kernelparameter idex=noprobe

kernelparameter idex=noprobe

[Markus Kossmann]

Hallo,
according to linux-2.6.4/Documentation/ide.txt :
"idex=noprobe"         : do not attempt to access/use this interface
SuSEs inofficial 2.6.4-0 kernel ( precompiled from rpm) comes with 
CONFIG_BLK_DEV_SIIMAGE=y
CONFIG_SCSI_SATA_SIL=m

Trying to use the sata_sil driver for my sil3112 controler I did give
ide4=noprobe ide5=noprobe hdi=noprobe hdj=noprobe hdk=noprobe hdl=noprobe
on the kernel commandline

The Bootup- messages show:
[...]
6>Using ACPI (MADT) for SMP configuration information
<4>Built 1 zonelists
<4>Kernel command line: root=/dev/hde2 vga=0x317 desktop hdc=ide-scsi hdclun=0 
splash=silent 3
ide4=noprobe ide5=noprobe hdi=noprobe hdj=noprobe hdk=noprobe hdl=noprobe
<6>ide_setup: hdc=ide-scsi
<6>ide_setup: hdclun=0
<6>bootsplash: silent mode.
<6>ide_setup: ide4=noprobe
<6>ide_setup: ide5=noprobe
<6>ide_setup: hdi=noprobe
<6>ide_setup: hdj=noprobe
<6>ide_setup: hdk=noprobe
<6>ide_setup: hdl=noprobe
[...]
6>ide: Assuming 33MHz system bus speed for PIO modes; override with idebus=xx
<6>AMD7411: IDE controller at PCI slot 0000:00:07.1
<6>AMD7411: chipset revision 1
<6>AMD7411: not 100%% native mode: will probe irqs later
<6>AMD7411: 0000:00:07.1 (rev 01) UDMA100 controller
<6>    ide0: BM-DMA at 0xf000-0xf007, BIOS settings: hda:DMA, hdb:pio
<6>    ide1: BM-DMA at 0xf008-0xf00f, BIOS settings: hdc:DMA, hdd:pio
<4>hda: DM166D DVD-ROM, ATAPI CD/DVD-ROM drive
<4>ide0 at 0x1f0-0x1f7,0x3f6 on irq 14
<4>hdc: PLEXTOR DVDR PX-708A, ATAPI CD/DVD-ROM drive
<4>ide1 at 0x170-0x177,0x376 on irq 15
<6>PDC20270: IDE controller at PCI slot 0000:00:09.0
<6>PDC20270: chipset revision 2
<6>PDC20270: 100%% native mode on irq 177
<6>    ide2: BM-DMA at 0x1830-0x1837, BIOS settings: hde:pio, hdf:pio
<6>    ide3: BM-DMA at 0x1838-0x183f, BIOS settings: hdg:pio, hdh:pio
<4>hde: SAMSUNG SP0612N, ATA DISK drive
<4>ide2 at 0x1868-0x186f,0x1862 on irq 177
<6>SiI3112 Serial ATA: IDE controller at PCI slot 0000:00:0a.0
<6>SiI3112 Serial ATA: chipset revision 2
<6>SiI3112 Serial ATA: 100%% native mode on irq 185
<6>    ide4: MMIO-DMA , BIOS settings: hdi:pio, hdj:pio
<6>    ide5: MMIO-DMA , BIOS settings: hdk:pio, hdl:pio
<4>hde: max request size: 1024KiB
<6>hde: 117304992 sectors (60060 MB) w/2048KiB Cache, CHS=16383/255/63, 
UDMA(100)
<6> hde: hde1 < hde5 hde6 hde7 hde8 hde9 > hde2 hde3 hde4
[...]

Why is the SIL3112 taken over by the siimage driver ?
Did I misinterpret ide.txt, is ide.txt wrong or is this a bug ? 

hdi=noprobe seems to work however. The attached SATA disk doesn't show up.  
And it does show up if I leave out hdi=noprobe. 

Is there any chance to use sata_sil with that kernel configuration ?
Or is recompiling with CONFIG_BLK_DEV_SIIMAGE=m or with siimage disabled the 
only option ? 

