Whamcloud - gitweb
girish [Sun, 3 May 2009 18:51:06 +0000 (18:51 +0000)]
b=19184
i=adilger
i=girish
o=brian behlendorf
add proper locking for creation of jbd_4k slab cache
shadow [Fri, 1 May 2009 11:42:55 +0000 (11:42 +0000)]
set notransno before readonly.
Branch b1_6
b=19085
i=rread
i=adilger
liuy [Thu, 30 Apr 2009 01:54:25 +0000 (01:54 +0000)]
Branch b1_6
b=18775
i=rread
i=nathan
Improve lctl set/get_param:
-handle the bad options
-support more than one arguments
-add '-F' option to show the indicator to the parameters, if '-N' specified
yangsheng [Wed, 29 Apr 2009 09:25:04 +0000 (09:25 +0000)]
Branch b1_6
b=19024
i=shadow, bobijam
Update to RHEL5.3 2.6.18-128.1.6.el5 kernel.
yangsheng [Wed, 29 Apr 2009 07:18:03 +0000 (07:18 +0000)]
Branch b1_6
b=18061
yangsheng [Wed, 29 Apr 2009 07:14:31 +0000 (07:14 +0000)]
Branch b1_6
b=18061
i=johann, bobijam
Update to SLES10 2.6.16.60-0.37 kernel.
wangyb [Tue, 28 Apr 2009 08:13:56 +0000 (08:13 +0000)]
branch b1_6
b=18015
i=zhenyu.xu
i=sheng.yang
check if synchronize_rcu is available, otherwise use synchronize_kernel
fanyong [Sat, 25 Apr 2009 10:58:38 +0000 (10:58 +0000)]
Branch b1_6
b=19171
i=tianzy
i=robert.read
Hold obd reference count when quota recovery.
fanyong [Sat, 25 Apr 2009 07:01:26 +0000 (07:01 +0000)]
Branch b1_6
b=15691
i=eric.mei
i=peng.ye
Another diagnostic patch to collect more log on both client and OSS.
bobijam [Fri, 24 Apr 2009 02:21:23 +0000 (02:21 +0000)]
Branch b1_6
b=18683
i=johann
i=sheng.yang
If no options are given, tunefs.lustre/mkfs.lustre just print info.
shadow [Thu, 23 Apr 2009 06:18:02 +0000 (06:18 +0000)]
correctly skip time estimate if in recovery
Branch b_release_1_8_0
b=19223
i=johann
i=nathan
fanyong [Thu, 23 Apr 2009 04:48:52 +0000 (04:48 +0000)]
Branch b1_6
b=19149
i=tianzy
i=panda
Make sure quota has been released to master in test_6.
tappro [Wed, 22 Apr 2009 18:38:09 +0000 (18:38 +0000)]
- set mod_close_req properly
b:3633
i:vitaly,adilger
shadow [Wed, 22 Apr 2009 18:06:05 +0000 (18:06 +0000)]
Allocate inodes for llog in last inode group for avoid broke recovery.
Branch b1_6
b=18192
i=adilger
i=shadow
i=girish
cvs2svn [Wed, 22 Apr 2009 17:55:30 +0000 (17:55 +0000)]
This commit was manufactured by cvs2svn to create branch 'b1_6'.
brian [Wed, 22 Apr 2009 17:55:29 +0000 (17:55 +0000)]
b=13908
i=yibin.wang
i=sheng.yang
Allow the name, version, kernel release and release of the lustre packages
to be defined on the command line.
With such a feature, actually properly name the patchless client packages in
our own build.
Tighten up some dependencies.
nathan [Wed, 22 Apr 2009 17:12:27 +0000 (17:12 +0000)]
b=18676
i=johann
i=manoj
replace obd ioctl with dir ioctl for changelogs
vitaly [Wed, 22 Apr 2009 13:37:05 +0000 (13:37 +0000)]
Branch HEAD
b=15977
i=deen
i=vs
lctl dk should zero the file before writing the debug log into,
otherwise, if a file was long enough and is not completely overwritten
we get a mix of several logs.
deen [Wed, 22 Apr 2009 12:08:34 +0000 (12:08 +0000)]
This patch implements separate function for handling OPEN requests from Lustre clients acting as
NFS servers, named mdt_open_anon_by_fid(). It searches for existing object and, if succeed, takes
an open lock on the object and its attributes. If a client didn't ask for the open lock, it releases
it at the end. We must always take an open lock at the beginning in order to protect object's
attributes.
