Whamcloud - gitweb
shadow [Thu, 15 Nov 2007 17:32:53 +0000 (17:32 +0000)]
sync kernel configs for sles10 with b1_6.
add support for sles10 in 1.8.
b=12052
i=johann
i=nikita
shadow [Thu, 15 Nov 2007 11:24:07 +0000 (11:24 +0000)]
set debug level on all nodes.
set max debug log size.
b=14087
i=grev
bwzhou [Thu, 15 Nov 2007 07:33:33 +0000 (07:33 +0000)]
Branch HEAD
b=13380
change timeout to 330 for server umount.
kalpak [Wed, 14 Nov 2007 22:29:42 +0000 (22:29 +0000)]
i=adilger
Update recommended e2fsprogs version to 1.40.2-cfs4
adilger [Wed, 14 Nov 2007 20:53:55 +0000 (20:53 +0000)]
Branch b1_6
Due to reported negative performance impact of this patch, it has been removed
from the ldiskfs patch series. The patch is not required functionality, and
was only added for an increase in reliability in unlikely corruption cases.
b=13144
i=kalpak
grev [Wed, 14 Nov 2007 19:52:48 +0000 (19:52 +0000)]
b=13769
i=Adilger
remove chmod 700
anserper [Tue, 13 Nov 2007 23:07:54 +0000 (23:07 +0000)]
b=13377
i=yury
i=nikita
girish [Tue, 13 Nov 2007 17:36:53 +0000 (17:36 +0000)]
JBD: Fix NULL pointer bh->b_data on NUMA box with journal checksumming.
Current journal checksumming patch failed fsstress test on NUMA. The
bh->b_data passed to the crc32_be () function could be NULL pointer,
which caused kernel oops immediately when running fsstress with -o
journal_checksum. It is because the page is part of highmem on NUMA box.
We need to kmap the page before access the bh->b_data to calculate
the checksums.
ericm [Tue, 13 Nov 2007 16:33:31 +0000 (16:33 +0000)]
b_colibri_devel use b1_8_iam of ldiskfs.
kalpak [Tue, 13 Nov 2007 14:51:36 +0000 (14:51 +0000)]
b=13858
i=fanyong
key may not be NULL terminated on HEAD, so change KEY_IS() macro to use strncmp() instead of strcmp.
grev [Tue, 13 Nov 2007 03:02:24 +0000 (03:02 +0000)]
minor message cleanup
tappro [Mon, 12 Nov 2007 19:26:07 +0000 (19:26 +0000)]
- exclude test33 due to bug 13710
tappro [Mon, 12 Nov 2007 19:14:55 +0000 (19:14 +0000)]
- check -ENODEV too
b:13223
i:shadow,fanyong
maxim [Mon, 12 Nov 2007 13:10:47 +0000 (13:10 +0000)]
b=14127
i=isaac
i=liang
Minor enhancement allowing lstclient to accept incoming connections.
yangsheng [Mon, 12 Nov 2007 07:30:37 +0000 (07:30 +0000)]
Branch HEAD
b=13969
i=scjody
i=johann
Update to RHEL5 kernel 2.6.18-8.1.15.el5.
yangsheng [Mon, 12 Nov 2007 07:14:09 +0000 (07:14 +0000)]
Branch HEAD
b=13874
i=scjody
i=johann
Update to SLES10 SP1 kernel 2.6.16.53-0.16.
yangsheng [Mon, 12 Nov 2007 03:57:04 +0000 (03:57 +0000)]
Branch HEAD
b=13889
i=scjody
i=johann
Update to SLES9 kernel-2.6.5-7.287.3.
kalpak [Sun, 11 Nov 2007 17:34:10 +0000 (17:34 +0000)]
b=13858
i=adilger,johann
cleanup in *_get_info methods to use the KEY_IS() macro and to abstract the datatypes for key and reply on the OST.
