Whamcloud - gitweb
fs/lustre-release.git
17 years agoDefensive check to avoid re-freeing export during force cleanup.
adilger [Thu, 5 Aug 2004 23:12:23 +0000 (23:12 +0000)]
Defensive check to avoid re-freeing export during force cleanup.

17 years agoRemove duplicate but unused declaration of mds_pack_open_flags().
adilger [Thu, 5 Aug 2004 23:11:16 +0000 (23:11 +0000)]
Remove duplicate but unused declaration of mds_pack_open_flags().

17 years agoMore verbose debugging for llog replay assertion.
adilger [Thu, 5 Aug 2004 23:07:30 +0000 (23:07 +0000)]
More verbose debugging for llog replay assertion.
b=4033

17 years ago- we need not that CWARN()s
alex [Wed, 4 Aug 2004 14:41:38 +0000 (14:41 +0000)]
- we need not that CWARN()s

17 years agofile lutil.h was initially added on branch b_newsysio.
ericm [Tue, 3 Aug 2004 00:52:49 +0000 (00:52 +0000)]
file lutil.h was initially added on branch b_newsysio.

17 years agofile lutil.c was initially added on branch b_newsysio.
ericm [Tue, 3 Aug 2004 00:52:47 +0000 (00:52 +0000)]
file lutil.c was initially added on branch b_newsysio.

17 years agob=4059
alex [Mon, 2 Aug 2004 23:39:02 +0000 (23:39 +0000)]
b=4059

- race in lmv_check_connect() has been fixed: few threads might try to
  establish LMV's connects to MDCs and some of them might find not-yet-
  -connected LMV operational

17 years agob=4030
alex [Mon, 2 Aug 2004 22:18:37 +0000 (22:18 +0000)]
b=4030
r=adilger

- new obd method ->o_connect_post() is called when reverse import
  gets initialized
- method ->o_llog_connect() gets obd_export, not obd_device
- MDS passes mdsnum in connect_flags slot
- OST saves last known group in LAST_GROUP file
- OST initializes all known groups upon startup, not on demand
- filter_connect_post() initializes context for given group upon
  connection from MDS
- filter_connect() initializes group from connect_flags
- osc_set_info() doesn't forward "mds_conn": no need to do this any more

17 years ago1)reorganize the smfs hook ops to make smfs walk a list of hooks ops in hook macro
wangdi [Mon, 2 Aug 2004 14:58:13 +0000 (14:58 +0000)]
1)reorganize the smfs hook ops to make smfs walk a list of hooks ops in hook macro
2)remove clonefs in llite, we do not need them now
3)add snap_dot_info in snap_super_info

17 years agob=4054
alex [Mon, 2 Aug 2004 14:39:03 +0000 (14:39 +0000)]
b=4054

- in RESENT case mds_getattr_name() doesn't take dentry for parent.
  so, we have no mea to check mds validness

17 years ago- comment fixed
alex [Sat, 31 Jul 2004 14:56:29 +0000 (14:56 +0000)]
- comment fixed

17 years agob=4051
alex [Sat, 31 Jul 2004 14:49:46 +0000 (14:49 +0000)]
b=4051
r=adilger

- lock inversion between mds_reint_unlink() and mdt_obj_create() fixed:
  mdt_obj_create() unlinks object, so it doesn't matter what directory
  to create objects in

17 years agob=4049
alex [Sat, 31 Jul 2004 14:43:41 +0000 (14:43 +0000)]
b=4049
r=adilger

- fix lock inversion by removing DLM lock acquiring in mdt_obj_create()
  instead just mark request difficult explicitly

17 years agob=4044
alex [Sat, 31 Jul 2004 14:35:46 +0000 (14:35 +0000)]
b=4044
r=adilger

- few clients connecting simultaneously might race in mds_lmv_connect().
  just take a sem around code that connects to LMV

17 years agor=adilger
alex [Sat, 31 Jul 2004 14:31:03 +0000 (14:31 +0000)]
r=adilger

- possible name collision in mdt_obj_create() fixed

17 years agoFix error handling.
adilger [Fri, 30 Jul 2004 21:53:55 +0000 (21:53 +0000)]
Fix error handling.

