This is a note to let you know that I've just added the patch titled
iommu/amd: Limit the IOVA page range to the specified addresses
to the 4.14-stable tree which can be found at:
http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum…
The filename of the patch is:
iommu-amd-limit-the-iova-page-range-to-the-specified-addresses.patch
and it can be found in the queue-4.14 subdirectory.
If you, or anyone else, feels it should not be added to the stable tree,
please let <stable(a)vger.kernel.org> know about it.
>From foo@baz Mon Dec 18 13:28:59 CET 2017
From: Gary R Hook <gary.hook(a)amd.com>
Date: Fri, 3 Nov 2017 10:50:34 -0600
Subject: iommu/amd: Limit the IOVA page range to the specified addresses
From: Gary R Hook <gary.hook(a)amd.com>
[ Upstream commit b92b4fb5c14257c0e7eae291ecc1f7b1962e1699 ]
The extent of pages specified when applying a reserved region should
include up to the last page of the range, but not the page following
the range.
Signed-off-by: Gary R Hook <gary.hook(a)amd.com>
Fixes: 8d54d6c8b8f3 ('iommu/amd: Implement apply_dm_region call-back')
Signed-off-by: Alex Williamson <alex.williamson(a)redhat.com>
Signed-off-by: Sasha Levin <alexander.levin(a)verizon.com>
Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org>
---
drivers/iommu/amd_iommu.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
--- a/drivers/iommu/amd_iommu.c
+++ b/drivers/iommu/amd_iommu.c
@@ -3155,7 +3155,7 @@ static void amd_iommu_apply_resv_region(
unsigned long start, end;
start = IOVA_PFN(region->start);
- end = IOVA_PFN(region->start + region->length);
+ end = IOVA_PFN(region->start + region->length - 1);
WARN_ON_ONCE(reserve_iova(&dma_dom->iovad, start, end) == NULL);
}
Patches currently in stable-queue which might be from gary.hook(a)amd.com are
queue-4.14/iommu-amd-limit-the-iova-page-range-to-the-specified-addresses.patch
This is a note to let you know that I've just added the patch titled
icmp: don't fail on fragment reassembly time exceeded
to the 4.14-stable tree which can be found at:
http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum…
The filename of the patch is:
icmp-don-t-fail-on-fragment-reassembly-time-exceeded.patch
and it can be found in the queue-4.14 subdirectory.
If you, or anyone else, feels it should not be added to the stable tree,
please let <stable(a)vger.kernel.org> know about it.
>From foo@baz Mon Dec 18 13:29:00 CET 2017
From: Matteo Croce <mcroce(a)redhat.com>
Date: Thu, 12 Oct 2017 16:12:37 +0200
Subject: icmp: don't fail on fragment reassembly time exceeded
From: Matteo Croce <mcroce(a)redhat.com>
[ Upstream commit 258bbb1b0e594ad5f5652cb526b3c63e6a7fad3d ]
The ICMP implementation currently replies to an ICMP time exceeded message
(type 11) with an ICMP host unreachable message (type 3, code 1).
However, time exceeded messages can either represent "time to live exceeded
in transit" (code 0) or "fragment reassembly time exceeded" (code 1).
Unconditionally replying to "fragment reassembly time exceeded" with
host unreachable messages might cause unjustified connection resets
which are now easily triggered as UFO has been removed, because, in turn,
sending large buffers triggers IP fragmentation.
The issue can be easily reproduced by running a lot of UDP streams
which is likely to trigger IP fragmentation:
# start netserver in the test namespace
ip netns add test
ip netns exec test netserver
# create a VETH pair
ip link add name veth0 type veth peer name veth0 netns test
ip link set veth0 up
ip -n test link set veth0 up
for i in $(seq 20 29); do
# assign addresses to both ends
ip addr add dev veth0 192.168.$i.1/24
ip -n test addr add dev veth0 192.168.$i.2/24
# start the traffic
netperf -L 192.168.$i.1 -H 192.168.$i.2 -t UDP_STREAM -l 0 &
done
# wait
send_data: data send error: No route to host (errno 113)
netperf: send_omni: send_data failed: No route to host
We need to differentiate instead: if fragment reassembly time exceeded
is reported, we need to silently drop the packet,
if time to live exceeded is reported, maintain the current behaviour.
In both cases increment the related error count "icmpInTimeExcds".
