[RFC PATCH v1 26/42] drivers/virt: add VBS secure-plane park loop

Sriram Nambakam <[email protected]> Wed, 5 Aug 2026 04:03:08 -0700
Newsgroups org.kernel.vger.kvm,org.kernel.vger.linux-kernel
Message-ID <[email protected]>
Add a minimal, self-contained driver that lets an otherwise ordinary
kernel act as the secure plane (plane >0) of a KVM VM-planes guest. When
the "vbs_park" command-line option is present, a kernel thread hands
control back to the normal plane via KVM_HC_VBS_VTL_RETURN and then
services VTL calls from a shared calling area, acknowledging each as a
no-op.

This is deliberately independent of the full VBS/HEKI stack (CONFIG_VBS):
it implements only the park/dispatch handshake so that any secure kernel
(or a future SVSM) can act as plane 1. Gated behind CONFIG_VBS_PARK.

Signed-off-by: Sriram Nambakam <[email protected]>
---
 drivers/virt/Kconfig    |  15 +++++
 drivers/virt/Makefile   |   1 +
 drivers/virt/vbs_park.c | 136 ++++++++++++++++++++++++++++++++++++++++
 3 files changed, 152 insertions(+)
 create mode 100644 drivers/virt/vbs_park.c

diff --git a/drivers/virt/Kconfig b/drivers/virt/Kconfig
index 52eb7e4ba71f..88e40eaba1c2 100644
--- a/drivers/virt/Kconfig
+++ b/drivers/virt/Kconfig
@@ -13,6 +13,21 @@ menuconfig VIRT_DRIVERS
 
 if VIRT_DRIVERS
 
+config VBS_PARK
+	bool "KVM VM-planes secure-plane park loop"
+	depends on X86 && KVM_GUEST
+	help
+	  Minimal in-kernel handler for the secure plane (plane >0) of a KVM
+	  VM-planes guest.  When enabled and the "vbs_park" kernel command-line
+	  option is present, a kernel thread hands control back to the normal
+	  plane via the KVM_HC_VBS_VTL_RETURN hypercall and then services VTL
+	  calls from a shared calling area.
+
+	  This is independent of the full VBS/HEKI stack (CONFIG_VBS): it
+	  implements only the park/dispatch handshake so that any secure kernel
+	  can act as plane 1.  Calls are acknowledged as no-ops.  Say N unless
+	  this kernel is used as a VM-planes secure plane.
+
 config VMGENID
 	tristate "Virtual Machine Generation ID driver"
 	default y
diff --git a/drivers/virt/Makefile b/drivers/virt/Makefile
index f29901bd7820..fa91899a356d 100644
--- a/drivers/virt/Makefile
+++ b/drivers/virt/Makefile
@@ -5,6 +5,7 @@
 
 obj-$(CONFIG_FSL_HV_MANAGER)	+= fsl_hypervisor.o
 obj-$(CONFIG_VMGENID)		+= vmgenid.o
+obj-$(CONFIG_VBS_PARK)		+= vbs_park.o
 obj-y				+= vboxguest/
 
