[PATCH v2 0/2] KVM/arm64: add fp/simd lazy switch support

[PATCH v2 0/2] KVM/arm64: add fp/simd lazy switch support

[Mario Smarduch]

This patch series is a followup to the armv7 fp/simd lazy switch
implementation, uses similar approach and depends on the series -
https://lists.cs.columbia.edu/pipermail/kvmarm/2015-September/016567.html
Patches are based on 4.3-rc2 commit 1f93e4a96c91093

Patches are based on earlier arm64 fp/simd optimization work -
https://lists.cs.columbia.edu/pipermail/kvmarm/2015-July/015748.html

And subsequent fixes by Marc and Christoffer at KVM Forum hackathon to handle
32-bit guest on 64 bit host - 
https://lists.cs.columbia.edu/pipermail/kvmarm/2015-August/016128.html

The patch series have been tested on Foundation Model arm64/arm64 and
arm32/arm64. The test program used can be found here 

https://github.com/mjsmar/arm-arm64-fpsimd-test

Launched upto 16 instances on 4-way Guest and another 16 on the host (both 
cases 1mS sleep), ran overnight. 

Changes v1->v2:
- Tested arm32/arm64
- rebased to 4.3-rc2
- changed a couple register accesses from 64 to 32 bit 

Mario Smarduch (2):
  add hooks for armv8 fp/simd lazy switch
  enable armv8 fp/simd lazy switch

 arch/arm/kvm/arm.c                |  2 --
 arch/arm64/include/asm/kvm_asm.h  |  1 +
 arch/arm64/include/asm/kvm_host.h |  3 ++
 arch/arm64/kernel/asm-offsets.c   |  1 +
 arch/arm64/kvm/hyp.S              | 59 ++++++++++++++++++++++++++-------------
 5 files changed, 45 insertions(+), 21 deletions(-)

-- 
1.9.1

[PATCH v2 1/2] add hooks for armv8 fp/simd lazy switch

[Mario Smarduch]

This patch adds hooks to support fp/simd lazy switch. A vcpu flag to track
fp/simd state, and flag offset in vcpu structure.

Signed-off-by: Mario Smarduch <m.smarduch@samsung.com>
---
 arch/arm64/include/asm/kvm_host.h | 3 +++
 arch/arm64/kernel/asm-offsets.c   | 1 +
 2 files changed, 4 insertions(+)

diff --git a/arch/arm64/include/asm/kvm_host.h b/arch/arm64/include/asm/kvm_host.h
index 4562459..03f25d0 100644
--- a/arch/arm64/include/asm/kvm_host.h
+++ b/arch/arm64/include/asm/kvm_host.h
@@ -157,6 +157,9 @@ struct kvm_vcpu_arch {
 	/* Interrupt related fields */
 	u64 irq_lines;		/* IRQ and FIQ levels */
 
+	/* Track fp/simd lazy switch */
+	u32 vfp_lazy;
+
 	/* Cache some mmu pages needed inside spinlock regions */
 	struct kvm_mmu_memory_cache mmu_page_cache;
 
diff --git a/arch/arm64/kernel/asm-offsets.c b/arch/arm64/kernel/asm-offsets.c
index 8d89cf8..8311da4 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_LAZY,         offsetof(struct kvm_vcpu, arch.vfp_lazy));
   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));
-- 
1.9.1

[PATCH v2 2/2] enable armv8 fp/simd lazy switch

[Mario Smarduch]

This patch enables arm64 lazy fp/simd switch. Removes the ARM constraint,
and follows the same approach as armv7 version - found here.

https://lists.cs.columbia.edu/pipermail/kvmarm/2015-September/016567.html

To summarize - provided the guest accesses fp/simd unit we limit number
of fp/simd context switches to two per vCPU execution schedule.

Signed-off-by: Mario Smarduch <m.smarduch@samsung.com>
---
 arch/arm/kvm/arm.c               |  2 --
 arch/arm64/include/asm/kvm_asm.h |  1 +
 arch/arm64/kvm/hyp.S             | 59 +++++++++++++++++++++++++++-------------
 3 files changed, 41 insertions(+), 21 deletions(-)

diff --git a/arch/arm/kvm/arm.c b/arch/arm/kvm/arm.c
index 1b1f9e9..fe609f1 100644
--- a/arch/arm/kvm/arm.c
+++ b/arch/arm/kvm/arm.c
@@ -112,12 +112,10 @@ void kvm_arch_check_processor_compat(void *rtn)
  */
 static void kvm_switch_fp_regs(struct kvm_vcpu *vcpu)
 {
-#ifdef CONFIG_ARM
 	if (vcpu->arch.vfp_lazy == 1) {
 		kvm_call_hyp(__kvm_restore_host_vfp_state, vcpu);
 		vcpu->arch.vfp_lazy = 0;
 	}
-#endif
 }
 
