diff options
author | Christoph Hellwig <hch@lst.de> | 2018-03-13 17:28:41 +0100 |
---|---|---|
committer | Jens Axboe <axboe@kernel.dk> | 2018-03-13 11:40:24 -0600 |
commit | 17cb960f29c29ee07bf6848ada3265f4be55972e (patch) | |
tree | 8571a25cac75a45261456dbe6711f324dc5b1044 /block | |
parent | ef6fa64f9b8e1611854077ea9213f2eef2428cd2 (diff) |
bsg: split handling of SCSI CDBs vs transport requeues
The current BSG design tries to shoe-horn the transport-specific
passthrough commands into the overall framework for SCSI passthrough
requests. This has a couple problems:
- each passthrough queue has to set the QUEUE_FLAG_SCSI_PASSTHROUGH flag
despite not dealing with SCSI commands at all. Because of that these
queues could also incorrectly accept SCSI commands from in-kernel
users or through the legacy SCSI_IOCTL_SEND_COMMAND ioctl.
- the real SCSI bsg queues also incorrectly accept bsg requests of the
BSG_SUB_PROTOCOL_SCSI_TRANSPORT type
- the bsg transport code is almost unredable because it tries to reuse
different SCSI concepts for its own purpose.
This patch instead adds a new bsg_ops structure to handle the two cases
differently, and thus solves all of the above problems. Another side
effect is that the bsg-lib queues also don't need to embedd a
struct scsi_request anymore.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Diffstat (limited to 'block')
-rw-r--r-- | block/bsg-lib.c | 158 | ||||
-rw-r--r-- | block/bsg.c | 262 |
2 files changed, 222 insertions, 198 deletions
diff --git a/block/bsg-lib.c b/block/bsg-lib.c index f2c2d54a61b4..fc2e5ff2c4b9 100644 --- a/block/bsg-lib.c +++ b/block/bsg-lib.c @@ -27,6 +27,94 @@ #include <linux/bsg-lib.h> #include <linux/export.h> #include <scsi/scsi_cmnd.h> +#include <scsi/sg.h> + +#define uptr64(val) ((void __user *)(uintptr_t)(val)) + +static int bsg_transport_check_proto(struct sg_io_v4 *hdr) +{ + if (hdr->protocol != BSG_PROTOCOL_SCSI || + hdr->subprotocol != BSG_SUB_PROTOCOL_SCSI_TRANSPORT) + return -EINVAL; + if (!capable(CAP_SYS_RAWIO)) + return -EPERM; + return 0; +} + +static int bsg_transport_fill_hdr(struct request *rq, struct sg_io_v4 *hdr, + fmode_t mode) +{ + struct bsg_job *job = blk_mq_rq_to_pdu(rq); + + job->request_len = hdr->request_len; + job->request = memdup_user(uptr64(hdr->request), hdr->request_len); + if (IS_ERR(job->request)) + return PTR_ERR(job->request); + return 0; +} + +static int bsg_transport_complete_rq(struct request *rq, struct sg_io_v4 *hdr) +{ + struct bsg_job *job = blk_mq_rq_to_pdu(rq); + int ret = 0; + + /* + * The assignments below don't make much sense, but are kept for + * bug by bug backwards compatibility: + */ + hdr->device_status = job->result & 0xff; + hdr->transport_status = host_byte(job->result); + hdr->driver_status = driver_byte(job->result); + hdr->info = 0; + if (hdr->device_status || hdr->transport_status || hdr->driver_status) + hdr->info |= SG_INFO_CHECK; + hdr->response_len = 0; + + if (job->result < 0) { + /* we're only returning the result field in the reply */ + job->reply_len = sizeof(u32); + ret = job->result; + } + + if (job->reply_len && hdr->response) { + int len = min(hdr->max_response_len, job->reply_len); + + if (copy_to_user(uptr64(hdr->response), job->reply, len)) + ret = -EFAULT; + else + hdr->response_len = len; + } + + /* we assume all request payload was transferred, residual == 0 */ + hdr->dout_resid = 0; + + if (rq->next_rq) { + unsigned int rsp_len = job->reply_payload.payload_len; + + if (WARN_ON(job->reply_payload_rcv_len > rsp_len)) + hdr->din_resid = 0; + else + hdr->din_resid = rsp_len - job->reply_payload_rcv_len; + } else { + hdr->din_resid = 0; + } + + return ret; +} + +static void bsg_transport_free_rq(struct request *rq) +{ + struct bsg_job *job = blk_mq_rq_to_pdu(rq); + + kfree(job->request); +} + +static const struct bsg_ops bsg_transport_ops = { + .check_proto = bsg_transport_check_proto, + .fill_hdr = bsg_transport_fill_hdr, + .complete_rq = bsg_transport_complete_rq, + .free_rq = bsg_transport_free_rq, +}; /** * bsg_teardown_job - routine to teardown a bsg job @@ -68,27 +156,9 @@ EXPORT_SYMBOL_GPL(bsg_job_get); void bsg_job_done(struct bsg_job *job, int result, unsigned int reply_payload_rcv_len) { - struct request *req = blk_mq_rq_from_pdu(job); - struct request *rsp = req->next_rq; - int err; - - err = job->sreq.result = result; - if (err < 0) - /* we're only returning the result field in the reply */ - job->sreq.sense_len = sizeof(u32); - else - job->sreq.sense_len = job->reply_len; - /* we assume all request payload was transferred, residual == 0 */ - job->sreq.resid_len = 0; - - if (rsp) { - WARN_ON(reply_payload_rcv_len > scsi_req(rsp)->resid_len); - - /* set reply (bidi) residual */ - scsi_req(rsp)->resid_len -= - min(reply_payload_rcv_len, scsi_req(rsp)->resid_len); - } - blk_complete_request(req); + job->result = result; + job->reply_payload_rcv_len = reply_payload_rcv_len; + blk_complete_request(blk_mq_rq_from_pdu(job)); } EXPORT_SYMBOL_GPL(bsg_job_done); @@ -113,7 +183,6 @@ static int bsg_map_buffer(struct bsg_buffer *buf, struct request *req) if (!buf->sg_list) return -ENOMEM; sg_init_table(buf->sg_list, req->nr_phys_segments); - scsi_req(req)->resid_len = blk_rq_bytes(req); buf->sg_cnt = blk_rq_map_sg(req->q, req, buf->sg_list); buf->payload_len = blk_rq_bytes(req); return 0; @@ -124,16 +193,13 @@ static int bsg_map_buffer(struct bsg_buffer *buf, struct request *req) * @dev: device that is being sent the bsg request * @req: BSG request that needs a job structure */ -static int bsg_prepare_job(struct device *dev, struct request *req) +static bool bsg_prepare_job(struct device *dev, struct request *req) { struct request *rsp = req->next_rq; - struct scsi_request *rq = scsi_req(req); struct bsg_job *job = blk_mq_rq_to_pdu(req); int ret; job->timeout = req->timeout; - job->request = rq->cmd; - job->request_len = rq->cmd_len; if (req->bio) { ret = bsg_map_buffer(&job->request_payload, req); @@ -149,12 +215,13 @@ static int bsg_prepare_job(struct device *dev, struct request *req) /* take a reference for the request */ get_device(job->dev); kref_init(&job->kref); - return 0; + return true; failjob_rls_rqst_payload: kfree(job->request_payload.sg_list); failjob_rls_job: - return -ENOMEM; + job->result = -ENOMEM; + return false; } /** @@ -183,9 +250,7 @@ static void bsg_request_fn(struct request_queue *q) break; spin_unlock_irq(q->queue_lock); - ret = bsg_prepare_job(dev, req); - if (ret) { - scsi_req(req)->result = ret; + if (!bsg_prepare_job(dev, req)) { blk_end_request_all(req, BLK_STS_OK); spin_lock_irq(q->queue_lock); continue; @@ -202,46 +267,34 @@ static void bsg_request_fn(struct request_queue *q) spin_lock_irq(q->queue_lock); } +/* called right after the request is allocated for the request_queue */ static int bsg_init_rq(struct request_queue *q, struct request *req, gfp_t gfp) { struct bsg_job *job = blk_mq_rq_to_pdu(req); - struct scsi_request *sreq = &job->sreq; - - /* called right after the request is allocated for the request_queue */ - sreq->sense = kzalloc(SCSI_SENSE_BUFFERSIZE, gfp); - if (!sreq->sense) + job->reply = kzalloc(SCSI_SENSE_BUFFERSIZE, gfp); + if (!job->reply) return -ENOMEM; - return 0; } +/* called right before the request is given to the request_queue user */ static void bsg_initialize_rq(struct request *req) { struct bsg_job *job = blk_mq_rq_to_pdu(req); - struct scsi_request *sreq = &job->sreq; - void *sense = sreq->sense; - - /* called right before the request is given to the request_queue user */ + void *reply = job->reply; memset(job, 0, sizeof(*job)); - - scsi_req_init(sreq); - - sreq->sense = sense; - sreq->sense_len = SCSI_SENSE_BUFFERSIZE; - - job->reply = sense; - job->reply_len = sreq->sense_len; + job->reply = reply; + job->reply_len = SCSI_SENSE_BUFFERSIZE; job->dd_data = job + 1; } static void bsg_exit_rq(struct request_queue *q, struct request *req) { struct bsg_job *job = blk_mq_rq_to_pdu(req); - struct scsi_request *sreq = &job->sreq; - kfree(sreq->sense); + kfree(job->reply); } /** @@ -275,11 +328,10 @@ struct request_queue *bsg_setup_queue(struct device *dev, const char *name, q->queuedata = dev; q->bsg_job_fn = job_fn; blk_queue_flag_set(QUEUE_FLAG_BIDI, q); - blk_queue_flag_set(QUEUE_FLAG_SCSI_PASSTHROUGH, q); blk_queue_softirq_done(q, bsg_softirq_done); blk_queue_rq_timeout(q, BLK_DEFAULT_SG_TIMEOUT); - ret = bsg_register_queue(q, dev, name, release); + ret = bsg_register_queue(q, dev, name, &bsg_transport_ops, release); if (ret) { printk(KERN_ERR "%s: bsg interface failed to " "initialize - register queue\n", dev->kobj.name); diff --git a/block/bsg.c b/block/bsg.c index 06dc96e1f670..defa06c11858 100644 --- a/block/bsg.c +++ b/block/bsg.c @@ -130,114 +130,120 @@ static inline struct hlist_head *bsg_dev_idx_hash(int index) return &bsg_device_list[index & (BSG_LIST_ARRAY_SIZE - 1)]; } -static int blk_fill_sgv4_hdr_rq(struct request_queue *q, struct request *rq, - struct sg_io_v4 *hdr, struct bsg_device *bd, - fmode_t mode) +#define uptr64(val) ((void __user *)(uintptr_t)(val)) + +static int bsg_scsi_check_proto(struct sg_io_v4 *hdr) +{ + if (hdr->protocol != BSG_PROTOCOL_SCSI || + hdr->subprotocol != BSG_SUB_PROTOCOL_SCSI_CMD) + return -EINVAL; + return 0; +} + +static int bsg_scsi_fill_hdr(struct request *rq, struct sg_io_v4 *hdr, + fmode_t mode) { - struct scsi_request *req = scsi_req(rq); + struct scsi_request *sreq = scsi_req(rq); - if (hdr->request_len > BLK_MAX_CDB) { - req->cmd = kzalloc(hdr->request_len, GFP_KERNEL); - if (!req->cmd) + sreq->cmd_len = hdr->request_len; + if (sreq->cmd_len > BLK_MAX_CDB) { + sreq->cmd = kzalloc(sreq->cmd_len, GFP_KERNEL); + if (!sreq->cmd) return -ENOMEM; } - if (copy_from_user(req->cmd, (void __user *)(unsigned long)hdr->request, - hdr->request_len)) + if (copy_from_user(sreq->cmd, uptr64(hdr->request), sreq->cmd_len)) return -EFAULT; - - if (hdr->subprotocol == BSG_SUB_PROTOCOL_SCSI_CMD) { - if (blk_verify_command(req->cmd, mode)) - return -EPERM; - } else if (!capable(CAP_SYS_RAWIO)) + if (blk_verify_command(sreq->cmd, mode)) return -EPERM; - - /* - * fill in request structure - */ - req->cmd_len = hdr->request_len; - - rq->timeout = msecs_to_jiffies(hdr->timeout); - if (!rq->timeout) - rq->timeout = q->sg_timeout; - if (!rq->timeout) - rq->timeout = BLK_DEFAULT_SG_TIMEOUT; - if (rq->timeout < BLK_MIN_SG_TIMEOUT) - rq->timeout = BLK_MIN_SG_TIMEOUT; - return 0; } -/* - * Check if sg_io_v4 from user is allowed and valid - */ -static int -bsg_validate_sgv4_hdr(struct sg_io_v4 *hdr, int *op) +static int bsg_scsi_complete_rq(struct request *rq, struct sg_io_v4 *hdr) { + struct scsi_request *sreq = scsi_req(rq); int ret = 0; - if (hdr->guard != 'Q') - return -EINVAL; + /* + * fill in all the output members + */ + hdr->device_status = sreq->result & 0xff; + hdr->transport_status = host_byte(sreq->result); + hdr->driver_status = driver_byte(sreq->result); + hdr->info = 0; + if (hdr->device_status || hdr->transport_status || hdr->driver_status) + hdr->info |= SG_INFO_CHECK; + hdr->response_len = 0; - switch (hdr->protocol) { - case BSG_PROTOCOL_SCSI: - switch (hdr->subprotocol) { - case BSG_SUB_PROTOCOL_SCSI_CMD: - case BSG_SUB_PROTOCOL_SCSI_TRANSPORT: - break; - default: - ret = -EINVAL; - } - break; - default: - ret = -EINVAL; + if (sreq->sense_len && hdr->response) { + int len = min_t(unsigned int, hdr->max_response_len, + sreq->sense_len); + + if (copy_to_user(uptr64(hdr->response), sreq->sense, len)) + ret = -EFAULT; + else + hdr->response_len = len; + } + + if (rq->next_rq) { + hdr->dout_resid = sreq->resid_len; + hdr->din_resid = scsi_req(rq->next_rq)->resid_len; + } else if (rq_data_dir(rq) == READ) { + hdr->din_resid = sreq->resid_len; + } else { + hdr->dout_resid = sreq->resid_len; } - *op = hdr->dout_xfer_len ? REQ_OP_SCSI_OUT : REQ_OP_SCSI_IN; return ret; } -/* - * map sg_io_v4 to a request. - */ +static void bsg_scsi_free_rq(struct request *rq) +{ + scsi_req_free_cmd(scsi_req(rq)); +} + +static const struct bsg_ops bsg_scsi_ops = { + .check_proto = bsg_scsi_check_proto, + .fill_hdr = bsg_scsi_fill_hdr, + .complete_rq = bsg_scsi_complete_rq, + .free_rq = bsg_scsi_free_rq, +}; + static struct request * -bsg_map_hdr(struct bsg_device *bd, struct sg_io_v4 *hdr, fmode_t mode) +bsg_map_hdr(struct request_queue *q, struct sg_io_v4 *hdr, fmode_t mode) { - struct request_queue *q = bd->queue; struct request *rq, *next_rq = NULL; int ret; - unsigned int op, dxfer_len; - void __user *dxferp = NULL; - struct bsg_class_device *bcd = &q->bsg_dev; - /* if the LLD has been removed then the bsg_unregister_queue will - * eventually be called and the class_dev was freed, so we can no - * longer use this request_queue. Return no such address. - */ - if (!bcd->class_dev) + if (!q->bsg_dev.class_dev) return