Phil Edwards wrote (edited):
> ... I've installed a 200GB Maxtor drive, and a Promise Ultra133
> TX2 card to let me actually use all of it.
>
> The mobo BIOS doesn't speak 48-bit LBA, so it sees a 137 GB drive.
> That's fine, I'm guessing, since (I'm told) Linux doesn't get its
> information from the BIOS.
>
> Windows 2000 sees the whole drive, and uses it with no problems. (Using
> Promise's supplied drivers.) I mention this only to point out that there
> doesn't /seem/ to be anything physically wrong with the drive, the card,
> the cable, etc.
>
> Booting 2.4.20 with "ide2=0x10d8,0x10d2" lets me see the new drive, along
> with a smaller drive on the same channel as slave:
I am not sure about your problem, but maybe I can add my ?0.02...
the second parameter of "ide2=0x10d8,0x10d2", i.e. 0x10d2, is the address
of an (only one) IDE register which is used for two things:
- If you write to it, you can enable the interrupt bit of the IDE
interface. Most of the time the IDE interface is used with its default
of "interrupt enabled" - so if this second address is false you will
not notice it there (you are writing to an invalid address, but that
does not usually trigger an error, and the default value of the register
is what you wanted to write anyways).
- Some software can decide to read this register because it is the
copy of the IDE status register - but unlike the real IDE status register
reading the copy does not acknowledge an IDE interrupt. One possible use
of this register is for the power saving drivers to know the state of
the disk without interfering with data read/written, for tools like
hdparm to get some information (special read polling modes)...
Unfortunately, most of the so-called IDE PCI boards (I am talking
here of my SIIG PCI card, maybe not yours) do not get this register
right (this SIIG returns 0 when you read this register, not the copy),
or implement an "upgraded" IDE interface to support RAID in an
undocumented way (writing another bit to this register?).
Testing on Redmond with their own driver doesn't prove that it is
a real IDE interface - the driver may volumtary never read the
copy register (to not acknowledge an interrupt from another request).
Also the driver may not care about power saving.
IMNSHO this copy register is absolutely needed in some cases
under Linux.
Unfortunately also, when this second parameter is completely wrong,
Linux continues to work approximately correctly.
What I can propose you is:
- to first double check with your documentation that you typed in
the right address 0x10d2 (I have also seen wrong documentation).
- to check under Redmond that the reserved I/O address for this
board is the documented one.
- to check under Linux the PCI description of reserved I/O address
- to check what the BIOS is thinking - not so easy...
Download Gujin:
http://sourceforge.net/projects/gujin
i.e.:
http://prdownloads.sourceforge.net/gujin/install-0.7.tgz?download
untar and and install it on a floppy:
./instboot boot.bin /dev/fd0 --full
(you need read/write access to /dev/fd0 if not root, read install.txt)
If you boot this floppy, it will display what is the address
reported by the Promise Ultra133 TX2 BIOS on the startup screen,
for the HD drive connected. Note that Gujin does not display an
IDE interface if no HD are connected.
As an extra information, my SIIG PCI card does not appear on this
screen, so I booted a very simple DOS to run dbgdisk.exe
(in standard.tgz) and it is easy to see the extra register always
reads as 0 on the "DBG" file created by dbgdisk.exe (read doc).
What a long ?0.02!
Etienne.
___________________________________________________________
Do You Yahoo!? -- Une adresse @yahoo.fr gratuite et en fran?ais !
Yahoo! Mail : http://fr.mail.yahoo.com
On Thu, May 29, 2003 at 01:09:03PM +0200, Etienne Lorrain wrote:
>
> What I can propose you is:
> - to first double check with your documentation that you typed in
> the right address 0x10d2 (I have also seen wrong documentation).
The steps I followed to arrive at those numbers was the Ultra-DMA HOWTO:
http://www.tldp.org/HOWTO/mini/Ultra-DMA-5.html#ss5.1
Once I had recompiled a new 2.4.20 with all the goodies, the ide2=
parameter became unneeded. At least, I can boot without it. Still having
GRUB problems, but no additional drive errors lately. (Knock on wooden
backup tapes.)
I will try and do the other steps this week.
Thanks,
Phil
--
If ye love wealth greater than liberty, the tranquility of servitude greater
than the animating contest for freedom, go home and leave us in peace. We seek
not your counsel, nor your arms. Crouch down and lick the hand that feeds you;
and may posterity forget that ye were our countrymen. - Samuel Adams