With the recent changes for how we compute the skb truesize it occurs to me
we are probably going to have a lot of calls to skb_end_pointer -
skb->head.  Instead of running all over the place doing that it would make
more sense to just make it a separate inline skb_end_offset(skb) that way
we can return the correct value without having gcc having to do all the
optimization to cancel out skb->head - skb->head.
Signed-off-by: NAlexander Duyck <alexander.h.duyck@intel.com>
Acked-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Since there is now only one spot that actually uses "fastpath" there isn't
much point in carrying it.  Instead we can just use a check for skb_cloned
to verify if we can perform the fast-path free for the head or not.
Signed-off-by: NAlexander Duyck <alexander.h.duyck@intel.com>
Acked-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The fast-path for pskb_expand_head contains a check where the size plus the
unaligned size of skb_shared_info is compared against the size of the data
buffer.  This code path has two issues.  First is the fact that after the
recent changes by Eric Dumazet to __alloc_skb and build_skb the shared info
is always placed in the optimal spot for a buffer size making this check
unnecessary.  The second issue is the fact that the check doesn't take into
account the aligned size of shared info.  As a result the code burns cycles
doing a memcpy with nothing actually being shifted.
Signed-off-by: NAlexander Duyck <alexander.h.duyck@intel.com>
Acked-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

It appears some networks play bad games with the two bits reserved for
ECN. This can trigger false congestion notifications and very slow
transferts.

Since RFC 3168 (6.1.1) forbids SYN packets to carry CT bits, we can
disable TCP ECN negociation if it happens we receive mangled CT bits in
the SYN packet.
Signed-off-by: NEric Dumazet <edumazet@google.com>
Cc: Perry Lorier <perryl@google.com>
Cc: Matt Mathis <mattmathis@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Wilmer van der Gaast <wilmer@google.com>
Cc: Ankur Jain <jankur@google.com>
Cc: Tom Herbert <therbert@google.com>
Cc: Dave Täht <dave.taht@bufferbloat.net>
Acked-by: NNeal Cardwell <ncardwell@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

With multiple cards is hard to figure out which port caused trouble
int the layer2 routines (e.g. got a timeout).
Now we have the informations in the log output.
Signed-off-by: NKarsten Keil <kkeil@linux-pingi.de>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The timer3 and the activation delay timer need to be independent.
If timer3 fires do not reqest power up we have to send only INFO 0.
Now layer1 pass TBR3 again.
Signed-off-by: NKarsten Keil <kkeil@linux-pingi.de>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

For certification test it is very useful to change the layer1
timer3 value on runtime.
Signed-off-by: NKarsten Keil <kkeil@linux-pingi.de>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

To be full preemptiv safe, we cannot handle a L2 timeout in the timer
context itself, we should do all actions via the D-channel thread.
Signed-off-by: NKarsten Keil <kkeil@linux-pingi.de>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Under some configs it was still not possible to unload the driver,
because the module use count was srewed up.
Signed-off-by: NKarsten Keil <keil@b1-systems.de>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Tei manager reports current layer 1 state on creation.
On state change it reports it to the socket interface.
Signed-off-by: NAndreas Eversberg <andreas@eversberg.eu>
Signed-off-by: NKarsten Keil <keil@b1-systems.de>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Use qdisc_drop() helper where possible.
Signed-off-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This change updates the link flow control configuration so that we
correctly set the link flow control settings for DCB.  Previously we would
have to call the fc_enable call 8 times, once for each packet buffer.  If
we move that logic into the fc_enable call itself we can avoid multiple
unnecessary register writes.

This change also corrects an issue in which we were only shifting the water
marks for 82599 parts by 6 instead of 10.  This was resulting in us only
using 1/16 of the packet buffer when flow control was enabled.
Signed-off-by: NAlexander Duyck <alexander.h.duyck@intel.com>
Tested-by: NRoss Brattain <ross.b.brattain@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

We can avoid many of the forward declarations found in ixgbe_common.c by
just reordering things so this patch does that to help cleanup the code.
Signed-off-by: NAlexander Duyck <alexander.h.duyck@intel.com>
Tested-by: NRoss Brattain <ross.b.brattain@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

This change replaces the calls to put_page with calls to __free_page.

Since the FCoE code is able to access order 1 pages I thought it would be a
good idea to change things over to using __free_pages since that is the
preferred approach for freeing pages.
Signed-off-by: NAlexander Duyck <alexander.h.duyck@intel.com>
Tested-by: NRoss Brattain <ross.b.brattain@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

This change makes it so that ixgbe_fc_autoneg is a void and always sets the
current_mode.  Previously if the link was down we would return an error,
however there is no harm in simply treating a link down case as a case in
which autoneg simply failed.  This allows us to rely on the return value of
the ixgbe_fc_enable call now since there should be no cases where it
returns an error that would normally be ignored.
Signed-off-by: NAlexander Duyck <alexander.h.duyck@intel.com>
Tested-by: NRoss Brattain <ross.b.brattain@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

This change reorders the mapping of rings to q_vectors in the case that the
number of rings exceeds the number of q_vectors.  Previously we would
allocate the first R/N queues to the first q_vector where R is the number
of rings and N is the number of q_vectors.  Instead of doing this we can do
a better job of interleaving the rings to the CPUs by assigning every Nth
ring to the q_vector.

