Whamcloud - gitweb
nikita [Mon, 30 Oct 2006 17:29:50 +0000 (17:29 +0000)]
0. increase OST_MIN_PRECREATE; 1. move fid_res_name_eq() to lustre_fid.h and use offset constants in it.
yury [Mon, 30 Oct 2006 16:19:24 +0000 (16:19 +0000)]
- use usual MDT name in mdd procfs stuff instead of mdd-0;
- make mdc/mdc_is_subdir() static to not confuse it with same name in cmm.
wangdi [Mon, 30 Oct 2006 15:44:13 +0000 (15:44 +0000)]
Branch:b_new_cmd
fix the len bug in cmm_split_entries, the original method has some problem
when entries are not full in the page
fanyong [Mon, 30 Oct 2006 10:31:28 +0000 (10:31 +0000)]
Small fix.
yury [Mon, 30 Oct 2006 10:21:05 +0000 (10:21 +0000)]
- fixed small typo in cmm_split_remove_page()
yury [Mon, 30 Oct 2006 10:07:45 +0000 (10:07 +0000)]
- handle empty page in cmm_split_remove_page() at Tom's suggestion;
- fixes about le_to_cpu in mdt_write_dir_page()
wangdi [Mon, 30 Oct 2006 09:54:23 +0000 (09:54 +0000)]
Branch: b_new_cmd
not root dir, it should be NOT expect splitted.
yury [Mon, 30 Oct 2006 09:01:43 +0000 (09:01 +0000)]
- cleanups in split - function names are changed to make split understandable not only for author;
- in cml_create() check if name came to correct MDT only if cmm_split_try() returned -EALREADY, that is dir is split. In other cases that is not needed and only slows things down;
- comments, allocating new fid and creating object for slaves is united to one function.
fanyong [Mon, 30 Oct 2006 03:27:22 +0000 (03:27 +0000)]
Since have called mdd_la_get before and held lock yet,
call __mdd_permission_internal directly.
fanyong [Mon, 30 Oct 2006 02:31:53 +0000 (02:31 +0000)]
Local client should not use "lfs getfacl / lfs setfacl".
yury [Sun, 29 Oct 2006 20:05:14 +0000 (20:05 +0000)]
- disable slow test
yury [Sun, 29 Oct 2006 19:48:28 +0000 (19:48 +0000)]
- more fixes in split, now 24o works and split is "counted" as bug free ;) at least I do not know any of them.
yury [Sun, 29 Oct 2006 19:03:49 +0000 (19:03 +0000)]
- fixes in split. Tom! Your changes in cmm_remove_entries() were wrong;
yury [Sun, 29 Oct 2006 14:56:53 +0000 (14:56 +0000)]
- removed debug stuff.
fanyong [Sun, 29 Oct 2006 14:19:19 +0000 (14:19 +0000)]
(1) "lfs getfacl" should return base file mode anyway.
(2) make "lfs setfacl" more like local client setfacl.
fanyong [Sun, 29 Oct 2006 13:12:19 +0000 (13:12 +0000)]
For remote client, return directly when lfs getfacl / lfs setfacl
if server do not support ACL.
fanyong [Sun, 29 Oct 2006 11:53:51 +0000 (11:53 +0000)]
Drop unnecessary warning message.
wangdi [Sun, 29 Oct 2006 11:07:02 +0000 (11:07 +0000)]
Branch:b_new_cmd
add group num when set attr to OST
wangdi [Sun, 29 Oct 2006 07:11:22 +0000 (07:11 +0000)]
Branch:b_new_cmd
cleanup in cmm split, after pdo added.
huanghua [Sat, 28 Oct 2006 15:37:19 +0000 (15:37 +0000)]
some code cleanup in split.
nikita [Sat, 28 Oct 2006 13:31:56 +0000 (13:31 +0000)]
procsleep patch: on statistics reset, allocate info if necessary.
yury [Sat, 28 Oct 2006 11:28:05 +0000 (11:28 +0000)]
- fixed ma_valid in cmm_mdsnum_check();
- fixes in cmm_try_to_split();
- added error messages to cmm_try_to_split();
- cache mdsnum for lmv slaves so that we need not to lookup them all the time. This reduces number of fld cache lookups from ~46000 to ~12000 on sanity 24o.
ericm [Sat, 28 Oct 2006 04:25:03 +0000 (04:25 +0000)]
branch: b_new_cmd
- replace spinlock with rwlock for policy table protection.
