Whamcloud - gitweb
fs/lustre-release.git
14 years agob=21547 add cascading_rw into lustre mpi tests
Elena Gryaznova [Thu, 17 Dec 2009 19:15:33 +0000 (22:15 +0300)]
b=21547 add cascading_rw into lustre mpi tests

o=James Simmons
i=Johann
i=grev

14 years agob=21252 add reference on extent lock during readahead
Hongchao.Zhang@Sun.COM [Fri, 18 Dec 2009 19:44:34 +0000 (03:44 +0800)]
b=21252 add reference on extent lock during readahead

add an extra LCK_PR reference on the extent lock
covering the readhead page to pin it during the readahead

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

14 years agob=21532 various build fixes for rhel4 and sles9
Brian J. Murrell [Thu, 17 Dec 2009 13:58:11 +0000 (08:58 -0500)]
b=21532 various build fixes for rhel4 and sles9

various build fixes for rhel4 and sles9 including:
- configure checks for and backports of:
  * atomic_inc_not_zero
  * atomic_cmpxchg
- don't include linux/pagevec.h in llite/lloop.c as it causes some duplicate
  definitions when it's later included by another source file
- use LPU64 portability formats

14 years agob=21586 clarify lbuild error reporting
Brian J. Murrell [Thu, 17 Dec 2009 15:09:47 +0000 (10:09 -0500)]
b=21586 clarify lbuild error reporting

clarify the reporting of build errors by:
- disabling xtrace during the stack trace printing
- moving the ccache statistics printing exit trap so that it's only
  installed if ccache has been enabled
- remove some useless $() constructs around function calls
- only send the untrapped error debuggery to me

i=wangyb
i=yangsheng

14 years agob=21527 don't use vendors own CFLAGS for the lustre build
johann [Thu, 17 Dec 2009 14:22:43 +0000 (15:22 +0100)]
b=21527 don't use vendors own CFLAGS for the lustre build

14 years agob=21582 make sure not to use cached statfs info in fail()
johann [Thu, 17 Dec 2009 13:22:10 +0000 (14:22 +0100)]
b=21582 make sure not to use cached statfs info in fail()

Use client_up which sleeps for 1s before running statfs.
It grants that we are not using cached statfs info, since patch
from bug 20065 has been landed.

i=grev

14 years agob=18630 avoid the race of acquiring/releasing quota
Landen [Thu, 17 Dec 2009 03:52:25 +0000 (11:52 +0800)]
b=18630 avoid the race of acquiring/releasing quota

Before this patch, checking and computing the info of lqs
are located in two different functions(check_cur_qunit()
and schedule_dqacq()). It is racy. With this patch, they
will all happen in one function(check_cur_qunit()).

i=johann
i=panda

14 years agob=16774 fix a issue caused by LDLM_POLICY_SKIP_LOCK
Hongchao.Zhang@Sun.COM [Wed, 16 Dec 2009 08:09:14 +0000 (16:09 +0800)]
b=16774 fix a issue caused by LDLM_POLICY_SKIP_LOCK

introduce a new lock flag LDLM_FL_SKIPPED to avoid
checking locks repeatly in ldlm_cancel_lru_local()

i=oleg.drokin@sun.com
i=hongchao.zhang@sun.com

14 years agob=20722 Make watchdog timer messages more clear and descriptive.
Dmitry Zogin [Wed, 16 Dec 2009 21:59:44 +0000 (16:59 -0500)]
b=20722 Make watchdog timer messages more clear and descriptive.

i=adilger

14 years agob=19884 add replay-dual test_14b to always except list
johann [Wed, 16 Dec 2009 23:13:40 +0000 (00:13 +0100)]
b=19884 add replay-dual test_14b to always except list

