Skip to contentSkip to conversation

KVM: arm64: FF-A proxy for pKVM

PATCH In review Quentin Perret 4 participants 22 messages Nov 16, 2022 – Apr 18, 2023

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

cover KVM: arm64: FF-A proxy for pKVM

Hi all,

pKVM's primary goal is to protect guest pages from a compromised host by enforcing access control restrictions using stage-2 page-tables. Sadly, this cannot prevent TrustZone from accessing non-secure memory, and a compromised host could, for example, perform a 'confused deputy' attack by asking TrustZone to use pages that have been donated to protected guests. This would effectively allow the host to have TrustZone exfiltrate guest secrets on its behalf, hence breaking the isolation that pKVM intends to provide.

This series addresses this problem by providing pKVM with the ability to monitor SMCs following the Arm FF-A protocol. FF-A provides (among other things) a set of memory management APIs allowing the Normal World to share, donate or lend pages with Secure. By monitoring these SMCs, pKVM can ensure that the pages that are shared, lent or donated to Secure by the host kernel are only pages that it owns.

It should be noted that the robustness of this approach relies on having all Secure Software on the device use the FF-A protocol for memory management transactions with the normal world, and not use vendor-specific SMCs that pKVM is unable to parse. This is an important software requirement to integrate pKVM correctly on a device. For the record, this work is tractable in real world scenarios -- it was done for the Pixel 7 and Pixel 7 Pro devices, both of which launched recently with pKVM enabled.

This series introduces support for monitoring FF-A SMCs in the form of a minimal proxy running in the EL2 hypervisor. This proxy aims to be entirely 'transparent' to both the host kernel and TrustZone when the system behaves correctly (the host is not compromised). The FF-A proxy intercepts the SMCs coming from the host, and verifies that the pages involved in the memory transition can be shared/lent/donated legally before forwarding the SMC to Secure. The proxy also tracks which pages have been made accessible to TrustZone at any point in time using software bits in the stage-2 page-table of the host, to ensure they can't be subsequently donated to guests.

The patch series is divided as follows:

- patches 01-02 refactor existing FF-A header to allow code re-use;

- patches 03-07 provide the initial infrastructure at EL2 to handle FF-A SMCs;

- patches 08-12 use the previously introduced infrastructure to intercept the main memory management operations to share, reclaim and lend memory to/with Secure, and implement the core of the memory tracking logic.

Since this series depends on Will's recent pKVM series [1], it is based on today's kvmarm/next:

    eb8be68e907e ("Merge branch kvm-arm64/misc-6.2 into kvmarm-master/next")

A branch with all the goodies applied can also be found here:

https://android-kvm.googlesource.com/linux qperret/ffa-proxy

Feedback welcome!

Cheers,
Quentin

[1] https://lore.kernel.org/kvm/[email protected]/

Fuad Tabba (1):
KVM: arm64: Handle FFA_FEATURES call from the host

Quentin Perret (1):
ANDROID: KVM: arm64: pkvm: Add support for fragmented FF-A descriptors

Will Deacon (10):
firmware: arm_ffa: Move constants to header file
firmware: arm_ffa: Move comment before the field it is documenting
KVM: arm64: Block unsafe FF-A calls from the host
KVM: arm64: Probe FF-A version and host/hyp partition ID during init
KVM: arm64: Allocate pages for hypervisor FF-A mailboxes
KVM: arm64: Handle FFA_RXTX_MAP and FFA_RXTX_UNMAP calls from the host
KVM: arm64: Add FF-A helpers to share/unshare memory with secure world
KVM: arm64: Handle FFA_MEM_SHARE calls from the host
KVM: arm64: Handle FFA_MEM_RECLAIM calls from the host
KVM: arm64: Handle FFA_MEM_LEND calls from the host

 arch/arm64/include/asm/kvm_host.h             |   1 +
 arch/arm64/include/asm/kvm_pkvm.h             |  21 +
 arch/arm64/kvm/arm.c                          |   1 +
 arch/arm64/kvm/hyp/include/nvhe/ffa.h         |  17 +
 arch/arm64/kvm/hyp/include/nvhe/mem_protect.h |   3 +
 arch/arm64/kvm/hyp/nvhe/Makefile              |   2 +-
 arch/arm64/kvm/hyp/nvhe/ffa.c                 | 741 ++++++++++++++++++
 arch/arm64/kvm/hyp/nvhe/hyp-main.c            |   3 +
 arch/arm64/kvm/hyp/nvhe/mem_protect.c         |  68 ++
 arch/arm64/kvm/hyp/nvhe/setup.c               |  11 +
 arch/arm64/kvm/pkvm.c                         |   1 +
 drivers/firmware/arm_ffa/driver.c             | 101 +--
 include/linux/arm_ffa.h                       |  93 ++-
 13 files changed, 970 insertions(+), 93 deletions(-)
 create mode 100644 arch/arm64/kvm/hyp/include/nvhe/ffa.h
 create mode 100644 arch/arm64/kvm/hyp/nvhe/ffa.c
Signature · 1 line
2.38.1.431.g37b22c650d-goog
#1 lore raw

PATCH 1/12 firmware: arm_ffa: Move constants to header file

From: Will Deacon <[email protected]>

FF-A function IDs and error codes will be needed in the hypervisor too, so move to them to the header file where they can be shared. Rename the version constants with an "FFA_" prefix so that they are less likely to clash with other code in the tree.

---
 drivers/firmware/arm_ffa/driver.c | 101 +++---------------------------
 include/linux/arm_ffa.h           |  83 ++++++++++++++++++++++++
 2 files changed, 93 insertions(+), 91 deletions(-)
drivers/firmware/arm_ffa/driver.c +10 -91
index d5e86ef40b89..fa85c64d3ded 100644
@@ -36,81 +36,6 @@
 #include "common.h"
 
 #define FFA_DRIVER_VERSION	FFA_VERSION_1_0
-
-#define FFA_SMC(calling_convention, func_num)				\
-	ARM_SMCCC_CALL_VAL(ARM_SMCCC_FAST_CALL, (calling_convention),	\
-			   ARM_SMCCC_OWNER_STANDARD, (func_num))
-
-#define FFA_SMC_32(func_num)	FFA_SMC(ARM_SMCCC_SMC_32, (func_num))
-#define FFA_SMC_64(func_num)	FFA_SMC(ARM_SMCCC_SMC_64, (func_num))
-
-#define FFA_ERROR			FFA_SMC_32(0x60)
-#define FFA_SUCCESS			FFA_SMC_32(0x61)
-#define FFA_INTERRUPT			FFA_SMC_32(0x62)
-#define FFA_VERSION			FFA_SMC_32(0x63)
-#define FFA_FEATURES			FFA_SMC_32(0x64)
-#define FFA_RX_RELEASE			FFA_SMC_32(0x65)
-#define FFA_RXTX_MAP			FFA_SMC_32(0x66)
-#define FFA_FN64_RXTX_MAP		FFA_SMC_64(0x66)
-#define FFA_RXTX_UNMAP			FFA_SMC_32(0x67)
-#define FFA_PARTITION_INFO_GET		FFA_SMC_32(0x68)
-#define FFA_ID_GET			FFA_SMC_32(0x69)
-#define FFA_MSG_POLL			FFA_SMC_32(0x6A)
-#define FFA_MSG_WAIT			FFA_SMC_32(0x6B)
-#define FFA_YIELD			FFA_SMC_32(0x6C)
-#define FFA_RUN				FFA_SMC_32(0x6D)
-#define FFA_MSG_SEND			FFA_SMC_32(0x6E)
-#define FFA_MSG_SEND_DIRECT_REQ		FFA_SMC_32(0x6F)
-#define FFA_FN64_MSG_SEND_DIRECT_REQ	FFA_SMC_64(0x6F)
-#define FFA_MSG_SEND_DIRECT_RESP	FFA_SMC_32(0x70)
-#define FFA_FN64_MSG_SEND_DIRECT_RESP	FFA_SMC_64(0x70)
-#define FFA_MEM_DONATE			FFA_SMC_32(0x71)
-#define FFA_FN64_MEM_DONATE		FFA_SMC_64(0x71)
-#define FFA_MEM_LEND			FFA_SMC_32(0x72)
-#define FFA_FN64_MEM_LEND		FFA_SMC_64(0x72)
-#define FFA_MEM_SHARE			FFA_SMC_32(0x73)
-#define FFA_FN64_MEM_SHARE		FFA_SMC_64(0x73)
-#define FFA_MEM_RETRIEVE_REQ		FFA_SMC_32(0x74)
-#define FFA_FN64_MEM_RETRIEVE_REQ	FFA_SMC_64(0x74)
-#define FFA_MEM_RETRIEVE_RESP		FFA_SMC_32(0x75)
-#define FFA_MEM_RELINQUISH		FFA_SMC_32(0x76)
-#define FFA_MEM_RECLAIM			FFA_SMC_32(0x77)
-#define FFA_MEM_OP_PAUSE		FFA_SMC_32(0x78)
-#define FFA_MEM_OP_RESUME		FFA_SMC_32(0x79)
-#define FFA_MEM_FRAG_RX			FFA_SMC_32(0x7A)
-#define FFA_MEM_FRAG_TX			FFA_SMC_32(0x7B)
-#define FFA_NORMAL_WORLD_RESUME		FFA_SMC_32(0x7C)
-
-/*
- * For some calls it is necessary to use SMC64 to pass or return 64-bit values.
- * For such calls FFA_FN_NATIVE(name) will choose the appropriate
- * (native-width) function ID.
- */
-#ifdef CONFIG_64BIT
-#define FFA_FN_NATIVE(name)	FFA_FN64_##name
-#else
-#define FFA_FN_NATIVE(name)	FFA_##name
-#endif
-
-/* FFA error codes. */
-#define FFA_RET_SUCCESS            (0)
-#define FFA_RET_NOT_SUPPORTED      (-1)
-#define FFA_RET_INVALID_PARAMETERS (-2)
-#define FFA_RET_NO_MEMORY          (-3)
-#define FFA_RET_BUSY               (-4)
-#define FFA_RET_INTERRUPTED        (-5)
-#define FFA_RET_DENIED             (-6)
-#define FFA_RET_RETRY              (-7)
-#define FFA_RET_ABORTED            (-8)
-
-#define MAJOR_VERSION_MASK	GENMASK(30, 16)
-#define MINOR_VERSION_MASK	GENMASK(15, 0)
-#define MAJOR_VERSION(x)	((u16)(FIELD_GET(MAJOR_VERSION_MASK, (x))))
-#define MINOR_VERSION(x)	((u16)(FIELD_GET(MINOR_VERSION_MASK, (x))))
-#define PACK_VERSION_INFO(major, minor)			\
-	(FIELD_PREP(MAJOR_VERSION_MASK, (major)) |	\
-	 FIELD_PREP(MINOR_VERSION_MASK, (minor)))
-#define FFA_VERSION_1_0		PACK_VERSION_INFO(1, 0)
 #define FFA_MIN_VERSION		FFA_VERSION_1_0
 
 #define SENDER_ID_MASK		GENMASK(31, 16)
@@ -120,12 +45,6 @@
 #define PACK_TARGET_INFO(s, r)		\
 	(FIELD_PREP(SENDER_ID_MASK, (s)) | FIELD_PREP(RECEIVER_ID_MASK, (r)))
 
