2005-04-17 00:20:36 +02:00
|
|
|
/*
|
|
|
|
* Copyright (c) 2004, 2005 Topspin Communications. All rights reserved.
|
2005-08-11 08:03:10 +02:00
|
|
|
* Copyright (c) 2005 Sun Microsystems, Inc. All rights reserved.
|
|
|
|
* Copyright (c) 2004 Voltaire, Inc. All rights reserved.
|
2005-04-17 00:20:36 +02:00
|
|
|
*
|
|
|
|
* 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.
|
|
|
|
*/
|
|
|
|
|
|
|
|
#include <linux/skbuff.h>
|
|
|
|
#include <linux/rtnetlink.h>
|
|
|
|
#include <linux/ip.h>
|
|
|
|
#include <linux/in.h>
|
|
|
|
#include <linux/igmp.h>
|
|
|
|
#include <linux/inetdevice.h>
|
|
|
|
#include <linux/delay.h>
|
|
|
|
#include <linux/completion.h>
|
include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h
percpu.h is included by sched.h and module.h and thus ends up being
included when building most .c files. percpu.h includes slab.h which
in turn includes gfp.h making everything defined by the two files
universally available and complicating inclusion dependencies.
percpu.h -> slab.h dependency is about to be removed. Prepare for
this change by updating users of gfp and slab facilities include those
headers directly instead of assuming availability. As this conversion
needs to touch large number of source files, the following script is
used as the basis of conversion.
http://userweb.kernel.org/~tj/misc/slabh-sweep.py
The script does the followings.
* Scan files for gfp and slab usages and update includes such that
only the necessary includes are there. ie. if only gfp is used,
gfp.h, if slab is used, slab.h.
* When the script inserts a new include, it looks at the include
blocks and try to put the new include such that its order conforms
to its surrounding. It's put in the include block which contains
core kernel includes, in the same order that the rest are ordered -
alphabetical, Christmas tree, rev-Xmas-tree or at the end if there
doesn't seem to be any matching order.
* If the script can't find a place to put a new include (mostly
because the file doesn't have fitting include block), it prints out
an error message indicating which .h file needs to be added to the
file.
The conversion was done in the following steps.
1. The initial automatic conversion of all .c files updated slightly
over 4000 files, deleting around 700 includes and adding ~480 gfp.h
and ~3000 slab.h inclusions. The script emitted errors for ~400
files.
2. Each error was manually checked. Some didn't need the inclusion,
some needed manual addition while adding it to implementation .h or
embedding .c file was more appropriate for others. This step added
inclusions to around 150 files.
3. The script was run again and the output was compared to the edits
from #2 to make sure no file was left behind.
4. Several build tests were done and a couple of problems were fixed.
e.g. lib/decompress_*.c used malloc/free() wrappers around slab
APIs requiring slab.h to be added manually.
5. The script was run on all .h files but without automatically
editing them as sprinkling gfp.h and slab.h inclusions around .h
files could easily lead to inclusion dependency hell. Most gfp.h
inclusion directives were ignored as stuff from gfp.h was usually
wildly available and often used in preprocessor macros. Each
slab.h inclusion directive was examined and added manually as
necessary.
6. percpu.h was updated not to include slab.h.
7. Build test were done on the following configurations and failures
were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my
distributed build env didn't work with gcov compiles) and a few
more options had to be turned off depending on archs to make things
build (like ipr on powerpc/64 which failed due to missing writeq).
* x86 and x86_64 UP and SMP allmodconfig and a custom test config.
* powerpc and powerpc64 SMP allmodconfig
* sparc and sparc64 SMP allmodconfig
* ia64 SMP allmodconfig
* s390 SMP allmodconfig
* alpha SMP allmodconfig
* um on x86_64 SMP allmodconfig
8. percpu.h modifications were reverted so that it could be applied as
a separate patch and serve as bisection point.
Given the fact that I had only a couple of failures from tests on step
6, I'm fairly confident about the coverage of this conversion patch.
If there is a breakage, it's likely to be something in one of the arch
headers which should be easily discoverable easily on most builds of
the specific arch.
Signed-off-by: Tejun Heo <tj@kernel.org>
Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
2010-03-24 09:04:11 +01:00
|
|
|
#include <linux/slab.h>
|
2005-04-17 00:20:36 +02:00
|
|
|
|
2005-12-27 05:43:12 +01:00
|
|
|
#include <net/dst.h>
|
|
|
|
|
2005-04-17 00:20:36 +02:00
|
|
|
#include "ipoib.h"
|
|
|
|
|
|
|
|
#ifdef CONFIG_INFINIBAND_IPOIB_DEBUG
|
|
|
|
static int mcast_debug_level;
|
|
|
|
|
|
|
|
module_param(mcast_debug_level, int, 0644);
|
|
|
|
MODULE_PARM_DESC(mcast_debug_level,
|
|
|
|
"Enable multicast debug tracing if > 0");
|
|
|
|
#endif
|
|
|
|
|
2006-01-13 23:51:39 +01:00
|
|
|
static DEFINE_MUTEX(mcast_mutex);
|
2005-04-17 00:20:36 +02:00
|
|
|
|
|
|
|
struct ipoib_mcast_iter {
|
|
|
|
struct net_device *dev;
|
|
|
|
union ib_gid mgid;
|
|
|
|
unsigned long created;
|
|
|
|
unsigned int queuelen;
|
|
|
|
unsigned int complete;
|
|
|
|
unsigned int send_only;
|
|
|
|
};
|
|
|
|
|
|
|
|
static void ipoib_mcast_free(struct ipoib_mcast *mcast)
|
|
|
|
{
|
|
|
|
struct net_device *dev = mcast->dev;
|
|
|
|
struct ipoib_dev_priv *priv = netdev_priv(dev);
|
|
|
|
struct ipoib_neigh *neigh, *tmp;
|
2006-01-17 21:19:40 +01:00
|
|
|
int tx_dropped = 0;
|
2005-04-17 00:20:36 +02:00
|
|
|
|
2008-10-29 20:52:50 +01:00
|
|
|
ipoib_dbg_mcast(netdev_priv(dev), "deleting multicast group %pI6\n",
|
2008-10-29 06:37:22 +01:00
|
|
|
mcast->mcmember.mgid.raw);
|
2005-04-17 00:20:36 +02:00
|
|
|
|
2008-09-30 19:36:21 +02:00
|
|
|
spin_lock_irq(&priv->lock);
|
2005-04-17 00:20:36 +02:00
|
|
|
|
|
|
|
list_for_each_entry_safe(neigh, tmp, &mcast->neigh_list, list) {
|
2006-01-10 16:43:02 +01:00
|
|
|
/*
|
|
|
|
* It's safe to call ipoib_put_ah() inside priv->lock
|
|
|
|
* here, because we know that mcast->ah will always
|
|
|
|
* hold one more reference, so ipoib_put_ah() will
|
|
|
|
* never do more than decrement the ref count.
