diff options
Diffstat (limited to 'include')
| -rw-r--r-- | include/hw/audio/virtio-snd.h | 235 | ||||
| -rw-r--r-- | include/hw/cxl/cxl.h | 6 | ||||
| -rw-r--r-- | include/hw/cxl/cxl_component.h | 6 | ||||
| -rw-r--r-- | include/hw/cxl/cxl_device.h | 109 | ||||
| -rw-r--r-- | include/hw/cxl/cxl_events.h | 3 | ||||
| -rw-r--r-- | include/hw/cxl/cxl_pci.h | 6 | ||||
| -rw-r--r-- | include/hw/pci-bridge/cxl_upstream_port.h | 19 | ||||
| -rw-r--r-- | include/hw/virtio/vhost-backend.h | 24 | ||||
| -rw-r--r-- | include/hw/virtio/vhost-user.h | 1 | ||||
| -rw-r--r-- | include/hw/virtio/vhost.h | 113 |
10 files changed, 504 insertions, 18 deletions
diff --git a/include/hw/audio/virtio-snd.h b/include/hw/audio/virtio-snd.h new file mode 100644 index 0000000000..c3767f442b --- /dev/null +++ b/include/hw/audio/virtio-snd.h @@ -0,0 +1,235 @@ +/* + * VIRTIO Sound Device conforming to + * + * "Virtual I/O Device (VIRTIO) Version 1.2 + * Committee Specification Draft 01 + * 09 May 2022" + * + * Copyright (c) 2023 Emmanouil Pitsidianakis <manos.pitsidianakis@linaro.org> + * Copyright (C) 2019 OpenSynergy GmbH + * + * This work is licensed under the terms of the GNU GPL, version 2 or + * (at your option) any later version. See the COPYING file in the + * top-level directory. + */ + +#ifndef QEMU_VIRTIO_SOUND_H +#define QEMU_VIRTIO_SOUND_H + +#include "hw/virtio/virtio.h" +#include "audio/audio.h" +#include "standard-headers/linux/virtio_ids.h" +#include "standard-headers/linux/virtio_snd.h" + +#define TYPE_VIRTIO_SND "virtio-sound-device" +#define VIRTIO_SND(obj) \ + OBJECT_CHECK(VirtIOSound, (obj), TYPE_VIRTIO_SND) + +/* CONFIGURATION SPACE */ + +typedef struct virtio_snd_config virtio_snd_config; + +/* COMMON DEFINITIONS */ + +/* common header for request/response*/ +typedef struct virtio_snd_hdr virtio_snd_hdr; + +/* event notification */ +typedef struct virtio_snd_event virtio_snd_event; + +/* common control request to query an item information */ +typedef struct virtio_snd_query_info virtio_snd_query_info; + +/* JACK CONTROL MESSAGES */ + +typedef struct virtio_snd_jack_hdr virtio_snd_jack_hdr; + +/* jack information structure */ +typedef struct virtio_snd_jack_info virtio_snd_jack_info; + +/* jack remapping control request */ +typedef struct virtio_snd_jack_remap virtio_snd_jack_remap; + +/* + * PCM CONTROL MESSAGES + */ +typedef struct virtio_snd_pcm_hdr virtio_snd_pcm_hdr; + +/* PCM stream info structure */ +typedef struct virtio_snd_pcm_info virtio_snd_pcm_info; + +/* set PCM stream params */ +typedef struct virtio_snd_pcm_set_params virtio_snd_pcm_set_params; + +/* I/O request header */ +typedef struct virtio_snd_pcm_xfer virtio_snd_pcm_xfer; + +/* I/O request status */ +typedef struct virtio_snd_pcm_status virtio_snd_pcm_status; + +/* device structs */ + +typedef struct VirtIOSound VirtIOSound; + +typedef struct VirtIOSoundPCMStream VirtIOSoundPCMStream; + +typedef struct virtio_snd_ctrl_command virtio_snd_ctrl_command; + +typedef struct VirtIOSoundPCM VirtIOSoundPCM; + +typedef struct VirtIOSoundPCMBuffer VirtIOSoundPCMBuffer; + +/* + * The VirtIO sound spec reuses layouts and