openwrt/staging/blogic.git
7 years agomlxsw: reg: Add MCIA register for cable info access
Arkadi Sharshevsky [Wed, 14 Jun 2017 07:27:39 +0000 (09:27 +0200)]
mlxsw: reg: Add MCIA register for cable info access

The MCIA register is used to access the SFP+ and QSFP connector's
EPROM. It will be used to query the cable info.

Signed-off-by: Arkadi Sharshevsky <arkadis@mellanox.com>
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agoMerge branch 'bpf-MIPS-infra'
David S. Miller [Wed, 14 Jun 2017 19:03:23 +0000 (15:03 -0400)]
Merge branch 'bpf-MIPS-infra'

David Daney says:

====================
bpf: Changes needed (or desired) for MIPS support

This is a grab bag of changes to the bpf testing infrastructure I
developed working on MIPS eBPF JIT support.  The change to
bpf_jit_disasm is probably universally beneficial, the others are more
MIPS specific.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agosamples/bpf: Fix tracex5 to work with MIPS syscalls.
David Daney [Tue, 13 Jun 2017 23:49:38 +0000 (16:49 -0700)]
samples/bpf: Fix tracex5 to work with MIPS syscalls.

There are two problems:

1) In MIPS the __NR_* macros expand to an expression, this causes the
   sections of the object file to be named like:

  .
  .
  .
  [ 5] kprobe/(5000 + 1) PROGBITS        0000000000000000 000160 ...
  [ 6] kprobe/(5000 + 0) PROGBITS        0000000000000000 000258 ...
  [ 7] kprobe/(5000 + 9) PROGBITS        0000000000000000 000348 ...
  .
  .
  .

The fix here is to use the "asm_offsets" trick to evaluate the macros
in the C compiler and generate a header file with a usable form of the
macros.

2) MIPS syscall numbers start at 5000, so we need a bigger map to hold
the sub-programs.

Signed-off-by: David Daney <david.daney@cavium.com>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agobpf: Add MIPS support to samples/bpf.
David Daney [Tue, 13 Jun 2017 23:49:37 +0000 (16:49 -0700)]
bpf: Add MIPS support to samples/bpf.

Signed-off-by: David Daney <david.daney@cavium.com>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agotest_bpf: Add test to make conditional jump cross a large number of insns.
David Daney [Tue, 13 Jun 2017 23:49:36 +0000 (16:49 -0700)]
test_bpf: Add test to make conditional jump cross a large number of insns.

On MIPS, conditional branches can only span 32k instructions.  To
exceed this limit in the JIT with the BPF maximum of 4k insns, we need
to choose eBPF insns that expand to more than 8 machine instructions.
Use BPF_LD_ABS as it is quite complex.  This forces the JIT to invert
the sense of the branch to branch around a long jump to the end.

This (somewhat) verifies that the branch inversion logic and target
address calculation of the long jumps are done correctly.

Signed-off-by: David Daney <david.daney@cavium.com>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agotools: bpf_jit_disasm: Handle large images.
David Daney [Tue, 13 Jun 2017 23:49:35 +0000 (16:49 -0700)]
tools: bpf_jit_disasm: Handle large images.

Dynamically allocate memory so that JIT images larger than the size of
the statically allocated array can be handled.

Signed-off-by: David Daney <david.daney@cavium.com>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agoMerge branch 'bpf-ctx-narrow'
David S. Miller [Wed, 14 Jun 2017 18:56:26 +0000 (14:56 -0400)]
Merge branch 'bpf-ctx-narrow'

Yonghong Song says:

====================
bpf: permit bpf program narrower loads for ctx fields

Today, if users try to access a ctx field through a narrower load, e.g.,
__be16 prot = __sk_buff->protocol, verifier will fail.
This set contains the verifier change to permit such loads for
certain ctx fields as well as the new test cases in selftests/bpf.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agoselftests/bpf: Add test cases to test narrower ctx field loads
Yonghong Song [Tue, 13 Jun 2017 22:52:14 +0000 (15:52 -0700)]
selftests/bpf: Add test cases to test narrower ctx field loads

Add test cases in test_verifier and test_progs.
Negative tests are added in test_verifier as well.
The test in test_progs will compare the value of narrower ctx field
load result vs. the masked value of normal full-field load result,
and will fail if they are not the same.

Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Yonghong Song <yhs@fb.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agobpf: permits narrower load from bpf program context fields
Yonghong Song [Tue, 13 Jun 2017 22:52:13 +0000 (15:52 -0700)]
bpf: permits narrower load from bpf program context fields

Currently, verifier will reject a program if it contains an
narrower load from the bpf context structure. For example,
        __u8 h = __sk_buff->hash, or
        __u16 p = __sk_buff->protocol
        __u32 sample_period = bpf_perf_event_data->sample_period
which are narrower loads of 4-byte or 8-byte field.

This patch solves the issue by:
  . Introduce a new parameter ctx_field_size to carry the
    field size of narrower load from prog type
    specific *__is_valid_access validator back to verifier.
  . The non-zero ctx_field_size for a memory access indicates
    (1). underlying prog type specific convert_ctx_accesses
         supporting non-whole-field access
    (2). the current insn is a narrower or whole field access.
  . In verifier, for such loads where load memory size is
    less than ctx_field_size, verifier transforms it
    to a full field load followed by proper masking.
  . Currently, __sk_buff and bpf_perf_event_data->sample_period
    are supporting narrowing loads.
  . Narrower stores are still not allowed as typical ctx stores
    are just normal stores.

Because of this change, some tests in verifier will fail and
these tests are removed. As a bonus, rename some out of bound
__sk_buff->cb access to proper field name and remove two
redundant "skb cb oob" tests.

Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Yonghong Song <yhs@fb.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agomacvlan: propagate the mac address change status for lowerdev
Zhang Shengju [Tue, 13 Jun 2017 14:45:11 +0000 (22:45 +0800)]
macvlan: propagate the mac address change status for lowerdev

The macvlan dev should propagate the return value of mac address change for
lower device in the passthru mode, instead of always return 0.

Signed-off-by: Zhang Shengju <zhangshengju@cmss.chinamobile.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agoMerge branch '10GbE' of git://git.kernel.org/pub/scm/linux/kernel/git/jkirsher/next...
David S. Miller [Wed, 14 Jun 2017 18:26:21 +0000 (14:26 -0400)]
Merge branch '10GbE' of git://git./linux/kernel/git/jkirsher/next-queue

Jeff Kirsher says:

====================
10GbE Intel Wired LAN Driver Updates 2017-06-13

This series contains updates to ixgbe and ixgbevf only.

Jake completes his fix ups for our drivers with the ixgbe changes to
resolve a race condition in processing timestamp requests.  These fixes
are the same fixes Jake applied earlier to the other drivers, including
the added statistic to help administrators know when an application
timestamp request is ignored.

With all the recent ixgbe/ixgbevf changes and fixes, Tony bumps the
the driver versions.  Then Tony provides a fix to resolve a static
analysis warning by changing a variable to unsigned integer since the
value can never be negative.

Emil fixes an issue for X550 devices where the qde parameter was being
ignored, so PFQDE.HIDE_VLAN was not being set.

Jeff Mahoney from SuSE fixes a possible kernel crash, where there was
a small window where tasks writing to the sriov_numvfs sysfs attribute
can sneak in after we call register_netdev().  So we need to call
pci_set_drvdata() before and not after register_netdev() to preserve the
intent of commit 0fb6a55cc31f ("ixgbe: fix crash on rmmod after probe
fail").
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agoixgbe: pci_set_drvdata must be called before register_netdev
Jeff Mahoney [Sat, 3 Jun 2017 22:01:17 +0000 (18:01 -0400)]
ixgbe: pci_set_drvdata must be called before register_netdev

We call pci_set_drvdata immediately after calling register_netdev,
which leaves a window where tasks writing to the sriov_numvfs sysfs
attribute can sneak in and crash the kernel.  register_netdev cleans
up after itself so placing pci_set_drvdata immediately before it
should preserve the intent of commit 0fb6a55cc31f ("ixgbe: fix crash
on rmmod after probe fail").

