platform/kernel/linux-rpi.git
6 years agoMerge tag 'batadv-next-for-davem-20180524' of git://git.open-mesh.org/linux-merge
David S. Miller [Thu, 24 May 2018 15:00:24 +0000 (11:00 -0400)]
Merge tag 'batadv-next-for-davem-20180524' of git://git.open-mesh.org/linux-merge

Simon Wunderlich says:

====================
This feature/cleanup patchset includes the following patches:

 - bump version strings, by Simon Wunderlich

 - Disable batman-adv debugfs by default, by Sven Eckelmann

 - Improve handling mesh nodes with multicast optimizations disabled,
   by Linus Luessing

 - Avoid bool in structs, by Sven Eckelmann

 - Allocate less memory when debugfs is disabled, by Sven Eckelmann

 - Fix batadv_interface_tx return data type, by Luc Van Oostenryck

 - improve link speed handling for virtual interfaces, by Marek Lindner

 - Enable BATMAN V algorithm by default, by Marek Lindner
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agobpfilter: don't pass O_CREAT when opening console for debug
Jakub Kicinski [Thu, 24 May 2018 07:41:19 +0000 (00:41 -0700)]
bpfilter: don't pass O_CREAT when opening console for debug

Passing O_CREAT (00000100) to open means we should also pass file
mode as the third parameter.  Creating /dev/console as a regular
file may not be helpful anyway, so simply drop the flag when
opening debug_fd.

Fixes: d2ba09c17a06 ("net: add skeleton of bpfilter kernel module")
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agobpfilter: fix build dependency
Alexei Starovoitov [Thu, 24 May 2018 04:29:05 +0000 (21:29 -0700)]
bpfilter: fix build dependency

BPFILTER could have been enabled without INET causing this build error:
ERROR: "bpfilter_process_sockopt" [net/bpfilter/bpfilter.ko] undefined!

Fixes: d2ba09c17a06 ("net: add skeleton of bpfilter kernel module")
Reported-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/pablo/nf-next
David S. Miller [Wed, 23 May 2018 20:37:11 +0000 (16:37 -0400)]
Merge git://git./linux/kernel/git/pablo/nf-next

Pablo Neira Ayuso says:

====================
Netfilter updates for net-next

The following patchset contains Netfilter updates for your net-next
tree, they are:

1) Remove obsolete nf_log tracing from nf_tables, from Florian Westphal.

2) Add support for map lookups to numgen, random and hash expressions,
   from Laura Garcia.

3) Allow to register nat hooks for iptables and nftables at the same
   time. Patchset from Florian Westpha.

4) Timeout support for rbtree sets.

5) ip6_rpfilter works needs interface for link-local addresses, from
   Vincent Bernat.

6) Add nf_ct_hook and nf_nat_hook structures and use them.

7) Do not drop packets on packets raceing to insert conntrack entries
   into hashes, this is particularly a problem in nfqueue setups.

8) Address fallout from xt_osf separation to nf_osf, patches
   from Florian Westphal and Fernando Mancera.

9) Remove reference to struct nft_af_info, which doesn't exist anymore.
   From Taehee Yoo.

This batch comes with is a conflict between 25fd386e0bc0 ("netfilter:
core: add missing __rcu annotation") in your tree and 2c205dd3981f
("netfilter: add struct nf_nat_hook and use it") coming in this batch.
This conflict can be solved by leaving the __rcu tag on
__netfilter_net_init() - added by 25fd386e0bc0 - and remove all code
related to nf_nat_decode_session_hook - which is gone after
2c205dd3981f, as described by:

diff --cc net/netfilter/core.c
index e0ae4aae96f5,206fb2c4c319..168af54db975
--- a/net/netfilter/core.c
+++ b/net/netfilter/core.c
@@@ -611,7 -580,13 +611,8 @@@ const struct nf_conntrack_zone nf_ct_zo
  EXPORT_SYMBOL_GPL(nf_ct_zone_dflt);
  #endif /* CONFIG_NF_CONNTRACK */

- static void __net_init __netfilter_net_init(struct nf_hook_entries **e, int max)
 -#ifdef CONFIG_NF_NAT_NEEDED
 -void (*nf_nat_decode_session_hook)(struct sk_buff *, struct flowi *);
 -EXPORT_SYMBOL(nf_nat_decode_session_hook);
 -#endif
 -
+ static void __net_init
+ __netfilter_net_init(struct nf_hook_entries __rcu **e, int max)
  {
   int h;

I can also merge your net-next tree into nf-next, solve the conflict and
resend the pull request if you prefer so.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoMerge branch 'amd-xgbe-next'
David S. Miller [Wed, 23 May 2018 20:33:01 +0000 (16:33 -0400)]
Merge branch 'amd-xgbe-next'

Tom Lendacky says:

====================
amd-xgbe: AMD XGBE driver updates 2018-05-21

The following updates are included in this driver update series:

- Fix the debug output for the max channels count
- Read (once) and save the port property registers during probe
- Remove the use of the comm_owned field
- Remove unused SFP diagnostic support indicator field
- Add ethtool --module-info support
- Add ethtool --show-ring/--set-ring support
- Update the driver in preparation for ethtool --set-channels support
- Add ethtool --show-channels/--set-channels support
- Update the driver to always perform link training in KR mode
- Advertise FEC support when using a KR re-driver
- Update the BelFuse quirk to now support SGMII
- Improve 100Mbps auto-negotiation for BelFuse parts

This patch series is based on net-next.

---

Changes since v1:
- Update the --set-channels support to the use of the combined, rx and
  tx options as specified in the ethtool man page (in other words, don't
  create combined channels based on the min of the tx and rx channels
  specified).
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoamd-xgbe: Improve SFP 100Mbps auto-negotiation
Tom Lendacky [Wed, 23 May 2018 16:39:47 +0000 (11:39 -0500)]
amd-xgbe: Improve SFP 100Mbps auto-negotiation

After changing speed to 100Mbps as a result of auto-negotiation (AN),
some 10/100/1000Mbps SFPs indicate a successful link (no faults or loss
of signal), but cannot successfully transmit or receive data.  These
SFPs required an extra auto-negotiation (AN) after the speed change in
order to operate properly.  Add a quirk for these SFPs so that if the
outcome of the AN actually results in changing to a new speed, re-initiate
AN at that new speed.

Signed-off-by: Tom Lendacky <thomas.lendacky@amd.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoamd-xgbe: Update the BelFuse quirk to support SGMII
Tom Lendacky [Wed, 23 May 2018 16:39:39 +0000 (11:39 -0500)]
amd-xgbe: Update the BelFuse quirk to support SGMII

Instead of using a quirk to make the BelFuse 1GBT-SFP06 part look like
a 1000baseX part, program the SFP PHY to support SGMII and 10/100/1000
baseT.

Signed-off-by: Tom Lendacky <thomas.lendacky@amd.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoamd-xgbe: Advertise FEC support with the KR re-driver
Tom Lendacky [Wed, 23 May 2018 16:39:31 +0000 (11:39 -0500)]
amd-xgbe: Advertise FEC support with the KR re-driver

When a KR re-driver is present, indicate the FEC support is available
during auto-negotiation.

Signed-off-by: Tom Lendacky <thomas.lendacky@amd.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoamd-xgbe: Always attempt link training in KR mode
Tom Lendacky [Wed, 23 May 2018 16:39:21 +0000 (11:39 -0500)]
amd-xgbe: Always attempt link training in KR mode

Link training is always attempted when in KR mode, but the code is
structured to check if link training has been enabled before attempting
to perform it.  Since that check will always be true, simplify the code
to always enable and start link training during KR auto-negotiation.

Signed-off-by: Tom Lendacky <thomas.lendacky@amd.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoamd-xgbe: Add ethtool show/set channels support
Tom Lendacky [Wed, 23 May 2018 16:39:13 +0000 (11:39 -0500)]
amd-xgbe: Add ethtool show/set channels support

Add ethtool support to show and set the device channel configuration.
Changing the channel configuration will result in a device restart.

