Whamcloud - gitweb
adilger [Mon, 24 Nov 2008 21:24:27 +0000 (21:24 +0000)]
Branch HEAD
Fix comment whitespace.
adilger [Mon, 24 Nov 2008 21:23:07 +0000 (21:23 +0000)]
Branch HEAD
Add patch description.
adilger [Mon, 24 Nov 2008 19:45:54 +0000 (19:45 +0000)]
Branch HEAD
Add description of the patch.
adilger [Mon, 24 Nov 2008 19:43:18 +0000 (19:43 +0000)]
Branch HEAD
Add brief description of the patch.
tappro [Mon, 24 Nov 2008 17:32:04 +0000 (17:32 +0000)]
- safe operations with recovery queues
b:16711
i:adilger,green
yury [Mon, 24 Nov 2008 16:44:04 +0000 (16:44 +0000)]
b=17631
r=shadow,panda
- fixes long sync bulk unlink in ptlrpcd which stops other rpcs from being handled and also causes asstion in umount time;
- make sure that long unlink wait is done with 1 sec interval to return quickly.
shadow [Mon, 24 Nov 2008 12:26:41 +0000 (12:26 +0000)]
Drop slow OSCs if we can, but not for requested start idx.
This means "if OSC is slow and it is not the requested
start OST, then it can be skipped, otherwise skip it only
if it is inactive/recovering/out-of-space.
Branch HEAD
b=16081
i=shadow
i=green
deshmukh [Mon, 24 Nov 2008 07:25:35 +0000 (07:25 +0000)]
Fixes related to mount failure path cleanup
b=17752
i=umka
i=shadow
nikita [Sun, 23 Nov 2008 20:27:27 +0000 (20:27 +0000)]
use LASSERT_SPIN_LOCKED() in LASSERT(spin_is_locked()) stead as the latter does not work correctly with the uniprocessor kernels. b=17664; i=robert.read
yury [Sun, 23 Nov 2008 13:04:54 +0000 (13:04 +0000)]
b=17750
r=shadow,deen
- fixes writing cookie beyond llcd's boundaries.
yury [Sun, 23 Nov 2008 12:37:46 +0000 (12:37 +0000)]
b=17738
r=zam,shadow
- kill old locks from ldlm pools in convert time before adding them back with new mode.
yury [Sun, 23 Nov 2008 12:27:39 +0000 (12:27 +0000)]
b=17690
r=shadow
- fixes in test_59b
yury [Sun, 23 Nov 2008 12:10:05 +0000 (12:10 +0000)]
b=17751
r=grev
- fixes and cleanups in test_124a
grev [Fri, 21 Nov 2008 21:48:07 +0000 (21:48 +0000)]
b=17735
i=Yury.Umanets
check_mem_leak fn: use echo instead of log fn
yangsheng [Fri, 21 Nov 2008 16:32:27 +0000 (16:32 +0000)]
Branch HEAD
b=17201
i=shadow, bobijam
Update to RHEL5 kernel-2.6.18-92.1.17.el5.
yangsheng [Fri, 21 Nov 2008 15:27:40 +0000 (15:27 +0000)]
file thread.h was initially added on branch b1_6.
yangsheng [Fri, 21 Nov 2008 15:27:39 +0000 (15:27 +0000)]
file thread.c was initially added on branch b1_6.
yangsheng [Fri, 21 Nov 2008 15:27:38 +0000 (15:27 +0000)]
file lshowmount.h was initially added on branch b1_6.
yangsheng [Fri, 21 Nov 2008 15:27:37 +0000 (15:27 +0000)]
file hostlist.h was initially added on branch b1_6.
yangsheng [Fri, 21 Nov 2008 15:27:36 +0000 (15:27 +0000)]
file hostlist.c was initially added on branch b1_6.
yangsheng [Fri, 21 Nov 2008 15:27:35 +0000 (15:27 +0000)]
file hash.h was initially added on branch b1_6.
yangsheng [Fri, 21 Nov 2008 15:27:34 +0000 (15:27 +0000)]
file hash.c was initially added on branch b1_6.
yangsheng [Fri, 21 Nov 2008 15:27:33 +0000 (15:27 +0000)]
file lshowmount.c was initially added on branch b1_6.
yangsheng [Fri, 21 Nov 2008 15:27:31 +0000 (15:27 +0000)]
file lshowmount.8 was initially added on branch b1_6.
vs [Thu, 20 Nov 2008 23:24:31 +0000 (23:24 +0000)]
Branch HEAD
b=17132
i=adilger
Use raid5/6 rhel5 improvements
shadow [Thu, 20 Nov 2008 21:21:23 +0000 (21:21 +0000)]
fix build on sles9.
