Re: [PATCH 1/1] powerpc/e500/qemu-e500: allow core to idle without waiting

2022-02-14 Thread Michael Ellerman 
On Wed, 12 Jan 2022 12:24:59 +0100, Joachim Wiberg wrote:
> From: Tobias Waldekranz 
> 
> This means an idle guest won't needlessly consume an entire core on
> the host, waiting for work to show up.
> 
> 

Applied to powerpc/next.

[1/1] powerpc/e500/qemu-e500: allow core to idle without waiting
  https://git.kernel.org/powerpc/c/f529edd1b69ddf832c3257dcd34e15100038d6b7

cheers

Re: [PATCH 1/1] powerpc/e500/qemu-e500: allow core to idle without waiting

2022-01-14 Thread Scott Wood 
On Wed, 2022-01-12 at 12:24 +0100, Joachim Wiberg wrote:
> From: Tobias Waldekranz 
> 
> This means an idle guest won't needlessly consume an entire core on
> the host, waiting for work to show up.
> 
> Signed-off-by: Tobias Waldekranz 
> Signed-off-by: Joachim Wiberg 
> ---
>  arch/powerpc/platforms/85xx/qemu_e500.c | 5 +
>  1 file changed, 5 insertions(+)
> 
> diff --git a/arch/powerpc/platforms/85xx/qemu_e500.c
> b/arch/powerpc/platforms/85xx/qemu_e500.c
> index a4127b0b161f..4c4d577effd9 100644
> --- a/arch/powerpc/platforms/85xx/qemu_e500.c
> +++ b/arch/powerpc/platforms/85xx/qemu_e500.c
> @@ -67,4 +67,9 @@ define_machine(qemu_e500) {
> .get_irq= mpic_get_coreint_irq,
> .calibrate_decr = generic_calibrate_decr,
> .progress   = udbg_progress,
> +#ifdef CONFIG_PPC64
> +   .power_save = book3e_idle,
> +#else
> +   .power_save = e500_idle,
> +#endif
>  };

Acked-by: Scott Wood 

-Scott

Re: [PATCH 1/1] powerpc/e500/qemu-e500: allow core to idle without waiting

2022-01-14 Thread Joachim Wiberg 
On Wed, Jan 12, 2022 at 06:36, Scott Wood  wrote:
> On Wed, 2022-01-12 at 12:24 +0100, Joachim Wiberg wrote:
>> [snip]
>> +#ifdef CONFIG_PPC64
>> +   .power_save = book3e_idle,
>> +#else
>> +   .power_save = e500_idle,
>> +#endif
>>  };
> In the 32-bit case shouldn't this already be getting added by
> setup_power_save()?  Though I see corenet_generic.c doing the same thing...

We went for consistency with the corenet_generic.c setup, which as you
noted does the same thing.

Best regards
 /Joachim

Re: [PATCH 1/1] powerpc/e500/qemu-e500: allow core to idle without waiting

2022-01-12 Thread Scott Wood 
On Wed, 2022-01-12 at 12:24 +0100, Joachim Wiberg wrote:
> From: Tobias Waldekranz 
> 
> This means an idle guest won't needlessly consume an entire core on
> the host, waiting for work to show up.
> 
> Signed-off-by: Tobias Waldekranz 
> Signed-off-by: Joachim Wiberg 
> ---
>  arch/powerpc/platforms/85xx/qemu_e500.c | 5 +
>  1 file changed, 5 insertions(+)
> 
> diff --git a/arch/powerpc/platforms/85xx/qemu_e500.c
> b/arch/powerpc/platforms/85xx/qemu_e500.c
> index a4127b0b161f..4c4d577effd9 100644
> --- a/arch/powerpc/platforms/85xx/qemu_e500.c
> +++ b/arch/powerpc/platforms/85xx/qemu_e500.c
> @@ -67,4 +67,9 @@ define_machine(qemu_e500) {
> .get_irq= mpic_get_coreint_irq,
> .calibrate_decr = generic_calibrate_decr,
> .progress   = udbg_progress,
> +#ifdef CONFIG_PPC64
> +   .power_save = book3e_idle,
> +#else
> +   .power_save = e500_idle,
> +#endif
>  };

In the 32-bit case shouldn't this already be getting added by
setup_power_save()?  Though I see corenet_generic.c doing the same thing...

-Scott

[PATCH 1/1] powerpc/e500/qemu-e500: allow core to idle without waiting

2022-01-12 Thread Joachim Wiberg 
From: Tobias Waldekranz 

This means an idle guest won't needlessly consume an entire core on
the host, waiting for work to show up.

Signed-off-by: Tobias Waldekranz 
Signed-off-by: Joachim Wiberg 
---
 arch/powerpc/platforms/85xx/qemu_e500.c | 5 +
 1 file changed, 5 insertions(+)

diff --git a/arch/powerpc/platforms/85xx/qemu_e500.c 
b/arch/powerpc/platforms/85xx/qemu_e500.c
index a4127b0b161f..4c4d577effd9 100644
--- a/arch/powerpc/platforms/85xx/qemu_e500.c
+++ b/arch/powerpc/platforms/85xx/qemu_e500.c
@@ -67,4 +67,9 @@ define_machine(qemu_e500) {
.get_irq= mpic_get_coreint_irq,
.calibrate_decr = generic_calibrate_decr,
.progress   = udbg_progress,
+#ifdef CONFIG_PPC64
+   .power_save = book3e_idle,
+#else
+   .power_save = e500_idle,
+#endif
 };
-- 
2.25.1