RDMA: Add netlink infrastructure

Add basic RDMA netlink infrastructure that allows for registration of
RDMA clients for which data is to be exported and supplies message
construction callbacks.

Signed-off-by: Nir Muchtar <nirm@voltaire.com>

[ Reorganize a few things, add CONFIG_NET dependency.  - Roland ]

Signed-off-by: Roland Dreier <roland@purestorage.com>

5 files changed, 205 insertions, 3 deletions

diff --git a/drivers/infiniband/Kconfig b/drivers/infiniband/Kconfig
index 6e35eccc9caa..0f9a84c1046a 100644
--- a/drivers/infiniband/Kconfig
+++ b/drivers/infiniband/Kconfig
@@ -2,6 +2,7 @@ menuconfig INFINIBAND
         tristate "InfiniBand support"
         depends on PCI || BROKEN
         depends on HAS_IOMEM
+        depends on NET
         ---help---
           Core support for InfiniBand (IB).  Make sure to also select
           any protocols you wish to use as well as drivers for your
diff --git a/drivers/infiniband/core/Makefile b/drivers/infiniband/core/Makefile
index cb1ab3ea4998..c8bbaef1becb 100644
--- a/drivers/infiniband/core/Makefile
+++ b/drivers/infiniband/core/Makefile
@@ -8,7 +8,7 @@ obj-$(CONFIG_INFINIBAND_USER_ACCESS) +=	ib_uverbs.o ib_ucm.o \
                                         $(user_access-y)
 
 ib_core-y :=                    packer.o ud_header.o verbs.o sysfs.o \
-                                device.o fmr_pool.o cache.o
+                                device.o fmr_pool.o cache.o netlink.o
 ib_core-$(CONFIG_INFINIBAND_USER_MEM) += umem.o
 
 ib_mad-y :=                     mad.o smi.o agent.o mad_rmpp.o
diff --git a/drivers/infiniband/core/device.c b/drivers/infiniband/core/device.c
index 46a7a3febc12..4007f721d25d 100644
--- a/drivers/infiniband/core/device.c
+++ b/drivers/infiniband/core/device.c
@@ -38,6 +38,7 @@
 #include <linux/slab.h>
 #include <linux/init.h>
 #include <linux/mutex.h>
+#include <rdma/rdma_netlink.h>
 
 #include "core_priv.h"
 
@@ -730,14 +731,23 @@ static int __init ib_core_init(void)
                 goto err;
         }
 
+        ret = ibnl_init();
+        if (ret) {
+                printk(KERN_WARNING "Couldn't init IB netlink interface\n");
+                goto err_sysfs;
+        }
+
         ret = ib_cache_setup();
         if (ret) {
                 printk(KERN_WARNING "Couldn't set up InfiniBand P_Key/GID cache\n");
-                goto err_sysfs;
+                goto err_nl;
         }
 
         return 0;
 
+err_nl:
+        ibnl_cleanup();
+
 err_sysfs:
         ib_sysfs_cleanup();
 
