From: Avi Kivity <avi@redhat.com>
To: "Nadav Har'El" <nyh@il.ibm.com>
Cc: kvm@vger.kernel.org, gleb@redhat.com
Subject: Re: [PATCH 08/28] nVMX: Hold a vmcs02 for each vmcs12
Date: Thu, 09 Dec 2010 14:41:37 +0200 [thread overview]
Message-ID: <4D00CE81.5050808@redhat.com> (raw)
In-Reply-To: <201012081704.oB8H43i2008629@rice.haifa.ibm.com>
On 12/08/2010 07:04 PM, Nadav Har'El wrote:
> In this patch we add a list of L0 (hardware) VMCSs, which we'll use to hold a
> hardware VMCS for each active vmcs12 (i.e., for each L2 guest).
>
> We call each of these L0 VMCSs a "vmcs02", as it is the VMCS that L0 uses
> to run its nested guest L2.
>
> Signed-off-by: Nadav Har'El<nyh@il.ibm.com>
> ---
> arch/x86/kvm/vmx.c | 96 +++++++++++++++++++++++++++++++++++++++++++
> 1 file changed, 96 insertions(+)
>
> --- .before/arch/x86/kvm/vmx.c 2010-12-08 18:56:49.000000000 +0200
> +++ .after/arch/x86/kvm/vmx.c 2010-12-08 18:56:49.000000000 +0200
> @@ -155,6 +155,12 @@ struct __packed vmcs12 {
> */
> #define VMCS12_REVISION 0x11e57ed0
>
> +struct vmcs_list {
> + struct list_head list;
> + gpa_t vmcs12_addr;
> + struct vmcs *vmcs02;
> +};
> +
> /*
> * The nested_vmx structure is part of vcpu_vmx, and holds information we need
> * for correct emulation of VMX (i.e., nested VMX) on this vcpu. For example,
> @@ -170,6 +176,10 @@ struct nested_vmx {
> /* The host-usable pointer to the above */
> struct page *current_vmcs12_page;
> struct vmcs12 *current_vmcs12;
> +
> + /* list of real (hardware) VMCS, one for each L2 guest of L1 */
> + struct list_head vmcs02_list; /* a vmcs_list */
> + int vmcs02_num;
> };
>
> struct vcpu_vmx {
> @@ -1736,6 +1746,85 @@ static void free_vmcs(struct vmcs *vmcs)
> free_pages((unsigned long)vmcs, vmcs_config.order);
> }
>
> +static struct vmcs *nested_get_current_vmcs(struct kvm_vcpu *vcpu)
> +{
> + struct vcpu_vmx *vmx = to_vmx(vcpu);
> + struct vmcs_list *list_item, *n;
> +
> + list_for_each_entry_safe(list_item, n,&vmx->nested.vmcs02_list, list)
> + if (list_item->vmcs12_addr == vmx->nested.current_vmptr)
> + return list_item->vmcs02;
> +
> + return NULL;
> +}
> +
> +/*
> + * Allocate an L0 VMCS (vmcs02) for the current L1 VMCS (vmcs12), if one
> + * does not already exist. The allocation is done in L0 memory, so to avoid
> + * denial-of-service attack by guests, we limit the number of concurrently-
> + * allocated vmcss. A well-behaving L1 will VMCLEAR unused vmcs12s and not
> + * trigger this limit.
> + */
> +static const int NESTED_MAX_VMCS = 256;
> +static int nested_create_current_vmcs(struct kvm_vcpu *vcpu)
> +{
> + struct vmcs_list *new_l2_guest;
> + struct vmcs *vmcs02;
> +
> + if (nested_get_current_vmcs(vcpu))
> + return 0; /* nothing to do - we already have a VMCS */
> +
> + if (to_vmx(vcpu)->nested.vmcs02_num>= NESTED_MAX_VMCS)
> + return -ENOMEM;
I asked for this to be fixed (say by freeing one vmcs02 from the list).
The guest can easily crash by running a lot of nested guests.
Actually you don't have to free it, simply reuse it for the new vmcs12.