-/*
- * FF-A specification mentions explicitly about '4K pages'. This should
- * not be confused with the kernel PAGE_SIZE, which is the translation
- * granule kernel is configured and may be one among 4K, 16K and 64K.
- */
-#define FFA_PAGE_SIZE		SZ_4K
 /*
  * Keeping RX TX buffer size as 4K for now
  * 64K may be preferred to keep it min a page in 64K PAGE_SIZE config
@@ -178,9 +97,9 @@ static struct ffa_drv_info *drv_info;
  */
 static u32 ffa_compatible_version_find(u32 version)
 {
-	u16 major = MAJOR_VERSION(version), minor = MINOR_VERSION(version);
-	u16 drv_major = MAJOR_VERSION(FFA_DRIVER_VERSION);
-	u16 drv_minor = MINOR_VERSION(FFA_DRIVER_VERSION);
+	u16 major = FFA_MAJOR_VERSION(version), minor = FFA_MINOR_VERSION(version);
+	u16 drv_major = FFA_MAJOR_VERSION(FFA_DRIVER_VERSION);
+	u16 drv_minor = FFA_MINOR_VERSION(FFA_DRIVER_VERSION);
 
 	if ((major < drv_major) || (major == drv_major && minor <= drv_minor))
 		return version;
@@ -204,16 +123,16 @@ static int ffa_version_check(u32 *version)
 
 	if (ver.a0 < FFA_MIN_VERSION) {
 		pr_err("Incompatible v%d.%d! Earliest supported v%d.%d\n",
-		       MAJOR_VERSION(ver.a0), MINOR_VERSION(ver.a0),
-		       MAJOR_VERSION(FFA_MIN_VERSION),
-		       MINOR_VERSION(FFA_MIN_VERSION));
+		       FFA_MAJOR_VERSION(ver.a0), FFA_MINOR_VERSION(ver.a0),
+		       FFA_MAJOR_VERSION(FFA_MIN_VERSION),
+		       FFA_MINOR_VERSION(FFA_MIN_VERSION));
 		return -EINVAL;
 	}
 
-	pr_info("Driver version %d.%d\n", MAJOR_VERSION(FFA_DRIVER_VERSION),
-		MINOR_VERSION(FFA_DRIVER_VERSION));
-	pr_info("Firmware version %d.%d found\n", MAJOR_VERSION(ver.a0),
-		MINOR_VERSION(ver.a0));
+	pr_info("Driver version %d.%d\n", FFA_MAJOR_VERSION(FFA_DRIVER_VERSION),
+		FFA_MINOR_VERSION(FFA_DRIVER_VERSION));
+	pr_info("Firmware version %d.%d found\n", FFA_MAJOR_VERSION(ver.a0),
+		FFA_MINOR_VERSION(ver.a0));
 	*version = ffa_compatible_version_find(ver.a0);
 
 	return 0;
include/linux/arm_ffa.h +83 -0

89 lines not on this page · raw message

Signature · 1 line
2.38.1.431.g37b22c650d-goog
Reviewed-by on v1 1/12
#2 lore raw

Re: [PATCH 01/12] firmware: arm_ffa: Move constants to header file

Quoted 7 lines · Quentin Perret

From: Will Deacon <[email protected]>

FF-A function IDs and error codes will be needed in the hypervisor too, so move to them to the header file where they can be shared. Rename the version constants with an "FFA_" prefix so that they are less likely to clash with other code in the tree.

Quoted 242 lines
---
 drivers/firmware/arm_ffa/driver.c | 101 +++---------------------------
 include/linux/arm_ffa.h           |  83 ++++++++++++++++++++++++
 2 files changed, 93 insertions(+), 91 deletions(-)
drivers/firmware/arm_ffa/driver.c +10 -91
index d5e86ef40b89..fa85c64d3ded 100644
@@ -36,81 +36,6 @@
 #include "common.h"
 
 #define FFA_DRIVER_VERSION	FFA_VERSION_1_0
-
-#define FFA_SMC(calling_convention, func_num)				\
-	ARM_SMCCC_CALL_VAL(ARM_SMCCC_FAST_CALL, (calling_convention),	\
-			   ARM_SMCCC_OWNER_STANDARD, (func_num))
-
-#define FFA_SMC_32(func_num)	FFA_SMC(ARM_SMCCC_SMC_32, (func_num))
-#define FFA_SMC_64(func_num)	FFA_SMC(ARM_SMCCC_SMC_64, (func_num))
-
-#define FFA_ERROR			FFA_SMC_32(0x60)
-#define FFA_SUCCESS			FFA_SMC_32(0x61)
-#define FFA_INTERRUPT			FFA_SMC_32(0x62)
-#define FFA_VERSION			FFA_SMC_32(0x63)
-#define FFA_FEATURES			FFA_SMC_32(0x64)
-#define FFA_RX_RELEASE			FFA_SMC_32(0x65)
-#define FFA_RXTX_MAP			FFA_SMC_32(0x66)
-#define FFA_FN64_RXTX_MAP		FFA_SMC_64(0x66)
-#define FFA_RXTX_UNMAP			FFA_SMC_32(0x67)
-#define FFA_PARTITION_INFO_GET		FFA_SMC_32(0x68)
-#define FFA_ID_GET			FFA_SMC_32(0x69)
-#define FFA_MSG_POLL			FFA_SMC_32(0x6A)
-#define FFA_MSG_WAIT			FFA_SMC_32(0x6B)
-#define FFA_YIELD			FFA_SMC_32(0x6C)
-#define FFA_RUN				FFA_SMC_32(0x6D)
-#define FFA_MSG_SEND			FFA_SMC_32(0x6E)
-#define FFA_MSG_SEND_DIRECT_REQ		FFA_SMC_32(0x6F)
-#define FFA_FN64_MSG_SEND_DIRECT_REQ	FFA_SMC_64(0x6F)
-#define FFA_MSG_SEND_DIRECT_RESP	FFA_SMC_32(0x70)
-#define FFA_FN64_MSG_SEND_DIRECT_RESP	FFA_SMC_64(0x70)
-#define FFA_MEM_DONATE			FFA_SMC_32(0x71)
-#define FFA_FN64_MEM_DONATE		FFA_SMC_64(0x71)
-#define FFA_MEM_LEND			FFA_SMC_32(0x72)
-#define FFA_FN64_MEM_LEND		FFA_SMC_64(0x72)
-#define FFA_MEM_SHARE			FFA_SMC_32(0x73)
-#define FFA_FN64_MEM_SHARE		FFA_SMC_64(0x73)
-#define FFA_MEM_RETRIEVE_REQ		FFA_SMC_32(0x74)
-#define FFA_FN64_MEM_RETRIEVE_REQ	FFA_SMC_64(0x74)
-#define FFA_MEM_RETRIEVE_RESP		FFA_SMC_32(0x75)
-#define FFA_MEM_RELINQUISH		FFA_SMC_32(0x76)
-#define FFA_MEM_RECLAIM			FFA_SMC_32(0x77)
-#define FFA_MEM_OP_PAUSE		FFA_SMC_32(0x78)
-#define FFA_MEM_OP_RESUME		FFA_SMC_32(0x79)
-#define FFA_MEM_FRAG_RX			FFA_SMC_32(0x7A)
-#define FFA_MEM_FRAG_TX			FFA_SMC_32(0x7B)
-#define FFA_NORMAL_WORLD_RESUME		FFA_SMC_32(0x7C)
-
-/*
- * For some calls it is necessary to use SMC64 to pass or return 64-bit values.
- * For such calls FFA_FN_NATIVE(name) will choose the appropriate
- * (native-width) function ID.
- */
-#ifdef CONFIG_64BIT
-#define FFA_FN_NATIVE(name)	FFA_FN64_##name
-#else
-#define FFA_FN_NATIVE(name)	FFA_##name
-#endif
-
-/* FFA error codes. */
-#define FFA_RET_SUCCESS            (0)
-#define FFA_RET_NOT_SUPPORTED      (-1)
-#define FFA_RET_INVALID_PARAMETERS (-2)
-#define FFA_RET_NO_MEMORY          (-3)
-#define FFA_RET_BUSY               (-4)
-#define FFA_RET_INTERRUPTED        (-5)
-#define FFA_RET_DENIED             (-6)
-#define FFA_RET_RETRY              (-7)
-#define FFA_RET_ABORTED            (-8)
-
-#define MAJOR_VERSION_MASK	GENMASK(30, 16)
-#define MINOR_VERSION_MASK	GENMASK(15, 0)
-#define MAJOR_VERSION(x)	((u16)(FIELD_GET(MAJOR_VERSION_MASK, (x))))
-#define MINOR_VERSION(x)	((u16)(FIELD_GET(MINOR_VERSION_MASK, (x))))
-#define PACK_VERSION_INFO(major, minor)			\
-	(FIELD_PREP(MAJOR_VERSION_MASK, (major)) |	\
-	 FIELD_PREP(MINOR_VERSION_MASK, (minor)))
-#define FFA_VERSION_1_0		PACK_VERSION_INFO(1, 0)
 #define FFA_MIN_VERSION		FFA_VERSION_1_0
 
 #define SENDER_ID_MASK		GENMASK(31, 16)
@@ -120,12 +45,6 @@
 #define PACK_TARGET_INFO(s, r)		\
 	(FIELD_PREP(SENDER_ID_MASK, (s)) | FIELD_PREP(RECEIVER_ID_MASK, (r)))
 
-/*
- * FF-A specification mentions explicitly about '4K pages'. This should
- * not be confused with the kernel PAGE_SIZE, which is the translation
- * granule kernel is configured and may be one among 4K, 16K and 64K.
- */
-#define FFA_PAGE_SIZE		SZ_4K
 /*
  * Keeping RX TX buffer size as 4K for now
  * 64K may be preferred to keep it min a page in 64K PAGE_SIZE config
@@ -178,9 +97,9 @@ static struct ffa_drv_info *drv_info;
  */
 static u32 ffa_compatible_version_find(u32 version)
 {
-	u16 major = MAJOR_VERSION(version), minor = MINOR_VERSION(version);
-	u16 drv_major = MAJOR_VERSION(FFA_DRIVER_VERSION);
-	u16 drv_minor = MINOR_VERSION(FFA_DRIVER_VERSION);
+	u16 major = FFA_MAJOR_VERSION(version), minor = FFA_MINOR_VERSION(version);
+	u16 drv_major = FFA_MAJOR_VERSION(FFA_DRIVER_VERSION);
+	u16 drv_minor = FFA_MINOR_VERSION(FFA_DRIVER_VERSION);
 
 	if ((major < drv_major) || (major == drv_major && minor <= drv_minor))
 		return version;
@@ -204,16 +123,16 @@ static int ffa_version_check(u32 *version)
 
 	if (ver.a0 < FFA_MIN_VERSION) {
 		pr_err("Incompatible v%d.%d! Earliest supported v%d.%d\n",
-		       MAJOR_VERSION(ver.a0), MINOR_VERSION(ver.a0),
-		       MAJOR_VERSION(FFA_MIN_VERSION),
-		       MINOR_VERSION(FFA_MIN_VERSION));
+		       FFA_MAJOR_VERSION(ver.a0), FFA_MINOR_VERSION(ver.a0),
+		       FFA_MAJOR_VERSION(FFA_MIN_VERSION),
+		       FFA_MINOR_VERSION(FFA_MIN_VERSION));
 		return -EINVAL;
 	}
 
