diff options
Diffstat (limited to 'hw/virtio/vhost-user.c')
| -rw-r--r-- | hw/virtio/vhost-user.c | 638 |
1 files changed, 539 insertions, 99 deletions
diff --git a/hw/virtio/vhost-user.c b/hw/virtio/vhost-user.c index ec21e8fbe8..4d6cd4e58a 100644 --- a/hw/virtio/vhost-user.c +++ b/hw/virtio/vhost-user.c @@ -35,11 +35,29 @@ #include <linux/userfaultfd.h> #endif -#define VHOST_MEMORY_MAX_NREGIONS 8 +#define VHOST_MEMORY_BASELINE_NREGIONS 8 #define VHOST_USER_F_PROTOCOL_FEATURES 30 #define VHOST_USER_SLAVE_MAX_FDS 8 /* + * Set maximum number of RAM slots supported to + * the maximum number supported by the target + * hardware plaform. + */ +#if defined(TARGET_X86) || defined(TARGET_X86_64) || \ + defined(TARGET_ARM) || defined(TARGET_ARM_64) +#include "hw/acpi/acpi.h" +#define VHOST_USER_MAX_RAM_SLOTS ACPI_MAX_RAM_SLOTS + +#elif defined(TARGET_PPC) || defined(TARGET_PPC_64) +#include "hw/ppc/spapr.h" +#define VHOST_USER_MAX_RAM_SLOTS SPAPR_MAX_RAM_SLOTS + +#else +#define VHOST_USER_MAX_RAM_SLOTS 512 +#endif + +/* * Maximum size of virtio device config space */ #define VHOST_USER_MAX_CONFIG_SIZE 256 @@ -59,6 +77,8 @@ enum VhostUserProtocolFeature { VHOST_USER_PROTOCOL_F_HOST_NOTIFIER = 11, VHOST_USER_PROTOCOL_F_INFLIGHT_SHMFD = 12, VHOST_USER_PROTOCOL_F_RESET_DEVICE = 13, + /* Feature 14 reserved for VHOST_USER_PROTOCOL_F_INBAND_NOTIFICATIONS. */ + VHOST_USER_PROTOCOL_F_CONFIGURE_MEM_SLOTS = 15, VHOST_USER_PROTOCOL_F_MAX }; @@ -100,6 +120,10 @@ typedef enum VhostUserRequest { VHOST_USER_SET_INFLIGHT_FD = 32, VHOST_USER_GPU_SET_SOCKET = 33, VHOST_USER_RESET_DEVICE = 34, + /* Message number 35 reserved for VHOST_USER_VRING_KICK. */ + VHOST_USER_GET_MAX_MEM_SLOTS = 36, + VHOST_USER_ADD_MEM_REG = 37, + VHOST_USER_REM_MEM_REG = 38, VHOST_USER_MAX } VhostUserRequest; @@ -121,9 +145,14 @@ typedef struct VhostUserMemoryRegion { typedef struct VhostUserMemory { uint32_t nregions; uint32_t padding; - VhostUserMemoryRegion regions[VHOST_MEMORY_MAX_NREGIONS]; + VhostUserMemoryRegion regions[VHOST_MEMORY_BASELINE_NREGIONS]; } VhostUserMemory; +typedef struct VhostUserMemRegMsg { + uint32_t padding; + VhostUserMemoryRegion region; +} VhostUserMemRegMsg; + typedef struct VhostUserLog { uint64_t mmap_size; uint64_t mmap_offset; @@ -182,6 +211,7 @@ typedef union { struct vhost_vring_state state; struct vhost_vring_addr addr; VhostUserMemory memory; + VhostUserMemRegMsg mem_reg; VhostUserLog log; struct vhost_iotlb_msg iotlb; VhostUserConfig config; @@ -210,7 +240,7 @@ struct vhost_user { int slave_fd; NotifierWithReturn postcopy_notifier; struct PostCopyFD postcopy_fd; - uint64_t postcopy_client_bases[VHOST_MEMORY_MAX_NREGIONS]; + uint64_t postcopy_client_bases[VHOST_USER_MAX_RAM_SLOTS]; /* Length of the region_rb and region_rb_offset arrays */ size_t region_rb_len; /* RAMBlock associated with a given region */ @@ -222,6 +252,16 @@ struct vhost_user { /* True once we've entered postcopy_listen */ bool postcopy_listen; + + /* Our current regions */ + int num_shadow_regions; + struct vhost_memory_region shadow_regions[VHOST_USER_MAX_RAM_SLOTS]; +}; + +struct scrub_regions { + struct vhost_memory_region *region; + int reg_idx; + int fd_idx; }; static bool ioeventfd_enabled(void) @@ -370,7 +410,7 @@ int vhost_user_gpu_set_socket(struct vhost_dev *dev, int fd) static int vhost_user_set_log_base(struct vhost_dev *dev, uint64_t base, struct vhost_log *log) { - int