|
|
|
|
*/
|
2005-04-17 00:20:36 +02:00
|
|
|
if (neigh->ah)
|
2006-01-10 16:43:02 +01:00
|
|
|
ipoib_put_ah(neigh->ah);
|
2006-11-16 13:16:47 +01:00
|
|
|
ipoib_neigh_free(dev, neigh);
|
2005-04-17 00:20:36 +02:00
|
|
|
}
|
|
|
|
|
2008-09-30 19:36:21 +02:00
|
|
|
spin_unlock_irq(&priv->lock);
|
2005-04-17 00:20:36 +02:00
|
|
|
|
|
|
|
if (mcast->ah)
|
|
|
|
ipoib_put_ah(mcast->ah);
|
|
|
|
|
2006-01-17 21:19:40 +01:00
|
|
|
while (!skb_queue_empty(&mcast->pkt_queue)) {
|
|
|
|
++tx_dropped;
|
2005-11-07 19:49:38 +01:00
|
|
|
dev_kfree_skb_any(skb_dequeue(&mcast->pkt_queue));
|
2006-01-17 21:19:40 +01:00
|
|
|
}
|
|
|
|
|
2008-09-30 19:36:21 +02:00
|
|
|
netif_tx_lock_bh(dev);
|
2007-09-29 00:33:51 +02:00
|
|
|
dev->stats.tx_dropped += tx_dropped;
|
2008-09-30 19:36:21 +02:00
|
|
|
netif_tx_unlock_bh(dev);
|
2005-04-17 00:20:36 +02:00
|
|
|
|
|
|
|
kfree(mcast);
|
|
|
|
}
|
|
|
|
|
|
|
|
static struct ipoib_mcast *ipoib_mcast_alloc(struct net_device *dev,
|
|
|
|
int can_sleep)
|
|
|
|
{
|
|
|
|
struct ipoib_mcast *mcast;
|
|
|
|
|
2005-11-02 16:23:14 +01:00
|
|
|
mcast = kzalloc(sizeof *mcast, can_sleep ? GFP_KERNEL : GFP_ATOMIC);
|
2005-04-17 00:20:36 +02:00
|
|
|
if (!mcast)
|
|
|
|
return NULL;
|
|
|
|
|
|
|
|
mcast->dev = dev;
|
|
|
|
mcast->created = jiffies;
|
2005-09-18 22:47:53 +02:00
|
|
|
mcast->backoff = 1;
|
2005-04-17 00:20:36 +02:00
|
|
|
|
|
|
|
INIT_LIST_HEAD(&mcast->list);
|
|
|
|
INIT_LIST_HEAD(&mcast->neigh_list);
|
|
|
|
skb_queue_head_init(&mcast->pkt_queue);
|
|
|
|
|
|
|
|
return mcast;
|
|
|
|
}
|
|
|
|
|
2006-05-29 18:14:05 +02:00
|
|
|
static struct ipoib_mcast *__ipoib_mcast_find(struct net_device *dev, void *mgid)
|
2005-04-17 00:20:36 +02:00
|
|
|
{
|
|
|
|
struct ipoib_dev_priv *priv = netdev_priv(dev);
|
|
|
|
struct rb_node *n = priv->multicast_tree.rb_node;
|
|
|
|
|
|
|
|
while (n) {
|
|
|
|
struct ipoib_mcast *mcast;
|
|
|
|
int ret;
|
|
|
|
|
|
|
|
mcast = rb_entry(n, struct ipoib_mcast, rb_node);
|
|
|
|
|
2006-05-29 18:14:05 +02:00
|
|
|
ret = memcmp(mgid, mcast->mcmember.mgid.raw,
|
2005-04-17 00:20:36 +02:00
|
|
|
sizeof (union ib_gid));
|
|
|
|
if (ret < 0)
|
|
|
|
n = n->rb_left;
|
|
|
|
else if (ret > 0)
|
|
|
|
n = n->rb_right;
|
|
|
|
else
|
|
|
|
return mcast;
|
|
|
|
}
|
|
|
|
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int __ipoib_mcast_add(struct net_device *dev, struct ipoib_mcast *mcast)
|
|
|
|
{
|
|
|
|
struct ipoib_dev_priv *priv = netdev_priv(dev);
|
|
|
|
struct rb_node **n = &priv->multicast_tree.rb_node, *pn = NULL;
|
|
|
|
|
|
|
|
while (*n) {
|
|
|
|
struct ipoib_mcast *tmcast;
|
|
|
|
int ret;
|
|
|
|
|
|
|
|
pn = *n;
|
|
|
|
tmcast = rb_entry(pn, struct ipoib_mcast, rb_node);
|
|
|
|
|
|
|
|
ret = memcmp(mcast->mcmember.mgid.raw, tmcast->mcmember.mgid.raw,
|
|
|
|
sizeof (union ib_gid));
|
|
|
|
if (ret < 0)
|
|
|
|
n = &pn->rb_left;
|
|
|
|
else if (ret > 0)
|
|
|
|
n = &pn->rb_right;
|
|
|
|
else
|
|
|
|
return -EEXIST;
|
|
|
|
}
|
|
|
|
|
|
|
|
rb_link_node(&mcast->rb_node, pn, n);
|
|
|
|
rb_insert_color(&mcast->rb_node, &priv->multicast_tree);
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int ipoib_mcast_join_finish(struct ipoib_mcast *mcast,
|
|
|
|
struct ib_sa_mcmember_rec *mcmember)
|
|
|
|
{
|
|
|
|
struct net_device *dev = mcast->dev;
|
|
|
|
struct ipoib_dev_priv *priv = netdev_priv(dev);
|
2006-02-28 05:47:43 +01:00
|
|
|
struct ipoib_ah *ah;
|
2005-04-17 00:20:36 +02:00
|
|
|
int ret;
|
2008-07-15 08:48:50 +02:00
|
|
|
int set_qkey = 0;
|
2005-04-17 00:20:36 +02:00
|
|
|
|
|
|
|
mcast->mcmember = *mcmember;
|
|
|
|
|
|
|
|
/* Set the cached Q_Key before we attach if it's the broadcast group */
|
|
|
|
if (!memcmp(mcast->mcmember.mgid.raw, priv->dev->broadcast + 4,
|
|
|
|
sizeof (union ib_gid))) {
|
2008-05-21 00:41:09 +02:00
|
|
|
spin_lock_irq(&priv->lock);
|
|
|
|
if (!priv->broadcast) {
|
|
|
|
spin_unlock_irq(&priv->lock);
|
|
|
|
return -EAGAIN;
|
|
|
|
}
|
2005-04-17 00:20:36 +02:00
|
|
|
priv->qkey = be32_to_cpu(priv->broadcast->mcmember.qkey);
|
2008-05-21 00:41:09 +02:00
|
|
|
spin_unlock_irq(&priv->lock);
|
2005-04-17 00:20:36 +02:00
|
|
|
priv->tx_wr.wr.ud.remote_qkey = priv->qkey;
|
2008-07-15 08:48:50 +02:00
|
|
|
set_qkey = 1;
|
2005-04-17 00:20:36 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
if (!test_bit(IPOIB_MCAST_FLAG_SENDONLY, &mcast->flags)) {
|
|
|
|
if (test_and_set_bit(IPOIB_MCAST_FLAG_ATTACHED, &mcast->flags)) {
|
2008-10-29 20:52:50 +01:00
|
|
|
ipoib_warn(priv, "multicast group %pI6 already attached\n",
|
2008-10-29 06:37:22 +01:00
|
|
|
mcast->mcmember.mgid.raw);
|
2005-04-17 00:20:36 +02:00
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
ret = ipoib_mcast_attach(dev, be16_to_cpu(mcast->mcmember.mlid),
|
2008-07-15 08:48:50 +02:00
|
|
|
&mcast->mcmember.mgid, set_qkey);
|
2005-04-17 00:20:36 +02:00
|
|
|
if (ret < 0) {
|
2008-10-29 20:52:50 +01:00
|
|
|
ipoib_warn(priv, "couldn't attach QP to multicast group %pI6\n",
|
2008-10-29 06:37:22 +01:00
|
|
|
mcast->mcmember.mgid.raw);
|
2005-04-17 00:20:36 +02:00
|
|
|
|
|
|
|
clear_bit(IPOIB_MCAST_FLAG_ATTACHED, &mcast->flags);
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
{
|
|
|
|
struct ib_ah_attr av = {
|
|
|
|
.dlid = be16_to_cpu(mcast->mcmember.mlid),
|
|
|
|
.port_num = priv->port,
|
|
|
|
.sl = mcast->mcmember.sl,
|
|
|
|
.ah_flags = IB_AH_GRH,
|
2006-04-10 18:43:47 +02:00
|
|
|
.static_rate = mcast->mcmember.rate,
|
2005-04-17 00:20:36 +02:00
|
|
|
.grh = {
|
|
|
|
.flow_label = be32_to_cpu(mcast->mcmember.flow_label),
|
|
|
|
.hop_limit = mcast->mcmember.hop_limit,
|
|
|
|
.sgid_index = 0,
|
|
|
|
.traffic_class = mcast->mcmember.traffic_class
|
|
|
|
}
|
|
|
|
};
|
|
|
|