-	pr_info("Driver version %d.%d\n", MAJOR_VERSION(FFA_DRIVER_VERSION),
-		MINOR_VERSION(FFA_DRIVER_VERSION));
-	pr_info("Firmware version %d.%d found\n", MAJOR_VERSION(ver.a0),
-		MINOR_VERSION(ver.a0));
+	pr_info("Driver version %d.%d\n", FFA_MAJOR_VERSION(FFA_DRIVER_VERSION),
+		FFA_MINOR_VERSION(FFA_DRIVER_VERSION));
+	pr_info("Firmware version %d.%d found\n", FFA_MAJOR_VERSION(ver.a0),
+		FFA_MINOR_VERSION(ver.a0));
 	*version = ffa_compatible_version_find(ver.a0);
 
 	return 0;
include/linux/arm_ffa.h +83 -0
index 5f02d2e6b9d9..daff44d777fa 100644
@@ -11,6 +11,89 @@
 #include <linux/types.h>
 #include <linux/uuid.h>
 
+#define FFA_SMC(calling_convention, func_num)				\
+	ARM_SMCCC_CALL_VAL(ARM_SMCCC_FAST_CALL, (calling_convention),	\
+			   ARM_SMCCC_OWNER_STANDARD, (func_num))
+
+#define FFA_SMC_32(func_num)	FFA_SMC(ARM_SMCCC_SMC_32, (func_num))
+#define FFA_SMC_64(func_num)	FFA_SMC(ARM_SMCCC_SMC_64, (func_num))
+
+#define FFA_ERROR			FFA_SMC_32(0x60)
+#define FFA_SUCCESS			FFA_SMC_32(0x61)
+#define FFA_INTERRUPT			FFA_SMC_32(0x62)
+#define FFA_VERSION			FFA_SMC_32(0x63)
+#define FFA_FEATURES			FFA_SMC_32(0x64)
+#define FFA_RX_RELEASE			FFA_SMC_32(0x65)
+#define FFA_RXTX_MAP			FFA_SMC_32(0x66)
+#define FFA_FN64_RXTX_MAP		FFA_SMC_64(0x66)
+#define FFA_RXTX_UNMAP			FFA_SMC_32(0x67)
+#define FFA_PARTITION_INFO_GET		FFA_SMC_32(0x68)
+#define FFA_ID_GET			FFA_SMC_32(0x69)
+#define FFA_MSG_POLL			FFA_SMC_32(0x6A)
+#define FFA_MSG_WAIT			FFA_SMC_32(0x6B)
+#define FFA_YIELD			FFA_SMC_32(0x6C)
+#define FFA_RUN				FFA_SMC_32(0x6D)
+#define FFA_MSG_SEND			FFA_SMC_32(0x6E)
+#define FFA_MSG_SEND_DIRECT_REQ		FFA_SMC_32(0x6F)
+#define FFA_FN64_MSG_SEND_DIRECT_REQ	FFA_SMC_64(0x6F)
+#define FFA_MSG_SEND_DIRECT_RESP	FFA_SMC_32(0x70)
+#define FFA_FN64_MSG_SEND_DIRECT_RESP	FFA_SMC_64(0x70)
+#define FFA_MEM_DONATE			FFA_SMC_32(0x71)
+#define FFA_FN64_MEM_DONATE		FFA_SMC_64(0x71)
+#define FFA_MEM_LEND			FFA_SMC_32(0x72)
+#define FFA_FN64_MEM_LEND		FFA_SMC_64(0x72)
+#define FFA_MEM_SHARE			FFA_SMC_32(0x73)
+#define FFA_FN64_MEM_SHARE		FFA_SMC_64(0x73)
+#define FFA_MEM_RETRIEVE_REQ		FFA_SMC_32(0x74)
+#define FFA_FN64_MEM_RETRIEVE_REQ	FFA_SMC_64(0x74)
+#define FFA_MEM_RETRIEVE_RESP		FFA_SMC_32(0x75)
+#define FFA_MEM_RELINQUISH		FFA_SMC_32(0x76)
+#define FFA_MEM_RECLAIM			FFA_SMC_32(0x77)
+#define FFA_MEM_OP_PAUSE		FFA_SMC_32(0x78)
+#define FFA_MEM_OP_RESUME		FFA_SMC_32(0x79)
+#define FFA_MEM_FRAG_RX			FFA_SMC_32(0x7A)
+#define FFA_MEM_FRAG_TX			FFA_SMC_32(0x7B)
+#define FFA_NORMAL_WORLD_RESUME		FFA_SMC_32(0x7C)
+
+/*
+ * For some calls it is necessary to use SMC64 to pass or return 64-bit values.
+ * For such calls FFA_FN_NATIVE(name) will choose the appropriate
+ * (native-width) function ID.
+ */
+#ifdef CONFIG_64BIT
+#define FFA_FN_NATIVE(name)	FFA_FN64_##name
+#else
+#define FFA_FN_NATIVE(name)	FFA_##name
+#endif
+
+/* FFA error codes. */
+#define FFA_RET_SUCCESS            (0)
+#define FFA_RET_NOT_SUPPORTED      (-1)
+#define FFA_RET_INVALID_PARAMETERS (-2)
+#define FFA_RET_NO_MEMORY          (-3)
+#define FFA_RET_BUSY               (-4)
+#define FFA_RET_INTERRUPTED        (-5)
+#define FFA_RET_DENIED             (-6)
+#define FFA_RET_RETRY              (-7)
+#define FFA_RET_ABORTED            (-8)
+
+/* FFA version encoding */
+#define FFA_MAJOR_VERSION_MASK	GENMASK(30, 16)
+#define FFA_MINOR_VERSION_MASK	GENMASK(15, 0)
+#define FFA_MAJOR_VERSION(x)	((u16)(FIELD_GET(FFA_MAJOR_VERSION_MASK, (x))))
+#define FFA_MINOR_VERSION(x)	((u16)(FIELD_GET(FFA_MINOR_VERSION_MASK, (x))))
+#define FFA_PACK_VERSION_INFO(major, minor)			\
+	(FIELD_PREP(FFA_MAJOR_VERSION_MASK, (major)) |		\
+	 FIELD_PREP(FFA_MINOR_VERSION_MASK, (minor)))
+#define FFA_VERSION_1_0		FFA_PACK_VERSION_INFO(1, 0)
+
+/**
+ * FF-A specification mentions explicitly about '4K pages'. This should
+ * not be confused with the kernel PAGE_SIZE, which is the translation
+ * granule kernel is configured and may be one among 4K, 16K and 64K.
+ */
+#define FFA_PAGE_SIZE		SZ_4K
+
 /* FFA Bus/Device/Driver related */
 struct ffa_device {
 	int vm_id;
Signature · 1 line
2.38.1.431.g37b22c650d-goog
Signature · 2 lines
Regards,
Sudeep
#3 lore raw

PATCH 2/12 firmware: arm_ffa: Move comment before the field it is documenting

In reply to Quentin Perret

From: Will Deacon <[email protected]>

This is consistent with the other comments in the struct.

---
 include/linux/arm_ffa.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
include/linux/arm_ffa.h +1 -1
index daff44d777fa..c87aeecaa9b2 100644
@@ -244,11 +244,11 @@ struct ffa_mem_region_attributes {
 	 */
 #define FFA_MEM_RETRIEVE_SELF_BORROWER	BIT(0)
 	u8 flag;
-	u32 composite_off;
 	/*
 	 * Offset in bytes from the start of the outer `ffa_memory_region` to
 	 * an `struct ffa_mem_region_addr_range`.
 	 */
+	u32 composite_off;
 	u64 reserved;
 };
 
Signature · 1 line
2.38.1.431.g37b22c650d-goog
Reviewed-by on v1 2/12
#4 lore raw

Re: [PATCH 02/12] firmware: arm_ffa: Move comment before the field it is documenting

Hi Quentin,

Quoted 4 lines · Quentin Perret

From: Will Deacon <[email protected]>

This is consistent with the other comments in the struct.

Not sure how that happened :). Anyways,

I am yet to look at the other patches and I would like to have a setup to test it as well. So I will look at the other patches and test it later. The reason for reviewing the first 2 patches moving the code out of the driver is to check if they can be merged for v6.2 itself.

I may start pushing FF-A v1.1 changes for v6.3 and trying to avoid conflicts or cross tree dependencies. I know it is quite late for v6.2 but these changes are trivial and good to get it in for v6.2 if possible.

Will, thoughts ? If you agree, please take it via arm64 for v6.2. I don't have any FF-A changes for v6.2 ATM, so there should be not conflicts.

--
Regards,
Sudeep

maintainer
#5 lore raw

Re: [PATCH 02/12] firmware: arm_ffa: Move comment before the field it is documenting

Quoted 14 lines · Sudeep Holla
Quoted 4 lines · Quentin Perret

From: Will Deacon <[email protected]>

This is consistent with the other comments in the struct.

Not sure how that happened :). Anyways,

I am yet to look at the other patches and I would like to have a setup to test it as well. So I will look at the other patches and test it later. The reason for reviewing the first 2 patches moving the code out of the driver is to check if they can be merged for v6.2 itself.

Quoted 6 lines

I may start pushing FF-A v1.1 changes for v6.3 and trying to avoid conflicts or cross tree dependencies. I know it is quite late for v6.2 but these changes are trivial and good to get it in for v6.2 if possible.

Will, thoughts ? If you agree, please take it via arm64 for v6.2. I don't have any FF-A changes for v6.2 ATM, so there should be not conflicts.

You're asking me to merge my own patches? Of course I can do that :D

Cheers,

Will

#6 lore raw

PATCH 3/12 KVM: arm64: Block unsafe FF-A calls from the host

In reply to Quentin Perret

From: Will Deacon <[email protected]>

When KVM is initialised in protected mode, we must take care to filter certain FFA calls from the host kernel so that the integrity of guest and hypervisor memory is maintained and is not made available to the secure world.

As a first step, intercept and block all memory-related FF-A SMC calls from the host to EL3. This puts the framework in place for handling them properly.

