A typical qdisc setup is the following :

bond0 : bonding device, using HTB hierarchy
eth1/eth2 : slaves, multiqueue NIC, using MQ + FQ qdisc

XPS allows to spread packets on specific tx queues, based on the cpu
doing the send.

Problem is that dequeues from bond0 qdisc can happen on random cpus,
due to the fact that qdisc_run() can dequeue a batch of packets.

CPUA -> queue packet P1 on bond0 qdisc, P1->ooo_okay=1
CPUA -> queue packet P2 on bond0 qdisc, P2->ooo_okay=0

CPUB -> dequeue packet P1 from bond0
        enqueue packet on eth1/eth2
CPUC -> dequeue packet P2 from bond0
        enqueue packet on eth1/eth2 using sk cache (ooo_okay is 0)

get_xps_queue() then might select wrong queue for P1, since current cpu
might be different than CPUA.

P2 might be sent on the old queue (stored in sk->sk_tx_queue_mapping),
if CPUC runs a bit faster (or CPUB spins a bit on qdisc lock)

Effect of this bug is TCP reorders, and more generally not optimal
TX queue placement. (A victim bulk flow can be migrated to the wrong TX
queue for a while)

To fix this, we have to record sender cpu number the first time
dev_queue_xmit() is called for one tx skb.

We can union napi_id (used on receive path) and sender_cpu,
granted we clear sender_cpu in skb_scrub_packet() (credit to Willem for
this union idea)
Signed-off-by: NEric Dumazet <edumazet@google.com>
Cc: Willem de Bruijn <willemb@google.com>
Cc: Nandita Dukkipati <nanditad@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Tx timestamps are looped onto the error queue on top of an skb. This
mechanism leaks packet headers to processes unless the no-payload
options SOF_TIMESTAMPING_OPT_TSONLY is set.

Add a sysctl that optionally drops looped timestamp with data. This
only affects processes without CAP_NET_RAW.

The policy is checked when timestamps are generated in the stack.
It is possible for timestamps with data to be reported after the
sysctl is set, if these were queued internally earlier.

No vulnerability is immediately known that exploits knowledge
gleaned from packet headers, but it may still be preferable to allow
administrators to lock down this path at the cost of possible
breakage of legacy applications.
Signed-off-by: NWillem de Bruijn <willemb@google.com>

----

Changes
  (v1 -> v2)
  - test socket CAP_NET_RAW instead of capable(CAP_NET_RAW)
  (rfc -> v1)
  - document the sysctl in Documentation/sysctl/net.txt
  - fix access control race: read .._OPT_TSONLY only once,
        use same value for permission check and skb generation.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Add timestamping option SOF_TIMESTAMPING_OPT_TSONLY. For transmit
timestamps, this loops timestamps on top of empty packets.

Doing so reduces the pressure on SO_RCVBUF. Payload inspection and
cmsg reception (aside from timestamps) are no longer possible. This
works together with a follow on patch that allows administrators to
only allow tx timestamping if it does not loop payload or metadata.
Signed-off-by: NWillem de Bruijn <willemb@google.com>

----

Changes (rfc -> v1)
  - add documentation
  - remove unnecessary skb->len test (thanks to Richard Cochran)
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The same macros are used for rx as well. So rename it.
Signed-off-by: NJiri Pirko <jiri@resnulli.us>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Not needed, only four cases:
 - kfree_skb (or one of its aliases).
   Don't need to zero, memory will be freed.
 - kfree_skb_partial and head was stolen:  memory will be freed.
 - skb_morph:  The skb header fields (including tc ones) will be
   copied over from the 'to-be-morphed' skb right after
   skb_release_head_state returns.
 - skb_segment:  Same as before, all the skb header
   fields are copied over from the original skb right away.
Signed-off-by: NFlorian Westphal <fw@strlen.de>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

skb_scrub_packet() is called when a packet switches between a context
such as between underlay and overlay, between namespaces, or between
L3 subnets.

While we already scrub the packet mark, connection tracking entry,
and cached destination, the security mark/context is left intact.

