diff options
Diffstat (limited to 'drivers/block/virtio_blk.c')
| -rw-r--r-- | drivers/block/virtio_blk.c | 866 |
1 files changed, 642 insertions, 224 deletions
diff --git a/drivers/block/virtio_blk.c b/drivers/block/virtio_blk.c index 6a77fa917428..357434bdae99 100644 --- a/drivers/block/virtio_blk.c +++ b/drivers/block/virtio_blk.c @@ -13,8 +13,8 @@ #include <linux/string_helpers.h> #include <linux/idr.h> #include <linux/blk-mq.h> -#include <linux/blk-mq-virtio.h> #include <linux/numa.h> +#include <linux/vmalloc.h> #include <uapi/linux/virtio_ring.h> #define PART_BITS 4 @@ -80,22 +80,49 @@ struct virtio_blk { int num_vqs; int io_queues[HCTX_MAX_TYPES]; struct virtio_blk_vq *vqs; + + /* For zoned device */ + unsigned int zone_sectors; }; struct virtblk_req { + /* Out header */ struct virtio_blk_outhdr out_hdr; - u8 status; + + /* In header */ + union { + u8 status; + + /* + * The zone append command has an extended in header. + * The status field in zone_append_in_hdr must always + * be the last byte. + */ + struct { + __virtio64 sector; + u8 status; + } zone_append; + } in_hdr; + + size_t in_hdr_len; + struct sg_table sg_table; struct scatterlist sg[]; }; -static inline blk_status_t virtblk_result(struct virtblk_req *vbr) +static inline blk_status_t virtblk_result(u8 status) { - switch (vbr->status) { + switch (status) { case VIRTIO_BLK_S_OK: return BLK_STS_OK; case VIRTIO_BLK_S_UNSUPP: return BLK_STS_NOTSUPP; + case VIRTIO_BLK_S_ZONE_OPEN_RESOURCE: + return BLK_STS_ZONE_OPEN_RESOURCE; + case VIRTIO_BLK_S_ZONE_ACTIVE_RESOURCE: + return BLK_STS_ZONE_ACTIVE_RESOURCE; + case VIRTIO_BLK_S_IOERR: + case VIRTIO_BLK_S_ZONE_UNALIGNED_WP: default: return BLK_STS_IOERR; } @@ -111,11 +138,11 @@ static inline struct virtio_blk_vq *get_virtio_blk_vq(struct blk_mq_hw_ctx *hctx static int virtblk_add_req(struct virtqueue *vq, struct virtblk_req *vbr) { - struct scatterlist hdr, status, *sgs[3]; + struct scatterlist out_hdr, in_hdr, *sgs[3]; unsigned int num_out = 0, num_in = 0; - sg_init_one(&hdr, &vbr->out_hdr, sizeof(vbr->out_hdr)); - sgs[num_out++] = &hdr; + sg_init_one(&out_hdr, &vbr->out_hdr, sizeof(vbr->out_hdr)); + sgs[num_out++] = &out_hdr; if (vbr->sg_table.nents) { if (vbr->out_hdr.type & cpu_to_virtio32(vq->vdev, VIRTIO_BLK_T_OUT)) @@ -124,8 +151,8 @@ static int virtblk_add_req(struct virtqueue *vq, struct virtblk_req *vbr) sgs[num_out + num_in++] = vbr->sg_table.sgl; } - sg_init_one(&status, &vbr->status, sizeof(vbr->status)); - sgs[num_out + num_in++] = &status; + sg_init_one(&in_hdr, &vbr->in_hdr.status, vbr->in_hdr_len); + sgs[num_out + num_in++] = &in_hdr; return virtqueue_add_sgs(vq, sgs, num_out, num_in, vbr, GFP_ATOMIC); } @@ -170,9 +197,7 @@ static int virtblk_setup_discard_write_zeroes_erase(struct request *req, bool un WARN_ON_ONCE(n != segments); - req->special_vec.bv_page = virt_to_page(range); - req->special_vec.bv_offset = offset_in_page(range); - req->special_vec.bv_len = sizeof(*range) * segments; + bvec_set_virt(&req->special_vec, range, sizeof(*range) * segments); req->rq_flags |= RQF_SPECIAL_PAYLOAD; return 0; @@ -201,7 +226,7 @@ static int virtblk_map_data(struct blk_mq_hw_ctx *hctx, struct request *req, if (unlikely(err)) return -ENOMEM; - return blk_rq_map_sg(hctx->queue, req, vbr->sg_table.sgl); + return blk_rq_map_sg(req, vbr->sg_table.sgl); } static void virtblk_cleanup_cmd(struct request *req) @@ -214,21 +239,25 @@ static blk_status_t virtblk_setup_cmd(struct virtio_device *vdev, struct request *req, struct virtblk_req *vbr) { + size_t in_hdr_len = sizeof(vbr->in_hdr.status); bool unmap = false; u32 type; + u64 sector = 0; - vbr->out_hdr.sector = 0; + if (!IS_ENABLED(CONFIG_BLK_DEV_ZONED) && op_is_zone_mgmt(req_op(req))) + return BLK_STS_NOTSUPP; + + /* Set fields for all request types */ + vbr->out_hdr.ioprio = cpu_to_virtio32(vdev, req_get_ioprio(req)); switch (req_op(req)) { case REQ_OP_READ: type = VIRTIO_BLK_T_IN; - vbr->out_hdr.sector = cpu_to_virtio64(vdev, - blk_rq_pos(req)); + sector = blk_rq_pos(req); break; case REQ_OP_WRITE: type = VIRTIO_BLK_T_OUT; - vbr->out_hdr.sector = cpu_to_virtio64(vdev, - blk_rq_pos(req)); + sector = blk_rq_pos(req); break; case REQ_OP_FLUSH: type = VIRTIO_BLK_T_FLUSH; @@ -243,16 +272,45 @@ static blk_status_t virtblk_setup_cmd(struct virtio_device *vdev, case REQ_OP_SECURE_ERASE: type = VIRTIO_BLK_T_SECURE_ERASE; break; - case REQ_OP_DRV_IN: - type = VIRTIO_BLK_T_GET_ID; + case REQ_OP_ZONE_OPEN: + type = VIRTIO_BLK_T_ZONE_OPEN; + sector = blk_rq_pos(req); break; + case REQ_OP_ZONE_CLOSE: + type = VIRTIO_BLK_T_ZONE_CLOSE; + sector = blk_rq_pos(req); + break; + case REQ_OP_ZONE_FINISH: + type = VIRTIO_BLK_T_ZONE_FINISH; + sector = blk_rq_pos(req); + break; + case REQ_OP_ZONE_APPEND: + type = VIRTIO_BLK_T_ZONE_APPEND; + sector = blk_rq_pos(req); + in_hdr_len = sizeof(vbr->in_hdr.zone_append); + break; + case REQ_OP_ZONE_RESET: + type = VIRTIO_BLK_T_ZONE_RESET; + sector = blk_rq_pos(req); + break; + case REQ_OP_ZONE_RESET_ALL: + type = VIRTIO_BLK_T_ZONE_RESET_ALL; + break; + case REQ_OP_DRV_IN: + /* + * Out header has already been prepared by the caller (virtblk_get_id() + * or virtblk_submit_zone_report()), nothing to do here. + */ + return 0; default: WARN_ON_ONCE(1); return BLK_STS_IOERR; } + /* Set fields for non-REQ_OP_DRV_IN request types */ + vbr->in_hdr_len = in_hdr_len; vbr->out_hdr.type = cpu_to_virtio32(vdev, type); - vbr->out_hdr.ioprio = cpu_to_virtio32(vdev, req_get_ioprio(req)); + vbr->out_hdr.sector = cpu_to_virtio64(vdev, sector); if (type == VIRTIO_BLK_T_DISCARD || type == VIRTIO_BLK_T_WRITE_ZEROES || type == VIRTIO_BLK_T_SECURE_ERASE) { @@ -263,13 +321,30 @@ static blk_status_t virtblk_setup_cmd(struct virtio_device *vdev, return 0; } +/* + * The status byte is always the last byte of the virtblk request + * in-header. This helper fetches its value for all in-header formats + * that are currently defined. + */ +static inline u8 virtblk_vbr_status(struct virtblk_req *vbr) +{ + return *((u8 *)&vbr->in_hdr + vbr->in_hdr_len - 1); +} + static inline void virtblk_request_done(struct request *req) { struct virtblk_req *vbr = blk_mq_rq_to_pdu(req); + blk_status_t status = virtblk_result(virtblk_vbr_status(vbr)); + struct virtio_blk *vblk = req->mq_hctx->queue->queuedata; virtblk_unmap_data(req, vbr); virtblk_cleanup_cmd(req); - blk_mq_end_request(req, virtblk_result(vbr)); + + if (req_op(req) == REQ_OP_ZONE_APPEND) + req->__sector = virtio64_to_cpu(vblk->vdev, + vbr->in_hdr.zone_append.sector); + + blk_mq_end_request(req, status); } static void virtblk_done(struct virtqueue *vq) @@ -291,8 +366,6 @@ static void virtblk_done(struct virtqueue *vq) blk_mq_complete_request(req); req_done = true; } - if (unlikely(virtqueue_is_broken(vq))) - break; } while (!virtqueue_enable_cb(vq)); /* In case queue is stopped waiting for more buffers. */ @@ -394,23 +467,21 @@ static bool virtblk_prep_rq_batch(struct request *req) struct virtio_blk *vblk = req->mq_hctx->queue->queuedata; struct virtblk_req *vbr = blk_mq_rq_to_pdu(req); - req->mq_hctx->tags->rqs[req->tag] = req; - return virtblk_prep_rq(req->mq_hctx, vblk, req, vbr) == BLK_STS_OK; } -static bool virtblk_add_req_batch(struct virtio_blk_vq *vq, - struct request **rqlist) +static void virtblk_add_req_batch(struct virtio_blk_vq *vq, + struct rq_list *rqlist) { + struct request *req; unsigned long flags; - int err; bool kick; spin_lock_irqsave(&vq->lock, flags); - while (!rq_list_empty(*rqlist)) { - struct request *req = rq_list_pop(rqlist); + while ((req = rq_list_pop(rqlist))) { struct virtblk_req *vbr = blk_mq_rq_to_pdu(req); + int err; err = virtblk_add_req(vq->vq, vbr); if (err) { @@ -423,40 +494,311 @@ static bool virtblk_add_req_batch(struct virtio_blk_vq *vq, kick = virtqueue_kick_prepare(vq->vq); spin_unlock_irqrestore(&vq->lock, flags); - return kick; + if (kick) + virtqueue_notify(vq->vq); } -static void virtio_queue_rqs(struct request **rqlist) +static void virtio_queue_rqs(struct rq_list *rqlist) { - struct request *req, *next, *prev = NULL; - struct request *requeue_list = NULL; - - rq_list_for_each_safe(rqlist, req, next) { - struct virtio_blk_vq *vq = get_virtio_blk_vq(req->mq_hctx); - bool kick; - - if (!virtblk_prep_rq_batch(req)) { - rq_list_move(rqlist, &requeue_list, req, prev); - req = prev; - if (!req) - continue; - } + struct rq_list submit_list = { }; + struct rq_list requeue_list = { }; + struct virtio_blk_vq *vq = NULL; + struct request *req; + + while ((req = rq_list_pop(rqlist))) { + struct virtio_blk_vq *this_vq = get_virtio_blk_vq(req->mq_hctx); - if (!next || req->mq_hctx != next->mq_hctx) { - req->rq_next = NULL; - kick = virtblk_add_req_batch(vq, rqlist); - if (kick) - virtqueue_notify(vq->vq); + if (vq && vq != this_vq) + virtblk_add_req_batch(vq, &submit_list); + vq = this_vq; - *rqlist = next; - prev = NULL; - } else - prev = req; + if (virtblk_prep_rq_batch(req)) + rq_list_add_tail(&submit_list, req); + else + rq_list_add_tail(&requeue_list, req); } + if (vq) + virtblk_add_req_batch(vq, &submit_list); *rqlist = requeue_list; } +#ifdef CONFIG_BLK_DEV_ZONED +static void *virtblk_alloc_report_buffer(struct virtio_blk *vblk, + unsigned int