nathan [Fri, 9 Nov 2007 18:47:59 +0000 (18:47 +0000)]
b=13709
since test 22 and 23 are excepted, test 24 must deal with the config from
test 21c
nathan [Thu, 8 Nov 2007 23:10:38 +0000 (23:10 +0000)]
b=13709
skipping test 22 means we don't reformat after test 21c
alex [Thu, 8 Nov 2007 15:13:09 +0000 (15:13 +0000)]
b=12797
- correct patch for sles10
alex [Thu, 8 Nov 2007 10:56:57 +0000 (10:56 +0000)]
b=12797
i=adilger
i=eeb
new data sd_iostats collects are:
* time requests spend in elevator's queue
* time requests are served by disk
* seek stats
* queue depth stats
shadow [Thu, 8 Nov 2007 08:48:15 +0000 (08:48 +0000)]
fix test as Andreas comment.
b=12052
tappro [Wed, 7 Nov 2007 22:40:29 +0000 (22:40 +0000)]
- the merge issue fix missed in commit for 13818.
b:13818
grev [Wed, 7 Nov 2007 22:30:36 +0000 (22:30 +0000)]
-- fixed typo
grev [Wed, 7 Nov 2007 18:48:06 +0000 (18:48 +0000)]
b=12900
i=alexey.lyashkov
i=nathan.rutman
ignore test_30 failure
grev [Wed, 7 Nov 2007 18:31:50 +0000 (18:31 +0000)]
b=12649
i=Adilger
t-f helper nodes_list(); add debug marker on all nodes instead of client only
b=13698
i=Tappro
i=Adilger
disable LFSCK test
yangsheng [Wed, 7 Nov 2007 10:45:37 +0000 (10:45 +0000)]
Branch HEAD
b=14041
i=scjody
i=johann
Update Lustre to RHEL4 latest kernel.
grev [Wed, 7 Nov 2007 01:17:15 +0000 (01:17 +0000)]
b=13798
i=Nathan
remove tdir creation from sanity subtests
grev [Tue, 6 Nov 2007 20:08:10 +0000 (20:08 +0000)]
b=13798
i=Nathan
create tdir with correct permissions
shadow [Tue, 6 Nov 2007 19:06:10 +0000 (19:06 +0000)]
support for sles10 kernel in lustre.
b=12052
i=johann
i=nikita
yury [Tue, 6 Nov 2007 13:17:50 +0000 (13:17 +0000)]
b=13372
r=adilger,nathan
- enable/disable lru resize via proc.
tappro [Tue, 6 Nov 2007 13:03:56 +0000 (13:03 +0000)]
- hash-based export handling
b:13815
i:green, shadow
bwzhou [Tue, 6 Nov 2007 10:02:59 +0000 (10:02 +0000)]
Branch HEAD
b=13933
i=adilger
i=johann
Set an upper limit for stripe count
yury [Mon, 5 Nov 2007 16:15:04 +0000 (16:15 +0000)]
b=13872
r=adilger,nathan
- lru resize and ldlm cleanups and fixes.
bwzhou [Mon, 5 Nov 2007 15:45:48 +0000 (15:45 +0000)]
Branch HEAD
b=11622
i=adilger
i=johann
1. add page allocation statistics for lustre;
2. use percpu counters for both memory and page accounting;
3. update leak-finder to understand the new format of debug info.
yury [Mon, 5 Nov 2007 13:37:21 +0000 (13:37 +0000)]
b=13854
r=adilger
- fixed LVF calculation in test_124a.
grev [Fri, 2 Nov 2007 22:32:53 +0000 (22:32 +0000)]
b=12845
i=Nathan
t-f fn wait_exit_ST()
b=13798
i=Adilger
Use TESTSUITE in tdir
b=12599
i=Nathan
lost fix
nathan [Fri, 2 Nov 2007 20:18:39 +0000 (20:18 +0000)]
b=13709
always except test 22 until 13709 is resolved
yangsheng [Fri, 2 Nov 2007 10:20:24 +0000 (10:20 +0000)]
Branch HEAD
b=13568
i=adilger
i=green
Remove 2.4 kernel code from include files.
yangsheng [Fri, 2 Nov 2007 10:03:55 +0000 (10:03 +0000)]
Branch HEAD
b=13568
i=shadow
i=johann
Remove lock_24kernel from lustre/lvfs/fsfilt_ext3.c.
yangsheng [Fri, 2 Nov 2007 09:57:35 +0000 (09:57 +0000)]
Branch HEAD
b=13568
i=nathan
i=green
Remove 2.4 kernel check from lustre/lmv/lproc_lmv.c.