 /**
diff --git a/arch/arm64/include/asm/kvm_asm.h b/arch/arm64/include/asm/kvm_asm.h
index 5e37710..83dcac5 100644
--- a/arch/arm64/include/asm/kvm_asm.h
+++ b/arch/arm64/include/asm/kvm_asm.h
@@ -117,6 +117,7 @@ extern char __kvm_hyp_vector[];
 extern void __kvm_flush_vm_context(void);
 extern void __kvm_tlb_flush_vmid_ipa(struct kvm *kvm, phys_addr_t ipa);
 extern void __kvm_tlb_flush_vmid(struct kvm *kvm);
+extern void __kvm_restore_host_vfp_state(struct kvm_vcpu *vcpu);
 
 extern int __kvm_vcpu_run(struct kvm_vcpu *vcpu);
 
diff --git a/arch/arm64/kvm/hyp.S b/arch/arm64/kvm/hyp.S
index e583613..ea99f66 100644
--- a/arch/arm64/kvm/hyp.S
+++ b/arch/arm64/kvm/hyp.S
@@ -385,14 +385,6 @@
 	tbz	\tmp, #KVM_ARM64_DEBUG_DIRTY_SHIFT, \target
 .endm
 
-/*
- * Branch to target if CPTR_EL2.TFP bit is set (VFP/SIMD trapping enabled)
- */
-.macro skip_fpsimd_state tmp, target
-	mrs	\tmp, cptr_el2
-	tbnz	\tmp, #CPTR_EL2_TFP_SHIFT, \target
-.endm
-
 .macro compute_debug_state target
 	// Compute debug state: If any of KDE, MDE or KVM_ARM64_DEBUG_DIRTY
 	// is set, we do a full save/restore cycle and disable trapping.
@@ -433,10 +425,6 @@
 	mrs	x5, ifsr32_el2
 	stp	x4, x5, [x3]
 
-	skip_fpsimd_state x8, 2f
-	mrs	x6, fpexc32_el2
-	str	x6, [x3, #16]
-2:
 	skip_debug_state x8, 1f
 	mrs	x7, dbgvcr32_el2
 	str	x7, [x3, #24]
@@ -481,8 +469,15 @@
 	isb
 99:
 	msr     hcr_el2, x2
-	mov	x2, #CPTR_EL2_TTA
+
+	mov     x2, #0
+	ldr     w3, [x0, #VCPU_VFP_LAZY]
+	tbnz    w3, #0, 98f
+
 	orr     x2, x2, #CPTR_EL2_TFP
+98:
+	orr     x2, x2, #CPTR_EL2_TTA
+
 	msr	cptr_el2, x2
 
 	mov	x2, #(1 << 15)	// Trap CP15 Cr=15
@@ -669,14 +664,12 @@ __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 +681,9 @@ switch_to_guest_fpsimd:
 
 	mrs	x0, tpidr_el2
 
+	mov     w2, #1
+	str     w2, [x0, #VCPU_VFP_LAZY]
+
 	ldr	x2, [x0, #VCPU_HOST_CONTEXT]
 	kern_hyp_va x2
 	bl __save_fpsimd
@@ -763,7 +759,6 @@ __kvm_vcpu_return:
 	add	x2, x0, #VCPU_CONTEXT
 
 	save_guest_regs
-	bl __save_fpsimd
 	bl __save_sysregs
 
 	skip_debug_state x3, 1f
@@ -784,7 +779,6 @@ __kvm_vcpu_return:
 	kern_hyp_va x2
 
 	bl __restore_sysregs
-	bl __restore_fpsimd
 	/* Clear FPSIMD and Trace trapping */
 	msr     cptr_el2, xzr
 
@@ -863,6 +857,33 @@ ENTRY(__kvm_flush_vm_context)
 	ret
 ENDPROC(__kvm_flush_vm_context)
 
+/**
+ * kvm_switch_fp_regs() - switch guest/host VFP/SIMD registers
+ * @vcpu:      pointer to vcpu structure.
+ *
+ */
+ENTRY(__kvm_restore_host_vfp_state)
+	push    x4, lr
+
+	kern_hyp_va x0
+	add     x2, x0, #VCPU_CONTEXT
+
+	// Load Guest HCR, determine if guest is 32 or 64 bit
+	ldr     x3, [x0, #VCPU_HCR_EL2]
+	tbnz    x3, #HCR_RW_SHIFT, 1f
+	mrs     x4, fpexc32_el2
+	str     x4, [x2, #CPU_SYSREG_OFFSET(FPEXC32_EL2)]
+1:
+	bl __save_fpsimd
+
+	ldr     x2, [x0, #VCPU_HOST_CONTEXT]
+	kern_hyp_va x2
+	bl __restore_fpsimd
+
+	pop     x4, lr
+	ret
+ENDPROC(__kvm_restore_host_vfp_state)
+
 __kvm_hyp_panic:
 	// Guess the context by looking@VTTBR:
 	// If zero, then we're already a host.
-- 
1.9.1