@@ -749,6 +759,7 @@ err:
 static void __exit ib_core_cleanup(void)
 {
         ib_cache_cleanup();
+        ibnl_cleanup();
         ib_sysfs_cleanup();
         /* Make sure that any pending umem accounting work is done. */
         destroy_workqueue(ib_wq);
diff --git a/drivers/infiniband/core/netlink.c b/drivers/infiniband/core/netlink.c
new file mode 100644
index 000000000000..4a5abaf0a25c
--- /dev/null
+++ b/drivers/infiniband/core/netlink.c
@@ -0,0 +1,190 @@
+/*
+ * Copyright (c) 2010 Voltaire Inc.  All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#define pr_fmt(fmt) "%s:%s: " fmt, KBUILD_MODNAME, __func__
+
+#include <net/netlink.h>
+#include <net/net_namespace.h>
+#include <net/sock.h>
+#include <rdma/rdma_netlink.h>
+
+struct ibnl_client {
+        struct list_head                list;
+        int                             index;
+        int                             nops;
+        const struct ibnl_client_cbs   *cb_table;
+};
+
+static DEFINE_MUTEX(ibnl_mutex);
+static struct sock *nls;
+static LIST_HEAD(client_list);
+
+int ibnl_add_client(int index, int nops,
+                    const struct ibnl_client_cbs cb_table[])
+{
+        struct ibnl_client *cur;
+        struct ibnl_client *nl_client;
+
+        nl_client = kmalloc(sizeof *nl_client, GFP_KERNEL);
+        if (!nl_client)
+                return -ENOMEM;
+
+        nl_client->index        = index;
+        nl_client->nops         = nops;
+        nl_client->cb_table     = cb_table;
+
+        mutex_lock(&ibnl_mutex);
+
+        list_for_each_entry(cur, &client_list, list) {
+                if (cur->index == index) {
+                        pr_warn("Client for %d already exists\n", index);
+                        mutex_unlock(&ibnl_mutex);
+                        kfree(nl_client);
+                        return -EINVAL;
+                }
+        }
+
+        list_add_tail(&nl_client->list, &client_list);
+
+        mutex_unlock(&ibnl_mutex);
+
+        return 0;
+}
+EXPORT_SYMBOL(ibnl_add_client);
+
+int ibnl_remove_client(int index)
+{
+        struct ibnl_client *cur, *next;
+
+        mutex_lock(&ibnl_mutex);
+        list_for_each_entry_safe(cur, next, &client_list, list) {
+                if (cur->index == index) {
+                        list_del(&(cur->list));
+                        mutex_unlock(&ibnl_mutex);
+                        kfree(cur);
+                        return 0;
+                }
+        }
+        pr_warn("Can't remove callback for client idx %d. Not found\n", index);
+        mutex_unlock(&ibnl_mutex);
+
+        return -EINVAL;
+}
+EXPORT_SYMBOL(ibnl_remove_client);
+
+void *ibnl_put_msg(struct sk_buff *skb, struct nlmsghdr **nlh, int seq,
+                   int len, int client, int op)
+{
+        unsigned char *prev_tail;
+
+        prev_tail = skb_tail_pointer(skb);
+        *nlh = NLMSG_NEW(skb, 0, seq, RDMA_NL_GET_TYPE(client, op),
+                        len, NLM_F_MULTI);
+        (*nlh)->nlmsg_len = skb_tail_pointer(skb) - prev_tail;
+        return NLMSG_DATA(*nlh);
+
+nlmsg_failure:
+        nlmsg_trim(skb, prev_tail);
+        return NULL;
+}
+EXPORT_SYMBOL(ibnl_put_msg);
+
+int ibnl_put_attr(struct sk_buff *skb, struct nlmsghdr *nlh,
+                  int len, void *data, int type)
+{
+        unsigned char *prev_tail;
+
+        prev_tail = skb_tail_pointer(skb);
+        NLA_PUT(skb, type, len, data);
+        nlh->nlmsg_len += skb_tail_pointer(skb) - prev_tail;
+        return 0;
+
+nla_put_failure:
+        nlmsg_trim(skb, prev_tail - nlh->nlmsg_len);
+        return -EMSGSIZE;
+}
+EXPORT_SYMBOL(ibnl_put_attr);
+
+static int ibnl_rcv_msg(struct sk_buff *skb, struct nlmsghdr *nlh)
+{
+        struct ibnl_client *client;
+        int type = nlh->nlmsg_type;
+        int index = RDMA_NL_GET_CLIENT(type);
+        int op = RDMA_NL_GET_OP(type);
+
+        list_for_each_entry(client, &client_list, list) {
+                if (client->index == index) {
+                        if (op < 0 || op >= client->nops ||
+                            !client->cb_table[RDMA_NL_GET_OP(op)].dump)
+                                return -EINVAL;
+                        return netlink_dump_start(nls, skb, nlh,
+                                                  client->cb_table[op].dump,
+                                                  NULL);
+                }
+        }
+
+        pr_info("Index %d wasn't found in client list\n", index);
+        return -EINVAL;
+}
+
+static void ibnl_rcv(struct sk_buff *skb)
+{
+        mutex_lock(&ibnl_mutex);
+        netlink_rcv_skb(skb, &ibnl_rcv_msg);
+        mutex_unlock(&ibnl_mutex);
+}
+
+int __init ibnl_init(void)
+{
+        nls = netlink_kernel_create(&init_net, NETLINK_RDMA, 0, ibnl_rcv,
+                                    NULL, THIS_MODULE);
+        if (!nls) {
+                pr_warn("Failed to create netlink socket\n");
+                return -ENOMEM;
+        }
+
+        return 0;
+}
+
+void ibnl_cleanup(void)
+{
+        struct ibnl_client *cur, *next;
+
+        mutex_lock(&ibnl_mutex);
+        list_for_each_entry_safe(cur, next, &client_list, list) {
+                list_del(&(cur->list));
+                kfree(cur);
+        }
+        mutex_unlock(&ibnl_mutex);
+
+        netlink_kernel_release(nls);
+}
diff --git a/drivers/infiniband/hw/qib/Kconfig b/drivers/infiniband/hw/qib/Kconfig
index 7c03a70c55a2..8349f9c5064c 100644
--- a/drivers/infiniband/hw/qib/Kconfig
+++ b/drivers/infiniband/hw/qib/Kconfig
@@ -1,6 +1,6 @@
 config INFINIBAND_QIB
         tristate "QLogic PCIe HCA support"
-        depends on 64BIT && NET
+        depends on 64BIT
         ---help---
         This is a low-level driver for QLogic PCIe QLE InfiniBand host
         channel adapters.  This driver does not support the QLogic