av.grh.dgid = mcast->mcmember.mgid;
|
|
|
|
|
2006-02-28 05:47:43 +01:00
|
|
|
ah = ipoib_create_ah(dev, priv->pd, &av);
|
|
|
|
if (!ah) {
|
2005-04-17 00:20:36 +02:00
|
|
|
ipoib_warn(priv, "ib_address_create failed\n");
|
|
|
|
} else {
|
2006-07-24 09:42:00 +02:00
|
|
|
spin_lock_irq(&priv->lock);
|
|
|
|
mcast->ah = ah;
|
|
|
|
spin_unlock_irq(&priv->lock);
|
|
|
|
|
2008-10-29 20:52:50 +01:00
|
|
|
ipoib_dbg_mcast(priv, "MGID %pI6 AV %p, LID 0x%04x, SL %d\n",
|
2008-10-29 06:37:22 +01:00
|
|
|
mcast->mcmember.mgid.raw,
|
2005-04-17 00:20:36 +02:00
|
|
|
mcast->ah->ah,
|
|
|
|
be16_to_cpu(mcast->mcmember.mlid),
|
|
|
|
mcast->mcmember.sl);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/* actually send any queued packets */
|
2008-09-30 19:36:21 +02:00
|
|
|
netif_tx_lock_bh(dev);
|
2005-04-17 00:20:36 +02:00
|
|
|
while (!skb_queue_empty(&mcast->pkt_queue)) {
|
|
|
|
struct sk_buff *skb = skb_dequeue(&mcast->pkt_queue);
|
2008-09-30 19:36:21 +02:00
|
|
|
netif_tx_unlock_bh(dev);
|
2005-04-17 00:20:36 +02:00
|
|
|
|
|
|
|
skb->dev = dev;
|
|
|
|
|
2009-06-02 07:19:30 +02:00
|
|
|
if (!skb_dst(skb) || !skb_dst(skb)->neighbour) {
|
2005-04-17 00:20:36 +02:00
|
|
|
/* put pseudoheader back on for next time */
|
|
|
|
skb_push(skb, sizeof (struct ipoib_pseudoheader));
|
|
|
|
}
|
|
|
|
|
|
|
|
if (dev_queue_xmit(skb))
|
|
|
|
ipoib_warn(priv, "dev_queue_xmit failed to requeue packet\n");
|
2008-09-30 19:36:21 +02:00
|
|
|
netif_tx_lock_bh(dev);
|
2005-04-17 00:20:36 +02:00
|
|
|
}
|
2008-09-30 19:36:21 +02:00
|
|
|
netif_tx_unlock_bh(dev);
|
2005-04-17 00:20:36 +02:00
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2007-02-16 02:00:17 +01:00
|
|
|
static int
|
2005-04-17 00:20:36 +02:00
|
|
|
ipoib_mcast_sendonly_join_complete(int status,
|
2007-02-16 02:00:17 +01:00
|
|
|
struct ib_sa_multicast *multicast)
|
2005-04-17 00:20:36 +02:00
|
|
|
{
|
2007-02-16 02:00:17 +01:00
|
|
|
struct ipoib_mcast *mcast = multicast->context;
|
2005-04-17 00:20:36 +02:00
|
|
|
struct net_device *dev = mcast->dev;
|
|
|
|
|
2007-02-16 02:00:17 +01:00
|
|
|
/* We trap for port events ourselves. */
|
|
|
|
if (status == -ENETRESET)
|
|
|
|
return 0;
|
|
|
|
|
2005-04-17 00:20:36 +02:00
|
|
|
if (!status)
|
2007-02-16 02:00:17 +01:00
|
|
|
status = ipoib_mcast_join_finish(mcast, &multicast->rec);
|
|
|
|
|
|
|
|
if (status) {
|
2005-04-17 00:20:36 +02:00
|
|
|
if (mcast->logcount++ < 20)
|
2008-10-29 20:52:50 +01:00
|
|
|
ipoib_dbg_mcast(netdev_priv(dev), "multicast join failed for %pI6, status %d\n",
|
2008-10-29 06:37:22 +01:00
|
|
|
mcast->mcmember.mgid.raw, status);
|
2005-04-17 00:20:36 +02:00
|
|
|
|
|
|
|
/* Flush out any queued packets */
|
2008-09-30 19:36:21 +02:00
|
|
|
netif_tx_lock_bh(dev);
|
2006-01-17 21:19:40 +01:00
|
|
|
while (!skb_queue_empty(&mcast->pkt_queue)) {
|
2007-09-29 00:33:51 +02:00
|
|
|
++dev->stats.tx_dropped;
|
2005-11-07 19:49:38 +01:00
|
|
|
dev_kfree_skb_any(skb_dequeue(&mcast->pkt_queue));
|
2006-01-17 21:19:40 +01:00
|
|
|
}
|
2008-09-30 19:36:21 +02:00
|
|
|
netif_tx_unlock_bh(dev);
|
2005-04-17 00:20:36 +02:00
|
|
|
|
|
|
|
/* Clear the busy flag so we try again */
|
2007-02-16 02:00:17 +01:00
|
|
|
status = test_and_clear_bit(IPOIB_MCAST_FLAG_BUSY,
|
|
|
|
&mcast->flags);
|
2005-04-17 00:20:36 +02:00
|
|
|
}
|
2007-02-16 02:00:17 +01:00
|
|
|
return status;
|
2005-04-17 00:20:36 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
static int ipoib_mcast_sendonly_join(struct ipoib_mcast *mcast)
|
|
|
|
{
|
|
|
|
struct net_device *dev = mcast->dev;
|
|
|
|
struct ipoib_dev_priv *priv = netdev_priv(dev);
|
|
|
|
struct ib_sa_mcmember_rec rec = {
|
|
|
|
#if 0 /* Some SMs don't support send-only yet */
|
|
|
|
.join_state = 4
|
|
|
|
#else
|
|
|
|
.join_state = 1
|
|
|
|
#endif
|
|
|
|
};
|
|
|
|
int ret = 0;
|
|
|
|
|
|
|
|
if (!test_bit(IPOIB_FLAG_OPER_UP, &priv->flags)) {
|
|
|
|
ipoib_dbg_mcast(priv, "device shutting down, no multicast joins\n");
|
|
|
|
return -ENODEV;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (test_and_set_bit(IPOIB_MCAST_FLAG_BUSY, &mcast->flags)) {
|
|
|
|
ipoib_dbg_mcast(priv, "multicast entry busy, skipping\n");
|
|
|
|
return -EBUSY;
|
|
|
|
}
|
|
|
|
|
|
|
|
rec.mgid = mcast->mcmember.mgid;
|
|
|
|
rec.port_gid = priv->local_gid;
|
2005-08-14 06:05:57 +02:00
|
|
|
rec.pkey = cpu_to_be16(priv->pkey);
|
2005-04-17 00:20:36 +02:00
|
|
|
|
2007-02-16 02:00:17 +01:00
|
|
|
mcast->mc = ib_sa_join_multicast(&ipoib_sa_client, priv->ca,
|
|
|
|
priv->port, &rec,
|
|
|
|
IB_SA_MCMEMBER_REC_MGID |
|
|
|
|
IB_SA_MCMEMBER_REC_PORT_GID |
|
|
|
|
IB_SA_MCMEMBER_REC_PKEY |
|
|
|
|
IB_SA_MCMEMBER_REC_JOIN_STATE,
|
|
|
|
GFP_ATOMIC,
|
|
|
|
ipoib_mcast_sendonly_join_complete,
|
|
|
|
mcast);
|
|
|
|
if (IS_ERR(mcast->mc)) {
|
|
|
|
ret = PTR_ERR(mcast->mc);
|
|
|
|
clear_bit(IPOIB_MCAST_FLAG_BUSY, &mcast->flags);
|
|
|
|
ipoib_warn(priv, "ib_sa_join_multicast failed (ret = %d)\n",
|
2005-04-17 00:20:36 +02:00
|
|
|
ret);
|
|
|
|
} else {
|
2008-10-29 20:52:50 +01:00
|
|
|
ipoib_dbg_mcast(priv, "no multicast record for %pI6, starting join\n",
|
2008-10-29 06:37:22 +01:00
|
|
|
mcast->mcmember.mgid.raw);
|
2005-04-17 00:20:36 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2008-09-16 20:57:45 +02:00
|
|
|
void ipoib_mcast_carrier_on_task(struct work_struct *work)
|
|
|
|
{
|
|
|
|
struct ipoib_dev_priv *priv = container_of(work, struct ipoib_dev_priv,
|
|
|
|
carrier_on_task);
|
2009-09-24 21:01:05 +02:00
|
|
|
struct ib_port_attr attr;
|
2008-09-16 20:57:45 +02:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Take rtnl_lock to avoid racing with ipoib_stop() and
|
|
|
|
* turning the carrier back on while a device is being
|
|
|
|
* removed.