Signed-off-by: Tom Lendacky <thomas.lendacky@amd.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoamd-xgbe: Prepare for ethtool set-channel support
Tom Lendacky [Wed, 23 May 2018 16:39:04 +0000 (11:39 -0500)]
amd-xgbe: Prepare for ethtool set-channel support

In order to support being able to dynamically set/change the number of
Rx and Tx channels, update the code to:
 - Move alloc and free of device memory into callable functions
 - Move setting of the real number of Rx and Tx channels to device startup
 - Move mapping of the RSS channels to device startup

Signed-off-by: Tom Lendacky <thomas.lendacky@amd.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoamd-xgbe: Add ethtool show/set ring parameter support
Tom Lendacky [Wed, 23 May 2018 16:38:56 +0000 (11:38 -0500)]
amd-xgbe: Add ethtool show/set ring parameter support

Add ethtool support to show and set the number of the Rx and Tx ring
descriptors.  Changing the ring configuration will result in a device
restart.

Signed-off-by: Tom Lendacky <thomas.lendacky@amd.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoamd-xgbe: Add ethtool support to retrieve SFP module info
Tom Lendacky [Wed, 23 May 2018 16:38:46 +0000 (11:38 -0500)]
amd-xgbe: Add ethtool support to retrieve SFP module info

Add support to get SFP module information using ethtool.

Signed-off-by: Tom Lendacky <thomas.lendacky@amd.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoamd-xgbe: Remove field that indicates SFP diagnostic support
Tom Lendacky [Wed, 23 May 2018 16:38:38 +0000 (11:38 -0500)]
amd-xgbe: Remove field that indicates SFP diagnostic support

The driver currently sets an indication of whether the SFP supports, and
that the driver can obtain, diagnostics data.  This isn't currently used
by the driver and the logic to set this indicator is flawed because the
field is cleared each time the SFP is checked and only set when a new SFP
is detected.  Remove this field and the logic supporting it.

Signed-off-by: Tom Lendacky <thomas.lendacky@amd.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoamd-xgbe: Remove use of comm_owned field
Tom Lendacky [Wed, 23 May 2018 16:38:29 +0000 (11:38 -0500)]
amd-xgbe: Remove use of comm_owned field

The comm_owned field can hide logic where double locking is attempted
and prevent multiple threads for the same device from accessing the
mutex properly.  Remove the comm_owned field and use the mutex API
exclusively for gaining ownership.  The current driver has been audited
and is obtaining communications ownership properly.

Signed-off-by: Tom Lendacky <thomas.lendacky@amd.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoamd-xgbe: Read and save the port property registers during probe
Tom Lendacky [Wed, 23 May 2018 16:38:20 +0000 (11:38 -0500)]
amd-xgbe: Read and save the port property registers during probe

Read and save the port property registers once during the device probe
and then use the saved values as they are needed.

Signed-off-by: Tom Lendacky <thomas.lendacky@amd.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoamd-xgbe: Fix debug output of max channel counts
Tom Lendacky [Wed, 23 May 2018 16:38:11 +0000 (11:38 -0500)]
amd-xgbe: Fix debug output of max channel counts

A debug output print statement uses the wrong variable to output the
maximum Rx channel count (cut and paste error, basically).  Fix the
statement to use the proper variable.

Signed-off-by: Tom Lendacky <thomas.lendacky@amd.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoMerge branch 'smc-next'
David S. Miller [Wed, 23 May 2018 20:02:36 +0000 (16:02 -0400)]
Merge branch 'smc-next'

Ursula Braun says:

====================
patches 2018-05-23

here are more smc-patches for net-next:

Patch 1 fixes an ioctl problem detected by syzbot.

Patch 2 improves smc_lgr_list locking in case of abnormal link
group termination. If you want to receive a version for the net-tree,
please let me know. It would look somewhat different, since the port
terminate code has been moved to smc_core.c on net-next.

Patch 3 enables SMC to deal with urgent data.

Patch 4 is a minor improvement to avoid out-of-sync linkgroups
between 2 peers.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agonet/smc: longer delay when freeing client link groups
Ursula Braun [Wed, 23 May 2018 14:38:12 +0000 (16:38 +0200)]
net/smc: longer delay when freeing client link groups

Client link group creation always follows the server linkgroup creation.
If peer creates a new server link group, client has to create a new
client link group. If peer reuses a server link group for a new
connection, client has to reuse its client link group as well. To
avoid out-of-sync conditions for link groups a longer delay for
for client link group removal is defined to make sure this link group
still exists, once the peer decides to reuse a server link group.

Currently the client link group delay time is just 10 jiffies larger
than the server link group delay time. This patch increases the delay
difference to 10 seconds to have a better protection against
out-of-sync link groups.

Signed-off-by: Ursula Braun <ubraun@linux.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agonet/smc: urgent data support
Stefan Raspl [Wed, 23 May 2018 14:38:11 +0000 (16:38 +0200)]
net/smc: urgent data support

Add support for out of band data send and receive.

Signed-off-by: Stefan Raspl <raspl@linux.ibm.com>
Signed-off-by: Ursula Braun <ubraun@linux.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agonet/smc: lock smc_lgr_list in port_terminate()
Hans Wippel [Wed, 23 May 2018 14:38:10 +0000 (16:38 +0200)]
net/smc: lock smc_lgr_list in port_terminate()

Currently, smc_port_terminate() is not holding the lock of the lgr list
while it is traversing the list. This patch adds locking to this
function and changes smc_lgr_terminate() accordingly.

Signed-off-by: Hans Wippel <hwippel@linux.ibm.com>
Signed-off-by: Ursula Braun <ubraun@linux.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agonet/smc: return 0 for ioctl calls in states INIT and CLOSED
Ursula Braun [Wed, 23 May 2018 14:38:09 +0000 (16:38 +0200)]
net/smc: return 0 for ioctl calls in states INIT and CLOSED

A connected SMC-socket contains addresses of descriptors for the
send buffer and the rmb (receive buffer). Fields of these descriptors
are used to determine the answer for certain ioctl requests.
Add extra handling for unconnected SMC socket states without valid
buffer descriptor addresses.

Signed-off-by: Ursula Braun <ubraun@linux.ibm.com>
Reported-by: syzbot+e6714328fda813fc670f@syzkaller.appspotmail.com
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agocxgb4: do L1 config when module is inserted
Ganesh Goudar [Wed, 23 May 2018 14:33:33 +0000 (20:03 +0530)]
cxgb4: do L1 config when module is inserted

trigger an L1 configure operation when a transceiver module
is inserted in order to cause current "sticky" options like
Requested Forward Error Correction to be reapplied.

Signed-off-by: Casey Leedom <leedom@chelsio.com>
Signed-off-by: Ganesh Goudar <ganeshgr@chelsio.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agocxgb4: change the port capability bits definition
Ganesh Goudar [Wed, 23 May 2018 14:32:58 +0000 (20:02 +0530)]
cxgb4: change the port capability bits definition

MDI Port Capabilities bit definitions were inconsistent with
regard to the MDI enum values. 2 bits used to define MDI in
the port capabilities are not really separable, it's a 2-bit
field with 4 different values. Change the port capability bit
definitions to be "AUTO" and "STRAIGHT" in order to get them
to line up with the enum's.

