Whamcloud - gitweb
fs/lustre-release.git
21 years agoFix minor leak of import struct, which we do not use yet.
adilger [Mon, 22 Jul 2002 18:45:22 +0000 (18:45 +0000)]
Fix minor leak of import struct, which we do not use yet.

21 years agoRemove some cruft that got into chaos10 patch
adilger [Mon, 22 Jul 2002 18:37:52 +0000 (18:37 +0000)]
Remove some cruft that got into chaos10 patch

21 years agoInclude lustre_mds.h to have definition for mdc_getlovinfo.
adilger [Mon, 22 Jul 2002 18:36:29 +0000 (18:36 +0000)]
Include lustre_mds.h to have definition for mdc_getlovinfo.

21 years agoCosmetic cleanup to use local variable for ll_inode_info.
adilger [Mon, 22 Jul 2002 18:35:56 +0000 (18:35 +0000)]
Cosmetic cleanup to use local variable for ll_inode_info.

21 years agoAdd beginning of symlink support. Not totally functional yet, but not totally
adilger [Mon, 22 Jul 2002 18:35:19 +0000 (18:35 +0000)]
Add beginning of symlink support.  Not totally functional yet, but not totally
broken either.

21 years agoAdd helper function ptlrpc_bulk_addref() to match ptlrpc_bulk_decref().
adilger [Mon, 22 Jul 2002 18:34:12 +0000 (18:34 +0000)]
Add helper function ptlrpc_bulk_addref() to match ptlrpc_bulk_decref().

21 years agoWe need an mdc_ref for a mountpoint, and not an mds_ref.
adilger [Mon, 22 Jul 2002 18:33:22 +0000 (18:33 +0000)]
We need an mdc_ref for a mountpoint, and not an mds_ref.

21 years agoCall bulk descriptor "desc" instead of "bulk" for consistency.
adilger [Mon, 22 Jul 2002 18:32:52 +0000 (18:32 +0000)]
Call bulk descriptor "desc" instead of "bulk" for consistency.

21 years agoThis should fix the extN bug. I have not enabled indexed directories by
adilger [Mon, 22 Jul 2002 18:27:07 +0000 (18:27 +0000)]
This should fix the extN bug.  I have not enabled indexed directories by
default yet, so it should not affect anything either way.

21 years agoAdd module refcounts for filesystem interface modules.
adilger [Mon, 22 Jul 2002 17:22:12 +0000 (17:22 +0000)]
Add module refcounts for filesystem interface modules.
Minor cleanups in MDS LOV handling (i.e. endianness, code duplication)

21 years agoHOWTO updates.
gord-fig [Sat, 20 Jul 2002 23:48:22 +0000 (23:48 +0000)]
HOWTO updates.

21 years agoMake the usermode patch apply cleanly.
gord-fig [Sat, 20 Jul 2002 23:47:23 +0000 (23:47 +0000)]
Make the usermode patch apply cleanly.

21 years agoAdd intent support for MKNOD.
adilger [Sat, 20 Jul 2002 16:21:15 +0000 (16:21 +0000)]
Add intent support for MKNOD.

Still to be done - SYMLINK and LINK.

21 years ago- new patch for new LLNL kernel. No changes.
braam [Sat, 20 Jul 2002 00:36:30 +0000 (00:36 +0000)]
- new patch for new LLNL kernel. No changes.

21 years agoPatch was missing IT_LOOKUP.
braam [Fri, 19 Jul 2002 21:53:06 +0000 (21:53 +0000)]
Patch was missing IT_LOOKUP.

21 years agoFixups - should work for single ost.
adilger [Fri, 19 Jul 2002 09:02:39 +0000 (09:02 +0000)]
Fixups - should work for single ost.

21 years agoChange mountpoint to be /mnt/lustre for LOV XML config.
adilger [Fri, 19 Jul 2002 09:00:08 +0000 (09:00 +0000)]
Change mountpoint to be /mnt/lustre for LOV XML config.

21 years agoFix linux version checking so that modules install correctly under UML also.
adilger [Fri, 19 Jul 2002 07:09:58 +0000 (07:09 +0000)]
Fix linux version checking so that modules install correctly under UML also.