14 years agob=21527 Use %configure in the RPM SPEC file
Brian J. Murrell [Wed, 16 Dec 2009 20:22:57 +0000 (15:22 -0500)]
b=21527 Use %configure in the RPM SPEC file

use the %configure macro defined by the RPM installation so that the
vendor's filesystem hierarchy preferences are observered

i=wangyb
i=yangsheng

14 years agob=20765 handle -EINTR in osc_interpret_create()
johann [Wed, 16 Dec 2009 21:43:12 +0000 (22:43 +0100)]
b=20765 handle -EINTR in osc_interpret_create()

i=tappro
i=zam

14 years agob=21600 Fix unitialized pointer
Brian J. Murrell [Wed, 16 Dec 2009 17:41:00 +0000 (12:41 -0500)]
b=21600 Fix unitialized pointer

gcc warns that ctx may be used prior to assignment, so initialize it
at declaration

i=rread
i=johann

14 years agob=21396 LNET soft lockups in socknal_connd thread
Liang Zhen [Wed, 16 Dec 2009 16:07:41 +0000 (00:07 +0800)]
b=21396 LNET soft lockups in socknal_connd thread

don't hog CPU for active-connecting if another connd is accepting
connecting-requst from the same peer
i=isaac
i=maxim

14 years agob=19887 force reconnect in test 26a
Johann Lombardi [Wed, 16 Dec 2009 13:29:56 +0000 (14:29 +0100)]
b=19887  force reconnect in test 26a

i=grev

14 years agob=17682 limit performance impact of rpctrace, dlmtrace & quota (real commit)
Dmitry Zogin [Tue, 15 Dec 2009 15:46:04 +0000 (10:46 -0500)]
b=17682 limit performance impact of rpctrace, dlmtrace & quota (real commit)

Performance improvements for debug messages with
D_RPCTRACE, D_LDLM, D_QUOTA
previous commit (995342ab73da4ac8e364f90d2b26a88a5eef4801)
only included the changelog entry.

i=alexey.lyashkov
i=andrew.perepechko
i=johann

14 years agob=17682 limit performance impact of rpctrace, dlmtrace & quota
Dmitry Zogin [Tue, 15 Dec 2009 15:46:04 +0000 (10:46 -0500)]
b=17682 limit performance impact of rpctrace, dlmtrace & quota

Performance improvements for debug messages with
D_RPCTRACE, D_LDLM, D_QUOTA.

i=alexey.lyashkov
i=andrew.perepechko
i=johann

14 years agob=21147 fix silent conflict with patch from bug 21379
Johann Lombardi [Wed, 16 Dec 2009 12:44:56 +0000 (13:44 +0100)]
b=21147 fix silent conflict with patch from bug 21379

14 years agob=21147 be tolerant to setting the same type and version of quota
Andrew Perepechko [Wed, 16 Dec 2009 00:31:19 +0000 (03:31 +0300)]
b=21147 be tolerant to setting the same type and version of quota

i=Johann Lombardi
i=ZhiYong Tian

14 years agoRemove lustrecvs. Not needed for git
Brian Reitz [Mon, 14 Dec 2009 06:49:06 +0000 (00:49 -0600)]
Remove lustrecvs.  Not needed for git

14 years agoIntroduce .gitignore files. GIT_EPOCH_B1_8
Brian Reitz [Sat, 12 Dec 2009 14:10:18 +0000 (08:10 -0600)]
Introduce .gitignore files.

The top level .gitignore file is new and is an attempt at
pulling in some of the common items that you might get for
free (by default)with CVS.  The other subdir/.gitignore files
are translated versions of their corresponding .cvsignore
file.  Because CVS does not descend into a subdir when
applying a ingore rule we have to prepend a "/" to
the pathname to get git to behave the same way.

14 years agoAdd ldiskfs/build symlink
Brian Reitz [Sat, 12 Dec 2009 14:06:41 +0000 (08:06 -0600)]
Add ldiskfs/build symlink

14 years agoThis commit was manufactured by cvs2svn to create branch 'git_cnv_b1_8'.
cvs2svn [Fri, 11 Dec 2009 04:40:35 +0000 (04:40 +0000)]
This commit was manufactured by cvs2svn to create branch 'git_cnv_b1_8'.

14 years agoBranch HEAD
robert.read [Fri, 11 Dec 2009 04:40:33 +0000 (04:40 +0000)]
Branch HEAD

Another test commit to ensure cvs is back to normal.

