Use built-in i2c bit-shift bus adapter to control the
i2c busses on the chip.

Cc: Jason Gunthorpe <jgunthorpe@obsidianresearch.com>
Reviewed-by: NEaswar Hariharan <easwar.hariharan@intel.com>
Reviewed-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NDean Luick <dean.luick@intel.com>
Signed-off-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NMike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

When performing process affinity recommendations for MPI ranks, the current
algorithm doesn't take into account multiple HFI units. Also, real
cores and HT cores are not distinguished from one another. Therefore,
all HT cores are recommended to be assigned first within the local NUMA
node before recommending the assignments of cores in other NUMA nodes.
It's ideal to assign all real cores across all NUMA nodes first, then all
HT 1 cores, then all HT 2 cores, and so on to balance CPU workload. CPU
cores in other NUMA nodes could be running interrupt handlers, and this is
not taken into account.

To balance the CPU workload for user processes, the following
recommendation algorithm is used:

 For each user process that is opening a context on HFI Y:
  a) If all cores are assigned to user processes, start assignments all
	 over from the first core
  b) Assign real cores first, then HT cores (First set of HT cores on
	 all physical cores, then second set of HT cores, and, so on) in the
	 following order:

	 1. Same NUMA node as HFI Y and not running an IRQ handler
	 2. Same NUMA node as HFI Y and running an IRQ handler
	 3. Different NUMA node to HFI Y and not running an IRQ handler
	 4. Different NUMA node to HFI Y and running an IRQ handler
  c) Mark core as assigned in the global affinity structure. As user
	 processes are done, remove core assignments from global affinity
	 structure.

This implementation allows an arbitrary number of HT cores and provides
support for multiple HFIs.

This is being included in the kernel rather than user space due to the
fact that user space has no way of knowing the CPU recommendations for
contexts running as part of other jobs.
Reviewed-by: NIra Weiny <ira.weiny@intel.com>
Reviewed-by: NMitko Haralanov <mitko.haralanov@intel.com>
Reviewed-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NSebastian Sanchez <sebastian.sanchez@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Kernel receive queues oversubscribe CPU cores on multi-HFI systems.
To prevent this, the kernel receive queues are separated onto
different cores, and the SDMA engine interrupts are constrained to
a lesser number of cores.

hfi1s_on_numa_node*krcvqs is the number of CPU cores that are
reserved for kernel receive queues for all HFIs. Each HFI initializes
its kernel receive queues to one of the reserved CPU cores. If there
ends up being 0 CPU cores leftover for SDMA engines, use the same
CPU cores as receive contexts.

In addition, general and control contexts are assigned to their own
CPU core, however, both types of contexts tend to have low traffic.
To save CPU cores, collapse general and control contexts to one CPU
core for all HFI units. This change prevents SDMA engine interrupts
from wrapping around general contexts.
Reviewed-by: NDean Luick <dean.luick@intel.com>
Signed-off-by: NSebastian Sanchez <sebastian.sanchez@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

When HFI units get initialized, they each use their own mask copy for
affinity assignments. On a multi-HFI system, affinity assignments
overbook CPU cores as each HFI doesn't have knowledge of affinity
assignments for other HFI units. Therefore, some CPU cores are never
used for interrupt handlers in systems with high number of CPU cores
per NUMA node.

For multi-HFI systems, SDMA engine interrupt assignments start all over
from the first CPU in the local NUMA node after the first HFI
initialization. This change allows assignments to continue where the
last HFI unit left off.

Add global structure for affinity assignments for multiple HFIs to share
affinity mask.
Reviewed-by: NJianxin Xiong <jianxin.xiong@intel.com>
Reviewed-by: NJubin John <jubin.john@intel.com>
Reviewed-by: NMike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: NSebastian Sanchez <sebastian.sanchez@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Add sw counter to track dropped unsupported packets.
Report unsupported packets drop as the RcvError.
Reviewed-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NJakub Pawlak <jakub.pawlak@intel.com>
Signed-off-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Add per VL XmitDiscards counters to the opapmaquery
status and error response.
Reviewed-by: NDean Luick <dean.luick@intel.com>
Signed-off-by: NJakub Pawlak <jakub.pawlak@intel.com>
Signed-off-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Fix sparse errors by making sure the fast assign destinations
are host cpu typed.

For the void __iomem *, just make the field match source
data.

