Whamcloud - gitweb
fs/lustre-release.git
21 years ago- temporarily disable building some dirs when cray portals are in use
zab [Wed, 3 Mar 2004 00:58:30 +0000 (00:58 +0000)]
- temporarily disable building some dirs when cray portals are in use
- revert the overly aggressive migration of portals/list.h
- make sure all builds see -DCRAY_PORTALS.  this is gross.
- protect the libcfs ioctl registration with __KERNEL__
- avoid portals_nid2str() as it breaks the compiler

21 years ago- more progress in moving towards building against cray's external
zab [Tue, 2 Mar 2004 23:12:39 +0000 (23:12 +0000)]
- more progress in moving towards building against cray's external
  portals includes.  This is a work in progress as I ran into
  internal compiler errors.  All builds are probably broken.

21 years ago* Merged HEAD
eeb [Sat, 21 Feb 2004 12:03:35 +0000 (12:03 +0000)]
*  Merged HEAD

21 years ago* Cray compatibility fixes.
eeb [Fri, 30 Jan 2004 23:43:14 +0000 (23:43 +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

21 years agomerge devel zcfg
rread [Tue, 28 Oct 2003 21:46:02 +0000 (21:46 +0000)]
merge devel zcfg

This merge broke config llog, so this branch doesn't mount right now:

llog_write_rec()) ASSERTION((buflen %% LLOG_MIN_REC_SIZE) == 0)

21 years agoThis commit was manufactured by cvs2svn to create branch
cvs2svn [Mon, 29 Sep 2003 20:49:22 +0000 (20:49 +0000)]
This commit was manufactured by cvs2svn to create branch
'unlabeled-1.1.4.3.6'.

21 years ago- bug fixed: variable was used being uninitialized
alex [Mon, 29 Sep 2003 20:49:21 +0000 (20:49 +0000)]
- bug fixed: variable was used being uninitialized

21 years ago- minor changes in order to get 2.6-compatibility
alex [Mon, 29 Sep 2003 16:09:24 +0000 (16:09 +0000)]
- minor changes in order to get 2.6-compatibility

NOTE: kdev_t and dev_t issue need to be investigated in depth

21 years ago- ext3-wantedi patch against 2.6.0-test5
alex [Sun, 28 Sep 2003 10:52:20 +0000 (10:52 +0000)]
- ext3-wantedi patch against 2.6.0-test5

21 years ago- ext3-raw-lookup.patch added into the pdirops series
alex [Thu, 25 Sep 2003 22:51:36 +0000 (22:51 +0000)]
- ext3-raw-lookup.patch added into the pdirops series

21 years agore-committing on b_devel:
phil [Thu, 25 Sep 2003 04:54:47 +0000 (04:54 +0000)]
re-committing on b_devel:
b=1997
r=alex
My previous fix for truncate/write lock inversion was close, but badly
flawed. I failed to remember that truncate will internally restart
the transaction -- start a new one, for all intents and purposes -- so
the ordering was backwards. "i_sem before transaction" is the
cardinal rule.

I tried to avoid that because I didn't want to hold the i_sem across
the entire disk I/O in filter_commitrw_write. After some discussion
with Alex, however, we decided that the i_sem need only be held for
the block allocation, and that the BKL suffices for updating i_size.
So the ordering in the write path is now: take i_sem, start transaction,
call filter_direct_io, f_d_io does block alloc, drops i_sem, does I/O.

21 years agob=1955
zab [Tue, 23 Sep 2003 17:39:26 +0000 (17:39 +0000)]
b=1955
export rpc stats from the osc so that sanity's 42 tests can test for write
rpcs leaving the osc rather than arriving at the ost.  sanity 42s should
pass on nodes without osts.