14 years agoBranch HEAD
robert.read [Fri, 11 Dec 2009 02:35:38 +0000 (02:35 +0000)]
Branch HEAD

Test commit to ensure cvs  is back to normal.

14 years agoBranch HEAD
robert.read [Fri, 11 Dec 2009 01:40:25 +0000 (01:40 +0000)]
Branch HEAD

Updating ChangeLog supported kernel section.

14 years agoBranch b1_8
johann [Thu, 10 Dec 2009 22:31:04 +0000 (22:31 +0000)]
Branch b1_8
b=19557
i=adilger
i=andrew

add cond_resched() calls to lustre_hash_for_each_empty() to avoid hogging
the cpu when many locks to process on disconnect/eviction.

14 years agoBranch b1_8
johann [Thu, 10 Dec 2009 22:31:00 +0000 (22:31 +0000)]
Branch b1_8
b=19557
i=adilger
i=andrew

add cond_resched() calls to lustre_hash_for_each_empty() to avoid hogging
the cpu when many locks to process on disconnect/eviction.

14 years agoUpdated for 1.8.2 build 07.
stinson1947 [Thu, 10 Dec 2009 17:32:03 +0000 (17:32 +0000)]
Updated for 1.8.2 build 07.

14 years agoUpdated for 2.0 build 32. v1_10_0_32
stinson1947 [Thu, 10 Dec 2009 17:30:44 +0000 (17:30 +0000)]
Updated for 2.0 build 32.

14 years agoBranch b1_8
johann [Thu, 10 Dec 2009 17:06:02 +0000 (17:06 +0000)]
Branch b1_8
b=16356

disable test 2 since it is known not to work until we have
version checking for lock replay

14 years agoBranch b1_8
johann [Thu, 10 Dec 2009 17:03:17 +0000 (17:03 +0000)]
Branch b1_8
b=19023

the patch from bug 19023 is landed now.

14 years ago Branch HEAD
dzogin [Thu, 10 Dec 2009 13:52:52 +0000 (13:52 +0000)]
 Branch HEAD
 b=21259
 i=andrew.perepechko
 i=alexey.lyashkov
----------------------------------------------------------------------
 Modified Files:
  lustre/ChangeLog lustre/obdclass/linux/linux-module.c
----------------------------------------------------------------------
Description: Allow non-root access for "lfs check".
Details    : Added a check in obd_class_ioctl() for OBD_IOC_PING_TARGET.

14 years ago Branch b1_8
dzogin [Thu, 10 Dec 2009 13:41:52 +0000 (13:41 +0000)]
 Branch b1_8
 b=21409
 i=grev
 ----------------------------------------------------------------------
 Modified Files:
  Tag: b1_8
  lustre/tests/test-framework.sh

The small fix for test-framework to make code in 1.8 and HEAD consistent
----------------------------------------------------------------------

14 years agob=19689
brian [Thu, 10 Dec 2009 12:59:39 +0000 (12:59 +0000)]
b=19689
i=adilger
i=bobijam
i=nathan
o=Jim Garlick

Change tunefs.lustre and mkfs.lustre --mountfsoptions so that
exactly the specified mount options are used.
Leaving off any "mandatory" mount options is an error.
Leaving off any default mount options causes a warning, but is allowed.
Change errors=remount-ro from mandatory to default.
Sanitize the mount string before storing it.
Update man pages accordingly.

This hopefully makes two awkward situations less so:
- making errors=panic the default (before we had to append errors=remount-ro
  and hope ldiskfs parsing caused the last option to override the first)
- setting other mount options on the OST's dropped mballoc,extents
  without warning.

