Whamcloud - gitweb
fs/lustre-release.git
21 years agomerge b_devel into b_eq
phil [Fri, 28 Nov 2003 08:12:52 +0000 (08:12 +0000)]
merge b_devel into b_eq

21 years agob=2254
phil [Fri, 28 Nov 2003 06:54:24 +0000 (06:54 +0000)]
b=2254
At truncate time, ext3 is zeroing indirect blocks and writing a
transaction to the journal.  At some point, perhaps when that
transaction commits, it gives the dirty buffer for that indirect block
to the buffer cache to write to disk.  Meanwhile, having marked that
block as unused, it reallocates it to us as a normal data block into
which IOR puts some data.

The obdfilter writes that block with brw_kiovec, which writes
immediately to the disk with no regard for what data might be in the
buffer cache.  Shortly thereafter, the buffer cache writes the block
of zeroes over top our valuable data.

The correct fix is to modify our special ext3 block allocation code to
look in the buffer cache for us, and discard any pending writes to the
newly-allocated blocks, much like the direct I/O code does.  As a
workaround, for kernels which do not yet have this change, I added
some code to the obdfilter to do this after the call to
ext3_map_inode_page returns.

This introduces kernel version 32, but doesn't force an upgrade.  I
updated the kernel patches for 2.4.18 and 2.4.20, but not 2.6.  I also:
 - tested the ext3_map_inode_page change on vanilla-2.4.20
 - tested the workaround change on chaos-2.4.18
 - compile-tested a version 32 chaos-2.4.18 kernel

21 years agoSlightly less obvious, but still very innocent, parts of b_eq:
phil [Thu, 27 Nov 2003 06:31:33 +0000 (06:31 +0000)]
Slightly less obvious, but still very innocent, parts of b_eq:

 - code or includes in the main build, but contained in #if __KERNEL__
 - userspace-only portals code
 - very few and minor other changes, such as renaming a function

21 years agoInnocent beginnings of the b_eq merge: the liblustre directory, which
phil [Thu, 27 Nov 2003 06:01:11 +0000 (06:01 +0000)]
Innocent beginnings of the b_eq merge: the liblustre directory, which
accounts for almost half of the patch and doesn't build in kernelspace.

21 years ago* Added a generic facility for callback to be called any time
eeb [Wed, 26 Nov 2003 18:40:20 +0000 (18:40 +0000)]
*   Added a generic facility for callback to be called any time
    l_wait_event() happens, to take the place of daemons in the kernel
    implementation

*   Changed liblustre_services() to use the new callback.

*   Backed out the liblustre OSC rpcd changes, and converted that to use
    the new callback too.

*   Disabled mmap in alloc_pages() FTTB (it seems to screw up under linux;
    dunno exactly why right now).

*   Left some hacking utensils behind in tcpnal_send() FTTB, just to make
    it easy to see what's getting sent.

*   Added some error checking in tcpnal_send(); it will abort if something
    screws up.

21 years agob=2296
rread [Wed, 26 Nov 2003 01:26:49 +0000 (01:26 +0000)]
b=2296
r=shaver

- Build ldlm into ptlrpc.
- Remove all the ptlrpc_ldlm_* hooks

This  almost certainly breaks building when $(SRDIR) != $(OBJDIR).

21 years ago1. move llog_origin_handle_cancel to llog_server.c
tianying [Tue, 25 Nov 2003 12:51:51 +0000 (12:51 +0000)]
1. move llog_origin_handle_cancel to llog_server.c
2. destroy useless plain logs during both llog_cleanup and llog_setup
3. change some CERRORs to CWARNs while just print out status information

21 years agoliblustre: follow up last merge from b_devel.
ericm [Tue, 25 Nov 2003 12:14:09 +0000 (12:14 +0000)]
liblustre: follow up last merge from b_devel.

21 years agomerge b_devel to b_eq: 20031125. kernel part only.
ericm [Tue, 25 Nov 2003 11:34:23 +0000 (11:34 +0000)]
merge b_devel to b_eq: 20031125. kernel part only.

21 years agocleanup clients correctly in insanity.sh
rread [Mon, 24 Nov 2003 23:57:01 +0000 (23:57 +0000)]
cleanup clients correctly in insanity.sh
set the mds upcall, so the mds can recovery from ost failures

21 years agotcpnal: use mutex to protect tcpnal_write, since in case of bulk put,
ericm [Mon, 24 Nov 2003 03:06:06 +0000 (03:06 +0000)]
tcpnal: use mutex to protect tcpnal_write, since in case of bulk put,
nal_thread will compete with upper thread to write into same socket.