|
|
|
|
*/
|
2009-09-24 21:01:05 +02:00
|
|
|
if (ib_query_port(priv->ca, priv->port, &attr) ||
|
|
|
|
attr.state != IB_PORT_ACTIVE) {
|
|
|
|
ipoib_dbg(priv, "Keeping carrier off until IB port is active\n");
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
2008-09-16 20:57:45 +02:00
|
|
|
rtnl_lock();
|
|
|
|
netif_carrier_on(priv->dev);
|
|
|
|
rtnl_unlock();
|
|
|
|
}
|
|
|
|
|
2007-02-16 02:00:17 +01:00
|
|
|
static int ipoib_mcast_join_complete(int status,
|
|
|
|
struct ib_sa_multicast *multicast)
|
2005-04-17 00:20:36 +02:00
|
|
|
{
|
2007-02-16 02:00:17 +01:00
|
|
|
struct ipoib_mcast *mcast = multicast->context;
|
2005-04-17 00:20:36 +02:00
|
|
|
struct net_device *dev = mcast->dev;
|
|
|
|
struct ipoib_dev_priv *priv = netdev_priv(dev);
|
|
|
|
|
2008-10-29 20:52:50 +01:00
|
|
|
ipoib_dbg_mcast(priv, "join completion for %pI6 (status %d)\n",
|
2008-10-29 06:37:22 +01:00
|
|
|
mcast->mcmember.mgid.raw, status);
|
2005-04-17 00:20:36 +02:00
|
|
|
|
2007-02-16 02:00:17 +01:00
|
|
|
/* We trap for port events ourselves. */
|
|
|
|
if (status == -ENETRESET)
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
if (!status)
|
|
|
|
status = ipoib_mcast_join_finish(mcast, &multicast->rec);
|
|
|
|
|
|
|
|
if (!status) {
|
2005-09-18 22:47:53 +02:00
|
|
|
mcast->backoff = 1;
|
2006-01-13 23:51:39 +01:00
|
|
|
mutex_lock(&mcast_mutex);
|
2005-04-17 00:20:36 +02:00
|
|
|
if (test_bit(IPOIB_MCAST_RUN, &priv->flags))
|
2006-11-22 15:57:56 +01:00
|
|
|
queue_delayed_work(ipoib_workqueue,
|
|
|
|
&priv->mcast_task, 0);
|
2006-01-13 23:51:39 +01:00
|
|
|
mutex_unlock(&mcast_mutex);
|
2007-03-08 23:59:30 +01:00
|
|
|
|
2008-09-16 20:57:45 +02:00
|
|
|
/*
|
|
|
|
* Defer carrier on work to ipoib_workqueue to avoid a
|
|
|
|
* deadlock on rtnl_lock here.
|
|
|
|
*/
|
|
|
|
if (mcast == priv->broadcast)
|
|
|
|
queue_work(ipoib_workqueue, &priv->carrier_on_task);
|
2007-03-08 23:59:30 +01:00
|
|
|
|
2007-02-16 02:00:17 +01:00
|
|
|
return 0;
|
2005-04-17 00:20:36 +02:00
|
|
|
}
|
|
|
|
|
2007-02-16 02:00:17 +01:00
|
|
|
if (mcast->logcount++ < 20) {
|
2009-01-16 22:42:59 +01:00
|
|
|
if (status == -ETIMEDOUT || status == -EAGAIN) {
|
2008-10-29 20:52:50 +01:00
|
|
|
ipoib_dbg_mcast(priv, "multicast join failed for %pI6, status %d\n",
|
2008-10-29 06:37:22 +01:00
|
|
|
mcast->mcmember.mgid.raw, status);
|
2005-04-17 00:20:36 +02:00
|
|
|
} else {
|
2008-10-29 20:52:50 +01:00
|
|
|
ipoib_warn(priv, "multicast join failed for %pI6, status %d\n",
|
2008-10-29 06:37:22 +01:00
|
|
|
mcast->mcmember.mgid.raw, status);
|
2005-04-17 00:20:36 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
mcast->backoff *= 2;
|
|
|
|
if (mcast->backoff > IPOIB_MAX_BACKOFF_SECONDS)
|
|
|
|
mcast->backoff = IPOIB_MAX_BACKOFF_SECONDS;
|
|
|
|
|
2007-02-16 02:00:17 +01:00
|
|
|
/* Clear the busy flag so we try again */
|
|
|
|
status = test_and_clear_bit(IPOIB_MCAST_FLAG_BUSY, &mcast->flags);
|
2006-03-02 20:07:47 +01:00
|
|
|
|
2007-02-16 02:00:17 +01:00
|
|
|
mutex_lock(&mcast_mutex);
|
2006-03-02 20:07:47 +01:00
|
|
|
spin_lock_irq(&priv->lock);
|
2007-02-16 02:00:17 +01:00
|
|
|
if (test_bit(IPOIB_MCAST_RUN, &priv->flags))
|
|
|
|
queue_delayed_work(ipoib_workqueue, &priv->mcast_task,
|
|
|
|
mcast->backoff * HZ);
|
2006-03-02 20:07:47 +01:00
|
|
|
spin_unlock_irq(&priv->lock);
|
2006-01-13 23:51:39 +01:00
|
|
|
mutex_unlock(&mcast_mutex);
|
2005-04-17 00:20:36 +02:00
|
|
|
|
2007-02-16 02:00:17 +01:00
|
|
|
return status;
|
2005-04-17 00:20:36 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
static void ipoib_mcast_join(struct net_device *dev, struct ipoib_mcast *mcast,
|
|
|
|
int create)
|
|
|
|
{
|
|
|
|
struct ipoib_dev_priv *priv = netdev_priv(dev);
|
|
|
|
struct ib_sa_mcmember_rec rec = {
|
|
|
|
.join_state = 1
|
|
|
|
};
|
|
|
|
ib_sa_comp_mask comp_mask;
|
|
|
|
int ret = 0;
|
|
|
|
|
2008-10-29 20:52:50 +01:00
|
|
|
ipoib_dbg_mcast(priv, "joining MGID %pI6\n", mcast->mcmember.mgid.raw);
|
2005-04-17 00:20:36 +02:00
|
|
|
|
|
|
|
rec.mgid = mcast->mcmember.mgid;
|
|
|
|
rec.port_gid = priv->local_gid;
|
2005-08-14 06:05:57 +02:00
|
|
|
rec.pkey = cpu_to_be16(priv->pkey);
|
2005-04-17 00:20:36 +02:00
|
|
|
|
|
|
|
comp_mask =
|
|
|
|
IB_SA_MCMEMBER_REC_MGID |
|
|
|
|
IB_SA_MCMEMBER_REC_PORT_GID |
|
|
|
|
IB_SA_MCMEMBER_REC_PKEY |
|
|
|
|
IB_SA_MCMEMBER_REC_JOIN_STATE;
|
|
|
|
|
|
|
|
if (create) {
|
|
|
|
comp_mask |=
|
2006-09-23 00:22:56 +02:00
|
|
|
IB_SA_MCMEMBER_REC_QKEY |
|
|
|
|
IB_SA_MCMEMBER_REC_MTU_SELECTOR |
|
|
|
|
IB_SA_MCMEMBER_REC_MTU |
|
|
|
|
IB_SA_MCMEMBER_REC_TRAFFIC_CLASS |
|
|
|
|
IB_SA_MCMEMBER_REC_RATE_SELECTOR |
|
|
|
|
IB_SA_MCMEMBER_REC_RATE |
|
|
|
|
IB_SA_MCMEMBER_REC_SL |
|
|
|
|
IB_SA_MCMEMBER_REC_FLOW_LABEL |
|
|
|
|
IB_SA_MCMEMBER_REC_HOP_LIMIT;
|
2005-04-17 00:20:36 +02:00
|
|
|
|
|
|
|
rec.qkey = priv->broadcast->mcmember.qkey;
|
2006-09-23 00:22:56 +02:00
|
|
|
rec.mtu_selector = IB_SA_EQ;
|
|
|
|
rec.mtu = priv->broadcast->mcmember.mtu;
|
|
|
|
rec.traffic_class = priv->broadcast->mcmember.traffic_class;
|
|
|
|
rec.rate_selector = IB_SA_EQ;
|
|
|
|
