mv is-router and flatifier to their own dir
reduce the amount of includes teshsuite/simdag
kill our _XBT_WIN32. Standard _WIN32 is sufficient
cleanups in routing - Make the link field of sg_platf_route_cbarg_t a vector<Link*> instead of a dynar. - This dynar used to contain sometimes links (when computing the RouteAndLatency) and sometimes link names (when parsing). - When parsing, the AS were in charge of checking that the links which names are given exist. Now, the parsing code has to deal with its own mess, to pass the expected Link*. The situation around this data structure is still very messy because: - src and dst are the names while src_gw and dst_gw are the netcards - not all the fields are used in all usages: - src/dst not used in data stored in floyd/dijkstra - gw_* not used when adding a route (only for ASroute) - gw_* not used in the data stored in floyd/dijkstra but in the return value of getRouteAndLatency for these models. Ultimately, this data structure should be limited to the xml world, and proper replacement designed for each situations.
our code is so entertaining
nicer API to the Netcard type
factor host samples, update flatifier
routing: inline method routing_get_network_element_type This was a C->C++ adapter (+ rename a netcard field in instr, and other cosmetics)
kill old cruft
I said there is no need to copy before strcmp
SimDag Revolution: SD_workstation becomes sg_host
body count++: routing_get_latency_and_bandwidth
SD_link_get_list -> sd_link_list
Thx green guy! power is speed.
Start the SimDag revolution: function factoring + SD_link_get_ -> sg_link_ for count/bandwidth/latency/name/data + SD_link_set_data -> sg_link_data_set + SD_link_is_shared -> sg_link_is_shared
Start the SimDag revolution: function renaming + number -> count (workstation/link lists) + remove 'current' in bandwidth/latency getters
remove completely the energy concern from the core - It is moved to the plugin only. - In particular, there is nothing in the core headers anymore - SMPI still defines some energy methods - the command-line config still needs to know the init method Also, I'm sorry, but I renamed simgrid::Host->(p_cpu/p_netcard) to pimpl_cpu/pimpl_netcard. I should have separated the rename, sorry.
Give a p_netcard to simgrid::Host instead of relying on extensions for that