Re: kernelparameter idex=noprobe

[Bartlomiej Zolnierkiewicz]

On Friday 12 of March 2004 18:26, Markus Kossmann wrote:
> Hallo,
> according to linux-2.6.4/Documentation/ide.txt :
> "idex=noprobe"         : do not attempt to access/use this interface
> SuSEs inofficial 2.6.4-0 kernel ( precompiled from rpm) comes with
> CONFIG_BLK_DEV_SIIMAGE=y
> CONFIG_SCSI_SATA_SIL=m
>
> Trying to use the sata_sil driver for my sil3112 controler I did give
> ide4=noprobe ide5=noprobe hdi=noprobe hdj=noprobe hdk=noprobe hdl=noprobe
> on the kernel commandline
>
> The Bootup- messages show:
> [...]
> 6>Using ACPI (MADT) for SMP configuration information
> <4>Built 1 zonelists
> <4>Kernel command line: root=/dev/hde2 vga=0x317 desktop hdc=ide-scsi
> hdclun=0 splash=silent 3
> ide4=noprobe ide5=noprobe hdi=noprobe hdj=noprobe hdk=noprobe hdl=noprobe
> <6>ide_setup: hdc=ide-scsi
> <6>ide_setup: hdclun=0
> <6>bootsplash: silent mode.
> <6>ide_setup: ide4=noprobe
> <6>ide_setup: ide5=noprobe
> <6>ide_setup: hdi=noprobe
> <6>ide_setup: hdj=noprobe
> <6>ide_setup: hdk=noprobe
> <6>ide_setup: hdl=noprobe
> [...]
> 6>ide: Assuming 33MHz system bus speed for PIO modes; override with
> idebus=xx <6>AMD7411: IDE controller at PCI slot 0000:00:07.1
> <6>AMD7411: chipset revision 1
> <6>AMD7411: not 100%% native mode: will probe irqs later
> <6>AMD7411: 0000:00:07.1 (rev 01) UDMA100 controller
> <6>    ide0: BM-DMA at 0xf000-0xf007, BIOS settings: hda:DMA, hdb:pio
> <6>    ide1: BM-DMA at 0xf008-0xf00f, BIOS settings: hdc:DMA, hdd:pio
> <4>hda: DM166D DVD-ROM, ATAPI CD/DVD-ROM drive
> <4>ide0 at 0x1f0-0x1f7,0x3f6 on irq 14
> <4>hdc: PLEXTOR DVDR PX-708A, ATAPI CD/DVD-ROM drive
> <4>ide1 at 0x170-0x177,0x376 on irq 15
> <6>PDC20270: IDE controller at PCI slot 0000:00:09.0
> <6>PDC20270: chipset revision 2
> <6>PDC20270: 100%% native mode on irq 177
> <6>    ide2: BM-DMA at 0x1830-0x1837, BIOS settings: hde:pio, hdf:pio
> <6>    ide3: BM-DMA at 0x1838-0x183f, BIOS settings: hdg:pio, hdh:pio
> <4>hde: SAMSUNG SP0612N, ATA DISK drive
> <4>ide2 at 0x1868-0x186f,0x1862 on irq 177
> <6>SiI3112 Serial ATA: IDE controller at PCI slot 0000:00:0a.0
> <6>SiI3112 Serial ATA: chipset revision 2
> <6>SiI3112 Serial ATA: 100%% native mode on irq 185
> <6>    ide4: MMIO-DMA , BIOS settings: hdi:pio, hdj:pio
> <6>    ide5: MMIO-DMA , BIOS settings: hdk:pio, hdl:pio
> <4>hde: max request size: 1024KiB
> <6>hde: 117304992 sectors (60060 MB) w/2048KiB Cache, CHS=16383/255/63,
> UDMA(100)
> <6> hde: hde1 < hde5 hde6 hde7 hde8 hde9 > hde2 hde3 hde4
> [...]
>
> Why is the SIL3112 taken over by the siimage driver ?
> Did I misinterpret ide.txt, is ide.txt wrong or is this a bug ?

It is not a bug - siimage driver is compiled in and PCI device is detected.
It has nothing to do with "idex=noprobe" or "hdx=noprobe".

> hdi=noprobe seems to work however. The attached SATA disk doesn't show up.
> And it does show up if I leave out hdi=noprobe.
>
> Is there any chance to use sata_sil with that kernel configuration ?
> Or is recompiling with CONFIG_BLK_DEV_SIIMAGE=m or with siimage disabled
> the only option ?

Yep.

Regards,
Bartlomiej

Booting off of IDE while using different libata drives on same southbridge

[J. Ryan Earl]

I am having a similar problem to what Markus Kossmann wrote about, but 
with the VIA Southbridge (Asus K8V).  My situation is similar, but a 
little different.  I would like to boot off a PATA drive attached to the 
Southbridge, but use libata for a couple SATA drives attached to the 
same Southbridge.

Is this still not possible?  I also tried hde/hdg=noprobe options, but 
they didn't help the situation.  It appears the only way to get the 
drives on sata_via is to boot off of them.  Am I correct in thinking 
this is the only way to go about this?

