summaryrefslogtreecommitdiff
path: root/drivers/infiniband
Commit message (Collapse)AuthorAgeFilesLines
* IB/ipoib: Add ability to set PKEY index to lower device driverAlex Vesker2017-10-141-2/+6
| | | | | | | | | | | To support passing child interfaces to the lower device a new rdma_netdev function was used, set_id. This will allow us to attach the PKEY index lower device resources such as TIS/QP. For devices that do not support offloads in IPoIB same logic will be used, setting the PKEY index to priv struct. Signed-off-by: Alex Vesker <valex@mellanox.com> Reviewed-by: Erez Shitrit <erezsh@mellanox.com>
* IB/ipoib: Grab rtnl lock on heavy flush when calling ndo_open/stopAlex Vesker2017-10-141-1/+6
| | | | | | | | | When ndo_open and ndo_stop are called RTNL lock should be held. In this specific case ipoib_ib_dev_open calls the offloaded ndo_open which re-sets the number of TX queue assuming RTNL lock is held. Since RTNL lock is not held, RTNL assert will fail. Signed-off-by: Alex Vesker <valex@mellanox.com>
* i40iw: Fix port number for query QPMustafa Ismail2017-10-041-0/+2
| | | | | | | | | | Port number 0 is an invalid port number. Set it to 1 as there is one port per i40iw device. Fixes: d37498417947 ("i40iw: add files for iwarp interface") Signed-off-by: Mustafa Ismail <mustafa.ismail@intel.com> Signed-off-by: Shiraz Saleem <shiraz.saleem@intel.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
* i40iw: Add missing memory barriersMustafa Ismail2017-10-043-8/+7
| | | | | | | | | | | | Remove duplicate set_64bit_val call to offset 24. Replace some instances of set_64bit_val with i40iw_insert_wqe_hdr as valid bit needs a write barrier and should be the last write operation for the WQE. Fixes: 786c6adb3a94 ("i40iw: add puda code") Signed-off-by: Mustafa Ismail <mustafa.ismail@intel.com> Signed-off-by: Shiraz Saleem <shiraz.saleem@intel.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
* RDMA/qedr: Parse vlan priority as slAmrani, Ram2017-10-041-0/+2
| | | | | | | | | | Parse the vlan priority from the vlan tag and configure it to the WC's sl field. Fixes: abd49676c707 ("qed: Add RoCE ll2 & GSI support") Signed-off-by: Ram Amrani <Ram.Amrani@cavium.com> Signed-off-by: Michal Kalderon <Michal.Kalderon@cavium.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
* RDMA/qedr: Parse VLAN ID correctly and ignore the value of zeroAmrani, Ram2017-10-042-4/+8
| | | | | | | | | | | | Rename vlan_id field name to vlan as it contains more than the vlan_id. Mask out non vlan id fields from vlan tag of the QED LL2 RX GSI vlan output. As it is expected to be vlan id only. Ignore vlan_id with value of zero. Fixes: abd49676c707 ("qed: Add RoCE ll2 & GSI support") Signed-off-by: Ram Amrani <Ram.Amrani@cavium.com> Signed-off-by: Michal Kalderon <Michal.Kalderon@cavium.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
* IB/mlx5: Fix label order in error path handlingParav Pandit2017-10-041-2/+2
| | | | | | | | | | | | | | | When UAR get_page fails, it needs to continue to cleanup debugfs for congestion control parameters. Labels for error path were incorrectly ordered. This patch fixes to do correct cleanup on debugfs init failure and uar get page failure. Fixes: 4a2da0b8c078 ("IB/mlx5: Add debug control parameters for congestion control") Signed-off-by: Parav Pandit <parav@mellanox.com> Reviewed-by: Mark Bloch <markb@mellanox.com> Signed-off-by: Leon Romanovsky <leon@kernel.org> Signed-off-by: Doug Ledford <dledford@redhat.com>
* RDMA/iwpm: Properly mark end of NL messagesShiraz Saleem2017-09-292-0/+13
| | | | | | | | | | | | | | | Commit 1a1c116f3dcf removes nlmsg_len calculation in ibnl_put_attr causing netlink messages to be rejected due to incorrect length. Add nlmsg_end after all attributes are appended to calculate the nlmsg_len. Fixes: 1a1c116f3dcf ("RDMA/netlink: Simplify the put_msg and put_attr") Signed-off-by: Shiraz Saleem <shiraz.saleem@intel.com> Signed-off-by: Tatyana Nikolova <tatyana.e.nikolova@intel.com> Reviewed-by: Leon Romanovsky <leonro@mellanox.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
* IB/hfi1: Unsuccessful PCIe caps tuning should not fail driver loadHarish Chegondi2017-09-271-29/+21
| | | | | | | | | | | | | | | | | | | | | | Failure to tune PCIe capabilities should not fail driver load. This can cause the driver load to fail on systems with any of the following: 1. HFI's parent is not root. Example: HFI card is behind a PCIe bridge. 2. HFI's parent is not PCI Express capable. In these situations, failure to tune PCIe capabilities should be logged in the system message logs but not cause the driver load to fail. This patch also ensures pcie capability word DevCtl is written only after a successful read and the capability tuning process continues even if read/write of the pcie capability word DevCtl fails. Fixes: c53df62c7a9a ("IB/hfi1: Check return values from PCI config API calls") Fixes: bf70a7757736 ("staging/rdma/hfi1: Enable WFR PCIe extended tags from the driver") Reviewed-by: Michael J. Ruhl <michael.j.ruhl@intel.com> Reviewed-by: Mike Marciniszyn <mike.marciniszyn@intel.com> Reviewed-by: Jakub Byczkowski <jakub.byczkowski@intel.com> Signed-off-by: Harish Chegondi <harish.chegondi@intel.com> Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
