On Fri, Mar 08, 2013 at 04:46:15AM +0100, Marc Zyngier wrote:

On Thu, 7 Mar 2013 10:09:03 +0200, "Michael S. Tsirkin" [off-list ref]
wrote:

quoted

On Tue, Mar 05, 2013 at 03:47:24AM +0000, Marc Zyngier wrote:

quoted

Provide the kvm.h file that defines the user space visible
interface.

Signed-off-by: Marc Zyngier <redacted>
---
 arch/arm64/include/uapi/asm/kvm.h | 112
 ++++++++++++++++++++++++++++++++++++++
 1 file changed, 112 insertions(+)
 create mode 100644 arch/arm64/include/uapi/asm/kvm.h

diff --git a/arch/arm64/include/uapi/asm/kvm.h

b/arch/arm64/include/uapi/asm/kvm.h
new file mode 100644
index 0000000..f5525f1

--- /dev/null
+++ b/arch/arm64/include/uapi/asm/kvm.h

@@ -0,0 +1,112 @@
+/*
+ * Copyright (C) 2012 - ARM Ltd
+ * Author: Marc Zyngier <marc.zyngier@arm.com>
+ *
+ * Derived from arch/arm/include/uapi/asm/kvm.h:
+ * Copyright (C) 2012 - Virtual Open Systems and Columbia University
+ * Author: Christoffer Dall <c.dall@virtualopensystems.com>
+ *
+ * This program is free software; you can redistribute it and/or

modify

quoted

quoted

+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see

<http://www.gnu.org/licenses/>.

quoted

quoted

+ */
+
+#ifndef __ARM_KVM_H__
+#define __ARM_KVM_H__
+
+#define KVM_SPSR_EL1	0
+#define KVM_NR_SPSR	1
+
+#ifndef __ASSEMBLY__
+#include <asm/types.h>
+#include <asm/ptrace.h>
+
+#define __KVM_HAVE_GUEST_DEBUG
+#define __KVM_HAVE_IRQ_LINE
+
+#define KVM_REG_SIZE(id)						\
+	(1U << (((id) & KVM_REG_SIZE_MASK) >> KVM_REG_SIZE_SHIFT))
+
+struct kvm_regs {
+	struct user_pt_regs regs;	/* sp = sp_el0 */
+
+	unsigned long	sp_el1;
+	unsigned long	elr_el1;
+
+	unsigned long	spsr[KVM_NR_SPSR];
+};
+

Using long in uapi is generally a mistake: with gcc it has
different size depending on whether you build in 64 and 32 bit mode.
I think it is better to make it __u64.

Ah, well spotted. Definitely a mistake.

quoted

quoted

+/* Supported Processor Types */
+#define KVM_ARM_TARGET_CORTEX_A57	0
+#define KVM_ARM_NUM_TARGETS		1
+
+/* KVM_ARM_SET_DEVICE_ADDR ioctl id encoding */
+#define KVM_ARM_DEVICE_TYPE_SHIFT	0
+#define KVM_ARM_DEVICE_TYPE_MASK	(0xffff << KVM_ARM_DEVICE_TYPE_SHIFT)
+#define KVM_ARM_DEVICE_ID_SHIFT		16
+#define KVM_ARM_DEVICE_ID_MASK		(0xffff << KVM_ARM_DEVICE_ID_SHIFT)
+
+/* Supported device IDs */
+#define KVM_ARM_DEVICE_VGIC_V2		0
+
+/* Supported VGIC address types  */
+#define KVM_VGIC_V2_ADDR_TYPE_DIST	0
+#define KVM_VGIC_V2_ADDR_TYPE_CPU	1
+
+#define KVM_VGIC_V2_DIST_SIZE		0x1000
+#define KVM_VGIC_V2_CPU_SIZE		0x2000
+
+struct kvm_vcpu_init {
+	__u32 target;
+	__u32 features[7];
+};
+
+struct kvm_sregs {
+};
+
+struct kvm_fpu {
+};
+
+struct kvm_guest_debug_arch {
+};
+
+struct kvm_debug_exit_arch {
+};
+

This is a problem too: structure alignment is different
in -m32 versus -m64 modes, which will affect the offset of
the following fields. I think it's best to add a "padding"
field in there, and size it to multiple of 8 bytes.

Sorry, which field are you referring to?

Sorry, I was wrong here: while it seems to be compler dependent gcc
returns 0 as size of an empty structure.

quoted

I think the same is a good idea for other empty stuctures,
since otherwise the padding is implicit and
not initialized by gcc, and it is hard not too leak
info to userspace when you copy these structures out.

And, it'll be handy if you want to extend the structures
down the line.

Sure, what would make some sense.

Thanks,

         M.
-- 
Fast, cheap, reliable. Pick two.