---
 arch/arm64/kvm/hyp/include/nvhe/ffa.h |  16 ++++
 arch/arm64/kvm/hyp/nvhe/Makefile      |   2 +-
 arch/arm64/kvm/hyp/nvhe/ffa.c         | 113 ++++++++++++++++++++++++++
 arch/arm64/kvm/hyp/nvhe/hyp-main.c    |   3 +
 4 files changed, 133 insertions(+), 1 deletion(-)
 create mode 100644 arch/arm64/kvm/hyp/include/nvhe/ffa.h
 create mode 100644 arch/arm64/kvm/hyp/nvhe/ffa.c
arch/arm64/kvm/hyp/include/nvhe/ffa.h new file +16 -0
new file mode 100644
index 000000000000..fc09ec671e24
@@ -0,0 +1,16 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Copyright (C) 2022 - Google LLC
+ * Author: Andrew Walbran <[email protected]>
+ */
+#ifndef __KVM_HYP_FFA_H
+#define __KVM_HYP_FFA_H
+
+#include <asm/kvm_host.h>
+
+#define FFA_MIN_FUNC_NUM 0x60
+#define FFA_MAX_FUNC_NUM 0x7F
+
+bool kvm_host_ffa_handler(struct kvm_cpu_context *host_ctxt);
+
+#endif /* __KVM_HYP_FFA_H */
arch/arm64/kvm/hyp/nvhe/Makefile +1 -1
index be0a2bc3e20d..1b34d3ff57f3 100644
@@ -22,7 +22,7 @@ lib-objs := $(addprefix ../../../lib/, $(lib-objs))
 
 hyp-obj-y := timer-sr.o sysreg-sr.o debug-sr.o switch.o tlb.o hyp-init.o host.o \
 	 hyp-main.o hyp-smp.o psci-relay.o early_alloc.o page_alloc.o \
-	 cache.o setup.o mm.o mem_protect.o sys_regs.o pkvm.o stacktrace.o
+	 cache.o setup.o mm.o mem_protect.o sys_regs.o pkvm.o stacktrace.o ffa.o
 hyp-obj-y += ../vgic-v3-sr.o ../aarch32.o ../vgic-v2-cpuif-proxy.o ../entry.o \
 	 ../fpsimd.o ../hyp-entry.o ../exception.o ../pgtable.o
 hyp-obj-$(CONFIG_DEBUG_LIST) += list_debug.o
arch/arm64/kvm/hyp/nvhe/ffa.c new file +113 -0

113 lines not on this page · raw message

arch/arm64/kvm/hyp/nvhe/hyp-main.c +3 -0
index 728e01d4536b..223611e43279 100644
@@ -13,6 +13,7 @@
 #include <asm/kvm_hyp.h>
 #include <asm/kvm_mmu.h>
 
+#include <nvhe/ffa.h>
 #include <nvhe/mem_protect.h>
 #include <nvhe/mm.h>
 #include <nvhe/pkvm.h>
@@ -373,6 +374,8 @@ static void handle_host_smc(struct kvm_cpu_context *host_ctxt)
 	bool handled;
 
 	handled = kvm_host_psci_handler(host_ctxt);
+	if (!handled)
+		handled = kvm_host_ffa_handler(host_ctxt);
 	if (!handled)
 		default_host_smc_handler(host_ctxt);
 
Signature · 1 line
2.38.1.431.g37b22c650d-goog
#7 lore raw

Re: [PATCH 03/12] KVM: arm64: Block unsafe FF-A calls from the host

Quoted 10 lines · Quentin Perret

From: Will Deacon <[email protected]>

When KVM is initialised in protected mode, we must take care to filter certain FFA calls from the host kernel so that the integrity of guest and hypervisor memory is maintained and is not made available to the secure world.

As a first step, intercept and block all memory-related FF-A SMC calls from the host to EL3. This puts the framework in place for handling them properly.

Shouldn't FFA_FEATURES interception actually precede this patch? At this point in the series we're outright lying about the supported features to the host.

--
Thanks,
Oliver

maintainer
#8 lore raw

Re: [PATCH 03/12] KVM: arm64: Block unsafe FF-A calls from the host

Quoted 8 lines · Oliver Upton
Quoted 7 lines · Quentin Perret

From: Will Deacon <[email protected]>

When KVM is initialised in protected mode, we must take care to filter certain FFA calls from the host kernel so that the integrity of guest and hypervisor memory is maintained and is not made available to the secure world.

Quoted 7 lines
Quoted 3 lines

As a first step, intercept and block all memory-related FF-A SMC calls from the host to EL3. This puts the framework in place for handling them properly.

Shouldn't FFA_FEATURES interception actually precede this patch? At this point in the series we're outright lying about the supported features to the host.

FF-A is in a pretty sorry state after this patch as we block all the memory transactions, but I take your point that we should be consistent and not advertise the features that we're blocking.

I'll return FFA_RET_NOT_SUPPORTED for all FFA_FEATURES calls until the interception patch comes in later and does something smarter.

Will

#9 lore raw

Re: [PATCH 03/12] KVM: arm64: Block unsafe FF-A calls from the host

In reply to Quentin Perret

Sorry, hit send a bit too early. Reviewing the patch itself:

On Wed, Nov 16, 2022 at 05:03:26PM +0000, Quentin Perret wrote:

[...]

Quoted 16 lines
arch/arm64/kvm/hyp/nvhe/ffa.c new file +16 -0
+static bool ffa_call_unsupported(u64 func_id)
+{
+	switch (func_id) {
+	/* Unsupported memory management calls */
+	case FFA_FN64_MEM_RETRIEVE_REQ:
+	case FFA_MEM_RETRIEVE_RESP:
+	case FFA_MEM_RELINQUISH:
+	case FFA_MEM_OP_PAUSE:
+	case FFA_MEM_OP_RESUME:
+	case FFA_MEM_FRAG_RX:
+	case FFA_FN64_MEM_DONATE:
+	/* Indirect message passing via RX/TX buffers */
+	case FFA_MSG_SEND:
+	case FFA_MSG_POLL:
+	case FFA_MSG_WAIT:
+	/* 32-bit variants of 64-bit calls */

arch/arm64/kvm/hyp/nvhe/ffa.c line 84

Quoted 10 lines
arch/arm64/kvm/hyp/nvhe/ffa.c new file +10 -0
+	case FFA_MSG_SEND_DIRECT_REQ:
+	case FFA_MSG_SEND_DIRECT_RESP:
+	case FFA_RXTX_MAP:
+	case FFA_MEM_DONATE:
+	case FFA_MEM_RETRIEVE_REQ:
+		return true;
+	}
+
+	return false;
+}

Wouldn't an allowlist behave better in this case? While unlikely, you wouldn't want EL3 implementing some FFA_BACKDOOR_PVM SMC that falls outside of the denylist and is passed through.

Quoted 10 lines
arch/arm64/kvm/hyp/nvhe/ffa.c new file +10 -0
+bool kvm_host_ffa_handler(struct kvm_cpu_context *host_ctxt)
+{
+	DECLARE_REG(u64, func_id, host_ctxt, 0);
+	struct arm_smccc_res res;
+
+	if (!is_ffa_call(func_id))
+		return false;
+
+	switch (func_id) {
+	/* Memory management */

arch/arm64/kvm/hyp/nvhe/ffa.c line 105

Quoted 10 lines
arch/arm64/kvm/hyp/nvhe/ffa.c new file +10 -0
+	case FFA_FN64_RXTX_MAP:
+	case FFA_RXTX_UNMAP:
+	case FFA_MEM_SHARE:
+	case FFA_FN64_MEM_SHARE:
+	case FFA_MEM_LEND:
+	case FFA_FN64_MEM_LEND:
+	case FFA_MEM_RECLAIM:
+	case FFA_MEM_FRAG_TX:
+		break;
+	}

What is the purpose of this switch?

arch/arm64/kvm/hyp/nvhe/ffa.c line 108

Quoted 3 lines
arch/arm64/kvm/hyp/nvhe/ffa.c new file +3 -0
+
+	if (!ffa_call_unsupported(func_id))
+		return false; /* Pass through */

Another (tiny) benefit of implementing an allowlist is that it avoids the use of double-negative logic like this.

--
Thanks,
Oliver

maintainer
#10 lore raw

Re: [PATCH 03/12] KVM: arm64: Block unsafe FF-A calls from the host

Quoted 28 lines · Oliver Upton

Sorry, hit send a bit too early. Reviewing the patch itself:

On Wed, Nov 16, 2022 at 05:03:26PM +0000, Quentin Perret wrote:

[...]

Quoted 22 lines
arch/arm64/kvm/hyp/nvhe/ffa.c new file +22 -0
+static bool ffa_call_unsupported(u64 func_id)
+{
+	switch (func_id) {
+	/* Unsupported memory management calls */
+	case FFA_FN64_MEM_RETRIEVE_REQ:
+	case FFA_MEM_RETRIEVE_RESP:
+	case FFA_MEM_RELINQUISH:
+	case FFA_MEM_OP_PAUSE:
+	case FFA_MEM_OP_RESUME:
+	case FFA_MEM_FRAG_RX:
+	case FFA_FN64_MEM_DONATE:
+	/* Indirect message passing via RX/TX buffers */
+	case FFA_MSG_SEND:
+	case FFA_MSG_POLL:
+	case FFA_MSG_WAIT:
+	/* 32-bit variants of 64-bit calls */
+	case FFA_MSG_SEND_DIRECT_REQ:
+	case FFA_MSG_SEND_DIRECT_RESP:
+	case FFA_RXTX_MAP:
+	case FFA_MEM_DONATE:
+	case FFA_MEM_RETRIEVE_REQ:
+		return true;
Quoted 8 lines

arch/arm64/kvm/hyp/nvhe/ffa.c line 84

Quoted 4 lines
arch/arm64/kvm/hyp/nvhe/ffa.c new file +4 -0
+	}
+
+	return false;
+}

Wouldn't an allowlist behave better in this case? While unlikely, you wouldn't want EL3 implementing some FFA_BACKDOOR_PVM SMC that falls outside of the denylist and is passed through.

Given that we're not intercepting all SMCs (rather, only those in the FF-A service range), I think the denylist works much better because the default action is "allow". We _have_ to trust EL3 regardless, as it could just use any allowed SMC for a backdoor if it wanted. Ultimately, EL3 runs the show in the hierarchical security model of the architecture and we sadly can't do much about that.

Quoted 14 lines
Quoted 14 lines
arch/arm64/kvm/hyp/nvhe/ffa.c new file +14 -0
+bool kvm_host_ffa_handler(struct kvm_cpu_context *host_ctxt)
+{
+	DECLARE_REG(u64, func_id, host_ctxt, 0);
+	struct arm_smccc_res res;
+
+	if (!is_ffa_call(func_id))
+		return false;
+
+	switch (func_id) {
+	/* Memory management */
+	case FFA_FN64_RXTX_MAP:
+	case FFA_RXTX_UNMAP:
+	case FFA_MEM_SHARE:
+	case FFA_FN64_MEM_SHARE:
Quoted 8 lines

arch/arm64/kvm/hyp/nvhe/ffa.c line 105

Quoted 6 lines
arch/arm64/kvm/hyp/nvhe/ffa.c new file +6 -0
+	case FFA_MEM_LEND:
+	case FFA_FN64_MEM_LEND:
+	case FFA_MEM_RECLAIM:
+	case FFA_MEM_FRAG_TX:
+		break;
+	}

What is the purpose of this switch?

As of this patch, it's not serving any functional purpose. The idea is that later patches hook in here to provide handling at EL2. I'll remove it and introduce it bit by bit.

Quoted 7 lines

arch/arm64/kvm/hyp/nvhe/ffa.c line 108

Quoted 4 lines
arch/arm64/kvm/hyp/nvhe/ffa.c new file +3 -0
+
+	if (!ffa_call_unsupported(func_id))
+		return false; /* Pass through */

Another (tiny) benefit of implementing an allowlist is that it avoids the use of double-negative logic like this.

I should just rework this to be ffa_call_supported().

Will

#11 lore raw

PATCH 4/12 KVM: arm64: Probe FF-A version and host/hyp partition ID during init

In reply to Quentin Perret

From: Will Deacon <[email protected]>

Probe FF-A during pKVM initialisation so that we can detect any inconsistencies in the version or partition ID early on.

