Whamcloud - gitweb
fs/lustre-release.git
13 years agob=22418 2.6.32 compat synchronize_rcu already defined
Brian J. Murrell [Wed, 25 Aug 2010 22:56:05 +0000 (18:56 -0400)]
b=22418 2.6.32 compat synchronize_rcu already defined

In 2.6.32 kernels when !CONFIG_TREE_PREEMPT_RCU synchronize_rcu is a #define
instead of being exported symbol.  In 2.6.33 kernels it  is an exported as
a symbol regardless of the configure option.  To handle this when the symbol
is not exported use the kernels provided define if available.  Otherwise
fallback to providing our own for older kernels.

o=brian.behlendorf
i=andreas
i=kalpak

13 years agob=20563 move definition of ll_need_32bit_api() after the one of ll_sb_info
Johann Lombardi [Thu, 26 Aug 2010 15:33:44 +0000 (17:33 +0200)]
b=20563 move definition of ll_need_32bit_api() after the one of ll_sb_info

13 years agob=18213 Handle the signal to avoid process hang.
yangsheng [Wed, 25 Aug 2010 16:22:31 +0000 (00:22 +0800)]
b=18213 Handle the signal to avoid process hang.

i=johann
i=hongchao.zhang

13 years agob=13752 Remove 2.4 kernel definition of cfs_waitq_wait_event_interruptible_timeout()
yangsheng [Wed, 25 Aug 2010 16:18:46 +0000 (00:18 +0800)]
b=13752 Remove 2.4 kernel definition of cfs_waitq_wait_event_interruptible_timeout()

i=issac

13 years agob=21610 Add lbuild-sles11sp1 file.
yangsheng [Wed, 25 Aug 2010 13:03:20 +0000 (21:03 +0800)]
b=21610 Add lbuild-sles11sp1 file.

i=johann

13 years agob=23515 recovery_*_scale tests need more than 2 clients
Elena Gryaznova [Wed, 25 Aug 2010 12:48:00 +0000 (16:48 +0400)]
b=23515 recovery_*_scale tests need more than 2 clients

i=Mikhail.Pershin

skip these tests by env for number of remote clients < 2

13 years agob=20563 add llite mount option to generate 32bit ino v2
pravin [Wed, 25 Aug 2010 11:50:43 +0000 (17:20 +0530)]
b=20563 add llite mount option to generate 32bit ino v2

i=vitaly
i=andreas
        patch adds mount option to generate 32bit ino, this can be used
for 32bit application compatibility.

13 years agob=22935 keep reference count for "lli_sai" to prevent it to be released when "statahe...
Fan Yong [Wed, 25 Aug 2010 05:15:41 +0000 (13:15 +0800)]
b=22935 keep reference count for "lli_sai" to prevent it to be released when "statahead_enter()" and "statahead_exit()"

keep reference count for "lli_sai" to prevent it to be released when "statahead_enter()" and "statahead_exit()"

i=eric.mei
i=di.wang

13 years agob=19151 use single-level test directory for large-scale test 3a
Jian Yu [Wed, 25 Aug 2010 02:42:23 +0000 (10:42 +0800)]
b=19151 use single-level test directory for large-scale test 3a

$tdir is a two-level directory which causes the mode setting issue
while running MPI programs. Let's use a single-level test directory
for large-scale test 3a.

i=grev
i=robert.read

13 years agodebian/patches and non-GA tags
Brian J. Murrell [Tue, 24 Aug 2010 21:53:17 +0000 (17:53 -0400)]
debian/patches and non-GA tags

Make the code to populate debian/patches work better with non-GA tags
such as v1_8_3_53.

13 years agob=22967 s/LB_LINUX_CONFTEST/AC_LANG_CONFTEST/
Brian J. Murrell [Tue, 24 Aug 2010 21:53:16 +0000 (17:53 -0400)]
b=22967 s/LB_LINUX_CONFTEST/AC_LANG_CONFTEST/

There was a macro update in more recent releases of autoconf that
requires us to use AC_LANG_CONFTEST instead of the LB_LINUX_CONFTEST
that we currently use.  The problem using LB_LINUX_CONFTEST causes
is that as configure is determining capabilities and setting capability
flags, these are not being used when compiling further conftest.c
programs.

