qp->r_aflags is already protected by qp->r_lock, therefore,
test_and_clear_bit() doesn't need to be atomic. Profile
shows this function call is costly.

Change the test_and_clear_bit() call to use the non-atomic
variant.
Reviewed-by: NMike Marciniszyn <mike.marciniszyn@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>

The code no longer uses tasklets for the send engine. However it does
use a tasklet for sdma but the send routines use a workqueue now days.
Update the comments to reflect that. Make things more generic with
saying "send engine" because that is what is being referred to.
Reviewed-by: NIra Weiny <ira.weiny@intel.com>
Signed-off-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

There is a a bug in defered ack stuff that causes a race with the
destroy of a QP.

A packet causes a defered ack to be pended by putting the QP
into an rcd queue.

A return from the driver interrupt processing will process that rcd
queue of QPs and attempt to do a direct send of the ack.   At this
point no locks are held and the above QP could now be put in the reset
state in the qp destroy logic.   A refcount protects the QP while it
is in the rcd queue so it isn't going anywhere yet.

If the direct send fails to allocate a pio buffer,
hfi1_schedule_send() is called to trigger sending an ack from the
send engine. There is no state test in that code path.

The refcount is then dropped from the driver.c caller
potentially allowing the qp destroy to continue from its
refcount wait in parallel with the workqueue scheduling of the qp.

Cc: stable@vger.kernel.org
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>

This allows for adding additional pages of adaptive pio
opcode control including manufacturer specific ones.
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>

This patch adds lockdep asserts in key code paths for
insuring lock correctness.
Reviewed-by: NIra Weiny <ira.weiny@intel.com>
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>

Use common header file structs, defines, and accessors
in the drivers.  The old declarations are removed.

The repositioning of the includes allows for the removal
of hfi1_message_header and replaces its use with ib_header.

Also corrected are two issues with set_armed_to_active():
- The "packet" parameter is now a pointer as it should have been
- The etype is validated to insure that the header is correct
Reviewed-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Reviewed-by: NDon Hiatt <don.hiatt@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>

This improves readability and hides the reference count
mechanism from the client drivers.
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>

Hanging has been observed while writing a file over NFSoRDMA. Dmesg on
the server contains messages like these:

[  931.992501] svcrdma: Error -22 posting RDMA_READ
[  952.076879] svcrdma: Error -22 posting RDMA_READ
[  982.154127] svcrdma: Error -22 posting RDMA_READ
[ 1012.235884] svcrdma: Error -22 posting RDMA_READ
[ 1042.319194] svcrdma: Error -22 posting RDMA_READ

Here is why:

With the base memory management extension enabled, FRMR is used instead
of FMR. The xprtrdma server issues each RDMA read request as the following
bundle:

(1)IB_WR_REG_MR, signaled;
(2)IB_WR_RDMA_READ, signaled;
(3)IB_WR_LOCAL_INV, signaled & fencing.

These requests are signaled. In order to generate completion, the fast
register work request is processed by the hfi1 send engine after being
posted to the work queue, and the corresponding lkey is not valid until
the request is processed. However, the rdmavt driver validates lkey when
the RDMA read request is posted and thus it fails immediately with error
-EINVAL (-22).

This patch changes the work flow of local operations (fast register and
local invalidate) so that fast register work requests are always
processed immediately to ensure that the corresponding lkey is valid
when subsequent work requests are posted. Local invalidate requests are
processed immediately if fencing is not required and no previous local
invalidate request is pending.

To allow completion generation for signaled local operations that have
been processed before posting to the work queue, an internal send flag
RVT_SEND_COMPLETION_ONLY is added. The hfi1 send engine checks this flag
and only generates completion for such requests.
Reviewed-by: NMike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: NJianxin Xiong <jianxin.xiong@intel.com>
Signed-off-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

In order to support extended memory management support, add send side
processing of work requests of type IB_WR_REG_MR, IB_WR_LOCAL_INV, and
IB_WR_SEND_WITH_INV. The first two are local operations and are supported
for both RC and UC. Send with invalidate is only supported for RC because
the corresponding IB opcodes are not defined for UC.
Reviewed-by: NMike Marciniszyn <mike.marciniszyn@intel.com>
Reviewed-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NJianxin Xiong <jianxin.xiong@intel.com>
Signed-off-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

As part of enabling extended memory management support, add the processing
of the RC send with invalidate.
Reviewed-by: NMike Marciniszyn <mike.marciniszyn@intel.com>
Reviewed-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NJianxin Xiong <jianxin.xiong@intel.com>
Signed-off-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

There were multiple places where FECN/BECN processing was
being done for the different types of QPs. All of that code
was very similar, which meant that it could be pulled into
a single function used by the different QP types.

