This bug was fixed in the package linux - 4.8.0-49.52

---------------
linux (4.8.0-49.52) yakkety; urgency=low

  * linux: 4.8.0-49.52 -proposed tracker (LP: #1684427)

  * [Hyper-V] hv: util: move waiting for release to hv_utils_transport itself
    (LP: #1682561)
    - Drivers: hv: util: move waiting for release to hv_utils_transport itself

linux (4.8.0-48.51) yakkety; urgency=low

  * linux: 4.8.0-48.51 -proposed tracker (LP: #1682034)

  * [Hyper-V] hv: vmbus: Raise retry/wait limits in vmbus_post_msg()
    (LP: #1681893)
    - Drivers: hv: vmbus: Raise retry/wait limits in vmbus_post_msg()

linux (4.8.0-47.50) yakkety; urgency=low

  * linux: 4.8.0-47.50 -proposed tracker (LP: #1679678)

  * CVE-2017-6353
    - sctp: deny peeloff operation on asocs with threads sleeping on it

  * CVE-2017-5986
    - sctp: avoid BUG_ON on sctp_wait_for_sndbuf

  * vfat: missing iso8859-1 charset (LP: #1677230)
    - [Config] NLS_ISO8859_1=y

  * [Hyper-V] pci-hyperv: Use device serial number as PCI domain (LP: #1667527)
    - net/mlx4_core: Use cq quota in SRIOV when creating completion EQs

  * Regression: KVM modules should be on main kernel package (LP: #1678099)
    - [Config] powerpc: Add kvm-hv and kvm-pr to the generic inclusion list

  * linux-lts-xenial 4.4.0-63.84~14.04.2 ADT test failure with linux-lts-xenial
    4.4.0-63.84~14.04.2 (LP: #1664912)
    - SAUCE: apparmor: fix link auditing failure due to, uninitialized var

  * regession tests failing after stackprofile test is run (LP: #1661030)
    - SAUCE: fix regression with domain change in complain mode

  * Permission denied and inconsistent behavior in complain mode with 'ip netns
    list' command (LP: #1648903)
    - SAUCE: fix regression with domain change in complain mode

  * unexpected errno=13 and disconnected path when trying to open /proc/1/ns/mnt
    from a unshared mount namespace (LP: #1656121)
    - SAUCE: apparmor: null profiles should inherit parent control flags

  * apparmor refcount leak of profile namespace when removing profiles
    (LP: #1660849)
    - SAUCE: apparmor: fix ns ref count link when removing profiles from policy

  * tor in lxd: apparmor="DENIED" operation="change_onexec"
    namespace="root//CONTAINERNAME_<var-lib-lxd>" profile="unconfined"
    name="system_tor" (LP: #1648143)
    - SAUCE: apparmor: Fix no_new_privs blocking change_onexec when using 
stacked
      namespaces

  * apparmor oops in bind_mnt when dev_path lookup fails (LP: #1660840)
    - SAUCE: apparmor: fix oops in bind_mnt when dev_path lookup fails

  * apparmor  auditing denied access of special apparmor .null fi\ le
    (LP: #1660836)
    - SAUCE: apparmor: Don't audit denied access of special apparmor .null file

  * apparmor label leak when new label is unused (LP: #1660834)
    - SAUCE: apparmor: fix label leak when new label is unused

  * apparmor reference count bug in label_merge_insert() (LP: #1660833)
    - SAUCE: apparmor: fix reference count bug in label_merge_insert()

  * apparmor's raw_data file in securityfs is sometimes truncated (LP: #1638996)
    - SAUCE: apparmor: fix replacement race in reading rawdata

  * unix domain socket cross permission check failing with nested namespaces
    (LP: #1660832)
    - SAUCE: apparmor: fix cross ns perm of unix domain sockets

  * [Hyper-V][Mellanox] net/mlx4_core: Avoid delays during VF driver device
    shutdown (LP: #1672785)
    - Revert "net/mlx4_en: Avoid unregister_netdev at shutdown flow"
    - net/mlx4_core: Avoid delays during VF driver device shutdown