So for example if a macro determines if foo is available and then sets
FOO if it is, and then a test in a subsequent macro tries to use FOO,
it will find it undefined.

i=michael.macdonald
i=andreas.dilger

13 years agob=22378 Correct MDS client stats
Dmitry Zogin [Tue, 24 Aug 2010 20:45:25 +0000 (16:45 -0400)]
b=22378 Correct MDS client stats

 Move some of the MDS stats from obdfilter layer to MDS layer.

 i=andreas.dilger
 i=amoly.liu

13 years agob=16890 llapi_quotactl man page update with additional explanation of igrace/bgrace...
Andrew Perepechko [Tue, 24 Aug 2010 20:35:54 +0000 (00:35 +0400)]
b=16890 llapi_quotactl man page update with additional explanation of igrace/bgrace/itime/btime

13 years agob=22378 Correct MDS client stats
Dmitry Zogin [Tue, 24 Aug 2010 20:30:07 +0000 (16:30 -0400)]
b=22378 Correct MDS client stats

 test_133 sanity.sh has been added

 i=andreas.dilger

13 years agob=23408 disable failure temporarily while we collect performance stats
Nathan Rutman [Tue, 24 Aug 2010 17:02:06 +0000 (10:02 -0700)]
b=23408 disable failure temporarily while we collect performance stats

i=vitaly

13 years agob=23595 fix broken patch
Johann Lombardi [Tue, 24 Aug 2010 20:09:06 +0000 (22:09 +0200)]
b=23595 fix broken patch

13 years agob=23595 return registration errors
Nathan Rutman [Mon, 23 Aug 2010 17:59:22 +0000 (10:59 -0700)]
b=23595 return registration errors

i=johann
i=vitaly

13 years agob=21174 fix tiny nit in previous landing
Johann Lombardi [Tue, 24 Aug 2010 16:26:33 +0000 (18:26 +0200)]
b=21174 fix tiny nit in previous landing

13 years agob=23216 a fix for a memory leak in echo_commitrw
Andrew Perepechko [Fri, 20 Aug 2010 21:51:56 +0000 (01:51 +0400)]
b=23216 a fix for a memory leak in echo_commitrw

i=Andreas Dilger
i=Zhen Liang

13 years agob=23573 skip conf-sanity fs2 tests for HARD failure mode
Elena Gryaznova [Fri, 20 Aug 2010 21:46:06 +0000 (01:46 +0400)]
b=23573 skip conf-sanity fs2 tests for HARD failure mode

i=Andrew.Perepechko

13 years agob=20407 replay-ost-single: do not skip for HARD mode and mixed_ost_devs
Elena Gryaznova [Fri, 20 Aug 2010 21:40:41 +0000 (01:40 +0400)]
b=20407 replay-ost-single: do not skip for HARD mode and mixed_ost_devs

i=Brian.Murrell

13 years agob=23595 fix conf-sanity 57 for remote ost
Nathan Rutman [Fri, 20 Aug 2010 21:16:04 +0000 (14:16 -0700)]
b=23595 fix conf-sanity 57 for remote ost

i=vitaly.fertman

13 years agob=22934 fix writeconf, redeux
Nathan Rutman [Fri, 20 Aug 2010 18:54:56 +0000 (11:54 -0700)]
b=22934 fix writeconf, redeux

13 years agob=21174 allow quotacheck over OSTs with sparse indices
Andrew Perepechko [Fri, 20 Aug 2010 11:15:25 +0000 (13:15 +0200)]
b=21174 allow quotacheck over OSTs with sparse indices

i=Elena Gryznova

13 years agob=23335 Allocate echo objects that can be mapped to a valid FID
Elena Gryaznova [Wed, 18 Aug 2010 13:31:59 +0000 (17:31 +0400)]
b=23335 Allocate echo objects that can be mapped to a valid FID

