b=15575
i=wangdi, johann

Description: Stack overflow during MDS log replay
Details    : ease stack pressure by using a thread dealing llog_process.

b=15278
i=johann
i=deen

	Added Changelog entry.

r=Nathan Rutman, Alexey Lyashkov

Ported the patch 'to organize the stats on per-nid basis instead of the existing per-export basis' to HEAD.
This patch also includes all its related bug fixes(bug 14016, bug 14370, bug 14873).

Add ChangeLog entries

b=14995
i=wangdi, rread

send activate event only for evicted imports.

i=johann
i=adilger

fix for configure: WARNING: linux/random.h: present but cannot be compiled

i=johann
i=alexey.lyashkov

fix error building lustre/liblustre/tests for Cray XT3 catamount

ChangeLog entries.

i=Jian.Yu
CLIENTMODSONLY part for HEAD

fix typo.

b=14803
i=johann, adilger

Description: Don't update lov_desc members until making sure they are valid
Details    : When updating lov_desc members via proc fs, need fix their
             validities before doing the real update.

i=adilger
i=rread

fixes error linking lustre/liblustre/tests/echo_test during catamount
build of b1_6 for Cray XT3 systems

i=cliffw
i=wangyb

Build OFED 1.3's kernel-ib and kernel-ib-devel with our patched lustre
server kernel for RHEL 4 and 5 and SLES 10.  SLES 9 is not supported by
OFED 1.3.
Since OFED 1.3 provides a newer open-iscsi stack, prefer those modules
from the kernel-ib over those provided by the vendor kernels.
Additionally, build Lustre with o2iblnd support.

i=grev

fix for test_16 failure due to trying to create mntpt dir on Read-Only fs.

i=adilger
i=johann

land attachment 16523

fixes this compile warning (which is treated as an error -Werror on XT3's):
during lustrefs-1.6-ss builds:

lustrefs-1.6-ss/lustre/mds/mds_lov.c: In function `mds_lov_read_objids':
mds_lov.c:198: warning: int format, different type arg (arg 10)

b=14410
i=bzzz
i=kalpak

Force q->max_phys_segments to MAX_PHYS_SEGMENTS on SLES10 to be
sure that 1MB requests are not fragmented by the block layer.

i=adilger
i=rread

add linux/random.h to AC_CHECK_HEADERS
add #if defined(HAVE_LINUX_RANDOM_H) to lustre/lvfs/prng.c

restore generation check.
b=15069
i=johan
i=rread

b=15240
i=adilger

Remove the RHEL4 comment in ChangeLog.

b=14929
i=zam

Document the maximum filesystem name length.
b=15486

b=15191
i=wangdi, shadow

during liblustre test during

mgc_llog_init() (lustre/mgc/libmgc.c)
  llog_initiator_connect(ctxt)

while in liblustre lib, it is defined as an empty function only return 0, while
in kernel lib, it set the context's loc_imp. And this context's loc_imp is
required after in llog_client_create().

This fix defines llog_initiator_connect() in both lib (kernel version and
liblustre version)

b=14714
i=adilger, johann

Fixed previous error.

b=14576
i=shadow, nikita

Add own nr_free_buffer_pages() function when it wasn't exported.

b=14856
i=green
i=johan

i=Adilger
skip test_37 if CLIENTMODSONLY is set

i=Adilger
force test_20a and test_20b to do write on the correct ost

b=15517
i=johan
i=deen

sys/syscall.h exist at all os'es instead of <syscall.h> which exist only
at linux.
b=14929
i=nikita

b=15412
i=johann
i=deen

move platform independed part of chksum into separate include
i=johann
i=green

i=nikita
i=johann

r=adilger,grev
- make synchronous fix for HEAD as in 1.6. Do not use RUNAS_ID + 1 in test_72.

r=shadow,adilger
- remove @ns from list using list_del_init() to makes sure that consequent list_empty() will really work right on removed list_head.

i=cliffw
i=wangyb

Allow a kernel target to specify a delimiter between the final "flavor"
specification and the version.
Specifically this allows RH kernels to be <version>smp and SLES10 kernels
to be <version>-smp.
Maintaining the vendors versioning specification means that when software
is trying to determine a vendors kernel using patterns (like OFED does), it
properly detects the SLES10 kernel.
This all lays the ground work for bug 15316.

i=Nathan.Rutman
skip test_37 for client only testing