adapt to VFIO live migration v2 (#782)

This commit adapts the vfio-user protocol specification and the libvfio-user
implementation to v2 of the VFIO live migration interface, as used in the kernel
and QEMU.

The differences between v1 and v2 are discussed in this email thread [1], and we
slightly differ from upstream VFIO v2 in that instead of transferring data over
a new FD, we use the existing UNIX socket with new commands
VFIO_USER_MIG_DATA_READ/WRITE. We also don't yet use P2P states.

The updated spec was submitted to qemu-devel [2].

[1] https://lore.kernel.org/all/20220130160826.32449-9-yishaih@nvidia.com/
[2] https://lore.kernel.org/all/20230718094150.110183-1-william.henderson@nutanix.com/

Signed-off-by: William Henderson <william.henderson@nutanix.com>

1 files changed, 87 insertions, 18 deletions

diff --git a/include/vfio-user.h b/include/vfio-user.h
index a749938..0b115d3 100644
--- a/include/vfio-user.h
+++ b/include/vfio-user.h
@@ -66,7 +66,10 @@ enum vfio_user_command {
     VFIO_USER_DMA_READ                  = 11,
     VFIO_USER_DMA_WRITE                 = 12,
     VFIO_USER_DEVICE_RESET              = 13,
-    VFIO_USER_DIRTY_PAGES               = 14,
+    VFIO_USER_REGION_WRITE_MULTI        = 15,
+    VFIO_USER_DEVICE_FEATURE            = 16,
+    VFIO_USER_MIG_DATA_READ             = 17,
+    VFIO_USER_MIG_DATA_WRITE            = 18,
     VFIO_USER_MAX,
 };
 
@@ -200,31 +203,97 @@ typedef struct vfio_user_region_io_fds_reply {
     } sub_regions[];
 } __attribute__((packed)) vfio_user_region_io_fds_reply_t;
 
+/* Analogous to struct vfio_device_feature_dma_logging_range */
+struct vfio_user_device_feature_dma_logging_range {
+    uint64_t iova;
+    uint64_t length;
+} __attribute__((packed));
 
-/* Analogous to vfio_iommu_type1_dirty_bitmap. */
-struct vfio_user_dirty_pages {
-    uint32_t    argsz;
-#ifndef VFIO_IOMMU_DIRTY_PAGES_FLAG_START
-#define VFIO_IOMMU_DIRTY_PAGES_FLAG_START      (1 << 0)
-#define VFIO_IOMMU_DIRTY_PAGES_FLAG_STOP       (1 << 1)
-#define VFIO_IOMMU_DIRTY_PAGES_FLAG_GET_BITMAP (1 << 2)
-#endif
-    uint32_t    flags;
+/* Analogous to struct vfio_device_feature_dma_logging_control */
+struct vfio_user_device_feature_dma_logging_control {
+    uint64_t page_size;
+    uint32_t num_ranges;
+    uint32_t reserved;
+    struct vfio_user_device_feature_dma_logging_range ranges[];
 } __attribute__((packed));
 
-/* Analogous to struct vfio_iommu_type1_dirty_bitmap_get. */
-struct vfio_user_bitmap_range {
+/* Analogous to struct vfio_device_feature_dma_logging_report */
+struct vfio_user_device_feature_dma_logging_report {
     uint64_t iova;
-    uint64_t size;
-    struct vfio_user_bitmap bitmap;
+    uint64_t length;
+    uint64_t page_size;
+    uint8_t  bitmap[];
+} __attribute__((packed));
+
+#ifndef VFIO_DEVICE_FEATURE_DMA_LOGGING_START
+#define VFIO_DEVICE_FEATURE_DMA_LOGGING_START  6
+#define VFIO_DEVICE_FEATURE_DMA_LOGGING_STOP   7
+#define VFIO_DEVICE_FEATURE_DMA_LOGGING_REPORT 8
+#endif
+
+/* Analogous to struct vfio_device_feature */
+struct vfio_user_device_feature {
+	uint32_t	argsz;
+	uint32_t	flags;
+#ifndef VFIO_DEVICE_FEATURE_MASK
+#define VFIO_DEVICE_FEATURE_MASK	(0xffff)  /* 16-bit feature index */
+#define VFIO_DEVICE_FEATURE_GET		(1 << 16) /* Get feature into data[] */
+#define VFIO_DEVICE_FEATURE_SET		(1 << 17) /* Set feature from data[] */
+#define VFIO_DEVICE_FEATURE_PROBE	(1 << 18) /* Probe feature support */
+#endif
+	uint8_t  	data[];
+} __attribute__((packed));
+
+/* Analogous to struct vfio_device_feature_migration */
+struct vfio_user_device_feature_migration {
+    uint64_t flags;
+#ifndef VFIO_REGION_TYPE_MIGRATION_DEPRECATED
+#define VFIO_MIGRATION_STOP_COPY    (1 << 0)
+#define VFIO_MIGRATION_P2P          (1 << 1)
+#endif
+/*
+ * PRE_COPY was added in a later kernel version, after
+ * VFIO_REGION_TYPE_MIGRATION_DEPRECATED had been introduced.
+ */
+#ifndef VFIO_MIGRATION_PRE_COPY
+#define VFIO_MIGRATION_PRE_COPY     (1 << 2)
+#endif
 } __attribute__((packed));
+#ifndef VFIO_REGION_TYPE_MIGRATION_DEPRECATED
+#define VFIO_DEVICE_FEATURE_MIGRATION 1
+#endif
+_Static_assert(sizeof(struct vfio_user_device_feature_migration) == 8,
+               "bad vfio_user_device_feature_migration size");
 
-#ifndef VFIO_REGION_TYPE_MIGRATION
+/* Analogous to struct vfio_device_feature_mig_state */
+struct vfio_user_device_feature_mig_state {
+    uint32_t    device_state;
+    uint32_t    data_fd;
+} __attribute__((packed));
+#ifndef VFIO_DEVICE_FEATURE_MIG_DEVICE_STATE
+#define VFIO_DEVICE_FEATURE_MIG_DEVICE_STATE 2
+#endif
+_Static_assert(sizeof(struct vfio_user_device_feature_migration) == 8,
+               "bad vfio_user_device_feature_mig_state size");
 
-#define VFIO_REGION_TYPE_MIGRATION (3)
-#define VFIO_REGION_SUBTYPE_MIGRATION (1)
+/* Analogous to enum vfio_device_mig_state */
+enum vfio_user_device_mig_state {
+    VFIO_USER_DEVICE_STATE_ERROR = 0,
+    VFIO_USER_DEVICE_STATE_STOP = 1,
+    VFIO_USER_DEVICE_STATE_RUNNING = 2,
+    VFIO_USER_DEVICE_STATE_STOP_COPY = 3,
+    VFIO_USER_DEVICE_STATE_RESUMING = 4,
+    VFIO_USER_DEVICE_STATE_RUNNING_P2P = 5,
+    VFIO_USER_DEVICE_STATE_PRE_COPY = 6,
+    VFIO_USER_DEVICE_STATE_PRE_COPY_P2P = 7,
+    VFIO_USER_DEVICE_NUM_STATES = 8,
+};
 
-#endif /* VFIO_REGION_TYPE_MIGRATION */
+struct vfio_user_mig_data {
+    uint32_t    argsz;
+    uint32_t    size;
+    uint8_t     data[];
+} __attribute__((packed));
 
 #ifdef __cplusplus
 }