Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

sendmsg()/recvmsg() fail when an unknown MSG_* flag argument is provided #162

Closed
pabeni opened this issue Feb 17, 2021 · 0 comments
Closed

sendmsg()/recvmsg() fail when an unknown MSG_* flag argument is provided #162

pabeni opened this issue Feb 17, 2021 · 0 comments
Assignees
@pabeni
Labels
bug
Projects
MPTCP Bugs

Comments

@pabeni
Copy link

pabeni commented Feb 17, 2021 •
edited
Loading

e.g.

recvmsg(10, {msg_namelen=0}, MSG_CMSG_CLOEXEC) = -1 EOPNOTSUPP

plain TCP just ignore unknown MSG_* flag, as in the above example.
@pabeni pabeni added the bug label Feb 17, 2021
@matttbe matttbe added this to Needs triage in MPTCP Bugs via automation Feb 20, 2021
@pabeni pabeni self-assigned this Apr 15, 2021
@pabeni pabeni changed the title sendmsg()/recvmsg() fail when an unknown CMSG argument is provided sendmsg()/recvmsg() fail when an unknown MSG_* flag argument is provided Apr 15, 2021
matttbe pushed a commit that referenced this issue Apr 19, 2021
@matttbe
mptcp: ignore unsupported msg flags
faf109a 
Currently mptcp_sendmsg() fails with EOPNOTSUPP if the
user-space provides some unsupported flag. That is unexpected
and may foul existing applications migrated to MPTCP, which
expect a different behavior.

Change the mentioned function to silently ignore the unsupported
flags except MSG_FASTOPEN. This is the only flags currently not
supported by MPTCP with user-space visible side-effects.

Closes: #162
Reviewed-by: Mat Martineau <mathew.j.martineau@linux.intel.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
MPTCP Bugs automation moved this from Needs triage to Closed Apr 19, 2021
jenkins-tessares pushed a commit that referenced this issue Apr 19, 2021
@jenkins-tessares
mptcp: ignore unsupported msg flags
00a8212 
Currently mptcp_sendmsg() fails with EOPNOTSUPP if the
user-space provides some unsupported flag. That is unexpected
and may foul existing applications migrated to MPTCP, which
expect a different behavior.

Change the mentioned function to silently ignore the unsupported
flags except MSG_FASTOPEN. This is the only flags currently not
supported by MPTCP with user-space visible side-effects.

Closes: #162
Reviewed-by: Mat Martineau <mathew.j.martineau@linux.intel.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
jenkins-tessares pushed a commit that referenced this issue Apr 20, 2021
@jenkins-tessares
mptcp: ignore unsupported msg flags
c4338af 
Currently mptcp_sendmsg() fails with EOPNOTSUPP if the
user-space provides some unsupported flag. That is unexpected
and may foul existing applications migrated to MPTCP, which
expect a different behavior.

Change the mentioned function to silently ignore the unsupported
flags except MSG_FASTOPEN. This is the only flags currently not
supported by MPTCP with user-space visible side-effects.

Closes: #162
Reviewed-by: Mat Martineau <mathew.j.martineau@linux.intel.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
jenkins-tessares pushed a commit that referenced this issue Apr 20, 2021
@jenkins-tessares
mptcp: ignore unsupported msg flags
9c7bbe4 
Currently mptcp_sendmsg() fails with EOPNOTSUPP if the
user-space provides some unsupported flag. That is unexpected
and may foul existing applications migrated to MPTCP, which
expect a different behavior.

Change the mentioned function to silently ignore the unsupported
flags except MSG_FASTOPEN. This is the only flags currently not
supported by MPTCP with user-space visible side-effects.

Closes: #162
Reviewed-by: Mat Martineau <mathew.j.martineau@linux.intel.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
matttbe pushed a commit that referenced this issue Apr 20, 2021
@matttbe
mptcp: ignore unsupported msg flags
b548696 
Currently mptcp_sendmsg() fails with EOPNOTSUPP if the
user-space provides some unsupported flag. That is unexpected
and may foul existing applications migrated to MPTCP, which
expect a different behavior.

Change the mentioned function to silently ignore the unsupported
flags except MSG_FASTOPEN. This is the only flags currently not
supported by MPTCP with user-space visible side-effects.

Closes: #162
Reviewed-by: Mat Martineau <mathew.j.martineau@linux.intel.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
matttbe pushed a commit that referenced this issue Apr 20, 2021
@matttbe
mptcp: ignore unsupported msg flags
e56fb58 
Currently mptcp_sendmsg() fails with EOPNOTSUPP if the
user-space provides some unsupported flag. That is unexpected
and may foul existing applications migrated to MPTCP, which
expect a different behavior.

Change the mentioned function to silently ignore the unsupported
flags except MSG_FASTOPEN. This is the only flags currently not
supported by MPTCP with user-space visible side-effects.

Closes: #162
Reviewed-by: Mat Martineau <mathew.j.martineau@linux.intel.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
jenkins-tessares pushed a commit that referenced this issue Apr 21, 2021
@jenkins-tessares
mptcp: ignore unsupported msg flags
d5c3978 
Currently mptcp_sendmsg() fails with EOPNOTSUPP if the
user-space provides some unsupported flag. That is unexpected
and may foul existing applications migrated to MPTCP, which
expect a different behavior.

Change the mentioned function to silently ignore the unsupported
flags except MSG_FASTOPEN. This is the only flags currently not
supported by MPTCP with user-space visible side-effects.

Closes: #162
Reviewed-by: Mat Martineau <mathew.j.martineau@linux.intel.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
matttbe pushed a commit that referenced this issue Apr 21, 2021
@matttbe
mptcp: ignore unsupported msg flags
f0d610c 
Currently mptcp_sendmsg() fails with EOPNOTSUPP if the
user-space provides some unsupported flag. That is unexpected
and may foul existing applications migrated to MPTCP, which
expect a different behavior.

Change the mentioned function to silently ignore the unsupported
flags except MSG_FASTOPEN. This is the only flags currently not
supported by MPTCP with user-space visible side-effects.

Closes: #162
Reviewed-by: Mat Martineau <mathew.j.martineau@linux.intel.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
matttbe pushed a commit that referenced this issue Apr 21, 2021
@matttbe
mptcp: ignore unsupported msg flags
0dbd6e8 
Currently mptcp_sendmsg() fails with EOPNOTSUPP if the
user-space provides some unsupported flag. That is unexpected
and may foul existing applications migrated to MPTCP, which
expect a different behavior.

Change the mentioned function to silently ignore the unsupported
flags except MSG_FASTOPEN. This is the only flags currently not
supported by MPTCP with user-space visible side-effects.

Closes: #162
Reviewed-by: Mat Martineau <mathew.j.martineau@linux.intel.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
jenkins-tessares pushed a commit that referenced this issue Apr 22, 2021
@jenkins-tessares
mptcp: ignore unsupported msg flags
7445733 
Currently mptcp_sendmsg() fails with EOPNOTSUPP if the
user-space provides some unsupported flag. That is unexpected
and may foul existing applications migrated to MPTCP, which
expect a different behavior.

Change the mentioned function to silently ignore the unsupported
flags except MSG_FASTOPEN. This is the only flags currently not
supported by MPTCP with user-space visible side-effects.

