- Store the length of the skb buffer mapped along with the handle and use it
   while unmapping the buffer.
Signed-off-by: NRasesh Mody <rmody@brocade.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Change details:
 - When bnad_setup_tx() returns NULL, the error is NOT returned to the caller.
   The caller will incorrectly assume success. So Return ENOMEM when bna_tx_create()
   fails.
 - If bnad_tx_msix_register() fails, call bna_tx_destroy() to free tx & to NULL
   the bnad reference to tcb.
Signed-off-by: NRasesh Mody <rmody@brocade.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Currently we already check to see whether the BNAD_TXQ_TX_STARTED cleared.
But if the tcb structure which contains this flag is also already freed by that
time, we would dereference the NULL pointer. This patch is to check tcb for NULL
pointer, before dereferencing it.
Signed-off-by: NRasesh Mody <rmody@brocade.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Valid bit check for completion needs read fence, so that it does not get
reordered with other loads.
Signed-off-by: NRasesh Mody <rmody@brocade.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Change Details:
 - Prefetch header in GRO path. This reduces napi_frags_skb time from 9% to 5%.
 - Changed the configurable limit of RxQ depth to 16384 (was 2048).
 - bnad_rx_unmap_q elements are cachealigned.
Signed-off-by: NRasesh Mody <rmody@brocade.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The CT2 HW supports multi-buffer Rx. This patch provides the necessary changes
for bnad to use multi-buffer Rx feature. For BNAD, multi-buffer Rx is by
default enabled when MTU is > 4096. For >4096 MTU, q0 data/large buffers are of
2048 size. As the resource requirements of multi-buffer Rx are different new Rx
needs to be created to use this feature. ASIC posts multiple completions if
frame exceeds buffer size. The last completion is marked with EOP flag.
 - Separate HQ and DQ enums for resource allocations and configurations.
 - rx_config and rxq structure changes to pass the correct info from bnad.
 - DQ depth need not be same as HQ depth. So CQ depth is adjusted accordingly.
 - Rx CFG frame size is taken from configured MTU.
 - Rx q0 buffer size is configured from bnad s rx_config when multi-buffer is
   enabled.
 - Poll for entire frame completion.
 - Once EOP completion is received gather the number of vectors used by the
   frame to submit it to the stack.
 - Changed MTU to frame size wherever necessary.
Signed-off-by: NRasesh Mody <rmody@brocade.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Change Details:
 - Added bna_rx_ucast_listset() for synchronous ucast listadd operation.
 - Clear mac->handle before adding it to free_q.
 - bnad_set_rx_mode() rewritten. bnad_set_rx_mode() adds the MACs in uc_list
   to UCAM. If it exceeds the max supported, DEFAULT mode is turned on. If
   MCAM limit is exceeded, ALLMULTI mode is turned on.
 - Clear CF flags, check for the new mode and reprogram the Rx approach.
 - Added bnad_set_rx_ucast_fltr() and bnad_set_rx_mcast_fltr().
 - Check for IFF_PROMISC to set the correct mode.
Signed-off-by: NRasesh Mody <rmody@brocade.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Change Details:
 - bna_rx_mcast_listset() API first looks at free_q only and not at other
   pending Qs rendering it non-deterministic of giving an upper limit.
   Modify bna_rx_mcast_listset() implementation to not use only half of the
   limit.
 - Allocate and initialize queue for deleting
 - Segregate the adding and deleting process by using separate queues.
 - The filter framework in bna does not let adding addresses to its max capacity
   due to asynchronous operations involved.
   Provide a synchronous option to set a given list.
Signed-off-by: NRasesh Mody <rmody@brocade.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Add APIs to set and get IOC currnet fw state and alt IOC fw state
 - bfa_ioc_ct_set_cur_ioc_fwstate()
 - bfa_ioc_ct_get_cur_ioc_fwstate()
 - bfa_ioc_ct_set_alt_ioc_fwstate()
 - bfa_ioc_ct_get_alt_ioc_fwstate()
Signed-off-by: NRasesh Mody <rmody@brocade.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

- Invoke skb_tx_timestamp() API just before invoking txq_doorbell()
 - Add ethtool (-T) support
Signed-off-by: NRasesh Mody <rmody@brocade.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Reported-by: NStephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