21 years agoAdd fsfilt_map_inode_page() instead of calling ext3_map_inode_page()
adilger [Sat, 22 Nov 2003 19:32:05 +0000 (19:32 +0000)]
Add fsfilt_map_inode_page() instead of calling ext3_map_inode_page()
directly from filter_direct_io().  Update fsfilt_extN.c.

21 years agoInclude add_page_private.patch in chaos kernel series.
adilger [Fri, 21 Nov 2003 23:27:33 +0000 (23:27 +0000)]
Include add_page_private.patch in chaos kernel series.

21 years ago- patch to fix 64-bit pointer arithmetic bug in the ext3 extended
alex [Fri, 21 Nov 2003 15:06:25 +0000 (15:06 +0000)]
- patch to fix 64-bit pointer arithmetic bug in the ext3 extended
  attributes code.

21 years agomove client_df to test-framework.
rread [Fri, 21 Nov 2003 09:10:56 +0000 (09:10 +0000)]
move client_df to test-framework.

21 years ago- integrate support for FAILURE_MODE into test-framework
rread [Fri, 21 Nov 2003 08:42:35 +0000 (08:42 +0000)]
- integrate support for FAILURE_MODE into test-framework
- add configs for mdev

21 years ago- cleanup old active files
rread [Fri, 21 Nov 2003 03:27:13 +0000 (03:27 +0000)]
- cleanup old active files
- cleanup insanity test_0
- add checks to make sure the nodes are up during setup

21 years agoadd usage() to test_framework, and use getopts cause getopt is broken
rread [Fri, 21 Nov 2003 01:04:55 +0000 (01:04 +0000)]
add usage() to test_framework, and use getopts cause getopt is broken
insanity fixes for hardware

21 years agoVerify data from a read that spans a failover.
shaver [Fri, 21 Nov 2003 00:19:11 +0000 (00:19 +0000)]
Verify data from a read that spans a failover.

21 years agoVerify data written by operations that span a failure.
shaver [Fri, 21 Nov 2003 00:05:31 +0000 (00:05 +0000)]
Verify data written by operations that span a failure.

21 years agoOverride the local_nid in the logs during zeroconf mount. If the
rread [Thu, 20 Nov 2003 21:42:24 +0000 (21:42 +0000)]
Override the local_nid in the logs during zeroconf mount.  If the
local_nid option is not specified, then mount.lustre will use
the hostname for socknal, or /proc/elan/device0/position for qswnal.

For nettype=elan, if remote_nid isn't set the mount.lustre will
attempt to parse the hostname and use the first series of numbers as
the nid.  So, mdev20-eth1 will become elan id 20.

21 years agoSmall test and upcall fixes; we fail over during a write now.
shaver [Thu, 20 Nov 2003 21:41:15 +0000 (21:41 +0000)]
Small test and upcall fixes; we fail over during a write now.

21 years agoCreate a per-fs lock for obdfilter block allocation. This is held only
adilger [Thu, 20 Nov 2003 19:05:02 +0000 (19:05 +0000)]
Create a per-fs lock for obdfilter block allocation.  This is held only
during actual block allocation and not during RPCs or writes.  This allows
us to allocate contiguous chunks of disk (if available) up to the size of
each RPC, instead of interleaving block allocations.

It slows down writes in the contention case, because we might be holding
the lock while waiting for a bitmap or something to be loaded from disk,
and in the current 2.4 IO code reads-behind-lots-of-writes can be punishing.
We might benefit here and elsewhere from AKPM's read priority patch.

The big benefit is that at read time, or after some amount of create-delete
we don't have a maximally fragmented block allocation to deal with, which
causes pathological seeking on the disks.

b=2260
r=peter,phil

21 years ago- Integrate insanity.sh with test-framework
rread [Thu, 20 Nov 2003 07:42:54 +0000 (07:42 +0000)]
- Integrate insanity.sh with test-framework
- Centralize the config data to files in tests/cfg/
- Different config file can be used by setting CONFIG=file or
  using command line option:   --config <file>
- FAILURE_MODE option for insanity.sh:
  FAILURE_MODE=SOFT - shutdown services with lconf --force
  FAILURE_MODE=HARD - power down the service nodes using $POWER_DOWN
     and $POWER_UP to reboot.
  This feature should be integrated with the other tests.

