Re: [PATCH 3/3] s390/mm: Enable gmap huge pmd support

[David Hildenbrand <david@redhat.com>]

On 15.02.2018 16:43, Janosch Frank wrote:
> Now that we have everything in place, let's allow huge (1m) pmds for
> gmap linking, effectively allowing hugetlbfs backed guests. Transparent
> huge pages and 2g huge pages are *not* supported through this change.
> 
> Signed-off-by: Janosch Frank <frankja@linux.vnet.ibm.com>
> ---
>  Documentation/virtual/kvm/api.txt   | 12 ++++++++++++
>  arch/s390/include/asm/mmu.h         |  2 ++
>  arch/s390/include/asm/mmu_context.h |  1 +
>  arch/s390/kvm/kvm-s390.c            | 17 ++++++++++++++++-
>  arch/s390/mm/gmap.c                 |  8 +++++---
>  include/uapi/linux/kvm.h            |  1 +
>  6 files changed, 37 insertions(+), 4 deletions(-)
> 
> diff --git a/Documentation/virtual/kvm/api.txt b/Documentation/virtual/kvm/api.txt
> index 792fa87..edf248a 100644
> --- a/Documentation/virtual/kvm/api.txt
> +++ b/Documentation/virtual/kvm/api.txt
> @@ -4270,6 +4270,18 @@ enables QEMU to build error log and branch to guest kernel registered
>  machine check handling routine. Without this capability KVM will
>  branch to guests' 0x200 interrupt vector.
>  
> +7.13 KVM_CAP_S390_HPAGE
> +
> +Architectures: s390
> +Parameters: none
> +
> +With this capability the KVM support for memory backing with 1m pages
> +through hugetlbfs can be enabled. This will disable cmm, cmma, pfmfi
> +and the storage key interpretation.
> +
> +While it is generally possible to create and start a huge page backed
> +VM without this capability, the VM will not be functional.
> +
>  8. Other capabilities.
>  ----------------------
>  
> diff --git a/arch/s390/include/asm/mmu.h b/arch/s390/include/asm/mmu.h
> index db35c41a..5ab9452 100644
> --- a/arch/s390/include/asm/mmu.h
> +++ b/arch/s390/include/asm/mmu.h
> @@ -24,6 +24,8 @@ typedef struct {
>  	unsigned int use_skey:1;
>  	/* The mmu context uses CMMA. */
>  	unsigned int use_cmma:1;
> +	/* The gmap associated with this context uses huge pages. */
> +	unsigned int use_gmap_hpage:1;
>  } mm_context_t;
>  
>  #define INIT_MM_CONTEXT(name)						   \
> diff --git a/arch/s390/include/asm/mmu_context.h b/arch/s390/include/asm/mmu_context.h
> index 65154ea..79f2159 100644
> --- a/arch/s390/include/asm/mmu_context.h
> +++ b/arch/s390/include/asm/mmu_context.h
> @@ -32,6 +32,7 @@ static inline int init_new_context(struct task_struct *tsk,
>  	mm->context.has_pgste = 0;
>  	mm->context.use_skey = 0;
>  	mm->context.use_cmma = 0;
> +	mm->context.use_gmap_hpage = 0;
>  #endif
>  	switch (mm->context.asce_limit) {
>  	case _REGION2_SIZE:
> diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
> index ddf3599..8ea7025 100644
> --- a/arch/s390/kvm/kvm-s390.c
> +++ b/arch/s390/kvm/kvm-s390.c
> @@ -406,6 +406,7 @@ int kvm_vm_ioctl_check_extension(struct kvm *kvm, long ext)
>  	case KVM_CAP_S390_CMMA_MIGRATION:
>  	case KVM_CAP_S390_AIS:
>  	case KVM_CAP_S390_AIS_MIGRATION:
> +	case KVM_CAP_S390_HPAGE:
>  		r = 1;
>  		break;
>  	case KVM_CAP_S390_MEM_OP:
> @@ -604,6 +605,19 @@ static int kvm_vm_ioctl_enable_cap(struct kvm *kvm, struct kvm_enable_cap *cap)
>  		VM_EVENT(kvm, 3, "ENABLE: CAP_S390_GS %s",
>  			 r ? "(not available)" : "(success)");
>  		break;
> +	case KVM_CAP_S390_HPAGE:
> +		mutex_lock(&kvm->lock);
> +		if (kvm->created_vcpus) {
> +			r = -EBUSY;
> +		}
> +		kvm->mm->context.use_gmap_hpage = 1;
> +		/* They would complicate matters too much. */
> +		kvm->arch.has_skf = 0;
> +		kvm->arch.has_cmma = 0;
> +		kvm->arch.has_pfmfi = 0;
> +		mutex_unlock(&kvm->lock);
> +		VM_EVENT(kvm, 3, "%s", "ENABLE: KVM_CAP_S390_HPAGE");
> +		break;
>  	case KVM_CAP_S390_USER_STSI:
>  		VM_EVENT(kvm, 3, "%s", "ENABLE: CAP_S390_USER_STSI");
>  		kvm->arch.user_stsi = 1;
> @@ -655,7 +669,8 @@ static int kvm_s390_set_mem_control(struct kvm *kvm, struct kvm_device_attr *att
>  		VM_EVENT(kvm, 3, "%s", "ENABLE: CMMA support");
>  		mutex_lock(&kvm->lock);
>  		if (!kvm->created_vcpus) {
> -			kvm->arch.has_cmma = 1;
> +			if (!kvm->mm->context.use_gmap_hpage)
> +				kvm->arch.has_cmma = 1;
>  			/* Not compatible with cmma. */
>  			kvm->arch.has_pfmfi = 0;