Fixes: 0fb6a55cc31f ("ixgbe: fix crash on rmmod after probe fail")
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
7 years agoixgbe: Resolve cppcheck format string warning
Tony Nguyen [Thu, 1 Jun 2017 19:06:05 +0000 (12:06 -0700)]
ixgbe: Resolve cppcheck format string warning

cppcheck warns that the format string is incorrect in the function
ixgbe_get_strings().  Since the value cannot be negative, change the
variable to unsigned which matches the format specifier.

Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
7 years agoixgbe: fix writes to PFQDE
Emil Tantilov [Tue, 23 May 2017 21:02:23 +0000 (14:02 -0700)]
ixgbe: fix writes to PFQDE

ixgbe_write_qde() was ignoring the qde parameter which resulted
in PFQDE.HIDE_VLAN not being set for X550.

Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
7 years agoixgbevf: Bump version number
Tony Nguyen [Thu, 18 May 2017 21:55:23 +0000 (14:55 -0700)]
ixgbevf: Bump version number

Update ixgbevf version number.

Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
7 years agoixgbe: Bump version number
Tony Nguyen [Thu, 18 May 2017 21:55:07 +0000 (14:55 -0700)]
ixgbe: Bump version number

Update ixgbe version number.

Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
7 years agoixgbe: check for Tx timestamp timeouts during watchdog
Jacob Keller [Wed, 3 May 2017 17:29:04 +0000 (10:29 -0700)]
ixgbe: check for Tx timestamp timeouts during watchdog

The ixgbe driver has logic to handle only one Tx timestamp at a time,
using a state bit lock to avoid multiple requests at once.

It may be possible, if incredibly unlikely, that a Tx timestamp event is
requested but never completes. Since we use an interrupt scheme to
determine when the Tx timestamp occurred we would never clear the state
bit in this case.

Add an ixgbe_ptp_tx_hang() function similar to the already existing
ixgbe_ptp_rx_hang() function. This function runs in the watchdog routine
and makes sure we eventually recover from this case instead of
permanently disabling Tx timestamps.

Note: there is no currently known way to cause this without hacking the
driver code to force it.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
7 years agoixgbe: add statistic indicating number of skipped Tx timestamps
Jacob Keller [Wed, 3 May 2017 17:29:00 +0000 (10:29 -0700)]
ixgbe: add statistic indicating number of skipped Tx timestamps

The ixgbe driver can only handle one Tx timestamp request at a time.
This means it is possible for an application timestamp request to be
ignored.

There is no easy way for an administrator to determine if this occurred.
Add a new statistic which tracks this, tx_hwtstamp_skipped.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
7 years agoixgbe: avoid permanent lock of *_PTP_TX_IN_PROGRESS
Jacob Keller [Wed, 3 May 2017 17:28:56 +0000 (10:28 -0700)]
ixgbe: avoid permanent lock of *_PTP_TX_IN_PROGRESS

The ixgbe driver uses a state bit lock to avoid handling more than one Tx
timestamp request at once. This is required because hardware is limited
to a single set of registers for Tx timestamps.

The state bit lock is not properly cleaned up during
ixgbe_xmit_frame_ring() if the transmit fails such as due to DMA or TSO
failure. In some hardware this results in blocking timestamps until the
service task times out. In other hardware this results in a permanent
lock of the timestamp bit because we never receive an interrupt
indicating the timestamp occurred, since indeed the packet was never
transmitted.

Fix this by checking for DMA and TSO errors in ixgbe_xmit_frame_ring() and
properly cleaning up after ourselves when these occur.

Reported-by: Reported-by: David Mirabito <davidm@metamako.com>
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
7 years agoixgbe: fix race condition with PTP_TX_IN_PROGRESS bits
Jacob Keller [Wed, 3 May 2017 17:28:53 +0000 (10:28 -0700)]
ixgbe: fix race condition with PTP_TX_IN_PROGRESS bits

Hardware related to the ixgbe driver is limited to handling a single Tx
timestamp request at a time. Thus, the driver ignores requests for Tx
timestamp while waiting for the current request to finish. It uses
a state bit lock which enforces that only one timestamp request is
honored at a time.

Unfortunately this suffers from a simple race condition. The bit lock is
not cleared until after skb_tstamp_tx() is called notifying applications
of a new Tx timestamp. Even a well behaved application sending only one
packet at a time and waiting for a response can wake up and send a new
packet before the bit lock is cleared. This results in needlessly
dropping some Tx timestamp requests.

We can fix this by unlocking the state bit as soon as we read the
Timestamp register, as this is the first point at which it is safe to
unlock.

To avoid issues with the skb pointer, we'll use a copy of the pointer
and set the global variable in the driver structure to NULL first. This
ensures that the next timestamp request does not modify our local copy
of the skb pointer.

This ensures that well behaved applications do not accidentally race
with the unlock bit. Obviously an application which sends multiple Tx
timestamp requests at once will still only timestamp one packet at
a time. Unfortunately there is nothing we can do about this.

Reported-by: David Mirabito <davidm@metamako.com>
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
7 years agoMerge branch 'net-dsa-Multi-CPU-ground-work'
David S. Miller [Tue, 13 Jun 2017 20:35:04 +0000 (16:35 -0400)]
Merge branch 'net-dsa-Multi-CPU-ground-work'

Florian Fainelli says:

====================
net: dsa: Multi-CPU ground work (v4)

This patch series prepares the ground for adding mutliple CPU port support to
DSA, and starts by removing redundant pieces of information such as
master_netdev which is cpu_dp->ethernet. Finally drivers are moved away from
directly accessing ds->dst->cpu_dp and use appropriate helper functions.

Note that if you have Device Tree blobs/platform configurations that are
currently listing multiple CPU ports, the proposed behavior in
dsa_ds_get_cpu_dp() will be to return the last bit set in ds->cpu_port_mask.

Future plans include:
- making dst->cpu_dp a flexible data structure (array, list, you name it)
- having the ability for drivers to return a default/preferred CPU port (if
  necessary)

Changes in v4:

- fixed build warning with NETPOLL enabled

Changes in v3:

- removed the last patch since it causes problems with bcm_sf2/b53 in a
  dual-CPU case (root cause known, proper fix underway)

- removed dsa_ds_get_cpu_dp()

Changes in v2:

- added Reviewed-by tags
- assign port->cpu_dp earlier before ops->setup() has run
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agonet: dsa: Introduce dsa_get_cpu_port()
Florian Fainelli [Tue, 13 Jun 2017 20:27:22 +0000 (13:27 -0700)]
net: dsa: Introduce dsa_get_cpu_port()

Introduce a helper function which will return a reference to the CPU
port used in a dsa_switch_tree. Right now this is a singleton, but this
will change once we introduce multi-CPU port support, so ease the
transition by converting the affected code paths.

Reviewed-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agonet: dsa: Associate slave network device with CPU port
Florian Fainelli [Tue, 13 Jun 2017 20:27:21 +0000 (13:27 -0700)]
net: dsa: Associate slave network device with CPU port

In preparation for supporting multiple CPU ports with DSA, have the
dsa_port structure know which CPU it is associated with. This will be
important in order to make sure the correct CPU is used for transmission
of the frames. If not for functional reasons, for performance (e.g: load
balancing) and forwarding decisions.

Reviewed-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agonet: dsa: Relocate master ethtool operations
Florian Fainelli [Tue, 13 Jun 2017 20:27:20 +0000 (13:27 -0700)]
net: dsa: Relocate master ethtool operations

Relocate master_ethtool_ops and master_orig_ethtool_ops into struct
dsa_port in order to be both consistent, and make things self contained
within the dsa_port structure.

This is a preliminary change to supporting multiple CPU port interfaces.

Reviewed-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agonet: dsa: Remove master_netdev and use dst->cpu_dp->netdev
Florian Fainelli [Tue, 13 Jun 2017 20:27:19 +0000 (13:27 -0700)]
net: dsa: Remove master_netdev and use dst->cpu_dp->netdev

In preparation for supporting multiple CPU ports, remove
dst->master_netdev and ds->master_netdev and replace them with only one
instance of the common object we have for a port: struct
dsa_port::netdev. ds->master_netdev is currently write only and would be
helpful in the case where we have two switches, both with CPU ports, and
also connected within each other, which the multi-CPU port patch series
would address.