Signed-off-by: Casey Leedom <leedom@chelsio.com>
Signed-off-by: Ganesh Goudar <ganeshgr@chelsio.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoMerge tag 'mac80211-next-for-davem-2018-05-23' of git://git.kernel.org/pub/scm/linux...
David S. Miller [Wed, 23 May 2018 19:53:00 +0000 (15:53 -0400)]
Merge tag 'mac80211-next-for-davem-2018-05-23' of git://git./linux/kernel/git/jberg/mac80211-next

Johannes Berg says:

For this round, we have various things all over the place, notably
 * a fix for a race in aggregation, which I want to let
   bake for a bit longer before sending to stable
 * some new statistics (ACK RSSI, TXQ)
 * TXQ configuration
 * preparations for HE, particularly radiotap
 * replace confusing "country IE" by "country element" since it's
   not referring to Ireland

Note that I merged net-next to get a fix from mac80211 that got
there via net, to apply one patch that would otherwise conflict.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoMerge branch 'qca8k-QCA8334-switch-support'
David S. Miller [Wed, 23 May 2018 19:46:50 +0000 (15:46 -0400)]
Merge branch 'qca8k-QCA8334-switch-support'

Michal Vokáč says:

====================
Add support for QCA8334 switch

This series basically adds support for a QCA8334 ethernet switch to the
qca8k driver. It is a four-port variant of the already supported seven
port QCA8337. Register map is the same for the whole familly and all chips
have the same device ID.

Major part of this series enhances the CPU port setting. Currently the CPU
port is not set to any sensible defaults compatible with the xGMII
interface. This series forces the CPU port to its maximum bandwidth and
also allows to adjust the new defaults using fixed-link device tree
sub-node.

Alongside these changes I fixed two checkpatch warnings regarding SPDX and
redundant parentheses.

Changes in v3:
 - Rebased on latest net-next/master.
 - Corrected fixed-link documentation.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agonet: dsa: qca8k: Remove redundant parentheses
Michal Vokáč [Wed, 23 May 2018 06:20:24 +0000 (08:20 +0200)]
net: dsa: qca8k: Remove redundant parentheses

Fix warning reported by checkpatch.

Signed-off-by: Michal Vokáč <michal.vokac@ysoft.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agonet: dsa: qca8k: Replace GPL boilerplate by SPDX
Michal Vokáč [Wed, 23 May 2018 06:20:23 +0000 (08:20 +0200)]
net: dsa: qca8k: Replace GPL boilerplate by SPDX

Replace the GPLv2 license boilerplate with the SPDX license identifier.

Signed-off-by: Michal Vokáč <michal.vokac@ysoft.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agonet: dsa: qca8k: Allow overwriting CPU port setting
Michal Vokáč [Wed, 23 May 2018 06:20:22 +0000 (08:20 +0200)]
net: dsa: qca8k: Allow overwriting CPU port setting

Implement adjust_link function that allows to overwrite default CPU port
setting using fixed-link device tree subnode.

Signed-off-by: Michal Vokáč <michal.vokac@ysoft.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agonet: dsa: qca8k: Force CPU port to its highest bandwidth
Michal Vokáč [Wed, 23 May 2018 06:20:21 +0000 (08:20 +0200)]
net: dsa: qca8k: Force CPU port to its highest bandwidth

By default autonegotiation is enabled to configure MAC on all ports.
For the CPU port autonegotiation can not be used so we need to set
some sensible defaults manually.

This patch forces the default setting of the CPU port to 1000Mbps/full
duplex which is the chip maximum capability.

Also correct size of the bit field used to configure link speed.

Fixes: 6b93fb46480a ("net-next: dsa: add new driver for qca8xxx family")
Signed-off-by: Michal Vokáč <michal.vokac@ysoft.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agonet: dsa: qca8k: Enable RXMAC when bringing up a port
Michal Vokáč [Wed, 23 May 2018 06:20:20 +0000 (08:20 +0200)]
net: dsa: qca8k: Enable RXMAC when bringing up a port

When a port is brought up/down do not enable/disable only the TXMAC
but the RXMAC as well. This is essential for the CPU port to work.

Fixes: 6b93fb46480a ("net-next: dsa: add new driver for qca8xxx family")
Signed-off-by: Michal Vokáč <michal.vokac@ysoft.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agonet: dsa: qca8k: Add support for QCA8334 switch
Michal Vokáč [Wed, 23 May 2018 06:20:19 +0000 (08:20 +0200)]
net: dsa: qca8k: Add support for QCA8334 switch

Add support for the four-port variant of the Qualcomm QCA833x switch.

Signed-off-by: Michal Vokáč <michal.vokac@ysoft.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agonet: dsa: qca8k: Add QCA8334 binding documentation
Michal Vokáč [Wed, 23 May 2018 06:20:18 +0000 (08:20 +0200)]
net: dsa: qca8k: Add QCA8334 binding documentation

Add support for the four-port variant of the Qualcomm QCA833x switch.

The CPU port default link settings can be reconfigured using
a fixed-link sub-node.

Signed-off-by: Michal Vokáč <michal.vokac@ysoft.com>
Reviewed-by: Rob Herring <robh@kernel.org>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agocxgb4: Add new T6 device ids
Ganesh Goudar [Wed, 23 May 2018 06:06:59 +0000 (11:36 +0530)]
cxgb4: Add new T6 device ids

Add 0x6088 and 0x6089 device ids for new T6 cards.

Signed-off-by: Ganesh Goudar <ganeshgr@chelsio.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoselftests: uevent filtering
Christian Brauner [Tue, 22 May 2018 19:34:21 +0000 (21:34 +0200)]
selftests: uevent filtering

Recent discussions around uevent filtering (cf. net-next commit [1], [2],
and [3] and discussions in [4], [5], and [6]) have shown that the semantics
around uevent filtering where not well understood.
Now that we have settled - at least for the moment - how uevent filtering
should look like let's add some selftests to ensure we don't regress
anything in the future.
Note, the semantics of uevent filtering are described in detail in my
commit message to [2] so I won't repeat them here.

[1]: https://git.kernel.org/pub/scm/linux/kernel/git/next/linux-next.git/commit/?id=90d52d4fd82007005125d9a8d2d560a1ca059b9d
[2]: https://git.kernel.org/pub/scm/linux/kernel/git/next/linux-next.git/commit/?id=a3498436b3a0f8ec289e6847e1de40b4123e1639
[3]: https://git.kernel.org/pub/scm/linux/kernel/git/next/linux-next.git/commit/?id=26045a7b14bc7a5455e411d820110f66557d6589
[4]: https://lkml.org/lkml/2018/4/4/739
[5]: https://lkml.org/lkml/2018/4/26/767
[6]: https://lkml.org/lkml/2018/4/26/738

Signed-off-by: Christian Brauner <christian@brauner.io>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoMerge branch 'fib-rule-selftest'
David S. Miller [Wed, 23 May 2018 19:14:30 +0000 (15:14 -0400)]
Merge branch 'fib-rule-selftest'

Roopa Prabhu says:

====================
fib rule selftest

This series adds a new test to test fib rules.
ip route get is used to test fib rule matches.
This series also extends ip route get to match on
sport and dport to test recent support of sport
and dport fib rule match.

v2 - address ido's commemt to make sport dport
ip route get to work correctly for input route
get. I don't support ip route get on ip-proto match yet.
ip route get creates a udp packet and i have left
it at that. We could extend ip route get to support
a few ip proto matches in followup patches.

v3 - Support ip_proto (only tcp and udp) match in getroute.
dropped printing of new match attrs in ip route get,
because ipv6 does not print it. And ipv6 currrently shares
the dump api with ipv6 notify and its better to not add them
to the notify api. dropped it to keep the api consistent between
ipv4 and ipv6 (though uid is already printed in the ipv4 case).
If we need it, both ipv4 and ipv6 can be enhanced to provide
a separate get api. Moved skb creation for ipv4 to a separate func.

v4 - drop separate skb for netlink and fix concerns around rcu and netlink
     reply (as pointed out by DaveM). I now try to reset the skb after the route
     lookup and before the netlink send (testing shows this is ok. More eyes and
     any feedback here will be helpful)

v5 - dropped RTA_TABLE ipv4_rtm_policy update from this series and posted
     it separately for net (feedback from Eric)
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoselftests: net: initial fib rule tests
Roopa Prabhu [Tue, 22 May 2018 21:03:29 +0000 (14:03 -0700)]
selftests: net: initial fib rule tests

This adds a first set of tests for fib rule match/action for
ipv4 and ipv6. Initial tests only cover action lookup table.
can be extended to cover other actions in the future.
Uses ip route get to validate the rule lookup.

Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoipv6: support sport, dport and ip_proto in RTM_GETROUTE
Roopa Prabhu [Tue, 22 May 2018 21:03:28 +0000 (14:03 -0700)]
ipv6: support sport, dport and ip_proto in RTM_GETROUTE

This is a followup to fib6 rules sport, dport and ipproto
match support. Only supports tcp, udp and icmp for ipproto.
Used by fib rule self tests.

Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoipv4: support sport, dport and ip_proto in RTM_GETROUTE
Roopa Prabhu [Tue, 22 May 2018 21:03:27 +0000 (14:03 -0700)]
ipv4: support sport, dport and ip_proto in RTM_GETROUTE

This is a followup to fib rules sport, dport and ipproto
match support. Only supports tcp, udp and icmp for ipproto.
Used by fib rule self tests.

Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agohv_netvsc: Add handlers for ethtool get/set msg level
Haiyang Zhang [Tue, 22 May 2018 18:29:34 +0000 (11:29 -0700)]
hv_netvsc: Add handlers for ethtool get/set msg level

The handlers for ethtool get/set msg level are missing from netvsc.
This patch adds them.

Signed-off-by: Haiyang Zhang <haiyangz@microsoft.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agonet: vxge: fix spelling mistake in macro VXGE_HW_ERR_PRIVILAGED_OPEARATION
Colin Ian King [Tue, 22 May 2018 16:18:09 +0000 (17:18 +0100)]
net: vxge: fix spelling mistake in macro VXGE_HW_ERR_PRIVILAGED_OPEARATION

Rename VXGE_HW_ERR_PRIVILAGED_OPEARATION to VXGE_HW_ERR_PRIVILEGED_OPERATION
to fix spelling mistake.

Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoMerge branch 'udp-gso-fixes'
David S. Miller [Wed, 23 May 2018 18:48:44 +0000 (14:48 -0400)]
Merge branch 'udp-gso-fixes'

Willem de Bruijn says:

====================
udp gso fixes

A few small fixes:
- disallow segmentation with XFRM
- do not leak gso packets into the ingress path

Changes
  v1 -> v2
  - fix build failure in team.c
  - drop scatter-gather fix:
      this is now fixed by commit 113f99c33585 ("net: test tailroom
      before appending to linear skb"). After this patch gso skbs are
      built non-linear regardless of NETIF_F_SG and skb_segment builds
      linear segs.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agogso: limit udp gso to egress-only virtual devices
Willem de Bruijn [Tue, 22 May 2018 15:34:40 +0000 (11:34 -0400)]
gso: limit udp gso to egress-only virtual devices

Until the udp receive stack supports large packets (UDP GRO), GSO
packets must not loop from the egress to the ingress path.

Revert the change that added NETIF_F_GSO_UDP_L4 to various virtual
devices through NETIF_F_GSO_ENCAP_ALL as this included devices that
may loop packets, such as veth and macvlan.

Instead add it to specific devices that forward to another device's
egress path, bonding and team.

Fixes: 83aa025f535f ("udp: add gso support to virtual devices")
CC: Alexander Duyck <alexander.duyck@gmail.com>
Signed-off-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoudp: exclude gso from xfrm paths
Willem de Bruijn [Tue, 22 May 2018 15:34:39 +0000 (11:34 -0400)]
udp: exclude gso from xfrm paths

UDP GSO delays final datagram construction to the GSO layer. This
conflicts with protocol transformations.

Fixes: bec1f6f69736 ("udp: generate gso with UDP_SEGMENT")
CC: Michal Kubecek <mkubecek@suse.cz>
Signed-off-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoMerge branch 'net-sfp-small-improvements'
David S. Miller [Wed, 23 May 2018 18:34:28 +0000 (14:34 -0400)]
Merge branch 'net-sfp-small-improvements'

Antoine Tenart says:

====================
net: sfp: small improvements

A small series of patches improving the SFP support by adding a warning
when no Tx disable pin is available, and making the i2c-bus property
mandatory.

Thanks!
Antoine

Since v1:
  - Removed the patch fixing the sfp driver when no i2c bus was described.
  - Made two new patches to make the i2c-bus property mandatory for sfp modules.

Since the phylink series:
  - s/-EOPNOTSUPP/-ENODEV/ in patch 1/2.
  - I added the acked-by tag in patch 2/2.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoDocumentation/bindings: net: the sfp i2c-bus property is now mandatory
Antoine Tenart [Tue, 22 May 2018 10:18:01 +0000 (12:18 +0200)]
Documentation/bindings: net: the sfp i2c-bus property is now mandatory

The i2c-bus property for sfp modules was made mandatory. Update the
documentation to keep it in sync with the driver's behaviour.

Signed-off-by: Antoine Tenart <antoine.tenart@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agonet: phy: sfp: make the i2c-bus dt property mandatory
Antoine Tenart [Tue, 22 May 2018 10:18:00 +0000 (12:18 +0200)]
net: phy: sfp: make the i2c-bus dt property mandatory

This patch makes the i2c-bus property mandatory when using a device
tree. If the sfp i2c bus isn't described it's impossible to guess the
protocol to use for a given module, and the sfp module would then not
work in most cases.

Signed-off-by: Antoine Tenart <antoine.tenart@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agonet: phy: sfp: warn the user when no tx_disable pin is available
Antoine Tenart [Tue, 22 May 2018 10:17:59 +0000 (12:17 +0200)]
net: phy: sfp: warn the user when no tx_disable pin is available

In case no Tx disable pin is available the SFP modules will always be
emitting. This could be an issue when using modules using laser as their
light source as we would have no way to disable it when the fiber is
removed. This patch adds a warning when registering an SFP cage which do
not have its tx_disable pin wired or available.

Signed-off-by: Antoine Tenart <antoine.tenart@bootlin.com>
Acked-by: Russell King <rmk+kernel@armlinux.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoMerge branch 'nfp-abm-add-basic-support-for-advanced-buffering-NIC'
David S. Miller [Wed, 23 May 2018 18:26:20 +0000 (14:26 -0400)]
Merge branch 'nfp-abm-add-basic-support-for-advanced-buffering-NIC'

Jakub Kicinski says:

====================
nfp: abm: add basic support for advanced buffering NIC

This series lays groundwork for advanced buffer management NIC feature.
It makes necessary NFP core changes, spawns representors and adds devlink
glue.  Following series will add the actual buffering configuration (patch
series size limit).

First three patches add support for configuring NFP buffer pools via a
mailbox.  The existing devlink APIs are used for the purpose.

Third patch allows us to perform small reads from the NFP memory.

The rest of the patch set adds eswitch mode change support and makes
the driver spawn appropriate representors.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agonfp: assign vNIC id as phys_port_name of vNICs which are not ports
Jakub Kicinski [Tue, 22 May 2018 05:12:55 +0000 (22:12 -0700)]
nfp: assign vNIC id as phys_port_name of vNICs which are not ports

When NFP is modelled as a switch we assign phys_port_name to respective
port(representor )s:

 vNIC0 - | - PF port (pf%d)     MAC/PHY (p%d[s%d]) - |E==

In most cases there is only one vNIC for communication with the switch.
If there is more than one we need to be able to identify them.  Use %d
as phys_port_name of the vNICs.

We don't have to pass ID to nfp_net_debugfs_vnic_add() separately any
more.

Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Dirk van der Merwe <dirk.vandermerwe@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agonfp: use split in naming of PCIe PF ports
Jakub Kicinski [Tue, 22 May 2018 05:12:54 +0000 (22:12 -0700)]
nfp: use split in naming of PCIe PF ports

PCI PFs can host more than one logical endpoint.  In NFP terms
this means having more than one vNIC for PCIe PF.  The vNICs
are usually corresponding 1:1 to Ethernet ports.  In core NIC
we use the legacy idea of vNIC *being* the Ethernet port,
hence netdevs put pX(sY) in their phys_port_name, like Ethernet
ports would.  When ASIC ports are fully represented we need to
be able to name different PCIe PF ports, too.  Use a scheme
similar to Ethernet ports - pfXsY, for PCIe PF number X,
sub-port Y.

Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Dirk van der Merwe <dirk.vandermerwe@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agonfp: abm: force Ethternet port up
Jakub Kicinski [Tue, 22 May 2018 05:12:53 +0000 (22:12 -0700)]
nfp: abm: force Ethternet port up

Current control firmware does not cater too well to multi-host
applications.  There is no way to check which hosts are up or
otherwise negotiate what the state of the external port (the
Ethernet port) should be.  Make sure the link is up when driver
loads, and don't take it down when Ethernet port netdev is
closed.

Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Dirk van der Merwe <dirk.vandermerwe@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agonfp: abm: spawn port netdevs
Jakub Kicinski [Tue, 22 May 2018 05:12:52 +0000 (22:12 -0700)]
nfp: abm: spawn port netdevs

To configure buffering points we need full set of netdevs:

                              ASIC

 user netdev  -- | -- PCIe port   MAC port -- | --

Configuring egrees qdiscs on user netdev configures standard
Linux TC software qdiscs, configuring PCIe port qdiscs will
provide a way of setting ASIC queuing parameters for PCIe block.
MAC port netdev egress qdiscs correspond to ASIC MAC Traffic
Manager block.

Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Dirk van der Merwe <dirk.vandermerwe@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agonfp: add devlink_eswitch_mode_set callback
Jakub Kicinski [Tue, 22 May 2018 05:12:51 +0000 (22:12 -0700)]
nfp: add devlink_eswitch_mode_set callback

Our previous apps all assumed to use only one eswitch mode (legacy
or switchdev) without the ability to change it.  ABM NIC will
want to support the switch so plumb devlink_eswitch_mode_set through.
The devlink_eswitch_mode_set is expected to spawn representors and
potentially devlink ports so it's called under big devlink lock and
pf->lock.

Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Dirk van der Merwe <dirk.vandermerwe@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agodevlink: don't take instance lock around eswitch mode set
Jakub Kicinski [Tue, 22 May 2018 05:12:50 +0000 (22:12 -0700)]
devlink: don't take instance lock around eswitch mode set

Changing switch mode may want to register and unregister devlink
ports.  Therefore similarly to DEVLINK_CMD_PORT_SPLIT/UNSPLIT it
should not take the instance lock.  Drivers don't depend on existing
locking since it's a very recent addition.

Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agonfp: add app pointer to port representors
Jakub Kicinski [Tue, 22 May 2018 05:12:49 +0000 (22:12 -0700)]
nfp: add app pointer to port representors

nfp_apps can currently associate their structures with vNICs but
not representors.  Add app priv pointer to representors as well.

Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Dirk van der Merwe <dirk.vandermerwe@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agonfp: abm: create project-specific vNIC structure
Jakub Kicinski [Tue, 22 May 2018 05:12:48 +0000 (22:12 -0700)]
nfp: abm: create project-specific vNIC structure

ABM NIC requires more complex vNIC handling, allocate
per-vNIC structure.  Find out RX queue base and PCI PF id.
There will be multiple PFs sharing the same MAC port, therefore
the MAC address assigned to the vNIC must be looked up in the
HWInfo database.

Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Dirk van der Merwe <dirk.vandermerwe@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agonfp: abm: add initial active buffer management NIC skeleton
Jakub Kicinski [Tue, 22 May 2018 05:12:47 +0000 (22:12 -0700)]
nfp: abm: add initial active buffer management NIC skeleton

Add a very rudimentary active buffer management NIC support.
For now it's like a core NIC without SR-IOV support.  Next
commits will extend its functionality.

Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Dirk van der Merwe <dirk.vandermerwe@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agonfp: core: allow 4-byte aligned accesses to Memory Units
Jakub Kicinski [Tue, 22 May 2018 05:12:46 +0000 (22:12 -0700)]
nfp: core: allow 4-byte aligned accesses to Memory Units

Current code doesn't enforce length requirements on 32bit accesses
with action NFP_CPP_ACTION_RW to memory units, but if the access
is only aligned to 4 bytes as well we will fall into the explicit
access case and error out.  Such accesses are correct, allow them
by lowering the width earlier.

While at it use a switch statement to improve readability.

Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Dirk van der Merwe <dirk.vandermerwe@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agonfp: add shared buffer configuration
Jakub Kicinski [Tue, 22 May 2018 05:12:45 +0000 (22:12 -0700)]
nfp: add shared buffer configuration

Allow app FW to advertise its shared buffer pool information.
Use the per-PF mailbox to configure them from devlink.

Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agonfp: add support for per-PCI PF mailbox
Jakub Kicinski [Tue, 22 May 2018 05:12:44 +0000 (22:12 -0700)]
nfp: add support for per-PCI PF mailbox

When working with devlink-related functionality for locking reasons
it's easier to create a new mailbox per-PCI PF device than try to
use one of the netdev/vNIC mailboxes.

Define new mailbox structure and resolve its symbol during probe.
For forward compatibility allow silent truncation of mailbox command
data.

Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agonfp: move rtsym helpers to pf code
Jakub Kicinski [Tue, 22 May 2018 05:12:43 +0000 (22:12 -0700)]
nfp: move rtsym helpers to pf code

nfp_net_pf_rtsym_read_optional() and nfp_net_pf_map_rtsym() are not
really related to networking code.  Move them to the PF code and
remove the net from their names.  They will soon be needed by code
outside of nfp_net_main.c anyway.

Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoMerge branch 'bpfilter'
David S. Miller [Wed, 23 May 2018 17:23:40 +0000 (13:23 -0400)]
Merge branch 'bpfilter'

Alexei Starovoitov says:

====================
bpfilter

v2->v3:
- followed Luis's suggestion and significantly simplied first patch
  with shmem_kernel_file_setup+kernel_write. Added kdoc for new helper
- fixed typos and race to access pipes with mutex
- tested with bpfilter being 'builtin'. CONFIG_BPFILTER_UMH=y|m both work.
  Interesting to see a usermode executable being embedded inside vmlinux.
- it doesn't hurt to enable bpfilter in .config.
  ip_setsockopt commands sent to usermode via pipes and -ENOPROTOOPT is
  returned from userspace, so kernel falls back to original iptables code

v1->v2:
this patch set is almost a full rewrite of the earlier umh modules approach
The v1 of patches and follow up discussion was covered by LWN:
https://lwn.net/Articles/749108/

I believe the v2 addresses all issues brought up by Andy and others.
Mainly there are zero changes to kernel/module.c
Instead of teaching module loading logic to recognize special
umh module, let normal kernel modules execute part of its own
.init.rodata as a new user space process (Andy's idea)
Patch 1 introduces this new helper:
int fork_usermode_blob(void *data, size_t len, struct umh_info *info);
Input:
  data + len == executable file
Output:
  struct umh_info {
       struct file *pipe_to_umh;
       struct file *pipe_from_umh;
       pid_t pid;
  };

Advantages vs v1:
- the embedded user mode executable is stored as .init.rodata inside
  normal kernel module. These pages are freed when .ko finishes loading
- the elf file is copied into tmpfs file. The user mode process is swappable.
- the communication between user mode process and 'parent' kernel module
  is done via two unix pipes, hence protocol is not exposed to
  user space
- impossible to launch umh on its own (that was the main issue of v1)
  and impossible to be man-in-the-middle due to pipes
- bpfilter.ko consists of tiny kernel part that passes the data
  between kernel and umh via pipes and much bigger umh part that
  doing all the work
- 'lsmod' shows bpfilter.ko as usual.
  'rmmod bpfilter' removes kernel module and kills corresponding umh
- signed bpfilter.ko covers the whole image including umh code

Few issues:
- the user can still attach to the process and debug it with
  'gdb /proc/pid/exe pid', but 'gdb -p pid' doesn't work.
  (a bit worse comparing to v1)
- tinyconfig will notice a small increase in .text
  +766 | TEXT | 7c8b94806bec umh: introduce fork_usermode_blob() helper
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agonet: add skeleton of bpfilter kernel module
Alexei Starovoitov [Tue, 22 May 2018 02:22:30 +0000 (19:22 -0700)]
net: add skeleton of bpfilter kernel module

bpfilter.ko consists of bpfilter_kern.c (normal kernel module code)
and user mode helper code that is embedded into bpfilter.ko

The steps to build bpfilter.ko are the following:
- main.c is compiled by HOSTCC into the bpfilter_umh elf executable file
- with quite a bit of objcopy and Makefile magic the bpfilter_umh elf file
  is converted into bpfilter_umh.o object file
  with _binary_net_bpfilter_bpfilter_umh_start and _end symbols
  Example:
  $ nm ./bld_x64/net/bpfilter/bpfilter_umh.o
  0000000000004cf8 T _binary_net_bpfilter_bpfilter_umh_end
  0000000000004cf8 A _binary_net_bpfilter_bpfilter_umh_size
  0000000000000000 T _binary_net_bpfilter_bpfilter_umh_start
- bpfilter_umh.o and bpfilter_kern.o are linked together into bpfilter.ko

bpfilter_kern.c is a normal kernel module code that calls
the fork_usermode_blob() helper to execute part of its own data
as a user mode process.

