From be1c63411addba3ad750eb4fdfc50b97bc82825e Mon Sep 17 00:00:00 2001 From: Olaf Kirch Date: Fri, 1 Dec 2006 10:39:12 +0100 Subject: [PATCH] blktrace: add timestamp message This adds a new timestamp message to blktrace, giving the timeofday when we starting tracing. This helps user space correlate block trace events with eg an application strace. This requires a (compatible) update to blkparse. The changed blkparse is still able to process traces generated by older kernels, and older versions of blkparse should silently ignore the new records (because they have a pid of 0). Signed-off-by: Olaf Kirch Signed-off-by: Jens Axboe --- block/blktrace.c | 57 ++++++++++++++++++++++++++++++++++++++++++++------------ 1 file changed, 45 insertions(+), 12 deletions(-) (limited to 'block') diff --git a/block/blktrace.c b/block/blktrace.c index 135593c8e45b..562ca7cbf858 100644 --- a/block/blktrace.c +++ b/block/blktrace.c @@ -22,30 +22,61 @@ #include #include #include +#include #include static DEFINE_PER_CPU(unsigned long long, blk_trace_cpu_offset) = { 0, }; static unsigned int blktrace_seq __read_mostly = 1; +/* + * Send out a notify message. + */ +static inline unsigned int trace_note(struct blk_trace *bt, + pid_t pid, int action, + const void *data, size_t len) +{ + struct blk_io_trace *t; + int cpu = smp_processor_id(); + + t = relay_reserve(bt->rchan, sizeof(*t) + len); + if (t == NULL) + return 0; + + t->magic = BLK_IO_TRACE_MAGIC | BLK_IO_TRACE_VERSION; + t->time = sched_clock() - per_cpu(blk_trace_cpu_offset, cpu); + t->device = bt->dev; + t->action = action; + t->pid = pid; + t->cpu = cpu; + t->pdu_len = len; + memcpy((void *) t + sizeof(*t), data, len); + return blktrace_seq; +} + /* * Send out a notify for this process, if we haven't done so since a trace * started */ static void trace_note_tsk(struct blk_trace *bt, struct task_struct *tsk) { - struct blk_io_trace *t; + tsk->btrace_seq = trace_note(bt, tsk->pid, + BLK_TN_PROCESS, + tsk->comm, sizeof(tsk->comm)); +} - t = relay_reserve(bt->rchan, sizeof(*t) + sizeof(tsk->comm)); - if (t) { - t->magic = BLK_IO_TRACE_MAGIC | BLK_IO_TRACE_VERSION; - t->device = bt->dev; - t->action = BLK_TC_ACT(BLK_TC_NOTIFY); - t->pid = tsk->pid; - t->cpu = smp_processor_id(); - t->pdu_len = sizeof(tsk->comm); - memcpy((void *) t + sizeof(*t), tsk->comm, t->pdu_len); - tsk->btrace_seq = blktrace_seq; - } +static void trace_note_time(struct blk_trace *bt) +{ + struct timespec now; + unsigned long flags; + u32 words[2]; + + getnstimeofday(&now); + words[0] = now.tv_sec; + words[1] = now.tv_nsec; + + local_irq_save(flags); + trace_note(bt, 0, BLK_TN_TIMESTAMP, words, sizeof(words)); + local_irq_restore(flags); } static int act_log_check(struct blk_trace *bt, u32 what, sector_t sector, @@ -394,6 +425,8 @@ static int blk_trace_startstop(request_queue_t *q, int start) blktrace_seq++; smp_mb(); bt->trace_state = Blktrace_running; + + trace_note_time(bt); ret = 0; } } else { -- cgit v1.2.3 From 0e75f9063f5c55fb0b0b546a7c356f8ec186825e Mon Sep 17 00:00:00 2001 From: Mike Christie Date: Fri, 1 Dec 2006 10:40:55 +0100 Subject: [PATCH] block: support larger block pc requests This patch modifies blk_rq_map/unmap_user() and the cdrom