Whamcloud - gitweb
fs/lustre-release.git
17 years agoBranch HEAD
scjody [Tue, 13 Mar 2007 19:28:16 +0000 (19:28 +0000)]
Branch HEAD

Merge SLES 10 build fix from b1_4.

17 years agofix patchless client build breakage with vanila 2.6.19 kernel.
shadow [Tue, 13 Mar 2007 06:34:41 +0000 (06:34 +0000)]
fix patchless client build breakage with vanila 2.6.19 kernel.
b=11804
r=adilger

17 years agoBranch HEAD
scjody [Tue, 13 Mar 2007 02:29:04 +0000 (02:29 +0000)]
Branch HEAD

Update SLES 10 kernel to 2.6.16.27-0.9.

17 years agoLand b1_6_bug11013 onto HEAD (20070313_0924)
yzy [Tue, 13 Mar 2007 01:43:45 +0000 (01:43 +0000)]
Land b1_6_bug11013 onto HEAD (20070313_0924)
b=11013
r=adilger

implement a common hash table, so it will be used as a hash class.
in this patch, finished :
1. " UUID <--> export " hash
2. " NID  <--> export " hash
3. "NetPeer <--> connect" hash

17 years agoBranch HEAD
scjody [Tue, 13 Mar 2007 00:23:26 +0000 (00:23 +0000)]
Branch HEAD

Update RHEL 4 kernel to 2.6.9-42.0.10.EL.

17 years agoBranch HEAD
scjody [Mon, 12 Mar 2007 22:52:29 +0000 (22:52 +0000)]
Branch HEAD

Merge from b_release_1_4_9: update SLES 9 kernel to 2.6.5-7.283.

17 years agoBranch HEAD
adilger [Sat, 10 Mar 2007 17:54:16 +0000 (17:54 +0000)]
Branch HEAD
Use intent directly instead of file->f_it, which doesn't exist on patchless
clients.
r=shadow (original patch)

17 years ago* Fixed bug noticed while Nic was investigating 11394 - kptllnd didn't
eeb [Sat, 10 Mar 2007 02:12:43 +0000 (02:12 +0000)]
*   Fixed bug noticed while Nic was investigating 11394 - kptllnd didn't
    propagate portals completion errors up to LNET.  The fix propagates the
    error and logs a D_NETERROR.

17 years agoclarify error message
nathan [Sat, 10 Mar 2007 00:03:14 +0000 (00:03 +0000)]
clarify error message

17 years ago* Changed syntax of "lctl ping target [timeout]", so "target" can be
eeb [Fri, 9 Mar 2007 21:38:32 +0000 (21:38 +0000)]
*   Changed syntax of "lctl ping target [timeout]", so "target" can be
    either a NID or a process ID.

17 years ago* removed PTL_EQ_HANDLER_NONE define now that cray portals defines it
eeb [Fri, 9 Mar 2007 21:08:17 +0000 (21:08 +0000)]
*    removed PTL_EQ_HANDLER_NONE define now that cray portals defines it

17 years agoBranch HEAD
bobijam [Fri, 9 Mar 2007 03:02:57 +0000 (03:02 +0000)]
Branch HEAD
b=11090
r=green

Description: Checking the version difference of client vs. server, report error if the gap is too big.

17 years agoBranch b_hd_newconfig
adilger [Fri, 9 Mar 2007 00:04:18 +0000 (00:04 +0000)]
Branch b_hd_newconfig
Add ChangeLog entry for bug 10889.

17 years agoBranch HEAD
adilger [Thu, 8 Mar 2007 23:54:31 +0000 (23:54 +0000)]
Branch HEAD
Have lustre/ChangeLog match b1_4 as much as possible.

17 years agoBranch HEAD
adilger [Thu, 8 Mar 2007 20:59:25 +0000 (20:59 +0000)]
Branch HEAD
Avoid duplicate definition of SYSIO_INTERFACE_NAME.
b=11175
r=nic (original patch)

17 years agoBranch HEAD
adilger [Thu, 8 Mar 2007 20:14:45 +0000 (20:14 +0000)]
Branch HEAD
Remove old, dangerous tcp zero copy patches.