14 years agob=20237
grev [Thu, 10 Dec 2009 12:40:19 +0000 (12:40 +0000)]
b=20237
i=Robert.Read
more tests gather logs changes

14 years agob=21115
grev [Thu, 10 Dec 2009 12:21:09 +0000 (12:21 +0000)]
b=21115
i=Robert.Read
mount the same config on all clients

14 years agob=21115
grev [Thu, 10 Dec 2009 12:06:26 +0000 (12:06 +0000)]
b=21115
i=Robert.Read
mount the same config on all clients

14 years agoBranch b1_8
tappro [Thu, 10 Dec 2009 11:29:36 +0000 (11:29 +0000)]
Branch b1_8
b=21240

revert wrongly committed code from another patch

14 years agoBranch b1_8
tappro [Thu, 10 Dec 2009 11:08:50 +0000 (11:08 +0000)]
Branch b1_8
b=21240
i=rread, johann

don't exit from test on recovery error

14 years agob=20580
grev [Thu, 10 Dec 2009 10:58:01 +0000 (10:58 +0000)]
b=20580
o=Johann
i=Andrew.Perepechko
i=grev
directio allocates the buffer twice, reduce test filesize

14 years agoBranch HEAD
tianzy [Thu, 10 Dec 2009 06:07:03 +0000 (06:07 +0000)]
Branch HEAD
When a reply is wrong, lquota will try to remove the qunit in order
to avoid other threads hanging on it.
b=21372
i=yong.fan

14 years agob=20853,i=isaac:
isaac [Thu, 10 Dec 2009 05:15:35 +0000 (05:15 +0000)]
b=20853,i=isaac:
- Mark portal 63 as reserved for Cray DVS.

14 years agob=20580
grev [Wed, 9 Dec 2009 21:05:40 +0000 (21:05 +0000)]
b=20580
i=Johann
skip test 78 for local OST

14 years agob=21142
grev [Wed, 9 Dec 2009 20:47:51 +0000 (20:47 +0000)]
b=21142
i=Andrew.Perepechko
test_cascading_rw: return if skip

14 years agob=21142
grev [Wed, 9 Dec 2009 17:59:32 +0000 (17:59 +0000)]
b=21142
i=Andrew.Perepechko
test_cascading_rw: return if skip

14 years ago- update mxlnd to upstream revision r224.
isaac [Wed, 9 Dec 2009 16:35:32 +0000 (16:35 +0000)]
-  update mxlnd to upstream revision r224.

14 years agob=21128
pravins [Wed, 9 Dec 2009 16:18:35 +0000 (16:18 +0000)]
b=21128
i=rread
i=adilger
get stack dump on stack usage overflow.

14 years ago- upstream mxlnd now requires MX library 1.2.10+
isaac [Wed, 9 Dec 2009 15:59:00 +0000 (15:59 +0000)]
- upstream mxlnd now requires MX library 1.2.10+

14 years ago- update mxlnd to upstream revision r224.
isaac [Wed, 9 Dec 2009 15:53:36 +0000 (15:53 +0000)]
- update mxlnd to upstream revision r224.

14 years agob=20580
grev [Wed, 9 Dec 2009 11:57:27 +0000 (11:57 +0000)]
b=20580
i=Johann
skip test_78 for local OST

14 years agob=20655
deshmukh [Wed, 9 Dec 2009 09:26:07 +0000 (09:26 +0000)]
b=20655
i=sheng.yang

- Fixed the patchless client build failure for ext4 on HEAD.

- And ldiskfs will be ext4 by default on HEAD for RHEL5.

14 years agob=21224
deshmukh [Wed, 9 Dec 2009 09:17:48 +0000 (09:17 +0000)]
b=21224
i=johann
i=yong.fan

Ported the quotacheck improvement patch to HEAD.

14 years agob=21379
zhanghc [Wed, 9 Dec 2009 07:25:40 +0000 (07:25 +0000)]
b=21379
1, fix a bug in osc_create
       there is no need to proceed after the orphans are deleted, which could
   cause the oscc->oscc_next_id is increased by 1 and break the balance between
   it and the last objid stored in "mds->mds_lov_page_array"

2, change "mds_osc_obd" and "mds_osc_exp" into "mds_lov_obd" and "mds_lov_exp"
   in mds_obd

3, check the specified osc's export instead lov's export in mds_lov_get_objid

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

14 years agob=21348
panda [Tue, 8 Dec 2009 21:28:04 +0000 (21:28 +0000)]
b=21348
i=Mike Pershin
i=Alexander Zarochentsev

adding proper maintainance of the evictor list

14 years agob=20953
panda [Tue, 8 Dec 2009 21:19:07 +0000 (21:19 +0000)]
b=20953

debug patch for sanity-quota test_30

14 years agoBranch HEAD
robert.read [Tue, 8 Dec 2009 21:06:56 +0000 (21:06 +0000)]
Branch HEAD
b=18357
i=adilger
i=rread
o=simmonsja@ornl.gov

Make newer versions of gcc happy.

