Whamcloud - gitweb
adilger [Tue, 10 Feb 2004 21:36:40 +0000 (21:36 +0000)]
Update b_bug974 from HEAD (20040210_1406) (b_smallfix landing)
cvs2svn [Tue, 10 Feb 2004 20:20:37 +0000 (20:20 +0000)]
This commit was manufactured by cvs2svn to create branch 'unlabeled-1.1.2'.
adilger [Tue, 10 Feb 2004 20:20:36 +0000 (20:20 +0000)]
Add cvsdiffclient and cvs-modified-files.pl to CVS (needs a usage note).
adilger [Tue, 10 Feb 2004 20:19:28 +0000 (20:19 +0000)]
Add land[12].sh and add some extra checks to merge[12].sh.
shaocheng [Tue, 10 Feb 2004 09:44:54 +0000 (09:44 +0000)]
- last commit failed to add and remove some files listed below.
shaocheng [Tue, 10 Feb 2004 09:44:50 +0000 (09:44 +0000)]
file Makefile.in was initially added on branch b_bug1414.
shaocheng [Tue, 10 Feb 2004 09:44:49 +0000 (09:44 +0000)]
file Makefile was initially added on branch b_bug1414.
shaocheng [Tue, 10 Feb 2004 09:44:48 +0000 (09:44 +0000)]
file lustre_user.h was initially added on branch b_bug1414.
shaocheng [Tue, 10 Feb 2004 09:44:47 +0000 (09:44 +0000)]
file liblustreapi.h was initially added on branch b_bug1414.
shaocheng [Tue, 10 Feb 2004 09:44:46 +0000 (09:44 +0000)]
file Makefile.am was initially added on branch b_bug1414.
shaocheng [Tue, 10 Feb 2004 09:34:15 +0000 (09:34 +0000)]
- ll_dirstripe_verify is used to verify if files are created using parent
directories' stripe info
- add set/get dir stripe info tests into sanity.sh
- create new header file include/lustre/liblustreapi.h and move
include/linux/lustre_user.h into include/lustre. pass "make rpms"
shaocheng [Tue, 10 Feb 2004 09:34:13 +0000 (09:34 +0000)]
file ll_dirstripe_verify.c was initially added on branch b_bug1414.
shaocheng [Tue, 10 Feb 2004 09:34:09 +0000 (09:34 +0000)]
file ll_dirstripe_verify was initially added on branch b_bug1414.
rread [Tue, 10 Feb 2004 08:59:13 +0000 (08:59 +0000)]
Fix a merge error in conf-sanity.sh. Using CMDVERBOSE this way was
passing "> /dev/null" as an extra argument to the functions, and this
is not really what we want. Since I don't think it's all that bad to
see the output of lconf, anyway, I just deleted the >/dev/null
entirely.
While looking into this, I also noticed that mount -f was still
attempting to fetch the log from the MDS. That is fixed, too.
wangdi [Tue, 10 Feb 2004 04:32:11 +0000 (04:32 +0000)]
update smfs, some minor fix about symlink, link and rename
adilger [Tue, 10 Feb 2004 04:15:04 +0000 (04:15 +0000)]
Correct description for bug 1934.
adilger [Tue, 10 Feb 2004 02:49:00 +0000 (02:49 +0000)]
Update b_bug974 from HEAD (20040209_1727) (b_recovery and 1.0.4 landing).
buffalo-pull [Tue, 10 Feb 2004 00:35:50 +0000 (00:35 +0000)]
- tagging RC_CURRENT
adilger [Tue, 10 Feb 2004 00:21:17 +0000 (00:21 +0000)]
Fix mkfsoptions support for osts.
b=2603, b=2604
adilger [Tue, 10 Feb 2004 00:15:19 +0000 (00:15 +0000)]
Additional debugging for MDS client eviction problem (from 1.0.4).
b=2443
adilger [Tue, 10 Feb 2004 00:11:08 +0000 (00:11 +0000)]
Print out dotted-quad IP addresses in the socknal (from 1.0.4)
b=2302
adilger [Tue, 10 Feb 2004 00:07:02 +0000 (00:07 +0000)]
Fix timeouts when evicting a client with a single lock held (from 1.0.4).
b=2642
adilger [Tue, 10 Feb 2004 00:02:17 +0000 (00:02 +0000)]
Don't leave stale dentries around after renames (from 1.0.4).