Fix a bug where the hw_free trace printed the pointer vs.
the dereferenced value.
Reviewed-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NMike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

The ftrace infrastructure used to evaluate the TRACE_SYSTEM
macro on every DEFINE_EVENT() macro. Now the TRACE_SYSTEM
macro only gets evaluated when trace/define_trace.h is
included, so the group event information is lost. This was
introduced in
commit acd388fd ("tracing: Give system name a pointer")
Therefore, each system tracepoint must be on its own file.
Reviewed-by: NMike Marciniszyn <mike.marciniszyn@intel.com>
Reviewed-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NSebastian Sanchez <sebastian.sanchez@intel.com>
Signed-off-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Fix a copy and paste typo in comment.
Reviewed-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NTadeusz Struk <tadeusz.struk@intel.com>
Signed-off-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Simple code clean up of hfi1_write_iter.
Reviewed-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NIra Weiny <ira.weiny@intel.com>
Signed-off-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

The definition of port state changed mid development and the
old structure was kept accidentally.  Remove this dead code.
Reviewed-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NIra Weiny <ira.weiny@intel.com>
Signed-off-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Fix to enable remote access rights when allocating stag.

Fixes: b7aee855 ("RDMA/i40iw: Add base memory management extensions")
Signed-off-by: NShiraz Saleem <shiraz.saleem@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

i40iw_create_cqp() printed the contents of variables maj_err and min_err
in an error message before they could be initialized (by calling
dev->cqp_ops->cqp_create).
Signed-off-by: NNicolas Iooss <nicolas.iooss_linux@m4x.org>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Add the missing port_xmit_wait counter. This counter is displayed through
some tools like perfquery but is not available via sysfs.

For the PORT_PMA_ATTR macro the _counter field is set to zero
allowing us to specify the offset directly like with PORT_PMA_ATTR_EXT

See also the earlier work in 2008 by Vladimir Skolovsky

https://www.mail-archive.com/general@lists.openfabrics.org/msg20313.htmlSigned-off-by: NVladimir Sokolvsky <vlad@mellanox.com>
Signed-off-by: NChristoph Lameter <cl@linux.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

The critical section should protect only the list traversal
and dd->asic_data modification, not the memory allocation.
The fix pulls the allocation out of the critical section.
Reviewed-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Reviewed-by: NSebastian Sanchez <sebastian.sanchez@intel.com>
Reviewed-by: NDean Luick <dean.luick@intel.com>
Signed-off-by: NTadeusz Struk <tadeusz.struk@intel.com>
Signed-off-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NMike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

There are several computatations of the sc in the
ud receive routine.

Besides the code duplication, all are wrong when the
sc is greater than 15.   In that case the code incorrectly
or's a 1 into the computed sc instead of 1 shifted left
by 4.

Fix precomputed sc5 by using an already implemented routine
hdr2sc() and deleting flawed duplicated code.

Cc: Stable <stable@vger.kernel.org> # 4.6+
Reviewed-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NMike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

The memory needed for the send and receive queues associated with
a QP is proportional to the max_sge parameter. The current value
of that parameter is such that with an mlx4 HCA the QP buffer size
is 8 MB. Since DMA is used for communication between HCA and CPU
that buffer either has to be allocated coherently or map_single()
must succeed for that buffer. Since large contiguous allocations
are fragile and since the maximum segment size for e.g. swiotlb
is 256 KB, reduce the max_sge parameter. This patch avoids that
the following text appears on the console after SRP logout and
relogin on a system equipped with multiple IB HCAs:

mlx4_core 0000:05:00.0: swiotlb buffer is full (sz: 8388608 bytes)
swiotlb: coherent allocation failed for device 0000:05:00.0 size=8388608
CPU: 11 PID: 148 Comm: kworker/11:1 Not tainted 4.7.0-rc4-dbg+ #1
Call Trace:
 [<ffffffff812c6d35>] dump_stack+0x67/0x92
 [<ffffffff812efe71>] swiotlb_alloc_coherent+0x141/0x150
 [<ffffffff810458be>] x86_swiotlb_alloc_coherent+0x3e/0x50
 [<ffffffffa03861fa>] mlx4_buf_direct_alloc.isra.5+0x9a/0x120 [mlx4_core]
 [<ffffffffa0386545>] mlx4_buf_alloc+0x165/0x1a0 [mlx4_core]
 [<ffffffffa035053d>] create_qp_common.isra.29+0x57d/0xff0 [mlx4_ib]
 [<ffffffffa03510da>] mlx4_ib_create_qp+0x12a/0x3f0 [mlx4_ib]
 [<ffffffffa031154a>] ib_create_qp+0x3a/0x250 [ib_core]
 [<ffffffffa055dd4b>] srpt_cm_handler+0x4bb/0xcad [ib_srpt]
 [<ffffffffa02c1ab0>] cm_process_work+0x20/0xf0 [ib_cm]
 [<ffffffffa02c3640>] cm_work_handler+0x1ac0/0x2059 [ib_cm]
 [<ffffffff810737ed>] process_one_work+0x19d/0x490
 [<ffffffff81073b29>] worker_thread+0x49/0x490
 [<ffffffff8107a0ea>] kthread+0xea/0x100
 [<ffffffff815b25af>] ret_from_fork+0x1f/0x40

