On Tue, Sep 28, 2021 at 2:31 AM Lorenz Bauer [off-list ref] wrote:
We have a unit test that invokes an XDP program with 1m different
inputs, aka 1m BPF_PROG_RUN syscalls. We run this test concurrently
with slight variations in how we generated the input.
Since commit f23c4b3924d2 ("bpf: Start using the BPF dispatcher in BPF_TEST_RUN")
the unit test has slowed down significantly. Digging deeper reveals that
the concurrent tests are serialised in the kernel on the XDP dispatcher.
This is a global resource that is protected by a mutex, on which we contend.
Fix this by not calling into the XDP dispatcher if we only want to perform
a single run of the BPF program.
See: https://lore.kernel.org/bpf/CACAyw9_y4QumOW35qpgTbLsJ532uGq-kVW-VESJzGyiZkypnvw@mail.gmail.com/ (local)
Signed-off-by: Lorenz Bauer <redacted>
Applied. Thanks