b=2428
adilger [Mon, 9 Feb 2004 23:40:41 +0000 (23:40 +0000)]
Zero end of page at obdfilter for partial page writes (from 1.0.4).
b=2648
r=phil
adilger [Mon, 9 Feb 2004 23:27:33 +0000 (23:27 +0000)]
RPM build fix from 1.0.4
r=jacob
adilger [Mon, 9 Feb 2004 23:10:25 +0000 (23:10 +0000)]
Don't flag the ptlrpcd thread with PF_MEMALLOC (from 1.0.4).
b=2636
adilger [Mon, 9 Feb 2004 23:07:20 +0000 (23:07 +0000)]
Fix truncate-vs-write corruption - truncate wasn't marking the last partial
page dirty in any lists (from 1.0.4).
b=2366
adilger [Mon, 9 Feb 2004 22:29:42 +0000 (22:29 +0000)]
Fix for failed assertion in iopen_connect_dentry (2.4 kernels only, from b1_0).
b=1792, b=2517
adilger [Mon, 9 Feb 2004 22:02:55 +0000 (22:02 +0000)]
Update b_smallfix from HEAD (20040209_1433)
(merge of b_recovery)
rread [Mon, 9 Feb 2004 21:59:58 +0000 (21:59 +0000)]
Fix "implicit declaration" warning
adilger [Mon, 9 Feb 2004 21:49:23 +0000 (21:49 +0000)]
Update ChangeLog with fixes from b_recovery landing.
wangdi [Mon, 9 Feb 2004 16:39:39 +0000 (16:39 +0000)]
update smfs 1)delete inode->i_list in smfs_delete_inode for the cache inode will have no chance to release it in iput 2)fix bugs in file read and write 3) some fixs for unlink and rmdir
yury [Mon, 9 Feb 2004 14:00:59 +0000 (14:00 +0000)]
Fixes and improvements in tmpfs-sanity.sh
yury [Mon, 9 Feb 2004 13:55:50 +0000 (13:55 +0000)]
Fixes in tmp-sanity.sh
yury [Mon, 9 Feb 2004 13:26:30 +0000 (13:26 +0000)]
added tmpfs-xattr to vanilla-2.4.20
yury [Mon, 9 Feb 2004 13:25:12 +0000 (13:25 +0000)]
Added CONFIG_TMPFS_XATTR=y in vanilla 2.4.20 config
cvs2svn [Mon, 9 Feb 2004 13:23:52 +0000 (13:23 +0000)]
This commit was manufactured by cvs2svn to create branch 'unlabeled-1.1.2'.
yury [Mon, 9 Feb 2004 13:23:49 +0000 (13:23 +0000)]
Added tmpfs-xattr patch for vanilla 2.4.20
yury [Mon, 9 Feb 2004 13:20:28 +0000 (13:20 +0000)]
Added tmpfs-sanity.sh which checks if xattr support for tmpfs works fine.
ericm [Mon, 9 Feb 2004 08:41:58 +0000 (08:41 +0000)]
merge HEAD to b_eq tag 20040209_1518 (kernel only)
major changes are from landing of b_recovery, which might have great impact
on liblustre.
wangdi [Mon, 9 Feb 2004 04:44:31 +0000 (04:44 +0000)]
minor fix about smfs in Makefile.am
rread [Sun, 8 Feb 2004 22:45:24 +0000 (22:45 +0000)]
- missed some calls to zconf_mount/umount in the test scripts when I
updated insanity
wangdi [Sun, 8 Feb 2004 18:35:26 +0000 (18:35 +0000)]
update smfs.c 1)fix bugs of inode and dentry in dir.c
rread [Sat, 7 Feb 2004 21:59:04 +0000 (21:59 +0000)]
Landing b_recovery
The principal change here is ptlrpc_connect_import and recovery are
merged and are now asynchronous and are run on ptlrpcd.
This branch also includes fixes for
2477 (this was the initial bug)
2355 recover for initial connections
1934 (b_1934 was merged directly to b_recovery)
1901 block all file creations until orphan recovery completes
2423 client remove rq_connection from request struct
2640 conf-sanity test_5, and proper cleanup in umount when
cleanup log is not availale.