rec.rate = priv->broadcast->mcmember.rate;
|
2005-04-17 00:20:36 +02:00
|
|
|
rec.sl = priv->broadcast->mcmember.sl;
|
|
|
|
rec.flow_label = priv->broadcast->mcmember.flow_label;
|
2006-09-23 00:22:56 +02:00
|
|
|
rec.hop_limit = priv->broadcast->mcmember.hop_limit;
|
2005-04-17 00:20:36 +02:00
|
|
|
}
|
|
|
|
|
2007-02-16 02:00:17 +01:00
|
|
|
set_bit(IPOIB_MCAST_FLAG_BUSY, &mcast->flags);
|
|
|
|
mcast->mc = ib_sa_join_multicast(&ipoib_sa_client, priv->ca, priv->port,
|
|
|
|
&rec, comp_mask, GFP_KERNEL,
|
|
|
|
ipoib_mcast_join_complete, mcast);
|
|
|
|
if (IS_ERR(mcast->mc)) {
|
|
|
|
clear_bit(IPOIB_MCAST_FLAG_BUSY, &mcast->flags);
|
|
|
|
ret = PTR_ERR(mcast->mc);
|
|
|
|
ipoib_warn(priv, "ib_sa_join_multicast failed, status %d\n", ret);
|
2005-04-17 00:20:36 +02:00
|
|
|
|
|
|
|
mcast->backoff *= 2;
|
|
|
|
if (mcast->backoff > IPOIB_MAX_BACKOFF_SECONDS)
|
|
|
|
mcast->backoff = IPOIB_MAX_BACKOFF_SECONDS;
|
|
|
|
|
2006-01-13 23:51:39 +01:00
|
|
|
mutex_lock(&mcast_mutex);
|
2005-04-17 00:20:36 +02:00
|
|
|
if (test_bit(IPOIB_MCAST_RUN, &priv->flags))
|
|
|
|
queue_delayed_work(ipoib_workqueue,
|
|
|
|
&priv->mcast_task,
|
2005-09-18 22:47:53 +02:00
|
|
|
mcast->backoff * HZ);
|
2006-01-13 23:51:39 +01:00
|
|
|
mutex_unlock(&mcast_mutex);
|
2007-02-16 02:00:17 +01:00
|
|
|
}
|
2005-04-17 00:20:36 +02:00
|
|
|
}
|
|
|
|
|
2006-11-22 15:57:56 +01:00
|
|
|
void ipoib_mcast_join_task(struct work_struct *work)
|
2005-04-17 00:20:36 +02:00
|
|
|
{
|
2006-11-22 15:57:56 +01:00
|
|
|
struct ipoib_dev_priv *priv =
|
|
|
|
container_of(work, struct ipoib_dev_priv, mcast_task.work);
|
|
|
|
struct net_device *dev = priv->dev;
|
2005-04-17 00:20:36 +02:00
|
|
|
|
|
|
|
if (!test_bit(IPOIB_MCAST_RUN, &priv->flags))
|
|
|
|
return;
|
|
|
|
|
|
|
|
if (ib_query_gid(priv->ca, priv->port, 0, &priv->local_gid))
|
2007-05-18 15:12:54 +02:00
|
|
|
ipoib_warn(priv, "ib_query_gid() failed\n");
|
2005-04-17 00:20:36 +02:00
|
|
|
else
|
|
|
|
memcpy(priv->dev->dev_addr + 4, priv->local_gid.raw, sizeof (union ib_gid));
|
|
|
|
|
|
|
|
{
|
|
|
|
struct ib_port_attr attr;
|
|
|
|
|
2007-02-22 05:28:05 +01:00
|
|
|
if (!ib_query_port(priv->ca, priv->port, &attr))
|
|
|
|
priv->local_lid = attr.lid;
|
|
|
|
else
|
2007-02-16 02:00:17 +01:00
|
|
|
ipoib_warn(priv, "ib_query_port failed\n");
|
2005-04-17 00:20:36 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
if (!priv->broadcast) {
|
2006-02-11 21:22:12 +01:00
|
|
|
struct ipoib_mcast *broadcast;
|
|
|
|
|
2009-01-13 04:28:42 +01:00
|
|
|
if (!test_bit(IPOIB_FLAG_ADMIN_UP, &priv->flags))
|
|
|
|
return;
|
|
|
|
|
2006-02-11 21:22:12 +01:00
|
|
|
broadcast = ipoib_mcast_alloc(dev, 1);
|
|
|
|
if (!broadcast) {
|
2005-04-17 00:20:36 +02:00
|
|
|
ipoib_warn(priv, "failed to allocate broadcast group\n");
|
2006-01-13 23:51:39 +01:00
|
|
|
mutex_lock(&mcast_mutex);
|
2005-04-17 00:20:36 +02:00
|
|
|
if (test_bit(IPOIB_MCAST_RUN, &priv->flags))
|
|
|
|
queue_delayed_work(ipoib_workqueue,
|
|
|
|
&priv->mcast_task, HZ);
|
2006-01-13 23:51:39 +01:00
|
|
|
mutex_unlock(&mcast_mutex);
|
2005-04-17 00:20:36 +02:00
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
2006-02-11 21:22:12 +01:00
|
|
|
spin_lock_irq(&priv->lock);
|
|
|
|
memcpy(broadcast->mcmember.mgid.raw, priv->dev->broadcast + 4,
|
2005-04-17 00:20:36 +02:00
|
|
|
sizeof (union ib_gid));
|
2006-02-11 21:22:12 +01:00
|
|
|
priv->broadcast = broadcast;
|
2005-04-17 00:20:36 +02:00
|
|
|
|
|
|
|
__ipoib_mcast_add(dev, priv->broadcast);
|
|
|
|
spin_unlock_irq(&priv->lock);
|
|
|
|
}
|
|
|
|
|
|
|
|
if (!test_bit(IPOIB_MCAST_FLAG_ATTACHED, &priv->broadcast->flags)) {
|
2007-02-16 02:00:17 +01:00
|
|
|
if (!test_bit(IPOIB_MCAST_FLAG_BUSY, &priv->broadcast->flags))
|
|
|
|
ipoib_mcast_join(dev, priv->broadcast, 0);
|
2005-04-17 00:20:36 +02:00
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
while (1) {
|
|
|
|
struct ipoib_mcast *mcast = NULL;
|
|
|
|
|
|
|
|
spin_lock_irq(&priv->lock);
|
|
|
|
list_for_each_entry(mcast, &priv->multicast_list, list) {
|
|
|
|
if (!test_bit(IPOIB_MCAST_FLAG_SENDONLY, &mcast->flags)
|
|
|
|
&& !test_bit(IPOIB_MCAST_FLAG_BUSY, &mcast->flags)
|
|
|
|
&& !test_bit(IPOIB_MCAST_FLAG_ATTACHED, &mcast->flags)) {
|
|
|
|
/* Found the next unjoined group */
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
spin_unlock_irq(&priv->lock);
|
|
|
|
|
|
|
|
if (&mcast->list == &priv->multicast_list) {
|
|
|
|
/* All done */
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
ipoib_mcast_join(dev, mcast, 1);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
2008-04-23 20:55:45 +02:00
|
|
|
priv->mcast_mtu = IPOIB_UD_MTU(ib_mtu_enum_to_int(priv->broadcast->mcmember.mtu));
|
2007-02-05 21:12:23 +01:00
|
|
|
|
2008-07-15 08:48:51 +02:00
|
|
|
if (!ipoib_cm_admin_enabled(dev)) {
|
|
|
|
rtnl_lock();
|
2008-07-15 08:48:51 +02:00
|
|
|
dev_set_mtu(dev, min(priv->mcast_mtu, priv->admin_mtu));
|
2008-07-15 08:48:51 +02:00
|
|
|
rtnl_unlock();
|
|
|
|
}
|
2005-04-17 00:20:36 +02:00
|
|
|
|
|
|
|
ipoib_dbg_mcast(priv, "successfully joined all multicast groups\n");
|
|
|
|
|
