On Mon, 15 Jun 2015, Alexander Duyck wrote: > So I can see the motivation behind bulk allocation, but I cannot see the > motivation behind bulk freeing. In the case of freeing the likelihood of the > memory regions all belonging to the same page just isn't as high. The likelyhood is high if the object are allocated in batch as well. In that case SLUB ensures that all objects from the same page are first allocated. > I don't really see the reason why you should be populating arrays. SLUB uses a > linked list and I don't see this implemented for SLOB or SLAB so maybe you > should look at making this into a linked list. Also as I stated in the other > comment maybe you should not do bulk allocation if you don't support it in > SLAB/SLOB and instead change this so that you return a count indicating that > only 1 value was allocated in this pass. It is extremely easy to just take the linked list off a page or a per cpu structure. Basically we would have constant cycle count for taking N objects available in a slab page. -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@xxxxxxxxx. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>