Re: UDMA

mlord (mlord@pobox.com)
Thu, 18 Dec 1997 08:50:06 -0500

Messages sorted by: [ date ][ thread ][ subject ][ author ]
Next message: Bill Hawes: "Re: 2.1.72 /proc, was: Re: /proc/memmap [name from inode?]"
Previous message: Richard B. Johnson: "Re: Pentium bug fix"
Maybe in reply to: Roy P. Turner: "UDMA"

Boris Seifert wrote:
>
> On Dec 17, 21:57, mlord wrote:
>
> > The 32-bit IO and multcount work only for PIO, not DMA.
> > On some drives (?), multcount might even slow down DMA,
> > though there is no evidence to support this theory.
> > Neither of these buys anything when DMA is used.
>
> How come it (seems to) work here ?
>
> [root@clito /root]# hdparm /dev/hda
>
> /dev/hda:
> multcount = 0 (off)
> I/O support = 1 (32-bit) <---here
> unmaskirq = 0 (off)
> using_dma = 1 (on) <---and here

In this case, "32-bit" I/O is not actually "working".
Sure, it's enabled, but having zero effect..

The concept of multcount and "32-bit" I/O just plain
has nothing to do with DMA transfers. No positive effect.

-- 
mlord@pobox.com
The Linux IDE guy

Next message: Bill Hawes: "Re: 2.1.72 /proc, was: Re: /proc/memmap [name from inode?]"
Previous message: Richard B. Johnson: "Re: Pentium bug fix"
Maybe in reply to: Roy P. Turner: "UDMA"