14 years agoBranch HEAD
zam [Tue, 8 Dec 2009 21:02:35 +0000 (21:02 +0000)]
Branch HEAD
b=21314
i=alexey.lyashkov
i=alexander.zarochentsev

keep lock reference when class_export_lock_put accesses the lock.

14 years agob=21530
brian [Tue, 8 Dec 2009 17:05:55 +0000 (17:05 +0000)]
b=21530
i=wangyb
i=yangsheng

Don't build RDAC for (patchless) clients.

14 years agoBranch b1_8
yangsheng [Tue, 8 Dec 2009 15:40:28 +0000 (15:40 +0000)]
Branch b1_8
b=21511

Update kernel to SLES11 2.6.27.39-0.3.1.

14 years agoBranch b1_8
yangsheng [Tue, 8 Dec 2009 06:25:39 +0000 (06:25 +0000)]
Branch b1_8
b=20065

i=adilger, bobijam

Add 1-second statfs caching

14 years agoBranch b1_8
liuy [Tue, 8 Dec 2009 02:32:50 +0000 (02:32 +0000)]
Branch b1_8
b=19518
i=adilger
i=bobijam

some fixes in lustre hash:
-no access to lh_rwlock if LH_REHASH is not set
-buckets are allocated one by one, to avoid unnecessary cacheline conflict
 (on bucket::lhb_rwlock) while searching.

14 years agoi=maxim,i=chris,b=21459:
isaac [Mon, 7 Dec 2009 19:31:38 +0000 (19:31 +0000)]
i=maxim,i=chris,b=21459:
- should update lp_alive for non-router peers.

14 years agoi=maxim,i=chris,b=21459:
isaac [Mon, 7 Dec 2009 14:39:47 +0000 (14:39 +0000)]
i=maxim,i=chris,b=21459:
- should update lp_alive for non-router peers.

14 years agob=21079
maxim [Sat, 5 Dec 2009 11:09:28 +0000 (11:09 +0000)]
b=21079
i=rread
i=liang
Sanity test_215 is rewritten to eliminate too optimistic assumptions
about values LNET exports to /proc/sys/lnet/*. Also, rewritten test
reflects latest modifications of /proc/sys/lnet/routers and
/proc/sys/lnet/nis (made in the scope of bug #20171).

14 years ago Branch HEAD
dzogin [Sat, 5 Dec 2009 06:19:00 +0000 (06:19 +0000)]
 Branch HEAD
 b=21513
 i=alexey.lyashkov
 i=johann
 ----------------------------------------------------------------------
 Modified Files:
  lustre/ChangeLog lustre/ptlrpc/import.c
 ----------------------------------------------------------------------
Description: Clear imp_force_reconnect correctly in ptlrpc_connect_interpret()

14 years agob=21460
jxiong [Sat, 5 Dec 2009 03:36:54 +0000 (03:36 +0000)]
b=21460
r=wangdi,eric.mei

- in ll_prepare_write, it initialises an env, but not hold a refcount against it, so it would have been freed in up-coming ll_commit_write.
- remove cui_oneshot

14 years agob=21514
brian [Fri, 4 Dec 2009 21:48:50 +0000 (21:48 +0000)]
b=21514
i=wangyb
i=yangsheng

If lbuild is interrupted (i.e. LBATS "Cancel") then it should not call
the ccache -s exit trap action because experience has shown that the
ccache binary is gone by the time lbuild gets to calling it.

Also, extend the API for the delete_exit_trap() to handle multiple
action handles in a single call.  This makes it easy to create a stack
of actions and then delete them all in a single command.

