These changes address comments by Jakub Kicinski on
commit 38b7e7f8 ("ixgbe: Do not allow LRO or MTU change with XDP").

Change the MTU check with XDP to allow any supported value and only
reject those outside of the range as opposed to rejecting any change
when XDP is active. In situations where MTU size is not supported,
return -EINVAL instead of -EPERM.

Add checks when enabling SRIOV, DCB, or adding L2FW offloaded device
as they are not supported with XDP.

CC: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Acked-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

ixgbe_fcoe_ddp_setup(), ixgbe_setup_fcoe_ddp_resources() and
ixgbe_sw_init() are never called in atomic context.
They call kmalloc(), dma_pool_alloc() and kzalloc() with GFP_ATOMIC,
which is not necessary.
GFP_ATOMIC can be replaced with GFP_KERNEL.

This is found by a static analysis tool named DCNS written by myself.
Signed-off-by: Jia-Ju Bai <baijiaju1990@gmail.com>
Acked-by: Sebastian Basierski <sebastianx.basierski@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

igb_integrated_phy_loopback() is never called in atomic context.
It calls mdelay() to busily wait, which is not necessary.
mdelay() can be replaced with msleep().

This is found by a static analysis tool named DCNS written by myself.
Signed-off-by: Jia-Ju Bai <baijiaju1990@gmail.com>
Tested-by: Aaron Brown <aaron.f.brown@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

igb_sw_init() is never called in atomic context.
It calls kzalloc() and kcalloc() with GFP_ATOMIC, which is not necessary.
GFP_ATOMIC can be replaced with GFP_KERNEL.

This is found by a static analysis tool named DCNS written by myself.
Signed-off-by: Jia-Ju Bai <baijiaju1990@gmail.com>
Tested-by: Aaron Brown <aaron.f.brown@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

On i210, Launchtime (TxTime) requires the usage of an "Advanced
Transmit Context Descriptor" for retrieving the timestamp of a packet.

The igb driver correctly builds such descriptor on the segmentation
flow (i.e. igb_tso()) or on the checksum one (i.e. igb_tx_csum()), but the
feature is broken for AF_PACKET if the IGB_TX_FLAGS_VLAN is not set,
which happens due to an early return on igb_tx_csum().

This flag is only set by the kernel when a VLAN interface is used,
thus we can't just rely on it. Here we are fixing this issue by checking
if launchtime is enabled for the current tx_ring before performing the
early return.
Signed-off-by: Jesus Sanchez-Palencia <jesus.sanchez-palencia@intel.com>
Tested-by: Aaron Brown <aaron.f.brown@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

In 'e1000_set_ringparam()', the tx_ring and rx_ring are updated with new value
and the old tx/rx rings are freed only when the device is up. There are resource
leaks on old tx/rx rings when the device is not up. This bug is reported by COD,
a tool for testing kernel module binaries I am building.

This patch fixes the bug by always calling 'kfree()' on old tx/rx rings in
'e1000_set_ringparam()'.
Signed-off-by: Bo Chen <chenbo@pdx.edu>
Reviewed-by: Alexander Duyck <alexander.h.duyck@intel.com>
Tested-by: Aaron Brown <aaron.f.brown@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

When the device is not up, the call to 'e1000_up()' from the error handling path
of 'e1000_set_ringparam()' causes a kernel oops with a null-pointer
dereference. The null-pointer dereference is triggered in function
'e1000_alloc_rx_buffers()' at line 'buffer_info = &rx_ring->buffer_info[i]'.

This bug was reported by COD, a tool for testing kernel module binaries I am
building. This bug was also detected by KFI from Dr. Kai Cong.

This patch fixes the bug by checking on 'netif_running()' before calling
'e1000_up()' in 'e1000_set_ringparam()'.
Signed-off-by: Bo Chen <chenbo@pdx.edu>
Acked-by: Alexander Duyck <alexander.h.duyck@intel.com>
Tested-by: Aaron Brown <aaron.f.brown@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

Use dma_zalloc_coherent instead of dma_alloc_coherent
followed by memset 0.
Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

Daniel Borkmann says:

====================
pull-request: bpf 2018-08-24

The following pull-request contains BPF updates for your *net* tree.

The main changes are:

1) Fix BPF sockmap and tls where we get a hang in do_tcp_sendpages()
   when sndbuf is full due to missing calls into underlying socket's
   sk_write_space(), from John.