While at it, introduce a helper function used in net/dsa/slave.c to
immediately get a reference on the master network device called
dsa_master_netdev().

Reviewed-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agocxgb4: handle serial flash interrupt
Ganesh Goudar [Tue, 13 Jun 2017 19:15:43 +0000 (00:45 +0530)]
cxgb4: handle serial flash interrupt

If SF bit is not cleared in PL_INT_CAUSE, subsequent non-data
interrupts are not raised.  Enable SF bit in Global Interrupt
Mask and handle it as non-fatal and hence eventually clear it.

Signed-off-by: Rahul Lakkireddy <rahul.lakkireddy@chelsio.com>
Signed-off-by: Ganesh Goudar <ganeshgr@chelsio.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agoof_mdio: move of_mdio_parse_addr to header file
Jon Mason [Tue, 13 Jun 2017 14:56:08 +0000 (10:56 -0400)]
of_mdio: move of_mdio_parse_addr to header file

The of_mdio_parse_addr() helper function is useful to other code, but
the module dependency chain causes issues.  To work around this, we can
move of_mdio_parse_addr() to be an inline function in the header file.
This gets rid of the dependencies and still allows for the reuse of
code.

Reported-by: Liviu Dudau <liviu@dudau.co.uk>
Signed-off-by: Jon Mason <jon.mason@broadcom.com>
Fixes: 342fa1964439 ("mdio: mux: make child bus walking more permissive and errors more verbose")
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agoselftests/bpf: make correct use of exit codes in bpf selftests
Jesper Dangaard Brouer [Tue, 13 Jun 2017 13:17:19 +0000 (15:17 +0200)]
selftests/bpf: make correct use of exit codes in bpf selftests

The selftests depend on using the shell exit code as a mean of
detecting the success or failure of test-binary executed.  The
appropiate output "[PASS]" or "[FAIL]" in generated by
tools/testing/selftests/lib.mk.

Notice that the exit code is masked with 255. Thus, be careful if
using the number of errors as the exit code, as 256 errors would be
seen as a success.

There are two standard defined exit(3) codes:
 /usr/include/stdlib.h
 #define EXIT_FAILURE    1       /* Failing exit status.  */
 #define EXIT_SUCCESS    0       /* Successful exit status.  */

Fix test_verifier.c to not use the negative value of variable
"results", but instead return EXIT_FAILURE.

Fix test_align.c and test_progs.c to actually use exit codes, before
they were always indicating success regardless of results.

Signed-off-by: Jesper Dangaard Brouer <brouer@redhat.com>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Reviewed-by: Fengguang Wu <fengguang.wu@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agofjes: remove duplicate set of flag IFF_BROADCAST
Zhang Shengju [Tue, 13 Jun 2017 12:49:49 +0000 (20:49 +0800)]
fjes: remove duplicate set of flag IFF_BROADCAST

Remove unnecessary setting of flag IFF_BROADCAST, since ether_setup
already does this.

Signed-off-by: Zhang Shengju <zhangshengju@cmss.chinamobile.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agonetworking: use skb_put_zero()
Johannes Berg [Tue, 13 Jun 2017 12:28:18 +0000 (14:28 +0200)]
networking: use skb_put_zero()

Use the recently introduced helper to replace the pattern of
skb_put() && memset(), this transformation was done with the
following spatch:

@@
identifier p;
expression len;
expression skb;
@@
-p = skb_put(skb, len);
-memset(p, 0, len);
+p = skb_put_zero(skb, len);

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agoMerge tag 'mac80211-next-for-davem-2017-06-13' of git://git.kernel.org/pub/scm/linux...
David S. Miller [Tue, 13 Jun 2017 17:52:37 +0000 (13:52 -0400)]
Merge tag 'mac80211-next-for-davem-2017-06-13' of git://git./linux/kernel/git/jberg/mac80211-next

Johannes Berg says:

====================
A couple of weeks worth of updates - looks like things are quiet:
 * merged net-next back to get a patch from net that another patch
   here depends on
 * various small improvements/cleanups across the board
 * 4-way handshake offload (many thanks to Arend for shepherding that)
 * mesh CSA/DFS support in mac80211
 * the skb_put_zero() we discussed previously
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agoMerge tag 'batadv-next-for-davem-20170613' of git://git.open-mesh.org/linux-merge
David S. Miller [Tue, 13 Jun 2017 17:47:16 +0000 (13:47 -0400)]
Merge tag 'batadv-next-for-davem-20170613' 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

 - decrease maximum fragment size, by Matthias Schiffer

 - Clean up seqfile writing, by Markus Elfring (2 patches)

 - use __func__ in debug messages, by Sven Eckelmann

 - Mark tpmeter initializers with __init, by Antonio Quartulli

 - ignore loop detection MAC addresses, by Simon Wunderlich

 - clean up some return handling, by Simon Wunderlich

 - improve ELP throughput value handling for WiFi neighbors
   in BATMAN V/ELP, by Sven Eckelmann (2 patches)
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agonet: phy: Make phy_ethtool_ksettings_get return void
yuval.shaia@oracle.com [Tue, 13 Jun 2017 07:09:46 +0000 (10:09 +0300)]
net: phy: Make phy_ethtool_ksettings_get return void

Make return value void since function never return meaningfull value

Signed-off-by: Yuval Shaia <yuval.shaia@oracle.com>
Acked-by: Sergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agoMerge branch 'MDIO-bus-reset-GPIO-cleanups'
David S. Miller [Tue, 13 Jun 2017 16:56:43 +0000 (12:56 -0400)]
Merge branch 'MDIO-bus-reset-GPIO-cleanups'

Sergei Shtylyov says:

====================
MDIO bus reset GPIO cleanups

Commit 4c5e7a2c0501 ("dt-bindings: mdio: Clarify binding document")
declared that a MDIO reset GPIO property should have only a single GPIO
reference/specifier, however the supporting code was left intact...
Here's a couple of the obvious cleanups to that code:

[1/2] mdio_bus: handle only single PHY reset GPIO
[2/2] mdio_bus: use devm_gpiod_get_optional()
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agomdio_bus: use devm_gpiod_get_optional()
Sergei Shtylyov [Mon, 12 Jun 2017 20:55:39 +0000 (23:55 +0300)]
mdio_bus: use devm_gpiod_get_optional()

The MDIO reset GPIO is really a classical optional GPIO property case,
so devm_gpiod_get_optional() should have been used, not devm_gpiod_get().
Doing this  saves several LoCs...

Signed-off-by: Sergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agomdio_bus: handle only single PHY reset GPIO
Sergei Shtylyov [Mon, 12 Jun 2017 20:55:38 +0000 (23:55 +0300)]
mdio_bus: handle only single PHY reset GPIO

Commit 4c5e7a2c0501 ("dt-bindings: mdio: Clarify binding document")
declared that a MDIO reset GPIO property should have only a single GPIO
reference/specifier, however the supporting code was left intact, still
burdening the kernel with now apparently useless loops -- get rid of them.

Signed-off-by: Sergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agoibmvnic: Remove netdev notify for failover resets
Nathan Fontenot [Tue, 13 Jun 2017 00:47:45 +0000 (20:47 -0400)]
ibmvnic: Remove netdev notify for failover resets

When handling a driver reset due to a failover of the backing
server on the vios, doing the netdev_notify_peers() can cause
network traffic to stall or halt. Remove the netdev notify call
for failover resets.

Signed-off-by: Nathan Fontenot <nfont@linux.vnet.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agoibmvnic: Client-initiated failover
Thomas Falcon [Mon, 12 Jun 2017 17:35:04 +0000 (12:35 -0500)]
ibmvnic: Client-initiated failover

The IBM vNIC protocol provides support for the user to initiate
a failover from the client LPAR in case the current backing infrastructure
is deemed inadequate or in an error state.