* IB/hfi1: On error, fix use after free during user context setupMichael J. Ruhl2017-09-271-19/+22
| | | | | | | | | | | | | | | During base context setup, if setup_base_ctxt() fails, the context is deallocated. This is incorrect because the context is referenced on return, to notify any waiting subcontext. If there are no subcontexts the pointer will be invalid. Reorganize the error path so that deallocate_ctxt() is called after all the possible subcontexts have been notified. Reviewed-by: Ira Weiny <ira.weiny@intel.com> Signed-off-by: Michael J. Ruhl <michael.j.ruhl@intel.com> Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
* Revert "IB/ipoib: Update broadcast object if PKey value was changed in index 0"Alex Estrin2017-09-271-13/+0
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | commit 9a9b8112699d will cause core to fail UD QP from being destroyed on ipoib unload, therefore cause resources leakage. On pkey change event above patch modifies mgid before calling underlying driver to detach it from QP. Drivers' detach_mcast() will fail to find modified mgid it was never given to attach in a first place. Core qp->usecnt will never go down, so ib_destroy_qp() will fail. IPoIB driver actually does take care of new broadcast mgid based on new pkey by destroying an old mcast object in ipoib_mcast_dev_flush()) .... if (priv->broadcast) { rb_erase(&priv->broadcast->rb_node, &priv->multicast_tree); list_add_tail(&priv->broadcast->list, &remove_list); priv->broadcast = NULL; } ... then in restarted ipoib_macst_join_task() creating a new broadcast mcast object, sending join request and on completion tells the driver to attach to reinitialized QP: ... if (!priv->broadcast) { ... broadcast = ipoib_mcast_alloc(dev, 0); ... memcpy(broadcast->mcmember.mgid.raw, priv->dev->broadcast + 4, sizeof (union ib_gid)); priv->broadcast = broadcast; ... Fixes: 9a9b8112699d ("IB/ipoib: Update broadcast object if PKey value was changed in index 0") Cc: stable@vger.kernel.org Reviewed-by: Mike Marciniszyn <mike.marciniszyn@intel.com> Reviewed-by: Dennis Dalessandro <dennis.dalessandro@intel.com> Signed-off-by: Alex Estrin <alex.estrin@intel.com> Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com> Reviewed-by: Feras Daoud <ferasda@mellanox.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
* IB/hfi1: Return correct value in general interrupt handlerKamenee Arumugam2017-09-271-1/+3
| | | | | | | | | | | | The general interrupt handler returns IRQ_HANDLED whether an IRQ was handled or not. Determine if an IRQ was handled and return the correct value. Reviewed-by: Dennis Dalessandro <dennis.dalessandro@intel.com> Reviewed-by: Michael J. Ruhl <michael.j.ruhl@intel.com> Signed-off-by: Kamenee Arumugam <kamenee.arumugam@intel.com> Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
* IB/hfi1: Check eeprom config partition validityJan Sokolowski2017-09-271-5/+15
| | | | | | | | | | | | | | Relying on a trailing magic value is incorrect. There are instances where this is not present as trailing magic value has a specific purpose which is not partition validation. Instead use the header magic value which is present in all variants of the platform configuration and is intended for validation. This is also used in other locations in the driver. Fixes: bc5214ee2922 (IB/hfi1: Handle missing magic values in config file) Reviewed-by: Jakub Byczkowski <jakub.byczkowski@intel.com> Signed-off-by: Jan Sokolowski <jan.sokolowski@intel.com> Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
* IB/hfi1: Only reset QSFP after link up and turn off AOC TXSebastian Sanchez2017-09-273-3/+15
| | | | | | | | | | | | QSFP reset enables AOC transmitters by default. They should be off before moving to high power mode to complete the setup. There is no need to reset the QSFP during LNI failure as it was reset at link down. Reviewed-by: Mike Marciniszyn <mike.marciniszyn@intel.com> Reviewed-by: Jakub Byczkowski <jakub.byczkowski@intel.com> Signed-off-by: Sebastian Sanchez <sebastian.sanchez@intel.com> Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
* IB/hfi1: Turn off AOC TX after offline substatesSebastian Sanchez2017-09-272-21/+65
| | | | | | | | | | | | | | | Offline.quietDuration was added in the 8051 firmware, and the driver only turns off the AOC transmitters when offline.quiet is reached. However, the AOC transmitters need to be turned off at the new state. Therefore, turn off the AOC transmitters at any offline substates including offline.quiet and offline.quietDuration, then recheck we reached offline.quiet to support backwards compatibility. Reviewed-by: Jakub Byczkowski <jakub.byczkowski@intel.com> Reviewed-by: Mike Marciniszyn <mike.marciniszyn@intel.com> Signed-off-by: Sebastian Sanchez <sebastian.sanchez@intel.com> Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
