Whamcloud - gitweb
fs/lustre-release.git
11 years agob=19051
manoj [Thu, 30 Apr 2009 20:55:08 +0000 (20:55 +0000)]
b=19051
r=yong.fan, jinshan.xiong

Use 64bits (loff_t) for holding file size instead of size_t

11 years agoBranch HEAD
johann [Thu, 30 Apr 2009 08:58:01 +0000 (08:58 +0000)]
Branch HEAD
b=18577
i=shadow
i=hongchao

create_count always drops to the min value (=32) because grow_count
is being changed before the precreate RPC completes.

11 years agodon't leak transno for recovery.
shadow [Thu, 30 Apr 2009 03:21:59 +0000 (03:21 +0000)]
don't leak transno for recovery.
Branch HEAD
b=19085
i=rread
i=adilger

11 years agoBranch HEAD
fanyong [Thu, 30 Apr 2009 01:43:19 +0000 (01:43 +0000)]
Branch HEAD
b=19171
i=tianzy
i=robert.read

Hold obd reference count when quota recovery.

11 years agoBranch HEAD
yangsheng [Thu, 30 Apr 2009 01:30:39 +0000 (01:30 +0000)]
Branch HEAD
b=18790

i=robert, panda

Check the types for PPC64 platform.

11 years agoBranch HEAD
vitaly [Wed, 29 Apr 2009 21:58:51 +0000 (21:58 +0000)]
Branch HEAD
b=18851
i=tappro
i=vs

walk through abort_list in abort_lock_replay_queue()

11 years agob=16312
brian [Wed, 29 Apr 2009 18:06:28 +0000 (18:06 +0000)]
b=16312
i=wangyb
i=jack

Bug fixes to fix:
- always naming the lustre packages as if they were patchless clients
- detect a partial cache commit and back it out

Provide a more clear e-mail subject line in the case of an untrapped error.
Remove an irrelevent comment.

11 years agoBranch HEAD
vitaly [Wed, 29 Apr 2009 15:37:44 +0000 (15:37 +0000)]
Branch HEAD
b=18571
i=girish
i=adilger

do not take an extra reference on journal bh which prevents it to be released.

11 years agob=18733
zhanghc [Wed, 29 Apr 2009 10:15:24 +0000 (10:15 +0000)]
b=18733

add protection (spin_lock(&lock->l_lock)) before
using lock->l_resource in "lock_res_and_lock"

i=johann
i=green

11 years agobranch: HEAD
ericm [Wed, 29 Apr 2009 03:47:57 +0000 (03:47 +0000)]
branch: HEAD
fix replayed meta sequence handling.
b=18982
r=rread
r=fanyong

11 years agoBranch HEAD
bobijam [Wed, 29 Apr 2009 02:07:28 +0000 (02:07 +0000)]
Branch HEAD
b=18690
o=adilger
i=johann
i=zhenyu.xu (bobijam)

Increase some hash table sizes.

11 years agoBranch HEAD
bobijam [Wed, 29 Apr 2009 01:45:10 +0000 (01:45 +0000)]
Branch HEAD
b=18073
i=johann
i=adilger
i=sheng.yang

use "\\" as escape character to prevent '.' from being converted to '/' in
lctl get/set_param.

11 years agoBranch HEAD
tianzy [Wed, 29 Apr 2009 01:29:49 +0000 (01:29 +0000)]
Branch HEAD
fix the printing of rq_xid(change into LPU)
b=18280
i=johann
i=tianzy

11 years agoBranch HEAD
vitaly [Tue, 28 Apr 2009 17:31:19 +0000 (17:31 +0000)]
Branch HEAD
b=19154
i=zam
i=rread

properly protect filter_destroy() with a DLM lock

11 years agob=16839
jxiong [Tue, 28 Apr 2009 05:02:36 +0000 (05:02 +0000)]
b=16839
r=shadow,rread

The rpc in __mds_lov_synchronize() should be sync.

11 years agocorrectly skip time estimate if in recovery
shadow [Tue, 28 Apr 2009 04:27:08 +0000 (04:27 +0000)]
correctly skip time estimate if in recovery

Branch HEAD
b=19223
i=johann
i=nathan

11 years agoUpdated for the next 2.0 bi-weekly build.
stinson1947 [Tue, 28 Apr 2009 01:35:16 +0000 (01:35 +0000)]
Updated for the next 2.0 bi-weekly build.