Notice that _binary_net_bpfilter_bpfilter_umh_start - end
is placed into .init.rodata section, so it's freed as soon as __init
function of bpfilter.ko is finished.
As part of __init the bpfilter.ko does first request/reply action
via two unix pipe provided by fork_usermode_blob() helper to
make sure that umh is healthy. If not it will kill it via pid.

Later bpfilter_process_sockopt() will be called from bpfilter hooks
in get/setsockopt() to pass iptable commands into umh via bpfilter.ko

If admin does 'rmmod bpfilter' the __exit code bpfilter.ko will
kill umh as well.

Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoumh: introduce fork_usermode_blob() helper
Alexei Starovoitov [Tue, 22 May 2018 02:22:29 +0000 (19:22 -0700)]
umh: introduce fork_usermode_blob() helper

Introduce helper:
int fork_usermode_blob(void *data, size_t len, struct umh_info *info);
struct umh_info {
       struct file *pipe_to_umh;
       struct file *pipe_from_umh;
       pid_t pid;
};

that GPLed kernel modules (signed or unsigned) can use it to execute part
of its own data as swappable user mode process.

The kernel will do:
- allocate a unique file in tmpfs
- populate that file with [data, data + len] bytes
- user-mode-helper code will do_execve that file and, before the process
  starts, the kernel will create two unix pipes for bidirectional
  communication between kernel module and umh
- close tmpfs file, effectively deleting it
- the fork_usermode_blob will return zero on success and populate
  'struct umh_info' with two unix pipes and the pid of the user process

As the first step in the development of the bpfilter project
the fork_usermode_blob() helper is introduced to allow user mode code
to be invoked from a kernel module. The idea is that user mode code plus
normal kernel module code are built as part of the kernel build
and installed as traditional kernel module into distro specified location,
such that from a distribution point of view, there is
no difference between regular kernel modules and kernel modules + umh code.
Such modules can be signed, modprobed, rmmod, etc. The use of this new helper
by a kernel module doesn't make it any special from kernel and user space
tooling point of view.

Such approach enables kernel to delegate functionality traditionally done
by the kernel modules into the user space processes (either root or !root) and
reduces security attack surface of the new code. The buggy umh code would crash
the user process, but not the kernel. Another advantage is that umh code
of the kernel module can be debugged and tested out of user space
(e.g. opening the possibility to run clang sanitizers, fuzzers or
user space test suites on the umh code).
In case of the bpfilter project such architecture allows complex control plane
to be done in the user space while bpf based data plane stays in the kernel.

Since umh can crash, can be oom-ed by the kernel, killed by the admin,
the kernel module that uses them (like bpfilter) needs to manage life
time of umh on its own via two unix pipes and the pid of umh.

The exit code of such kernel module should kill the umh it started,
so that rmmod of the kernel module will cleanup the corresponding umh.
Just like if the kernel module does kmalloc() it should kfree() it
in the exit code.

Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agonetfilter: nf_tables: remove nft_af_info.
Taehee Yoo [Tue, 15 May 2018 12:23:31 +0000 (21:23 +0900)]
netfilter: nf_tables: remove nft_af_info.

The struct nft_af_info was removed.

Signed-off-by: Taehee Yoo <ap420073@gmail.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
6 years agonl80211: Reject disconnect commands except from conn_owner
Andrew Zaborowski [Tue, 22 May 2018 00:46:02 +0000 (02:46 +0200)]
nl80211: Reject disconnect commands except from conn_owner

Reject NL80211_CMD_DISCONNECT, NL80211_CMD_DISASSOCIATE,
NL80211_CMD_DEAUTHENTICATE and NL80211_CMD_ASSOCIATE commands
from clients other than the connection owner set in the connect,
authenticate or associate commands, if it was set.

The main point of this check is to prevent chaos when two processes
try to use nl80211 at the same time, it's not a security measure.
The same thing should possibly be done for JOIN_IBSS/LEAVE_IBSS and
START_AP/STOP_AP.

Signed-off-by: Andrew Zaborowski <andrew.zaborowski@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
6 years agorfkill: Create rfkill-none LED trigger
João Paulo Rechi Vita [Tue, 22 May 2018 21:29:32 +0000 (14:29 -0700)]
rfkill: Create rfkill-none LED trigger

Creates a new trigger rfkill-none, as a complement to rfkill-any, which
drives LEDs when any radio is enabled. The new trigger is meant to turn
a LED ON whenever all radios are OFF, and turn it OFF otherwise.

Signed-off-by: João Paulo Rechi Vita <jprvita@endlessm.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
6 years agorfkill: Rename rfkill_any_led_trigger* functions
João Paulo Rechi Vita [Tue, 22 May 2018 21:29:31 +0000 (14:29 -0700)]
rfkill: Rename rfkill_any_led_trigger* functions

Rename these functions to rfkill_global_led_trigger*, as they are going
to be extended to handle another global rfkill led trigger.

This commit does not change any functionality.

Signed-off-by: João Paulo Rechi Vita <jprvita@endlessm.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
6 years agonl80211: Update ERP info using NL80211_CMD_UPDATE_CONNECT_PARAMS
Vidyullatha Kanchanapally [Tue, 22 May 2018 08:19:08 +0000 (10:19 +0200)]
nl80211: Update ERP info using NL80211_CMD_UPDATE_CONNECT_PARAMS

Use NL80211_CMD_UPDATE_CONNECT_PARAMS to update new ERP information,
Association IEs and the Authentication type to driver / firmware which
will be used in subsequent roamings.

Signed-off-by: Vidyullatha Kanchanapally <vidyullatha@codeaurora.org>
[arend: extended fils-sk kernel doc and added check in wiphy_register()]
Reviewed-by: Jithu Jance <jithu.jance@broadcom.com>
Reviewed-by: Eylon Pedinovsky <eylon.pedinovsky@broadcom.com>
Signed-off-by: Arend van Spriel <arend.vanspriel@broadcom.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
6 years agonl80211: add FILS related parameters to ROAM event
Arend Van Spriel [Tue, 22 May 2018 08:19:07 +0000 (10:19 +0200)]
nl80211: add FILS related parameters to ROAM event

In case of FILS shared key offload the parameters can change
upon roaming of which user-space needs to be notified.

Reviewed-by: Jithu Jance <jithu.jance@broadcom.com>
Reviewed-by: Eylon Pedinovsky <eylon.pedinovsky@broadcom.com>
Signed-off-by: Arend van Spriel <arend.vanspriel@broadcom.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
6 years agocfg80211: use separate struct for FILS parameters
Arend Van Spriel [Tue, 22 May 2018 08:19:06 +0000 (10:19 +0200)]
cfg80211: use separate struct for FILS parameters

Put FILS related parameters into their own struct definition so
it can be reused for roam events in subsequent change.