2) Two BPF sockmap fixes to reject invalid parameters on map creation
   and to fix a map element miscount on allocation failure. Another fix
   for BPF hash tables to use per hash table salt for jhash(), from Daniel.

3) Fix for bpftool's command line parsing in order to terminate on bad
   arguments instead of keeping looping in some border cases, from Quentin.

4) Fix error value of xdp_umem_assign_dev() in order to comply with
   expected bind ops error codes, from Prashant.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

Jeff Kirsher says:

====================
Intel Wired LAN Driver Fixes 2018-08-23

This series contains bug fixes to the ice driver.

Anirudh provides several fixes, starting with static analysis fixes by
replacing a bitwise-and with a constant value and replace "magic"
numbers with defines.  Fixed the control queue processing by removing
unnecessary read/writes to registers, as well as getting a accurate
value for "pending".  Added additional checks to avoid NULL pointer
dereferences.  Fixed up code formatting issues, by cleaning up code
comments and coding style.

Bruce cleans up a duplicate check for owner, within the same function.
Also cleans up interrupt causes that are not handled or applicable.
Fix checkpatch warning about the use of bool in structures due to the
wasted space and size of bool, so convert struct members to u8 instead.

Jake fixes a number of potential bugs in the reporting of stats via
ethtool, by simply reporting all the queue statistics, even for the
queues that are not activated.  Fixed a compiler warning, as well as
make the code a bit cleaner but just using order_base_2() for
calculating the power-of-2.

Preethi adds a check to avoid a NULL pointer dereference crash during
initialization.

Brett clarifies the code when it comes to port VLANs and regular VLANs,
by renaming defines and field values to match their intended use and
purpose.

Jesse initializes a variable to avoid garbage values being returned to
the caller.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

1) Add missing "\n" when printing link event error message.

2) Update dev_err statement in probe.

3) Add function description for ice_clear_pf_cfg.

4) Fix coding style for ice_acquire_nvm.

5) netdev->mtu is unsigned so use %u.
Signed-off-by: Anirudh Venkataramanan <anirudh.venkataramanan@intel.com>
Tested-by: Tony Brelinski <tonyx.brelinski@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

Recent versions of checkpatch have a new warning based on a documented
preference of Linus to not use bool in structures due to wasted space and
the size of bool is implementation dependent.  For more information, see
the email thread at https://lkml.org/lkml/2017/11/21/384.
Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Signed-off-by: Anirudh Venkataramanan <anirudh.venkataramanan@intel.com>
Tested-by: Tony Brelinski <tonyx.brelinski@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

In ice_vsi_setup_[tx|rx]_rings, err is uninitialized which can result in
a garbage value return to the caller. Fix that.
Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
Signed-off-by: Anirudh Venkataramanan <anirudh.venkataramanan@intel.com>
Tested-by: Tony Brelinski <tonyx.brelinski@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

1) When ice_ena_msix_range() fails to reserve vectors, a devm_kfree()
   warning was seen in the error flow path. So check pf->irq_tracker
   before use in ice_clear_interrupt_scheme().

2) In ice_vsi_cfg(), check vsi->netdev before use.

3) In ice_get_link_status, check link_up before use.
Signed-off-by: Anirudh Venkataramanan <anirudh.venkataramanan@intel.com>
Tested-by: Tony Brelinski <tonyx.brelinski@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

Remove the following interrupt causes that are not applicable or not
handled:
- PFINT_OICR_HLP_RDY_M
- PFINT_OICR_CPM_RDY_M
- PFINT_OICR_GPIO_M
- PFINT_OICR_STORM_DETECT_M

Add the following interrupt cause that's actually handled in ice_misc_intr:
- PFINT_OICR_PE_CRITERR_M
Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Signed-off-by: Anirudh Venkataramanan <anirudh.venkataramanan@intel.com>
Tested-by: Tony Brelinski <tonyx.brelinski@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

When command line parsing fails in the while loop in do_event_pipe()
because the number of arguments is incorrect or because the keyword is
unknown, an error message is displayed, but bpftool remains stuck in
the loop. Make sure we exit the loop upon failure.

Fixes: f412eed9 ("tools: bpftool: add simple perf event output reader")
Signed-off-by: Quentin Monnet <quentin.monnet@netronome.com>
Reviewed-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>

In the struct ice_aqc_vsi_props the field port_vlan_flags is an
overloaded term because it is used for both port VLANs (PVLANs) and
regular VLANs. This is an issue and is very confusing especially when
dealing with VFs because normal VLANs and port VLANs are not the same.
To fix this the field was renamed to vlan_flags and all of the #define's
labeled *_PVLAN_* were renamed to *_VLAN_* if they are not specific to
port VLANs.

