Whamcloud - gitweb
fs/lustre-release.git
13 years agob=24388 remove files inadvertently added by previous commit
Johann Lombardi [Thu, 27 Jan 2011 13:54:48 +0000 (14:54 +0100)]
b=24388 remove files inadvertently added by previous commit

13 years agob=24388 sgpdd-survey fix: use node_var_name () for variables
Elena Gryaznova [Thu, 27 Jan 2011 13:52:47 +0000 (14:52 +0100)]
b=24388 sgpdd-survey fix: use node_var_name () for variables

i=andrew

13 years agob=21776 Set PF_MEMALLOC on outgoing path to prevent deadlock on memory allocation...
Isaac Huang [Thu, 27 Jan 2011 13:49:27 +0000 (14:49 +0100)]
b=21776 Set PF_MEMALLOC on outgoing path to prevent deadlock on memory allocation under pressure

i=johann
i=dmitry
i=maxim

13 years agob=22980 init_logging does not exist in 1.8
Johann Lombardi [Thu, 27 Jan 2011 13:38:37 +0000 (14:38 +0100)]
b=22980 init_logging does not exist in 1.8

13 years agob=24417 Update Build-Depends
Brian J. Murrell [Wed, 26 Jan 2011 19:38:18 +0000 (14:38 -0500)]
b=24417 Update Build-Depends

- remove texlive-latex-recommended as a build requirement
- add missing "| automake1.7 | automake1.8 | automake1.9" to
  debian/control.main

Issue: LU-52
Change-Id: I53b54a068e2dc5cf8704235031ca198784bab6c9

13 years agob=24416 debian packaging fixes
Brian J. Murrell [Wed, 26 Jan 2011 19:38:17 +0000 (14:38 -0500)]
b=24416 debian packaging fixes

- don't make a patch out of anything in /debian
- exclude noise files from the debian built source tarball
- fake debian/patche{s,d} for make dist
- a few more reasons to run autogen.sh
- figure out if dist tarball needs autogen.shs and include it if so
- look for and run autogen.sh in the build subdir
- make debdiff as part of make dist
- add a debian/source/format file
- mv the orig tarball and the debdiff to the debs dir
- don't try to dist /debian for non-dpkg-using build targets

Issue: LU-51

13 years agob=24413 fix for automake > 1.9.6
Brian J. Murrell [Wed, 26 Jan 2011 16:05:41 +0000 (11:05 -0500)]
b=24413 fix for automake > 1.9.6

We seem to be using a Makefile variable that does not exist in more
recent versions of automake.  This fixes that problem.

i=oleg
v=hudson@whamcloud

Issue: LU-48
Change-Id: I886b54583b444c4911e71ae6ba6836a8ce075dd5

13 years agoRevert "b=24413 fix for automake > 1.9.6"
Terry Rutledge [Wed, 26 Jan 2011 19:02:44 +0000 (11:02 -0800)]
Revert "b=24413 fix for automake > 1.9.6"

This reverts commit 9aaebdf27eefe38f846b89ea738dd0f4f3e16d8e.

13 years agob=24413 fix for automake > 1.9.6
Brian J. Murrell [Wed, 26 Jan 2011 16:05:41 +0000 (11:05 -0500)]
b=24413 fix for automake > 1.9.6

We seem to be using a Makefile variable that does not exist in more
recent versions of automake.  This fixes that problem.

i=oleg
v=hudson@whamcloud

Issue: LU-48
Change-Id: I886b54583b444c4911e71ae6ba6836a8ce075dd5

13 years agob=22980 Support unlocked_ioctl
Niu Yawei [Wed, 26 Jan 2011 07:28:59 +0000 (23:28 -0800)]
b=22980 Support unlocked_ioctl

Adding 'unlocked_ioctl' for performance sensitive ioctls, such as OBD_IOC_BRW_READ/WRITE

i=andreas
i=johann

Change-Id: I963aaad6bf9e2fdf9b70f15659e5ad47180e0d5a

13 years agob=24320 do not fork a new thread in mem pressure
Vitaly Fertman [Wed, 26 Jan 2011 15:22:15 +0000 (18:22 +0300)]
b=24320 do not fork a new thread in mem pressure

i=johann
i=green

we already check for PF_MEMALLOC in ldlm shrinker and pass this flag
to the blocking thread, but a new thread start was still done with
no check for this flag.