21 years ago- assert condition have been changed a bit
alex [Tue, 23 Sep 2003 15:31:23 +0000 (15:31 +0000)]
- assert condition have been changed a bit

21 years ago- chaos-2.4.18-pdirops series uses ext3-htree.patch instead of ext-2.4-patch*
alex [Tue, 23 Sep 2003 10:36:59 +0000 (10:36 +0000)]
- chaos-2.4.18-pdirops series uses ext3-htree.patch instead of ext-2.4-patch*

21 years ago- few patches from chaos-2.4.18-pdirops series are updated due to ext3-htree.patch
alex [Tue, 23 Sep 2003 10:33:32 +0000 (10:33 +0000)]
- few patches from chaos-2.4.18-pdirops series are updated due to ext3-htree.patch

21 years ago- needless assert have been removed
alex [Tue, 23 Sep 2003 10:18:11 +0000 (10:18 +0000)]
- needless assert have been removed

21 years agob=1825
girishc [Mon, 22 Sep 2003 17:34:42 +0000 (17:34 +0000)]
b=1825
copying code refactored (ll_prep_inode) in b_devel to b_llp_hp and for nfs
(-Niu's changes)

21 years agoLand latest b_llp_hp changes onto b_flock.
dmilos [Mon, 22 Sep 2003 15:01:18 +0000 (15:01 +0000)]
Land latest b_llp_hp changes onto b_flock.

21 years agob=1825
girishc [Mon, 22 Sep 2003 12:15:14 +0000 (12:15 +0000)]
b=1825
moving lustre related changes for to single file, llite_nfs.c

21 years ago- few bugs fixed
alex [Mon, 22 Sep 2003 11:45:59 +0000 (11:45 +0000)]
- few bugs fixed
- additional asserts added

NOTE: those fixes need to be pushed in the another series

21 years agoupdating from parent branch(b_llp_hp) to b_nfsllphp
girishc [Mon, 22 Sep 2003 08:36:46 +0000 (08:36 +0000)]
updating from parent branch(b_llp_hp) to b_nfsllphp

21 years ago- land b_log on b_devel
braam [Sat, 20 Sep 2003 17:45:50 +0000 (17:45 +0000)]
- land b_log on b_devel

21 years ago* use records that have a tail for the simple record tests.
rread [Fri, 19 Sep 2003 21:28:45 +0000 (21:28 +0000)]
* use records that have a tail for the simple record tests.
* change verify_handle to scan the bitmap to deterime the correct
last_idx and active rec count.
* test 4 now uses verify_handle, and fails the last_idx check after
writing one rec.

* run-llog.sh now tells lctl to ignore_errors, so it can cleanup
after a failed test.

21 years ago* add another loop to test 3 to write some records with a 16 byte
rread [Fri, 19 Sep 2003 19:18:40 +0000 (19:18 +0000)]
* add another loop to test 3 to write some records with a 16 byte
buf. Test now fails after 3c:

3c: handle->last_idx is 1012, expected 1011 after write:(llog_test.c:45:verify_handle() 773 | 9178+2020)

* moved all the code that verifies the handle against the
expected record count to verfiy_handle().

21 years ago- the log processing function appears to work (test added to llog_test).
braam [Fri, 19 Sep 2003 01:37:23 +0000 (01:37 +0000)]
- the log processing function appears to work (test added to llog_test).

21 years ago- a variety of log fixes to the catalog system
braam [Thu, 18 Sep 2003 02:11:12 +0000 (02:11 +0000)]
- a variety of log fixes to the catalog system
  - can now create catalog entries and secondary logs
  - some obdo is not being cleaned up.

21 years ago- merge b_devel into b_llog
braam [Wed, 17 Sep 2003 19:14:16 +0000 (19:14 +0000)]
- merge b_devel into b_llog

21 years agob=1825
girishc [Wed, 17 Sep 2003 18:34:36 +0000 (18:34 +0000)]
b=1825
changes to make NFS export of lustre compatible with nfsv2
and fixes in access mode flag

21 years agoDiscard lock-associated data (dirty pages, at present) when cancelling locks
shaver [Wed, 17 Sep 2003 17:19:05 +0000 (17:19 +0000)]
Discard lock-associated data (dirty pages, at present) when cancelling locks
due to an eviction or OST fail-out.

b=1902.

