On Fri, 27 Oct 2017 09:27:30 +0200, Jiri Pirko wrote:

quoted

quoted

quoted

2.       Deleting the ingress qdisc fails to remove filters added in
HW. Filters in SW gets deleted.

We haven’t exactly root-caused this, the changes being extensive, but
our guess is again something wrong with the offload check or similar
while unregistering the block callback (tcf_block_cb_unregister) and
further to the classifier (CLS_U32/CLS_FLOWER etc.) with the
DESTROY/REMOVE command.  

Hmm. How does this worked previously. I mean, do you see change of
behaviour? I'm asking because I don't see how rules added only to HW
could be removed, driver should care of it. Or are you talking about
rules added to both SW and HW?  

These are rules added to both SW and HW. Previously all cls_* had
ndo_setup_tc calls based on the offload capability.

commit 8d26d5636d "net: sched: avoid ndo_setup_tc calls for
TC_SETUP_CLS*" removed this bit to work with the new block callback. Is
there something similar in the block callback flow while acting on the
tcf_proto destroy call initiated when the qdisc is cleared?  

Yes, it is the same.

FWIW I also see what Amritha and Alex are describing here, for cls_bpf
there are no DESTROYs coming on rmmod or qdisc del.  There is a DESTROY
if I manually remove the filter (or if an ADD with skip_sw fails).