13 years agob=24245 fix SA perf test to support SA disabled by default
Johann Lombardi [Tue, 25 Jan 2011 17:07:20 +0000 (18:07 +0100)]
b=24245 fix SA perf test to support SA disabled by default

13 years agob=17275 make lustre client less verbose at startup time for Cray
Bob Glossman [Mon, 24 Jan 2011 14:17:35 +0000 (15:17 +0100)]
b=17275 make lustre client less verbose at startup time for Cray

i=johann

13 years agob=24360 fix NULL pointer deref in mds_verify_child() when ll_lookup_one_len() fails
niu [Mon, 24 Jan 2011 13:10:07 +0000 (14:10 +0100)]
b=24360 fix NULL pointer deref in mds_verify_child() when ll_lookup_one_len() fails

i=johann

13 years agob=20563 Fix fid_flatten() after 1 trillion SEQ numbers
Andreas Dilger [Tue, 30 Nov 2010 23:22:14 +0000 (16:22 -0700)]
b=20563 Fix fid_flatten() after 1 trillion SEQ numbers

Fix the fid_flatten() function to properly handle FID mapping to
64-bit inode numbers, after the first 1 trillion SEQ numbers have
been granted out.  Even with CMD this would only happen after
1024 MDTs have each had 1B client mounts, so there is little risk
of introducing collisions as a result of this change, and at worst
this is a client-local phenomenon that is not persistent.

13 years agob=20563 Fix fid_flatten32() to not lose OID bits
Andreas Dilger [Tue, 30 Nov 2010 22:58:29 +0000 (15:58 -0700)]
b=20563 Fix fid_flatten32() to not lose OID bits

The original implementation of fid_flatten32() was broken due to an
error in the shift calculation (note to self - "0x00" is 8 bits, not
16 bits).  This could negatively impact 32-bit clients that were
creating more than 64k files in the same directory.

This 32-bit inode number is visible only within a single client mount,
is not used in any persistent storage, and only if a 2.x server is in
use (which is basically none today) by a 32-bit client, so there is no
issue to change it at this time.

13 years agob=22660 Return kernel's locking return code to when lustre reports success
Joseph Herring [Mon, 24 Jan 2011 12:58:03 +0000 (13:58 +0100)]
b=22660 Return kernel's locking return code to when lustre reports success

i=johann
i=andrew

13 years agob=23352 modified value of at_min is not taken into account
Zhang HongChao [Mon, 24 Jan 2011 12:53:25 +0000 (13:53 +0100)]
b=23352 modified value of at_min is not taken into account

i=Johann
i=Eri Mei

13 years agob=22378 Correct MDS client stats
Dmitry Zogin [Mon, 18 Oct 2010 19:20:47 +0000 (15:20 -0400)]
b=22378 Correct MDS client stats

sanity test_133b fails with "The getattr counter on mds is wrong"
message.

i=Andrew Perepechko

13 years agob=15962 disable statahead by default due to important races found in the code
Johann Lombardi [Thu, 20 Jan 2011 13:15:47 +0000 (14:15 +0100)]
b=15962 disable statahead by default due to important races found in the code

13 years agob=22882 MMP might sleep negative time
Bernd Schubert [Thu, 20 Jan 2011 12:39:18 +0000 (13:39 +0100)]
b=22882 MMP might sleep negative time

i=adilger
i=girish

13 years agob=21456 Patch to support lnet v1 pings in 'lctl ping'
Christopher Morrone [Thu, 20 Jan 2011 12:24:48 +0000 (13:24 +0100)]
b=21456 Patch to support lnet v1 pings in 'lctl ping'

i=maxim
i=isaac

13 years agobump version to 1.8.5.53
Johann Lombardi [Wed, 19 Jan 2011 14:14:22 +0000 (15:14 +0100)]
bump version to 1.8.5.53

13 years agob=23988 Remove sd iostats patch from sles11 patch series
Johann Lombardi [Wed, 19 Jan 2011 14:12:17 +0000 (15:12 +0100)]
b=23988 Remove sd iostats patch from sles11 patch series

13 years agoUpdated version for 1.8.6 build 02. 1.8.5.52 v1_8_5_52
Terry Rutledge [Tue, 11 Jan 2011 16:37:12 +0000 (08:37 -0800)]
Updated version for 1.8.6 build 02.