  * Update ENA driver to 1.1.2 from net-next (LP: #1664312)
    - net: ena: Remove unnecessary pci_set_drvdata()
    - net: ena: Fix error return code in ena_device_init()
    - net: ena: change the return type of ena_set_push_mode() to be void.
    - net: ena: use setup_timer() and mod_timer()
    - net/ena: remove ntuple filter support from device feature list
    - net/ena: fix queues number calculation
    - net/ena: fix ethtool RSS flow configuration
    - net/ena: fix RSS default hash configuration
    - net/ena: fix NULL dereference when removing the driver after device reset
      failed
    - net/ena: refactor ena_get_stats64 to be atomic context safe
    - net/ena: fix potential access to freed memory during device reset
    - net/ena: use READ_ONCE to access completion descriptors
    - net/ena: reduce the severity of ena printouts
    - net/ena: change driver's default timeouts
    - net/ena: change condition for host attribute configuration
    - net/ena: update driver version to 1.1.2

  * ISST-LTE:pVM:roselp4:ubuntu16.04.2: number of numa_miss and numa_foreign
    wrong in numastat (LP: #1672953)
    - mm: fix remote numa hits statistics
    - mm: get rid of __GFP_OTHER_NODE

  * Using an NVMe drive causes huge power drain (LP: #1664602)
    - nvme/scsi: Remove power management support
    - nvme: Pass pointers, not dma addresses, to nvme_get/set_features()
    - nvme: introduce struct nvme_request
    - nvme: Add a quirk mechanism that uses identify_ctrl
    - nvme: Enable autonomous power state transitions

  * POWER9: Additional patches for TTY and CPU_IDLE (LP: #1674325)
    - tty: Fix ldisc crash on reopened tty
    - SAUCE: powerpc/powernv/cpuidle: Pass correct drv->cpumask for registration

  * Ubuntu 16.10: Network checksum fixes needed for IPoIB for Mellanox CX4/CX5
    card (LP: #1670247)
    - Revert "powerpc: port 64 bits pgtable_cache to 32 bits"
    - powerpc/Makefile: Drop CONFIG_WORD_SIZE for BITS
    - powerpc: port 64 bits pgtable_cache to 32 bits
    - [Config] CONFIG_WORD_SIZE disappeared
    - powerpc/64: Fix checksum folding in csum_tcpudp_nofold and
      ip_fast_csum_nofold
    - powerpc/64: Use optimized checksum routines on little-endian
    - CONFIG_GENERIC_CSUM=n for ppc64el
    - powerpc/64: Fix checksum folding in csum_add()

  * [Hyper-V] Rebase Hyper-V to the upstream 4.10 kernel (LP: #1670544)
    - PCI: hv: Use device serial number as PCI domain
    - PCI: hv: Fix wslot_to_devfn() to fix warnings on device removal
    - PCI: hv: Use the correct buffer size in new_pcichild_device()
    - scsi: storvsc: Payload buffer incorrectly sized for 32 bit kernels.
    - hv_netvsc: remove excessive logging on MTU change
    - net: centralize net_device min/max MTU checking
    - net: deprecate eth_change_mtu, remove usage
    - net: use core MTU range checking in virt drivers
    - hv_netvsc: fix a race between netvsc_send() and netvsc_init_buf()
    - net: use core MTU range checking in virt drivers
    - tools: hv: fix a compile warning in snprintf
    - tools: hv: remove unnecessary header files and netlink related code
    - vmbus: add support for dynamic device id's
    - Drivers: hv: utils: reduce HV_UTIL_NEGO_TIMEOUT timeout
    - Drivers: hv: utils: Fix the mapping between host version and protocol to 
use
    - Drivers: hv: vss: Improve log messages.
    - hv: change clockevents unbind tactics
    - Drivers: hv: balloon: Disable hot add when CONFIG_MEMORY_HOTPLUG is not 
set
    - Drivers: hv: balloon: Fix info request to show max page count
    - Drivers: hv: balloon: Add logging for dynamic memory operations
    - [Config] CONFIG_UIO_HV_GENERIC=m
    - uio-hv-generic: new userspace i/o driver for VMBus
    - hyperv: Fix spelling of HV_UNKOWN
    - Drivers: hv: ring_buffer: count on wrap around mappings in
      get_next_pkt_raw() (v2)
    - ethernet: use net core MTU range checking in more drivers

