<feed xmlns='http://www.w3.org/2005/Atom'>
<title>kernel/linux.git/include/linux/mlx5, branch v4.7.3</title>
<subtitle>Linux kernel stable tree (mirror)</subtitle>
<id>https://git.radix-linux.su/kernel/linux.git/atom?h=v4.7.3</id>
<link rel='self' href='https://git.radix-linux.su/kernel/linux.git/atom?h=v4.7.3'/>
<link rel='alternate' type='text/html' href='https://git.radix-linux.su/kernel/linux.git/'/>
<updated>2016-08-20T16:11:02+00:00</updated>
<entry>
<title>IB/mlx5: Fix MODIFY_QP command input structure</title>
<updated>2016-08-20T16:11:02+00:00</updated>
<author>
<name>Artemy Kovalyov</name>
<email>artemyko@mellanox.com</email>
</author>
<published>2016-06-17T12:33:31+00:00</published>
<link rel='alternate' type='text/html' href='https://git.radix-linux.su/kernel/linux.git/commit/?id=8be3949fcc62090a2ad6d77ca73db238c1866f04'/>
<id>urn:sha1:8be3949fcc62090a2ad6d77ca73db238c1866f04</id>
<content type='text'>
commit e3353c268b06236d6c40fa1714c114f21f44451c upstream.

Make MODIFY_QP command input structure compliant to specification

Fixes: e126ba97dba9 ('mlx5: Add driver for Mellanox Connect-IB adapters')
Signed-off-by: Artemy Kovalyov &lt;artemyko@mellanox.com&gt;
Signed-off-by: Leon Romanovsky &lt;leon@kernel.org&gt;
Signed-off-by: Doug Ledford &lt;dledford@redhat.com&gt;
Signed-off-by: Greg Kroah-Hartman &lt;gregkh@linuxfoundation.org&gt;

</content>
</entry>
<entry>
<title>net/mlx5: Add timeout handle to commands with callback</title>
<updated>2016-07-01T10:12:03+00:00</updated>
<author>
<name>Mohamad Haj Yahia</name>
<email>mohamad@mellanox.com</email>
</author>
<published>2016-06-30T14:34:43+00:00</published>
<link rel='alternate' type='text/html' href='https://git.radix-linux.su/kernel/linux.git/commit/?id=65ee67084589c1783a74b4a4a5db38d7264ec8b5'/>
<id>urn:sha1:65ee67084589c1783a74b4a4a5db38d7264ec8b5</id>
<content type='text'>
The current implementation does not handle timeout in case of command
with callback request, and this can lead to deadlock if the command
doesn't get fw response.
Add delayed callback timeout work before posting the command to fw.
In case of real fw command completion we will cancel the delayed work.
In case of fw command timeout the callback timeout handler will be
called and it will simulate fw completion with timeout error.

Fixes: e126ba97dba9 ('mlx5: Add driver for Mellanox Connect-IB adapters')
Signed-off-by: Mohamad Haj Yahia &lt;mohamad@mellanox.com&gt;
Signed-off-by: Saeed Mahameed &lt;saeedm@mellanox.com&gt;
Signed-off-by: David S. Miller &lt;davem@davemloft.net&gt;
</content>
</entry>
<entry>
<title>Merge tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/dledford/rdma</title>
<updated>2016-06-25T01:52:31+00:00</updated>
<author>
<name>Linus Torvalds</name>
<email>torvalds@linux-foundation.org</email>
</author>
<published>2016-06-25T01:52:31+00:00</published>
<link rel='alternate' type='text/html' href='https://git.radix-linux.su/kernel/linux.git/commit/?id=aebe9bb85e6358dc85f81533b14f5c2dfe14c8b4'/>
<id>urn:sha1:aebe9bb85e6358dc85f81533b14f5c2dfe14c8b4</id>
<content type='text'>
Pull rdma fixes from Doug Ledford:
 "This is the second batch of queued up rdma patches for this rc cycle.

  There isn't anything really major in here.  It's passed 0day,
  linux-next, and local testing across a wide variety of hardware.
  There are still a few known issues to be tracked down, but this should
  amount to the vast majority of the rdma RC fixes.

