RE: (U)DMA Explained (Re: Kernel freeze on boot (.123-.125) in I

Daniel Nash (danash@students.wisc.edu)
Thu, 22 Oct 1998 11:30:35 -0500 (CDT)


On 22-Oct-98 Andre M. Hedrick wrote:
> On Wed, 21 Oct 1998, Daniel Nash wrote:
>
>> hda: IBM-DAQA-33240, 3098MB w/96kB Cache, CHS=787/128/63, DMA
>> hdb: WDC AC31600H, 1549MB w/128kB Cache, CHS=787/64/63, DMA
>
> First Question:
>
> Are IBM-DAQA-33240 and WDC AC31600H are UDMA by default???
>
> If they are, you need to move them off the 430HX PIIX3.
> This silly controller thinks that it can do UDMA, because it can decode
> the difference in the hardware. However, you would have seen (U)DMA if
> you could turn off in the BIOS timing settings.

Both are older, non-UDMA drives. That's why I left them on the motherboard.

> What is the mainboard nad bios revision?

Abit IT5H rev 1.5 I'll check the BIOS revision when I get home later.

>> hde: IBM-DHEA-38451, 8063MB w/472kB Cache, CHS=16383/16/63, UDMA
>
> There are now three DMA reporting states of IDE-devices.
>
> A trailing tag with UDMA is considered, 100% native.
> A trailing tag with (U)DMA is considered, native capable (forgot to enable)
> A trailing tag with DMA is considered, straight Multi/Single Word native.
>
> There are currently no reporting point to date that report the mode level.
> That is DMA can be DMA mode 2/1/0 as multi or single word, thus 6 possible
> mode levels are blanketed under this trailing TAG. Note that (U)DMA is
> reverted to any of these mode levels, also. UDMA will be UDMA mode 2/1/0,
> thus 3possiblemode levels are blanketed under this trailing TAG.

Ok, are you suggesting that I should somehow find out which mode is correct
for each drive and set it with something like hdparm? How might I do this?

-- Daniel Nash
"Waiter, Waiter, there's an avocado in my guacamole." - Me

-
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/