21 years agoSo, it turns out that the CVS head was wrong - kptrouter is unloaded last.
adilger [Fri, 19 Jul 2002 06:33:57 +0000 (06:33 +0000)]
So, it turns out that the CVS head was wrong - kptrouter is unloaded last.

21 years agoGet the correct return code from the setup, not from echo.
adilger [Fri, 19 Jul 2002 05:40:58 +0000 (05:40 +0000)]
Get the correct return code from the setup, not from echo.

21 years agoRemove intersync stuff from /etc/init.d/lustre.
adilger [Fri, 19 Jul 2002 05:21:24 +0000 (05:21 +0000)]
Remove intersync stuff from /etc/init.d/lustre.

21 years agoChange the default config so that it doesn't do anything out of the box.
adilger [Fri, 19 Jul 2002 05:15:03 +0000 (05:15 +0000)]
Change the default config so that it doesn't do anything out of the box.

21 years agoSame fixes on head as branch.
adilger [Fri, 19 Jul 2002 04:49:14 +0000 (04:49 +0000)]
Same fixes on head as branch.

21 years agoPeter's module fixes on the tip
pschwan [Fri, 19 Jul 2002 04:19:07 +0000 (04:19 +0000)]
Peter's module fixes on the tip

21 years ago- fix / problem
braam [Thu, 18 Jul 2002 23:52:04 +0000 (23:52 +0000)]
- fix / problem

21 years ago- added recovery.lyx
pschwan [Thu, 18 Jul 2002 23:37:14 +0000 (23:37 +0000)]
- added recovery.lyx
- changed the 'date' commands to use %x
- added DOCVERSION which will hopefully be included in make dist

21 years agoadd config files to EXTRA_DIST
pschwan [Thu, 18 Jul 2002 23:14:13 +0000 (23:14 +0000)]
add config files to EXTRA_DIST

21 years agorevert configure.in
braam [Thu, 18 Jul 2002 23:12:10 +0000 (23:12 +0000)]
revert configure.in

21 years agoMakefile now puts the version and date on front of lustre.pdf
braam [Thu, 18 Jul 2002 22:49:09 +0000 (22:49 +0000)]
Makefile now puts the version and date on front of lustre.pdf

21 years agoSplit the documentation and sample config files into a separate lustre-doc RPM
pschwan [Thu, 18 Jul 2002 22:46:19 +0000 (22:46 +0000)]
Split the documentation and sample config files into a separate lustre-doc RPM

21 years agoAdd Phil's patch-2.4.18-um change.
gord-fig [Thu, 18 Jul 2002 22:42:29 +0000 (22:42 +0000)]
Add Phil's patch-2.4.18-um change.
Add --with-linux and --with-portals as aliases for --enable-linuxdir and --enable-portalsdir.

21 years agoAdd version information to lustre-HOWTO.
gord-fig [Thu, 18 Jul 2002 22:41:15 +0000 (22:41 +0000)]
Add version information to lustre-HOWTO.

21 years agochangelog update, head
pschwan [Thu, 18 Jul 2002 21:42:32 +0000 (21:42 +0000)]
changelog update, head

21 years agoa small coherence test with 2 mountpoints
pschwan [Thu, 18 Jul 2002 17:15:52 +0000 (17:15 +0000)]
a small coherence test with 2 mountpoints

21 years agoOne more.
pschwan [Thu, 18 Jul 2002 14:18:56 +0000 (14:18 +0000)]
One more.

21 years agoSmall cleanups to make it look and behave more like the runfailure-net
pschwan [Thu, 18 Jul 2002 14:13:46 +0000 (14:13 +0000)]
Small cleanups to make it look and behave more like the runfailure-net

21 years agoA mostly-fix for "mknod /mnt/lustre/foofo p". It doesn't fail outright
adilger [Thu, 18 Jul 2002 10:07:04 +0000 (10:07 +0000)]
A mostly-fix for "mknod /mnt/lustre/foofo p".  It doesn't fail outright
anymore (S_IFREG only for regular files), it doesn't oops (pass in
dentry->d_it to ll_create), it only complains "file already exists" after
it has already created the file.  Definitely an intent thing, but it is
too late to fix that part.

Also includes the LOV changes from today, as they don't seem to harm
the current setup (runtests finishes).

