Apply the fix from:
"tcp_cubic: fix spurious HYSTART_DELAY exit upon drop in min RTT"
to the BPF implementation of TCP CUBIC congestion control.
Repeating the commit description here for completeness:
Mirja Kuehlewind reported a bug in Linux TCP CUBIC Hystart, where
Hystart HYSTART_DELAY mechanism can exit Slow Start spuriously on an
ACK when the minimum rtt of a connection goes down. From inspection it
is clear from the existing code that this could happen in an example
like the following:
o The first 8 RTT samples in a round trip are 150ms, resulting in a
curr_rtt of 150ms and a delay_min of 150ms.
o The 9th RTT sample is 100ms. The curr_rtt does not change after the
first 8 samples, so curr_rtt remains 150ms. But delay_min can be
lowered at any time, so delay_min falls to 100ms. The code executes
the HYSTART_DELAY comparison between curr_rtt of 150ms and delay_min
of 100ms, and the curr_rtt is declared far enough above delay_min to
force a (spurious) exit of Slow start.
The fix here is simple: allow every RTT sample in a round trip to
lower the curr_rtt.
Fixes: 6de4a9c430 ("bpf: tcp: Add bpf_cubic example")
Reported-by: Mirja Kuehlewind <mirja.kuehlewind@ericsson.com>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
==================
BPF Selftest Notes
==================
General instructions on running selftests can be found in
`Documentation/bpf/bpf_devel_QA.rst`_.
Additional information about selftest failures are
documented here.
bpf_iter test failures with clang/llvm 10.0.0
=============================================
With clang/llvm 10.0.0, the following two bpf_iter tests failed:
* ``bpf_iter/ipv6_route``
* ``bpf_iter/netlink``
The symptom for ``bpf_iter/ipv6_route`` looks like
.. code-block:: c
2: (79) r8 = *(u64 *)(r1 +8)
...
14: (bf) r2 = r8
15: (0f) r2 += r1
; BPF_SEQ_PRINTF(seq, "%pi6 %02x ", &rt->fib6_dst.addr, rt->fib6_dst.plen);
16: (7b) *(u64 *)(r8 +64) = r2
only read is supported
The symptom for ``bpf_iter/netlink`` looks like
.. code-block:: c
; struct netlink_sock *nlk = ctx->sk;
2: (79) r7 = *(u64 *)(r1 +8)
...
15: (bf) r2 = r7
16: (0f) r2 += r1
; BPF_SEQ_PRINTF(seq, "%pK %-3d ", s, s->sk_protocol);
17: (7b) *(u64 *)(r7 +0) = r2
only read is supported
This is due to a llvm BPF backend bug. The fix
https://reviews.llvm.org/D78466
has been pushed to llvm 10.x release branch and will be
available in 10.0.1. The fix is available in llvm 11.0.0 trunk.
7d21d54d62