21 years agoExpose an API for callers which are themselves called with the imp_lock held to
shaver [Wed, 17 Sep 2003 16:59:55 +0000 (16:59 +0000)]
Expose an API for callers which are themselves called with the imp_lock held to
safely drop a possibly-final request reference.

b=1890.

21 years agob=1825
girishc [Wed, 17 Sep 2003 15:49:59 +0000 (15:49 +0000)]
b=1825
sending appropriate  error return value

21 years agob=1825
girishc [Wed, 17 Sep 2003 10:03:27 +0000 (10:03 +0000)]
b=1825
mds_create_objects() checks for FMODE_WRITE mode in rec->ur_flags,
embeding the appropriate mode in flags

21 years agob=1825
girishc [Wed, 17 Sep 2003 09:51:19 +0000 (09:51 +0000)]
b=1825
temporary flag setting in mds for mds_create_objects removed(Done in a better way from NFSD)
Some error code cleanup dione as per review comments

21 years ago- minor cleanups in order to get chaos-2.4.18-pdirops series working
alex [Tue, 16 Sep 2003 21:24:37 +0000 (21:24 +0000)]
- minor cleanups in order to get chaos-2.4.18-pdirops series working

21 years ago- ext3_prep_san_write() and ext3_map_inode_page() call
alex [Mon, 15 Sep 2003 15:42:25 +0000 (15:42 +0000)]
- ext3_prep_san_write() and ext3_map_inode_page() call
  ext3_get_block_wrap() instead of ext3_get_block_handle() for
  extents to be used

21 years ago- raw lookup patch added into chaos-2.4.18 series
alex [Mon, 15 Sep 2003 06:59:32 +0000 (06:59 +0000)]
- raw lookup patch added into chaos-2.4.18 series

21 years ago- lstripe code for b_devel
braam [Mon, 15 Sep 2003 05:04:06 +0000 (05:04 +0000)]
- lstripe code for b_devel
  - error handling is not complete yet (sanity 27f fails)
  - lfind interaction (27g 34's) untested

21 years ago- extents initialization routine is called upon mount now
alex [Sun, 14 Sep 2003 15:05:21 +0000 (15:05 +0000)]
- extents initialization routine is called upon mount now

NOTE: tested on 2.4.18-67chaos (dual P3) with dbench/fsx/etc for
      several hours

21 years agoremove kernel.spec.in
wangdi [Sun, 14 Sep 2003 11:37:48 +0000 (11:37 +0000)]
remove kernel.spec.in

21 years ago- make passing lmm's into open enqueue's possible
braam [Sun, 14 Sep 2003 08:20:48 +0000 (08:20 +0000)]
- make passing lmm's into open enqueue's possible
  - also removed erroneous passing of target.
  - lstripe should be easy from here.
- 1525 appears fixed (probably because of yesterdays fix)
- a new bug (oops with rmdir . ; ls .) was fixed
- recheck 1381 and closed it.

21 years ago- ext3_writepage_trans_blocks() accounts that page may need several blocks.
alex [Sat, 13 Sep 2003 18:31:41 +0000 (18:31 +0000)]
- ext3_writepage_trans_blocks() accounts that page may need several blocks.
  thus we need not account it twice

21 years ago- fix agaist bug #1322 'clients looping in revalidate after recovery failure'
alex [Sat, 13 Sep 2003 12:17:39 +0000 (12:17 +0000)]
- fix agaist bug #1322 'clients looping in revalidate after recovery failure'

21 years agob=1825
girishc [Sat, 13 Sep 2003 11:28:36 +0000 (11:28 +0000)]
b=1825
fix minor bug during rename

21 years ago- i_filterdata added into struct inode
alex [Fri, 12 Sep 2003 20:51:07 +0000 (20:51 +0000)]
- i_filterdata added into struct inode

21 years agob=1807
phil [Fri, 12 Sep 2003 19:13:46 +0000 (19:13 +0000)]
b=1807
- backport oscc_create rewrite from b_llp_hp
- send osc_creates to a new OST thread, so that time-sensitive
  creations will not have to compete with billions of read/write requests

