Received-SPF: pass (google.com: domain of linux-kernel+bounces-97586-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:4601:e00::3 as permitted sender) client-ip=2604:1380:4601:e00::3;
Date: Fri, 8 Mar 2024 13:01:16 -0800
From: Isaku Yamahata <isaku.yamahata@intel.com>
To: Yin Fengwei <fengwei.yin@intel.com>
Cc: isaku.yamahata@intel.com, kvm@vger.kernel.org,
	linux-kernel@vger.kernel.org, isaku.yamahata@gmail.com,
	Paolo Bonzini <pbonzini@redhat.com>, erdemaktas@google.com,
	Sean Christopherson <seanjc@google.com>,
	Sagi Shahar <sagis@google.com>, Kai Huang <kai.huang@intel.com>,
	chen.bo@intel.com, hang.yuan@intel.com, tina.zhang@intel.com,
	isaku.yamahata@linux.intel.com
Subject: Re: [PATCH v19 014/130] KVM: Add KVM vcpu ioctl to pre-populate
 guest memory
Message-ID: <20240308210116.GB713729@ls.amr.corp.intel.com>
References: <cover.1708933498.git.isaku.yamahata@intel.com>
 <8b7380f1b02f8e3995f18bebb085e43165d5d682.1708933498.git.isaku.yamahata@intel.com>
 <6b453972-2723-47c5-981e-56c150f217d7@intel.com>
Precedence: bulk
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Disposition: inline
In-Reply-To: <6b453972-2723-47c5-981e-56c150f217d7@intel.com>

On Thu, Mar 07, 2024 at 03:01:11PM +0800,
Yin Fengwei <fengwei.yin@intel.com> wrote:

> > diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
> > index 0349e1f241d1..2f0a8e28795e 100644
> > --- a/virt/kvm/kvm_main.c
> > +++ b/virt/kvm/kvm_main.c
> > @@ -4409,6 +4409,62 @@ static int kvm_vcpu_ioctl_get_stats_fd(struct kvm_vcpu *vcpu)
> >  	return fd;
> >  }
> >  
> > +__weak void kvm_arch_vcpu_pre_memory_mapping(struct kvm_vcpu *vcpu)
> > +{
> > +}
> > +
> > +__weak int kvm_arch_vcpu_memory_mapping(struct kvm_vcpu *vcpu,
> > +					struct kvm_memory_mapping *mapping)
> > +{
> > +	return -EOPNOTSUPP;
> > +}
> > +
> > +static int kvm_vcpu_memory_mapping(struct kvm_vcpu *vcpu,
> > +				   struct kvm_memory_mapping *mapping)
> > +{
> > +	bool added = false;
> > +	int idx, r = 0;
> > +
> > +	/* flags isn't used yet. */
> > +	if (mapping->flags)
> > +		return -EINVAL;
> > +
> > +	/* Sanity check */
> > +	if (!IS_ALIGNED(mapping->source, PAGE_SIZE) ||
> > +	    !mapping->nr_pages ||
> > +	    mapping->nr_pages & GENMASK_ULL(63, 63 - PAGE_SHIFT) ||
> > +	    mapping->base_gfn + mapping->nr_pages <= mapping->base_gfn)
> I suppose !mapping->nr_pages can be deleted as this line can cover it.
> > +		return -EINVAL;
> > +
> > +	vcpu_load(vcpu);
> > +	idx = srcu_read_lock(&vcpu->kvm->srcu);
> > +	kvm_arch_vcpu_pre_memory_mapping(vcpu);
> > +
> > +	while (mapping->nr_pages) {
> > +		if (signal_pending(current)) {
> > +			r = -ERESTARTSYS;
> > +			break;
> > +		}
> > +
> > +		if (need_resched())
> > +			cond_resched();
> > +
> > +		r = kvm_arch_vcpu_memory_mapping(vcpu, mapping);
> > +		if (r)
> > +			break;
> > +
> > +		added = true;
> > +	}
> > +
> > +	srcu_read_unlock(&vcpu->kvm->srcu, idx);
> > +	vcpu_put(vcpu);
> > +
> > +	if (added && mapping->nr_pages > 0)
> > +		r = -EAGAIN;
> > +
> > +	return r;
> > +}
> > +
> >  static long kvm_vcpu_ioctl(struct file *filp,
> >  			   unsigned int ioctl, unsigned long arg)
> >  {
> > @@ -4610,6 +4666,17 @@ static long kvm_vcpu_ioctl(struct file *filp,
> >  		r = kvm_vcpu_ioctl_get_stats_fd(vcpu);
> >  		break;
> >  	}
> > +	case KVM_MEMORY_MAPPING: {
> > +		struct kvm_memory_mapping mapping;
> > +
> > +		r = -EFAULT;
> > +		if (copy_from_user(&mapping, argp, sizeof(mapping)))
> > +			break;
> > +		r = kvm_vcpu_memory_mapping(vcpu, &mapping);
> return value r should be checked before copy_to_user

That's intentional to tell the mapping is partially or fully processed
regardless that error happened or not.

> 
> 
> Regards
> Yin, Fengwei
> 
> > +		if (copy_to_user(argp, &mapping, sizeof(mapping)))
> > +			r = -EFAULT;
> > +		break;
> > +	}
> >  	default:
> >  		r = kvm_arch_vcpu_ioctl(filp, ioctl, arg);
> >  	}
> 

-- 
Isaku Yamahata <isaku.yamahata@intel.com>