2670 recovery timer race
2532 mdc_close recovey bug
And few others that werent' filed.
wangdi [Sat, 7 Feb 2004 18:52:59 +0000 (18:52 +0000)]
oops! just forget commit a fix in the last commit
wangdi [Sat, 7 Feb 2004 18:46:05 +0000 (18:46 +0000)]
update smfs 1)fix read file bugs 2)add some superblock operations 3)some minor fix
eeb [Sat, 7 Feb 2004 11:44:41 +0000 (11:44 +0000)]
* Fixed 2634 ("instant" socknal timeout?)
* Fixed 2684 ("static declaration of ep_nmd_merge...)
buffalo-pull [Sat, 7 Feb 2004 04:18:54 +0000 (04:18 +0000)]
- tagging RC_CURRENT
adilger [Sat, 7 Feb 2004 01:26:30 +0000 (01:26 +0000)]
Quiet warning on ia64 compiles.
adilger [Sat, 7 Feb 2004 01:19:35 +0000 (01:19 +0000)]
Fix type mismatch errors found on ia64.
phil [Sat, 7 Feb 2004 01:18:19 +0000 (01:18 +0000)]
don't f_put(dentry) until we're finished using it, kthx.
zab [Sat, 7 Feb 2004 00:37:01 +0000 (00:37 +0000)]
b=2683
In the echo_client's prep/commit test ioo_bufcnt could be out of
sync with the npages given to obd_{prep,commit}rw.
alex [Fri, 6 Feb 2004 22:54:44 +0000 (22:54 +0000)]
- use-after-free debugging tool against 2.4.24
- additional functionality: free routine checks for possible corruptions
after given size
alex [Fri, 6 Feb 2004 22:51:28 +0000 (22:51 +0000)]
- we have to get undo access for block bitmaps
- check for invalid goald in ext3_new_blocks() and correct if needed
adilger [Fri, 6 Feb 2004 21:14:38 +0000 (21:14 +0000)]
Update ChangeLog
b=2370
adilger [Fri, 6 Feb 2004 21:03:22 +0000 (21:03 +0000)]
Make it clear which error was hit in test 62 (failed in buffalo testing of 974).
wangdi [Fri, 6 Feb 2004 16:38:06 +0000 (16:38 +0000)]
update smfs, 1)add dir and files methods of smfs 2) some fixes of it
shaocheng [Fri, 6 Feb 2004 09:19:05 +0000 (09:19 +0000)]
- has been tested against lov.sh configuration.
- add test script tests/dirstripe.sh
shaocheng [Fri, 6 Feb 2004 09:19:03 +0000 (09:19 +0000)]
file dirstripe.sh was initially added on branch b_bug1414.
alex [Fri, 6 Feb 2004 07:21:20 +0000 (07:21 +0000)]
- few additional asserts added
buffalo-pull [Fri, 6 Feb 2004 03:12:30 +0000 (03:12 +0000)]
- tagging RC_CURRENT
nic [Fri, 6 Feb 2004 02:20:29 +0000 (02:20 +0000)]
b=2623
r=ccooper
Fix problem where none of the tests were checking for a successfull cleanup.
Error in cleanup when modules are still loaded.
zab [Fri, 6 Feb 2004 01:48:56 +0000 (01:48 +0000)]
- add an overwrite pass
- clean up output directory creation
adilger [Thu, 5 Feb 2004 23:46:43 +0000 (23:46 +0000)]
Update b_eq from HEAD (20040205_1502)
Most of the changes are from the landing of b_orphan onto HEAD.
wangdi [Thu, 5 Feb 2004 18:25:41 +0000 (18:25 +0000)]
add some methods in file.c
cvs2svn [Thu, 5 Feb 2004 17:58:12 +0000 (17:58 +0000)]
This commit was manufactured by cvs2svn to create branch 'unlabeled-1.1.2'.
alex [Thu, 5 Feb 2004 17:58:11 +0000 (17:58 +0000)]
- ext3 batching allocation against suse2 series
- ext3-mballoc* patches export ext3_new_blocks()
alex [Thu, 5 Feb 2004 17:56:30 +0000 (17:56 +0000)]
- export few extents-related functions
wangdi [Thu, 5 Feb 2004 17:37:36 +0000 (17:37 +0000)]
update smfs 1)some fix in cache mount 2)some fix in lookup and touch, create
eeb [Thu, 5 Feb 2004 11:51:17 +0000 (11:51 +0000)]
* Removed support for old (pre open source) QSW installations
adilger [Thu, 5 Feb 2004 09:35:57 +0000 (09:35 +0000)]
Update b_bug974 from HEAD (20040205_0202) (to get lmc isspace fix).