values from the High Definition + * Audio spec (virtio/v1.2: 5.14 Sound Device). This struct handles each I/O + * message's buffer (virtio/v1.2: 5.14.6.8 PCM I/O Messages). + * + * In the case of TX (i.e. playback) buffers, we defer reading the raw PCM data + * from the virtqueue until QEMU's sound backsystem calls the output callback. + * This is tracked by the `bool populated;` field, which is set to true when + * data has been read into our own buffer for consumption. + * + * VirtIOSoundPCMBuffer has a dynamic size since it includes the raw PCM data + * in its allocation. It must be initialized and destroyed as follows: + * + * size_t size = [[derived from owned VQ element descriptor sizes]]; + * buffer = g_malloc0(sizeof(VirtIOSoundPCMBuffer) + size); + * buffer->elem = [[owned VQ element]]; + * + * [..] + * + * g_free(buffer->elem); + * g_free(buffer); + */ +struct VirtIOSoundPCMBuffer { + QSIMPLEQ_ENTRY(VirtIOSoundPCMBuffer) entry; + VirtQueueElement *elem; + VirtQueue *vq; + size_t size; + /* + * In TX / Plaback, `offset` represents the first unused position inside + * `data`. If `offset == size` then there are no unused data left. + */ + uint64_t offset; + /* Used for the TX queue for lazy I/O copy from `elem` */ + bool populated; + /* + * VirtIOSoundPCMBuffer is an unsized type because it ends with an array of + * bytes. The size of `data` is determined from the I/O message's read-only + * or write-only size when allocating VirtIOSoundPCMBuffer. + */ + uint8_t data[]; +}; + +struct VirtIOSoundPCM { + VirtIOSound *snd; + /* + * PCM parameters are a separate field instead of a VirtIOSoundPCMStream + * field, because the operation of PCM control requests is first + * VIRTIO_SND_R_PCM_SET_PARAMS and then VIRTIO_SND_R_PCM_PREPARE; this + * means that some times we get parameters without having an allocated + * stream yet. + */ + virtio_snd_pcm_set_params *pcm_params; + VirtIOSoundPCMStream **streams; +}; + +struct VirtIOSoundPCMStream { + VirtIOSoundPCM *pcm; + virtio_snd_pcm_info info; + virtio_snd_pcm_set_params params; + uint32_t id; + /* channel position values (VIRTIO_SND_CHMAP_XXX) */ + uint8_t positions[VIRTIO_SND_CHMAP_MAX_SIZE]; + VirtIOSound *s; + bool flushing; + audsettings as; + union { + SWVoiceIn *in; + SWVoiceOut *out; + } voice; + QemuMutex queue_mutex; + bool active; + QSIMPLEQ_HEAD(, VirtIOSoundPCMBuffer) queue; + QSIMPLEQ_HEAD(, VirtIOSoundPCMBuffer) invalid; +}; + +/* + * PCM stream state machine. + * ------------------------- + * + * 5.14.6.6.1 PCM Command Lifecycle + * ================================ + * + * A PCM stream has the following command lifecycle: + * - `SET PARAMETERS` + * The driver negotiates the stream parameters (format, transport, etc) with + * the device. + * Possible valid transitions: `SET PARAMETERS`, `PREPARE`. + * - `PREPARE` + * The device prepares the stream (allocates resources, etc). + * Possible valid transitions: `SET PARAMETERS`, `PREPARE`, `START`, + * `RELEASE`. Output only: the driver transfers data for pre-buffing. + * - `START` + * The device starts the stream (unmute, putting into running state, etc). + * Possible