13 years agoadd missing changelog entries
Johann Lombardi [Fri, 7 Jan 2011 08:26:20 +0000 (09:26 +0100)]
add missing changelog entries

13 years agob=24039 actually add exit_traps.sh to EXTRA_DIST
Andreas Dilger [Thu, 6 Jan 2011 04:41:02 +0000 (21:41 -0700)]
b=24039 actually add exit_traps.sh to EXTRA_DIST

13 years agob=23122 make exit_traps.sh executable
Andreas Dilger [Thu, 6 Jan 2011 04:41:01 +0000 (21:41 -0700)]
b=23122 make exit_traps.sh executable

While bug 24093 added exit_traps.sh to the make dist list, it is
not an executable file to start with.  Fix this in the git repo.

13 years agob=24093 not all build files/scripts being distributed
Brian J. Murrell [Fri, 26 Nov 2010 21:08:25 +0000 (05:08 +0800)]
b=24093 not all build files/scripts being distributed

Some files that need to be are not being included in the tarball
when make dist is being run.

i=cliff
i=minh

13 years agob=24087 reverse order of $LINUX{,_OBJ}/include
Brian J. Murrell [Fri, 26 Nov 2010 21:08:26 +0000 (05:08 +0800)]
b=24087 reverse order of $LINUX{,_OBJ}/include

It is important that /usr/src/linux-...-obj/include is searched for
includes before /usr/src/linux-.../include so that the inclusion of
"include/linux/autoconf.h" picks up the one for the kernel we are
trying to build against, and not the one for the currently running
kernel, which is what is in /usr/src/linux-.../ copy is.

i=wangyb
i=cliff

13 years agob=24294 test_pios: take the ost-s sizes into account
Elena Gryaznova [Fri, 24 Dec 2010 14:56:02 +0000 (17:56 +0300)]
b=24294 test_pios: take the ost-s sizes into account

i=Vitaly.Fertman

remove obsolete workaround bug19657 part

13 years agob=23793 MOUNTOPT "-o" cleanup
Elena Gryaznova [Thu, 23 Dec 2010 20:37:13 +0000 (23:37 +0300)]
b=23793 MOUNTOPT "-o" cleanup

i=Minh.Diep

13 years agob=23051 improve summary of acc-sm to include test times
Elena Gryaznova [Wed, 22 Dec 2010 22:43:49 +0000 (01:43 +0300)]
b=23051 improve summary of acc-sm to include test times

i=Andrew.Perepechko

acceptance-small test suites name cleanup:
- rename sanityN -> sanityn, lfscktest -> lfsck
- add racer.sh, liblustre.sh scripts
- remove fsx,bonnie,dbench,iozone.lfsck parts

13 years agob=23051 improve summary of acc-sm to include test times
Elena Gryaznova [Wed, 22 Dec 2010 20:16:47 +0000 (23:16 +0300)]
b=23051 improve summary of acc-sm to include test times

i=Andreas.Dilger
i=Andrew.Perepechko

13 years agob=23081 Move llap page to tail instead of head.
Jeremy Filizetti [Wed, 22 Dec 2010 03:21:00 +0000 (11:21 +0800)]
b=23081 Move llap page to tail instead of head.