For OSA devices that support the QUERY_CARD_INFO command, supply
accurate data based on the card type, port mode and link speed
via the 'ethtool' interface.
Signed-off-by: NEugene Crosser <Eugene.Crosser@ru.ibm.com>
Signed-off-by: NFrank Blaschka <frank.blaschka@de.ibm.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

state checking in conn_action_txdone() is inconsistent.
This patch makes it consistent and issues a trace message
if an unexpected state is detected for the netiucv device.
Signed-off-by: NUrsula Braun <ursula.braun@de.ibm.com>
Signed-off-by: NFrank Blaschka <frank.blaschka@de.ibm.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Valgrind found that extracted labels that are passed from the lexer
weren't freed upon exit. Therefore, add a small helper function that
walks label tables and frees them. Since also NULL can be passed to
free(3), we do not need to take care of that here. While at it, fix
up a spacing error in bpf_set_curr_label().
Signed-off-by: NDaniel Borkmann <dborkman@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

We must not leave the socket intact in bpf_runnable(). The socket
is used to test if the filter code is being accepted by the kernel
or not. So right after we do the setsockopt(2), we need to close
it again.
Signed-off-by: NDaniel Borkmann <dborkman@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Manish Chopra says:

====================
qlcnic: Refactoring and enhancements

This patch series includes follwing changes

* Refactor DCBX code. Do not allow DCBX operations for VFs
* Issue INIT_NIC mailbox command only once
* Refactor initialize nic code path
* Allow configuration for single TX/RX queue
* VLAN enhancement for 84xx adapters
* Support for 16 virtual NIC functions for 84XX series adapters

Please apply to net-next
====================
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NManish Chopra <manish.chopra@qlogic.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Extend virtual NIC functions from 8 to 16 for 84xx adapter.
Signed-off-by: NJitendra Kalsaria <jitendra.kalsaria@qlogic.com>
Signed-off-by: NManish Chopra <manish.chopra@qlogic.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

o Support multiple VLANs on 84xx VF devices
Signed-off-by: NManish Chopra <manish.chopra@qlogic.com>
Signed-off-by: NSucheta Chakraborty <sucheta.chakraborty@qlogic.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

o Allow user to set sigle Tx/Rx queue in MSI-x mode,
  for ALL supported adapters.
Signed-off-by: NHimanshu Madhani <himanshu.madhani@qlogic.com>
Signed-off-by: NManish Chopra <manish.chopra@qlogic.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

o Change function name from qlcnic_83xx_register_nic_idc_func to
  qlcnic_83xx_initialize_nic
Signed-off-by: NSucheta Chakraborty <sucheta.chakraborty@qlogic.com>
Signed-off-by: NManish Chopra <manish.chopra@qlogic.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

o DCB AEN registration was reissuing INIT_NIC command. Instead, club
  all options of INIT NIC command and issue this command only once.
Signed-off-by: NSucheta Chakraborty <sucheta.chakraborty@qlogic.com>
Signed-off-by: NManish Chopra <manish.chopra@qlogic.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

o These operations will be supported only through PFs (SR-IOV and non-SR-IOV).
Signed-off-by: NSucheta Chakraborty <sucheta.chakraborty@qlogic.com>
Signed-off-by: NManish Chopra <manish.chopra@qlogic.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Ben Hutchings says:

====================
Miscellaneous changes for 3.14:

1. Add more information to some WARN messages.
2. Refactor pushing of RSS configuration, from Andrew Rybchenko.
3. Refactor handling of automatic (device address list) vs manual (RX
NFC) MAC filters.
4. Implement clearing of manual RX filters on EF10 when ntuple offload
is disabled.
5. Remove definitions that are unused since the RX buffer allocation
changes, from Andrew Rybchenko.
6. Improve naming of some statistics, from Shradha Shah.
7. Add statistics for PTP support code.
8. Fix insertion of RX drop filters on EF10.
====================
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Tom Herbert says:

====================
net: Add rxhash utility hash functions

v3:

There's really nothing specific about rxhash that constrains
it to be a value just for these receive path. Drop the 'rx'
part in utility functions, including skb_get_rxhash. In subsequent
patches, we can change the rxhash and l4_rxhash names also, as
well as abstracting out the interface to the hash.

Added comments about hash types per feedback.

In this version I'm omitting the changes to drivers to make the
patch set manageable. Will add those changes in followup pathes.