Reviewed-by: Jithu Jance <jithu.jance@broadcom.com>
Reviewed-by: Eylon Pedinovsky <eylon.pedinovsky@broadcom.com>
Signed-off-by: Arend van Spriel <arend.vanspriel@broadcom.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
6 years agonl80211: Optimize cfg80211_bss_expire invocations
Denis Kenzior [Mon, 21 May 2018 15:31:13 +0000 (10:31 -0500)]
nl80211: Optimize cfg80211_bss_expire invocations

Only invoke cfg80211_bss_expire on the first nl80211_dump_scan
invocation to avoid (likely) redundant processing.

Signed-off-by: Denis Kenzior <denkenz@gmail.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
6 years agomac80211: Support adding duration for prepare_tx() callback
Ilan Peer [Fri, 20 Apr 2018 10:49:25 +0000 (13:49 +0300)]
mac80211: Support adding duration for prepare_tx() callback

There are specific cases, such as SAE authentication exchange, that
might require long duration to complete. For such cases, add support
for indicating to the driver the required duration of the prepare_tx()
operation, so the driver would still be able to complete the frame
exchange.

Currently, indicate the duration only for SAE authentication exchange,
as SAE authentication can take up to 2000 msec (as defined in IEEE
P802.11-REVmd D1.0 p. 3504).

As the patch modified the prepare_tx() callback API, also modify
the relevant code in iwlwifi.

Signed-off-by: Ilan Peer <ilan.peer@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
6 years agoMerge remote-tracking branch 'net-next/master' into mac80211-next
Johannes Berg [Wed, 23 May 2018 09:05:26 +0000 (11:05 +0200)]
Merge remote-tracking branch 'net-next/master' into mac80211-next

Bring in net-next which had pulled in net, so I have the changes
from mac80211 and can apply a patch that would otherwise conflict.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
6 years agonetfilter: nfnetlink_queue: resolve clash for unconfirmed conntracks
Pablo Neira Ayuso [Wed, 23 May 2018 07:17:24 +0000 (09:17 +0200)]
netfilter: nfnetlink_queue: resolve clash for unconfirmed conntracks

In nfqueue, two consecutive skbuffs may race to create the conntrack
entry. Hence, the one that loses the race gets dropped due to clash in
the insertion into the hashes from the nf_conntrack_confirm() path.

This patch adds a new nf_conntrack_update() function which searches for
possible clashes and resolve them. NAT mangling for the packet losing
race is corrected by using the conntrack information that won race.

In order to avoid direct module dependencies with conntrack and NAT, the
nf_ct_hook and nf_nat_hook structures are used for this purpose.

Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
6 years agonetfilter: add struct nf_nat_hook and use it
Pablo Neira Ayuso [Wed, 23 May 2018 07:17:19 +0000 (09:17 +0200)]
netfilter: add struct nf_nat_hook and use it

Move decode_session() and parse_nat_setup_hook() indirections to struct
nf_nat_hook structure.

Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
6 years agonetfilter: add struct nf_ct_hook and use it
Pablo Neira Ayuso [Wed, 23 May 2018 07:17:12 +0000 (09:17 +0200)]
netfilter: add struct nf_ct_hook and use it

Move the nf_ct_destroy indirection to the struct nf_ct_hook.

Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
6 years agonetfilter: ip6t_rpfilter: provide input interface for route lookup
Vincent Bernat [Sun, 20 May 2018 11:03:38 +0000 (13:03 +0200)]
netfilter: ip6t_rpfilter: provide input interface for route lookup

In commit 47b7e7f82802, this bit was removed at the same time the
RT6_LOOKUP_F_IFACE flag was removed. However, it is needed when
link-local addresses are used, which is a very common case: when
packets are routed, neighbor solicitations are done using link-local
addresses. For example, the following neighbor solicitation is not
matched by "-m rpfilter":

    IP6 fe80::5254:33ff:fe00:1 > ff02::1:ff00:3: ICMP6, neighbor
    solicitation, who has 2001:db8::5254:33ff:fe00:3, length 32

Commit 47b7e7f82802 doesn't quite explain why we shouldn't use
RT6_LOOKUP_F_IFACE in the rpfilter case. I suppose the interface check
later in the function would make it redundant. However, the remaining
of the routing code is using RT6_LOOKUP_F_IFACE when there is no
source address (which matches rpfilter's case with a non-unicast
destination, like with neighbor solicitation).

Signed-off-by: Vincent Bernat <vincent@bernat.im>
Fixes: 47b7e7f82802 ("netfilter: don't set F_IFACE on ipv6 fib lookups")
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
6 years agonetfilter: nft_set_rbtree: add timeout support
Pablo Neira Ayuso [Wed, 16 May 2018 20:58:34 +0000 (22:58 +0200)]
netfilter: nft_set_rbtree: add timeout support

Add garbage collection logic to expire elements stored in the rb-tree
representation.

Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
6 years agonetfilter: make NF_OSF non-visible symbol
Fernando Fernandez Mancera [Tue, 15 May 2018 23:07:58 +0000 (01:07 +0200)]
netfilter: make NF_OSF non-visible symbol

Signed-off-by: Fernando Fernandez Mancera <ffmancera@riseup.net>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
6 years agonetfilter: lift one-nat-hook-only restriction
Florian Westphal [Mon, 14 May 2018 21:46:59 +0000 (23:46 +0200)]
netfilter: lift one-nat-hook-only restriction

This reverts commit f92b40a8b2645
("netfilter: core: only allow one nat hook per hook point"), this
limitation is no longer needed.  The nat core now invokes these
functions and makes sure that hook evaluation stops after a mapping is
created and a null binding is created otherwise.

Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
6 years agonetfilter: nf_nat: add nat type hooks to nat core
Florian Westphal [Mon, 14 May 2018 21:46:58 +0000 (23:46 +0200)]
netfilter: nf_nat: add nat type hooks to nat core

Currently the packet rewrite and instantiation of nat NULL bindings
happens from the protocol specific nat backend.

Invocation occurs either via ip(6)table_nat or the nf_tables nat chain type.

Invocation looks like this (simplified):
NF_HOOK()
   |
   `---iptable_nat
 |
 `---> nf_nat_l3proto_ipv4 -> nf_nat_packet
               |
          new packet? pass skb though iptables nat chain
                       |
       `---> iptable_nat: ipt_do_table

In nft case, this looks the same (nft_chain_nat_ipv4 instead of
iptable_nat).

This is a problem for two reasons:
1. Can't use iptables nat and nf_tables nat at the same time,
   as the first user adds a nat binding (nf_nat_l3proto_ipv4 adds a
   NULL binding if do_table() did not find a matching nat rule so we
   can detect post-nat tuple collisions).
2. If you use e.g. nft_masq, snat, redir, etc. uses must also register
   an empty base chain so that the nat core gets called fro NF_HOOK()
   to do the reverse translation, which is neither obvious nor user
   friendly.

After this change, the base hook gets registered not from iptable_nat or
nftables nat hooks, but from the l3 nat core.

iptables/nft nat base hooks get registered with the nat core instead:

NF_HOOK()
   |
   `---> nf_nat_l3proto_ipv4 -> nf_nat_packet
|
         new packet? pass skb through iptables/nftables nat chains
                |
+-> iptables_nat: ipt_do_table
        +-> nft nat chain x
        `-> nft nat chain y

The nat core deals with null bindings and reverse translation.
When no mapping exists, it calls the registered nat lookup hooks until
one creates a new mapping.
If both iptables and nftables nat hooks exist, the first matching
one is used (i.e., higher priority wins).

Also, nft users do not need to create empty nat hooks anymore,
nat core always registers the base hooks that take care of reverse/reply
translation.

Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
6 years agonetfilter: nf_nat: add nat hook register functions to nf_nat
Florian Westphal [Mon, 14 May 2018 21:46:57 +0000 (23:46 +0200)]
netfilter: nf_nat: add nat hook register functions to nf_nat

This adds the infrastructure to register nat hooks with the nat core
instead of the netfilter core.

nat hooks are used to configure nat bindings.  Such hooks are registered
from ip(6)table_nat or by the nftables core when a nat chain is added.

