diff options
Diffstat (limited to 'drivers/block')
-rw-r--r-- | drivers/block/mtip32xx/mtip32xx.c | 6 | ||||
-rw-r--r-- | drivers/block/null_blk.c | 317 | ||||
-rw-r--r-- | drivers/block/rbd.c | 1 | ||||
-rw-r--r-- | drivers/block/xen-blkback/blkback.c | 15 | ||||
-rw-r--r-- | drivers/block/xen-blkback/common.h | 8 |
5 files changed, 261 insertions, 86 deletions
diff --git a/drivers/block/mtip32xx/mtip32xx.c b/drivers/block/mtip32xx/mtip32xx.c index a28a562f7b7f..3457ac8c03e2 100644 --- a/drivers/block/mtip32xx/mtip32xx.c +++ b/drivers/block/mtip32xx/mtip32xx.c @@ -3810,7 +3810,6 @@ static int mtip_block_initialize(struct driver_data *dd) sector_t capacity; unsigned int index = 0; struct kobject *kobj; - unsigned char thd_name[16]; if (dd->disk) goto skip_create_disk; /* hw init done, before rebuild */ @@ -3958,10 +3957,9 @@ skip_create_disk: } start_service_thread: - sprintf(thd_name, "mtip_svc_thd_%02d", index); dd->mtip_svc_handler = kthread_create_on_node(mtip_service_thread, - dd, dd->numa_node, "%s", - thd_name); + dd, dd->numa_node, + "mtip_svc_thd_%02d", index); if (IS_ERR(dd->mtip_svc_handler)) { dev_err(&dd->pdev->dev, "service thread failed to start\n"); diff --git a/drivers/block/null_blk.c b/drivers/block/null_blk.c index 6255d1c4bba4..a428e4ef71fd 100644 --- a/drivers/block/null_blk.c +++ b/drivers/block/null_blk.c @@ -8,6 +8,7 @@ #include <linux/slab.h> #include <linux/blk-mq.h> #include <linux/hrtimer.h> +#include <linux/lightnvm.h> struct nullb_cmd { struct list_head list; @@ -17,6 +18,7 @@ struct nullb_cmd { struct bio *bio; unsigned int tag; struct nullb_queue *nq; + struct hrtimer timer; }; struct nullb_queue { @@ -39,23 +41,14 @@ struct nullb { struct nullb_queue *queues; unsigned int nr_queues; + char disk_name[DISK_NAME_LEN]; }; static LIST_HEAD(nullb_list); static struct mutex lock; static int null_major; static int nullb_indexes; - -struct completion_queue { - struct llist_head list; - struct hrtimer timer; -}; - -/* - * These are per-cpu for now, they will need to be configured by the - * complete_queues parameter and appropriately mapped. - */ -static DEFINE_PER_CPU(struct completion_queue, completion_queues); +static struct kmem_cache *ppa_cache; enum { NULL_IRQ_NONE = 0, @@ -119,6 +112,10 @@ static int nr_devices = 2; module_param(nr_devices, int, S_IRUGO); MODULE_PARM_DESC(nr_devices, "Number of devices to register"); +static bool use_lightnvm; +module_param(use_lightnvm, bool, S_IRUGO); +MODULE_PARM_DESC(use_lightnvm, "Register as a LightNVM device"); + static int irqmode = NULL_IRQ_SOFTIRQ; static int null_set_irqmode(const char *str, const struct kernel_param *kp) @@ -135,8 +132,8 @@ static const struct kernel_param_ops null_irqmode_param_ops = { device_param_cb(irqmode, &null_irqmode_param_ops, &irqmode, S_IRUGO); MODULE_PARM_DESC(irqmode, "IRQ completion handler. 