This bug was fixed in the package linux-azure - 4.15.0-1096.106~16.04.1
---------------
linux-azure (4.15.0-1096.106~16.04.1) xenial; urgency=medium
* xenial/linux-azure: 4.15.0-1096.106~16.04.1 -proposed tracker (LP:
#1894681)
[ Ubuntu: 4.15.0-1096.106 ]
* bionic/linux-azure-4.15: 4.15.0-1096.106 -proposed tracker (LP: #1894683)
* Mellanox Check For Write Combining Support (LP: #1874503)
- RDMA/rxe: Close a race after ib_register_device
- IB/mlx5: Align usage of QP1 create flags with rest of mlx5 defines
- IB/mlx5: Test write combining support
* Enable Invariant TSC Support (LP: #1875467)
- x86/hyperv: Allow guests to enable InvariantTSC
- x86/hyperv: Set TSC clocksource as default w/ InvariantTSC
* Only notify Hyper-V for die events that are oops (LP: #1891222)
- Drivers: hv: vmbus: Only notify Hyper-V for die events that are oops
* speed for CX4 VF showing as unknown in ethtool output (LP: #1876770)
- net/mlx5: Expose link speed directly
- net/mlx5: Expose port speed when possible
- net/mlx5: Tidy up and fix reverse christmas ordring
* bionic/linux: 4.15.0-118.119 -proposed tracker (LP: #1894697)
* Packaging resync (LP: #1786013)
- update dkms package versions
* Introduce the new NVIDIA 450-server and the 450 UDA series (LP: #1887674)
- [packaging] add signed modules for nvidia 450 and 450-server
* cgroup refcount is bogus when cgroup_sk_alloc is disabled (LP: #1886860)
- cgroup: add missing skcd->no_refcnt check in cgroup_sk_clone()
* CVE-2020-12888
- vfio/type1: Support faulting PFNMAP vmas
- vfio-pci: Fault mmaps to enable vma tracking
- vfio-pci: Invalidate mmaps and block MMIO access on disabled memory
* [Hyper-V] VSS and File Copy daemons intermittently fails to start
(LP: #1891224)
- [Packaging] Bind hv_vss_daemon startup to hv_vss device
- [Packaging] bind hv_fcopy_daemon startup to hv_fcopy device
* KVM: Fix zero_page reference counter overflow when using KSM on KVM compute
host (LP: #1837810)
- KVM: fix overflow of zero page refcount with ksm running
* Fix false-negative return value for rtnetlink.sh in kselftests/net
(LP: #1890136)
- selftests: rtnetlink: correct the final return value for the test
- selftests: rtnetlink: make kci_test_encap() return sub-test result
* Bionic update: upstream stable patchset 2020-08-18 (LP: #1892091)
- USB: serial: qcserial: add EM7305 QDL product ID
- USB: iowarrior: fix up report size handling for some devices
- usb: xhci: define IDs for various ASMedia host controllers
- usb: xhci: Fix ASMedia ASM1142 DMA addressing
- Revert "ALSA: hda: call runtime_allow() for all hda controllers"
- ALSA: seq: oss: Serialize ioctls
- staging: android: ashmem: Fix lockdep warning for write operation
- Bluetooth: Fix slab-out-of-bounds read in
hci_extended_inquiry_result_evt()
- Bluetooth: Prevent out-of-bounds read in hci_inquiry_result_evt()
- Bluetooth: Prevent out-of-bounds read in
hci_inquiry_result_with_rssi_evt()
- omapfb: dss: Fix max fclk divider for omap36xx
- binder: Prevent context manager from incrementing ref 0
- vgacon: Fix for missing check in scrollback handling
- mtd: properly check all write ioctls for permissions
- leds: wm831x-status: fix use-after-free on unbind
- leds: da903x: fix use-after-free on unbind
- leds: lm3533: fix use-after-free on unbind
- leds: 88pm860x: fix use-after-free on unbind
- net/9p: validate fds in p9_fd_open
- drm/nouveau/fbcon: fix module unload when fbcon init has failed for some
reason
- drm/nouveau/fbcon: zero-initialise the mode_cmd2 structure
- i2c: slave: improve sanity check when registering
- i2c: slave: add sanity check when unregistering
- usb: hso: check for return value in hso_serial_common_create()
- firmware: Fix a reference count leak.