-----
This patch series introduce skb_set_rxhash and skb_clear_rxhash
which are called to set the rxhash (from network drivers) and
to clear the rxhash. This API should be used instead of updating
fields in the skbuff directly.
====================
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Adds skb_copy_hash to copy rxhash and l4_rxhash from one skb to another.
Signed-off-by: NTom Herbert <therbert@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The function skb_set_rxash was added for drivers to call to set
the rxhash in an skb. The type of hash is also specified as
a parameter (L2, L3, L4, or unknown type).
Signed-off-by: NTom Herbert <therbert@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

In several places 'skb->rxhash = 0' is being done to clear the
rxhash value in an skb.  This does not clear l4_rxhash which could
still be set so that the rxhash wouldn't be recalculated on subsequent
call to skb_get_rxhash.  This patch adds an explict function to clear
all the rxhash related information in the skb properly.

skb_clear_hash_if_not_l4 clears the rxhash only if it is not marked as
l4_rxhash.

Fixed up places where 'skb->rxhash = 0' was being called.
Signed-off-by: NTom Herbert <therbert@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Changing name of function as part of making the hash in skbuff to be
generic property, not just for receive path.
Signed-off-by: NTom Herbert <therbert@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The callback function of call_rcu() just calls a kfree(), so we
can use kfree_rcu() instead of call_rcu() + callback function.
Signed-off-by: NWei Yongjun <yongjun_wei@trendmicro.com.cn>
Acked-by: NArvid Brodin <arvid.brodin@alten.se>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Scott Feldman says:

====================
bonding: add some more netlink attributes

The following series implements five more bonding netlink attributes:

	primary
	primary_reselect
	fail_over_mac
	xmit_hash_policy
	resend_igmp

Tested with modified iproute2 to verify attributes can be set at bond creation
time or set later.  Verified sysfs interface to attributes continues to work.
====================
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Add IFLA_BOND_RESEND_IGMP to allow get/set of bonding parameter
resend_igmp via netlink.
Signed-off-by: NScott Feldman <sfeldma@cumulusnetworks.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Add IFLA_BOND_XMIT_HASH_POLICY to allow get/set of bonding parameter
xmit_hash_policy via netlink.
Signed-off-by: NScott Feldman <sfeldma@cumulusnetworks.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Add IFLA_BOND_FAIL_OVER_MAC to allow get/set of bonding parameter
fail_over_mac via netlink.
Signed-off-by: NScott Feldman <sfeldma@cumulusnetworks.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Add IFLA_BOND_PRIMARY_SELECT to allow get/set of bonding parameter
primary_select via netlink.
Signed-off-by: NScott Feldman <sfeldma@cumulusnetworks.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Add IFLA_BOND_PRIMARY to allow get/set of bonding parameter
primary via netlink.
Signed-off-by: NScott Feldman <sfeldma@cumulusnetworks.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch brings NUMA support and automatic fallback to vmalloc()
in case kmalloc() failed to allocate FQ hash table.

NUMA support depends on XPS being setup for the device before
qdisc allocation. After a XPS change, it might be worth creating
qdisc hierarchy again.
Signed-off-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NJiri Pirko <jiri@resnulli.us>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Compile tested only.
Signed-off-by: NStephen Hemminger <stephen@networkplumber.org>
Acked-by: NRasesh Mody <rmody@brocade.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

While investigating performance problems on small RPC workloads,
I noticed linux TCP stack was always splitting the last TSO skb
into two parts (skbs). One being a multiple of MSS, and a small one
with the Push flag. This split is done even if TCP_NODELAY is set,
or if no small packet is in flight.

Example with request/response of 4K/4K

IP A > B: . ack 68432 win 2783 <nop,nop,timestamp 6524593 6525001>
IP A > B: . 65537:68433(2896) ack 69632 win 2783 <nop,nop,timestamp 6524593 6525001>
IP A > B: P 68433:69633(1200) ack 69632 win 2783 <nop,nop,timestamp 6524593 6525001>
IP B > A: . ack 68433 win 2768 <nop,nop,timestamp 6525001 6524593>
IP B > A: . 69632:72528(2896) ack 69633 win 2768 <nop,nop,timestamp 6525001 6524593>
IP B > A: P 72528:73728(1200) ack 69633 win 2768 <nop,nop,timestamp 6525001 6524593>
IP A > B: . ack 72528 win 2783 <nop,nop,timestamp 6524593 6525001>
IP A > B: . 69633:72529(2896) ack 73728 win 2783 <nop,nop,timestamp 6524593 6525001>
IP A > B: P 72529:73729(1200) ack 73728 win 2783 <nop,nop,timestamp 6524593 6525001>

We can avoid this split by including the Nagle tests at the right place.

Note : If some NIC had trouble sending TSO packets with a partial
last segment, we would have hit the problem in GRO/forwarding workload already.

tcp_minshall_update() is moved to tcp_output.c and is updated as we might
feed a TSO packet with a partial last segment.

