[PATCH 3/3] arm/arm64: Enable Dirty Page logging for ARMv8 move log read, tlb flush to generic code

[Mario Smarduch]

On 11/21/2014 02:09 AM, Christoffer Dall wrote:
> On Wed, Nov 19, 2014 at 12:15:55PM -0800, Mario Smarduch wrote:
>> On 11/19/2014 06:39 AM, Christoffer Dall wrote:
>>> Hi Mario,
>>>
>>> On Fri, Nov 07, 2014 at 12:51:39PM -0800, Mario Smarduch wrote:
>>>> On 11/07/2014 12:20 PM, Christoffer Dall wrote:
>>>>> On Thu, Oct 09, 2014 at 07:34:07PM -0700, Mario Smarduch wrote:
>>>>>> This patch enables ARMv8 dirty page logging and unifies ARMv7/ARMv8 code.
>>>>>>
>>>>>> Signed-off-by: Mario Smarduch <m.smarduch at samsung.com>
>>>>>> ---
>>>>>>  arch/arm/include/asm/kvm_host.h | 12 ------------
>>>>>>  arch/arm/kvm/arm.c              |  9 ---------
>>>>>>  arch/arm/kvm/mmu.c              | 17 +++++++++++------
>>>>>>  arch/arm64/kvm/Kconfig          |  2 +-
>>>>>>  4 files changed, 12 insertions(+), 28 deletions(-)
>>>>>>
>>>>>> diff --git a/arch/arm/include/asm/kvm_host.h b/arch/arm/include/asm/kvm_host.h
>>>>>> index 12311a5..59565f5 100644
>>>>>> --- a/arch/arm/include/asm/kvm_host.h
>>>>>> +++ b/arch/arm/include/asm/kvm_host.h
>>>>>> @@ -220,18 +220,6 @@ static inline void __cpu_init_hyp_mode(phys_addr_t boot_pgd_ptr,
>>>>>>  	kvm_call_hyp((void*)hyp_stack_ptr, vector_ptr, pgd_ptr);
>>>>>>  }
>>>>>>  
>>>>>> -/**
>>>>>> - * kvm_arch_flush_remote_tlbs() - flush all VM TLB entries
>>>>>> - * @kvm:       pointer to kvm structure.
>>>>>> - *
>>>>>> - * Interface to HYP function to flush all VM TLB entries without address
>>>>>> - * parameter.
>>>>>> - */
>>>>>> -static inline void kvm_arch_flush_remote_tlbs(struct kvm *kvm)
>>>>>> -{
>>>>>> -	kvm_call_hyp(__kvm_tlb_flush_vmid, kvm);
>>>>>> -}
>>>>>> -
>>>>>>  static inline int kvm_arch_dev_ioctl_check_extension(long ext)
>>>>>>  {
>>>>>>  	return 0;
>>>>>> diff --git a/arch/arm/kvm/arm.c b/arch/arm/kvm/arm.c
>>>>>> index 0546fa3..6a6fd6b 100644
>>>>>> --- a/arch/arm/kvm/arm.c
>>>>>> +++ b/arch/arm/kvm/arm.c
>>>>>> @@ -242,7 +242,6 @@ void kvm_arch_commit_memory_region(struct kvm *kvm,
>>>>>>  				   const struct kvm_memory_slot *old,
>>>>>>  				   enum kvm_mr_change change)
>>>>>>  {
>>>>>> -#ifdef CONFIG_ARM
>>>>>>  	/*
>>>>>>  	 * At this point memslot has been committed and there is an
>>>>>>  	 * allocated dirty_bitmap[], dirty pages will be be tracked while the
>>>>>> @@ -250,7 +249,6 @@ void kvm_arch_commit_memory_region(struct kvm *kvm,
>>>>>>  	 */
>>>>>>  	if ((change != KVM_MR_DELETE) && (mem->flags & KVM_MEM_LOG_DIRTY_PAGES))
>>>>>>  		kvm_mmu_wp_memory_region(kvm, mem->slot);
>>>>>> -#endif
>>>>>>  }
>>>>>>  
>>>>>>  void kvm_arch_flush_shadow_all(struct kvm *kvm)
>>>>>> @@ -783,13 +781,6 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
>>>>>>  	}
>>>>>>  }
>>>>>>  
>>>>>> -#ifdef CONFIG_ARM64
>>>>>> -int kvm_arch_vm_ioctl_get_dirty_log(struct kvm *kvm, struct kvm_dirty_log *log)
>>>>>> -{
>>>>>> -	return -EINVAL;
>>>>>> -}
>>>>>> -#endif
>>>>>> -
>>>>>>  static int kvm_vm_ioctl_set_device_addr(struct kvm *kvm,
