[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <Y8HU45aISOPwX76D@google.com>
Date: Fri, 13 Jan 2023 22:02:11 +0000
From: Sean Christopherson <seanjc@...gle.com>
To: Chao Peng <chao.p.peng@...ux.intel.com>
Cc: kvm@...r.kernel.org, linux-kernel@...r.kernel.org,
linux-mm@...ck.org, linux-fsdevel@...r.kernel.org,
linux-arch@...r.kernel.org, linux-api@...r.kernel.org,
linux-doc@...r.kernel.org, qemu-devel@...gnu.org,
Paolo Bonzini <pbonzini@...hat.com>,
Jonathan Corbet <corbet@....net>,
Vitaly Kuznetsov <vkuznets@...hat.com>,
Wanpeng Li <wanpengli@...cent.com>,
Jim Mattson <jmattson@...gle.com>,
Joerg Roedel <joro@...tes.org>,
Thomas Gleixner <tglx@...utronix.de>,
Ingo Molnar <mingo@...hat.com>, Borislav Petkov <bp@...en8.de>,
Arnd Bergmann <arnd@...db.de>,
Naoya Horiguchi <naoya.horiguchi@....com>,
Miaohe Lin <linmiaohe@...wei.com>, x86@...nel.org,
"H . Peter Anvin" <hpa@...or.com>, Hugh Dickins <hughd@...gle.com>,
Jeff Layton <jlayton@...nel.org>,
"J . Bruce Fields" <bfields@...ldses.org>,
Andrew Morton <akpm@...ux-foundation.org>,
Shuah Khan <shuah@...nel.org>, Mike Rapoport <rppt@...nel.org>,
Steven Price <steven.price@....com>,
"Maciej S . Szmigiero" <mail@...iej.szmigiero.name>,
Vlastimil Babka <vbabka@...e.cz>,
Vishal Annapurve <vannapurve@...gle.com>,
Yu Zhang <yu.c.zhang@...ux.intel.com>,
"Kirill A . Shutemov" <kirill.shutemov@...ux.intel.com>,
luto@...nel.org, jun.nakajima@...el.com, dave.hansen@...el.com,
ak@...ux.intel.com, david@...hat.com, aarcange@...hat.com,
ddutile@...hat.com, dhildenb@...hat.com,
Quentin Perret <qperret@...gle.com>, tabba@...gle.com,
Michael Roth <michael.roth@....com>, mhocko@...e.com,
wei.w.wang@...el.com
Subject: Re: [PATCH v10 2/9] KVM: Introduce per-page memory attributes
On Fri, Dec 02, 2022, Chao Peng wrote:
> diff --git a/arch/x86/kvm/Kconfig b/arch/x86/kvm/Kconfig
> index fbeaa9ddef59..a8e379a3afee 100644
> --- a/arch/x86/kvm/Kconfig
> +++ b/arch/x86/kvm/Kconfig
> @@ -49,6 +49,7 @@ config KVM
> select SRCU
> select INTERVAL_TREE
> select HAVE_KVM_PM_NOTIFIER if PM
> + select HAVE_KVM_MEMORY_ATTRIBUTES
I would prefer to call this KVM_GENERIC_MEMORY_ATTRIBUTES. Similar to
KVM_GENERIC_HARDWARE_ENABLING, ARM does need/have hardware enabling, it just
doesn't want KVM's generic implementation. In this case, pKVM does support memory
attributes, but uses stage-2 tables to track ownership and doesn't need/want the
overhead of the generic implementation.
> help
...
> +#define KVM_MEMORY_ATTRIBUTE_READ (1ULL << 0)
> +#define KVM_MEMORY_ATTRIBUTE_WRITE (1ULL << 1)
> +#define KVM_MEMORY_ATTRIBUTE_EXECUTE (1ULL << 2)
> +#define KVM_MEMORY_ATTRIBUTE_PRIVATE (1ULL << 3)
I think we should carve out bits 0-2 for RWX, but I don't think we should actually
define them until they're actually accepted by KVM.
> +static int kvm_vm_ioctl_set_mem_attributes(struct kvm *kvm,
> + struct kvm_memory_attributes *attrs)
> +{
> + gfn_t start, end;
> + unsigned long i;
> + void *entry;
> + u64 supported_attrs = kvm_supported_mem_attributes(kvm);
> +
> + /* flags is currently not used. */
> + if (attrs->flags)
> + return -EINVAL;
> + if (attrs->attributes & ~supported_attrs)
Nit, no need for "supported_attrs", just consume kvm_supported_mem_attributes()
directly.
> + return -EINVAL;
> + if (attrs->size == 0 || attrs->address + attrs->size < attrs->address)
> + return -EINVAL;
> + if (!PAGE_ALIGNED(attrs->address) || !PAGE_ALIGNED(attrs->size))
> + return -EINVAL;
> +
> + start = attrs->address >> PAGE_SHIFT;
> + end = (attrs->address + attrs->size - 1 + PAGE_SIZE) >> PAGE_SHIFT;
> +
> + entry = attrs->attributes ? xa_mk_value(attrs->attributes) : NULL;
> +
> + mutex_lock(&kvm->lock);
Peeking forward multiple patches, this needs to take kvm->slots_lock, not kvm->lock.
There's a bug in the lpage_disallowed patch that I believe can most easily be
solved by making this mutually exclusive with memslot changes.
When a memslot is created, KVM needs to walk through the attributes to detect
whether or not the attributes are identical for the entire slot. To avoid races,
that means taking slots_lock.
The alternative would be to query the attributes when adjusting the hugepage level
and avoid lpage_disallowed entirely, but in the (very brief) time I've thought
about this I haven't come up with a way to do that in a performant manner.
> + for (i = start; i < end; i++)
Curly braces needed on the for-loop.
Powered by blists - more mailing lists