Skip to content
  1. Jul 20, 2012
    • Mike Marciniszyn's avatar
      IB/qib: Add congestion control agent implementation · 36a8f01c
      Mike Marciniszyn authored 
      

Add a congestion control agent in the driver that handles gets and
sets from the congestion control manager in the fabric for the
Performance Scale Messaging (PSM) library.

Signed-off-by: default avatarMike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: default avatarRoland Dreier <roland@purestorage.com>
      36a8f01c
    • Mike Marciniszyn's avatar
      IB/qib: Reduce sdma_lock contention · 551ace12
      Mike Marciniszyn authored 
      

Profiling has shown that sdma_lock is proving a bottleneck for
performance. The situations include:
 - RDMA reads when krcvqs > 1
 - post sends from multiple threads

For RDMA read the current global qib_wq mechanism runs on all CPUs
and contends for the sdma_lock when multiple RMDA read requests are
fielded on differenct CPUs. For post sends, the direct call to
qib_do_send() from multiple threads causes the contention.

Since the sdma mechanism is per port, this fix converts the existing
workqueue to a per port single thread workqueue to reduce the lock
contention in the RDMA read case, and for any other case where the QP
is scheduled via the workqueue mechanism from more than 1 CPU.

For the post send case, This patch modifies the post send code to test
for a non empty sdma engine.  If the sdma is not idle the (now single
thread) workqueue will be used to trigger the send engine instead of
the direct call to qib_do_send().

Signed-off-by: default avatarMike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: default avatarRoland Dreier <roland@purestorage.com>
      551ace12
    • Betty Dall's avatar
      IB/qib: Fix an incorrect log message · f3331f88
      Betty Dall authored 
      

There is a cut-and-paste typo in the function qib_pci_slot_reset()
where it prints that the "link_reset" function is called rather than
the "slot_reset" function.  This makes the message misleading.

Signed-off-by: default avatarBetty Dall <betty.dall@hp.com>
Signed-off-by: default avatarRoland Dreier <roland@purestorage.com>
      f3331f88
  3. Jul 18, 2012
  5. Jul 11, 2012
  7. Jul 09, 2012
    • Mike Marciniszyn's avatar
      IB/qib: RCU locking for MR validation · 8aac4cc3
      Mike Marciniszyn authored 
      

Profiling indicates that MR validation locking is expensive.  The MR
table is largely read-only and is a suitable candidate for RCU locking.

The patch uses RCU locking during validation to eliminate one
lock/unlock during that validation.

Reviewed-by: default avatarMike Heinz <michael.william.heinz@intel.com>
Signed-off-by: default avatarMike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: default avatarRoland Dreier <roland@purestorage.com>
      8aac4cc3
    • Mike Marciniszyn's avatar
      IB/qib: Avoid returning EBUSY from MR deregister · 6a82649f
      Mike Marciniszyn authored 
      

A timing issue can occur where qib_mr_dereg can return -EBUSY if the
MR use count is not zero.

This can occur if the MR is de-registered while RDMA read response
packets are being progressed from the SDMA ring.  The suspicion is
that the peer sent an RDMA read request, which has already been copied
across to the peer.  The peer sees the completion of his request and
then communicates to the responder that the MR is not needed any
longer.  The responder tries to de-register the MR, catching some
responses remaining in the SDMA ring holding the MR use count.

The code now uses a get/put paradigm to track MR use counts and
coordinates with the MR de-registration process using a completion
when the count has reached zero.  A timeout on the delay is in place
to catch other EBUSY issues.

The reference count protocol is as follows:
- The return to the user counts as 1
- A reference from the lk_table or the qib_ibdev counts as 1.
- Transient I/O operations increase/decrease as necessary

A lot of code duplication has been folded into the new routines
init_qib_mregion() and deinit_qib_mregion().  Additionally, explicit
initialization of fields to zero is now handled by kzalloc().

Also, duplicated code 'while.*num_sge' that decrements reference
counts have been consolidated in qib_put_ss().

Reviewed-by: default avatarRamkrishna Vepa <ramkrishna.vepa@intel.com>
Signed-off-by: default avatarMike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: default avatarRoland Dreier <roland@purestorage.com>
      6a82649f
    • Mike Marciniszyn's avatar
      IB/qib: Fix UC MR refs for immediate operations · 354dff1b
      Mike Marciniszyn authored 
      

An MR reference leak exists when handling UC RDMA writes with
immediate data because we manipulate the reference counts as if the
operation had been a send.

This patch moves the last_imm label so that the RDMA write operations
with immediate data converge at the cq building code.  The copy/mr
deref code is now done correctly prior to the branch to last_imm.

Reviewed-by: default avatarEdward Mascarenhas <edward.mascarenhas@intel.com>
Signed-off-by: default avatarMike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: default avatarRoland Dreier <roland@purestorage.com>
      354dff1b
  9. Jul 01, 2012
  11. Jun 30, 2012
  13. Jun 29, 2012