---
 arch/arm64/include/asm/kvm_host.h     |  1 +
 arch/arm64/kvm/arm.c                  |  1 +
 arch/arm64/kvm/hyp/include/nvhe/ffa.h |  1 +
 arch/arm64/kvm/hyp/nvhe/ffa.c         | 30 +++++++++++++++++++++++++++
 arch/arm64/kvm/hyp/nvhe/setup.c       |  5 +++++
 5 files changed, 38 insertions(+)
arch/arm64/include/asm/kvm_host.h +1 -0
index 63307e7dc9c5..ab5ccf47c05f 100644
@@ -347,6 +347,7 @@ struct kvm_host_data {
 struct kvm_host_psci_config {
 	/* PSCI version used by host. */
 	u32 version;
+	u32 smccc_version;
 
 	/* Function IDs used by host if version is v0.1. */
 	struct psci_0_1_function_ids function_ids_0_1;
arch/arm64/kvm/arm.c +1 -0
index 1d1a3b93a3de..8b93af23f327 100644
@@ -1780,6 +1780,7 @@ static bool init_psci_relay(void)
 	}
 
 	kvm_host_psci_config.version = psci_ops.get_version();
+	kvm_host_psci_config.smccc_version = arm_smccc_get_version();
 
 	if (kvm_host_psci_config.version == PSCI_VERSION(0, 1)) {
 		kvm_host_psci_config.function_ids_0_1 = get_psci_0_1_function_ids();
arch/arm64/kvm/hyp/include/nvhe/ffa.h +1 -0
index fc09ec671e24..5c9b92430ff3 100644
@@ -11,6 +11,7 @@
 #define FFA_MIN_FUNC_NUM 0x60
 #define FFA_MAX_FUNC_NUM 0x7F
 
+int hyp_ffa_init(void);
 bool kvm_host_ffa_handler(struct kvm_cpu_context *host_ctxt);
 
 #endif /* __KVM_HYP_FFA_H */
arch/arm64/kvm/hyp/nvhe/ffa.c +30 -0
index 6ccf935d3b41..969d72390844 100644
@@ -31,6 +31,12 @@
 #include <nvhe/ffa.h>
 #include <nvhe/trap_handler.h>
 
+/*
+ * "ID value 0 must be returned at the Non-secure physical FF-A instance"
+ * We share this ID with the host.
+ */
+#define HOST_FFA_ID	0
+
 static void ffa_to_smccc_error(struct arm_smccc_res *res, u64 ffa_errno)
 {
 	*res = (struct arm_smccc_res) {
@@ -111,3 +117,27 @@ bool kvm_host_ffa_handler(struct kvm_cpu_context *host_ctxt)
 	ffa_set_retval(host_ctxt, &res);
 	return true;
 }
+
+int hyp_ffa_init(void)
+{
+	struct arm_smccc_res res;
+
+	if (kvm_host_psci_config.smccc_version < ARM_SMCCC_VERSION_1_2)
+		return 0;
+
+	arm_smccc_1_1_smc(FFA_VERSION, FFA_VERSION_1_0, 0, 0, 0, 0, 0, 0, &res);
+	if (res.a0 == FFA_RET_NOT_SUPPORTED)
+		return 0;
+
+	if (res.a0 != FFA_VERSION_1_0)
+		return -EOPNOTSUPP;
+
+	arm_smccc_1_1_smc(FFA_ID_GET, 0, 0, 0, 0, 0, 0, 0, &res);
+	if (res.a0 != FFA_SUCCESS)
+		return -EOPNOTSUPP;
+
+	if (res.a2 != HOST_FFA_ID)
+		return -EINVAL;
+
+	return 0;
+}
arch/arm64/kvm/hyp/nvhe/setup.c +5 -0
index 110f04627785..c4ca174a0592 100644
@@ -11,6 +11,7 @@
 #include <asm/kvm_pkvm.h>
 
 #include <nvhe/early_alloc.h>
+#include <nvhe/ffa.h>
 #include <nvhe/fixed_config.h>
 #include <nvhe/gfp.h>
 #include <nvhe/memory.h>
@@ -314,6 +315,10 @@ void __noreturn __pkvm_init_finalise(void)
 	if (ret)
 		goto out;
 
+	ret = hyp_ffa_init();
+	if (ret)
+		goto out;
+
 	pkvm_hyp_vm_table_init(vm_table_base);
 out:
 	/*
Signature · 1 line
2.38.1.431.g37b22c650d-goog
#12 lore raw

PATCH 5/12 KVM: arm64: Handle FFA_FEATURES call from the host

In reply to Quentin Perret

From: Fuad Tabba <[email protected]>

Filter out advertising unsupported features, and only advertise features and properties that are supported by the hypervisor proxy.

---
 arch/arm64/kvm/hyp/nvhe/ffa.c | 44 +++++++++++++++++++++++++++++++++++
 1 file changed, 44 insertions(+)
arch/arm64/kvm/hyp/nvhe/ffa.c +44 -0
index 969d72390844..d199f868583e 100644
@@ -45,6 +45,16 @@ static void ffa_to_smccc_error(struct arm_smccc_res *res, u64 ffa_errno)
 	};
 }
 
+static void ffa_to_smccc_res_prop(struct arm_smccc_res *res, int ret, u64 prop)
+{
+	if (ret == FFA_RET_SUCCESS) {
+		*res = (struct arm_smccc_res) { .a0 = FFA_SUCCESS,
+						.a2 = prop };
+	} else {
+		ffa_to_smccc_error(res, ret);
+	}
+}
+
 static void ffa_set_retval(struct kvm_cpu_context *ctxt,
 			   struct arm_smccc_res *res)
 {
@@ -89,6 +99,35 @@ static bool ffa_call_unsupported(u64 func_id)
 	return false;
 }
 
+static bool do_ffa_features(struct arm_smccc_res *res,
+			    struct kvm_cpu_context *ctxt)
+{
+	DECLARE_REG(u32, id, ctxt, 1);
+	u64 prop = 0;
+	int ret = 0;
+
+	if (ffa_call_unsupported(id)) {
+		ret = FFA_RET_NOT_SUPPORTED;
+		goto out_handled;
+	}
+
+	switch (id) {
+	case FFA_MEM_SHARE:
+	case FFA_FN64_MEM_SHARE:
+	case FFA_MEM_LEND:
+	case FFA_FN64_MEM_LEND:
+		ret = FFA_RET_SUCCESS;
+		prop = 0; /* No support for dynamic buffers */
+		goto out_handled;
+	default:
+		return false;
+	}
+
+out_handled:
+	ffa_to_smccc_res_prop(res, ret, prop);
+	return true;
+}
+
 bool kvm_host_ffa_handler(struct kvm_cpu_context *host_ctxt)
 {
 	DECLARE_REG(u64, func_id, host_ctxt, 0);
@@ -98,6 +137,10 @@ bool kvm_host_ffa_handler(struct kvm_cpu_context *host_ctxt)
 		return false;
 
 	switch (func_id) {
+	case FFA_FEATURES:
+		if (!do_ffa_features(&res, host_ctxt))
+			return false;
+		goto out_handled;
 	/* Memory management */
 	case FFA_FN64_RXTX_MAP:
 	case FFA_RXTX_UNMAP:
@@ -114,6 +157,7 @@ bool kvm_host_ffa_handler(struct kvm_cpu_context *host_ctxt)
 		return false; /* Pass through */
 
 	ffa_to_smccc_error(&res, FFA_RET_NOT_SUPPORTED);
+out_handled:
 	ffa_set_retval(host_ctxt, &res);
 	return true;
 }
Signature · 1 line
2.38.1.431.g37b22c650d-goog
#13 lore raw

PATCH 6/12 KVM: arm64: Allocate pages for hypervisor FF-A mailboxes

In reply to Quentin Perret

From: Will Deacon <[email protected]>

The FF-A proxy code needs to allocate its own buffer pair for communication with EL3 and for forwarding calls from the host at EL1.

Reserve a couple of pages for this purpose and use them to initialise the hypervisor's FF-A buffer structure.

---
 arch/arm64/include/asm/kvm_pkvm.h     |  8 ++++++++
 arch/arm64/kvm/hyp/include/nvhe/ffa.h |  2 +-
 arch/arm64/kvm/hyp/nvhe/ffa.c         | 24 +++++++++++++++++++++++-
 arch/arm64/kvm/hyp/nvhe/setup.c       |  8 +++++++-
 arch/arm64/kvm/pkvm.c                 |  1 +
 5 files changed, 40 insertions(+), 3 deletions(-)
arch/arm64/include/asm/kvm_pkvm.h +8 -0
index 01129b0d4c68..2b495ec59deb 100644
@@ -106,4 +106,12 @@ static inline unsigned long host_s2_pgtable_pages(void)
 	return res;
 }
 
+#define KVM_FFA_MBOX_NR_PAGES	1
+
+static inline unsigned long hyp_ffa_proxy_pages(void)
+{
+	/* A page each for the hypervisor's RX and TX mailboxes. */
+	return 2 * KVM_FFA_MBOX_NR_PAGES;
+}
+
 #endif	/* __ARM64_KVM_PKVM_H__ */
arch/arm64/kvm/hyp/include/nvhe/ffa.h +1 -1
index 5c9b92430ff3..1becb10ecd80 100644
@@ -11,7 +11,7 @@
 #define FFA_MIN_FUNC_NUM 0x60
 #define FFA_MAX_FUNC_NUM 0x7F
 
-int hyp_ffa_init(void);
+int hyp_ffa_init(void *pages);
 bool kvm_host_ffa_handler(struct kvm_cpu_context *host_ctxt);
 
 #endif /* __KVM_HYP_FFA_H */
arch/arm64/kvm/hyp/nvhe/ffa.c +23 -1
index d199f868583e..7cb9095e00ff 100644
@@ -28,8 +28,11 @@
 
 #include <linux/arm-smccc.h>
 #include <linux/arm_ffa.h>
+#include <asm/kvm_pkvm.h>
+
 #include <nvhe/ffa.h>
 #include <nvhe/trap_handler.h>
+#include <nvhe/spinlock.h>
 
 /*
  * "ID value 0 must be returned at the Non-secure physical FF-A instance"
@@ -37,6 +40,19 @@
  */
 #define HOST_FFA_ID	0
 
+struct kvm_ffa_buffers {
+	hyp_spinlock_t lock;
+	void *tx;
+	void *rx;
+};
+
+/*
+ * Note that we don't currently lock these buffers explicitly, instead
+ * relying on the locking of the host FFA buffers as we only have one
+ * client.
+ */
+static struct kvm_ffa_buffers hyp_buffers;
+
 static void ffa_to_smccc_error(struct arm_smccc_res *res, u64 ffa_errno)
 {
 	*res = (struct arm_smccc_res) {
@@ -162,7 +178,7 @@ bool kvm_host_ffa_handler(struct kvm_cpu_context *host_ctxt)
 	return true;
 }
 
-int hyp_ffa_init(void)
+int hyp_ffa_init(void *pages)
 {
 	struct arm_smccc_res res;
 
@@ -183,5 +199,11 @@ int hyp_ffa_init(void)
 	if (res.a2 != HOST_FFA_ID)
 		return -EINVAL;
 