It seems wrong to inherit the security context of a packet when going
from overlay to underlay or across forwarding paths.
Signed-off-by: NThomas Graf <tgraf@suug.ch>
Acked-by: NFlavio Leitner <fbl@sysclose.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This change pulls the core functionality out of __netdev_alloc_skb and
places them in a new function named __alloc_rx_skb.  The reason for doing
this is to make these bits accessible to a new function __napi_alloc_skb.
In addition __alloc_rx_skb now has a new flags value that is used to
determine which page frag pool to allocate from.  If the SKB_ALLOC_NAPI
flag is set then the NAPI pool is used.  The advantage of this is that we
do not have to use local_irq_save/restore when accessing the NAPI pool from
NAPI context.

In my test setup I saw at least 11ns of savings using the napi_alloc_skb
function versus the netdev_alloc_skb function, most of this being due to
the fact that we didn't have to call local_irq_save/restore.

The main use case for napi_alloc_skb would be for things such as copybreak
or page fragment based receive paths where an skb is allocated after the
data has been received instead of before.
Signed-off-by: NAlexander Duyck <alexander.h.duyck@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch splits the netdev_alloc_frag function up so that it can be used
on one of two page frag pools instead of being fixed on the
netdev_alloc_cache.  By doing this we can add a NAPI specific function
__napi_alloc_frag that accesses a pool that is only used from softirq
context.  The advantage to this is that we do not need to call
local_irq_save/restore which can be a significant savings.

I also took the opportunity to refactor the core bits that were placed in
__alloc_page_frag.  First I updated the allocation to do either a 32K
allocation or an order 0 page.  This is based on the changes in commmit
d9b2938a where it was found that latencies could be reduced in case of
failures.  Then I also rewrote the logic to work from the end of the page to
the start.  By doing this the size value doesn't have to be used unless we
have run out of space for page fragments.  Finally I cleaned up the atomic
bits so that we just do an atomic_sub_and_test and if that returns true then
we set the page->_count via an atomic_set.  This way we can remove the extra
conditional for the atomic_read since it would have led to an atomic_inc in
the case of success anyway.
Signed-off-by: NAlexander Duyck <alexander.h.duyck@redhat.com>
Acked-by: NAlexei Starovoitov <ast@plumgrid.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Commit ce1a4ea3 ("net: avoid one atomic operation in skb_clone()")
took the wrong way to save one atomic operation.

It is actually possible to avoid two atomic operations, if we
do not change skb->fclone values, and only rely on clone_ref
content to signal if the clone is available or not.

skb_clone() can simply use the fast clone if clone_ref is 1.

kfree_skbmem() can avoid the atomic_dec_and_test() if clone_ref is 1.

Note that because we usually free the clone before the original skb,
this particular attempt is only done for the original skb to have better
branch prediction.

SKB_FCLONE_FREE is removed.
Signed-off-by: NEric Dumazet <edumazet@google.com>
Cc: Chris Mason <clm@fb.com>
Cc: Sabrina Dubroca <sd@queasysnail.net>
Cc: Vijay Subramanian <subramanian.vijay@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Not sure what I was thinking, but doing anything after
releasing a refcount is suicidal or/and embarrassing.

By the time we set skb->fclone to SKB_FCLONE_FREE, another cpu
could have released last reference and freed whole skb.

We potentially corrupt memory or trap if CONFIG_DEBUG_PAGEALLOC is set.
Reported-by: NChris Mason <clm@fb.com>
Fixes: ce1a4ea3 ("net: avoid one atomic operation in skb_clone()")
Signed-off-by: NEric Dumazet <edumazet@google.com>
Cc: Sabrina Dubroca <sd@queasysnail.net>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

So it can be used from out of openvswitch code.
Did couple of cosmetic changes on the way, namely variable naming and
adding support for 8021AD proto.
Signed-off-by: NJiri Pirko <jiri@resnulli.us>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

note that skb_make_writable already exists in net/netfilter/core.c
but does something slightly different.
Suggested-by: NEric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: NJiri Pirko <jiri@resnulli.us>
Acked-by: NPravin B Shelar <pshelar@nicira.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Add a new GSO type, SKB_GSO_TUNNEL_REMCSUM, which indicates remote
checksum offload being done (in this case inner checksum must not
be offloaded to the NIC).