While at it, fix a typo in a comment, and convert the if statement
into a switch to mate it more readable.
Signed-off-by: Matteo Croce <mcroce(a)redhat.com>
Signed-off-by: David S. Miller <davem(a)davemloft.net>
Signed-off-by: Sasha Levin <alexander.levin(a)verizon.com>
Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org>
---
net/ipv4/icmp.c | 15 ++++++++++++---
1 file changed, 12 insertions(+), 3 deletions(-)
--- a/net/ipv4/icmp.c
+++ b/net/ipv4/icmp.c
@@ -782,7 +782,7 @@ static bool icmp_tag_validation(int prot
}
/*
- * Handle ICMP_DEST_UNREACH, ICMP_TIME_EXCEED, ICMP_QUENCH, and
+ * Handle ICMP_DEST_UNREACH, ICMP_TIME_EXCEEDED, ICMP_QUENCH, and
* ICMP_PARAMETERPROB.
*/
@@ -810,7 +810,8 @@ static bool icmp_unreach(struct sk_buff
if (iph->ihl < 5) /* Mangled header, drop. */
goto out_err;
- if (icmph->type == ICMP_DEST_UNREACH) {
+ switch (icmph->type) {
+ case ICMP_DEST_UNREACH:
switch (icmph->code & 15) {
case ICMP_NET_UNREACH:
case ICMP_HOST_UNREACH:
@@ -846,8 +847,16 @@ static bool icmp_unreach(struct sk_buff
}
if (icmph->code > NR_ICMP_UNREACH)
goto out;
- } else if (icmph->type == ICMP_PARAMETERPROB)
+ break;
+ case ICMP_PARAMETERPROB:
info = ntohl(icmph->un.gateway) >> 24;
+ break;
+ case ICMP_TIME_EXCEEDED:
+ __ICMP_INC_STATS(net, ICMP_MIB_INTIMEEXCDS);
+ if (icmph->code == ICMP_EXC_FRAGTIME)
+ goto out;
+ break;
+ }
/*
* Throw it at our lower layers
Patches currently in stable-queue which might be from mcroce(a)redhat.com are
queue-4.14/icmp-don-t-fail-on-fragment-reassembly-time-exceeded.patch
This is a note to let you know that I've just added the patch titled
IB/mlx4: Fix RSS's QPC attributes assignments
to the 4.14-stable tree which can be found at:
http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum…
The filename of the patch is:
ib-mlx4-fix-rss-s-qpc-attributes-assignments.patch
and it can be found in the queue-4.14 subdirectory.
If you, or anyone else, feels it should not be added to the stable tree,
please let <stable(a)vger.kernel.org> know about it.
>From foo@baz Mon Dec 18 13:28:59 CET 2017
From: Guy Levi <guyle(a)mellanox.com>
Date: Wed, 25 Oct 2017 22:39:35 +0300
Subject: IB/mlx4: Fix RSS's QPC attributes assignments
From: Guy Levi <guyle(a)mellanox.com>
[ Upstream commit 108809a0571cd1e1b317c5c083a371e163e1f8f9 ]
In the modify QP handler the base_qpn_udp field in the RSS QPC is
overwrite later by irrelevant value assignment. Hence, ingress packets
which gets to the RSS QP will be steered then to a garbage QPN.
The patch fixes this by skipping the above assignment when a RSS QP is
modified, also, the RSS context's attributes assignments are relocated
just before the context is posted to avoid future issues like this.
Additionally, this patch takes the opportunity to change the code to be
disciplined to the device's manual and assigns the RSS QP context just at
RESET to INIT transition.