  Round two of 4.7 rc fixes:

   - A couple minor fixes to the rdma core
   - Multiple minor fixes to hfi1
   - Multiple minor fixes to mlx4/mlx4
   - A few minor fixes to i40iw"

* tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/dledford/rdma: (31 commits)
  IB/srpt: Reduce QP buffer size
  i40iw: Enable level-1 PBL for fast memory registration
  i40iw: Return correct max_fast_reg_page_list_len
  i40iw: Correct status check on i40iw_get_pble
  i40iw: Correct CQ arming
  IB/rdmavt: Correct qp_priv_alloc() return value test
  IB/hfi1: Don't zero out qp-&gt;s_ack_queue in rvt_reset_qp
  IB/hfi1: Fix deadlock with txreq allocation slow path
  IB/mlx4: Prevent cross page boundary allocation
  IB/mlx4: Fix memory leak if QP creation failed
  IB/mlx4: Verify port number in flow steering create flow
  IB/mlx4: Fix error flow when sending mads under SRIOV
  IB/mlx4: Fix the SQ size of an RC QP
  IB/mlx5: Fix wrong naming of port_rcv_data counter
  IB/mlx5: Fix post send fence logic
  IB/uverbs: Initialize ib_qp_init_attr with zeros
  IB/core: Fix false search of the IB_SA_WELL_KNOWN_GUID
  IB/core: Fix RoCE v1 multicast join logic issue
  IB/core: Fix no default GIDs when netdevice reregisters
  IB/hfi1: Send a pkey change event on driver pkey update
  ...
</content>
</entry>
<entry>
<title>IB/mlx5: Fix post send fence logic</title>
<updated>2016-06-23T14:03:57+00:00</updated>
<author>
<name>Eli Cohen</name>
<email>eli@mellanox.com</email>
</author>
<published>2016-06-22T14:27:26+00:00</published>
<link rel='alternate' type='text/html' href='https://git.radix-linux.su/kernel/linux.git/commit/?id=c9b254955b9f8814966f5dabd34c39d0e0a2b437'/>
<id>urn:sha1:c9b254955b9f8814966f5dabd34c39d0e0a2b437</id>
<content type='text'>
If the caller specified IB_SEND_FENCE in the send flags of the work
request and no previous work request stated that the successive one
should be fenced, the work request would be executed without a fence.
This could result in RDMA read or atomic operations failure due to a MR
being invalidated. Fix this by adding the mlx5 enumeration for fencing
RDMA/atomic operations and fix the logic to apply this.

Fixes: e126ba97dba9 ('mlx5: Add driver for Mellanox Connect-IB adapters')
Signed-off-by: Eli Cohen &lt;eli@mellanox.com&gt;
Signed-off-by: Leon Romanovsky &lt;leon@kernel.org&gt;
Signed-off-by: Doug Ledford &lt;dledford@redhat.com&gt;
</content>
</entry>
<entry>
<title>Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net</title>
<updated>2016-06-10T15:32:24+00:00</updated>
<author>
<name>Linus Torvalds</name>
<email>torvalds@linux-foundation.org</email>
</author>
<published>2016-06-10T15:32:24+00:00</published>
<link rel='alternate' type='text/html' href='https://git.radix-linux.su/kernel/linux.git/commit/?id=698ea54dde6768d4a96080d0fb796cb3a4eadaf8'/>
<id>urn:sha1:698ea54dde6768d4a96080d0fb796cb3a4eadaf8</id>
<content type='text'>
Pull networking fixes from David Miller:

 1) nfnetlink timestamp taken from wrong skb, fix from Florian Westphal.

 2) Revert some msleep conversions in rtlwifi as these spots are in
    atomic context, from Larry Finger.

 3) Validate that NFTA_SET_TABLE attribute is actually specified when we
    call nf_tables_getset().  From Phil Turnbull.

 4) Don't do mdio_reset in stmmac driver with spinlock held as that can
    sleep, from Vincent Palatin.

 5) sk_filter() does things other than run a BPF filter, so we should
    not elide it's call just because sk-&gt;sk_filter is NULL.  Fix from
    Eric Dumazet.

 6) Fix missing backlog updates in several packet schedulers, from Cong
    Wang.

 7) bnx2x driver should allow VLAN add/remove while the interface is
    down, from Michal Schmidt.