  * Kernel linux-image-4.4.0-67-generic prevent the boot on Microsoft Hyper-v
    2012r2 Gen2 VM (LP: #1674635)
    - scsi: storvsc: Workaround for virtual DVD SCSI version

  * Enable lspcon on i915 (LP: #1676747)
    - drm: Helper for lspcon in drm_dp_dual_mode
    - drm/i915: Add lspcon support for I915 driver
    - drm/i915: Parse VBT data for lspcon
    - drm/i915: Enable lspcon initialization
    - drm/i915: Add lspcon resume function

  * stress_smoke_test passing and exiting rc=9 (linux 4.9.0-12.13 ADT test
    failure with linux 4.9.0-12.13) (LP: #1658633)
    - ext4: lock the xattr block before checksuming it

  * ip_rcv_finish() NULL pointer kernel panic (LP: #1672470)
    - (upstream) bridge: drop netfilter fake rtable unconditionally

  * dm-queue-length module is not included in installer/initramfs (LP: #1673350)
    - d-i: Also add dm-queue-length to multipath modules

  * Broadcom bluetooth modules sometimes fail to initialize (LP: #1483101)
    - Bluetooth: btbcm: Add a delay for module reset

  * Need support of Broadcom bluetooth device [413c:8143] (LP: #1166113)
    - Bluetooth: btusb: Add support for 413c:8143

  * Unable to Connect Third HDD via USB Hub (LP: #1663991)
    - mm/slub.c: fix random_seq offset destruction

  * POWER9 : Enable Stop 0-2 with ESL=EC=0 (LP: #1666197)
    - powernv:idle: Add IDLE_STATE_ENTER_SEQ_NORET macro
    - powernv:stop: Rename pnv_arch300_idle_init to pnv_power9_idle_init
    - cpuidle:powernv: Add helper function to populate powernv idle states.
    - powernv: Pass PSSCR value and mask to power9_idle_stop
    - Documentation:powerpc: Add device-tree bindings for power-mgt
    - powerpc/powernv: Fix bug due to labeling ambiguity in power_enter_stop

  * Nvlink2: Additional patches (LP: #1667081)
    - mm: enable CONFIG_MOVABLE_NODE on non-x86 arches
    - of/fdt: mark hotpluggable memory
    - dt: add documentation of "hotpluggable" memory property
    - powerpc/mm: Fix memory hotplug BUG() on radix
    - powerpc/powernv: Initialise nest mmu
    - powerpc/powernv: Use OPAL call for TCE kill on NVLink2
    - powerpc/mm: refactor radix physical page mapping
    - powerpc/mm: add radix__create_section_mapping()
    - powerpc/mm: add radix__remove_section_mapping()
    - powerpc/mm: unstub radix__vmemmap_remove_mapping()
    - [Config] Update CONFIG_MOVABLE_NODE values and annotations
    - [Config] CONFIG_MOVABLE_NODE=n for s390x

  * FC Adapter (LPe32000-based) prints "iotag out of range", goes offline, and
    delays boot a lot (Ubuntu17.04/Emulex/lpfc)) (LP: #1670490)
    - scsi: lpfc: Correct WQ creation for pagesize
    - scsi: lpfc: Add missing memory barrier

  * CIFS: Call echo service immediately after socket reconnect (LP: #1669941)
    - Call echo service immediately after socket reconnect

