On Thu, 16 May 2002, Stefan Bambach wrote:

> - What is the FORCE_IDE_CONTROLLER_ON for ?
>   Do you ignore the pci configuration register ? So a
>   disable IDE channel in BIOS (normal Award, ...) is
>   ignored and activated anyway ? So all IDE channels will
>   be activated and usable from within linux ?

The problem is that Linux interprets "BIOS did not enable IDE controller"
as meaning "BIOS DISABLED IDE CONTROLLER". This is a bad assumption.
FORCE_IDE_CONTROLLER_ON is a way of making linux enable the controller.

This is just a problem in controllers. How should linux interpret the
enable bits. What we really need is a command line way to tell linux what
to turn on. But the IDE command line parameters are a bit of a mess.

> - What is the WAIT_HDA_SPINUP for ?

This forces linux to wait for hda to spin up. Otherwise it decides that
hda does not exist. Personally I think this is a mistake in IDE. Drive
present should not be the same as drive spun up. Their mistake.

> - Everyone using IDE drives have to apply the ide-patch and
>   wait 10-15 seconds until the drive is ready ?

10-15? it's more like 4. You should be able to say this:
hda=0,0,0
and linux will force a wait on hda spin-up. But that seems to fail
sometimes.

> - Is there a IDE drive that does not have this long spin up
>   time ?
don't know.

> - Do I need the spinup patch, if I boot from DoC and only
>   whats to use the ide drive for data ?

maybe. What do you mean, data?

> - Does this IDE Problem mean, that I cannot use IDE drives,
>   if I want a usable system (music box) within 5 seconds ?

You should boot linux from Flash, and have a user-mode program insmod the
ide module.

IDE needs repair in Linux. But it is close.

ron

Reply via email to