1. 13 Nov, 2017 4 commits
  2. 10 Nov, 2017 22 commits
  3. 01 Nov, 2017 1 commit
  4. 30 Oct, 2017 7 commits
  5. 25 Oct, 2017 6 commits
    • Somnath Kotur's avatar
      bnxt_re: Implement the shutdown hook of the L2-RoCE driver interface · 5455e73a
      Somnath Kotur authored
      When host is shutting down, it invokes the shutdown hook of the
      L2 driver where it would attempt to free the MSI-X vectors, but would fail
      because some vectors are held by the RoCE driver.
      Implement the new hook in the L2 -> RoCE interface which will be invoked so that
      the RoCE driver can unregister the device and free up the MSI-X vectors it had
      claimed so that L2 can proceed with it's shutdown without failure.
      Signed-off-by: default avatarSomnath Kotur <somnath.kotur@broadcom.com>
      Signed-off-by: default avatarDoug Ledford <dledford@redhat.com>
      5455e73a
    • Leon Romanovsky's avatar
      RDMA/cxgb4: Declare stag as __be32 · 35fb2a88
      Leon Romanovsky authored
      The scqe.stag is actually __b32, fix it.
      
        drivers/infiniband/hw/cxgb4/cq.c:754:52: warning: cast to restricted __be32
      
      Cc: Steve Wise <swise@opengridcomputing.com>
      Signed-off-by: default avatarLeon Romanovsky <leon@kernel.org>
      Reviewed-by: default avatarSteve Wise <swise@opengridcomputing.com>
      Signed-off-by: default avatarDoug Ledford <dledford@redhat.com>
      35fb2a88
    • Kees Cook's avatar
      IB/rxe: Convert timers to use timer_setup() · 3bfbea74
      Kees Cook authored
      In preparation for unconditionally passing the struct timer_list pointer to
      all timer callbacks, switch to using the new timer_setup() and from_timer()
      to pass the timer pointer explicitly.
      
      Cc: Moni Shoua <monis@mellanox.com>
      Cc: Doug Ledford <dledford@redhat.com>
      Cc: Sean Hefty <sean.hefty@intel.com>
      Cc: Hal Rosenstock <hal.rosenstock@gmail.com>
      Cc: linux-rdma@vger.kernel.org
      Signed-off-by: default avatarKees Cook <keescook@chromium.org>
      Signed-off-by: default avatarDoug Ledford <dledford@redhat.com>
      3bfbea74
    • Michael J. Ruhl's avatar
      RDMA/netlink: OOPs in rdma_nl_rcv_msg() from misinterpreted flag · b4d91aeb
      Michael J. Ruhl authored
      rdma_nl_rcv_msg() checks to see if it should use the .dump() callback
      or the .doit() callback.  The check is done with this check:
      
      if (flags & NLM_F_DUMP) ...
      
      The NLM_F_DUMP flag is two bits (NLM_F_ROOT | NLM_F_MATCH).
      
      When an RDMA_NL_LS message (response) is received, the bit used for
      indicating an error is the same bit as NLM_F_ROOT.
      
      NLM_F_ROOT == (0x100) == RDMA_NL_LS_F_ERR.
      
      ibacm sends a response with the RDMA_NL_LS_F_ERR bit set if an error
      occurs in the service.  The current code then misinterprets the
      NLM_F_DUMP bit and trys to call the .dump() callback.
      