To retain the performance in the fastpath, the common code
starts with an inline function, which only calls the slow
path if the packet has any of the [FB]ECN bits set.
Reviewed-by: NMike Marciniszyn <mike.marciniszyn@intel.com>
Reviewed-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NMitko Haralanov <mitko.haralanov@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>

The TODO list for the hfi1 driver was completed during 4.6. In addition
other objections raised (which are far beyond what was in the TODO list)
have been addressed as well. It is now time to remove the driver from
staging and into the drivers/infiniband sub-tree.
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>

hrtimer functions do not guarantee serialization, so we extend the
cca_timer_lock to cover the hrtimer_forward_now() in the hrtimer
callback handler and the hrtimer_start() in process_becn(). This
prevents races between these 2 functions to update the hrtimer state
leading to problems such as:
kernel BUG at kernel/hrtimer.c:1282!
encountered during validation of the CCA feature.
Reviewed-by: NMike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: NJubin John <jubin.john@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Remove unreachable code from RC ack handling to fix an
smatch error.

Fixes: 633d2739 ("staging/rdma/hfi1: use mod_timer when appropriate")
Reported-by: NDan Carpenter <dan.carpenter@oracle.com>
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>

This allows for separately enabling pio and sdma
tracepoints to cut the volume of trace information.
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>

Implement changes recommended by the Coccinelle tool to move constant to
the right in bitwise operations

-bash-4.2$ make coccicheck MODE=report M=drivers/infiniband/hw/hfi1/

drivers/infiniband/hw/hfi1/pio.c:765:4-16: Move constant to right.
drivers/infiniband/hw/hfi1/rc.c:2503:19-29: Move constant to right.
drivers/infiniband/hw/hfi1/chip.c:9813:11-22: Move constant to right.
drivers/infiniband/hw/hfi1/chip.c:14468:29-40: Move constant to right.
Reviewed-by: NJubin John <jubin.john@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>

There is a header size counter in both the QP struture and the txreq
structure. The counter in the txreq structure is not updated properly
for RC and UC queue pairs with GRH enabled, and thus causing SDMA
send to fail. This patch fixes the RC and UC path.
Reviewed-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Reviewed-by: NDean Luick <dean.luick@intel.com>
Reviewed-by: NMike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: NJianxin Xiong <jianxin.xiong@intel.com>
Signed-off-by: NJubin John <jubin.john@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Fix the header by moving the copyright notice out of the license text
and to the top of the header. Also, update the copyright date.
Reviewed-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NJubin John <jubin.john@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Add braces on all arms of statements to fix checkpatch check:
CHECK: braces {} should be used on all arms of this statement
Reviewed-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Reviewed-by: NIra Weiny <ira.weiny@intel.com>
Reviewed-by: NMike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: NJubin John <jubin.john@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Fix code alignment to fix checkpatch check:
CHECK: Alignment should match open parenthesis
Reviewed-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Reviewed-by: NIra Weiny <ira.weiny@intel.com>
Reviewed-by: NMike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: NJubin John <jubin.john@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Convert pointer comparisons to NULL to !pointer
to fix checkpatch check:
CHECK: Comparison to NULL could be written "!pointer"
Reviewed-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Reviewed-by: NIra Weiny <ira.weiny@intel.com>
Reviewed-by: NMike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: NJubin John <jubin.john@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Remove the space after a cast to fix checkpatch check:
CHECK: No space is necessary after a cast
Reviewed-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Reviewed-by: NIra Weiny <ira.weiny@intel.com>
Reviewed-by: NMike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: NJubin John <jubin.john@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Add spaces around binary operators.

Fixes checkpatch check:
CHECK: spaces preferred around that 'x'
where x is a binary operator
Reviewed-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Reviewed-by: NIra Weiny <ira.weiny@intel.com>
Reviewed-by: NMike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: NJubin John <jubin.john@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

The change requires a new pio_busy field in the iowait structure to
track the number of outstanding pios.  The new counter together
with the sdma counter serve as the basis for a packet by packet decision
as to which egress mechanism to use.  Since packets given to different
egress mechanisms are not ordered, this scheme will preserve the order.