 obj-$(CONFIG_NITRO_ENCLAVES)	+= nitro_enclaves/
diff --git a/drivers/virt/vbs_park.c b/drivers/virt/vbs_park.c
new file mode 100644
index 000000000000..fabb6beeea7b
--- /dev/null
+++ b/drivers/virt/vbs_park.c
@@ -0,0 +1,136 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * vbs_park - minimal KVM VM-planes secure-plane park loop
+ *
+ * This provides only the secure-plane (plane >0) side of the VM-planes
+ * park/dispatch handshake so that an otherwise ordinary kernel can act as
+ * plane 1.  It is deliberately independent of the full VBS/HEKI stack
+ * (CONFIG_VBS): it implements no security policy.  Its single job is to hand
+ * control back to the normal plane (plane 0) via the KVM_HC_VBS_VTL_RETURN
+ * hypercall and then service VTL calls from the shared calling area.
+ *
+ * Control flow (all within plane 0's single KVM_RUN; see
+ * arch/x86/kvm/x86.c __kvm_emulate_hypercall):
+ *
+ *   plane 0                        KVM                       plane 1 (here)
+ *   -------                        ---                       --------------
+ *   fill calling area
+ *   HC_VBS_VTL_CALL(ca_gpa) ─────▶ switch_plane ───────────▶ resume in
+ *                                  (RAX := ca_gpa)            vtl_return()
+ *                                                            handle call_id
+ *                                                            write ca->status
+ *   resume after VTL_CALL ◀─────── switch_plane ◀─────────── HC_VBS_VTL_RETURN
+ *
+ * Activated by the "vbs_park" kernel command-line option; without it this
+ * kernel boots normally and never parks.
+ */
+
+#define pr_fmt(fmt) "vbs-park: " fmt
+
+#include <linux/kernel.h>
+#include <linux/init.h>
+#include <linux/kthread.h>
+#include <linux/io.h>
+#include <linux/mm.h>
+#include <linux/types.h>
+#include <linux/errno.h>
+#include <linux/err.h>
+#include <linux/kvm_para.h>
+#include <asm/kvm_para.h>
+
+/*
+ * Shared-memory calling area.  MUST match struct vbs_kvm_ca in
+ * security/vbs/kvm_planes.c (the normal-plane <-> secure-plane wire ABI):
+ *
+ *   [ call_pending | call_id | status | arg_size | resp_size | buffer ]
+ */
+struct vtl_ca {
+	__u8	call_pending;	/* 1 while call is in flight		*/
+	__u8	rsvd[3];
+	__u32	call_id;	/* request id (set by caller)		*/
+	__s32	status;		/* return code (set by responder)	*/
+	__u32	arg_size;	/* request payload size			*/
+	__u32	resp_size;	/* response payload size		*/
+	__u8	buffer[];	/* request data in, response data out	*/
+} __packed;
+
+/* Set from the "vbs_park" kernel command-line option. */
+static bool vbs_park_active __ro_after_init;
+
+static int __init vbs_park_setup(char *str)
+{
+	vbs_park_active = true;
+	return 1;
+}
+__setup("vbs_park", vbs_park_setup);
+
+/*
+ * Park the secure plane and hand control back to the normal plane.  On the
+ * next VTL call KVM resumes us here with the calling-area GPA in the
+ * hypercall return value (RAX).  @status is carried for tracing only; the
+ * real result is already in the calling area.
+ */
+static u64 vtl_return(long status)
+{
+	return kvm_hypercall1(KVM_HC_VBS_VTL_RETURN, (unsigned long)status);
+}
+
+static int vbs_park_fn(void *unused)
+{
+	long status = 0;
+
+	pr_info("secure-plane park loop started\n");
+
+	for (;;) {
+		struct vtl_ca *ca;
+		u64 ca_gpa;
+
+		/* Park; resume with the next request's calling-area GPA. */
+		ca_gpa = vtl_return(status);
+		if (!ca_gpa) {
+			status = -EINVAL;
+			continue;
+		}
+
+		ca = memremap(ca_gpa, PAGE_SIZE, MEMREMAP_WB);
+		if (!ca) {
+			pr_err_ratelimited("failed to map calling area 0x%llx\n",
+					   ca_gpa);
+			status = -EFAULT;
+			continue;
+		}
+
+		/*
+		 * No security policy lives here: acknowledge the call as a
+		 * no-op so the normal plane can make progress.  Replace this
+		 * with real handlers (or move plane 1 to a dedicated SVSM) to
+		 * enforce actual VBS semantics.
+		 */
+		pr_info_ratelimited("VTL call id=0x%x arg_size=%u (no-op)\n",
+				    ca->call_id, ca->arg_size);
+		ca->status    = 0;
+		ca->resp_size = 0;
+		status = 0;
+
+		memunmap(ca);
+	}
+
+	return 0;
+}
+
+static int __init vbs_park_init(void)
+{
+	struct task_struct *t;
+
+	if (!vbs_park_active)
+		return 0;
+
+	t = kthread_run(vbs_park_fn, NULL, "vbs-park");
+	if (IS_ERR(t)) {
+		pr_err("failed to start park loop: %ld\n", PTR_ERR(t));
+		return PTR_ERR(t);
+	}
+
+	return 0;
+}
+late_initcall(vbs_park_init);
-- 
2.55.0