The below tables illustrate this change for the R = 16 N = 4 case.
          Before patch  After patch
q_vector:  0  1  2  3    0  1  2  3
Rings:     0  4  8 12    0  1  2  3
           1  5  9 13    4  5  6  7
           3  6 10 14    8  9 10 11
           4  7 11 15   12 13 14 15

This should improve the performance for both DCB or ATR when the number of
rings exceeds the number of q_vectors allocated by the adapter.
Signed-off-by: NAlexander Duyck <alexander.h.duyck@intel.com>
Tested-by: NRoss Brattain <ross.b.brattain@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

This change makes it so that we can track instances of where a packet was
dropped due to a packet being received when there are no DMA buffers
available in the ring.

For some reason this was only being enabled with RSC, however it makes
more sense to always have this feature on so that we can track any cases
where we might drop a buffer due to an Rx ring being full.
Signed-off-by: NAlexander Duyck <alexander.h.duyck@intel.com>
Tested-by: NRoss Brattain <ross.b.brattain@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

i217 is the next-generation LOM that will be available on systems with the
Lynx Point Platform Controller Hub (PCH) chipset from Intel.  This patch
provides the initial support for the device.
Signed-off-by: NBruce Allan <bruce.w.allan@intel.com>
Tested-by: NJeff Pieper <jeffrey.e.pieper@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

Version bump to 1.11.3-k.
Signed-off-by: NMatthew Vick <matthew.vick@intel.com>
Tested-by: NAaron Brown <aaron.f.brown@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

The idea here seems to be to get a 44bit DMA mask working and if this
fails it should fallback to a 32bit DMA mask. The dma_mask variable is
assigned once to 44bit and never updated. pci_set_dma_mask() and
pci_set_consistent_dma_mask() are both implemented as functions so there
is no evil macro which might update dma_mask. Looking at the assembly, I
see a call to dma_set_mask() followed by dma_supported() and then a jump
passed the second dma_set_mask(). The only way to get to second
dma_set_mask() call is by an error code in the first one.

So I hereby remove the check since it looks superfluous. Please ignore
the path if there is black magic involved.
Signed-off-by: NSebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch adds support for a skb_head_is_locked helper function.  It is
meant to be used any time we are considering transferring the head from
skb->head to a paged frag.  If the head is locked it means we cannot remove
the head from the skb so it must be copied or we must take the skb as a
whole.
Signed-off-by: NAlexander Duyck <alexander.h.duyck@intel.com>
Acked-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

GRO is very optimistic in skb truesize estimates, only taking into
account the used part of fragments.

Be conservative, and use more precise computation, so that bloated GRO
skbs can be collapsed eventually.
Signed-off-by: NEric Dumazet <edumazet@google.com>
Cc: Alexander Duyck <alexander.h.duyck@intel.com>
Cc: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Acked-by: NAlexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NGreg Rose <gregory.v.rose@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

If the Physical Function (PF) resets after the VF has set jumbo
frame MTU then the VF jumbo frame is overwritten.  Make sure the
VF driver always requests proper MTU size after reset
synchronization.
Signed-off-by: NGreg Rose <gregory.v.rose@intel.com>
Tested-by: NSibai Li <sibai.li@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

The X540 10Gig controller is capable of linking at 100Mbits - add
support for reporting that link speed.
Signed-off-by: NGreg Rose <gregory.v.rose@intel.com>
Tested-by: NSibai Li <sibai.li@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

For the 82573, ASPM L1 gets disabled wholesale so this special-case code
is not required. For the 82574 the previous patch does the same as for
the 82573, disabling L1 on the adapter. Thus, this code is no longer
required and can be removed.
Signed-off-by: NChris Boot <bootc@bootc.net>
Tested-by: NJeff Pieper <jeffrey.e.pieper@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

ASPM on the 82574 causes trouble. Currently the driver disables L0s for
this NIC but only disables L1 if the MTU is >1500. This patch simply
causes L1 to be disabled regardless of the MTU setting.
Signed-off-by: NChris Boot <bootc@bootc.net>
Cc: "Wyborny, Carolyn" <carolyn.wyborny@intel.com>
Cc: Nix <nix@esperi.org.uk>
Link: https://lkml.org/lkml/2012/3/19/362Tested-by: NJeff Pieper <jeffrey.e.pieper@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

Previously, IPv6 extension header parsing was disabled for all devices
supported by e1000e when using packet split mode. However, as per a
silicon errata, only certain devices need this restriction and will need
to disable IPv6 extension header parsing for all modes.
Signed-off-by: NMatthew Vick <matthew.vick@intel.com>
Tested-by: NAaron Brown <aaron.f.brown@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