17 years agoGrab a lov llog lock so we ensure the llog records for a single transno
adilger [Fri, 30 Jul 2004 21:53:14 +0000 (21:53 +0000)]
Grab a lov llog lock so we ensure the llog records for a single transno
are allocated in the same order they will be replayed in.  Concurrent
unlinks might otherwise have misordered llog records and not be replayable.
b=4033
r=andreas (for Tian)

17 years ago- ext3_check_dir_entry shouldn't panic if inum in cross-node dentry
alex [Fri, 30 Jul 2004 20:20:25 +0000 (20:20 +0000)]
- ext3_check_dir_entry shouldn't panic if inum in cross-node dentry
  larger than local filesystem may have

17 years agob=3825
rread [Fri, 30 Jul 2004 18:08:56 +0000 (18:08 +0000)]
b=3825

typo

17 years agoMake recovery on clients less verbose and misleading.
rread [Thu, 29 Jul 2004 20:29:03 +0000 (20:29 +0000)]
Make recovery on clients less verbose and misleading.

17 years agob=3554
rread [Thu, 29 Jul 2004 20:23:26 +0000 (20:23 +0000)]
b=3554

Use cond_resched() ll_pgcache_remove_extent, and make sure we've got
the definition for it.

17 years agob=3986
alex [Thu, 29 Jul 2004 09:07:47 +0000 (09:07 +0000)]
b=3986
r=adilger

- we can't initialize few groups in parallel because
  filter_read_group_internal() reallocates bunch of
  global-per-obd structures. so, here is simple patch
  to serialize it. indeed, this isn's very scalable
  to hundreds of MDSs, but I think it's enough for a
  while. tested on mountain and my home cluster

17 years ago- just to avoid gcc warning
alex [Wed, 28 Jul 2004 21:25:47 +0000 (21:25 +0000)]
- just to avoid gcc warning

17 years agob=3825
alex [Wed, 28 Jul 2004 21:22:30 +0000 (21:22 +0000)]
b=3825
r=robert

  landed on HEAD

17 years agob=4019
alex [Wed, 28 Jul 2004 19:17:32 +0000 (19:17 +0000)]
b=4019
r=adilger

 - obd_connect() gets one more arg: connect flags
 - client_connect_import() puts this flag in imp_connect_flags
 - ptlrpc_connect_import() sends this flag to peer
 - mds_lmv_connect() passes OBD_OPT_MDS_CONNECTION when connects to MDS
 - mds_connect() looks for OBD_OPT_MDS_CONNECTION and treats
   connection with this flag as MDS-to-MDS and doesn't try
   to connect to own LMV

 this way connection to LMV gets delayed and MDS reconnects to LMV
 after recovery abort

17 years agofile small_scatterlist-2.4.21-chaos.patch was initially added on branch b1_4_eeb_perf.
eeb [Wed, 28 Jul 2004 18:09:53 +0000 (18:09 +0000)]
file small_scatterlist-2.4.21-chaos.patch was initially added on branch b1_4_eeb_perf.

17 years agofile blkdev_tunables-2.4.21-chaos.patch was initially added on branch b1_4_eeb_perf.
eeb [Wed, 28 Jul 2004 18:09:52 +0000 (18:09 +0000)]
file blkdev_tunables-2.4.21-chaos.patch was initially added on branch b1_4_eeb_perf.

17 years ago- lmv.sh config generator may add several OST nodes to setup as specified
alex [Tue, 27 Jul 2004 15:37:43 +0000 (15:37 +0000)]
- lmv.sh config generator may add several OST nodes to setup as specified
  in OSTCOUNT environment variable

17 years agob=3999
alex [Tue, 27 Jul 2004 15:35:06 +0000 (15:35 +0000)]
b=3999

- IT_UNLINK doesn't use 4rd buffer

17 years agob=3550
alex [Tue, 27 Jul 2004 12:38:12 +0000 (12:38 +0000)]
b=3550
r=adilger