Branch HEAD
shadow [Thu, 20 Nov 2008 15:19:26 +0000 (15:19 +0000)]
fix 'make rpm'
shadow [Thu, 20 Nov 2008 12:51:23 +0000 (12:51 +0000)]
fix typo.
Branch HEAD
shadow [Thu, 20 Nov 2008 09:25:26 +0000 (09:25 +0000)]
*** empty log message ***
shadow [Thu, 20 Nov 2008 09:24:35 +0000 (09:24 +0000)]
first part of 2.6.26 support (lnet/libcfs part)
Branch HEAD
b=14250
i=maxim
i=deen
liangzhen [Thu, 20 Nov 2008 06:54:16 +0000 (06:54 +0000)]
Branch HEAD
o2iblnd asserts when ib_poll_cq() returns an error
b=17533
i=maxim
i=Christopher Morrone
fanyong [Thu, 20 Nov 2008 05:51:00 +0000 (05:51 +0000)]
Branch HEAD
b=16947
i=h.huang
i=yury.umanets
Hold lli_lock when access lli_sai to prevent NULL pointer.
shadow [Thu, 20 Nov 2008 05:07:25 +0000 (05:07 +0000)]
not send LOV EA under replay, we can't know about they size at this
time. Don't allow client connect to mds before any ost connected,
for avoid problems with LOV EA size and returning EIO to client.
Branch HEAD
b=16080
i=umka
i=tappro
fanyong [Thu, 20 Nov 2008 02:57:49 +0000 (02:57 +0000)]
Branch HEAD
b=13058
i=alexey.lyashkov
i=tianzy
64 bit quota support check when autoconf.
anserper [Thu, 20 Nov 2008 01:19:08 +0000 (01:19 +0000)]
Branch HEAD
b=17371
i=Johann Lombardi
testcase for 17371
vitaly [Wed, 19 Nov 2008 21:28:19 +0000 (21:28 +0000)]
Branch HEAD
b=16129
the test number fix
vitaly [Wed, 19 Nov 2008 21:14:46 +0000 (21:14 +0000)]
Branch b1_8_gate
b=16129
i=adilger
i=green
- a high priority request list is added into service;
- once a lock is canceled, all the IO requests, including coming
ones, under this lock, are moved into this list;
- PING is also added into this list;
- once a lock cancel timeout occurs, the timeout is prolonged
if there is an IO rpc under this lock;
- another request list is added into the export, used to speed up
the rpc-lock matching.
stinson1947 [Wed, 19 Nov 2008 20:03:19 +0000 (20:03 +0000)]
Updated for 2.0 build 07
pravins [Wed, 19 Nov 2008 19:54:00 +0000 (19:54 +0000)]
b=17740
allocate separate dentry for Iterator context.
pravins [Wed, 19 Nov 2008 19:29:49 +0000 (19:29 +0000)]
b=17740
pass NULL to bypass capa while destroying orphan lov object.
grev [Wed, 19 Nov 2008 19:20:13 +0000 (19:20 +0000)]
b=17653
i=Adilger
test_21c fix: restore config
grev [Wed, 19 Nov 2008 19:04:56 +0000 (19:04 +0000)]
b=17477
i=Yury.Umanets
run acc-sm:formatall() only if forced
grev [Wed, 19 Nov 2008 17:56:07 +0000 (17:56 +0000)]
b=17477
i=Adilger
force replay-dual to check and mount MOUNT2
yury [Wed, 19 Nov 2008 14:18:37 +0000 (14:18 +0000)]
b=17686
r=shadow,panda
- fixes race in ptlrpcd caused busy obd and inability to cleanup;
- cleanups and debugs in llcd code.
