Re: [OFFTOPIC] Re: IDE-DMA strangeness (another one)

Mark Lord (mlord@pobox.com)
Fri, 04 Dec 1998 00:11:30 +0000


"Alex A.M.R. Slingerland" wrote:
...
> On my ASUS P5A with Ali 5 chipset, a Quantum EL 5.1
> (and K6-2-300, 128Mb PC100 SDRAM), with hdparm 3.5 on Linux
> 2.1.126 and later, I get something like:
>
> $ hdparm -tT /dev/hda
> bla-bla 128Mb: 51 Mb/s
> bla-bla 64Mb : 11.19 Mb/s
>
> It's using PIO 4 (as linux doesn't currently support
...
> Is this drive actually doing 11 Mb/s in PIO (not that

Yes, it is. For sequential reads. If you check the
manufacturers website for specs, you'll find that a media
rate of 88mb/sec (8 * 11MB/sec) is well within the capabilities
of the drive.

> Does this mean I'm using (roughly, at least, on
> average) 11.19/16.66 ~= 2/3 of the CPU's processing
> power PIO-ing things off the drive (in e.g. a long

No. It just means that the drive can deliver data
at at least 11MByte/sec from the platters. Nothing more.

>...could I still expect a
> (significant) increase in
> _hdparm_measured_throughput_ when using UDMA/DMA

Maybe. If the drive had "ideal" firmware, then you shouldn't
see any difference with hdparm (lightly loaded system).
But some drives are better optimized for DMA than others.

My own drives deliver >13MByte/sec in PIO4, DMA2, and UDMA2.
Other drives I have seen are slower when not in UDMA2.

With DMA, though, the CPU consumption drops by a substantial
margin during I/O, a plus under any circumstances.

-- 
mlord@pobox.com

- 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/