LU-6261 gnilnd: Changes for small message rate improvment

Change number of threads on Aries service nodes to 7 for better
performance of small messages.
Change the max_immediate value to 8k instead of 2k. FMA has better
performance up to 8k.
lnet passes an niov of 256 reflecting the size of the router buffer
number of pages in the buffer pool when receiving data from the ib
interface when the message size is over one page. Adjust the niov to
reflect the message size/offset for sends.
Use the cookie defined in gni_pub.h instead of value from gnilnd.h.

Signed-off-by: Chris Horn <hornc@cray.com>
Change-Id: I84270653efcff56f06da7de7fb9674ae319800dd
Reviewed-on: http://review.whamcloud.com/15431
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: James Simmons <uja.ornl@yahoo.com>
Reviewed-by: Doug Oucharek <doug.s.oucharek@intel.com>
Reviewed-by: James Shimek <jshimek@cray.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>

diff --git a/lnet/klnds/gnilnd/gnilnd.c b/lnet/klnds/gnilnd/gnilnd.c
index 4961d44..c911944 100644 (file)
--- a/lnet/klnds/gnilnd/gnilnd.c
+++ b/lnet/klnds/gnilnd/gnilnd.c
@@ -1955,7 +1955,7 @@ kgnilnd_dev_init(kgn_device_t *dev)
        cq_size = *kgnilnd_tunables.kgn_credits * 2 * 3;
 
        rrc = kgnilnd_cdm_create(dev->gnd_id, *kgnilnd_tunables.kgn_ptag,
-                                GNILND_COOKIE, 0,
+                                GNI_JOB_CREATE_COOKIE(GNI_PKEY_LND, 0), 0,
                                 &dev->gnd_domain);
        if (rrc != GNI_RC_SUCCESS) {
                CERROR("Can't create CDM %d (%d)\n", dev->gnd_id, rrc);

diff --git a/lnet/klnds/gnilnd/gnilnd.h b/lnet/klnds/gnilnd/gnilnd.h
index d82e971..cb9c20b 100644 (file)
--- a/lnet/klnds/gnilnd/gnilnd.h
+++ b/lnet/klnds/gnilnd/gnilnd.h
@@ -99,7 +99,6 @@
 /* fixed constants */
 #define GNILND_MAXDEVS         1               /* max # of GNI devices currently supported */
 #define GNILND_MBOX_CREDITS    256             /* number of credits per mailbox */
-#define GNILND_COOKIE          0xa3579         /* cookie used by along with ptag by GNI */
 #define GNILND_CONN_MAGIC         0xa100f       /* magic value for verifying connection validity */
 /* checksum values */
 #define GNILND_CHECKSUM_OFF            0       /* checksum turned off */
@@ -120,7 +119,6 @@
 #define GNILND_COMPUTE            1             /* compute image */
 #define GNILND_FAST_RECONNECT     1             /* Fast Reconnect option */
 #else
-#define GNILND_SCHED_THREADS      3             /* default # of kgnilnd_scheduler threads */
 #define GNILND_FMABLK             1024          /* default number of mboxes per fmablk */
 #define GNILND_SCHED_NICE         -20          /* default nice value for scheduler threads */
 #define GNILND_COMPUTE            0             /* service image */

diff --git a/lnet/klnds/gnilnd/gnilnd_aries.h b/lnet/klnds/gnilnd/gnilnd_aries.h
index e017fe3..ce187ed 100644 (file)
--- a/lnet/klnds/gnilnd/gnilnd_aries.h
+++ b/lnet/klnds/gnilnd/gnilnd_aries.h
@@ -51,6 +51,7 @@
 #else
 #define GNILND_REVERSE_RDMA        GNILND_REVERSE_GET
 #define GNILND_RDMA_DLVR_OPTION    GNI_DLVMODE_PERFORMANCE
+#define GNILND_SCHED_THREADS       7             /* scheduler threads */
 #endif
 
 /* Thread-safe kgni implemented in minor ver 45, code rev 0xb9 */

diff --git a/lnet/klnds/gnilnd/gnilnd_cb.c b/lnet/klnds/gnilnd/gnilnd_cb.c
index 92aaf57..7175f0d 100644 (file)
--- a/lnet/klnds/gnilnd/gnilnd_cb.c
+++ b/lnet/klnds/gnilnd/gnilnd_cb.c
@@ -505,8 +505,14 @@ kgnilnd_setup_immediate_buffer(kgn_tx_t *tx, unsigned int niov, struct iovec *io
        if (nob == 0) {
                tx->tx_buffer = NULL;
        } else if (kiov != NULL) {
+
+               if ((niov > 0) && unlikely(niov > (nob/PAGE_SIZE))) {
+                       niov = ((nob + offset + PAGE_SIZE - 1) / PAGE_SIZE);
+               }
+
                LASSERTF(niov > 0 && niov < GNILND_MAX_IMMEDIATE/PAGE_SIZE,
-                        "bad niov %d\n", niov);
+                       "bad niov %d msg %p kiov %p iov %p offset %d nob%d\n",
+                       niov, msg, kiov, iov, offset, nob);
 
                while (offset >= kiov->kiov_len) {
                        offset -= kiov->kiov_len;

diff --git a/lnet/klnds/gnilnd/gnilnd_gemini.h b/lnet/klnds/gnilnd/gnilnd_gemini.h
index 79be5b3..1e7e2f6 100644 (file)
--- a/lnet/klnds/gnilnd/gnilnd_gemini.h
+++ b/lnet/klnds/gnilnd/gnilnd_gemini.h
@@ -34,6 +34,10 @@
 #define GNILND_REVERSE_RDMA        GNILND_REVERSE_NONE
 #define GNILND_RDMA_DLVR_OPTION     GNI_DLVMODE_PERFORMANCE
 
+#if !defined(CONFIG_CRAY_COMPUTE)
+#define GNILND_SCHED_THREADS        3            /* scheduler threads */
+#endif
+
 /* Thread-safe kgni implemented in minor ver 44, code rev 0xb9 */
 #define GNILND_KGNI_TS_MINOR_VER 0x44
 

diff --git a/lnet/klnds/gnilnd/gnilnd_modparams.c b/lnet/klnds/gnilnd/gnilnd_modparams.c
index 8bef922..7d1c39f 100644 (file)
--- a/lnet/klnds/gnilnd/gnilnd_modparams.c
+++ b/lnet/klnds/gnilnd/gnilnd_modparams.c
@@ -66,7 +66,7 @@ static int max_reconnect_interval = GNILND_MAX_RECONNECT_TO;
 CFS_MODULE_PARM(max_reconnect_interval, "i", int, 0644,
                "maximum connection retry interval (seconds)");
 
-static int max_immediate = (2<<10);
+static int max_immediate = 8192;
 CFS_MODULE_PARM(max_immediate, "i", int, 0644,
                "immediate/RDMA breakpoint");