Yuchung discovered tcp_is_cwnd_limited() was returning false in
slow start phase even if the application filled the socket write queue.

All congestion modules take into account tcp_is_cwnd_limited()
before increasing cwnd, so this behavior limits slow start from
probing the bandwidth at full speed.

The problem is that even if write queue is full (aka we are _not_
application limited), cwnd can be under utilized if TSO should auto
defer or TCP Small queues decided to hold packets.

So the in_flight can be kept to smaller value, and we can get to the
point tcp_is_cwnd_limited() returns false.

With TCP Small Queues and FQ/pacing, this issue is more visible.

We fix this by having tcp_cwnd_validate(), which is supposed to track
such things, take into account unsent_segs, the number of segs that we
are not sending at the moment due to TSO or TSQ, but intend to send
real soon. Then when we are cwnd-limited, remember this fact while we
are processing the window of ACKs that comes back.

For example, suppose we have a brand new connection with cwnd=10; we
are in slow start, and we send a flight of 9 packets. By the time we
have received ACKs for all 9 packets we want our cwnd to be 18.
We implement this by setting tp->lsnd_pending to 9, and
considering ourselves to be cwnd-limited while cwnd is less than
twice tp->lsnd_pending (2*9 -> 18).

This makes tcp_is_cwnd_limited() more understandable, by removing
the GSO/TSO kludge, that tried to work around the issue.

Note the in_flight parameter can be removed in a followup cleanup
patch.
Signed-off-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NNeal Cardwell <ncardwell@google.com>
Signed-off-by: NYuchung Cheng <ycheng@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This switches a few remaining capable(CAP_NET_ADMIN) to ns_capable so
that root in a user namespace may set tc rules inside that namespace.
Signed-off-by: NStéphane Graber <stgraber@ubuntu.com>
Acked-by: NSerge E. Hallyn <serge.hallyn@ubuntu.com>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Jamal Hadi Salim <jhs@mojatatu.com>
Cc: "David S. Miller" <davem@davemloft.net>
Acked-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Grygorii Strashko says:

====================
introduce devm_mdiobus_alloc/free and clean up davinci mdio

Introduce a resource managed devm_mdiobus_alloc[_size]()/devm_mdiobus_free()
to automatically clean up MDIO bus alocations made by MDIO drivers,
thus leading to simplified MDIO drivers code.

Clean up Davinci MDIO driver and use new devm API.

Changes in v3:
- added devm_mdiobus_alloc_size() and
  devm_mdiobus_alloc() converted to be just a simple wrapper now.

Changes in v2:
- minor comments taken into account
- additional patches added for cleaning up Davinci MDIO driver
====================

Acked-by: Santosh Shilimkar<santosh.shilimkar@ti.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Simplify IO memory mapping by using devm_ioremap_resource()
which will do all errors handling and reporting for us.
Acked-and-tested-by: NLad, Prabhakar <prabhakar.csengg@gmail.com>
Signed-off-by: NGrygorii Strashko <grygorii.strashko@ti.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The "default" pinctrl state is set by Drivers core now before
calling the driver's probe.
Hence, it's safe to drop pinctrl_pm_select_default_state() call
from Davinci mdio driver probe.

Cc: Florian Fainelli <f.fainelli@gmail.com>
Cc: Linus Walleij <linus.walleij@linaro.org>
Acked-and-tested-by: NLad, Prabhakar <prabhakar.csengg@gmail.com>
Signed-off-by: NGrygorii Strashko <grygorii.strashko@ti.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Use devm_* API for memory allocation and to get device's clock
to simplify driver's code.

Cc: Florian Fainelli <f.fainelli@gmail.com>
Cc: Sergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Acked-and-tested-by: NLad, Prabhakar <prabhakar.csengg@gmail.com>
Signed-off-by: NGrygorii Strashko <grygorii.strashko@ti.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Add a resource managed devm_mdiobus_alloc[_size]()/devm_mdiobus_free()
to automatically clean up MDIO bus alocations made by MDIO drivers,
thus leading to simplified MDIO drivers code.