 8) Several RDS/TCP race fixes from Sowmini Varadhan.

 9) fq_codel scheduler doesn't return correct queue length in dumps,
    from Eric Dumazet.

10) Fix TCP stats for tail loss probe and early retransmit in ipv6, from
    Yuchung Cheng.

11) Properly initialize udp_tunnel_socket_cfg in l2tp_tunnel_create(),
    from Guillaume Nault.

12) qfq scheduler leaks SKBs if a kzalloc fails, fix from Florian
    Westphal.

13) sock_fprog passed into PACKET_FANOUT_DATA needs compat handling,
    from Willem de Bruijn.

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net: (85 commits)
  vmxnet3: segCnt can be 1 for LRO packets
  packet: compat support for sock_fprog
  stmmac: fix parameter to dwmac4_set_umac_addr()
  net/mlx5e: Fix blue flame quota logic
  net/mlx5e: Use ndo_stop explicitly at shutdown flow
  net/mlx5: E-Switch, always set mc_promisc for allmulti vports
  net/mlx5: E-Switch, Modify node guid on vf set MAC
  net/mlx5: E-Switch, Fix vport enable flow
  net/mlx5: E-Switch, Use the correct error check on returned pointers
  net/mlx5: E-Switch, Use the correct free() function
  net/mlx5: Fix E-Switch flow steering capabilities check
  net/mlx5: Fix flow steering NIC capabilities check
  net/mlx5: Fix root flow table update
  net/mlx5: Fix MLX5_CMD_OP_MAX to be defined correctly
  net/mlx5: Fix masking of reserved bits in XRCD number
  net/mlx5: Fix the size of modify QP mailbox
  mlxsw: spectrum: Don't sleep during ndo_get_phys_port_name()
  mlxsw: spectrum: Make split flow match firmware requirements
  wext: Fix 32 bit iwpriv compatibility issue with 64 bit Kernel
  cfg80211: remove get/set antenna and tx power warnings
  ...
</content>
</entry>
<entry>
<title>net/mlx5: E-Switch, Modify node guid on vf set MAC</title>
<updated>2016-06-10T05:06:26+00:00</updated>
<author>
<name>Noa Osherovich</name>
<email>noaos@mellanox.com</email>
</author>
<published>2016-06-09T21:07:37+00:00</published>
<link rel='alternate' type='text/html' href='https://git.radix-linux.su/kernel/linux.git/commit/?id=23898c763f4af6f5c80b0230b1ea788a0ce3cf73'/>
<id>urn:sha1:23898c763f4af6f5c80b0230b1ea788a0ce3cf73</id>
<content type='text'>
In RoCE, the RDMA-CM needs the node guid to establish connection
between nodes.
Today, the node guid exposed to mlx5 Ethernet VFs is zero, therefore
RDMA-CM on the VF is broken.

Whenever the administrator sets a MAC for a VF, derive the node guid
from it and set it as well in the following way:
MAC: e4:1d:2d:b3:f4:01 -&gt; node_guid: e4:1d:2d:ff:fe:b3:f4:01

Fixes: 77256579c6b43 ('net/mlx5: E-Switch, Introduce Vport...')
Signed-off-by: Noa Osherovich &lt;noaos@mellanox.com&gt;
Signed-off-by: Majd Dibbiny &lt;majd@mellanox.com&gt;
Signed-off-by: Saeed Mahameed &lt;saeedm@mellanox.com&gt;
Signed-off-by: David S. Miller &lt;davem@davemloft.net&gt;
</content>
</entry>
<entry>
<title>net/mlx5: Fix flow steering NIC capabilities check</title>
<updated>2016-06-10T05:06:25+00:00</updated>
<author>
<name>Maor Gottlieb</name>
<email>maorg@mellanox.com</email>
</author>
<published>2016-06-09T21:07:32+00:00</published>
<link rel='alternate' type='text/html' href='https://git.radix-linux.su/kernel/linux.git/commit/?id=876d634d19e41603aab91455f2c52a78a28372d5'/>
<id>urn:sha1:876d634d19e41603aab91455f2c52a78a28372d5</id>
<content type='text'>
Flow steering infrastructure is currently used only on link layer
ethernet, therefore the driver should initialize the flow steering
when the device link layer is ethernet.