i=adilger
i=johann

13 years agob=24226 typo fix for sanity test 72
James Simmons [Wed, 22 Dec 2010 19:56:56 +0000 (22:56 +0300)]
b=24226 typo fix for sanity test 72

i=Elena Gryaznova

13 years agob=20394 correct check for transno value in filter_finish_transno
Mikhail Pershin [Tue, 21 Dec 2010 07:01:38 +0000 (10:01 +0300)]
b=20394 correct check for transno value in filter_finish_transno

i=tappro,johann

13 years agob=24048 Set body->eadatasize in mdc_getattr_pack()
Mikhail Pershin [Tue, 21 Dec 2010 06:54:27 +0000 (09:54 +0300)]
b=24048 Set body->eadatasize in mdc_getattr_pack()

i=adilger, johann

13 years agob=18717 make "lfs check" output consistent on stdout
Elena Gryaznova [Mon, 20 Dec 2010 21:13:17 +0000 (00:13 +0300)]
b=18717 make "lfs check" output consistent on stdout

o=Andreas.Dilger
i=Brian.Murrell
i=Elena.Gryaznova

13 years agob=23049 canonicalize disk names
Elena Gryaznova [Mon, 20 Dec 2010 16:40:09 +0000 (19:40 +0300)]
b=23049 canonicalize disk names

o=Brian.Murrell
i=grev

13 years agob=23049 various t-f.sh patches
Elena Gryaznova [Mon, 20 Dec 2010 16:11:01 +0000 (19:11 +0300)]
b=23049 various t-f.sh patches

o=Brian.Murrell
i=grev

rundbench is a bash script;
obdfilter-survey is a bash script;
don't su if MPI_USER == "";

13 years agob=23049 set path to truncate
Elena Gryaznova [Mon, 20 Dec 2010 16:00:58 +0000 (19:00 +0300)]
b=23049 set path to truncate

i=Brian.Murrell

13 years agob=22544 delete module_setup.sh
Elena Gryaznova [Mon, 20 Dec 2010 12:07:25 +0000 (15:07 +0300)]
b=22544 delete module_setup.sh

i=Brian.Murrell

13 years agob=24039 lfs setstripe --pool broken
Elena Gryaznova [Mon, 20 Dec 2010 11:48:38 +0000 (14:48 +0300)]
b=24039 lfs setstripe --pool broken

i=Andreas.Dilger
i=Vladimir.Saveliev

13 years agob=24239 use SAMPLE_FILE instead of termcap
Elena Gryaznova [Mon, 20 Dec 2010 09:28:37 +0000 (12:28 +0300)]
b=24239 use SAMPLE_FILE instead of termcap

i=Vitaly.Fertman

13 years agob=24266 increase replay-single test_70d dbench duration for HARD failure mode
Elena Gryaznova [Mon, 20 Dec 2010 09:14:19 +0000 (12:14 +0300)]
b=24266 increase replay-single test_70d dbench duration for HARD failure mode

i=Mikhail.Pershin

13 years agob=24226 Only force the mode change if we're changing the size as well 1.8.5.51 v1_8_5_51
James Simmons [Fri, 17 Dec 2010 22:57:43 +0000 (23:57 +0100)]
b=24226 Only force the mode change if we're changing the size as well

i=johann

The offending code was added by commit 77ba4b2141d04180211efa8a75c11ab0abf7fafb to remove
setgid/setuid bits when do_truncate() is called on the file. We should only force the change when
that occurs, similarly to ll_setattr() in lustre/llite/llite_lib.c

13 years agoRevert "b=23787 make struct lprocfs_percpu definition C99 compliant"
Johann Lombardi [Fri, 17 Dec 2010 22:47:38 +0000 (23:47 +0100)]
Revert "b=23787 make struct lprocfs_percpu definition C99 compliant"

This reverts commit 46ff2dd66e91530f804d5b2321e58827f71e0251.
This is causing build errors.