- replay-single.sh, replay-dual.sh, and conf-sanity.sh all pass on hardware
- replay-ost-dual.sh and insanity.sh don't pass yet

21 years ago- print the correct strings
rread [Thu, 20 Nov 2003 00:49:00 +0000 (00:49 +0000)]
- print the correct strings

21 years ago- Remove last references to OBD_CLASS_UUID. Now the obd->obd_uuid is
rread [Tue, 18 Nov 2003 20:36:15 +0000 (20:36 +0000)]
- Remove last references to OBD_CLASS_UUID.  Now the obd->obd_uuid is
used as the client uuid for self and lctl exports.

21 years agoinsanity.sh - this is test17 turned into a regular test, if you could
rread [Tue, 18 Nov 2003 09:52:36 +0000 (09:52 +0000)]
insanity.sh - this is test17 turned into a regular test, if you could
call this test regular. Still a work in progress, but does at least
setup/cleanup.

21 years ago- give the user an error message when the dump file can't be opened
zab [Mon, 17 Nov 2003 18:46:30 +0000 (18:46 +0000)]
- give the user an error message when the dump file can't be opened

21 years ago- kernel part of b1933
alex [Mon, 17 Nov 2003 17:32:25 +0000 (17:32 +0000)]
- kernel part of b1933

21 years agoreturn true value from zconf_mount
rread [Sat, 15 Nov 2003 01:49:00 +0000 (01:49 +0000)]
return true value from zconf_mount

21 years agob=2250
rread [Sat, 15 Nov 2003 01:17:53 +0000 (01:17 +0000)]
b=2250

create lctl commands set_lustre_upcall and set_timeout, so they
can be saved as part of the 0conf log.

21 years agoCode is dead, dead, dead, and keeps showing up in my greps.
adilger [Fri, 14 Nov 2003 22:50:30 +0000 (22:50 +0000)]
Code is dead, dead, dead, and keeps showing up in my greps.

21 years agoMore test cleanups.
rread [Fri, 14 Nov 2003 22:32:26 +0000 (22:32 +0000)]
More test cleanups.

21 years agoIn osc_interpret_create, don't overwrite rc if it's already set.
rread [Fri, 14 Nov 2003 22:24:19 +0000 (22:24 +0000)]
In osc_interpret_create, don't overwrite rc if it's already set.

21 years agoInteroperability for different PAGE_SIZE/wordsize machines. Tested on ia64
adilger [Fri, 14 Nov 2003 20:12:23 +0000 (20:12 +0000)]
Interoperability for different PAGE_SIZE/wordsize machines.  Tested on ia64
and i386 separately, and with ia64 client + i386 MDS/OST.

Mostly aligning structs to have 64-bit fields aligned on 64-bit boundaries.
Remove some VFS constants and replace them with Lustre constants on the wire.
Since the MDS doesn't really open files itself, we don't need to convert from
the wire constants back to local flags at all.

Frobbing of niobufs on the targets to split them into PAGE_SIZE chunks (this
may be a problem on large PAGE_SIZE servers with small PAGE_SIZE clients,
not sure yet).

I have tested and this appears to be compatible with old filesystems.
At worst we should only need another --write_conf on the MDS.

b=686, b=1821, b=1343, b=2042

21 years agoI touched one thing in lustre_mds.h, and I was dismayed to see the
phil [Fri, 14 Nov 2003 09:24:33 +0000 (09:24 +0000)]
I touched one thing in lustre_mds.h, and I was dismayed to see the
entire source tree rebuild!

It took just a few minutes to remove a number of #includes which
violated all manner of abstraction boundary.  ericm assures me that I
have not broken anything major in liblustre.

Also, I moved a good chunk of inline functions into llite_internal.h
(which is where I secretly believe that most of lustre_lite.h will end up),
and renamed ll_ino2fid (which no longer takes an inode) to mdc_pack_fid.