21 years agoMore connection-level hackery for the DLM
pschwan [Thu, 18 Jul 2002 07:12:16 +0000 (07:12 +0000)]
More connection-level hackery for the DLM

21 years agoBe more explicit to aid debugging
pschwan [Thu, 18 Jul 2002 07:03:45 +0000 (07:03 +0000)]
Be more explicit to aid debugging

21 years agosmall cleanup
pschwan [Thu, 18 Jul 2002 06:26:43 +0000 (06:26 +0000)]
small cleanup

21 years agoBig diff, mostly cosmetic.
adilger [Wed, 17 Jul 2002 20:26:53 +0000 (20:26 +0000)]
Big diff, mostly cosmetic.

The two fixes are:
- osc_con2dlmcl() now fills in rconn properly.
- lov_connect() gets the mdc_conn data passed in via the conn param

The cosmetic:
- add a new field to the export struct exp_rconnh.  Don't use the
  import field, it is for the LDLM to use later.
- make the export struct elements exp_foo instead of export_foo.

21 years ago- Attempted fix for the crash-during-unload-after-ldlm-test bug
pschwan [Wed, 17 Jul 2002 18:47:59 +0000 (18:47 +0000)]
- Attempted fix for the crash-during-unload-after-ldlm-test bug

21 years ago- Temporary fix for the problem that the MDS needs to make lock requests to
pschwan [Wed, 17 Jul 2002 15:55:10 +0000 (15:55 +0000)]
- Temporary fix for the problem that the MDS needs to make lock requests to
itself, but doesn't go through the whole connection setup nonsense, so its
connection never advances above LUSTRE_CONN_NEW level
- Right after we failed to get a lock in the I/O patch, we tried to debug-print
it.  Fixed.

21 years agoTweak autogen.sh and Makefiles to be compatible with Automake 1.6.2.
gord-fig [Wed, 17 Jul 2002 13:26:34 +0000 (13:26 +0000)]
Tweak autogen.sh and Makefiles to be compatible with Automake 1.6.2.

21 years ago- fix default rule
rread [Wed, 17 Jul 2002 08:18:46 +0000 (08:18 +0000)]
- fix default rule

21 years ago- move LLNL kernel patch forward to chaos9
braam [Wed, 17 Jul 2002 02:07:21 +0000 (02:07 +0000)]
- move LLNL kernel patch forward to chaos9

21 years ago- mountpoint uses mdc, not mds
rread [Wed, 17 Jul 2002 01:31:13 +0000 (01:31 +0000)]
- mountpoint uses mdc, not mds
- still not getting connectoin to lov, though

21 years ago- mountpoints are coded, but ost uuid is not found, at least for lov
rread [Wed, 17 Jul 2002 00:57:48 +0000 (00:57 +0000)]
- mountpoints are coded, but ost uuid is not found, at least for lov

21 years ago- lov config works now
rread [Wed, 17 Jul 2002 00:10:09 +0000 (00:10 +0000)]
- lov config works now

21 years ago- clean up usage help
rread [Tue, 16 Jul 2002 22:55:06 +0000 (22:55 +0000)]
- clean up usage help
- use normal traceback

21 years ago- look for profile for the current hostname, or localhost for default
rread [Tue, 16 Jul 2002 22:42:15 +0000 (22:42 +0000)]
- look for profile for the current hostname, or localhost for default

21 years ago- look for lctl in ../utils or in path
rread [Tue, 16 Jul 2002 21:50:12 +0000 (21:50 +0000)]
- look for lctl in ../utils or in path

21 years ago- simple config, mostly to load modules
rread [Tue, 16 Jul 2002 21:06:44 +0000 (21:06 +0000)]
- simple config, mostly to load modules

21 years ago- config and cleanup is working from obd to lov
rread [Tue, 16 Jul 2002 20:57:57 +0000 (20:57 +0000)]
- config and cleanup is working from obd to lov
- modules, rpc, and mountpoints not working yet

21 years ago- initial checkin
rread [Tue, 16 Jul 2002 20:56:51 +0000 (20:56 +0000)]
- initial checkin
- lov test config for lconf

21 years ago- use stdout for regular messages
rread [Tue, 16 Jul 2002 20:52:12 +0000 (20:52 +0000)]
- use stdout for regular messages

21 years agoFix cleanup in llite/super.c
adilger [Tue, 16 Jul 2002 17:20:41 +0000 (17:20 +0000)]
Fix cleanup in llite/super.c

21 years ago- improve error handling for command line use
rread [Tue, 16 Jul 2002 16:32:01 +0000 (16:32 +0000)]
- improve error handling for command line use

21 years agoFix problem with multiple connections to the same local device.
adilger [Tue, 16 Jul 2002 11:52:07 +0000 (11:52 +0000)]
Fix problem with multiple connections to the same local device.

