Whamcloud - gitweb
b=24017 fix the performance issue of lu_cache_shrink
authorJian Yu <jian.yu@oracle.com>
Tue, 16 Nov 2010 15:38:24 +0000 (23:38 +0800)
committerVitaly Fertman <vitaly.fertman@sun.com>
Tue, 16 Nov 2010 22:28:47 +0000 (01:28 +0300)
commit4a2abc926bf7408bbdd13fc0d99dc8abd54d8e37
treeba89b707c55a4e627659a8a4123f3167204451cd
parent4a99c80813924fdcd3871a166394734fb4cd045a
b=24017 fix the performance issue of lu_cache_shrink

- increase busy counter in lu_obj_hop_get, so that we don't have to hack out
  this in llite
- make some changes to cfs_hash:
  . do not need to return anything by cfs_hash_get/put, they are not used by
    anyone, cfs_hash_object also can do the same thing
  . pass cfs_hash to cfs_hash_get/put

o=liang
i=vitaly.fertman
libcfs/include/libcfs/libcfs_hash.h
libcfs/libcfs/hash.c
lustre/include/lu_object.h
lustre/ldlm/ldlm_lockd.c
lustre/lov/lov_pool.c
lustre/obdclass/cl_object.c
lustre/obdclass/lu_object.c
lustre/obdclass/obd_config.c
lustre/ptlrpc/connection.c
lustre/quota/quota_context.c