Also in ice_vsi_manage_vlan_stripping, set the ICE_AQ_VSI_VLAN_MODE_ALL
bit to allow the driver to add a VLAN tag to all packets it sends.
Signed-off-by: Brett Creeley <brett.creeley@intel.com>
Signed-off-by: Anirudh Venkataramanan <anirudh.venkataramanan@intel.com>
Tested-by: Tony Brelinski <tonyx.brelinski@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

Currently, we use a combination of ilog2 and is_power_of_2() to
calculate the next power of 2 for the qcount. This appears to be causing
a warning on some combinations of GCC and the Linux kernel:

MODPOST 1 modules
WARNING: "____ilog2_NaN" [ice.ko] undefined!

This appears to because because GCC realizes that qcount could be zero
in some circumstances and thus attempts to link against the
intentionally undefined ___ilog2_NaN function.

The order_base_2 function is intentionally defined to return 0 when
passed 0 as an argument, and thus will be safe to use here.

This not only fixes the warning but makes the resulting code slightly
cleaner, and is really what we should have used originally.

Also update the comment to make it more clear that we are rounding up,
not just incrementing the ilog2 of qcount unconditionally.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Anirudh Venkataramanan <anirudh.venkataramanan@intel.com>
Tested-by: Tony Brelinski <tonyx.brelinski@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

This patch is a consolidation of multiple bug fixes for control queue
processing.

1)  In ice_clean_adminq_subtask() remove unnecessary reads/writes to
    registers. The bits PFINT_FW_CTL, PFINT_MBX_CTL and PFINT_SB_CTL
    are not set when an interrupt arrives, which means that clearing them
    again can be omitted.

2)  Get an accurate value in "pending" by re-reading the control queue
    head register from the hardware.

3)  Fix a corner case involving lost control queue messages by checking
    for new control messages (using ice_ctrlq_pending) before exiting the
    cleanup routine.
Signed-off-by: Anirudh Venkataramanan <anirudh.venkataramanan@intel.com>
Tested-by: Tony Brelinski <tonyx.brelinski@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

Clean control queues only when they are initialized. One of the ways to
validate if the basic initialization is done is by checking value of
cq->sq.head and cq->rq.head variables that specify the register address.
This patch adds a check to avoid NULL pointer dereference crash when tried
to shutdown uninitialized control queue.
Signed-off-by: Preethi Banala <preethi.banala@intel.com>
Signed-off-by: Anirudh Venkataramanan <anirudh.venkataramanan@intel.com>
Tested-by: Tony Brelinski <tonyx.brelinski@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

It is not safe to have the string table for statistics change order or
size over the lifetime of a given netdevice. This is because of the
nature of the 3-step process for obtaining stats. First, user space
performs a request for the size of the strings table. Second it performs
a separate request for the strings themselves, after allocating space
for the table. Third, it requests the stats themselves, also allocating
space for the table.

If the size decreased, there is potential to see garbage data or stats
values. In the worst case, we could potentially see stats values become
mis-aligned with their strings, so that it looks like a statistic is
being reported differently than it actually is.

Even worse, if the size increased, there is potential that the strings
table or stats table was not allocated large enough and the stats code
could access and write to memory it should not, potentially resulting in
undefined behavior and system crashes.

It isn't even safe if the size always changes under the RTNL lock. This
is because the calls take place over multiple user space commands, so it
is not possible to hold the RTNL lock for the entire duration of
obtaining strings and stats. Further, not all consumers of the ethtool
API are the user space ethtool program, and it is possible that one
assumes the strings will not change (valid under the current contract),
and thus only requests the stats values when requesting stats in a loop.

Finally, it's not possible in the general case to detect when the size
changes, because it is quite possible that one value which could impact
the stat size increased, while another decreased. This would result in
the same total number of stats, but reordering them so that stats no
longer line up with the strings they belong to. Since only size changes
aren't enough, we would need some sort of hash or token to determine
when the strings no longer match. This would require extending the
ethtool stats commands, but there is no more space in the relevant
structures.

The real solution to resolve this would be to add a completely new API
for stats, probably over netlink.

In the ice driver, the only thing impacting the stats that is not
constant is the number of queues. Instead of reporting stats for each
used queue, report stats for each allocated queue. We do not change the
number of queues allocated for a given netdevice, as we pass this into
the alloc_etherdev_mq() function to set the num_tx_queues and
num_rx_queues.