Cc: Florian Fainelli <f.fainelli@gmail.com>
Cc: Sergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Acked-and-tested-by: NLad, Prabhakar <prabhakar.csengg@gmail.com>
Signed-off-by: NGrygorii Strashko <grygorii.strashko@ti.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This adds two new flags to quirks and thus removes the need to carry
revision in rhine_private. As a result, the init logic is simplified
a bit.

This also fixes a compiler warning in OF code on 64bit due to pointer
casting:

        drivers/net/ethernet/via/via-rhine.c: In function ‘rhine_init_one_platform’:
        drivers/net/ethernet/via/via-rhine.c:1132:13: warning: cast from pointer to integer of different size [-Wpointer-to-int-cast]
          revision = (u32)match->data;
                     ^

That code was added in commit 2d283862
("net: via-rhine: add OF bus binding").

Tested in platform configuration on a VIA WM8950 APC Rock board.
Reported-by: NJan Moskyto Matejka <mq@suse.cz>
Signed-off-by: NAlexey Charkov <alchark@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

We send packets using a copy-free mechanism (this is the Guest to Host transport
via VMBUS). While this is obviously optimal for large packets,
it may not be optimal for small packets. Hyper-V host supports
a second mechanism for sending packets that is "copy based". We implement that
mechanism in this patch.

In this version of the patch I have addressed a comment from David Miller.

With this patch (and all of the other offload and VRSS patches), we are now able
to almost saturate a 10G interface between Linux VMs on Hyper-V
on different hosts - close to  9 Gbps as measured via iperf.
Signed-off-by: NK. Y. Srinivasan <kys@microsoft.com>
Reviewed-by: NHaiyang Zhang <haiyangz@microsoft.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

mii_irq[] array is never initialized anywhere in the driver, thus mii_irq[]
will always equate to zero. So, for the case where the PHY does not have an
irq, we should use PHY_POLL for that situation.
Signed-off-by: NDinh Nguyen <dinguyen@altera.com>
Tested-by: NVince Bridgers <vbridger@altera.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This is a small supplement for commit e7428e95
("virtio-net: put virtio-net header inline with data"). TCP packages have
enough room to put virtio-net header in, but UDP packages do not. By
setting dev->needed_headroom for virtio-net device, UDP packages could have
enough room.

For UDP packages, sk_buff is alloced in fun __ip_append_data. The size is
"alloclen + hh_len + 15", and "hh_len = LL_RESERVED_SPACE(rt-dst.dev);".
The Macro is defined as follows:
#define LL_RESERVED_SPACE(dev) \
     ((((dev)->hard_header_len+(dev)->needed_headroom)\
     &~(HH_DATA_MOD - 1)) + HH_DATA_MOD)
By default, for UDP packages, after skb is allocated, only 16 bytes
reserved. And 2 bytes remained after mac header is set. That is not enough
to put virtio-net header in. If we set dev->needed_headroom to 12 or 10
(according to mergeable_rx_bufs is on or off ), more room can be reserved.
Then there is enough room for UDP packages to put the header in.

test result list as below:
guest and host: suse11sp3, netperf, intel 2.4GHz
+-------+---------+---------+---------+---------+
|       |   old             |   new             |
+-------+---------+---------+---------+---------+
| UDP   |  Gbit/s | pps     |  Gbit/s | pps     |
| 64    |  0.57   | 692232  |  0.61   | 742420  |
| 256   |  1.60   | 686860  |  1.71   | 733331  |
| 512   |  2.92   | 674576  |  3.07   | 710446  |
| 1024  |  4.99   | 598977  |  5.17   | 620821  |
| 1460  |  5.68   | 483757  |  7.16   | 610519  |
| 4096  |  6.98   | 637468  |  7.21   | 658471  |
+-------+---------+---------+---------+---------+
Signed-off-by: NZhang Jie <zhangjie14@huawei.com>
Acked-by: NRusty Russell <rusty@rustcorp.com.au>
Acked-by: NJason Wang <jasowang@redhat.com>
Acked-by: NMichael S. Tsirkin <mst@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Commit 1bb8dce5 ("tipc: fix memory
leak during module removal") introduced a memory leak issue: when
name table is stopped, it's forgotten that publication instances are
freed properly. Additionally the useless "continue" statement in
tipc_nametbl_stop() is removed as well.
Reported-by: NJason <huzhijiang@gmail.com>
Signed-off-by: NYing Xue <ying.xue@windriver.com>
Acked-by: NErik Hugne <erik.hugne@ericsson.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This replaces 6 identical code snippets with a call to a new
static inline function.
Signed-off-by: NLorenzo Colitti <lorenzo@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Use the helper function to retrieve the driver private context instead of
using (void *)(ds + 1).
Signed-off-by: NFlorian Fainelli <f.fainelli@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