>>>>>>  					struct kvm_arm_device_addr *dev_addr)
>>>>>>  {
>>>>>> diff --git a/arch/arm/kvm/mmu.c b/arch/arm/kvm/mmu.c
>>>>>> index df1a5a3..8c0f9f2 100644
>>>>>> --- a/arch/arm/kvm/mmu.c
>>>>>> +++ b/arch/arm/kvm/mmu.c
>>>>>> @@ -49,11 +49,18 @@ static phys_addr_t hyp_idmap_vector;
>>>>>>  
>>>>>>  static bool kvm_get_logging_state(struct kvm_memory_slot *memslot)
>>>>>>  {
>>>>>> -#ifdef CONFIG_ARM
>>>>>>  	return !!memslot->dirty_bitmap;
>>>>>> -#else
>>>>>> -	return false;
>>>>>> -#endif
>>>>>> +}
>>>>>> +
>>>>>> +/**
>>>>>> + * kvm_arch_flush_remote_tlbs() - flush all VM TLB entries for ARMv7/8
>>>>>> + * @kvm:	pointer to kvm structure.
>>>>>> + *
>>>>>> + * Interface to HYP function to flush all VM TLB entries
>>>>>> + */
>>>>>> +inline void kvm_arch_flush_remote_tlbs(struct kvm *kvm)
>>>>>> +{
>>>>>> +	kvm_call_hyp(__kvm_tlb_flush_vmid, kvm);
>>>>>>  }
>>>>>>  
>>>>>>  static void kvm_tlb_flush_vmid_ipa(struct kvm *kvm, phys_addr_t ipa)
>>>>>> @@ -769,7 +776,6 @@ static bool transparent_hugepage_adjust(pfn_t *pfnp, phys_addr_t *ipap)
>>>>>>  	return false;
>>>>>>  }
>>>>>>  
>>>>>> -#ifdef CONFIG_ARM
>>>>>>  /**
>>>>>>   * stage2_wp_ptes - write protect PMD range
>>>>>>   * @pmd:	pointer to pmd entry
>>>>>> @@ -917,7 +923,6 @@ void kvm_mmu_write_protect_pt_masked(struct kvm *kvm,
>>>>>>  
>>>>>>  	stage2_wp_range(kvm, start, end);
>>>>>>  }
>>>>>> -#endif
>>>>>>  
>>>>>>  static int user_mem_abort(struct kvm_vcpu *vcpu, phys_addr_t fault_ipa,
>>>>>>  			  struct kvm_memory_slot *memslot,
>>>>>> diff --git a/arch/arm64/kvm/Kconfig b/arch/arm64/kvm/Kconfig
>>>>>> index 40a8d19..a1a35809 100644
>>>>>> --- a/arch/arm64/kvm/Kconfig
>>>>>> +++ b/arch/arm64/kvm/Kconfig
>>>>>> @@ -26,7 +26,7 @@ config KVM
>>>>>>  	select KVM_ARM_HOST
>>>>>>  	select KVM_ARM_VGIC
>>>>>>  	select KVM_ARM_TIMER
>>>>>> -	select HAVE_KVM_ARCH_DIRTY_LOG
>>>>>> +	select HAVE_KVM_ARCH_TLB_FLUSH_ALL
>>>>>>  	---help---
>>>>>>  	  Support hosting virtualized guest machines.
>>>>>>  
>>>>>> -- 
>>>>>> 1.9.1
>>>>>>
>>>>>
>>>>> this patch needs to be rebased onto the newest version of your v7 patch,
>>>>> but otherwise it looks roughly ok.
>>>>
>>> Just discussed with Marc, that we'll try to merge the v7 patches asap,
>>
>> Hi Christoffer,
>>    that's good to hear, hopefully v7 patches are not to far behind to merge.
>>
>>> but when you send out the rebased series for v8, you need to handle the
>>> pud_huge case in the write-protect patch.
>>
>> Making the addition should be straightforward, but testing I'm not
>> sure how at this time, 2MB is only huge page size supported
>> by the kernel on arm64 (plus many other platforms). Maybe something
>> rigged up using memreserve ... but I'm not sure.
> 
> Let's start by making sure we're not regressing setups that we can
> easily test and we'll review the code.

Great will do.
> 
>>
>> The target for that would be subsequent (not next) merge window correct?
>>
> Probably, it's getting tight already.

That works well for me too, I'll be back to opensource work Dec. 1st.

Thanks,
  Mario


> 
> -Christoffer
>