Added logic in __skb_udp_tunnel_segment to handle remote checksum
offload case.
Signed-off-by: NTom Herbert <therbert@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch generalizes commit d6a4a104 ("tcp: GSO should be TSQ
friendly") to protocols using skb_set_owner_w()

TCP uses its own destructor (tcp_wfree) and needs a more complex scheme
as explained in commit 6ff50cd5 ("tcp: gso: do not generate out of
order packets")

This allows UDP sockets using UFO to get proper backpressure,
thus avoiding qdisc drops and excessive cpu usage.

Here are performance test results (macvlan on vlan):

- Before
# netperf -t UDP_STREAM ...
Socket  Message  Elapsed      Messages
Size    Size     Time         Okay Errors   Throughput
bytes   bytes    secs            #      #   10^6bits/sec

212992   65507   60.00      144096 1224195    1258.56
212992           60.00          51              0.45

Average:        CPU     %user     %nice   %system   %iowait    %steal     %idle
Average:        all      0.23      0.00     25.26      0.08      0.00     74.43

- After
# netperf -t UDP_STREAM ...
Socket  Message  Elapsed      Messages
Size    Size     Time         Okay Errors   Throughput
bytes   bytes    secs            #      #   10^6bits/sec

212992   65507   60.00      109593      0     957.20
212992           60.00      109593            957.20

Average:        CPU     %user     %nice   %system   %iowait    %steal     %idle
Average:        all      0.18      0.00      8.38      0.02      0.00     91.43

[edumazet] Rewrote patch and changelog.
Signed-off-by: NToshiaki Makita <makita.toshiaki@lab.ntt.co.jp>
Signed-off-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

if ->encapsulation is set we have to use inner_tcp_hdrlen and add the
size of the inner network headers too.

This is 'mostly harmless'; tbf might send skb that is slightly over
quota or drop skb even if it would have fit.
Signed-off-by: NFlorian Westphal <fw@strlen.de>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This is illegal to use atomic_set(&page->_count, ...) even if we 'own'
the page. Other entities in the kernel need to use get_page_unless_zero()
to get a reference to the page before testing page properties, so we could
loose a refcount increment.

The only case it is valid is when page->_count is 0

Fixes: 540eb7bf ("net: Update alloc frag to reduce get/put page usage and recycle pages")
Signed-off-by: NEric Dumaze <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch fix following warning.
Warning(.//net/core/skbuff.c:4142): No description found for parameter 'header_len'
Warning(.//net/core/skbuff.c:4142): No description found for parameter 'data_len'
Warning(.//net/core/skbuff.c:4142): No description found for parameter 'max_page_order'
Warning(.//net/core/skbuff.c:4142): No description found for parameter 'errcode'
Warning(.//net/core/skbuff.c:4142): No description found for parameter 'gfp_mask'

Acutually the descriptions exist, but missing "@" in front.

This problem start to happen when following commit was merged
into Linus's tree during 3.18-rc1 merge period.
commit 2e4e4410
net: add alloc_skb_with_frags() helper
Signed-off-by: NMasanari Iida <standby24x7@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Its unfortunate we have to walk again skb list to find the tail
after segmentation, even if data is probably hot in cpu caches.

skb_segment() can store the tail of the list into segs->prev,
and validate_xmit_skb_list() can immediately get the tail.
Signed-off-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

SKB_FCLONE_UNAVAILABLE has overloaded meaning depending on type of skb.
1: If skb is allocated from head_cache, it indicates fclone is not available.
2: If skb is a companion fclone skb (allocated from fclone_cache), it indicates
it is available to be used.

To avoid confusion for case 2 above, this patch  replaces
SKB_FCLONE_UNAVAILABLE with SKB_FCLONE_FREE where appropriate. For fclone
companion skbs, this indicates it is free for use.

SKB_FCLONE_UNAVAILABLE will now simply indicate skb is from head_cache and
cannot / will not have a companion fclone.
Signed-off-by: NVijay Subramanian <subramanian.vijay@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

skb_gro_receive() is only called from tcp_gro_receive() which is
not in a module.
Signed-off-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

In 34666d46 ("netfilter: bridge: move br_netfilter out of the core"),
the bridge netfilter code has been modularized.