Closes: #162
Reviewed-by: Mat Martineau <mathew.j.martineau@linux.intel.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
matttbe pushed a commit that referenced this issue Apr 22, 2021
@matttbe
mptcp: ignore unsupported msg flags
4e61c73 
Currently mptcp_sendmsg() fails with EOPNOTSUPP if the
user-space provides some unsupported flag. That is unexpected
and may foul existing applications migrated to MPTCP, which
expect a different behavior.

Change the mentioned function to silently ignore the unsupported
flags except MSG_FASTOPEN. This is the only flags currently not
supported by MPTCP with user-space visible side-effects.

Closes: #162
Reviewed-by: Mat Martineau <mathew.j.martineau@linux.intel.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
matttbe pushed a commit that referenced this issue Apr 22, 2021
@matttbe
mptcp: ignore unsupported msg flags
b2c639b 
Currently mptcp_sendmsg() fails with EOPNOTSUPP if the
user-space provides some unsupported flag. That is unexpected
and may foul existing applications migrated to MPTCP, which
expect a different behavior.

Change the mentioned function to silently ignore the unsupported
flags except MSG_FASTOPEN. This is the only flags currently not
supported by MPTCP with user-space visible side-effects.

Closes: #162
Reviewed-by: Mat Martineau <mathew.j.martineau@linux.intel.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
matttbe pushed a commit that referenced this issue Apr 22, 2021
@matttbe
mptcp: ignore unsupported msg flags
138086e 
Currently mptcp_sendmsg() fails with EOPNOTSUPP if the
user-space provides some unsupported flag. That is unexpected
and may foul existing applications migrated to MPTCP, which
expect a different behavior.

Change the mentioned function to silently ignore the unsupported
flags except MSG_FASTOPEN. This is the only flags currently not
supported by MPTCP with user-space visible side-effects.

Closes: #162
Reviewed-by: Mat Martineau <mathew.j.martineau@linux.intel.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
jenkins-tessares pushed a commit that referenced this issue Apr 23, 2021
@jenkins-tessares
mptcp: ignore unsupported msg flags
7bd491b 
Currently mptcp_sendmsg() fails with EOPNOTSUPP if the
user-space provides some unsupported flag. That is unexpected
and may foul existing applications migrated to MPTCP, which
expect a different behavior.

Change the mentioned function to silently ignore the unsupported
flags except MSG_FASTOPEN. This is the only flags currently not
supported by MPTCP with user-space visible side-effects.

Closes: #162
Reviewed-by: Mat Martineau <mathew.j.martineau@linux.intel.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
matttbe pushed a commit that referenced this issue Apr 23, 2021
@matttbe
mptcp: ignore unsupported msg flags
745334a 
Currently mptcp_sendmsg() fails with EOPNOTSUPP if the
user-space provides some unsupported flag. That is unexpected
and may foul existing applications migrated to MPTCP, which
expect a different behavior.

Change the mentioned function to silently ignore the unsupported
flags except MSG_FASTOPEN. This is the only flags currently not
supported by MPTCP with user-space visible side-effects.

Closes: #162
Reviewed-by: Mat Martineau <mathew.j.martineau@linux.intel.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
fengguang pushed a commit to 0day-ci/linux that referenced this issue Apr 23, 2021
@intel-lab-lkp
mptcp: ignore unsupported msg flags
4aefac9 
Currently mptcp_sendmsg() fails with EOPNOTSUPP if the
user-space provides some unsupported flag. That is unexpected
and may foul existing applications migrated to MPTCP, which
expect a different behavior.

Change the mentioned function to silently ignore the unsupported
flags except MSG_FASTOPEN. This is the only flags currently not
supported by MPTCP with user-space visible side-effects.

Closes: multipath-tcp/mptcp_net-next#162
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Mat Martineau <mathew.j.martineau@linux.intel.com>
fengguang pushed a commit to 0day-ci/linux that referenced this issue Apr 24, 2021
@davem330
mptcp: ignore unsupported msg flags
987858e 
Currently mptcp_sendmsg() fails with EOPNOTSUPP if the
user-space provides some unsupported flag. That is unexpected
and may foul existing applications migrated to MPTCP, which
expect a different behavior.

Change the mentioned function to silently ignore the unsupported
flags except MSG_FASTOPEN. This is the only flags currently not
supported by MPTCP with user-space visible side-effects.

Closes: multipath-tcp/mptcp_net-next#162
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Mat Martineau <mathew.j.martineau@linux.intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
geliangtang pushed a commit to geliangtang/mptcp_net-next that referenced this issue May 7, 2024
mptcp: fix the max value of scaling_ratio
a06be60 
BPF tests fail sometimes (a probability of approximately 1%) with
"bytes != total_bytes" errors:

 test_burst:PASS:open_and_load:burst 0 nsec
 test_bpf_sched:PASS:Scheduler name too long 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 create_netns:PASS:ip netns add mptcp_ns 0 nsec
 create_netns:PASS:ip -net mptcp_ns link set dev lo up 0 nsec
 sched_init:PASS:create_netns 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link add veth1 type veth peer name
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.1/24 dev veth1 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth1 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.2/24 dev veth2 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth2 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns mptcp endpoint add 10.0.1.2 subflow
 sched_init:PASS:endpoint_init 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 (network_helpers.c:613: errno: Resource temporarily unavailable) \
                                        send 5608500 expected 10485760
 (network_helpers.c:661: errno: None) recv 2755984 expected 10485760
 (network_helpers.c:669: errno: None) Failed in thread_ret -11
 send_data_and_verify:FAIL:send_recv_data unexpected error: -4 (errno 0)
 multipath-tcp#162/9   mptcp/burst:FAIL
 multipath-tcp#162     mptcp:FAIL

In this case, mptcp_recvmsg() gets EAGAIN errors. This issue introduces
by commit b8dc6d6 ("mptcp: fix rcv buffer auto-tuning"). The max
value of scaling_ratio should be TCP_DEFAULT_SCALING_RATIO (128), not
U8_MAX (255). Otherwise, scaling_ratio is assigned to a too high value.

Fixes: b8dc6d6 ("mptcp: fix rcv buffer auto-tuning")
Closes: multipath-tcp#487
Signed-off-by: Geliang Tang <tanggeliang@kylinos.cn>
geliangtang pushed a commit to geliangtang/mptcp_net-next that referenced this issue May 7, 2024
mptcp: fix the max value of scaling_ratio
45b6cb9 
BPF tests fail sometimes (a probability of approximately 1%) with
"bytes != total_bytes" errors:

 test_burst:PASS:open_and_load:burst 0 nsec
 test_bpf_sched:PASS:Scheduler name too long 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 create_netns:PASS:ip netns add mptcp_ns 0 nsec
 create_netns:PASS:ip -net mptcp_ns link set dev lo up 0 nsec
 sched_init:PASS:create_netns 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link add veth1 type veth peer name
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.1/24 dev veth1 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth1 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.2/24 dev veth2 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth2 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns mptcp endpoint add 10.0.1.2 subflow
 sched_init:PASS:endpoint_init 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 (network_helpers.c:613: errno: Resource temporarily unavailable) \
                                        send 5608500 expected 10485760
 (network_helpers.c:661: errno: None) recv 2755984 expected 10485760
 (network_helpers.c:669: errno: None) Failed in thread_ret -11
 send_data_and_verify:FAIL:send_recv_data unexpected error: -4 (errno 0)
 multipath-tcp#162/9   mptcp/burst:FAIL
 multipath-tcp#162     mptcp:FAIL

In this case, mptcp_recvmsg() gets EAGAIN errors. This issue introduces
by commit b8dc6d6 ("mptcp: fix rcv buffer auto-tuning"). The max
value of scaling_ratio should be TCP_DEFAULT_SCALING_RATIO (128), not
U8_MAX (255). Otherwise, scaling_ratio is assigned to a too high value.

