Before upgrading our PowerEdge 2650 (dual HT Xeons, Tigon3,
aic7899, workspace on sw raid5 over 4 disks, ext3) to RH9,
I gave 2.6.0-test2 a spin. Worked fine, except for one thing.
In 2.4, CONFIG_SMP automatically uses acpitable.c to detect
secondary threads via the MADT (since MPS doesn't handle them).
In 2.6.0-test2, with CONFIG_SMP and CONFIG_ACPI_HT_ONLY, this
doesn't happen, _unless_ I also pass acpismp=force on the command
line. Without acpismp=force, it only finds two CPUs.
The logic in arch/i386/kernel/setup.c, which defaults acpi to
disabled if HT_ONLY is chosen, seems backwards. Surely if I
configure HT_ONLY it's because I want to use it, no?
/Mikael
You're right.
This was an ill-fated attempt at backwards compatibility.
I removed acpismp=force in an ACPI cleanup a short time ago, and it
should
hit the tree via the ACPI maintainer after Andy returns from vacation.
Cheers,
-Len
> -----Original Message-----
> From: Mikael Pettersson [mailto:[email protected]]
> Sent: Wednesday, August 06, 2003 10:40 AM
> To: [email protected]
> Subject: 2.6.0-test2 on Dell PE2650, ACPI_HT_ONLY strangeness
>
>
> Before upgrading our PowerEdge 2650 (dual HT Xeons, Tigon3,
> aic7899, workspace on sw raid5 over 4 disks, ext3) to RH9,
> I gave 2.6.0-test2 a spin. Worked fine, except for one thing.
>
> In 2.4, CONFIG_SMP automatically uses acpitable.c to detect
> secondary threads via the MADT (since MPS doesn't handle them).
>
> In 2.6.0-test2, with CONFIG_SMP and CONFIG_ACPI_HT_ONLY, this
> doesn't happen, _unless_ I also pass acpismp=force on the command
> line. Without acpismp=force, it only finds two CPUs.
>
> The logic in arch/i386/kernel/setup.c, which defaults acpi to
> disabled if HT_ONLY is chosen, seems backwards. Surely if I
> configure HT_ONLY it's because I want to use it, no?
>
> /Mikael
> -
> To unsubscribe from this list: send the line "unsubscribe
> linux-kernel" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/
>
Brown, Len writes:
> You're right.
>
> This was an ill-fated attempt at backwards compatibility.
> I removed acpismp=force in an ACPI cleanup a short time ago, and it
> should
> hit the tree via the ACPI maintainer after Andy returns from vacation.
Great! Thanks.
/Mikael
>
> Cheers,
> -Len
>
> > -----Original Message-----
> > From: Mikael Pettersson [mailto:[email protected]]
> > Sent: Wednesday, August 06, 2003 10:40 AM
> > To: [email protected]
> > Subject: 2.6.0-test2 on Dell PE2650, ACPI_HT_ONLY strangeness
> >
> >
> > Before upgrading our PowerEdge 2650 (dual HT Xeons, Tigon3,
> > aic7899, workspace on sw raid5 over 4 disks, ext3) to RH9,
> > I gave 2.6.0-test2 a spin. Worked fine, except for one thing.
> >
> > In 2.4, CONFIG_SMP automatically uses acpitable.c to detect
> > secondary threads via the MADT (since MPS doesn't handle them).
> >
> > In 2.6.0-test2, with CONFIG_SMP and CONFIG_ACPI_HT_ONLY, this
> > doesn't happen, _unless_ I also pass acpismp=force on the command
> > line. Without acpismp=force, it only finds two CPUs.
> >
> > The logic in arch/i386/kernel/setup.c, which defaults acpi to
> > disabled if HT_ONLY is chosen, seems backwards. Surely if I
> > configure HT_ONLY it's because I want to use it, no?
> >
> > /Mikael
> > -
> > To unsubscribe from this list: send the line "unsubscribe
> > linux-kernel" in
> > the body of a message to [email protected]
> > More majordomo info at http://vger.kernel.org/majordomo-info.html
> > Please read the FAQ at http://www.tux.org/lkml/
> >