11 years agoBranch HEAD
bobijam [Tue, 28 Apr 2009 01:22:24 +0000 (01:22 +0000)]
Branch HEAD
b=18683
i=joahnn
i=sheng.yang

If no options are given, tunefs.lustre/mkfs.lustre just print info.

11 years agoBranch HEAD
robert.read [Mon, 27 Apr 2009 21:49:25 +0000 (21:49 +0000)]
Branch HEAD

b=19104
r=wangdi,fanyong

we have to replace an entirely new environment for ptlrpc callbacks. Otherwise, in liblustre, the env running for IO would be used for callback.

Relanding Jay's patch.

11 years agoBranch HEAD
robert.read [Mon, 27 Apr 2009 21:45:04 +0000 (21:45 +0000)]
Branch HEAD
b=18818
i=manoj
i=rread
fix RMF_SETINFO_VAL type and verify all setinfo lengths

Relanding Nathan's patch.

11 years agoBranch HEAD
robert.read [Mon, 27 Apr 2009 21:43:36 +0000 (21:43 +0000)]
Branch HEAD

port lazystatfs to HEAD.

Branch HEAD
b=17974
i=rread

Re-landing shadow's patch.

11 years ago- back out peer_buffer_credits changes, as requested by rread.
isaac [Mon, 27 Apr 2009 18:56:09 +0000 (18:56 +0000)]
- back out peer_buffer_credits changes, as requested by rread.

11 years agob=19262
grev [Sat, 25 Apr 2009 09:40:28 +0000 (09:40 +0000)]
b=19262
i=Robert.Read
include functions.sh properly

11 years agoBranch HEAD
robert.read [Sat, 25 Apr 2009 01:42:31 +0000 (01:42 +0000)]
Branch HEAD

b=17974

Backing out change due to sanity failures.

11 years agoBranch HEAD
robert.read [Fri, 24 Apr 2009 23:42:25 +0000 (23:42 +0000)]
Branch HEAD
b=18818

Backing changes due to sanity failure.

11 years agoBranch HEAD
robert.read [Fri, 24 Apr 2009 23:26:11 +0000 (23:26 +0000)]
Branch HEAD
b=16312

Revert cause of build failure.

11 years agoBranch HEAD
robert.read [Fri, 24 Apr 2009 23:09:28 +0000 (23:09 +0000)]
Branch HEAD

b=19104

Revertting patches due to sanity failure.

11 years agob=16312
brian [Fri, 24 Apr 2009 20:12:13 +0000 (20:12 +0000)]
b=16312

Remedial cleanup of an lbuild feature that changed while this patch was
waiting to land.

11 years agob=16312
brian [Fri, 24 Apr 2009 18:33:17 +0000 (18:33 +0000)]
b=16312
i=robert.read
i=yibin.wang

Make lbuild build kernels based on the vendor's own SRPM rather than all of
the gyrations we used to go through to build the kernel tarballs and so on.

This should give people a kernel (and associated packages) that looks much
more like the kernel (and packages) they are used to getting from their
vendor.

This has resulted in a much easier to understand lbuild and once we finally
fully drop support for rhel4 and sles9 we can get rid of the
lustre-kernel-2.4.spec.in, build/lbuild.old_school and build/lmake files.
Anyone know knows anything about the build system will understand what kind
of simplicity in understanding of how the build works that will yield.

11 years ago*** empty log message ***
brian [Fri, 24 Apr 2009 17:59:56 +0000 (17:59 +0000)]
*** empty log message ***

11 years agoWith 1.6 heading towards maintenance mode, use a dedicated branch of the
brian [Fri, 24 Apr 2009 17:10:03 +0000 (17:10 +0000)]
With 1.6 heading towards maintenance mode, use a dedicated branch of the
build system for it to isolate future build changes away from it.

11 years agofix typo founded by maxim.
shadow [Fri, 24 Apr 2009 09:04:38 +0000 (09:04 +0000)]
fix typo founded by maxim.

Branch HEAD

11 years agoBranch HEAD
fanyong [Fri, 24 Apr 2009 02:03:43 +0000 (02:03 +0000)]
Branch HEAD
b=19149
i=tianzy
i=panda

Make sure quota has been released to master in test_6.

11 years agoBranch HEAD
robert.read [Fri, 24 Apr 2009 01:18:01 +0000 (01:18 +0000)]
Branch HEAD
b=13908
Also build the source package, for the same reason.