- fix concurrent load ptlrpc_gss module.
- minor message adjustment.
nikita [Fri, 27 Oct 2006 23:48:55 +0000 (23:48 +0000)]
kernel_patches: add statistics clearing for /proc/sleep.
nikita [Fri, 27 Oct 2006 23:48:18 +0000 (23:48 +0000)]
iam: add -EAGAIN handling for index split.
tappro [Fri, 27 Oct 2006 21:38:53 +0000 (21:38 +0000)]
- filter MDS_DONE_WRITING during recovery
yury [Fri, 27 Oct 2006 17:56:59 +0000 (17:56 +0000)]
- removed obsolete debug stuff.
huanghua [Fri, 27 Oct 2006 16:14:47 +0000 (16:14 +0000)]
more efficient unlinkmany
wangdi [Fri, 27 Oct 2006 10:22:50 +0000 (10:22 +0000)]
Branch:b_new_cmd
add debug info for b11123
huanghua [Fri, 27 Oct 2006 08:59:39 +0000 (08:59 +0000)]
Be sure not to take PDO lock for non-existing or remote directory.
yury [Fri, 27 Oct 2006 07:11:39 +0000 (07:11 +0000)]
- drop not directory dentries in ll_drop_dentry(). Let's see if it helps with OOM on clients.
wangdi [Fri, 27 Oct 2006 06:10:24 +0000 (06:10 +0000)]
Branch: b_new_cmd
add this ldlm debug info temporarily
wangdi [Fri, 27 Oct 2006 06:06:28 +0000 (06:06 +0000)]
Branch: b_new_cmd
add gr in these 2 places
nikita [Fri, 27 Oct 2006 00:14:46 +0000 (00:14 +0000)]
iam: fix possible cause of dynlock-double-unlock.
nikita [Fri, 27 Oct 2006 00:14:13 +0000 (00:14 +0000)]
dynlocks: a bit of cleanup.
yury [Thu, 26 Oct 2006 14:51:51 +0000 (14:51 +0000)]
- cleanups in mdt_object_set_mode()
nikita [Thu, 26 Oct 2006 14:00:30 +0000 (14:00 +0000)]
osd: remove no longer needed commentary.
nikita [Thu, 26 Oct 2006 13:59:55 +0000 (13:59 +0000)]
move LU_OBJECT_HEARD_BANSHEE into mdd_object.mod_flags.
yury [Thu, 26 Oct 2006 13:39:49 +0000 (13:39 +0000)]
- fixes in split:
- do not take pdo lock for non-splittable dirs if not modification is foreseen;
- added assert to split code which checks that split should only be done in EX lock is taken.
yury [Thu, 26 Oct 2006 11:35:10 +0000 (11:35 +0000)]
- use LCK_XX locks instead of MDT specified macros.
yury [Thu, 26 Oct 2006 10:28:40 +0000 (10:28 +0000)]
- debug message in split code. Cleanups.
- use less controversial MDL -> DLM locks converting style.
yury [Thu, 26 Oct 2006 07:32:29 +0000 (07:32 +0000)]
- remove commentted out stuff about ll_iget(), check for I_FREEING before updating inode by data from server.
fanyong [Thu, 26 Oct 2006 06:10:08 +0000 (06:10 +0000)]
Do little-endian and big-endian convert for ACL, make it more compatible.
nikita [Thu, 26 Oct 2006 00:36:03 +0000 (00:36 +0000)]
osd: fix thinko.
nikita [Thu, 26 Oct 2006 00:18:13 +0000 (00:18 +0000)]
osd: 0. remove mp longer used functions; 1. add osd-level per-object locking necessary due to pdirops; 2. add locking description.
nikita [Thu, 26 Oct 2006 00:16:04 +0000 (00:16 +0000)]
more silencing.
tappro [Wed, 25 Oct 2006 19:00:33 +0000 (19:00 +0000)]
- fix mds-mds connection. There was missed IBITS, ACL connetion flags and missed
ibits_known field
- remove second connection to the mdc0 from seq client, use the same export as
mdc_device does.
wangdi [Wed, 25 Oct 2006 18:06:17 +0000 (18:06 +0000)]
Branch: b_new_cmd
serval group number fixes in ost, which may fix read-ahead bug in cmd37
nikita [Wed, 25 Oct 2006 17:26:41 +0000 (17:26 +0000)]
iam/pdirops: remove obsolete comment.