DSA drivers have a trick which consists in allocating "priv_size" more
bytes to account for the DSA driver private context. Add a helper
function to access that private context instead of open-coding it in
drivers with (void *)(ds + 1).
Signed-off-by: NFlorian Fainelli <f.fainelli@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Jeff Kirsher says:

====================
Intel Wired LAN Driver Updates

This series contains updates to i40e and i40evf.

Jacob provides a i40e patch to remove ptp_tx_work() because it conflicts
with the interrupt-based handler and will cause spurious warning messages
to be displayed in the kernel log.  Since the hardware can properly trigger
an interrupt, we do not need to poll for an available timestamp in a work
queue any more.

Greg provides two patches for i40e, first to retain the MAC filter list
when changing a port VLAN.  Second fixes an issue where the function to
set the VF MAC address was not taking the port VLAN filter into account
when setting/clearing/resetting the VF's host administered MAC address.

Mitch provides three patches for i40evf, first adds support for getting
and setting RSS hashing options, RSS LUT entries, and getting the number
of device channels from ethtool.  Fixes a panic that woudl occur in the VF
if the PF driver failed or was removed from the host kernel.

Shannon adds driver version string to the driver version command so that
can be passed down through the firmware to low-level NC-SI functions.

Neerav provides several patches, first redistributes queue vectors after
DCB reconfiguration because the total number of queues that are configured
for a VSI may change based on a change in the total number of TCs enabled
as a result of a change in the DCB configuration based on DCBX.  Second,
allows the driver probe to continue when querying DCB config fails.
Lastly, provides changes to retrieve and store missing link config
information that was not being cached in the struct i40e_link_status.

Kevin provides a couple of patches, first to change a int variable type
to u16 to avoid a typecheck failure.  Second to update a function
parameters to remove an argument which is no longer used.
====================
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Commit a89778d8 ("tipc: add support
for link state subscriptions") introduced below possible deadlock
scenario:

       CPU0                          CPU1
T0:   tipc_publish()                 link_timeout()
T1:   tipc_nametbl_publish()         [grab node lock]*
T2:   [grab nametbl write lock]*     link_state_event()
T3:   named_cluster_distribute()     link_activate()
T4:   [grab node lock]*              tipc_node_link_up()
T5:                                  tipc_nametbl_publish()
T6:                                  [grab nametble write lock]*

The opposite order of holding nametbl write lock and node lock on
above two different paths may result in a deadlock. If we move the
the delivery of named messages via link out of name nametbl lock,
the reverse order of holding locks will be eliminated, as a result,
the deadlock will be killed as well.
Signed-off-by: NYing Xue <ying.xue@windriver.com>
Reviewed-by: NErik Hugne <erik.hugne@ericsson.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Commit 78acb1f9 ("tipc: add
ioctl to fetch link names") introduced a buffer overflow bug where
specially crafted ioctl requests could cause out-of-bounds indexing
of the node->links array. This was caused by an incorrect check vs
MAX_BEARERS, and the static code checker complaint is:
net/tipc/node.c:459 tipc_node_get_linkname() error: buffer overflow 'node->links' 2 <= 2
Signed-off-by: NErik Hugne <erik.hugne@ericsson.com>
Reported-by: NDan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

PTR_RET is deprecated. Do not recommend its usage anymore.
Use PTR_ERR_OR_ZERO instead.
Signed-off-by: NFrank Blaschka <frank.blaschka@de.ibm.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