21 years agoll_setattr not supported in b_llp_hp
girishc [Fri, 12 Sep 2003 13:25:11 +0000 (13:25 +0000)]
ll_setattr not supported in b_llp_hp

21 years agob=1792: Fix iopen_connect_dentry to properly rehash the __iopen__ dentry when
shaver [Thu, 11 Sep 2003 22:04:20 +0000 (22:04 +0000)]
b=1792: Fix iopen_connect_dentry to properly rehash the __iopen__ dentry when
        coverting it into a real, "connected" dentry.

21 years agoBringing NFS export code uptodate with b_llp_hp
girishc [Thu, 11 Sep 2003 17:05:34 +0000 (17:05 +0000)]
Bringing NFS export code uptodate with b_llp_hp

21 years agomerge b_devel -> b_eq: 20030909
ericm [Tue, 9 Sep 2003 14:37:10 +0000 (14:37 +0000)]
merge b_devel -> b_eq: 20030909
only kernel pass sanity, liblustre still broken

21 years agoFix for missing open flags for vfs_intent-2.4.20-rh kernel patch.
adilger [Tue, 9 Sep 2003 11:19:11 +0000 (11:19 +0000)]
Fix for missing open flags for vfs_intent-2.4.20-rh kernel patch.
This change was also needed for the vfs_intent-2.4.20-hp patch.  b=1877
Also add bug 1344 revalidate loop fixes to some of the vfs_intent patches.

The diff is fairly large, so that the files in the 2.4.20-rh patch are
in the same order as in all of the other vfs_intent patches to make
comparison and maintenance easier.  Some other minor changes are made to
keep the patches consistent.

There is still a difference between vfs_intent-2.4.20-vanilla and -rh
in follow_down, but I don't know which one is the "keeper".

21 years agoland b_llog on b_devel:
phil [Tue, 9 Sep 2003 04:49:12 +0000 (04:49 +0000)]
land b_llog on b_devel:
reorganization and API update of the logging infrastructure. minimal
changes to the other code.

21 years agomerge b_llpmd into b_devel. the major highlights:
phil [Tue, 9 Sep 2003 03:55:05 +0000 (03:55 +0000)]
merge b_llpmd into b_devel.  the major highlights:
- new I/O backend
- new client page cache and llite/lov/osc plumbing
- pre-creation of OST objects
- most of the OBD protocol now revolves around exports, not obd_devices

21 years ago- export llog_cat_put
phil [Tue, 9 Sep 2003 01:51:07 +0000 (01:51 +0000)]
- export llog_cat_put
- disable the get/put_catalog bits, until they're a little more cooked

21 years ago- clear 'rc' if rc == LLOG_EMPTY in llog_init_handle, so we don't free
phil [Tue, 9 Sep 2003 00:39:05 +0000 (00:39 +0000)]
- clear 'rc' if rc == LLOG_EMPTY in llog_init_handle, so we don't free
  the header and return an error that our caller doesn't expect
- don't call llog_lvfs_pad if left == reclen, only if 0 < left < reclen
- fix the uuid magic in llog_test
- call llog_init_handle in the right places, and don't forget to close
  the log if one of the 1a checks fail

21 years ago- make catalog api's available again, now based on the other layers.
braam [Mon, 8 Sep 2003 23:37:44 +0000 (23:37 +0000)]
- make catalog api's available again, now based on the other layers.