nr_zones, + size_t *buflen) +{ + struct request_queue *q = vblk->disk->queue; + size_t bufsize; + void *buf; + + nr_zones = min_t(unsigned int, nr_zones, + get_capacity(vblk->disk) >> ilog2(vblk->zone_sectors)); + + bufsize = sizeof(struct virtio_blk_zone_report) + + nr_zones * sizeof(struct virtio_blk_zone_descriptor); + bufsize = min_t(size_t, bufsize, + queue_max_hw_sectors(q) << SECTOR_SHIFT); + bufsize = min_t(size_t, bufsize, queue_max_segments(q) << PAGE_SHIFT); + + while (bufsize >= sizeof(struct virtio_blk_zone_report)) { + buf = __vmalloc(bufsize, GFP_KERNEL | __GFP_NORETRY); + if (buf) { + *buflen = bufsize; + return buf; + } + bufsize >>= 1; + } + + return NULL; +} + +static int virtblk_submit_zone_report(struct virtio_blk *vblk, + char *report_buf, size_t report_len, + sector_t sector) +{ + struct request_queue *q = vblk->disk->queue; + struct request *req; + struct virtblk_req *vbr; + int err; + + req = blk_mq_alloc_request(q, REQ_OP_DRV_IN, 0); + if (IS_ERR(req)) + return PTR_ERR(req); + + vbr = blk_mq_rq_to_pdu(req); + vbr->in_hdr_len = sizeof(vbr->in_hdr.status); + vbr->out_hdr.type = cpu_to_virtio32(vblk->vdev, VIRTIO_BLK_T_ZONE_REPORT); + vbr->out_hdr.sector = cpu_to_virtio64(vblk->vdev, sector); + + err = blk_rq_map_kern(req, report_buf, report_len, GFP_KERNEL); + if (err) + goto out; + + blk_execute_rq(req, false); + err = blk_status_to_errno(virtblk_result(vbr->in_hdr.status)); +out: + blk_mq_free_request(req); + return err; +} + +static int virtblk_parse_zone(struct virtio_blk *vblk, + struct virtio_blk_zone_descriptor *entry, + unsigned int idx, + struct blk_report_zones_args *args) +{ + struct blk_zone zone = { }; + + zone.start = virtio64_to_cpu(vblk->vdev, entry->z_start); + if (zone.start + vblk->zone_sectors <= get_capacity(vblk->disk)) + zone.len = vblk->zone_sectors; + else + zone.len = get_capacity(vblk->disk) - zone.start; + zone.capacity = virtio64_to_cpu(vblk->vdev, entry->z_cap); + zone.wp = virtio64_to_cpu(vblk->vdev, entry->z_wp); + + switch (entry->z_type) { + case VIRTIO_BLK_ZT_SWR: + zone.type = BLK_ZONE_TYPE_SEQWRITE_REQ; + break; + case VIRTIO_BLK_ZT_SWP: + zone.type = BLK_ZONE_TYPE_SEQWRITE_PREF; + break; + case VIRTIO_BLK_ZT_CONV: + zone.type = BLK_ZONE_TYPE_CONVENTIONAL; + break; + default: + dev_err(&vblk->vdev->dev, "zone %llu: invalid type %#x\n", + zone.start, entry->z_type); + return -EIO; + } + + switch (entry->z_state) { + case VIRTIO_BLK_ZS_EMPTY: + zone.cond = BLK_ZONE_COND_EMPTY; + break; + case VIRTIO_BLK_ZS_CLOSED: + zone.cond = BLK_ZONE_COND_CLOSED; + break; + case VIRTIO_BLK_ZS_FULL: + zone.cond = BLK_ZONE_COND_FULL; + zone.wp = zone.start + zone.len; + break; + case VIRTIO_BLK_ZS_EOPEN: + zone.cond = BLK_ZONE_COND_EXP_OPEN; + break; + case VIRTIO_BLK_ZS_IOPEN: + zone.cond = BLK_ZONE_COND_IMP_OPEN; + break; + case VIRTIO_BLK_ZS_NOT_WP: + zone.cond = BLK_ZONE_COND_NOT_WP; + break; + case VIRTIO_BLK_ZS_RDONLY: + zone.cond = BLK_ZONE_COND_READONLY; + zone.wp = ULONG_MAX; + break; + case VIRTIO_BLK_ZS_OFFLINE: + zone.cond = BLK_ZONE_COND_OFFLINE; + zone.wp = ULONG_MAX; + break; + default: + dev_err(&vblk->vdev->dev, "zone %llu: invalid condition %#x\n", + zone.start, entry->z_state); + return -EIO; + } + + /* + * The callback below checks the validity of the reported + * entry data, no need to further validate it here. + */ + return disk_report_zone(vblk->disk, &zone, idx, args); +} + +static int virtblk_report_zones(struct gendisk *disk, sector_t sector, + unsigned int nr_zones, + struct blk_report_zones_args *args) +{ + struct virtio_blk *vblk = disk->private_data; + struct virtio_blk_zone_report *report; + unsigned long long nz, i; + size_t buflen; + unsigned int zone_idx = 0; + int ret; + + if (WARN_ON_ONCE(!vblk->zone_sectors)) + return -EOPNOTSUPP; + + report = virtblk_alloc_report_buffer(vblk, nr_zones, &buflen); + if (!report) + return -ENOMEM; + + mutex_lock(&vblk->vdev_mutex); + + if (!vblk->vdev) { + ret = -ENXIO; + goto fail_report; + } + + while (zone_idx < nr_zones && sector < get_capacity(vblk->disk)) { + memset(report, 0, buflen); + + ret = virtblk_submit_zone_report(vblk, (char *)report, + buflen, sector); + if (ret) + goto fail_report; + + nz = min_t(u64, virtio64_to_cpu(vblk->vdev, report->nr_zones), + nr_zones); + if (!nz) + break; + + for (i = 0; i < nz && zone_idx < nr_zones; i++) { + ret = virtblk_parse_zone(vblk, &report->zones[i], + zone_idx, args); + if (ret) + goto fail_report; + + sector = virtio64_to_cpu(vblk->vdev, + report->zones[i].z_start) + + vblk->zone_sectors; + zone_idx++; + } + } + + if (zone_idx > 