fds[VHOST_MEMORY_MAX_NREGIONS]; + int fds[VHOST_USER_MAX_RAM_SLOTS]; size_t fd_num = 0; bool shmfd = virtio_has_feature(dev->protocol_features, VHOST_USER_PROTOCOL_F_LOG_SHMFD); @@ -407,6 +447,27 @@ static int vhost_user_set_log_base(struct vhost_dev *dev, uint64_t base, return 0; } +static MemoryRegion *vhost_user_get_mr_data(uint64_t addr, ram_addr_t *offset, + int *fd) +{ + MemoryRegion *mr; + + assert((uintptr_t)addr == addr); + mr = memory_region_from_host((void *)(uintptr_t)addr, offset); + *fd = memory_region_get_fd(mr); + + return mr; +} + +static void vhost_user_fill_msg_region(VhostUserMemoryRegion *dst, + struct vhost_memory_region *src) +{ + assert(src != NULL && dst != NULL); + dst->userspace_addr = src->userspace_addr; + dst->memory_size = src->memory_size; + dst->guest_phys_addr = src->guest_phys_addr; +} + static int vhost_user_fill_set_mem_table_msg(struct vhost_user *u, struct vhost_dev *dev, VhostUserMsg *msg, @@ -417,19 +478,17 @@ static int vhost_user_fill_set_mem_table_msg(struct vhost_user *u, ram_addr_t offset; MemoryRegion *mr; struct vhost_memory_region *reg; + VhostUserMemoryRegion region_buffer; msg->hdr.request = VHOST_USER_SET_MEM_TABLE; for (i = 0; i < dev->mem->nregions; ++i) { reg = dev->mem->regions + i; - assert((uintptr_t)reg->userspace_addr == reg->userspace_addr); - mr = memory_region_from_host((void *)(uintptr_t)reg->userspace_addr, - &offset); - fd = memory_region_get_fd(mr); + mr = vhost_user_get_mr_data(reg->userspace_addr, &offset, &fd); if (fd > 0) { if (track_ramblocks) { - assert(*fd_num < VHOST_MEMORY_MAX_NREGIONS); + assert(*fd_num < VHOST_MEMORY_BASELINE_NREGIONS); trace_vhost_user_set_mem_table_withfd(*fd_num, mr->name, reg->memory_size, reg->guest_phys_addr, @@ -437,16 +496,12 @@ static int vhost_user_fill_set_mem_table_msg(struct vhost_user *u, offset); u->region_rb_offset[i] = offset; u->region_rb[i] = mr->ram_block; - } else if (*fd_num == VHOST_MEMORY_MAX_NREGIONS) { + } else if (*fd_num == VHOST_MEMORY_BASELINE_NREGIONS) { error_report("Failed preparing vhost-user memory table msg"); return -1; } - msg->payload.memory.regions[*fd_num].userspace_addr = - reg->userspace_addr; - msg->payload.memory.regions[*fd_num].memory_size = - reg->memory_size; - msg->payload.memory.regions[*fd_num].guest_phys_addr = - reg->guest_phys_addr; + vhost_user_fill_msg_region(®ion_buffer, reg); + msg->payload.memory.regions[*fd_num] = region_buffer; msg->payload.memory.regions[*fd_num].mmap_offset = offset; fds[(*fd_num)++] = fd; } else if (track_ramblocks) { @@ -470,11 +525,335 @@ static int vhost_user_fill_set_mem_table_msg(struct vhost_user *u, return 1; } +static inline bool reg_equal(struct vhost_memory_region *shadow_reg, + struct vhost_memory_region *vdev_reg) +{ + return shadow_reg->guest_phys_addr == vdev_reg->guest_phys_addr && + shadow_reg->userspace_addr == vdev_reg->userspace_addr && + shadow_reg->memory_size == vdev_reg->memory_size; +} + +static void scrub_shadow_regions(struct vhost_dev *dev, + struct scrub_regions *add_reg, + int *nr_add_reg, + struct scrub_regions *rem_reg, + int *nr_rem_reg, uint64_t *shadow_pcb, + bool track_ramblocks) +{ + struct vhost_user *u = dev->opaque; + bool found[VHOST_USER_MAX_RAM_SLOTS] = {}; + struct vhost_memory_region *reg, *shadow_reg; + int i, j, fd, add_idx = 0, rm_idx = 0, fd_num = 0; + ram_addr_t offset; + MemoryRegion *mr; + bool matching; + + /* + * Find