Fixes:3078f5f1bd8b ("IB/mlx4: Add support for RSS QP")
Signed-off-by: Guy Levi <guyle(a)mellanox.com>
Reviewed-by: Yishai Hadas <yishaih(a)mellanox.com>
Signed-off-by: Leon Romanovsky <leon(a)kernel.org>
Signed-off-by: Doug Ledford <dledford(a)redhat.com>
Signed-off-by: Sasha Levin <alexander.levin(a)verizon.com>
Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org>
---
drivers/infiniband/hw/mlx4/qp.c | 16 +++++++++-------
1 file changed, 9 insertions(+), 7 deletions(-)
--- a/drivers/infiniband/hw/mlx4/qp.c
+++ b/drivers/infiniband/hw/mlx4/qp.c
@@ -2182,11 +2182,6 @@ static int __mlx4_ib_modify_qp(void *src
context->flags = cpu_to_be32((to_mlx4_state(new_state) << 28) |
(to_mlx4_st(dev, qp->mlx4_ib_qp_type) << 16));
- if (rwq_ind_tbl) {
- fill_qp_rss_context(context, qp);
- context->flags |= cpu_to_be32(1 << MLX4_RSS_QPC_FLAG_OFFSET);
- }
-
if (!(attr_mask & IB_QP_PATH_MIG_STATE))
context->flags |= cpu_to_be32(MLX4_QP_PM_MIGRATED << 11);
else {
@@ -2387,6 +2382,7 @@ static int __mlx4_ib_modify_qp(void *src
context->pd = cpu_to_be32(pd->pdn);
if (!rwq_ind_tbl) {
+ context->params1 = cpu_to_be32(MLX4_IB_ACK_REQ_FREQ << 28);
get_cqs(qp, src_type, &send_cq, &recv_cq);
} else { /* Set dummy CQs to be compatible with HV and PRM */
send_cq = to_mcq(rwq_ind_tbl->ind_tbl[0]->cq);
@@ -2394,7 +2390,6 @@ static int __mlx4_ib_modify_qp(void *src
}
context->cqn_send = cpu_to_be32(send_cq->mcq.cqn);
context->cqn_recv = cpu_to_be32(recv_cq->mcq.cqn);
- context->params1 = cpu_to_be32(MLX4_IB_ACK_REQ_FREQ << 28);
/* Set "fast registration enabled" for all kernel QPs */
if (!ibuobject)
@@ -2513,7 +2508,7 @@ static int __mlx4_ib_modify_qp(void *src
MLX4_IB_LINK_TYPE_ETH;
if (dev->dev->caps.tunnel_offload_mode == MLX4_TUNNEL_OFFLOAD_MODE_VXLAN) {
/* set QP to receive both tunneled & non-tunneled packets */
- if (!(context->flags & cpu_to_be32(1 << MLX4_RSS_QPC_FLAG_OFFSET)))
+ if (!rwq_ind_tbl)
context->srqn = cpu_to_be32(7 << 28);
}
}
@@ -2562,6 +2557,13 @@ static int __mlx4_ib_modify_qp(void *src
}
}
+ if (rwq_ind_tbl &&
+ cur_state == IB_QPS_RESET &&
+ new_state == IB_QPS_INIT) {
+ fill_qp_rss_context(context, qp);
+ context->flags |= cpu_to_be32(1 << MLX4_RSS_QPC_FLAG_OFFSET);
+ }
+
err = mlx4_qp_modify(dev->dev, &qp->mtt, to_mlx4_state(cur_state),
to_mlx4_state(new_state), context, optpar,
sqd_event, &qp->mqp);
Patches currently in stable-queue which might be from guyle(a)mellanox.com are
queue-4.14/ib-mlx4-fix-rss-s-qpc-attributes-assignments.patch
This is a note to let you know that I've just added the patch titled
IB/ipoib: Grab rtnl lock on heavy flush when calling ndo_open/stop
to the 4.14-stable tree which can be found at:
http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum…
The filename of the patch is:
ib-ipoib-grab-rtnl-lock-on-heavy-flush-when-calling-ndo_open-stop.patch
and it can be found in the queue-4.14 subdirectory.
If you, or anyone else, feels it should not be added to the stable tree,
please let <stable(a)vger.kernel.org> know about it.
>From foo@baz Mon Dec 18 13:29:00 CET 2017
From: Alex Vesker <valex(a)mellanox.com>
Date: Tue, 10 Oct 2017 10:36:41 +0300
Subject: IB/ipoib: Grab rtnl lock on heavy flush when calling ndo_open/stop
From: Alex Vesker <valex(a)mellanox.com>
[ Upstream commit b4b678b06f6eef18bff44a338c01870234db0bc9 ]
When ndo_open and ndo_stop are called RTNL lock should be held.
In this specific case ipoib_ib_dev_open calls the offloaded ndo_open
which re-sets the number of TX queue assuming RTNL lock is held.
Since RTNL lock is not held, RTNL assert will fail.