In addition, add missing capability check before initializing the
namespace of NIC RX flow tables.

Fixes: 2530236303d9 ('net/mlx5_core: Flow steering tree initialization')
Signed-off-by: Maor Gottlieb &lt;maorg@mellanox.com&gt;
Signed-off-by: Saeed Mahameed &lt;saeedm@mellanox.com&gt;
Signed-off-by: David S. Miller &lt;davem@davemloft.net&gt;
</content>
</entry>
<entry>
<title>net/mlx5: Fix MLX5_CMD_OP_MAX to be defined correctly</title>
<updated>2016-06-10T05:06:25+00:00</updated>
<author>
<name>Shahar Klein</name>
<email>shahark@mellanox.com</email>
</author>
<published>2016-06-09T21:07:30+00:00</published>
<link rel='alternate' type='text/html' href='https://git.radix-linux.su/kernel/linux.git/commit/?id=86d56a1a6b7352542661d8a9463758c7f285fce3'/>
<id>urn:sha1:86d56a1a6b7352542661d8a9463758c7f285fce3</id>
<content type='text'>
Having MLX5_CMD_OP_MAX on another file causes us to repeatedly miss
accounting new commands added to the driver and hence there're no entries
for them in debugfs. To solve that, we integrate it into the commands enum
as the last entry.

Fixes: 34a40e689393 ('net/mlx5_core: Introduce modify flow table command')
Signed-off-by: Shahar Klein &lt;shahark@mellanox.com&gt;
Signed-off-by: Or Gerlitz &lt;ogerlitz@mellanox.com&gt;
Signed-off-by: Saeed Mahameed &lt;saeedm@mellanox.com&gt;
Signed-off-by: David S. Miller &lt;davem@davemloft.net&gt;
</content>
</entry>
<entry>
<title>net/mlx5: Fix the size of modify QP mailbox</title>
<updated>2016-06-10T05:06:25+00:00</updated>
<author>
<name>Majd Dibbiny</name>
<email>majd@mellanox.com</email>
</author>
<published>2016-06-09T21:07:28+00:00</published>
<link rel='alternate' type='text/html' href='https://git.radix-linux.su/kernel/linux.git/commit/?id=418f8399a8bedf376ec13eb01088f04a76ebdd6f'/>
<id>urn:sha1:418f8399a8bedf376ec13eb01088f04a76ebdd6f</id>
<content type='text'>
Add 16 reserved bytes at the end of mlx5_modify_qp_mbox_in to
match the hardware spec definition.

Fixes: e126ba97dba9 ('mlx5: Add driver for Mellanox Connect-IB adapters')
Signed-off-by: Majd Dibbiny &lt;majd@mellanox.com&gt;
Signed-off-by: Saeed Mahameed &lt;saeedm@mellanox.com&gt;
Signed-off-by: David S. Miller &lt;davem@davemloft.net&gt;
</content>
</entry>
<entry>
<title>IB/mlx5: Fix pkey_index length in the QP path record</title>
<updated>2016-06-07T14:03:50+00:00</updated>
<author>
<name>Noa Osherovich</name>
<email>noaos@mellanox.com</email>
</author>
<published>2016-06-04T12:15:36+00:00</published>
<link rel='alternate' type='text/html' href='https://git.radix-linux.su/kernel/linux.git/commit/?id=d3ae2bdeba9bad8cb95301451aeaf03ce31e82f0'/>
<id>urn:sha1:d3ae2bdeba9bad8cb95301451aeaf03ce31e82f0</id>
<content type='text'>
Pkey index fields in the QP context path record are extended to 16
bits, as required by IB spec (version 1.3).
This change affects all QP commands which include path records.

To enable this change, moved the free adaptive routing flag bit
(free_ar) to the most significant byte of the QP path record.

Fixes: e126ba97dba9e ('mlx5: Add driver for Mellanox Connect-IB ...')
Signed-off-by: Noa Osherovich &lt;noaos@mellanox.com&gt;
Reviewed-by: Jack Morgenstein &lt;jackm@dev.mellanox.co.il&gt;
Signed-off-by: Leon Romanovsky &lt;leon@kernel.org&gt;
Signed-off-by: Doug Ledford &lt;dledford@redhat.com&gt;
</content>
</entry>
</feed>