* IB/mlx5: Fix NULL deference on mlx5_ib_update_xlt failureIlya Lesokhin2017-09-251-10/+17
| | | | | | | | | | | | | | | mlx5_ib_reg_user_mr called mlx5_ib_dereg_mr in case of MR population failure. This resulted in a NULL dereference as ibmr->device wasn't initialized yet. We address this by adding an internal dereg_mr function that can handle partially initialized MRs, and fixing clean_mr to work on partially initialized MRs. Fixes: ff740aefecb9 ("IB/mlx5: Decouple MR allocation and population flows") Signed-off-by: Ilya Lesokhin <ilyal@mellanox.com> Signed-off-by: Leon Romanovsky <leon@kernel.org> Signed-off-by: Doug Ledford <dledford@redhat.com>
* IB/mlx5: Simplify mlx5_ib_cont_pagesIlya Lesokhin2017-09-251-30/+17
| | | | | | | | | | | | | | | | | | | | | | | | The patch simplifies mlx5_ib_cont_pages and fixes the following issues in the original implementation: First issues is related to alignment of the PFNs. After the check base + p != PFN, the alignment of the PFN wasn't checked. So the PFN sequence 0, 1, 1, 2 would result in a page_shift of 13 even though the 3rd PFN is not 8KB aligned. This wasn't actually a bug because it was supported by all the existing mlx5 compatible device, but we don't want to require this support in all future devices. Another issue is because the inner loop didn't advance PFN so the test "if (base + p != pfn)" always failed for SGE with len > (1<<page_shift). Fixes: e126ba97dba9 ("mlx5: Add driver for Mellanox Connect-IB adapters") Signed-off-by: Ilya Lesokhin <ilyal@mellanox.com> Reviewed-by: Eli Cohen <eli@mellanox.com> Signed-off-by: Leon Romanovsky <leon@kernel.org> Signed-off-by: Doug Ledford <dledford@redhat.com>
* IB/ipoib: Fix inconsistency with free_netdev and free_rdma_netdevAlex Vesker2017-09-252-6/+19
| | | | | | | | | | | | | | Call free_rdma_netdev instead of free_netdev each time we want to release a netdevice. This call is also relevant for future freeing of offloaded child interfaces. This patch also adds a missing call for free netdevice when releasing a parent interface that has child interfaces using ipoib_remove_one. Fixes: cd565b4b51e5 ('IB/IPoIB: Support acceleration options callbacks') Signed-off-by: Alex Vesker <valex@mellanox.com> Signed-off-by: Leon Romanovsky <leon@kernel.org> Signed-off-by: Doug Ledford <dledford@redhat.com>
* IB/ipoib: Fix sysfs Pkey create<->remove possible deadlockShalom Lagziel2017-09-251-6/+14
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | A possible ABBA lock can happen with RTNL and vlan_rwsem. For example: Flow A: Device Flush __ipoib_ib_dev_flush down_read(vlan_rwsem) // Lock A ipoib_flush_ah flush_workqueue(priv->wq) // Wait for completion A work on shared WQ (Mcast carrier) ipoib_mcast_carrier_on_task while (!rtnl_trylock()) // Wait for lock B Flow B: Sysfs PKEY delete ipoib_vlan_delete lock(RTNL) // Lock B down_write(vlan_rwsem) // Wait for lock A This can happen with PKEY creates as well. The solution is to release the RTNL lock in sysfs functions in case it is not possible to lock VLAN RW semaphore and reset the SYS call. Fixes: 69956d83267e ("IB/ipoib: Sync between remove_one to sysfs calls that use rtnl_lock") Signed-off-by: Shalom Lagziel <shaloml@mellanox.com> Signed-off-by: Alex Vesker <valex@mellanox.com> Signed-off-by: Leon Romanovsky <leon@kernel.org> Signed-off-by: Doug Ledford <dledford@redhat.com>
* IB: Correct MR length field to be 64-bitParav Pandit2017-09-252-3/+3
| | | | | | | | | | | | | | | | | | | | | The ib_mr->length represents the length of the MR in bytes as per the IBTA spec 1.3 section 11.2.10.3 (REGISTER PHYSICAL MEMORY REGION). Currently ib_mr->length field is defined as only 32-bits field. This might result into truncation and failed WRs of consumers who registers more than 4GB bytes memory regions and whose WRs accessing such MRs. This patch makes the length 64-bit to avoid such truncation. Cc: Sagi Grimberg <sagi@grimberg.me> Cc: Chuck Lever <chuck.lever@oracle.com> Cc: Faisal Latif <faisal.latif@intel.com> Fixes: 4c67e2bfc8b7 ("IB/core: Introduce new fast registration API") Signed-off-by: Ilya Lesokhin <ilyal@mellanox.com> Signed-off-by: Parav Pandit <parav@mellanox.com> Signed-off-by: Leon Romanovsky <leon@kernel.org> Signed-off-by: Doug Ledford <dledford@redhat.com>
* IB/core: Fix qp_sec use after free accessParav Pandit2017-09-251-1/+3
| | | | | | | | | | | | When security_ib_alloc_security fails, qp->qp_sec memory is freed. However ib_destroy_qp still tries to access this memory which result in kernel crash. So its initialized to NULL to avoid such access. Fixes: d291f1a65232 ("IB/core: Enforce PKey security on QPs") Signed-off-by: Parav Pandit <parav@mellanox.com> Reviewed-by: Daniel Jurgens <danielj@mellanox.com> Signed-off-by: Leon Romanovsky <leon@kernel.org> Signed-off-by: Doug Ledford <dledford@redhat.com>
* IB/core: Fix typo in the name of the tag-matching cap structLeon Romanovsky2017-09-252-12/+12