This resolves the potential bugs at the slight cost of displaying many
queue statistics which will not be activated.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Anirudh Venkataramanan <anirudh.venkataramanan@intel.com>
Tested-by: Tony Brelinski <tonyx.brelinski@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

Use define for the unit size shift of the Rx LAN context descriptor base
address instead of the magic number 7.
Signed-off-by: Anirudh Venkataramanan <anirudh.venkataramanan@intel.com>
Tested-by: Tony Brelinski <tonyx.brelinski@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

All BPF hash and LRU maps currently have a known and global seed
we feed into jhash() which is 0. This is suboptimal, thus fix it
by generating a random seed upon hashtab setup time which we can
later on feed into jhash() on lookup, update and deletions.

Fixes: 0f8e4bd8 ("bpf: add hashtable type of eBPF maps")
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Song Liu <songliubraving@fb.com>
Reviewed-by: Eduardo Valentin <eduval@amazon.com>

There is already a check for owner == ICE_SCHED_NODE_OWNER_LAN at the
beginning of ice_sched_update_vsi_child_nodes. Remove the additional
check to address the static analysis tool smatch issue "warn: we tested
'owner' before and it was 'false'".
Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Signed-off-by: Anirudh Venkataramanan <anirudh.venkataramanan@intel.com>
Tested-by: Tony Brelinski <tonyx.brelinski@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

This patch fixes the following smatch errors:

1) Fix "odd binop '0x0 & 0xc'" when performing the bitwise-and with a
   constant value of zero (ICE_AQC_GSET_RSS_LUT_TABLE_SIZE_128_FLAG).
   Remove a similar bitwise-and with 0 in ice_add_marker_act() and use the
   right mask ICE_LG_ACT_GENERIC_OFFSET_M in the expression.

2) Fix a similar issue "odd binop '0x0 & 0x1800' in ice_req_irq_msix_misc.

3) Fix "odd binop '0x380000 & 0x7fff8'" in ice_add_marker_act(). Also, use
   a new define ICE_LG_ACT_GENERIC_OFF_RX_DESC_PROF_IDX instead of magic
   number '7'.

4) Fix warn: odd binop '0x0 & 0x18' in ice_set_dflt_vsi_ctx() by removing
   unnecessary logic to explicitly unset bits 3 and 4 in port_vlan_bits.
   These bits are unset already by the memset on ctxt->info.
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Anirudh Venkataramanan <anirudh.venkataramanan@intel.com>
Tested-by: Tony Brelinski <tonyx.brelinski@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

Johan Hedberg says:

====================
pull request: bluetooth 2018-08-23

Here are two important Bluetooth fixes for the MediaTek and RealTek HCI
drivers.

Please let me know if there are any issues pulling, thanks.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

Huazhong Tan says:

====================
net: hns3: bug fix & optimization for HNS3 driver

This patchset presents a bug fix found out when CONFIG_ARM64_64K_PAGES
enable and an optimization for HNS3 driver.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

'truesize' is supposed to be u32, not int, so fix it.
Signed-off-by: Huazhong tan <tanhuazhong@huawei.com>
Signed-off-by: Salil Mehta <salil.mehta@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

When enable the config item "CONFIG_ARM64_64K_PAGES", the size of
PAGE_SIZE is 65536(64K). But the type of page_offset is u16, it will
overflow. So change it to u32, when "CONFIG_ARM64_64K_PAGES" enabled.

Fixes: 76ad4f0e ("net: hns3: Add support of HNS3 Ethernet Driver for hip08 SoC")
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: Salil Mehta <salil.mehta@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Commit 6edb3c96 ("net/ipv6: Defer initialization of dst to data path")
forgot to handle anycast route and init anycast rt->dst.input to ip6_forward.
Fix it by setting anycast rt->dst.input back to ip6_input.

Fixes: 6edb3c96 ("net/ipv6: Defer initialization of dst to data path")
Signed-off-by: Hangbin Liu <liuhangbin@gmail.com>
Reviewed-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Huazhong Tan says:

====================
net: hns: bug fixes & optimization for HNS driver

This patchset presents some bug fixes found out when CONFIG_ARM64_64K_PAGES
enable and an optimization for HNS driver.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

Update hns to drop the hns_nic_get_headlen function in favour of
eth_get_headlen, and hence also removes now redundant hns_nic_get_headlen.
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: Salil Mehta <salil.mehta@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

skb->truesize is not meant to be tracking amount of used bytes in a skb,
but amount of reserved/consumed bytes in memory.