-ryan

Bartlomiej Zolnierkiewicz wrote:

>>Is there any chance to use sata_sil with that kernel configuration ?
>>Or is recompiling with CONFIG_BLK_DEV_SIIMAGE=m or with siimage disabled
>>the only option ?
>>    
>>
>
>Yep.
>
>Regards,
>Bartlomiej
>

Re: Booting off of IDE while using different libata drives on same southbridge

[Bartlomiej Zolnierkiewicz]

On Monday 03 of May 2004 23:19, J. Ryan Earl wrote:
> I am having a similar problem to what Markus Kossmann wrote about, but
> with the VIA Southbridge (Asus K8V).  My situation is similar, but a
> little different.  I would like to boot off a PATA drive attached to the
> Southbridge, but use libata for a couple SATA drives attached to the
> same Southbridge.
>
> Is this still not possible?  I also tried hde/hdg=noprobe options, but
> they didn't help the situation.  It appears the only way to get the
> drives on sata_via is to boot off of them.  Am I correct in thinking
> this is the only way to go about this?

Did you actually tried it (booting off of them)?
[ I can't see how this can help. ]

Just don't compile-in generic IDE PCI driver which controls your SATA drives
(or don't load this module if you're using initrd).

> -ryan
>
> Bartlomiej Zolnierkiewicz wrote:
> >>Is there any chance to use sata_sil with that kernel configuration ?
> >>Or is recompiling with CONFIG_BLK_DEV_SIIMAGE=m or with siimage disabled
> >>the only option ?
> >
> >Yep.
> >
> >Regards,
> >Bartlomiej

Re: Booting off of IDE while using different libata drives on same southbridge

[J. Ryan Earl]

Bartlomiej Zolnierkiewicz wrote:

>On Monday 03 of May 2004 23:19, J. Ryan Earl wrote:
>  
>
>>I am having a similar problem to what Markus Kossmann wrote about, but
>>with the VIA Southbridge (Asus K8V).  My situation is similar, but a
>>little different.  I would like to boot off a PATA drive attached to the
>>Southbridge, but use libata for a couple SATA drives attached to the
>>same Southbridge.
>>
>>Is this still not possible?  I also tried hde/hdg=noprobe options, but
>>they didn't help the situation.  It appears the only way to get the
>>drives on sata_via is to boot off of them.  Am I correct in thinking
>>this is the only way to go about this?
>>    
>>
>
>Did you actually tried it (booting off of them)?
>[ I can't see how this can help. ]
>  
>
With libata linked into the kernel, and ide as a module, I can boot off 
of libata.

>Just don't compile-in generic IDE PCI driver which controls your SATA drives
>(or don't load this module if you're using initrd).
>  
>
As I said, I want to boot off of the PATA drives attached to the 
southbridge.  To boot of them, I have to have the IDE driver compiled 
into the kernel; if I do this I can't use libata on that southbridge, 
the IDE driver take precedence for the serial ata channels.

-ryan

Re: Booting off of IDE while using different libata drives on same southbridge

[Bartlomiej Zolnierkiewicz]

On Tuesday 04 of May 2004 02:56, you wrote:
> Bartlomiej Zolnierkiewicz wrote:
> >On Monday 03 of May 2004 23:19, J. Ryan Earl wrote:
> >>I am having a similar problem to what Markus Kossmann wrote about, but
> >>with the VIA Southbridge (Asus K8V).  My situation is similar, but a
> >>little different.  I would like to boot off a PATA drive attached to the
> >>Southbridge, but use libata for a couple SATA drives attached to the
> >>same Southbridge.
> >>
> >>Is this still not possible?  I also tried hde/hdg=noprobe options, but
> >>they didn't help the situation.  It appears the only way to get the
> >>drives on sata_via is to boot off of them.  Am I correct in thinking
> >>this is the only way to go about this?
> >
> >Did you actually tried it (booting off of them)?
> >[ I can't see how this can help. ]
>
> With libata linked into the kernel, and ide as a module, I can boot off
> of libata.

Therefore you changed order in which these drivers are initialized,
it has nothing to do with booting off of SATA drives.

> >Just don't compile-in generic IDE PCI driver which controls your SATA
> > drives (or don't load this module if you're using initrd).
>
> As I said, I want to boot off of the PATA drives attached to the
> southbridge.  To boot of them, I have to have the IDE driver compiled
> into the kernel; if I do this I can't use libata on that southbridge,
> the IDE driver take precedence for the serial ata channels.

Just don't compile generic IDE PCI driver!!
CONFIG_BLK_DEV_GENERIC=n

VIA PATA controller uses via82cxxx driver not generic IDE PCI one.

> -ryan

Re: Booting off of IDE while using different libata drives on same southbridge

[J. Ryan Earl]

Bartlomiej Zolnierkiewicz wrote:

> Just don't compile generic IDE PCI driver!!
>
>CONFIG_BLK_DEV_GENERIC=n
>
>VIA PATA controller uses via82cxxx driver not generic IDE PCI one.
>  
>
Ahhh, I knew I was making a mistake.  I didn't realize I was using the 
generic IDE driver.

Thank you sir!
-ryan