21 years ago- don't crash if there are no options
rread [Thu, 13 Nov 2003 19:22:32 +0000 (19:22 +0000)]
- don't crash if there are no options

21 years ago* Changed liblustre/libtest to call into lctl for interactive echo tests
eeb [Thu, 13 Nov 2003 18:49:21 +0000 (18:49 +0000)]
*    Changed liblustre/libtest to call into lctl for interactive echo tests

*    Changed some global names in the utilities to avoid conflicts

*    Added set_ioc_handler() to allow portals/obd ioctl redirects to
     the liblustre handler (as well as dumping to a file)

21 years ago- kmem_cache_validate patch has been removed from all the series
alex [Thu, 13 Nov 2003 15:06:14 +0000 (15:06 +0000)]
- kmem_cache_validate patch has been removed from all the series

NOTE: all the supported series still build

21 years agoopts could be NULL, cause segfault.
ericm [Thu, 13 Nov 2003 12:06:04 +0000 (12:06 +0000)]
opts could be NULL, cause segfault.

21 years agoliblustre: minor fix for last merge.
ericm [Thu, 13 Nov 2003 08:02:59 +0000 (08:02 +0000)]
liblustre: minor fix for last merge.

21 years agoagain merge b_devel to b_eq 20031113
ericm [Thu, 13 Nov 2003 07:54:51 +0000 (07:54 +0000)]
again merge b_devel to b_eq 20031113
yesterday's merge brought in some nasty bugs.

21 years ago b:2215 - OSTs fetch unlink llog records from MDS post replay
tianying [Thu, 13 Nov 2003 05:56:59 +0000 (05:56 +0000)]
  b:2215 - OSTs fetch unlink llog records from MDS post replay

     1. add lop_connect and lop_precleanup to llog_operations
     2. rename llog_obd_ctxt to llog_ctxt; llog_commit_data to llog_canceld_ctxt
     3. split out llog functions in llog_client.c and llog_server.c and remove llogd.c
     4. add one test-59 to sanity.sh to verify cancellation of llog records async
     5. fix calling of mds_cleanup_orphans, add test-34 to replay-single.sh
     6. fix some codes about llog

21 years agobraino
rread [Thu, 13 Nov 2003 01:52:01 +0000 (01:52 +0000)]
braino

21 years ago- return replay-ost-single to runable state
rread [Thu, 13 Nov 2003 01:27:35 +0000 (01:27 +0000)]
- return replay-ost-single to runable state
  now just need to pass

- commonize the zconf mount/umount

21 years ago- include pcfg's in dump_log
rread [Thu, 13 Nov 2003 00:04:13 +0000 (00:04 +0000)]
- include pcfg's in dump_log

21 years agoRemove straggler .pc file.
adilger [Wed, 12 Nov 2003 23:04:30 +0000 (23:04 +0000)]
Remove straggler .pc file.

21 years ago- remove extra dec in the error cleanup case.
rread [Wed, 12 Nov 2003 21:15:12 +0000 (21:15 +0000)]
- remove extra dec in the error cleanup case.

21 years agob=2239
shaver [Wed, 12 Nov 2003 14:10:30 +0000 (14:10 +0000)]
b=2239
- rq_timeout halving for imp_server_timeout imps

- set sending_error in the resent-RPC case in check_set

- wake and error out if a create fails due to invalidated OSC
r=phik

21 years agomerge b_devel to b_eq: 20031112
ericm [Wed, 12 Nov 2003 11:03:26 +0000 (11:03 +0000)]
merge b_devel to b_eq: 20031112
for robert's final zeroconf code.

21 years agoFix compile warning.
adilger [Wed, 12 Nov 2003 09:40:51 +0000 (09:40 +0000)]
Fix compile warning.

21 years ago- add /sbin/mount.lustre to rpm
rread [Wed, 12 Nov 2003 06:18:47 +0000 (06:18 +0000)]
- add /sbin/mount.lustre to rpm

21 years agofix cut-n-paste error
rread [Wed, 12 Nov 2003 06:16:30 +0000 (06:16 +0000)]
fix cut-n-paste error

21 years agoAdd new "lustre" fs type which supports only zeroconf mounts.
rread [Tue, 11 Nov 2003 23:13:17 +0000 (23:13 +0000)]
Add new "lustre" fs type which supports only zeroconf mounts.
 - old zeroconf client code removed from lustre-lite, and the lconf
   --zeroconf option deleted
 - common code factored out for ll/lustre fill_super and put_super
 - lconf still uses lustre_lite (but not for long)

llmount will be used by mount for lustre filesystems if copied to
/sbin/mount.lustre:

  mount -t lustre mds_host:/mds_servicee/profile  /mnt/lustre

Multiple mounts of the same filesystem are supported.