11 years agob=19104
jxiong [Fri, 24 Apr 2009 01:05:55 +0000 (01:05 +0000)]
b=19104
r=wangdi,fanyong

we have to replace an entirely new environment for ptlrpc callbacks. Otherwise, in liblustre, the env running for IO would be used for callback.

11 years agoBranch HEAD
robert.read [Fri, 24 Apr 2009 00:57:07 +0000 (00:57 +0000)]
Branch HEAD

Whitespace cleanup and silence persistent error message.

11 years agoBranch HEAD
robert.read [Fri, 24 Apr 2009 00:41:59 +0000 (00:41 +0000)]
Branch HEAD
b=13908

Re-enable building lustre-client-tests, so oslo can find tests for all clients,
especially clients we don't build  a server for like ia64.

Tested with "make rpms"

11 years agoBranch HEAD
robert.read [Fri, 24 Apr 2009 00:11:26 +0000 (00:11 +0000)]
Branch HEAD
b=19190
i=nathan

Wait for recovery to complete before finishing test.

11 years agoBranch HEAD
robert.read [Thu, 23 Apr 2009 21:01:22 +0000 (21:01 +0000)]
Branch HEAD
b=19280,17974

Temporarily disable test 50e and 50f, which are broken in multi-node configurations.

11 years agob=19047
nathan [Thu, 23 Apr 2009 20:54:00 +0000 (20:54 +0000)]
b=19047
i=adilger
i=nathan
add llog_reader man page

11 years agob=18818
nathan [Thu, 23 Apr 2009 19:35:49 +0000 (19:35 +0000)]
b=18818
i=manoj
i=rread
fix RMF_SETINFO_VAL type and verify all setinfo lengths

11 years agoport lazystatfs to HEAD.
shadow [Thu, 23 Apr 2009 18:14:02 +0000 (18:14 +0000)]
port lazystatfs to HEAD.

Branch HEAD
b=17974
i=rread

11 years agoBranch HEAD
adilger [Thu, 23 Apr 2009 08:01:43 +0000 (08:01 +0000)]
Branch HEAD
Fix error message to match actual filenames.
Fix OBD_FAIL_TGT_REPLAY_DROP comment to have correct fail_loc value.
b=19114

11 years agob=15332,i=liang:
isaac [Thu, 23 Apr 2009 04:29:19 +0000 (04:29 +0000)]
b=15332,i=liang:
- add a new LND optiion to control peer buffer credits on routers.

11 years agoallocate inodes for llog files in last inodes group, for avoid broking
shadow [Wed, 22 Apr 2009 18:03:59 +0000 (18:03 +0000)]
allocate inodes for llog files in last inodes group, for avoid broking
recovery.

Branch HEAD
b=18192
i=adilger
i=shadow
i=girish

11 years agob=13908
brian [Wed, 22 Apr 2009 17:55:29 +0000 (17:55 +0000)]
b=13908
i=yibin.wang
i=sheng.yang

Allow the name, version, kernel release and release of the lustre packages
to be defined on the command line.
With such a feature, actually properly name the patchless client packages in
our own build.
Tighten up some dependencies.

11 years agob=18676
nathan [Wed, 22 Apr 2009 17:12:27 +0000 (17:12 +0000)]
b=18676
i=johann
i=manoj
replace obd ioctl with dir ioctl for changelogs

11 years agoBranch HEAD
vitaly [Wed, 22 Apr 2009 13:37:05 +0000 (13:37 +0000)]
Branch HEAD
b=15977
i=deen
i=vs

lctl dk should zero the file before writing the debug log into,
otherwise, if a file was long enough and is not completely overwritten
we get a mix of several logs.

11 years agoThis patch implements separate function for handling OPEN requests from Lustre client...
deen [Wed, 22 Apr 2009 12:08:34 +0000 (12:08 +0000)]
This patch implements separate function for handling OPEN requests from Lustre clients acting as
NFS servers, named mdt_open_anon_by_fid(). It searches for existing object and, if succeed, takes
an open lock on the object and its attributes. If a client didn't ask for the open lock, it releases
it at the end. We must always take an open lock at the beginning in order to protect object's
attributes.

