Page 1 of 1

ACPI 2.17 & Kernel 14.105

Posted: Wed Feb 07, 2007 16:33
by mikeforester
I just discovered the last fixpak, XR_C006, and installed it. Suddenly I can't boot with ACPI.PSD in CONFIG.SYS -- even using the W4 kernel. I get the message "OS/2 is unable to operate your hard drive or diskette. Please correct the problem and restart the system."

This happens with both the 14.105_W4 and 14.105_SMP kernels. It seems to occur just after OS2ASPI.DMD loads, which is just after OS2LVM.DMD. (I do not have SCSI hard drives.)

I've also tried different switches on ACPI.PSD, including /FS. I have not tried re-enabling apica.add. Marking IRQs 3, 5 and 7 as "Reserved" in the BIOS didn't help either.

I would have thought the presence of a PSD in CONFIG.SYS would have no effect on the W4 kernel.

My motherboard has an nForce4 chipset, and most previous versions of ACPI.PSD worked well with kernel 14.104a_SMP. Any ideas?

Posted: Wed Feb 07, 2007 16:35
by mikeforester
PS, I also just upgraded to DaniS506.add ver 1.7.9.

Posted: Wed Feb 07, 2007 16:40
by Eugene Gorbunoff
AMD X2? Don't forget add /APIC switch

Posted: Wed Feb 07, 2007 16:57
by mikeforester
Eugene Gorbunoff wrote:AMD X2? Don't forget add /APIC switch

It's there. Starting with 2.17 I'm using:

Code: Select all

PSD=ACPI.PSD /SMP /APIC /R

And yes, it's an AMD X2.

Posted: Wed Feb 07, 2007 17:12
by Eugene Gorbunoff
And.. don't use Fixpak6. IBM doesn't have testers and they are unable release working software for OS/2 anymore. Please use official eComStation kernel.

Posted: Wed Feb 07, 2007 18:02
by mikeforester
Eugene Gorbunoff wrote:And.. don't use Fixpak6. IBM doesn't have testers and they are unable release working software for OS/2 anymore. Please use official eComStation kernel.

Well, I realize that. I was just going by the thread "Initial XR_C006 experience" on comp.os.os2.misc. I can always back it out. But I have noticed others in this forum are using 14.105.

Kernel 14.105

Posted: Wed Feb 07, 2007 18:11
by Alan Beagley
mikeforester wrote:
Eugene Gorbunoff wrote:And.. don't use Fixpak6. IBM doesn't have testers and they are unable release working software for OS/2 anymore. Please use official eComStation kernel.

Well, I realize that. I was just going by the thread "Initial XR_C006 experience" on comp.os.os2.misc. I can always back it out. But I have noticed others in this forum are using 14.105.

My system with nForce4 (Asus A8N-E) now seems happy in SMP mode with Kernel 14.105 (from XR_C006) and ACPI 2.16. And using the suggestions in the 2.17 docs. to reserve IRQs 3, 5 and 7 in the mobo BIOS allows me to renable the OHCI driver and use AMouse again.

Re: Kernel 14.105

Posted: Wed Feb 07, 2007 18:33
by magog96
Alan Beagley wrote:My system with nForce4 (Asus A8N-E) now seems happy in SMP mode with Kernel 14.105 (from XR_C006) and ACPI 2.16. And using the suggestions in the 2.17 docs. to reserve IRQs 3, 5 and 7 in the mobo BIOS allows me to renable the OHCI driver and use AMouse again.


Could you post your ACPI.PSD parameters please?

Many thanks in advance!

Re: Kernel 14.105

Posted: Wed Feb 07, 2007 19:48
by Alan Beagley
magog96 wrote:
Alan Beagley wrote:My system with nForce4 (Asus A8N-E) now seems happy in SMP mode with Kernel 14.105 (from XR_C006) and ACPI 2.16. And using the suggestions in the 2.17 docs. to reserve IRQs 3, 5 and 7 in the mobo BIOS allows me to renable the OHCI driver and use AMouse again.

Could you post your ACPI.PSD parameters please?

Many thanks in advance!

PSD=ACPI.PSD /DV:0x57 /SMP /TMR /R

I also have the PSCOM.SYS driver instead of the standard COM.SYS.

Re: Kernel 14.105

Posted: Wed Feb 07, 2007 20:25
by mikeforester
Alan Beagley wrote:My system with nForce4 (Asus A8N-E) now seems happy in SMP mode with Kernel 14.105 (from XR_C006) and ACPI 2.16. And using the suggestions in the 2.17 docs. to reserve IRQs 3, 5 and 7 in the mobo BIOS allows me to renable the OHCI driver and use AMouse again.

Thanks, Alan. I went back to 2.16 but had the same problem, "unable to operate hard disk." I've now rolled back to 14.104a. However, later I ran CHKDSK and found a couple of hosed EA's on the C: drive; maybe that had something to do with it, maybe not.

I'll play around with it more another time.

However, I would point out to Evgeny and Pasha that in that same Usenet thread, Chuck McKinnis said he was testing XR_C006, which suggests he may release it for eCS.

Posted: Thu Feb 08, 2007 17:39
by Eugene Gorbunoff
NVidia requires:
ACPI.PSD /APIC /SMP /TMR

and don't update BIOS, don't change status of IRQs

Posted: Thu Feb 08, 2007 18:10
by Alan Beagley
Eugene Gorbunoff wrote:NVidia requires:
ACPI.PSD /APIC /SMP /TMR

and don't update BIOS, don't change status of IRQs

i've asked this question before somewhere but haven't seen an answeer:

My Asus A8N-E board came with BIOS ver. 1010. In the process of trying to get eCS SMP working on it, I had updated the BIOS to 1013 (the latest), but with no success. When I read that one should NOT update the BIOS, I backleveled it to 1004, the oldest that supports dual-core CPUs.

Everything does work now, but I made other changes too (e.g., a later version of the ACPI package and different parameters), so I cannot be sure that changing to the old BIOS is what solved the problem or was even necessary.

Which BIOS levels work, and which ones do not? What is likely to be the effect of updating mine again to the original 1010 or to 1013?

Posted: Fri Mar 09, 2007 19:10
by Alan Beagley
Alan Beagley wrote:
Eugene Gorbunoff wrote:NVidia requires:
ACPI.PSD /APIC /SMP /TMR

and don't update BIOS, don't change status of IRQs

i've asked this question before somewhere but haven't seen an answeer:

My Asus A8N-E board came with BIOS ver. 1010. In the process of trying to get eCS SMP working on it, I had updated the BIOS to 1013 (the latest), but with no success. When I read that one should NOT update the BIOS, I backleveled it to 1004, the oldest that supports dual-core CPUs.

Everything does work now, but I made other changes too (e.g., a later version of the ACPI package and different parameters), so I cannot be sure that changing to the old BIOS is what solved the problem or was even necessary.

Which BIOS levels work, and which ones do not? What is likely to be the effect of updating mine again to the original 1010 or to 1013?

I updated the motherboard BIOS to 1013 again, knowing that I could always backlevel it to 1004 again if necessary.

Everything seems to be working just fine still, so I don't understand the warning against updating the BIOS.