21 years ago- If the first user command isn't ENQUEUE, postsetup would be run too
phil [Mon, 8 Sep 2003 23:02:24 +0000 (23:02 +0000)]
- If the first user command isn't ENQUEUE, postsetup would be run too
  late, which causes the 'sanity' 1a hang that coop sees.  Fixed by
  running postsetup if any of the non-setup RPCs arrives.
- Remove overwriting of lgh_last_idx in llog_init_handle
- init the handle from the test code (not quite right yet)
- #if 0 out the filter catalog functions until they build

21 years ago- new abstraction in the llog_* function wrappers
phil [Mon, 8 Sep 2003 22:18:37 +0000 (22:18 +0000)]
- new abstraction in the llog_* function wrappers
- fix double-allocation of handle headers
- added llog_read_header()
- more llog tests

21 years ago- initialize handles
braam [Mon, 8 Sep 2003 22:07:47 +0000 (22:07 +0000)]
- initialize handles

21 years ago- unify llog_open and llog_create
braam [Mon, 8 Sep 2003 18:31:11 +0000 (18:31 +0000)]
- unify llog_open and llog_create

21 years ago- set the mds's and ost's obd_logops
phil [Mon, 8 Sep 2003 17:08:32 +0000 (17:08 +0000)]
- set the mds's and ost's obd_logops
- initial oa to null in llog_lvfs_create, avoids crashing
- in llog test 1a, don't try to close if the open failed
- 'lctl modules' hook for llog_test

21 years ago- steal from the reserved bits instead of the bitmap bits, another
phil [Mon, 8 Sep 2003 07:51:02 +0000 (07:51 +0000)]
- steal from the reserved bits instead of the bitmap bits, another
  piece of sage Andreas advice
- add an llh_flags field to the header
- rename confusing lgh_index to lgh_last_idx
- if the log is marked ZAP_WHEN_EMPTY, the log is full, and all of the
  records are cancelled, call llog_destroy from llog_cancel_rec
- extend llog_open to also support reopen-by-name
- add llog_vfs_read_blob and llog_vfs_read_header
- removed a bunch of unnecessary EXPORT_SYMBOLs
- add llog_lvfs_open()

21 years ago- pad llog_* body structures correctly (thanks Andreas)
phil [Mon, 8 Sep 2003 06:25:28 +0000 (06:25 +0000)]
- pad llog_* body structures correctly (thanks Andreas)
- countless build fixes

21 years ago- few compile fixes
braam [Mon, 8 Sep 2003 05:35:19 +0000 (05:35 +0000)]
- few compile fixes

21 years ago- next steps towards a working llogging api
braam [Mon, 8 Sep 2003 05:14:19 +0000 (05:14 +0000)]
- next steps towards a working llogging api

21 years ago- remove obd_run_ctxt from obd->u.mds and obd->u.filter; put it in the
phil [Mon, 8 Sep 2003 05:10:41 +0000 (05:10 +0000)]
- remove obd_run_ctxt from obd->u.mds and obd->u.filter; put it in the
  obd_device
- add lvfs_callback_ops, and the first one, the l_fid2dentry
- add callback ops to obd_run_ctxt
- change the 100 uses of mds->mds_ctxt and filter->fo_ctxt
- add a handful of llog tests

21 years ago- fixes to the api's
braam [Mon, 8 Sep 2003 04:07:01 +0000 (04:07 +0000)]
- fixes to the api's

21 years agoFix a few build errors
phil [Mon, 8 Sep 2003 03:12:41 +0000 (03:12 +0000)]
Fix a few build errors

21 years ago- the same llog reorg again, now on the right branch...
braam [Mon, 8 Sep 2003 02:39:13 +0000 (02:39 +0000)]
- the same llog reorg again, now on the right branch...

21 years agoFixups to checkstack target (get diffs in right order, show top 30 abusers).
adilger [Sun, 7 Sep 2003 08:46:05 +0000 (08:46 +0000)]
Fixups to checkstack target (get diffs in right order, show top 30 abusers).
Don't show more functions with less than 200 bytes stack on ia64.

21 years agoUpdate checkstack program, to output more useful "stack_usage function name".
adilger [Sun, 7 Sep 2003 05:24:07 +0000 (05:24 +0000)]
Update checkstack program, to output more useful "stack_usage function name".

Add a "make checkstack" target which shows the current "worst offenders" list
of stack users, or the difference in this list since the last time that target
was run.

There is also a "make checkstack-clean" tarket which resets the "saved" stack
values so that the stack usage diff doesn't show old items.