Support for two H_VIOCTL sub-commands for vNIC devices are required
to implement this function. These commands are H_GET_SESSION_TOKEN
and H_SESSION_ERR_DETECTED.

"[H_GET_SESSION_TOKEN] is used to obtain a session token from a VNIC client
adapter.  This token is opaque to the caller and is intended to be used in
tandem with the SESSION_ERROR_DETECTED vioctl subfunction."

"[H_SESSION_ERR_DETECTED] is used to report that the currently active
backing device for a VNIC client adapter is behaving poorly, and that
the hypervisor should attempt to fail over to a different backing device,
if one is available."

To provide tools access to this functionality the vNIC driver creates a
sysfs file that, when written to, will send a request to pHyp to failover
to a different backing device.

Signed-off-by: Thomas Falcon <tlfalcon@linux.vnet.ibm.com>
Reviewed-by: Nathan Fontenot <nfont@linux.vnet.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agonet: mvpp2: enable basic 10G support
Antoine Ténart [Mon, 12 Jun 2017 14:01:39 +0000 (16:01 +0200)]
net: mvpp2: enable basic 10G support

On GOP port 0 two MAC modes are available: GMAC and XLG. The XLG MAC is
used for 10G connectivity. This patch adds a basic 10G support by
allowing to use the XLG MAC on port 0 and by reworking the
port_enable/disable functions so that the XLG MAC is configured when
using 10G.

Signed-off-by: Antoine Tenart <antoine.tenart@free-electrons.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agoMerge branch 'dsa-mv88e6xxx-port-macros-cosmetics'
David S. Miller [Tue, 13 Jun 2017 15:23:13 +0000 (11:23 -0400)]
Merge branch 'dsa-mv88e6xxx-port-macros-cosmetics'

Vivien Didelot says:

====================
net: dsa: mv88e6xxx: port macros cosmetics

This patch series brings no functional changes.

It prefixes all common port registers macros with MV88E6XXX_PORT.
If registers or some bits differs between switch models, a reference
model is chosen (e.g. MV88E6390_PORT_MAC_CTL_SPEED_10000.)

The register names are documented as found in the datasheets.

Avoid BIT() and shifts defines and prefer a better representation of the
Marvell switch registers with ordered, hexadecimal, 16-bit values.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agonet: dsa: mv88e6xxx: prefix remaining port macros
Vivien Didelot [Mon, 12 Jun 2017 16:37:45 +0000 (12:37 -0400)]
net: dsa: mv88e6xxx: prefix remaining port macros

For implicit namespacing and clarity, prefix the remaining common Port
Registers macros with MV88E6XXX_PORT.

Document the register and prefer ordered hex masks values for all
Marvell 16-bit registers.

Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agonet: dsa: mv88e6xxx: prefix Port IEEE Priority mapping macros
Vivien Didelot [Mon, 12 Jun 2017 16:37:44 +0000 (12:37 -0400)]
net: dsa: mv88e6xxx: prefix Port IEEE Priority mapping macros

For implicit namespacing and clarity, prefix the common Port IEEE
Priority Remapping registers macros with MV88E6095_PORT_IEEE_PRIO.

The 88E6390 family turned the 0x18 register into a single indirect
table, document that at the same time.

Document the register and prefer ordered hex masks values for all
Marvell 16-bit registers.

Also fix the following checkpatch checks with a temporary variable:

    CHECK: Alignment should match open parenthesis
    #65: FILE: drivers/net/dsa/mv88e6xxx/port.c:932:
    + err = mv88e6xxx_port_ieeepmt_write(chip, port,
    +    MV88E6390_PORT_IEEE_PRIO_MAP_TABLE_INGRESS_PCP,

Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agonet: dsa: mv88e6xxx: prefix Port Association Vector macros
Vivien Didelot [Mon, 12 Jun 2017 16:37:43 +0000 (12:37 -0400)]
net: dsa: mv88e6xxx: prefix Port Association Vector macros

For implicit namespacing and clarity, prefix the common Port Association
Vector Register macros with MV88E6XXX_PORT_ASSOC_VECTOR.

Document the register and prefer ordered hex masks values for all
Marvell 16-bit registers.

Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agonet: dsa: mv88e6xxx: prefix Port Egress Rate Control macros
Vivien Didelot [Mon, 12 Jun 2017 16:37:42 +0000 (12:37 -0400)]
net: dsa: mv88e6xxx: prefix Port Egress Rate Control macros

For implicit namespacing and clarity, prefix the common Port Egress Rate
Control and Port Egress Rate Control 2 registers macros with
MV88E6XXX_PORT_EGRESS_RATE_CTL1 and MV88E6XXX_PORT_EGRESS_RATE_CTL2.

Document the register and prefer ordered hex masks values for all
Marvell 16-bit registers.

Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agonet: dsa: mv88e6xxx: prefix Port Control 2 macros
Vivien Didelot [Mon, 12 Jun 2017 16:37:41 +0000 (12:37 -0400)]
net: dsa: mv88e6xxx: prefix Port Control 2 macros

For implicit namespacing and clarity, prefix the common Port Control 2
Register macros with MV88E6XXX_PORT_CTL2 and the ones which differ
between implementations with a chosen reference model
(e.g. MV88E6095_PORT_CTL2_CPU_PORT_MASK.)

Document the register and prefer ordered hex masks values for all
Marvell 16-bit registers.

Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agonet: dsa: mv88e6xxx: prefix Port Default VLAN macros
Vivien Didelot [Mon, 12 Jun 2017 16:37:40 +0000 (12:37 -0400)]
net: dsa: mv88e6xxx: prefix Port Default VLAN macros

For implicit namespacing and clarity, prefix the common Port Default
VLAN Register macros with MV88E6XXX_PORT_DEFAULT_VLAN.

Document the register and prefer ordered hex masks values for all
Marvell 16-bit registers.

Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agonet: dsa: mv88e6xxx: prefix Port Based VLAN macros
Vivien Didelot [Mon, 12 Jun 2017 16:37:39 +0000 (12:37 -0400)]
net: dsa: mv88e6xxx: prefix Port Based VLAN macros

For implicit namespacing and clarity, prefix the common Port Based VLAN
Register macros with MV88E6XXX_PORT_BASE_VLAN.

Document the register and prefer ordered hex masks values for all
Marvell 16-bit registers.

Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agonet: dsa: mv88e6xxx: prefix Port Control 1 macros
Vivien Didelot [Mon, 12 Jun 2017 16:37:38 +0000 (12:37 -0400)]
net: dsa: mv88e6xxx: prefix Port Control 1 macros

For implicit namespacing and clarity, prefix the common Port Control 1
Register macros with MV88E6XXX_PORT_CTL1.

Document the register and prefer ordered hex masks values for all
Marvell 16-bit registers.

Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agonet: dsa: mv88e6xxx: prefix Port Control macros
Vivien Didelot [Mon, 12 Jun 2017 16:37:37 +0000 (12:37 -0400)]
net: dsa: mv88e6xxx: prefix Port Control macros

For implicit namespacing and clarity, prefix the common Port Control
Register macros with MV88E6XXX_PORT_CTL0 and the ones which differ
between implementations with a chosen reference model
(e.g. MV88E6185_PORT_CTL0_USE_TAG.)

The reason for CTL0 is to make it clear between the badly named
"Port Control", "Port Control 1" and "Port Control 2" registers.

Document the register and prefer ordered hex masks values for all
Marvell 16-bit registers.

Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agonet: dsa: mv88e6xxx: prefix Port Switch ID macros
Vivien Didelot [Mon, 12 Jun 2017 16:37:36 +0000 (12:37 -0400)]
net: dsa: mv88e6xxx: prefix Port Switch ID macros

For implicit namespacing and clarity, prefix the common Switch ID
Register macros with MV88E6XXX_PORT_SWITCH_ID.

Document the register and prefer ordered hex masks values for all
Marvell 16-bit registers, this means shifting their values by 4.

Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agonet: dsa: mv88e6xxx: prefix Port Jamming macros
Vivien Didelot [Mon, 12 Jun 2017 16:37:35 +0000 (12:37 -0400)]
net: dsa: mv88e6xxx: prefix Port Jamming macros

For implicit namespacing and clarity, prefix the common Port Jamming
Control Register macros with MV88E6XXX_PORT_JAM_CTL and the ones which
differ between implementations with a chosen reference model
(e.g. MV88E6097_PORT_JAM_CTL.)

The 88E6390 family renamed the register to Flow Control and turned it
into an indirect table. Document that as well.

Document the register and prefer ordered hex masks values for all
Marvell 16-bit registers.

Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agonet: dsa: mv88e6xxx: prefix Port MAC Control macros
Vivien Didelot [Mon, 12 Jun 2017 16:37:34 +0000 (12:37 -0400)]
net: dsa: mv88e6xxx: prefix Port MAC Control macros

For implicit namespacing and clarity, prefix the common MAC Control
Register macros with MV88E6XXX_PORT_MAC_CTL and the ones which differ
between implementations with a chosen reference model
(e.g. MV88E6065_PORT_MAC_CTL_SPEED_200.)

Document the register and prefer ordered hex masks values for all
Marvell 16-bit registers.

Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agonet: dsa: mv88e6xxx: prefix Port Status macros
Vivien Didelot [Mon, 12 Jun 2017 16:37:33 +0000 (12:37 -0400)]
net: dsa: mv88e6xxx: prefix Port Status macros

For implicit namespacing and clarity, prefix the common Port Status
Register macros with MV88E6XXX_PORT_STS and the ones which differ
between implementations with a chosen reference model
(e.g. MV88E6352_PORT_STS_EEE.)

Document the register and prefer ordered hex masks values for all
Marvell 16-bit registers.

Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agobatman-adv: Accept only filled wifi station info
Sven Eckelmann [Fri, 9 Jun 2017 15:06:51 +0000 (17:06 +0200)]
batman-adv: Accept only filled wifi station info

The wifi driver can decide to not provide parts of the station info. For
example, the expected throughput of the station can be omitted when the
used rate control doesn't provide this kind of information.

The B.A.T.M.A.N. V implementation must therefore check the filled bitfield
before it tries to access the expected_throughput of the returned
station_info.

Reported-by: Alvaro Antelo <alvaro.antelo@gmail.com>
Fixes: c833484e5f38 ("batman-adv: ELP - compute the metric based on the estimated throughput")
Signed-off-by: Sven Eckelmann <sven@narfation.org>
Reviewed-by: Marek Lindner <mareklindner@neomailbox.ch>
Signed-off-by: Simon Wunderlich <sw@simonwunderlich.de>
7 years agobatman-adv: Use default throughput value on cfg80211 error
Sven Eckelmann [Fri, 9 Jun 2017 15:06:50 +0000 (17:06 +0200)]
batman-adv: Use default throughput value on cfg80211 error

A wifi interface should never be handled like an ethernet devices. The
parser of the cfg80211 output must therefore skip the ethtool code when
cfg80211_get_station returned an error.

Fixes: f44a3ae9a281 ("batman-adv: refactor wifi interface detection")
Signed-off-by: Sven Eckelmann <sven@narfation.org>
Reviewed-by: Marek Lindner <mareklindner@neomailbox.ch>
Signed-off-by: Simon Wunderlich <sw@simonwunderlich.de>
7 years agonet: rfkill: gpio: Switch to devm_acpi_dev_add_driver_gpios()
Andy Shevchenko [Sat, 10 Jun 2017 19:10:34 +0000 (22:10 +0300)]
net: rfkill: gpio: Switch to devm_acpi_dev_add_driver_gpios()

Switch to use managed variant of acpi_dev_add_driver_gpios() to simplify
error path and fix potentially wrong assingment if ->probe() fails.

Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
7 years agomac80211: add the action to the drv_ampdu_action tracepoint
Emmanuel Grumbach [Sat, 10 Jun 2017 10:52:46 +0000 (13:52 +0300)]
mac80211: add the action to the drv_ampdu_action tracepoint

It is very useful to know what ampdu action is currently
happening. Add this information to the tracepoint.

Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
7 years agonl80211: remove desciption about request from NL80211_CMD_ROAM
Arend Van Spriel [Fri, 9 Jun 2017 12:08:46 +0000 (13:08 +0100)]
nl80211: remove desciption about request from NL80211_CMD_ROAM

The description of NL80211_CMD_ROAM indicated possibility for a
request to roam issued by user-space. However, it also states that
as not being implemented right now. This has been so since commit
b23aa676ab9d ("cfg80211: connect/disconnect API") added in 2009.
So it seems safe to assume it will not be added any time soon and
thus remove it.

Signed-off-by: Arend van Spriel <arend.vanspriel@broadcom.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
7 years agonl80211: add authorized flag to ROAM event
Avraham Stern [Fri, 9 Jun 2017 12:08:45 +0000 (13:08 +0100)]
nl80211: add authorized flag to ROAM event

Drivers that initiate roaming while being connected to a network that
uses 802.1X authentication need to inform user space if 802.1X
authentication is further required after roaming.
For example, when using the Fast transition protocol, roaming within
the mobility domain does not require new 802.1X authentication, but
roaming to another mobility domain does.
In addition, some drivers may not support 802.1X authentication
(so it has to be done in user space), while other drivers do.

Add a flag to the roaming notification to indicate if user space is
required to do 802.1X authentication after the roaming or not.
This flag will only be used for networks that use 802.1X
authentication. For networks that do not use 802.1X authentication it
is assumed that no further action is required from user space after
the roaming notification.

Signed-off-by: Avraham Stern <avraham.stern@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
[arend.vanspriel@broadcom.com reuse NL80211_ATTR_PORT_AUTHORIZED]
Signed-off-by: Arend van Spriel <arend.vanspriel@broadcom.com>
[rebase to apply w/o the flag in CONNECT]
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
7 years agocfg80211: support 4-way handshake offloading for 802.1X
Avraham Stern [Fri, 9 Jun 2017 12:08:43 +0000 (13:08 +0100)]
cfg80211: support 4-way handshake offloading for 802.1X

Add API for setting the PMK to the driver. For FT support, allow
setting also the PMK-R0 Name.

This can be used by drivers that support 4-Way handshake offload
while IEEE802.1X authentication is managed by upper layers.

Signed-off-by: Avraham Stern <avraham.stern@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
[arend.vanspriel@broadcom.com: add WANT_1X_4WAY_HS attribute]
Signed-off-by: Arend van Spriel <arend.vanspriel@broadcom.com>
[reword NL80211_EXT_FEATURE_4WAY_HANDSHAKE_STA_1X docs a bit to
say that the device may require it]
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
7 years agocfg80211: support 4-way handshake offloading for WPA/WPA2-PSK
Eliad Peller [Fri, 9 Jun 2017 12:08:42 +0000 (13:08 +0100)]
cfg80211: support 4-way handshake offloading for WPA/WPA2-PSK

Let drivers advertise support for station-mode 4-way handshake
offloading with a new NL80211_EXT_FEATURE_4WAY_HANDSHAKE_STA_PSK flag.

Extend use of NL80211_ATTR_PMK attribute indicating it might be passed
as part of NL80211_CMD_CONNECT command, and contain the PSK (which is
the PMK, hence the name.)

The driver/device is assumed to handle the 4-way handshake by
itself in this case (including key derivations, etc.), instead
of relying on the supplicant.

This patch is somewhat based on this one (by Vladimir Kondratiev):
https://patchwork.kernel.org/patch/1309561/.

Signed-off-by: Vladimir Kondratiev <qca_vkondrat@qca.qualcomm.com>
Signed-off-by: Eliad Peller <eliadx.peller@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
[arend.vanspriel@broadcom.com rebase dealing with existing ATTR_PMK]
Signed-off-by: Arend van Spriel <arend.vanspriel@broadcom.com>
[reword NL80211_EXT_FEATURE_4WAY_HANDSHAKE_STA_PSK docs to indicate
that this offload might be required]
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
7 years agonet: phy: marvell: Show complete link partner advertising
Thomas Bogendoerfer [Mon, 12 Jun 2017 12:54:57 +0000 (14:54 +0200)]
net: phy: marvell: Show complete link partner advertising

Give back all modes advertised by the link partner. This change brings
the marvell phy driver in line with all other phy drivers.