b=16715
i=oleg.drokin
i=alexey.lyashkov
zhanghc [Wed, 22 Apr 2009 02:07:14 +0000 (02:07 +0000)]
b=18689
change thread name "ll_log_commit_*" in
llog_recov_thread_init into "lcm_*" so it
can be fitted into task struct
i=adilger
anserper [Wed, 22 Apr 2009 01:08:07 +0000 (01:08 +0000)]
b=16890
the missing part from a previous commit
jxiong [Wed, 22 Apr 2009 00:45:23 +0000 (00:45 +0000)]
b=18711
r=jay, adilger
in lu_context_key_deregister
- to not print message inside a spinlock
- assert when the key has extra references
anserper [Wed, 22 Apr 2009 00:26:47 +0000 (00:26 +0000)]
b=16890
i=Johann Lombardi
i=ZhiYong Tian
avoid several unnecessary memsets
grev [Tue, 21 Apr 2009 19:03:40 +0000 (19:03 +0000)]
b=19116
i=Robert.Read
convert "-" to X to create client variable name
grev [Tue, 21 Apr 2009 18:50:50 +0000 (18:50 +0000)]
b=19116
i=Robert.Read
convert "-" to X to create client variable name
grev [Tue, 21 Apr 2009 18:13:18 +0000 (18:13 +0000)]
b=19175
i=Robert.Read
check_catastrophe fix
yangsheng [Tue, 21 Apr 2009 14:42:04 +0000 (14:42 +0000)]
Branch HEAD
b=18810
i=bzzz, adilger
Use sb_bgl_lock to protect bitmaps.
Author: adilger
tianzy [Tue, 21 Apr 2009 09:22:08 +0000 (09:22 +0000)]
Branch HEAD
1. delete the lc_watchdog_touch_ms()
2. changing to seconds instead of ms
3. delete lcw_time in lc_watchdog
b=18582
i=nathan
i=johann
tianzy [Tue, 21 Apr 2009 09:21:35 +0000 (09:21 +0000)]
Branch b1_6
1. delete the lc_watchdog_touch_ms()
2. changing to seconds instead of ms
3. delete lcw_time in lc_watchdog
b=18582
i=nathan
i=johann
grev [Mon, 20 Apr 2009 21:11:47 +0000 (21:11 +0000)]
b=19175
i=Robert.Read
check_catastrophe fix
anserper [Mon, 20 Apr 2009 21:02:04 +0000 (21:02 +0000)]
b=18951
grev [Mon, 20 Apr 2009 18:49:35 +0000 (18:49 +0000)]
b=19140
i=Alexander.Zarochentsev
PTLDEBUG/SUBSYSTEM t-f set_param fix:
accept the string value
anserper [Mon, 20 Apr 2009 18:29:09 +0000 (18:29 +0000)]
b=18951
i=Nathan Rutman
i=Vladimir Saveliev
fix the race between fflush and pause
zhanghc [Mon, 20 Apr 2009 11:56:24 +0000 (11:56 +0000)]
b=18689
change thread name "ll_log_commit_*" in
llog_recov_thread_init into "lcm_*" so it
can be fitted into the task struct
i=adilger
grev [Fri, 17 Apr 2009 12:01:21 +0000 (12:01 +0000)]
b=19107
i=Maxim.Patlasov
acc-sm liblustre skip list
zhanghc [Thu, 16 Apr 2009 15:41:39 +0000 (15:41 +0000)]
b=17485
don't reuse orphan objects in OST during
establishing connection in MDS
i=shadow
i=bobijam
zam [Thu, 16 Apr 2009 08:15:02 +0000 (08:15 +0000)]
Branch b1_6
b=18988
i=Alexey.Lyashkov
new sanity test for for O_APPEND.
zam [Thu, 16 Apr 2009 07:54:44 +0000 (07:54 +0000)]
Branch HEAD
b=18988
i=Alexey.Lyashkov
new sanity test for for O_APPEND.
deshmukh [Thu, 16 Apr 2009 05:42:38 +0000 (05:42 +0000)]
b=17931
i=fonyoung
i=grev
conf_sanity_33a test-case is fixed now.
maxim [Wed, 15 Apr 2009 19:38:54 +0000 (19:38 +0000)]
b=18844
i=isaac
Landing a patch fixing deadlock in usocklnd on HEAD. The patch also includes:
- minor code cleanup suggested by Isaac in comment #6 of bug #18844
- trivial cleanup patch for acceptor.c (att #21983 of bug #14132)
- minor one-line fix for handling EINTR error of poll(2) - inspected by Shadow.
