Signed-off-by: NDavid S. Miller <davem@davemloft.net>

ndo_start_xmit never returns it to stack, but nes_nic_send helper used it if
skb could not be queued to hardware.  Switch to bool instead.

Cc: <linux-rdma@vger.kernel.org>
Signed-off-by: NFlorian Westphal <fw@strlen.de>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

For set/query MTU port firmware commands the MTU field
is 16 bits, here I changed all the "int mtu" parameters
of the functions wrapping those firmware commands to be u16.
Signed-off-by: NSaeed Mahameed <saeedm@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

gcc finds that the i40iw_make_cm_node() function in the recently added
i40iw driver uses an uninitilized variable as an index into an array
if CONFIG_IPV6 is disabled and the driver uses IPv6 mode:

drivers/infiniband/hw/i40iw/i40iw_cm.c: In function 'i40iw_make_cm_node':
drivers/infiniband/hw/i40iw/i40iw_cm.c:2206:52: error: 'arpindex' may be used uninitialized in this function [-Werror=maybe-uninitialized]
  ether_addr_copy(cm_node->rem_mac, iwdev->arp_table[arpindex].mac_addr);

As far as I can tell, this code path can not be used because the ipv4
variable is always set with CONFIG_IPV6 is disabled, but it's better
to be sure and prevent the undefined behavior, as well as shut up
that warning in a proper way.

This adds an 'else' clause for the case we get the warning about,
causing the function to return an error in a controlled way.
To avoid adding extra mess with combined io()/#ifdef clauses,
I'm also converting the existing #ifdef into a more readable
if(IS_ENABLED()) check.
Signed-off-by: NArnd Bergmann <arnd@arndb.de>
Fixes: f27b4746 ("i40iw: add connection management code")
Acked-by: NMustafa Ismail <Mustafa.ismail@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

The previous patch that added a couple of callback functions put
the declarations inside of an #ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING,
which causes the build to fail if that option is disabled:

drivers/infiniband/hw/mlx5/main.c: In function 'mlx5_ib_add':
drivers/infiniband/hw/mlx5/main.c:2358:31: error: 'mlx5_ib_get_vf_config' undeclared (first use in this function)

This moves the four declarations below the #ifdef section so they
are always available.
Signed-off-by: NArnd Bergmann <arnd@arndb.de>
Fixes: eff901d3 ("IB/mlx5: Implement callbacks for manipulating VFs")
Reviewed-by: NLeon Romanovsky <leonro@mellanox.com>
Reviewed-by: NEli Cohen <eli@mellanox.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

move delayed ack macro definitions to common
header file t4_msg.h.
Signed-off-by: NVarun Prakash <varun@chelsio.com>
Acked-by: NHariprasad Shenai <hariprasad@chelsio.com>
Signed-off-by: NNicholas Bellinger <nab@linux-iscsi.org>

move struct ulptx_idata definition to
common header file t4_msg.h.
Signed-off-by: NVarun Prakash <varun@chelsio.com>
Acked-by: NHariprasad Shenai <hariprasad@chelsio.com>
Signed-off-by: NNicholas Bellinger <nab@linux-iscsi.org>

move struct tcp_options, struct cpl_pass_accept_req,
enum defining congestion control algorithms
and associated macros to common header file t4_msg.h
Signed-off-by: NVarun Prakash <varun@chelsio.com>
Acked-by: NHariprasad Shenai <hariprasad@chelsio.com>
Signed-off-by: NNicholas Bellinger <nab@linux-iscsi.org>

Implement the IB defined callbacks used to manipulate the policy for the
link state, set GUIDs or get statistics information. This functionality
is added into a new file that will be used to add any SRIOV related
functionality to the mlx5 IB layer.

The following callbacks have been added:

mlx5_ib_get_vf_config
mlx5_ib_set_vf_link_state
mlx5_ib_get_vf_stats
mlx5_ib_set_vf_guid

In addition, publish whether this device is based on a virtual function.