0) + ret = zone_idx; + else + ret = -EINVAL; +fail_report: + mutex_unlock(&vblk->vdev_mutex); + kvfree(report); + return ret; +} + +static int virtblk_read_zoned_limits(struct virtio_blk *vblk, + struct queue_limits *lim) +{ + struct virtio_device *vdev = vblk->vdev; + u32 v, wg; + + dev_dbg(&vdev->dev, "probing host-managed zoned device\n"); + + lim->features |= BLK_FEAT_ZONED; + + virtio_cread(vdev, struct virtio_blk_config, + zoned.max_open_zones, &v); + lim->max_open_zones = v; + dev_dbg(&vdev->dev, "max open zones = %u\n", v); + + virtio_cread(vdev, struct virtio_blk_config, + zoned.max_active_zones, &v); + lim->max_active_zones = v; + dev_dbg(&vdev->dev, "max active zones = %u\n", v); + + virtio_cread(vdev, struct virtio_blk_config, + zoned.write_granularity, &wg); + if (!wg) { + dev_warn(&vdev->dev, "zero write granularity reported\n"); + return -ENODEV; + } + lim->physical_block_size = wg; + lim->io_min = wg; + + dev_dbg(&vdev->dev, "write granularity = %u\n", wg); + + /* + * virtio ZBD specification doesn't require zones to be a power of + * two sectors in size, but the code in this driver expects that. + */ + virtio_cread(vdev, struct virtio_blk_config, zoned.zone_sectors, + &vblk->zone_sectors); + if (vblk->zone_sectors == 0 || !is_power_of_2(vblk->zone_sectors)) { + dev_err(&vdev->dev, + "zoned device with non power of two zone size %u\n", + vblk->zone_sectors); + return -ENODEV; + } + lim->chunk_sectors = vblk->zone_sectors; + dev_dbg(&vdev->dev, "zone sectors = %u\n", vblk->zone_sectors); + + if (virtio_has_feature(vdev, VIRTIO_BLK_F_DISCARD)) { + dev_warn(&vblk->vdev->dev, + "ignoring negotiated F_DISCARD for zoned device\n"); + lim->max_hw_discard_sectors = 0; + } + + virtio_cread(vdev, struct virtio_blk_config, + zoned.max_append_sectors, &v); + if (!v) { + dev_warn(&vdev->dev, "zero max_append_sectors reported\n"); + return -ENODEV; + } + if ((v << SECTOR_SHIFT) < wg) { + dev_err(&vdev->dev, + "write granularity %u exceeds max_append_sectors %u limit\n", + wg, v); + return -ENODEV; + } + lim->max_hw_zone_append_sectors = v; + dev_dbg(&vdev->dev, "max append sectors = %u\n", v); + + return 0; +} +#else +/* + * Zoned block device support is not configured in this kernel, host-managed + * zoned devices can't be supported. + */ +#define virtblk_report_zones NULL +static inline int virtblk_read_zoned_limits(struct virtio_blk *vblk, + struct queue_limits *lim) +{ + dev_err(&vblk->vdev->dev, + "virtio_blk: zoned devices are not supported"); + return -EOPNOTSUPP; +} +#endif /* CONFIG_BLK_DEV_ZONED */ + /* return id (s/n) string for *disk to *id_str */ static int virtblk_get_id(struct gendisk *disk, char *id_str) @@ -464,27 +806,33 @@ static int virtblk_get_id(struct gendisk *disk, char *id_str) struct virtio_blk *vblk = disk->private_data; struct request_queue *q = vblk->disk->queue; struct request *req; + struct virtblk_req *vbr; int err; req = blk_mq_alloc_request(q, REQ_OP_DRV_IN, 0); if (IS_ERR(req)) return PTR_ERR(req); - err = blk_rq_map_kern(q, req, id_str, VIRTIO_BLK_ID_BYTES, GFP_KERNEL); + vbr = blk_mq_rq_to_pdu(req); + vbr->in_hdr_len = sizeof(vbr->in_hdr.status); + vbr->out_hdr.type = cpu_to_virtio32(vblk->vdev, VIRTIO_BLK_T_GET_ID); + vbr->out_hdr.sector = 0; + + err = blk_rq_map_kern(req, id_str, VIRTIO_BLK_ID_BYTES, GFP_KERNEL); if (err) goto out; blk_execute_rq(req, false); - err = blk_status_to_errno(virtblk_result(blk_mq_rq_to_pdu(req))); + err = blk_status_to_errno(virtblk_result(vbr->in_hdr.status)); out: blk_mq_free_request(req); return err; } /* We provide getgeo only to please some old bootloader/partitioning tools */ -static int virtblk_getgeo(struct block_device *bd, struct hd_geometry *geo) +static int virtblk_getgeo(struct gendisk *disk, struct hd_geometry *geo) { - struct virtio_blk *vblk = bd->bd_disk->private_data; + struct virtio_blk *vblk = disk->private_data; int ret = 0; mutex_lock(&vblk->vdev_mutex); @@ -506,7 +854,7 @@ static int virtblk_getgeo(struct block_device *bd, struct hd_geometry *geo) /* some standard values, similar to sd */ geo->heads = 1 << 6; geo->sectors = 1 << 5; - geo->cylinders = get_capacity(bd->bd_disk) >> 11; + geo->cylinders = get_capacity(disk) >> 11; } out: mutex_unlock(&vblk->vdev_mutex); @@ -526,6 +874,7 @@ static const struct block_device_operations virtblk_fops = { .owner = THIS_MODULE, .getgeo = virtblk_getgeo, .free_disk = virtblk_free_disk, + .report_zones = virtblk_report_zones, }; static int index_to_minor(int index) @@ -609,12 +958,11 @@ static void virtblk_config_changed(struct virtio_device *vdev) static int