GRO test is very timing-sensitive, packets may be delayed
by the network or just sent slowly. Because of this we retry
each test case up to 6 times.
This makes perfect sense for positive cases, in which we want
to see coalescing. Negative test cases, which modify headers
and expect no coalescing should have opposite treatment.
We should really try 6 times and make sure that each time
the test failed. This would, however, require that we annotate
each test to indicate whether its positive or negative.
Let's start with a simpler improvement. Do not allow
retries if we detected over-coalescing. Previously the negative
case would have to get lucky at least once in 6 tries to pass.
Now the first failure breaks the retry loop.
For background - NICs tend to ignore the contents of the TCP
timestamp option, so that test case commonly fails. In NIPA
having 6 attempts, however, was enough for some NICs to get
multiple successful runs in a row, getting the test cases
auto-classified as expected to pass, even tho the NIC does
not comply with the expectations.
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Link: https://patch.msgid.link/20260607002401.212976-1-kuba@kernel.org
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Add PPPoE test-cases to the GRO selftest. Only run a subset of
common_tests to avoid changing the hardcoded L3 offsets everywhere.
Add a new "pppoe_sid" test case to verify that packets with different
PPPoE session IDs are correctly identified as separate flows and not
coalesced.
Signed-off-by: Qingfang Deng <qingfang.deng@linux.dev>
Link: https://patch.msgid.link/20260513013400.7467-2-qingfang.deng@linux.dev
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
We have a test for coalescing with bad TCP checksum, let's also
test bad IPv4 header checksum.
Reviewed-by: Willem de Bruijn <willemb@google.com>
Link: https://patch.msgid.link/20260402210000.1512696-9-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
We explicitly test ipip encap. Let's add ip6ip6, too. Having
just ipip seems like favoring IPv4 which we should not do :)
Testing all combinations is left for future work, not sure
it's actually worth it.
Reviewed-by: Willem de Bruijn <willemb@google.com>
Link: https://patch.msgid.link/20260402210000.1512696-8-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
When constructing the packets for large_* test cases we use
a static value for packet count and MSS. It works okay for
ipv4 vs ipv6 but the gap between ipv4 and ip6ip6 is going to
be quite significant.
Make the defines calculate the worst case values, those
are only used for sizing stack arrays. Create helpers for
calculating precise values based on the exact test case.
Reviewed-by: Willem de Bruijn <willemb@google.com>
Link: https://patch.msgid.link/20260402210000.1512696-7-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Willem points out TOTAL_HDR_LEN is identical to MAX_HDR_LEN.
This seems to have been the case ever since the test was added.
Replace the uses of TOTAL_HDR_LEN with MAX_HDR_LEN, MAX seems
more common for what this value is.
Reviewed-by: Willem de Bruijn <willemb@google.com>
Link: https://patch.msgid.link/20260402210000.1512696-6-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Try to use already calculated offsets and not depend on the ipip
flag as much. This patch should not change any functionality,
it's just a cleanup to make ip6ip6 support easier.
Reviewed-by: Willem de Bruijn <willemb@google.com>
Link: https://patch.msgid.link/20260402210000.1512696-5-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The new capacity/order test exits as soon as it sees the expected
packet sequence. This may allow the "flushing" FIN packet to spill
over to the next test. Let's always wait for the FIN before exiting.
Reviewed-by: Willem de Bruijn <willemb@google.com>
Link: https://patch.msgid.link/20260402210000.1512696-4-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Small IPv4 packets get padded to 60B, this may break / confuse
some buggy implementations. Add a test to coalesce a 1B payload.
Keep this separate from the lrg_sml test because I suspect some
implementations may not handle this case (treat padded frames
as ineligible for coalescing).
Reviewed-by: Willem de Bruijn <willemb@google.com>
Link: https://patch.msgid.link/20260402210000.1512696-3-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Add a test trying to induce a GRO context timeout followed
by another sequence of packets for the same flow. The second
burst arrives 100ms after the first one so any implementation
(SW or HW) must time out waiting at that point. We expect both
bursts to be aggregated successfully but separately.
Reviewed-by: Willem de Bruijn <willemb@google.com>
Link: https://patch.msgid.link/20260402210000.1512696-2-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Test accuracy of GRO stats. We want to cover two potentially tricky
cases:
- single segment GRO
- packets which were eligible but didn't get GRO'd
The first case is trivial, teach gro.c to send one packet, and check
GRO stats didn't move.
Second case requires gro.c to send a lot of flows expecting the NIC
to run out of GRO flow capacity.
To avoid system traffic noise we steer the packets to a dedicated
queue and operate on qstat.
Link: https://patch.msgid.link/20260318033819.1469350-5-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Longer packet sequence tests are quite flaky when the test is run
over a real network. Try to avoid at least the jitter on the sender
side by scheduling all the packets to be sent at once using SO_TXTIME.
Use hardcoded tx time of 5msec in the future. In my test increasing
this time past 2msec makes no difference so 5msec is plenty of margin.
Since we now expect more output buffering make sure to raise SNDBUF.
Note that this is an opportunistic reliability improvement which
will only work if the qdisc can schedule Tx time for us (fq).
Fiddling with qdisc config was deemed too complex, so it's not
part of the patch.
Reviewed-by: Willem de Bruijn <willemb@google.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Link: https://patch.msgid.link/20260318033819.1469350-4-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The gro.c packet sender is used for SW testing but bulk of incoming
new tests will be HW-specific. So it's better to put them under
drivers/net/hw/, to avoid tip-toeing around netdevsim. Move gro.c
to lib so we can reuse it.
Reviewed-by: Petr Machata <petrm@nvidia.com>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Link: https://patch.msgid.link/20260318033819.1469350-2-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