In mlx5 supported devices, virtual functions are implemented as vHCAs.
vHCAs have their own QP number space so it is possible that two vHCAs
will use a QP with the same number at the same time.
Signed-off-by: NEli Cohen <eli@mellanox.com>
Reviewed-by: NOr Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Add a vf parameter to mlx5_core_query_vport_counter so we can call it to
query counters of virtual functions. Also update current users of the
API.

PFs may call mlx5_core_query_vport_counter with other_vport set to
indicate that they are querying a virtual function. The virtual
function to be queried is given by the vf parameter. Virtual function
numbering is zero based so the first VF is 0 and so on. When a PF
queries its own function, the other_vport parameter is cleared.
Signed-off-by: NEli Cohen <eli@mellanox.com>
Reviewed-by: NOr Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Fix the condition that dictates when MAD_IFC should be used. According
to firmware specifications, MAD_IFC commands must be used only if the
ib_virt capability is off.
Signed-off-by: NEli Cohen <eli@mellanox.com>
Reviewed-by: NOr Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

This patch replaces the obsolete crypto hash interface with shash
and resolves a build failure after merge of the rdma tree
which is caused by the removal of crypto hash interface

Removing CRYPTO_ALG_ASYNC from crypto_alloc_shash(),
because it is by definition sync only
Signed-off-by: NMustafa Ismail <mustafa.ismail@intel.com>
Signed-off-by: NTatyana Nikolova <tatyana.e.nikolova@intel.com>
Acked-by: NHerbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

MAINTAINERS, Kconfig, and Makefile to build i40iw module
Signed-off-by: NFaisal Latif <faisal.latif@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Kconfig and Makefile needed to build iwarp module.

Changes since v2:
	moved from Kbuild to Makefile
Signed-off-by: NFaisal Latif <faisal.latif@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

i40iw_vf.[ch] and i40iw_virtchnl[ch] are used for virtual
channel support for iWARP VF module.

Changes since v2:
	code cleanup
Acked-by: NAnjali Singhai Jain <anjali.singhai@intel.com>
Acked-by: NShannon Nelson <shannon.nelson@intel.com>
Signed-off-by: NFaisal Latif <faisal.latif@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

i40iw_user.h and i40iw_uk.c are used by both user library as well as
kernel requests.
Acked-by: NAnjali Singhai Jain <anjali.singhai@intel.com>
Acked-by: NShannon Nelson <shannon.nelson@intel.com>
Signed-off-by: NFaisal Latif <faisal.latif@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

X722 Hardware registers defines for iWARP component.
Acked-by: NAnjali Singhai Jain <anjali.singhai@intel.com>
Acked-by: NShannon Nelson <shannon.nelson@intel.com>
Signed-off-by: NFaisal Latif <faisal.latif@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

header files for hardware accesses
Signed-off-by: NFaisal Latif <faisal.latif@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

i40iw_ctrl.c provides for hardware wqe support and cqp.

Changes since v2:
	cleanup coccinelle error reported by Julia Lawall

Changes since v1:
	reported by Christoph Hellwig's review
        -remove unnecessary casts
Acked-by: NAnjali Singhai Jain <anjali.singhai@intel.com>
Acked-by: NShannon Nelson <shannon.nelson@intel.com>
Signed-off-by: NFaisal Latif <faisal.latif@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Removei/change for port mapper code which has been moved to iwcm.
Signed-off-by: NMustafa Ismail <mustafa.ismail@intel.com>
Signed-off-by: NFaisal Latif <faisal.latif@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Now with the new iWARP port mapping service in the iwcm, it is
trivial to add cxgb3 support.
Signed-off-by: NSteve Wise <swise@opengridcomputing.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Now that most of the port mapper code been moved to iwcm, we can remove
it from iw_cxgb4.
Signed-off-by: NSteve Wise <swise@opengridcomputing.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Now that most of the port mapper code been moved to iwcm, we can
remove it from port mapper service user drivers.
Signed-off-by: NMustafa Ismail <mustafa.ismail@intel.com>
Signed-off-by: NTatyana E. Nikolova <tatyana.e.nikolova@intel.com>
Signed-off-by: NFaisal Latif <faisal.latif@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

The non-rdamvt versions of qib and hfi1 allow for a differing
heuristic to override a schedule progress in favor of a direct
call the the progress routine.