and scsi_ioctl.c users so that it supports requests larger than bio by chaining them together. Signed-off-by: Mike Christie Signed-off-by: Jens Axboe --- block/ll_rw_blk.c | 166 +++++++++++++++++++++++++++++++++++++------------ block/scsi_ioctl.c | 53 ++++++++-------- drivers/cdrom/cdrom.c | 6 +- fs/bio.c | 18 +----- include/linux/blkdev.h | 7 ++- 5 files changed, 160 insertions(+), 90 deletions(-) (limited to 'block') diff --git a/block/ll_rw_blk.c b/block/ll_rw_blk.c index 9eaee6640535..0f82e12f7b67 100644 --- a/block/ll_rw_blk.c +++ b/block/ll_rw_blk.c @@ -2322,6 +2322,84 @@ void blk_insert_request(request_queue_t *q, struct request *rq, EXPORT_SYMBOL(blk_insert_request); +static int __blk_rq_unmap_user(struct bio *bio) +{ + int ret = 0; + + if (bio) { + if (bio_flagged(bio, BIO_USER_MAPPED)) + bio_unmap_user(bio); + else + ret = bio_uncopy_user(bio); + } + + return ret; +} + +static int __blk_rq_map_user(request_queue_t *q, struct request *rq, + void __user *ubuf, unsigned int len) +{ + unsigned long uaddr; + struct bio *bio, *orig_bio; + int reading, ret; + + reading = rq_data_dir(rq) == READ; + + /* + * if alignment requirement is satisfied, map in user pages for + * direct dma. else, set up kernel bounce buffers + */ + uaddr = (unsigned long) ubuf; + if (!(uaddr & queue_dma_alignment(q)) && !(len & queue_dma_alignment(q))) + bio = bio_map_user(q, NULL, uaddr, len, reading); + else + bio = bio_copy_user(q, uaddr, len, reading); + + if (IS_ERR(bio)) { + return PTR_ERR(bio); + } + + orig_bio = bio; + blk_queue_bounce(q, &bio); + /* + * We link the bounce buffer in and could have to traverse it + * later so we have to get a ref to prevent it from being freed + */ + bio_get(bio); + + /* + * for most (all? don't know of any) queues we could + * skip grabbing the queue lock here. only drivers with + * funky private ->back_merge_fn() function could be + * problematic. + */ + spin_lock_irq(q->queue_lock); + if (!rq->bio) + blk_rq_bio_prep(q, rq, bio); + else if (!q->back_merge_fn(q, rq, bio)) { + ret = -EINVAL; + spin_unlock_irq(q->queue_lock); + goto unmap_bio; + } else { + rq->biotail->bi_next = bio; + rq->biotail = bio; + + rq->nr_sectors += bio_sectors(bio); + rq->hard_nr_sectors = rq->nr_sectors; + rq->data_len += bio->bi_size; + } + spin_unlock_irq(q->queue_lock); + + return bio->bi_size; + +unmap_bio: + /* if it was boucned we must call the end io function */ + bio_endio(bio, bio->bi_size, 0); + __blk_rq_unmap_user(orig_bio); + bio_put(bio); + return ret; +} + /** * blk_rq_map_user - map user data to a request, for REQ_BLOCK_PC usage * @q: request queue where request should be inserted @@ -2343,42 +2421,44 @@ EXPORT_SYMBOL(blk_insert_request); * unmapping. */ int blk_rq_map_user(request_queue_t *q, struct request *rq, void __user *ubuf, - unsigned int len) + unsigned long len) { - unsigned long uaddr; - struct bio *bio; - int reading; + unsigned long bytes_read = 0; + int ret; if (len > (q->max_hw_sectors << 9)) return -EINVAL; if (!len || !ubuf) return -EINVAL; - reading = rq_data_dir(rq) == READ; + while (bytes_read != len) { + unsigned long map_len, end, start; - /* - * if alignment requirement is satisfied, map in user pages for - * direct dma. else, set up kernel bounce buffers - */ - uaddr = (unsigned long) ubuf; - if (!