{inet,ipv6}_gro_receive functions perform flush checks (ttl, flags,
iph->id, ...) against all packets in a loop. These flush checks are used in
all merging UDP and TCP flows.
These checks need to be done only once and only against the found p skb,
since they only affect flush and not same_flow.
This patch leverages correct network header offsets from the cb for both
outer and inner network headers - allowing these checks to be done only
once, in tcp_gro_receive and udp_gro_receive_segment. As a result,
NAPI_GRO_CB(p)->flush is not used at all. In addition, flush_id checks are
more declarative and contained in inet_gro_flush, thus removing the need
for flush_id in napi_gro_cb.
This results in less parsing code for non-loop flush tests for TCP and UDP
flows.
To make sure results are not within noise range - I've made netfilter drop
all TCP packets, and measured CPU performance in GRO (in this case GRO is
responsible for about 50% of the CPU utilization).
perf top while replaying 64 parallel IP/TCP streams merging in GRO:
(gro_receive_network_flush is compiled inline to tcp_gro_receive)
net-next:
6.94% [kernel] [k] inet_gro_receive
3.02% [kernel] [k] tcp_gro_receive
patch applied:
4.27% [kernel] [k] tcp_gro_receive
4.22% [kernel] [k] inet_gro_receive
perf top while replaying 64 parallel IP/IP/TCP streams merging in GRO (same
results for any encapsulation, in this case inet_gro_receive is top
offender in net-next)
net-next:
10.09% [kernel] [k] inet_gro_receive
2.08% [kernel] [k] tcp_gro_receive
patch applied:
6.97% [kernel] [k] inet_gro_receive
3.68% [kernel] [k] tcp_gro_receive
Signed-off-by: Richard Gobert <richardbgobert@gmail.com>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Link: https://lore.kernel.org/r/20240509190819.2985-3-richardbgobert@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Commits a602456 ("udp: Add GRO functions to UDP socket") and 57c67ff ("udp:
additional GRO support") introduce incorrect usage of {ip,ipv6}_hdr in the
complete phase of gro. The functions always return skb->network_header,
which in the case of encapsulated packets at the gro complete phase, is
always set to the innermost L3 of the packet. That means that calling
{ip,ipv6}_hdr for skbs which completed the GRO receive phase (both in
gro_list and *_gro_complete) when parsing an encapsulated packet's _outer_
L3/L4 may return an unexpected value.
This incorrect usage leads to a bug in GRO's UDP socket lookup.
udp{4,6}_lib_lookup_skb functions use ip_hdr/ipv6_hdr respectively. These
*_hdr functions return network_header which will point to the innermost L3,
resulting in the wrong offset being used in __udp{4,6}_lib_lookup with
encapsulated packets.
This patch adds network_offset and inner_network_offset to napi_gro_cb, and
makes sure both are set correctly.
To fix the issue, network_offsets union is used inside napi_gro_cb, in
which both the outer and the inner network offsets are saved.
Reproduction example:
Endpoint configuration example (fou + local address bind)
# ip fou add port 6666 ipproto 4
# ip link add name tun1 type ipip remote 2.2.2.1 local 2.2.2.2 encap fou encap-dport 5555 encap-sport 6666 mode ipip
# ip link set tun1 up
# ip a add 1.1.1.2/24 dev tun1
Netperf TCP_STREAM result on net-next before patch is applied:
net-next main, GRO enabled:
$ netperf -H 1.1.1.2 -t TCP_STREAM -l 5
Recv Send Send
Socket Socket Message Elapsed
Size Size Size Time Throughput
bytes bytes bytes secs. 10^6bits/sec
131072 16384 16384 5.28 2.37
net-next main, GRO disabled:
$ netperf -H 1.1.1.2 -t TCP_STREAM -l 5
Recv Send Send
Socket Socket Message Elapsed
Size Size Size Time Throughput
bytes bytes bytes secs. 10^6bits/sec
131072 16384 16384 5.01 2745.06
patch applied, GRO enabled:
$ netperf -H 1.1.1.2 -t TCP_STREAM -l 5
Recv Send Send
Socket Socket Message Elapsed
Size Size Size Time Throughput
bytes bytes bytes secs. 10^6bits/sec
131072 16384 16384 5.01 2877.38
Fixes: a6024562ff ("udp: Add GRO functions to UDP socket")
Signed-off-by: Richard Gobert <richardbgobert@gmail.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Currently the so-called GRO fast path is only enabled for
napi_frags_skb() callers.
After the prior patch, we no longer have to clear frag0 whenever
we pulled bytes to skb->head.
We therefore can initialize frag0 to skb->data so that GRO
fast path can be used in the following additional cases:
- Drivers using header split (populating skb->data with headers,
and having payload in one or more page fragments).