With the change to using valid FIDs for all OST objects in bug 19427,
the echo objid needs to be below 2^32, because regular FID numbers
are limited to 2^32 objects in a single sequence number.

o=andreas.dilger
i=aleksandr.guzovskiy
i=mikhail.pershin

13 years agob=23278 replay-single test 86 does not remount client
Elena Gryaznova [Tue, 17 Aug 2010 15:46:03 +0000 (19:46 +0400)]
b=23278 replay-single test 86 does not remount client

o=Oleg.Drokin
i=Elena.Gryaznova

13 years agob=20407 TF: "HARD" failovers with multiple targets per server
Elena Gryaznova [Mon, 16 Aug 2010 16:23:10 +0000 (20:23 +0400)]
b=20407 TF: "HARD" failovers with multiple targets per server

i=Brian.Murrell
i=Li.Wei

13 years agob=21174 allow quotacheck over OSTs with sparse indices
Andrew Perepechko [Mon, 16 Aug 2010 12:00:02 +0000 (16:00 +0400)]
b=21174 allow quotacheck over OSTs with sparse indices

i=Johann Lombardi
i=ZhiYong Tian

13 years agob=21610 Update changelog & which_kernel for sles11 sp1.
yangsheng [Mon, 16 Aug 2010 05:07:58 +0000 (23:07 -0600)]
b=21610 Update changelog & which_kernel for sles11 sp1.

13 years agob=21610 Kernel update for SLES11 SP1. v1_8_4_50
yangsheng [Fri, 13 Aug 2010 20:26:18 +0000 (04:26 +0800)]
b=21610 Kernel update for SLES11 SP1.

Update SLES11 SP1 to 2.6.32.13-0.5.1.

i=adilger, johann, kalpak
i=girish, rahul, whitebear
i=brian, wangyb

13 years agob=23206 performance-sanity test_8 FAIL
Dmitry Zogin [Wed, 11 Aug 2010 14:27:14 +0000 (10:27 -0400)]
b=23206 performance-sanity test_8 FAIL

 Debug patch

13 years agob=23409 add -i to the setstripe usage and man page
Andreas Dilger [Tue, 10 Aug 2010 05:57:50 +0000 (01:57 -0400)]
b=23409 add -i to the setstripe usage and man page

Add the "-i" option to the "lfs setstripe" usage and man page.
Fix nroff formatting in the "lfs setstripe" and "lfs getstripe".

i=sheila.barthel

13 years agob=21720 fix test 18 to interleave tests increase pass margin
Nathan Rutman [Mon, 9 Aug 2010 19:20:29 +0000 (12:20 -0700)]
b=21720 fix test 18 to interleave tests increase pass margin

i=rread

backport 2.0's fractional-second createmany

13 years agob=23270 simplify "lctl osts" (llapi_ostlist) code
Andreas Dilger [Mon, 9 Aug 2010 14:25:34 +0000 (10:25 -0400)]
b=23270 simplify "lctl osts" (llapi_ostlist) code

Simplify "lfs osts" command so that it avoids the filesystem traversal code
entirely, and just calls setup_osts() to print the OST list.

13 years agob=23316 BUG: soft lockup - CPU#2 stuck for 10s! [ll_cfg_requeue:2851]
Dmitry Zogin [Mon, 9 Aug 2010 12:59:45 +0000 (08:59 -0400)]
b=23316 BUG: soft lockup - CPU#2 stuck for 10s! [ll_cfg_requeue:2851]

 Use SHARED_DIR_LOGS in error_noexit().
 i=grev

13 years agob=22891 Objects not getting deleted for files which have been removed
Dmitry Zogin [Mon, 9 Aug 2010 12:51:27 +0000 (08:51 -0400)]
b=22891 Objects not getting deleted for files which have been removed

 ll_have_md_lock() should differentiate between CR and CW OPEN locks.
 Also sanityN.sh test_36b was added.

 i=oleg.drokin
 i=johann.lombardi

13 years agob=22979 ignore the case of zero unused lock before recovery for replay-single test_85
Fan Yong [Mon, 9 Aug 2010 09:02:57 +0000 (17:02 +0800)]
b=22979 ignore the case of zero unused lock before recovery for replay-single test_85

Ignore the case of zero unused lock before recovery for replay-single test_85.

i=johann.lombardi
i=Hongchao.zhang

13 years agob=22596 Avoid test failed on single OST.
yangsheng [Mon, 9 Aug 2010 08:54:59 +0000 (16:54 +0800)]
b=22596 Avoid test failed on single OST.

i=adilger, grev

Use OSTCOUNT instead hard code 2 to adapt single OST case. This test
isn't necessary OSTCOUNT > 1.

