Re: [PATCH 3/3] KVM/arm64: enable enhanced armv8 fp/simd lazy switch

[Mario Smarduch <m.smarduch@samsung.com>]

On 11/6/2015 3:29 AM, Christoffer Dall wrote:
> On Thu, Nov 05, 2015 at 04:57:12PM -0800, Mario Smarduch wrote:
>>
>>
>> On 11/5/2015 7:02 AM, Christoffer Dall wrote:
>>> On Fri, Oct 30, 2015 at 02:56:33PM -0700, Mario Smarduch wrote:
>>>> This patch enables arm64 lazy fp/simd switch, similar to arm described in
>>>> second patch. Change from previous version - restore function is moved to
>>>> host. 
>>>>
>>>> Signed-off-by: Mario Smarduch <m.smarduch@samsung.com>
>>>> ---
>>>>  arch/arm64/include/asm/kvm_host.h |  2 +-
>>>>  arch/arm64/kernel/asm-offsets.c   |  1 +
>>>>  arch/arm64/kvm/hyp.S              | 37 +++++++++++++++++++++++++++++++------
>>>>  3 files changed, 33 insertions(+), 7 deletions(-)
>>>>
>>>> diff --git a/arch/arm64/include/asm/kvm_host.h b/arch/arm64/include/asm/kvm_host.h
>>>> index 26a2347..dcecf92 100644
>>>> --- a/arch/arm64/include/asm/kvm_host.h
>>>> +++ b/arch/arm64/include/asm/kvm_host.h
>>>> @@ -251,11 +251,11 @@ static inline void kvm_arch_hardware_unsetup(void) {}
>>>>  static inline void kvm_arch_sync_events(struct kvm *kvm) {}
>>>>  static inline void kvm_arch_vcpu_uninit(struct kvm_vcpu *vcpu) {}
>>>>  static inline void kvm_arch_sched_in(struct kvm_vcpu *vcpu, int cpu) {}
>>>> -static inline void kvm_restore_host_vfp_state(struct kvm_vcpu *vcpu) {}
>>>>  
>>>>  void kvm_arm_init_debug(void);
>>>>  void kvm_arm_setup_debug(struct kvm_vcpu *vcpu);
>>>>  void kvm_arm_clear_debug(struct kvm_vcpu *vcpu);
>>>>  void kvm_arm_reset_debug_ptr(struct kvm_vcpu *vcpu);
>>>> +void kvm_restore_host_vfp_state(struct kvm_vcpu *vcpu);
>>>>  
>>>>  #endif /* __ARM64_KVM_HOST_H__ */
>>>> diff --git a/arch/arm64/kernel/asm-offsets.c b/arch/arm64/kernel/asm-offsets.c
>>>> index 8d89cf8..c9c5242 100644
>>>> --- a/arch/arm64/kernel/asm-offsets.c
>>>> +++ b/arch/arm64/kernel/asm-offsets.c
>>>> @@ -124,6 +124,7 @@ int main(void)
>>>>    DEFINE(VCPU_HCR_EL2,		offsetof(struct kvm_vcpu, arch.hcr_el2));
>>>>    DEFINE(VCPU_MDCR_EL2,	offsetof(struct kvm_vcpu, arch.mdcr_el2));
>>>>    DEFINE(VCPU_IRQ_LINES,	offsetof(struct kvm_vcpu, arch.irq_lines));
>>>> +  DEFINE(VCPU_VFP_DIRTY,	offsetof(struct kvm_vcpu, arch.vfp_dirty));
>>>>    DEFINE(VCPU_HOST_CONTEXT,	offsetof(struct kvm_vcpu, arch.host_cpu_context));
>>>>    DEFINE(VCPU_HOST_DEBUG_STATE, offsetof(struct kvm_vcpu, arch.host_debug_state));
>>>>    DEFINE(VCPU_TIMER_CNTV_CTL,	offsetof(struct kvm_vcpu, arch.timer_cpu.cntv_ctl));
>>>> diff --git a/arch/arm64/kvm/hyp.S b/arch/arm64/kvm/hyp.S
>>>> index e583613..ed2c4cf 100644
>>>> --- a/arch/arm64/kvm/hyp.S
>>>> +++ b/arch/arm64/kvm/hyp.S
>>>> @@ -36,6 +36,28 @@
>>>>  #define CPU_SYSREG_OFFSET(x)	(CPU_SYSREGS + 8*x)
>>>>  
>>>>  	.text
>>>> +
>>>> +/**
>>>> + * void kvm_restore_host_vfp_state(struct vcpu *vcpu) - Executes lazy
>>>> + *	fp/simd switch, saves the guest, restores host. Called from host
>>>> + *	mode, placed outside of hyp section.
>>>
>>> same comments on style as previous patch
>> Got it.
>>>
>>>> + */
>>>> +ENTRY(kvm_restore_host_vfp_state)
>>>> +	push    xzr, lr
>>>> +
>>>> +	add     x2, x0, #VCPU_CONTEXT
>>>> +	mov     w3, #0
>>>> +	strb    w3, [x0, #VCPU_VFP_DIRTY]
>>>
>>> I've been discussing with myself if it would make more sense to clear
>>> the dirty flag in the C-code...
>> Since all the work is done here I placed it here.
> 
> yeah, that's what I thought first, but then I thought it's typically
> easier to understand the logic in the C-code and this is technically a
> side-effect from the name of the function "kvm_restore_host_vfp_state"
> which should then be "kvm_restore_host_vfp_state_and_clear_dirty_flag"
> :)
> 