deshmukh [Wed, 19 Nov 2008 13:22:35 +0000 (13:22 +0000)]
Removing unwanted patches
deshmukh [Wed, 19 Nov 2008 12:59:17 +0000 (12:59 +0000)]
Reverting the changes
deshmukh [Wed, 19 Nov 2008 08:55:54 +0000 (08:55 +0000)]
Land b_head_interop_disk on HEAD (20081119_1314)
b=11826
i=nikita
i=adilger
deshmukh [Wed, 19 Nov 2008 06:47:42 +0000 (06:47 +0000)]
Exported some ldiskfs functions which are required for iop
i=nikita
i=adilger
b=11826
huanghua [Wed, 19 Nov 2008 02:29:18 +0000 (02:29 +0000)]
Branch HEAD
b=17083
i=johann
add warning message while waiting for exports to be destroyed.
nathan [Wed, 19 Nov 2008 01:20:00 +0000 (01:20 +0000)]
b=15699
libcfs proc debug change requires this
yury [Tue, 18 Nov 2008 10:25:14 +0000 (10:25 +0000)]
b=17037
- fixes merge issue in HEAD which caused deadlock on ctxt->loc_sem
anserper [Tue, 18 Nov 2008 04:08:21 +0000 (04:08 +0000)]
Branch HEAD
b=12433
i=Oleg Drokin
i=Yury Umanets
fix imp_inval message
tianzy [Mon, 17 Nov 2008 07:09:51 +0000 (07:09 +0000)]
Branch HEAD
fix a compile error in sles9 and sles10
i=johann
i=nathan
tianzy [Mon, 17 Nov 2008 07:06:45 +0000 (07:06 +0000)]
Branch HEAD
decay qos ost/oss penalties if MDS is not creating objects
i=nathan
i=johann
liangzhen [Mon, 17 Nov 2008 03:10:28 +0000 (03:10 +0000)]
Branch HEAD
remove LASSERT on message type from network
b=17372
i=isaac
i=Maxim
fanyong [Sat, 15 Nov 2008 18:39:35 +0000 (18:39 +0000)]
Land b_head_quota onto HEAD (20081116_0105)
b=13058
i=johann
i=yury.umanets
fanyong [Sat, 15 Nov 2008 15:33:22 +0000 (15:33 +0000)]
Branch b_head_quota
b=13058
i=nikita.danilov
i=johann
i=robert.read
Debug patch for quota on HEAD.
ericm [Fri, 14 Nov 2008 19:54:21 +0000 (19:54 +0000)]
branch: HEAD
reserve some space in mdt_body and mdt_rec_xxx structures and add
MDS_SETATTR64_REC to allow possible expansion of uid/gid to 64 bits
in the future.
b=17601
r=adilger
r=huanghua
r=fanyong
grev [Fri, 14 Nov 2008 19:20:44 +0000 (19:20 +0000)]
b=16488
i=Oleg.Drokin
RACER accs-m test suit
nathan [Fri, 14 Nov 2008 17:24:56 +0000 (17:24 +0000)]
b=15699
i=adilger
i=manoj
move string-based mask setting into a standalone fn; will also be used by changelogs
grev [Fri, 14 Nov 2008 10:28:02 +0000 (10:28 +0000)]
b=17122
i=Adilger
skip sanity test_100 for NETTYPE != tcp
nikita [Thu, 13 Nov 2008 13:44:41 +0000 (13:44 +0000)]
cl_lock_free(): remove assertion that does not work in liblustre.
yury [Thu, 13 Nov 2008 10:15:44 +0000 (10:15 +0000)]
b=17479
r=adilger,behlendorf1
- avoid div/mod in lustre_hash code
green [Thu, 13 Nov 2008 00:42:02 +0000 (00:42 +0000)]
b=16074
r=adilger, Lee Ward
Do not get extra reference on source inode since we already hold the reference
from previous name.
nikita [Wed, 12 Nov 2008 21:00:37 +0000 (21:00 +0000)]
vvp: always write page from the beginning: sptlrpc depends on kiov_offset being 0.
nikita [Wed, 12 Nov 2008 20:59:18 +0000 (20:59 +0000)]
lov_object: simplify and document lov_object life-time rules.
nikita [Wed, 12 Nov 2008 20:58:06 +0000 (20:58 +0000)]
cl_page_list_own(): if page was truncated concurrently, remove it from the queue at once.
