Messages in this thread | | | Date | Tue, 3 Apr 2012 12:52:19 -0400 | From | Boris Ostrovsky <> | Subject | Re: [PATCH] Use safe_halt() rather than halt() in acpi_idle_play_deay() |
| |
On 04/03/12 12:37, Luck, Tony: <tony.luck@intel.com wrote: > ACPI code is shared by arch/x86 and arch/ia64. ia64 doesn't provide a plain > "halt()" function. Use safe_halt() instead. > > Signed-off-by: Tony Luck<tony.luck@intel.com> > > --- > > E-mail discussion indicated this would be OK. Please check on x86 > before applying.
Tested-by: Boris Ostrovsky <boris.ostrovsky@amd.com>
-boris
> > diff --git a/drivers/acpi/processor_idle.c b/drivers/acpi/processor_idle.c > index b3447f6..f3decb3 100644 > --- a/drivers/acpi/processor_idle.c > +++ b/drivers/acpi/processor_idle.c > @@ -786,7 +786,7 @@ static int acpi_idle_play_dead(struct cpuidle_device *dev, int index) > while (1) { > > if (cx->entry_method == ACPI_CSTATE_HALT) > - halt(); > + safe_halt(); > else if (cx->entry_method == ACPI_CSTATE_SYSTEMIO) { > inb(cx->address); > /* See comment in acpi_idle_do_entry() */ >
| |