summaryrefslogtreecommitdiff
path: root/drivers/infiniband/hw
AgeCommit message (Collapse)AuthorFilesLines
2014-04-04Merge tag 'rdma-for-linus' of ↵Linus Torvalds60-1009/+3166
git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband Pull infiniband updates from Roland Dreier: "Main batch of InfiniBand/RDMA changes for 3.15: - The biggest change is core API extensions and mlx5 low-level driver support for handling DIF/DIX-style protection information, and the addition of PI support to the iSER initiator. Target support will be arriving shortly through the SCSI target tree. - A nice simplification to the "umem" memory pinning library now that we have chained sg lists. Kudos to Yishai Hadas for realizing our code didn't have to be so crazy. - Another nice simplification to the sg wrappers used by qib, ipath and ehca to handle their mapping of memory to adapter. - The usual batch of fixes to bugs found by static checkers etc. from intrepid people like Dan Carpenter and Yann Droneaud. - A large batch of cxgb4, ocrdma, qib driver updates" * tag 'rdma-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband: (102 commits) RDMA/ocrdma: Unregister inet notifier when unloading ocrdma RDMA/ocrdma: Fix warnings about pointer <-> integer casts RDMA/ocrdma: Code clean-up RDMA/ocrdma: Display FW version RDMA/ocrdma: Query controller information RDMA/ocrdma: Support non-embedded mailbox commands RDMA/ocrdma: Handle CQ overrun error RDMA/ocrdma: Display proper value for max_mw RDMA/ocrdma: Use non-zero tag in SRQ posting RDMA/ocrdma: Memory leak fix in ocrdma_dereg_mr() RDMA/ocrdma: Increment abi version count RDMA/ocrdma: Update version string be2net: Add abi version between be2net and ocrdma RDMA/ocrdma: ABI versioning between ocrdma and be2net RDMA/ocrdma: Allow DPP QP creation RDMA/ocrdma: Read ASIC_ID register to select asic_gen RDMA/ocrdma: SQ and RQ doorbell offset clean up RDMA/ocrdma: EQ full catastrophe avoidance RDMA/cxgb4: Disable DSGL use by default RDMA/cxgb4: rx_data() needs to hold the ep mutex ...
2014-04-03Merge branches 'core', 'cxgb4', 'ip-roce', 'iser', 'misc', 'mlx4', 'nes', ↵Roland Dreier46-488/+2033
'ocrdma', 'qib', 'sgwrapper', 'srp' and 'usnic' into for-next
2014-04-03RDMA/ocrdma: Unregister inet notifier when unloading ocrdmaSelvin Xavier1-2/+14
Unregister the inet notifier during ocrdma unload to avoid a panic after driver unload. Signed-off-by: Selvin Xavier <selvin.xavier@emulex.com> Signed-off-by: Devesh Sharma <devesh.sharma@emulex.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2014-04-03RDMA/ocrdma: Fix warnings about pointer <-> integer castsRoland Dreier1-3/+4
We should cast pointers to and from unsigned long to turn them into ints. Signed-off-by: Roland Dreier <roland@purestorage.com>
2014-04-03RDMA/ocrdma: Code clean-upDevesh Sharma7-47/+24
Clean up code. Also modifying GSI QP to error during ocrdma_close is fixed. Signed-off-by: Devesh Sharma <devesh.sharma@emulex.com> Signed-off-by: Selvin Xavier <selvin.xavier@emulex.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2014-04-03RDMA/ocrdma: Display FW versionSelvin Xavier1-1/+41
Adding a sysfs file for getting the FW version. Signed-off-by: Selvin Xavier <selvin.xavier@emulex.com> Signed-off-by: Devesh Sharma <devesh.sharma@emulex.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2014-04-03RDMA/ocrdma: Query controller informationSelvin Xavier8-2/+1135
Issue mailbox commands to query ocrdma controller information and phy information and print them while adding ocrdma device. Signed-off-by: Selvin Xavier <selvin.xavier@emulex.com> Signed-off-by: Devesh Sharma <devesh.sharma@emulex.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2014-04-03RDMA/ocrdma: Support non-embedded mailbox commandsSelvin Xavier1-8/+47
Added a routine to issue non-embedded mailbox commands for handling large mailbox request/response data. Signed-off-by: Selvin Xavier <selvin.xavier@emulex.com> Signed-off-by: Devesh Sharma <devesh.sharma@emulex.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2014-04-03RDMA/ocrdma: Handle CQ overrun errorSelvin Xavier1-1/+4
