RDS: Explicitly allocate rm in sendmsg()

r_m_copy_from_user used to allocate the rm as well as kernel
buffers for the data, and then copy the data in. Now, sendmsg()
allocates the rm, although the data buffer alloc still happens
in r_m_copy_from_user.

SGs are still allocated with rm, but now r_m_alloc_sgs() is
used to reserve them. This allows multiple SG lists to be
allocated from the one rm -- this is important once we also
want to alloc our rdma sgl from this pool.

Signed-off-by: Andy Grover <andy.grover@oracle.com>

1 files changed, 27 insertions, 4 deletions

diff --git a/net/rds/send.c b/net/rds/send.c
index 19dfd025498e..28d09447207b 100644
--- a/net/rds/send.c
+++ b/net/rds/send.c
@@ -758,6 +758,19 @@ out:
         return *queued;
 }
 
+/*
+ * rds_message is getting to be quite complicated, and we'd like to allocate
+ * it all in one go. This figures out how big it needs to be up front.
+ */
+static int rds_rm_size(struct msghdr *msg, int data_len)
+{
+        int size = 0;
+
+        size += ceil(data_len, PAGE_SIZE) * sizeof(struct scatterlist);
+
+        return size;
+}
+
 static int rds_cmsg_send(struct rds_sock *rs, struct rds_message *rm,
                          struct msghdr *msg, int *allocated_mr)
 {
@@ -845,13 +858,23 @@ int rds_sendmsg(struct kiocb *iocb, struct socket *sock, struct msghdr *msg,
                 goto out;
         }
 
-        rm = rds_message_copy_from_user(msg->msg_iov, payload_len);
-        if (IS_ERR(rm)) {
-                ret = PTR_ERR(rm);
-                rm = NULL;
+        /* size of rm including all sgs */
+        ret = rds_rm_size(msg, payload_len);
+        if (ret < 0)
+                goto out;
+
+        rm = rds_message_alloc(ret, GFP_KERNEL);
+        if (!rm) {
+                ret = -ENOMEM;
                 goto out;
         }
 
+        rm->data.m_sg = rds_message_alloc_sgs(rm, ceil(payload_len, PAGE_SIZE));
+        /* XXX fix this to not allocate memory */
+        ret = rds_message_copy_from_user(rm, msg->msg_iov, payload_len);
+        if (ret)
+                goto out;
+
         rm->m_daddr = daddr;
 
         /* rds_conn_create has a spinlock that runs with IRQ off.