (02:53 +0000)]
b=19928
r=eric.mei,rread
Fixed the deadlock problem of simultaneous truncate
bobijam [Thu, 6 Aug 2009 01:40:48 +0000 (01:40 +0000)]
Branch HEAD
b=19854
o=johann
i=oleg.drokin (green)
i=robert.read
o=oleg.drokin(green)
i=johann
i=zhenyu.xu (bobijam)
Description: enable client interface failover
Details : When a child reconnects from another NID, properly update export
nid hash position and ldlm reverse import.
ericm [Wed, 5 Aug 2009 22:08:09 +0000 (22:08 +0000)]
branch: HEAD
fix a case of recursive call loop.
b=20235
r=rread
r=wangdi
nathan [Wed, 5 Aug 2009 21:00:10 +0000 (21:00 +0000)]
b=20372
i=nathan
i=rread
fix compile on XT3
stinson1947 [Wed, 5 Aug 2009 16:01:43 +0000 (16:01 +0000)]
Updated version for a respin of build 23.
pravins [Wed, 5 Aug 2009 11:37:51 +0000 (11:37 +0000)]
b=20339
i=he.huang
i=sheng.yang
fix ia64 and ppc build failure.
shadow [Wed, 5 Aug 2009 07:22:58 +0000 (07:22 +0000)]
don't send revalidate rpc if client have LOOKUP lock already.
Branch HEAD
b=17545
i=green
i=panda
bobijam [Mon, 3 Aug 2009 01:47:57 +0000 (01:47 +0000)]
Branch b1_8
b=20219
o=johann
i=andrew.perepechko (panda)
i=alexander.zarochentsev (zam)
Release lock ref for lockless i/o if allocation failure.
dzogin [Sun, 2 Aug 2009 13:55:49 +0000 (13:55 +0000)]
Branch b1_8
b=20321
i=dmitry.zogin
i=adilger
Description: Deadlock between filter_destroy() and filter_commitrw_write().
Details : filter_destroy() does not hold the DLM lock over the whole
operation. If the DLM lock is dropped, filter_commitrw() can go
through, causing the deadlock between page lock and i_mutex.
tappro [Sun, 2 Aug 2009 09:25:09 +0000 (09:25 +0000)]
Branch HEAD
b=18374
i=rread
i=johann
Port fixes and tests from 1.8 to the HEAD
adilger [Sat, 1 Aug 2009 19:00:54 +0000 (19:00 +0000)]
Branch HEAD
Add ChangeLog entry
b=18539
adilger [Sat, 1 Aug 2009 19:00:20 +0000 (19:00 +0000)]
Branch b1_8
Add ChangeLog entry
b=18539
anserper [Sat, 1 Aug 2009 16:49:50 +0000 (16:49 +0000)]
added a comment wrt unaligned dio and write grants
fanyong [Sat, 1 Aug 2009 04:55:51 +0000 (04:55 +0000)]
Branch HEAD
b=18902
i=alexey.lyashkov
i=vladimir.saveliev
1) do not reset statahead if conflict was found when opendir
2) do not release "sai" if someone else want to use it for race condition
3) patch from bug 20139
4) increase timeout value for racer killing threads to process client evicted case
There are race conditions for statahead:
1) just before statahead thread is triggered, the first item of the dentry many moved to other
directory, which will cause the statahead thread to be triggered for different directory, and then
cause the original "ls -l" thread which is waiting for the expected directory can not be waken up.
2) when some items are in statahead, such items maybe move to other directory, then the interpret
call maybe find different dentry under such changed directory, and left the original original "ls
-l" thread which is waiting for the expected directory can not be waken up.
3) lli_fid should be set just after vfs inode created on client-side, to prevent the same inode was
created multiple times by statahead thread and other stat thread.
robert.read [Fri, 31 Jul 2009 18:08:08 +0000 (18:08 +0000)]
Branch HEAD
b=19963
i=johann
i=andrew
Delete test file.
isaac [Fri, 31 Jul 2009 18:01:30 +0000 (18:01 +0000)]
b=19782,i=isaac:
- fixed a bug in sample script.
nathan [Fri, 31 Jul 2009 17:29:08 +0000 (17:29 +0000)]
b=20001
i=nathan
i=manoj
a=thomas.leibovici@cea.fr
Add llapi calls for retrieving the list of pools and pool members as structures
isaac [Fri, 31 Jul 2009 17:19:50 +0000 (17:19 +0000)]
b=16186,i=isaac:
- enable peer_timeout by default for ko2iblnd and ksocklnd.
stinson1947 [Fri, 31 Jul 2009 16:35:39 +0000 (16:35 +0000)]
Updated for 2.0 Build 23.
anserper [Fri, 31 Jul 2009 16:21:57 +0000 (16:21 +0000)]
b=20314
i=Johann Lombardi
i=Oleg Drokin
properly handle rc<0 from osc_brw_prep_request
grev [Fri, 31 Jul 2009 16:16:46 +0000 (16:16 +0000)]
b=19418
i=Brian
cmd3-38 port to acc-sm: new metadata updates test
anserper [Fri, 31 Jul 2009 16:03:55 +0000 (16:03 +0000)]
b=20317
i=Johann Lombardi
i=Alexander Zarochentsev
do not update kms without extent locks on DIO path
grev [Fri, 31 Jul 2009 15:38:55 +0000 (15:38 +0000)]
b=16476
i=Brian
wait_for_host fix for nodes which have no lustre-tests installed
grev [Fri, 31 Jul 2009 15:33:43 +0000 (15:33 +0000)]
b=16476
i=Brian
wait_for_host fix for nodes which have no lustre-tests installed
grev [Fri, 31 Jul 2009 15:03:39 +0000 (15:03 +0000)]
b=19887
i=Nathan
test_26b: increase wait client evicted value
yangsheng [Fri, 31 Jul 2009 10:38:09 +0000 (10:38 +0000)]
Branch b1_8
b=19808
Add a entry of ChangLog to notice new kernel support.
shadow [Fri, 31 Jul 2009 10:28:03 +0000 (10:28 +0000)]
correctly handling request reference in error cases.
Branch b1_8
b=20122
i=adilger
i=panda
shadow [Fri, 31 Jul 2009 10:23:31 +0000 (10:23 +0000)]
correctly handling request reference in error cases
Branch HEAD
b=20122
i=adilger
i=panda
girish [Fri, 31 Jul 2009 04:57:33 +0000 (04:57 +0000)]
b=20190
i=adilger
i=rahul
fix deadlock due to wrong locking in ext4_mb_free_metadata()
anserper [Thu, 30 Jul 2009 21:17:07 +0000 (21:17 +0000)]
b=18362
i=Johann Lombardi
cleanup HAVE_I_ALLOC_SEM ifndef as we dont have any autoconf test for it anymore
nathan [Thu, 30 Jul 2009 19:29:04 +0000 (19:29 +0000)]
b=20030
i=thomas.leibovici@cea.fr
i=adilger
return changelog entries as a struct from llapi
grev [Thu, 30 Jul 2009 19:25:24 +0000 (19:25 +0000)]
b=20259
i=Brian
get rid of useless rm PI*
robert.read [Thu, 30 Jul 2009 18:38:06 +0000 (18:38 +0000)]
Branch HEAD
b=19338
Reverting the patch from 19338 since it has caused 20278, which is worse than the original issue.