This patch adds that for both drivers and rdmavt.
Reviewed-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NMike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

If SMI AH is not destroyed before de-allocating the PD, it would result in
non-zero PD use count when de-allocating the PD, triggering a WARN_ON() at
drivers/infiniband/core/verbs.c:284 ib_dealloc_pd+0x69/0xb0 [ib_core]()
when unloading the qib driver on systems with dual-port card.

This problem has always been there in qib and was detected only after the
commit 7dd78647 ("IB/core: Make ib_dealloc_pd return void") introduced
a WARN_ON in ib_dealloc_pd() that triggers if a PD's use count is non-zero
before de-allocating the PD.

Below is the call trace from the dmesg log.

[ 7264.966129] Call Trace:
[ 7264.969652]  [<ffffffff81338470>] dump_stack+0x44/0x64
[ 7264.976181]  [<ffffffff81086bb6>] warn_slowpath_common+0x86/0xc0
[ 7264.983656]  [<ffffffff81086cfa>] warn_slowpath_null+0x1a/0x20
[ 7264.990961]  [<ffffffffa025c2d9>] ib_dealloc_pd+0x69/0xb0 [ib_core]
[ 7264.998717]  [<ffffffffa0044de8>] ib_mad_port_close+0xb8/0x120 [ib_mad]
[ 7265.006866]  [<ffffffffa0044ebf>] ib_mad_remove_device+0x6f/0xc0 [ib_mad]
[ 7265.015224]  [<ffffffffa025fc87>] ib_unregister_device+0xa7/0x140 [ib_core]
[ 7265.023738]  [<ffffffffa04b5b79>] rvt_unregister_device+0x29/0x80 [rdmavt]
[ 7265.032181]  [<ffffffffa088d2a2>] qib_unregister_ib_device+0x22/0x210 [ib_qib]
[ 7265.040993]  [<ffffffffa085f73f>] qib_remove_one+0x1f/0x250 [ib_qib]
[ 7265.048823]  [<ffffffff8137a319>] pci_device_remove+0x39/0xc0
[ 7265.055984]  [<ffffffff81466a1a>] __device_release_driver+0x9a/0x140
[ 7265.063821]  [<ffffffff81466bc8>] driver_detach+0xb8/0xc0
[ 7265.070579]  [<ffffffff81465a15>] bus_remove_driver+0x55/0xd0
[ 7265.077717]  [<ffffffff8146732c>] driver_unregister+0x2c/0x50
[ 7265.084849]  [<ffffffff813789ba>] pci_unregister_driver+0x2a/0x80
[ 7265.092366]  [<ffffffffa08921bd>] qib_ib_cleanup+0x37/0x65 [ib_qib]
[ 7265.100068]  [<ffffffff811096d0>] SyS_delete_module+0x190/0x220
[ 7265.107379]  [<ffffffff816a7bae>] entry_SYSCALL_64_fastpath+0x12/0x71
Reviewed-by: NMike Marciniszyn <mike.marciniszyn@intel.com>
Reviewed-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NHarish Chegondi <harish.chegondi@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Rdmavt adopted an smi_ah from qib which is not needed by hfi1. Move this
back to qib and get it out of the common library.
Reviewed-by: NIra Weiny <ira.weiny@intel.com>
Reviewed-by: NJubin John <jubin.john@intel.com>
Signed-off-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Qib needs to be notified when mad agents are created and freed, there is
some counter maintenance that needs to be performed. Add those callbacks at
registration time with rdmavt.
Reviewed-by: NIra Weiny <ira.weiny@intel.com>
Reviewed-by: NJubin John <jubin.john@intel.com>
Signed-off-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

This patch adds an additional lock to reduce contention on the s_lock.