Ok I'll set in C.
>>>
>>>> +
>>>> +	bl __save_fpsimd
>>>> +
>>>> +	ldr     x2, [x0, #VCPU_HOST_CONTEXT]
>>>> +	bl __restore_fpsimd
>>>> +
>>>> +	pop     xzr, lr
>>>> +	ret
>>>> +ENDPROC(kvm_restore_host_vfp_state)
>>>> +
>>>>  	.pushsection	.hyp.text, "ax"
>>>>  	.align	PAGE_SHIFT
>>>>  
>>>> @@ -482,7 +504,11 @@
>>>>  99:
>>>>  	msr     hcr_el2, x2
>>>>  	mov	x2, #CPTR_EL2_TTA
>>>> +
>>>> +	ldrb	w3, [x0, #VCPU_VFP_DIRTY]
>>>> +	tbnz    w3, #0, 98f
>>>>  	orr     x2, x2, #CPTR_EL2_TFP
>>>> +98:
>>>
>>> mmm, don't you need to only set the fpexc32 when you're actually going
>>> to trap the guest accesses?
>>
>> My understanding is you always need to set enable in fpexec32 for 32 bit guests,
>> otherwise EL1 would get the trap instead of EL2. Not sure if that's the point
>> you're making.
>>
> 
> If you're expecting to trap all accesses by setting CPTR_EL2_TFP and
> you're running a 32-bit guest, you must also enable in fpexec32, because
> otherwise the trap is not taken to EL2, but to EL1 instead.
> 
> Before these patches, we *always* expected to trap VFP accesses after
> entering the guest, but since that has now changed, you should only
> fiddle with fpexec32 if you are indeed trapping (first entry after
> vcpu_load) and if it's a 32-bit VM.
> 
> Does that help?
Yes sure does! Puts the vcpu_load jump to EL2 vs always into perspective.
> 
>>>
>>> also, you can consider only setting this in vcpu_load (jumping quickly
>>> to EL2 to do so) if we're running a 32-bit guest.  Probably worth
>>> measuring the difference between the extra EL2 jump on vcpu_load
>>> compared to hitting this register on every entry/exit.
>>
>> Sure, makes sense since this is a hot code path.
>>>
>>> Code-wise, it will be nicer to do it on vcpu_load.
>>>
>>>>  	msr	cptr_el2, x2
>>>>  
>>>>  	mov	x2, #(1 << 15)	// Trap CP15 Cr=15
>>>> @@ -669,14 +695,12 @@ __restore_debug:
>>>>  	ret
>>>>  
>>>>  __save_fpsimd:
>>>> -	skip_fpsimd_state x3, 1f
>>>>  	save_fpsimd
>>>> -1:	ret
>>>> +	ret
>>>>  
>>>>  __restore_fpsimd:
>>>> -	skip_fpsimd_state x3, 1f
>>>>  	restore_fpsimd
>>>> -1:	ret
>>>> +	ret
>>>>  
>>>>  switch_to_guest_fpsimd:
>>>>  	push	x4, lr
>>>> @@ -688,6 +712,9 @@ switch_to_guest_fpsimd:
>>>>  
>>>>  	mrs	x0, tpidr_el2
>>>>  
>>>> +	mov     w2, #1
>>>> +	strb    w2, [x0, #VCPU_VFP_DIRTY]
>>>
>>> hmm, just noticing this.  Are you not writing a 32-bit value to a
>>> potentially 8-bit field (ignoring padding in the struct), as the dirty
>>> flag is declared a bool.
>>
>> From the manuals byte stores require a word size registers on arm64/arm32.
>> The interconnect probably drops the remaining bytes.
>> Also double checked and the compiler uses same instructions.
>>
> 
> duh, I didn't see the 'b' in strb - I guess I was too tired to review
> patches.
> 
> -Christoffer
>