memory regions present in our shadow state which are not in + * the device's current memory state. + * + * Mark regions in both the shadow and device state as "found". + */ + for (i = 0; i < u->num_shadow_regions; i++) { + shadow_reg = &u->shadow_regions[i]; + matching = false; + + for (j = 0; j < dev->mem->nregions; j++) { + reg = &dev->mem->regions[j]; + + mr = vhost_user_get_mr_data(reg->userspace_addr, &offset, &fd); + + if (reg_equal(shadow_reg, reg)) { + matching = true; + found[j] = true; + if (track_ramblocks) { + /* + * Reset postcopy client bases, region_rb, and + * region_rb_offset in case regions are removed. + */ + if (fd > 0) { + u->region_rb_offset[j] = offset; + u->region_rb[j] = mr->ram_block; + shadow_pcb[j] = u->postcopy_client_bases[i]; + } else { + u->region_rb_offset[j] = 0; + u->region_rb[j] = NULL; + } + } + break; + } + } + + /* + * If the region was not found in the current device memory state + * create an entry for it in the removed list. + */ + if (!matching) { + rem_reg[rm_idx].region = shadow_reg; + rem_reg[rm_idx++].reg_idx = i; + } + } + + /* + * For regions not marked "found", create entries in the added list. + * + * Note their indexes in the device memory state and the indexes of their + * file descriptors. + */ + for (i = 0; i < dev->mem->nregions; i++) { + reg = &dev->mem->regions[i]; + mr = vhost_user_get_mr_data(reg->userspace_addr, &offset, &fd); + if (fd > 0) { + ++fd_num; + } + + /* + * If the region was in both the shadow and device state we don't + * need to send a VHOST_USER_ADD_MEM_REG message for it. + */ + if (found[i]) { + continue; + } + + add_reg[add_idx].region = reg; + add_reg[add_idx].reg_idx = i; + add_reg[add_idx++].fd_idx = fd_num; + } + *nr_rem_reg = rm_idx; + *nr_add_reg = add_idx; + + return; +} + +static int send_remove_regions(struct vhost_dev *dev, + struct scrub_regions *remove_reg, + int nr_rem_reg, VhostUserMsg *msg, + bool reply_supported) +{ + struct vhost_user *u = dev->opaque; + struct vhost_memory_region *shadow_reg; + int i, fd, shadow_reg_idx, ret; + ram_addr_t offset; + VhostUserMemoryRegion region_buffer; + + /* + * The regions in remove_reg appear in the same order they do in the + * shadow table. Therefore we can minimize memory copies by iterating + * through remove_reg backwards. + */ + for (i = nr_rem_reg - 1; i >= 0; i--) { + shadow_reg = remove_reg[i].region; + shadow_reg_idx = remove_reg[i].reg_idx; + + vhost_user_get_mr_data(shadow_reg->userspace_addr, &offset, &fd); + + if (fd > 0) { + msg->hdr.request = VHOST_USER_REM_MEM_REG; + vhost_user_fill_msg_region(®ion_buffer, shadow_reg); + msg->payload.mem_reg.region = region_buffer; + + if (vhost_user_write(dev, msg, &fd, 1) < 0) { + return -1; + } + + if (reply_supported) { + ret = process_message_reply(dev, msg); + if (ret) { + return ret; + } + } + } + + /* + * At this point we know the backend has unmapped the region. It is now + * safe to remove it from the shadow table. + */ + memmove(&u->shadow_regions[shadow_reg_idx], + &u->shadow_regions[shadow_reg_idx + 1], + sizeof(struct vhost_memory_region) * + (u->num_shadow_regions - shadow_reg_idx)); + u->num_shadow_regions--; + } + + return 0; +} + +static int send_add_regions(struct vhost_dev *dev, + struct scrub_regions *add_reg, int nr_add_reg, + VhostUserMsg *msg, uint64_t *shadow_pcb, + bool reply_supported, bool track_ramblocks) +{ + struct