21 years agomerge b_devel into b_llpmd
phil [Fri, 5 Sep 2003 18:23:40 +0000 (18:23 +0000)]
merge b_devel into b_llpmd

21 years agoAdd jbd-* fixes to some of the series files where I know they apply:
adilger [Fri, 5 Sep 2003 00:16:36 +0000 (00:16 +0000)]
Add jbd-* fixes to some of the series files where I know they apply:
- vanilla-2.4.20 and hp-pnnl-2.4.20 need all three
- chaos-2.4.18 doesn't need the flushtime patch (it is in their tree already)

I'm not sure whether chaos-2.4.20 and rh-2.4.20 need them or not (patches
were originally from RH kernels, so they may already be in rh-2.4.20).

21 years agorevert the last fix for cygwin, which make ogdb-uml always null.
ericm [Wed, 3 Sep 2003 16:21:34 +0000 (16:21 +0000)]
revert the last fix for cygwin, which make ogdb-uml always null.

21 years agoadd kgdb over udp patch
wangdi [Tue, 2 Sep 2003 13:57:04 +0000 (13:57 +0000)]
add kgdb over udp patch

21 years ago[portals]: in handler of SIGALRM, we must reenable SIGALRM before longjmp,
ericm [Thu, 28 Aug 2003 15:25:38 +0000 (15:25 +0000)]
[portals]: in handler of SIGALRM, we must reenable SIGALRM before longjmp,
because it's automatically disabled when entering handler.

21 years ago[tcpnal]: use passed-in port number to start listening socket.
ericm [Thu, 28 Aug 2003 09:09:13 +0000 (09:09 +0000)]
[tcpnal]: use passed-in port number to start listening socket.

21 years ago[portals]:
ericm [Tue, 26 Aug 2003 03:28:43 +0000 (03:28 +0000)]
[portals]:
  PtlEQWait_timeout(): if left_over is 0, don't call alarm with it,
  otherwise we'll not got timeout notification.

21 years agomerge b_devel into b_llpio, including v24 kernel patch
phil [Sat, 23 Aug 2003 22:21:18 +0000 (22:21 +0000)]
merge b_devel into b_llpio, including v24 kernel patch

21 years agoThis commit was manufactured by cvs2svn to create branch 'unlabeled-1.1.4'.
cvs2svn [Fri, 22 Aug 2003 21:40:01 +0000 (21:40 +0000)]
This commit was manufactured by cvs2svn to create branch 'unlabeled-1.1.4'.

21 years agofix socknal build on vanilla kernels by adding socket exports
phil [Fri, 22 Aug 2003 21:39:57 +0000 (21:39 +0000)]
fix socknal build on vanilla kernels by adding socket exports
bumped the kernel patch version to 23, but really only vanilla-2.4.20 changed

21 years agomerge b_devel -> b_eq: 20030822
ericm [Fri, 22 Aug 2003 15:02:33 +0000 (15:02 +0000)]
merge b_devel -> b_eq: 20030822

21 years agofile Makefile.am was initially added on branch b_devel.
wangdi [Fri, 22 Aug 2003 09:05:16 +0000 (09:05 +0000)]
file Makefile.am was initially added on branch b_devel.

21 years agob=1803
rread [Fri, 22 Aug 2003 02:19:00 +0000 (02:19 +0000)]
b=1803
r=shaver

New import state machine, as documented on the lustre wiki in
ImportStates.

A new function, ptlrpc_connect_import, performs all import connects
and moves the import from the DISCON state to either FULL, EVICTED,
REPLAY, or RECOVER, depending on the situation. Unlike the levels, the
states are now exact, and the request->rq_send_state much match the
import state to be sent.

Passes recovery/01, replay-small, and replay-dual.

21 years agofile import.c was initially added on branch b_devel.
rread [Fri, 22 Aug 2003 02:18:55 +0000 (02:18 +0000)]
file import.c was initially added on branch b_devel.