|
|
|
clear_bit(IPOIB_MCAST_RUN, &priv->flags);
|
|
|
|
}
|
|
|
|
|
|
|
|
int ipoib_mcast_start_thread(struct net_device *dev)
|
|
|
|
{
|
|
|
|
struct ipoib_dev_priv *priv = netdev_priv(dev);
|
|
|
|
|
|
|
|
ipoib_dbg_mcast(priv, "starting multicast thread\n");
|
|
|
|
|
2006-01-13 23:51:39 +01:00
|
|
|
mutex_lock(&mcast_mutex);
|
2005-04-17 00:20:36 +02:00
|
|
|
if (!test_and_set_bit(IPOIB_MCAST_RUN, &priv->flags))
|
2006-11-22 15:57:56 +01:00
|
|
|
queue_delayed_work(ipoib_workqueue, &priv->mcast_task, 0);
|
2006-01-13 23:51:39 +01:00
|
|
|
mutex_unlock(&mcast_mutex);
|
2005-04-17 00:20:36 +02:00
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2005-09-20 19:52:04 +02:00
|
|
|
int ipoib_mcast_stop_thread(struct net_device *dev, int flush)
|
2005-04-17 00:20:36 +02:00
|
|
|
{
|
|
|
|
struct ipoib_dev_priv *priv = netdev_priv(dev);
|
|
|
|
|
|
|
|
ipoib_dbg_mcast(priv, "stopping multicast thread\n");
|
|
|
|
|
2006-01-13 23:51:39 +01:00
|
|
|
mutex_lock(&mcast_mutex);
|
2005-04-17 00:20:36 +02:00
|
|
|
clear_bit(IPOIB_MCAST_RUN, &priv->flags);
|
|
|
|
cancel_delayed_work(&priv->mcast_task);
|
2006-01-13 23:51:39 +01:00
|
|
|
mutex_unlock(&mcast_mutex);
|
2005-04-17 00:20:36 +02:00
|
|
|
|
2005-09-20 19:52:04 +02:00
|
|
|
if (flush)
|
|
|
|
flush_workqueue(ipoib_workqueue);
|
2005-04-17 00:20:36 +02:00
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int ipoib_mcast_leave(struct net_device *dev, struct ipoib_mcast *mcast)
|
|
|
|
{
|
|
|
|
struct ipoib_dev_priv *priv = netdev_priv(dev);
|
|
|
|
int ret = 0;
|
|
|
|
|
2007-03-19 23:31:36 +01:00
|
|
|
if (test_and_clear_bit(IPOIB_MCAST_FLAG_BUSY, &mcast->flags))
|
|
|
|
ib_sa_free_multicast(mcast->mc);
|
|
|
|
|
2007-02-16 02:00:17 +01:00
|
|
|
if (test_and_clear_bit(IPOIB_MCAST_FLAG_ATTACHED, &mcast->flags)) {
|
2008-10-29 20:52:50 +01:00
|
|
|
ipoib_dbg_mcast(priv, "leaving MGID %pI6\n",
|
2008-10-29 06:37:22 +01:00
|
|
|
mcast->mcmember.mgid.raw);
|
2005-04-17 00:20:36 +02:00
|
|
|
|
2007-02-16 02:00:17 +01:00
|
|
|
/* Remove ourselves from the multicast group */
|
2008-07-15 08:48:50 +02:00
|
|
|
ret = ib_detach_mcast(priv->qp, &mcast->mcmember.mgid,
|
|
|
|
be16_to_cpu(mcast->mcmember.mlid));
|
2007-02-16 02:00:17 +01:00
|
|
|
if (ret)
|
2008-07-15 08:48:50 +02:00
|
|
|
ipoib_warn(priv, "ib_detach_mcast failed (result = %d)\n", ret);
|
2007-02-16 02:00:17 +01:00
|
|
|
}
|
2005-04-17 00:20:36 +02:00
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2006-05-29 18:14:05 +02:00
|
|
|
void ipoib_mcast_send(struct net_device *dev, void *mgid, struct sk_buff *skb)
|
2005-04-17 00:20:36 +02:00
|
|
|
{
|
|
|
|
struct ipoib_dev_priv *priv = netdev_priv(dev);
|
|
|
|
struct ipoib_mcast *mcast;
|
2008-09-30 19:36:21 +02:00
|
|
|
unsigned long flags;
|
2005-04-17 00:20:36 +02:00
|
|
|
|
2008-09-30 19:36:21 +02:00
|
|
|
spin_lock_irqsave(&priv->lock, flags);
|
2005-04-17 00:20:36 +02:00
|
|
|
|
2008-03-11 15:10:02 +01:00
|
|
|
if (!test_bit(IPOIB_FLAG_OPER_UP, &priv->flags) ||
|
2006-02-11 21:22:12 +01:00
|
|
|
!priv->broadcast ||
|
|
|
|
!test_bit(IPOIB_MCAST_FLAG_ATTACHED, &priv->broadcast->flags)) {
|
2007-09-29 00:33:51 +02:00
|
|
|
++dev->stats.tx_dropped;
|
2006-02-08 01:37:08 +01:00
|
|
|
dev_kfree_skb_any(skb);
|
|
|
|
goto unlock;
|
|
|
|
}
|
|
|
|
|
2005-04-17 00:20:36 +02:00
|
|
|
mcast = __ipoib_mcast_find(dev, mgid);
|
|
|
|
if (!mcast) {
|
|
|
|
/* Let's create a new send only group now */
|
2008-10-29 20:52:50 +01:00
|
|
|
ipoib_dbg_mcast(priv, "setting up send only multicast group for %pI6\n",
|
2008-10-29 06:37:22 +01:00
|
|
|
mgid);
|
2005-04-17 00:20:36 +02:00
|
|
|
|
|
|
|
mcast = ipoib_mcast_alloc(dev, 0);
|
|
|
|
if (!mcast) {
|
|
|
|
ipoib_warn(priv, "unable to allocate memory for "
|
|
|
|
"multicast structure\n");
|
2007-09-29 00:33:51 +02:00
|
|
|
++dev->stats.tx_dropped;
|
2005-04-17 00:20:36 +02:00
|
|
|
dev_kfree_skb_any(skb);
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
|
|
|
set_bit(IPOIB_MCAST_FLAG_SENDONLY, &mcast->flags);
|
2006-05-29 18:14:05 +02:00
|
|
|
memcpy(mcast->mcmember.mgid.raw, mgid, sizeof (union ib_gid));
|
2005-04-17 00:20:36 +02:00
|
|
|
__ipoib_mcast_add(dev, mcast);
|
|
|
|
list_add_tail(&mcast->list, &priv->multicast_list);
|
|
|
|
}
|
|
|
|
|
|
|
|
if (!mcast->ah) {
|
|
|
|
if (skb_queue_len(&mcast->pkt_queue) < IPOIB_MAX_MCAST_QUEUE)
|
|
|
|
skb_queue_tail(&mcast->pkt_queue, skb);
|
2006-01-17 21:19:40 +01:00
|
|
|
else {
|
2007-09-29 00:33:51 +02:00
|
|
|
++dev->stats.tx_dropped;
|
2005-04-17 00:20:36 +02:00
|
|
|
dev_kfree_skb_any(skb);
|
2006-01-17 21:19:40 +01:00
|
|
|
}
|
2005-04-17 00:20:36 +02:00
|
|
|
|
2007-02-16 02:00:17 +01:00
|
|
|
if (test_bit(IPOIB_MCAST_FLAG_BUSY, &mcast->flags))
|
2005-04-17 00:20:36 +02:00
|
|
|
ipoib_dbg_mcast(priv, "no address vector, "
|
|
|
|
"but multicast join already started\n");
|
|
|
|
else if (test_bit(IPOIB_MCAST_FLAG_SENDONLY, &mcast->flags))
|
|
|
|
ipoib_mcast_sendonly_join(mcast);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* If lookup completes between here and out:, don't
|
|
|
|
* want to send packet twice.