Signed-off-by: Selvin Xavier <selvin.xavier@emulex.com> Signed-off-by: Devesh Sharma <devesh.sharma@emulex.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2014-04-03RDMA/ocrdma: Display proper value for max_mwSelvin Xavier3-1/+3
Signed-off-by: Selvin Xavier <selvin.xavier@emulex.com> Signed-off-by: Devesh Sharma <devesh.sharma@emulex.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2014-04-03RDMA/ocrdma: Use non-zero tag in SRQ postingSelvin Xavier1-10/+18
As part of SRQ receive buffers posting we populate a non-zero tag which will be returned in SRQ receive completions. Signed-off-by: Selvin Xavier <selvin.xavier@emulex.com> Signed-off-by: Devesh Sharma <devesh.sharma@emulex.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2014-04-03RDMA/ocrdma: Memory leak fix in ocrdma_dereg_mr()Selvin Xavier1-2/+1
Signed-off-by: Selvin Xavier <selvin.xavier@emulex.com> Signed-off-by: Devesh Sharma <devesh.sharma@emulex.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2014-04-03RDMA/ocrdma: Increment abi version countDevesh Sharma1-1/+1
Increment the ABI version count for driver/library interface. Signed-off-by: Devesh Sharma <devesh.sharma@emulex.com> Signed-off-by: Selvin Xavier <selvin.xavier@emulex.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2014-04-03RDMA/ocrdma: Update version stringDevesh Sharma2-3/+5
Update the driver vrsion string and node description string Signed-off-by: Devesh Sharma <devesh.sharma@emulex.com> Signed-off-by: Selvin Xavier <selvin.xavier@emulex.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2014-04-03RDMA/ocrdma: ABI versioning between ocrdma and be2netDevesh Sharma2-0/+2
While loading RoCE driver be2net driver should check for ABI version to catch functional incompatibilities. Signed-off-by: Devesh Sharma <devesh.sharma@emulex.com> Signed-off-by: Selvin Xavier <selvin.xavier@emulex.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2014-04-03RDMA/ocrdma: Allow DPP QP creationDevesh Sharma1-2/+1
Allow creating DPP QP even if inline-data is not requested. This is an optimization to lower latency. Signed-off-by: Devesh Sharma <devesh.sharma@emulex.com> Signed-off-by: Selvin Xavier <selvin.xavier@emulex.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2014-04-03RDMA/ocrdma: Read ASIC_ID register to select asic_genDevesh Sharma5-9/+35
ocrdma driver selects execution path based on sli_family and asic generation number. This introduces code to read the asic gen number from pci register instead of obtaining it from the Emulex NIC driver. Signed-off-by: Devesh Sharma <devesh.sharma@emulex.com> Signed-off-by: Selvin Xavier <selvin.xavier@emulex.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2014-04-03RDMA/ocrdma: SQ and RQ doorbell offset clean upDevesh Sharma3-24/+11
Introducing new macros to define SQ and RQ doorbell offset. Signed-off-by: Devesh Sharma <devesh.sharma@emulex.com> Signed-off-by: Selvin Xavier <selvin.xavier@emulex.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2014-04-03RDMA/ocrdma: EQ full catastrophe avoidanceDevesh Sharma4-50/+103
Stale entries in the CQ being destroyed causes hardware to generate EQEs indefinitely for a given CQ. Thus causing uncontrolled execution of irq_handler. This patch fixes this using following sementics: * irq_handler will ring EQ doorbell atleast once and implement budgeting scheme. * cq_destroy will count number of valid entires during destroy and ring cq-db so that hardware does not generate uncontrolled EQE. * cq_destroy will synchronize with last running irq_handler instance. * arm_cq will always defer arming CQ till poll_cq, except for the first arm_cq call. * poll_cq will always ring cq-db with arm=SET if arm_cq was called prior to enter poll_cq. * poll_cq will always ring cq-db with arm=UNSET if arm_cq was not called prior to enter poll_cq. Signed-off-by: Devesh Sharma <devesh.sharma@emulex.com> Signed-off-by: Selvin Xavier <selvin.xavier@emulex.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2014-04-02RDMA/cxgb4: Disable DSGL use by defaultSteve Wise1-2/+2