21 years agoLiuKai's another Cygwin fix for data alignment, add #pragma
ericm [Thu, 21 Aug 2003 10:37:15 +0000 (10:37 +0000)]
LiuKai's another Cygwin fix for data alignment, add #pragma

21 years ago[tcpnal]: service socket listening on port 988
ericm [Thu, 21 Aug 2003 04:54:47 +0000 (04:54 +0000)]
[tcpnal]: service socket listening on port 988

21 years ago[tcpnal]: merge LiuKai's cygwin patch
ericm [Thu, 21 Aug 2003 04:16:54 +0000 (04:16 +0000)]
[tcpnal]: merge LiuKai's cygwin patch

21 years ago[tcpnal]: enlarge iov array in tcpnal_send().
ericm [Thu, 21 Aug 2003 02:53:50 +0000 (02:53 +0000)]
[tcpnal]: enlarge iov array in tcpnal_send().

21 years ago[tcpnal]:
ericm [Wed, 20 Aug 2003 11:23:33 +0000 (11:23 +0000)]
[tcpnal]:
 - don't consider pid when doing connection search, it's conform to the
   ksocknal behavior.

21 years ago[tcpnal]:
ericm [Wed, 20 Aug 2003 11:09:01 +0000 (11:09 +0000)]
[tcpnal]:
  - protect tcpnal internal data structure

21 years ago[tcpnal]:
ericm [Wed, 20 Aug 2003 08:46:12 +0000 (08:46 +0000)]
[tcpnal]:
  - add nal_cb locking, fix define of state_lock in user space.

21 years ago[tcpnal}:
ericm [Wed, 20 Aug 2003 08:22:54 +0000 (08:22 +0000)]
[tcpnal}:
  temporarily work around for nal thread endless waiting problem.

21 years agofile nfs_export_kernel-2.4.20.pc was initially added on branch b_nfsdevel.
girishc [Wed, 20 Aug 2003 07:43:23 +0000 (07:43 +0000)]
file nfs_export_kernel-2.4.20.pc was initially added on branch b_nfsdevel.

21 years ago[tcpnal]: rewrite the data out path.
ericm [Wed, 20 Aug 2003 07:22:07 +0000 (07:22 +0000)]
[tcpnal]: rewrite the data out path.
  Get rid of pipe as a bridge between user threads and nal threads, which
  is not so efficient and don't work well in Cygwin. Now user thread might
  directly call into portals lib via lib_dispatch.

  2 issues remain:
  - sometimes during startup low level nal threads could run into endless
    waiting.
  - need more sycronous facilities to prevent race.

21 years agofile LLNL_Changelog was initially added on branch b_llnl_stable.
behlendo [Tue, 19 Aug 2003 19:13:24 +0000 (19:13 +0000)]
file LLNL_Changelog was initially added on branch b_llnl_stable.

21 years agomerge b_multinet into HEAD
phil [Tue, 19 Aug 2003 17:38:30 +0000 (17:38 +0000)]
merge b_multinet into HEAD

21 years agob=1505
phil [Tue, 19 Aug 2003 17:10:03 +0000 (17:10 +0000)]
b=1505
Disable some console messages from failed statfs(), reconnect

21 years ago[liblustre]: merge LiuKai's cygwin patch
ericm [Tue, 19 Aug 2003 12:08:53 +0000 (12:08 +0000)]
[liblustre]: merge LiuKai's cygwin patch
 - add portals/include/cygwin-ioctl.h
 - don't compile accepter.c in liblustre
 - add missing le-cpu convert macro
 - use windows API directly at some places

21 years agofile cygwin-ioctl.h was initially added on branch b_eq.
ericm [Tue, 19 Aug 2003 12:08:49 +0000 (12:08 +0000)]
file cygwin-ioctl.h was initially added on branch b_eq.

21 years agofile uml_2.6.0_test3 was initially added on branch b_llpmd.
alex [Mon, 18 Aug 2003 09:40:33 +0000 (09:40 +0000)]
file uml_2.6.0_test3 was initially added on branch b_llpmd.