PTR_RET is deprecated. Do not recommend its usage anymore.
Use PTR_ERR_OR_ZERO instead.
Signed-off-by: NFrank Blaschka <frank.blaschka@de.ibm.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The existing options for priority queueing within OSA devices were based on
the now partially defunct TOS field. This patch adds two new options as
follows:
   - prio_queueing_skb: bases priority queueing on skb-priority
   - prio_queueing_vlan: uses the priority code point in the 802.1Q header
Signed-off-by: NStefan Raspl <raspl@linux.vnet.ibm.com>
Signed-off-by: NFrank Blaschka <frank.blaschka@de.ibm.com>
Reviewed-by: NUrsula Braun <ursula.braun@de.ibm.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Make the current priority queueing logic apply to IPv6 traffic.
Signed-off-by: NStefan Raspl <raspl@linux.vnet.ibm.com>
Signed-off-by: NFrank Blaschka <frank.blaschka@de.ibm.com>
Reviewed-by: NUrsula Braun <ursula.braun@de.ibm.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

PTR_RET is deprecated. Do not recommend its usage anymore.
Use PTR_ERR_OR_ZERO instead.
Signed-off-by: NDuan Jiong <duanj.fnst@cn.fujitsu.com>
Signed-off-by: NMartin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: NFrank Blaschka <frank.blaschka@de.ibm.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

PTR_RET is deprecated. Do not recommend its usage anymore.
Use PTR_ERR_OR_ZERO instead.
Signed-off-by: NDuan Jiong <duanj.fnst@cn.fujitsu.com>
Signed-off-by: NMartin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: NFrank Blaschka <frank.blaschka@de.ibm.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The default queue for queueing modes prio_queueing_prec and
prio_queueing_tos is supposed to be QETH_DEFAULT_QUEUE. However, neither
mode will reset the default default_out_queue value when enabled, leaving
a previously set value activate as the default.
E.g. enabling one of the fixed queueing modes, e.g. no_prio_queueing:0,
and successively switching to any of the prio_queueing_* modes will
leave default_out_queue at the previous (wrong) value 0.
Signed-off-by: NStefan Raspl <raspl@linux.vnet.ibm.com>
Signed-off-by: NFrank Blaschka <frank.blaschka@de.ibm.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

PTP Hardware Clock drivers specify a maximum frequency adjustment that
their clocks can accommodate. Normally, user space programs will want to
respect the advertised limits. However, no kernel or driver code checks
that the dialed frequency offset is within the bounds, and out of range
values can lead to surprising results.

This patch fixes the issue by rejecting bad values.
Signed-off-by: NRichard Cochran <richardcochran@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NHisao Tanabe <xtanabe@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The commit 3de0b592 ("ethtool: Support for configurable RSS hash
key") introduced a new function ethtool_copy_validate_indir() with
full iteration of the loop to validate the ring indices, which could
be an overkill. To minimize the impact, we ought to exit the loop as
soon as the invalid index occurs for the very first time. The
remaining loop simply doesn't serve any more purpose.
Signed-off-by: NJean Sacren <sakiwit@gmail.com>
Cc: Venkata Duvvuru <VenkatKumar.Duvvuru@Emulex.Com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Bump i40e to 0.3.46 and i40evf to 0.9.23.

Change-ID: Ia604ae6d513d9aaa8bfdac79665d9a3a72507df7
Signed-off-by: NCatherine Sullivan <catherine.sullivan@intel.com>
Tested-by: NKavindya Deegala <kavindya.s.deegala@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

Some information returned via "Get Link Status" command was not
being cached in the struct i40e_link_status.
Add this so the driver can utilize this information as needed.

Change-ID: If084a0ae2a63b16b97572196b993742af2c67772
Signed-off-by: NNeerav Parikh <neerav.parikh@intel.com>
Acked-by: NGreg Rose <gregory.v.rose@intel.com>
Tested-by: NKavindya Deegala <kavindya.s.deegala@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

Change the formal parameters to remove argument which is no longer used.

Change-ID: I493b56eba74a1f396fa1b7d24c0a4acbe536b5bf
Signed-off-by: NKevin Scott <kevin.c.scott@intel.com>
Acked-by: NShannon Nelson <shannon.nelson@intel.com>
Tested-by: NKavindya Deegala <kavindya.s.deegala@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

Ethtool -k/-K can handle this so we do not need it in debugfs.