17 years agoBranch HEAD
adilger [Thu, 8 Mar 2007 20:09:21 +0000 (20:09 +0000)]
Branch HEAD
Remove old, dangerous tcp-zero-copy patch from 2.6.18 series.

17 years ago* Fixed #if typo
eeb [Thu, 8 Mar 2007 19:33:45 +0000 (19:33 +0000)]
*   Fixed #if typo

17 years agocatch extra args without dashes
nathan [Thu, 8 Mar 2007 19:32:21 +0000 (19:32 +0000)]
catch extra args without dashes

17 years agocheck for dangling upgrade flag
nathan [Thu, 8 Mar 2007 18:37:18 +0000 (18:37 +0000)]
check for dangling upgrade flag

17 years agomust reset the upgrade flag or we try to upgrade every time the server
nathan [Thu, 8 Mar 2007 17:13:23 +0000 (17:13 +0000)]
must reset the upgrade flag or we try to upgrade every time the server
mounts. (recently introduced bug.)

17 years agoadd plot-llstat, move from sbin to bin
nathan [Wed, 7 Mar 2007 02:03:43 +0000 (02:03 +0000)]
add plot-llstat, move from sbin to bin

17 years agoChange name from --noformat to --dryrun for clarity
nathan [Tue, 6 Mar 2007 01:04:02 +0000 (01:04 +0000)]
Change name from --noformat to --dryrun for clarity

17 years agoupdate vers and supported kernels
nathan [Tue, 6 Mar 2007 00:09:13 +0000 (00:09 +0000)]
update vers and supported kernels

17 years agoBranch HEAD
scjody [Mon, 5 Mar 2007 18:05:08 +0000 (18:05 +0000)]
Branch HEAD

Revert my fix to bug 11636 since it broke the build.

b=11786

17 years ago* Support "lustrecvs b_self_test" to check out the lnet self-test development
eeb [Mon, 5 Mar 2007 12:38:06 +0000 (12:38 +0000)]
*   Support "lustrecvs b_self_test" to check out the lnet self-test development
    branch.

17 years agoBranch HEAD
adilger [Sat, 3 Mar 2007 14:23:10 +0000 (14:23 +0000)]
Branch HEAD
Remove the broken multi-mount-protection patch.

17 years agoBranch HEAD
adilger [Sat, 3 Mar 2007 14:21:20 +0000 (14:21 +0000)]
Branch HEAD
Add 1.4.9 release date.

17 years agoupdate tunefs examples
nathan [Fri, 2 Mar 2007 19:39:59 +0000 (19:39 +0000)]
update tunefs examples

17 years agob=11149
nathan [Fri, 2 Mar 2007 18:08:49 +0000 (18:08 +0000)]
b=11149
wrong end condition

17 years agob=11149
nathan [Fri, 2 Mar 2007 17:19:39 +0000 (17:19 +0000)]
b=11149
r=adilger
Add checks for null indicies in QOS code.
Also fix 'lfs df' and /proc...lov/target_obd.

17 years agoBranch HEAD
adilger [Fri, 2 Mar 2007 14:13:07 +0000 (14:13 +0000)]
Branch HEAD
Merge readlink return type configure check from b1_4.
b=11731
r=mjmac (original patch)

17 years agoBranch HEAD
adilger [Fri, 2 Mar 2007 13:56:21 +0000 (13:56 +0000)]
Branch HEAD
Enable CONFIG_E1000_DISABLE_PACKET_SPLIT to match b1_4.
Enable CONFIG_IPV6=m in UML kernel so that tcp symbols are exported.

17 years agob=11752
nathan [Thu, 1 Mar 2007 19:12:57 +0000 (19:12 +0000)]
b=11752
assert possibly caused by uninitialized var

17 years ago* fixed copy/paste error in previous 11684 commit
eeb [Thu, 1 Mar 2007 17:11:52 +0000 (17:11 +0000)]
*   fixed copy/paste error in previous 11684 commit

17 years agoBranch HEAD
scjody [Thu, 1 Mar 2007 14:58:57 +0000 (14:58 +0000)]
Branch HEAD
b=11636
r=green,adilger

Remove useless kernel configs.

