Commit Graph

14 Commits

Author SHA1 Message Date
Dennis Dalessandro f44728d69a IB/qib: Use rdmavt protection domain
Remove protection domain datastructure from qib and use rdmavts version.

Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Reviewed-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
2016-03-10 20:37:16 -05:00
Sagi Grimberg b8533eccc8 IB/qib: Remove old FRWR API
No ULP uses it anymore, go ahead and remove it.

Signed-off-by: Sagi Grimberg <sagig@mellanox.com>
Acked-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-10-28 22:32:29 -04:00
Sagi Grimberg 38071a461f IB/qib: Support the new memory registration API
Support the new memory registration API by allocating a
private page list array in qib_mr and populate it when
qib_map_mr_sg is invoked. Also, support IB_WR_REG_MR
by duplicating qib_fastreg_mr just take the needed information
from different places:
- page_size, iova, length (ib_mr)
- page array (qib_mr)
- key, access flags (ib_reg_wr)

The IB_WR_FAST_REG_MR handlers will be removed later when
all the ULPs will be converted.

Signed-off-by: Sagi Grimberg <sagig@mellanox.com>
Acked-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-10-28 22:27:18 -04:00
Christoph Hellwig e622f2f4ad IB: split struct ib_send_wr
This patch split up struct ib_send_wr so that all non-trivial verbs
use their own structure which embedds struct ib_send_wr.  This dramaticly
shrinks the size of a WR for most common operations:

sizeof(struct ib_send_wr) (old):	96

sizeof(struct ib_send_wr):		48
sizeof(struct ib_rdma_wr):		64
sizeof(struct ib_atomic_wr):		96
sizeof(struct ib_ud_wr):		88
sizeof(struct ib_fast_reg_wr):		88
sizeof(struct ib_bind_mw_wr):		96
sizeof(struct ib_sig_handover_wr):	80

And with Sagi's pending MR rework the fast registration WR will also be
down to a reasonable size:

sizeof(struct ib_fastreg_wr):		64

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Bart Van Assche <bart.vanassche@sandisk.com> [srp, srpt]
Reviewed-by: Chuck Lever <chuck.lever@oracle.com> [sunrpc]
Tested-by: Haggai Eran <haggaie@mellanox.com>
Tested-by: Sagi Grimberg <sagig@mellanox.com>
Tested-by: Steve Wise <swise@opengridcomputing.com>
2015-10-08 11:09:10 +01:00
Mike Marciniszyn d6f1c17e16 IB/qib: Change lkey table allocation to support more MRs
The lkey table is allocated with with a get_user_pages() with an
order based on a number of index bits from a module parameter.

The underlying kernel code cannot allocate that many contiguous pages.

There is no reason the underlying memory needs to be physically
contiguous.

This patch:
- switches the allocation/deallocation to vmalloc/vfree
- caps the number of bits to 23 to insure at least 1 generation bit
  o this matches the module parameter description

Cc: stable@vger.kernel.org
Reviewed-by: Vinit Agnihotri <vinit.abhay.agnihotri@intel.com>
Signed-off-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-08-28 22:54:46 -04:00
Andreea-Cristina Bernat 590c3fec2d IB/qib: Replace rcu_assign_pointer() with RCU_INIT_POINTER() in qib_keys.c
The uses of "rcu_assign_pointer()" are NULLing out the pointers.
According to RCU_INIT_POINTER()'s block comment:
"1.   This use of RCU_INIT_POINTER() is NULLing out the pointer"
it is better to use it instead of rcu_assign_pointer() because it has a
smaller overhead.

The following Coccinelle semantic patch was used:
@@
@@

- rcu_assign_pointer
+ RCU_INIT_POINTER
  (..., NULL)

[Derived from http://marc.info/?l=linux-rdma&m=140836519219236&w=2]

Tested-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Andreea-Cristina Bernat <bernat.ada@gmail.com>
Signed-off-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
2015-02-17 10:26:19 -08:00
Mike Marciniszyn f3bdf34465 IB/qib: Fix lockdep splat in qib_alloc_lkey()
The following backtrace is reported with CONFIG_PROVE_RCU:

    drivers/infiniband/hw/qib/qib_keys.c:64 suspicious rcu_dereference_check() usage!
    other info that might help us debug this:
    rcu_scheduler_active = 1, debug_locks = 1
    4 locks held by kworker/0:1/56:
    #0:  (events){.+.+.+}, at: [<ffffffff8107a4f5>] process_one_work+0x165/0x4a0
    #1:  ((&wfc.work)){+.+.+.}, at: [<ffffffff8107a4f5>] process_one_work+0x165/0x4a0
    #2:  (device_mutex){+.+.+.}, at: [<ffffffffa0148dd8>] ib_register_device+0x38/0x220 [ib_core]
    #3:  (&(&dev->lk_table.lock)->rlock){......}, at: [<ffffffffa017e81c>] qib_alloc_lkey+0x3c/0x1b0 [ib_qib]