Fixes: b8dc6d6 ("mptcp: fix rcv buffer auto-tuning")
Closes: multipath-tcp#487
Signed-off-by: Geliang Tang <tanggeliang@kylinos.cn>
geliangtang pushed a commit to geliangtang/mptcp_net-next that referenced this issue May 10, 2024
mptcp: fix the max value of scaling_ratio
66cfade 
BPF tests fail sometimes (a probability of approximately 1%) with
"bytes != total_bytes" errors:

 test_burst:PASS:open_and_load:burst 0 nsec
 test_bpf_sched:PASS:Scheduler name too long 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 create_netns:PASS:ip netns add mptcp_ns 0 nsec
 create_netns:PASS:ip -net mptcp_ns link set dev lo up 0 nsec
 sched_init:PASS:create_netns 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link add veth1 type veth peer name
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.1/24 dev veth1 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth1 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.2/24 dev veth2 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth2 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns mptcp endpoint add 10.0.1.2 subflow
 sched_init:PASS:endpoint_init 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 (network_helpers.c:613: errno: Resource temporarily unavailable) \
                                        send 5608500 expected 10485760
 (network_helpers.c:661: errno: None) recv 2755984 expected 10485760
 (network_helpers.c:669: errno: None) Failed in thread_ret -11
 send_data_and_verify:FAIL:send_recv_data unexpected error: -4 (errno 0)
 multipath-tcp#162/9   mptcp/burst:FAIL
 multipath-tcp#162     mptcp:FAIL

In this case, mptcp_recvmsg() gets EAGAIN errors. This issue introduces
by commit b8dc6d6 ("mptcp: fix rcv buffer auto-tuning"). The max
value of scaling_ratio should be TCP_DEFAULT_SCALING_RATIO (128), not
U8_MAX (255). Otherwise, scaling_ratio is assigned to a too high value.

Fixes: b8dc6d6 ("mptcp: fix rcv buffer auto-tuning")
Closes: multipath-tcp#487
Signed-off-by: Geliang Tang <tanggeliang@kylinos.cn>
geliangtang pushed a commit to geliangtang/mptcp_net-next that referenced this issue May 10, 2024
mptcp: fix the max value of scaling_ratio
e8e0610 
BPF tests fail sometimes (a probability of approximately 1%) with
"bytes != total_bytes" errors:

 test_burst:PASS:open_and_load:burst 0 nsec
 test_bpf_sched:PASS:Scheduler name too long 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 create_netns:PASS:ip netns add mptcp_ns 0 nsec
 create_netns:PASS:ip -net mptcp_ns link set dev lo up 0 nsec
 sched_init:PASS:create_netns 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link add veth1 type veth peer name
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.1/24 dev veth1 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth1 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.2/24 dev veth2 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth2 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns mptcp endpoint add 10.0.1.2 subflow
 sched_init:PASS:endpoint_init 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 (network_helpers.c:613: errno: Resource temporarily unavailable) \
                                        send 5608500 expected 10485760
 (network_helpers.c:661: errno: None) recv 2755984 expected 10485760
 (network_helpers.c:669: errno: None) Failed in thread_ret -11
 send_data_and_verify:FAIL:send_recv_data unexpected error: -4 (errno 0)
 multipath-tcp#162/9   mptcp/burst:FAIL
 multipath-tcp#162     mptcp:FAIL

In this case, mptcp_recvmsg() gets EAGAIN errors. This issue introduces
by commit b8dc6d6 ("mptcp: fix rcv buffer auto-tuning"). The max
value of scaling_ratio should be TCP_DEFAULT_SCALING_RATIO (128), not
U8_MAX (255). Otherwise, scaling_ratio is assigned to a too high value.

Fixes: b8dc6d6 ("mptcp: fix rcv buffer auto-tuning")
Closes: multipath-tcp#487
Signed-off-by: Geliang Tang <tanggeliang@kylinos.cn>
geliangtang pushed a commit to geliangtang/mptcp_net-next that referenced this issue May 11, 2024
mptcp: fix the max value of scaling_ratio
7ad75d1 
BPF tests fail sometimes (a probability of approximately 1%) with
"bytes != total_bytes" errors:

 test_burst:PASS:open_and_load:burst 0 nsec
 test_bpf_sched:PASS:Scheduler name too long 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 create_netns:PASS:ip netns add mptcp_ns 0 nsec
 create_netns:PASS:ip -net mptcp_ns link set dev lo up 0 nsec
 sched_init:PASS:create_netns 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link add veth1 type veth peer name
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.1/24 dev veth1 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth1 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.2/24 dev veth2 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth2 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns mptcp endpoint add 10.0.1.2 subflow
 sched_init:PASS:endpoint_init 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 (network_helpers.c:613: errno: Resource temporarily unavailable) \
                                        send 5608500 expected 10485760
 (network_helpers.c:661: errno: None) recv 2755984 expected 10485760
 (network_helpers.c:669: errno: None) Failed in thread_ret -11
 send_data_and_verify:FAIL:send_recv_data unexpected error: -4 (errno 0)
 multipath-tcp#162/9   mptcp/burst:FAIL
 multipath-tcp#162     mptcp:FAIL

In this case, mptcp_recvmsg() gets EAGAIN errors. This issue introduces
by commit b8dc6d6 ("mptcp: fix rcv buffer auto-tuning"). The max
value of scaling_ratio should be TCP_DEFAULT_SCALING_RATIO (128), not
U8_MAX (255). Otherwise, scaling_ratio is assigned to a too high value.

Fixes: b8dc6d6 ("mptcp: fix rcv buffer auto-tuning")
Closes: multipath-tcp#487
Signed-off-by: Geliang Tang <tanggeliang@kylinos.cn>
geliangtang pushed a commit to geliangtang/mptcp_net-next that referenced this issue May 11, 2024
mptcp: fix the max value of scaling_ratio
5216df7 
BPF tests fail sometimes (a probability of approximately 1%) with
"bytes != total_bytes" errors:

 test_burst:PASS:open_and_load:burst 0 nsec
 test_bpf_sched:PASS:Scheduler name too long 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 create_netns:PASS:ip netns add mptcp_ns 0 nsec
 create_netns:PASS:ip -net mptcp_ns link set dev lo up 0 nsec
 sched_init:PASS:create_netns 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link add veth1 type veth peer name
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.1/24 dev veth1 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth1 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.2/24 dev veth2 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth2 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns mptcp endpoint add 10.0.1.2 subflow
 sched_init:PASS:endpoint_init 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 (network_helpers.c:613: errno: Resource temporarily unavailable) \
                                        send 5608500 expected 10485760
 (network_helpers.c:661: errno: None) recv 2755984 expected 10485760
 (network_helpers.c:669: errno: None) Failed in thread_ret -11
 send_data_and_verify:FAIL:send_recv_data unexpected error: -4 (errno 0)
 multipath-tcp#162/9   mptcp/burst:FAIL
 multipath-tcp#162     mptcp:FAIL

In this case, mptcp_recvmsg() gets EAGAIN errors. This issue introduces
by commit b8dc6d6 ("mptcp: fix rcv buffer auto-tuning"). The max
value of scaling_ratio should be TCP_DEFAULT_SCALING_RATIO (128), not
U8_MAX (255). Otherwise, scaling_ratio is assigned to a too high value.