Signed-off-by: Alex Vesker <valex(a)mellanox.com>
Signed-off-by: Sasha Levin <alexander.levin(a)verizon.com>
Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org>
---
drivers/infiniband/ulp/ipoib/ipoib_ib.c | 7 ++++++-
1 file changed, 6 insertions(+), 1 deletion(-)
--- a/drivers/infiniband/ulp/ipoib/ipoib_ib.c
+++ b/drivers/infiniband/ulp/ipoib/ipoib_ib.c
@@ -1203,10 +1203,15 @@ static void __ipoib_ib_dev_flush(struct
ipoib_ib_dev_down(dev);
if (level == IPOIB_FLUSH_HEAVY) {
+ rtnl_lock();
if (test_bit(IPOIB_FLAG_INITIALIZED, &priv->flags))
ipoib_ib_dev_stop(dev);
- if (ipoib_ib_dev_open(dev) != 0)
+
+ result = ipoib_ib_dev_open(dev);
+ rtnl_unlock();
+ if (result)
return;
+
if (netif_queue_stopped(dev))
netif_start_queue(dev);
}
Patches currently in stable-queue which might be from valex(a)mellanox.com are
queue-4.14/ib-ipoib-grab-rtnl-lock-on-heavy-flush-when-calling-ndo_open-stop.patch
This is a note to let you know that I've just added the patch titled
Ib/hfi1: Return actual operational VLs in port info query
to the 4.14-stable tree which can be found at:
http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum…
The filename of the patch is:
ib-hfi1-return-actual-operational-vls-in-port-info-query.patch
and it can be found in the queue-4.14 subdirectory.
If you, or anyone else, feels it should not be added to the stable tree,
please let <stable(a)vger.kernel.org> know about it.
>From foo@baz Mon Dec 18 13:28:59 CET 2017
From: Patel Jay P <jay.p.patel(a)intel.com>
Date: Mon, 23 Oct 2017 06:05:53 -0700
Subject: Ib/hfi1: Return actual operational VLs in port info query
From: Patel Jay P <jay.p.patel(a)intel.com>
[ Upstream commit 00f9203119dd2774564407c7a67b17d81916298b ]
__subn_get_opa_portinfo stores value returned by hfi1_get_ib_cfg() as
operational vls. hfi1_get_ib_cfg() returns vls_operational field in
hfi1_pportdata. The problem with this is that the value is always equal
to vls_supported field in hfi1_pportdata.
The logic to calculate operational_vls is to set value passed by FM
(in __subn_set_opa_portinfo routine). If no value is passed then
default value is stored in operational_vls.
Field actual_vls_operational is calculated on the basis of buffer
control table. Hence, modifying hfi1_get_ib_cfg() to return
actual_operational_vls when used with HFI1_IB_CFG_OP_VLS parameter
Reviewed-by: Mike Marciniszyn <mike.marciniszyn(a)intel.com>
Reviewed-by: Dennis Dalessandro <dennis.dalessandro(a)intel.com>
Signed-off-by: Patel Jay P <jay.p.patel(a)intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro(a)intel.com>
Signed-off-by: Doug Ledford <dledford(a)redhat.com>
Signed-off-by: Sasha Levin <alexander.levin(a)verizon.com>
Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org>
---
drivers/infiniband/hw/hfi1/chip.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
--- a/drivers/infiniband/hw/hfi1/chip.c
+++ b/drivers/infiniband/hw/hfi1/chip.c
@@ -9952,7 +9952,7 @@ int hfi1_get_ib_cfg(struct hfi1_pportdat
goto unimplemented;
case HFI1_IB_CFG_OP_VLS:
- val = ppd->vls_operational;
+ val = ppd->actual_vls_operational;
break;
case HFI1_IB_CFG_VL_HIGH_CAP: /* VL arb high priority table size */
val = VL_ARB_HIGH_PRIO_TABLE_SIZE;
Patches currently in stable-queue which might be from jay.p.patel(a)intel.com are
queue-4.14/ib-hfi1-return-actual-operational-vls-in-port-info-query.patch
This is a note to let you know that I've just added the patch titled
IB/hfi1: Mask out A bit from psn trace
to the 4.14-stable tree which can be found at:
http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum…
The filename of the patch is:
ib-hfi1-mask-out-a-bit-from-psn-trace.patch
and it can be found in the queue-4.14 subdirectory.