b=974
buffalo-pull [Thu, 5 Feb 2004 01:15:24 +0000 (01:15 +0000)]
- tagging RC_CURRENT
zab [Thu, 5 Feb 2004 01:02:07 +0000 (01:02 +0000)]
- collect basic oprofile results when it is available
- make the output directory if it doesn't exist
- add a forgotten item to the key
adilger [Thu, 5 Feb 2004 00:43:53 +0000 (00:43 +0000)]
Re-enable the statfs cache on the OSTs, unless there isn't much space left,
in order to hopefully remove the perf hit from the 974 changes.
We compensate for potentially fast devices consuming a lot of free space by
slightly over-consuming available space from the cached statfs data as each
BRW is handled. Not enough to throw out the statfs results by any meaningful
amount (esp. given that we include the slightly-old results from the client
cache when replying to clients). This overconsumption is thrown out each time
we refresh the statfs data (every second or every RPC if almost full).
DEBUG_REQ() now shows the req->rq_repmsg->status in addition to req->rq_status
since the latter is usually zero and the former has the real reply status.
b=2644
zab [Wed, 4 Feb 2004 18:21:40 +0000 (18:21 +0000)]
- require iozone with the -+o patch
- simplify iozone file cleanup a little
- scale each thread's io contribution based on the number of threads
- fix sloppy shell arg passing that accidentally worked
wangdi [Wed, 4 Feb 2004 17:08:26 +0000 (17:08 +0000)]
update smfs 1.fix bugs of read_super. 2.some minor fix
buffalo-pull [Wed, 4 Feb 2004 00:30:28 +0000 (00:30 +0000)]
- tagging RC_CURRENT
zab [Tue, 3 Feb 2004 19:47:14 +0000 (19:47 +0000)]
- clean up last_block cleanup a little
- teach the iozone test to use multiple threads
- add a -e argument to specify the obdecho module by hand
- the echo_filter test performs reads and on the right number of pages
- use more representative io sizes
- use 'getconf PAGE_SIZE' to find the page size on the machine
cvs2svn [Tue, 3 Feb 2004 18:21:49 +0000 (18:21 +0000)]
This commit was manufactured by cvs2svn to create branch 'unlabeled-1.1.2'.
alex [Tue, 3 Feb 2004 18:21:47 +0000 (18:21 +0000)]
- series for uml on 2.6.0
NOTE: subsytem initialize order problem is to be solved yet
wangdi [Tue, 3 Feb 2004 16:39:38 +0000 (16:39 +0000)]
update smfs. 1)fix bugs in sm_read_super 2) add some methods in symlink ops
rread [Tue, 3 Feb 2004 07:43:47 +0000 (07:43 +0000)]
python 1.5 doesn't have isspace() (I know, how barbaric), but this should
do just as well.
wangchao [Tue, 3 Feb 2004 03:18:20 +0000 (03:18 +0000)]
b=2603
r=shaver
work with --mkfsoptions of ost in lmc and lconf
buffalo-pull [Tue, 3 Feb 2004 02:03:28 +0000 (02:03 +0000)]
- tagging RC_CURRENT
zab [Mon, 2 Feb 2004 22:06:26 +0000 (22:06 +0000)]
- add a helper to call bc with a specified precision
- run an iostat -x on each block device during each test
- reformat a bit to make room for the iostat reports
eeb [Mon, 2 Feb 2004 20:16:23 +0000 (20:16 +0000)]
* Modified socknal to...
. NOT bump the module refcount when a new peer is found
. Clean up all open connections and autoconnect route table entries on
module unload.
zab [Mon, 2 Feb 2004 19:42:40 +0000 (19:42 +0000)]
- try to verify that a bdev is alive by reading from it
- more carefully cleanup to avoid error messages in early cleanup
zab [Mon, 2 Feb 2004 18:31:22 +0000 (18:31 +0000)]
- add an argument to select which tests to run
wangdi [Mon, 2 Feb 2004 16:48:15 +0000 (16:48 +0000)]
update smfs
tianying [Mon, 2 Feb 2004 13:56:38 +0000 (13:56 +0000)]
Update b_bug1934 from HEAD (20040202_2102).