Remove unused mds and filter nspath code.

21 years ago* removed some more build warnings
eeb [Tue, 11 Nov 2003 16:45:19 +0000 (16:45 +0000)]
*   removed some more build warnings

21 years ago* reduced the number of warnings for liblustre compilation
eeb [Tue, 11 Nov 2003 16:21:56 +0000 (16:21 +0000)]
*  reduced the number of warnings for liblustre compilation

21 years agomerge b_devel to b_eq: 20031111, for recent kernel patches updates.
ericm [Tue, 11 Nov 2003 10:05:28 +0000 (10:05 +0000)]
merge b_devel to b_eq: 20031111, for recent kernel patches updates.

21 years agoDespite Zach's assurances, it seems that conditional_schedule() is a
adilger [Mon, 10 Nov 2003 16:01:27 +0000 (16:01 +0000)]
Despite Zach's assurances, it seems that conditional_schedule() is a
low-latency RHism, so define a compat macro for non-RH systems.
b=2227

21 years agoliblustre: temprorily add accepter installation, for Jerrifer build rpm.
ericm [Mon, 10 Nov 2003 10:45:40 +0000 (10:45 +0000)]
liblustre: temprorily add accepter installation, for Jerrifer build rpm.

21 years agotcpnal: need exchange NID when tcp connection created, ksocknal now default
ericm [Sat, 8 Nov 2003 15:22:46 +0000 (15:22 +0000)]
tcpnal: need exchange NID when tcp connection created, ksocknal now default
require this.

21 years agotransparently convert '-' to '_' in command line options.
rread [Fri, 7 Nov 2003 21:45:46 +0000 (21:45 +0000)]
transparently convert '-' to '_' in command line options.

so, for example, you can use --write-conf  if you want

21 years ago- fix typo so replay-single test_32 runs again.
rread [Fri, 7 Nov 2003 20:58:06 +0000 (20:58 +0000)]
- fix typo so replay-single test_32 runs again.

21 years agoi really should make this a compile time option, at least.
rread [Fri, 7 Nov 2003 18:56:51 +0000 (18:56 +0000)]
i really should make this a compile time option, at least.

21 years ago- remove use of "lctl device_list" from lconf, now using
rread [Fri, 7 Nov 2003 18:52:22 +0000 (18:52 +0000)]
- remove use of "lctl device_list" from lconf, now using
    /proc/fs/lustre/devices

21 years agob=2225 fix obd_self_export issues
rread [Fri, 7 Nov 2003 18:52:21 +0000 (18:52 +0000)]
b=2225 fix obd_self_export issues

21 years agomerge b_devel to b_eq: 20031107
ericm [Fri, 7 Nov 2003 15:50:56 +0000 (15:50 +0000)]
merge b_devel to b_eq: 20031107
kernel passed sanity, but liblustre broken since lconf changes. need fix
by robert.

21 years agoAdd define of LCONF LMC LCTL in replay-dual.sh replay-single.sh replay-ost-single...
sice [Fri, 7 Nov 2003 08:22:16 +0000 (08:22 +0000)]
Add define of LCONF LMC LCTL in replay-dual.sh replay-single.sh replay-ost-single.sh test-framework.sh

21 years agoRemind the user to use 'cfg_device' to set target device for config commands.
rread [Fri, 7 Nov 2003 06:19:32 +0000 (06:19 +0000)]
Remind the user to use 'cfg_device' to set target device for config commands.

21 years ago- change NETWORKTYPE to NETTYPE, and make sure it can be overridden.
rread [Wed, 5 Nov 2003 08:35:46 +0000 (08:35 +0000)]
- change NETWORKTYPE to NETTYPE, and make sure it can be overridden.

21 years agor=2152
rread [Wed, 5 Nov 2003 05:57:47 +0000 (05:57 +0000)]
r=2152

replay-single.sh now supports multiple nodes. The script must be run
on the client, but the mds, mdsfailover, and ost can all be different
nodes. If mdsfailover_HOST is set, then the MDS service  will be
failed between the two mds nodes.

uml1# PDSH="pdsh -w" mds_HOST=uml2 ost_HOST=uml3 mdsfailover_HOST=uml4
  ./replay-single.sh

It still runs on a single node as before.

21 years agob=1028
phil [Wed, 5 Nov 2003 00:47:41 +0000 (00:47 +0000)]
b=1028
Andreas pointed out that we already have a function obdo_from_inode,
and that we might as well pack all valid fields, and let the client
take what it can.