grev [Wed, 12 Nov 2008 20:52:41 +0000 (20:52 +0000)]
b=17634
i=Yury.Umanets
insanity cleanup (remove dup fn)
yury [Wed, 12 Nov 2008 19:32:36 +0000 (19:32 +0000)]
b=17310
r=shadow,vitaly
- check rq_phase correctly in ptlrpc_expired_set() and couple of other places.
yury [Wed, 12 Nov 2008 17:33:02 +0000 (17:33 +0000)]
- trivial doxygen fixes
yury [Wed, 12 Nov 2008 17:31:10 +0000 (17:31 +0000)]
b=17037
r=tappro,wangdi
- fixes ost cleanup issue due to missed llcd_put() in the case ost does not receive disconnect from mds;
- do not sleep on hanging llcd. Instead assert on it _after_ stopping recov_thread's ptlrpcd which should kill any remeining llcds;
- fixes and cleanups, comments.
grev [Wed, 12 Nov 2008 16:07:34 +0000 (16:07 +0000)]
b=16488
i=Oleg.Drokin
new runracer script
grev [Wed, 12 Nov 2008 16:06:06 +0000 (16:06 +0000)]
file runracer was initially added on branch b1_8_gate.
yangsheng [Wed, 12 Nov 2008 14:36:24 +0000 (14:36 +0000)]
Branch HEAD
b=17379
As adilger asked, remove the "out_release:" entirely.
grev [Wed, 12 Nov 2008 11:19:59 +0000 (11:19 +0000)]
b=17555
i=Adilger
use current config instead of reformat fs to have single ost
nikita [Wed, 12 Nov 2008 11:16:25 +0000 (11:16 +0000)]
osc_ldlm_glimpse_ast(): keep lock mutex over glimpse processing, for uniformity with other ASTs.
nikita [Wed, 12 Nov 2008 11:15:46 +0000 (11:15 +0000)]
cl_lock_request(): add detailed logging for 17665.
alex [Wed, 12 Nov 2008 08:09:08 +0000 (08:09 +0000)]
b=16918
i=adilger
- fix test checking ROC cache invalidation
nikita [Tue, 11 Nov 2008 11:48:35 +0000 (11:48 +0000)]
cl_object_glimpse(): do not use loop control variable after the loop has terminated.
nikita [Tue, 11 Nov 2008 11:47:41 +0000 (11:47 +0000)]
peacify lockdep during mount.
liangzhen [Tue, 11 Nov 2008 08:47:59 +0000 (08:47 +0000)]
Branch HEAD
Assign msg_ev for GET before lnet_prep_send for REPLY
i=isaac
tianzy [Tue, 11 Nov 2008 07:54:34 +0000 (07:54 +0000)]
Branch HEAD
Remove usage of trace_call_on_all_cpus()
b=15878
i=adilger
i=robert.read
liangzhen [Tue, 11 Nov 2008 07:41:59 +0000 (07:41 +0000)]
Branch HEAD
Allow user to specify small value for o2iblnd::concurrent_sends
b=15983
i=isaac
zhanghc [Tue, 11 Nov 2008 07:12:49 +0000 (07:12 +0000)]
branch HEAD
b=17176
fixed a bug in 14774 patch -- compare peer's nid instead of self's nid
in ptlrpc_connection during select failover MDS/OST nodes
i=deen
pravins [Tue, 11 Nov 2008 06:51:43 +0000 (06:51 +0000)]
b=16461
i=umka
i=h.huang
patch to add lr_padding to lu_range struct. this field will be used by
compact fld work.
yangsheng [Tue, 11 Nov 2008 06:06:02 +0000 (06:06 +0000)]
Branch HEAD
b=17458
i=shadow, bobijam
Update kernel to SLES10 SP2 2.6.16.60-0.31.
zhanghc [Tue, 11 Nov 2008 03:47:05 +0000 (03:47 +0000)]
branch HEAD
b=17495
move the check of recovering state of the OST in osc_precreate
out of "if (oscc->oscc_last_id < oscc->oscc_next_id)" condition
so create operation don't use recovering OST
i=adilger
i=nathan.rutman
nikita [Mon, 10 Nov 2008 18:27:54 +0000 (18:27 +0000)]
remove forgotten debugging code that was causing 17643.
