Linux Block IO - present and future
One of the primary focus points of 2.5 was fixing up the bit rotting block layer, and as a result 2.6 now sports a brand new implementation of basically anything that has to do with passing IO around in the kernel, from producer to disk driver. The talk will feature an in-depth look at the IO core system in 2.6 comparing to 2.4, looking at performance, flexibility, and added functionality. The rewrite of the IO scheduler API and the new IO schedulers will get a fair treatment as well.
No 2.6 talk would be complete without 2.7 speculations, so I shall try to predict what changes the future holds for the world of Linux block I/O.