What would be even nicer would be (shell and pdsh style) glob handling
so that one could do delete_exit_trap foo[1-10], assuming one pushed 10
actions on a stack to be unwound during an operation that is supposed to
be atomic.  Or more simply delete_exit_trap foo*.

14 years agob=19616
grev [Fri, 4 Dec 2009 21:01:18 +0000 (21:01 +0000)]
b=19616
o=Walter.Poxon
i=grev
read the list of tests to except from the file if it exist

14 years agocat /tmp/msg
grev [Fri, 4 Dec 2009 20:58:44 +0000 (20:58 +0000)]
cat /tmp/msg

14 years agob=21041
brian [Fri, 4 Dec 2009 20:22:53 +0000 (20:22 +0000)]
b=21041
i=dmitry.zogin
i=johann

On RHEL5 at least, the sg_set_page autoconf macro needs the addition of
<asm/types.h> in order to properly work.

14 years agob=21516
nathan [Fri, 4 Dec 2009 19:35:03 +0000 (19:35 +0000)]
b=21516
i=rread
fix acc-sm for new lustre_rsync name
fix stripe count determination in test_7

14 years agob=21348
panda [Fri, 4 Dec 2009 11:30:30 +0000 (11:30 +0000)]
b=21348
i=Mike Pershin
i=Alexander Zarochentsev

allow adding several instances to evictor list

14 years agoBranch b1_8
yangsheng [Fri, 4 Dec 2009 10:30:49 +0000 (10:30 +0000)]
Branch b1_8
b=21375

i=johann, adilger

Cleanup for FS_HAS_FIEMAP.

14 years agob=21426
brian [Thu, 3 Dec 2009 19:44:59 +0000 (19:44 +0000)]
b=21426

Re-land now that the missing directory is created in CVS (and a .cvsignore
file placed in it so that a checkout will create the directory).

14 years ago Updated version for 1.8.2 Build 06.
stinson1947 [Thu, 3 Dec 2009 19:36:46 +0000 (19:36 +0000)]
 Updated version for 1.8.2 Build 06.

14 years agoCreate a .cvsignore for this directory.
brian [Thu, 3 Dec 2009 18:59:19 +0000 (18:59 +0000)]
Create a .cvsignore for this directory.

14 years agoBack out 20315 landing due to breakage with SLES10.
brian [Thu, 3 Dec 2009 18:44:00 +0000 (18:44 +0000)]
Back out 20315 landing due to breakage with SLES10.
I suspect that we should be using RPMs %configure macro in our lustre.spec
and we are not.

14 years agoBack out 20315 landing due to breakage with SLES10.
brian [Thu, 3 Dec 2009 18:41:43 +0000 (18:41 +0000)]
Back out 20315 landing due to breakage with SLES10.
I suspect that we should be using RPMs %configure macro in our lustre.spec
and we are not.

14 years agoBranch b1_8
johann [Thu, 3 Dec 2009 17:14:53 +0000 (17:14 +0000)]
Branch b1_8
b=20668
i=johann (patch from eeb)

Remove trailing semicolons in do{}while(0) macro definitions.

14 years agoBranch b1_8
johann [Thu, 3 Dec 2009 17:14:47 +0000 (17:14 +0000)]
Branch b1_8
b=20668
i=johann (patch from eeb)

Remove trailing semicolons in do{}while(0) macro definitions.

14 years agob=21477
panda [Thu, 3 Dec 2009 16:55:49 +0000 (16:55 +0000)]
b=21477
i=Johann

disable pinger when appropriate

14 years agob=21477
panda [Thu, 3 Dec 2009 16:52:52 +0000 (16:52 +0000)]
b=21477

i=Johann Lombardi
i=Dmitry Zogin

check connection for NULL (for the obdecho client)

14 years ago Branch b1_8
dzogin [Thu, 3 Dec 2009 16:01:37 +0000 (16:01 +0000)]
 Branch b1_8
 b=21513
 i=alexey.lyashkov
 i=johann
----------------------------------------------------------------------
 Modified Files:
  Tag: b1_8
  lustre/ChangeLog lustre/ptlrpc/import.c
----------------------------------------------------------------------
Description: Clear imp_force_reconnect correctly in ptlrpc_connect_interpret()

