On 12/26/2019 08:03 AM, Jouni Malinen wrote:
On Thu, Dec 26, 2019 at 08:35:16AM -0800, Ben Greear wrote:
On 12/26/2019 02:05 AM, Jouni Malinen wrote:
On Mon, Mar 18, 2019 at 01:46:34PM -0700, greearb@xxxxxxxxxxxxxxx wrote:
And, if user tries to delete a neigh entry that is already deleted,
return OK instead of failure.
But I dropped this part since it seems to be completely separate item
from the main change and would change the existing interface in a manner
that does not seem to be necessary and may be unexpected and
furthermore, it breaks the rrm_neighbor_db test case.
Would you consider a version of this patch that took an extra 'force'
or similar argument? While automating this, we would rather the
delete not fail if the target is already gone: The system is already in
the desired state, so automation should not need to deal with any
extra fault recovery in this state.
Why would this extra complexity be needed in hostapd instead of the tool
that issues these remove operations for not existing entries simply
ignore the result code? The "target is already gone" is not really
accurate, i.e., the target might never have been in the first place..
This sounds like a pretty special case and I'd have the caller ignore
the result for such a special case; adding " force" or something similar
to the command just to hardcode the result value to be "OK" seems
excessive.
I can check the code again when I get a chance, but from what I recall,
it was not convenient to make sure we removed entries exactly once in
all cases, so our code ended up doing duplicated removes in some cases.
It was easier to make hostap ignore the duplicate removes instead
of parsing errors to figure out if the error was related to removing
a non-existing entity or not.
Either way, it will not be a problem for us to carry this additional patch
in our own tree.
Thanks,
Ben
--
Ben Greear <greearb@xxxxxxxxxxxxxxx>
Candela Technologies Inc http://www.candelatech.com
_______________________________________________
Hostap mailing list
Hostap@xxxxxxxxxxxxxxxxxxx
http://lists.infradead.org/mailman/listinfo/hostap