This patch tremendously improves performance, as the traffic now looks
like :

IP A > B: . ack 98304 win 2783 <nop,nop,timestamp 6834277 6834685>
IP A > B: P 94209:98305(4096) ack 98304 win 2783 <nop,nop,timestamp 6834277 6834685>
IP B > A: . ack 98305 win 2768 <nop,nop,timestamp 6834686 6834277>
IP B > A: P 98304:102400(4096) ack 98305 win 2768 <nop,nop,timestamp 6834686 6834277>
IP A > B: . ack 102400 win 2783 <nop,nop,timestamp 6834279 6834686>
IP A > B: P 98305:102401(4096) ack 102400 win 2783 <nop,nop,timestamp 6834279 6834686>
IP B > A: . ack 102401 win 2768 <nop,nop,timestamp 6834687 6834279>
IP B > A: P 102400:106496(4096) ack 102401 win 2768 <nop,nop,timestamp 6834687 6834279>
IP A > B: . ack 106496 win 2783 <nop,nop,timestamp 6834280 6834687>
IP A > B: P 102401:106497(4096) ack 106496 win 2783 <nop,nop,timestamp 6834280 6834687>
IP B > A: . ack 106497 win 2768 <nop,nop,timestamp 6834688 6834280>
IP B > A: P 106496:110592(4096) ack 106497 win 2768 <nop,nop,timestamp 6834688 6834280>

Before :

lpq83:~# nstat >/dev/null;perf stat ./super_netperf 200 -t TCP_RR -H lpq84 -l 20 -- -r 4K,4K
280774

 Performance counter stats for './super_netperf 200 -t TCP_RR -H lpq84 -l 20 -- -r 4K,4K':

     205719.049006 task-clock                #    9.278 CPUs utilized
         8,449,968 context-switches          #    0.041 M/sec
         1,935,997 CPU-migrations            #    0.009 M/sec
           160,541 page-faults               #    0.780 K/sec
   548,478,722,290 cycles                    #    2.666 GHz                     [83.20%]
   455,240,670,857 stalled-cycles-frontend   #   83.00% frontend cycles idle    [83.48%]
   272,881,454,275 stalled-cycles-backend    #   49.75% backend  cycles idle    [66.73%]
   166,091,460,030 instructions              #    0.30  insns per cycle
                                             #    2.74  stalled cycles per insn [83.39%]
    29,150,229,399 branches                  #  141.699 M/sec                   [83.30%]
     1,943,814,026 branch-misses             #    6.67% of all branches         [83.32%]

      22.173517844 seconds time elapsed

lpq83:~# nstat | egrep "IpOutRequests|IpExtOutOctets"
IpOutRequests                   16851063           0.0
IpExtOutOctets                  23878580777        0.0

After patch :

lpq83:~# nstat >/dev/null;perf stat ./super_netperf 200 -t TCP_RR -H lpq84 -l 20 -- -r 4K,4K
280877

 Performance counter stats for './super_netperf 200 -t TCP_RR -H lpq84 -l 20 -- -r 4K,4K':

     107496.071918 task-clock                #    4.847 CPUs utilized
         5,635,458 context-switches          #    0.052 M/sec
         1,374,707 CPU-migrations            #    0.013 M/sec
           160,920 page-faults               #    0.001 M/sec
   281,500,010,924 cycles                    #    2.619 GHz                     [83.28%]
   228,865,069,307 stalled-cycles-frontend   #   81.30% frontend cycles idle    [83.38%]
   142,462,742,658 stalled-cycles-backend    #   50.61% backend  cycles idle    [66.81%]
    95,227,712,566 instructions              #    0.34  insns per cycle
                                             #    2.40  stalled cycles per insn [83.43%]
    16,209,868,171 branches                  #  150.795 M/sec                   [83.20%]
       874,252,952 branch-misses             #    5.39% of all branches         [83.37%]

      22.175821286 seconds time elapsed

lpq83:~# nstat | egrep "IpOutRequests|IpExtOutOctets"
IpOutRequests                   11239428           0.0
IpExtOutOctets                  23595191035        0.0

Indeed, the occupancy of tx skbs (IpExtOutOctets/IpOutRequests) is higher :
2099 instead of 1417, thus helping GRO to be more efficient when using FQ packet
scheduler.

Many thanks to Neal for review and ideas.
Signed-off-by: NEric Dumazet <edumazet@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Nandita Dukkipati <nanditad@google.com>
Cc: Van Jacobson <vanj@google.com>
Acked-by: NNeal Cardwell <ncardwell@google.com>
Tested-by: NNeal Cardwell <ncardwell@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>