Anyone familiar with IRQ11?

[Rod Pemberton]

Anyone familiar with IRQ11?

My understanding is that IRQ14 and IRQ15 are to be triggered
for IDE harddisks. However, this "new" 2009 motherboard is
triggering IRQ11 ... I'm thinking this may be due to IDE emulation
for SATA drives. I don't believe my earlier 2006 motherboard
did this, but I didn't have code to check for this at that time.

The BIOS PCI IRQ list upon boot up lists:

...
Native IDE Controller IRQ11
...
IDE Controller IRQ14
...

Has anyone else seen this? Could this be a BIOS setting?

Nothing is listed for IRQ15.

I haven't worked on my old OS in a while but was wondering if I
now need to enable IRQ11 and a harddisk routine for it.


Rod Pemberton

[CN]

On 10/18/2014 2:47 PM, Rod Pemberton wrote:
>
> Anyone familiar with IRQ11?

Yeah, it's an IRQ available for arbitrary PCI devices.

> My understanding is that IRQ14 and IRQ15 are to be triggered
> for IDE harddisks. However, this "new" 2009 motherboard is
> triggering IRQ11 ... I'm thinking this may be due to IDE emulation
> for SATA drives. I don't believe my earlier 2006 motherboard
> did this, but I didn't have code to check for this at that time.
>
> The BIOS PCI IRQ list upon boot up lists:
>
> ...
> Native IDE Controller IRQ11
> ...
> IDE Controller IRQ14
> ...

It looks like your system has a PCI IDE controller in Native mode in
addition to an IDE controller in Legacy mode. Native mode is different
from Legacy mode as follows: instead of being hardwired to legacy
resources (IRQ 14/15, ports 0x1f0 etc), the device fully conforms to the
PCI spec, so its I/O ranges are determined by its BAR registers, and its
interrupt line is a regularly routed PCI interrupt, which could be
shared with other PCI devices.

> Has anyone else seen this? Could this be a BIOS setting?

It's most likely a BIOS setting.

>
> Nothing is listed for IRQ15.

Check the IDE/SATA mode settings, usually it's a subset of
Legacy/Native/AHCI/RAID.

>
> I haven't worked on my old OS in a while but was wondering if I
> now need to enable IRQ11 and a harddisk routine for it.

Depends on what you want to get in the end. E.g. do you claim support
for PCI IDE in Native mode? If yes, then yes. If no, then no. Here's a
document, which may help you to understand the difference between Native
PCI and Legacy IDE:

http://www.bswd.com/pciide.pdf

Some basic knowledge of PCI config space may be required.

[Rod Pemberton]

Thanks. I probably have that document somewhere. I just haven't read it.

I was about to develop the IDE routines when I stopped development
some years ago. These would've been for "Legacy" IDE, i.e., portmapped.
That machine and the earlier one both had IDE drives. This machine
has SATA drives. I'm assuming the drives on this machine are emulated
or legacy when DOS boots, native for Linux. I'm still not quite sure
why IRQ11 would be triggering for emulated or legacy though when DOS
boots. I.e., I would expect IRQ11 in Linux, but not for DOS. Anyway,
if IRQ11 is correct for DOS or emulated IDE, that adds an extra issue,
i.e., no IRQ14 or some PCI setup to enable/keep it (?). It's unfortunate
that both of those machines had electrical failures. I'll have to pull
another machine out of storage to confirm proper IDE.


Rod Pemberton

[wolfgang kern]

Rod Pemberton wrote:

> Anyone familiar with IRQ11?

I see it very common used for PCI devices of all kind.

> My understanding is that IRQ14 and IRQ15 are to be triggered
> for IDE harddisks. However, this "new" 2009 motherboard is
> triggering IRQ11 ... I'm thinking this may be due to IDE emulation
> for SATA drives. I don't believe my earlier 2006 motherboard
> did this, but I didn't have code to check for this at that time.

> The BIOS PCI IRQ list upon boot up lists:
>
> ...
> Native IDE Controller IRQ11
> ...
> IDE Controller IRQ14
> ...

My PCI list show two harddisk devices:
classcode 01018A and 01018F and both are assigned to IRQ0A (10decimal).

The 8A type (legacy) doesn't show port number 01F0 nor 0170 in the BARs
while the 8F type lists two pairs of (IDE-base + control)-ports which
are the native SATA ports which I use the same way as the legacy ports
for up to four SATA drives in addition to four possible IDE-drives.

> Has anyone else seen this? Could this be a BIOS setting?

Not sure for an editable BIOS-setting.
The BIOS enumerates all PCI-devices and may assign IRQs at own will.

> Nothing is listed for IRQ15.

> I haven't worked on my old OS in a while but was wondering if I
> now need to enable IRQ11 and a harddisk routine for it.

What could such a HD routine do ?
we had a discussion about the sense of HD-IRQs in the past :)

I enable all IRQs and just count their occurance if there is nothing
else to do. But I disable HD-IRQs already in the controllers, so
there is nothing to count and this also saves on shared IRQ handling.

__
wolfgang

[Ross Ridge]

Rod Pemberton <buz...@nonamewhichexists.cmm> wrote:
>I was about to develop the IDE routines when I stopped development
>some years ago. These would've been for "Legacy" IDE, i.e., portmapped.
>That machine and the earlier one both had IDE drives. This machine
>has SATA drives. I'm assuming the drives on this machine are emulated
>or legacy when DOS boots, native for Linux. I'm still not quite sure
>why IRQ11 would be triggering for emulated or legacy though when DOS
>boots.

MS-DOS doesn't care what ports or interrupts the hard drive controller
uses. It accesses disk through the solely through BIOS, so it doesn't
matter whether the controller is in legacy or native mode. Or whether
it's using IDE, SATA, SCSI, ESDI or something that hasn't even been
invented yet. The BIOS deals with the controller, not MS-DOS.

So your assumpt that the controller switchs from legacy to native during
the boot process is false. If you set the controller to native mode
it will remain in native mode throughout the boot process. It's not
necessary nor helpful for it to switch to legacy mode to boot MS-DOS.

Ross Ridge

--
l/ // Ross Ridge -- The Great HTMU
[oo][oo] rri...@csclub.uwaterloo.ca
-()-/()/ http://www.csclub.uwaterloo.ca/~rridge/
db //