On 2026-11-09 08:23, Sashiko wrote: > [Medium] In test_dp_features, the skip path does not call ovs_exit_sig. > [Low] CONFIG_NET_TC_SKB_EXT is not in sort order. > [Low] Should CONFIG_NET_SCHED=y be added to the fragment too?
All three are in v3. The recirc round is a test of its own now, and it calls ovs_exit_sig before it returns the skip code, like the other skips in the middle of a test in that file. The fragment has CONFIG_NET_SCHED=y, CONFIG_NET_CLS_ACT=y and CONFIG_NET_TC_SKB_EXT=y, with the last one after CONFIG_NET_IPGRE_DEMUX in the order that ignores underscores. Two more changes on the way. The skip reason is printed to stdout, so a run that skipped the recirc round no longer ends as a plain pass with no sign that the round did not run. And the expected errno is read from python rather than written as 95, because EOPNOTSUPP is 45 on alpha and sparc, 122 on mips and 223 on parisc, where a hardcoded number would turn a real failure into a skip. _______________________________________________ dev mailing list [email protected] https://mail.openvswitch.org/mailman/listinfo/ovs-dev