[PATCH v2 2/2] enable armv8 fp/simd lazy switch

[Christoffer Dall]

On Mon, Oct 05, 2015 at 07:33:14PM -0700, Mario Smarduch wrote:
> This patch enables arm64 lazy fp/simd switch. Removes the ARM constraint,
> and follows the same approach as armv7 version - found here.
> 
> https://lists.cs.columbia.edu/pipermail/kvmarm/2015-September/016567.html
> 
> To summarize - provided the guest accesses fp/simd unit we limit number
> of fp/simd context switches to two per vCPU execution schedule.
> 
> Signed-off-by: Mario Smarduch <m.smarduch@samsung.com>
> ---
>  arch/arm/kvm/arm.c               |  2 --
>  arch/arm64/include/asm/kvm_asm.h |  1 +
>  arch/arm64/kvm/hyp.S             | 59 +++++++++++++++++++++++++++-------------
>  3 files changed, 41 insertions(+), 21 deletions(-)
> 
> diff --git a/arch/arm/kvm/arm.c b/arch/arm/kvm/arm.c
> index 1b1f9e9..fe609f1 100644
> --- a/arch/arm/kvm/arm.c
> +++ b/arch/arm/kvm/arm.c
> @@ -112,12 +112,10 @@ void kvm_arch_check_processor_compat(void *rtn)
>   */
>  static void kvm_switch_fp_regs(struct kvm_vcpu *vcpu)
>  {
> -#ifdef CONFIG_ARM
>  	if (vcpu->arch.vfp_lazy == 1) {
>  		kvm_call_hyp(__kvm_restore_host_vfp_state, vcpu);
>  		vcpu->arch.vfp_lazy = 0;
>  	}
> -#endif

I'm assuming you'll rework this like for the 32-bit series to just do
this in EL1, yes?

>  }
>  
>  /**
> diff --git a/arch/arm64/include/asm/kvm_asm.h b/arch/arm64/include/asm/kvm_asm.h
> index 5e37710..83dcac5 100644
> --- a/arch/arm64/include/asm/kvm_asm.h
> +++ b/arch/arm64/include/asm/kvm_asm.h
> @@ -117,6 +117,7 @@ extern char __kvm_hyp_vector[];
>  extern void __kvm_flush_vm_context(void);
>  extern void __kvm_tlb_flush_vmid_ipa(struct kvm *kvm, phys_addr_t ipa);
>  extern void __kvm_tlb_flush_vmid(struct kvm *kvm);
> +extern void __kvm_restore_host_vfp_state(struct kvm_vcpu *vcpu);
>  
>  extern int __kvm_vcpu_run(struct kvm_vcpu *vcpu);
>  
> diff --git a/arch/arm64/kvm/hyp.S b/arch/arm64/kvm/hyp.S
> index e583613..ea99f66 100644
> --- a/arch/arm64/kvm/hyp.S
> +++ b/arch/arm64/kvm/hyp.S
> @@ -385,14 +385,6 @@
>  	tbz	\tmp, #KVM_ARM64_DEBUG_DIRTY_SHIFT, \target
>  .endm
>  
> -/*
> - * Branch to target if CPTR_EL2.TFP bit is set (VFP/SIMD trapping enabled)
> - */
> -.macro skip_fpsimd_state tmp, target
> -	mrs	\tmp, cptr_el2
> -	tbnz	\tmp, #CPTR_EL2_TFP_SHIFT, \target
> -.endm
> -
>  .macro compute_debug_state target
>  	// Compute debug state: If any of KDE, MDE or KVM_ARM64_DEBUG_DIRTY
>  	// is set, we do a full save/restore cycle and disable trapping.
> @@ -433,10 +425,6 @@
>  	mrs	x5, ifsr32_el2
>  	stp	x4, x5, [x3]
>  
> -	skip_fpsimd_state x8, 2f
> -	mrs	x6, fpexc32_el2
> -	str	x6, [x3, #16]
> -2:
>  	skip_debug_state x8, 1f
>  	mrs	x7, dbgvcr32_el2
>  	str	x7, [x3, #24]
> @@ -481,8 +469,15 @@
>  	isb
>  99:
>  	msr     hcr_el2, x2
> -	mov	x2, #CPTR_EL2_TTA
> +
> +	mov     x2, #0
> +	ldr     w3, [x0, #VCPU_VFP_LAZY]
> +	tbnz    w3, #0, 98f
> +
>  	orr     x2, x2, #CPTR_EL2_TFP
> +98:
> +	orr     x2, x2, #CPTR_EL2_TTA
> +

this can be reworked more simply by using an extra register and leaving
the initial mov x2, #CPTR_EL2_TTA alone and just skip setting the TFP
bit if we've already switched the state.