| | | | | | | | | | | | | The tag matching functionality is implemented by mlx5 driver by extending XRQ, however this internal kernel information was exposed to user space applications with *xrq* name instead of *tm*. This patch renames *xrq* to *tm* to handle that. Fixes: 8d50505ada72 ("IB/uverbs: Expose XRQ capabilities") Signed-off-by: Leon Romanovsky <leonro@mellanox.com> Reviewed-by: Yishai Hadas <yishaih@mellanox.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
* Merge tag 'for-linus' of ↵Linus Torvalds2017-09-2317-161/+281
|\ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | git://git.kernel.org/pub/scm/linux/kernel/git/dledford/rdma Pull rdma fixes from Doug Ledford: - Smattering of miscellanous fixes - A five patch series for i40iw that had a patch (5/5) that was larger than I would like, but I took it because it's needed for large scale users - An 8 patch series for bnxt_re that landed right as I was leaving on PTO and so had to wait until now...they are all appropriate fixes for -rc IMO * tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/dledford/rdma: (22 commits) bnxt_re: Don't issue cmd to delete GID for QP1 GID entry before the QP is destroyed bnxt_re: Fix memory leak in FRMR path bnxt_re: Remove RTNL lock dependency in bnxt_re_query_port bnxt_re: Fix race between the netdev register and unregister events bnxt_re: Free up devices in module_exit path bnxt_re: Fix compare and swap atomic operands bnxt_re: Stop issuing further cmds to FW once a cmd times out bnxt_re: Fix update of qplib_qp.mtu when modified i40iw: Add support for port reuse on active side connections i40iw: Add missing VLAN priority i40iw: Call i40iw_cm_disconn on modify QP to disconnect i40iw: Prevent multiple netdev event notifier registrations i40iw: Fail open if there are no available MSI-X vectors RDMA/vmw_pvrdma: Fix reporting correct opcodes for completion IB/bnxt_re: Fix frame stack compilation warning IB/mlx5: fix debugfs cleanup IB/ocrdma: fix incorrect fall-through on switch statement IB/ipoib: Suppress the retry related completion errors iw_cxgb4: remove the stid on listen create failure iw_cxgb4: drop listen destroy replies if no ep found ...
| * bnxt_re: Don't issue cmd to delete GID for QP1 GID entry before the QP is ↵Somnath Kotur2017-09-221-3/+20
| | | | | | | | | | | | | | | | | | | | | | | | destroyed FW needs the 0th GID Entry in the Table to be preserved before it's corresponding QP1 is deleted, else it will fail the cmd. Check for the same and return to prevent error msg being logged for cmd failure. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
| * bnxt_re: Fix memory leak in FRMR pathSelvin Xavier2017-09-221-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This patch fixes a memory leak issue when alloc_mr is used. mr->pages and mr->npages are used only in alloc_mr path. mr->pages is allocated when alloc_mr is called or in the case of FRMR, while creating the MR. mr->npages is updated only when the MR created is used i.e. after invoking map_mr_sg verb, before data transfer. In the dereg_mr path, if mr->npages is 0, driver ends up not freeing the memory created. Removing the npages check from the dereg_mr path for kernel consumers. Signed-off-by: Selvin Xavier <selvin.xavier@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
| * bnxt_re: Remove RTNL lock dependency in bnxt_re_query_portSomnath Kotur2017-09-223-8/+9
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | When there is a NETDEV_UNREGISTER event, bnxt_re driver calls ib_unregister_device() (RTNL lock held). ib_unregister_device attempts to flush a worker queue scheduled by ib_core and that queue might have a pending ib_query_port(). ib_query_port in turn calls bnxt_re_query_port(), which while querying the link speed using ib_get_eth_speed(), tries to acquire the rtnl_lock() which was already held by NETDEV_UNREGISTER. Fixing the issue by removing the link speed query from bnxt_re_query_port() Now the speed is queried post a successful ib_register_device or whenever there is a NETDEV_CHANGE event. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
| * bnxt_re: Fix race between the netdev register and unregister eventsSomnath Kotur2017-09-222-5/+15
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Upon receipt of the NETDEV_REGISTER event from the netdev notifier chain, the IB stack registration is spawned off to a workqueue since that also requires an rtnl lock. There could be 2 kinds of races between the NETDEV_REGISTER and the NETDEV_UNREGISTER event handling. a)The NETDEV_UNREGISTER event is received in rapid succession after the NETDEV_REGISTER event even before the work queue got a chance to run. b)The NETDEV_UNREGISTER event is received while the workqueue that handles registration with the IB stack is still in progress. Handle both the races with a bit flag that is set just before the work item is queued and cleared in the workqueue after the event is handled just before the workqueue item is freed. While adding the new flag, it was noted that the flags are all used in *_bit() operations which expect a bit number and not a literal constant with a bit set. So change the numbers to be bit numbers. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
