Messages in this thread | | | Date | Sat, 20 Mar 1999 22:20:17 -0500 (EST) | From | Mathew G Monroe <> | Subject | Re: disk head scheduling |
| |
Excerpts from internet.computing.linux-kernel: 21-Mar-99 Re: disk head scheduling by Jelle Foks@flying.demon. > > Suppose someone writes a tool that measures seek-times of a specific disk > in detail, wouldn't that program be able to discover the physical layout? > And then, wouldn't there be the possibility to feed that information to > the kernel? Even if the benchmark takes long, it only has to be run once > on each disk. >
No. Modern hard rivers are dar more complex than that. All the ones I have seen in the research labs are including at least a 200 MHZ ARM processor on board. Basically because of silicon feature size designers have found themselves trying to fill put space. So now have imbeeded processors on many hard drivers. They are really smart now days. From the discussion I have seen on this board it appears most people are stuck talking baout 5 to 10 year old drive technology. If people really want to try and squezzy that last 5-10% performance out of the drivers, go talk to the engineers, someone who actually designs them. The preformance increases will be a lot more and a lot easier to come by.
Mathew Monroe
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
| |