13 years agochange version to 1.8.5.51
Johann Lombardi [Fri, 17 Dec 2010 16:04:20 +0000 (17:04 +0100)]
change version to 1.8.5.51

13 years agob=19808 fix d_obtain_alias() misuse due to compat macro
Andreas Dilger [Fri, 17 Dec 2010 16:01:15 +0000 (17:01 +0100)]
b=19808 fix d_obtain_alias() misuse due to compat macro

i=oleg
i=yangsheng

13 years agob=24055 a patch to detect if quota is turned on properly
Landen [Fri, 17 Dec 2010 06:46:28 +0000 (14:46 +0800)]
b=24055 a patch to detect if quota is turned on properly

i=johann
i=panda

13 years agob=22546 fix errors in test_18c
Landen [Fri, 17 Dec 2010 06:46:29 +0000 (14:46 +0800)]
b=22546 fix errors in test_18c

i=johann
i=panda

13 years agob=24245 skip sanity test 123 under 1.8 <-> 2.x interoperability mode
Jian Yu [Fri, 17 Dec 2010 01:25:35 +0000 (09:25 +0800)]
b=24245 skip sanity test 123 under 1.8 <-> 2.x interoperability mode

statahead is disabled automatically under 1.8 <-> 2.x interoperability
mode

i=grev

13 years agob=23821 Limit bio_alloc() to BIO_MAX_PAGES iovecs.
Andreas Dilger [Thu, 23 Sep 2010 19:17:19 +0000 (13:17 -0600)]
b=23821 Limit bio_alloc() to BIO_MAX_PAGES iovecs.

Fix logic error when patch was originally landed from b=9945.
i=alex.zhuravlev
i=kwestneat

13 years agofix the formatting in filter_do_bio
Andrew Perepechko [Fri, 17 Dec 2010 00:32:35 +0000 (03:32 +0300)]
fix the formatting in filter_do_bio

13 years agob=23787 make struct lprocfs_percpu definition C99 compliant
Dmitry Zogin [Fri, 17 Dec 2010 00:23:15 +0000 (03:23 +0300)]
b=23787 make struct lprocfs_percpu definition C99 compliant

i=Andreas Dilger
i=Andrew Perepechko

13 years agob=23786 make lh_exit code C99 compliant
Dmitry Zogin [Fri, 17 Dec 2010 00:18:07 +0000 (03:18 +0300)]
b=23786 make lh_exit code C99 compliant

Based on the patch from Kenneth D. Matney, Sr. <matneykdsr@ornl.gov>

i=Andreas Dilger
i=Andrew Perepechko

13 years agob=23157 do not crash on wrong network message in filter_connect_internal
Bernd Schubert [Fri, 17 Dec 2010 00:09:07 +0000 (03:09 +0300)]
b=23157 do not crash on wrong network message in filter_connect_internal

13 years agob=24270 need to mkdir mntpt before mount
Minh Diep [Thu, 16 Dec 2010 23:06:13 +0000 (16:06 -0700)]
b=24270 need to mkdir mntpt before mount

i=Elena

13 years agob=16605 don't LASSERT on unverified client data in filter_parent
Andreas Dilger [Thu, 16 Dec 2010 17:36:38 +0000 (20:36 +0300)]
b=16605 don't LASSERT on unverified client data in filter_parent

i=Johann Lombardi
i=Andrew Perepechko

13 years agob=13698 llapi_get_version
Vladimir Saveliev [Tue, 14 Dec 2010 12:30:12 +0000 (15:30 +0300)]
b=13698 llapi_get_version

this uses OBD_GET_VERSION ioctl to obtain lustre version

i=andreas.dilger

13 years agob=23961 fix for setup with several network interfaces
Elena Gryaznova [Mon, 13 Dec 2010 13:32:55 +0000 (16:32 +0300)]
b=23961 fix for setup with several network interfaces

i=Andrew.Perepechko

- metadata-updates fix for setup when several interfaces are UP on host;
  hostname could be assigned to IP which is different from lnet network used,
  the hostname-s of NODES_TO_USE are now stored in HOSTS
- new SHUTDOWN_ATTEMPTS: the tunable number of attepts to shutdown node
- shutdown_node_hard () fix: do not call "power off" each time,
  wait that the node is not pingable before the next "power off" attempt
- unused check_port() is removed

13 years agob=4424 Reserve obd_connect_data.ocd_max_easize field
root [Fri, 10 Dec 2010 09:34:05 +0000 (02:34 -0700)]
b=4424 Reserve obd_connect_data.ocd_max_easize field

To avoid potential incompatible changes between b1_8 and master,
reserve the ocd_max_easize field.  The corresponding connect flag
OBD_CONNECT_MAX_EASIZE has been reserved for some time already.