| * bnxt_re: Free up devices in module_exit pathSomnath Kotur2017-09-221-0/+16
| | | | | | | | | | | | | | | | Clean up all devices added to the bnxt_re_dev_list in the module_exit entry point. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
| * bnxt_re: Fix compare and swap atomic operandsDevesh Sharma2017-09-221-0/+1
| | | | | | | | | | | | | | | | | | | | Driver must assign the user supplied compare/swap values in the wqe to successfully complete the atomic compare and swap operation. Signed-off-by: Devesh Sharma <devesh.sharma@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
| * bnxt_re: Stop issuing further cmds to FW once a cmd times outSomnath Kotur2017-09-222-1/+6
| | | | | | | | | | | | | | | | | | | | | | Once a cmd to FW times out(after 20s) it is reasonable to assume the FW or atleast the control path is dead. No point issuing further cmds to the FW as each subsequent cmd with another 20s timeout will cascade resulting in unnecessary traces and/or NMI Lockups. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
| * bnxt_re: Fix update of qplib_qp.mtu when modifiedDevesh Sharma2017-09-221-0/+3
| | | | | | | | | | | | | | | | | | | | | | | | The MTU value in the qplib_qp.mtu should be consistent with whatever mtu was set during INIT to RTR.The Next PSN and number of packets are calculated based on this member in the qplib_qp structure. Signed-off-by: Narender Reddy <narender.reddy@broadcom.com> Signed-off-by: Devesh Sharma <devesh.sharma@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
| * i40iw: Add support for port reuse on active side connectionsShiraz Saleem2017-09-222-76/+78
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | During OpenMPI scale up testing, we observe rdma_connect failures if ports are reused on multiple connections. This is because the Control Queue-Pair (CQP) command to add the reused port to Accelerated Port Bit VectorTable (APBVT) fails as there already exists an entry. Check for duplicate port before invoking the CQP command to add APBVT entry and delete the entry only if the port is not in use. Signed-off-by: Shiraz Saleem <shiraz.saleem@intel.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
| * i40iw: Add missing VLAN priorityMustafa Ismail2017-09-222-1/+4
| | | | | | | | | | | | | | | | | | Set the VLAN priority which is in the upper 3 bits of the VLAN tag field in the QP context. Signed-off-by: Mustafa Ismail <mustafa.ismail@intel.com> Signed-off-by: Shiraz Saleem <shiraz.saleem@intel.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
| * i40iw: Call i40iw_cm_disconn on modify QP to disconnectShiraz Saleem2017-09-221-0/+12
| | | | | | | | | | | | | | | | | | | | | | | | If QP modify to closing/terminate/error fails, connection is not torn down as there is no corresponding asynchronous event that will initiate the teardown. Add explicit call to i40iw_cm_disconn if not waiting in modify QP, otherwise schedule it in CM timer. Signed-off-by: Shiraz Saleem <shiraz.saleem@intel.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
| * i40iw: Prevent multiple netdev event notifier registrationsShiraz Saleem2017-09-223-19/+20
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Netdev event notifier registration/de-registration is not synchronized with a lock and there is a possibility of a duplicate registration of notifier before the unregister completes. Register netdev event notifiers during module init and de-register them at module exit. This avoids the need to tie the registration to first netdev client interface open and de-registration to last client interface close and the synchronization to achieve it. This also fixes a crash due to duplicate registration. BUG: unable to handle kernel paging request at ffffffffa0d60388 IP: [<ffffffff8160f75d>] notifier_call_chain+0x3d/0x70 PGD 190d067 PUD 190e063 PMD 76c840067 PTE 0 Oops: 0000 [#1] SMP Modules linked in: i40e(OF-) fuse btrfs zlib_deflate raid6_pq xor vfat msdos [..] e1000e vxlan ip_tunnel ptp pps_core i2c_core video [last unloaded: i40iw] CPU: 1 PID: 27101 Comm: modprobe Tainted: GF W O-------------- 3.10.0-229.el7.x86_64 #1 Hardware name: Gigabyte Technology Co., Ltd. To be filled by O.E.M./Q87M-D2H, BIOS F7 01/17/2014 task: ffff88076e8a96c0 ti: ffff8806959c8000 task.ti: ffff8806959c8000 RIP: 0010:[<ffffffff8160f75d>] [<ffffffff8160f75d>] notifier_call_chain+0x3d/0x70 RSP: 0018:ffff8806959cbb38 EFLAGS: 00010282 RAX: ffffffffa0d60380 RBX: 00000000fffffffd RCX: 0000000000000000 0708] RDX: 0000000000000000 RSI: ffff88081227a000 RDI: 0000000000000002 RBP: ffff8806959cbb60 R08: 0000000000000246 R09: 000000000000700c R10: ffff88080e16ea40 R11: 00000000000ae8df R12: ffffffffa0d60380 R13: 0000000000000002 R14: ffff88076e738800 R15: 0000000000000000 FS: 00007f604ef4a740(0000) GS:ffff88083e240000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: ffffffffa0d60388 CR3: 0000000753cd2000 CR4: 00000000001407e0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 