Branch HEAD
b=17461 i=adilger i=alex.zhuravlev Ptlrpc optimizations to minimize COS overhead: . reply state object use own locking instead of ptlrpc service one . per-cpu dedicated threads for reply handling . resource hash table size increase
Showing
- lustre/include/lustre_dlm.h 1 addition, 1 deletionlustre/include/lustre_dlm.h
- lustre/include/lustre_net.h 14 additions, 3 deletionslustre/include/lustre_net.h
- lustre/ldlm/ldlm_lib.c 4 additions, 4 deletionslustre/ldlm/ldlm_lib.c
- lustre/mdt/mdt_handler.c 2 additions, 0 deletionslustre/mdt/mdt_handler.c
- lustre/mdt/mdt_recovery.c 2 additions, 0 deletionslustre/mdt/mdt_recovery.c
- lustre/mgs/mgs_handler.c 2 additions, 0 deletionslustre/mgs/mgs_handler.c
- lustre/obdecho/echo.c 2 additions, 0 deletionslustre/obdecho/echo.c
- lustre/ptlrpc/events.c 2 additions, 0 deletionslustre/ptlrpc/events.c
- lustre/ptlrpc/pack_generic.c 2 additions, 0 deletionslustre/ptlrpc/pack_generic.c
- lustre/ptlrpc/ptlrpc_module.c 7 additions, 0 deletionslustre/ptlrpc/ptlrpc_module.c
- lustre/ptlrpc/service.c 414 additions, 70 deletionslustre/ptlrpc/service.c
Loading
Please register or sign in to comment