Add several other OBD_CONNECT_ flags that have already been defined
to the wirecheck/wiretest tools.

13 years agob=22376 sanity test for non-root exec-only file execution
Elena Gryaznova [Wed, 8 Dec 2010 20:32:43 +0000 (23:32 +0300)]
b=22376 sanity test for non-root exec-only file execution

o=Andreas.Dilger
i=grev

13 years agob=23766 interop bits for sanity/203
Andrew Perepechko [Thu, 2 Dec 2010 23:47:38 +0000 (02:47 +0300)]
b=23766 interop bits for sanity/203

i=Elena Gryaznova

13 years agob=24118 test_70b rundbench load failed
Elena Gryaznova [Thu, 2 Dec 2010 16:27:50 +0000 (19:27 +0300)]
b=24118 test_70b rundbench load failed

i=Brian.Murrell
i=Jian.Yu

- give rundbench a chance to start before the dbench load check
- new check_for_process () and killall_process () to check/kill any defined progs
  instead of "dbench" only
- fix 70a, 70b to mount the clients on MOUNT instead of DIR

13 years agoRevert "b=23913 fix "ASSERTION(!cfs_list_empty(&dquot->dq_hash)) failed""
Andrew Perepechko [Wed, 1 Dec 2010 22:49:02 +0000 (01:49 +0300)]
Revert "b=23913 fix "ASSERTION(!cfs_list_empty(&dquot->dq_hash)) failed""

This reverts commit 9491d5d931407ccf68d95932ca4ea3c545c6f86d.

13 years agob=24228 fix test duration check to be more accurate
Elena Gryaznova [Wed, 1 Dec 2010 17:25:24 +0000 (20:25 +0300)]
b=24228 fix test duration check to be more accurate

i=vitaly

13 years agob=23535 sgpdd-survey.sh should check for sg_map
Elena Gryaznova [Tue, 30 Nov 2010 16:20:43 +0000 (19:20 +0300)]
b=23535 sgpdd-survey.sh should check for sg_map

i=Brian.Murrell

check that iokit sgpdd-survey and sg_map are installed

13 years agob=22157 combined mgs/mds fix for single node setup
Elena Gryaznova [Tue, 30 Nov 2010 14:34:46 +0000 (17:34 +0300)]
b=22157 combined mgs/mds fix for single node setup

i=Mikhail.Pershin

for configuration combined mgs/mds on single node setup
we do not need to unload the modules because conf-sanity keeps
the mgs mounted during all tests

13 years agob=23402 mmp_fini () multiple oss fix
Elena Gryaznova [Tue, 30 Nov 2010 12:51:40 +0000 (15:51 +0300)]
b=23402 mmp_fini () multiple oss fix

i=Yu.Jian

13 years agob=23575 O2iblnd credit deadlock regression
Isaac Huang [Mon, 29 Nov 2010 04:22:42 +0000 (21:22 -0700)]
b=23575 O2iblnd credit deadlock regression

This fixed a regression of bug 14425.

i=liang

13 years agob=23913 fix "ASSERTION(!cfs_list_empty(&dquot->dq_hash)) failed"
Landen [Mon, 29 Nov 2010 08:07:11 +0000 (16:07 +0800)]
b=23913 fix "ASSERTION(!cfs_list_empty(&dquot->dq_hash)) failed"

1. In an error handling, free_dquot() should be used.
2. Check if quota is on in dqacq_handler()

i=johann
i=panda

13 years agob=23868 fix "sanity-quota test_18c: @@@@@@ FAIL: quotaon failed!"
Landen [Mon, 29 Nov 2010 08:07:12 +0000 (16:07 +0800)]
b=23868 fix "sanity-quota test_18c: @@@@@@ FAIL: quotaon failed!"