valid transitions: `STOP`. + * The driver transfers data to/from the stream. + * - `STOP` + * The device stops the stream (mute, putting into non-running state, etc). + * Possible valid transitions: `START`, `RELEASE`. + * - `RELEASE` + * The device releases the stream (frees resources, etc). + * Possible valid transitions: `SET PARAMETERS`, `PREPARE`. + * + * +---------------+ +---------+ +---------+ +-------+ +-------+ + * | SetParameters | | Prepare | | Release | | Start | | Stop | + * +---------------+ +---------+ +---------+ +-------+ +-------+ + * |- | | | | + * || | | | | + * |< | | | | + * |------------->| | | | + * |<-------------| | | | + * | |- | | | + * | || | | | + * | |< | | | + * | |--------------------->| | + * | |---------->| | | + * | | | |-------->| + * | | | |<--------| + * | | |<-------------------| + * |<-------------------------| | | + * | |<----------| | | + * + * CTRL in the VirtIOSound device + * ============================== + * + * The control messages that affect the state of a stream arrive in the + * `virtio_snd_handle_ctrl()` queue callback and are of type `struct + * virtio_snd_ctrl_command`. They are stored in a queue field in the device + * type, `VirtIOSound`. This allows deferring the CTRL request completion if + * it's not immediately possible due to locking/state reasons. + * + * The CTRL message is finally handled in `process_cmd()`. + */ +struct VirtIOSound { + VirtIODevice parent_obj; + + VirtQueue *queues[VIRTIO_SND_VQ_MAX]; + uint64_t features; + VirtIOSoundPCM *pcm; + QEMUSoundCard card; + VMChangeStateEntry *vmstate; + virtio_snd_config snd_conf; + QemuMutex cmdq_mutex; + QTAILQ_HEAD(, virtio_snd_ctrl_command) cmdq; + bool processing_cmdq; +}; + +struct virtio_snd_ctrl_command { + VirtQueueElement *elem; + VirtQueue *vq; + virtio_snd_hdr ctrl; + virtio_snd_hdr resp; + QTAILQ_ENTRY(virtio_snd_ctrl_command) next; +}; +#endif diff --git a/include/hw/cxl/cxl.h b/include/hw/cxl/cxl.h index 4944725849..75e47b6864 100644 --- a/include/hw/cxl/cxl.h +++ b/include/hw/cxl/cxl.h @@ -61,4 +61,10 @@ OBJECT_DECLARE_SIMPLE_TYPE(CXLHost, PXB_CXL_HOST) typedef struct CXLUpstreamPort CXLUpstreamPort; DECLARE_INSTANCE_CHECKER(CXLUpstreamPort, CXL_USP, TYPE_CXL_USP) CXLComponentState *cxl_usp_to_cstate(CXLUpstreamPort *usp); + +#define TYPE_CXL_DSP "cxl-downstream" + +typedef struct CXLDownstreamPort CXLDownstreamPort; +DECLARE_INSTANCE_CHECKER(CXLDownstreamPort, CXL_DSP, TYPE_CXL_DSP) + #endif diff --git a/include/hw/cxl/cxl_component.h b/include/hw/cxl/cxl_component.h index 3c795a6278..5227a8e833 100644 --- a/include/hw/cxl/cxl_component.h +++ b/include/hw/cxl/cxl_component.h @@ -26,7 +26,8 @@ enum reg_type { CXL2_LOGICAL_DEVICE, CXL2_ROOT_PORT, CXL2_UPSTREAM_PORT, - CXL2_DOWNSTREAM_PORT + CXL2_DOWNSTREAM_PORT, + CXL3_SWITCH_MAILBOX_CCI, }; /* @@ -175,7 +176,8 @@ HDM_DECODER_INIT(3); (CXL_IDE_REGISTERS_OFFSET + CXL_IDE_REGISTERS_SIZE) #define CXL_SNOOP_REGISTERS_SIZE 0x8 -QEMU_BUILD_BUG_MSG((CXL_SNOOP_REGISTERS_OFFSET + CXL_SNOOP_REGISTERS_SIZE) >= 0x1000, +QEMU_BUILD_BUG_MSG((CXL_SNOOP_REGISTERS_OFFSET + + CXL_SNOOP_REGISTERS_SIZE) >= 0x1000, "No