Re: [PATCH v2 03/21] arm64: KVM: Implement vgic-v2 save/restore

[Marc Zyngier <marc.zyngier@arm.com>]

On 30/11/15 20:00, Christoffer Dall wrote:
> On Fri, Nov 27, 2015 at 06:49:57PM +0000, Marc Zyngier wrote:
>> Implement the vgic-v2 save restore (mostly) as a direct translation
>> of the assembly code version.
>>
>> Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
>> ---
>>  arch/arm64/kvm/Makefile         |  1 +
>>  arch/arm64/kvm/hyp/Makefile     |  5 +++
>>  arch/arm64/kvm/hyp/hyp.h        |  3 ++
>>  arch/arm64/kvm/hyp/vgic-v2-sr.c | 89 +++++++++++++++++++++++++++++++++++++++++
>>  4 files changed, 98 insertions(+)
>>  create mode 100644 arch/arm64/kvm/hyp/Makefile
>>  create mode 100644 arch/arm64/kvm/hyp/vgic-v2-sr.c
>>
>> diff --git a/arch/arm64/kvm/Makefile b/arch/arm64/kvm/Makefile
>> index 1949fe5..d31e4e5 100644
>> --- a/arch/arm64/kvm/Makefile
>> +++ b/arch/arm64/kvm/Makefile
>> @@ -10,6 +10,7 @@ KVM=../../../virt/kvm
>>  ARM=../../../arch/arm/kvm
>>  
>>  obj-$(CONFIG_KVM_ARM_HOST) += kvm.o
>> +obj-$(CONFIG_KVM_ARM_HOST) += hyp/
>>  
>>  kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/kvm_main.o $(KVM)/coalesced_mmio.o $(KVM)/eventfd.o $(KVM)/vfio.o
>>  kvm-$(CONFIG_KVM_ARM_HOST) += $(ARM)/arm.o $(ARM)/mmu.o $(ARM)/mmio.o
>> diff --git a/arch/arm64/kvm/hyp/Makefile b/arch/arm64/kvm/hyp/Makefile
>> new file mode 100644
>> index 0000000..d8d5968
>> --- /dev/null
>> +++ b/arch/arm64/kvm/hyp/Makefile
>> @@ -0,0 +1,5 @@
>> +#
>> +# Makefile for Kernel-based Virtual Machine module, HYP part
>> +#
>> +
>> +obj-$(CONFIG_KVM_ARM_HOST) += vgic-v2-sr.o
>> diff --git a/arch/arm64/kvm/hyp/hyp.h b/arch/arm64/kvm/hyp/hyp.h
>> index dac843e..78f25c4 100644
>> --- a/arch/arm64/kvm/hyp/hyp.h
>> +++ b/arch/arm64/kvm/hyp/hyp.h
>> @@ -27,5 +27,8 @@
>>  
>>  #define kern_hyp_va(v) (typeof(v))((unsigned long)v & HYP_PAGE_OFFSET_MASK)
>>  
>> +void __vgic_v2_save_state(struct kvm_vcpu *vcpu);
>> +void __vgic_v2_restore_state(struct kvm_vcpu *vcpu);
> 
> should we call these flush/sync here now ?
> 
>> +
>>  #endif /* __ARM64_KVM_HYP_H__ */
>>  
>> diff --git a/arch/arm64/kvm/hyp/vgic-v2-sr.c b/arch/arm64/kvm/hyp/vgic-v2-sr.c
>> new file mode 100644
>> index 0000000..29a5c1d
>> --- /dev/null
>> +++ b/arch/arm64/kvm/hyp/vgic-v2-sr.c
>> @@ -0,0 +1,89 @@
>> +/*
>> + * Copyright (C) 2012-2015 - ARM Ltd
>> + * Author: Marc Zyngier <marc.zyngier@arm.com>
>> + *
>> + * This program is free software; you can redistribute it and/or modify
>> + * it under the terms of the GNU General Public License version 2 as
>> + * published by the Free Software Foundation.
>> + *
>> + * This program is distributed in the hope that it will be useful,
>> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
>> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
>> + * GNU General Public License for more details.
>> + *
>> + * You should have received a copy of the GNU General Public License
>> + * along with this program.  If not, see <http://www.gnu.org/licenses/>.
>> + */
>> +
>> +#include <linux/compiler.h>
>> +#include <linux/irqchip/arm-gic.h>
>> +#include <linux/kvm_host.h>
>> +
>> +#include <asm/kvm_mmu.h>
>> +
>> +#include "hyp.h"
>> +
>> +/* vcpu is already in the HYP VA space */
> 
> should we annotate hyp pointers similarly to __user or will that be
> confusing when VHE enters the scene ?