Use IS_ENABLED instead of ifdef to cover the module case.

Fixes: 34666d46 ("netfilter: bridge: move br_netfilter out of the core")
Signed-off-by: NPablo Neira Ayuso <pablo@netfilter.org>

Fast clone cloning can actually avoid an atomic_inc(), if we
guarantee prior clone_ref value is 1.

This requires a change kfree_skbmem(), to perform the
atomic_dec_and_test() on clone_ref before setting fclone to
SKB_FCLONE_UNAVAILABLE.
Signed-off-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Lets use a proper structure to clearly document and implement
skb fast clones.

Then, we might experiment more easily alternative layouts.

This patch adds a new skb_fclone_busy() helper, used by tcp and xfrm,
to stop leaking of implementation details.
Signed-off-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

In commit 8a29111c ("net: gro: allow to build full sized skb")
I added a regression for linear skb that traditionally force GRO
to use the frag_list fallback.

Erez Shitrit found that at most two segments were aggregated and
the "if (skb_gro_len(p) != pinfo->gso_size)" test was failing.

This is because pinfo at this spot still points to the last skb in the
chain, instead of the first one, where we find the correct gso_size
information.
Signed-off-by: NEric Dumazet <edumazet@google.com>
Fixes: 8a29111c ("net: gro: allow to build full sized skb")
Reported-by: NErez Shitrit <erezsh@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

With proliferation of bit fields in sk_buff, __copy_skb_header() became
quite expensive, showing as the most expensive function in a GSO
workload.

__copy_skb_header() performance is also critical for non GSO TCP
operations, as it is used from skb_clone()

This patch carefully moves all the fields that were not copied in a
separate zone : cloned, nohdr, fclone, peeked, head_frag, xmit_more

Then I moved all other fields and all other copied fields in a section
delimited by headers_start[0]/headers_end[0] section so that we
can use a single memcpy() call, inlined by compiler using long
word load/stores.

I also tried to make all copies in the natural orders of sk_buff,
to help hardware prefetching.

I made sure sk_buff size did not change.
Signed-off-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Cache skb_shinfo(skb) in a variable to avoid computing it multiple
times.

Reorganize the tests to remove one indentation level.
Signed-off-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

While profiling TCP stack, I noticed one useless atomic operation
in tcp_sendmsg(), caused by skb_header_release().

It turns out all current skb_header_release() users have a fresh skb,
that no other user can see, so we can avoid one atomic operation.

Introduce __skb_header_release() to clearly document this.

This gave me a 1.5 % improvement on TCP_RR workload.
Signed-off-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Extract from sock_alloc_send_pskb() code building skb with frags,
so that we can reuse this in other contexts.

Intent is to use it from tcp_send_rcvq(), tcp_collapse(), ...

We also want to replace some skb_linearize() calls to a more reliable
strategy in pathological cases where we need to reduce number of frags.
Signed-off-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

We can allow a segment with FIN to be aggregated,
if we take care to add tcp flags,
and if skb_try_coalesce() takes care of zero sized skbs.
Signed-off-by: NEric Dumazet <edumazet@google.com>
Acked-by: NNeal Cardwell <ncardwell@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

There is a possible issue with the use, or lack thereof of sk_refcnt and
sk_wmem_alloc in the wifi ack status functionality.

Specifically if a socket were to request acknowledgements, and the socket
were to have sk_refcnt drop to 0 resulting in it waiting on sk_wmem_alloc
to reach 0 it would be possible to have sock_queue_err_skb orphan the last
buffer, resulting in __sk_free being called on the socket.  After this the
buffer is enqueued on sk_error_queue, however the queue has already been
flushed resulting in at least a memory leak, if not a data corruption.
Signed-off-by: NAlexander Duyck <alexander.h.duyck@intel.com>
Acked-by: NJohannes Berg <johannes@sipsolutions.net>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This change adds some documentation to the call skb_clone_sk.  This is
meant to help clarify the purpose of the function for other developers.
Signed-off-by: NAlexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The phy timestamping takes a different path than the regular timestamping
does in that it will create a clone first so that the packets needing to be
timestamped can be placed in a queue, or the context block could be used.