- replay of unlink. landed for testing on china cluster

17 years ago- don't show error message in replay case
alex [Tue, 27 Jul 2004 12:32:15 +0000 (12:32 +0000)]
- don't show error message in replay case

17 years agob=3995
alex [Tue, 27 Jul 2004 12:17:06 +0000 (12:17 +0000)]
b=3995

- mdc_enqueue() saved wrong easize for IT_OPEN replay

17 years agob=3983
alex [Mon, 26 Jul 2004 22:50:10 +0000 (22:50 +0000)]
b=3983

brown paper bug have been fixed

17 years agob=3550
rread [Mon, 26 Jul 2004 16:20:25 +0000 (16:20 +0000)]
b=3550

Preemptive LASSERT to make sure replay creates the correct
inode here.

17 years agoUpdate snapfs: 1) some fix on clonefs read in lustre
wangdi [Mon, 26 Jul 2004 13:03:28 +0000 (13:03 +0000)]
Update snapfs: 1) some fix on clonefs read in lustre

17 years agob=3983
alex [Mon, 26 Jul 2004 11:20:43 +0000 (11:20 +0000)]
b=3983

- adds a bit more balanced hash that takes all chars into account
- stores hash type in mea
- converts old non-marked mea to new one with magic

17 years ago- more info in test 1c
alex [Mon, 26 Jul 2004 11:14:15 +0000 (11:14 +0000)]
- more info in test 1c

17 years ago- mds_reint_unlink() shouldn't kill slaves if master directory can't be unlinked
alex [Mon, 26 Jul 2004 10:59:34 +0000 (10:59 +0000)]
- mds_reint_unlink() shouldn't kill slaves if master directory can't be unlinked

17 years agob=3874
alex [Sat, 24 Jul 2004 12:10:35 +0000 (12:10 +0000)]
b=3874

- this BKL collides with jbd a lot. remove it.

17 years agob=3874
alex [Sat, 24 Jul 2004 12:06:50 +0000 (12:06 +0000)]
b=3874

- avoid needless fsfilt_get_md() and taking i_sem on a non-splittable parent

17 years agob=3633,3934
rread [Fri, 23 Jul 2004 18:06:05 +0000 (18:06 +0000)]
b=3633,3934

debugging

17 years agob=3899
rread [Thu, 22 Jul 2004 19:45:34 +0000 (19:45 +0000)]
b=3899

Handle server recovery in a new thread, instead of tying up request
processing threads. Centralizes the recovery management, including the
abort recovery checking.

17 years agob=3934
rread [Thu, 22 Jul 2004 19:33:03 +0000 (19:33 +0000)]
b=3934

Add some more debugging.

17 years agob=3892
rread [Thu, 22 Jul 2004 17:43:21 +0000 (17:43 +0000)]
b=3892

Move the transno update so it's done after we check for the -ENOTCONN
reply.

17 years agob=2525
rread [Thu, 22 Jul 2004 16:30:26 +0000 (16:30 +0000)]
b=2525

I've seen the "server went back in time" message while testing HEAD,
so this will fix that.  Patch already on 1.4.

17 years agob=3920
alex [Thu, 22 Jul 2004 15:27:42 +0000 (15:27 +0000)]
b=3920

bits from 3599 landed to help debugging

17 years ago- PTLDEBUG var to control debug set
alex [Thu, 22 Jul 2004 15:24:25 +0000 (15:24 +0000)]
- PTLDEBUG var to control debug set

17 years agob=2185
alex [Thu, 22 Jul 2004 09:19:49 +0000 (09:19 +0000)]
b=2185

- ext3-inode-reuse is part of vanilla-2.4.24 series
- MDS doesn't take lock to prevent inode recreation
  (this improves performance a bit)

17 years ago- ost_brw_write() should return an error rather than uninitialized npages
alex [Thu, 22 Jul 2004 09:08:10 +0000 (09:08 +0000)]
- ost_brw_write() should return an error rather than uninitialized npages

17 years agob=3940
rread [Thu, 22 Jul 2004 06:36:02 +0000 (06:36 +0000)]
b=3940

Corrected. OBD_MD_MDS should only be set when the directory was
created on a different node.