vhost_user *u = dev->opaque; + int i, fd, ret, reg_idx, reg_fd_idx; + struct vhost_memory_region *reg; + MemoryRegion *mr; + ram_addr_t offset; + VhostUserMsg msg_reply; + VhostUserMemoryRegion region_buffer; + + for (i = 0; i < nr_add_reg; i++) { + reg = add_reg[i].region; + reg_idx = add_reg[i].reg_idx; + reg_fd_idx = add_reg[i].fd_idx; + + mr = vhost_user_get_mr_data(reg->userspace_addr, &offset, &fd); + + if (fd > 0) { + if (track_ramblocks) { + trace_vhost_user_set_mem_table_withfd(reg_fd_idx, mr->name, + reg->memory_size, + reg->guest_phys_addr, + reg->userspace_addr, + offset); + u->region_rb_offset[reg_idx] = offset; + u->region_rb[reg_idx] = mr->ram_block; + } + msg->hdr.request = VHOST_USER_ADD_MEM_REG; + vhost_user_fill_msg_region(®ion_buffer, reg); + msg->payload.mem_reg.region = region_buffer; + msg->payload.mem_reg.region.mmap_offset = offset; + + if (vhost_user_write(dev, msg, &fd, 1) < 0) { + return -1; + } + + if (track_ramblocks) { + uint64_t reply_gpa; + + if (vhost_user_read(dev, &msg_reply) < 0) { + return -1; + } + + reply_gpa = msg_reply.payload.mem_reg.region.guest_phys_addr; + + if (msg_reply.hdr.request != VHOST_USER_ADD_MEM_REG) { + error_report("%s: Received unexpected msg type." + "Expected %d received %d", __func__, + VHOST_USER_ADD_MEM_REG, + msg_reply.hdr.request); + return -1; + } + + /* + * We're using the same structure, just reusing one of the + * fields, so it should be the same size. + */ + if (msg_reply.hdr.size != msg->hdr.size) { + error_report("%s: Unexpected size for postcopy reply " + "%d vs %d", __func__, msg_reply.hdr.size, + msg->hdr.size); + return -1; + } + + /* Get the postcopy client base from the backend's reply. */ + if (reply_gpa == dev->mem->regions[reg_idx].guest_phys_addr) { + shadow_pcb[reg_idx] = + msg_reply.payload.mem_reg.region.userspace_addr; + trace_vhost_user_set_mem_table_postcopy( + msg_reply.payload.mem_reg.region.userspace_addr, + msg->payload.mem_reg.region.userspace_addr, + reg_fd_idx, reg_idx); + } else { + error_report("%s: invalid postcopy reply for region. " + "Got guest physical address %" PRIX64 ", expected " + "%" PRIX64, __func__, reply_gpa, + dev->mem->regions[reg_idx].guest_phys_addr); + return -1; + } + } else if (reply_supported) { + ret = process_message_reply(dev, msg); + if (ret) { + return ret; + } + } + } else if (track_ramblocks) { + u->region_rb_offset[reg_idx] = 0; + u->region_rb[reg_idx] = NULL; + } + + /* + * At this point, we know the backend has mapped in the new + * region, if the region has a valid file descriptor. + * + * The region should now be added to the shadow table. + */ + u->shadow_regions[u->num_shadow_regions].guest_phys_addr = + reg->guest_phys_addr; + u->shadow_regions[u->num_shadow_regions].userspace_addr = + reg->userspace_addr; + u->shadow_regions[u->num_shadow_regions].memory_size = + reg->memory_size; + u->num_shadow_regions++; + } + + return 0; +} + +static int vhost_user_add_remove_regions(struct vhost_dev *dev, + VhostUserMsg *msg, + bool reply_supported, + bool track_ramblocks) +{ + struct vhost_user *u = dev->opaque; + struct scrub_regions add_reg[VHOST_USER_MAX_RAM_SLOTS]; + struct scrub_regions rem_reg[VHOST_USER_MAX_RAM_SLOTS]; + uint64_t shadow_pcb[VHOST_USER_MAX_RAM_SLOTS] = {}; + int nr_add_reg, nr_rem_reg; + + msg->hdr.size = sizeof(msg->payload.mem_reg.padding) + + sizeof(VhostUserMemoryRegion); + + /* Find