Version: 1
-Previous: 736d4dfe2bf524bc9d965c3a3c229278333f6ea8
+Previous: 6b65f4b2e13fa8248b9a20712264c56fdef7486f
Head: 5a39fae0e595b08ff2b5ca58896ee0ecd51ab66c
Applied:
rs-shutdown: 110fcad2ecce39d6888448c19b08791ea193e7f9
ip6-opt: 7e43a759255e9890d1e41f1edf71792836f53941
buf-addr: eb70620389ced3c0c705d44b69e691a47cd3e4fd
rs-1sge: 85e4c4a0da0b501b60a1035d7a003ee20a749511
- rs-def-inline: 6ad70d3b80178e40fea97f662a6b31e51cddc0af
Hidden:
+++ /dev/null
-Bottom: e9e6fbdeef17d4c02f8c8aa86981e6965cd2d120
-Top: ca334c09004719e129d514631e1e81d0bd614f71
-Author: Sean Hefty <sean.hefty@intel.com>
-Date: 2012-05-24 18:18:43 -0700
-
-rsockets: Reduce the default inline size
-
-Inline data consumes the same space used by the SGL. Since
-we reduced the default number of SGEs per SQ entry to 1,
-also reduce the default inline data size to 16 bytes.
-Otherwise, the SQ size won't actually be reduced.
-
-Although this increases the latency of small messages over
-16 bytes, tests show that decreasing the inline data size
-from 64 bytes to 32 or 16 bytes improves large message
-bandwidth 8-10%.
-
-Sample rstream 64k_bw test results:
-
-inline size 64: 24.85 Gbps
-inline size 16: 26.5 Gbps
-
-Signed-off-by: Sean Hefty <sean.hefty@intel.com>
-
-
----
-
-diff --git a/src/rsocket.c b/src/rsocket.c
-index bd4c8dd..6634e97 100644
---- a/src/rsocket.c
-+++ b/src/rsocket.c
-@@ -51,7 +51,6 @@
- #include "cma.h"
- #include "indexer.h"
-
--#define RS_INLINE 64
- #define RS_OLAP_START_SIZE 2048
- #define RS_MAX_TRANSFER 65536
- #define RS_QP_SIZE 384
-@@ -223,7 +222,7 @@ static struct rsocket *rs_alloc(struct rsocket *inherited_rs)
- rs->ctrl_avail = inherited_rs->ctrl_avail;
- } else {
- rs->sbuf_size = rs->rbuf_size = RS_BUF_SIZE;
-- rs->sq_inline = RS_INLINE;
-+ rs->sq_inline = RS_MIN_INLINE;
- rs->sq_size = rs->rq_size = RS_QP_SIZE;
- rs->ctrl_avail = RS_QP_CTRL_SIZE;
- }