cvs2svn [Sun, 1 Feb 2004 17:42:37 +0000 (17:42 +0000)]
This commit was manufactured by cvs2svn to create branch 'unlabeled-1.1.2'.
wangdi [Sun, 1 Feb 2004 17:42:36 +0000 (17:42 +0000)]
update smfs, add some files dir.c cache.c inode.c still many problems, will fix them soon
alex [Sat, 31 Jan 2004 20:58:36 +0000 (20:58 +0000)]
- ldlm builds as part of ptlrpc
wangdi [Sat, 31 Jan 2004 17:00:59 +0000 (17:00 +0000)]
add mount cache in sm_read_super
wangdi [Sat, 31 Jan 2004 09:00:38 +0000 (09:00 +0000)]
add smfs
buffalo-pull [Sat, 31 Jan 2004 00:52:01 +0000 (00:52 +0000)]
- tagging RC_CURRENT
adilger [Sat, 31 Jan 2004 00:47:46 +0000 (00:47 +0000)]
Update b_orphan from HEAD (20040130_1601)
buffalo-pull [Sat, 31 Jan 2004 00:46:14 +0000 (00:46 +0000)]
- tagging RC_CURRENT
adilger [Sat, 31 Jan 2004 00:33:58 +0000 (00:33 +0000)]
Land b_orphan on HEAD (20040130_1601)
eeb [Fri, 30 Jan 2004 23:42:59 +0000 (23:42 +0000)]
* Cray compatibility fixes.
- Added #if CRAY_PORTALS for compiling against cray portals.
- PTLRPC_MTU, PTLRPC_MAX_IOV derived from PTL_MTU, PTL_MAX_IOV if
defined.
- PTLRPC_MAX_BRW_PAGES, PTLRPC_MAX_IOV defined consistently from
PTLRPC_MTU, PTLRPC_MAX_IOV, and are the constants used by the OSC.
- Changed bulk descriptors to merge contiguous fragments. Bulk I/O only
uses iovs/kiovs if # frags > 1.
If compiling with cray kernel portals, PTL_MD_PHYS is expected to be
defined and bulk descriptors contain struct iovec with physical
addresses. NB. Cray requested this method, rather than the ptl_kiov_t
our portals uses. If they can't use physical addresses (we can always
turn PTLRPC_MAX_BRW_PAGES down to 1 until they can handle multiple
frags), we have a problem, since we'll have to re-write the old
kmap/kunmap code.
- Added PtlSnprintHandle(char *str, int str_len, ptl_handle_any_t handle)
to keep handles opaque but provide useful debug messages.
- Implemented PtlEQPoll(), and PtlEQGet()/PtlEQWait() in terms of it.
Fixed tcpnal non-busy wait hack by changing nal::yield() to take a
timeout and do the right locking, and cb_callback() to schedule
waiting processes.
Implemented non-busy wait for socknal and qswnal as programming
examples (we don't actually drain event queues in the kernel) and
cleaned up tcpnal to only schedule when required.
- Conformed to the portals spec on...
. PTL_TIME_FOREVER
. PTL_EQ_HANDLER_NONE
. PTL_NI_OK
. event field names
. PTL_MD_MAX_SIZE usage
- Defined PTL_EVENT_XXX_{START,END}. We don't actually implement START
events, but rather LASSERT() our callers have disabled them.
- Added liblustre/cray callback handling.
- #define PTLRPC_MD_OPTIONS as lustre's MD options. Includes the "Hello
Cray portals; lustre calling" flag and disables start events.
- Removed PTL_MD_AUTO_UNLINK; it's specified in the 'unlink' parameter
of PtlMDBind/Attach() etc.
* Other fixes/cleanups
- Removed 'struct lustre_peer'
- Added liblustre/tests/.cvsignore
- Removed PtlNIBarrier() and some cruddy portals globals.
- Fixed lib_msg_alloc() to allocate ATOMIC if necessary (NB socknal
never needs ATOMIC here, neither does qswnal if EKC is tuned
properly).
- Fixed lib_finalize() to only unlink if the MD was created with
PTL_UNLINK (lustre always does, so we never noticed this before).
- Cleaned up our asynchronous PtlMDUnlink()....>PTL_EVENT_UNLINK to
invalidate the MD handle immediately on the unlink (manual and
automatic).
- factored some common code in lib-move.c