17 years agoadd wait command to make this scriptable
nathan [Wed, 28 Feb 2007 01:34:42 +0000 (01:34 +0000)]
add wait command to make this scriptable

17 years ago* fixed cut/paste error on adding darwin sysctl for "debug_peer_on_timeout"
eeb [Tue, 27 Feb 2007 21:04:22 +0000 (21:04 +0000)]
*  fixed cut/paste error on adding darwin sysctl for "debug_peer_on_timeout"

17 years agofind lctl and mkfs.lustre if not local build
nathan [Tue, 27 Feb 2007 20:35:56 +0000 (20:35 +0000)]
find lctl and mkfs.lustre if not local build

17 years agomake sure liblustre test can start
nathan [Mon, 26 Feb 2007 22:55:32 +0000 (22:55 +0000)]
make sure liblustre test can start

17 years agoFix an error where the garbage may be returned by this function because
deen [Mon, 26 Feb 2007 22:06:30 +0000 (22:06 +0000)]
Fix an error where the garbage may be returned by this function because
of uninitialized pointer.

r=rmg@ (nathan)

17 years agoBranch HEAD
adilger [Mon, 26 Feb 2007 20:29:26 +0000 (20:29 +0000)]
Branch HEAD
Don't try to report stats on an incoming request we can't decode
(trips LASSERT in since RPCv2 support was added).

17 years agoBad incoming messages should never cause an assertion.
nathan [Mon, 26 Feb 2007 20:24:48 +0000 (20:24 +0000)]
Bad incoming messages should never cause an assertion.
Assertions are for verifying logic.

17 years agob=10316,r=eeb:
isaac [Mon, 26 Feb 2007 19:57:57 +0000 (19:57 +0000)]
b=10316,r=eeb:
-   fixed console chatter in case of -ETIMEDOUT.

17 years agob=10894
nathan [Mon, 26 Feb 2007 19:11:33 +0000 (19:11 +0000)]
b=10894
r=adilger
Clarify watchdog timer units.
Add ability to change MDT ldlm timeout.

17 years agob=11028
green [Sun, 25 Feb 2007 22:51:00 +0000 (22:51 +0000)]
b=11028
r=adilger

Fix cfs_timeval_sub to properly fill struct timeval.
Correctly assign microseconds to microsecondsand seconds to seconds.

17 years agoBranch:HEAD
wangdi [Sun, 25 Feb 2007 09:24:57 +0000 (09:24 +0000)]
Branch:HEAD
typo fix about join_lsm unpack, caused by 11471 landing

17 years agob=11267
nathan [Sat, 24 Feb 2007 23:45:53 +0000 (23:45 +0000)]
b=11267
add echo server starter
fix pthread stderr probs

17 years agofix typo.
shadow [Fri, 23 Feb 2007 07:57:17 +0000 (07:57 +0000)]
fix typo.

17 years agofix run test large amount free memory.
shadow [Thu, 22 Feb 2007 18:34:56 +0000 (18:34 +0000)]
fix run test large amount free memory.
r=adilger

17 years agoremove ext3-wantedi-misc patch.
shadow [Thu, 22 Feb 2007 17:53:17 +0000 (17:53 +0000)]
remove ext3-wantedi-misc patch.
struct dentry_params renamed to lvfs_dentry_params and moved into
lvfs_linux.h

b=11248
r=adilger

17 years agotypo
nathan [Thu, 22 Feb 2007 16:04:26 +0000 (16:04 +0000)]
typo

17 years ago1.6b8
nathan [Wed, 21 Feb 2007 22:17:47 +0000 (22:17 +0000)]
1.6b8

17 years agob=2258
nathan [Wed, 21 Feb 2007 21:25:50 +0000 (21:25 +0000)]
b=2258
r=adilger
Insure min dynamic threads < max threads
Also make absolutely sure we don't start more than MAX
on the biggest machines, we limit the MDS auto-threads to start at 32 and go to 128, nowhere near the 512 max limit.  On e.g.
ilter.c lustre/ost/ost_handler.c
on the biggest machines, we limit the MDS auto-threads to start at 32 and go to 128, nowhere near the 512 max limit.  On e.g.
    (8 * 1024)
on the biggest machines, we limit the MDS auto-threads to start at 32 and go to 128, nowhere near the 512 max limit.  On e.g.
s/lustre-core/mds/handler.c,v
on the biggest machines, we limit the MDS auto-threads to start at 32 and go to 128, nowhere near the 512 max limit.  On e.g.
threads = MDS_THREADS_AUTO_MAX;
on the biggest machines, we limit the MDS auto-threads to start at 32 and go to 128, nowhere near the 512 max limit.  On e.g.
andler.c 21 Feb 2007 21:22:08 -0000
on the biggest machines, we limit the MDS auto-threads to start at 32 and go to 128, nowhere near the 512 max limit.  On e.g.