In order to support these use cases I am pulling the core of the code out
so it can be used in other drivers beyond just phy devices.

In addition I have added a destructor named sock_efree which is meant to
provide a simple way for dropping the reference to skb exceptions that
aren't part of either the receive or send windows for the socket, and I
have removed some duplication in spots where this destructor could be used
in place of sock_edemux.
Signed-off-by: NAlexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This change merges the shared bits that exist between skb_tx_tstamp and
skb_complete_tx_timestamp.  By doing this we can avoid the two diverging as
there were already changes pushed into skb_tx_tstamp that hadn't made it
into the other function.

In addition this resolves issues with the fact that
skb_complete_tx_timestamp was included in linux/skbuff.h even though it was
only compiled in if phy timestamping was enabled.
Signed-off-by: NAlexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch fix spelling typo found in DocBook/networking.xml.
It is because the neworking.xml is generated from comments
in the source, I have to fix typo in comments within the source.
Signed-off-by: NMasanari Iida <standby24x7@gmail.com>
Acked-by: NRandy Dunlap <rdunlap@infradead.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

sk->sk_error_queue is dequeued in four locations. All share the
exact same logic. Deduplicate.

Also collapse the two critical sections for dequeue (at the top of
the recv handler) and signal (at the bottom).

This moves signal generation for the next packet forward, which should
be harmless.

It also changes the behavior if the recv handler exits early with an
error. Previously, a signal for follow-up packets on the errqueue
would then not be scheduled. The new behavior, to always signal, is
arguably a bug fix.

For rxrpc, the change causes the same function to be called repeatedly
for each queued packet (because the recv handler == sk_error_report).
It is likely that all packets will fail for the same reason (e.g.,
memory exhaustion).

This code runs without sk_lock held, so it is not safe to trust that
sk->sk_err is immutable inbetween releasing q->lock and the subsequent
test. Introduce int err just to avoid this potential race.
Signed-off-by: NWillem de Bruijn <willemb@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Replace uses of get_cpu_var for address calculation through this_cpu_ptr.

Cc: netdev@vger.kernel.org
Cc: Eric Dumazet <edumazet@google.com>
Acked-by: NDavid S. Miller <davem@davemloft.net>
Signed-off-by: NChristoph Lameter <cl@linux.com>
Signed-off-by: NTejun Heo <tj@kernel.org>

Currently the functionality to untag traffic on input resides
as part of the vlan module and is build only when VLAN support
is enabled in the kernel.  When VLAN is disabled, the function
vlan_untag() turns into a stub and doesn't really untag the
packets.  This seems to create an interesting interaction
between VMs supporting checksum offloading and some network drivers.

There are some drivers that do not allow the user to change
tx-vlan-offload feature of the driver.  These drivers also seem
to assume that any VLAN-tagged traffic they transmit will
have the vlan information in the vlan_tci and not in the vlan
header already in the skb.  When transmitting skbs that already
have tagged data with partial checksum set, the checksum doesn't
appear to be updated correctly by the card thus resulting in a
failure to establish TCP connections.

The following is a packet trace taken on the receiver where a
sender is a VM with a VLAN configued.  The host VM is running on
doest not have VLAN support and the outging interface on the
host is tg3:
10:12:43.503055 52:54:00:ae:42:3f > 28:d2:44:7d:c2:de, ethertype 802.1Q
(0x8100), length 78: vlan 100, p 0, ethertype IPv4, (tos 0x0, ttl 64, id 27243,
offset 0, flags [DF], proto TCP (6), length 60)
    10.0.100.1.58545 > 10.0.100.10.ircu-2: Flags [S], cksum 0xdc39 (incorrect
-> 0x48d9), seq 1069378582, win 29200, options [mss 1460,sackOK,TS val
4294837885 ecr 0,nop,wscale 7], length 0
10:12:44.505556 52:54:00:ae:42:3f > 28:d2:44:7d:c2:de, ethertype 802.1Q
(0x8100), length 78: vlan 100, p 0, ethertype IPv4, (tos 0x0, ttl 64, id 27244,
offset 0, flags [DF], proto TCP (6), length 60)
    10.0.100.1.58545 > 10.0.100.10.ircu-2: Flags [S], cksum 0xdc39 (incorrect
-> 0x44ee), seq 1069378582, win 29200, options [mss 1460,sackOK,TS val
4294838888 ecr 0,nop,wscale 7], length 0

This connection finally times out.

