Merge branch 's390-next'

Julian Wiedmann says:

====================
s390/qeth: updates 2019-04-17

please apply some additional qeth patches to net-next. This patchset
converts the driver to use th

Merge branch 's390-next'

Julian Wiedmann says:

====================
s390/qeth: updates 2019-04-17

please apply some additional qeth patches to net-next. This patchset
converts the driver to use the kernel's multiqueue model.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>

show more ...

s390/qeth: stop/wake TX queues based on their fill level

Current xmit code only stops the txq after attempting to fill an
IO buffer that hasn't been TX-completed yet. In many-connection
scenarios, t

s390/qeth: stop/wake TX queues based on their fill level

Current xmit code only stops the txq after attempting to fill an
IO buffer that hasn't been TX-completed yet. In many-connection
scenarios, this can result in frequent rejected TX attempts, requeuing
of skbs with NETDEV_TX_BUSY and extra overhead.

Now that we have a proper 1-to-1 relation between stack-side txqs and
our HW Queues, overhaul the stop/wake logic so that the xmit code
stops the txq as needed.
Given that we might map multiple skbs into a single buffer, it's crucial
to ensure that the queue always provides an _entirely_ empty IO buffer.
Otherwise large skbs (eg TSO) might not fit into the last available
buffer. So whenever qeth_do_send_packet() first utilizes an _empty_
buffer, it updates & checks the used_buffers count.

This now ensures that an skb passed to qeth_xmit() can always be mapped
into an IO buffer, so remove all of the -EBUSY roll-back handling in the
TX path. We preserve the minimal safety-checks ("Is this IO buffer
really available?"), just in case some nasty future bug ever attempts to
corrupt an in-use buffer.

Signed-off-by: Julian Wiedmann <jwi@linux.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

show more ...

s390/qeth: add TX multiqueue support for IQD devices

qeth has been supporting multiple HW Output Queues for a long time. But
rather than exposing those queues to the stack, it uses its own queue
sel

s390/qeth: add TX multiqueue support for IQD devices

qeth has been supporting multiple HW Output Queues for a long time. But
rather than exposing those queues to the stack, it uses its own queue
selection logic in .ndo_start_xmit... with all the drawbacks that
entails.
Start off by switching IQD devices over to a proper mqs net_device,
and converting all the netdev_queue management code.

One oddity with IQD devices is the requirement to place all mcast
traffic on the _highest_ established HW queue. Doing so via
.ndo_select_queue seems straight-forward - but that won't work if only
some of the HW queues are active
(ie. when dev->real_num_tx_queues < dev->num_tx_queues), since
netdev_cap_txqueue() will not allow us to put skbs on the higher queues.

To make this work, we
1. let .ndo_select_queue() map all mcast traffic to netdev_queue 0, and
2. later re-map the netdev_queue and HW queue indices in
.ndo_start_xmit and the TX completion handler.

With this patch we default to a fixed set of 1 ucast and 1 mcast queue.
Support for dynamic reconfiguration is added at a later time.

Signed-off-by: Julian Wiedmann <jwi@linux.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

show more ...

Merge tag 'asoc-fix-v5.1-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/sound into for-linus

ASoC: Fixes for v5.1

A few core fixes along with the driver specific ones, mainly fixing

Merge tag 'asoc-fix-v5.1-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/sound into for-linus

ASoC: Fixes for v5.1

A few core fixes along with the driver specific ones, mainly fixing
small issues that only affect x86 platforms for various reasons (their
unusual machine enumeration mechanisms mainly, plus a fix for error
handling in topology).

There's some of the driver fixes that look larger than they are, like
the hdmi-codec changes which resulted in an indentation change, and most
of the other large changes are for new drivers like the STM32 changes.

show more ...

Merge tag 'v5.1-rc1' into asoc-5.1

Linux 5.1-rc1

Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next

Pull networking updates from David Miller:
"Here we go, another merge window full of networking and #ebpf changes:

1) Snoop DH

Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next

Pull networking updates from David Miller:
"Here we go, another merge window full of networking and #ebpf changes:

1) Snoop DHCPACKS in batman-adv to learn MAC/IP pairs in the DHCP
range without dealing with floods of ARP traffic, from Linus
Lüssing.

2) Throttle buffered multicast packet transmission in mt76, from
Felix Fietkau.

3) Support adaptive interrupt moderation in ice, from Brett Creeley.

4) A lot of struct_size conversions, from Gustavo A. R. Silva.

5) Add peek/push/pop commands to bpftool, as well as bash completion,
from Stanislav Fomichev.

6) Optimize sk_msg_clone(), from Vakul Garg.

7) Add SO_BINDTOIFINDEX, from David Herrmann.

8) Be more conservative with local resends due to local congestion,
from Yuchung Cheng.

9) Allow vetoing of unsupported VXLAN FDBs, from Petr Machata.

10) Add health buffer support to devlink, from Eran Ben Elisha.

11) Add TXQ scheduling API to mac80211, from Toke Høiland-Jørgensen.

12) Add statistics to basic packet scheduler filter, from Cong Wang.

13) Add GRE tunnel support for mlxsw Spectrum-2, from Nir Dotan.

14) Lots of new IP tunneling forwarding tests, also from Nir Dotan.

