linux.git/net/core/gro.c, branch v6.10 Linux kernel source tree net: gro: move L3 flush checks to tcp_gro_receive and udp_gro_receive_segment 2024-05-13T21:44:06+00:00 Richard Gobert richardbgobert@gmail.com 2024-05-09T19:08:18+00:00 4b0ebbca3e1679765c06d5c466ee7f3228d4b156 {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> 
{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>
net: move skb_gro_receive_list from udp to core 2024-05-06T09:54:03+00:00 Felix Fietkau nbd@nbd.name 2024-05-02T08:44:42+00:00 8928756d53d5b99dcd18073dc7738b8ebdbe7d96 This helper function will be used for TCP fraglist GRO support Acked-by: Paolo Abeni <pabeni@redhat.com> Reviewed-by: Eric Dumazet <edumazet@google.com> Signed-off-by: Felix Fietkau <nbd@nbd.name> Reviewed-by: David Ahern <dsahern@kernel.org> Reviewed-by: Willem de Bruijn <willemb@google.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com> 
This helper function will be used for TCP fraglist GRO support

Acked-by: Paolo Abeni <pabeni@redhat.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Reviewed-by: David Ahern <dsahern@kernel.org>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Merge git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net 2024-05-02T19:06:25+00:00 Jakub Kicinski kuba@kernel.org 2024-05-02T19:05:13+00:00 e958da0ddbe831197a0023251880a4a09d5ba268 Cross-merge networking fixes after downstream PR. Conflicts: include/linux/filter.h kernel/bpf/core.c 66e13b615a0c ("bpf: verifier: prevent userspace memory access") d503a04f8bc0 ("bpf: Add support for certain atomics in bpf_arena to x86 JIT") https://lore.kernel.org/all/20240429114939.210328b0@canb.auug.org.au/ No adjacent changes. Signed-off-by: Jakub Kicinski <kuba@kernel.org> 
Cross-merge networking fixes after downstream PR.

Conflicts:

include/linux/filter.h
kernel/bpf/core.c
  66e13b615a0c ("bpf: verifier: prevent userspace memory access")
  d503a04f8bc0 ("bpf: Add support for certain atomics in bpf_arena to x86 JIT")
https://lore.kernel.org/all/20240429114939.210328b0@canb.auug.org.au/

No adjacent changes.

Signed-off-by: Jakub Kicinski <kuba@kernel.org>
net: gro: fix udp bad offset in socket lookup by adding {inner_}network_offset to napi_gro_cb 2024-05-02T09:02:48+00:00 Richard Gobert richardbgobert@gmail.com 2024-04-30T14:35:54+00:00 5ef31ea5d053a8f493a772ebad3f3ce82c35d845 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: a6024562ffd7 ("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> 
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: a6024562ffd7 ("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>
net: move skb ref helpers to new header 2024-04-12T02:29:22+00:00 Mina Almasry almasrymina@google.com 2024-04-10T19:05:01+00:00 f6d827b180bda01f8805bf5e85307419b0d6f890 Add a new header, linux/skbuff_ref.h, which contains all the skb_*_ref() helpers. Many of the consumers of skbuff.h do not actually use any of the skb ref helpers, and we can speed up compilation a bit by minimizing this header file. Additionally in the later patch in the series we add page_pool support to skb_frag_ref(), which requires some page_pool dependencies. We can now add these dependencies to skbuff_ref.h instead of a very ubiquitous skbuff.h Signed-off-by: Mina Almasry <almasrymina@google.com> Link: https://lore.kernel.org/r/20240410190505.1225848-2-almasrymina@google.com Signed-off-by: Jakub Kicinski <kuba@kernel.org> 
Add a new header, linux/skbuff_ref.h, which contains all the skb_*_ref()
helpers. Many of the consumers of skbuff.h do not actually use any of
the skb ref helpers, and we can speed up compilation a bit by minimizing
this header file.

Additionally in the later patch in the series we add page_pool support
to skb_frag_ref(), which requires some page_pool dependencies. We can
now add these dependencies to skbuff_ref.h instead of a very ubiquitous
skbuff.h