Current hardware doesn't correctly support DSGL. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2014-04-02RDMA/cxgb4: rx_data() needs to hold the ep mutexSteve Wise1-7/+9
To avoid racing with other threads doing close/flush/whatever, rx_data() should hold the endpoint mutex. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2014-04-02RDMA/cxgb4: Drop RX_DATA packets if the endpoint is goneSteve Wise1-0/+2
Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2014-04-02RDMA/cxgb4: Lock around accept/reject downcallsSteve Wise1-10/+21
There is a race between ULP threads doing an accept/reject, and the ingress processing thread handling close/abort for the same connection. The accept/reject path needs to hold the lock to serialize these paths. Signed-off-by: Steve Wise <swise@opengridcomputing.com> [ Fold in locking fix found by Dan Carpenter <dan.carpenter@oracle.com>. - Roland ] Signed-off-by: Roland Dreier <roland@purestorage.com>
2014-04-01IB/ehca: Remove ib_sg_dma_address() and ib_sg_dma_len() overloadsMike Marciniszyn1-12/+0
These methods appear to only mimic the sg_dma_address() and sg_dma_len() behavior. They can be safely removed. Suggested-by: Bart Van Assche <bvanassche@acm.org> Cc: Bart Van Assche <bvanassche@acm.org> Cc: Hoang-Nam Nguyen <hnguyen@de.ibm.com> Cc: Christoph Raisch <raisch@de.ibm.com> Reviewed-by: Dennis Dalessandro <dennis.dalessandro@intel.com> Signed-off-by: Mike Marciniszyn <mike.marciniszyn@intel.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2014-04-01IB/ipath: Remove ib_sg_dma_address() and ib_sg_dma_len() overloadsMike Marciniszyn1-28/+15
The removal of these methods is compensated for by code changes to .map_sg to insure that the vanilla sg_dma_address() and sg_dma_len() will do the same thing as the equivalent former ib_sg_dma_address() and ib_sg_dma_len() calls into the drivers. The introduction of this patch required that the struct ipath_dma_mapping_ops be converted to a C99 initializer. Suggested-by: Bart Van Assche <bvanassche@acm.org> Reviewed-by: Dennis Dalessandro <dennis.dalessandro@intel.com> Signed-off-by: Mike Marciniszyn <mike.marciniszyn@intel.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2014-04-01IB/qib: Remove ib_sg_dma_address() and ib_sg_dma_len() overloadsMike Marciniszyn1-17/+4
Remove the overload for .dma_len and .dma_address The removal of these methods is compensated for by code changes to .map_sg to insure that the vanilla sg_dma_address() and sg_dma_len() will do the same thing as the equivalent former ib_sg_dma_address() and ib_sg_dma_len() calls into the drivers. Suggested-by: Bart Van Assche <bvanassche@acm.org> Reviewed-by: Dennis Dalessandro <dennis.dalessandro@intel.com> Tested-by: Vinod Kumar <vinod.kumar@intel.com> Signed-off-by: Mike Marciniszyn <mike.marciniszyn@intel.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2014-04-01mlx4_core: Make buffer larger to avoid overflow warningDan Carpenter1-3/+3
My static checker complains that the sprintf() here can overflow. drivers/infiniband/hw/mlx4/main.c:1836 mlx4_ib_alloc_eqs() error: format string overflow. buf_size: 32 length: 69 This seems like a valid complaint. The "dev->pdev->bus->name" string can be 48 characters long. I just made the buffer 80 characters instead of 69 and I changed the sprintf() to snprintf(). Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2014-04-01mlx4_core: Fix some indenting in mlx4_ib_add()Dan Carpenter1-2/+3
The code was indented too far and also kernel style says we should have curly braces. Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2014-04-01IB/ehca: Returns an error on ib_copy_to_udata() failureYann Droneaud1-0/+1
In case of error when writing to userspace, function ehca_create_cq() does not set an error code before following its error path. This patch sets the error code to -EFAULT when ib_copy_to_udata() fails. This was caught when using spatch (aka. coccinelle) to rewrite call to ib_copy_{from,to}_udata(). Link: https://www.gitorious.org/opteya/coccib/source/75ebf2c1033c64c1d81df13e4ae44ee99c989eba:ib_copy_udata.cocci Link: http://marc.info/?i=cover.1394485254.git.ydroneaud@opteya.com Cc: <stable@vger.kernel.org> Signed-off-by: Yann Droneaud <ydroneaud@opteya.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2014-04-01IB/mthca: Return an error on ib_copy_to_udata() failureYann Droneaud1-0/+1