- Drivers not using any page frag (entire packet is in skb->data)
Add a likely() in skb_gro_may_pull() to help the compiler
to generate better code if possible.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Change skb_gro_network_header() to accept a const sk_buff
and to no longer check if frag0 is NULL or not.
This allows to remove skb_gro_frag0_invalidate()
which is seen in profiles when header-split is enabled.
sk_buff parameter is constified for skb_gro_header_fast(),
inet_gro_compute_pseudo() and ip6_gro_compute_pseudo().
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
skb_gro_header_hard() is renamed to skb_gro_may_pull() to match
the convention used by common helpers like pskb_may_pull().
This means the condition is inverted:
if (skb_gro_header_hard(skb, hlen))
slow_path();
becomes:
if (!skb_gro_may_pull(skb, hlen))
slow_path();
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
This patch enables the GRO codepath for IPv4 ESP in UDP encapsulated
packets. Decapsulation happens at L2 and saves a full round through
the stack for each packet. This is also needed to support HW offload
for ESP in UDP encapsulation.
Enabling this would imporove performance for ESP in UDP datapath, i.e
IPsec with NAT in between.
By default GRP for ESP-in-UDP is disabled for UDP sockets.
To enable this feature for an ESP socket, the following two options
need to be set:
1. enable ESP-in-UDP: (this is already set by an IKE daemon).
int type = UDP_ENCAP_ESPINUDP;
setsockopt(fd, SOL_UDP, UDP_ENCAP, &type, sizeof(type));
2. To enable GRO for ESP in UDP socket:
type = true;
setsockopt(fd, SOL_UDP, UDP_GRO, &type, sizeof(type));
Enabling ESP-in-UDP has the side effect of preventing the Linux stack from
seeing ESP packets at the L3 (when ESP OFFLOAD is disabled), as packets are
immediately decapsulated from UDP and decrypted.
This change may affect nftable rules that match on ESP packets at L3.
Also tcpdump won't see the ESP packet.
Developers/admins are advised to review and adapt any nftable rules
accordingly before enabling this feature to prevent potential rule breakage.
Also tcpdump will not see from ESP packets from a ESP in UDP flow, when this
is enabled.
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
Co-developed-by: Antony Antony <antony.antony@secunet.com>
Signed-off-by: Antony Antony <antony.antony@secunet.com>
Reviewed-by: Eyal Birger <eyal.birger@gmail.com>
This patch fixes a misuse of IP{6}CB(skb) in GRO, while calling to
`udp6_lib_lookup2` when handling udp tunnels. `udp6_lib_lookup2` fetch the
device from CB. The fix changes it to fetch the device from `skb->dev`.
l3mdev case requires special attention since it has a master and a slave
device.
Fixes: a6024562ff ("udp: Add GRO functions to UDP socket")
Reported-by: Gal Pressman <gal@nvidia.com>
Signed-off-by: Richard Gobert <richardbgobert@gmail.com>
Reviewed-by: David Ahern <dsahern@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
The GRO control block (NAPI_GRO_CB) is currently at its maximum size.
This commit reduces its size by putting two groups of fields that are
used only at different times into a union.
Specifically, the fields frag0 and frag0_len are the fields that make up
the frag0 optimisation mechanism, which is used during the initial
parsing of the SKB.
The fields last and age are used after the initial parsing, while the
SKB is stored in the GRO list, waiting for other packets to arrive.
There was one location in dev_gro_receive that modified the frag0 fields
after setting last and age. I changed this accordingly without altering
the code behaviour.
Signed-off-by: Richard Gobert <richardbgobert@gmail.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Link: https://lore.kernel.org/r/20230601161407.GA9253@debian
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Introduce a simple helper function to replace a common pattern.
When accessing the GRO header, we fetch the pointer from frag0,
then test its validity and fetch it from the skb when necessary.
This leads to the pattern
skb_gro_header_fast -> skb_gro_header_hard -> skb_gro_header_slow
recurring many times throughout GRO code.
This patch replaces these patterns with a single inlined function
call, improving code readability.
Signed-off-by: Richard Gobert <richardbgobert@gmail.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Link: https://lore.kernel.org/r/20220823071034.GA56142@debian
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
While reading gro_normal_batch, it can be changed concurrently.
Thus, we need to add READ_ONCE() to its reader.
Fixes: 323ebb61e3 ("net: use listified RX for handling GRO_NORMAL skbs")
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Acked-by: Edward Cree <ecree.xilinx@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Following the cited commit, sparse started complaining about:
../include/net/gro.h:58:1: warning: directive in macro's argument list
../include/net/gro.h:59:1: warning: directive in macro's argument list
Fix that by moving the defines out of the struct_group() macro.