yangsheng [Fri, 2 Nov 2007 09:54:22 +0000 (09:54 +0000)]
Branch HEAD
b=13568
i=green
i=adilger
Remove 2.4 kernel check from lustre/lov/lproc_lov.c.
yangsheng [Fri, 2 Nov 2007 07:49:33 +0000 (07:49 +0000)]
Branch HEAD
b=13568
i=shadow
i=adilger
Remove 2.4 kernel code from lustre/mdc/mdc_request.c.
yangsheng [Fri, 2 Nov 2007 07:45:30 +0000 (07:45 +0000)]
Branch HEAD
b=13568
i=johann
i=green
Remove 2.4 kernel check from lustre/mdt/mdt_lproc.c.
yangsheng [Fri, 2 Nov 2007 07:41:40 +0000 (07:41 +0000)]
Branch HEAD
b=13568
i=nathan
i=scjody
Remove 2.4 kernel code from lustre/mgc/mgc_request.c.
yangsheng [Fri, 2 Nov 2007 06:58:30 +0000 (06:58 +0000)]
Branch HEAD
b=13568
i=adilger
i=johann
Remove the 2.4 kernel check from mgs subsystem.
yangsheng [Fri, 2 Nov 2007 05:48:14 +0000 (05:48 +0000)]
Branch HEAD
b=13568
i=adilger
i=scjody
Remove 2.4 kernel code from lustre/obdecho/echo_client.c.
yangsheng [Fri, 2 Nov 2007 04:21:37 +0000 (04:21 +0000)]
Branch HEAD
b=12568
i=adilger
i=scjody
Remove 2.4 kernel check from obdfilter/filter.c, obdfilter/lproc_obdfilter.c.
yangsheng [Fri, 2 Nov 2007 04:14:51 +0000 (04:14 +0000)]
Branch HEAD
b=13568
i=adilger
i=johann
Remove the 2.4 kernel check from lproc_osc.c,osc_request.c in osc subsystem.
scjody [Thu, 1 Nov 2007 13:47:08 +0000 (13:47 +0000)]
Branch HEAD
Also remove Module.symvers (used on some SLES releases) when removing
Modules.symvers.
b=12842
i=brian
i=mjmac
yangsheng [Thu, 1 Nov 2007 09:26:28 +0000 (09:26 +0000)]
Branch HEAD
b=13690
i=brian
i=scjody
Use absolute path for include file when run configure with argument
--with-linux-obj=....
yangsheng [Thu, 1 Nov 2007 09:13:20 +0000 (09:13 +0000)]
Branch b1_6
b=13690
i=brian
i=scjody
Use absolute path for include file when run configure with argument
--with-linux-obj=....
liuy [Thu, 1 Nov 2007 02:59:57 +0000 (02:59 +0000)]
b=13884
i=adilger, deen
change the error message in llu_file_prwv()
nathan [Thu, 1 Nov 2007 01:55:50 +0000 (01:55 +0000)]
b=12860
i=adilger
i=green
simultaneous MDT->OST connections at startup can cause the
sync to abort, leaving the OSC in a bad state
tappro [Wed, 31 Oct 2007 20:12:52 +0000 (20:12 +0000)]
- missed file in MDT_FAIL fix commit
tappro [Wed, 31 Oct 2007 19:03:13 +0000 (19:03 +0000)]
- replace MDT_FAIL_* with OBD_FAIL_*
b:14001
i:nikita, yury
grev [Wed, 31 Oct 2007 00:17:36 +0000 (00:17 +0000)]
b=13922
i=Nathan
skip 32a, 32b tests if MDS is remote
johann [Tue, 30 Oct 2007 21:29:15 +0000 (21:29 +0000)]
Branch HEAD
b=14007
i=nathan
i=tappro
check for MDS_OPEN_OWNEROVERRIDE in attr_pack()/attr_unpack()
since it is used in mds_fix_attr() (see bug 5781) for NFS.
