Hi David, On 02/16/2016 03:48 PM, David Hunt wrote: > Adds the new rte_mempool_create_ext api and callback mechanism for > external mempool handlers > > Modifies the existing rte_mempool_create to set up the handler_idx to > the relevant mempool handler based on the handler name: > ring_sp_sc > ring_mp_mc > ring_sp_mc > ring_mp_sc > > v2: merges the duplicated code in rte_mempool_xmem_create and > rte_mempool_create_ext into one common function. The old functions > now call the new common function with the relevant parameters. > > Signed-off-by: David Hunt <david.hunt at intel.com>
I think the refactoring of rte_mempool_create() (adding of mempool_create()) should go in another commit. It will make the patches much easier to read. Also, I'm sorry but it seems that several comments or question I've made in http://dpdk.org/ml/archives/dev/2016-February/032706.html are not addressed. Examples: - putting some part of the patch in separate commits - meaning of "rt_pool" - put_pool_bulk unclear comment - should we also have get_pool_bulk stats? - missing _MEMPOOL_STAT_ADD() in mempool_bulk() - why internal in rte_mempool_internal.h? - why default in rte_mempool_default.c? - remaining references to stack handler (in a comment) - ...? As you know, doing a proper code review takes a lot of time. If I have to re-check all of my previous comments, it will take even more. I'm not saying all my comments require a code change, but in case you don't agree, please at least explain your opinion so we can debate on the list. Regards, Olivier