      If the .dump() callback for the specified request is not available
      (which is true for the RDMA_NL_LS messages) the following Oops occurs:
      
      [ 4555.960256] BUG: unable to handle kernel NULL pointer dereference at
         (null)
      [ 4555.969046] IP:           (null)
      [ 4555.972664] PGD 10543f1067 P4D 10543f1067 PUD 1033f93067 PMD 0
      [ 4555.979287] Oops: 0010 [#1] SMP
      [ 4555.982809] Modules linked in: rpcrdma ib_isert iscsi_target_mod
      target_core_mod ib_iser libiscsi scsi_transport_iscsi ib_ipoib rdma_ucm ib_ucm
      ib_uverbs ib_umad rdma_cm ib_cm iw_cm dm_mirror dm_region_hash dm_log dm_mod
      dax sb_edac x86_pkg_temp_thermal intel_powerclamp coretemp kvm irqbypass
      crct10dif_pclmul crc32_pclmul ghash_clmulni_intel pcbc aesni_intel crypto_simd
      glue_helper cryptd hfi1 rdmavt iTCO_wdt iTCO_vendor_support ib_core mei_me
      lpc_ich pcspkr mei ioatdma sg shpchp i2c_i801 mfd_core wmi ipmi_si ipmi_devintf
      ipmi_msghandler acpi_power_meter acpi_pad nfsd auth_rpcgss nfs_acl lockd grace
      sunrpc ip_tables ext4 mbcache jbd2 sd_mod mgag200 drm_kms_helper syscopyarea
      sysfillrect sysimgblt fb_sys_fops ttm igb ahci crc32c_intel ptp libahci
      pps_core drm dca libata i2c_algo_bit i2c_core
      [ 4556.061190] CPU: 54 PID: 9841 Comm: ibacm Tainted: G          I
      4.14.0-rc2+ #6
      [ 4556.069667] Hardware name: Intel Corporation S2600WT2/S2600WT2, BIOS
      SE5C610.86B.01.01.0008.021120151325 02/11/2015
      [ 4556.081339] task: ffff880855f42d00 task.stack: ffffc900246b4000
      [ 4556.087967] RIP: 0010:          (null)
      [ 4556.092166] RSP: 0018:ffffc900246b7bc8 EFLAGS: 00010246
      [ 4556.098018] RAX: ffffffff81dbe9e0 RBX: ffff881058bb1000 RCX:
      0000000000000000
      [ 4556.105997] RDX: 0000000000001100 RSI: ffff881058bb1320 RDI:
      ffff881056362000
      [ 4556.113984] RBP: ffffc900246b7bf8 R08: 0000000000000ec0 R09:
      0000000000001100
      [ 4556.121971] R10: ffff8810573a5000 R11: 0000000000000000 R12:
      ffff881056362000
      [ 4556.129957] R13: 0000000000000ec0 R14: ffff881058bb1320 R15:
      0000000000000ec0
      [ 4556.137945] FS:  00007fe0ba5a38c0(0000) GS:ffff88105f080000(0000)
      knlGS:0000000000000000
      [ 4556.147000] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
      [ 4556.153433] CR2: 0000000000000000 CR3: 0000001056f5d003 CR4:
      00000000001606e0
      [ 4556.161419] Call Trace:
      [ 4556.164167]  ? netlink_dump+0x12c/0x290
      [ 4556.168468]  __netlink_dump_start+0x186/0x1f0
      [ 4556.173357]  rdma_nl_rcv_msg+0x193/0x1b0 [ib_core]
      [ 4556.178724]  rdma_nl_rcv+0xdc/0x130 [ib_core]
      [ 4556.183604]  netlink_unicast+0x181/0x240
      [ 4556.187998]  netlink_sendmsg+0x2c2/0x3b0
      [ 4556.192392]  sock_sendmsg+0x38/0x50
      [ 4556.196299]  SYSC_sendto+0x102/0x190
      [ 4556.200308]  ? __audit_syscall_entry+0xaf/0x100
      [ 4556.205387]  ? syscall_trace_enter+0x1d0/0x2b0
      [ 4556.210366]  ? __audit_syscall_exit+0x209/0x290
      [ 4556.215442]  SyS_sendto+0xe/0x10
      [ 4556.219060]  do_syscall_64+0x67/0x1b0
      [ 4556.223165]  entry_SYSCALL64_slow_path+0x25/0x25
      [ 4556.228328] RIP: 0033:0x7fe0b9db2a63
      [ 4556.232333] RSP: 002b:00007ffc55edc260 EFLAGS: 00000293 ORIG_RAX:
      000000000000002c
      [ 4556.240808] RAX: ffffffffffffffda RBX: 0000000000000010 RCX:
      00007fe0b9db2a63
      [ 4556.248796] RDX: 0000000000000010 RSI: 00007ffc55edc280 RDI:
      000000000000000d
      [ 4556.256782] RBP: 00007ffc55edc670 R08: 00007ffc55edc270 R09:
      000000000000000c
      [ 4556.265321] R10: 0000000000000000 R11: 0000000000000293 R12:
      00007ffc55edc280
      [ 4556.273846] R13: 000000000260b400 R14: 000000000000000d R15:
      0000000000000001
      [ 4556.282368] Code:  Bad RIP value.
      [ 4556.286629] RIP:           (null) RSP: ffffc900246b7bc8
      [ 4556.293013] CR2: 0000000000000000
      [ 4556.297292] ---[ end trace 8d67abcfd10ec209 ]---
      [ 4556.305465] Kernel panic - not syncing: Fatal exception
      [ 4556.313786] Kernel Offset: disabled
      [ 4556.321563] ---[ end Kernel panic - not syncing: Fatal exception
      [ 4556.328960] ------------[ cut here ]------------
      
