hda=none hda=noprobe is ignored by <=2.6.15-26

hda=none hda=noprobe is ignored by <=2.6.15-26

[koko]

First some old Ignored(sadly) links to the same bug :
https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=132708
http://lkml.org/lkml/2003/6/16/29
http://www.gatago.com/linux/kernel/6116284.html

This bug is ASUS VIA MB specific. Asus A7V133, hda hdb empty.

>From Alan Cox (alan@redhat.com) 	on 2005-04-07 04:41 EST	[reply]	 	
>BIOS problem I believe not kernel. We only probe hda/hdb because the BIOS
>claims the slot may have a drive on it.

So why kernel probes IDE channel I just ordered him to IGNORE?

[snip]
[17179569.184000] Kernel command line: root=/dev/hdc3 ro quiet splash
hda=noprobe hdb=noprobe
[17179569.184000] ide_setup: hda=noprobe
[17179569.184000] ide_setup: hdb=noprobe
[snip]
[17179573.696000] VP_IDE: VIA vt82c686b (rev 40) IDE UDMA100
controller on pci0000:00:04.1
[17179573.696000]     ide0: BM-DMA at 0xd800-0xd807, BIOS settings:
hda:pio, hdb:pio
[17179573.696000]     ide1: BM-DMA at 0xd808-0xd80f, BIOS settings:
hdc:DMA, hdd:pio
[17179573.696000] Probing IDE interface ide0...
[snip]
 Here we go, and ....
[snip]
[17179577.060000] Probing IDE interface ide0...
[17179611.848000] ide0: Wait for ready failed before probe !
[snip]

almost a minute here :/

Heres how it looks without noprobe :
[17179572.788000] VP_IDE: VIA vt82c686b (rev 40) IDE UDMA100
controller on pci0000:00:04.1
[17179572.788000]     ide0: BM-DMA at 0xd800-0xd807, BIOS settings:
hda:pio, hdb:pio
[17179572.788000]     ide1: BM-DMA at 0xd808-0xd80f, BIOS settings:
hdc:DMA, hdd:pio
[17179572.788000] Probing IDE interface ide0...
[17179578.248000] hda: IRQ probe failed (0xfffffcf8)
[17179583.472000] hda: IRQ probe failed (0xfffffcf8)
[17179583.472000] hda: no response (status = 0x0a), resetting drive
[17179588.808000] hda: IRQ probe failed (0xfffffcf8)
[17179588.808000] hda: no response (status = 0x0a)
[17179594.312000] hdb: IRQ probe failed (0xfffffcf8)
[17179599.536000] hdb: IRQ probe failed (0xfffffcf8)
[17179599.536000] hdb: no response (status = 0x0a), resetting drive
[17179604.872000] hdb: IRQ probe failed (0xfffffcf8)
[17179604.872000] hdb: no response (status = 0x0a)
[17179604.928000] Probing IDE interface ide1...
[snip]
[17179608.240000] Probing IDE interface ide0...
[17179643.028000] ide0: Wait for ready failed before probe !
[17179648.476000] hda: IRQ probe failed (0xffff7cf0)
[17179653.700000] hda: IRQ probe failed (0xffff7cf0)
[17179653.700000] hda: no response (status = 0x0a), resetting drive
[17179659.036000] hda: IRQ probe failed (0xffff7cf0)
[17179659.036000] hda: no response (status = 0x0a)
[17179664.540000] hdb: IRQ probe failed (0xffff7cf0)
[17179669.764000] hdb: IRQ probe failed (0xffff7cf0)
[17179669.764000] hdb: no response (status = 0x0a), resetting drive
[17179675.100000] hdb: IRQ probe failed (0xffff7cf0)
[17179675.100000] hdb: no response (status = 0x0a)
[17179675.156000] Probing IDE interface ide2...

That one takes almost 3 minutes.

Why kernel probes IDE channels twice? Why it ignores noprobe in the
second probe? How is Asus A7V133 Bios broken (how can I check how its
broken) and how can I fix it?
Please dont ignore this message like kernel noprobe :)

PS:distro Ubuntu 6.06 and 2 random liveCD ones so its not distro specific.

-- 
Who logs in to gdm? Not I, said the duck.

Re: hda=none hda=noprobe is ignored by <=2.6.15-26

[Adam Henley]

As I understand it..reading Documentation/ide.txt:
"hdx=noprobe" :  drive may be present, but do not probe for it

If the CMOS/BIOS is saying there is a drive there then the kernel may
be respecting that and looking for it there...I guess that would be
why Alan gave you the response he did.

There is an option:
"hdx=none"              : drive is NOT present, ignore cmos and do not probe

Did that work for you when you tried it? It looks like that option
will explicitly ignore the CMOS.

Apologies if I'm misinterpreting the docs/misunderstanding how the
kernel works, I'm inexperienced and worse the wear for beer.

Thanks,
adam

Re: hda=none hda=noprobe is ignored by <=2.6.15-26

[Alan Cox]

Ar Sad, 2006-08-05 am 00:29 +0200, ysgrifennodd koko:
> So why kernel probes IDE channel I just ordered him to IGNORE?

hda=noprobe or failing that hda=none ought to be sufficient. If you need
the latter make sure your CMOS drive settings are correct too.

> [17179577.060000] Probing IDE interface ide0...
> [17179611.848000] ide0: Wait for ready failed before probe !
> [snip]
> 
> almost a minute here :/

That indicates a possible pull-up resistor problem too. The kernel is
seeing the cable as having a drive on it that is not responding so waits
the required spin up time.