The problem was that we stored the remote export addr+cookie in the
device array.  When we disconnected any connection, it would drop the
remote export from the most recent connection.

What we do now is store the remote export addr+cookie in the local
export/import struct and always use that (the old mdc_connh and osc_connh
are gone).  We could do some further cleanup of the API at this point,
but I will leave it until there is some discussion on the matter.  I was
able to successfully do multiple simultaneous connect and disconnect steps,
as well as run multi-thread test_getattrs with the change.  I'm just doing
a full runregression-net test.

I'm still not 100% confident that this is what Peter wants, so I tagged
the pre-commit tree with "before_rconn" to make it easy to generate
a diff (I will also tag the post-commit with "after_rconn").

21 years ago- remove obdfs/Makefile
rread [Tue, 16 Jul 2002 01:53:37 +0000 (01:53 +0000)]
- remove obdfs/Makefile
- get kernel version from version.h, fix uml kversion bug

21 years ago- configuring an obd seems to work
rread [Tue, 16 Jul 2002 01:52:16 +0000 (01:52 +0000)]
- configuring an obd seems to work

21 years ago- enable command line flags for lctl
rread [Tue, 16 Jul 2002 01:49:31 +0000 (01:49 +0000)]
- enable command line flags for lctl
- device <name> now works

21 years agoFixed brw_finish to kunmap the pointers in the bulk descriptor instead of the
pschwan [Mon, 15 Jul 2002 20:55:57 +0000 (20:55 +0000)]
Fixed brw_finish to kunmap the pointers in the bulk descriptor instead of the
callback data.  The cb_data had a pointer to stack data which might not be
valid anymore, and instead of allocating a new array to hold it, we can just
walk the bulk pages.

21 years agoThe ioctl handler for test_brw was corrupting kernel memory and giving an
pschwan [Mon, 15 Jul 2002 20:32:19 +0000 (20:32 +0000)]
The ioctl handler for test_brw was corrupting kernel memory and giving an
empty array to osc_brw_read, which cheerfully tried to kmap(0).  Fixed.

21 years ago- change id to ref
rread [Mon, 15 Jul 2002 19:23:33 +0000 (19:23 +0000)]
- change id to ref
- change ref attributes to just uuidref
- changing service_ref to more specific references where appropriate.

21 years ago- Fix for the 16kb page directory handling (thanks Andreas, it appears to
pschwan [Mon, 15 Jul 2002 18:04:24 +0000 (18:04 +0000)]
- Fix for the 16kb page directory handling (thanks Andreas, it appears to
  work perfectly)
- Fix for Eric's page_array compilation error
- Increase event queue sizes to 1024
- Re-enable bits of the runfailure-ost test

21 years ago- add Andreas's suggestion for the LLNL Oops's to the patch
braam [Mon, 15 Jul 2002 15:40:45 +0000 (15:40 +0000)]
- add Andreas's suggestion for the LLNL Oops's to the patch

21 years ago- see previous message
braam [Sun, 14 Jul 2002 17:27:12 +0000 (17:27 +0000)]
- see previous message

21 years ago- lookup2: drop a lock in an unlikely error case; clarify interrupt
braam [Sun, 14 Jul 2002 17:26:49 +0000 (17:26 +0000)]
- lookup2: drop a lock in an unlikely error case; clarify interrupt
  handling in lookup2
- osc_request.c: bring connection level to full after connect
- mds_reint.c: in unlink, don't cancel locks that you don't have for
  negative dentries (bug fix).

21 years agoThis commit contains probably 92% of the striping infrastructure
braam [Sun, 14 Jul 2002 00:13:54 +0000 (00:13 +0000)]
This commit contains probably 92% of the striping infrastructure
we need initially.  The most pervasive change is the introduction
of "lov_stripe_md" throughout the code.