grev [Tue, 30 Oct 2007 19:01:31 +0000 (19:01 +0000)]
b=13990
i=Adilger
i=Nathan
force 24a, 24b, 33, 36 tests to work properly with remote mds
grev [Mon, 29 Oct 2007 21:33:43 +0000 (21:33 +0000)]
b=13534
i=Nathan
i=Huanghua
test_50 fix for remote mds
isaac [Mon, 29 Oct 2007 18:15:47 +0000 (18:15 +0000)]
b=12494,i=eeb:
- increase send queue size for ciblnd/openiblnd.
grev [Mon, 29 Oct 2007 16:25:24 +0000 (16:25 +0000)]
b=13774
i=Adilger
test_54e fix
johann [Mon, 29 Oct 2007 16:08:06 +0000 (16:08 +0000)]
Branch HEAD
b=13930
i=johann,adilger
ll_inode_fill_obdo() expects OBD_BRW_WRITE or OBD_BRW_READ as rw argument
rather than READ/WRITE.
brian [Fri, 26 Oct 2007 22:15:50 +0000 (22:15 +0000)]
b=13639
i=yangsheng,mjmac
More correctly set the EXTRAVERSION in the sles10 kernel-lustre-source.
brian [Fri, 26 Oct 2007 20:52:07 +0000 (20:52 +0000)]
b=13639
i=yangsheng,mjmac
More correctly set the EXTRAVERSION in the sles10 kernel-lustre-source.
yury [Fri, 26 Oct 2007 10:53:21 +0000 (10:53 +0000)]
- fixed typo in test_124a
johann [Thu, 25 Oct 2007 19:48:30 +0000 (19:48 +0000)]
Branch HEAD
b=13828
i=adilger,shadow
attachment #13289:
Don't use kernel's ATTR_* attributes over the network since the client and
MDS may run different kernels. Instead, use MDS_ATTR_* attributes (defined
in lustre_idl.h) for sa_valid.
attachment #13342:
- use a higher value for ATTR_CTIME_SET
- define MDS_ATTR_* in hex
- improve attr_unpack/attr_pack/mds_pack_open_flags
shadow [Thu, 25 Oct 2007 17:37:23 +0000 (17:37 +0000)]
interrupt oig_wait can produce painc on resend.
b=13888
i=nikita
i=alex
shadow [Thu, 25 Oct 2007 11:32:27 +0000 (11:32 +0000)]
fix tests 118 to work with remote ost.
b=12858
i=green
i=adilger
vitaly [Thu, 25 Oct 2007 10:55:29 +0000 (10:55 +0000)]
Branch HEAD
b=13622
i=tappro
i=green
1) make class_handle_hash_back() to work;
2) ll_update_inode() checks UPDATE ibit lock still presents before setting LLIF_MDS_SIZE_LOCK;
3) ldlm_lock_match() returns the matched mode instead of 1; pass bitwise sets of modes into;
komaln [Thu, 25 Oct 2007 08:40:04 +0000 (08:40 +0000)]
b=11230
r=Adilger, Nathan
Patch to handle symlinks while setting tunables.
huanghua [Thu, 25 Oct 2007 08:03:39 +0000 (08:03 +0000)]
file mdc_fid.c was initially added on branch b1_6_cli_reqs.
bobijam [Thu, 25 Oct 2007 05:59:56 +0000 (05:59 +0000)]
Branch HEAD
b=13564
i=johann
i=nathan
error handling cleanup.
bobijam [Thu, 25 Oct 2007 04:37:38 +0000 (04:37 +0000)]
Branch HEAD
b=12584
i=johann
i=nathan
add sub-routine declaration before use it.
bobijam [Thu, 25 Oct 2007 03:03:57 +0000 (03:03 +0000)]
Branch HEAD
b=13497
i=deen
i=johann
Description: LASSERT_{REQ,REP}SWAB macros are buggy
Details : If SWAB_PARANOIA is disabled, the LASSERT_REQSWAB and
LASSERT_REPSWAB macros become no-ops, which is incorrect. Drop
these macros and replace them with their difinitions instead.