> +
> + new_l2_guest = (struct vmcs_list *)
> + kmalloc(sizeof(struct vmcs_list), GFP_KERNEL);
> + if (!new_l2_guest)
> + return -ENOMEM;
> +
> + vmcs02 = alloc_vmcs();
> + if (!vmcs02) {
> + kfree(new_l2_guest);
> + return -ENOMEM;
> + }
> +
> + new_l2_guest->vmcs12_addr = to_vmx(vcpu)->nested.current_vmptr;
> + new_l2_guest->vmcs02 = vmcs02;
> + list_add(&(new_l2_guest->list),&(to_vmx(vcpu)->nested.vmcs02_list));
> + to_vmx(vcpu)->nested.vmcs02_num++;
> + return 0;
> +}
> +
--
error compiling committee.c: too many arguments to function
next prev parent reply other threads:[~2010-12-09 12:41 UTC|newest]
Thread overview: 40+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-12-08 16:59 [PATCH 0/28] nVMX: Nested VMX, v7 Nadav Har'El
2010-12-08 17:00 ` [PATCH 01/28] nVMX: Add "nested" module option to vmx.c Nadav Har'El
2010-12-08 17:00 ` [PATCH 02/28] nVMX: Add VMX and SVM to list of supported cpuid features Nadav Har'El
2010-12-09 11:38 ` Joerg Roedel
2010-12-15 13:25 ` Nadav Har'El
2010-12-08 17:01 ` [PATCH 03/28] nVMX: Implement VMXON and VMXOFF Nadav Har'El
2010-12-08 17:02 ` [PATCH 04/28] nVMX: Allow setting the VMXE bit in CR4 Nadav Har'El
2010-12-08 17:02 ` [PATCH 05/28] nVMX: Introduce vmcs12: a VMCS structure for L1 Nadav Har'El
2010-12-08 17:03 ` [PATCH 06/28] nVMX: Implement reading and writing of VMX MSRs Nadav Har'El
2010-12-09 11:04 ` Avi Kivity
2010-12-08 17:03 ` [PATCH 07/28] nVMX: Decoding memory operands of VMX instructions Nadav Har'El
2010-12-09 11:08 ` Avi Kivity
2010-12-08 17:04 ` [PATCH 08/28] nVMX: Hold a vmcs02 for each vmcs12 Nadav Har'El
2010-12-09 12:41 ` Avi Kivity [this message]
2010-12-08 17:04 ` [PATCH 09/28] nVMX: Add VMCS fields to the vmcs12 Nadav Har'El
2010-12-09 12:43 ` Avi Kivity
2010-12-10 12:10 ` Nadav Har'El
2010-12-08 17:05 ` [PATCH 10/28] nVMX: Success/failure of VMX instructions Nadav Har'El
2010-12-08 17:05 ` [PATCH 11/28] nVMX: Implement VMCLEAR Nadav Har'El
2010-12-08 17:06 ` [PATCH 12/28] nVMX: Implement VMPTRLD Nadav Har'El
2010-12-08 17:06 ` [PATCH 13/28] nVMX: Implement VMPTRST Nadav Har'El
2010-12-08 17:07 ` [PATCH 14/28] nVMX: Implement VMREAD and VMWRITE Nadav Har'El
2010-12-08 17:07 ` [PATCH 15/28] nVMX: Prepare vmcs02 from vmcs01 and vmcs12 Nadav Har'El
2010-12-08 17:08 ` [PATCH 16/28] nVMX: Move register-syncing to a function Nadav Har'El
2010-12-08 17:08 ` [PATCH 17/28] nVMX: Implement VMLAUNCH and VMRESUME Nadav Har'El
2010-12-08 17:09 ` [PATCH 18/28] nVMX: No need for handle_vmx_insn function any more Nadav Har'El
2010-12-08 17:09 ` [PATCH 19/28] nVMX: Exiting from L2 to L1 Nadav Har'El
2010-12-09 12:55 ` Avi Kivity
2010-12-08 17:10 ` [PATCH 20/28] nVMX: Deciding if L0 or L1 should handle an L2 exit Nadav Har'El
2010-12-08 17:10 ` [PATCH 21/28] nVMX: Correct handling of interrupt injection Nadav Har'El
2010-12-08 17:11 ` [PATCH 22/28] nVMX: Correct handling of exception injection Nadav Har'El
2010-12-08 17:11 ` [PATCH 23/28] nVMX: Correct handling of idt vectoring info Nadav Har'El
2010-12-08 17:12 ` [PATCH 24/28] nVMX: Handling of CR0 and CR4 modifying instructions Nadav Har'El
2010-12-09 13:19 ` Avi Kivity
2010-12-08 17:12 ` [PATCH 25/28] nVMX: Further fixes for lazy FPU loading Nadav Har'El
2010-12-09 13:05 ` Avi Kivity
2010-12-08 17:13 ` [PATCH 26/28] nVMX: Additional TSC-offset handling Nadav Har'El
2010-12-08 17:13 ` [PATCH 27/28] nVMX: Miscellenous small corrections Nadav Har'El
2010-12-08 17:14 ` [PATCH 28/28] nVMX: Documentation Nadav Har'El
2010-12-09 12:44 ` [PATCH 0/28] nVMX: Nested VMX, v7 Avi Kivity
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=4D00CE81.5050808@redhat.com \
--to=avi@redhat.com \
--cc=gleb@redhat.com \
--cc=kvm@vger.kernel.org \
--cc=nyh@il.ibm.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox