[ewg] [GIT PULL] cxgb4 drivers for ofed-1.6

Tue Mar 15 11:14:29 PDT 2011

Hey Vlad,

Please pull from:

ssh://vlad@www.openfabrics.org/~swise/scm/ofed_kernel ofed_kernel_1_6

This pull will bring the cxgb4 NIC and IWARP drivers up to the latest upstream code, and enables cxgb4 in the ofa kernel 
build scripts.  I'll be adding libcxgb4 and fixing put the ofed packaging soon to finish cxgb4 inclusion in OFED-1.6.

Thanks,

Steve.

-----

commit cb075d05851e4f5db08be91652e252ac9a425074
Author: Steve Wise<swise at opengridcomputing.com>
Date:   Thu Mar 10 12:19:13 2011 -0600

     Add cxgb4 drivers.

     Signed-off-by: Steve Wise<swise at opengridcomputing.com>

commit 656f75cc4615e35936764e0060e1624c99d2648c
Author: Steve Wise<swise at opengridcomputing.com>
Date:   Thu Mar 10 12:19:13 2011 -0600

     RDMA/cxgb4: Dispatch FATAL event on EEH errors.

     This at least kicks the user mode applications that are watching
     for device events.

     Signed-off-by: Steve Wise<swise at opengridcomputing.com>

commit e4d28388515f1546bb9b1fa9d29cef38205bd1e0
Author: Steve Wise<swise at opengridcomputing.com>
Date:   Thu Mar 10 12:19:13 2011 -0600

     RDMA/cxgb4: Do CIDX_INC updates every 1/16 cq depth cqe reaps.

     This avoids the CIDX_INC overflow issue with T4A2 when running
     kernel RDMA applications.

     Signed-off-by: Steve Wise<swise at opengridcomputing.com>

commit d5758ce407995d2ff3160238669d4a9d08c0a956
Author: Steve Wise<swise at opengridcomputing.com>
Date:   Thu Mar 10 12:19:12 2011 -0600

     RDMA/cxgb4: remove db_drop_task.

     Unloading iw_cxgb4 can crash due to the unload code trying to use
     db_drop_task, which is uninitialized.  So remove this dead code.

     Signed-off-by: Steve Wise<swise at opengridcomputing.com>

commit 38a506012b6ed40960c2f4b4fc8b7d71d833ec30
Author: Steve Wise<swise at opengridcomputing.com>
Date:   Thu Mar 10 12:19:12 2011 -0600

     RDMA/cxgb4: Use ULP_MODE_TCPDDP.

     Set the ULP mode for initial RDMA connection setup to
     the proper DDP mode.  This avoids wasting some HW
     resources while in streaming mode.

     Signed-off-by: Steve Wise<swise at opengridcomputing.com>

commit 3ab285e94df20d5ffd1543d4e7a201e5116f0c9f
Author: Steve Wise<swise at opengridcomputing.com>
Date:   Thu Mar 10 12:19:11 2011 -0600

     RDMA/cxgb4: Turn on delayed ACK.

     Set the default to on.

     Signed-off-by: Steve Wise<swise at opengridcomputing.com>

commit b7fc03b673391b6490f6a2432742bfa65673a57c
Author: Steve Wise<swise at opengridcomputing.com>
Date:   Thu Mar 10 12:19:11 2011 -0600

     RDMA/cxgb4: debugfs dump_qp() updates.

     - Show whether the SQ is in onchip memory or not.
     - Dump both SQ and RQ QIDs.

     Signed-off-by: Steve Wise<swise at opengridcomputing.com>

commit 9cd44bac8b400e9108d99cbdd236f41efc5d65ee
Author: Dimitris Michailidis<dm at chelsio.com>
Date:   Fri Mar 4 15:00:37 2011 -0600

     cxgb4: fix reported state of interfaces without link

     Currently tools like ip and ifconfig report incorrect state for cxgb4
     interfaces that are up but do not have link and do so until first link
     establishment.  This is because the initial netif_carrier_off call is
     before register_netdev and it needs to be after to be fully effective.
     Fix this by moving netif_carrier_off into .ndo_open.

     Signed-off-by: Dimitris Michailidis<dm at chelsio.com>
     Signed-off-by: David S. Miller<davem at davemloft.net>

commit e671aa04a090454e680e28a1f5877d07566f4c2a
Author: Dimitris Michailidis<dm at chelsio.com>
Date:   Tue Dec 14 21:36:55 2010 +0000

     cxgb4: NUMA-aware Tx queue allocations

     Allocate Tx queue memory on the node indicated by the new
     netdev_queue_numa_node_read.

     Signed-off-by: Dimitris Michailidis<dm at chelsio.com>
     Signed-off-by: David S. Miller<davem at davemloft.net>

commit a18da08f0bd51f395248374637f46d884aa78c71
Author: Dimitris Michailidis<dm at chelsio.com>
Date:   Tue Dec 14 21:36:54 2010 +0000

     cxgb4: extend VPD parsing

     Current code parses the VPD RO section for keywords but makes static
     assumptions about the location of the section.  Remove them and parse
     the VPD to find it.

     Signed-off-by: Dimitris Michailidis<dm at chelsio.com>
     Signed-off-by: David S. Miller<davem at davemloft.net>

commit 7e3c4a969586aa3d457fdfd4015cf4db1b7fcedc
Author: Dimitris Michailidis<dm at chelsio.com>
Date:   Tue Dec 14 21:36:52 2010 +0000

     cxgb4: remove a bitmap

     The driver keeps a bitmap of the netdevs it registered so it knows what to
     unregister later.  Remove that and look at reg_state instead.

     Signed-off-by: Dimitris Michailidis<dm at chelsio.com>
     Signed-off-by: David S. Miller<davem at davemloft.net>

commit 33456f01cdc2856584dfbc20d36460a3ec52f39a
Author: Dimitris Michailidis<dm at chelsio.com>
Date:   Tue Dec 14 21:36:51 2010 +0000

     cxgb4: remove the name field from the adapter structure

     Remove a field the driver uses to keep track of the name of the first
     netdev it manages to register.  Do this by changing the registration
     loop to stop the first time it fails so the first registered device is
     trivial to tell.

     Signed-off-by: Dimitris Michailidis<dm at chelsio.com>
     Signed-off-by: David S. Miller<davem at davemloft.net>

commit 08376e9660af8938abc202772eef4a8baa2ea717
Author: Dimitris Michailidis<dm at chelsio.com>
Date:   Tue Dec 14 21:36:50 2010 +0000

     cxgb4: correct formatting of MSI-X interrupt names

     The last byte of the buffer for MSI-X names could not be used due to a
     bogus -1.  Also do not explicitly clear the last byte, snprintf will do
     the right thing.

     Signed-off-by: Dimitris Michailidis<dm at chelsio.com>
     Signed-off-by: David S. Miller<davem at davemloft.net>

commit 0f1f94f3b24f719f701751a7117d67474e553a47
Author: Dimitris Michailidis<dm at chelsio.com>
Date:   Tue Dec 14 21:36:49 2010 +0000

     cxgb4: allocate more space for MSI-X interrupt names

     Currently MSI-X names for netdevs with long names are truncated in
     /proc/interrupts due to insufficient space.  Use IFNAMSIZ to size the
     needed space.

     Signed-off-by: Dimitris Michailidis<dm at chelsio.com>
     Signed-off-by: David S. Miller<davem at davemloft.net>

commit a65a919ad24185b79fc1c5ec9196e5f02ebb4c0e
Author: Dimitris Michailidis<dm at chelsio.com>
Date:   Tue Dec 14 21:36:48 2010 +0000

     cxgb4: print port information after registering each netdev

     Print information about each port when its netdev is registered instead
     of looping separately over the ports at the end.  The bulk of this patch
     is due to indentation change.

     Signed-off-by: Dimitris Michailidis<dm at chelsio.com>
     Signed-off-by: David S. Miller<davem at davemloft.net>

commit d130bdf3c727992c862e1c97e8f9eb9c3fcdb330
Author: Dimitris Michailidis<dm at chelsio.com>
Date:   Tue Dec 14 21:36:47 2010 +0000

     cxgb4: distinguish between 1-lane KR/KX and 4-lane KR/KX/KX4 ports

     And fix the supported flags ethtool reports for the two cases.

     Signed-off-by: Dimitris Michailidis<dm at chelsio.com>
     Signed-off-by: David S. Miller<davem at davemloft.net>

commit 30dc30d4d9e240588ecd527e989c3e9f81985bca
Author: Dimitris Michailidis<dm at chelsio.com>
Date:   Tue Dec 14 21:36:46 2010 +0000

     cxgb4: set the number of queues before device registration

     The number of queues is known early, move the calls to
     netif_set_real_num_[rt]x_queues before register_netdev.

     Signed-off-by: Dimitris Michailidis<dm at chelsio.com>
     Signed-off-by: David S. Miller<davem at davemloft.net>

commit aa950568488f3d40b3ff7ab83ccbdba1dff8268d
Author: Dimitris Michailidis<dm at chelsio.com>
Date:   Tue Dec 14 21:36:45 2010 +0000

     cxgb4: do not read the clock frequency from VPD

     No need to read the clock frequency from VPD, we already get it a bit
     later from FW, after any potential adjustments.

     Signed-off-by: Dimitris Michailidis<dm at chelsio.com>
     Signed-off-by: David S. Miller<davem at davemloft.net>

commit 780f80816be51dfcc7f123a5a68c2cd637dc908a
Author: Dimitris Michailidis<dm at chelsio.com>
Date:   Tue Dec 14 21:36:44 2010 +0000

     cxgb4: enable PCIe relaxed ordering

     Enable relaxed ordering for descriptor reads and packet I/O.

     Signed-off-by: Dimitris Michailidis<dm at chelsio.com>
     Signed-off-by: David S. Miller<davem at davemloft.net>

commit 60f5ad40e49782d657a2bfb9d41f48fe5c626df2
Author: Steve Wise<swise at opengridcomputing.com>
Date:   Fri Jan 21 17:00:34 2011 +0000

     RDMA/cxgb4: Set the correct device physical function for iWARP connections

     The PF passed to FW was 0, causing PCI failures in an SR-IOV environment.

     Signed-off-by: Steve Wise<swise at opengridcomputing.com>
     Cc:<stable at kernel.org>
     Signed-off-by: Roland Dreier<roland at purestorage.com>

commit e9840a89d5d91dea2affd641f2a9014ffdb39527
Author: Steve Wise<swise at opengridcomputing.com>
Date:   Fri Jan 21 17:00:29 2011 +0000

     RDMA/cxgb4: Limit MAXBURST EQ context field to 256B

     MAXBURST cannot exceed 256B for on-chip queues.  With a 512B MAXBURST,
     we can lock up the chip.

     Signed-off-by: Steve Wise<swise at opengridcomputing.com>
     Cc:<stable at kernel.org>
     Signed-off-by: Roland Dreier<roland at purestorage.com>

commit 293f6c225dd4588cdcfdbd76cd31eb3303475271
Author: Steve Wise<swise at opengridcomputing.com>
Date:   Mon Jan 10 17:41:43 2011 -0800

     RDMA/cxgb4: Don't re-init wait object in init/fini paths

     Re-initializing the wait object in rdma_init()/rdma_fini() causes a
     timing window which can lead to a deadlock during close.  Once this
     deadlock hits, all RDMA activity over the T4 device will be stuck.

     There's no need to re-init the wait object, so remove it.

     Signed-off-by: Steve Wise<swise at opengridcomputing.com>
     Cc:<stable at kernel.org>
     Signed-off-by: Roland Dreier<rolandd at cisco.com>