Fixes: b8dc6d6 ("mptcp: fix rcv buffer auto-tuning")
Closes: multipath-tcp#487
Signed-off-by: Geliang Tang <tanggeliang@kylinos.cn>
geliangtang pushed a commit to geliangtang/mptcp_net-next that referenced this issue May 12, 2024
mptcp: fix the max value of scaling_ratio
ddc10a6 
BPF tests fail sometimes (a probability of approximately 1%) with
"bytes != total_bytes" errors:

 test_burst:PASS:open_and_load:burst 0 nsec
 test_bpf_sched:PASS:Scheduler name too long 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 create_netns:PASS:ip netns add mptcp_ns 0 nsec
 create_netns:PASS:ip -net mptcp_ns link set dev lo up 0 nsec
 sched_init:PASS:create_netns 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link add veth1 type veth peer name
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.1/24 dev veth1 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth1 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.2/24 dev veth2 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth2 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns mptcp endpoint add 10.0.1.2 subflow
 sched_init:PASS:endpoint_init 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 (network_helpers.c:613: errno: Resource temporarily unavailable) \
                                        send 5608500 expected 10485760
 (network_helpers.c:661: errno: None) recv 2755984 expected 10485760
 (network_helpers.c:669: errno: None) Failed in thread_ret -11
 send_data_and_verify:FAIL:send_recv_data unexpected error: -4 (errno 0)
 multipath-tcp#162/9   mptcp/burst:FAIL
 multipath-tcp#162     mptcp:FAIL

In this case, mptcp_recvmsg() gets EAGAIN errors. This issue introduces
by commit b8dc6d6 ("mptcp: fix rcv buffer auto-tuning"). The max
value of scaling_ratio should be TCP_DEFAULT_SCALING_RATIO (128), not
U8_MAX (255). Otherwise, scaling_ratio is assigned to a too high value.

Fixes: b8dc6d6 ("mptcp: fix rcv buffer auto-tuning")
Closes: multipath-tcp#487
Signed-off-by: Geliang Tang <tanggeliang@kylinos.cn>
geliangtang pushed a commit to geliangtang/mptcp_net-next that referenced this issue May 13, 2024
mptcp: fix the max value of scaling_ratio
965af6e 
BPF tests fail sometimes (a probability of approximately 1%) with
"bytes != total_bytes" errors:

 test_burst:PASS:open_and_load:burst 0 nsec
 test_bpf_sched:PASS:Scheduler name too long 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 create_netns:PASS:ip netns add mptcp_ns 0 nsec
 create_netns:PASS:ip -net mptcp_ns link set dev lo up 0 nsec
 sched_init:PASS:create_netns 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link add veth1 type veth peer name
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.1/24 dev veth1 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth1 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.2/24 dev veth2 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth2 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns mptcp endpoint add 10.0.1.2 subflow
 sched_init:PASS:endpoint_init 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 (network_helpers.c:613: errno: Resource temporarily unavailable) \
                                        send 5608500 expected 10485760
 (network_helpers.c:661: errno: None) recv 2755984 expected 10485760
 (network_helpers.c:669: errno: None) Failed in thread_ret -11
 send_data_and_verify:FAIL:send_recv_data unexpected error: -4 (errno 0)
 multipath-tcp#162/9   mptcp/burst:FAIL
 multipath-tcp#162     mptcp:FAIL

In this case, mptcp_recvmsg() gets EAGAIN errors. This issue introduces
by commit b8dc6d6 ("mptcp: fix rcv buffer auto-tuning"). The max
value of scaling_ratio should be TCP_DEFAULT_SCALING_RATIO (128), not
U8_MAX (255). Otherwise, scaling_ratio is assigned to a too high value.

Fixes: b8dc6d6 ("mptcp: fix rcv buffer auto-tuning")
Closes: multipath-tcp#487
Signed-off-by: Geliang Tang <tanggeliang@kylinos.cn>
geliangtang pushed a commit to geliangtang/mptcp_net-next that referenced this issue May 13, 2024
mptcp: fix the max value of scaling_ratio
af0284b 
BPF tests fail sometimes (a probability of approximately 1%) with
"bytes != total_bytes" errors:

 test_burst:PASS:open_and_load:burst 0 nsec
 test_bpf_sched:PASS:Scheduler name too long 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 create_netns:PASS:ip netns add mptcp_ns 0 nsec
 create_netns:PASS:ip -net mptcp_ns link set dev lo up 0 nsec
 sched_init:PASS:create_netns 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link add veth1 type veth peer name
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.1/24 dev veth1 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth1 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.2/24 dev veth2 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth2 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns mptcp endpoint add 10.0.1.2 subflow
 sched_init:PASS:endpoint_init 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 (network_helpers.c:613: errno: Resource temporarily unavailable) \
                                        send 5608500 expected 10485760
 (network_helpers.c:661: errno: None) recv 2755984 expected 10485760
 (network_helpers.c:669: errno: None) Failed in thread_ret -11
 send_data_and_verify:FAIL:send_recv_data unexpected error: -4 (errno 0)
 multipath-tcp#162/9   mptcp/burst:FAIL
 multipath-tcp#162     mptcp:FAIL

In this case, mptcp_recvmsg() gets EAGAIN errors. This issue introduces
by commit b8dc6d6 ("mptcp: fix rcv buffer auto-tuning"). The max
value of scaling_ratio should be TCP_DEFAULT_SCALING_RATIO (128), not
U8_MAX (255). Otherwise, scaling_ratio is assigned to a too high value.

Fixes: b8dc6d6 ("mptcp: fix rcv buffer auto-tuning")
Closes: multipath-tcp#487
Signed-off-by: Geliang Tang <tanggeliang@kylinos.cn>
geliangtang pushed a commit to geliangtang/mptcp_net-next that referenced this issue May 14, 2024
mptcp: fix the max value of scaling_ratio
b4a7b47 
BPF tests fail sometimes (a probability of approximately 1%) with
"bytes != total_bytes" errors:

 test_burst:PASS:open_and_load:burst 0 nsec
 test_bpf_sched:PASS:Scheduler name too long 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 create_netns:PASS:ip netns add mptcp_ns 0 nsec
 create_netns:PASS:ip -net mptcp_ns link set dev lo up 0 nsec
 sched_init:PASS:create_netns 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link add veth1 type veth peer name
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.1/24 dev veth1 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth1 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.2/24 dev veth2 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth2 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns mptcp endpoint add 10.0.1.2 subflow
 sched_init:PASS:endpoint_init 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 (network_helpers.c:613: errno: Resource temporarily unavailable) \
                                        send 5608500 expected 10485760
 (network_helpers.c:661: errno: None) recv 2755984 expected 10485760
 (network_helpers.c:669: errno: None) Failed in thread_ret -11
 send_data_and_verify:FAIL:send_recv_data unexpected error: -4 (errno 0)
 multipath-tcp#162/9   mptcp/burst:FAIL
 multipath-tcp#162     mptcp:FAIL

In this case, mptcp_recvmsg() gets EAGAIN errors. This issue introduces
by commit b8dc6d6 ("mptcp: fix rcv buffer auto-tuning"). The max
value of scaling_ratio should be TCP_DEFAULT_SCALING_RATIO (128), not
U8_MAX (255). Otherwise, scaling_ratio is assigned to a too high value.