13 years agob=13585 Remove i_filterdata patches.
yangsheng [Mon, 9 Aug 2010 08:51:31 +0000 (16:51 +0800)]
b=13585 Remove i_filterdata patches.

i=adilger

13 years agob=22514 Update to latest RHEL5.5 kernel.
yangsheng [Mon, 9 Aug 2010 08:45:33 +0000 (16:45 +0800)]
b=22514 Update to latest RHEL5.5 kernel.

13 years agob=22906 mke2fs needs ~16TB LUNs to be 16TB-1 block
Andreas Dilger [Fri, 6 Aug 2010 19:43:04 +0000 (13:43 -0600)]
b=22906 mke2fs needs ~16TB LUNs to be 16TB-1 block

Adjust LUNs at or just over 16TB to be 1 block below 16TB, to avoid
problems with current (1.41) mke2fs being unhappy with this size.

i=zhiqi.tao
i=girish.shilamkar

13 years agob=13698 LL_IOC_RECREATE_FID (1.8)
Vladimir Saveliev [Sun, 1 Aug 2010 05:21:42 +0000 (09:21 +0400)]
b=13698 LL_IOC_RECREATE_FID (1.8)

define new ioctl for object replicate
it uses IDIF FID instead of truct ll_recreate_obj
old LL_IOC_RECREATE is kept for compatibility

i=andreas.dilger
i=di.wang

13 years agob=22481 man page for lfs_migrate
Andreas Dilger [Wed, 4 Aug 2010 19:56:43 +0000 (13:56 -0600)]
b=22481 man page for lfs_migrate

Add a manual page for lfs_migrate.
Minor formatting changes of the lfs_migrate::usage() message.

13 years agob=12197 use absolute path in mtab
minhdiep [Tue, 3 Aug 2010 21:53:36 +0000 (15:53 -0600)]
b=12197 use absolute path in mtab

i=johann

The mount point passed into mtab should be the absolute path
Convert the mount point to real path.

13 years agob=22944 skip conf-sanity test 16 in interop mode.
Eric Mei [Tue, 3 Aug 2010 13:36:58 +0000 (07:36 -0600)]
b=22944 skip conf-sanity test 16 in interop mode.

r=andreas.dilger
r=jian.yu
r=grev

13 years agob=23234 use a regular expression to parse ip_addr from ret_str in lc_net
Andrew Perepechko [Mon, 2 Aug 2010 19:43:38 +0000 (23:43 +0400)]
b=23234 use a regular expression to parse ip_addr from ret_str in lc_net

lc_net does not parse unexpected output from pdsh well

a=Chris Horn (CRAY)
i=Elena Gryaznova
i=Andrew Perepechko

13 years agob=23310 Partly matching the out message against the expected one to resolve the diffe...
Fan Yong [Mon, 2 Aug 2010 15:40:28 +0000 (23:40 +0800)]
b=23310 Partly matching the out message against the expected one to resolve the different output message format for getfacl/setfal on different Linux distributions

i=tappro
i=wangyb

Partly matching the out message against the expected one to resolve the different output message format for getfacl/setfal on different Linux distributions.

13 years agob=22107 pin object's inode in memory to avoid certain timeouts
Andrew Perepechko [Sat, 31 Jul 2010 20:30:34 +0000 (00:30 +0400)]
b=22107 pin object's inode in memory to avoid certain timeouts

i=Andreas Dilger
i=Johann Lombardi

13 years agob=14242 test_6g fails when b_release_1_6_4 is run on Cray XT3
Elena Gryaznova [Fri, 30 Jul 2010 15:00:45 +0000 (19:00 +0400)]
b=14242 test_6g fails when b_release_1_6_4 is run on Cray XT3

i=Andreas.Dilger

acceptance-small test-framework RUNAS_GID changes

13 years agob=21760 Application hung in direct I/O
Dmitry Zogin [Fri, 30 Jul 2010 13:48:09 +0000 (09:48 -0400)]
b=21760 Application hung in direct I/O

 Make sure that the bulk is aborted, if a request has been aborted in flight.
 Call ptlrpc_abort_bulk() out of ptlrpc_check_set()

 i=oleg.drokin
 i=andrew.perepechko

13 years agoRevert "b=23139 give the required grant for reconnection"
Johann Lombardi [Mon, 2 Aug 2010 07:49:53 +0000 (09:49 +0200)]
Revert "b=23139 give the required grant for reconnection"

This reverts commit 307f1ef16b4f32b9deeefff4b0aa5a1f0f0d2efa.
Revert patch from bug 23139 since it causes build failure on i686
and it also contains a bogus LASSERT.

13 years agoUpdate changelog section to 1.8.5
Johann Lombardi [Fri, 30 Jul 2010 13:59:23 +0000 (15:59 +0200)]
Update changelog section to 1.8.5

13 years agob=23139 give the required grant for reconnection
hongchao.zhang [Thu, 22 Jul 2010 04:25:27 +0000 (12:25 +0800)]
b=23139 give the required grant for reconnection

if a client is reconnecting to the filter, the grant
required by the client should be honored

i=oleg.drokin
i=eric.mei

13 years agob=20433 we should recycle dentries and inodes if only cancelling locks existing
Landen [Fri, 30 Jul 2010 07:35:08 +0000 (15:35 +0800)]
b=20433 we should recycle dentries and inodes if only cancelling locks existing

i=green
i=adilger

13 years agob=23382 t-f: do_nodes(): wrong sed RE
Elena Gryaznova [Thu, 29 Jul 2010 12:25:57 +0000 (16:25 +0400)]
b=23382 t-f: do_nodes(): wrong sed RE

i=Andrew.Perepechko

13 years agob=23352 print the arrival time of late cancel RPC
hongchao.zhang [Wed, 21 Jul 2010 05:35:15 +0000 (13:35 +0800)]
b=23352 print the arrival time of late cancel RPC

in "ldlm_cancel_handler", print the arrival time of RPCs, which
cancel the lock but the corresponding export has disappeared

i=nathan.rutman
i=hongchao.zhang

13 years agob=23371 Aviod deadlock with i_data_sem.
yangsheng [Thu, 29 Jul 2010 09:53:54 +0000 (17:53 +0800)]
b=23371 Aviod deadlock with i_data_sem.

i=adilger
i=andrew

ext4_ext_walk_space() take i_data_sem at present. So we have to detect this case to avoid deadlock.

13 years agob=23064 improve bdi usage
yangsheng [Thu, 29 Jul 2010 09:49:18 +0000 (17:49 +0800)]
b=23064 improve bdi usage

i=andreas
i=kalpak
i=johann

13 years agob=23439 fix some recovery debug messages
Johann Lombardi [Thu, 29 Jul 2010 08:56:33 +0000 (10:56 +0200)]
b=23439 fix some recovery debug messages

i=andrew

13 years agob=23228 handle any previous state in test_59
Nathan Rutman [Wed, 28 Jul 2010 23:02:38 +0000 (16:02 -0700)]
b=23228 handle any previous state in test_59

i=johann

13 years agob=22632 also build mptlinux on SLES11 1.8.4 v1_8_4 v1_8_4_RC2
Johann Lombardi [Fri, 23 Jul 2010 22:49:59 +0000 (00:49 +0200)]
b=22632 also build mptlinux on SLES11

i=johann

Add SLES11 to the list of platforms we build mptlinux on.
FWIW, RDAC fails to build on SLES11 so it has not been added here.