Stack: ffffffff819e73a0 0000000000000000 0000000000000002 ffff88076e738800 00000000ffffffff ffff8806959cbba0 ffffffff8109d61d 0000000000000000 0000000000000000 ffff88076e738800 0000000000000000 ffff88076e738800 Call Trace: [<ffffffff8109d61d>] __blocking_notifier_call_chain+0x4d/0x70 [<ffffffff8109d656>] blocking_notifier_call_chain+0x16/0x20 [<ffffffff8156b9e4>] __inet_del_ifa+0x154/0x2b0 [<ffffffff8156d102>] inetdev_event+0x182/0x530 [<ffffffff8160f76c>] notifier_call_chain+0x4c/0x70 [<ffffffff8109d446>] raw_notifier_call_chain+0x16/0x20 [<ffffffff814f71fd>] call_netdevice_notifiers+0x2d/0x60 [<ffffffff814f8845>] rollback_registered_many+0x105/0x220 [<ffffffff814f89a0>] rollback_registered+0x40/0x70 [<ffffffff814f9c88>] unregister_netdevice_queue+0x48/0x80 [<ffffffff814f9cdc>] unregister_netdev+0x1c/0x30 [<ffffffffa0067139>] i40e_vsi_release+0x2a9/0x2b0 [i40e] [<ffffffffa00674e8>] i40e_remove+0x128/0x2b0 [i40e] [<ffffffff813092db>] pci_device_remove+0x3b/0xb0 [<ffffffff813d26ef>] __device_release_driver+0x7f/0xf0 [<ffffffff813d3068>] driver_detach+0xb8/0xc0 [<ffffffff813d22db>] bus_remove_driver+0x9b/0x120 [<ffffffff813d36dc>] driver_unregister+0x2c/0x50 [<ffffffff81307d4c>] pci_unregister_driver+0x2c/0x90 [<ffffffffa008f9d0>] i40e_exit_module+0x10/0x23 [i40e] [<ffffffff810dad0b>] SyS_delete_module+0x16b/0x2d0 [<ffffffff81013b0c>] ? do_notify_resume+0x9c/0xb0 [<ffffffff81613da9>] system_call_fastpath+0x16/0x1b Code: e5 41 57 4d 89 c7 41 56 49 89 d6 41 55 49 89 f5 41 54 53 89 cb 75 14 eb 3d 0f 1f 44 00 00 83 eb 01 74 25 4d 85 e4 74 20 4c 89 e0 <4c> 8b 60 08 4c 89 f2 4c 89 ee 48 89 c7 ff 10 4d 85 ff 74 04 41 RIP [<ffffffff8160f75d>] notifier_call_chain+0x3d/0x70 Signed-off-by: Shiraz Saleem <shiraz.saleem@intel.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
| * i40iw: Fail open if there are no available MSI-X vectorsShiraz Saleem2017-09-221-1/+6
| | | | | | | | | | | | | | | | Check number of available MSI-X vectors for i40iw. If there are no available vectors, fail the open. Signed-off-by: Shiraz Saleem <shiraz.saleem@intel.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
| * RDMA/vmw_pvrdma: Fix reporting correct opcodes for completionAdit Ranadive2017-09-221-3/+28
| | | | | | | | | | | | | | | | | | | | | | | | Since the IB_WC_BIND_MW opcode has been dropped, set the correct IB WC opcode explicitly. Fixes: 29c8d9eba550 ("IB: Add vmw_pvrdma driver") Reviewed-by: Aditya Sarwade <asarwade@vmware.com> Reviewed-by: Jorgen Hansen <jhansen@vmware.com> Signed-off-by: Adit Ranadive <aditr@vmware.com> Signed-off-by: Bryan Tan <bryantan@vmware.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
| * IB/bnxt_re: Fix frame stack compilation warningLeon Romanovsky2017-09-221-31/+36
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Reduce stack size by dynamically allocating memory instead of declaring large struct on the stack: drivers/infiniband/hw/bnxt_re/ib_verbs.c: In function ‘bnxt_re_query_qp’: drivers/infiniband/hw/bnxt_re/ib_verbs.c:1600:1: warning: the frame size of 1216 bytes is larger than 1024 bytes [-Wframe-larger-than=] } ^ Cc: Selvin Xavier <selvin.xavier@broadcom.com> Fixes: 1ac5a4047975 ("RDMA/bnxt_re: Add bnxt_re RoCE driver") Signed-off-by: Leon Romanovsky <leon@kernel.org> Acked-by: Selvin Xavier <selvin.xavier@broadcom.com> Reviewed-by: Jonathan Toppins <jtoppins@redhat.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
| * IB/mlx5: fix debugfs cleanupSudip Mukherjee2017-09-221-3/+3
| | | | | | | | | | | | | | | | | | | | | | | | | | | | If delay_drop_debugfs_init() fails in any of the operations to create debugfs, it is calling delay_drop_debugfs_cleanup() as part of its cleanup. But delay_drop_debugfs_cleanup() checks for 'dbg' and since we have not yet pointed 'dbg' to the debugfs we need to cleanup, the cleanup fails and we are left with stray debugfs elements and also a memory leak. Fixes: 4a5fd5d2965c ("IB/mlx5: Add necessary delay drop assignment") Signed-off-by: Sudip Mukherjee <sudipm.mukherjee@gmail.com> Acked-by: Leon Romanovsky <leonro@mellanox.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
| * IB/ocrdma: fix incorrect fall-through on switch statementColin Ian King2017-09-221-0/+3
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | In the case where mbox_status is OCRDMA_MBX_STATUS_FAILED and add_status is OCRDMA_MBX_STATUS_FAILED err_num is assigned -EAGAIN however the case OCRDMA_MBX_STATUS_FAILED is missing a break and falls through to the default case which then re-assigns err_num to -EFAULT. Fix this so that err_num is assigned to -EAGAIN for the add_status OCRDMA_MBX_STATUS_FAILED case and -EFAULT otherwise. Detected by CoverityScan CID#703125 ("Missing break in switch") Fixes: fe2caefcdf58 ("RDMA/ocrdma: Add driver for Emulex OneConnect IBoE RDMA adapter") Signed-off-by: Colin Ian King <colin.king@canonical.com> Reviewed-by: Leon Romanovsky <leonro@mellanox.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
| * IB/ipoib: Suppress the retry related completion errorsSantosh Shilimkar2017-09-221-5/+11