17 years agob=3362
wangchao [Thu, 22 Jul 2004 06:07:20 +0000 (06:07 +0000)]
b=3362
r=braam
update lctl.lyx

17 years agob=3399
wangchao [Thu, 22 Jul 2004 06:04:07 +0000 (06:04 +0000)]
b=3399
r=braam
update lmc.lyx and lconf.lyx

17 years agob=3940
rread [Thu, 22 Jul 2004 06:02:01 +0000 (06:02 +0000)]
b=3940

revert while i figure out what i did wrong.

17 years agob=3940
rread [Thu, 22 Jul 2004 05:25:49 +0000 (05:25 +0000)]
b=3940

Also reconstruct what we can in the body, which is hopefully enough.

17 years agoadd uml_clone.sh for clonefs test
wangdi [Wed, 21 Jul 2004 18:32:29 +0000 (18:32 +0000)]
add uml_clone.sh for clonefs test

17 years ago1)add clonefs read in lustre
wangdi [Wed, 21 Jul 2004 18:28:32 +0000 (18:28 +0000)]
1)add clonefs read in lustre
2)add clientfsoption in lmc and lconf
3)some fix in cobd for cache miss handler

17 years agofile lustrefs was initially added on branch b1_2_smallfix.
jacob [Wed, 21 Jul 2004 00:32:57 +0000 (00:32 +0000)]
file lustrefs was initially added on branch b1_2_smallfix.

17 years agofile rhel-2.4.target was initially added on branch b1_2_smallfix.
jacob [Wed, 21 Jul 2004 00:26:04 +0000 (00:26 +0000)]
file rhel-2.4.target was initially added on branch b1_2_smallfix.

17 years agofile x86-fpu-crash.patch was initially added on branch b1_2_smallfix.
jacob [Wed, 21 Jul 2004 00:26:03 +0000 (00:26 +0000)]
file x86-fpu-crash.patch was initially added on branch b1_2_smallfix.

17 years agofile qlogic-suse-2.4.21-2.patch was initially added on branch b1_2_smallfix.
jacob [Wed, 21 Jul 2004 00:26:02 +0000 (00:26 +0000)]
file qlogic-suse-2.4.21-2.patch was initially added on branch b1_2_smallfix.

17 years agofile kernel-2.4.21-rhel-2.4-i686.config was initially added on branch b1_2_smallfix.
jacob [Wed, 21 Jul 2004 00:26:01 +0000 (00:26 +0000)]
file kernel-2.4.21-rhel-2.4-i686.config was initially added on branch b1_2_smallfix.

17 years agofile kernel-2.4.21-rhel-2.4-i686-smp.config was initially added on branch b1_2_smallfix.
jacob [Wed, 21 Jul 2004 00:26:00 +0000 (00:26 +0000)]
file kernel-2.4.21-rhel-2.4-i686-smp.config was initially added on branch b1_2_smallfix.

17 years agob=3896
alex [Tue, 20 Jul 2004 18:01:05 +0000 (18:01 +0000)]
b=3896

oscc_internal_create() increase pre-creation window smoothly

17 years agob=3869
rread [Tue, 20 Jul 2004 16:55:07 +0000 (16:55 +0000)]
b=3869
r=bzzz

Reconstruct create reply for CROSS_REF directories from the dentry.

17 years agob=3869
rread [Tue, 20 Jul 2004 15:44:20 +0000 (15:44 +0000)]
b=3869

use the correct bug#

17 years agob=3892
rread [Tue, 20 Jul 2004 06:29:01 +0000 (06:29 +0000)]
b=3892

more diagnostics.

17 years agob=3488
rread [Tue, 20 Jul 2004 04:54:12 +0000 (04:54 +0000)]
b=3488
more debugging

17 years agob=3887
rread [Mon, 19 Jul 2004 22:55:28 +0000 (22:55 +0000)]
b=3887

renmae 2nd ptlrcd "ptlrpcd-recov"

17 years agob=3886
rread [Mon, 19 Jul 2004 22:49:58 +0000 (22:49 +0000)]
b=3886

A small races exists between receiving a reply and timeout. This diff
prevents ptlrpc_expire_one_request from timing out a req that has
already received the reply.