17 years agoset CONFIG_LOCALVERISON="" to solve lbuild failures
zam [Wed, 21 Feb 2007 12:49:24 +0000 (12:49 +0000)]
set CONFIG_LOCALVERISON="" to solve lbuild failures

17 years agob=11692
nathan [Tue, 20 Feb 2007 22:04:00 +0000 (22:04 +0000)]
b=11692
play nicer with parameters on logs upgraded from 1.4.x
- allows updates for old logs whenever we can (i.e. ok for mdc, not for osc)
- allows updates for everyone if --writeconf is specified at tunefs upgrade
time (requires entire fs to be stopped though)
- allows updates of failover partners if --writeconf is specified
- fixes various other little bugs and error messages
- adds a couple of regression tests for upgrading from 1.4.6

17 years agomissing linux kernel config file for the SLES10 kernel,
zam [Tue, 20 Feb 2007 20:48:45 +0000 (20:48 +0000)]
missing linux kernel config file for the SLES10 kernel,
the config borrowed from b_release_1_4_9

17 years agob=11461
zam [Tue, 20 Feb 2007 19:22:09 +0000 (19:22 +0000)]
b=11461
r=scjody,vitaly
a fix for extra suffixes for kernel version string which
confuse lbuild.

17 years agob=10719
nathan [Tue, 20 Feb 2007 17:18:36 +0000 (17:18 +0000)]
b=10719
r=adilger
set external journal device read-only

17 years agofail test 11 after 30 min
nathan [Tue, 20 Feb 2007 16:23:07 +0000 (16:23 +0000)]
fail test 11 after 30 min

17 years agowait a little after turning off fail_loc to make sure files show up
nathan [Tue, 20 Feb 2007 16:20:53 +0000 (16:20 +0000)]
wait a little after turning off fail_loc to make sure files show up

17 years agob=3244
kalpak [Mon, 19 Feb 2007 13:48:36 +0000 (13:48 +0000)]
b=3244
r=Andreas Dilger

Description: Addition of EXT3_FEATURE_RO_COMPAT_DIR_NLINKS flag for
             > 32000 subdirectories

Details    : Add EXT3_FEATURE_RO_COMPAT_DIR_NLINK flag to
             EXT3_FEATURE_RO_COMPAT_SUPP. This flag will be set whenever
             subdirectory count crosses 32000. This will aid e2fsck to
             correctly handle more than 32000 subdirectories.

17 years agoquota test 11 takes forever on my machine when run as part of acceptance-small.
nathan [Sun, 18 Feb 2007 22:26:46 +0000 (22:26 +0000)]
quota test 11 takes forever on my machine when run as part of acceptance-small.
- Add some progress indication.
- Fix "cycles" (rm -f erases test dir, so must recreate)

17 years agoBranch HEAD
adilger [Sat, 17 Feb 2007 18:06:19 +0000 (18:06 +0000)]
Branch HEAD
Remove trailing whitespace.

17 years agoBranch HEAD
adilger [Sat, 17 Feb 2007 09:14:44 +0000 (09:14 +0000)]
Branch HEAD
Use lprocfs_rd_atomic() because ns_locks is atomic in HEAD.  Otherwise we get
bogus values from /proc/fs/lustre/ldlm/namespaces/*/lock_count.

17 years agoA __u64 is not always a long long. On some architectures it is only a
yujian [Sat, 17 Feb 2007 07:52:44 +0000 (07:52 +0000)]
A __u64 is not always a long long.  On some architectures it is only a
long.  Let's declare write_bytes as unsigned long long.

17 years agoBranch HEAD
adilger [Sat, 17 Feb 2007 07:22:32 +0000 (07:22 +0000)]
Branch HEAD
We don't support LUSTRE_KERNEL_VERSION < 45 anymore.