+	hyp_buffers = (struct kvm_ffa_buffers) {
+		.lock	= __HYP_SPIN_LOCK_UNLOCKED,
+		.tx	= pages,
+		.rx	= pages + (KVM_FFA_MBOX_NR_PAGES * PAGE_SIZE),
+	};
+
 	return 0;
 }
arch/arm64/kvm/hyp/nvhe/setup.c +7 -1
index c4ca174a0592..bb98630dfeaf 100644
@@ -29,6 +29,7 @@ static void *vmemmap_base;
 static void *vm_table_base;
 static void *hyp_pgt_base;
 static void *host_s2_pgt_base;
+static void *ffa_proxy_pages;
 static struct kvm_pgtable_mm_ops pkvm_pgtable_mm_ops;
 static struct hyp_pool hpool;
 
@@ -58,6 +59,11 @@ static int divide_memory_pool(void *virt, unsigned long size)
 	if (!host_s2_pgt_base)
 		return -ENOMEM;
 
+	nr_pages = hyp_ffa_proxy_pages();
+	ffa_proxy_pages = hyp_early_alloc_contig(nr_pages);
+	if (!ffa_proxy_pages)
+		return -ENOMEM;
+
 	return 0;
 }
 
@@ -315,7 +321,7 @@ void __noreturn __pkvm_init_finalise(void)
 	if (ret)
 		goto out;
 
-	ret = hyp_ffa_init();
+	ret = hyp_ffa_init(ffa_proxy_pages);
 	if (ret)
 		goto out;
 
arch/arm64/kvm/pkvm.c +1 -0
index cf56958b1492..bb405e274df6 100644
@@ -74,6 +74,7 @@ void __init kvm_hyp_reserve(void)
 	hyp_mem_pages += host_s2_pgtable_pages();
 	hyp_mem_pages += hyp_vm_table_pages();
 	hyp_mem_pages += hyp_vmemmap_pages(STRUCT_HYP_PAGE_SIZE);
+	hyp_mem_pages += hyp_ffa_proxy_pages();
 
 	/*
 	 * Try to allocate a PMD-aligned region to reduce TLB pressure once
Signature · 1 line
2.38.1.431.g37b22c650d-goog
#14 lore raw

PATCH 7/12 KVM: arm64: Handle FFA_RXTX_MAP and FFA_RXTX_UNMAP calls from the host

In reply to Quentin Perret

From: Will Deacon <[email protected]>

Handle FFA_RXTX_MAP and FFA_RXTX_UNMAP calls from the host by sharing the host's mailbox memory with the hypervisor and establishing a separate pair of mailboxes between the hypervisor and the SPMD at EL3.

---
 arch/arm64/kvm/hyp/nvhe/ffa.c | 173 ++++++++++++++++++++++++++++++++++
 include/linux/arm_ffa.h       |   8 ++
 2 files changed, 181 insertions(+)
arch/arm64/kvm/hyp/nvhe/ffa.c +173 -0
index 7cb9095e00ff..8388cc901c97 100644
@@ -31,6 +31,8 @@
 #include <asm/kvm_pkvm.h>
 
 #include <nvhe/ffa.h>
+#include <nvhe/mem_protect.h>
+#include <nvhe/memory.h>
 #include <nvhe/trap_handler.h>
 #include <nvhe/spinlock.h>
 
@@ -52,6 +54,7 @@ struct kvm_ffa_buffers {
  * client.
  */
 static struct kvm_ffa_buffers hyp_buffers;
+static struct kvm_ffa_buffers host_buffers;
 
 static void ffa_to_smccc_error(struct arm_smccc_res *res, u64 ffa_errno)
 {
@@ -71,6 +74,11 @@ static void ffa_to_smccc_res_prop(struct arm_smccc_res *res, int ret, u64 prop)
 	}
 }
 
+static void ffa_to_smccc_res(struct arm_smccc_res *res, int ret)
+{
+	ffa_to_smccc_res_prop(res, ret, 0);
+}
+
 static void ffa_set_retval(struct kvm_cpu_context *ctxt,
 			   struct arm_smccc_res *res)
 {
@@ -88,6 +96,140 @@ static bool is_ffa_call(u64 func_id)
 	       ARM_SMCCC_FUNC_NUM(func_id) <= FFA_MAX_FUNC_NUM;
 }
 
+static int spmd_map_ffa_buffers(u64 ffa_page_count)
+{
+	struct arm_smccc_res res;
+
+	arm_smccc_1_1_smc(FFA_FN64_RXTX_MAP,
+			  hyp_virt_to_phys(hyp_buffers.tx),
+			  hyp_virt_to_phys(hyp_buffers.rx),
+			  ffa_page_count,
+			  0, 0, 0, 0,
+			  &res);
+
+	return res.a0 == FFA_SUCCESS ? FFA_RET_SUCCESS : res.a2;
+}
+
+static int spmd_unmap_ffa_buffers(void)
+{
+	struct arm_smccc_res res;
+
+	arm_smccc_1_1_smc(FFA_RXTX_UNMAP,
+			  HOST_FFA_ID,
+			  0, 0, 0, 0, 0, 0,
+			  &res);
+
+	return res.a0 == FFA_SUCCESS ? FFA_RET_SUCCESS : res.a2;
+}
+
+static void do_ffa_rxtx_map(struct arm_smccc_res *res,
+			    struct kvm_cpu_context *ctxt)
+{
+	DECLARE_REG(phys_addr_t, tx, ctxt, 1);
+	DECLARE_REG(phys_addr_t, rx, ctxt, 2);
+	DECLARE_REG(u32, npages, ctxt, 3);
+	int ret = 0;
+	void *rx_virt, *tx_virt;
+
+	if (npages != (KVM_FFA_MBOX_NR_PAGES * PAGE_SIZE) / FFA_PAGE_SIZE) {
+		ret = FFA_RET_INVALID_PARAMETERS;
+		goto out;
+	}
+
+	if (!PAGE_ALIGNED(tx) || !PAGE_ALIGNED(rx)) {
+		ret = FFA_RET_INVALID_PARAMETERS;
+		goto out;
+	}
+
+	hyp_spin_lock(&host_buffers.lock);
+	if (host_buffers.tx) {
+		ret = FFA_RET_DENIED;
+		goto out_unlock;
+	}
+
+	ret = spmd_map_ffa_buffers(npages);
+	if (ret)
+		goto out_unlock;
+
+	ret = __pkvm_host_share_hyp(hyp_phys_to_pfn(tx));
+	if (ret) {
+		ret = FFA_RET_INVALID_PARAMETERS;
+		goto err_unmap;
+	}
+
+	ret = __pkvm_host_share_hyp(hyp_phys_to_pfn(rx));
+	if (ret) {
+		ret = FFA_RET_INVALID_PARAMETERS;
+		goto err_unshare_tx;
+	}
+
+	tx_virt = hyp_phys_to_virt(tx);
+	ret = hyp_pin_shared_mem(tx_virt, tx_virt + 1);
+	if (ret) {
+		ret = FFA_RET_INVALID_PARAMETERS;
+		goto err_unshare_rx;
+	}
+
+	rx_virt = hyp_phys_to_virt(rx);
+	ret = hyp_pin_shared_mem(rx_virt, rx_virt + 1);
+	if (ret) {
+		ret = FFA_RET_INVALID_PARAMETERS;
+		goto err_unpin_tx;
+	}
+
+	host_buffers.tx = tx_virt;
+	host_buffers.rx = rx_virt;
+
+out_unlock:
+	hyp_spin_unlock(&host_buffers.lock);
+out:
+	ffa_to_smccc_res(res, ret);
+	return;
+
+err_unpin_tx:
+	hyp_unpin_shared_mem(tx_virt, tx_virt + 1);
+err_unshare_rx:
+	__pkvm_host_unshare_hyp(hyp_phys_to_pfn(rx));
+err_unshare_tx:
+	__pkvm_host_unshare_hyp(hyp_phys_to_pfn(tx));
+err_unmap:
+	spmd_unmap_ffa_buffers();
+	goto out_unlock;
+}
+
+static void do_ffa_rxtx_unmap(struct arm_smccc_res *res,
+			      struct kvm_cpu_context *ctxt)
+{
+	DECLARE_REG(u32, id, ctxt, 1);
+	int ret = 0;
+
+	if (id != HOST_FFA_ID) {
+		ret = FFA_RET_INVALID_PARAMETERS;
+		goto out;
+	}
+
+	hyp_spin_lock(&host_buffers.lock);
+	if (!host_buffers.tx) {
+		ret = FFA_RET_INVALID_PARAMETERS;
+		goto out_unlock;
+	}
+
+	hyp_unpin_shared_mem(host_buffers.tx, host_buffers.tx + 1);
+	WARN_ON(__pkvm_host_unshare_hyp(hyp_virt_to_pfn(host_buffers.tx)));
+	host_buffers.tx = NULL;
+
+	hyp_unpin_shared_mem(host_buffers.rx, host_buffers.rx + 1);
+	WARN_ON(__pkvm_host_unshare_hyp(hyp_virt_to_pfn(host_buffers.rx)));
+	host_buffers.rx = NULL;
+
+	spmd_unmap_ffa_buffers();
+
+out_unlock:
+	hyp_spin_unlock(&host_buffers.lock);
+out:
+	ffa_to_smccc_res(res, ret);
+}
+
 static bool ffa_call_unsupported(u64 func_id)
 {
 	switch (func_id) {
@@ -159,7 +301,11 @@ bool kvm_host_ffa_handler(struct kvm_cpu_context *host_ctxt)
 		goto out_handled;
 	/* Memory management */
 	case FFA_FN64_RXTX_MAP:
+		do_ffa_rxtx_map(&res, host_ctxt);
+		goto out_handled;
 	case FFA_RXTX_UNMAP:
+		do_ffa_rxtx_unmap(&res, host_ctxt);
+		goto out_handled;
 	case FFA_MEM_SHARE:
 	case FFA_FN64_MEM_SHARE:
 	case FFA_MEM_LEND:
@@ -181,6 +327,7 @@ bool kvm_host_ffa_handler(struct kvm_cpu_context *host_ctxt)
 int hyp_ffa_init(void *pages)
 {
 	struct arm_smccc_res res;
+	size_t min_rxtx_sz;
 
 	if (kvm_host_psci_config.smccc_version < ARM_SMCCC_VERSION_1_2)
 		return 0;
@@ -199,11 +346,37 @@ int hyp_ffa_init(void *pages)
 	if (res.a2 != HOST_FFA_ID)
 		return -EINVAL;
 
+	arm_smccc_1_1_smc(FFA_FEATURES, FFA_FN64_RXTX_MAP,
+			  0, 0, 0, 0, 0, 0, &res);
+	if (res.a0 != FFA_SUCCESS)
+		return -EOPNOTSUPP;
+
+	switch (res.a2) {
+	case FFA_FEAT_RXTX_MIN_SZ_4K:
+		min_rxtx_sz = SZ_4K;
+		break;
+	case FFA_FEAT_RXTX_MIN_SZ_16K:
+		min_rxtx_sz = SZ_16K;
+		break;
+	case FFA_FEAT_RXTX_MIN_SZ_64K:
+		min_rxtx_sz = SZ_64K;
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	if (min_rxtx_sz > PAGE_SIZE)
+		return -EOPNOTSUPP;
+
 	hyp_buffers = (struct kvm_ffa_buffers) {
 		.lock	= __HYP_SPIN_LOCK_UNLOCKED,
 		.tx	= pages,
 		.rx	= pages + (KVM_FFA_MBOX_NR_PAGES * PAGE_SIZE),
 	};
 
+	host_buffers = (struct kvm_ffa_buffers) {
+		.lock	= __HYP_SPIN_LOCK_UNLOCKED,
+	};
+
 	return 0;
 }
include/linux/arm_ffa.h +8 -0
index c87aeecaa9b2..b9f81035eb41 100644
@@ -94,6 +94,14 @@
  */
 #define FFA_PAGE_SIZE		SZ_4K
 
+/*
+ * Minimum buffer size/alignment encodings returned by an FFA_FEATURES
+ * query for FFA_RXTX_MAP.
+ */
+#define FFA_FEAT_RXTX_MIN_SZ_4K		0
+#define FFA_FEAT_RXTX_MIN_SZ_64K	1
+#define FFA_FEAT_RXTX_MIN_SZ_16K	2
+
 /* FFA Bus/Device/Driver related */
 struct ffa_device {
 	int vm_id;
Signature · 1 line
2.38.1.431.g37b22c650d-goog
#15 lore raw

PATCH 8/12 KVM: arm64: Add FF-A helpers to share/unshare memory with secure world

In reply to Quentin Perret

From: Will Deacon <[email protected]>

Extend pKVM's memory protection code so that we can update the host's stage-2 page-table to track pages shared with secure world by the host using FF-A and prevent those pages from being mapped into a guest.