The iowait drain/wait mechanisms are extended for a pio case.  An
additional qp wait flag is added for the PIO drain wait case.

Currently the only pio wait is for buffers, so the no_bufs_available()
routine name is changed to pio_wait() and a third argument is passed
with one of the two pio wait flags to generalize the routine.  A module
parameter is added to hold a configurable threshold. For now, the
module parameter is zero.

A heuristic routine is added to return the func pointer of the proper
egress routine to use.

The heuristic is as follows:
- SMI always uses pio
- GSI,UD qps <= threshold use pio
- UD qps > threadhold use sdma
  o No coordination with sdma is required because order is not required
    and this qp pio count is not maintained for UD
- RC/UC ONLY packets <= threshold chose as follows:
  o If sdmas pending, use SDMA
  o Otherwise use pio and enable the pio tracking count at
    the time the pio buffer is allocated
- RC/UC ONLY packets > threshold use SDMA
  o If pio's are pending the pio_wait with the new wait flag is
    called to delay for pios to drain

The threshold is potentially reduced by the QP's mtu.

The sc_buffer_alloc() has two additional args (a callback, a void *)
which are exploited by the RC/UC cases to pass a new complete routine
and a qp *.

When the shadow ring completes the credit associated with a packet,
the new complete routine is called.  The verbs_pio_complete() will then
decrement the busy count and trigger any drain waiters in qp destroy
or reset.
Reviewed-by: NJubin John <jubin.john@intel.com>
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>

The send engine wasn't correctly handling
pre-built packets, and worse, the pointer to
a packet state's txreq wasn't initialized correctly.

To fix:
- all waiters need to save any prebuilt packets
  (smda waits already did)
- the progress routine needs to handle a QPs prebuilt packet
  and initialize the txreq pointer properly

To keep SDMA working, the dma send code needs to see if
a packet has been built already. If not the code will build
it.
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>

It is in the qp.
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>

It can be conveyed in the verbs_txreq.
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>

Instead of writing the header into a buffer then copying it into another
buffer to be sent, remove that memcpy and instead build the header directly
into the tx request that will be sent.
Reviewed-by: NMike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: NVennila Megavannan <vennila.megavannan@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>

A busy_jiffies variable is maintained and updated when rc qps are
created and deleted. busy_jiffies is a scaled value of the number
of rc qps in the device. busy_jiffies is incremented every rc qp
scaling interval. busy_jiffies is added to the rc timeout
in add_retry_timer and mod_retry_timer. The rc qp scaling interval
is selected based on extensive performance evaluation of targeted
workloads.
Reviewed-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Reviewed-by: NMike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: NVennila Megavannan <vennila.megavannan@intel.com>
Signed-off-by: NJubin John <jubin.john@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

The patch series which added a new API for the RNR timer did not include an
updated call in the loopback path. RC/UC RNR loopback would be broken
without this.
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>

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>

Add a new rnr timer to hfi1.

This allows for future optimizations having the
retry and rnr timers separate.
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>

Use new timer API to optimize maintenance of
timers during ACK processing.

When we are still expecting ACKs, mod the timer
to avoid a heavyweight delete/add. Otherwise, insure
do_rc_ack() maintains the timer as it had.
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>

Use the new timer routines.
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>

Centralize disparate timer maintenance.

This allow for central control and changes to the RC
timer handling including future optimizations.
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>

Copy the last 8 bytes of user mode RC WRITE_ONLY and WRITE_LAST
opcodes separately from the rest of the data.

It is a de-facto standard for some MPI implementations to use a
poll on the last few bytes of a verbs message to indicate that
the message has been received rather than follow the required
function method.  The driver uses the kernel memcpy routine, which
becomes "rep movsb" on modern machines.  This copy, while very
fast, does not guarantee in-order copy completion and the result
is an occasional perceived corrupted packet.  Avoid the issue by
splitting the last 8 bytes to copy from the verbs opcodes where it
matters and performing an in-order byte copy.
Reviewed-by: NMike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: NDean Luick <dean.luick@intel.com>
Signed-off-by: NJubin John <jubin.john@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

In addition to removing the modify queue pair verb from hfi1 we also
remove ancillary functions which existed only for modify queue pair and
are also already present in hfi1.
Reviewed-by: NMike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>