In case of error when writing to userspace, the function mthca_create_cq() does not set an error code before following its error path. This patch sets the error code to -EFAULT when ib_copy_to_udata() fails. This was caught when using spatch (aka. coccinelle) to rewrite call to ib_copy_{from,to}_udata(). Link: https://www.gitorious.org/opteya/coccib/source/75ebf2c1033c64c1d81df13e4ae44ee99c989eba:ib_copy_udata.cocci Link: http://marc.info/?i=cover.1394485254.git.ydroneaud@opteya.com Cc: <stable@vger.kernel.org> Signed-off-by: Yann Droneaud <ydroneaud@opteya.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2014-03-28RDMA/cxgb4: set error code on kmalloc() failureYann Droneaud1-1/+3
If kmalloc() fails in c4iw_alloc_ucontext(), the function leaves but does not set an error code in ret variable: it will return 0 to the caller. This patch set ret to -ENOMEM in such case. Cc: Steve Wise <swise@opengridcomputing.com> Cc: Steve Wise <swise@chelsio.com> Signed-off-by: Yann Droneaud <ydroneaud@opteya.com> Acked-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-03-26mlx4: Use actual number of PCI functions (PF + VFs) for alias GUID logicMatan Barak1-1/+1
The code which is dealing with SRIOV alias GUIDs in the mlx4 IB driver has some logic which operated according to the maximal possible active functions (PF + VFs). After the single port VFs code integration this resulted in a flow of false-positive warnings going to the kernel log after the PF driver started the alias GUID work. Fix it by referring to the actual number of functions. Signed-off-by: Matan Barak <matanb@mellanox.com> Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-03-24RDMA/cxgb4: Update snd_seq when sending MPA messagesSteve Wise1-0/+3
Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2014-03-24RDMA/cxgb4: Connect_request_upcall fixesSteve Wise1-16/+24
When processing an MPA Start Request, if the listening endpoint is DEAD, then abort the connection. If the IWCM returns an error, then we must abort the connection and release resources. Also abort_connection() should not post a CLOSE event, so clean that up too. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2014-03-24RDMA/cxgb4: Ignore read reponse type 1 CQEsSteve Wise1-4/+20
These are generated by HW in some error cases and need to be silently discarded. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2014-03-24RDMA/cxgb4: Fix possible memory leak in RX_PKT processingSteve Wise1-1/+8
If cxgb4_ofld_send() returns < 0, then send_fw_pass_open_req() must free the request skb and the saved skb with the tcp header. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2014-03-24RDMA/cxgb4: Don't leak skb in c4iw_uld_rx_handler()Steve Wise1-2/+4
Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2014-03-21net/mlx4: Adapt code for N-Port VFMatan Barak3-18/+31
Adds support for N-Port VFs, this includes: 1. Adding support in the wrapped FW command In wrapped commands, we need to verify and convert the slave's port into the real physical port. Furthermore, when sending the response back to the slave, a reverse conversion should be made. 2. Adjusting sqpn for QP1 para-virtualization The slave assumes that sqpn is used for QP1 communication. If the slave is assigned to a port != (first port), we need to adjust the sqpn that will direct its QP1 packets into the correct endpoint. 3. Adjusting gid[5] to modify the port for raw ethernet In B0 steering, gid[5] contains the port. It needs to be adjusted into the physical port. 4. Adjusting number of ports in the query / ports caps in the FW commands When a slave queries the hardware, it needs to view only the physical ports it's assigned to. 5. Adjusting the sched_qp according to the port number The QP port is encoded in the sched_qp, thus in modify_qp we need to encode the correct port in sched_qp. Signed-off-by: Matan Barak <matanb@mellanox.com> Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-03-21IB/mlx4_ib: Adapt code to use caps.num_ports instead of a constantMatan Barak1-4/+4