21 years agob=1028
phil [Wed, 5 Nov 2003 00:00:19 +0000 (00:00 +0000)]
b=1028
r=zab
Return the object's block count from write requests, and store it in
the client inode.

21 years agoDon't do setattr after transaction handle has been committed.
adilger [Mon, 3 Nov 2003 23:41:32 +0000 (23:41 +0000)]
Don't do setattr after transaction handle has been committed.
Combine size and timestamp setattrs, and update size under i_sem.
Use client timestamps instead of server timestamps for files.

21 years agoreturn the osc create grow counts to 2000, which sadly doesn't work
rread [Mon, 3 Nov 2003 07:00:17 +0000 (07:00 +0000)]
return the osc create grow counts to 2000, which sadly doesn't work
very well on my machine.

21 years ago- filter_finish_transno() moved in filter_direct_io()
alex [Sat, 1 Nov 2003 11:48:07 +0000 (11:48 +0000)]
- filter_finish_transno() moved in filter_direct_io()
- fsfilt_commit_async() is called unconditionally in filter_direct_io()

21 years agoland zcfg on devel
rread [Fri, 31 Oct 2003 19:39:27 +0000 (19:39 +0000)]
land zcfg on devel

- includes changes from b_llogging

- the MDS -> LOV connection is created using the MDS config log, so the
  log must exist. Lconf will create the log automatically when
  --reformat is used. To create the config logs on an existing
  filesystem, run lconf on the MDS with --write_conf. This only needs to
  be done on the MDS.

- LOV does not connect to the MDS during setup.  Instead, the MDS and
  MDC use obd_get_info("lovdesc") to get the stripe info. The LOVDESC
  and LOVTGTS files on the MDS are no longer used (and not created
  on new filesystems.)

- Zeroconf clients are new support by lconf --zeroconf, and
  replay-single.sh uses this to mount the client.  The exact arguments
  needed for zeroconf will be changing quickly, so don't use
  lconf --zeroconf in other scripts yet.  Instead, once the dust
  clears, lconf will be changed to always do zeroconf mounts, and
  eventually when don't need lconf anymore is when we start changing
  the test scripts.

21 years agomerge devel zcfg
rread [Fri, 31 Oct 2003 18:50:49 +0000 (18:50 +0000)]
merge devel zcfg

21 years agomerge devel zcfg
rread [Fri, 31 Oct 2003 06:53:23 +0000 (06:53 +0000)]
merge devel zcfg

21 years ago- async commit/wait API have been changed a bit. it used oti_handler, but
alex [Wed, 29 Oct 2003 22:38:26 +0000 (22:38 +0000)]
- async commit/wait API have been changed a bit. it used oti_handler, but
  it must not

21 years agomerge devel zcfg
rread [Wed, 29 Oct 2003 19:35:58 +0000 (19:35 +0000)]
merge devel zcfg

21 years agob=2200: perform delorphan recovery when an OST is reintegrated.
shaver [Wed, 29 Oct 2003 17:35:57 +0000 (17:35 +0000)]
b=2200: perform delorphan recovery when an OST is reintegrated.
        - use generic per-obd notification system
- relay LOV's notifications to MDS
- support running delorphan on only one OST, if an OST
  UUID is specified
- suppress CERRORs and refailing for -EIO in interpret_create,
  because we handle it more gracefully now

r=phik.

21 years agoAdd extN compat.
adilger [Wed, 29 Oct 2003 08:37:23 +0000 (08:37 +0000)]
Add extN compat.

21 years agoMade a mistake in commit dsp.patch, just revert it, sorry
wangdi [Wed, 29 Oct 2003 08:02:47 +0000 (08:02 +0000)]
Made a mistake in commit dsp.patch, just revert it, sorry

21 years agofix patch confilcts on dsp.patch for 2.4.20-rh kernel
wangdi [Wed, 29 Oct 2003 07:52:31 +0000 (07:52 +0000)]
fix patch confilcts on dsp.patch for 2.4.20-rh kernel

21 years agomerge b_devel -> b_eq: 20031029
ericm [Wed, 29 Oct 2003 05:14:23 +0000 (05:14 +0000)]
merge b_devel -> b_eq: 20031029

kerenl pass sanity.sh, liblustre is broken

21 years agob_2036
youfeng [Wed, 29 Oct 2003 03:47:04 +0000 (03:47 +0000)]
b_2036
formats nids for readability, from "nid" to "nid (format)"
nid    : using LPX64 everywhere
format : "%u:%d.%d.%d.%d" for IP NIDS (leading part iscluster ID)
         "%u:%u" for non-IP NIDs (leading part is cluster ID)

21 years agob: 1990
niu [Wed, 29 Oct 2003 02:28:00 +0000 (02:28 +0000)]
b: 1990
r: Andreas

Padding record header to 64 bits size.