nikita [Wed, 25 Oct 2006 17:26:20 +0000 (17:26 +0000)]
iam: little cleanup.
nikita [Wed, 25 Oct 2006 17:25:50 +0000 (17:25 +0000)]
dynlocks: dynlock_unlock(): print invalid magic before panicking.
nikita [Wed, 25 Oct 2006 16:58:06 +0000 (16:58 +0000)]
mdt: add const to make compiler happy.
nikita [Wed, 25 Oct 2006 16:55:18 +0000 (16:55 +0000)]
disable expensive assert.
nikita [Wed, 25 Oct 2006 16:55:01 +0000 (16:55 +0000)]
fix comment to match reality.
nikita [Wed, 25 Oct 2006 16:54:39 +0000 (16:54 +0000)]
make lustre less noisy during normal operation.
wangdi [Wed, 25 Oct 2006 13:28:40 +0000 (13:28 +0000)]
Branch: b_new_cmd
change this debug level to VFSTRACE temporarily for debuging read-ahead
wangdi [Wed, 25 Oct 2006 13:24:34 +0000 (13:24 +0000)]
Branch: b_new_cmd
unlink dir still need some credits for del the index and unlink the object
yury [Wed, 25 Oct 2006 13:01:23 +0000 (13:01 +0000)]
- fixes in mdt_rename_lock()
yury [Wed, 25 Oct 2006 12:50:57 +0000 (12:50 +0000)]
- use FL_LOCAL for big rename lock, it is never going to be sent to other nodes.
tappro [Wed, 25 Oct 2006 10:37:59 +0000 (10:37 +0000)]
- use req_xid_is_last()
- fix mdt_req_from_mcd() for MDS_DONE_WRITING
yury [Wed, 25 Oct 2006 10:09:09 +0000 (10:09 +0000)]
- added two asserts to catch test 11 issue.
fanyong [Wed, 25 Oct 2006 07:30:05 +0000 (07:30 +0000)]
Small fix.
fanyong [Wed, 25 Oct 2006 03:01:24 +0000 (03:01 +0000)]
Make the code more compatible.
vitaly [Tue, 24 Oct 2006 20:44:41 +0000 (20:44 +0000)]
Branch b_new_cmd
do not take credits for unlinking directories.
nikita [Tue, 24 Oct 2006 13:21:19 +0000 (13:21 +0000)]
iam: iam_it_collision(): release path before repeating lookup (fid to buffalo failure).
yury [Tue, 24 Oct 2006 13:00:30 +0000 (13:00 +0000)]
- fixed test 26 in recovery-small
yury [Tue, 24 Oct 2006 12:27:15 +0000 (12:27 +0000)]
- do not call any obd_fid_init() from llite, do that from lmv only;
- handle adding new mdc in lvm correctly from POV of seq init;
- move tgt disconnect to lmv_disconnect_mdc(), call obd_fid_fini() from it;
- deny adding new fld target on fly, continue to work with old targets set (hack until DHT is not used);
- fix sanity 24g, use mdd_is_root() for root comparisons as lu_fid_eq() can't be used fo root fids.
wangdi [Tue, 24 Oct 2006 11:50:28 +0000 (11:50 +0000)]
Branch: b_new_cmd
we should update cat log hdr only after we successfully create the new log,
otherwise the index will be wrong, when disk full, which maybe cause lbug later
yury [Tue, 24 Oct 2006 07:46:04 +0000 (07:46 +0000)]
- remove pdirops defines. Now it is active all the time without any keys in configure.
huanghua [Tue, 24 Oct 2006 07:33:24 +0000 (07:33 +0000)]
(1) more restriction on fid_is_sane();
(2) use precise symlink length.
nikita [Tue, 24 Oct 2006 06:38:01 +0000 (06:38 +0000)]
lvfs: fsfilt_ext3_read(): return actually read size.
nikita [Tue, 24 Oct 2006 05:21:32 +0000 (05:21 +0000)]
osd,mdd: move osd_readpage() and friends into mdd layer. Remove ->do_readpage() method from dt_object_operations.
ericm [Mon, 23 Oct 2006 22:49:31 +0000 (22:49 +0000)]
branch: b_new_cmd
move configuration of capa/pag/gss back to lustre-core.
nikita [Mon, 23 Oct 2006 22:26:01 +0000 (22:26 +0000)]
osd: implement osd_{read,write}() on top of fsfilt_ldiskfs buffer-head-based versions.