  * Kernel: Fix Transactional memory config typo (LP: #1669023)
    - powerpc/process: Fix CONFIG_ALIVEC typo in restore_tm_state()

  * h-prod does not function across cores (LP: #1670726)
    - KVM: PPC: Book3S HV: Fix H_PROD to actually wake the target vcpu

  * [Hyper-V] Missing PCI patches breaking SR-IOV hot remove (LP: #1670518)
    - PCI: hv: Fix hv_pci_remove() for hot-remove
    - PCI: hv: Delete the device earlier from hbus->children for hot-remove
    - PCI: hv: Make unnecessarily global IRQ masking functions static
    - PCI: hv: Allocate physically contiguous hypercall params buffer

  * move aufs.ko from -extra to linux-image package (LP: #1673498)
    - [config] aufs.ko moved to linux-image package

  * POWER9: Improve CAS negotiation (LP: #1671169)
    - powerpc: Parse the command line before calling CAS
    - powerpc: Add missing error check to prom_find_boot_cpu()
    - powerpc/pseries: Advertise HPT resizing support via CAS
    - powerpc/64: Disable use of radix under a hypervisor
    - powerpc/pseries: Advertise Hot Plug Event support to firmware
    - powerpc: Update to new option-vector-5 format for CAS

  * Power9 kernel: add virtualization patches (LP: #1670800)
    - powerpc/fadump: Set core e_flags using kernel's ELF ABI version
    - powerpc/sparse: Add more assembler prototypes
    - powerpc/pasemi: Fix Nemo SB600 i8259 interrupts.
    - powerpc/pasemi: Fix device_type of Nemo SB600 node.
    - powerpc/pseries: Use H_CLEAR_HPT to clear MMU hash table during kexec
    - powerpc/pseries: Move CMO code from plapr_wrappers.h to platforms/pseries
    - powerpc: Fix old style declaration GCC warnings
    - powerpc/pseries: add definitions for new H_SIGNAL_SYS_RESET hcall
    - powerpc/prom: Define structs for client architecture vectors
    - powerpc/prom: Switch to using structs for ibm_architecture_vec
    - tracing: Have the reg function allow to fail
    - powerpc: port 64 bits pgtable_cache to 32 bits
    - powerpc/64: Don't try to use radix MMU under a hypervisor
    - powerpc/pseries: Fixes for the "ibm,architecture-vec-5" options
    - powerpc/64: Enable use of radix MMU under hypervisor on POWER9

  * lsattr 32bit does not work on 64bit kernel (Inappropriate ioctl error)
    (LP: #1619918)
    - btrfs: fix btrfs_compat_ioctl failures on non-compat ioctls

  * linux-tools-common should Depends: lsb-release (LP: #1667571)
    - [Config] linux-tools-common depends on lsb-release

  * CAPI:Ubuntu: Kernel panic while rebooting (LP: #1667599)
    - pci/hotplug/pnv-php: Remove WARN_ON() in pnv_php_put_slot()

  * Add Use-After-Free Patch for Ubuntu16.10 - EEH on BELL3 adapter fails to
    recover (serial/tty) (LP: #1669153)
    - 8250_pci: Fix potential use-after-free in error path

  * Request to backport cxlflash patches to Xenial SRU stream (LP: #1623750)
    - scsi: cxlflash: Scan host only after the port is ready for I/O
    - scsi: cxlflash: Fix to avoid EEH and host reset collisions
    - scsi: cxlflash: Improve EEH recovery time

  * FlashGT Integration and Setup: fsbmc30: After 17th reboot of soft bootme,
    HTX & Linux errors seen with 256 virtual LUNs (LP: #1667239)
    - cxl: Fix coredump generation when cxl_get_fd() is used

  * POWER9: Additional patches for 17.04 and 16.04.2 (LP: #1667116)
    - powerpc/mm: Update PROTFAULT handling in the page fault path
    - powerpc/mm/radix: Update pte update sequence for pte clear case
    - powerpc/mm/radix: Use ptep_get_and_clear_full when clearing pte for full 
mm
    - powerpc/mm/radix: Skip ptesync in pte update helpers
    - SAUCE: powerpc/mm/hash: Always clear UPRT and Host Radix bits when setting
      up CPU

