Hi Sergey, On Mon, 16 Mar 2015, Sergey Popovich wrote: > I have a lot of changes to the kernel code of ipset subsystem for current > nf-next (and upstream kernel therefore) split into the following major > categories in order I wish to supply them to the upstream. > > * Bugfixes > There is not too much, but exists in upstream, few of which is stability > critical. > > * Code cleanups > Changes to make code more readable, and other things. > > * Improvements to the ipset core set/type/variant > registration/deregistration and other parts of the core. > > * Improvements for bitmap:* set types where two variants implemented: > one for set without extensions and another for set with extensions. > > * Improvements for hash:*, where hash buckets still stores array > of set elements, but all element iterations is done via bitmap using > ffs()/ffz(). There is no error prone memcpy() of elements within bucket. > > * Improvements for list:set, where main change is to move to the > standard linux linked lists implementation. > > * Improve counters extension implementation to move from > atomic64_t variables used to store statistics to per-CPU variables. > > All of these changes are made with primary goal in mind: prepare > code for conversion from read/write locking to the RCU. > > Currently my patch series consists of nearly 170 patches. I wish to > supply each of them divided into categories I described above. > > I choose to send patches against current nf-next instead of > ipset upstream as last one seems already constains some attempts > to implement RCU in the ipset. On this mailing list I found discussion > about some weakness in proposed RCU implementation. > > I think my implementation after all code prepare is ready for review. > > All commits contains meaningful descriptions (I hope) and code > is well commented where it is really necessary. > > Main development is done for 3.2.x series of kernel, and adopted > to the current upstream (actually cosmetics changes to apply patches). > > As userspace ipset 6.23 utility is used and all changes does not > include any kernel<->user space API/ABI breaks. > > Tests are performed on both 32/64 bit arches with kernel compiled > for debugging running ipset regression test suite. Only i386/x86_64 > architectures are used for tests. > > Thanks for your patience for reading this, please take time to review > and possibly accept my work. Please rebase your patches against the ipset git tree: the development happens there and patches are submitted from the ipset git tree to nf-next. Also, do not forget to create numbered patches in the future. Thanks! I'm able to apply one single patch from your current series. Best regards, Jozsef - E-mail : kadlec@xxxxxxxxxxxxxxxxx, kadlecsik.jozsef@xxxxxxxxxxxxx PGP key : http://www.kfki.hu/~kadlec/pgp_public_key.txt Address : Wigner Research Centre for Physics, Hungarian Academy of Sciences H-1525 Budapest 114, POB. 49, Hungary -- To unsubscribe from this list: send the line "unsubscribe netfilter-devel" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
