On Mon, Jul 15, 2024 at 11:53:45AM +0200, Alexis Lothoré (eBPF Foundation) wrote:
test_xdp_veth.sh tests that XDP return codes work as expected, by bringing up multiple veth pairs isolated in different namespaces, attaching specific xdp programs to each interface, and ensuring that the whole chain allows to ping one end interface from the first one. The test runs well but is currently not integrated in test_progs, which prevents it from being run automatically in the CI infrastructure.
Rewrite it as a C test relying on libbpf to allow running it in the CI infrastructure. The new code brings up the same network infrastructure and reuses the same eBPF programs as test_xdp_veth.sh, for which skeletons are already generated by the bpf tests makefile.
Signed-off-by: Alexis Lothoré (eBPF Foundation) alexis.lothore@bootlin.com
The new code has been tested in an aarch64 qemu instance: Summary: 1/0 PASSED, 0 SKIPPED, 0 FAILED
I have also checked that some minor alterations in the network configuration (altering the redirect map, or not loading one of the xdp programs) make the test fail.
On my testing setup, the test takes a bit more than 3 seconds to run on average.
Changes in v2:
- fix many formatting issues raised by checkpatch
- use static namespaces instead of random ones
- use SYS_NOFAIL instead of snprintf() + system ()
- squashed the new test addition patch and the old test removal patch
tools/testing/selftests/bpf/Makefile | 1 - .../selftests/bpf/prog_tests/test_xdp_veth.c | 211 +++++++++++++++++++++ tools/testing/selftests/bpf/test_xdp_veth.sh | 121 ------------ 3 files changed, 211 insertions(+), 122 deletions(-)
diff --git a/tools/testing/selftests/bpf/Makefile b/tools/testing/selftests/bpf/Makefile index a7932bead77d..2864a0dc04d5 100644 --- a/tools/testing/selftests/bpf/Makefile +++ b/tools/testing/selftests/bpf/Makefile @@ -117,7 +117,6 @@ TEST_PROGS := test_kmod.sh \ test_xdp_redirect.sh \ test_xdp_redirect_multi.sh \ test_xdp_meta.sh \
- test_xdp_veth.sh \ test_tunnel.sh \ test_lwt_seg6local.sh \ test_lirc_mode2.sh \
diff --git a/tools/testing/selftests/bpf/prog_tests/test_xdp_veth.c b/tools/testing/selftests/bpf/prog_tests/test_xdp_veth.c new file mode 100644 index 000000000000..3ffeb411c131 --- /dev/null +++ b/tools/testing/selftests/bpf/prog_tests/test_xdp_veth.c @@ -0,0 +1,211 @@ +// SPDX-License-Identifier: GPL-2.0 +/**
- Create 3 namespaces with 3 veth peers, and
- forward packets in-between using native XDP
XDP_TX
- NS1(veth11) NS2(veth22) NS3(veth33)
| | |
| | |
- (veth1, (veth2, (veth3,
- id:111) id:122) id:133)
^ | ^ | ^ |
| | XDP_REDIRECT | | XDP_REDIRECT | |
| ------------------ ------------------ |
-----------------------------------------
XDP_REDIRECT- */
Hi Alexis,
A minor nit from my side.
The comment above starts with '/**' but otherwise it is not a Kernel doc. It's probably best if it's a networking-style multi-line comment instead.
/* Create 3 namespaces with 3 veth peers, and * ... */
...