13 years agob=21587 add debug patch
Johann Lombardi [Fri, 23 Jul 2010 22:46:54 +0000 (00:46 +0200)]
b=21587 add debug patch

i=andrew

13 years agob=23402 mmp_init() fix
E.Gryaznova [Fri, 23 Jul 2010 22:39:17 +0000 (00:39 +0200)]
b=23402 mmp_init() fix

i=jian.yu

13 years agoadd changelog entries
Johann Lombardi [Fri, 23 Jul 2010 22:35:11 +0000 (00:35 +0200)]
add changelog entries

13 years agob=23368 fix conflicting ext4 mount flags
Johann Lombardi [Fri, 23 Jul 2010 22:25:55 +0000 (00:25 +0200)]
b=23368 fix conflicting ext4 mount flags

i=adilger

13 years agob=23368 fix bug in mainline rhel5/ext4 causing slab corruption when mount failed
Johann Lombardi [Fri, 23 Jul 2010 21:57:44 +0000 (23:57 +0200)]
b=23368 fix bug in mainline rhel5/ext4 causing slab corruption when mount failed

i=adilger

13 years agob=23368 disable DELALLOC by default for RHEL5/ext4
Johann Lombardi [Fri, 23 Jul 2010 21:54:55 +0000 (23:54 +0200)]
b=23368 disable DELALLOC by default for RHEL5/ext4

i=yangsheng

As for SLES11, we should disabled delayed allocation by default
since it is known to be buggy.

13 years agoset version to 1.8.4 for rc1 v1_8_4_RC1
johann [Fri, 9 Jul 2010 22:36:13 +0000 (00:36 +0200)]
set version to 1.8.4 for rc1

13 years agoset expected release date in the changelogs
johann [Fri, 9 Jul 2010 22:34:57 +0000 (00:34 +0200)]
set expected release date in the changelogs

13 years agoUpdate copyright
johann [Fri, 9 Jul 2010 22:26:48 +0000 (00:26 +0200)]
Update copyright

13 years agob=23305 fix changelog entry to point to public bug
johann [Fri, 9 Jul 2010 22:15:29 +0000 (00:15 +0200)]
b=23305 fix changelog entry to point to public bug

13 years agoMove ext4-remove-extents-warning-rhel5.patch to correct place
johann [Fri, 9 Jul 2010 21:12:30 +0000 (23:12 +0200)]
Move ext4-remove-extents-warning-rhel5.patch to correct place

13 years agob=23302 Remove "extents disabled" warning
Girish Shilamkar [Fri, 9 Jul 2010 10:41:48 +0000 (16:11 +0530)]
b=23302 Remove "extents disabled" warning

i=johann

13 years agob=23122 Change config check for sles11 sp1.
yangsheng [Fri, 9 Jul 2010 14:47:09 +0000 (22:47 +0800)]
b=23122 Change config check for sles11 sp1.

i=johann

13 years agob=22771 add changelog entry
johann [Fri, 9 Jul 2010 11:14:20 +0000 (13:14 +0200)]
b=22771 add changelog entry

13 years agob=22771 Patches to disable mb_cache
Girish Shilamkar [Thu, 8 Jul 2010 18:34:52 +0000 (00:04 +0530)]
b=22771 Patches to disable mb_cache

i=adilger

13 years agoAdd missing changelog entries
johann [Thu, 8 Jul 2010 22:04:08 +0000 (00:04 +0200)]
Add missing changelog entries

13 years agob=19529 Avoid deadlock for local client writes.
Dmitry Zogin [Thu, 8 Jul 2010 17:54:32 +0000 (13:54 -0400)]
b=19529 Avoid deadlock for local client writes.

Check the OBD_BRW_MEMALLOC flag correctly in the remote buffer.

 i=johann
 i=andreas.dilger

13 years agob=23235 Reintroduce ext4_dquot_initialize() and ext4_dquot_drop() to avoid deadlock.
yangsheng [Thu, 8 Jul 2010 14:10:04 +0000 (22:10 +0800)]
b=23235 Reintroduce ext4_dquot_initialize() and ext4_dquot_drop() to avoid deadlock.