b=16715
i=oleg.drokin
i=alexey.lyashkov

11 years agob=18689
zhanghc [Wed, 22 Apr 2009 02:07:14 +0000 (02:07 +0000)]
b=18689

change thread name "ll_log_commit_*" in
llog_recov_thread_init into "lcm_*" so it
can be fitted into task struct

i=adilger

11 years agob=16890
anserper [Wed, 22 Apr 2009 01:08:07 +0000 (01:08 +0000)]
b=16890

the missing part from a previous commit

11 years agob=18711
jxiong [Wed, 22 Apr 2009 00:45:23 +0000 (00:45 +0000)]
b=18711
r=jay, adilger

in lu_context_key_deregister
- to not print message inside a spinlock
- assert when the key has extra references

11 years agob=16890
anserper [Wed, 22 Apr 2009 00:26:47 +0000 (00:26 +0000)]
b=16890
i=Johann Lombardi
i=ZhiYong Tian

avoid several unnecessary memsets

11 years agob=19116
grev [Tue, 21 Apr 2009 19:03:40 +0000 (19:03 +0000)]
b=19116
i=Robert.Read
convert "-" to X to create client variable name

11 years agoBranch HEAD
yangsheng [Tue, 21 Apr 2009 14:42:04 +0000 (14:42 +0000)]
Branch HEAD
b=18810

i=bzzz, adilger

Use sb_bgl_lock to protect bitmaps.

Author: adilger

11 years agoBranch HEAD
tianzy [Tue, 21 Apr 2009 09:22:08 +0000 (09:22 +0000)]
Branch HEAD
1. delete the lc_watchdog_touch_ms()
2. changing to seconds instead of ms
3. delete lcw_time in lc_watchdog
b=18582
i=nathan
i=johann

11 years agob=19175
grev [Mon, 20 Apr 2009 21:11:47 +0000 (21:11 +0000)]
b=19175
i=Robert.Read
check_catastrophe fix

11 years agob=18951
anserper [Mon, 20 Apr 2009 21:02:04 +0000 (21:02 +0000)]
b=18951

11 years agob=19140
grev [Mon, 20 Apr 2009 18:49:35 +0000 (18:49 +0000)]
b=19140
i=Alexander.Zarochentsev
PTLDEBUG/SUBSYSTEM t-f set_param fix:
accept the string value

11 years agob=18951
anserper [Mon, 20 Apr 2009 18:29:09 +0000 (18:29 +0000)]
b=18951
i=Nathan Rutman
i=Vladimir Saveliev

fix the race between fflush and pause

11 years agoBranch HEAD
zam [Thu, 16 Apr 2009 07:54:44 +0000 (07:54 +0000)]
Branch HEAD
b=18988
i=Alexey.Lyashkov

new sanity test for for O_APPEND.

11 years agob=17931
deshmukh [Thu, 16 Apr 2009 05:42:38 +0000 (05:42 +0000)]
b=17931
i=fonyoung
i=grev

conf_sanity_33a test-case is fixed now.

11 years agob=18844
maxim [Wed, 15 Apr 2009 19:38:54 +0000 (19:38 +0000)]
b=18844
i=isaac
Landing a patch fixing deadlock in usocklnd on HEAD. The patch also includes:
 - minor code cleanup suggested by Isaac in comment #6 of bug #18844
 - trivial cleanup patch for acceptor.c (att #21983 of bug #14132)
 - minor one-line fix for handling EINTR error of poll(2) - inspected by Shadow.

11 years agob=18414
maxim [Wed, 15 Apr 2009 19:22:44 +0000 (19:22 +0000)]
b=18414
A few words to lnet/ChangeLog about 'running out of ports' fix.

11 years agoBranch HEAD
yangsheng [Wed, 15 Apr 2009 10:35:49 +0000 (10:35 +0000)]
Branch HEAD
b=18033

Fix the typo.

11 years agob=19123
zam [Wed, 15 Apr 2009 06:25:13 +0000 (06:25 +0000)]
b=19123
i=robert.read

test-framework.sh:get_mds_dir:
Add another sleep 1 before "lfs df -i" to be sure that statfs cache is
invalidated.

11 years agobranch: HEAD
ericm [Tue, 14 Apr 2009 20:36:46 +0000 (20:36 +0000)]
branch: HEAD
Fix null pointer reference of rq_export, which maybe null in case of gss.
b=18983
r=fanyong
r=wangdi

11 years agoBranch b1_8
adilger [Tue, 14 Apr 2009 07:02:57 +0000 (07:02 +0000)]
Branch b1_8
Use EXT_UNSET_BLOCK to avoid confusion with EXT_MAX_BLOCK on 16TB-1 block
filesystems.

