For memory regions that are allocated using reg_umr, the suffix of
mlx5_core_create_mkey isn't being called. Instead the creation is completed in
a callback function (reg_mr_callback). This means that these MRs aren't being
added to the MR radix tree. The patch add them in the callback.
A later patch that adds page fault handling uses the MR radix tree to find the
MR pointer given an key that is read from the page fault event from hardware,
or from the WQE that caused the page fault. Therefore we need all MRs,
including those created with the reg_mr_callback to be in that tree.
Signed-off-by: Haggai Eran <hagg...@mellanox.com>
---
drivers/infiniband/hw/mlx5/mr.c | 9 +++++++++
1 file changed, 9 insertions(+)
diff --git a/drivers/infiniband/hw/mlx5/mr.c b/drivers/infiniband/hw/mlx5/mr.c
index 24a68aa..f447257 100644
--- a/drivers/infiniband/hw/mlx5/mr.c
+++ b/drivers/infiniband/hw/mlx5/mr.c
@@ -73,6 +73,8 @@ static void reg_mr_callback(int status, void *context)
struct mlx5_cache_ent *ent = &cache->ent[c];
u8 key;
unsigned long flags;
+ struct mlx5_mr_table *table = &dev->mdev.priv.mr_table;
+ int err;
spin_lock_irqsave(&ent->lock, flags);
ent->pending--;
@@ -107,6 +109,13 @@ static void reg_mr_callback(int status, void *context)
ent->cur++;
ent->size++;
spin_unlock_irqrestore(&ent->lock, flags);
+
+ write_lock_irq(&table->lock);
+ err = radix_tree_insert(&table->tree, mlx5_base_mkey(mr->mmr.key),
+ &mr->mmr);
+ if (err)
+ pr_err("Error inserting to mr tree. 0x%x\n", -err);
+ write_unlock_irq(&table->lock);
}
static int add_keys(struct mlx5_ib_dev *dev, int c, int num)
--
1.7.11.2
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