14 years agoBranch b1_8
tappro [Thu, 3 Dec 2009 15:03:45 +0000 (15:03 +0000)]
Branch b1_8
b=21430
i=rread
i=johann

seeds PRNG with current time on server, check export is connected before request
handling

14 years agoBranch HEAD
tappro [Thu, 3 Dec 2009 10:28:50 +0000 (10:28 +0000)]
Branch HEAD
b=21240
i=rread

don't check result of lctl because it is allowed to fail during that test

14 years agoBranch HEAD
tappro [Thu, 3 Dec 2009 10:27:01 +0000 (10:27 +0000)]
Branch HEAD
b=21430
i=adilger
i=johann

seeds PRNG with current time on server, checks export is connected during
request handling

14 years agob=21265
grev [Thu, 3 Dec 2009 09:35:57 +0000 (09:35 +0000)]
b=21265
i=Robert.Read
skip last sleep if ELAPSED + sleep > DURATION

14 years agob=21142
grev [Thu, 3 Dec 2009 08:33:40 +0000 (08:33 +0000)]
b=21142
i=Andrew.Perepechko
skip cascading_rw for NFSCLIENT mode

14 years agoBranch HEAD
yangsheng [Thu, 3 Dec 2009 06:20:14 +0000 (06:20 +0000)]
Branch HEAD
b=21375

i=robert, johann

Cleanup code.

14 years agoReverting the following patch to avoid the lbat head build failure.
deshmukh [Thu, 3 Dec 2009 05:37:23 +0000 (05:37 +0000)]
Reverting the following patch to avoid the lbat head build failure.

b=21426
[lustre-cvs] HEAD: lustre-build autoMakefile.am.toplevel lbuild

14 years agob=16721
jxiong [Thu, 3 Dec 2009 03:28:30 +0000 (03:28 +0000)]
b=16721
r=wangdi,tappro

Use global rename lock at mdt to avoid deadlock

14 years agob=21434
jxiong [Thu, 3 Dec 2009 03:22:59 +0000 (03:22 +0000)]
b=21434
r=wangdi,eric.mei

We can't use env and io in osc page operations because they are not correctly setup. To support lockless IO, I choose to set srvlock flag at page initialization time.

14 years agob=21156,i=liang:
isaac [Wed, 2 Dec 2009 23:35:59 +0000 (23:35 +0000)]
b=21156,i=liang:
- simple code cleanup.

14 years agob=21156,i=liang:
isaac [Wed, 2 Dec 2009 23:31:16 +0000 (23:31 +0000)]
b=21156,i=liang:
- simple code cleanup.

14 years agob=21426
brian [Wed, 2 Dec 2009 22:00:56 +0000 (22:00 +0000)]
b=21426
i=wangyb
i=yangsheng

Allow the easy dropping in of patches to OFED.  This is mainly meant as a
mechanism to test patches from OFA prior to them committing them to their
upstream.

We should likely never actually release a version of Lustre and OFED built
with any OFED patches.

14 years agob=20101
panda [Wed, 2 Dec 2009 21:42:20 +0000 (21:42 +0000)]
b=20101
i=Andreas Dilger
i=Robert Read

a=James Simmons

Fixed lfs to maintain old behavior and support new functionality

14 years agob=20315
brian [Wed, 2 Dec 2009 21:35:56 +0000 (21:35 +0000)]
b=20315
i=adilger
o=Christopher Morrone

Use the more standard libexecdir for scripts.

14 years agob=20315
brian [Wed, 2 Dec 2009 21:33:10 +0000 (21:33 +0000)]
b=20315
i=adilger
o=Christopher Morrone

Use the more standard libexecdir for scripts.

14 years agob=21457
brian [Wed, 2 Dec 2009 20:44:37 +0000 (20:44 +0000)]
b=21457
i=<inspected_by>
i=<inspected_by>

Refactor the tarball downloading code so that we remove the duplicated
(in fact, at least triplicated) code and avoid races between multiple
lbuilds downloading the same thing.  Now parallel lbuilds wait for the
first lbuild to complete the download and use it's product.