maxim [Wed, 15 Apr 2009 19:22:44 +0000 (19:22 +0000)]
b=18414
A few words to lnet/ChangeLog about 'running out of ports' fix.
yangsheng [Wed, 15 Apr 2009 10:35:49 +0000 (10:35 +0000)]
Branch HEAD
b=18033
Fix the typo.
zhanghc [Wed, 15 Apr 2009 08:32:13 +0000 (08:32 +0000)]
b=18733
add protection (spin_lock(&lock->l_loc)
before using lock->l_resource in "lock_res_and_lock"
i=johann
i=green
zam [Wed, 15 Apr 2009 06:25:13 +0000 (06:25 +0000)]
b=19123
i=robert.read
test-framework.sh:get_mds_dir:
Add another sleep 1 before "lfs df -i" to be sure that statfs cache is
invalidated.
ericm [Tue, 14 Apr 2009 20:36:46 +0000 (20:36 +0000)]
branch: HEAD
Fix null pointer reference of rq_export, which maybe null in case of gss.
b=18983
r=fanyong
r=wangdi
shadow [Tue, 14 Apr 2009 18:46:41 +0000 (18:46 +0000)]
revert
b=16839
yangsheng [Tue, 14 Apr 2009 12:54:49 +0000 (12:54 +0000)]
Branch b1_6
b=18399
Commit the new patch.
yangsheng [Tue, 14 Apr 2009 12:53:41 +0000 (12:53 +0000)]
Branch b1_6
b=18399
i=adilger, johann
Patch the kernel to avoid the deadlock in prune_icache().
Author: ZhangHongChao
adilger [Tue, 14 Apr 2009 07:02:57 +0000 (07:02 +0000)]
Branch b1_8
Use EXT_UNSET_BLOCK to avoid confusion with EXT_MAX_BLOCK on 16TB-1 block
filesystems.
This patch was committed to b_release_1_8_0, but was missed on ldiskfs HEAD
for b1_8.
b=16114
o=girish.shilamkar
i=adilger
i=kalpak.shah
adilger [Tue, 14 Apr 2009 04:45:18 +0000 (04:45 +0000)]
Branch b1_8
Update ldiskfs build version to 3.0.9 for next release.
adilger [Tue, 14 Apr 2009 04:40:46 +0000 (04:40 +0000)]
Branch b1_8
Description: fix racy locking of mballoc block bitmaps causing BUG
Details : The locking of the mballoc buddy bitmap and the in-memory
block bitmap was using two different spin locks in some
cases. This made it possible to incorrectly access the
mballoc bitmap while another process was modifying it,
causing a sanity assertion to fail. While no on-disk corruption
was reported, there was some risk of this happening.
b=18810
i=alex
anserper [Tue, 14 Apr 2009 00:02:35 +0000 (00:02 +0000)]
b=18728
i=Alex Zhuravlev
i=Alexander Zarochentsev
decrease response time for write-waiting threads
anserper [Mon, 13 Apr 2009 23:57:46 +0000 (23:57 +0000)]
b=16890
i=ZhiYong Tian
i=Johann Lombardi
admin qf memset reduction, backport from the kernel patch (5.93 KB, patch)
anserper [Mon, 13 Apr 2009 23:46:03 +0000 (23:46 +0000)]
b=18951
i=Nathan Rutman
i=Vladimir Saveliev
fix for a specific race in multiop
anserper [Mon, 13 Apr 2009 23:32:29 +0000 (23:32 +0000)]
b=19063
i=Elena Gryaznova
quota_save_version must wait until switch to the new version has happened
shadow [Mon, 13 Apr 2009 18:58:32 +0000 (18:58 +0000)]
add lazystatfs mount option to allow statfs(2) to skip down OSTs
Branch b1_6
b=17974
i=green
i=vitaly
liangzhen [Mon, 13 Apr 2009 17:45:26 +0000 (17:45 +0000)]
Branch HEAD
b=13621, 15983
i=isaac
i=maxim
o2iblnd protocol version 2
liangzhen [Mon, 13 Apr 2009 16:31:42 +0000 (16:31 +0000)]
Branch HEAD
Tag for performance test
bobijam [Mon, 13 Apr 2009 06:26:21 +0000 (06:26 +0000)]
Branch HEAD
b=16417
sleep some time before getting 'lfs df' value.