      Special case RDMA_NL_LS response messages to call the appropriate
      callback.
      
      Additionally, make sure that the .dump() callback is not NULL
      before calling it.
      
      Fixes: 647c75ac ("RDMA/netlink: Convert LS to doit callback")
      Reviewed-by: default avatarMike Marciniszyn <mike.marciniszyn@intel.com>
      Reviewed-by: default avatarKaike Wan <kaike.wan@intel.com>
      Reviewed-by: default avatarAlex Estrin <alex.estrin@intel.com>
      Signed-off-by: default avatarMichael J. Ruhl <michael.j.ruhl@intel.com>
      Reviewed-by: default avatarShiraz Saleem <shiraz.saleem@intel.com>
      Signed-off-by: default avatarDoug Ledford <dledford@redhat.com>
      b4d91aeb
    • Parav Pandit's avatar
      IB/cm: Fix memory corruption in handling CM request · 5a3dc323
      Parav Pandit authored
      In recent code, two path record entries are alwasy cleared while
      allocated could be either one or two path record entries.
      This leads to zero out of unallocated memory.
      
      This fix initializes alternative path record only when alternative path
      is set.
      
      While we are at it, path record allocation doesn't check for OPA
      alternative path, but rest of the code checks for OPA alternative path.
      Path record allocation code doesn't check for OPA alternative LID.
      This can further lead to memory corruption when only one path record is
      allocated, but there is actually alternative OPA path record present in CM
      request.
      
      Cc: <stable@vger.kernel.org> # v4.12+
      Fixes: 9fdca4da ("IB/SA: Split struct sa_path_rec based on IB and ROCE specific fields")
      Signed-off-by: default avatarParav Pandit <parav@mellanox.com>
      Reviewed-by: default avatarMoni Shoua <monis@mellanox.com>
      Signed-off-by: default avatarLeon Romanovsky <leon@kernel.org>
      Signed-off-by: default avatarDoug Ledford <dledford@redhat.com>
      5a3dc323
    • Maor Gottlieb's avatar
      IB/mlx5: Add support for RSS on the inner packet · 309fa347
      Maor Gottlieb authored
      Some user space application would like to do RSS on the inner
      packet fields instead on the outer.
      When MLX5_RX_HASH_INNER is set with one or more of the other
      hash fields, then the RSS will be done using the inner packet.
      Signed-off-by: default avatarMaor Gottlieb <maorg@mellanox.com>
      Reviewed-by: default avatarMark Bloch <markb@mellanox.com>
      Signed-off-by: default avatarLeon Romanovsky <leon@kernel.org>
      Signed-off-by: default avatarDoug Ledford <dledford@redhat.com>
      309fa347