diff --git a/doc/guides/nics/mlx4.rst b/doc/guides/nics/mlx4.rst
index cd34838f41..c8a02be4dd 100644
--- a/doc/guides/nics/mlx4.rst
+++ b/doc/guides/nics/mlx4.rst
@@ -119,6 +119,17 @@ Run-time configuration
   times for additional ports. All ports are probed by default if left
   unspecified.
 
+- ``mr_ext_memseg_en`` parameter [int]
+
+  A nonzero value enables extending memseg when registering DMA memory. If
+  enabled, the number of entries in MR (Memory Region) lookup table on datapath
+  is minimized and it benefits performance. On the other hand, it worsens memory
+  utilization because registered memory is pinned by kernel driver. Even if a
+  page in the extended chunk is freed, that doesn't become reusable until the
+  entire memory is freed.
+
+  Enabled by default.
+
 Kernel module parameters
 ~~~~~~~~~~~~~~~~~~~~~~~~
 
diff --git a/drivers/net/mlx4/mlx4.c b/drivers/net/mlx4/mlx4.c
index 315640a6d7..252658fc6a 100644
--- a/drivers/net/mlx4/mlx4.c
+++ b/drivers/net/mlx4/mlx4.c
@@ -66,11 +66,14 @@ struct mlx4_conf {
 		uint32_t present; /**< Bit-field for existing ports. */
 		uint32_t enabled; /**< Bit-field for user-enabled ports. */
 	} ports;
+	int mr_ext_memseg_en;
+	/** Whether memseg should be extended for MR creation. */
 };
 
 /* Available parameters list. */
 const char *pmd_mlx4_init_params[] = {
 	MLX4_PMD_PORT_KVARG,
+	MLX4_MR_EXT_MEMSEG_EN_KVARG,
 	NULL,
 };
 
@@ -509,6 +512,8 @@ mlx4_arg_parse(const char *key, const char *val, struct mlx4_conf *conf)
 			return -rte_errno;
 		}
 		conf->ports.enabled |= 1 << tmp;
+	} else if (strcmp(MLX4_MR_EXT_MEMSEG_EN_KVARG, key) == 0) {
+		conf->mr_ext_memseg_en = !!tmp;
 	} else {
 		rte_errno = EINVAL;
 		WARN("%s: unknown parameter", key);
@@ -544,10 +549,10 @@ mlx4_args(struct rte_devargs *devargs, struct mlx4_conf *conf)
 	}
 	/* Process parameters. */
 	for (i = 0; pmd_mlx4_init_params[i]; ++i) {
-		arg_count = rte_kvargs_count(kvlist, MLX4_PMD_PORT_KVARG);
+		arg_count = rte_kvargs_count(kvlist, pmd_mlx4_init_params[i]);
 		while (arg_count-- > 0) {
 			ret = rte_kvargs_process(kvlist,
-						 MLX4_PMD_PORT_KVARG,
+						 pmd_mlx4_init_params[i],
 						 (int (*)(const char *,
 							  const char *,
 							  void *))
@@ -876,6 +881,7 @@ mlx4_pci_probe(struct rte_pci_driver *pci_drv, struct rte_pci_device *pci_dev)
 	struct ibv_device_attr_ex device_attr_ex;
 	struct mlx4_conf conf = {
 		.ports.present = 0,
+		.mr_ext_memseg_en = 1,
 	};
 	unsigned int vf;
 	int i;
@@ -1100,6 +1106,7 @@ mlx4_pci_probe(struct rte_pci_driver *pci_drv, struct rte_pci_device *pci_dev)
 					device_attr_ex.tso_caps.max_tso;
 		DEBUG("TSO is %ssupported",
 		      priv->tso ? "" : "not ");
+		priv->mr_ext_memseg_en = conf.mr_ext_memseg_en;
 		/* Configure the first MAC address by default. */
 		err = mlx4_get_mac(priv, &mac.addr_bytes);
 		if (err) {
diff --git a/drivers/net/mlx4/mlx4.h b/drivers/net/mlx4/mlx4.h
index 1a7b1fb541..4ff98d772b 100644
--- a/drivers/net/mlx4/mlx4.h
+++ b/drivers/net/mlx4/mlx4.h
@@ -53,6 +53,9 @@
 /** Port parameter. */
 #define MLX4_PMD_PORT_KVARG "port"
 
+/** Enable extending memsegs when creating a MR. */
+#define MLX4_MR_EXT_MEMSEG_EN_KVARG "mr_ext_memseg_en"
+
 /* Reserved address space for UAR mapping. */
 #define MLX4_UAR_SIZE (1ULL << (sizeof(uintptr_t) * 4))
 
@@ -165,6 +168,8 @@ struct mlx4_priv {
 	uint32_t hw_csum_l2tun:1; /**< Checksum support for L2 tunnels. */
 	uint32_t hw_fcs_strip:1; /**< FCS stripping toggling is supported. */
 	uint32_t tso:1; /**< Transmit segmentation offload is supported. */
+	uint32_t mr_ext_memseg_en:1;
+	/** Whether memseg should be extended for MR creation. */
 	uint32_t tso_max_payload_sz; /**< Max supported TSO payload size. */
 	uint32_t hw_rss_max_qps; /**< Max Rx Queues supported by RSS. */
 	uint64_t hw_rss_sup; /**< Supported RSS hash fields (Verbs format). */
diff --git a/drivers/net/mlx4/mlx4_mr.c b/drivers/net/mlx4/mlx4_mr.c
index 0ba55fda04..6db917a092 100644
--- a/drivers/net/mlx4/mlx4_mr.c
+++ b/drivers/net/mlx4/mlx4_mr.c
@@ -580,14 +580,24 @@ mlx4_mr_create(struct rte_eth_dev *dev, struct mlx4_mr_cache *entry,
 	 */
 	mlx4_mr_garbage_collect(dev);
 	/*
-	 * Find out a contiguous virtual address chunk in use, to which the
-	 * given address belongs, in order to register maximum range. In the
-	 * best case where mempools are not dynamically recreated and
+	 * If enabled, find out a contiguous virtual address chunk in use, to
+	 * which the given address belongs, in order to register maximum range.
+	 * In the best case where mempools are not dynamically recreated and
 	 * '--socket-mem' is specified as an EAL option, it is very likely to
 	 * have only one MR(LKey) per a socket and per a hugepage-size even
-	 * though the system memory is highly fragmented.
+	 * though the system memory is highly fragmented. As the whole memory
+	 * chunk will be pinned by kernel, it can't be reused unless entire
+	 * chunk is freed from EAL.
+	 *
+	 * If disabled, just register one memseg (page). Then, memory
+	 * consumption will be minimized but it may drop performance if there
+	 * are many MRs to lookup on the datapath.
 	 */
-	if (!rte_memseg_contig_walk(mr_find_contig_memsegs_cb, &data)) {
+	if (!priv->mr_ext_memseg_en) {
+		data.msl = rte_mem_virt2memseg_list((void *)addr);
+		data.start = RTE_ALIGN_FLOOR(addr, data.msl->page_sz);
+		data.end = data.start + data.msl->page_sz;
+	} else if (!rte_memseg_contig_walk(mr_find_contig_memsegs_cb, &data)) {
 		WARN("port %u unable to find virtually contiguous"
 		     " chunk for address (%p)."
 		     " rte_memseg_contig_walk() failed.",