For 82574 and 82583 devices, resolve an intermittent link issue where
the link negotiates to 100Mbps rather than 1Gbps when powering off the
PHY and powering on the PHY after several seconds.
Signed-off-by: NMatthew Vick <matthew.vick@intel.com>
Tested-by: NAaron Brown <aaron.f.brown@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

Calling the locked versions of the read/write PHY ops function pointers
often produces excessively long lines.  Shorten these as is done with
the non-locked versions of the PHY register read/write functions.
Signed-off-by: NBruce Allan <bruce.w.allan@intel.com>
Tested-by: NAaron Brown <aaron.f.brown@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

There is a known issue in the 82577 and 82578 device that can cause a hang
in the device hardware during traffic stress; the current workaround in the
driver is to disable transmit flow control by default.  If the user enables
transmit flow control and the device hang occurs, provide a message in the
syslog suggesting to re-enable the workaround.
Signed-off-by: NBruce Allan <bruce.w.allan@intel.com>
Tested-by: NJeff Pieper <jeffrey.e.pieper@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

While testing the TCP changes I had to fix an issue in order to be able to
load and unload the module.

The recent patch that added thermal sensor support added a use after free
bug on module unload with an 82598 adapter in the system.  To resolve the
issue I have updated the code so that when we free the info_kobj we set it
back to NULL.

I suspect there are likely other bugs present, but I will leave that for
another patch that can undergo more testing.

I am submitting this directly to net-next since this fixes a fairly serious
bug that will lock up the ixgbe module until the system is rebooted.
Signed-off-by: NAlexander Duyck <alexander.h.duyck@intel.com>
Cc: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This change cleans up the last bits of tcp_try_coalesce so that we only
need one goto which jumps to the end of the function.  The idea is to make
the code more readable by putting things in a linear order so that we start
execution at the top of the function, and end it at the bottom.

I also made a slight tweak to the code for handling frags when we are a
clone.  Instead of making it an if (clone) loop else nr_frags = 0 I changed
the logic so that if (!clone) we just set the number of frags to 0 which
disables the for loop anyway.
Signed-off-by: NAlexander Duyck <alexander.h.duyck@intel.com>
Cc: Eric Dumazet <edumazet@google.com>
Cc: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Acked-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This change reorders the code related to the use of an skb->head_frag so it
is placed before we check the rest of the frags.  This allows the code to
read more linearly instead of like some sort of loop.
Signed-off-by: NAlexander Duyck <alexander.h.duyck@intel.com>
Cc: Eric Dumazet <edumazet@google.com>
Cc: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Acked-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch addresses several issues in the way we were tracking the
truesize in tcp_try_coalesce.

First it was using ksize which prevents us from having a 0 sized head frag
and getting a usable result.  To resolve that this patch uses the end
pointer which is set based off either ksize, or the frag_size supplied in
build_skb.  This allows us to compute the original truesize of the entire
buffer and remove that value leaving us with just what was added as pages.

The second issue was the use of skb->len if there is a mergeable head frag.
We should only need to remove the size of an data aligned sk_buff from our
current skb->truesize to compute the delta for a buffer with a reused head.
By using skb->len the value of truesize was being artificially reduced
which means that head frags could use more memory than buffers using
standard allocations.
Signed-off-by: NAlexander Duyck <alexander.h.duyck@intel.com>
Cc: Eric Dumazet <edumazet@google.com>
Cc: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Acked-by: NEric Dumazet <edumazet@google.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>

This change is meant ot prevent stealing the skb->head to use as a page in
the event that the skb->head was cloned.  This allows the other clones to
track each other via shinfo->dataref.

Without this we break down to two methods for tracking the reference count,
one being dataref, the other being the page count.  As a result it becomes
difficult to track how many references there are to skb->head.
Signed-off-by: NAlexander Duyck <alexander.h.duyck@intel.com>
Cc: Eric Dumazet <edumazet@google.com>
Cc: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Acked-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Extend tcp coalescing implementing it from tcp_queue_rcv(), the main
receiver function when application is not blocked in recvmsg().

Function tcp_queue_rcv() is moved a bit to allow its call from
tcp_data_queue()

This gives good results especially if GRO could not kick, and if skb
head is a fragment.
Signed-off-by: NEric Dumazet <edumazet@google.com>
Cc: Alexander Duyck <alexander.h.duyck@intel.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Tom Herbert <therbert@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Before stealing fragments or skb head, we must make sure skbs are not
cloned.

Alexander was worried about destination skb being cloned : In bridge
setups, a driver could be fooled if skb->data_len would not match skb
nr_frags.

If source skb is cloned, we must take references on pages instead.

Bug happened using tcpdump (if not using mmap())

Introduce kfree_skb_partial() helper to cleanup code.
Reported-by: NAlexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>