linux-kernel - Re: [PATCH 1/5] KVM: gmem: Don't rely on __kvm_gmem_get

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <80a5a52d-68d9-b4d3-e243-7720a097a3a1@amd.com>
Date: Wed, 22 Jan 2025 08:39:37 -0600
From: Tom Lendacky <thomas.lendacky@....com>
To: Michael Roth <michael.roth@....com>, kvm@...r.kernel.org
Cc: linux-coco@...ts.linux.dev, linux-mm@...ck.org,
 linux-kernel@...r.kernel.org, jroedel@...e.de, pbonzini@...hat.com,
 seanjc@...gle.com, vbabka@...e.cz, amit.shah@....com,
 pratikrajesh.sampat@....com, ashish.kalra@....com, liam.merwick@...cle.com,
 david@...hat.com, vannapurve@...gle.com, ackerleytng@...gle.com,
 quic_eberman@...cinc.com
Subject: Re: [PATCH 1/5] KVM: gmem: Don't rely on __kvm_gmem_get_pfn() for
 preparedness

On 12/12/24 00:36, Michael Roth wrote:
> Currently __kvm_gmem_get_pfn() sets 'is_prepared' so callers can skip
> calling kvm_gmem_prepare_folio(). However, subsequent patches will
> introduce some locking constraints around setting/checking preparedness
> that will require filemap_invalidate_lock*() to be held while checking
> for preparedness. This locking could theoretically be done inside
> __kvm_gmem_get_pfn(), or by requiring that filemap_invalidate_lock*() is
> held while calling __kvm_gmem_get_pfn(), but that places unnecessary
> constraints around when __kvm_gmem_get_pfn() can be called, whereas
> callers could just as easily call kvm_gmem_is_prepared() directly.
> 
> So, in preparation for these locking changes, drop the 'is_prepared'
> argument, and leave it up to callers to handle checking preparedness
> where needed and with the proper locking constraints.
> 
> Signed-off-by: Michael Roth <michael.roth@....com>
> ---
>  virt/kvm/guest_memfd.c | 13 +++++--------
>  1 file changed, 5 insertions(+), 8 deletions(-)
> 
> diff --git a/virt/kvm/guest_memfd.c b/virt/kvm/guest_memfd.c
> index b69af3580bef..aa0038ddf4a4 100644
> --- a/virt/kvm/guest_memfd.c
> +++ b/virt/kvm/guest_memfd.c
> @@ -773,7 +773,7 @@ void kvm_gmem_unbind(struct kvm_memory_slot *slot)
>  static struct folio *__kvm_gmem_get_pfn(struct file *file,
>  					struct kvm_memory_slot *slot,
>  					pgoff_t index, kvm_pfn_t *pfn,
> -					bool *is_prepared, int *max_order)
> +					int *max_order)
>  {
>  	struct kvm_gmem *gmem = file->private_data;
>  	struct folio *folio;
> @@ -803,7 +803,6 @@ static struct folio *__kvm_gmem_get_pfn(struct file *file,
>  	if (max_order)
>  		*max_order = 0;
>  
> -	*is_prepared = kvm_gmem_is_prepared(file, index, folio);
>  	return folio;
>  }
>  
> @@ -814,19 +813,18 @@ int kvm_gmem_get_pfn(struct kvm *kvm, struct kvm_memory_slot *slot,
>  	pgoff_t index = kvm_gmem_get_index(slot, gfn);
>  	struct file *file = kvm_gmem_get_file(slot);
>  	struct folio *folio;
> -	bool is_prepared = false;
>  	int r = 0;
>  
>  	if (!file)
>  		return -EFAULT;
>  
> -	folio = __kvm_gmem_get_pfn(file, slot, index, pfn, &is_prepared, max_order);
> +	folio = __kvm_gmem_get_pfn(file, slot, index, pfn, max_order);
>  	if (IS_ERR(folio)) {
>  		r = PTR_ERR(folio);
>  		goto out;
>  	}
>  
> -	if (!is_prepared)
> +	if (kvm_gmem_is_prepared(file, index, folio))

Shouldn't this be !kvm_gmem_is_prepared() ?

Thanks,
Tom

>  		r = kvm_gmem_prepare_folio(kvm, file, slot, gfn, folio);
>  
>  	folio_unlock(folio);
> @@ -872,7 +870,6 @@ long kvm_gmem_populate(struct kvm *kvm, gfn_t start_gfn, void __user *src, long
>  		struct folio *folio;
>  		gfn_t gfn = start_gfn + i;
>  		pgoff_t index = kvm_gmem_get_index(slot, gfn);
> -		bool is_prepared = false;
>  		kvm_pfn_t pfn;
>  
>  		if (signal_pending(current)) {
> @@ -880,13 +877,13 @@ long kvm_gmem_populate(struct kvm *kvm, gfn_t start_gfn, void __user *src, long
>  			break;
>  		}
>  
> -		folio = __kvm_gmem_get_pfn(file, slot, index, &pfn, &is_prepared, &max_order);
> +		folio = __kvm_gmem_get_pfn(file, slot, index, &pfn, &max_order);
>  		if (IS_ERR(folio)) {
>  			ret = PTR_ERR(folio);
>  			break;
>  		}
>  
> -		if (is_prepared) {
> +		if (kvm_gmem_is_prepared(file, index, folio)) {
>  			folio_unlock(folio);
>  			folio_put(folio);
>  			ret = -EEXIST;