grev [Sat, 11 Apr 2009 09:35:15 +0000 (09:35 +0000)]
b=18779
i=Vladimir.Saveliev
i=Nathan.Rutman
test_34c fix
grev [Fri, 10 Apr 2009 21:07:05 +0000 (21:07 +0000)]
b=18271
i=Nathan.Rutman
writeconf fn fix
stinson1947 [Fri, 10 Apr 2009 20:26:29 +0000 (20:26 +0000)]
Updated version.
stinson1947 [Fri, 10 Apr 2009 20:25:20 +0000 (20:25 +0000)]
Updated version string for the next bi-weekly build. This is also a 2.0 Alpha candidate.
vitaly [Fri, 10 Apr 2009 19:43:32 +0000 (19:43 +0000)]
Branch HEAD
b=18863
i=shadow
i=bzzz
lu cache shrinker bugfix, return the proper amount of cached objects.
vitaly [Fri, 10 Apr 2009 19:40:39 +0000 (19:40 +0000)]
Branch HEAD
b=18792
i=jay
i=huanghua
1. remove the assert for inode reference in cl_inode_fini().
2. as ldlm_cli_cancel() takes a "long" lock, i.e. without a reference,
we should not drop this reference later. --enable-lu_ref fix.
robert.read [Fri, 10 Apr 2009 19:18:37 +0000 (19:18 +0000)]
Branch HEAD
b=19025
i=jay
Upstream patch, this time with the patch.
robert.read [Fri, 10 Apr 2009 19:01:16 +0000 (19:01 +0000)]
Branch HEAD
b=19025
i=jay
i=rread
Upstream patch to fix panic on i/o errors.
robert.read [Fri, 10 Apr 2009 18:48:35 +0000 (18:48 +0000)]
Branch HEAD
b=18556
o=tappro
i=rread
i=nathan
Reoganize recovery timer managing.
robert.read [Fri, 10 Apr 2009 18:36:42 +0000 (18:36 +0000)]
Branch HEAD
b=18978
o=zam
i=tappro
i=rread
Don't LBUG in CMD cmr_version_get/set stubs.
grev [Fri, 10 Apr 2009 15:02:02 +0000 (15:02 +0000)]
b=19139
i=Jian.Yu
typo END_RUN_FILE name fix
grev [Fri, 10 Apr 2009 14:56:35 +0000 (14:56 +0000)]
b=19139
i=Jian.Yu
typo END_RUN_FILE name fix
grev [Fri, 10 Apr 2009 14:48:15 +0000 (14:48 +0000)]
b=19129
i=Alexander.Zarochentsev
zconf_mount_clients remote command fix
grev [Fri, 10 Apr 2009 14:45:28 +0000 (14:45 +0000)]
b=19129
i=Alexander.Zarochentsev
zconf_mount_clients remote command fix
grev [Fri, 10 Apr 2009 14:40:25 +0000 (14:40 +0000)]
b=18080
i=Alexander.Zarochentsev
add sanityN test_35 to A_E list
yujian [Fri, 10 Apr 2009 08:50:03 +0000 (08:50 +0000)]
Branch b1_6
b=18501
i=nathan
Document the limitation for lustre_config in the upgrading case and
specify "--nomgs" option to a non-co-located MDT.
jxiong [Fri, 10 Apr 2009 04:04:16 +0000 (04:04 +0000)]
b=19016
r=wangdi,fan.yong
Fix an issue about enqueuing a lock in clio.
jxiong [Fri, 10 Apr 2009 03:32:21 +0000 (03:32 +0000)]
b=18293
i=fan.yong,rread
Revise the handling of IO error in clio:
- don't set page error for sync and direct IO;
- when IO error occurs, take the page out from SoM queue since we don't resend
the page right now.
bobijam [Fri, 10 Apr 2009 02:45:19 +0000 (02:45 +0000)]
Branch b1_6
b=18918
i=nathan.rutman
fix service name on (un)mount for MGS.
bobijam [Fri, 10 Apr 2009 01:51:30 +0000 (01:51 +0000)]
Branch b1_6
b=18638
i=jinshan.xiong (jay)
i=andrew.perepachko (panda)
ldlm_lock_put() use RCU to call ldlm_lock_free, so need call synchronize_rcu()
to wait a grace period elapsed, so that ldlm_lock_free() get a chance to be
called.
ericm [Thu, 9 Apr 2009 20:50:31 +0000 (20:50 +0000)]
branch: HEAD
OSD doing batch read of dir entries in compatiblility mode iterator.
b=17560
o=pravin
i=oleg,rahul,pravin,rread
ericm [Thu, 9 Apr 2009 20:43:11 +0000 (20:43 +0000)]
branch: HEAD
add an assertion trying to catch the rarely happened case of possible
import reference over dropping.
