acpi problems with IBM/Lenovo/ThinkCentre

[John Baldwin]

On Tuesday 14 March 2006 09:11, Danny Braniss wrote:
> > On Tuesday 14 March 2006 02:34 am, Danny Braniss wrote:
> > > this is an intel/xeon dual core, booting with a 32bit kernel:
> > >
> > > FreeBSD 6.1-PRERELEASE #4: Mon Mar 13 09:52:47 IST 2006
> > >     danny at bsd:/r+d/obj/bsd/i386/r+d/6.1/src/sys/HUJI
> > > ACPI APIC Table: <PTLTD          APIC  >
> > > Timecounter "i8254" frequency 1193182 Hz quality 0
> > > CPU: Intel(R) Pentium(R) D CPU 2.80GHz (2793.01-MHz 686-class CPU)
> > >   Origin = "GenuineIntel"  Id = 0xf44  Stepping = 4
> > >  
> > > Features=0xbfebfbff<FPU,VME,DE,PSE,TSC,MSR,PAE,MCE,CX8,APIC,SEP,MTRR,PGE,MC
> > >A, C
> > > MOV,PAT,PSE36,CLFLUSH,DTS,ACPI,MMX,FXSR,SSE,SSE2,SS,HTT,TM,PBE>
> > >   Features2=0x641d<SSE3,RSVD2,MON,DS_CPL,CNTX-ID,CX16,<b14>>
> > >   AMD Features=0x20000000<LM>
> > >   Hyperthreading: 2 logical CPUs
> > > ...
> > > acpi0: <PTLTD   RSDT> on motherboard
> > >     ACPI-1304: *** Error: Method execution failed [\\_SB_.PCI0.USB1._PRW]
> > > (Node 0xc32861c0), AE_AML_NO_RETURN_VALUE
> > >     ACPI-0239: *** Error: Method execution failed [\\_SB_.PCI0.USB1._PRW]
> > > (Node 0xc32861c0), AE_AML_NO_RETURN_VALUE
> > >     ACPI-1304: *** Error: Method execution failed [\\_SB_.PCI0.USB2._PRW]
> > > (Node 0xc32860a0), AE_AML_NO_RETURN_VALUE
> > >     ACPI-0239: *** Error: Method execution failed [\\_SB_.PCI0.USB2._PRW]
> > > (Node 0xc32860a0), AE_AML_NO_RETURN_VALUE
> > >     ACPI-1304: *** Error: Method execution failed [\\_SB_.PCI0.USB3._PRW]
> > > (Node 0xc3247da0), AE_AML_NO_RETURN_VALUE
> > >     ACPI-0239: *** Error: Method execution failed [\\_SB_.PCI0.USB3._PRW]
> > > (Node 0xc3247da0), AE_AML_NO_RETURN_VALUE
> > >     ACPI-1304: *** Error: Method execution failed [\\_SB_.PCI0.USB4._PRW]
> > > (Node 0xc3247c80), AE_AML_NO_RETURN_VALUE
> > >     ACPI-0239: *** Error: Method execution failed [\\_SB_.PCI0.USB4._PRW]
> > > (Node 0xc3247c80), AE_AML_NO_RETURN_VALUE
> > >     ACPI-1304: *** Error: Method execution failed [\\_SB_.PCI0.USBE._PRW]
> > > (Node 0xc3247be0), AE_AML_NO_RETURN_VALUE
> > >     ACPI-0239: *** Error: Method execution failed [\\_SB_.PCI0.USBE._PRW]
> > > (Node 0xc3247be0), AE_AML_NO_RETURN_VALUE
> > > ...
> > > which probably break the usb stuff,
> > > but with a 64kernel:
> > > ...
> > >     ACPI-1304: *** Error: Method execution failed
> > > [\\_SB_.PCI0.LPC0.SIO_.COMA._ STA] (Node 0xffffff0000b8fc40),
> > > AE_AML_ALIGNMENT
> > >     ACPI-0239: *** Error: Method execution failed
> > > [\\_SB_.PCI0.LPC0.SIO_.COMA._ STA] (Node 0xffffff0000b8fc40),
> > > AE_AML_ALIGNMENT
> > > ...
> > 
> > Fixing these requires a MFC of changes to ACPI-CA to not require strict
> > alignment for amd64.  The ACPI-CA in 6.x requires strict alignment for
> > all 64-bit archs because ia64 requires it, but that has since been fixed
> > in current, just not MFC'd apparently.
> 
> assuming that there will be more amd64 than ia64, can the 
> ACPI_MISALIGNED_TRANSFERS
> be defined by default, and only undefed for ia64?

As I mentioned, it's already fixed in current and in the vendor ACPI-CA
sources.  We just need to backport that change to RELENG_6.  Hmm, Intel
actually changed the whole flag around (changed it to a different one
with an inverted sense).  Probably best to just sync up ACPI-CA in
RELENG_6 with ACPI-CA in head.  Nate, what do you think?

-- 
John Baldwin <jhb at FreeBSD.org>  <><  http://www.FreeBSD.org/~jhb/
"Power Users Use the Power to Serve"  =  http://www.FreeBSD.org