Removed - Thermonuclear device not required ;-) On 13/10/15 12:53, Dave Taht wrote: > nuke it then. > > On Tue, Oct 13, 2015 at 1:26 PM, Kevin Darbyshire-Bryant > wrote: >> I can't find anything that uses "TCA_CAKE_ACTIVE_FLOWS" in the current >> codebase of ether cake or tc-adv, so if we're in the current business of >> breaking API/ABI stuff then shouldn't it go? >> >> The output of 'sparse & bulk' flows is still there in the output stats >> shown by tc, so is obviously in use. >> >> On 13/10/15 11:34, Dave Taht wrote: >>> active_flows was a parameter that was in one version or another that >>> was exposed to the API. It was very useful and interesting. >>> >>> On Tue, Oct 13, 2015 at 12:22 PM, Kevin Darbyshire-Bryant >>> wrote: >>>> Hi List, >>>> >>>> In pkt_sched.h we see an enumeration for 'TCA_CAKE_ACTIVE_FLOWS': >>>> >>>> /* CAKE */ >>>> enum { >>>> TCA_CAKE_UNSPEC, >>>> TCA_CAKE_BASE_RATE, >>>> TCA_CAKE_DIFFSERV_MODE, >>>> TCA_CAKE_ATM, >>>> TCA_CAKE_FLOW_MODE, >>>> TCA_CAKE_OVERHEAD, >>>> TCA_CAKE_ACTIVE_FLOWS, >>>> TCA_CAKE_RTT, >>>> TCA_CAKE_TARGET, >>>> __TCA_CAKE_MAX >>>> }; >>>> #define TCA_CAKE_MAX (__TCA_CAKE_MAX - 1) >>>> >>>> These enumerations are normally used for passing options from tc into >>>> the cake module, but at the present time, nowhere in either cake or tc >>>> can I see use of 'ACTIVE_FLOWS'. It's not read as an input into cake, >>>> nor is it sent out of cake back to tc via the nla_policy. >>>> >>>> Is this a 'typo' from when bulk/sparse flows reporting was put into the >>>> stats, or is it some as yet unimplemented feature? >>>> >>>> I've meant to ask about this before but it sort of got lost. >>>> >>>> Kevin >>>> >>>> >>>> _______________________________________________ >>>> Cake mailing list >>>> Cake@lists.bufferbloat.net >>>> https://lists.bufferbloat.net/listinfo/cake >>>> >>> >> > >