Signed-off-by: Thomas Bogendoerfer <tbogendoerfer@suse.de>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agovxlan: dont migrate permanent fdb entries during learn
Roopa Prabhu [Sun, 11 Jun 2017 23:32:50 +0000 (16:32 -0700)]
vxlan: dont migrate permanent fdb entries during learn

This patch fixes vxlan_snoop to not move permanent fdb entries
on learn events. This is consistent with the bridge fdb
handling of permanent entries.

Fixes: 26a41ae60438 ("vxlan: only migrate dynamic FDB entries")
Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agoMerge tag 'wireless-drivers-next-for-davem-2017-06-12' of git://git.kernel.org/pub...
David S. Miller [Mon, 12 Jun 2017 14:14:29 +0000 (10:14 -0400)]
Merge tag 'wireless-drivers-next-for-davem-2017-06-12' of git://git./linux/kernel/git/kvalo/wireless-drivers-next

Kalle Valo says:

====================
wireless-drivers-next patches for 4.13

The first pull request for 4.13. We have a new driver qtnfmac, but
also rsi driver got a support for new firmware and supporting ath10k
SDIO devices was started.

Major changes:

ath10k

* add initial SDIO support (still work in progress)

rsi

* new loading for the new firmware version

rtlwifi

* final patches for the new btcoex support

rt2x00

* add device ID for Epson WN7512BEP

qtnfmac

* new driver for Quantenna QSR10G chipsets
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agoMerge branch 'udp-reduce-cache-pressure'
David S. Miller [Mon, 12 Jun 2017 14:01:30 +0000 (10:01 -0400)]
Merge branch 'udp-reduce-cache-pressure'

Paolo Abeni says:

====================
udp: reduce cache pressure

In the most common use case, many skb fields are not used by recvmsg(), and
the few ones actually accessed lays on cold cachelines, which leads to several
cache miss per packet.

This patch series attempts to reduce such misses with different strategies:
* caching the interesting fields in the scratched space
* avoid accessing at all uninteresting fields
* prefetching

Tested using the udp_sink program by Jesper[1] as the receiver, an h/w l4 rx
hash on the ingress nic, so that the number of ingress nic rx queues hit by the
udp traffic could be controlled via ethtool -L.

The udp_sink program was bound to the first idle cpu, to get more
stable numbers.

On a single numa node receiver:

nic rx queues           vanilla                 patched kernel      delta
1                       1850 kpps               1850 kpps           0%
2                       2370 kpps               2700 kpps           13.9%
16                      2000 kpps               2220 kpps           11%

[1] https://github.com/netoptimizer/network-testing/blob/master/src/udp_sink.c

v1 -> v2:
  - replaced secpath_reset() with skb_release_head_state()
  - changed udp_dev_scratch fields types to u{32,16} variant,
    replaced bitfield with bool

v2 -> v3:
  - no changes, tested against apachebench for performances regression
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agoudp: try to avoid 2 cache miss on dequeue
Paolo Abeni [Mon, 12 Jun 2017 09:23:43 +0000 (11:23 +0200)]
udp: try to avoid 2 cache miss on dequeue

when udp_recvmsg() is executed, on x86_64 and other archs, most skb
fields are on cold cachelines.
If the skb are linear and the kernel don't need to compute the udp
csum, only a handful of skb fields are required by udp_recvmsg().
Since we already use skb->dev_scratch to cache hot data, and
there are 32 bits unused on 64 bit archs, use such field to cache
as much data as we can, and try to prefetch on dequeue the relevant
fields that are left out.

This can save up to 2 cache miss per packet.

v1 -> v2:
  - changed udp_dev_scratch fields types to u{32,16} variant,
    replaced bitfiled with bool

Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Acked-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agoudp: avoid a cache miss on dequeue
Paolo Abeni [Mon, 12 Jun 2017 09:23:42 +0000 (11:23 +0200)]
udp: avoid a cache miss on dequeue

Since UDP no more uses sk->destructor, we can clear completely
the skb head state before enqueuing. Amend and use
skb_release_head_state() for that.

All head states share a single cacheline, which is not
normally used/accesses on dequeue. We can avoid entirely accessing
such cacheline implementing and using in the UDP code a specialized
skb free helper which ignores the skb head state.

This saves a cacheline miss at skb deallocation time.

v1 -> v2:
  replaced secpath_reset() with skb_release_head_state()

Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Acked-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agonet: factor out a helper to decrement the skb refcount
Paolo Abeni [Mon, 12 Jun 2017 09:23:41 +0000 (11:23 +0200)]
net: factor out a helper to decrement the skb refcount

The same code is replicated in 3 different places; move it to a
common helper.

Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Acked-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agosh_eth: add support for changing MTU
Niklas Söderlund [Mon, 12 Jun 2017 08:39:03 +0000 (10:39 +0200)]
sh_eth: add support for changing MTU

The hardware supports the MTU to be changed and the driver it self is
somewhat prepared to support this. This patch hooks up the callbacks to
be able to change the MTU from user-space.

Signed-off-by: Niklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Acked-by: Sergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agobpf, arm64: take advantage of stack_depth tracking
Daniel Borkmann [Sun, 11 Jun 2017 01:55:27 +0000 (03:55 +0200)]
bpf, arm64: take advantage of stack_depth tracking

Make use of recently implemented stack_depth tracking for arm64 JIT,
so that stack usage can be reduced heavily for programs not using
tail calls at least.

Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agoMerge tag 'mlx5-updates-2017-06-11' of git://git.kernel.org/pub/scm/linux/kernel...
David S. Miller [Sun, 11 Jun 2017 22:10:42 +0000 (18:10 -0400)]
Merge tag 'mlx5-updates-2017-06-11' of git://git./linux/kernel/git/saeed/linux

Saeed Mahameed says:

====================
mlx5-updates-2017-06-11

This series provides updates to mlx5 header rewrite feature, from Or Gerlitz.
and three more small updates From maor and eran.

-------
Or says:

Packets belonging to flows which are different by matching may still need
to go through the same header re-writes (e.g set the current routing hop
MACs and issue TTL decrement).  To minimize the number of modify header
IDs, we add a cache for header re-write IDs which is keyed by the binary
chain of modify header actions.

The caching is supported for both eswitch and NIC use-cases, where the
actual conversion of the code to use caching comes in separate patches,
one per use-case.

Using a per field mask field, the TC pedit action supports modifying
partial fields. The last patch enables offloading that.
-------

From Maor, update flow table commands layout to the latest HW spec.
From Eran, ethtool connector type reporting updates.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agonet: ethernet: ti: cpdma: do not enable host error misc irq
Grygorii Strashko [Thu, 8 Jun 2017 18:51:52 +0000 (13:51 -0500)]
net: ethernet: ti: cpdma: do not enable host error misc irq

CPSW driver does not handle this interrupt, so there are no reasons to enable
it in hardware.

Signed-off-by: Grygorii Strashko <grygorii.strashko@ti.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agonet: ethernet: ti: cpsw: enable HWTSTAMP_FILTER_PTP_V1_L4_EVENT filter
Grygorii Strashko [Thu, 8 Jun 2017 18:51:31 +0000 (13:51 -0500)]
net: ethernet: ti: cpsw: enable HWTSTAMP_FILTER_PTP_V1_L4_EVENT filter

CPSW driver supports PTP v1 messages, but for unknown reasons this filter
is not advertised. As result,
./tools/testing/selftests/networking/timestamping/timestamping utility
can't be used for testing of CPSW RX timestamping with option
SOF_TIMESTAMPING_RX_HARDWARE, because it uses
HWTSTAMP_FILTER_PTP_V1_L4_SYNC filter.

Hence, fix it by advertising HWTSTAMP_FILTER_PTP_V1_L4_XXX filters
in CPSW driver.