I've only access to the TG3 hardware in this configuration thus have
only tested this with TG3 driver.  There are a lot of other drivers
that do not permit user changes to vlan acceleration features, and
I don't know if they all suffere from a similar issue.

The patch attempt to fix this another way.  It moves the vlan header
stipping code out of the vlan module and always builds it into the
kernel network core.  This way, even if vlan is not supported on
a virtualizatoin host, the virtual machines running on top of such
host will still work with VLANs enabled.

CC: Patrick McHardy <kaber@trash.net>
CC: Nithin Nayak Sujir <nsujir@broadcom.com>
CC: Michael Chan <mchan@broadcom.com>
CC: Jiri Pirko <jiri@resnulli.us>
Signed-off-by: NVladislav Yasevich <vyasevic@redhat.com>
Acked-by: NJiri Pirko <jiri@resnulli.us>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

TCP timestamping extends SO_TIMESTAMPING to bytestreams.

Bytestreams do not have a 1:1 relationship between send() buffers and
network packets. The feature interprets a send call on a bytestream as
a request for a timestamp for the last byte in that send() buffer.

The choice corresponds to a request for a timestamp when all bytes in
the buffer have been sent. That assumption depends on in-order kernel
transmission. This is the common case. That said, it is possible to
construct a traffic shaping tree that would result in reordering.
The guarantee is strong, then, but not ironclad.

This implementation supports send and sendpages (splice). GSO replaces
one large packet with multiple smaller packets. This patch also copies
the option into the correct smaller packet.

This patch does not yet support timestamping on data in an initial TCP
Fast Open SYN, because that takes a very different data path.

If ID generation in ee_data is enabled, bytestream timestamps return a
byte offset, instead of the packet counter for datagrams.

The implementation supports a single timestamp per packet. It silenty
replaces requests for previous timestamps. To avoid missing tstamps,
flush the tcp queue by disabling Nagle, cork and autocork. Missing
tstamps can be detected by offset when the ee_data ID is enabled.

Implementation details:

- On GSO, the timestamping code can be included in the main loop. I
moved it into its own loop to reduce the impact on the common case
to a single branch.

- To avoid leaking the absolute seqno to userspace, the offset
returned in ee_data must always be relative. It is an offset between
an skb and sk field. The first is always set (also for GSO & ACK).
The second must also never be uninitialized. Only allow the ID
option on sockets in the ESTABLISHED state, for which the seqno
is available. Never reset it to zero (instead, move it to the
current seqno when reenabling the option).
Signed-off-by: NWillem de Bruijn <willemb@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Kernel transmit latency is often incurred in the packet scheduler.
Introduce a new timestamp on transmission just before entering the
scheduler. When data travels through multiple devices (bonding,
tunneling, ...) each device will export an individual timestamp.
Signed-off-by: NWillem de Bruijn <willemb@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Datagrams timestamped on transmission can coexist in the kernel stack
and be reordered in packet scheduling. When reading looped datagrams
from the socket error queue it is not always possible to unique
correlate looped data with original send() call (for application
level retransmits). Even if possible, it may be expensive and complex,
requiring packet inspection.

Introduce a data-independent ID mechanism to associate timestamps with
send calls. Pass an ID alongside the timestamp in field ee_data of
sock_extended_err.

The ID is a simple 32 bit unsigned int that is associated with the
socket and incremented on each send() call for which software tx
timestamp generation is enabled.

The feature is enabled only if SOF_TIMESTAMPING_OPT_ID is set, to
avoid changing ee_data for existing applications that expect it 0.
The counter is reset each time the flag is reenabled. Reenabling
does not change the ID of already submitted data. It is possible
to receive out of order IDs if the timestamp stream is not quiesced
first.
Signed-off-by: NWillem de Bruijn <willemb@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>