Fixes: b8dc6d6 ("mptcp: fix rcv buffer auto-tuning")
Closes: multipath-tcp#487
Signed-off-by: Geliang Tang <tanggeliang@kylinos.cn>
geliangtang pushed a commit to geliangtang/mptcp_net-next that referenced this issue May 15, 2024
mptcp: fix the max value of scaling_ratio
d7fbef7 
BPF tests fail sometimes (a probability of approximately 1%) with
"bytes != total_bytes" errors:

 test_burst:PASS:open_and_load:burst 0 nsec
 test_bpf_sched:PASS:Scheduler name too long 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 create_netns:PASS:ip netns add mptcp_ns 0 nsec
 create_netns:PASS:ip -net mptcp_ns link set dev lo up 0 nsec
 sched_init:PASS:create_netns 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link add veth1 type veth peer name
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.1/24 dev veth1 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth1 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.2/24 dev veth2 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth2 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns mptcp endpoint add 10.0.1.2 subflow
 sched_init:PASS:endpoint_init 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 (network_helpers.c:613: errno: Resource temporarily unavailable) \
                                        send 5608500 expected 10485760
 (network_helpers.c:661: errno: None) recv 2755984 expected 10485760
 (network_helpers.c:669: errno: None) Failed in thread_ret -11
 send_data_and_verify:FAIL:send_recv_data unexpected error: -4 (errno 0)
 multipath-tcp#162/9   mptcp/burst:FAIL
 multipath-tcp#162     mptcp:FAIL

In this case, mptcp_recvmsg() gets EAGAIN errors. This issue introduces
by commit b8dc6d6 ("mptcp: fix rcv buffer auto-tuning"). The max
value of scaling_ratio should be TCP_DEFAULT_SCALING_RATIO (128), not
U8_MAX (255). Otherwise, scaling_ratio is assigned to a too high value.

Fixes: b8dc6d6 ("mptcp: fix rcv buffer auto-tuning")
Closes: multipath-tcp#487
Signed-off-by: Geliang Tang <tanggeliang@kylinos.cn>
geliangtang pushed a commit to geliangtang/mptcp_net-next that referenced this issue May 15, 2024
mptcp: fix the max value of scaling_ratio
2de434d 
BPF tests fail sometimes (a probability of approximately 1%) with
"bytes != total_bytes" errors:

 test_burst:PASS:open_and_load:burst 0 nsec
 test_bpf_sched:PASS:Scheduler name too long 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 create_netns:PASS:ip netns add mptcp_ns 0 nsec
 create_netns:PASS:ip -net mptcp_ns link set dev lo up 0 nsec
 sched_init:PASS:create_netns 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link add veth1 type veth peer name
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.1/24 dev veth1 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth1 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.2/24 dev veth2 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth2 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns mptcp endpoint add 10.0.1.2 subflow
 sched_init:PASS:endpoint_init 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 (network_helpers.c:613: errno: Resource temporarily unavailable) \
                                        send 5608500 expected 10485760
 (network_helpers.c:661: errno: None) recv 2755984 expected 10485760
 (network_helpers.c:669: errno: None) Failed in thread_ret -11
 send_data_and_verify:FAIL:send_recv_data unexpected error: -4 (errno 0)
 multipath-tcp#162/9   mptcp/burst:FAIL
 multipath-tcp#162     mptcp:FAIL

In this case, mptcp_recvmsg() gets EAGAIN errors. This issue introduces
by commit b8dc6d6 ("mptcp: fix rcv buffer auto-tuning"). The max
value of scaling_ratio should be TCP_DEFAULT_SCALING_RATIO (128), not
U8_MAX (255). Otherwise, scaling_ratio is assigned to a too high value.

Fixes: b8dc6d6 ("mptcp: fix rcv buffer auto-tuning")
Closes: multipath-tcp#487
Signed-off-by: Geliang Tang <tanggeliang@kylinos.cn>
geliangtang pushed a commit to geliangtang/mptcp_net-next that referenced this issue May 18, 2024
@geliangtang
mptcp: fix the max value of scaling_ratio
6700eaf 
BPF tests fail sometimes (a probability of approximately 1%) with
"bytes != total_bytes" errors:

 test_burst:PASS:open_and_load:burst 0 nsec
 test_bpf_sched:PASS:Scheduler name too long 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 create_netns:PASS:ip netns add mptcp_ns 0 nsec
 create_netns:PASS:ip -net mptcp_ns link set dev lo up 0 nsec
 sched_init:PASS:create_netns 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link add veth1 type veth peer name
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.1/24 dev veth1 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth1 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.2/24 dev veth2 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth2 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns mptcp endpoint add 10.0.1.2 subflow
 sched_init:PASS:endpoint_init 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 (network_helpers.c:613: errno: Resource temporarily unavailable) \
                                        send 5608500 expected 10485760
 (network_helpers.c:661: errno: None) recv 2755984 expected 10485760
 (network_helpers.c:669: errno: None) Failed in thread_ret -11
 send_data_and_verify:FAIL:send_recv_data unexpected error: -4 (errno 0)
 multipath-tcp#162/9   mptcp/burst:FAIL
 multipath-tcp#162     mptcp:FAIL

In this case, mptcp_recvmsg() gets EAGAIN errors. This issue introduces
by commit b8dc6d6 ("mptcp: fix rcv buffer auto-tuning"). The max
value of scaling_ratio should be TCP_DEFAULT_SCALING_RATIO (128), not
U8_MAX (255). Otherwise, scaling_ratio is assigned to a too high value.

Fixes: b8dc6d6 ("mptcp: fix rcv buffer auto-tuning")
Closes: multipath-tcp#487
Signed-off-by: Geliang Tang <tanggeliang@kylinos.cn>
geliangtang pushed a commit to geliangtang/mptcp_net-next that referenced this issue May 22, 2024
mptcp: fix the max value of scaling_ratio
81f74a9 
BPF tests fail sometimes (a probability of approximately 1%) with
"bytes != total_bytes" errors:

 test_burst:PASS:open_and_load:burst 0 nsec
 test_bpf_sched:PASS:Scheduler name too long 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 create_netns:PASS:ip netns add mptcp_ns 0 nsec
 create_netns:PASS:ip -net mptcp_ns link set dev lo up 0 nsec
 sched_init:PASS:create_netns 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link add veth1 type veth peer name
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.1/24 dev veth1 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth1 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.2/24 dev veth2 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth2 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns mptcp endpoint add 10.0.1.2 subflow
 sched_init:PASS:endpoint_init 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 (network_helpers.c:613: errno: Resource temporarily unavailable) \
                                        send 5608500 expected 10485760
 (network_helpers.c:661: errno: None) recv 2755984 expected 10485760
 (network_helpers.c:669: errno: None) Failed in thread_ret -11
 send_data_and_verify:FAIL:send_recv_data unexpected error: -4 (errno 0)
 multipath-tcp#162/9   mptcp/burst:FAIL
 multipath-tcp#162     mptcp:FAIL

In this case, mptcp_recvmsg() gets EAGAIN errors. This issue introduces
by commit b8dc6d6 ("mptcp: fix rcv buffer auto-tuning"). The max
value of scaling_ratio should be TCP_DEFAULT_SCALING_RATIO (128), not
U8_MAX (255). Otherwise, scaling_ratio is assigned to a too high value.