space for registers"); typedef struct component_registers { diff --git a/include/hw/cxl/cxl_device.h b/include/hw/cxl/cxl_device.h index 51cd0d9ce3..61b7f897f7 100644 --- a/include/hw/cxl/cxl_device.h +++ b/include/hw/cxl/cxl_device.h @@ -111,6 +111,20 @@ typedef enum { CXL_MBOX_MAX = 0x17 } CXLRetCode; +typedef struct CXLCCI CXLCCI; +typedef struct cxl_device_state CXLDeviceState; +struct cxl_cmd; +typedef CXLRetCode (*opcode_handler)(const struct cxl_cmd *cmd, + uint8_t *payload_in, size_t len_in, + uint8_t *payload_out, size_t *len_out, + CXLCCI *cci); +struct cxl_cmd { + const char *name; + opcode_handler handler; + ssize_t in; + uint16_t effect; /* Reported in CEL */ +}; + typedef struct CXLEvent { CXLEventRecordRaw data; QSIMPLEQ_ENTRY(CXLEvent) node; @@ -127,6 +141,31 @@ typedef struct CXLEventLog { QSIMPLEQ_HEAD(, CXLEvent) events; } CXLEventLog; +typedef struct CXLCCI { + const struct cxl_cmd (*cxl_cmd_set)[256]; + struct cel_log { + uint16_t opcode; + uint16_t effect; + } cel_log[1 << 16]; + size_t cel_size; + + /* background command handling (times in ms) */ + struct { + uint16_t opcode; + uint16_t complete_pct; + uint16_t ret_code; /* Current value of retcode */ + uint64_t starttime; + /* set by each bg cmd, cleared by the bg_timer when complete */ + uint64_t runtime; + QEMUTimer *timer; + } bg; + size_t payload_max; + /* Pointer to device hosting the CCI */ + DeviceState *d; + /* Pointer to the device hosting the protocol conversion */ + DeviceState *intf; +} CXLCCI; + typedef struct cxl_device_state { MemoryRegion device_registers; @@ -154,17 +193,13 @@ typedef struct cxl_device_state { struct { MemoryRegion mailbox; uint16_t payload_size; + uint8_t mbox_msi_n; union { uint8_t mbox_reg_state[CXL_MAILBOX_REGISTERS_LENGTH]; uint16_t mbox_reg_state16[CXL_MAILBOX_REGISTERS_LENGTH / 2]; uint32_t mbox_reg_state32[CXL_MAILBOX_REGISTERS_LENGTH / 4]; uint64_t mbox_reg_state64[CXL_MAILBOX_REGISTERS_LENGTH / 8]; }; - struct cel_log { - uint16_t opcode; - uint16_t effect; - } cel_log[1 << 16]; - size_t cel_size; }; struct { @@ -178,21 +213,26 @@ typedef struct cxl_device_state { uint64_t pmem_size; uint64_t vmem_size; + const struct cxl_cmd (*cxl_cmd_set)[256]; CXLEventLog event_logs[CXL_EVENT_TYPE_MAX]; } CXLDeviceState; /* Initialize the register block for a device */ -void cxl_device_register_block_init(Object *obj, CXLDeviceState *dev); +void cxl_device_register_block_init(Object *obj, CXLDeviceState *dev, + CXLCCI *cci); +typedef struct CXLType3Dev CXLType3Dev; +typedef struct CSWMBCCIDev CSWMBCCIDev; /* Set up default values for the register block */ -void cxl_device_register_init_common(CXLDeviceState *dev); +void cxl_device_register_init_t3(CXLType3Dev *ct3d); +void cxl_device_register_init_swcci(CSWMBCCIDev *sw); /* * CXL 2.0 - 8.2.8.1 including errata F4 * Documented as a 128 bit register, but 64 bit accesses and the second * 64 bits are currently reserved. */ -REG64(CXL_DEV_CAP_ARRAY, 0) /* Documented as 128 bit register but 64 byte accesses */ +REG64(CXL_DEV_CAP_ARRAY, 0) FIELD(CXL_DEV_CAP_ARRAY, CAP_ID, 0, 16) FIELD(CXL_DEV_CAP_ARRAY, CAP_VERSION, 16, 