After next patch, nat hooks will be registered with nf_nat instead of
netfilter core.  This allows to use many nat lookup functions at the
same time while doing the real packet rewrite (nat transformation) in
one place.

This change doesn't convert the intended users yet to ease review.

Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
6 years agonetfilter: core: export raw versions of add/delete hook functions
Florian Westphal [Mon, 14 May 2018 21:46:56 +0000 (23:46 +0200)]
netfilter: core: export raw versions of add/delete hook functions

This will allow the nat core to reuse the nf_hook infrastructure
to maintain nat lookup functions.

The raw versions don't assume a particular hook location, the
functions get added/deleted from the hook blob that is passed to the
functions.

Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
6 years agonetfilter: nf_tables: allow chain type to override hook register
Florian Westphal [Mon, 14 May 2018 21:46:55 +0000 (23:46 +0200)]
netfilter: nf_tables: allow chain type to override hook register

Will be used in followup patch when nat types no longer
use nf_register_net_hook() but will instead register with the nat core.

Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
6 years agonetfilter: xtables: allow table definitions not backed by hook_ops
Florian Westphal [Mon, 14 May 2018 21:46:54 +0000 (23:46 +0200)]
netfilter: xtables: allow table definitions not backed by hook_ops

The ip(6)tables nat table is currently receiving skbs from the netfilter
core, after a followup patch skbs will be coming from the netfilter nat
core instead, so the table is no longer backed by normal hook_ops.

Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
6 years agonetfilter: nf_nat: move common nat code to nat core
Florian Westphal [Mon, 14 May 2018 21:46:53 +0000 (23:46 +0200)]
netfilter: nf_nat: move common nat code to nat core

Copy-pasted, both l3 helpers almost use same code here.
Split out the common part into an 'inet' helper.

Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
6 years agoMerge branch 'qed-firmware-TLV'
David S. Miller [Wed, 23 May 2018 03:29:55 +0000 (23:29 -0400)]
Merge branch 'qed-firmware-TLV'

Sudarsana Reddy Kalluru says:

====================
qed*: Add support for management firmware TLV request.

Management firmware (MFW) requires config and state information from
the driver. It queries this via TLV (type-length-value) request wherein
mfw specificies the list of required TLVs. Driver fills the TLV data
and responds back to MFW.
This patch series adds qed/qede/qedf/qedi driver implementation for
supporting the TLV queries from MFW.

Changes from previous versions:
-------------------------------
v2: Split patch (2) into multiple simpler patches.
v2: Update qed_tlv_parsed_buf->p_val datatype to void pointer to avoid
    bunch of unnecessary typecasts.

Please consider applying this series to "net-next".
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoqedi: Add get_generic_tlv_data handler.
Manish Rangankar [Tue, 22 May 2018 07:28:46 +0000 (00:28 -0700)]
qedi: Add get_generic_tlv_data handler.

Signed-off-by: Manish Rangankar <manish.rangankar@cavium.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoqedi: Add support for populating ethernet TLVs.
Manish Rangankar [Tue, 22 May 2018 07:28:45 +0000 (00:28 -0700)]
qedi: Add support for populating ethernet TLVs.

This patch adds callbacks for providing the ethernet protocol driver TLVs.

Signed-off-by: Manish Rangankar <manish.rangankar@cavium.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoqedf: Add get_generic_tlv_data handler.
Chad Dupuis [Tue, 22 May 2018 07:28:44 +0000 (00:28 -0700)]
qedf: Add get_generic_tlv_data handler.

Signed-off-by: Chad Dupuis <chad.dupuis@cavium.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoqedf: Add support for populating ethernet TLVs.
Chad Dupuis [Tue, 22 May 2018 07:28:43 +0000 (00:28 -0700)]
qedf: Add support for populating ethernet TLVs.

This patch adds callbacks for providing the ethernet protocol driver TLVs.

Signed-off-by: Chad Dupuis <chad.dupuis@cavium.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoqede: Add support for populating ethernet TLVs.
Sudarsana Reddy Kalluru [Tue, 22 May 2018 07:28:42 +0000 (00:28 -0700)]
qede: Add support for populating ethernet TLVs.

This patch adds callbacks for providing the ethernet protocol driver TLVs.

Signed-off-by: Sudarsana Reddy Kalluru <Sudarsana.Kalluru@cavium.com>
Signed-off-by: Ariel Elior <ariel.elior@cavium.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoqed: Add driver infrastucture for handling mfw requests.
Sudarsana Reddy Kalluru [Tue, 22 May 2018 07:28:41 +0000 (00:28 -0700)]
qed: Add driver infrastucture for handling mfw requests.

MFW requests the TLVs in interrupt context. Extracting of the required
data from upper layers and populating of the TLVs require process context.
The patch adds work-queues for processing the tlv requests. It also adds
the implementation for requesting the tlv values from appropriate protocol
driver.

Signed-off-by: Sudarsana Reddy Kalluru <Sudarsana.Kalluru@cavium.com>
Signed-off-by: Ariel Elior <ariel.elior@cavium.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoqed: Add support for processing iscsi tlv request.
Sudarsana Reddy Kalluru [Tue, 22 May 2018 07:28:40 +0000 (00:28 -0700)]
qed: Add support for processing iscsi tlv request.

Signed-off-by: Sudarsana Reddy Kalluru <Sudarsana.Kalluru@cavium.com>
Signed-off-by: Ariel Elior <ariel.elior@cavium.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoqed: Add support for processing fcoe tlv request.
Sudarsana Reddy Kalluru [Tue, 22 May 2018 07:28:39 +0000 (00:28 -0700)]
qed: Add support for processing fcoe tlv request.

Signed-off-by: Sudarsana Reddy Kalluru <Sudarsana.Kalluru@cavium.com>
Signed-off-by: Ariel Elior <ariel.elior@cavium.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoqed: Add support for tlv request processing.
Sudarsana Reddy Kalluru [Tue, 22 May 2018 07:28:38 +0000 (00:28 -0700)]
qed: Add support for tlv request processing.

The patch adds driver support for processing TLV requests/repsonses
from the mfw and upper driver layers respectively. The implementation
reads the requested TLVs from the shared memory, requests the values
from upper layer drivers, populates this info (TLVs) shared memory and
notifies MFW about the TLV values.

Signed-off-by: Sudarsana Reddy Kalluru <Sudarsana.Kalluru@cavium.com>
Signed-off-by: Ariel Elior <ariel.elior@cavium.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoqed: Add MFW interfaces for TLV request support.
Sudarsana Reddy Kalluru [Tue, 22 May 2018 07:28:37 +0000 (00:28 -0700)]
qed: Add MFW interfaces for TLV request support.

The patch adds required management firmware (MFW) interfaces such as
mailbox commands, TLV types etc.

Signed-off-by: Sudarsana Reddy Kalluru <Sudarsana.Kalluru@cavium.com>
Signed-off-by: Ariel Elior <ariel.elior@cavium.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoMerge branch '40GbE' of git://git.kernel.org/pub/scm/linux/kernel/git/jkirsher/next...
David S. Miller [Tue, 22 May 2018 19:45:06 +0000 (15:45 -0400)]
Merge branch '40GbE' of git://git./linux/kernel/git/jkirsher/next-queue

Jeff Kirsher says:

====================
40GbE Intel Wired LAN Driver Updates 2018-05-22

This series contains updates to i40e only.

Jake provides all the changes in this series starting with making it
consistent in how we approach the bit lock.  Fixed the reporting of the
VEB statistics and the queue statistics to always return every queue
even if it is not currently in use.  Use WARN_ONCE() so that the first
time we end up with an incorrect size we will dump a stack trace and a
message to help highlight the issue early in testing.  Folded the fixed
string prefix into the stat string definition.  Instead of using a
separate char *p pointer when copying strings, use the data pointer
directly.  Added code comments for several of the statistic functions to
better explain the number and ordering of statistics.
====================

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