Fixes: b8dc6d6 ("mptcp: fix rcv buffer auto-tuning")
Closes: multipath-tcp#487
Signed-off-by: Geliang Tang <tanggeliang@kylinos.cn>
geliangtang pushed a commit to geliangtang/mptcp_net-next that referenced this issue May 23, 2024
mptcp: fix the max value of scaling_ratio
ed48e84 
BPF tests fail sometimes (a probability of approximately 1%) with
"bytes != total_bytes" errors:

 test_burst:PASS:open_and_load:burst 0 nsec
 test_bpf_sched:PASS:Scheduler name too long 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 create_netns:PASS:ip netns add mptcp_ns 0 nsec
 create_netns:PASS:ip -net mptcp_ns link set dev lo up 0 nsec
 sched_init:PASS:create_netns 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link add veth1 type veth peer name
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.1/24 dev veth1 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth1 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.2/24 dev veth2 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth2 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns mptcp endpoint add 10.0.1.2 subflow
 sched_init:PASS:endpoint_init 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 (network_helpers.c:613: errno: Resource temporarily unavailable) \
                                        send 5608500 expected 10485760
 (network_helpers.c:661: errno: None) recv 2755984 expected 10485760
 (network_helpers.c:669: errno: None) Failed in thread_ret -11
 send_data_and_verify:FAIL:send_recv_data unexpected error: -4 (errno 0)
 multipath-tcp#162/9   mptcp/burst:FAIL
 multipath-tcp#162     mptcp:FAIL

In this case, mptcp_recvmsg() gets EAGAIN errors. This issue introduces
by commit b8dc6d6 ("mptcp: fix rcv buffer auto-tuning"). The max
value of scaling_ratio should be TCP_DEFAULT_SCALING_RATIO (128), not
U8_MAX (255). Otherwise, scaling_ratio is assigned to a too high value.

Fixes: b8dc6d6 ("mptcp: fix rcv buffer auto-tuning")
Closes: multipath-tcp#487
Signed-off-by: Geliang Tang <tanggeliang@kylinos.cn>
geliangtang pushed a commit to geliangtang/mptcp_net-next that referenced this issue May 23, 2024
mptcp: fix the max value of scaling_ratio
746d4aa 
BPF tests fail sometimes (a probability of approximately 1%) with
"bytes != total_bytes" errors:

 test_burst:PASS:open_and_load:burst 0 nsec
 test_bpf_sched:PASS:Scheduler name too long 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 create_netns:PASS:ip netns add mptcp_ns 0 nsec
 create_netns:PASS:ip -net mptcp_ns link set dev lo up 0 nsec
 sched_init:PASS:create_netns 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link add veth1 type veth peer name
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.1/24 dev veth1 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth1 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.2/24 dev veth2 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth2 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns mptcp endpoint add 10.0.1.2 subflow
 sched_init:PASS:endpoint_init 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 (network_helpers.c:613: errno: Resource temporarily unavailable) \
                                        send 5608500 expected 10485760
 (network_helpers.c:661: errno: None) recv 2755984 expected 10485760
 (network_helpers.c:669: errno: None) Failed in thread_ret -11
 send_data_and_verify:FAIL:send_recv_data unexpected error: -4 (errno 0)
 multipath-tcp#162/9   mptcp/burst:FAIL
 multipath-tcp#162     mptcp:FAIL

In this case, mptcp_recvmsg() gets EAGAIN errors. This issue introduces
by commit b8dc6d6 ("mptcp: fix rcv buffer auto-tuning"). The max
value of scaling_ratio should be TCP_DEFAULT_SCALING_RATIO (128), not
U8_MAX (255). Otherwise, scaling_ratio is assigned to a too high value.

Fixes: b8dc6d6 ("mptcp: fix rcv buffer auto-tuning")
Closes: multipath-tcp#487
Signed-off-by: Geliang Tang <tanggeliang@kylinos.cn>
geliangtang pushed a commit to geliangtang/mptcp_net-next that referenced this issue May 24, 2024
mptcp: fix the max value of scaling_ratio
4bdb90c 
BPF tests fail sometimes (a probability of approximately 1%) with
"bytes != total_bytes" errors:

 test_burst:PASS:open_and_load:burst 0 nsec
 test_bpf_sched:PASS:Scheduler name too long 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 create_netns:PASS:ip netns add mptcp_ns 0 nsec
 create_netns:PASS:ip -net mptcp_ns link set dev lo up 0 nsec
 sched_init:PASS:create_netns 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link add veth1 type veth peer name
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.1/24 dev veth1 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth1 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.2/24 dev veth2 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth2 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns mptcp endpoint add 10.0.1.2 subflow
 sched_init:PASS:endpoint_init 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 (network_helpers.c:613: errno: Resource temporarily unavailable) \
                                        send 5608500 expected 10485760
 (network_helpers.c:661: errno: None) recv 2755984 expected 10485760
 (network_helpers.c:669: errno: None) Failed in thread_ret -11
 send_data_and_verify:FAIL:send_recv_data unexpected error: -4 (errno 0)
 multipath-tcp#162/9   mptcp/burst:FAIL
 multipath-tcp#162     mptcp:FAIL

In this case, mptcp_recvmsg() gets EAGAIN errors. This issue introduces
by commit b8dc6d6 ("mptcp: fix rcv buffer auto-tuning"). The max
value of scaling_ratio should be TCP_DEFAULT_SCALING_RATIO (128), not
U8_MAX (255). Otherwise, scaling_ratio is assigned to a too high value.

Fixes: b8dc6d6 ("mptcp: fix rcv buffer auto-tuning")
Closes: multipath-tcp#487
Signed-off-by: Geliang Tang <tanggeliang@kylinos.cn>
geliangtang pushed a commit to geliangtang/mptcp_net-next that referenced this issue May 27, 2024
mptcp: fix the max value of scaling_ratio
23531a8 
BPF tests fail sometimes (a probability of approximately 1%) with
"bytes != total_bytes" errors:

 test_burst:PASS:open_and_load:burst 0 nsec
 test_bpf_sched:PASS:Scheduler name too long 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 create_netns:PASS:ip netns add mptcp_ns 0 nsec
 create_netns:PASS:ip -net mptcp_ns link set dev lo up 0 nsec
 sched_init:PASS:create_netns 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link add veth1 type veth peer name
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.1/24 dev veth1 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth1 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.2/24 dev veth2 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth2 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns mptcp endpoint add 10.0.1.2 subflow
 sched_init:PASS:endpoint_init 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 (network_helpers.c:613: errno: Resource temporarily unavailable) \
                                        send 5608500 expected 10485760
 (network_helpers.c:661: errno: None) recv 2755984 expected 10485760
 (network_helpers.c:669: errno: None) Failed in thread_ret -11
 send_data_and_verify:FAIL:send_recv_data unexpected error: -4 (errno 0)
 multipath-tcp#162/9   mptcp/burst:FAIL
 multipath-tcp#162     mptcp:FAIL

In this case, mptcp_recvmsg() gets EAGAIN errors. This issue introduces
by commit b8dc6d6 ("mptcp: fix rcv buffer auto-tuning"). The max
value of scaling_ratio should be TCP_DEFAULT_SCALING_RATIO (128), not
U8_MAX (255). Otherwise, scaling_ratio is assigned to a too high value.