Signed-off-by: Mina Almasry <almasrymina@google.com>
Link: https://lore.kernel.org/r/20240410190505.1225848-2-almasrymina@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
gro: fix ownership transfer 2024-03-29T11:30:44+00:00 Antoine Tenart atenart@kernel.org 2024-03-26T11:33:59+00:00 ed4cccef64c1d0d5b91e69f7a8a6697c3a865486 If packets are GROed with fraglist they might be segmented later on and continue their journey in the stack. In skb_segment_list those skbs can be reused as-is. This is an issue as their destructor was removed in skb_gro_receive_list but not the reference to their socket, and then they can't be orphaned. Fix this by also removing the reference to the socket. For example this could be observed, kernel BUG at include/linux/skbuff.h:3131! (skb_orphan) RIP: 0010:ip6_rcv_core+0x11bc/0x19a0 Call Trace: ipv6_list_rcv+0x250/0x3f0 __netif_receive_skb_list_core+0x49d/0x8f0 netif_receive_skb_list_internal+0x634/0xd40 napi_complete_done+0x1d2/0x7d0 gro_cell_poll+0x118/0x1f0 A similar construction is found in skb_gro_receive, apply the same change there. Fixes: 5e10da5385d2 ("skbuff: allow 'slow_gro' for skb carring sock reference") Signed-off-by: Antoine Tenart <atenart@kernel.org> Reviewed-by: Willem de Bruijn <willemb@google.com> Signed-off-by: David S. Miller <davem@davemloft.net> 
If packets are GROed with fraglist they might be segmented later on and
continue their journey in the stack. In skb_segment_list those skbs can
be reused as-is. This is an issue as their destructor was removed in
skb_gro_receive_list but not the reference to their socket, and then
they can't be orphaned. Fix this by also removing the reference to the
socket.

For example this could be observed,

  kernel BUG at include/linux/skbuff.h:3131!  (skb_orphan)
  RIP: 0010:ip6_rcv_core+0x11bc/0x19a0
  Call Trace:
   ipv6_list_rcv+0x250/0x3f0
   __netif_receive_skb_list_core+0x49d/0x8f0
   netif_receive_skb_list_internal+0x634/0xd40
   napi_complete_done+0x1d2/0x7d0
   gro_cell_poll+0x118/0x1f0

A similar construction is found in skb_gro_receive, apply the same
change there.

Fixes: 5e10da5385d2 ("skbuff: allow 'slow_gro' for skb carring sock reference")
Signed-off-by: Antoine Tenart <atenart@kernel.org>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
net: introduce struct net_hotdata 2024-03-08T05:12:41+00:00 Eric Dumazet edumazet@google.com 2024-03-06T16:00:14+00:00 2658b5a8a4eee5fad378d0bde2f221deacbc58f1 Instead of spreading networking critical fields all over the places, add a custom net_hotdata structure so that we can precisely control its layout. In this first patch, move : - gro_normal_batch used in rx (GRO stack) - offload_base used in rx and tx (GRO and TSO stacks) Signed-off-by: Eric Dumazet <edumazet@google.com> Acked-by: Soheil Hassas Yeganeh <soheil@google.com> Reviewed-by: David Ahern <dsahern@kernel.org> Link: https://lore.kernel.org/r/20240306160031.874438-2-edumazet@google.com Signed-off-by: Jakub Kicinski <kuba@kernel.org> 
Instead of spreading networking critical fields
all over the places, add a custom net_hotdata
structure so that we can precisely control its layout.

In this first patch, move :

- gro_normal_batch used in rx (GRO stack)
- offload_base used in rx and tx (GRO and TSO stacks)

Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Reviewed-by: David Ahern <dsahern@kernel.org>
Link: https://lore.kernel.org/r/20240306160031.874438-2-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
net: gro: enable fast path for more cases 2024-03-05T12:30:11+00:00 Eric Dumazet edumazet@google.com 2024-03-01T19:37:39+00:00 c7583e9f768eeb82f2531c8372584ba89cfade8b 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> 
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>
net: gro: rename skb_gro_header_hard() 2024-03-05T12:30:11+00:00 Eric Dumazet edumazet@google.com 2024-03-01T19:37:37+00:00 93e16ea025d234d0ed01d9dc9c819257a2159bb6 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> 
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>
gro: move the tc_ext comparison to a helper 2023-06-18T17:08:35+00:00 Jakub Kicinski kuba@kernel.org 2023-06-16T20:49:39+00:00 2dc6af8be002d16715485372ce1e02b65faf9283 The double ifdefs (one for the variable declaration and one around the code) are quite aesthetically displeasing. Factor this code out into a helper for easier wrapping. This will become even more ugly when another skb ext comparison is added in the future. The resulting machine code looks the same, the compiler seems to try to use %rax more and some blocks more around but I haven't spotted minor differences. Reviewed-by: Simon Horman <simon.horman@corigine.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: David S. Miller <davem@davemloft.net> 
The double ifdefs (one for the variable declaration and
one around the code) are quite aesthetically displeasing.
Factor this code out into a helper for easier wrapping.

This will become even more ugly when another skb ext
comparison is added in the future.

The resulting machine code looks the same, the compiler
seems to try to use %rax more and some blocks more around
but I haven't spotted minor differences.

Reviewed-by: Simon Horman <simon.horman@corigine.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>