|
|
|
|
*/
|
|
|
|
mcast = NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
out:
|
|
|
|
if (mcast && mcast->ah) {
|
2009-06-02 07:19:30 +02:00
|
|
|
if (skb_dst(skb) &&
|
|
|
|
skb_dst(skb)->neighbour &&
|
|
|
|
!*to_ipoib_neigh(skb_dst(skb)->neighbour)) {
|
|
|
|
struct ipoib_neigh *neigh = ipoib_neigh_alloc(skb_dst(skb)->neighbour,
|
2007-10-10 04:43:36 +02:00
|
|
|
skb->dev);
|
2005-04-17 00:20:36 +02:00
|
|
|
|
|
|
|
if (neigh) {
|
|
|
|
kref_get(&mcast->ah->ref);
|
2007-10-24 04:57:54 +02:00
|
|
|
neigh->ah = mcast->ah;
|
2005-04-17 00:20:36 +02:00
|
|
|
list_add_tail(&neigh->list, &mcast->neigh_list);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2009-09-06 05:23:40 +02:00
|
|
|
spin_unlock_irqrestore(&priv->lock, flags);
|
2005-04-17 00:20:36 +02:00
|
|
|
ipoib_send(dev, skb, mcast->ah, IB_MULTICAST_QPN);
|
2009-09-06 05:23:40 +02:00
|
|
|
return;
|
2005-04-17 00:20:36 +02:00
|
|
|
}
|
|
|
|
|
2006-02-08 01:37:08 +01:00
|
|
|
unlock:
|
2008-09-30 19:36:21 +02:00
|
|
|
spin_unlock_irqrestore(&priv->lock, flags);
|
2005-04-17 00:20:36 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
void ipoib_mcast_dev_flush(struct net_device *dev)
|
|
|
|
{
|
|
|
|
struct ipoib_dev_priv *priv = netdev_priv(dev);
|
|
|
|
LIST_HEAD(remove_list);
|
2006-01-12 23:32:20 +01:00
|
|
|
struct ipoib_mcast *mcast, *tmcast;
|
2005-04-17 00:20:36 +02:00
|
|
|
unsigned long flags;
|
|
|
|
|
|
|
|
ipoib_dbg_mcast(priv, "flushing multicast list\n");
|
|
|
|
|
|
|
|
spin_lock_irqsave(&priv->lock, flags);
|
|
|
|
|
2006-01-12 23:32:20 +01:00
|
|
|
list_for_each_entry_safe(mcast, tmcast, &priv->multicast_list, list) {
|
|
|
|
list_del(&mcast->list);
|
|
|
|
rb_erase(&mcast->rb_node, &priv->multicast_tree);
|
|
|
|
list_add_tail(&mcast->list, &remove_list);
|
2005-04-17 00:20:36 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
if (priv->broadcast) {
|
2006-09-23 00:22:46 +02:00
|
|
|
rb_erase(&priv->broadcast->rb_node, &priv->multicast_tree);
|
2006-01-12 23:32:20 +01:00
|
|
|
list_add_tail(&priv->broadcast->list, &remove_list);
|
|
|
|
priv->broadcast = NULL;
|
2005-04-17 00:20:36 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
spin_unlock_irqrestore(&priv->lock, flags);
|
|
|
|
|
|
|
|
list_for_each_entry_safe(mcast, tmcast, &remove_list, list) {
|
|
|
|
ipoib_mcast_leave(dev, mcast);
|
|
|
|
ipoib_mcast_free(mcast);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2010-03-22 04:21:39 +01:00
|
|
|
static int ipoib_mcast_addr_is_valid(const u8 *addr, const u8 *broadcast)
|
2009-09-06 05:23:40 +02:00
|
|
|
{
|
|
|
|
/* reserved QPN, prefix, scope */
|
|
|
|
if (memcmp(addr, broadcast, 6))
|
|
|
|
return 0;
|
|
|
|
/* signature lower, pkey */
|
|
|
|
if (memcmp(addr + 7, broadcast + 7, 3))
|
|
|
|
return 0;
|
|
|
|
return 1;
|
|
|
|
}
|
|
|
|
|
2006-11-22 15:57:56 +01:00
|
|
|
void ipoib_mcast_restart_task(struct work_struct *work)
|
2005-04-17 00:20:36 +02:00
|
|
|
{
|
2006-11-22 15:57:56 +01:00
|
|
|
struct ipoib_dev_priv *priv =
|
|
|
|
container_of(work, struct ipoib_dev_priv, restart_task);
|
|
|
|
struct net_device *dev = priv->dev;
|
2010-04-01 23:22:57 +02:00
|
|
|
struct netdev_hw_addr *ha;
|
2005-04-17 00:20:36 +02:00
|
|
|
struct ipoib_mcast *mcast, *tmcast;
|
|
|
|
LIST_HEAD(remove_list);
|
|
|
|
unsigned long flags;
|
IPoIB: Allow setting policy to ignore multicast groups
The kernel IB stack allows (through the RDMA CM) userspace
applications to join and use multicast groups from the IPoIB MGID
range. This allows multicast traffic to be handled directly from
userspace QPs, without going through the kernel stack, which gives
better performance for some applications.
However, to fully interoperate with IP multicast, such userspace
applications need to participate in IGMP reports and queries, or else
routers may not forward the multicast traffic to the system where the
application is running. The simplest way to do this is to share the
kernel IGMP implementation by using the IP_ADD_MEMBERSHIP option to
join multicast groups that are being handled directly in userspace.
However, in such cases, the actual multicast traffic should not also
be handled by the IPoIB interface, because that would burn resources
handling multicast packets that will just be discarded in the kernel.
To handle this, this patch adds lookup on the database used for IB
multicast group reference counting when IPoIB is joining multicast
groups, and if a multicast group is already handled by user space,
then the IPoIB kernel driver ignores the group. This is controlled by
a per-interface policy flag. When the flag is set, IPoIB will not
join and attach its QP to a multicast group which already has an entry
in the database; when the flag is cleared, IPoIB will behave as before
this change.
For each IPoIB interface, the /sys/class/net/$intf/umcast attribute
controls the policy flag. The default value is off/0.
Signed-off-by: Or Gerlitz <ogerlitz@voltaire.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
2007-10-08 10:13:00 +02:00
|
|
|
struct ib_sa_mcmember_rec rec;
|
2005-04-17 00:20:36 +02:00
|
|
|
|
|
|
|
ipoib_dbg_mcast(priv, "restarting multicast task\n");
|
|
|
|
|
2005-09-20 19:52:04 +02:00
|
|
|
ipoib_mcast_stop_thread(dev, 0);
|
2005-04-17 00:20:36 +02:00
|
|
|
|
2006-06-09 21:20:56 +02:00
|
|
|
local_irq_save(flags);
|
2008-07-15 09:13:44 +02:00
|
|
|
netif_addr_lock(dev);
|
2006-01-11 20:47:34 +01:00
|
|
|
spin_lock(&priv->lock);
|
2005-04-17 00:20:36 +02:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Unfortunately, the networking core only gives us a list of all of
|
|
|
|
* the multicast hardware addresses. We need to figure out which ones
|
|
|
|
* are new and which ones have been removed
|
|
|
|
*/
|
|
|
|
|
|
|
|
/* Clear out the found flag */
|
|
|
|
list_for_each_entry(mcast, &priv->multicast_list, list)
|
|
|
|
clear_bit(IPOIB_MCAST_FLAG_FOUND, &mcast->flags);
|
|
|
|
|
|
|
|
/* Mark all of the entries that are found or don't exist */
|
2010-04-01 23:22:57 +02:00
|
|
|
netdev_for_each_mc_addr(ha, dev) {
|
2005-04-17 00:20:36 +02:00
|
|
|
union ib_gid mgid;
|
|
|
|
|
2010-04-01 23:22:57 +02:00
|
|
|
if (!ipoib_mcast_addr_is_valid(ha->addr, dev->broadcast))
|
2009-09-06 05:23:40 +02:00
|
|
|
continue;
|
|
|
|
|
2010-04-01 23:22:57 +02:00
|
|
|
memcpy(mgid.raw, ha->addr + 4, sizeof mgid);
|
2005-04-17 00:20:36 +02:00
|
|
|
|
|
|
|
mcast = __ipoib_mcast_find(dev, &mgid);
|
|
|
|
if (!mcast || test_bit(IPOIB_MCAST_FLAG_SENDONLY, &mcast->flags)) {
|
|
|
|
struct ipoib_mcast *nmcast;
|
|
|
|
|
IPoIB: Allow setting policy to ignore multicast groups
The kernel IB stack allows (through the RDMA CM) userspace
applications to join and use multicast groups from the IPoIB MGID
range. This allows multicast traffic to be handled directly from
userspace QPs, without going through the kernel stack, which gives
better performance for some applications.
However, to fully interoperate with IP multicast, such userspace
applications need to participate in IGMP reports and queries, or else
routers may not forward the multicast traffic to the system where the
application is running. The simplest way to do this is to share the
kernel IGMP implementation by using the IP_ADD_MEMBERSHIP option to
join multicast groups that are being handled directly in userspace.
However, in such cases, the actual multicast traffic should not also
be handled by the IPoIB interface, because that would burn resources
handling multicast packets that will just be discarded in the kernel.
To handle this, this patch adds lookup on the database used for IB
multicast group reference counting when IPoIB is joining multicast
groups, and if a multicast group is already handled by user space,
then the IPoIB kernel driver ignores the group. This is controlled by
a per-interface policy flag. When the flag is set, IPoIB will not
join and attach its QP to a multicast group which already has an entry
in the database; when the flag is cleared, IPoIB will behave as before
this change.
For each IPoIB interface, the /sys/class/net/$intf/umcast attribute
controls the policy flag. The default value is off/0.
Signed-off-by: Or Gerlitz <ogerlitz@voltaire.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
2007-10-08 10:13:00 +02:00
|
|
|
/* ignore group which is directly joined by userspace */
|
|
|
|
if (test_bit(IPOIB_FLAG_UMCAST, &priv->flags) &&
|
|
|
|
!ib_sa_get_mcmember_rec(priv->ca, priv->port, &mgid, &rec)) {
|
2008-10-29 20:52:50 +01:00
|
|
|
ipoib_dbg_mcast(priv, "ignoring multicast entry for mgid %pI6\n",
|
2008-10-29 06:37:22 +01:00
|
|
|
mgid.raw);
|
IPoIB: Allow setting policy to ignore multicast groups
The kernel IB stack allows (through the RDMA CM) userspace
applications to join and use multicast groups from the IPoIB MGID
range. This allows multicast traffic to be handled directly from
userspace QPs, without going through the kernel stack, which gives
better performance for some applications.