Some code in the mlx4 IB driver stack assumed MLX4_MAX_PORTS ports. Instead, we should only loop until the number of actual ports in i the device, which is stored in dev->caps.num_ports. Signed-off-by: Matan Barak <matanb@mellanox.com> Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-03-20IB/qib: Cleanup qib_register_observer()Dan Carpenter1-15/+10
Returning directly is easier to read than do-nothing gotos. Remove the duplicative check on "olp" and pull the code in one indent level. Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> Acked-by: Mike Marciniszyn <mike.marciniszyn@intel.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2014-03-20IB/qib: Change SDMA progression mode depending on single- or multi-railCQ Tang1-13/+123
Improve performance by changing the behavour of the driver when all SDMA descriptors are in use, and the processes adding new descriptors are single- or multi-rail. For single-rail processes, the driver will block the call and finish posting all SDMA descriptors onto the hardware queue before returning back to PSM. Repeated kernel calls are slower than blocking. For multi-rail processes, the driver will return to PSM as quick as possible so PSM can feed packets to other rail. If all hardware queues are full, PSM will buffer the remaining SDMA descriptors until notified by interrupt that space is available. This patch builds a red-black tree to track the number rails opened by a particular PID. If the number is more than one, it is a multi-rail PSM process, otherwise, it is a single-rail process. Reviewed-by: Dean Luick <dean.luick@intel.com> Reviewed-by: John A Gregor <john.a.gregor@intel.com> Reviewed-by: Mitko Haralanov <mitko.haralanov@intel.com> Signed-off-by: CQ Tang <cq.tang@intel.com> Signed-off-by: Mike Marciniszyn <mike.marciniszyn@intel.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2014-03-20RDMA/cxgb4: Save the correct map length for fast_reg_page_listsSteve Wise2-2/+11
We cannot save the mapped length using the rdma max_page_list_len field of the ib_fast_reg_page_list struct because the core code uses it. This results in an incorrect unmap of the page list in c4iw_free_fastreg_pbl(). I found this with dma mapping debugging enabled in the kernel. The fix is to save the length in the c4iw_fr_page_list struct. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2014-03-20RDMA/cxgb4: Default peer2peer mode to 1Steve Wise1-2/+2
Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2014-03-20RDMA/cxgb4: Mind the sq_sig_all/sq_sig_type QP attributesSteve Wise2-2/+5
Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2014-03-20RDMA/cxgb4: Fix incorrect BUG_ON conditionsSteve Wise1-2/+2
Based on original work from Jay Hernandez <jay@chelsio.com> Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2014-03-20RDMA/cxgb4: Always release neigh entrySteve Wise1-1/+1
Always release the neigh entry in rx_pkt(). Based on original work by Santosh Rastapur <santosh@chelsio.com>. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2014-03-20RDMA/cxgb4: Allow loopback connectionsSteve Wise1-1/+2
find_route() must treat loopback as a valid egress interface. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2014-03-20RDMA/cxgb4: Cap CQ size at T4_MAX_IQ_SIZESteve Wise1-1/+1
Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2014-03-20RDMA/cxgb4: Fix four byte info leak in c4iw_create_cq()Dan Carpenter1-0/+1
There is a four byte hole at the end of the "uresp" struct after the ->qid_mask member. Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2014-03-20RDMA/cxgb4: Fix underflows in c4iw_create_qp()Dan Carpenter1-1/+1
These sizes should be unsigned so we don't allow negative values and have underflow bugs. These can come from the user so there may be security implications, but I have not tested this. Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
generated by cgit v1.2.3 (git 2.39.0) at 2025-03-12 22:27:37 +0300