Fixes: b8dc6d6 ("mptcp: fix rcv buffer auto-tuning")
Closes: multipath-tcp#487
Signed-off-by: Geliang Tang <tanggeliang@kylinos.cn>
geliangtang pushed a commit to geliangtang/mptcp_net-next that referenced this issue May 29, 2024
mptcp: fix the max value of scaling_ratio
bb7579a 
BPF tests fail sometimes (a probability of approximately 1%) with
"bytes != total_bytes" errors:

 test_burst:PASS:open_and_load:burst 0 nsec
 test_bpf_sched:PASS:Scheduler name too long 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 create_netns:PASS:ip netns add mptcp_ns 0 nsec
 create_netns:PASS:ip -net mptcp_ns link set dev lo up 0 nsec
 sched_init:PASS:create_netns 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link add veth1 type veth peer name
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.1/24 dev veth1 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth1 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.2/24 dev veth2 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth2 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns mptcp endpoint add 10.0.1.2 subflow
 sched_init:PASS:endpoint_init 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 (network_helpers.c:613: errno: Resource temporarily unavailable) \
                                        send 5608500 expected 10485760
 (network_helpers.c:661: errno: None) recv 2755984 expected 10485760
 (network_helpers.c:669: errno: None) Failed in thread_ret -11
 send_data_and_verify:FAIL:send_recv_data unexpected error: -4 (errno 0)
 multipath-tcp#162/9   mptcp/burst:FAIL
 multipath-tcp#162     mptcp:FAIL

In this case, mptcp_recvmsg() gets EAGAIN errors. This issue introduces
by commit b8dc6d6 ("mptcp: fix rcv buffer auto-tuning"). The max
value of scaling_ratio should be TCP_DEFAULT_SCALING_RATIO (128), not
U8_MAX (255). Otherwise, scaling_ratio is assigned to a too high value.

Fixes: b8dc6d6 ("mptcp: fix rcv buffer auto-tuning")
Closes: multipath-tcp#487
Signed-off-by: Geliang Tang <tanggeliang@kylinos.cn>
geliangtang pushed a commit to geliangtang/mptcp_net-next that referenced this issue May 29, 2024
mptcp: fix the max value of scaling_ratio
fb933bc 
BPF tests fail sometimes (a probability of approximately 1%) with
"bytes != total_bytes" errors:

 test_burst:PASS:open_and_load:burst 0 nsec
 test_bpf_sched:PASS:Scheduler name too long 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 create_netns:PASS:ip netns add mptcp_ns 0 nsec
 create_netns:PASS:ip -net mptcp_ns link set dev lo up 0 nsec
 sched_init:PASS:create_netns 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link add veth1 type veth peer name
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.1/24 dev veth1 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth1 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.2/24 dev veth2 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth2 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns mptcp endpoint add 10.0.1.2 subflow
 sched_init:PASS:endpoint_init 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 (network_helpers.c:613: errno: Resource temporarily unavailable) \
                                        send 5608500 expected 10485760
 (network_helpers.c:661: errno: None) recv 2755984 expected 10485760
 (network_helpers.c:669: errno: None) Failed in thread_ret -11
 send_data_and_verify:FAIL:send_recv_data unexpected error: -4 (errno 0)
 multipath-tcp#162/9   mptcp/burst:FAIL
 multipath-tcp#162     mptcp:FAIL

In this case, mptcp_recvmsg() gets EAGAIN errors. This issue introduces
by commit b8dc6d6 ("mptcp: fix rcv buffer auto-tuning"). The max
value of scaling_ratio should be TCP_DEFAULT_SCALING_RATIO (128), not
U8_MAX (255). Otherwise, scaling_ratio is assigned to a too high value.

Fixes: b8dc6d6 ("mptcp: fix rcv buffer auto-tuning")
Closes: multipath-tcp#487
Signed-off-by: Geliang Tang <tanggeliang@kylinos.cn>
geliangtang pushed a commit to geliangtang/mptcp_net-next that referenced this issue May 29, 2024
mptcp: fix the max value of scaling_ratio
c23ddaf 
BPF tests fail sometimes (a probability of approximately 1%) with
"bytes != total_bytes" errors:

 test_burst:PASS:open_and_load:burst 0 nsec
 test_bpf_sched:PASS:Scheduler name too long 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 create_netns:PASS:ip netns add mptcp_ns 0 nsec
 create_netns:PASS:ip -net mptcp_ns link set dev lo up 0 nsec
 sched_init:PASS:create_netns 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link add veth1 type veth peer name
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.1/24 dev veth1 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth1 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.2/24 dev veth2 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth2 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns mptcp endpoint add 10.0.1.2 subflow
 sched_init:PASS:endpoint_init 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 (network_helpers.c:613: errno: Resource temporarily unavailable) \
                                        send 5608500 expected 10485760
 (network_helpers.c:661: errno: None) recv 2755984 expected 10485760
 (network_helpers.c:669: errno: None) Failed in thread_ret -11
 send_data_and_verify:FAIL:send_recv_data unexpected error: -4 (errno 0)
 multipath-tcp#162/9   mptcp/burst:FAIL
 multipath-tcp#162     mptcp:FAIL

In this case, mptcp_recvmsg() gets EAGAIN errors. This issue introduces
by commit b8dc6d6 ("mptcp: fix rcv buffer auto-tuning"). The max
value of scaling_ratio should be TCP_DEFAULT_SCALING_RATIO (128), not
U8_MAX (255). Otherwise, scaling_ratio is assigned to a too high value.

Fixes: b8dc6d6 ("mptcp: fix rcv buffer auto-tuning")
Closes: multipath-tcp#487
Signed-off-by: Geliang Tang <tanggeliang@kylinos.cn>
geliangtang pushed a commit to geliangtang/mptcp_net-next that referenced this issue May 29, 2024
mptcp: fix the max value of scaling_ratio
19ab595 
BPF tests fail sometimes (a probability of approximately 1%) with
"bytes != total_bytes" errors:

 test_burst:PASS:open_and_load:burst 0 nsec
 test_bpf_sched:PASS:Scheduler name too long 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 create_netns:PASS:ip netns add mptcp_ns 0 nsec
 create_netns:PASS:ip -net mptcp_ns link set dev lo up 0 nsec
 sched_init:PASS:create_netns 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link add veth1 type veth peer name
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.1/24 dev veth1 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth1 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.2/24 dev veth2 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth2 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns mptcp endpoint add 10.0.1.2 subflow
 sched_init:PASS:endpoint_init 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 (network_helpers.c:613: errno: Resource temporarily unavailable) \
                                        send 5608500 expected 10485760
 (network_helpers.c:661: errno: None) recv 2755984 expected 10485760
 (network_helpers.c:669: errno: None) Failed in thread_ret -11
 send_data_and_verify:FAIL:send_recv_data unexpected error: -4 (errno 0)
 multipath-tcp#162/9   mptcp/burst:FAIL
 multipath-tcp#162     mptcp:FAIL

In this case, mptcp_recvmsg() gets EAGAIN errors. This issue introduces
by commit b8dc6d6 ("mptcp: fix rcv buffer auto-tuning"). The max
value of scaling_ratio should be TCP_DEFAULT_SCALING_RATIO (128), not
U8_MAX (255). Otherwise, scaling_ratio is assigned to a too high value.