---
 arch/arm64/kvm/hyp/include/nvhe/mem_protect.h |  3 +
 arch/arm64/kvm/hyp/nvhe/mem_protect.c         | 68 +++++++++++++++++++
 2 files changed, 71 insertions(+)
arch/arm64/kvm/hyp/include/nvhe/mem_protect.h +3 -0
index b7bdbe63deed..0972faccc2af 100644
@@ -57,6 +57,7 @@ extern struct host_mmu host_mmu;
 enum pkvm_component_id {
 	PKVM_ID_HOST,
 	PKVM_ID_HYP,
+	PKVM_ID_FFA,
 };
 
 extern unsigned long hyp_nr_cpus;
@@ -66,6 +67,8 @@ int __pkvm_host_share_hyp(u64 pfn);
 int __pkvm_host_unshare_hyp(u64 pfn);
 int __pkvm_host_donate_hyp(u64 pfn, u64 nr_pages);
 int __pkvm_hyp_donate_host(u64 pfn, u64 nr_pages);
+int __pkvm_host_share_ffa(u64 pfn, u64 nr_pages);
+int __pkvm_host_unshare_ffa(u64 pfn, u64 nr_pages);
 
 bool addr_is_memory(phys_addr_t phys);
 int host_stage2_idmap_locked(phys_addr_t addr, u64 size, enum kvm_pgtable_prot prot);
arch/arm64/kvm/hyp/nvhe/mem_protect.c +68 -0

95 lines not on this page · raw message

Signature · 1 line
2.38.1.431.g37b22c650d-goog
#16 lore raw

PATCH 9/12 KVM: arm64: Handle FFA_MEM_SHARE calls from the host

In reply to Quentin Perret

From: Will Deacon <[email protected]>

Intercept FFA_MEM_SHARE/FFA_FN64_MEM_SHARE calls from the host and transition the host stage-2 page-table entries from the OWNED state to the SHARED_OWNED state prior to forwarding the call onto EL3.

---
 arch/arm64/kvm/hyp/nvhe/ffa.c | 153 ++++++++++++++++++++++++++++++++++
 1 file changed, 153 insertions(+)
arch/arm64/kvm/hyp/nvhe/ffa.c +153 -0
index 8388cc901c97..e6d85849c617 100644
@@ -122,6 +122,14 @@ static int spmd_unmap_ffa_buffers(void)
 	return res.a0 == FFA_SUCCESS ? FFA_RET_SUCCESS : res.a2;
 }
 
+static void spmd_mem_share(struct arm_smccc_res *res, u32 len, u32 fraglen)
+{
+	arm_smccc_1_1_smc(FFA_FN64_MEM_SHARE,
+			  len, fraglen,
+			  0, 0, 0, 0, 0,
+			  res);
+}
+
 static void do_ffa_rxtx_map(struct arm_smccc_res *res,
 			    struct kvm_cpu_context *ctxt)
 {
@@ -230,6 +238,149 @@ static void do_ffa_rxtx_unmap(struct arm_smccc_res *res,
 	ffa_to_smccc_res(res, ret);
 }
 
+static u32 __ffa_host_share_ranges(struct ffa_mem_region_addr_range *ranges,
+				   u32 nranges)
+{
+	u32 i;
+
+	for (i = 0; i < nranges; ++i) {
+		struct ffa_mem_region_addr_range *range = &ranges[i];
+		u64 sz = (u64)range->pg_cnt * FFA_PAGE_SIZE;
+		u64 pfn = hyp_phys_to_pfn(range->address);
+
+		if (!PAGE_ALIGNED(sz))
+			break;
+
+		if (__pkvm_host_share_ffa(pfn, sz / PAGE_SIZE))
+			break;
+	}
+
+	return i;
+}
+
+static u32 __ffa_host_unshare_ranges(struct ffa_mem_region_addr_range *ranges,
+				     u32 nranges)
+{
+	u32 i;
+
+	for (i = 0; i < nranges; ++i) {
+		struct ffa_mem_region_addr_range *range = &ranges[i];
+		u64 sz = (u64)range->pg_cnt * FFA_PAGE_SIZE;
+		u64 pfn = hyp_phys_to_pfn(range->address);
+
+		if (!PAGE_ALIGNED(sz))
+			break;
+
+		if (__pkvm_host_unshare_ffa(pfn, sz / PAGE_SIZE))
+			break;
+	}
+
+	return i;
+}
+
+static int ffa_host_share_ranges(struct ffa_mem_region_addr_range *ranges,
+				 u32 nranges)
+{
+	u32 nshared = __ffa_host_share_ranges(ranges, nranges);
+	int ret = 0;
+
+	if (nshared != nranges) {
+		WARN_ON(__ffa_host_unshare_ranges(ranges, nshared) != nshared);
+		ret = FFA_RET_DENIED;
+	}
+
+	return ret;
+}
+
+static int ffa_host_unshare_ranges(struct ffa_mem_region_addr_range *ranges,
+				   u32 nranges)
+{
+	u32 nunshared = __ffa_host_unshare_ranges(ranges, nranges);
+	int ret = 0;
+
+	if (nunshared != nranges) {
+		WARN_ON(__ffa_host_share_ranges(ranges, nunshared) != nunshared);
+		ret = FFA_RET_DENIED;
+	}
+
+	return ret;
+}
+
+static void do_ffa_mem_share(struct arm_smccc_res *res,
+			     struct kvm_cpu_context *ctxt)
+{
+	DECLARE_REG(u32, len, ctxt, 1);
+	DECLARE_REG(u32, fraglen, ctxt, 2);
+	DECLARE_REG(u64, addr_mbz, ctxt, 3);
+	DECLARE_REG(u32, npages_mbz, ctxt, 4);
+	struct ffa_composite_mem_region *reg;
+	struct ffa_mem_region *buf;
+	int ret = 0;
+	u32 offset;
+
+	if (addr_mbz || npages_mbz || fraglen > len ||
+	    fraglen > KVM_FFA_MBOX_NR_PAGES * PAGE_SIZE) {
+		ret = FFA_RET_INVALID_PARAMETERS;
+		goto out;
+	}
+
+	if (fraglen < len) {
+		ret = FFA_RET_ABORTED;
+		goto out;
+	}
+
+	if (fraglen < sizeof(struct ffa_mem_region) +
+		      sizeof(struct ffa_mem_region_attributes)) {
+		ret = FFA_RET_INVALID_PARAMETERS;
+		goto out;
+	}
+
+	hyp_spin_lock(&host_buffers.lock);
+	if (!host_buffers.tx) {
+		ret = FFA_RET_INVALID_PARAMETERS;
+		goto out_unlock;
+	}
+
+	buf = hyp_buffers.tx;
+	memcpy(buf, host_buffers.tx, fraglen);
+
+	offset = buf->ep_mem_access[0].composite_off;
+	if (!offset || buf->ep_count != 1 || buf->sender_id != HOST_FFA_ID) {
+		ret = FFA_RET_INVALID_PARAMETERS;
+		goto out_unlock;
+	}
+
+	if (fraglen < offset + sizeof(struct ffa_composite_mem_region)) {
+		ret = FFA_RET_INVALID_PARAMETERS;
+		goto out_unlock;
+	}
+
+	reg = (void *)buf + offset;
+	if (fraglen < offset + sizeof(struct ffa_composite_mem_region) +
+		      reg->addr_range_cnt *
+		      sizeof(struct ffa_mem_region_addr_range)) {
+		ret = FFA_RET_INVALID_PARAMETERS;
+		goto out_unlock;
+	}
+
+	ret = ffa_host_share_ranges(reg->constituents, reg->addr_range_cnt);
+	if (ret)
+		goto out_unlock;
+
+	spmd_mem_share(res, len, fraglen);
+	if (res->a0 != FFA_SUCCESS) {
+		WARN_ON(ffa_host_unshare_ranges(reg->constituents,
+						reg->addr_range_cnt));
+	}
+
+out_unlock:
+	hyp_spin_unlock(&host_buffers.lock);
+out:
+	if (ret)
+		ffa_to_smccc_res(res, ret);
+	return;
+}
+
 static bool ffa_call_unsupported(u64 func_id)
 {
 	switch (func_id) {
@@ -308,6 +459,8 @@ bool kvm_host_ffa_handler(struct kvm_cpu_context *host_ctxt)
 		goto out_handled;
 	case FFA_MEM_SHARE:
 	case FFA_FN64_MEM_SHARE:
+		do_ffa_mem_share(&res, host_ctxt);
+		goto out_handled;
 	case FFA_MEM_LEND:
 	case FFA_FN64_MEM_LEND:
 	case FFA_MEM_RECLAIM:
Signature · 1 line
2.38.1.431.g37b22c650d-goog
#17 lore raw

PATCH 10/12 KVM: arm64: Handle FFA_MEM_RECLAIM calls from the host

In reply to Quentin Perret

From: Will Deacon <[email protected]>

Intecept FFA_MEM_RECLAIM calls from the host and transition the host stage-2 page-table entries from the SHARED_OWNED state back to the OWNED state once EL3 has confirmed that the secure mapping has been reclaimed.