8) FIELD(CXL_DEV_CAP_ARRAY, CAP_COUNT, 32, 16) @@ -231,8 +271,20 @@ CXL_DEVICE_CAPABILITY_HEADER_REGISTER(MEMORY_DEVICE, CXL_DEVICE_CAP_HDR1_OFFSET + CXL_DEVICE_CAP_REG_SIZE * 2) -void cxl_initialize_mailbox(CXLDeviceState *cxl_dstate); -void cxl_process_mailbox(CXLDeviceState *cxl_dstate); +void cxl_initialize_mailbox_t3(CXLCCI *cci, DeviceState *d, size_t payload_max); +void cxl_initialize_mailbox_swcci(CXLCCI *cci, DeviceState *intf, + DeviceState *d, size_t payload_max); +void cxl_init_cci(CXLCCI *cci, size_t payload_max); +int cxl_process_cci_message(CXLCCI *cci, uint8_t set, uint8_t cmd, + size_t len_in, uint8_t *pl_in, + size_t *len_out, uint8_t *pl_out, + bool *bg_started); +void cxl_initialize_t3_fm_owned_ld_mctpcci(CXLCCI *cci, DeviceState *d, + DeviceState *intf, + size_t payload_max); + +void cxl_initialize_t3_ld_cci(CXLCCI *cci, DeviceState *d, + DeviceState *intf, size_t payload_max); #define cxl_device_cap_init(dstate, reg, cap_id, ver) \ do { \ @@ -297,6 +349,23 @@ REG64(CXL_MEM_DEV_STS, 0) FIELD(CXL_MEM_DEV_STS, MBOX_READY, 4, 1) FIELD(CXL_MEM_DEV_STS, RESET_NEEDED, 5, 3) +static inline void __toggle_media(CXLDeviceState *cxl_dstate, int val) +{ + uint64_t dev_status_reg; + + dev_status_reg = FIELD_DP64(0, CXL_MEM_DEV_STS, MEDIA_STATUS, val); + cxl_dstate->mbox_reg_state64[R_CXL_MEM_DEV_STS] = dev_status_reg; +} +#define cxl_dev_disable_media(cxlds) \ + do { __toggle_media((cxlds), 0x3); } while (0) +#define cxl_dev_enable_media(cxlds) \ + do { __toggle_media((cxlds), 0x1); } while (0) + +static inline bool sanitize_running(CXLCCI *cci) +{ + return !!cci->bg.runtime && cci->bg.opcode == 0x4400; +} + typedef struct CXLError { QTAILQ_ENTRY(CXLError) node; int type; /* Error code as per FE definition */ @@ -333,6 +402,10 @@ struct CXLType3Dev { AddressSpace hostpmem_as; CXLComponentState cxl_cstate; CXLDeviceState cxl_dstate; + CXLCCI cci; /* Primary PCI mailbox CCI */ + /* Always intialized as no way to know if a VDM might show up */ + CXLCCI vdm_fm_owned_ld_mctp_cci; + CXLCCI ld0_cci; /* DOE */ DOECap doe_cdat; @@ -361,9 +434,21 @@ struct CXLType3Class { uint64_t offset); void (*set_lsa)(CXLType3Dev *ct3d, const void *buf, uint64_t size, uint64_t offset); - bool (*set_cacheline)(CXLType3Dev *ct3d, uint64_t dpa_offset, uint8_t *data); + bool (*set_cacheline)(CXLType3Dev *ct3d, uint64_t dpa_offset, + uint8_t *data); }; +struct CSWMBCCIDev { + PCIDevice parent_obj; + PCIDevice *target; + CXLComponentState cxl_cstate; + CXLDeviceState cxl_dstate; + CXLCCI *cci; +}; + +#define TYPE_CXL_SWITCH_MAILBOX_CCI "cxl-switch-mailbox-cci" +OBJECT_DECLARE_TYPE(CSWMBCCIDev, CSWMBCCIClass, CXL_SWITCH_MAILBOX_CCI) + MemTxResult cxl_type3_read(PCIDevice *d, hwaddr host_addr, uint64_t *data, unsigned size, MemTxAttrs attrs); MemTxResult cxl_type3_write(PCIDevice *d, hwaddr host_addr, uint64_t data, @@ -376,7 +461,7 @@ bool cxl_event_insert(CXLDeviceState *cxlds, CXLEventLogType log_type, CXLEventRecordRaw *event); CXLRetCode cxl_event_get_records(CXLDeviceState *cxlds, CXLGetEventPayload *pl, uint8_t log_type, int max_recs, - uint16_t *len); + size_t *len); CXLRetCode cxl_event_clear_records(CXLDeviceState *cxlds, CXLClearEventPayload *pl); diff --git a/include/hw/cxl/cxl_events.h b/include/hw/cxl/cxl_events.h index 089ba2091f..d778487b7e 100644 --- a/include/hw/cxl/cxl_events.h +++ b/include/hw/cxl/cxl_events.h @@ -92,7 +92,8 @@ typedef enum CXLEventIntMode { CXL_INT_RES = 0x03, } CXLEventIntMode; #define CXL_EVENT_INT_MODE_MASK 0x3 -#define CXL_EVENT_INT_SETTING(vector) ((((uint8_t)vector & 0xf) << 4) | CXL_INT_MSI_MSIX) +#define CXL_EVENT_INT_SETTING(vector) \ + ((((uint8_t)vector & 0xf) << 4) | CXL_INT_MSI_MSIX) typedef struct CXLEventInterruptPolicy { uint8_t info_settings; uint8_t warn_settings; diff --git a/include/hw/cxl/cxl_pci.h b/include/hw/cxl/cxl_pci.h index 407be95b9e..ddf01a543b 100644 --- a/include/hw/cxl/cxl_pci.h +++ b/include/hw/cxl/cxl_pci.h @@ -86,7 +86,7 @@ typedef struct CXLDVSECDevice { QEMU_BUILD_BUG_ON(sizeof(CXLDVSECDevice) != 0x38); /* CXL 2.0 - 8.1.5 (ID 0003) */ -typedef struct CXLDVSECPortExtensions { +typedef struct CXLDVSECPortExt { DVSECHeader hdr; uint16_t status; uint16_t control; @@ -100,8 +100,8 @@ typedef struct CXLDVSECPortExtensions { uint32_t alt_prefetch_limit_high; uint32_t rcrb_base; uint32_t rcrb_base_high; -} CXLDVSECPortExtensions; -QEMU_BUILD_BUG_ON(sizeof(CXLDVSECPortExtensions) != 0x28); +} CXLDVSECPortExt; +QEMU_BUILD_BUG_ON(sizeof(CXLDVSECPortExt) != 0x28); #define PORT_CONTROL_OFFSET 0xc #define PORT_CONTROL_UNMASK_SBR 1 diff --git a/include/hw/pci-bridge/cxl_upstream_port.h b/include/hw/pci-bridge/cxl_upstream_port.h new file mode 100644 index 0000000000..12635139f6 --- /dev/null +++ b/include/hw/pci-bridge/cxl_upstream_port.h @@ -0,0 +1,19 @@ + +#ifndef CXL_USP_H +#define CXL_USP_H +#include "hw/pci/pcie.h" +#include "hw/pci/pcie_port.h" +#include "hw/cxl/cxl.h" + +typedef struct CXLUpstreamPort { + /*< private >*/ + PCIEPort parent_obj; + + /*< public >*/ + CXLComponentState cxl_cstate; + CXLCCI swcci; + DOECap doe_cdat; + uint64_t sn; +} CXLUpstreamPort; + +#endif /* CXL_SUP_H */ diff --git a/include/hw/virtio/vhost-backend.h b/include/hw/virtio/vhost-backend.h index 96ccc18cd3..a86d103f82 100644 --- a/include/hw/virtio/vhost-backend.h +++ b/include/hw/virtio/vhost-backend.h @@ -26,6 +26,18 @@ typedef enum VhostSetConfigType { VHOST_SET_CONFIG_TYPE_MIGRATION = 1, } VhostSetConfigType; +typedef enum VhostDeviceStateDirection { + /* Transfer state from back-end (device) to front-end */ + VHOST_TRANSFER_STATE_DIRECTION_SAVE = 0, + /* Transfer state from front-end to back-end (device) */ + VHOST_TRANSFER_STATE_DIRECTION_LOAD = 1, +} VhostDeviceStateDirection; + +typedef enum VhostDeviceStatePhase { + /* The device (and all its vrings) is stopped */ + VHOST_TRANSFER_STATE_PHASE_STOPPED = 0, +} VhostDeviceStatePhase; + struct vhost_inflight; struct vhost_dev; struct vhost_log; @@ -129,6 +141,15 @@ typedef int (*vhost_set_config_call_op)(struct vhost_dev *dev, typedef void (*vhost_reset_status_op)(struct vhost_dev *dev); +typedef bool (*vhost_supports_device_state_op)(struct vhost_dev *dev); +typedef int (*vhost_set_device_state_fd_op)(struct vhost_dev *dev, + VhostDeviceStateDirection direction, + VhostDeviceStatePhase