I looked at doing that. That's a possibility, and I don't think that
would be too bad as long as we have kern_hyp_va() doing the (potentially
NOP) conversion. The only issue is that this is only enforced with
sparse, not by a usual compilation.

Still, this is a valid use case, and I'll try to invest some time doing
that.

> 
>> +void __hyp_text __vgic_v2_save_state(struct kvm_vcpu *vcpu)
>> +{
>> +	struct kvm *kvm = kern_hyp_va(vcpu->kvm);
>> +	struct vgic_v2_cpu_if *cpu_if = &vcpu->arch.vgic_cpu.vgic_v2;
>> +	struct vgic_dist *vgic = &kvm->arch.vgic;
>> +	void __iomem *base = kern_hyp_va(vgic->vctrl_base);
>> +	u32 __iomem *lr_base;
>> +	u32 eisr0, eisr1, elrsr0, elrsr1;
>> +	int i = 0, nr_lr;
>> +
>> +	if (!base)
>> +		return;
>> +
>> +	nr_lr = vcpu->arch.vgic_cpu.nr_lr;
>> +	cpu_if->vgic_vmcr = readl_relaxed(base + GICH_VMCR);
>> +	cpu_if->vgic_misr = readl_relaxed(base + GICH_MISR);
>> +	eisr0  = readl_relaxed(base + GICH_EISR0);
>> +	elrsr0 = readl_relaxed(base + GICH_ELRSR0);
>> +	if (unlikely(nr_lr > 32)) {
>> +		eisr1  = readl_relaxed(base + GICH_EISR1);
>> +		elrsr1 = readl_relaxed(base + GICH_ELRSR1);
>> +	} else {
>> +		eisr1 = elrsr1 = 0;
>> +	}
>> +#ifdef CONFIG_CPU_BIG_ENDIAN
>> +	cpu_if->vgic_eisr  = ((u64)eisr0 << 32) | eisr1;
>> +	cpu_if->vgic_elrsr = ((u64)elrsr0 << 32) | elrsr1;
>> +#else
>> +	cpu_if->vgic_eisr  = ((u64)eisr1 << 32) | eisr0;
>> +	cpu_if->vgic_elrsr = ((u64)elrsr1 << 32) | elrsr0;
>> +#endif
>> +	cpu_if->vgic_apr    = readl_relaxed(base + GICH_APR);
>> +
>> +	writel_relaxed(0, base + GICH_HCR);
>> +
>> +	lr_base = base + GICH_LR0;
>> +	do {
>> +		cpu_if->vgic_lr[i++] = readl_relaxed(lr_base++);
>> +	} while (--nr_lr);
> 
> why not a simple for-loop?

Good question. I blame coding from 30000 feet.

>> +}
>> +
> 
> copy the vcpu HYP VA comment down here.
> 
>> +void __hyp_text __vgic_v2_restore_state(struct kvm_vcpu *vcpu)
>> +{
>> +	struct kvm *kvm = kern_hyp_va(vcpu->kvm);
>> +	struct vgic_v2_cpu_if *cpu_if = &vcpu->arch.vgic_cpu.vgic_v2;
>> +	struct vgic_dist *vgic = &kvm->arch.vgic;
>> +	void __iomem *base = kern_hyp_va(vgic->vctrl_base);
>> +	u32 __iomem *lr_base;
>> +	unsigned int i = 0, nr_lr;
>> +
>> +	if (!base)
>> +		return;
>> +
>> +	writel_relaxed(cpu_if->vgic_hcr, base + GICH_HCR);
>> +	writel_relaxed(cpu_if->vgic_vmcr, base + GICH_VMCR);
>> +	writel_relaxed(cpu_if->vgic_apr, base + GICH_APR);
>> +
>> +	lr_base = base + GICH_LR0;
>> +	nr_lr = vcpu->arch.vgic_cpu.nr_lr;
>> +	do {
>> +		writel_relaxed(cpu_if->vgic_lr[i++], lr_base++);
>> +	} while (--nr_lr);
> 
> same question as above.
> 
>> +}
>> -- 
>> 2.1.4
>>
> Otherwise looks good.

Thanks,

	M.
-- 
Jazz is not dead. It just smells funny...