Signed-off-by: Grygorii Strashko <grygorii.strashko@ti.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agoMerge branch 'bpf-misc-updates'
David S. Miller [Sat, 10 Jun 2017 23:05:47 +0000 (19:05 -0400)]
Merge branch 'bpf-misc-updates'

Daniel Borkmann says:

====================
Misc BPF updates

This set contains a couple of misc updates: stack usage reduction
for perf_sample_data in tracing progs, reduction of stale data in
verifier on register state transitions that I still had in my queue
and few selftest improvements as well as bpf_set_hash() helper for
tc programs.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agobpf: add bpf_set_hash helper for tc progs
Daniel Borkmann [Sat, 10 Jun 2017 22:50:47 +0000 (00:50 +0200)]
bpf: add bpf_set_hash helper for tc progs

Allow for tc BPF programs to set a skb->hash, apart from clearing
and triggering a recalc that we have right now. It allows for BPF
to implement a custom hashing routine for skb_get_hash().

Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agobpf: remove cg_skb_func_proto and use sk_filter_func_proto directly
Daniel Borkmann [Sat, 10 Jun 2017 22:50:46 +0000 (00:50 +0200)]
bpf: remove cg_skb_func_proto and use sk_filter_func_proto directly

Since cg_skb_func_proto() doesn't do anything else than just calling
into sk_filter_func_proto(), remove it and set sk_filter_func_proto()
directly for .get_func_proto callback.

Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agobpf, tests: set rlimit also for test_align, so it doesn't fail
Daniel Borkmann [Sat, 10 Jun 2017 22:50:45 +0000 (00:50 +0200)]
bpf, tests: set rlimit also for test_align, so it doesn't fail

When running all the tests, through 'make run_tests', I had
test_align failing due to insufficient rlimit. Set it the same
way as all other test cases from BPF selftests do, so that
test case properly loads everything.

  [...]
  Summary: 7 PASSED, 1 FAILED
  selftests: test_progs [PASS]
  /home/foo/net-next/tools/testing/selftests/bpf
  Test   0: mov ... Failed to load program.
  FAIL
  Test   1: shift ... Failed to load program.
  FAIL
  Test   2: addsub ... Failed to load program.
  FAIL
  Test   3: mul ... Failed to load program.
  FAIL
  Test   4: unknown shift ... Failed to load program.
  FAIL
  Test   5: unknown mul ... Failed to load program.
  FAIL
  Test   6: packet const offset ... Failed to load program.
  FAIL
  Test   7: packet variable offset ... Failed to load program.
  FAIL
  Results: 0 pass 8 fail
  selftests: test_align [PASS]
  [...]

Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agobpf, tests: add a test for htab lookup + update traversal
Daniel Borkmann [Sat, 10 Jun 2017 22:50:44 +0000 (00:50 +0200)]
bpf, tests: add a test for htab lookup + update traversal

Add a test case to track behaviour when traversing and updating the
htab map. We recently used such traversal, so it's quite useful to
keep it as an example in selftests.

Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agobpf: reset id on spilled regs in clear_all_pkt_pointers
Daniel Borkmann [Sat, 10 Jun 2017 22:50:43 +0000 (00:50 +0200)]
bpf: reset id on spilled regs in clear_all_pkt_pointers

Right now, we don't reset the id of spilled registers in case of
clear_all_pkt_pointers(). Given pkt_pointers are highly likely to
contain an id, do so by reusing __mark_reg_unknown_value().

Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agobpf: reset id on CONST_IMM transition
Daniel Borkmann [Sat, 10 Jun 2017 22:50:42 +0000 (00:50 +0200)]
bpf: reset id on CONST_IMM transition

Whenever we set the register to the type CONST_IMM, we currently don't
reset the id to 0. id member is not used in CONST_IMM case, so don't
let it become stale, where pruning won't be able to match later on.

Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agobpf: don't check spilled reg state for non-STACK_SPILLed type slots
Daniel Borkmann [Sat, 10 Jun 2017 22:50:41 +0000 (00:50 +0200)]
bpf: don't check spilled reg state for non-STACK_SPILLed type slots

spilled_regs[] state is only used for stack slots of type STACK_SPILL,
never for STACK_MISC. Right now, in states_equal(), even if we have
old and current stack state of type STACK_MISC, we compare spilled_regs[]
for that particular offset. Just skip these like we do everywhere else.

Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agobpf: avoid excessive stack usage for perf_sample_data
Daniel Borkmann [Sat, 10 Jun 2017 22:50:40 +0000 (00:50 +0200)]
bpf: avoid excessive stack usage for perf_sample_data

perf_sample_data consumes 386 bytes on stack, reduce excessive stack
usage and move it to per cpu buffer. It's allowed due to preemption
being disabled for tracing, xdp and tc programs, thus at all times
only one program can run on a specific CPU and programs cannot run
from interrupt. We similarly also handle bpf_pt_regs.

Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agonet: fec: Add a fec_enet_clear_ethtool_stats() stub for CONFIG_M5272
Fabio Estevam [Sat, 10 Jun 2017 01:37:22 +0000 (22:37 -0300)]
net: fec: Add a fec_enet_clear_ethtool_stats() stub for CONFIG_M5272

Commit 2b30842b23b9 ("net: fec: Clear and enable MIB counters on imx51")
introduced fec_enet_clear_ethtool_stats(), but missed to add a stub
for the CONFIG_M5272=y case, causing build failure for the
m5272c3_defconfig.

Add the missing empty stub to fix the build failure.

Fixes: Commit 2b30842b23b9 ("net: fec: Clear and enable MIB counters on imx51")
Reported-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: Fabio Estevam <fabio.estevam@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agoRemove the redundant skb->dev initialization in ip6_fragment
Chenbo Feng [Sat, 10 Jun 2017 19:35:38 +0000 (12:35 -0700)]
Remove the redundant skb->dev initialization in ip6_fragment

After moves the skb->dev and skb->protocol initialization into
ip6_output, setting the skb->dev inside ip6_fragment is unnecessary.

Fixes: 97a7a37a7b7b("ipv6: Initial skb->dev and skb->protocol in ip6_output")
Signed-off-by: Chenbo Feng <fengc@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agosctp: no need to check assoc id before calling sctp_assoc_set_id
Xin Long [Sat, 10 Jun 2017 07:27:12 +0000 (15:27 +0800)]
sctp: no need to check assoc id before calling sctp_assoc_set_id

sctp_assoc_set_id does the assoc id check in the beginning when
processing dupcookie, no need to do the same check before calling
it.

v1->v2:
  fix some typo errs Marcelo pointed in changelog.

Signed-off-by: Xin Long <lucien.xin@gmail.com>
Acked-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agosctp: use read_lock_bh in sctp_eps_seq_show
Xin Long [Sat, 10 Jun 2017 07:13:32 +0000 (15:13 +0800)]
sctp: use read_lock_bh in sctp_eps_seq_show

This patch is to use read_lock_bh instead of local_bh_disable
and read_lock in sctp_eps_seq_show.