phase, + int fd, + int *reply_fd, + Error **errp); +typedef int (*vhost_check_device_state_op)(struct vhost_dev *dev, Error **errp); + typedef struct VhostOps { VhostBackendType backend_type; vhost_backend_init vhost_backend_init; @@ -176,6 +197,9 @@ typedef struct VhostOps { vhost_force_iommu_op vhost_force_iommu; vhost_set_config_call_op vhost_set_config_call; vhost_reset_status_op vhost_reset_status; + vhost_supports_device_state_op vhost_supports_device_state; + vhost_set_device_state_fd_op vhost_set_device_state_fd; + vhost_check_device_state_op vhost_check_device_state; } VhostOps; int vhost_backend_update_device_iotlb(struct vhost_dev *dev, diff --git a/include/hw/virtio/vhost-user.h b/include/hw/virtio/vhost-user.h index 20b69d8e85..d7c09ffd34 100644 --- a/include/hw/virtio/vhost-user.h +++ b/include/hw/virtio/vhost-user.h @@ -31,6 +31,7 @@ enum VhostUserProtocolFeature { VHOST_USER_PROTOCOL_F_STATUS = 16, /* Feature 17 reserved for VHOST_USER_PROTOCOL_F_XEN_MMAP. */ VHOST_USER_PROTOCOL_F_SHARED_OBJECT = 18, + VHOST_USER_PROTOCOL_F_DEVICE_STATE = 19, VHOST_USER_PROTOCOL_F_MAX }; diff --git a/include/hw/virtio/vhost.h b/include/hw/virtio/vhost.h index 5e8183f64a..05d7204a08 100644 --- a/include/hw/virtio/vhost.h +++ b/include/hw/virtio/vhost.h @@ -351,4 +351,117 @@ static inline int vhost_reset_device(struct vhost_dev *hdev) } #endif /* CONFIG_VHOST */ +/** + * vhost_supports_device_state(): Checks whether the back-end supports + * transferring internal device state for the purpose of migration. + * Support for this feature is required for vhost_set_device_state_fd() + * and vhost_check_device_state(). + * + * @dev: The vhost device + * + * Returns true if the device supports these commands, and false if it + * does not. + */ +bool vhost_supports_device_state(struct vhost_dev *dev); + +/** + * vhost_set_device_state_fd(): Begin transfer of internal state from/to + * the back-end for the purpose of migration. Data is to be transferred + * over a pipe according to @direction and @phase. The sending end must + * only write to the pipe, and the receiving end must only read from it. + * Once the sending end is done, it closes its FD. The receiving end + * must take this as the end-of-transfer signal and close its FD, too. + * + * @fd is the back-end's end of the pipe: The write FD for SAVE, and the + * read FD for LOAD. This function transfers ownership of @fd to the + * back-end, i.e. closes it in the front-end. + * + * The back-end may optionally reply with an FD of its own, if this + * improves efficiency on its end. In this case, the returned FD is + * stored in *reply_fd. The back-end will discard the FD sent to it, + * and the front-end must use *reply_fd for transferring state to/from + * the back-end. + * + * @dev: The vhost device + * @direction: The direction in which the state is to be transferred. + * For outgoing migrations, this is SAVE, and data is read + * from the back-end and stored by the front-end in the + * migration stream. + * For incoming migrations, this is LOAD, and data is read + * by the front-end from the migration stream and sent