stinson1947 [Mon, 10 Nov 2008 17:03:30 +0000 (17:03 +0000)]
Updated for the next bi-weekly b2.0 build.
yury [Mon, 10 Nov 2008 14:07:22 +0000 (14:07 +0000)]
b=17617 (in fact this is part of 17310 for HEAD)
- fixes manual_umount_client() in conf_sanity.sh to handle --force correctly as it is used by number of tests (this part should be missed in tests porting);
- add assert before decref mgc refcounter;
- fixed possible leaked niduuid in case of error;
- kill @rc in server_put_super() as it is not used;
- use OBD_FAIL_CHECK_ORSET in ptlrpc_check_set() to avoid assert;
- enable test_45 as it is working now.
liuy [Mon, 10 Nov 2008 12:53:52 +0000 (12:53 +0000)]
Branch HEAD
add b_hd_params_tree branch
grev [Mon, 10 Nov 2008 09:49:42 +0000 (09:49 +0000)]
b=17235
i=Alexander.Zarochentsev
cos new test_33a
yury [Sat, 8 Nov 2008 13:31:43 +0000 (13:31 +0000)]
b=17310
r=shadow,johann
- make sure that no new inflight rpcs may come after ptlrpcd_deactivate_import() for both
synchronous and asynchronous sending. To do so we make sure that imp_inflight++ is done only when
permission is granted by ptlrpc_import_delay_req() which makes decision should req be sent,
deferred or killed as import is not in the state to send it in observable future. For async
sending, rpc is only counted inflight when its added to sending or delaying list instead of just
adding it to set for processing.
This fixes assert in ptlrpc_invalidate_import() and as number of other issues;
- synchronize imp_inflight and the presence on sending or delaying list for ptlrpc_queue_wait()
case. So that, now it is guaranteed that if im_inflight != 0 we may always find hanging rpc either
in sending or in delaying list;
- make sure that in ptlrcp_queue_wait() we remove rpc from sending or delaying list and dec
inflight only after ptlrpc_unregister_reply() is done. This way we make sure that accounting is
correct. Rpc can't be returned to the pool or counted finished until lnet lets us go with finished
reply unlink;
- check for inflight and rq_list in pinger;
- comments, cleanups;
nikita [Sat, 8 Nov 2008 13:18:30 +0000 (13:18 +0000)]
recovery-small.sh: test_55: fix obvious typos.
nikita [Fri, 7 Nov 2008 23:54:43 +0000 (23:54 +0000)]
land clio.
b=14166
yury [Fri, 7 Nov 2008 22:37:57 +0000 (22:37 +0000)]
b=17511
r=johann,adilger
- removes deadlock possibility by disabling rehash in hash_del() operations and moving hash_add()
out of spin_locks when calling. Hash table has own mechanisms for protecting its structures and it
also has hash_add_unique() method for using in concurrent run contexts;
- fixed missed lh_put() in hash_add_unique() which led to extra refs in some cases (extra ref to
export) and inability to cleanup;
- fixed __lustre_hash_set_theta() which set @max theta into ->lh_min_theta;
- in lustre_hash_rehash_size() disable rehash also for the case when new and old hash sizes equal
in corner cases (max_size or min_size). Before this fix it could be possible to do needless
rehashes when size is actually did not change but we do this expensive operation;
- disable rehash in hash_add_unique() if no actual add happened since entry with the same key is
already found in the table;
- some cleanups in hash table code;
grev [Fri, 7 Nov 2008 20:35:25 +0000 (20:35 +0000)]
b=17477
i=Adilger
replace cleanup_and_setup_lustre fn by check_and_setup_lustre fn
zam [Fri, 7 Nov 2008 17:57:56 +0000 (17:57 +0000)]
Branch HEAD
b=17235
i=adigler@sun.com
add -r 'deletepath' option to createmany.c; createmany.c code cleanup.
grev [Fri, 7 Nov 2008 17:15:13 +0000 (17:15 +0000)]
b=17477
i=Adilger
check config if lustre is mounted before acc-sm run
grev [Fri, 7 Nov 2008 16:34:28 +0000 (16:34 +0000)]
b=14384
i=Brian
assert_DIR cleanup