0-none, 1-softirq, 2-timer"); -static int completion_nsec = 10000; -module_param(completion_nsec, int, S_IRUGO); +static unsigned long completion_nsec = 10000; +module_param(completion_nsec, ulong, S_IRUGO); MODULE_PARM_DESC(completion_nsec, "Time in ns to complete a request in hardware. Default: 10,000ns"); static int hw_queue_depth = 64; @@ -173,6 +170,8 @@ static void free_cmd(struct nullb_cmd *cmd) put_tag(cmd->nq, cmd->tag); } +static enum hrtimer_restart null_cmd_timer_expired(struct hrtimer *timer); + static struct nullb_cmd *__alloc_cmd(struct nullb_queue *nq) { struct nullb_cmd *cmd; @@ -183,6 +182,11 @@ static struct nullb_cmd *__alloc_cmd(struct nullb_queue *nq) cmd = &nq->cmds[tag]; cmd->tag = tag; cmd->nq = nq; + if (irqmode == NULL_IRQ_TIMER) { + hrtimer_init(&cmd->timer, CLOCK_MONOTONIC, + HRTIMER_MODE_REL); + cmd->timer.function = null_cmd_timer_expired; + } return cmd; } @@ -213,6 +217,11 @@ static struct nullb_cmd *alloc_cmd(struct nullb_queue *nq, int can_wait) static void end_cmd(struct nullb_cmd *cmd) { + struct request_queue *q = NULL; + + if (cmd->rq) + q = cmd->rq->q; + switch (queue_mode) { case NULL_Q_MQ: blk_mq_end_request(cmd->rq, 0); @@ -223,55 +232,34 @@ static void end_cmd(struct nullb_cmd *cmd) break; case NULL_Q_BIO: bio_endio(cmd->bio); - break; + goto free_cmd; } + /* Restart queue if needed, as we are freeing a tag */ + if (q && !q->mq_ops && blk_queue_stopped(q)) { + unsigned long flags; + + spin_lock_irqsave(q->queue_lock, flags); + if (blk_queue_stopped(q)) + blk_start_queue(q); + spin_unlock_irqrestore(q->queue_lock, flags); + } +free_cmd: free_cmd(cmd); } static enum hrtimer_restart null_cmd_timer_expired(struct hrtimer *timer) { - struct completion_queue *cq; - struct llist_node *entry; - struct nullb_cmd *cmd; - - cq = &per_cpu(completion_queues, smp_processor_id()); - - while ((entry = llist_del_all(&cq->list)) != NULL) { - entry = llist_reverse_order(entry); - do { - struct request_queue *q = NULL; - - cmd = container_of(entry, struct nullb_cmd, ll_list); - entry = entry->next; - if (cmd->rq) - q = cmd->rq->q; - end_cmd(cmd); - - if (q && !q->mq_ops && blk_queue_stopped(q)) { - spin_lock(q->queue_lock); - if (blk_queue_stopped(q)) - blk_start_queue(q); - spin_unlock(q->queue_lock); - } - } while (entry); - } + end_cmd(container_of(timer, struct nullb_cmd, timer)); return HRTIMER_NORESTART; } static void null_cmd_end_timer(struct nullb_cmd *cmd) { - struct completion_queue *cq = &per_cpu(completion_queues, get_cpu()); + ktime_t kt = ktime_set(0, completion_nsec); - cmd->ll_list.next = NULL; - if (llist_add(&cmd->ll_list, &cq->list)) { - ktime_t kt = ktime_set(0, completion_nsec); - - hrtimer_start(&cq->timer, kt, HRTIMER_MODE_REL_PINNED); - } - - put_cpu(); + hrtimer_start(&cmd->timer, kt, HRTIMER_MODE_REL); } static void null_softirq_done_fn(struct request *rq) @@ -369,6 +357,10 @@ static int null_queue_rq(struct blk_mq_hw_ctx *hctx, { struct nullb_cmd *cmd = blk_mq_rq_to_pdu(bd->rq); + if (irqmode == NULL_IRQ_TIMER) { + hrtimer_init(&cmd->timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL); + cmd->timer.function = null_cmd_timer_expired; + } cmd->rq = bd->rq; cmd->nq = hctx->driver_data; @@ -427,15 +419,157 @@ static void null_del_dev(struct nullb *nullb) { list_del_init(&nullb->list); - del_gendisk(nullb->disk); + if (use_lightnvm) + nvm_unregister(nullb->disk_name); + else + del_gendisk(nullb->disk); blk_cleanup_queue(nullb->q); if (queue_mode == NULL_Q_MQ) blk_mq_free_tag_set(&nullb->tag_set); - put_disk(nullb->disk); + if (!use_lightnvm) + put_disk(nullb->disk); cleanup_queues(nullb); kfree(nullb); } +#ifdef CONFIG_NVM + +static void null_lnvm_end_io(struct request *rq, int error) +{ + struct nvm_rq *rqd = rq->end_io_data; + struct nvm_dev *dev = rqd->dev; + + dev->mt->end_io(rqd, error); + + blk_put_request(rq); +} + +static int null_lnvm_submit_io(struct nvm_dev *dev, struct nvm_rq *rqd) +{ + struct request_queue *q = dev->q; + struct request *rq; + struct bio *bio = rqd->bio; + + rq = blk_mq_alloc_request(q, bio_rw(bio), GFP_KERNEL, 0); + if (IS_ERR(rq)) + return -ENOMEM; + + rq->cmd_type = REQ_TYPE_DRV_PRIV; + rq->__sector = bio->bi_iter.bi_sector; + rq->ioprio = bio_prio(bio); + + if (bio_has_data(bio)) + rq->nr_phys_segments = bio_phys_segments(q, bio); + + rq->__data_len = bio->bi_iter.bi_size; + rq->bio = rq->biotail = bio; + + rq->end_io_data = rqd; + + blk_execute_rq_nowait(q, NULL, rq, 0, null_lnvm_end_io); + + return 0; +} + +static int null_lnvm_id(struct nvm_dev *dev, struct nvm_id *id) +{ + sector_t size = gb * 1024 * 1024 * 1024ULL; + sector_t blksize; + struct nvm_id_group *grp; + + id->ver_id = 0x1; + id->vmnt = 0; + id->cgrps = 1; + id->cap = 0x3; + id->dom = 0x1; + + id->ppaf.blk_offset = 0; + id->ppaf.blk_len = 16; + id->ppaf.pg_offset = 16; + id->ppaf.pg_len = 16; + id->ppaf.sect_offset = 32; + id->ppaf.sect_len = 8; + id->ppaf.pln_offset = 40; + id->ppaf.pln_len = 8; + id->ppaf.lun_offset = 48; + id->ppaf.lun_len = 8; + id->ppaf.ch_offset = 56; + id->ppaf.ch_len = 8; + + do_div(size, bs); /* convert size to pages */ + do_div(size, 256); /* concert size to pgs pr blk */ + grp = &id->groups[0]; + grp->mtype = 0; + grp->fmtype = 0; + grp->num_ch = 1; + grp->num_pg = 256; + blksize = size; + do_div(size, (1 << 16)); + grp->num_lun = size + 1; + do_div(blksize, grp->num_lun); + grp->num_blk = blksize; + grp->num_pln = 1; + + grp->fpg_sz = bs; + grp->csecs = bs; + grp->trdt = 25000; + grp->trdm = 25000; + grp->tprt = 500000; + grp->tprm = 500000; + grp->tbet = 1500000; + grp->tbem = 1500000; + grp->mpos = 0x010101; /* single plane rwe */ + grp->cpar = hw_queue_depth; + + return 0; +} + +static void *null_lnvm_create_dma_pool(struct nvm_dev *dev, char *name) +{ + mempool_t *virtmem_pool; + + virtmem_pool = mempool_create_slab_pool(64, ppa_cache); + if (!virtmem_pool) { + pr_err("null_blk: Unable to create virtual memory pool\n"); + return NULL; + } + + return virtmem_pool; +} + +static void null_lnvm_destroy_dma_pool(void *pool) +{ + mempool_destroy(pool); +} + +static void *null_lnvm_dev_dma_alloc(struct nvm_dev *dev, void *pool, + gfp_t mem_flags, dma_addr_t *dma_handler) +{ + return mempool_alloc(pool, mem_flags); +} + +static void null_lnvm_dev_dma_free(void *pool, void *entry, + dma_addr_t dma_handler) +{ + mempool_free(entry, pool); +} + +static struct nvm_dev_ops null_lnvm_dev_ops = { + .identity = null_lnvm_id, + .submit_io = null_lnvm_submit_io, + + .create_dma_pool = null_lnvm_create_dma_pool, + .destroy_dma_pool = null_lnvm_destroy_dma_pool, + .dev_dma_alloc = null_lnvm_dev_dma_alloc, + .dev_dma_free = null_lnvm_dev_dma_free, + + /* Simulate