nikita [Mon, 23 Oct 2006 18:34:38 +0000 (18:34 +0000)]
kernel_patches: dynlocks: a bit of cleanups.
nikita [Mon, 23 Oct 2006 18:33:30 +0000 (18:33 +0000)]
ldlm: ldlm_lock_debug(): fix it so it prints something (from Andreas via Oleg).
nikita [Mon, 23 Oct 2006 18:29:47 +0000 (18:29 +0000)]
pdirops: quick fix to enable pdirops unconditionally.
yury [Mon, 23 Oct 2006 16:56:16 +0000 (16:56 +0000)]
- more cleanups in pdo
nikita [Mon, 23 Oct 2006 15:59:57 +0000 (15:59 +0000)]
dynlocks: fix assertion that does not take read-locking into account.
yury [Mon, 23 Oct 2006 15:48:14 +0000 (15:48 +0000)]
- small cleanups in mdt_object_lock()
nikita [Mon, 23 Oct 2006 15:36:49 +0000 (15:36 +0000)]
iam: tentative fix for b=11098.
wangdi [Mon, 23 Oct 2006 13:57:00 +0000 (13:57 +0000)]
Branch: b_new_cmd
since we already set no_need_trans for split. so remove this ugly hack
for split.
tappro [Mon, 23 Oct 2006 13:37:51 +0000 (13:37 +0000)]
- fix qos_maxage issue. It was uninitialized.
yury [Mon, 23 Oct 2006 12:08:59 +0000 (12:08 +0000)]
- small fixes, cleanups, comments in pdirops code in mdt.
nikita [Mon, 23 Oct 2006 10:18:22 +0000 (10:18 +0000)]
osd: osd_dir_page_build(): cleanup BE/cpu/LE conversion.
nikita [Mon, 23 Oct 2006 10:17:54 +0000 (10:17 +0000)]
mdd: cleanup naming: fid is record rather than key.
nikita [Mon, 23 Oct 2006 10:17:24 +0000 (10:17 +0000)]
iam: lvar_leaf(): zero out dummy record.
wangdi [Mon, 23 Oct 2006 10:02:39 +0000 (10:02 +0000)]
Branch: b_new_cmd
fix previous commit bugs, since lookup also use it.
huanghua [Mon, 23 Oct 2006 09:24:48 +0000 (09:24 +0000)]
fid need convertion.
wangdi [Mon, 23 Oct 2006 09:11:02 +0000 (09:11 +0000)]
Branch: b_new_cmd
fix typo here, we should use BE for rec in iam
yury [Mon, 23 Oct 2006 08:34:54 +0000 (08:34 +0000)]
- fixed typo in fid_be_to_cpu()
huanghua [Mon, 23 Oct 2006 07:49:04 +0000 (07:49 +0000)]
using big endian format for fid in (name->fid) index.
vitaly [Sun, 22 Oct 2006 20:43:43 +0000 (20:43 +0000)]
Branch b_new_cmd
do not pack MA_LOV on unlink if nlink > 1
yury [Sun, 22 Oct 2006 18:50:12 +0000 (18:50 +0000)]
- a bit more sanity checks
yury [Sun, 22 Oct 2006 18:43:45 +0000 (18:43 +0000)]
- added last chunk of ldlm part of pdiros, enabled by default, use --disable-pdirops key to disble it;
- fixes in mdt_is_subdir();
- added few asserts checking f_ver invariants.
vitaly [Sun, 22 Oct 2006 18:14:08 +0000 (18:14 +0000)]
Branch b_new_cmd
'./stat' command could not be found
wangdi [Sun, 22 Oct 2006 09:28:41 +0000 (09:28 +0000)]
Branch:b_new_cmd
when stripe_count == -1, we should retrieve the ost_count as the
stripe_count when computing the log credits
tappro [Sun, 22 Oct 2006 08:23:47 +0000 (08:23 +0000)]
- use full_name_hash()
wangdi [Sun, 22 Oct 2006 05:33:35 +0000 (05:33 +0000)]
Branch: b_new_cmd
clear MA_LOV | MA_COOKIE, if we do not unlink it. since we get it
in the begining of unlink for retrieving the lsm_stripe_count for
unlink credits
wangdi [Sun, 22 Oct 2006 04:08:07 +0000 (04:08 +0000)]
Branch: b_new_cmd
correct handling mdd pdir locks in mdd_rename