init_vq(struct virtio_blk *vblk) { int err; - int i; - vq_callback_t **callbacks; - const char **names; + unsigned short i; + struct virtqueue_info *vqs_info; struct virtqueue **vqs; unsigned short num_vqs; - unsigned int num_poll_vqs; + unsigned short num_poll_vqs; struct virtio_device *vdev = vblk->vdev; struct irq_affinity desc = { 0, }; @@ -629,9 +977,8 @@ static int init_vq(struct virtio_blk *vblk) return -EINVAL; } - num_vqs = min_t(unsigned int, - min_not_zero(num_request_queues, nr_cpu_ids), - num_vqs); + num_vqs = blk_mq_num_possible_queues( + min_not_zero(num_request_queues, num_vqs)); num_poll_vqs = min_t(unsigned int, poll_queues, num_vqs - 1); @@ -648,28 +995,26 @@ static int init_vq(struct virtio_blk *vblk) if (!vblk->vqs) return -ENOMEM; - names = kmalloc_array(num_vqs, sizeof(*names), GFP_KERNEL); - callbacks = kmalloc_array(num_vqs, sizeof(*callbacks), GFP_KERNEL); + vqs_info = kcalloc(num_vqs, sizeof(*vqs_info), GFP_KERNEL); vqs = kmalloc_array(num_vqs, sizeof(*vqs), GFP_KERNEL); - if (!names || !callbacks || !vqs) { + if (!vqs_info || !vqs) { err = -ENOMEM; goto out; } for (i = 0; i < num_vqs - num_poll_vqs; i++) { - callbacks[i] = virtblk_done; - snprintf(vblk->vqs[i].name, VQ_NAME_LEN, "req.%d", i); - names[i] = vblk->vqs[i].name; + vqs_info[i].callback = virtblk_done; + snprintf(vblk->vqs[i].name, VQ_NAME_LEN, "req.%u", i); + vqs_info[i].name = vblk->vqs[i].name; } for (; i < num_vqs; i++) { - callbacks[i] = NULL; - snprintf(vblk->vqs[i].name, VQ_NAME_LEN, "req_poll.%d", i); - names[i] = vblk->vqs[i].name; + snprintf(vblk->vqs[i].name, VQ_NAME_LEN, "req_poll.%u", i); + vqs_info[i].name = vblk->vqs[i].name; } /* Discover virtqueues and write information to configuration. */ - err = virtio_find_vqs(vdev, num_vqs, vqs, callbacks, names, &desc); + err = virtio_find_vqs(vdev, num_vqs, vqs, vqs_info, &desc); if (err) goto out; @@ -681,10 +1026,14 @@ static int init_vq(struct virtio_blk *vblk) out: kfree(vqs); - kfree(callbacks); - kfree(names); - if (err) + kfree(vqs_info); + if (err) { kfree(vblk->vqs); + /* + * Set to NULL to prevent freeing vqs again during freezing. + */ + vblk->vqs = NULL; + } return err; } @@ -735,14 +1084,6 @@ static int virtblk_get_cache_mode(struct virtio_device *vdev) return writeback; } -static void virtblk_update_cache_mode(struct virtio_device *vdev) -{ - u8 writeback = virtblk_get_cache_mode(vdev); - struct virtio_blk *vblk = vdev->priv; - - blk_queue_write_cache(vblk->disk->queue, writeback, false); -} - static const char *const virtblk_cache_types[] = { "write through", "write back" }; @@ -754,6 +1095,7 @@ cache_type_store(struct device *dev, struct device_attribute *attr, struct gendisk *disk = dev_to_disk(dev); struct virtio_blk *vblk = disk->private_data; struct virtio_device *vdev = vblk->vdev; + struct queue_limits lim; int i; BUG_ON(!virtio_has_feature(vblk->vdev, VIRTIO_BLK_F_CONFIG_WCE)); @@ -762,7 +1104,15 @@ cache_type_store(struct device *dev, struct device_attribute *attr, return i; virtio_cwrite8(vdev, offsetof(struct virtio_blk_config, wce), i); - virtblk_update_cache_mode(vdev); + + lim = queue_limits_start_update(disk->queue); + if (virtblk_get_cache_mode(vdev)) + lim.features |= BLK_FEAT_WRITE_CACHE; + else + lim.features &= ~BLK_FEAT_WRITE_CACHE; + i = queue_limits_commit_update_frozen(disk->queue, &lim); + if (i) + return i; return count; } @@ -833,7 +1183,8 @@ static void virtblk_map_queues(struct blk_mq_tag_set *set) if (i == HCTX_TYPE_POLL) blk_mq_map_queues(&set->map[i]); else - blk_mq_virtio_map_queues(&set->map[i], vblk->vdev, 0); + blk_mq_map_hw_queues(&set->map[i], + &vblk->vdev->dev, 0); } } @@ -861,11 +1212,13 @@ static int virtblk_poll(struct blk_mq_hw_ctx *hctx, struct io_comp_batch *iob) while ((vbr = virtqueue_get_buf(vq->vq, &len)) != NULL) { struct request *req = blk_mq_rq_from_pdu(vbr); + u8 status = virtblk_vbr_status(vbr); found++; - if (!blk_mq_add_to_batch(req, iob, vbr->status, - virtblk_complete_batch)) - blk_mq_complete_request(req); + if (!blk_mq_complete_request_remote(req) && + !blk_mq_add_to_batch(req, iob, status != VIRTIO_BLK_S_OK, + virtblk_complete_batch)) + virtblk_request_done(req); } if (found) @@ -888,30 +1241,17 @@ static const struct blk_mq_ops virtio_mq_ops = { static unsigned int virtblk_queue_depth; module_param_named(queue_depth, virtblk_queue_depth, uint, 0444); -static int virtblk_probe(struct virtio_device *vdev) +static int virtblk_read_limits(struct virtio_blk *vblk, + struct queue_limits *lim) { - struct virtio_blk *vblk; - struct request_queue *q; - int err, index; - - u32 