This patch was committed to b_release_1_8_0, but was missed on ldiskfs HEAD
for b1_8.
b=16114
o=girish.shilamkar
i=adilger
i=kalpak.shah

11 years agoBranch b1_8
adilger [Tue, 14 Apr 2009 04:45:18 +0000 (04:45 +0000)]
Branch b1_8
Update ldiskfs build version to 3.0.9 for next release.

11 years agoBranch b1_8
adilger [Tue, 14 Apr 2009 04:40:46 +0000 (04:40 +0000)]
Branch b1_8
Description: fix racy locking of mballoc block bitmaps causing BUG
Details    : The locking of the mballoc buddy bitmap and the in-memory
     block bitmap was using two different spin locks in some
     cases.  This made it possible to incorrectly access the
     mballoc bitmap while another process was modifying it,
     causing a sanity assertion to fail.  While no on-disk corruption
     was reported, there was some risk of this happening.
b=18810
i=alex

11 years agoBranch HEAD
liangzhen [Mon, 13 Apr 2009 17:45:26 +0000 (17:45 +0000)]
Branch HEAD
b=13621, 15983
i=isaac
i=maxim

o2iblnd protocol version 2

11 years agoBranch HEAD
liangzhen [Mon, 13 Apr 2009 16:31:42 +0000 (16:31 +0000)]
Branch HEAD

Tag for performance test

11 years agoBranch HEAD
bobijam [Mon, 13 Apr 2009 06:26:21 +0000 (06:26 +0000)]
Branch HEAD
b=16417

sleep some time before getting 'lfs df' value.

11 years agob=18271
grev [Fri, 10 Apr 2009 21:07:05 +0000 (21:07 +0000)]
b=18271
i=Nathan.Rutman
writeconf fn fix

11 years agoUpdated version.
stinson1947 [Fri, 10 Apr 2009 20:26:29 +0000 (20:26 +0000)]
Updated version.

11 years agoUpdated version string for the next bi-weekly build. This is also a 2.0 Alpha candidate.
stinson1947 [Fri, 10 Apr 2009 20:25:20 +0000 (20:25 +0000)]
Updated version string for the next bi-weekly build. This is also a 2.0 Alpha candidate.

11 years agoBranch HEAD
vitaly [Fri, 10 Apr 2009 19:43:32 +0000 (19:43 +0000)]
Branch HEAD
b=18863
i=shadow
i=bzzz

lu cache shrinker bugfix, return the proper amount of cached objects.

11 years agoBranch HEAD
vitaly [Fri, 10 Apr 2009 19:40:39 +0000 (19:40 +0000)]
Branch HEAD
b=18792
i=jay
i=huanghua

1. remove the assert for inode reference in cl_inode_fini().
2. as ldlm_cli_cancel() takes a "long" lock, i.e. without a reference,
we should not drop this reference later. --enable-lu_ref fix.

11 years agoBranch HEAD
robert.read [Fri, 10 Apr 2009 19:18:37 +0000 (19:18 +0000)]
Branch HEAD
b=19025
i=jay

Upstream patch, this time with the patch.

11 years agoBranch HEAD
robert.read [Fri, 10 Apr 2009 19:01:16 +0000 (19:01 +0000)]
Branch HEAD
b=19025
i=jay
i=rread

Upstream patch to fix panic on i/o errors.

11 years agoBranch HEAD
robert.read [Fri, 10 Apr 2009 18:48:35 +0000 (18:48 +0000)]
Branch HEAD
b=18556
o=tappro
i=rread
i=nathan

Reoganize recovery timer managing.

11 years agoBranch HEAD
robert.read [Fri, 10 Apr 2009 18:36:42 +0000 (18:36 +0000)]
Branch HEAD
b=18978
o=zam
i=tappro
i=rread

Don't LBUG in CMD cmr_version_get/set stubs.

11 years agob=19139
grev [Fri, 10 Apr 2009 15:02:02 +0000 (15:02 +0000)]
b=19139
i=Jian.Yu
typo END_RUN_FILE name fix

11 years agob=19129
grev [Fri, 10 Apr 2009 14:45:28 +0000 (14:45 +0000)]
b=19129
i=Alexander.Zarochentsev
zconf_mount_clients remote command fix

11 years agob=18080
grev [Fri, 10 Apr 2009 14:40:25 +0000 (14:40 +0000)]
b=18080
i=Alexander.Zarochentsev
add sanityN test_35 to A_E list

11 years agob=19016
jxiong [Fri, 10 Apr 2009 04:04:16 +0000 (04:04 +0000)]
b=19016
r=wangdi,fan.yong

Fix an issue about enqueuing a lock in clio.