>  	msr	cptr_el2, x2
>  
>  	mov	x2, #(1 << 15)	// Trap CP15 Cr=15
> @@ -669,14 +664,12 @@ __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 +681,9 @@ switch_to_guest_fpsimd:
>  
>  	mrs	x0, tpidr_el2
>  
> +	mov     w2, #1
> +	str     w2, [x0, #VCPU_VFP_LAZY]
> +
>  	ldr	x2, [x0, #VCPU_HOST_CONTEXT]
>  	kern_hyp_va x2
>  	bl __save_fpsimd
> @@ -763,7 +759,6 @@ __kvm_vcpu_return:
>  	add	x2, x0, #VCPU_CONTEXT
>  
>  	save_guest_regs
> -	bl __save_fpsimd
>  	bl __save_sysregs
>  
>  	skip_debug_state x3, 1f
> @@ -784,7 +779,6 @@ __kvm_vcpu_return:
>  	kern_hyp_va x2
>  
>  	bl __restore_sysregs
> -	bl __restore_fpsimd
>  	/* Clear FPSIMD and Trace trapping */
>  	msr     cptr_el2, xzr
>  
> @@ -863,6 +857,33 @@ ENTRY(__kvm_flush_vm_context)
>  	ret
>  ENDPROC(__kvm_flush_vm_context)
>  
> +/**
> + * kvm_switch_fp_regs() - switch guest/host VFP/SIMD registers
> + * @vcpu:      pointer to vcpu structure.
> + *
> + */
> +ENTRY(__kvm_restore_host_vfp_state)
> +	push    x4, lr

even when you move this to el1, I don't think you need to do the
additional store of the registers as this is going to be called from
C-code and you should have plenty of caller-save registers, if I read
the aarch64 PCS correctly.

> +
> +	kern_hyp_va x0
> +	add     x2, x0, #VCPU_CONTEXT
> +
> +	// Load Guest HCR, determine if guest is 32 or 64 bit
> +	ldr     x3, [x0, #VCPU_HCR_EL2]
> +	tbnz    x3, #HCR_RW_SHIFT, 1f
> +	mrs     x4, fpexc32_el2
> +	str     x4, [x2, #CPU_SYSREG_OFFSET(FPEXC32_EL2)]
> +1:
> +	bl __save_fpsimd
> +
> +	ldr     x2, [x0, #VCPU_HOST_CONTEXT]
> +	kern_hyp_va x2
> +	bl __restore_fpsimd
> +
> +	pop     x4, lr
> +	ret
> +ENDPROC(__kvm_restore_host_vfp_state)
> +
>  __kvm_hyp_panic:
>  	// Guess the context by looking at VTTBR:
>  	// If zero, then we're already a host.
> -- 
> 1.9.1
> 

Otherwise this looks reasonably good.


Thanks,
-Christoffer

[PATCH v2 1/2] add hooks for armv8 fp/simd lazy switch

[Christoffer Dall]

On Mon, Oct 05, 2015 at 07:33:13PM -0700, Mario Smarduch wrote:
> This patch adds hooks to support fp/simd lazy switch. A vcpu flag to track
> fp/simd state, and flag offset in vcpu structure.
> 
> Signed-off-by: Mario Smarduch <m.smarduch@samsung.com>
> ---
>  arch/arm64/include/asm/kvm_host.h | 3 +++
>  arch/arm64/kernel/asm-offsets.c   | 1 +
>  2 files changed, 4 insertions(+)
> 
> diff --git a/arch/arm64/include/asm/kvm_host.h b/arch/arm64/include/asm/kvm_host.h
> index 4562459..03f25d0 100644
> --- a/arch/arm64/include/asm/kvm_host.h
> +++ b/arch/arm64/include/asm/kvm_host.h
> @@ -157,6 +157,9 @@ struct kvm_vcpu_arch {
>  	/* Interrupt related fields */
>  	u64 irq_lines;		/* IRQ and FIQ levels */
>  
> +	/* Track fp/simd lazy switch */
> +	u32 vfp_lazy;
> +


like for the 32-bit patch set I think this could have a more meaningful
name with more clearly defined semantics, e.g. vfp_host_dirty.  You
could also call it vfp_flags like the debug flags and define a specific
flag for your use...

Thanks,
-Christoffer


>  	/* Cache some mmu pages needed inside spinlock regions */
>  	struct kvm_mmu_memory_cache mmu_page_cache;
>  
> diff --git a/arch/arm64/kernel/asm-offsets.c b/arch/arm64/kernel/asm-offsets.c
> index 8d89cf8..8311da4 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_LAZY,         offsetof(struct kvm_vcpu, arch.vfp_lazy));
>    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));
> -- 
> 1.9.1
>