This lock is used in post_send() so that the post_send is not
serialized with the send engine and other send related processing.

To do this the s_next_psn is now maintained on post_send() while
post_send() related fields are moved to a new cache line.  There is
an s_avail maintained for the post_send() to mitigate trading cache
lines with the send engine.  The lock is released/acquired around
releasing the just built packet to the egress mechanism.
Reviewed-by: NJubin John <jubin.john@intel.com>
Reviewed-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NDean Luick <dean.luick@intel.com>
Signed-off-by: NHarish Chegondi <harish.chegondi@intel.com>
Signed-off-by: NMike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: NIra Weiny <ira.weiny@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

This would avoid conflict with the functions in hfi1 that have similar
names when both qib and hfi1 drivers are configured to be built into
the kernel. This issue came up in the 0-day build report.
Reviewed-by: NMike Marciniszyn <mike.marciniszyn@intel.com>
Reviewed-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NHarish Chegondi <harish.chegondi@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Replace the timer API's to initialize a timer & then assign the callback
function by the setup_timer() API.
Signed-off-by: NHari Prasath Gujulan Elango <hgujulan@visteon.com>
Signed-off-by: NMike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

This patch is a prerequisite for adding a separate lock
for post send.

The timing of updating s_last needs to be before returning
any send completion to avoid a race between a poll cq seeing
a completion and the post send checking for a full queue.
Reviewed-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NMike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Use the new RNR timer for hfi1.

For qib, this timer doesn't exist, so exploit driver
callbacks to use the new timer as appropriate.
Reviewed-by: NJubin John <jubin.john@intel.com>
Signed-off-by: NMike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Delete code from query_port which has been moved into rvt_query_port
Create a call back function to shut down a port which may be called from
rvt_modify_port
Signed-off-by: NHarish Chegondi <harish.chegondi@intel.com>
Signed-off-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Query gid is in rdmavt, but still relies on the driver to maintain the
guid table. Add the necessary driver call back and remove the existing
verb handler.
Reviewed-by: NHarish Chegondi <harish.chegondi@intel.com>
Signed-off-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Destroy QP functionality in rdmavt will be used instead.
Remove the remove_qp function being called exclusively by destroy qp code.
Reviewed-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NHarish Chegondi <harish.chegondi@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Modify queue pair functionality in rdmavt will be used instead.
Remove ancillary functions which are being used by modify QP code.
Reviewed-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NHarish Chegondi <harish.chegondi@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Add calls to rcu_read_lock()/rcu_read_unlock() as rvt_lookup_qpn callers
must hold the rcu_read_lock before calling and keep the lock until the
returned qp is no longer in use.

Remove lookaside qp and some qp refcount atomics in the sdma send code
that is redundant with the s_dma_busy refcount, which will also stall
the state processing to the reset state.

Change the qpn hash function to hash_32 which is hash function used
in rvt_lookup_qpn. qpn_hash function would be eliminated in later patches.
Reviewed-by: NIra Weiny <ira.weiny@intel.com>
Reviewed-by: NMike Marciniszyn <mike.marciniszyn@intel.com>
Reviewed-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NHarish Chegondi <harish.chegondi@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Remove some of the unnecessary code from qib_register_ib_device.
Reviewed-by: NIra Weiny <ira.weiny@intel.com>
Reviewed-by: NHarish Chegondi <harish.chegondi@intel.com>
Signed-off-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

srq functionality is now in rdmavt. Remove it from the qib driver.
Reviewed-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Reviewed-by: NHarish Chegondi <harish.chegondi@intel.com>
Signed-off-by: NJubin John <jubin.john@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Rely on rvt_query_qp function defined in rdmavt
Reviewed-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NHarish Chegondi <harish.chegondi@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>