Fixes: de5a1f3ce4 ("net: gro: minor optimization for dev_gro_receive()")
Reviewed-by: Maxim Mikityanskiy <maximmi@nvidia.com>
Signed-off-by: Gal Pressman <gal@nvidia.com>
Acked-by: Alexander Lobakin <alexandr.lobakin@intel.com>
Acked-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
While inspecting some perf report, I noticed that the compiler
emits suboptimal code for the napi CB initialization, fetching
and storing multiple times the memory for flags bitfield.
This is with gcc 10.3.1, but I observed the same with older compiler
versions.
We can help the compiler to do a nicer work clearing several
fields at once using an u32 alias. The generated code is quite
smaller, with the same number of conditional.
Before:
objdump -t net/core/gro.o | grep " F .text"
0000000000000bb0 l F .text 0000000000000357 dev_gro_receive
After:
0000000000000bb0 l F .text 000000000000033c dev_gro_receive
v1 -> v2:
- use struct_group (Alexander and Alex)
RFC -> v1:
- use __struct_group to delimit the zeroed area (Alexander)
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Vladimir reported csum issues after my recent change in skb_postpull_rcsum()
Issue here is the following:
initial skb->csum is the csum of
[part to be pulled][rest of packet]
Old code:
skb->csum = csum_sub(skb->csum, csum_partial(pull, pull_length, 0));
New code:
skb->csum = ~csum_partial(pull, pull_length, ~skb->csum);
This is broken if the csum of [pulled part]
happens to be equal to skb->csum, because end
result of skb->csum is 0 in new code, instead
of being 0xffffffff
David Laight suggested to use
skb->csum = -csum_partial(pull, pull_length, -skb->csum);
I based my patches on existing code present in include/net/seg6.h,
update_csum_diff4() and update_csum_diff16() which might need
a similar fix.
I guess that my tests, mostly pulling 40 bytes of IPv6 header
were not providing enough entropy to hit this bug.
v2: added wsum_negate() to make sparse happy.
Fixes: 29c3002644 ("net: optimize skb_postpull_rcsum()")
Fixes: 0bd28476f6 ("gro: optimize skb_gro_postpull_rcsum()")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Suggested-by: David Laight <David.Laight@ACULAB.COM>
Cc: David Lebrun <dlebrun@google.com>
Tested-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Link: https://lore.kernel.org/r/20211204045356.3659278-1-eric.dumazet@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
We can leverage third argument to csum_partial():
X = csum_sub(X, csum_partial(start, len, 0));
-->
X = csum_add(X, ~csum_partial(start, len, 0));
-->
X = ~csum_partial(start, len, ~X);
This removes one add/adc pair and its dependency against the carry flag.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Move gro code and data from net/core/dev.c to net/core/gro.c
to ease maintenance.
gro_normal_list() and gro_normal_one() are inlined
because they are called from both files.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
net/core/gro.c will contain all core gro functions,
to shrink net/core/skbuff.c and net/core/dev.c
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
include/linux/netdevice.h became too big, move gro stuff
into include/net/gro.h
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
call_gro_receive() is used to limit GRO recursion, but it works only
with callback pointers.
There's a combined version of call_gro_receive() + INDIRECT_CALL_2()
in <net/inet_common.h>, but it doesn't check for IPv6 modularity.
Add a similar new helper to cover both of these. It can and will be
used to avoid retpoline overhead when IP header lies behind another
offloaded proto.
Signed-off-by: Alexander Lobakin <alobakin@pm.me>
Signed-off-by: David S. Miller <davem@davemloft.net>
If some source file includes <net/gro.h>, but doesn't include
<linux/indirect_call_wrapper.h>:
In file included from net/8021q/vlan_core.c:7:
./include/net/gro.h:6:1: warning: data definition has no type or storage class
6 | INDIRECT_CALLABLE_DECLARE(struct sk_buff *ipv6_gro_receive(struct list_head *,
| ^~~~~~~~~~~~~~~~~~~~~~~~~
./include/net/gro.h:6:1: error: type defaults to ‘int’ in declaration of ‘INDIRECT_CALLABLE_DECLARE’ [-Werror=implicit-int]
[...]
Include <linux/indirect_call_wrapper.h> directly. It's small and
won't pull lots of dependencies.
Also add some incomplete struct declarations to be fully stacked.
Fixes: 04f00ab227 ("net/core: move gro function declarations to separate header ")
Signed-off-by: Alexander Lobakin <alobakin@pm.me>
Signed-off-by: David S. Miller <davem@davemloft.net>
