[Click] Queue overflow bug?
Eddie Kohler
kohler at cs.ucla.edu
Wed May 5 19:39:53 EDT 2010
You are missing nothing. Thanks for the report - it's fixed!
Eddie
On 5/5/10 11:07 AM, Ian Rose wrote:
> Hi -
>
> I think I have found a bug in the Queue element, although this would
> surprise me considering how often this element is used by people...
>
> line 126 of fullnotequeue.hh (in the push_failure method) kills the
> packet regardless instead of calling "checked_output_push(1, p)" like
> the other queue implementations do (e.g. line 165 of simpleeueue.cc).
>
> Am I missing something?
> - Ian
>
>
> Here is a config that demonstrates the problem (sorry about the
> weirdness with the Unqueue elements - I'm not sure if there is a simpler
> way to make this happen):
>
> InfiniteSource(\<0800>, LIMIT 20) -> Unqueue -> q::Queue(10) ->
> Unqueue(ACTIVE false) -> Discard;
> q[1] -> Script(TYPE PACKET, print "I see a drop", return 0) -> Discard;
>
> If you run this you should ONLY see "q :: Queue: overflow", but if you
> change q to a SimpleQueue instead, then you should see:
>
> q :: SimpleQueue: overflow
> I see a drop
> I see a drop
> I see a drop
> I see a drop
> I see a drop
> I see a drop
> I see a drop
> I see a drop
> I see a drop
> I see a drop
> _______________________________________________
> click mailing list
> click at amsterdam.lcs.mit.edu
> https://amsterdam.lcs.mit.edu/mailman/listinfo/click
More information about the click
mailing list