- cfg80211: check vendor command doit pointer before use
- igb: reinit_locked() should be called with rtnl_lock
- atm: fix atm_dev refcnt leaks in atmtcp_remove_persistent
- tools lib traceevent: Fix memory leak in process_dynamic_array_len
- Drivers: hv: vmbus: Ignore CHANNELMSG_TL_CONNECT_RESULT(23)
- xattr: break delegations in {set,remove}xattr
- ipv4: Silence suspicious RCU usage warning
- ipv6: fix memory leaks on IPV6_ADDRFORM path
- net: ethernet: mtk_eth_soc: fix MTU warnings
- vxlan: Ensure FDB dump is performed under RCU
- net: lan78xx: replace bogus endpoint lookup
- hv_netvsc: do not use VF device if link is down
- net: gre: recompute gre csum for sctp over gre tunnels
- openvswitch: Prevent kernel-infoleak in ovs_ct_put_key()
- Revert "vxlan: fix tos value before xmit"
- selftests/net: relax cpu affinity requirement in msg_zerocopy test
- rxrpc: Fix race between recvmsg and sendmsg on immediate call failure
- i40e: add num_vectors checker in iwarp handler
- i40e: Wrong truncation from u16 to u8
- i40e: Memory leak in i40e_config_iwarp_qvlist
- Smack: fix use-after-free in smk_write_relabel_self()
* Bionic update: upstream stable patchset 2020-08-11 (LP: #1891228)
- AX.25: Fix out-of-bounds read in ax25_connect()
- AX.25: Prevent out-of-bounds read in ax25_sendmsg()
- dev: Defer free of skbs in flush_backlog
- drivers/net/wan/x25_asy: Fix to make it work
- net-sysfs: add a newline when printing 'tx_timeout' by sysfs
- net: udp: Fix wrong clean up for IS_UDPLITE macro
- rxrpc: Fix sendmsg() returning EPIPE due to recvmsg() returning ENODATA
- AX.25: Prevent integer overflows in connect and sendmsg
- ip6_gre: fix null-ptr-deref in ip6gre_init_net()
- rtnetlink: Fix memory(net_device) leak when ->newlink fails
- tcp: allow at most one TLP probe per flight
- regmap: debugfs: check count when read regmap file
- qrtr: orphan socket in qrtr_release()
- sctp: shrink stream outq only when new outcnt < old outcnt
- sctp: shrink stream outq when fails to do addstream reconf
- crypto: ccp - Release all allocated memory if sha type is invalid
- media: rc: prevent memory leak in cx23888_ir_probe
- iio: imu: adis16400: fix memory leak
- ath9k_htc: release allocated buffer if timed out
- ath9k: release allocated buffer if timed out
- PCI/ASPM: Disable ASPM on ASMedia ASM1083/1085 PCIe-to-PCI bridge
- wireless: Use offsetof instead of custom macro.