Fixes: b99f8e4d ("IB/srpt: convert to the generic RDMA READ/WRITE API")
Signed-off-by: NBart Van Assche <bart.vanassche@sandisk.com>
Cc: Laurence Oberman <loberman@redhat.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Set the chunk_size to enable level-1 PBL support when the fast memory
page count is more than one.
Signed-off-by: NShiraz Saleem <shiraz.saleem@intel.com>
Signed-off-by: NFaisal Latif <faisal.latif@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Return correct value for max_fast_reg_page_list_len from
i40iw_query_device().
Signed-off-by: NFaisal Latif <faisal.latif@intel.com>
Signed-off-by: NShiraz Saleem <shiraz.saleem@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

i40iw_get_pble returns 0 on success. Correct the check on return
code.
Signed-off-by: NFaisal Latif <faisal.latif@intel.com>
Signed-off-by: NShiraz Saleem <shiraz.saleem@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

CQ is armed for solicited events only, ignoring other notification
flags. Correct this by arming for next and arming for solicited
event if IB_CQ_SOLICITED is set. Also protect CQ shadow area update
with spinlock.
Signed-off-by: NShiraz Saleem <shiraz.saleem@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

The current drivers return errors from this calldown
wrapped in an ERR_PTR().

The rdmavt code incorrectly tests for NULL.

The code is fixed to use IS_ERR() and change ret according
to the driver return value.

Cc: Stable <stable@vger.kernel.org> # 4.6+
Reviewed-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NMike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Since rvt_reset_qp already zero's out qp->s_ack_queue head and tail
pointers, there is no need to zero out qp->s_ack_queue itself.
Reviewed-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Reviewed-by: NMike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: NAshutosh Dixit <ashutosh.dixit@intel.com>
Signed-off-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

A failure in the get_txreq() inline will result in a
slow path retry using __get_txreq().

__get_txreq() attempts to procure the qp s_lock, which
is already held in all callers.

Fix by deleting the s_lock maintenance in __get_txreq()
and add sparse syntax hooks to future proof the code.

Cc: Stable <stable@vger.kernel.org> # 4.6+
Reviewed-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NMike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Prevent cross page boundary allocation by allocating
new page, this is required to be aligned with ConnectX-3 HW
requirements.

Not doing that might cause to "RDMA read local protection" error.

Fixes: 1b2cd0fc ('IB/mlx4: Support the new memory registration API')
Suggested-by: NChristoph Hellwig <hch@infradead.org>
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Reviewed-by: NSagi Grimberg <sagi@grimberg.me>
Signed-off-by: NYishai Hadas <yishaih@mellanox.com>
Signed-off-by: NLeon Romanovsky <leonro@mellanox.com>
Signed-off-by: NLeon Romanovsky <leon@kernel.org>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

When RC, UC, or RAW QPs are created, a qp object is allocated (kzalloc).
If at a later point (in procedure create_qp_common) the qp creation fails,
this qp object must be freed.

Fixes: 1ffeb2eb ("IB/mlx4: SR-IOV IB context objects and proxy/tunnel SQP support")
Signed-off-by: NDotan Barak <dotanb@dev.mellanox.co.il>
Signed-off-by: NJack Morgenstein <jackm@dev.mellanox.co.il>
Signed-off-by: NLeon Romanovsky <leon@kernel.org>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

In procedure mlx4_ib_create_flow, passing an invalid port number
will cause an out-of-bounds array access. Data passed to this procedure
can come from user-space.  Therefore, need to validate port number
before proceeding onwards.

Note that we check against the number of physical ports declared at
the verbs (ib core) level; When bonding is active, the verbs level
sees one physical port, even though the low-level driver sees two ports.

Fixes: f77c0162 ("IB/mlx4: Add receive flow steering support")
Signed-off-by: NYishai Hadas <yishaih@mellanox.com>
Reviewed-by: NJack Morgenstein <jackm@dev.mellanox.co.il>
Reviewed-by: NMoni Shoua <monis@mellanox.com>
Signed-off-by: NLeon Romanovsky <leon@kernel.org>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Fix mad send error flow to prevent double freeing address handles,
and leaking tx_ring entries when SRIOV is active.

If ib_mad_post_send fails, the address handle pointer in the tx_ring entry
must be set to NULL (or there will be a double-free) and tx_tail must be
incremented (or there will be a leak of tx_ring entries).
The tx_ring is handled the same way in the send-completion handler.