15) Add 3ad stats to bonding, from Nikolay Aleksandrov.

16) Lots of probing improvements for bpftool, from Quentin Monnet.

17) Various nfp drive #ebpf JIT improvements from Jakub Kicinski.

18) Allow #ebpf programs to access gso_segs from skb shared info, from
Eric Dumazet.

19) Add sock_diag support for AF_XDP sockets, from Björn Töpel.

20) Support 22260 iwlwifi devices, from Luca Coelho.

21) Use rbtree for ipv6 defragmentation, from Peter Oskolkov.

22) Add JMP32 instruction class support to #ebpf, from Jiong Wang.

23) Add spinlock support to #ebpf, from Alexei Starovoitov.

24) Support 256-bit keys and TLS 1.3 in ktls, from Dave Watson.

25) Add device infomation API to devlink, from Jakub Kicinski.

26) Add new timestamping socket options which are y2038 safe, from
Deepa Dinamani.

27) Add RX checksum offloading for various sh_eth chips, from Sergei
Shtylyov.

28) Flow offload infrastructure, from Pablo Neira Ayuso.

29) Numerous cleanups, improvements, and bug fixes to the PHY layer
and many drivers from Heiner Kallweit.

30) Lots of changes to try and make packet scheduler classifiers run
lockless as much as possible, from Vlad Buslov.

31) Support BCM957504 chip in bnxt_en driver, from Erik Burrows.

32) Add concurrency tests to tc-tests infrastructure, from Vlad
Buslov.

33) Add hwmon support to aquantia, from Heiner Kallweit.

34) Allow 64-bit values for SO_MAX_PACING_RATE, from Eric Dumazet.

And I would be remiss if I didn't thank the various major networking
subsystem maintainers for integrating much of this work before I even
saw it. Alexei Starovoitov, Daniel Borkmann, Pablo Neira Ayuso,
Johannes Berg, Kalle Valo, and many others. Thank you!"

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next: (2207 commits)
net/sched: avoid unused-label warning
net: ignore sysctl_devconf_inherit_init_net without SYSCTL
phy: mdio-mux: fix Kconfig dependencies
net: phy: use phy_modify_mmd_changed in genphy_c45_an_config_aneg
net: dsa: mv88e6xxx: add call to mv88e6xxx_ports_cmode_init to probe for new DSA framework
selftest/net: Remove duplicate header
sky2: Disable MSI on Dell Inspiron 1545 and Gateway P-79
net/mlx5e: Update tx reporter status in case channels were successfully opened
devlink: Add support for direct reporter health state update
devlink: Update reporter state to error even if recover aborted
sctp: call iov_iter_revert() after sending ABORT
team: Free BPF filter when unregistering netdev
ip6mr: Do not call __IP6_INC_STATS() from preemptible context
isdn: mISDN: Fix potential NULL pointer dereference of kzalloc
net: dsa: mv88e6xxx: support in-band signalling on SGMII ports with external PHYs
cxgb4/chtls: Prefix adapter flags with CXGB4
net-sysfs: Switch to bitmap_zalloc()
mellanox: Switch to bitmap_zalloc()
bpf: add test cases for non-pointer sanitiation logic
mlxsw: i2c: Extend initialization by querying resources data
...

show more ...

Merge remote-tracking branch 'net-next/master' into mac80211-next

Merge net-next to resolve a conflict and to get the mac80211
rhashtable fixes so further patches can be applied on top.

Signed-off-

Merge remote-tracking branch 'net-next/master' into mac80211-next

Merge net-next to resolve a conflict and to get the mac80211
rhashtable fixes so further patches can be applied on top.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>

show more ...

Merge branch 's390-next'

Julian Wiedmann says:

====================
s390/qeth: updates 2019-02-15

please apply a few more qeth patches to net-next. Along with some smaller
improvements, this revam

Merge branch 's390-next'

Julian Wiedmann says:

====================
s390/qeth: updates 2019-02-15

please apply a few more qeth patches to net-next. Along with some smaller
improvements, this revamps our code for the SW statistics that are exposed
through ETHTOOL_GSTATS.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>

show more ...

s390/qeth: add support for ETHTOOL_GRINGPARAM

Implement a trivial callback that exposes the queue sizes.

Signed-off-by: Julian Wiedmann <jwi@linux.ibm.com>
Signed-off-by: David S. Miller <davem@dav

s390/qeth: add support for ETHTOOL_GRINGPARAM

Implement a trivial callback that exposes the queue sizes.

Signed-off-by: Julian Wiedmann <jwi@linux.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

show more ...

s390/qeth: overhaul ethtool statistics

Accumulate per-TX queue statistics, and increase their size to 64 bit.
Don't bother with enabling/disabling the statistics, the overhead is
negligible.

Signed

s390/qeth: overhaul ethtool statistics

Accumulate per-TX queue statistics, and increase their size to 64 bit.
Don't bother with enabling/disabling the statistics, the overhead is
negligible.

Signed-off-by: Julian Wiedmann <jwi@linux.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

show more ...

s390/qeth: move ethtool code into its own file

Most of this is self-contained code.

Signed-off-by: Julian Wiedmann <jwi@linux.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>