Change-ID: I9df692f10a60b71805f8f48d3b87c7da3820b2aa
Signed-off-by: NAnjali Singhai Jain <anjali.singhai@intel.com>
Signed-off-by: NCatherine Sullivan <catherine.sullivan@intel.com>
Tested-by: NKavindya Deegala <kavindya.s.deegala@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

Although NTUPLE feature control is disabled by default, do not
allow user to be able to change it.

Change-ID: I4c3f95fb6dbc56b1b6cc47116ea1f72de02bc99c
Signed-off-by: NAnjali Singhai Jain <anjali.singhai@intel.com>
Signed-off-by: NCatherine Sullivan <catherine.sullivan@intel.com>
Tested-by: NKavindya Deegala <kavindya.s.deegala@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

Change the variable type to avoid compiler warning about int to u16
possible data truncation.

Change-ID: I5eb3b578c86513c9625ca32d2f0b57cc01d7dc98
Signed-off-by: NKevin Scott <kevin.c.scott@intel.com>
Acked-by: NGreg Rose <gregory.v.rose@intel.com>
Acked-by: NShannon Nelson <shannon.nelson@intel.com>
Tested-by: NKavindya Deegala <kavindya.s.deegala@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

If the PF driver fails or is removed from the host, the VF driver will
fill up its log with this message.

Change-ID: I67045f987f7c0d444d21ded403adc509343cdb8f
Signed-off-by: NMitch Williams <mitch.a.williams@intel.com>
Signed-off-by: NCatherine Sullivan <catherine.sullivan@intel.com>
Tested-by: NSibai Li <sibai.li@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

Fix a panic that would occur in the VF if the PF driver failed or was
removed from the host kernel. In this case, the VF driver calls
i40evf_close(), but this function does nothing because the driver is in
the resetting state. Because of this, the driver doesn't free its irqs
and causes a kernel panic when it tries to disable MSI-X.

Change-ID: If95644a89e554b4d7be0dca1b6add26f63047129
Signed-off-by: NMitch Williams <mitch.a.williams@intel.com>
Signed-off-by: NCatherine Sullivan <catherine.sullivan@intel.com>
Tested-by: NSibai Li <sibai.li@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

In case of any AQ command to query port's DCB configuration fails
during driver's probe time; the probe fails and returns an error.

This patch prevents this issue by continuing the driver probe even
when an error is returned.

Also, added an error message to dump the AQ error status to show what
error caused the failure to get the DCB configuration from firmware.

Change-ID: Ifd5663512588bca684069bb7d4fb586dd72221af
Signed-off-by: NNeerav Parikh <Neerav.Parikh@intel.com>
Signed-off-by: NCatherine Sullivan <catherine.sullivan@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

The queue vectors are originally distributed among the total
number of queues that are configured for a given VSI. The number
of queues that are configured for a VSI depends on the number of
TCs that are enabled for the VSI at VSI setup/creation time.

But, the total number of queues that are configured for a VSI may
change based on change in the total number of TCs enabled as a
result of change in DCB configuration based on DCBX.

Hence, it is required for the queue vectors to be redistributed
when the total queues that are configured for a given VSI changes.

Without redistributing the vectors the queues that are configured
for the newly enabled TCs may not be able to do any Tx/Rx.

Change-ID: I4e780903db019e6199c7ce43627cef22c916acfe
Signed-off-by: NNeerav Parikh <Neerav.Parikh@intel.com>
Signed-off-by: NCatherine Sullivan <catherine.sullivan@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

The driver version string was added to this struct to be passed
down through the firmware to low-level NC-SI functions. We tell
the firmware about the length of the ASCII string not counting
any terminating null.

Change-ID: I09ac98ff9b869e8661c55fc6a5c98808fc280c91
Signed-off-by: NShannon Nelson <shannon.nelson@intel.com>
Acked-by: NAnjali Singhai <anjali.singhai@intel.com>
Tested-by: NKavindya Deegala <kavindya.s.deegala@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

The function to set the VF MAC address was not taking the port VLAN
filter into account when setting/clearing/resetting the VF's host
administered MAC address.  Be sure to use the port VLAN for VF MAC
filtering configurations.

Change-ID: I12595331981c79529738f041dcdbdb667ef8f5e6
Signed-off-by: NGreg Rose <gregory.v.rose@intel.com>
Signed-off-by: NCatherine Sullivan <catherine.sullivan@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>