mlx4_core: Add module parameter for number of MTTs per segment

The current MTT allocator uses kmalloc() to allocate a buffer for its
buddy allocator, and thus is limited in the amount of MTT segments
that it can control.  As a result, the size of memory that can be
registered is limited too.  This patch uses a module parameter to
control the number of MTT entries that each segment represents,
allowing more memory to be registered with the same number of
segments.

Signed-off-by: Eli Cohen <eli@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>

4 files changed, 17 insertions, 6 deletions

diff --git a/drivers/net/mlx4/main.c b/drivers/net/mlx4/main.c
index 30bea9689694..018348c01193 100644
--- a/drivers/net/mlx4/main.c
+++ b/drivers/net/mlx4/main.c
@@ -100,6 +100,10 @@ module_param_named(use_prio, use_prio, bool, 0444);
 MODULE_PARM_DESC(use_prio, "Enable steering by VLAN priority on ETH ports "
 		  "(0/1, default 0)");
 
+static int log_mtts_per_seg = ilog2(MLX4_MTT_ENTRY_PER_SEG);
+module_param_named(log_mtts_per_seg, log_mtts_per_seg, int, 0444);
+MODULE_PARM_DESC(log_mtts_per_seg, "Log2 number of MTT entries per segment (1-5)");
+
 int mlx4_check_port_params(struct mlx4_dev *dev,
 			   enum mlx4_port_type *port_type)
 {
@@ -203,12 +207,13 @@ static int mlx4_dev_cap(struct mlx4_dev *dev, struct mlx4_dev_cap *dev_cap)
 	dev->caps.max_cqes	     = dev_cap->max_cq_sz - 1;
 	dev->caps.reserved_cqs	     = dev_cap->reserved_cqs;
 	dev->caps.reserved_eqs	     = dev_cap->reserved_eqs;
+	dev->caps.mtts_per_seg	     = 1 << log_mtts_per_seg;
 	dev->caps.reserved_mtts	     = DIV_ROUND_UP(dev_cap->reserved_mtts,
-						    MLX4_MTT_ENTRY_PER_SEG);
+						    dev->caps.mtts_per_seg);
 	dev->caps.reserved_mrws	     = dev_cap->reserved_mrws;
 	dev->caps.reserved_uars	     = dev_cap->reserved_uars;
 	dev->caps.reserved_pds	     = dev_cap->reserved_pds;
-	dev->caps.mtt_entry_sz	     = MLX4_MTT_ENTRY_PER_SEG * dev_cap->mtt_entry_sz;
+	dev->caps.mtt_entry_sz	     = dev->caps.mtts_per_seg * dev_cap->mtt_entry_sz;
 	dev->caps.max_msg_sz         = dev_cap->max_msg_sz;
 	dev->caps.page_size_cap	     = ~(u32) (dev_cap->min_page_sz - 1);
 	dev->caps.flags		     = dev_cap->flags;
@@ -1304,6 +1309,11 @@ static int __init mlx4_verify_params(void)
 		return -1;
 	}
 
+	if ((log_mtts_per_seg < 1) || (log_mtts_per_seg > 5)) {
+		printk(KERN_WARNING "mlx4_core: bad log_mtts_per_seg: %d\n", log_mtts_per_seg);
+		return -1;
+	}
+
 	return 0;
 }
 
diff --git a/drivers/net/mlx4/mr.c b/drivers/net/mlx4/mr.c
index 0caf74cae8bc..3b8973d19933 100644
--- a/drivers/net/mlx4/mr.c
+++ b/drivers/net/mlx4/mr.c
@@ -209,7 +209,7 @@ int mlx4_mtt_init(struct mlx4_dev *dev, int npages, int page_shift,
 	} else
 		mtt->page_shift = page_shift;
 
-	for (mtt->order = 0, i = MLX4_MTT_ENTRY_PER_SEG; i < npages; i <<= 1)
+	for (mtt->order = 0, i = dev->caps.mtts_per_seg; i < npages; i <<= 1)
 		++mtt->order;
 
 	mtt->first_seg = mlx4_alloc_mtt_range(dev, mtt->order);
@@ -350,7 +350,7 @@ int mlx4_mr_enable(struct mlx4_dev *dev, struct mlx4_mr *mr)
 		mpt_entry->pd_flags |= cpu_to_be32(MLX4_MPT_PD_FLAG_FAST_REG |
 						   MLX4_MPT_PD_FLAG_RAE);
 		mpt_entry->mtt_sz    = cpu_to_be32((1 << mr->mtt.order) *
-						   MLX4_MTT_ENTRY_PER_SEG);
+						   dev->caps.mtts_per_seg);
 	} else {
 		mpt_entry->flags    |= cpu_to_be32(MLX4_MPT_FLAG_SW_OWNS);
 	}
@@ -391,7 +391,7 @@ static int mlx4_write_mtt_chunk(struct mlx4_dev *dev, struct mlx4_mtt *mtt,
 	    (start_index + npages - 1) / (PAGE_SIZE / sizeof (u64)))
 		return -EINVAL;
 
-	if (start_index & (MLX4_MTT_ENTRY_PER_SEG - 1))
+	if (start_index & (dev->caps.mtts_per_seg - 1))
 		return -EINVAL;
 
 	mtts = mlx4_table_find(&priv->mr_table.mtt_table, mtt->first_seg +
diff --git a/drivers/net/mlx4/profile.c b/drivers/net/mlx4/profile.c
index cebdf3243ca1..bd22df95adf9 100644
--- a/drivers/net/mlx4/profile.c
+++ b/drivers/net/mlx4/profile.c
@@ -98,7 +98,7 @@ u64 mlx4_make_profile(struct mlx4_dev *dev,
 	profile[MLX4_RES_EQ].size     = dev_cap->eqc_entry_sz;
 	profile[MLX4_RES_DMPT].size   = dev_cap->dmpt_entry_sz;
 	profile[MLX4_RES_CMPT].size   = dev_cap->cmpt_entry_sz;
-	profile[MLX4_RES_MTT].size    = MLX4_MTT_ENTRY_PER_SEG * dev_cap->mtt_entry_sz;
+	profile[MLX4_RES_MTT].size    = dev->caps.mtts_per_seg * dev_cap->mtt_entry_sz;
 	profile[MLX4_RES_MCG].size    = MLX4_MGM_ENTRY_SIZE;
 
 	profile[MLX4_RES_QP].num      = request->num_qp;
diff --git a/include/linux/mlx4/device.h b/include/linux/mlx4/device.h
index 3aff8a6a389e..ce7cc6c7bcbb 100644
--- a/include/linux/mlx4/device.h
+++ b/include/linux/mlx4/device.h
@@ -210,6 +210,7 @@ struct mlx4_caps {
 	int			num_comp_vectors;
 	int			num_mpts;
 	int			num_mtt_segs;
+	int			mtts_per_seg;
 	int			fmr_reserved_mtts;
 	int			reserved_mtts;
 	int			reserved_mrws;