If you, or anyone else, feels it should not be added to the stable tree,
please let <stable(a)vger.kernel.org> know about it.
>From foo@baz Mon Dec 18 13:29:00 CET 2017
From: Don Hiatt <don.hiatt(a)intel.com>
Date: Mon, 9 Oct 2017 12:38:12 -0700
Subject: IB/hfi1: Mask out A bit from psn trace
From: Don Hiatt <don.hiatt(a)intel.com>
[ Upstream commit d0a2f454713a42447ee4007582c0e43c47bcf230 ]
The trace logic prior to the fixes below used to mask the
A bit from the psn. It now mistakenly displays the A bit,
which is already displayed separately.
Fix by adding the appropriate mask to the psn tracing.
Fixes: 228d2af1b723 ("IB/hfi1: Separate input/output header tracing")
Fixes: 863cf89d472f ("IB/hfi1: Add 16B trace support")
Reviewed-by: Mike Marciniszyn <mike.marciniszyn(a)intel.com>
Signed-off-by: Don Hiatt <don.hiatt(a)intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro(a)intel.com>
Signed-off-by: Doug Ledford <dledford(a)redhat.com>
Signed-off-by: Sasha Levin <alexander.levin(a)verizon.com>
Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org>
---
drivers/infiniband/hw/hfi1/trace.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
--- a/drivers/infiniband/hw/hfi1/trace.c
+++ b/drivers/infiniband/hw/hfi1/trace.c
@@ -154,7 +154,7 @@ void hfi1_trace_parse_9b_bth(struct ib_o
*opcode = ib_bth_get_opcode(ohdr);
*tver = ib_bth_get_tver(ohdr);
*pkey = ib_bth_get_pkey(ohdr);
- *psn = ib_bth_get_psn(ohdr);
+ *psn = mask_psn(ib_bth_get_psn(ohdr));
*qpn = ib_bth_get_qpn(ohdr);
}
@@ -169,7 +169,7 @@ void hfi1_trace_parse_16b_bth(struct ib_
*pad = ib_bth_get_pad(ohdr);
*se = ib_bth_get_se(ohdr);
*tver = ib_bth_get_tver(ohdr);
- *psn = ib_bth_get_psn(ohdr);
+ *psn = mask_psn(ib_bth_get_psn(ohdr));
*qpn = ib_bth_get_qpn(ohdr);
}
Patches currently in stable-queue which might be from don.hiatt(a)intel.com are
queue-4.14/ib-hfi1-mask-out-a-bit-from-psn-trace.patch
This is a note to let you know that I've just added the patch titled
IB/core: Fix use workqueue without WQ_MEM_RECLAIM
to the 4.14-stable tree which can be found at:
http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum…
The filename of the patch is:
ib-core-fix-use-workqueue-without-wq_mem_reclaim.patch
and it can be found in the queue-4.14 subdirectory.
If you, or anyone else, feels it should not be added to the stable tree,
please let <stable(a)vger.kernel.org> know about it.
>From foo@baz Mon Dec 18 13:29:00 CET 2017
From: Parav Pandit <parav(a)mellanox.com>
Date: Mon, 16 Oct 2017 08:45:15 +0300
Subject: IB/core: Fix use workqueue without WQ_MEM_RECLAIM
From: Parav Pandit <parav(a)mellanox.com>
[ Upstream commit 39baf10310e6669564a485b55267fae70a4e44ae ]
The IB/core provides address resolution service and invokes callback
handler when address resolve request completes of requester in worker
thread context.
Such caller might allocate or free memory in callback handler
depending on the completion status to make further progress or to
terminate a connection. Most ULPs resolve route which involves
allocating route entry and path record elements in callback event handler.
It has been noticed that WQ_MEM_RECLAIM flag should not be used for
workers that tend to allocate memory in this [1] thread discussion.
In order to mitigate this situation, WQ_MEM_RECLAIM flag was dropped for
other such WQs in this [2] patch.
Similar problem might arise with address resolution path, though its not
yet noticed. The ib_addr workqueue is not memory reclaim path due to its
nature of invoking callback that might allocate memory or don't free any
memory under memory pressure.