b=18849
vitaly [Thu, 9 Apr 2009 16:15:57 +0000 (16:15 +0000)]
Branch HEAD
b=17926
i=rread
i=shadow
limit the default journal size with 400MB
bobijam [Thu, 9 Apr 2009 09:53:34 +0000 (09:53 +0000)]
Branch HEAD
b=16417
i=grev
i=alexey.lyashkov (shadow)
revise sanityN.sh test_36()
- need sleep some time before getting 'after' value from 'lfs df', because
if 'multiop & kill' happens too fast, 'lfs df' sometimes only get old df
value from cache w/o reissuing statfs RPC request.
shadow [Thu, 9 Apr 2009 05:36:17 +0000 (05:36 +0000)]
don't allow connect to already connected import
Branch HEAD
b=16839
i=rread
zhanghc [Wed, 8 Apr 2009 17:20:51 +0000 (17:20 +0000)]
b=17491
remove "wrapped times" stuff in original patch
i=nathan.rutman
i=robert.read
zhanghc [Wed, 8 Apr 2009 16:02:55 +0000 (16:02 +0000)]
b=17491
remove "wrapped times" stuff in original patch
i=nathan.rutman
i=robert.read
shadow [Wed, 8 Apr 2009 11:05:18 +0000 (11:05 +0000)]
kill one configure check
Branch b1_6
b=14250
i=shadow
tappro [Wed, 8 Apr 2009 09:59:33 +0000 (09:59 +0000)]
- fix wrong flag check in ptlrpc_at_set_reply()
- remove staled clients before finish_recovery stage
- increase time for recovery if switched to vbr
b:18556
i:rread,zam,nathan
shadow [Wed, 8 Apr 2009 06:24:43 +0000 (06:24 +0000)]
Don't allow reconnect to already connected import.
Branch b1_6
b=16839
i=green
i=bzzz
deshmukh [Wed, 8 Apr 2009 05:28:39 +0000 (05:28 +0000)]
Branch HEAD
b=18882
i=oleg.drokin
bobijam [Wed, 8 Apr 2009 02:43:10 +0000 (02:43 +0000)]
Branch b1_6
b=17378
i=robert.read
add test case: sanity.sh test_17h().
anserper [Tue, 7 Apr 2009 22:15:47 +0000 (22:15 +0000)]
b=18619
i=Vitaly Fertman
i=Oleg Drokin
make oap_async_flags access atomic
grev [Tue, 7 Apr 2009 18:20:09 +0000 (18:20 +0000)]
b=18080
i=Alexander.Zarochentsev
add sanityN test_35 to A_E list
shadow [Tue, 7 Apr 2009 16:34:54 +0000 (16:34 +0000)]
correct count of recoverable clients
Branch b1_6
b=18480
i=tappro
i=johann
shadow [Tue, 7 Apr 2009 15:41:06 +0000 (15:41 +0000)]
allow killing process which wait statahead result
Branch HEAD
b=18902
i=fanyong
i=tappro
isaac [Tue, 7 Apr 2009 15:32:17 +0000 (15:32 +0000)]
b=19087,i=liang:
- removed unused field md_addrkey so that sizeof lnet_libmd_t with one frag would be
exactly 128 on x86_64, very sweet.
maxim [Tue, 7 Apr 2009 11:09:13 +0000 (11:09 +0000)]
b=18414
i=isaac
i=liang
Landing a patch fixing 'running out of ports' issue on HEAD. The patch
essentially does:
- add a delay before next reconnect attempt in ksocklnd in the case of lost race;
- limit the frequency of query-requests in lnet;
- special handling of 'dead peer' notifications in lnet.
zam [Tue, 7 Apr 2009 06:39:30 +0000 (06:39 +0000)]
Branch HEAD
b=18988
i=alexey.lyashkov
i=jinshan.xiong
i=adilger
If append write, vvp_io_write_start sets pos to the end of the file and
doesn't touch kiocb->ki_pos, it triggers BUG_ON(iocb->ki_pos != pos) in Linux
kernel code. This patch adds missing ki_pos adjustment.
robert.read [Tue, 7 Apr 2009 05:05:20 +0000 (05:05 +0000)]
Branch HEAD
b=18554
o=shadow,rread
i=zam
i=eric.mei
Full fix for 18554. Includes original patch, Shadow's new fixes, and a test.
bobijam [Tue, 7 Apr 2009 01:31:54 +0000 (01:31 +0000)]
Branch HEAD
b=17378
i=robert.read
add test case: sanity.sh test_17h()