In addtion several small little bugs were nailed in the locking --
more are outstanding.

The setup scripts are not yet capable of running this code.

Kernel patches were updated to include LOOKUP (to let runtests.sh
work).

21 years agoDon't load /etc/lustre/lustre.cfg if config files were passed on the command
pschwan [Sat, 13 Jul 2002 23:42:21 +0000 (23:42 +0000)]
Don't load /etc/lustre/lustre.cfg if config files were passed on the command
line.

21 years ago- Use refcounting to control lifetime of bulk descriptors.
shaver [Sat, 13 Jul 2002 23:38:45 +0000 (23:38 +0000)]
- Use refcounting to control lifetime of bulk descriptors.
- Add some LDLM timeout #warnings for future reference.
- Make all bulk-desc cleanup happen asynchronously, for cleaner interrupt (and
  later timeout) handling.
- Have OSC wait for brw_write reply before beginning bulk operations, for
  symmetry with brw_read and general niceness.
- Fix reply-for-freed-request check to use proper XID width.
- Use l_killable_pending in ptlrpc_check_bulk_sent, for consistency. (Also in the
  new ptlrpc_check_bulk_received sibling.)

21 years agoosc_connect needs to go with level LUSTRE_CONN_NEW. Fixed.
pschwan [Sat, 13 Jul 2002 19:10:22 +0000 (19:10 +0000)]
osc_connect needs to go with level LUSTRE_CONN_NEW.  Fixed.

21 years ago- Added match_or_enqueue helper function
pschwan [Sat, 13 Jul 2002 19:03:23 +0000 (19:03 +0000)]
- Added match_or_enqueue helper function
- fixed userspace build problem in lustre_lib.h
- added intent-based lookup code
- fixed intent-based setattr
- added mds_fid2locked_dentry and mds_name2locked_dentry helpers
- don't crash in ptlrpc_reply, just warn of API violation
- update create.pl to open instead of mcreate

21 years agoFixed warning introduced yesterday
pschwan [Sat, 13 Jul 2002 18:51:02 +0000 (18:51 +0000)]
Fixed warning introduced yesterday

21 years agoDon't try to auto-load obd module when we are just in the process of
adilger [Fri, 12 Jul 2002 20:44:28 +0000 (20:44 +0000)]
Don't try to auto-load obd module when we are just in the process of
registering it.

21 years agoUse unique OST fault-injection code instead of duplicate MDS code.
adilger [Fri, 12 Jul 2002 20:43:44 +0000 (20:43 +0000)]
Use unique OST fault-injection code instead of duplicate MDS code.

21 years agolist_mods() should not return an error if we are not waiting.
adilger [Fri, 12 Jul 2002 20:38:06 +0000 (20:38 +0000)]
list_mods() should not return an error if we are not waiting.

21 years ago- Added some temporary LDLM_LOCK_PUT/LDLM_LOCK_GET macros to aid in debugging
pschwan [Fri, 12 Jul 2002 16:48:13 +0000 (16:48 +0000)]
- Added some temporary LDLM_LOCK_PUT/LDLM_LOCK_GET macros to aid in debugging
- Fixed a couple more places where we dereference 'lock' after ldlm_lock_put()
- Fixed a showstopper in ldlm_handle2lock that was causing serious problems.
- The DLM in UML with 3 mountpoints now survives pretty much any load that
  create.pl can throw at it

21 years ago- Fixed up some OST handler functions to always return a nonzero error when
pschwan [Fri, 12 Jul 2002 16:23:03 +0000 (16:23 +0000)]
- Fixed up some OST handler functions to always return a nonzero error when
  they don't set request->rq_repmsg (this should fix Eric's RPC crash)

21 years ago- add lustre initscript
rread [Fri, 12 Jul 2002 09:21:30 +0000 (09:21 +0000)]
- add lustre initscript

21 years ago- change version of HEAD to 0.5
rread [Fri, 12 Jul 2002 08:53:35 +0000 (08:53 +0000)]
- change version of HEAD to 0.5

21 years ago- fix typos
rread [Fri, 12 Jul 2002 08:53:03 +0000 (08:53 +0000)]
- fix typos

21 years ago- add module aliases to /modules.conf
rread [Fri, 12 Jul 2002 07:01:47 +0000 (07:01 +0000)]
- add module aliases to /modules.conf