    stack backtrace:
    Pid: 56, comm: kworker/0:1 Not tainted 3.10.0-rc1+ #6
    Call Trace:
    [<ffffffff810c0b85>] lockdep_rcu_suspicious+0xe5/0x130
    [<ffffffffa017e8e1>] qib_alloc_lkey+0x101/0x1b0 [ib_qib]
    [<ffffffffa0184886>] qib_get_dma_mr+0xa6/0xd0 [ib_qib]
    [<ffffffffa01461aa>] ib_get_dma_mr+0x1a/0x50 [ib_core]
    [<ffffffffa01678dc>] ib_mad_port_open+0x12c/0x390 [ib_mad]
    [<ffffffff810c2c55>] ?  trace_hardirqs_on_caller+0x105/0x190
    [<ffffffffa0167b92>] ib_mad_init_device+0x52/0x110 [ib_mad]
    [<ffffffffa01917c0>] ?  sl2vl_attr_show+0x30/0x30 [ib_qib]
    [<ffffffffa0148f49>] ib_register_device+0x1a9/0x220 [ib_core]
    [<ffffffffa01b1685>] qib_register_ib_device+0x735/0xa40 [ib_qib]
    [<ffffffff8106ba98>] ? mod_timer+0x118/0x220
    [<ffffffffa017d425>] qib_init_one+0x1e5/0x400 [ib_qib]
    [<ffffffff812ce86e>] local_pci_probe+0x4e/0x90
    [<ffffffff81078118>] work_for_cpu_fn+0x18/0x30
    [<ffffffff8107a566>] process_one_work+0x1d6/0x4a0
    [<ffffffff8107a4f5>] ?  process_one_work+0x165/0x4a0
    [<ffffffff8107c9c9>] worker_thread+0x119/0x370
    [<ffffffff8107c8b0>] ?  manage_workers+0x180/0x180
    [<ffffffff8108294e>] kthread+0xee/0x100
    [<ffffffff81082860>] ?  __init_kthread_worker+0x70/0x70
    [<ffffffff815c04ac>] ret_from_fork+0x7c/0xb0
    [<ffffffff81082860>] ?  __init_kthread_worker+0x70/0x70

Per Documentation/RCU/lockdep-splat.txt, the code now uses rcu_access_pointer()
vs. rcu_dereference().

Reported-by: Jay Fenlason <fenlason@redhat.com>
Reviewed-by: Dean Luick <dean.luick@intel.com>
Signed-off-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
2013-06-04 17:05:20 -07:00
Mike Marciniszyn c00aaa1a02 IB/qib: Fix local access validation for user MRs
Commit 8aac4cc3a9 ("IB/qib: RCU locking for MR validation") introduced
a bug that broke user post sends.  The proper validation of the MR
was lost in the patch.

This patch corrects that validation.

Reviewed-by: Dean Luick <dean.luick@intel.com>
Signed-off-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Roland Dreier <roland@purestorage.com>
2012-10-01 09:59:26 -07:00
Mike Marciniszyn 7e23017704 IB/qib: Fix sparse RCU warnings in qib_keys.c
Commit 8aac4cc3a9 ("IB/qib: RCU locking for MR validation") introduced
new sparse warnings in qib_keys.c.

Acked-by: Fengguang Wu <fengguang.wu@intel.com>
Signed-off-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
2012-07-10 10:01:56 -07:00
Mike Marciniszyn 8aac4cc3a9 IB/qib: RCU locking for MR validation
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: Mike Heinz <michael.william.heinz@intel.com>
Signed-off-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
2012-07-08 18:05:19 -07:00
Mike Marciniszyn 6a82649f21 IB/qib: Avoid returning EBUSY from MR deregister
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: Ramkrishna Vepa <ramkrishna.vepa@intel.com>
Signed-off-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
2012-07-08 18:05:19 -07:00
Mike Marciniszyn 4db62d4786 IB/qib: Fix refcount leak in lkey/rkey validation
The mr optimization introduced a reference count leak on an exception
test.  The lock/refcount manipulation is moved down and the problematic
exception test now calls bail to insure that the lock is released.

Additional fixes as suggested by Ralph Campbell <ralph.campbell@qlogic.org>:
- reduce lock scope of dma regions
- use explicit values on returns vs. automatic ret value

Signed-off-by: Mike Marciniszyn <mike.marciniszyn@qlogic.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
2011-01-10 17:42:23 -08:00
Mike Marciniszyn 2a600f14d2 IB/qib: RDMA lkey/rkey validation is inefficient for large MRs
The current code loops during rkey/lkey validiation to isolate the MR
for the RDMA, which is expensive when the current operation is inside
a very large memory region.

This fix optimizes rkey/lkey validation routines for user memory
regions and fast memory regions.  The MR entry can be isolated by
shifts/mods instead of looping.  The existing loop is preserved for
phys memory regions for now.

Signed-off-by: Mike Marciniszyn <mike.marciniszyn@qlogic.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
2011-01-10 17:42:22 -08:00
Ralph Campbell f931551baf IB/qib: Add new qib driver for QLogic PCIe InfiniBand adapters
Add a low-level IB driver for QLogic PCIe adapters.

Signed-off-by: Ralph Campbell <ralph.campbell@qlogic.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
2010-05-23 21:44:54 -07:00