21 years agomerge devel zcfg
rread [Tue, 28 Oct 2003 21:46:17 +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 ago- brown paper bug fixed: it seems I pressed dd randomly :(
alex [Tue, 28 Oct 2003 17:49:06 +0000 (17:49 +0000)]
- brown paper bug fixed: it seems I pressed dd randomly :(

21 years ago- b2188: filter fixes: sync transaction, commit earlier
alex [Tue, 28 Oct 2003 13:19:17 +0000 (13:19 +0000)]
- b2188: filter fixes: sync transaction, commit earlier

21 years agob: 1990
niu [Tue, 28 Oct 2003 09:29:05 +0000 (09:29 +0000)]
b: 1990
r: Andreas

Eliminate u8/u16 from log header, move padding from record header/tail to
record body.

21 years agob: 1990
niu [Mon, 27 Oct 2003 03:28:43 +0000 (03:28 +0000)]
b: 1990
r: Peter

 * rewrite llog_cat_cancel_rec, zap the empty canceled plain log.
 * add test: open & destroy plain log.
 * add test: verify catalog log entries after cancel many many records.

21 years ago- all obd operations should return int
braam [Sat, 25 Oct 2003 19:19:49 +0000 (19:19 +0000)]
- all obd operations should return int
- in the process capture a double free in a prep_async_page error case
  in rw.c
- remove unused logging code
- add a mini fix from Tian to not fail in llog_obd_setup when
  re-entering it.

21 years agomerge1.sh devel zcfg
rread [Fri, 24 Oct 2003 18:13:07 +0000 (18:13 +0000)]
merge1.sh devel zcfg

Resolved conflicts with Tian Ying's recent checkins.  Also forced to
fix local export cleanup problem by adding an obd_init_export()
method. This is called from class_new_export(), and is for devices to
initialize their export data. Allos local (as opposed to network)
connections to be used and cleaned up correctly. Only
mds_init_export() implemented.

21 years ago- trusted EAs support in ext3
alex [Thu, 23 Oct 2003 21:42:40 +0000 (21:42 +0000)]
- trusted EAs support in ext3
- MAX_VERSION increase

21 years agoThis is a bugfix for #1838. Most of the patch by A. Dilger, I only
green [Thu, 23 Oct 2003 20:18:30 +0000 (20:18 +0000)]
This is a bugfix for #1838. Most of the patch by A. Dilger, I only
propagated parental time change to all operations, not only mkdir.

21 years agob=2150: handle ENOSPC encountered during precreation
shaver [Thu, 23 Oct 2003 20:11:35 +0000 (20:11 +0000)]
b=2150: handle ENOSPC encountered during precreation
r=zab

21 years agoBack out a small LLP change: instead of deleting a bunch of files just
phil [Thu, 23 Oct 2003 18:46:03 +0000 (18:46 +0000)]
Back out a small LLP change: instead of deleting a bunch of files just
to re-create them, I was doing an initial OST create which includes
the initial set of precreated files.  Stop doing that.

Those files might have a little data in them, or some might have been
created and removed already, and both of those are bad.

21 years agob=2110
tianying [Thu, 23 Oct 2003 13:33:28 +0000 (13:33 +0000)]
b=2110
r=Peter

1.Unlink log record will be created when mds_reint_unlink, mds_mfd_close or mds_cleanup_orphans is called. And that record will be cancelled by MDS when it receives the commit callback for filter_destroy.

2.It resolves the resetup problem for lustre built with enable-orphans on.

3.It resolves the cleanup problem of mds when recovery procedure is abort.

21 years agoAdd back hlist compatibility to lustre_compat25.h until we update the kernel
adilger [Wed, 22 Oct 2003 16:49:25 +0000 (16:49 +0000)]
Add back hlist compatibility to lustre_compat25.h until we update the kernel
so Coop can continue testing at LLNL.