ERR_PTR(-ENXIO); - bsg_dbg(bd, "map hdr %llx/%u %llx/%u\n", - (unsigned long long) hdr->dout_xferp, - hdr->dout_xfer_len, (unsigned long long) hdr->din_xferp, - hdr->din_xfer_len); + if (hdr->guard != 'Q') + return ERR_PTR(-EINVAL); - ret = bsg_validate_sgv4_hdr(hdr, &op); + ret = q->bsg_dev.ops->check_proto(hdr); if (ret) return ERR_PTR(ret); - /* - * map scatter-gather elements separately and string them to request - */ - rq = blk_get_request(q, op, GFP_KERNEL); + rq = blk_get_request(q, hdr->dout_xfer_len ? + REQ_OP_SCSI_OUT : REQ_OP_SCSI_IN, + GFP_KERNEL); if (IS_ERR(rq)) return rq; - ret = blk_fill_sgv4_hdr_rq(q, rq, hdr, bd, mode); + ret = q->bsg_dev.ops->fill_hdr(rq, hdr, mode); if (ret) goto out; - if (op == REQ_OP_SCSI_OUT && hdr->din_xfer_len) { + rq->timeout = msecs_to_jiffies(hdr->timeout); + if (!rq->timeout) + rq->timeout = q->sg_timeout; + if (!rq->timeout) + rq->timeout = BLK_DEFAULT_SG_TIMEOUT; + if (rq->timeout < BLK_MIN_SG_TIMEOUT) + rq->timeout = BLK_MIN_SG_TIMEOUT; + + if (hdr->dout_xfer_len && hdr->din_xfer_len) { if (!test_bit(QUEUE_FLAG_BIDI, &q->queue_flags)) { ret = -EOPNOTSUPP; goto out; @@ -246,42 +252,39 @@ bsg_map_hdr(struct bsg_device *bd, struct sg_io_v4 *hdr, fmode_t mode) next_rq = blk_get_request(q, REQ_OP_SCSI_IN, GFP_KERNEL); if (IS_ERR(next_rq)) { ret = PTR_ERR(next_rq); - next_rq = NULL; goto out; } - rq->next_rq = next_rq; - dxferp = (void __user *)(unsigned long)hdr->din_xferp; - ret = blk_rq_map_user(q, next_rq, NULL, dxferp, + rq->next_rq = next_rq; + ret = blk_rq_map_user(q, next_rq, NULL, uptr64(hdr->din_xferp), hdr->din_xfer_len, GFP_KERNEL); if (ret) - goto out; + goto out_free_nextrq; } if (hdr->dout_xfer_len) { - dxfer_len = hdr->dout_xfer_len; - dxferp = (void __user *)(unsigned long)hdr->dout_xferp; + ret = blk_rq_map_user(q, rq, NULL, uptr64(hdr->dout_xferp), + hdr->dout_xfer_len, GFP_KERNEL); } else if (hdr->din_xfer_len) { - dxfer_len = hdr->din_xfer_len; - dxferp = (void __user *)(unsigned long)hdr->din_xferp; - } else - dxfer_len = 0; - - if (dxfer_len) { - ret = blk_rq_map_user(q, rq, NULL, dxferp, dxfer_len, - GFP_KERNEL); - if (ret) - goto out; + ret = blk_rq_map_user(q, rq, NULL, uptr64(hdr->din_xferp), + hdr->din_xfer_len, GFP_KERNEL); + } else { + ret = blk_rq_map_user(q, rq, NULL, NULL, 0, GFP_KERNEL); } + if (ret) + goto out_unmap_nextrq; return rq; + +out_unmap_nextrq: + if (rq->next_rq) + blk_rq_unmap_user(rq->next_rq->bio); +out_free_nextrq: + if (rq->next_rq) + blk_put_request(rq->next_rq); out: - scsi_req_free_cmd(scsi_req(rq)); + q->bsg_dev.ops->free_rq(rq); blk_put_request(rq); - if (next_rq) { - blk_rq_unmap_user(next_rq->bio); - blk_put_request(next_rq); - } return ERR_PTR(ret); } @@ -383,56 +386,18 @@ static struct bsg_command *bsg_get_done_cmd(struct bsg_device *bd) static int blk_complete_sgv4_hdr_rq(struct request *rq, struct sg_io_v4 *hdr, struct bio *bio, struct bio *bidi_bio) { - struct scsi_request *req = scsi_req(rq); - int ret = 0; - - pr_debug("rq %p bio %p 0x%x\n", rq, bio, req->result); - /* - * fill in all the output members - */ - hdr->device_status = req->result & 0xff; - hdr->transport_status = host_byte(req->result); - hdr->driver_status = driver_byte(req->result); - hdr->info = 0; - if (hdr->device_status || hdr->transport_status || hdr->driver_status) - hdr->info |= SG_INFO_CHECK; - hdr->response_len = 0; - - if (req->sense_len && hdr->response) { - int len = min_t(unsigned int, hdr->max_response_len, - req->sense_len); + int ret; - ret = copy_to_user((void __user *)(unsigned long)hdr->response, - req->sense, len); - if (!ret) - hdr->response_len = len; - else - ret = -EFAULT; - } + ret = rq->q->bsg_dev.ops->complete_rq(rq, hdr); if (rq->next_rq) { - hdr->dout_resid = req->resid_len; - hdr->din_resid = scsi_req(rq->next_rq)->resid_len; blk_rq_unmap_user(bidi_bio); blk_put_request(rq->next_rq); - } else if (rq_data_dir(rq) == READ) - hdr->din_resid = req->resid_len; - else - hdr->dout_resid = req->resid_len; - - /* - * If the request generated a negative error number, return it - * (providing we aren't already returning an error); if it's - * just a protocol response (i.e. non negative), that gets - * processed above. - */ - if (!ret && req->result < 0) - ret = req->result; + } blk_rq_unmap_user(bio); - scsi_req_free_cmd(req); + rq->q->bsg_dev.ops->free_rq(rq); blk_put_request(rq); - return ret; } @@ -614,7 +579,7 @@ static int __bsg_write(struct bsg_device *bd, const char __user *buf, /* * get a request, fill in the blanks, and add to request queue */ - rq = bsg_map_hdr(bd, &bc->hdr, mode); + rq = bsg_map_hdr(bd->queue, &bc->hdr, mode); if (IS_ERR(rq)) { ret = PTR_ERR(rq); rq = NULL; @@ -742,11 +707,6 @@ static struct bsg_device *bsg_add_device(struct inode *inode, struct bsg_device *bd; unsigned char buf[32]; - if (!blk_queue_scsi_passthrough(rq)) { - WARN_ONCE(true, "Attempt to register a non-SCSI queue\n"); - return ERR_PTR(-EINVAL); - } - if (!blk_get_queue(rq)) return ERR_PTR(-ENXIO); @@ -907,7 +867,7 @@ static long bsg_ioctl(struct file *file, unsigned int cmd, unsigned long arg) if (copy_from_user(&hdr, uarg, sizeof(hdr))) return -EFAULT; - rq = bsg_map_hdr(bd, &hdr, file->f_mode); + rq = bsg_map_hdr(bd->queue, &hdr, file->f_mode); if (IS_ERR(rq)) return PTR_ERR(rq); @@ -959,7 +919,8 @@ void bsg_unregister_queue(struct request_queue *q) EXPORT_SYMBOL_GPL(bsg_unregister_queue); int bsg_register_queue(struct request_queue *q, struct device *parent, - const char *name, void (*release)(struct device *)) + const char *name, const struct bsg_ops *ops, + void (*release)(struct device *)) { struct bsg_class_device *bcd; dev_t dev; @@ -996,6 +957,7 @@ int bsg_register_queue(struct request_queue *q, struct device *parent, bcd->queue = q; bcd->parent = get_device(parent); bcd->release = release; + bcd->ops = ops; kref_init(&bcd->ref); dev = MKDEV(bsg_major, bcd->minor); class_dev = device_create(bsg_class, parent, dev, NULL, "%s", devname); @@ -1023,7 +985,17 @@ unlock: mutex_unlock(&bsg_mutex); return ret; } -EXPORT_SYMBOL_GPL(bsg_register_queue); + +int bsg_scsi_register_queue(struct request_queue *q, struct device *parent) +{ + if (!blk_queue_scsi_passthrough(q)) { + WARN_ONCE(true, "Attempt to register a non-SCSI queue\n"); + return -EINVAL; + } + + return bsg_register_queue(q, parent, NULL, &bsg_scsi_ops, NULL); +} +EXPORT_SYMBOL_GPL(bsg_scsi_register_queue); static struct cdev bsg_cdev; |