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | IPoIB doesn't support transport/rnr retry schemes as per RFC so those errors are expected. No need to flood the log files with them. Tested-by: Michael Nowak <michael.nowak@oracle.com> Tested-by: Rafael Alejandro Peralez <rafael.peralez@oracle.com> Tested-by: Liwen Huang <liwen.huang@oracle.com> Tested-by: Hong Liu <hong.x.liu@oracle.com> Reviewed-by: Mukesh Kacker <mukesh.kacker@oracle.com> Reported-by: Rajiv Raja <rajiv.raja@oracle.com> Signed-off-by: Santosh Shilimkar <santosh.shilimkar@oracle.com> Signed-off-by: Yuval Shaia <yuval.shaia@oracle.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
| * iw_cxgb4: remove the stid on listen create failureSteve Wise2017-09-221-1/+1
| | | | | | | | | | | | | | | | | | | | | | If a listen create fails, then the server tid (stid) is incorrectly left in the stid idr table, which can cause a touch-after-free if the stid is looked up and the already freed endpoint is touched. So make sure and remove it in the error path. Cc: stable@vger.kernel.org Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
| * iw_cxgb4: drop listen destroy replies if no ep foundSteve Wise2017-09-221-0/+5
| | | | | | | | | | | | | | | | | | | | If the thread waiting for a CLOSE_LISTSRV_RPL times out and bails, then we need to handle a subsequent CPL if it arrives and the stid has been released. In this case silently drop it. Cc: stable@vger.kernel.org Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
| * iw_cxgb4: put ep reference in pass_accept_req()Steve Wise2017-09-221-1/+1
| | | | | | | | | | | | | | | | | | | | | | The listening endpoint should always be dereferenced at the end of pass_accept_req(). Fixes: f86fac79afec ("RDMA/iw_cxgb4: atomic find and reference for listening endpoints") Cc: stable@vger.kernel.org Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
| * IB/core: Fix for core panicAlex Estrin2017-09-221-2/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Build with the latest patches resulted in panic: 11384.486289] BUG: unable to handle kernel NULL pointer dereference at (null) [11384.486293] IP: (null) [11384.486295] PGD 0 [11384.486295] P4D 0 [11384.486296] [11384.486299] Oops: 0010 [#1] SMP ......... snip ...... [11384.486401] CPU: 0 PID: 968 Comm: kworker/0:1H Tainted: G W O 4.13.0-a-stream-20170825 #1 [11384.486402] Hardware name: Intel Corporation S2600WT2R/S2600WT2R, BIOS SE5C610.86B.01.01.0014.121820151719 12/18/2015 [11384.486418] Workqueue: ib-comp-wq ib_cq_poll_work [ib_core] [11384.486419] task: ffff880850579680 task.stack: ffffc90007fec000 [11384.486420] RIP: 0010: (null) [11384.486420] RSP: 0018:ffffc90007fef970 EFLAGS: 00010206 [11384.486421] RAX: ffff88084cfe8000 RBX: ffff88084dce4000 RCX: ffffc90007fef978 [11384.486422] RDX: 0000000000000000 RSI: 0000000000000001 RDI: ffff88084cfe8000 [11384.486422] RBP: ffffc90007fefab0 R08: 0000000000000000 R09: ffff88084dce4080 [11384.486423] R10: ffffffffa02d7f60 R11: 0000000000000000 R12: ffff88105af65a00 [11384.486423] R13: ffff88084dce4000 R14: 000000000000c000 R15: 000000000000c000 [11384.486424] FS: 0000000000000000(0000) GS:ffff88085f400000(0000) knlGS:0000000000000000 [11384.486425] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [11384.486425] CR2: 0000000000000000 CR3: 0000000001c09000 CR4: 00000000001406f0 [11384.486426] Call Trace: [11384.486431] ? is_valid_mcast_lid.isra.21+0xfb/0x110 [ib_core] [11384.486436] ib_attach_mcast+0x6f/0xa0 [ib_core] [11384.486441] ipoib_mcast_attach+0x81/0x190 [ib_ipoib] [11384.486443] ipoib_mcast_join_complete+0x354/0xb40 [ib_ipoib] [11384.486448] mcast_work_handler+0x330/0x6c0 [ib_core] [11384.486452] join_handler+0x101/0x220 [ib_core] [11384.486455] ib_sa_mcmember_rec_callback+0x54/0x80 [ib_core] [11384.486459] recv_handler+0x3a/0x60 [ib_core] [11384.486462] ib_mad_recv_done+0x423/0x9b0 [ib_core] [11384.486466] __ib_process_cq+0x5d/0xb0 [ib_core] [11384.486469] ib_cq_poll_work+0x20/0x60 [ib_core] [11384.486472] process_one_work+0x149/0x360 [11384.486474] worker_thread+0x4d/0x3c0 [11384.486487] kthread+0x109/0x140 [11384.486488] ? rescuer_thread+0x380/0x380 [11384.486489] ? kthread_park+0x60/0x60 [11384.486490] ? kthread_park+0x60/0x60 [11384.486493] ret_from_fork+0x25/0x30 [11384.486493] Code: Bad RIP value. [11384.486493] Code: Bad RIP value. [11384.486496] RIP: (null) RSP: ffffc90007fef970 [11384.486497] CR2: 0000000000000000 [11384.486531] ---[ end trace b1acec6fb4ff6e75 ]--- [11384.532133] Kernel panic - not syncing: Fatal exception [11384.536541] Kernel Offset: disabled [11384.969491] ---[ end Kernel panic - not syncing: Fatal exception [11384.976875] sched: Unexpected reschedule of offline CPU#1! [11384.983646] ------------[ cut here ]------------ Rdma device driver may not have implemented (*get_link_layer)() so it can not be called directly. Should use appropriate helper function. Reviewed-by: Yuval Shaia <yuval.shaia@oracle.com> Fixes: 523633359224 ("IB/core: Fix the validations of a multicast LID in attach or detach operations") Cc: stable@kernel.org # 4.13 Reviewed-by: Dennis Dalessandro <dennis.dalessandro@intel.com> Signed-off-by: Alex Estrin <alex.estrin@intel.com> Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com> Reviewed-by: Leon Romanovsky <leonro@mellanox.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