17 years agob=3885
rread [Mon, 19 Jul 2004 22:38:40 +0000 (22:38 +0000)]
b=3885

THe hack on b1_2 branch to call set_info("mds_conn") on a specific OSC
instead of all of them was removed on HEAD, and this broke
mds_lov_synchronize.  Instead of introducing a new hack to do this,
for now this changes m_l_s to send the set_info() to all the connected
OSTs. This should be harmless.

17 years agob=3874
alex [Mon, 19 Jul 2004 11:57:30 +0000 (11:57 +0000)]
b=3874

- with CMD each MDS has own group on OST, so group != 0 always. this is
  why filter_should_precreate() did return 1 always, so precreation path
  didn't take a place. I've comment out check for group out for a while

17 years ago- lmv_disconnect() should drop connected flag upon real disconnecting only
alex [Mon, 19 Jul 2004 09:45:03 +0000 (09:45 +0000)]
- lmv_disconnect() should drop connected flag upon real disconnecting only
- lmv_disconnect() should drop connected flag upon real disconnecting only
- lmv_disconnect() should drop connected flag upon real disconnecting only
- lmv_disconnect() should drop connected flag upon real disconnecting only
- lmv_disconnect() should drop connected flag upon real disconnecting only
- lmv_disconnect() should drop connected flag upon real disconnecting only
- lmv_disconnect() should drop connected flag upon real disconnecting only
- lmv_disconnect() should drop connected flag upon real disconnecting only

17 years ago- minor fixes to make the patch compatible with quadrics patches
alex [Sun, 18 Jul 2004 12:32:35 +0000 (12:32 +0000)]
- minor fixes to make the patch compatible with quadrics patches

17 years ago- more fixes
alex [Sun, 18 Jul 2004 12:29:24 +0000 (12:29 +0000)]
- more fixes

17 years ago- minor fixes to the patch. previous version couldn't be popped cleanly
alex [Sun, 18 Jul 2004 12:21:22 +0000 (12:21 +0000)]
- minor fixes to the patch. previous version couldn't be popped cleanly

17 years ago- teached to catch double-free messy
alex [Sun, 18 Jul 2004 08:36:01 +0000 (08:36 +0000)]
- teached to catch double-free messy

17 years agob=3772
alex [Sun, 18 Jul 2004 08:24:45 +0000 (08:24 +0000)]
b=3772

- race in dynlock_unlock() fixed: several threads may find refcount=0 and
  try to free lock member at the same time. thanks to Andreas!

17 years ago- if MDS aborts recovery it disconnects all the exports and lmv_disconnect()
alex [Sun, 18 Jul 2004 05:41:57 +0000 (05:41 +0000)]
- if MDS aborts recovery it disconnects all the exports and lmv_disconnect()
  is called. the last one didn't drop lmv->connected to 0 and this prevent
  subsequent lmv_connect() to work properly

17 years agob=3869,1742
rread [Sun, 18 Jul 2004 00:29:37 +0000 (00:29 +0000)]
b=3869,1742

These are the remaining fixes from 1742 that are needed to allow a
client to reconnect during recover.  The request that triggered 3869
on the last run was actually a RESENT request that was sent before
recovery had finished, which this patch will fix.

17 years ago1 add snap read in fsfilt_getpage
wangdi [Sat, 17 Jul 2004 15:26:39 +0000 (15:26 +0000)]
1 add snap read in fsfilt_getpage
2 some fix for regular file cow

17 years agob=3869
alex [Sat, 17 Jul 2004 12:39:41 +0000 (12:39 +0000)]
b=3869

- more debug info

17 years agob=1451
rread [Sat, 17 Jul 2004 00:57:13 +0000 (00:57 +0000)]
b=1451

Don't allow client to reconnect if an RPC is already in progress for
that client. Added an extra counter for this, because the export
refcount is also held by locks and who knows what else.