to + * the back-end to restore the saved state. + * @phase: Which migration phase we are in. Currently, there is only + * STOPPED (device and all vrings are stopped), in the future, + * more phases such as PRE_COPY or POST_COPY may be added. + * @fd: Back-end's end of the pipe through which to transfer state; note + * that ownership is transferred to the back-end, so this function + * closes @fd in the front-end. + * @reply_fd: If the back-end wishes to use a different pipe for state + * transfer, this will contain an FD for the front-end to + * use. Otherwise, -1 is stored here. + * @errp: Potential error description + * + * Returns 0 on success, and -errno on failure. + */ +int vhost_set_device_state_fd(struct vhost_dev *dev, + VhostDeviceStateDirection direction, + VhostDeviceStatePhase phase, + int fd, + int *reply_fd, + Error **errp); + +/** + * vhost_set_device_state_fd(): After transferring state from/to the + * back-end via vhost_set_device_state_fd(), i.e. once the sending end + * has closed the pipe, inquire the back-end to report any potential + * errors that have occurred on its side. This allows to sense errors + * like: + * - During outgoing migration, when the source side had already started + * to produce its state, something went wrong and it failed to finish + * - During incoming migration, when the received state is somehow + * invalid and cannot be processed by the back-end + * + * @dev: The vhost device + * @errp: Potential error description + * + * Returns 0 when the back-end reports successful state transfer and + * processing, and -errno when an error occurred somewhere. + */ +int vhost_check_device_state(struct vhost_dev *dev, Error **errp); + +/** + * vhost_save_backend_state(): High-level function to receive a vhost + * back-end's state, and save it in @f. Uses + * `vhost_set_device_state_fd()` to get the data from the back-end, and + * stores it in consecutive chunks that are each prefixed by their + * respective length (be32). The end is marked by a 0-length chunk. + * + * Must only be called while the device and all its vrings are stopped + * (`VHOST_TRANSFER_STATE_PHASE_STOPPED`). + * + * @dev: The vhost device from which to save the state + * @f: Migration stream in which to save the state + * @errp: Potential error message + * + * Returns 0 on success, and -errno otherwise. + */ +int vhost_save_backend_state(struct vhost_dev *dev, QEMUFile *f, Error **errp); + +/** + * vhost_load_backend_state(): High-level function to load a vhost + * back-end's state from @f, and send it over to the back-end. Reads + * the data from @f in the format used by `vhost_save_state()`, and uses + * `vhost_set_device_state_fd()` to transfer it to the back-end. + * + * Must only be called while the device and all its vrings are stopped + * (`VHOST_TRANSFER_STATE_PHASE_STOPPED`). + * + * @dev: The vhost device to which to send the sate + * @f: Migration stream from which to load the state + * @errp: Potential error message + * + * Returns 0 on success, and -errno otherwise. + */ +int vhost_load_backend_state(struct vhost_dev *dev, QEMUFile *f, Error **errp); + #endif |