* | IB/mlx4: fix sprintf format warningArnd Bergmann2017-09-131-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | gcc-7 points out that a negative port_num value would overflow the string buffer: drivers/infiniband/hw/mlx4/sysfs.c: In function 'mlx4_ib_device_register_sysfs': drivers/infiniband/hw/mlx4/sysfs.c:251:16: error: 'sprintf' may write a terminating nul past the end of the destination [-Werror=format-overflow=] drivers/infiniband/hw/mlx4/sysfs.c:251:2: note: 'sprintf' output between 2 and 11 bytes into a destination of size 10 drivers/infiniband/hw/mlx4/sysfs.c:303:17: error: 'sprintf' may write a terminating nul past the end of the destination [-Werror=format-overflow=] drivers/infiniband/hw/mlx4/sysfs.c:303:3: note: 'sprintf' output between 2 and 11 bytes into a destination of size 10 While we should be able to assume that port_num is positive here, making the buffer one byte longer has no downsides and avoids the warning. Fixes: c1e7e466120b ("IB/mlx4: Add iov directory in sysfs under the ib device") Link: http://lkml.kernel.org/r/20170714120720.906842-23-arnd@arndb.de Signed-off-by: Arnd Bergmann <arnd@arndb.de> Reviewed-by: Leon Romanovsky <leonro@mellanox.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* | Merge tag 'nfsd-4.14' of git://linux-nfs.org/~bfields/linuxLinus Torvalds2017-09-091-0/+24
|\ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Pull nfsd updates from Bruce Fields: "More RDMA work and some op-structure constification from Chuck Lever, and a small cleanup to our xdr encoding" * tag 'nfsd-4.14' of git://linux-nfs.org/~bfields/linux: svcrdma: Estimate Send Queue depth properly rdma core: Add rdma_rw_mr_payload() svcrdma: Limit RQ depth svcrdma: Populate tail iovec when receiving nfsd: Incoming xdr_bufs may have content in tail buffer svcrdma: Clean up svc_rdma_build_read_chunk() sunrpc: Const-ify struct sv_serv_ops nfsd: Const-ify NFSv4 encoding and decoding ops arrays sunrpc: Const-ify instances of struct svc_xprt_ops nfsd4: individual encoders no longer see error cases nfsd4: skip encoder in trivial error cases nfsd4: define ->op_release for compound ops nfsd4: opdesc will be useful outside nfs4proc.c nfsd4: move some nfsd4 op definitions to xdr4.h
| * | rdma core: Add rdma_rw_mr_payload()Chuck Lever2017-09-051-0/+24
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The amount of payload per MR depends on device capabilities and the memory registration mode in use. The new rdma_rw API hides both, making it difficult for ULPs to determine how large their transport send queues need to be. Expose the MR payload information via a new API. Signed-off-by: Chuck Lever <chuck.lever@oracle.com> Acked-by: Doug Ledford <dledford@redhat.com> Signed-off-by: J. Bruce Fields <bfields@redhat.com>
* | | lib/interval_tree: fast overlap detectionDavidlohr Bueso2017-09-087-25/+26
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Allow interval trees to quickly check for overlaps to avoid unnecesary tree lookups in interval_tree_iter_first(). As of this patch, all interval tree flavors will require using a 'rb_root_cached' such that we can have the leftmost node easily available. While most users will make use of this feature, those with special functions (in addition to the generic insert, delete, search calls) will avoid using the cached option as they can do funky things with insertions -- for example, vma_interval_tree_insert_after(). [jglisse@redhat.com: fix deadlock from typo vm_lock_anon_vma()] Link: http://lkml.kernel.org/r/20170808225719.20723-1-jglisse@redhat.com Link: http://lkml.kernel.org/r/20170719014603.19029-12-dave@stgolabs.net Signed-off-by: Davidlohr Bueso <dbueso@suse.de> Signed-off-by: Jérôme Glisse <jglisse@redhat.com> Acked-by: Christian König <christian.koenig@amd.com> Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Acked-by: Doug Ledford <dledford@redhat.com> Acked-by: Michael S. Tsirkin <mst@redhat.com> Cc: David Airlie <airlied@linux.ie> Cc: Jason Wang <jasowang@redhat.com> Cc: Christian Benvenuti <benve@cisco.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* | | RDMA/netlink: clean up message validity array initializerLinus Torvalds2017-09-081-6/+5
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The fix in the parent made me look at that function, and react to how illogical and illegible the array initializer was. Use named array indexes to make it clearer what is going on, and make the initializer not depend silently on the exact index numbers. [ The initializer now also shows an odd inconsistency in the naming: note the IWCM vs IWPM.. - Linus ] Cc: Leon Romanovsky <leonro@mellanox.com> Cc: Doug Ledford <dledford@redhat.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
View Lorry Controller Status