(uaddr & queue_dma_alignment(q)) && !(len & queue_dma_alignment(q))) - bio = bio_map_user(q, NULL, uaddr, len, reading); - else - bio = bio_copy_user(q, uaddr, len, reading); + map_len = min_t(unsigned long, len - bytes_read, BIO_MAX_SIZE); + end = ((unsigned long)ubuf + map_len + PAGE_SIZE - 1) + >> PAGE_SHIFT; + start = (unsigned long)ubuf >> PAGE_SHIFT; - if (!IS_ERR(bio)) { - rq->bio = rq->biotail = bio; - blk_rq_bio_prep(q, rq, bio); + /* + * A bad offset could cause us to require BIO_MAX_PAGES + 1 + * pages. If this happens we just lower the requested + * mapping len by a page so that we can fit + */ + if (end - start > BIO_MAX_PAGES) + map_len -= PAGE_SIZE; - rq->buffer = rq->data = NULL; - rq->data_len = len; - return 0; + ret = __blk_rq_map_user(q, rq, ubuf, map_len); + if (ret < 0) + goto unmap_rq; + bytes_read += ret; + ubuf += ret; } - /* - * bio is the err-ptr - */ - return PTR_ERR(bio); + rq->buffer = rq->data = NULL; + return 0; +unmap_rq: + blk_rq_unmap_user(rq); + return ret; } EXPORT_SYMBOL(blk_rq_map_user); @@ -2404,7 +2484,7 @@ EXPORT_SYMBOL(blk_rq_map_user); * unmapping. */ int blk_rq_map_user_iov(request_queue_t *q, struct request *rq, - struct sg_iovec *iov, int iov_count) + struct sg_iovec *iov, int iov_count, unsigned int len) { struct bio *bio; @@ -2418,10 +2498,15 @@ int blk_rq_map_user_iov(request_queue_t *q, struct request *rq, if (IS_ERR(bio)) return PTR_ERR(bio); - rq->bio = rq->biotail = bio; + if (bio->bi_size != len) { + bio_endio(bio, bio->bi_size, 0); + bio_unmap_user(bio); + return -EINVAL; + } + + bio_get(bio); blk_rq_bio_prep(q, rq, bio); rq->buffer = rq->data = NULL; - rq->data_len = bio->bi_size; return 0; } @@ -2429,23 +2514,26 @@ EXPORT_SYMBOL(blk_rq_map_user_iov); /** * blk_rq_unmap_user - unmap a request with user data - * @bio: bio to be unmapped - * @ulen: length of user buffer + * @rq: rq to be unmapped * * Description: - * Unmap a bio previously mapped by blk_rq_map_user(). + * Unmap a rq previously mapped by blk_rq_map_user(). + * rq->bio must be set to the original head of the request. */ -int blk_rq_unmap_user(struct bio *bio, unsigned int ulen) +int blk_rq_unmap_user(struct request *rq) { - int ret = 0; + struct bio *bio, *mapped_bio; - if (bio) { - if (bio_flagged(bio, BIO_USER_MAPPED)) - bio_unmap_user(bio); + while ((bio = rq->bio)) { + if (bio_flagged(bio, BIO_BOUNCED)) + mapped_bio = bio->bi_private; else - ret = bio_uncopy_user(bio); - } + mapped_bio = bio; + __blk_rq_unmap_user(mapped_bio); + rq->bio = bio->bi_next; + bio_put(bio); + } return 0; } @@ -2476,11 +2564,8 @@ int blk_rq_map_kern(request_queue_t *q, struct request *rq, void *kbuf, if (rq_data_dir(rq) == WRITE) bio->bi_rw |= (1 << BIO_RW); - rq->bio = rq->biotail = bio; blk_rq_bio_prep(q, rq, bio); - rq->buffer = rq->data = NULL; - rq->data_len = len; return 0; } @@ -3495,6 +3580,7 @@ void blk_rq_bio_prep(request_queue_t *q, struct request *rq, struct bio *bio) rq->hard_cur_sectors = rq->current_nr_sectors; rq->hard_nr_sectors = rq->nr_sectors = bio_sectors(bio); rq->buffer = bio_data(bio); + rq->data_len = bio->bi_size; rq->bio = rq->biotail = bio; } diff --git a/block/scsi_ioctl.c b/block/scsi_ioctl.c index e55a75621437..5493c2fbbab1 100644 --- a/block/scsi_ioctl.c +++ b/block/scsi_ioctl.c @@ -226,7 +226,6 @@ static int sg_io(struct file *file, request_queue_t *q, unsigned long start_time; int writing = 0, ret = 0; struct request *rq; - struct bio *bio; char sense[SCSI_SENSE_BUFFERSIZE]; unsigned char cmd[BLK_MAX_CDB]; @@ -258,30 +257,6 @@ static int sg_io(struct file *file, request_queue_t *q, if (!rq) return -ENOMEM; - if (hdr->iovec_count) { - const int size = sizeof(struct sg_iovec) * hdr->iovec_count; - struct sg_iovec *iov; - - iov = kmalloc(size, GFP_KERNEL); - if (!iov) { - ret = -ENOMEM; - goto out; - } - - if (copy_from_user(iov, hdr->dxferp, size)) { - kfree(iov); - ret = -EFAULT; - goto out; - } - - ret = blk_rq_map_user_iov(q, rq, iov, hdr->iovec_count); - kfree(iov); - } else if (hdr->dxfer_len) - ret = blk_rq_map_user(q, rq, hdr->dxferp, hdr->dxfer_len); - - if (ret) - goto out; - /* * fill in request structure */ @@ -294,7 +269,6 @@ static int sg_io(struct file *file, request_queue_t *q, rq->sense_len = 0; rq->cmd_type = REQ_TYPE_BLOCK_PC; - bio = rq->bio; /* * bounce this after holding a reference to the original bio, it's @@ -309,6 +283,31 @@ static int sg_io(struct file *file, request_queue_t *q, if (!rq->timeout) rq->timeout = BLK_DEFAULT_TIMEOUT; + if (hdr->iovec_count) { + const int size = sizeof(struct sg_iovec) * hdr->iovec_count; + struct sg_iovec *iov; + + iov = kmalloc(size, GFP_KERNEL); + if (!iov) { + ret = -ENOMEM; + goto out; + } + + if (copy_from_user(iov, hdr->dxferp, size)) { + kfree(iov); + ret = -EFAULT; + goto out; + } + + ret = blk_rq_map_user_iov(q, rq, iov, hdr->iovec_count, + hdr->dxfer_len); + kfree(iov); + } else if (hdr->dxfer_len) + ret = blk_rq_map_user(q, rq, hdr->dxferp, hdr->dxfer_len); + + if (ret) + goto out; + rq->retries = 0; start_time = jiffies; @@ -339,7 +338,7 @@ static int sg_io(struct file *file, request_queue_t *q, hdr->sb_len_wr = len; } - if (blk_rq_unmap_user(bio, hdr->dxfer_len)) + if (blk_rq_unmap_user(rq)) ret = -EFAULT; /* may not have succeeded, but output values written to control diff --git a/drivers/cdrom/cdrom.c b/drivers/cdrom/cdrom.c index 7ea0f48f8fa6..2df5cf4ec743 100644 --- a/drivers/cdrom/cdrom.c +++ b/drivers/cdrom/cdrom.c @@ -2133,16 +2133,14 @@ static int cdrom_read_cdda_bpc(struct cdrom_device_info *cdi, __u8 __user *ubuf, rq->timeout = 60 * HZ; bio = rq->bio; - if (rq->bio) - blk_queue_bounce(q, &rq->bio); - if (blk_execute_rq(q, cdi->disk, rq, 0)) { struct request_sense *s = rq->sense; ret = -EIO; cdi->last_sense = s->sense_key; } - if (blk_rq_unmap_user(bio, len)) + rq->bio = bio; + if (blk_rq_unmap_user(rq)) ret = -EFAULT; if (ret) diff --git a/fs/bio.c b/fs/bio.c index d91cfbf7ebc4..aa4d09bd4e71 100644 --- a/fs/bio.c +++ b/fs/bio.c @@ -560,10 +560,8 @@ struct bio *bio_copy_user(request_queue_t *q, unsigned long uaddr, break; } - if (bio_add_pc_page(q, bio, page, bytes, 0) < bytes) { - ret = -EINVAL; + if (bio_add_pc_page(q, bio, page, bytes, 0) < bytes) break; - } len -= bytes; } @@ -750,7 +748,6 @@ struct bio *bio_map_user_iov(request_queue_t *q, struct block_device *bdev, int write_to_vm) { struct bio *bio; - int len = 0, i; bio = __bio_map_user_iov(q, bdev, iov, iov_count, write_to_vm); @@ -765,18 +762,7 @@ struct bio *bio_map_user_iov(request_queue_t *q, struct block_device *bdev, */ bio_get(bio); - for (i = 0; i < iov_count; i++) - len += iov[i].iov_len; - - if (bio->bi_size == len) - return bio; - - /* - * don't support partial mappings - */ - bio_endio(bio, bio->bi_size, 0); - bio_unmap_user(bio); - return ERR_PTR(-EINVAL); + return bio; } static void __bio_unmap_user(struct bio *bio) diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h index 7bfcde2d5578..e1c7286165ff 100644 --- a/include/linux/blkdev.h +++ b/include/linux/blkdev.h @@ -678,10 +678,11 @@ extern void __blk_stop_queue(request_queue_t *q); extern void blk_run_queue(request_queue_t *); extern void blk_start_queueing(request_queue_t *); extern void blk_queue_activity_fn(request_queue_t *, activity_fn *, void *); -extern int blk_rq_map_user(request_queue_t *, struct request *, void __user *, unsigned int); -extern int blk_rq_unmap_user(struct bio *, unsigned int); +extern int blk_rq_map_user(request_queue_t *, struct request *, void __user *, unsigned long); +extern int blk_rq_unmap_user(struct request *); extern int blk_rq_map_kern(request_queue_t *, struct request *, void *, unsigned int, gfp_t); -extern int blk_rq_map_user_iov(request_queue_t *, struct request *, struct sg_iovec *, int); +extern int blk_rq_map_user_iov(request_queue_t *, struct request *, + struct sg_iovec *, int, unsigned int); extern int blk_execute_rq(request_queue_t *, struct gendisk *, struct request *, int); extern void blk_execute_rq_nowait(request_queue_t *, struct gendisk *, -- cgit v1.2.3 From bb37b94c68e7b37eecea8576039ae9396ca07839 Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Fri, 1 Dec 2006 10:42:33 +0100 Subject: [BLOCK] Cleanup unused variable passing - ->init_queue() does not need the elevator passed in - ->put_request() is a hot path and need not have the queue passed in - cfq_update_io_seektime() does not need cfqd passed in Signed-off-by: Jens Axboe --- block/as-iosched.c | 2 +- block/cfq-iosched.c | 9 ++++----- block/deadline-iosched.c | 2 +- block/elevator.c | 4 ++-- block/noop-iosched.c | 2 +- include/linux/elevator.h | 4 ++-- 6 files changed, 11 insertions(+), 12 deletions(-) (limited to 'block') diff --git a/block/as-iosched.c b/block/as-iosched.c index 50b95e4c1425..00242111a457 100644 --- a/block/as-iosched.c +++ b/block/as-iosched.c @@ -1317,7 +1317,7 @@ static void as_exit_queue(elevator_t *e) /* * initialize elevator private data (as_data). */ -static void *as_init_queue(request_queue_t *q, elevator_t *e) +static void *as_init_queue(request_queue_t *q) { struct as_data *ad; diff --git a/block/cfq-iosched.c b/block/cfq-iosched.c index 1d9c3c70a9a0..e9019ed39b73 100644 --- a/block/cfq-iosched.c +++ b/block/cfq-iosched.c @@ -1464,8 +1464,7 @@ cfq_update_io_thinktime(struct cfq_data *cfqd, struct cfq_io_context *cic) } static void -cfq_update_io_seektime(struct cfq_data *cfqd, struct cfq_io_context *cic, - struct request *rq) +cfq_update_io_seektime(struct cfq_io_context *cic, struct request *rq) { sector_t sdist; u64 total; @@ -1617,7 +1616,7 @@ cfq_rq_enqueued(struct cfq_data *cfqd, struct cfq_queue *cfqq, } cfq_update_io_thinktime(cfqd, cic); - cfq_update_io_seektime(cfqd, cic, rq); + cfq_update_io_seektime(cic, rq); cfq_update_idle_window(cfqd, cfqq, cic); cic->last_queue = jiffies; @@ -1770,7 +1769,7 @@ static int cfq_may_queue(request_queue_t *q, int rw) /* * queue lock held here */ -static void cfq_put_request(request_queue_t *q, struct request *rq) +static void cfq_put_request(struct request *rq) { struct cfq_queue *cfqq = RQ_CFQQ(rq); @@ -1951,7 +1950,7 @@ static void cfq_exit_queue(elevator_t *e) kfree(cfqd); } -static void *cfq_init_queue(request_queue_t *q, elevator_t *e) +static void *cfq_init_queue(request_queue_t *q) { struct cfq_data *cfqd; int i; diff --git a/block/deadline-iosched.c b/block/deadline-iosched.c index b7c5b34cb7b4..6d673e938d3e 100644 --- a/block/deadline-iosched.c +++ b/block/deadline-iosched.c @@ -356,7 +356,7 @@ static void deadline_exit_queue(elevator_t *e) /* * initialize elevator private data (deadline_data). */ -static void *deadline_init_queue(request_queue_t *q, elevator_t *e) +static void *deadline_init_queue(request_queue_t *q) { struct deadline_data *dd; diff --git a/block/elevator.c b/block/elevator.c index 8ccd163254b8..c0063f345c5d 100644 --- a/block/elevator.c +++ b/block/elevator.c @@ -129,7 +129,7 @@ static struct elevator_type *elevator_get(const char *name) static void *elevator_init_queue(request_queue_t *q, struct elevator_queue *eq) { - return eq->ops->elevator_init_fn(q, eq); + return eq->ops->elevator_init_fn(q); } static void elevator_attach(request_queue_t *q, struct elevator_queue *eq, @@ -810,7 +810,7 @@ void elv_put_request(request_queue_t *q, struct request *rq) elevator_t *e = q->elevator; if (e->ops->elevator_put_req_fn) - e->ops->elevator_put_req_fn(q, rq); + e->ops->elevator_put_req_fn(rq); } int elv_may_queue(request_queue_t *q, int rw) diff --git a/block/noop-iosched.c b/block/noop-iosched.c index 79af43179421..1c3de2b9a6b5 100644 --- a/block/noop-iosched.c +++ b/block/noop-iosched.c @@ -65,7 +65,7 @@ noop_latter_request(request_queue_t *q, struct request *rq) return list_entry(rq->queuelist.next, struct request, queuelist); } -static void *noop_init_queue(request_queue_t *q, elevator_t *e) +static void *noop_init_queue(request_queue_t *q) { struct noop_data *nd; diff --git a/include/linux/elevator.h b/include/linux/elevator.h index 2fa9f1144228..a24931d24404 100644 --- a/include/linux/elevator.h +++ b/include/linux/elevator.h @@ -21,11 +21,11 @@ typedef void (elevator_completed_req_fn) (request_queue_t *, struct request *); typedef int (elevator_may_queue_fn) (request_queue_t *, int); typedef int (elevator_set_req_fn) (request_queue_t *, struct request *, gfp_t); -typedef void (elevator_put_req_fn) (request_queue_t *, struct request *); +typedef void (elevator_put_req_fn) (struct request *); typedef void (elevator_activate_req_fn) (request_queue_t *, struct request *); typedef void (elevator_deactivate_req_fn) (request_queue_t *, struct request *); -typedef void *(elevator_init_fn) (request_queue_t *, elevator_t *); +typedef void *(elevator_init_fn) (request_queue_t *); typedef void (elevator_exit_fn) (elevator_t *); struct elevator_ops -- cgit v1.2.3