[U-Boot-Users] ATA driver in U-Boot

[U-Boot-Users] ATA driver in U-Boot

[John W. Linville]

On Fri, Oct 08, 2004 at 12:17:55PM -0800, Victor Wren wrote:

> Given that the kernel is being loaded out of flash (with all drivers compiled 
> in), and that the only reference to the IDE drive that U-Boot really knows 
> about is a commandline parameter passed to the kernel, would it cause any 
> conspicuous problems to remove ATA support from U-Boot, or does the kernel 
> depend on hardware information that U-Boot is providing (I notice there's a 
> "U-Boot" section in the kernel config)?

Victor,

It's been a little while since I looked at it, but I don't believe there
is any reason why U-Boot needs to have the IDE/ATA stuff built-in unless
U-Boot is going to use it.  I'm pretty sure it doesn't communicate w/
the kernel about it (i.e. IDE) at all.  Of course, YMMV... :-)

John
-- 
John W. Linville
linville at tuxdriver.com

[U-Boot-Users] ATA driver in U-Boot

[Victor Wren]

I am not sure if this is off-topic, because I'm unsure of the division of 
labor.

Is there any reason that U-Boot needs ATA support compiled in?  I ask because 
I'm trying to speed up booting on a small system with a hard drive.  U-Boot 
waits for the devices to become available, scanning the ATA bus, before 
proceeding to load the Linux kernel.  It appears to me that the kernel then 
does the same thing again, whereas if it went ahead and went straight to 
loading the kernel, it would give the drive another second or two to spin up 
and be ready for a bus scan.

Given that the kernel is being loaded out of flash (with all drivers compiled 
in), and that the only reference to the IDE drive that U-Boot really knows 
about is a commandline parameter passed to the kernel, would it cause any 
conspicuous problems to remove ATA support from U-Boot, or does the kernel 
depend on hardware information that U-Boot is providing (I notice there's a 
"U-Boot" section in the kernel config)?

Thanks,
Victor Wren

[U-Boot-Users] ATA driver in U-Boot

[Wolfgang Denk]

In message <41668573.16343.3D098D@localhost> you wrote:
> 
> Is there any reason that U-Boot needs ATA support compiled in?  I ask because 

No, there is not. Which is the reason that this is an option which is
configurable. And actually most of the boards don't enable it.

> I'm trying to speed up booting on a small system with a hard drive.  U-Boot 
> waits for the devices to become available, scanning the ATA bus, before 
> proceeding to load the Linux kernel.  It appears to me that the kernel then 
> does the same thing again, whereas if it went ahead and went straight to 
> loading the kernel, it would give the drive another second or two to spin up 
> and be ready for a bus scan.

If you enable ATA support in U-Boot you do this because you intend to
use the disk in U-Boot. If this will never be the  case  then  simply
don't enable it. You get what you ask for.

> Given that the kernel is being loaded out of flash (with all drivers compiled 
> in), and that the only reference to the IDE drive that U-Boot really knows 
> about is a commandline parameter passed to the kernel, would it cause any 
> conspicuous problems to remove ATA support from U-Boot, or does the kernel 
> depend on hardware information that U-Boot is providing (I notice there's a 
> "U-Boot" section in the kernel config)?

The kernel does not know which boot loader was running. Just  deiable
ATA support (or anything else) in U-Boot if you don't need it.

Best regards,

Wolfgang Denk

-- 
Software Engineering:  Embedded and Realtime Systems,  Embedded Linux
Phone: (+49)-8142-4596-87  Fax: (+49)-8142-4596-88  Email: wd at denx.de
1 1 was a race-horse, 2 2 was 1 2. When 1 1 1 1 race, 2 2 1 1 2.