[1] https://www.spinics.net/lists/linux-rdma/msg53239.html
[2] https://www.spinics.net/lists/linux-rdma/msg53416.html
Fixes: f54816261c2b ("IB/addr: Remove deprecated create_singlethread_workqueue")
Fixes: 5fff41e1f89d ("IB/core: Fix race condition in resolving IP to MAC")
Signed-off-by: Parav Pandit <parav(a)mellanox.com>
Reviewed-by: Daniel Jurgens <danielj(a)mellanox.com>
Signed-off-by: Leon Romanovsky <leon(a)kernel.org>
Signed-off-by: Doug Ledford <dledford(a)redhat.com>
Signed-off-by: Sasha Levin <alexander.levin(a)verizon.com>
Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org>
---
drivers/infiniband/core/addr.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
--- a/drivers/infiniband/core/addr.c
+++ b/drivers/infiniband/core/addr.c
@@ -852,7 +852,7 @@ static struct notifier_block nb = {
int addr_init(void)
{
- addr_wq = alloc_ordered_workqueue("ib_addr", WQ_MEM_RECLAIM);
+ addr_wq = alloc_ordered_workqueue("ib_addr", 0);
if (!addr_wq)
return -ENOMEM;
Patches currently in stable-queue which might be from parav(a)mellanox.com are
queue-4.14/ib-core-don-t-enforce-pkey-security-on-smi-mads.patch
queue-4.14/ib-core-bound-check-alternate-path-port-number.patch
queue-4.14/ib-core-fix-calculation-of-maximum-roce-mtu.patch
queue-4.14/ib-core-fix-use-workqueue-without-wq_mem_reclaim.patch
This is a note to let you know that I've just added the patch titled
IB/core: Fix endianness annotation in rdma_is_multicast_addr()
to the 4.14-stable tree which can be found at:
http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum…
The filename of the patch is:
ib-core-fix-endianness-annotation-in-rdma_is_multicast_addr.patch
and it can be found in the queue-4.14 subdirectory.
If you, or anyone else, feels it should not be added to the stable tree,
please let <stable(a)vger.kernel.org> know about it.
>From foo@baz Mon Dec 18 13:29:00 CET 2017
From: Bart Van Assche <bart.vanassche(a)wdc.com>
Date: Wed, 11 Oct 2017 10:48:43 -0700
Subject: IB/core: Fix endianness annotation in rdma_is_multicast_addr()
From: Bart Van Assche <bart.vanassche(a)wdc.com>
[ Upstream commit 1c3aea2bc8f0b2e5b57375ead40457ff75a3a2ec ]
Since ipv4_addr is a big endian 32-bit number, annotate it as such.
Fixes: commit be1d325a3358 ("IB/core: Set RoCEv2 MGID according to spec")
Signed-off-by: Bart Van Assche <bart.vanassche(a)wdc.com>
Reviewed-by: Leon Romanovsky <leonro(a)mellanox.com>
Signed-off-by: Doug Ledford <dledford(a)redhat.com>
Signed-off-by: Sasha Levin <alexander.levin(a)verizon.com>
Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org>
---
include/rdma/ib_addr.h | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
--- a/include/rdma/ib_addr.h
+++ b/include/rdma/ib_addr.h
@@ -306,12 +306,12 @@ static inline void rdma_get_ll_mac(struc
static inline int rdma_is_multicast_addr(struct in6_addr *addr)
{
- u32 ipv4_addr;
+ __be32 ipv4_addr;
if (addr->s6_addr[0] == 0xff)
return 1;
- memcpy(&ipv4_addr, addr->s6_addr + 12, 4);
+ ipv4_addr = addr->s6_addr32[3];
return (ipv6_addr_v4mapped(addr) && ipv4_is_multicast(ipv4_addr));
}
Patches currently in stable-queue which might be from bart.vanassche(a)wdc.com are
queue-4.14/ib-core-fix-endianness-annotation-in-rdma_is_multicast_addr.patch
queue-4.14/target-iscsi-detect-conn_cmd_list-corruption-early.patch
queue-4.14/target-iscsi-fix-a-race-condition-in-iscsit_add_reject_from_cmd.patch
queue-4.14/rdma-cma-avoid-triggering-undefined-behavior.patch
queue-4.14/scsi-core-fix-a-scsi_show_rq-null-pointer-dereference.patch
queue-4.14/blk-mq-sched-dispatch-from-scheduler-iff-progress-is-made-in-dispatch.patch
This is a note to let you know that I've just added the patch titled
IB/core: Fix calculation of maximum RoCE MTU
to the 4.14-stable tree which can be found at:
http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum…
The filename of the patch is:
ib-core-fix-calculation-of-maximum-roce-mtu.patch
and it can be found in the queue-4.14 subdirectory.
