This patchset enables the usage of the batman-adv multicast optimizations for scenarios involving bridges on top of e.g. bat0, too.
The first one alters the forwarding behaviour for IGMP and MLD reports which is a prerequisite before enabling multicast optimizations in bridged setups. It also increases the MCAST TVLV version number to 2. More on the issue of insufficient report handling with the v1 MCAST TVLV can be found here:
http://www.open-mesh.org/projects/batman-adv/wiki/Multicast-optimizations-li...
The second patch finally enables the multicast optimizations for bridged setups.
Along come two more patches adding according debugging facilities to make it possible for the user to check why the multicast optimizations might not work ideally to give hints about what they might change about their topology.
This patchset can be found in the current linus/multicast-bridge branch.
Cheers, Linus
-----
Thanks for the many valuable feedback at the battlemesh and beyond!
Changes in v11: * Rewording of commit messages (PATCH 1/4+2/4) * Rebasing to master (solving conflict with kernel doc "return" changes) * Updating kerneldoc to proper "Return:" syntax * Using eth_zero_addr() over memset (fixes a new checkpatch warning) * Changing uint8_t to u8 and uint32_t to u32 (fixes new checkpatch warnings) * Turning some batadv_dbg() into batadv_info(), making querier presence more visible (i.e. in dmesg too) (PATCH 3/4)
Changes in v10: - PATCH 1/4: * Former "[PATCH 4/4] batman-adv: Forward IGMP/MLD reports to selected querier (only)" substituted by the interim branch "batman-adv: Always flood IGMP/MLD reports" with the following changes: * Increased compatibility to mcast-v1-tvlv nodes by registering a v1 tvlv container and handler, too (note: PATCH 2/4 needs to still unregister the handler if the node is bridged bc. it can't "trust" v1 nodes) * Removed skb_set_network_header() call (extra patch pending for review for maint) * Compared to patchset v9 this substitution makes things compilable again for kernels < 2.6.35 - PATCH 2/4: * Moved enum introduction of BATADV_DBG_MCAST to PATCH 3/4, not needed in PATCH 2/4 yet * Adjusted compat code to new compat layout * Adjustments to batadv_mcast_mla_tvlv_update() to fit now preceding PATCH 1/4 * bat_priv->mcast.bridged flag already introduced here (instead of in PATCH 3/4) because we need it for the new mcast-v1-tvlv handling
--- Interim Changlog of "batman-adv: Always flood IGMP/MLD reports"
v1: * Removed query snooping and state * Squashed all three patches into one * Renamed "batadv_mcast_tvlv_ogm_handler_v1()" to *_v2() * Added explicit icmpv6.h include * Rebased on top of master
----- Interim Interim Changelog of "batman-adv: Unicasting multicast reports to querier-node only" v6: * compat: copied copyright headers from original upstream c files * compat: unified ordering in compat c files: -> copyright header, then includes, then kernel specific functions v5: * Removed RFC tag: Needed exports got merged to net-next and are going to be available with Linux 4.2 * Redid compat solution - now fully backwards compatible down to 2.6.33 v4: * excluded bridge part from this patchset, they should hopefully be added to net-next soon * Added a compat solution (PATCH 3/3) * Removed Kconfig-depends as by David's suggestion the needed parsing functions for MLD are going to be forced built-ins even if IPv6 is going to be built as a module * Removed unused variable 'int ret' in batadv_mcast_is_report_ipv6() * Adjusted to new folder structure v3: * Adding Kconfig-depends and #if's (so basically adding similar dependancy constraints as the bridge code has, except that there are no depends if batman-adv gets compiled without multicast optimizations) -> the case of IPv6=M and batman-adv=y is still impossible if multicast optimizations are enabled; but I don't see the practical demand for that either - people who use IPv6 as a module will probably also want to use batman-adv as a module v2: * various bugfixes (now runtime tested, too - should(tm) work) * added netdev+bridge mailinglists ----- ---
Changes in v9: - PATCH 1/4: * fix: added compat code for pr_warn_once() * compat fix for bridge export stubs: fixes compile error with kernels < 3.16 without bridge (snooping) support - PATCH 2/4: * perform updates of variables within bat_priv->mcast.querier_ipv{4,6} individually (there's a new, third member in 4/4 which shouldn't be overriden) * PATCH 4/4: NEW
Changes in v8 (thanks to Simon's suggestions): - PATCH 2/3: * print shadowing status log of an appearing and shadowing querier, too (the bridge-querier-existence call has an additional 10s delay to ensure reports had their time to arrive - the bridge-querier-port call doesn't have that) - PATCH 3/3: * changing debugfs output from "+" and "-" to "U/4/6" and "." * fixing "no querier present" logic (introduced in [PATCHv7 3/3])
Changes in v7 (thanks to Simon's suggestions): - PATCH 2/3: * renaming old/new_querier to old/new_state * slightly extended kerneldoc of batadv_mcast_querier_log() * removing words "good" and "bad" from debug output * simplified batadv_mcast_flags_log() * assignment instead of memset in batadv_mcast_mla_tvlv_update() and batadv_softif_init_late() * simple struct member assignments instead of one complex struct assigment * removing unnecessary memcmp's * substituting return statement for an if-block in batadv_mcast_querier_log() and batadv_mcast_bridge_log() * print "Unsnoopables(U)-flag" instead of just "U-flag" - PATCH 3/3: * use bat_priv values instead of querying bridge ABI in batadv_mcast_flags_print_header()
Changes in v6: * New PATCH 2/3 inserted, moving logging to separate patch * More verbose logging added to PATCH 2/3: Bridge and querier state changes are logged too * upper case to lower case for kernel doc of batadv_mcast_flags_open (PATCH 2/3) * Adding note to kernel doc of batadv_mcast_get_bridge about increased refcount (PATCH 1/3) * Printing some lines about current bridge and querier state to debugfs too (PATCH 3/3)
Changes in v5 (PATCH 2/2 only): * s/dat_cache/mcast_flags/ in kerneldoc (copy&paste error)
Changes in v4 (PATCH 2/2 only): * initial {ad,e}dition of this patch
Changes in v3 (PATCH 1/2 only): * Removed "RFC" tag in title again: The stubs and new export are upstream in net-next and therefore going to be included in 3.17 * Added some debug output: * Two warning messages: -> Old kernel version or no bridge IGMP/MLD snooping compiled * New batman-adv log-level "mcast": -> Logging mcast flag changes (a third debugging facility, a new table for debugfs for a global mcast flag overview will be added in a separate patch later as discussed with Simon)
Changes in v2 (PATCH 1/2 only): * fetching local (= on this same kernel) multicast listeners from the bridge instead of the bat0 interface if a bridge is present - just like ip addresses and routes should be used from br0, the same goes for multicast listeners * beautification of batadv_mcast_mla_br_addr_cpy(), now using already present functions from the kernel instead of own, hackish approach * changed names of some goto-labels (not using "skip" anymore) * using new, third bridge multicast export (because this export is not upstream yet, I've added the "RFC" in the title): br_multicast_has_querier_anywhere() * adding compat stubs for two bridge multicast exports, to make batman-adv compile- and usable even if a 3.16 kernel was compiled without bridge code - the stubs are supposed to be upstream in the bridge code in 3.17 (therefore just 'compat') * updated kerneldocs for batadv_mcast_mla_bridge_get() and batadv_mcast_mla_softif_get() * The two sentences in the commit message starting with "Queriers: ..." were slightly modified to include the third bridge multicast export