lwn.git/net/batman-adv/network-coding.c, branch v4.4.30

batman-adv: Avoid recursive call_rcu for batadv_nc_node

2016-01-31T19:29:02+00:00

[ Upstream commit 44e8e7e91d6c7c7ab19688750f7257292640d1a0 ] The batadv_nc_node_free_ref function uses call_rcu to delay the free of the batadv_nc_node object until no (already started) rcu_read_lock is enabled anymore. This makes sure that no context is still trying to access the object which should be removed. But batadv_nc_node also contains a reference to orig_node which must be removed. The reference drop of orig_node was done in the call_rcu function batadv_nc_node_free_rcu but should actually be done in the batadv_nc_node_release function to avoid nested call_rcus. This is important because rcu_barrier (e.g. batadv_softif_free or batadv_exit) will not detect the inner call_rcu as relevant for its execution. Otherwise this barrier will most likely be inserted in the queue before the callback of the first call_rcu was executed. The caller of rcu_barrier will therefore continue to run before the inner call_rcu callback finished. Fixes: d56b1705e28c ("batman-adv: network coding - detect coding nodes and remove these after timeout") Signed-off-by: Sven Eckelmann Signed-off-by: Marek Lindner Signed-off-by: Antonio Quartulli Signed-off-by: Greg Kroah-Hartman

batman-adv: Annotate deleting functions with external lock via lockdep

2015-08-24T22:12:23+00:00

Functions which use (h)list_del* are requiring correct locking when they operate on global lists. Most of the time the search in the list and the delete are done in the same function. All other cases should have it visible that they require a special lock to avoid race conditions. Lockdep asserts can be used to check these problem during runtime when the lockdep functionality is enabled. Signed-off-by: Sven Eckelmann Signed-off-by: Marek Lindner Signed-off-by: Antonio Quartulli

batman-adv: Remove multiple assignment per line

2015-08-24T22:12:18+00:00

The Linux CodingStyle disallows multiple assignments in a single line. (see chapter 1) Reported-by: Markus Pargmann Signed-off-by: Sven Eckelmann Signed-off-by: Marek Lindner Signed-off-by: Antonio Quartulli

batman-adv: Fix kerneldoc over 80 column lines

2015-08-24T22:12:18+00:00

Kerneldoc required single line documentation in the past (before 2009). Therefore, the 80 columns limit per line check of checkpatch was disabled for kerneldoc. But kerneldoc is not excluded anymore from it and checkpatch now enabled the check again. Signed-off-by: Sven Eckelmann Signed-off-by: Marek Lindner Signed-off-by: Antonio Quartulli

batman-adv: Replace C99 int types with kernel type

2015-08-24T22:12:17+00:00

(s|u)(8|16|32|64) are the preferred types in the kernel. The use of the standard C99 types u?int(8|16|32|64)_t are objected by some people and even checkpatch now warns about using them. Signed-off-by: Sven Eckelmann Signed-off-by: Marek Lindner Signed-off-by: Antonio Quartulli

batman-adv: Make NC capability changes atomic

2015-08-14T20:50:43+00:00

Bitwise OR/AND assignments in C aren't guaranteed to be atomic. One OGM handler might undo the set/clear of a specific bit from another handler run in between. Fix this by using the atomic set_bit()/clear_bit()/test_bit() functions. Fixes: 3f4841ffb336 ("batman-adv: tvlv - add network coding container") Signed-off-by: Linus Lüssing Signed-off-by: Marek Lindner Signed-off-by: Antonio Quartulli

batman-adv: Add required includes to all files

2015-06-07T15:07:19+00:00

The header files could not be build indepdent from each other. This is happened because headers didn't include the files for things they've used. This was problematic because the success of a build depended on the knowledge about the right order of local includes. Also source files were not including everything they've used explicitly. Instead they required that transitive includes are always stable. This is problematic because some transitive includes are not obvious, depend on config settings and may not be stable in the future. The order for include blocks are: * primary headers (main.h and the *.h file of a *.c file) * global linux headers * required local headers * extra forward declarations for pointers in function/struct declarations The only exceptions are linux/bitops.h and linux/if_ether.h in packet.h. This header file is shared with userspace applications like batctl and must therefore build together with userspace applications. The header linux/bitops.h is not part of the uapi headers and linux/if_ether.h conflicts with the musl implementation of netinet/if_ether.h. The maintainers rejected the use of __KERNEL__ preprocessor checks and thus these two headers are only in main.h. All files using packet.h first have to include main.h to work correctly. Reported-by: Markus Pargmann Signed-off-by: Sven Eckelmann Signed-off-by: Marek Lindner

batman-adv: Use common Jenkins Hash implementation

2015-06-07T15:07:17+00:00

An unoptimized version of the Jenkins one-at-a-time hash function is used and partially copied all over the code wherever an hashtable is used. Instead the optimized version shared between the whole kernel should be used to reduce code duplication and use better optimized code. Only the DAT code must use the old implementation because it is used as distributed hash function which has to be common for all nodes. Signed-off-by: Sven Eckelmann Signed-off-by: Marek Lindner

batman-adv: checkpatch - spaces preferred around that '*'

2015-05-29T08:13:37+00:00

Signed-off-by: Marek Lindner

batman-adv: Use safer default config for optional features

2015-05-29T08:13:36+00:00

The current default settings for optional features in batman-adv seems to be based around the idea that the user only compiles what he requires. They will automatically enabled when they are compiled in. For example the network coding part of batman-adv is by default disabled in the out-of-tree module but will be enabled when the code is compiled during the module build. But distributions like Debian just enable all features of the batman-adv kernel module and hope that more experimental features or features with possible negative effects have to be enabled using some runtime configuration interface. The network_coding feature can help in specific setups but also has drawbacks and is not disabled by default in the out-of-tree module. Disabling by default in the runtime config seems to be also quite sane. The bridge_loop_avoidance is the only feature which is disabled by default but may be necessary even in simple setups. Packet loops may even be created during the initial node setup when this is not enabled. This is different than STP on bridges because mesh is usually used on Adhoc WiFi. Having two nodes (by accident) in the same LAN segment and in the same mesh network is rather common in this situation. Signed-off-by: Sven Eckelmann Acked-by: Martin Hundebøll Signed-off-by: Marek Lindner