The problem is that lustre already starts a transaction before calling the ldiskfs/quota functions
most of the time, so we still need quota drop & initialize to start the transaction first to avoid
ordering issue with the other quota operations.

i=johann
i=landen
i=panda

13 years agoBump version to 1.8.3.58 v1_8_3_58
johann [Wed, 7 Jul 2010 21:02:37 +0000 (23:02 +0200)]
Bump version to 1.8.3.58

13 years agob=23216 a fix for a possible memory leak in ldiskfs_mb_load_buddy
Andrew Perepechko [Wed, 7 Jul 2010 20:59:09 +0000 (00:59 +0400)]
b=23216 a fix for a possible memory leak in ldiskfs_mb_load_buddy

i=Alex Zhuravlev
i=Johann Lombardi

13 years agob=23175 disable lockless truncate
johann [Wed, 7 Jul 2010 17:20:50 +0000 (19:20 +0200)]
b=23175 disable lockless truncate

lockless truncate is suspected to cause bug 23175. Disable it by
default for now to see if the problem happens again.

13 years agob=23139 workaround to avoid assertion in osc_init_grant
hongchao.zhang [Wed, 30 Jun 2010 15:57:49 +0000 (23:57 +0800)]
b=23139 workaround to avoid assertion in osc_init_grant

workaround for 1.6 servers which don't have
the patch from bug20278 applied

i=oleg.drokin
i=eric.mei

13 years agob=23216 prevent memory leak in ost_brw_read and ost_brw_write
Andrew Perepechko [Tue, 6 Jul 2010 21:51:13 +0000 (01:51 +0400)]
b=23216 prevent memory leak in ost_brw_read and ost_brw_write

i=Alexander Zarochentsev
i=Oleg Drokin

13 years agob=21980 cache `ll_obdo_cache': Can't free all objects
Dmitry Zogin [Mon, 5 Jul 2010 23:17:46 +0000 (19:17 -0400)]
b=21980 cache `ll_obdo_cache': Can't free all objects

Always use OBDO_ALLOC/FREE for obdo allocations to prevent slab fragmentation.
 Other related fixes.

 o=johann
 i=di.wang
 i=dmitry.zoguine

13 years agob=20563 flush seq fix
pravin [Mon, 5 Jul 2010 14:24:34 +0000 (19:54 +0530)]
b=20563 flush seq fix

i=alexander.zarochentsev
i=rahul

always flush seq on OBD_NOTIFY_INACTIVE event.

13 years agob=23235 Init lnb[n].lnb_grant_used on recoverable_resent.
yangsheng [Mon, 5 Jul 2010 13:22:48 +0000 (21:22 +0800)]
b=23235 Init lnb[n].lnb_grant_used on recoverable_resent.

i=green
i=wangdi

13 years agob=23237 mount.lustre dies with SIGSEGV: Unable to read 1.8 config /tmp/lustre_tmp...
Dmitry Zogin [Fri, 2 Jul 2010 22:11:37 +0000 (18:11 -0400)]
b=23237 mount.lustre dies with SIGSEGV: Unable to read 1.8 config /tmp/lustre_tmp.IfgmBK/mountdata

 Do not try to close non-open file and return the error in get_mountdata()

 i=andreas.dilger
 i=johann

13 years agob=23175 Use MPI_Bcast to send random value to all ranks The file size will be differe...
minhdiep [Fri, 2 Jul 2010 21:23:56 +0000 (15:23 -0600)]
b=23175 Use MPI_Bcast to send random value to all ranks The file size will be different if each rank generate its own chunksize. We need rank 0 to generate the random and pass it to all ranks i=Johann

13 years agoBump version to 1.8.3.57 v1_8_3_57
Johann Lombardi [Thu, 1 Jul 2010 13:16:56 +0000 (15:16 +0200)]
Bump version to 1.8.3.57

13 years agob=23192 ping_evictor_main() should skip export eviction only we are still in recovery
Johann Lombardi [Wed, 30 Jun 2010 23:00:27 +0000 (01:00 +0200)]
b=23192 ping_evictor_main() should skip export eviction only we are still in recovery

i=panda
i=tappro

target_recovery_check_and_stop() now returns 1 when the ost is not in
recovery. This confuses the ping evictor which decides not to process
export that needs to be evicted.