For instance, if we use a single byte in last page fragment, we have to
account the full size of the fragment.

So skb_add_rx_frag needs to calculate the length of the entire buffer into
turesize.

Fixes: 9cbe9fd5 ("net: hns: optimize XGE capability by reducing cpu usage")
Signed-off-by: Huazhong tan <tanhuazhong@huawei.com>
Signed-off-by: Salil Mehta <salil.mehta@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

'truesize' is supposed to be u32, not int, so fix it.
Signed-off-by: Huazhong tan <tanhuazhong@huawei.com>
Signed-off-by: Salil Mehta <salil.mehta@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

When enable the config item "CONFIG_ARM64_64K_PAGES", the size of PAGE_SIZE
is 65536(64K). But the  type of length and page_offset are u16, they will
overflow. So change them to u32.

Fixes: 6fe6611f ("net: add Hisilicon Network Subsystem hnae framework support")
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: Salil Mehta <salil.mehta@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Kevin Yang says:

====================
tcp_bbr: PROBE_RTT minor bug fixes

This series includes two minor bug fixes for the TCP BBR PROBE_RTT
mechanism, and one preparatory patch:

(1) A preparatory patch to reorganize the PROBE_RTT logic by refactoring
    (into its own function) the code to exit PROBE_RTT, since the next
    patch will be using that code in a new context.

(2) Fix: When BBR restarts from idle and if BBR is in PROBE_RTT mode,
    BBR should check if it's time to exit PROBE_RTT. If yes, then BBR
    should exit PROBE_RTT mode and restore the cwnd to its full value.

(3) Fix: Apply the PROBE_RTT cwnd cap even if the count of fully-ACKed
    packets is 0.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

This commit fixes a corner case where TCP BBR would enter PROBE_RTT
mode but not reduce its cwnd. If a TCP receiver ACKed less than one
full segment, the number of delivered/acked packets was 0, so that
bbr_set_cwnd() would short-circuit and exit early, without cutting
cwnd to the value we want for PROBE_RTT.

The fix is to instead make sure that even when 0 full packets are
ACKed, we do apply all the appropriate caps, including the cap that
applies in PROBE_RTT mode.

Fixes: 0f8782ea ("tcp_bbr: add BBR congestion control")
Signed-off-by: Kevin Yang <yyd@google.com>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Reviewed-by: Yuchung Cheng <ycheng@google.com>
Reviewed-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This patch fix the case where BBR does not exit PROBE_RTT mode when
it restarts from idle. When BBR restarts from idle and if BBR is in
PROBE_RTT mode, BBR should check if it's time to exit PROBE_RTT. If
yes, then BBR should exit PROBE_RTT mode and restore the cwnd to its
full value.

Fixes: 0f8782ea ("tcp_bbr: add BBR congestion control")
Signed-off-by: Kevin Yang <yyd@google.com>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Reviewed-by: Yuchung Cheng <ycheng@google.com>
Reviewed-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This patch add a helper function bbr_check_probe_rtt_done() to
  1. check the condition to see if bbr should exit probe_rtt mode;
  2. process the logic of exiting probe_rtt mode.

Fixes: 0f8782ea ("tcp_bbr: add BBR congestion control")
Signed-off-by: Kevin Yang <yyd@google.com>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Yuchung Cheng <ycheng@google.com>
Reviewed-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

tcp uses per-cpu (and per namespace) sockets (net->ipv4.tcp_sk) internally
to send some control packets.

1) RST packets, through tcp_v4_send_reset()
2) ACK packets in SYN-RECV and TIME-WAIT state, through tcp_v4_send_ack()

These packets assert IP_DF, and also use the hashed IP ident generator
to provide an IPv4 ID number.

Geoff Alexander reported this could be used to build off-path attacks.

These packets should not be fragmented, since their size is smaller than
IPV4_MIN_MTU. Only some tunneled paths could eventually have to fragment,
regardless of inner IPID.

We really can use zero IPID, to address the flaw, and as a bonus,
avoid a couple of atomic operations in ip_idents_reserve()
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: Geoff Alexander <alexandg@cs.unm.edu>
Tested-by: Geoff Alexander <alexandg@cs.unm.edu>
Signed-off-by: David S. Miller <davem@davemloft.net>