the regions which need to be removed or added. */ + scrub_shadow_regions(dev, add_reg, &nr_add_reg, rem_reg, &nr_rem_reg, + shadow_pcb, track_ramblocks); + + if (nr_rem_reg && send_remove_regions(dev, rem_reg, nr_rem_reg, msg, + reply_supported) < 0) + { + goto err; + } + + if (nr_add_reg && send_add_regions(dev, add_reg, nr_add_reg, msg, + shadow_pcb, reply_supported, track_ramblocks) < 0) + { + goto err; + } + + if (track_ramblocks) { + memcpy(u->postcopy_client_bases, shadow_pcb, + sizeof(uint64_t) * VHOST_USER_MAX_RAM_SLOTS); + /* + * Now we've registered this with the postcopy code, we ack to the + * client, because now we're in the position to be able to deal with + * any faults it generates. + */ + /* TODO: Use this for failure cases as well with a bad value. */ + msg->hdr.size = sizeof(msg->payload.u64); + msg->payload.u64 = 0; /* OK */ + + if (vhost_user_write(dev, msg, NULL, 0) < 0) { + return -1; + } + } + + return 0; + +err: + if (track_ramblocks) { + memcpy(u->postcopy_client_bases, shadow_pcb, + sizeof(uint64_t) * VHOST_USER_MAX_RAM_SLOTS); + } + + return -1; +} + static int vhost_user_set_mem_table_postcopy(struct vhost_dev *dev, - struct vhost_memory *mem) + struct vhost_memory *mem, + bool reply_supported, + bool config_mem_slots) { struct vhost_user *u = dev->opaque; - int fds[VHOST_MEMORY_MAX_NREGIONS]; + int fds[VHOST_MEMORY_BASELINE_NREGIONS]; size_t fd_num = 0; VhostUserMsg msg_reply; int region_i, msg_i; @@ -494,71 +873,84 @@ static int vhost_user_set_mem_table_postcopy(struct vhost_dev *dev, u->region_rb_len = dev->mem->nregions; } - if (vhost_user_fill_set_mem_table_msg(u, dev, &msg, fds, &fd_num, + if (config_mem_slots) { + if (vhost_user_add_remove_regions(dev, &msg, reply_supported, true) < 0) { - return -1; - } + return -1; + } + } else { + if (vhost_user_fill_set_mem_table_msg(u, dev, &msg, fds, &fd_num, + true) < 0) { + return -1; + } - if (vhost_user_write(dev, &msg, fds, fd_num) < 0) { - return -1; - } + if (vhost_user_write(dev, &msg, fds, fd_num) < 0) { + return -1; + } - if (vhost_user_read(dev, &msg_reply) < 0) { - return -1; - } + if (vhost_user_read(dev, &msg_reply) < 0) { + return -1; + } - if (msg_reply.hdr.request != VHOST_USER_SET_MEM_TABLE) { - error_report("%s: Received unexpected msg type." - "Expected %d received %d", __func__, - VHOST_USER_SET_MEM_TABLE, msg_reply.hdr.request); - return -1; - } - /* We're using the same structure, just reusing one of the - * fields, so it should be the same size. - */ - if (msg_reply.hdr.size != msg.hdr.size) { - error_report("%s: Unexpected size for postcopy reply " - "%d vs %d", __func__, msg_reply.hdr.size, msg.hdr.size); - return -1; - } - - memset(u->postcopy_client_bases, 0, - sizeof(uint64_t) * VHOST_MEMORY_MAX_NREGIONS); - - /* They're in the same order as the regions that were sent - * but some of the regions were skipped (above) if they - * didn't have fd's - */ - for (msg_i = 0, region_i = 0; - region_i < dev->mem->nregions; - region_i++) { - if (msg_i < fd_num && - msg_reply.payload.memory.regions[msg_i].guest_phys_addr == - dev->mem->regions[region_i].guest_phys_addr) { - u->postcopy_client_bases[region_i] = - msg_reply.payload.memory.regions[msg_i].userspace_addr; - trace_vhost_user_set_mem_table_postcopy( - msg_reply.payload.memory.regions[msg_i].userspace_addr, - msg.payload.memory.regions[msg_i].userspace_addr, - msg_i, region_i); - msg_i++; + if (msg_reply.hdr.request != VHOST_USER_SET_MEM_TABLE) { + error_report("%s: Received unexpected msg type." + "Expected %d received %d", __func__, + VHOST_USER_SET_MEM_TABLE, msg_reply.hdr.request); + return -1; + } + + /* + * We're using the same structure, just reusing one of the + * fields, so it should be the same size. + */ + if (msg_reply.hdr.size != msg.hdr.size) { + error_report("%s: Unexpected size for postcopy reply " + "%d vs %d", __func__, msg_reply.hdr.size, + msg.hdr.size); + return -1; + } + + memset(u->postcopy_client_bases, 0, + sizeof(uint64_t) * VHOST_USER_MAX_RAM_SLOTS); + + /* + * They're in the same order as the regions that were sent + * but some of the regions were skipped (above) if they + * didn't have fd's + */ + for (msg_i = 0, region_i = 0; + region_i < dev->mem->nregions; + region_i++) { + if (msg_i < fd_num && + msg_reply.payload.memory.regions[msg_i].guest_phys_addr == + dev->mem->regions[region_i].guest_phys_addr) { + u->postcopy_client_bases[region_i] = + msg_reply.payload.memory.regions[msg_i].userspace_addr; + trace_vhost_user_set_mem_table_postcopy( + msg_reply.payload.memory.regions[msg_i].userspace_addr, + msg.payload.memory.regions[msg_i].userspace_addr, + msg_i, region_i); + msg_i++; + } + } + if (msg_i != fd_num) { + error_report("%s: postcopy reply not fully consumed " + "%d vs %zd", + __func__, msg_i, fd_num); + return -1; + } + + /* + * Now we've registered this with the postcopy code, we ack to the + * client, because now we're in the position to be able to deal + * with any faults it generates. + */ + /* TODO: Use this for failure cases as well with a bad value. */ + msg.hdr.size = sizeof(msg.payload.u64); + msg.payload.u64 = 0; /* OK */ + if (vhost_user_write(dev, &msg, NULL, 0) < 0) { + return -1; } - } - if (msg_i != fd_num) { - error_report("%s: postcopy reply not fully consumed " - "%d vs %zd", - __func__, msg_i, fd_num); - return -1; - } - /* Now we've registered this with the postcopy code, we ack to the client, - * because now we're in the position to be able to deal with any faults - * it generates. - */ - /* TODO: Use this for failure cases as well with a bad value */ - msg.hdr.size = sizeof(msg.payload.u64); - msg.payload.u64 = 0; /* OK */ - if (vhost_user_write(dev, &msg, NULL, 0) < 0) { - return -1; } return 0; @@ -568,17 +960,22 @@ static int vhost_user_set_mem_table(struct vhost_dev *dev, struct vhost_memory *mem) { struct vhost_user *u = dev->opaque; - int fds[VHOST_MEMORY_MAX_NREGIONS]; + int fds[VHOST_MEMORY_BASELINE_NREGIONS]; size_t fd_num = 0; bool do_postcopy = u->postcopy_listen && u->postcopy_fd.handler; bool reply_supported = virtio_has_feature(dev->protocol_features, VHOST_USER_PROTOCOL_F_REPLY_ACK); + bool config_mem_slots = + virtio_has_feature(dev->protocol_features, + VHOST_USER_PROTOCOL_F_CONFIGURE_MEM_SLOTS); if (do_postcopy) { - /* Postcopy has enough differences that it's best done in it's own + /* + * Postcopy has enough differences that it's best done in it's own * version */ - return vhost_user_set_mem_table_postcopy(dev, mem); + return vhost_user_set_mem_table_postcopy(dev, mem, reply_supported, + config_mem_slots); } VhostUserMsg msg = { @@ -589,17 +986,23 @@ static int vhost_user_set_mem_table(struct vhost_dev *dev, msg.hdr.flags |= VHOST_USER_NEED_REPLY_MASK; } - if (vhost_user_fill_set_mem_table_msg(u, dev, &msg, fds, &fd_num, + if (config_mem_slots) { + if (vhost_user_add_remove_regions(dev, &msg, reply_supported, false) < 0) { - return -1; - } - - if (vhost_user_write(dev, &msg, fds, fd_num) < 0) { - return -1; - } + return -1; + } + } else { + if (vhost_user_fill_set_mem_table_msg(u, dev, &msg, fds, &fd_num, + false) < 0) { + return -1; + } + if (vhost_user_write(dev, &msg, fds, fd_num) < 0) { + return -1; + } - if (reply_supported) { - return process_message_reply(dev, &msg); + if (reply_supported) { + return process_message_reply(dev, &msg); + } } return 0; @@ -764,7 +1167,7 @@ static int vhost_set_vring_file(struct vhost_dev *dev, VhostUserRequest request, struct vhost_vring_file *file) { - int fds[VHOST_MEMORY_MAX_NREGIONS]; + int fds[VHOST_USER_MAX_RAM_SLOTS]; size_t fd_num = 0; VhostUserMsg msg = { .hdr.request = request, @@ -880,6 +1283,23 @@ static int vhost_user_set_owner(struct vhost_dev *dev) return 0; } +static int vhost_user_get_max_memslots(struct vhost_dev *dev, + uint64_t *max_memslots) +{ + uint64_t backend_max_memslots; + int err; + + err = vhost_user_get_u64(dev, VHOST_USER_GET_MAX_MEM_SLOTS, + &backend_max_memslots); + if (err < 0) { + return err; + } + + *max_memslots = backend_max_memslots; + + return 0; +} + static int vhost_user_reset_device(struct vhost_dev *dev) { VhostUserMsg msg = { @@ -1377,7 +1797,7 @@ static int vhost_user_postcopy_notifier(NotifierWithReturn *notifier, static int vhost_user_backend_init(struct vhost_dev *dev, void *opaque) { - uint64_t features, protocol_features; + uint64_t features, protocol_features, ram_slots; struct vhost_user *u; int err; @@ -1439,6 +1859,27 @@ static int vhost_user_backend_init(struct vhost_dev *dev, void *opaque) "slave-req protocol features."); return -1; } + + /* get max memory regions if backend supports configurable RAM slots */ + if (!virtio_has_feature(dev->protocol_features, + VHOST_USER_PROTOCOL_F_CONFIGURE_MEM_SLOTS)) { + u->user->memory_slots = VHOST_MEMORY_BASELINE_NREGIONS; + } else { + err = vhost_user_get_max_memslots(dev, &ram_slots); + if (err < 0) { + return err; + } + + if (ram_slots < u->user->memory_slots) { + error_report("The backend specified a max ram slots limit " + "of %" PRIu64", when the prior validated limit was %d. " + "This limit should never decrease.", ram_slots, + u->user->memory_slots); + return -1; + } + + u->user->memory_slots = MIN(ram_slots, VHOST_USER_MAX_RAM_SLOTS); + } } if (dev->migration_blocker == NULL && @@ -1504,7 +1945,9 @@ static int vhost_user_get_vq_index(struct vhost_dev *dev, int idx) static int vhost_user_memslots_limit(struct vhost_dev *dev) { - return VHOST_MEMORY_MAX_NREGIONS; + struct vhost_user *u = dev->opaque; + + return u->user->memory_slots; } static bool vhost_user_requires_shm_log(struct vhost_dev *dev) @@ -1545,13 +1988,9 @@ static bool vhost_user_can_merge(struct vhost_dev *dev, { ram_addr_t offset; int mfd, rfd; - MemoryRegion *mr; - - mr = memory_region_from_host((void *)(uintptr_t)start1, &offset); - mfd = memory_region_get_fd(mr); - mr = memory_region_from_host((void *)(uintptr_t)start2, &offset); - rfd = memory_region_get_fd(mr); + (void)vhost_user_get_mr_data(start1, &offset, &mfd); + (void)vhost_user_get_mr_data(start2, &offset, &rfd); return mfd == rfd; } @@ -1893,6 +2332,7 @@ bool vhost_user_init(VhostUserState *user, CharBackend *chr, Error **errp) return false; } user->chr = chr; + user->memory_slots = 0; return true; } |