i=johann
i=mikhail.pershin

13 years agob=23954 MGS device has stopped when we try to start the second mgs
Elena Gryaznova [Fri, 26 Nov 2010 18:18:36 +0000 (21:18 +0300)]
b=23954 MGS device has stopped when we try to start the second mgs

i=Mikhail.Pershin

add test_24b to ALWAYS_EXCEPT list for configuration mgs/mds are not combined

13 years agob=23869 HARD failure mode fixes
Elena Gryaznova [Fri, 26 Nov 2010 17:43:09 +0000 (20:43 +0300)]
b=23869 HARD failure mode fixes

i=Brian.Murrel
i=Wei.G.Li

facet_failover() has to restart only those affected facets which were UP
before the node failure

replay-single tests which use shutdown_facet() && reboot_facet() instead of
facet_failover() have to take care about the affected facets

13 years agob=23956 change conf-sanity test_37 to be functional on remote setup
Elena Gryaznova [Fri, 26 Nov 2010 17:12:33 +0000 (20:12 +0300)]
b=23956 change conf-sanity test_37 to be functional on remote setup

i=Mikhail.Pershin

fix test_37 to not be skipped on remote setup;
use the existing mds device instead of create a new one

13 years agob=24020 lustre doesn't start with ext4 based ldiskfs.
Elena Gryaznova [Fri, 26 Nov 2010 16:36:06 +0000 (19:36 +0300)]
b=24020 lustre doesn't start with ext4 based ldiskfs.

i=Andrew.Perepechko

13 years agob=24201 add procfs tunable to enable/disable lockless direct I/O
Andrew Perepechko [Fri, 26 Nov 2010 01:29:34 +0000 (04:29 +0300)]
b=24201 add procfs tunable to enable/disable lockless direct I/O

llite.lustre-*.lockless_direct_io=0 will disable default semantics
of direct I/O that forces it to be lockless. lockless_direct_io value,
however, will be ignored if per-file LL_FILE_LOCKED_DIRECTIO bit
is set.

i=Johann Lombardi

13 years agob=21804 make sure the request is protected by rq_refcount while
Johann Lombardi [Thu, 25 Nov 2010 17:36:38 +0000 (18:36 +0100)]
b=21804 make sure the request is protected by rq_refcount while

i=vitaly
i=oleg

13 years agob=23820 add missing changelog entry
Johann Lombardi [Thu, 25 Nov 2010 17:22:45 +0000 (18:22 +0100)]
b=23820 add missing changelog entry

13 years agob=21760 start bulk unregistering at the same time as reply unlink
Johann Lombardi [Thu, 25 Nov 2010 17:21:16 +0000 (18:21 +0100)]
b=21760 start bulk unregistering at the same time as reply unlink

i=vitaly
i=oleg

13 years agob=23820 ptlrpc_check_set()) ASSERTION(req->rq_phase == RQ_PHASE_BULK) failed
Dmitry Zogin [Sat, 2 Oct 2010 03:33:18 +0000 (23:33 -0400)]
b=23820 ptlrpc_check_set()) ASSERTION(req->rq_phase == RQ_PHASE_BULK) failed

i=oleg.drokin
i=johann.lombardi

Handle unsent requests with rq_net_err in ptlrpc_check_set().

13 years agomove to 1.8.6
Johann Lombardi [Thu, 25 Nov 2010 15:56:42 +0000 (16:56 +0100)]
move to 1.8.6

13 years agofixed a typo from previous commit
Andrew Perepechko [Thu, 18 Nov 2010 12:40:18 +0000 (15:40 +0300)]
fixed a typo from previous commit

13 years agob=23766 atime is not updated properly
Andrew Perepechko [Thu, 18 Nov 2010 12:35:02 +0000 (15:35 +0300)]
b=23766 atime is not updated properly

a test case for atime updates on an MDS

i=Johann Lombardi
i=Andreas Dilger

13 years agob=23985 Fix upstream jbd2 code issue. 1.8.5 v1_8_5_RC4
yangsheng [Mon, 15 Nov 2010 14:43:46 +0000 (22:43 +0800)]
b=23985 Fix upstream jbd2 code issue.