- ARM: 8986/1: hw_breakpoint: Don't invoke overflow handler on uaccess
watchpoints
- drm/amdgpu: Prevent kernel-infoleak in amdgpu_info_ioctl()
- drm: hold gem reference until object is no longer accessed
- f2fs: check memory boundary by insane namelen
- f2fs: check if file namelen exceeds max value
- 9p/trans_fd: abort p9_read_work if req status changed
- 9p/trans_fd: Fix concurrency del of req_list in
p9_fd_cancelled/p9_read_work
- x86/build/lto: Fix truncated .bss with -fdata-sections
- rds: Prevent kernel-infoleak in rds_notify_queue_get()
- xfs: fix missed wakeup on l_flush_wait
- net/x25: Fix x25_neigh refcnt leak when x25 disconnect
- net/x25: Fix null-ptr-deref in x25_disconnect
- selftests/net: rxtimestamp: fix clang issues for target arch PowerPC
- sh: Fix validation of system call number
- net: lan78xx: add missing endpoint sanity check
- net: lan78xx: fix transfer-buffer memory leak
- mlx4: disable device on shutdown
- mlxsw: core: Increase scope of RCU read-side critical section
- mlxsw: core: Free EMAD transactions using kfree_rcu()
- ibmvnic: Fix IRQ mapping disposal in error path
- bpf: Fix map leak in HASH_OF_MAPS map
- mac80211: mesh: Free ie data when leaving mesh
- mac80211: mesh: Free pending skb when destroying a mpath
- arm64/alternatives: move length validation inside the subsection
- arm64: csum: Fix handling of bad packets
- usb: hso: Fix debug compile warning on sparc32
- qed: Disable "MFW indication via attention" SPAM every 5 minutes
- nfc: s3fwrn5: add missing release on skb in s3fwrn5_recv_frame
- parisc: add support for cmpxchg on u8 pointers
- net: ethernet: ravb: exit if re-initialization fails in tx timeout
- Revert "i2c: cadence: Fix the hold bit setting"
- x86/unwind/orc: Fix ORC for newly forked tasks
- cxgb4: add missing release on skb in uld_send()
- xen-netfront: fix potential deadlock in xennet_remove()
- KVM: LAPIC: Prevent setting the tscdeadline timer if the lapic is hw
disabled
- x86/i8259: Use printk_deferred() to prevent deadlock
- drm/amdgpu: fix multiple memory leaks in acp_hw_init
- selftests/net: psock_fanout: fix clang issues for target arch PowerPC
- net/mlx5: Verify Hardware supports requested ptp function on a given pin
- random32: update the net random state on interrupt and activity
- ARM: percpu.h: fix build error
- random: fix circular include dependency on arm64 after addition of
percpu.h
- random32: remove net_rand_state from the latent entropy gcc plugin
- random32: move the pseudo-random 32-bit definitions to prandom.h
- ext4: fix direct I/O read error
-- Kelsey Skunberg <kelsey.skunb...@canonical.com> Thu, 10 Sep 2020
11:06:27 -0600
** Changed in: linux-azure (Ubuntu)
Status: New => Fix Released
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux-azure in Ubuntu.
https://bugs.launchpad.net/bugs/1874503
Title:
Mellanox Check For Write Combining Support
Status in linux-azure package in Ubuntu:
Fix Released
Status in linux-azure-4.15 package in Ubuntu:
New
Status in linux-azure source package in Bionic:
Invalid
Status in linux-azure-4.15 source package in Bionic:
Fix Released
Status in linux-azure source package in Focal:
Fix Committed
Status in linux-azure-4.15 source package in Focal:
Invalid
Bug description:
[Impact]
Microsoft and Mellanox would like to request the following two commits
in the releases that run on Azure:
3f89b01f4bba IB/mlx5: Align usage of QP1 create flags with rest of mlx5
defines
11f552e21755 IB/mlx5: Test write combining support
These commits landed in mainline as of v5.5-rc1.
These commits will improved network performance of RDMA out of the VF
when WC can be used. The highest benefit is to RDMA and DPDK, but UCX
should also see improved latency.
The Mellanox driver uses WC to optimize posting work to the HCA, and
getting this wrong in either direction can cause a significant
performance loss. These patches prevent the possible performance
loss.
[Test Case]
Simply measure the network performance with the patches applied to
check if there was some gain.
[Regression Potential]
It affects mainly the mlx5 driver and IB, while many of the instances
do not make use of those, they are used by the high performance users.
To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/linux-azure/+bug/1874503/+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