grev [Wed, 24 Oct 2007 19:27:45 +0000 (19:27 +0000)]
b=12649
i=Adilger
force runtests to use t-f log()
grev [Wed, 24 Oct 2007 19:20:52 +0000 (19:20 +0000)]
b=13798
i=Nathan
i=Tianzy
remove run test_0 from test_14 to pre_test_14
vitaly [Wed, 24 Oct 2007 16:59:05 +0000 (16:59 +0000)]
Branch HEAD
b=12702
rallback the patch as it breaks runtests
anserper [Wed, 24 Oct 2007 09:19:04 +0000 (09:19 +0000)]
reverting back accidental lustrecvs update
anserper [Wed, 24 Oct 2007 09:14:10 +0000 (09:14 +0000)]
b=10600
small fix - inline specifier is stripped from quote_get_qunit_data_size definition and declaration
tappro [Tue, 23 Oct 2007 21:20:15 +0000 (21:20 +0000)]
- lost #define changes from 1_6
b:13816
i:shadow,yury
liangzhen [Tue, 23 Oct 2007 07:23:22 +0000 (07:23 +0000)]
Branch HEAD
Support both ofed 1.2.0 and 1.2.5
b=13378
i=isaac
tappro [Tue, 23 Oct 2007 06:35:14 +0000 (06:35 +0000)]
- lost fixes from 1_6
b:13818
i:adilger, alex
jxiong [Tue, 23 Oct 2007 05:40:47 +0000 (05:40 +0000)]
Add test_68(using lustre file as swap space) into slow test set because it always takes a longer to finish.
grev [Mon, 22 Oct 2007 23:04:24 +0000 (23:04 +0000)]
b=13572
i=Adilger
t-f wait_delete_complete() function; force test_79 to wait;
quiet test_65l output.
nikita [Mon, 22 Oct 2007 15:44:31 +0000 (15:44 +0000)]
libcfs: add cpu<->be conversion functions to the client.
liuy [Mon, 22 Oct 2007 09:17:00 +0000 (09:17 +0000)]
Branch HEAD
b=13556
i=scjody
change mgsnode
yangsheng [Mon, 22 Oct 2007 02:42:40 +0000 (02:42 +0000)]
Branch HEAD
b=13521
i=scjody
i=nathan
Update which_patch & target file for SLES10 SP1 kernel-2.6.16.53-0.8.
scjody [Fri, 19 Oct 2007 20:23:56 +0000 (20:23 +0000)]
Branch b1_6
Update version
scjody [Fri, 19 Oct 2007 20:08:24 +0000 (20:08 +0000)]
Branch b1_6
Update release dates
yury [Fri, 19 Oct 2007 09:54:48 +0000 (09:54 +0000)]
b=13854
r=adilger,nathan
- fixed LVF calculation in sanity test 124a.
grev [Thu, 18 Oct 2007 18:15:53 +0000 (18:15 +0000)]
b=13829
i=Adilger
skip test_125 if mds acl disabled
nathan [Thu, 18 Oct 2007 17:42:47 +0000 (17:42 +0000)]
b=5494
i=adilger
add error_ignore method to run test but ignore failures
johann [Thu, 18 Oct 2007 14:57:29 +0000 (14:57 +0000)]
Branch b1_6
b=13828
i=adilger,shadow
Don't use kernel's ATTR_* attributes over the network since the client and MDS
may run different kernels. Instead, use MDS_ATTR_* attributes (defined in
lustre_idl.h) for sa_valid.
bobijam [Thu, 18 Oct 2007 11:21:55 +0000 (11:21 +0000)]
Branch HEAD
b=13626
i=johann, adilger
use k(un)map_atomic instead of k(un)map for short/tight code paths.
grev [Thu, 18 Oct 2007 08:28:18 +0000 (08:28 +0000)]
b=12560
lost test_7 fix
bobijam [Thu, 18 Oct 2007 08:22:55 +0000 (08:22 +0000)]
Branch HEAD
b=13626
reverse the patch since been reported broken test on box with 2G ram.
alex [Thu, 18 Oct 2007 06:55:50 +0000 (06:55 +0000)]
b=12797
- revert the patch landed by mistake
alex [Thu, 18 Oct 2007 06:53:00 +0000 (06:53 +0000)]
b=12797
i=adilger
i=eeb
- improved sd iostats:
* time requests spend in elevator's queue
* time requests are served by disk
* seek stats
* queue depth stats