Fixes: b8dc6d6 ("mptcp: fix rcv buffer auto-tuning")
Closes: multipath-tcp#487
Signed-off-by: Geliang Tang <tanggeliang@kylinos.cn>
geliangtang pushed a commit to geliangtang/mptcp_net-next that referenced this issue May 30, 2024
mptcp: fix the max value of scaling_ratio
f9979bc 
BPF tests fail sometimes (a probability of approximately 1%) with
"bytes != total_bytes" errors:

 test_burst:PASS:open_and_load:burst 0 nsec
 test_bpf_sched:PASS:Scheduler name too long 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 create_netns:PASS:ip netns add mptcp_ns 0 nsec
 create_netns:PASS:ip -net mptcp_ns link set dev lo up 0 nsec
 sched_init:PASS:create_netns 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link add veth1 type veth peer name
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.1/24 dev veth1 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth1 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.2/24 dev veth2 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth2 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns mptcp endpoint add 10.0.1.2 subflow
 sched_init:PASS:endpoint_init 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 (network_helpers.c:613: errno: Resource temporarily unavailable) \
                                        send 5608500 expected 10485760
 (network_helpers.c:661: errno: None) recv 2755984 expected 10485760
 (network_helpers.c:669: errno: None) Failed in thread_ret -11
 send_data_and_verify:FAIL:send_recv_data unexpected error: -4 (errno 0)
 multipath-tcp#162/9   mptcp/burst:FAIL
 multipath-tcp#162     mptcp:FAIL

In this case, mptcp_recvmsg() gets EAGAIN errors. This issue introduces
by commit b8dc6d6 ("mptcp: fix rcv buffer auto-tuning"). The max
value of scaling_ratio should be TCP_DEFAULT_SCALING_RATIO (128), not
U8_MAX (255). Otherwise, scaling_ratio is assigned to a too high value.

Fixes: b8dc6d6 ("mptcp: fix rcv buffer auto-tuning")
Closes: multipath-tcp#487
Signed-off-by: Geliang Tang <tanggeliang@kylinos.cn>
geliangtang pushed a commit to geliangtang/mptcp_net-next that referenced this issue May 30, 2024
mptcp: fix the max value of scaling_ratio
021156a 
BPF tests fail sometimes (a probability of approximately 1%) with
"bytes != total_bytes" errors:

 test_burst:PASS:open_and_load:burst 0 nsec
 test_bpf_sched:PASS:Scheduler name too long 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 create_netns:PASS:ip netns add mptcp_ns 0 nsec
 create_netns:PASS:ip -net mptcp_ns link set dev lo up 0 nsec
 sched_init:PASS:create_netns 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link add veth1 type veth peer name
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.1/24 dev veth1 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth1 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.2/24 dev veth2 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth2 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns mptcp endpoint add 10.0.1.2 subflow
 sched_init:PASS:endpoint_init 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 (network_helpers.c:613: errno: Resource temporarily unavailable) \
                                        send 5608500 expected 10485760
 (network_helpers.c:661: errno: None) recv 2755984 expected 10485760
 (network_helpers.c:669: errno: None) Failed in thread_ret -11
 send_data_and_verify:FAIL:send_recv_data unexpected error: -4 (errno 0)
 multipath-tcp#162/9   mptcp/burst:FAIL
 multipath-tcp#162     mptcp:FAIL

In this case, mptcp_recvmsg() gets EAGAIN errors. This issue introduces
by commit b8dc6d6 ("mptcp: fix rcv buffer auto-tuning"). The max
value of scaling_ratio should be TCP_DEFAULT_SCALING_RATIO (128), not
U8_MAX (255). Otherwise, scaling_ratio is assigned to a too high value.

Fixes: b8dc6d6 ("mptcp: fix rcv buffer auto-tuning")
Closes: multipath-tcp#487
Signed-off-by: Geliang Tang <tanggeliang@kylinos.cn>
geliangtang pushed a commit to geliangtang/mptcp_net-next that referenced this issue May 31, 2024
mptcp: fix the max value of scaling_ratio
728d179 
BPF tests fail sometimes (a probability of approximately 1%) with
"bytes != total_bytes" errors:

 test_burst:PASS:open_and_load:burst 0 nsec
 test_bpf_sched:PASS:Scheduler name too long 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 create_netns:PASS:ip netns add mptcp_ns 0 nsec
 create_netns:PASS:ip -net mptcp_ns link set dev lo up 0 nsec
 sched_init:PASS:create_netns 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link add veth1 type veth peer name
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.1/24 dev veth1 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth1 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.2/24 dev veth2 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth2 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns mptcp endpoint add 10.0.1.2 subflow
 sched_init:PASS:endpoint_init 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 (network_helpers.c:613: errno: Resource temporarily unavailable) \
                                        send 5608500 expected 10485760
 (network_helpers.c:661: errno: None) recv 2755984 expected 10485760
 (network_helpers.c:669: errno: None) Failed in thread_ret -11
 send_data_and_verify:FAIL:send_recv_data unexpected error: -4 (errno 0)
 multipath-tcp#162/9   mptcp/burst:FAIL
 multipath-tcp#162     mptcp:FAIL

In this case, mptcp_recvmsg() gets EAGAIN errors. This issue introduces
by commit b8dc6d6 ("mptcp: fix rcv buffer auto-tuning"). The max
value of scaling_ratio should be TCP_DEFAULT_SCALING_RATIO (128), not
U8_MAX (255). Otherwise, scaling_ratio is assigned to a too high value.

Fixes: b8dc6d6 ("mptcp: fix rcv buffer auto-tuning")
Closes: multipath-tcp#487
Signed-off-by: Geliang Tang <tanggeliang@kylinos.cn>
geliangtang pushed a commit to geliangtang/mptcp_net-next that referenced this issue Jun 4, 2024
@geliangtang
mptcp: fix the max value of scaling_ratio
26074c8 
BPF tests fail sometimes (a probability of approximately 1%) with
"bytes != total_bytes" errors:

 test_burst:PASS:open_and_load:burst 0 nsec
 test_bpf_sched:PASS:Scheduler name too long 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 create_netns:PASS:ip netns add mptcp_ns 0 nsec
 create_netns:PASS:ip -net mptcp_ns link set dev lo up 0 nsec
 sched_init:PASS:create_netns 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link add veth1 type veth peer name
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.1/24 dev veth1 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth1 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns addr add 10.0.1.2/24 dev veth2 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns link set dev veth2 up 0 nsec
 endpoint_init:PASS:ip -net mptcp_ns mptcp endpoint add 10.0.1.2 subflow
 sched_init:PASS:endpoint_init 0 nsec
 test_bpf_sched:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 send_data_and_verify:PASS:burst 0 nsec
 (network_helpers.c:613: errno: Resource temporarily unavailable) \
                                        send 5608500 expected 10485760
 (network_helpers.c:661: errno: None) recv 2755984 expected 10485760
 (network_helpers.c:669: errno: None) Failed in thread_ret -11
 send_data_and_verify:FAIL:send_recv_data unexpected error: -4 (errno 0)
 multipath-tcp#162/9   mptcp/burst:FAIL
 multipath-tcp#162     mptcp:FAIL

In this case, mptcp_recvmsg() gets EAGAIN errors. This issue introduces
by commit b8dc6d6 ("mptcp: fix rcv buffer auto-tuning"). The max
value of scaling_ratio should be TCP_DEFAULT_SCALING_RATIO (128), not
U8_MAX (255). Otherwise, scaling_ratio is assigned to a too high value.

Fixes: b8dc6d6 ("mptcp: fix rcv buffer auto-tuning")
Closes: multipath-tcp#487
Signed-off-by: Geliang Tang <tanggeliang@kylinos.cn>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees

@pabeni pabeni

Labels
bug
Projects
No open projects
MPTCP Bugs
  
Closed
Milestone
No milestone
Development

No branches or pull requests
1 participant
@pabeni