Signed-off-by: Xin Long <lucien.xin@gmail.com>
Acked-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agosctp: fix recursive locking warning in sctp_do_peeloff
Xin Long [Sat, 10 Jun 2017 06:56:56 +0000 (14:56 +0800)]
sctp: fix recursive locking warning in sctp_do_peeloff

Dmitry got the following recursive locking report while running syzkaller
fuzzer, the Call Trace:
 __dump_stack lib/dump_stack.c:16 [inline]
 dump_stack+0x2ee/0x3ef lib/dump_stack.c:52
 print_deadlock_bug kernel/locking/lockdep.c:1729 [inline]
 check_deadlock kernel/locking/lockdep.c:1773 [inline]
 validate_chain kernel/locking/lockdep.c:2251 [inline]
 __lock_acquire+0xef2/0x3430 kernel/locking/lockdep.c:3340
 lock_acquire+0x2a1/0x630 kernel/locking/lockdep.c:3755
 lock_sock_nested+0xcb/0x120 net/core/sock.c:2536
 lock_sock include/net/sock.h:1460 [inline]
 sctp_close+0xcd/0x9d0 net/sctp/socket.c:1497
 inet_release+0xed/0x1c0 net/ipv4/af_inet.c:425
 inet6_release+0x50/0x70 net/ipv6/af_inet6.c:432
 sock_release+0x8d/0x1e0 net/socket.c:597
 __sock_create+0x38b/0x870 net/socket.c:1226
 sock_create+0x7f/0xa0 net/socket.c:1237
 sctp_do_peeloff+0x1a2/0x440 net/sctp/socket.c:4879
 sctp_getsockopt_peeloff net/sctp/socket.c:4914 [inline]
 sctp_getsockopt+0x111a/0x67e0 net/sctp/socket.c:6628
 sock_common_getsockopt+0x95/0xd0 net/core/sock.c:2690
 SYSC_getsockopt net/socket.c:1817 [inline]
 SyS_getsockopt+0x240/0x380 net/socket.c:1799
 entry_SYSCALL_64_fastpath+0x1f/0xc2

This warning is caused by the lock held by sctp_getsockopt() is on one
socket, while the other lock that sctp_close() is getting later is on
the newly created (which failed) socket during peeloff operation.

This patch is to avoid this warning by use lock_sock with subclass
SINGLE_DEPTH_NESTING as Wang Cong and Marcelo's suggestion.

Reported-by: Dmitry Vyukov <dvyukov@google.com>
Suggested-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Suggested-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: Xin Long <lucien.xin@gmail.com>
Acked-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agonet/packet: remove unneeded declaraion of tpacket_snd().
Rosen, Rami [Sat, 10 Jun 2017 00:22:48 +0000 (03:22 +0300)]
net/packet: remove unneeded declaraion of tpacket_snd().

This patch removes unneeded forward declaration of tpacket_snd()
in net/packet/af_packet.c.

Signed-off-by: Rami Rosen <rami.rosen@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agobpf: Remove duplicate tcp_filter hook in ipv6
Chenbo Feng [Fri, 9 Jun 2017 19:17:37 +0000 (12:17 -0700)]
bpf: Remove duplicate tcp_filter hook in ipv6

There are two tcp_filter hooks in tcp_ipv6 ingress path currently.
One is at tcp_v6_rcv and another is in tcp_v6_do_rcv. It seems the
tcp_filter() call inside tcp_v6_do_rcv is redundent and some packet
will be filtered twice in this situation. This will cause trouble
when using eBPF filters to account traffic data.

Signed-off-by: Chenbo Feng <fengc@google.com>
Acked-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agobonding: warn user when 802.3ad speed is unknown
Nicolas Dichtel [Fri, 9 Jun 2017 15:58:08 +0000 (17:58 +0200)]
bonding: warn user when 802.3ad speed is unknown

Goal is to advertise the user when ethtool speeds and 802.3ad speeds are
desynchronized.
When this case happens, the kernel needs to be patched.

Suggested-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agonetns: fix error code when the nsid is already used
Nicolas Dichtel [Fri, 9 Jun 2017 12:41:57 +0000 (14:41 +0200)]
netns: fix error code when the nsid is already used

When the user tries to assign a specific nsid, idr_alloc() is called with
the range [nsid, nsid+1]. If this nsid is already used, idr_alloc() returns
ENOSPC (No space left on device). In our case, it's better to return
EEXIST to make it clear that the nsid is not available.

CC: Jamal Hadi Salim <jhs@mojatatu.com>
Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agonetns: define extack error msg for nsis cmds
Nicolas Dichtel [Fri, 9 Jun 2017 12:41:56 +0000 (14:41 +0200)]
netns: define extack error msg for nsis cmds

It helps the user to identify errors.

CC: Jamal Hadi Salim <jhs@mojatatu.com>
Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agocxgb4: fix memory leak in init_one()
Ganesh Goudar [Fri, 9 Jun 2017 13:56:24 +0000 (19:26 +0530)]
cxgb4: fix memory leak in init_one()

Free up mbox_log allocated for PF0 to PF3.

Fixes: 7829451c695e ("cxgb4: Add control net_device for configuring PCIe VF")
Signed-off-by: Ganesh Goudar <ganeshgr@chelsio.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agoqed: add qed_int_sb_init() stub function
Arnd Bergmann [Fri, 9 Jun 2017 10:37:35 +0000 (12:37 +0200)]
qed: add qed_int_sb_init() stub function

When CONFIG_QED_SRIOV is disabled, we get a build error:

drivers/net/ethernet/qlogic/qed/qed_int.c: In function 'qed_int_sb_init':
drivers/net/ethernet/qlogic/qed/qed_int.c:1499:4: error: implicit declaration of function 'qed_vf_set_sb_info'; did you mean 'qed_mcp_get_resc_info'? [-Werror=implicit-function-declaration]

All the other declarations have a 'static inline' stub as an alternative
here, so this adds one more for qed_int_sb_init.

Fixes: 50a207147fce ("qed: Hold a single array for SBs")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: Yuval Mintz <Yuval.Mintz@cavium.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agoMerge branch 'qed-Light-L2-updates'
David S. Miller [Fri, 9 Jun 2017 19:49:03 +0000 (15:49 -0400)]
Merge branch 'qed-Light-L2-updates'

Yuval Mintz says:

====================
qed*: Light L2 updates

This series does a major overhaul of the LL2 logic in qed.
The single biggest change done here is in #5 where we're changing
the API qed provides for LL2 [both internally in case of storage and
externally in case of RoCE] to become callback-based to allow cleaner
scalability in preperation to the future iWARP submission which would
aadd additional flavors of LL2. It's also the only patch in series
to modify !qed logic [qedr].

Patches prior to that mostly deal with refactoring LL2 code,
encapsulating varaious parameters into structure and re-ordering
of LL2 code. The latter patches add some small missing bits of LL2
ffunctionality.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agoqed: collect GSI port statistics
Mintz, Yuval [Fri, 9 Jun 2017 14:13:25 +0000 (17:13 +0300)]
qed: collect GSI port statistics

The LL2 statistics already have place holders for these, but haven't
populated them so far.

Signed-off-by: Yuval Mintz <Yuval.Mintz@cavium.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agoqed: Call rx_release_cb() when flushing LL2
Mintz, Yuval [Fri, 9 Jun 2017 14:13:24 +0000 (17:13 +0300)]
qed: Call rx_release_cb() when flushing LL2

Driver to inform the connection owner that the its buffers are being
released as part of a flush.

Signed-off-by: Yuval Mintz <Yuval.Mintz@cavium.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agoqed: No need for LL2 frags indication
Mintz, Yuval [Fri, 9 Jun 2017 14:13:23 +0000 (17:13 +0300)]
qed: No need for LL2 frags indication

This is a legacy leftover; There's no current flow where 'frags_mapped'
would be set.

Signed-off-by: Yuval Mintz <Yuval.Mintz@cavium.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agoqed*: LL2 callback operations
Michal Kalderon [Fri, 9 Jun 2017 14:13:22 +0000 (17:13 +0300)]
qed*: LL2 callback operations

LL2 today is interrupt driven - when tx/rx completion arrives [or any
other indication], qed needs to operate on the connection and pass
the information to the protocol-driver [or internal qed consumer].
Since we have several flavors of ll2 employeed by the driver,
each handler needs to do an if-else to determine the right functionality
to use based on the connection type.

In order to make things more scalable [given that we're going to add
additional types of ll2 flavors] move the infrastrucutre into using
a callback-based approach - the callbacks would be provided as part
of the connection's initialization parameters.

Signed-off-by: Michal Kalderon <Michal.Kalderon@cavium.com>
Signed-off-by: Yuval Mintz <Yuval.Mintz@cavium.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agoqed: LL2 code relocations
Mintz, Yuval [Fri, 9 Jun 2017 14:13:21 +0000 (17:13 +0300)]
qed: LL2 code relocations

Instead of having the OOO logic packetd, divide it with rest of code
according to establish/release flows.

Signed-off-by: Yuval Mintz <Yuval.Mintz@cavium.com>
Signed-off-by: David S. Miller <davem@davemloft.net>