---
 arch/arm64/kvm/hyp/nvhe/ffa.c | 80 ++++++++++++++++++++++++++++++++++-
 1 file changed, 79 insertions(+), 1 deletion(-)
arch/arm64/kvm/hyp/nvhe/ffa.c +79 -1
index e6d85849c617..8f690b80af60 100644
@@ -130,6 +130,23 @@ static void spmd_mem_share(struct arm_smccc_res *res, u32 len, u32 fraglen)
 			  res);
 }
 
+static void spmd_mem_reclaim(struct arm_smccc_res *res, u32 handle_lo,
+			     u32 handle_hi, u32 flags)
+{
+	arm_smccc_1_1_smc(FFA_MEM_RECLAIM,
+			  handle_lo, handle_hi, flags,
+			  0, 0, 0, 0,
+			  res);
+}
+
+static void spmd_retrieve_req(struct arm_smccc_res *res, u32 len)
+{
+	arm_smccc_1_1_smc(FFA_FN64_MEM_RETRIEVE_REQ,
+			  len, len,
+			  0, 0, 0, 0, 0,
+			  res);
+}
+
 static void do_ffa_rxtx_map(struct arm_smccc_res *res,
 			    struct kvm_cpu_context *ctxt)
 {
@@ -381,6 +398,65 @@ static void do_ffa_mem_share(struct arm_smccc_res *res,
 	return;
 }
 
+static void do_ffa_mem_reclaim(struct arm_smccc_res *res,
+			       struct kvm_cpu_context *ctxt)
+{
+	DECLARE_REG(u32, handle_lo, ctxt, 1);
+	DECLARE_REG(u32, handle_hi, ctxt, 2);
+	DECLARE_REG(u32, flags, ctxt, 3);
+	struct ffa_composite_mem_region *reg;
+	struct ffa_mem_region *buf;
+	int ret = 0;
+	u32 offset;
+	u64 handle;
+
+	handle = PACK_HANDLE(handle_lo, handle_hi);
+
+	hyp_spin_lock(&host_buffers.lock);
+
+	buf = hyp_buffers.tx;
+	*buf = (struct ffa_mem_region) {
+		.sender_id	= HOST_FFA_ID,
+		.handle		= handle,
+	};
+
+	spmd_retrieve_req(res, sizeof(*buf));
+	buf = hyp_buffers.rx;
+	if (res->a0 != FFA_MEM_RETRIEVE_RESP)
+		goto out_unlock;
+
+	/* Check for fragmentation */
+	if (res->a1 != res->a2) {
+		ret = FFA_RET_ABORTED;
+		goto out_unlock;
+	}
+
+	offset = buf->ep_mem_access[0].composite_off;
+	/*
+	 * We can trust the SPMD to get this right, but let's at least
+	 * check that we end up with something that doesn't look _completely_
+	 * bogus.
+	 */
+	if (WARN_ON(offset > KVM_FFA_MBOX_NR_PAGES * PAGE_SIZE)) {
+		ret = FFA_RET_ABORTED;
+		goto out_unlock;
+	}
+
+	reg = (void *)buf + offset;
+	spmd_mem_reclaim(res, handle_lo, handle_hi, flags);
+	if (res->a0 != FFA_SUCCESS)
+		goto out_unlock;
+
+	/* If the SPMD was happy, then we should be too. */
+	WARN_ON(ffa_host_unshare_ranges(reg->constituents,
+					reg->addr_range_cnt));
+out_unlock:
+	hyp_spin_unlock(&host_buffers.lock);
+
+	if (ret)
+		ffa_to_smccc_res(res, ret);
+}
+
 static bool ffa_call_unsupported(u64 func_id)
 {
 	switch (func_id) {
@@ -461,9 +537,11 @@ bool kvm_host_ffa_handler(struct kvm_cpu_context *host_ctxt)
 	case FFA_FN64_MEM_SHARE:
 		do_ffa_mem_share(&res, host_ctxt);
 		goto out_handled;
+	case FFA_MEM_RECLAIM:
+		do_ffa_mem_reclaim(&res, host_ctxt);
+		goto out_handled;
 	case FFA_MEM_LEND:
 	case FFA_FN64_MEM_LEND:
-	case FFA_MEM_RECLAIM:
 	case FFA_MEM_FRAG_TX:
 		break;
 	}
Signature · 1 line
2.38.1.431.g37b22c650d-goog
#18 lore raw

PATCH 11/12 KVM: arm64: Handle FFA_MEM_LEND calls from the host

In reply to Quentin Perret

From: Will Deacon <[email protected]>

Handle FFA_MEM_LEND calls from the host by treating them identically to FFA_MEM_SHARE calls for the purposes of the host stage-2 page-table, but forwarding on the original request to EL3.

---
 arch/arm64/kvm/hyp/nvhe/ffa.c | 20 +++++++++++++-------
 1 file changed, 13 insertions(+), 7 deletions(-)
arch/arm64/kvm/hyp/nvhe/ffa.c +13 -7
index 8f690b80af60..84024cba12ff 100644
@@ -122,10 +122,10 @@ static int spmd_unmap_ffa_buffers(void)
 	return res.a0 == FFA_SUCCESS ? FFA_RET_SUCCESS : res.a2;
 }
 
-static void spmd_mem_share(struct arm_smccc_res *res, u32 len, u32 fraglen)
+static void spmd_mem_xfer(struct arm_smccc_res *res, u64 func_id, u32 len,
+			  u32 fraglen)
 {
-	arm_smccc_1_1_smc(FFA_FN64_MEM_SHARE,
-			  len, fraglen,
+	arm_smccc_1_1_smc(func_id, len, fraglen,
 			  0, 0, 0, 0, 0,
 			  res);
 }
@@ -323,8 +323,9 @@ static int ffa_host_unshare_ranges(struct ffa_mem_region_addr_range *ranges,
 	return ret;
 }
 
-static void do_ffa_mem_share(struct arm_smccc_res *res,
-			     struct kvm_cpu_context *ctxt)
+static __always_inline void do_ffa_mem_xfer(const u64 func_id,
+					    struct arm_smccc_res *res,
+					    struct kvm_cpu_context *ctxt)
 {
 	DECLARE_REG(u32, len, ctxt, 1);
 	DECLARE_REG(u32, fraglen, ctxt, 2);
@@ -335,6 +336,9 @@ static void do_ffa_mem_share(struct arm_smccc_res *res,
 	int ret = 0;
 	u32 offset;
 
+	BUILD_BUG_ON(func_id != FFA_FN64_MEM_SHARE &&
+		     func_id != FFA_FN64_MEM_LEND);
+
 	if (addr_mbz || npages_mbz || fraglen > len ||
 	    fraglen > KVM_FFA_MBOX_NR_PAGES * PAGE_SIZE) {
 		ret = FFA_RET_INVALID_PARAMETERS;
@@ -384,7 +388,7 @@ static void do_ffa_mem_share(struct arm_smccc_res *res,
 	if (ret)
 		goto out_unlock;
 
-	spmd_mem_share(res, len, fraglen);
+	spmd_mem_xfer(res, func_id, len, fraglen);
 	if (res->a0 != FFA_SUCCESS) {
 		WARN_ON(ffa_host_unshare_ranges(reg->constituents,
 						reg->addr_range_cnt));
@@ -535,13 +539,15 @@ bool kvm_host_ffa_handler(struct kvm_cpu_context *host_ctxt)
 		goto out_handled;
 	case FFA_MEM_SHARE:
 	case FFA_FN64_MEM_SHARE:
-		do_ffa_mem_share(&res, host_ctxt);
+		do_ffa_mem_xfer(FFA_FN64_MEM_SHARE, &res, host_ctxt);
 		goto out_handled;
 	case FFA_MEM_RECLAIM:
 		do_ffa_mem_reclaim(&res, host_ctxt);
 		goto out_handled;
 	case FFA_MEM_LEND:
 	case FFA_FN64_MEM_LEND:
+		do_ffa_mem_xfer(FFA_FN64_MEM_LEND, &res, host_ctxt);
+		goto out_handled;
 	case FFA_MEM_FRAG_TX:
 		break;
 	}
Signature · 1 line
2.38.1.431.g37b22c650d-goog
#19 lore raw

PATCH 12/12 ANDROID: KVM: arm64: pkvm: Add support for fragmented FF-A descriptors

In reply to Quentin Perret

FF-A memory descriptors may need to be sent in fragments when they don't fit in the mailboxes. Doing so involves using the FRAG_TX and FRAG_RX primitives defined in the FF-A protocol.

Add support in the pKVM FF-A relayer for fragmented descriptors by monitoring outgoing FRAG_TX transactions and by buffering large descriptors on the reclaim path.

---
 arch/arm64/include/asm/kvm_pkvm.h |  17 ++-
 arch/arm64/kvm/hyp/nvhe/ffa.c     | 170 +++++++++++++++++++++++++-----
 2 files changed, 161 insertions(+), 26 deletions(-)
arch/arm64/include/asm/kvm_pkvm.h +15 -2
index 2b495ec59deb..e46250a02017 100644
@@ -6,7 +6,9 @@
 #ifndef __ARM64_KVM_PKVM_H__
 #define __ARM64_KVM_PKVM_H__
 
+#include <linux/arm_ffa.h>
 #include <linux/memblock.h>
+#include <linux/scatterlist.h>
 #include <asm/kvm_pgtable.h>
 
 /* Maximum number of VMs that can co-exist under pKVM. */
@@ -110,8 +112,19 @@ static inline unsigned long host_s2_pgtable_pages(void)
 
 static inline unsigned long hyp_ffa_proxy_pages(void)
 {
-	/* A page each for the hypervisor's RX and TX mailboxes. */
-	return 2 * KVM_FFA_MBOX_NR_PAGES;
+	size_t desc_max;
+
+	/*
+	 * The hypervisor FFA proxy needs enough memory to buffer a fragmented
+	 * descriptor returned from EL3 in response to a RETRIEVE_REQ call.
+	 */
+	desc_max = sizeof(struct ffa_mem_region) +
+		   sizeof(struct ffa_mem_region_attributes) +
+		   sizeof(struct ffa_composite_mem_region) +
+		   SG_MAX_SEGMENTS * sizeof(struct ffa_mem_region_addr_range);
+
+	/* Plus a page each for the hypervisor's RX and TX mailboxes. */
+	return (2 * KVM_FFA_MBOX_NR_PAGES) + DIV_ROUND_UP(desc_max, PAGE_SIZE);
 }
 
 #endif	/* __ARM64_KVM_PKVM_H__ */
arch/arm64/kvm/hyp/nvhe/ffa.c +146 -24

269 lines not on this page · raw message

Signature · 1 line
2.38.1.431.g37b22c650d-goog
#20 lore raw

Re: [PATCH 12/12] ANDROID: KVM: arm64: pkvm: Add support for fragmented FF-A descriptors

Quoted 7 lines · Quentin Perret

FF-A memory descriptors may need to be sent in fragments when they don't fit in the mailboxes. Doing so involves using the FRAG_TX and FRAG_RX primitives defined in the FF-A protocol.

Add support in the pKVM FF-A relayer for fragmented descriptors by monitoring outgoing FRAG_TX transactions and by buffering large descriptors on the reclaim path.

And it seems that I have forgotten to remove the ANDROID: tag from this patch prior to posting ... Please ignore, it'll be gone in v2.

Cheers,
Quentin

maintainer
#21 lore raw
In reply to Quentin Perret
Quoted 10 lines · Quentin Perret

pKVM's primary goal is to protect guest pages from a compromised host by enforcing access control restrictions using stage-2 page-tables. Sadly, this cannot prevent TrustZone from accessing non-secure memory, and a compromised host could, for example, perform a 'confused deputy' attack by asking TrustZone to use pages that have been donated to protected guests. This would effectively allow the host to have TrustZone exfiltrate guest secrets on its behalf, hence breaking the isolation that pKVM intends to provide.

[...]

Applied first two cleanups to arm64 (for-next/ffa), thanks!

[01/12] firmware: arm_ffa: Move constants to header file

        https://git.kernel.org/arm64/c/229d58e31678
[02/12] firmware: arm_ffa: Move comment before the field it is documenting
        https://git.kernel.org/arm64/c/c8e320b00a2a

Cheers,

Signature · 5 lines
Will

https://fixes.arm64.dev
https://next.arm64.dev
https://will.arm64.dev