13 years agob=22360 Check for file_operations .flush fl_owner_t id parameter
Andreas Dilger [Wed, 30 Jun 2010 21:09:54 +0000 (23:09 +0200)]
b=22360  Check for file_operations .flush fl_owner_t id parameter

i=johann

Since 2.6.18 (commit 75e1fcc0b18df0a65ab113198e9dc0e98999a08c) the
file_operations .flush() method has taken an fl_owner_t id parameter.
This is backported to some older vendor kernels, so a simple kernel
version check, as usual, is not sufficient to determine whether this
parameter is present or not.

13 years agob=21528 move "not all requested locks are canceled" message to dlmtrace
Wally Wang [Wed, 30 Jun 2010 14:16:43 +0000 (16:16 +0200)]
b=21528 move "not all requested locks are canceled" message to dlmtrace

13 years agob=22658 a conf-sanity test case for the proper missing llogs handling
Andrew Perepechko [Tue, 29 Jun 2010 15:33:51 +0000 (19:33 +0400)]
b=22658 a conf-sanity test case for the proper missing llogs handling

a=Johann Lombardi

i=Johann Lombardi
i=Andrew Perepechko

13 years agoremove LC_EXPORT___IGET since it is no longer used
Johann Lombardi [Wed, 30 Jun 2010 13:01:55 +0000 (15:01 +0200)]
remove LC_EXPORT___IGET since it is no longer used

./configure: line 11646: LC_EXPORT___IGET: command not found

13 years agob=23210 don't update obd->obd_osfs if target is gone already
Johann Lombardi [Wed, 30 Jun 2010 08:53:33 +0000 (10:53 +0200)]
b=23210 don't update obd->obd_osfs if target is gone already

i=wangdi
i=landen

lov_disconnect() can clean up lov->lov_tgts while the statfs interpret
routine of an rpc in flight has not been executed yet.

13 years agob=23131 issue OBD_NOTIFY_CREATE event in lov
Johann Lombardi [Wed, 30 Jun 2010 08:28:46 +0000 (10:28 +0200)]
b=23131 issue OBD_NOTIFY_CREATE event in lov

i=nathan
i=andrew

(mds_lov.c:573:mds_lov_update_desc()) Process entered
(obd_class.h:376:obd_get_info()) Process entered
(obd_class.h:378:obd_get_info()) obd_get_info: NULL export
(obd_class.h:378:obd_get_info()) Process leaving (rc=18446744073709551597 : -19
ffffffffffffffed)

The problem is that mds_lov_connect() calls obd_notify(OBD_NOTIFY_CREATE)
*before* obd_connect(...,&mds->mds_lov_exp). Since mds_notify(OBD_NOTIFY_CREATE)
requires mds->mds_lov_exp which is not yet initialized, it fails.

13 years agob=23192 fix race when ping evictor and srv thread executes target_recovery_check_and_...
Johann Lombardi [Tue, 29 Jun 2010 11:59:03 +0000 (13:59 +0200)]
b=23192 fix race when ping evictor and srv thread executes target_recovery_check_and_stop() simultaneously

i=tappro
i=andrew

target_recovery_expired() wakes up both the ping evictor and the service thread
waiting in process_recovery_queue() which can race in
target_recovery_check_and_stop().

13 years agob=23185 check for both arches
Brian J. Murrell [Mon, 28 Jun 2010 16:33:30 +0000 (12:33 -0400)]
b=23185 check for both arches

When we build our version of the SLES kernel, we optimize it for i686
whereas the SUSE kernel is i386.  The actual arch makes a difference in
where the Module.symvers can be found, so just look in both locations
to cover both the upstream vendor kernel as well as our patched kernel.

13 years agob=17485 fix replay-single test 86 to remount client at the end
ZhangHongChao [Mon, 28 Jun 2010 16:25:42 +0000 (18:25 +0200)]
b=17485 fix replay-single test 86 to remount client at the end

i=johann