However, to fully interoperate with IP multicast, such userspace
applications need to participate in IGMP reports and queries, or else
routers may not forward the multicast traffic to the system where the
application is running. The simplest way to do this is to share the
kernel IGMP implementation by using the IP_ADD_MEMBERSHIP option to
join multicast groups that are being handled directly in userspace.
However, in such cases, the actual multicast traffic should not also
be handled by the IPoIB interface, because that would burn resources
handling multicast packets that will just be discarded in the kernel.
To handle this, this patch adds lookup on the database used for IB
multicast group reference counting when IPoIB is joining multicast
groups, and if a multicast group is already handled by user space,
then the IPoIB kernel driver ignores the group. This is controlled by
a per-interface policy flag. When the flag is set, IPoIB will not
join and attach its QP to a multicast group which already has an entry
in the database; when the flag is cleared, IPoIB will behave as before
this change.
For each IPoIB interface, the /sys/class/net/$intf/umcast attribute
controls the policy flag. The default value is off/0.
Signed-off-by: Or Gerlitz <ogerlitz@voltaire.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
2007-10-08 10:13:00 +02:00
|
|
|
continue;
|
|
|
|
}
|
|
|
|
|
2005-04-17 00:20:36 +02:00
|
|
|
/* Not found or send-only group, let's add a new entry */
|
2008-10-29 20:52:50 +01:00
|
|
|
ipoib_dbg_mcast(priv, "adding multicast entry for mgid %pI6\n",
|
2008-10-29 06:37:22 +01:00
|
|
|
mgid.raw);
|
2005-04-17 00:20:36 +02:00
|
|
|
|
|
|
|
nmcast = ipoib_mcast_alloc(dev, 0);
|
|
|
|
if (!nmcast) {
|
|
|
|
ipoib_warn(priv, "unable to allocate memory for multicast structure\n");
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
|
|
|
|
set_bit(IPOIB_MCAST_FLAG_FOUND, &nmcast->flags);
|
|
|
|
|
|
|
|
nmcast->mcmember.mgid = mgid;
|
|
|
|
|
|
|
|
if (mcast) {
|
|
|
|
/* Destroy the send only entry */
|
2006-06-26 09:24:41 +02:00
|
|
|
list_move_tail(&mcast->list, &remove_list);
|
2005-04-17 00:20:36 +02:00
|
|
|
|
|
|
|
rb_replace_node(&mcast->rb_node,
|
|
|
|
&nmcast->rb_node,
|
|
|
|
&priv->multicast_tree);
|
|
|
|
} else
|
|
|
|
__ipoib_mcast_add(dev, nmcast);
|
|
|
|
|
|
|
|
list_add_tail(&nmcast->list, &priv->multicast_list);
|
|
|
|
}
|
|
|
|
|
|
|
|
if (mcast)
|
|
|
|
set_bit(IPOIB_MCAST_FLAG_FOUND, &mcast->flags);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Remove all of the entries don't exist anymore */
|
|
|
|
list_for_each_entry_safe(mcast, tmcast, &priv->multicast_list, list) {
|
|
|
|
if (!test_bit(IPOIB_MCAST_FLAG_FOUND, &mcast->flags) &&
|
|
|
|
!test_bit(IPOIB_MCAST_FLAG_SENDONLY, &mcast->flags)) {
|
2008-10-29 20:52:50 +01:00
|
|
|
ipoib_dbg_mcast(priv, "deleting multicast group %pI6\n",
|
2008-10-29 06:37:22 +01:00
|
|
|
mcast->mcmember.mgid.raw);
|
2005-04-17 00:20:36 +02:00
|
|
|
|
|
|
|
rb_erase(&mcast->rb_node, &priv->multicast_tree);
|
|
|
|
|
|
|
|
/* Move to the remove list */
|
2006-06-26 09:24:41 +02:00
|
|
|
list_move_tail(&mcast->list, &remove_list);
|
2005-04-17 00:20:36 +02:00
|
|
|
}
|
|
|
|
}
|
2006-01-11 20:47:34 +01:00
|
|
|
|
|
|
|
spin_unlock(&priv->lock);
|
2008-07-15 09:13:44 +02:00
|
|
|
netif_addr_unlock(dev);
|
2006-06-09 21:20:56 +02:00
|
|
|
local_irq_restore(flags);
|
2005-04-17 00:20:36 +02:00
|
|
|
|
|
|
|
/* We have to cancel outside of the spinlock */
|
|
|
|
list_for_each_entry_safe(mcast, tmcast, &remove_list, list) {
|
|
|
|
ipoib_mcast_leave(mcast->dev, mcast);
|
|
|
|
ipoib_mcast_free(mcast);
|
|
|
|
}
|
|
|
|
|
|
|
|
if (test_bit(IPOIB_FLAG_ADMIN_UP, &priv->flags))
|
|
|
|
ipoib_mcast_start_thread(dev);
|
|
|
|
}
|
|
|
|
|
2005-11-03 05:51:01 +01:00
|
|
|
#ifdef CONFIG_INFINIBAND_IPOIB_DEBUG
|
|
|
|
|
2005-04-17 00:20:36 +02:00
|
|
|
struct ipoib_mcast_iter *ipoib_mcast_iter_init(struct net_device *dev)
|
|
|
|
{
|
|
|
|
struct ipoib_mcast_iter *iter;
|
|
|
|
|
|
|
|
iter = kmalloc(sizeof *iter, GFP_KERNEL);
|
|
|
|
if (!iter)
|
|
|
|
return NULL;
|
|
|
|
|
|
|
|
iter->dev = dev;
|
2005-11-07 19:33:11 +01:00
|
|
|
memset(iter->mgid.raw, 0, 16);
|
2005-04-17 00:20:36 +02:00
|
|
|
|
|
|
|
if (ipoib_mcast_iter_next(iter)) {
|
2005-11-07 19:33:11 +01:00
|
|
|
kfree(iter);
|
2005-04-17 00:20:36 +02:00
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
return iter;
|
|
|
|
}
|
|
|
|
|
|
|
|
int ipoib_mcast_iter_next(struct ipoib_mcast_iter *iter)
|
|
|
|
{
|
|
|
|
struct ipoib_dev_priv *priv = netdev_priv(iter->dev);
|
|
|
|
struct rb_node *n;
|
|
|
|
struct ipoib_mcast *mcast;
|
|
|
|
int ret = 1;
|
|
|
|
|
|
|
|
spin_lock_irq(&priv->lock);
|
|
|
|
|
|
|
|
n = rb_first(&priv->multicast_tree);
|
|
|
|
|
|
|
|
while (n) {
|
|
|
|
mcast = rb_entry(n, struct ipoib_mcast, rb_node);
|
|
|
|
|
|
|
|
if (memcmp(iter->mgid.raw, mcast->mcmember.mgid.raw,
|
|
|
|
sizeof (union ib_gid)) < 0) {
|
|
|
|
iter->mgid = mcast->mcmember.mgid;
|
|
|
|
iter->created = mcast->created;
|
|
|
|
iter->queuelen = skb_queue_len(&mcast->pkt_queue);
|
|
|
|
iter->complete = !!mcast->ah;
|
|
|
|
iter->send_only = !!(mcast->flags & (1 << IPOIB_MCAST_FLAG_SENDONLY));
|
|
|
|
|
|
|
|
ret = 0;
|
|
|
|
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
n = rb_next(n);
|
|
|
|
}
|
|
|
|
|
|
|
|
spin_unlock_irq(&priv->lock);
|
|
|
|
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
|
|
|
void ipoib_mcast_iter_read(struct ipoib_mcast_iter *iter,
|
|
|
|
union ib_gid *mgid,
|
|
|
|
unsigned long *created,
|
|
|
|
unsigned int *queuelen,
|
|
|
|
unsigned int *complete,
|
|
|
|
unsigned int *send_only)
|
|
|
|
{
|
|
|
|
*mgid = iter->mgid;
|
|
|
|
*created = iter->created;
|
|
|
|
*queuelen = iter->queuelen;
|
|
|
|
*complete = iter->complete;
|
|
|
|
*send_only = iter->send_only;
|
|
|
|
}
|
2005-11-03 05:51:01 +01:00
|
|
|
|
|
|
|
#endif /* CONFIG_INFINIBAND_IPOIB_DEBUG */
|