Fixes: 37bfc7c1 ("IB/mlx4: SR-IOV multiplex and demultiplex MADs")
Signed-off-by: NYishai Hadas <yishaih@mellanox.com>
Reviewed-by: NJack Morgenstein <jackm@dev.mellanox.co.il>
Signed-off-by: NLeon Romanovsky <leon@kernel.org>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

When calculating the required size of an RC QP send queue, leave
enough space for masked atomic operations, which require more space than
"regular" atomic operation.

Fixes: 6fa8f719 ("IB/mlx4: Add support for masked atomic operations")
Signed-off-by: NYishai Hadas <yishaih@mellanox.com>
Reviewed-by: NJack Morgenstein <jackm@mellanox.co.il>
Reviewed-by: NEran Ben Elisha <eranbe@mellanox.com>
Signed-off-by: NLeon Romanovsky <leon@kernel.org>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

port_xmit_data is written instead of port_rcv_data.

Fixes: 3efd9a11 ('IB/mlx5: Modify MAD reading counters method to use counter registers')
Signed-off-by: NTalat Batheesh <talatb@mellanox.com>
Signed-off-by: NLeon Romanovsky <leon@kernel.org>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

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: e126ba97 ('mlx5: Add driver for Mellanox Connect-IB adapters')
Signed-off-by: NEli Cohen <eli@mellanox.com>
Signed-off-by: NLeon Romanovsky <leon@kernel.org>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Initialize ib_qp_init_attr with zeros in order to avoid from garbage
in fields that won't be set with user values.

Fixes: a060b562 ('IB/core: generic RDMA READ/WRITE API')
Signed-off-by: NMaor Gottlieb <maorg@mellanox.com>
Signed-off-by: NLeon Romanovsky <leon@kernel.org>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

When virtualziation is supported, VFs may send SA MADs to a GID formed
by the concatenation of the subnet prefix with the
IB_SA_WELL_KNOWN_GUID. When a response is required, the current code
will search the local HCA's port for the received GID to figure out the
GID index of the entry containing this GID. However, since this is not a
real GID it will not be found and error will be printed.

We change the logic to check if the destination GID is this special GID
and avoid lookup in this case and use GID index 0.

Fixes: a0c1b2a3 ('IB/core: Support accessing SA in virtualized environment')
Signed-off-by: NEli Cohen <eli@mellanox.com>
Signed-off-by: NLeon Romanovsky <leon@kernel.org>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

During multicast join of RoCEv1, IGMP join state and max hop limit
were updated incorrectly. IGMP join should be sent and marked as
joined only on RoCEv2 after a successful join. Max hops should be
updated to the hop limit on RoCEv2 regardless of the join state.

Fixes: bee3c3c9 ('IB/cma: Join and leave multicast groups...')
Signed-off-by: NAlex Vesker <valex@mellanox.com>
Signed-off-by: NLeon Romanovsky <leon@kernel.org>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Currently, when the netdevice returned by get_netdev is unregistered,
we delete all GIDs (including the default GIDs) and reset their
attributes. Therefore, when we re-register it, no default GIDs
will be assigned (as their "default GID") attribute will be reset.
Fixing this by keeping "default GID" attribute.

Fixes: 03db3a2d ('IB/core: Add RoCE GID table management')
Signed-off-by: NTalat Batheesh <talatb@mellanox.com>
Signed-off-by: NLeon Romanovsky <leon@kernel.org>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Swapping a cable from a "Mgmt Allowed=No" switch port to a
"Mgmt Allowed=Yes" switch port doesn't send a pkey change
notification. Therefore, the link doesn't become active as
the oib_utils layer uses an old pkey table cache.

Fix by ensuring the pkey change notification is sent when
the table is changed both explicitly by the FM and implicitly
by the driver via a cable swap.
Reviewed-by: NMike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: NSebastian Sanchez <sebastian.sanchez@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

FULL_MGMT_P_KEY doesn't get cleared from the pkey table at link bounce
because the link down and link bounce code paths are different when
moving a QSFP cable on a switch. This causes an HFI unit connected to a
switch to try to be initialized to an FM node when the QSFP cable is
moved from a MgmtAllowed=NO port to a MgmtAllowed=YES port and back to a
MgmtAllowed=NO port. Remove FULL_MGMT_P_KEY from pkey table at link up.
Reviewed-by: NDean Luick <dean.luick@intel.com>
Signed-off-by: NSebastian Sanchez <sebastian.sanchez@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

This fixes potential buffer overflow because the sprintf function
doesn't check buffer boundaries. Use snprintf instead.
Reviewed-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NTadeusz Struk <tadeusz.struk@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

This fixes potential NULL ptr dereference because IS_ERR(dd) doesn't
handle NULL. Fix the issue by initializing the pointer with a not NULL
error code.
Reviewed-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NTadeusz Struk <tadeusz.struk@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

If a context has already been assigned to an FD, prevent
another assignment.
Reviewed-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NIra Weiny <ira.weiny@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>