11 years agob=18293
jxiong [Fri, 10 Apr 2009 03:32:21 +0000 (03:32 +0000)]
b=18293
i=fan.yong,rread

Revise the handling of IO error in clio:
- don't set page error for sync and direct IO;
- when IO error occurs, take the page out from SoM queue since we don't resend
  the page right now.

11 years agobranch: HEAD
ericm [Thu, 9 Apr 2009 20:50:31 +0000 (20:50 +0000)]
branch: HEAD
OSD doing batch read of dir entries in compatiblility mode iterator.

b=17560
o=pravin
i=oleg,rahul,pravin,rread

11 years agobranch: HEAD
ericm [Thu, 9 Apr 2009 20:43:11 +0000 (20:43 +0000)]
branch: HEAD
add an assertion trying to catch the rarely happened case of possible
import reference over dropping.
b=18849

11 years agoBranch HEAD
vitaly [Thu, 9 Apr 2009 16:15:57 +0000 (16:15 +0000)]
Branch HEAD
b=17926
i=rread
i=shadow

limit the default journal size with 400MB

11 years agoBranch HEAD
bobijam [Thu, 9 Apr 2009 09:53:34 +0000 (09:53 +0000)]
Branch HEAD
b=16417
i=grev
i=alexey.lyashkov (shadow)

revise sanityN.sh test_36()
- need sleep some time before getting 'after' value from 'lfs df', because
  if 'multiop & kill' happens too fast, 'lfs df' sometimes only get old df
  value from cache w/o reissuing statfs RPC request.

11 years agodon't allow connect to already connected import
shadow [Thu, 9 Apr 2009 05:36:17 +0000 (05:36 +0000)]
don't allow connect to already connected import

Branch HEAD
b=16839
i=rread

11 years agob=17491
zhanghc [Wed, 8 Apr 2009 17:20:51 +0000 (17:20 +0000)]
b=17491

remove "wrapped times" stuff in original patch

i=nathan.rutman
i=robert.read

11 years ago- fix wrong flag check in ptlrpc_at_set_reply()
tappro [Wed, 8 Apr 2009 09:59:33 +0000 (09:59 +0000)]
- fix wrong flag check in ptlrpc_at_set_reply()
- remove staled clients before finish_recovery stage
- increase time for recovery if switched to vbr
b:18556
i:rread,zam,nathan

11 years agoBranch HEAD
deshmukh [Wed, 8 Apr 2009 05:28:39 +0000 (05:28 +0000)]
Branch HEAD
b=18882
i=oleg.drokin

11 years agoallow killing process which wait statahead result
shadow [Tue, 7 Apr 2009 15:41:06 +0000 (15:41 +0000)]
allow killing process which wait statahead result

Branch HEAD
b=18902
i=fanyong
i=tappro

11 years agob=19087,i=liang:
isaac [Tue, 7 Apr 2009 15:32:17 +0000 (15:32 +0000)]
b=19087,i=liang:
- removed unused field md_addrkey so that sizeof lnet_libmd_t with one frag would be
  exactly 128 on x86_64, very sweet.

11 years agob=18414
maxim [Tue, 7 Apr 2009 11:09:13 +0000 (11:09 +0000)]
b=18414
i=isaac
i=liang
Landing a patch fixing 'running out of ports' issue on HEAD. The patch
essentially does:
 - add a delay before next reconnect attempt in ksocklnd in the case of lost race;
 - limit the frequency of query-requests in lnet;
 - special handling of 'dead peer' notifications in lnet.

11 years agoBranch HEAD
zam [Tue, 7 Apr 2009 06:39:30 +0000 (06:39 +0000)]
Branch HEAD
b=18988
i=alexey.lyashkov
i=jinshan.xiong
i=adilger

If append write, vvp_io_write_start sets pos to the end of the file and
doesn't touch kiocb->ki_pos, it triggers BUG_ON(iocb->ki_pos != pos) in Linux
kernel code.  This patch adds missing ki_pos adjustment.

11 years agoBranch HEAD
robert.read [Tue, 7 Apr 2009 05:05:20 +0000 (05:05 +0000)]
Branch HEAD
b=18554
o=shadow,rread
i=zam
i=eric.mei

Full fix for 18554. Includes original patch, Shadow's new fixes, and a test.