The jbd2_stats_proc_init() was placed on wrong location in
  jbd2_journal_init_dev().  This may cause /proc/fs/jdb2/<dev>/*
  cannot be created when using external journal device.

i=adilger
i=johann

13 years agoChangelog update
Andrew Perepechko [Mon, 15 Nov 2010 12:43:06 +0000 (15:43 +0300)]
Changelog update

13 years agob=24051 sanity test_76 fix
Andrew Perepechko [Mon, 15 Nov 2010 12:29:06 +0000 (15:29 +0300)]
b=24051 sanity test_76 fix

Due to per-cpu slab caches, "active" stat from /proc/slabinfo
is not always accurate.

i=ZhiYong Tian

13 years agob=23766 atime isn't properly updated on the MDS
Andrew Perepechko [Mon, 15 Nov 2010 12:21:11 +0000 (15:21 +0300)]
b=23766 atime isn't properly updated on the MDS

Based on the patch from Robin Humble (ANU)

i=Andreas Dilger
i=Vitaly Fertman

13 years agob=21587 don't LBUG if transno has changed during replay v1_8_5_RC3
Andrew Perepechko [Wed, 27 Oct 2010 04:54:18 +0000 (12:54 +0800)]
b=21587 don't LBUG if transno has changed during replay

i=johann
i=tappro

13 years agob=23988 Changes sd-iostats patch for 2.6.32 kernel.
yangsheng [Fri, 22 Oct 2010 09:43:38 +0000 (17:43 +0800)]
b=23988 Changes sd-iostats patch for 2.6.32 kernel.

Also statistic with REQ_TYPE_BLOCK_PC command.

13 years agoRevert "b=23988 Changes sd-iostats patch for 2.6.32 kernel."
Johann Lombardi [Thu, 21 Oct 2010 12:27:41 +0000 (14:27 +0200)]
Revert "b=23988 Changes sd-iostats patch for 2.6.32 kernel."

This reverts commit 50a915fdbb340f04a850778870d027c4c3303aab.

13 years agob=23965 Add fiemap patch for ldiskfs base on ext4(sles11 sp1).
yangsheng [Thu, 21 Oct 2010 12:06:52 +0000 (20:06 +0800)]
b=23965 Add fiemap patch for ldiskfs base on ext4(sles11 sp1).

i=girish
i=johann

13 years agob=23988 Changes sd-iostats patch for 2.6.32 kernel.
yangsheng [Thu, 21 Oct 2010 12:00:24 +0000 (20:00 +0800)]
b=23988 Changes sd-iostats patch for 2.6.32 kernel.

13 years agob=24003 Shows stripe_count: 65535 but should be '-1'.
yangsheng [Thu, 21 Oct 2010 10:42:00 +0000 (18:42 +0800)]
b=24003 Shows stripe_count: 65535 but should be '-1'.

i=johann

13 years agob=21587 don't clean up echo_client is refcount is not 0
Johann Lombardi [Wed, 20 Oct 2010 17:46:23 +0000 (19:46 +0200)]
b=21587 don't clean up echo_client is refcount is not 0

i=adilger

13 years agob=23966 Revert "b=21137 sles11 with 1.8 is slower than 1.6 sles10 for O_DIRECT single...
Johann Lombardi [Wed, 20 Oct 2010 09:37:53 +0000 (11:37 +0200)]
b=23966 Revert "b=21137 sles11 with 1.8 is slower than 1.6 sles10 for O_DIRECT single file IOR writes"

This reverts commit 1242fac546a8c252dee5715bb218133795211028.
This introduces issues with SATA drivers. See bug 23966.

13 years agob=23721 download OFED from OFA v1_8_5_RC2
Brian J. Murrell [Thu, 14 Oct 2010 18:34:15 +0000 (20:34 +0200)]
b=23721 download OFED from OFA

Download the OFED distribution directly from OFA to ameliorate
re-distribution complications.

i=yangsheng
i=wangyb

13 years agofix nit in lbuild-sles10
Johann Lombardi [Thu, 14 Oct 2010 16:42:27 +0000 (18:42 +0200)]
fix nit in lbuild-sles10