nvme protocol restriction */ + .max_phys_sect = 64, +}; +#else +static struct nvm_dev_ops null_lnvm_dev_ops; +#endif /* CONFIG_NVM */ + static int null_open(struct block_device *bdev, fmode_t mode) { return 0; @@ -575,11 +709,6 @@ static int null_add_dev(void) queue_flag_set_unlocked(QUEUE_FLAG_NONROT, nullb->q); queue_flag_clear_unlocked(QUEUE_FLAG_ADD_RANDOM, nullb->q); - disk = nullb->disk = alloc_disk_node(1, home_node); - if (!disk) { - rv = -ENOMEM; - goto out_cleanup_blk_queue; - } mutex_lock(&lock); list_add_tail(&nullb->list, &nullb_list); @@ -589,6 +718,21 @@ static int null_add_dev(void) blk_queue_logical_block_size(nullb->q, bs); blk_queue_physical_block_size(nullb->q, bs); + sprintf(nullb->disk_name, "nullb%d", nullb->index); + + if (use_lightnvm) { + rv = nvm_register(nullb->q, nullb->disk_name, + &null_lnvm_dev_ops); + if (rv) + goto out_cleanup_blk_queue; + goto done; + } + + disk = nullb->disk = alloc_disk_node(1, home_node); + if (!disk) { + rv = -ENOMEM; + goto out_cleanup_lightnvm; + } size = gb * 1024 * 1024 * 1024ULL; set_capacity(disk, size >> 9); @@ -598,10 +742,15 @@ static int null_add_dev(void) disk->fops = &null_fops; disk->private_data = nullb; disk->queue = nullb->q; - sprintf(disk->disk_name, "nullb%d", nullb->index); + strncpy(disk->disk_name, nullb->disk_name, DISK_NAME_LEN); + add_disk(disk); +done: return 0; +out_cleanup_lightnvm: + if (use_lightnvm) + nvm_unregister(nullb->disk_name); out_cleanup_blk_queue: blk_cleanup_queue(nullb->q); out_cleanup_tags: @@ -617,7 +766,9 @@ out: static int __init null_init(void) { + int ret = 0; unsigned int i; + struct nullb *nullb; if (bs > PAGE_SIZE) { pr_warn("null_blk: invalid block size\n"); @@ -625,6 +776,18 @@ static int __init null_init(void) bs = PAGE_SIZE; } + if (use_lightnvm && bs != 4096) { + pr_warn("null_blk: LightNVM only supports 4k block size\n"); + pr_warn("null_blk: defaults block size to 4k\n"); + bs = 4096; + } + + if (use_lightnvm && queue_mode != NULL_Q_MQ) { + pr_warn("null_blk: LightNVM only supported for blk-mq\n"); + pr_warn("null_blk: defaults queue mode to blk-mq\n"); + queue_mode = NULL_Q_MQ; + } + if (queue_mode == NULL_Q_MQ && use_per_node_hctx) { if (submit_queues < nr_online_nodes) { pr_warn("null_blk: submit_queues param is set to %u.", @@ -638,32 +801,38 @@ static int __init null_init(void) mutex_init(&lock); - /* Initialize a separate list for each CPU for issuing softirqs */ - for_each_possible_cpu(i) { - struct completion_queue *cq = &per_cpu(completion_queues, i); - - init_llist_head(&cq->list); - - if (irqmode != NULL_IRQ_TIMER) - continue; - - hrtimer_init(&cq->timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL); - cq->timer.function = null_cmd_timer_expired; - } - null_major = register_blkdev(0, "nullb"); if (null_major < 0) return null_major; - for (i = 0; i < nr_devices; i++) { - if (null_add_dev()) { - unregister_blkdev(null_major, "nullb"); - return -EINVAL; + if (use_lightnvm) { + ppa_cache = kmem_cache_create("ppa_cache", 64 * sizeof(u64), + 0, 0, NULL); + if (!ppa_cache) { + pr_err("null_blk: unable to create ppa cache\n"); + ret = -ENOMEM; + goto err_ppa; } } + for (i = 0; i < nr_devices; i++) { + ret = null_add_dev(); + if (ret) + goto err_dev; + } + pr_info("null: module loaded\n"); return 0; + +err_dev: + while (!list_empty(&nullb_list)) { + nullb = list_entry(nullb_list.next, struct nullb, list); + null_del_dev(nullb); + } + kmem_cache_destroy(ppa_cache); +err_ppa: + unregister_blkdev(null_major, "nullb"); + return ret; } static void __exit null_exit(void) @@ -678,6 +847,8 @@ static void __exit null_exit(void) null_del_dev(nullb); } mutex_unlock(&lock); + + kmem_cache_destroy(ppa_cache); } module_init(null_init); diff --git a/drivers/block/rbd.c b/drivers/block/rbd.c index 235708c7c46e..81ea69fee7ca 100644 --- a/drivers/block/rbd.c +++ b/drivers/block/rbd.c @@ -3442,6 +3442,7 @@ static void rbd_queue_workfn(struct work_struct *work) goto err_rq; } img_request->rq = rq; + snapc = NULL; /* img_request consumes a ref */ if (op_type == OBJ_OP_DISCARD) result = rbd_img_request_fill(img_request, OBJ_REQUEST_NODATA, diff --git a/drivers/block/xen-blkback/blkback.c b/drivers/block/xen-blkback/blkback.c index f9099940c272..41fb1a917b17 100644 --- a/drivers/block/xen-blkback/blkback.c +++ b/drivers/block/xen-blkback/blkback.c @@ -950,6 +950,8 @@ static int xen_blkbk_parse_indirect(struct blkif_request *req, goto unmap; for (n = 0, i = 0; n < nseg; n++) { + uint8_t first_sect, last_sect; + if ((n % SEGS_PER_INDIRECT_FRAME) == 0) { /* Map indirect segments */ if (segments) @@ -957,15 +959,18 @@ static int xen_blkbk_parse_indirect(struct blkif_request *req, segments = kmap_atomic(pages[n/SEGS_PER_INDIRECT_FRAME]->page); } i = n % SEGS_PER_INDIRECT_FRAME; + pending_req->segments[n]->gref = segments[i].gref; - seg[n].nsec = segments[i].last_sect - - segments[i].first_sect + 1; - seg[n].offset = (segments[i].first_sect << 9); - if ((segments[i].last_sect >= (XEN_PAGE_SIZE >> 9)) || - (segments[i].last_sect < segments[i].first_sect)) { + + first_sect = READ_ONCE(segments[i].first_sect); + last_sect = READ_ONCE(segments[i].last_sect); + if (last_sect >= (XEN_PAGE_SIZE >> 9) || last_sect < first_sect) { rc = -EINVAL; goto unmap; } + + seg[n].nsec = last_sect - first_sect + 1; + seg[n].offset = first_sect << 9; preq->nr_sects += seg[n].nsec; } diff --git a/drivers/block/xen-blkback/common.h b/drivers/block/xen-blkback/common.h index 68e87a037b99..c929ae22764c 100644 --- a/drivers/block/xen-blkback/common.h +++ b/drivers/block/xen-blkback/common.h @@ -408,8 +408,8 @@ static inline void blkif_get_x86_32_req(struct blkif_request *dst, struct blkif_x86_32_request *src) { int i, n = BLKIF_MAX_SEGMENTS_PER_REQUEST, j; - dst->operation = src->operation; - switch (src->operation) { + dst->operation = READ_ONCE(src->operation); + switch (dst->operation) { case BLKIF_OP_READ: case BLKIF_OP_WRITE: case BLKIF_OP_WRITE_BARRIER: @@ -456,8 +456,8 @@ static inline void blkif_get_x86_64_req(struct blkif_request *dst, struct blkif_x86_64_request *src) { int i, n = BLKIF_MAX_SEGMENTS_PER_REQUEST, j; - dst->operation = src->operation; - switch (src->operation) { + dst->operation = READ_ONCE(src->operation); + switch (dst->operation) { case BLKIF_OP_READ: case BLKIF_OP_WRITE: case BLKIF_OP_WRITE_BARRIER: |