21 years ago- fix rpm build
rread [Fri, 12 Jul 2002 07:01:01 +0000 (07:01 +0000)]
- fix rpm build

21 years ago- fixes from santa fe
rread [Fri, 12 Jul 2002 06:59:55 +0000 (06:59 +0000)]
- fixes from santa fe

21 years ago- run config in production
rread [Fri, 12 Jul 2002 03:02:27 +0000 (03:02 +0000)]
- run config in production

21 years agoRemove hard-coded mountpoint from runtests script.
adilger [Thu, 11 Jul 2002 22:55:00 +0000 (22:55 +0000)]
Remove hard-coded mountpoint from runtests script.

21 years agoFix statfs for new ptlrpc interface in head.
adilger [Thu, 11 Jul 2002 22:53:34 +0000 (22:53 +0000)]
Fix statfs for new ptlrpc interface in head.

21 years agoRemove sgid #warning. This is already done in the filesystem and the MDS.
adilger [Thu, 11 Jul 2002 22:52:24 +0000 (22:52 +0000)]
Remove sgid #warning.  This is already done in the filesystem and the MDS.
It does not really make sense to do it in the MDC.

21 years agoI broke llcleanup earlier. This fixes it.
pschwan [Thu, 11 Jul 2002 22:43:00 +0000 (22:43 +0000)]
I broke llcleanup earlier.  This fixes it.

21 years ago- The server side of the DLM wasn't always handling the invalid handle case
pschwan [Thu, 11 Jul 2002 15:37:29 +0000 (15:37 +0000)]
- The server side of the DLM wasn't always handling the invalid handle case
very gracefully.  Since there are races that make invalid handles not hugely
uncommon, that's fixed now.
- Even on the client side, we could get cancelled between ldlm_lock_decref() and
ldlm_lock_cancel(), thus passing an invalid handle into cancel().  This is
handled now.
- Fixed another leaked request.
- Fixed an IT_SETATTR bug, even though we don't currently exercise that code
path.
- Fixed common.sh for the multiple mount cleanup case
- llcleanup.sh was unloading things in the wrong order (MDS is pinned by LDLM
and needs to be unloaded later in the process).  Fixed.

21 years agomdc_connect needs to use a level of LUSTRE_CONN_NEW.
shaver [Wed, 10 Jul 2002 20:31:23 +0000 (20:31 +0000)]
mdc_connect needs to use a level of LUSTRE_CONN_NEW.
(I broke mounting on the trunk before, this fixes it.)

21 years agoDefault to an rq_level of LUSTRE_CONN_FULL.
shaver [Mon, 8 Jul 2002 19:23:02 +0000 (19:23 +0000)]
Default to an rq_level of LUSTRE_CONN_FULL.
Use l_killable_pending in place of explicit lists of tests.

21 years ago- updates for rpm
rread [Sat, 6 Jul 2002 23:00:46 +0000 (23:00 +0000)]
- updates for rpm

21 years agoChange TCP port number.
adilger [Fri, 5 Jul 2002 23:26:23 +0000 (23:26 +0000)]
Change TCP port number.

21 years agoChange TCP acceptor port number to be 2432, an unused Coda port. You need
adilger [Fri, 5 Jul 2002 23:21:40 +0000 (23:21 +0000)]
Change TCP acceptor port number to be 2432, an unused Coda port.  You need
to update all of your <myname>.cfg files (or XML if you use it) to use the
new port number if you want to use tcpdump, and to maintain consistency
across scripts.

21 years agoMove the on-wire stucts/definitions into lustre_idl.h, for the CVS head.
uid34591 [Fri, 5 Jul 2002 22:50:42 +0000 (22:50 +0000)]
Move the on-wire stucts/definitions into lustre_idl.h, for the CVS head.

21 years agoFix "tags" target if portals is a symlink
uid34591 [Fri, 5 Jul 2002 22:39:03 +0000 (22:39 +0000)]
Fix "tags" target if portals is a symlink

21 years agoInclude headers explicitly that were previously included from lustre_net.h.
uid34591 [Fri, 5 Jul 2002 21:53:27 +0000 (21:53 +0000)]
Include headers explicitly that were previously included from lustre_net.h.