v, blk_size, max_size, sg_elems, opt_io_size; + struct virtio_device *vdev = vblk->vdev; + u32 v, max_size, sg_elems, opt_io_size; u32 max_discard_segs = 0; u32 discard_granularity = 0; u16 min_io_size; u8 physical_block_exp, alignment_offset; - unsigned int queue_depth; - - if (!vdev->config->get) { - dev_err(&vdev->dev, "%s failure: config access disabled\n", - __func__); - return -EINVAL; - } - - err = ida_alloc_range(&vd_index_ida, 0, - minor_to_index(1 << MINORBITS) - 1, GFP_KERNEL); - if (err < 0) - goto out; - index = err; + size_t max_dma_size; + int err; /* We need to know how many segments before we allocate. */ err = virtio_cread_feature(vdev, VIRTIO_BLK_F_SEG_MAX, @@ -925,80 +1265,14 @@ static int virtblk_probe(struct virtio_device *vdev) /* Prevent integer overflows and honor max vq size */ sg_elems = min_t(u32, sg_elems, VIRTIO_BLK_MAX_SG_ELEMS - 2); - vdev->priv = vblk = kmalloc(sizeof(*vblk), GFP_KERNEL); - if (!vblk) { - err = -ENOMEM; - goto out_free_index; - } - - mutex_init(&vblk->vdev_mutex); - - vblk->vdev = vdev; - - INIT_WORK(&vblk->config_work, virtblk_config_changed_work); - - err = init_vq(vblk); - if (err) - goto out_free_vblk; - - /* Default queue sizing is to fill the ring. */ - if (!virtblk_queue_depth) { - queue_depth = vblk->vqs[0].vq->num_free; - /* ... but without indirect descs, we use 2 descs per req */ - if (!virtio_has_feature(vdev, VIRTIO_RING_F_INDIRECT_DESC)) - queue_depth /= 2; - } else { - queue_depth = virtblk_queue_depth; - } - - memset(&vblk->tag_set, 0, sizeof(vblk->tag_set)); - vblk->tag_set.ops = &virtio_mq_ops; - vblk->tag_set.queue_depth = queue_depth; - vblk->tag_set.numa_node = NUMA_NO_NODE; - vblk->tag_set.flags = BLK_MQ_F_SHOULD_MERGE; - vblk->tag_set.cmd_size = - sizeof(struct virtblk_req) + - sizeof(struct scatterlist) * VIRTIO_BLK_INLINE_SG_CNT; - vblk->tag_set.driver_data = vblk; - vblk->tag_set.nr_hw_queues = vblk->num_vqs; - vblk->tag_set.nr_maps = 1; - if (vblk->io_queues[HCTX_TYPE_POLL]) - vblk->tag_set.nr_maps = 3; - - err = blk_mq_alloc_tag_set(&vblk->tag_set); - if (err) - goto out_free_vq; - - vblk->disk = blk_mq_alloc_disk(&vblk->tag_set, vblk); - if (IS_ERR(vblk->disk)) { - err = PTR_ERR(vblk->disk); - goto out_free_tags; - } - q = vblk->disk->queue; - - virtblk_name_format("vd", index, vblk->disk->disk_name, DISK_NAME_LEN); - - vblk->disk->major = major; - vblk->disk->first_minor = index_to_minor(index); - vblk->disk->minors = 1 << PART_BITS; - vblk->disk->private_data = vblk; - vblk->disk->fops = &virtblk_fops; - vblk->index = index; - - /* configure queue flush support */ - virtblk_update_cache_mode(vdev); - - /* If disk is read-only in the host, the guest should obey */ - if (virtio_has_feature(vdev, VIRTIO_BLK_F_RO)) - set_disk_ro(vblk->disk, 1); - /* We can handle whatever the host told us to handle. */ - blk_queue_max_segments(q, sg_elems); + lim->max_segments = sg_elems; /* No real sector limit. */ - blk_queue_max_hw_sectors(q, UINT_MAX); + lim->max_hw_sectors = UINT_MAX; - max_size = virtio_max_dma_size(vdev); + max_dma_size = virtio_max_dma_size(vdev); + max_size = max_dma_size > U32_MAX ? U32_MAX : max_dma_size; /* Host can optionally specify maximum segment size and number of * segments. */ @@ -1007,50 +1281,39 @@ static int virtblk_probe(struct virtio_device *vdev) if (!err) max_size = min(max_size, v); - blk_queue_max_segment_size(q, max_size); + lim->max_segment_size = max_size; /* Host can optionally specify the block size of the device */ - err = virtio_cread_feature(vdev, VIRTIO_BLK_F_BLK_SIZE, + virtio_cread_feature(vdev, VIRTIO_BLK_F_BLK_SIZE, struct virtio_blk_config, blk_size, - &blk_size); - if (!err) { - err = blk_validate_block_size(blk_size); - if (err) { - dev_err(&vdev->dev, - "virtio_blk: invalid block size: 0x%x\n", - blk_size); - goto out_cleanup_disk; - } - - blk_queue_logical_block_size(q, blk_size); - } else - blk_size = queue_logical_block_size(q); + &lim->logical_block_size); /* Use topology information if available */ err = virtio_cread_feature(vdev, VIRTIO_BLK_F_TOPOLOGY, struct virtio_blk_config, physical_block_exp, &physical_block_exp); if (!err && physical_block_exp) - blk_queue_physical_block_size(q, - blk_size * (1 << physical_block_exp)); + lim->physical_block_size = + lim->logical_block_size * (1 << physical_block_exp); err = virtio_cread_feature(vdev, VIRTIO_BLK_F_TOPOLOGY, struct virtio_blk_config, alignment_offset, &alignment_offset); if (!err && alignment_offset) - blk_queue_alignment_offset(q, blk_size * alignment_offset); + lim->alignment_offset = + lim->logical_block_size * alignment_offset; err = virtio_cread_feature(vdev, VIRTIO_BLK_F_TOPOLOGY, struct virtio_blk_config, min_io_size, &min_io_size); if (!err && min_io_size) - blk_queue_io_min(q, blk_size * min_io_size); + lim->io_min = lim->logical_block_size * min_io_size; err = virtio_cread_feature(vdev, VIRTIO_BLK_F_TOPOLOGY, struct virtio_blk_config, opt_io_size, &opt_io_size); if (!err && opt_io_size) - blk_queue_io_opt(q, blk_size * opt_io_size); + lim->io_opt = lim->logical_block_size * opt_io_size; if (virtio_has_feature(vdev, VIRTIO_BLK_F_DISCARD)) { virtio_cread(vdev, struct virtio_blk_config, @@ -1058,7 +1321,7 @@ static int virtblk_probe(struct virtio_device *vdev) virtio_cread(vdev, struct virtio_blk_config, max_discard_sectors, &v); - blk_queue_max_discard_sectors(q, v ? v : UINT_MAX); + lim->max_hw_discard_sectors = v ? v : UINT_MAX; virtio_cread(vdev, struct virtio_blk_config, max_discard_seg, &max_discard_segs); @@ -1067,7 +1330,7 @@ static int virtblk_probe(struct virtio_device *vdev) if (virtio_has_feature(vdev, VIRTIO_BLK_F_WRITE_ZEROES)) { virtio_cread(vdev, struct virtio_blk_config, max_write_zeroes_sectors, &v); - blk_queue_max_write_zeroes_sectors(q, v ? v : UINT_MAX); + lim->max_write_zeroes_sectors = v ? v : UINT_MAX; } /* The discard and secure erase limits are combined since the Linux @@ -1093,8 +1356,7 @@ static int virtblk_probe(struct virtio_device *vdev) if (!v) { dev_err(&vdev->dev, "virtio_blk: secure_erase_sector_alignment can't be 0\n"); - err = -EINVAL; - goto out_cleanup_disk; + return -EINVAL; } discard_granularity = min_not_zero(discard_granularity, v); @@ -1108,11 +1370,10 @@ static int virtblk_probe(struct virtio_device *vdev) if (!v) { dev_err(&vdev->dev, "virtio_blk: max_secure_erase_sectors can't be 0\n"); - err = -EINVAL; - goto out_cleanup_disk; + return -EINVAL; } - blk_queue_max_secure_erase_sectors(q, v); + lim->max_secure_erase_sectors = v; virtio_cread(vdev, struct virtio_blk_config, max_secure_erase_seg, &v); @@ -1123,8 +1384,7 @@ static int virtblk_probe(struct virtio_device *vdev) if (!v) { dev_err(&vdev->dev, "virtio_blk: max_secure_erase_seg can't be 0\n"); - err = -EINVAL; - goto out_cleanup_disk; + return -EINVAL; } max_discard_segs = min_not_zero(max_discard_segs, v); @@ -1140,18 +1400,144 @@ static int virtblk_probe(struct virtio_device *vdev) if (!max_discard_segs) max_discard_segs = sg_elems; - blk_queue_max_discard_segments(q, - min(max_discard_segs, MAX_DISCARD_SEGMENTS)); + lim->max_discard_segments = + min(max_discard_segs, MAX_DISCARD_SEGMENTS); if (discard_granularity) - q->limits.discard_granularity = discard_granularity << SECTOR_SHIFT; + lim->discard_granularity = + discard_granularity << SECTOR_SHIFT; else - q->limits.discard_granularity = blk_size; + lim->discard_granularity = lim->logical_block_size; + } + + if (virtio_has_feature(vdev, VIRTIO_BLK_F_ZONED)) { + u8 model; + + virtio_cread(vdev, struct virtio_blk_config, zoned.model, &model); + switch (model) { + case VIRTIO_BLK_Z_NONE: + case VIRTIO_BLK_Z_HA: + /* treat host-aware devices as non-zoned */ + return 0; + case VIRTIO_BLK_Z_HM: + err = virtblk_read_zoned_limits(vblk, lim); + if (err) + return err; + break; + default: + dev_err(&vdev->dev, "unsupported zone model %d\n", model); + return -EINVAL; + } + } + + return 0; +} + +static int virtblk_probe(struct virtio_device *vdev) +{ + struct virtio_blk *vblk; + struct queue_limits lim = { + .features = BLK_FEAT_ROTATIONAL, + .logical_block_size = SECTOR_SIZE, + }; + int err, index; + unsigned int queue_depth; + + if (!vdev->config->get) { + dev_err(&vdev->dev, "%s failure: config access disabled\n", + __func__); + return -EINVAL; + } + + err = ida_alloc_range(&vd_index_ida, 0, + minor_to_index(1 << MINORBITS) - 1, GFP_KERNEL); + if (err < 0) + goto out; + index = err; + + vdev->priv = vblk = kmalloc(sizeof(*vblk), GFP_KERNEL); + if (!vblk) { + err = -ENOMEM; + goto out_free_index; + } + + mutex_init(&vblk->vdev_mutex); + + vblk->vdev = vdev; + + INIT_WORK(&vblk->config_work, virtblk_config_changed_work); + + err = init_vq(vblk); + if (err) + goto out_free_vblk; + + /* Default queue sizing is to fill the ring. */ + if (!virtblk_queue_depth) { + queue_depth = vblk->vqs[0].vq->num_free; + /* ... but without indirect descs, we use 2 descs per req */ + if (!virtio_has_feature(vdev, VIRTIO_RING_F_INDIRECT_DESC)) + queue_depth /= 2; + } else { + queue_depth = virtblk_queue_depth; } + memset(&vblk->tag_set, 0, sizeof(vblk->tag_set)); + vblk->tag_set.ops = &virtio_mq_ops; + vblk->tag_set.queue_depth = queue_depth; + vblk->tag_set.numa_node = NUMA_NO_NODE; + vblk->tag_set.cmd_size = + sizeof(struct virtblk_req) + + sizeof(struct scatterlist) * VIRTIO_BLK_INLINE_SG_CNT; + vblk->tag_set.driver_data = vblk; + vblk->tag_set.nr_hw_queues = vblk->num_vqs; + vblk->tag_set.nr_maps = 1; + if (vblk->io_queues[HCTX_TYPE_POLL]) + vblk->tag_set.nr_maps = 3; + + err = blk_mq_alloc_tag_set(&vblk->tag_set); + if (err) + goto out_free_vq; + + err = virtblk_read_limits(vblk, &lim); + if (err) + goto out_free_tags; + + if (virtblk_get_cache_mode(vdev)) + lim.features |= BLK_FEAT_WRITE_CACHE; + + vblk->disk = blk_mq_alloc_disk(&vblk->tag_set, &lim, vblk); + if (IS_ERR(vblk->disk)) { + err = PTR_ERR(vblk->disk); + goto out_free_tags; + } + + virtblk_name_format("vd", index, vblk->disk->disk_name, DISK_NAME_LEN); + + vblk->disk->major = major; + vblk->disk->first_minor = index_to_minor(index); + vblk->disk->minors = 1 << PART_BITS; + vblk->disk->private_data = vblk; + vblk->disk->fops = &virtblk_fops; + vblk->index = index; + + /* If disk is read-only in the host, the guest should obey */ + if (virtio_has_feature(vdev, VIRTIO_BLK_F_RO)) + set_disk_ro(vblk->disk, 1); + virtblk_update_capacity(vblk, false); virtio_device_ready(vdev); + /* + * All steps that follow use the VQs therefore they need to be + * placed after the virtio_device_ready() call above. + */ + if (IS_ENABLED(CONFIG_BLK_DEV_ZONED) && + (lim.features & BLK_FEAT_ZONED)) { + err = blk_revalidate_disk_zones(vblk->disk); + if (err) + goto out_cleanup_disk; + } + err = device_add_disk(&vdev->dev, vblk->disk, virtblk_attr_groups); if (err) goto out_cleanup_disk; @@ -1199,10 +1585,16 @@ static void virtblk_remove(struct virtio_device *vdev) put_disk(vblk->disk); } -#ifdef CONFIG_PM_SLEEP -static int virtblk_freeze(struct virtio_device *vdev) +static int virtblk_freeze_priv(struct virtio_device *vdev) { struct virtio_blk *vblk = vdev->priv; + struct request_queue *q = vblk->disk->queue; + unsigned int memflags; + + /* Ensure no requests in virtqueues before deleting vqs. */ + memflags = blk_mq_freeze_queue(q); + blk_mq_quiesce_queue_nowait(q); + blk_mq_unfreeze_queue(q, memflags); /* Ensure we don't receive any more interrupts */ virtio_reset_device(vdev); @@ -1210,15 +1602,19 @@ static int virtblk_freeze(struct virtio_device *vdev) /* Make sure no work handler is accessing the device. */ flush_work(&vblk->config_work); - blk_mq_quiesce_queue(vblk->disk->queue); - vdev->config->del_vqs(vdev); kfree(vblk->vqs); + /* + * Set to NULL to prevent freeing vqs again after a failed vqs + * allocation during resume. Note that kfree() already handles NULL + * pointers safely. + */ + vblk->vqs = NULL; return 0; } -static int virtblk_restore(struct virtio_device *vdev) +static int virtblk_restore_priv(struct virtio_device *vdev) { struct virtio_blk *vblk = vdev->priv; int ret; @@ -1228,12 +1624,33 @@ static int virtblk_restore(struct virtio_device *vdev) return ret; virtio_device_ready(vdev); - blk_mq_unquiesce_queue(vblk->disk->queue); + return 0; } + +#ifdef CONFIG_PM_SLEEP +static int virtblk_freeze(struct virtio_device *vdev) +{ + return virtblk_freeze_priv(vdev); +} + +static int virtblk_restore(struct virtio_device *vdev) +{ + return virtblk_restore_priv(vdev); +} #endif +static int virtblk_reset_prepare(struct virtio_device *vdev) +{ + return virtblk_freeze_priv(vdev); +} + +static int virtblk_reset_done(struct virtio_device *vdev) +{ + return virtblk_restore_priv(vdev); +} + static const struct virtio_device_id id_table[] = { { VIRTIO_ID_BLOCK, VIRTIO_DEV_ANY_ID }, { 0 }, @@ -1252,7 +1669,7 @@ static unsigned int features[] = { VIRTIO_BLK_F_RO, VIRTIO_BLK_F_BLK_SIZE, VIRTIO_BLK_F_FLUSH, VIRTIO_BLK_F_TOPOLOGY, VIRTIO_BLK_F_CONFIG_WCE, VIRTIO_BLK_F_MQ, VIRTIO_BLK_F_DISCARD, VIRTIO_BLK_F_WRITE_ZEROES, - VIRTIO_BLK_F_SECURE_ERASE, + VIRTIO_BLK_F_SECURE_ERASE, VIRTIO_BLK_F_ZONED, }; static struct virtio_driver virtio_blk = { @@ -1261,7 +1678,6 @@ static struct virtio_driver virtio_blk = { .feature_table_legacy = features_legacy, .feature_table_size_legacy = ARRAY_SIZE(features_legacy), .driver.name = KBUILD_MODNAME, - .driver.owner = THIS_MODULE, .id_table = id_table, .probe = virtblk_probe, .remove = virtblk_remove, @@ -1270,13 +1686,15 @@ static struct virtio_driver virtio_blk = { .freeze = virtblk_freeze, .restore = virtblk_restore, #endif + .reset_prepare = virtblk_reset_prepare, + .reset_done = virtblk_reset_done, }; static int __init virtio_blk_init(void) { int error; - virtblk_wq = alloc_workqueue("virtio-blk", 0, 0); + virtblk_wq = alloc_workqueue("virtio-blk", WQ_PERCPU, 0); if (!virtblk_wq) return -ENOMEM; |