17 years agofile uml-2.6.7-01-bb2.patch was initially added on branch b1_4_smallfix.
green [Fri, 16 Jul 2004 18:57:47 +0000 (18:57 +0000)]
file uml-2.6.7-01-bb2.patch was initially added on branch b1_4_smallfix.

17 years agosome fix for obdfilter cow and snap ext3 patch
wangdi [Fri, 16 Jul 2004 17:41:10 +0000 (17:41 +0000)]
some fix for obdfilter cow and snap ext3 patch

17 years ago- vmlist needs to be exported for modules support
alex [Fri, 16 Jul 2004 17:08:04 +0000 (17:08 +0000)]
- vmlist needs to be exported for modules support

17 years ago- slab-use-after-free debug tool added to vanilla-2.4.24 series to debug 3772
alex [Fri, 16 Jul 2004 12:40:54 +0000 (12:40 +0000)]
- slab-use-after-free debug tool added to vanilla-2.4.24 series to debug 3772
- dynlocks use own slab: slab-use-after-free may be enabled for only dynlocks

17 years agob=3772
alex [Fri, 16 Jul 2004 08:07:25 +0000 (08:07 +0000)]
b=3772

- more debug checks and printks

17 years agob=3843
rread [Fri, 16 Jul 2004 00:27:46 +0000 (00:27 +0000)]
b=3843
r=bzzz

Initialize req before passing it to down.

17 years ago1)add more snap cow hook for dir ops.
wangdi [Thu, 15 Jul 2004 17:59:18 +0000 (17:59 +0000)]
1)add more snap cow hook for dir ops.
2)add snap cow for obdfilter
3)some fix in smfs_cow

17 years agob=3772
alex [Thu, 15 Jul 2004 14:53:37 +0000 (14:53 +0000)]
b=3772

- debug info for corrupted lock case

17 years agob=3844
rread [Wed, 14 Jul 2004 20:37:55 +0000 (20:37 +0000)]
b=3844
Add debug LASSERT to ldlm_replay_locks()

17 years agob=3824
alex [Wed, 14 Jul 2004 15:04:46 +0000 (15:04 +0000)]
b=3824

- new inode flag EXT3_EXTRA_ISIZE to mark inodes having i_extra_isize
  initialized properly

17 years agob=3829
alex [Wed, 14 Jul 2004 14:04:27 +0000 (14:04 +0000)]
b=3829

- mds_copy_unlink_reply() shouldn't be called if md_unlink() returns an error

17 years agofile Makefile.mk was initially added on branch b1_4_smallfix_iib.
zab [Wed, 14 Jul 2004 00:12:09 +0000 (00:12 +0000)]
file Makefile.mk was initially added on branch b1_4_smallfix_iib.

17 years agofile autoMakefile.am was initially added on branch b1_4_smallfix_iib.
zab [Wed, 14 Jul 2004 00:12:08 +0000 (00:12 +0000)]
file autoMakefile.am was initially added on branch b1_4_smallfix_iib.

17 years agofile Makefile.in was initially added on branch b1_4_smallfix_iib.
zab [Wed, 14 Jul 2004 00:12:07 +0000 (00:12 +0000)]
file Makefile.in was initially added on branch b1_4_smallfix_iib.

17 years agofile .cvsignore was initially added on branch b1_4_smallfix_iib.
zab [Wed, 14 Jul 2004 00:12:06 +0000 (00:12 +0000)]
file .cvsignore was initially added on branch b1_4_smallfix_iib.

17 years agob=3405
rread [Tue, 13 Jul 2004 01:57:19 +0000 (01:57 +0000)]
b=3405

Use lmv by default for insanity.sh.

17 years agob=3405
rread [Tue, 13 Jul 2004 00:57:58 +0000 (00:57 +0000)]
b=3405

update replay-dual.sh and replay-ost-single.sh

17 years agob=3405
rread [Mon, 12 Jul 2004 22:19:10 +0000 (22:19 +0000)]
b=3405

- Fix insanity.sh, still using the insanity-local.sh config by default.
- More cleanups in replay-single.sh. We need to generalize failures a
bit more, so we can choose a different mds each time, if more than one
is available.