17 years agoRemove the 4G limitation of maximum writing bytes.
yujian [Sat, 17 Feb 2007 04:05:14 +0000 (04:05 +0000)]
Remove the 4G limitation of maximum writing bytes.

17 years agofix a typo
tianzy [Sat, 17 Feb 2007 01:54:57 +0000 (01:54 +0000)]
fix a typo
b=11471

17 years agob=11691
nathan [Fri, 16 Feb 2007 23:14:06 +0000 (23:14 +0000)]
b=11691
r=adilger
Don't print regular error messages about the MGS when we're connecting to
an old MDT.

17 years agoqos test: don't fail if we write 0 bytes to smaller ost
nathan [Fri, 16 Feb 2007 22:12:39 +0000 (22:12 +0000)]
qos test: don't fail if we write 0 bytes to smaller ost

17 years agob=11667
nathan [Fri, 16 Feb 2007 20:54:26 +0000 (20:54 +0000)]
b=11667
missing symbol

17 years ago* landing 11667 (debug_peer_on_timeout) and 11684 (liblustre block for asynch
eeb [Fri, 16 Feb 2007 18:01:23 +0000 (18:01 +0000)]
*   landing 11667 (debug_peer_on_timeout) and 11684 (liblustre block for asynch
    OST_CONNECTs to complete at startup) on HEAD

17 years agofixes for running in more user environments
nathan [Fri, 16 Feb 2007 17:17:40 +0000 (17:17 +0000)]
fixes for running in more user environments

17 years agoSeverity : enhancement
eeb [Fri, 16 Feb 2007 15:44:57 +0000 (15:44 +0000)]
Severity   : enhancement
Bugzilla   : 11684
Description: Added D_NETTRACE for recording network packet history
     (initially only for ptllnd).  Also a separate userspace
     ptllnd facility to gather history which should really be
     covered by D_NETTRACE too, if only CDEBUG recorded history in
     userspace.

17 years agoBranch HEAD
adilger [Fri, 16 Feb 2007 10:37:35 +0000 (10:37 +0000)]
Branch HEAD
Avoid synchronous disconnects for liblustre clients, already in Cray
upstream.
b=10803
r=green

17 years agoBranch HEAD
adilger [Fri, 16 Feb 2007 08:44:17 +0000 (08:44 +0000)]
Branch HEAD
Description: large O_DIRECT requests fail under memory pressure/fragmentation
Details    : Large single O_DIRECT read and write calls can fail to allocate
     a sufficiently large buffer to process the request.  In case of
     allocation failure the allocation is retried with a smaller
     buffer and broken into smaller requests.
b=10901
r=vitaly r=green

17 years agoBranch HEAD
adilger [Fri, 16 Feb 2007 08:29:36 +0000 (08:29 +0000)]
Branch HEAD
Remove warning about non-portable initializer.

17 years agoBranch HEAD
adilger [Fri, 16 Feb 2007 08:27:03 +0000 (08:27 +0000)]
Branch HEAD
Reorder include files so that the cookie check doesn't always fail.

17 years agofix the warnings
tianzy [Fri, 16 Feb 2007 07:00:50 +0000 (07:00 +0000)]
fix the warnings
b=11471

17 years agoadding 2.6-sles10 target and kernel config files for the sles10 kernel
zam [Thu, 15 Feb 2007 20:45:13 +0000 (20:45 +0000)]
adding 2.6-sles10 target and kernel config files for the sles10 kernel
(borrowed from b1_4).  the tcp-zerocopy-2.6-sles10.patch patch removed for
fixing an lbuild failure and because lnet does not need that patch anymore.

b=10947

17 years agob=6221
nathan [Thu, 15 Feb 2007 20:09:41 +0000 (20:09 +0000)]
b=6221
r=green
don't allocate inside a spinlock

17 years agoremove use old rd only API.
shadow [Thu, 15 Feb 2007 18:31:33 +0000 (18:31 +0000)]
remove use old rd only API.
b=11248

17 years agoRemove old rdonly API which unsed from at least lustre 1.4.3
shadow [Thu, 15 Feb 2007 12:38:18 +0000 (12:38 +0000)]
Remove old rdonly API which unsed from at least lustre 1.4.3
b=11248
r=adilger