If you, or anyone else, feels it should not be added to the stable tree,
please let <stable(a)vger.kernel.org> know about it.
>From foo@baz Mon Dec 18 13:29:00 CET 2017
From: Parav Pandit <parav(a)mellanox.com>
Date: Mon, 16 Oct 2017 08:45:16 +0300
Subject: IB/core: Fix calculation of maximum RoCE MTU
From: Parav Pandit <parav(a)mellanox.com>
[ Upstream commit 99260132fde7bddc6e0132ce53da94d1c9ccabcb ]
The original code only took into consideration the largest header
possible after the IB_BTH_BYTES. This was incorrect, as the largest
possible header size is the largest possible combination of headers we
might run into. The new code accounts for all possible headers in the
largest possible combination and subtracts that from the MTU to make
sure that all packets will fit on the wire.
Link: https://www.spinics.net/lists/linux-rdma/msg54558.html
Fixes: 3c86aa70bf67 ("RDMA/cm: Add RDMA CM support for IBoE devices")
Signed-off-by: Parav Pandit <parav(a)mellanox.com>
Reviewed-by: Daniel Jurgens <danielj(a)mellanox.com>
Reported-by: Roland Dreier <roland(a)purestorage.com>
Signed-off-by: Leon Romanovsky <leon(a)kernel.org>
Signed-off-by: Doug Ledford <dledford(a)redhat.com>
Signed-off-by: Sasha Levin <alexander.levin(a)verizon.com>
Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org>
---
include/rdma/ib_addr.h | 7 ++++---
include/rdma/ib_pack.h | 19 +++++++++++--------
2 files changed, 15 insertions(+), 11 deletions(-)
--- a/include/rdma/ib_addr.h
+++ b/include/rdma/ib_addr.h
@@ -245,10 +245,11 @@ static inline void rdma_addr_set_dgid(st
static inline enum ib_mtu iboe_get_mtu(int mtu)
{
/*
- * reduce IB headers from effective IBoE MTU. 28 stands for
- * atomic header which is the biggest possible header after BTH
+ * Reduce IB headers from effective IBoE MTU.
*/
- mtu = mtu - IB_GRH_BYTES - IB_BTH_BYTES - 28;
+ mtu = mtu - (IB_GRH_BYTES + IB_UDP_BYTES + IB_BTH_BYTES +
+ IB_EXT_XRC_BYTES + IB_EXT_ATOMICETH_BYTES +
+ IB_ICRC_BYTES);
if (mtu >= ib_mtu_enum_to_int(IB_MTU_4096))
return IB_MTU_4096;
--- a/include/rdma/ib_pack.h
+++ b/include/rdma/ib_pack.h
@@ -37,14 +37,17 @@
#include <uapi/linux/if_ether.h>
enum {
- IB_LRH_BYTES = 8,
- IB_ETH_BYTES = 14,
- IB_VLAN_BYTES = 4,
- IB_GRH_BYTES = 40,
- IB_IP4_BYTES = 20,
- IB_UDP_BYTES = 8,
- IB_BTH_BYTES = 12,
- IB_DETH_BYTES = 8
+ IB_LRH_BYTES = 8,
+ IB_ETH_BYTES = 14,
+ IB_VLAN_BYTES = 4,
+ IB_GRH_BYTES = 40,
+ IB_IP4_BYTES = 20,
+ IB_UDP_BYTES = 8,
+ IB_BTH_BYTES = 12,
+ IB_DETH_BYTES = 8,
+ IB_EXT_ATOMICETH_BYTES = 28,
+ IB_EXT_XRC_BYTES = 4,
+ IB_ICRC_BYTES = 4
};
struct ib_field {
Patches currently in stable-queue which might be from parav(a)mellanox.com are
queue-4.14/ib-core-don-t-enforce-pkey-security-on-smi-mads.patch
queue-4.14/ib-core-bound-check-alternate-path-port-number.patch
queue-4.14/ib-core-fix-calculation-of-maximum-roce-mtu.patch
queue-4.14/ib-core-fix-use-workqueue-without-wq_mem_reclaim.patch