Re: [PATCH v4 2/9] powerpc/powernv/idle: Restore AMR/UAMOR/AMOR after idle

[Akshay Adiga <akshay.adiga@linux.vnet.ibm.com>]

On Fri, Mar 01, 2019 at 01:49:10AM +1100, Michael Ellerman wrote:
> In order to implement KUAP (Kernel Userspace Access Protection) on
> Power9 we will be using the AMR, and therefore indirectly the
> UAMOR/AMOR.
> 
> So save/restore these regs in the idle code.
> 
> Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
> ---
>  arch/powerpc/kernel/idle_book3s.S | 27 +++++++++++++++++++++++----
>  1 file changed, 23 insertions(+), 4 deletions(-)
> 
> v4: New.
> 
> diff --git a/arch/powerpc/kernel/idle_book3s.S b/arch/powerpc/kernel/idle_book3s.S
> index 36178000a2f2..4a860d3b9229 100644
> --- a/arch/powerpc/kernel/idle_book3s.S
> +++ b/arch/powerpc/kernel/idle_book3s.S
> @@ -170,8 +170,11 @@ END_FTR_SECTION_IFCLR(CPU_FTR_ARCH_300)
>  	bne-	core_idle_lock_held
>  	blr
> 
> -/* Reuse an unused pt_regs slot for IAMR */
> +/* Reuse some unused pt_regs slots for AMR/IAMR/UAMOR/UAMOR */
> +#define PNV_POWERSAVE_AMR	_TRAP
>  #define PNV_POWERSAVE_IAMR	_DAR
> +#define PNV_POWERSAVE_UAMOR	_DSISR
> +#define PNV_POWERSAVE_AMOR	RESULT
> 
>  /*
>   * Pass requested state in r3:
> @@ -205,8 +208,16 @@ END_FTR_SECTION_IFCLR(CPU_FTR_ARCH_300)
>  	SAVE_NVGPRS(r1)
> 
>  BEGIN_FTR_SECTION
> +	mfspr	r4, SPRN_AMR
>  	mfspr	r5, SPRN_IAMR
> +	mfspr	r6, SPRN_UAMOR
> +	std	r4, PNV_POWERSAVE_AMR(r1)
>  	std	r5, PNV_POWERSAVE_IAMR(r1)
> +	std	r6, PNV_POWERSAVE_UAMOR(r1)
> +BEGIN_FTR_SECTION_NESTED(42)
> +	mfspr	r7, SPRN_AMOR
> +	std	r7, PNV_POWERSAVE_AMOR(r1)
> +END_FTR_SECTION_NESTED_IFSET(CPU_FTR_HVMODE, 42)
>  END_FTR_SECTION_IFSET(CPU_FTR_ARCH_207S)

Can we save/restore for ARCH_300S only ?

I checked the user manual "POWER8 Processor User’s Manual for the
Single-Chip Module" v1.3 dated 16 March 2016. In section 9.7.2.2,
It mentions that AMOR and UAMOR will be preserved by hardware.

Excrept : 
" When an SMT thread switch is enabled, the napping thread’s
 resources can be given to other threads to improve core
 performance. Software must restore the architected state
 of the dormant thread upon exiting nap mode except for
 the following architected facilities, which are preserved
 by the hardware:
 - SLB State
 - All Hypervisor Special Registers (includes PURR, SPURR, AMOR,
 UAMOR, AMR, ACOP)
"

Keeping 207S has no harm, its just that we may spend a few
more cycles saving and restoring the sprs.

> 
>  	mfcr	r5
> @@ -935,12 +946,20 @@ END_FTR_SECTION_IFSET(CPU_FTR_HVMODE)
>  	REST_GPR(2, r1)
> 
>  BEGIN_FTR_SECTION
> -	/* IAMR was saved in pnv_powersave_common() */
> +	/* These regs were saved in pnv_powersave_common() */
> +	ld	r4, PNV_POWERSAVE_AMR(r1)
>  	ld	r5, PNV_POWERSAVE_IAMR(r1)
> +	ld	r6, PNV_POWERSAVE_UAMOR(r1)
> +	mtspr	SPRN_AMR, r4
>  	mtspr	SPRN_IAMR, r5
> +	mtspr	SPRN_UAMOR, r6
> +BEGIN_FTR_SECTION_NESTED(42)
> +	ld	r7, PNV_POWERSAVE_AMOR(r1)
> +	mtspr	SPRN_AMOR, r7
> +END_FTR_SECTION_NESTED_IFSET(CPU_FTR_HVMODE, 42)
>  	/*
> -	 * We don't need an isync here because the upcoming mtmsrd is
> -	 * execution synchronizing.
> +	 * We don't need an isync here after restoring IAMR because the upcoming
> +	 * mtmsrd is execution synchronizing.
>  	 */
>  END_FTR_SECTION_IFSET(CPU_FTR_ARCH_207S)
> 
> -- 
> 2.20.1
>