17 years agoport the patch(9414) on HEAD(b1_5)
tianzy [Thu, 15 Feb 2007 07:11:12 +0000 (07:11 +0000)]
port the patch(9414) on HEAD(b1_5)
b=11471

17 years agoupdate rhel4 kernel to 2.6.9-42.0.8.EL
wufei [Thu, 15 Feb 2007 06:29:25 +0000 (06:29 +0000)]
update rhel4 kernel to 2.6.9-42.0.8.EL

17 years agoBranch HEAD
bobijam [Thu, 15 Feb 2007 00:23:59 +0000 (00:23 +0000)]
Branch HEAD
b=9829
r=adilger

Description: client incorrectly hits assertion in ptlrpc_replay_req()
Details    : for a short time RPCs with bulk IO are in the replay list,
       but replay of bulk IOs is unimplemented.  If the OST filesystem
       is corrupted due to disk cache incoherency and then replay is
       started it is possible to trip an assertion.  Avoid putting
       committed RPCs into the replay list at all to avoid this issue.

17 years agoBranch HEAD
adilger [Thu, 15 Feb 2007 00:23:02 +0000 (00:23 +0000)]
Branch HEAD
Some minor cleanups to make extents, mballoc kernel patches the same.
Use ext3_fsblk_t as appropriate for 2.6.18.
b=11675

17 years agoAdded supported kernel list
nathan [Wed, 14 Feb 2007 21:11:41 +0000 (21:11 +0000)]
Added supported kernel list

17 years agoBranch HEAD
adilger [Wed, 14 Feb 2007 20:34:02 +0000 (20:34 +0000)]
Branch HEAD
Add ChangeLog entry
b=11669

17 years agoBranch HEAD
adilger [Wed, 14 Feb 2007 04:56:39 +0000 (04:56 +0000)]
Branch HEAD
Fix 2.6.18 ext3 mballoc, extents patches for several omissions.
Reorder patch hunks to match other patches to facilitate comparisons.
b=10090  ldiskfs corruption under memory pressure, coverity fixes
b=6191   mballoc fails on x86_64 > 2TB
b=10634  mballoc using wrong find_next_bit() on big endian systems
b=6449   don't print "too long searching" message to console

Compile tested.

17 years agoBranch HEAD
adilger [Wed, 14 Feb 2007 00:21:55 +0000 (00:21 +0000)]
Branch HEAD
Quiet -Wundef compiler warning.

17 years agoBranch HEAD
adilger [Wed, 14 Feb 2007 00:20:17 +0000 (00:20 +0000)]
Branch HEAD
Line wrap comments at 80 columns.

17 years agoBranch HEAD
adilger [Wed, 14 Feb 2007 00:19:00 +0000 (00:19 +0000)]
Branch HEAD
Better comment.

17 years agoBranch HEAD
adilger [Wed, 14 Feb 2007 00:17:09 +0000 (00:17 +0000)]
Branch HEAD
Clarify comment.

17 years agoBranch HEAD
adilger [Wed, 14 Feb 2007 00:16:34 +0000 (00:16 +0000)]
Branch HEAD
Fix compiler warning.

17 years agothread id starts counting at 1
nathan [Tue, 13 Feb 2007 21:03:49 +0000 (21:03 +0000)]
thread id starts counting at 1

17 years agoBranch HEAD
adilger [Tue, 13 Feb 2007 19:31:12 +0000 (19:31 +0000)]
Branch HEAD
Description: RPCs being resent when they shouldn't be
Details    : Some RPCs that should not be resent are being resent.  This
     can cause inconsistencies in the RPC state machine.  Do not
     resend such requests.
b=11203
r=nathan r=green

Fix 2.4 kernel compile error introduced by 10866 landing (missing typecheck).

17 years agoalias b1_5 and b1_6 to HEAD
nathan [Tue, 13 Feb 2007 18:31:58 +0000 (18:31 +0000)]
alias b1_5 and b1_6 to HEAD

17 years agoremove debug assert.
shadow [Tue, 13 Feb 2007 18:28:22 +0000 (18:28 +0000)]
remove debug assert.
b=11669
r=green,adilger