  * [Hyper-V] Ubuntu 14.04.2 LTS Generation 2 SCSI Errors on VSS Based Backups
    (LP: #1470250)
    - Drivers: hv: vss: Operation timeouts should match host expectation
    - SAUCE: Tools: hv: vss: Thaw the filesystem and continue after freeze fails

  * PowerNV: No rate limit for kernel error "KVM can't copy data from"
    (LP: #1667416)
    - SAUCE: KVM: PPC: Book3S: Ratelimit copy data failure error messages

  * kernel 4.4.0-63 with USB WLAN RTL8192CU freezes desktop (LP: #1666421)
    - rtlwifi: rtl_usb: Fix missing entry in USB driver's private data

  * Export symbol "dev_pm_qos_update_user_latency_tolerance" (LP: #1666401)
    - PM / QoS: Export dev_pm_qos_update_user_latency_tolerance

  * Linux ZFS port doesn't respect RLIMIT_FSIZE (LP: #1656259)
    - SAUCE: (noup) Update zfs to 0.6.5.8-0ubuntu4.2

 -- Stefan Bader <stefan.ba...@canonical.com>  Thu, 20 Apr 2017 09:38:36
+0200

-- 
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux in Ubuntu.
https://bugs.launchpad.net/bugs/1672785

Title:
  [Hyper-V][Mellanox] net/mlx4_core: Avoid delays during VF driver
  device shutdown

Status in linux package in Ubuntu:
  Fix Released
Status in linux source package in Xenial:
  Fix Released
Status in linux source package in Yakkety:
  Fix Released
Status in linux source package in Zesty:
  Fix Released

Bug description:
  Mellanox has submitted the following patch upstream that's important
  for SR-IOV in Azure.

  Please integrate it into the Mellanox mlx4 drivers for lts-xenial,
  HWE, Zesty, and Azure custom.

  https://patchwork.ozlabs.org/patch/738305/

  From: Jack Morgenstein <ja...@dev.mellanox.co.il>

  Some Hypervisors detach VFs from VMs by instantly causing an FLR event
  to be generated for a VF.

  In the mlx4 case, this will cause that VF's comm channel to be disabled
  before the VM has an opportunity to invoke the VF device's "shutdown"
  method.

  For such Hypervisors, there is a race condition between the VF's
  shutdown method and its internal-error detection/reset thread.

  The internal-error detection/reset thread (which runs every 5 seconds) also
  detects a disabled comm channel. If the internal-error detection/reset
  flow wins the race, we still get delays (while that flow tries repeatedly
  to detect comm-channel recovery).

  The cited commit fixed the command timeout problem when the
  internal-error detection/reset flow loses the race.

  This commit avoids the unneeded delays when the internal-error
  detection/reset flow wins.

  Fixes: d585df1c5ccf ("net/mlx4_core: Avoid command timeouts during VF driver 
device shutdown")
  Signed-off-by: Jack Morgenstein <ja...@dev.mellanox.co.il>
  Reported-by: Simon Xiao <six...@microsoft.com>
  Signed-off-by: Tariq Toukan <tar...@mellanox.com>
  ---
   drivers/net/ethernet/mellanox/mlx4/cmd.c  | 11 +++++++++++
   drivers/net/ethernet/mellanox/mlx4/main.c | 11 +++++++++++
   include/linux/mlx4/device.h               |  1 +
   3 files changed, 23 insertions(+)

To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1672785/+subscriptions

-- 
Mailing list: https://launchpad.net/~kernel-packages
Post to     : kernel-packages@lists.launchpad.net
Unsubscribe : https://launchpad.net/~kernel-packages
More help   : https://help.launchpad.net/ListHelp

Reply via email to