As Christian said, this should be an XOR.

Either KVM_CAP_S390_HPAGE or CMMA can be enabled, not both (-EINVAL).

>  			ret = 0;
> diff --git a/arch/s390/mm/gmap.c b/arch/s390/mm/gmap.c
> index 2cafcba..6937853 100644
> --- a/arch/s390/mm/gmap.c
> +++ b/arch/s390/mm/gmap.c
> @@ -2,8 +2,10 @@
>  /*
>   *  KVM guest address space mapping code
>   *
> - *    Copyright IBM Corp. 2007, 2016
> + *    Copyright IBM Corp. 2007, 2016, 2017
>   *    Author(s): Martin Schwidefsky <schwidefsky@de.ibm.com>
> + *		 David Hildenbrand <david@redhat.com>
> + *		 Janosch Frank <frankja@linux.vnet.ibm.com>
>   */
>  
>  #include <linux/kernel.h>
> @@ -595,8 +597,8 @@ int __gmap_link(struct gmap *gmap, unsigned long gaddr, unsigned long vmaddr)
>  		return -EFAULT;
>  	pmd = pmd_offset(pud, vmaddr);
>  	VM_BUG_ON(pmd_none(*pmd));
> -	/* large pmds cannot yet be handled */
> -	if (pmd_large(*pmd))
> +	/* Are we allowed to use huge pages? */
> +	if (pmd_large(*pmd) && !gmap->mm->context.use_gmap_hpage)
>  		return -EFAULT;
>  	/* Link gmap segment table entry location to page table. */
>  	rc = radix_tree_preload(GFP_KERNEL);
> diff --git a/include/uapi/linux/kvm.h b/include/uapi/linux/kvm.h
> index 0fb5ef9..4e397d7 100644
> --- a/include/uapi/linux/kvm.h
> +++ b/include/uapi/linux/kvm.h
> @@ -934,6 +934,7 @@ struct kvm_ppc_resize_hpt {
>  #define KVM_CAP_S390_AIS_MIGRATION 150
>  #define KVM_CAP_PPC_GET_CPU_CHAR 151
>  #define KVM_CAP_S390_BPB 152
> +#define KVM_CAP_S390_HPAGE 153
>  
>  #ifdef KVM_CAP_IRQ_ROUTING
>  
> 


-- 

Thanks,

David / dhildenb