From 1200e07f3ad4b9d976cf2fff3a0c3d9a1faecb3e Mon Sep 17 00:00:00 2001 From: Ming Lei Date: Mon, 8 Apr 2019 19:02:38 +0800 Subject: block: don't use for-inside-for in bio_for_each_segment_all Commit 6dc4f100c175 ("block: allow bio_for_each_segment_all() to iterate over multi-page bvec") changes bio_for_each_segment_all() to use for-inside-for. This way breaks all bio_for_each_segment_all() call with error out branch via 'break', since now 'break' can only break from the inner loop. Fixes this issue by implementing bio_for_each_segment_all() via single 'for' loop, and now the logic is very similar with normal bvec iterator. Cc: Qu Wenruo Cc: linux-btrfs@vger.kernel.org Cc: linux-fsdevel@vger.kernel.org Cc: Omar Sandoval Reviewed-by: Johannes Thumshirn Reported-and-Tested-by: Qu Wenruo Fixes: 6dc4f100c175 ("block: allow bio_for_each_segment_all() to iterate over multi-page bvec") Reviewed-by: Christoph Hellwig Signed-off-by: Ming Lei Signed-off-by: Jens Axboe --- include/linux/bio.h | 20 ++++++++++++-------- include/linux/bvec.h | 14 ++++++++++---- 2 files changed, 22 insertions(+), 12 deletions(-) (limited to 'include') diff --git a/include/linux/bio.h b/include/linux/bio.h index bb6090aa165d..e584673c1881 100644 --- a/include/linux/bio.h +++ b/include/linux/bio.h @@ -120,19 +120,23 @@ static inline bool bio_full(struct bio *bio) return bio->bi_vcnt >= bio->bi_max_vecs; } -#define mp_bvec_for_each_segment(bv, bvl, i, iter_all) \ - for (bv = bvec_init_iter_all(&iter_all); \ - (iter_all.done < (bvl)->bv_len) && \ - (mp_bvec_next_segment((bvl), &iter_all), 1); \ - iter_all.done += bv->bv_len, i += 1) +static inline bool bio_next_segment(const struct bio *bio, + struct bvec_iter_all *iter) +{ + if (iter->idx >= bio->bi_vcnt) + return false; + + bvec_advance(&bio->bi_io_vec[iter->idx], iter); + return true; +} /* * drivers should _never_ use the all version - the bio may have been split * before it got to the driver and the driver won't own all of it */ -#define bio_for_each_segment_all(bvl, bio, i, iter_all) \ - for (i = 0, iter_all.idx = 0; iter_all.idx < (bio)->bi_vcnt; iter_all.idx++) \ - mp_bvec_for_each_segment(bvl, &((bio)->bi_io_vec[iter_all.idx]), i, iter_all) +#define bio_for_each_segment_all(bvl, bio, i, iter) \ + for (i = 0, bvl = bvec_init_iter_all(&iter); \ + bio_next_segment((bio), &iter); i++) static inline void bio_advance_iter(struct bio *bio, struct bvec_iter *iter, unsigned bytes) diff --git a/include/linux/bvec.h b/include/linux/bvec.h index f6275c4da13a..3bc91879e1e2 100644 --- a/include/linux/bvec.h +++ b/include/linux/bvec.h @@ -145,18 +145,18 @@ static inline bool bvec_iter_advance(const struct bio_vec *bv, static inline struct bio_vec *bvec_init_iter_all(struct bvec_iter_all *iter_all) { - iter_all->bv.bv_page = NULL; iter_all->done = 0; + iter_all->idx = 0; return &iter_all->bv; } -static inline void mp_bvec_next_segment(const struct bio_vec *bvec, - struct bvec_iter_all *iter_all) +static inline void bvec_advance(const struct bio_vec *bvec, + struct bvec_iter_all *iter_all) { struct bio_vec *bv = &iter_all->bv; - if (bv->bv_page) { + if (iter_all->done) { bv->bv_page = nth_page(bv->bv_page, 1); bv->bv_offset = 0; } else { @@ -165,6 +165,12 @@ static inline void mp_bvec_next_segment(const struct bio_vec *bvec, } bv->bv_len = min_t(unsigned int, PAGE_SIZE - bv->bv_offset, bvec->bv_len - iter_all->done); + iter_all->done += bv->bv_len; + + if (iter_all->done == bvec->bv_len) { + iter_all->idx++; + iter_all->done = 0; + } } /* -- cgit v1.2.3 From 1b8f21b74c3c9c82fce5a751d7aefb7cc0b8d33d Mon Sep 17 00:00:00 2001 From: Ming Lei Date: Tue, 9 Apr 2019 06:31:21 +0800 Subject: blk-mq: introduce blk_mq_complete_request_sync() In NVMe's error handler, follows the typical steps of tearing down hardware for recovering controller: 1) stop blk_mq hw queues 2) stop the real hw queues 3) cancel in-flight requests via blk_mq_tagset_busy_iter(tags, cancel_request, ...) cancel_request(): mark the request as abort blk_mq_complete_request(req); 4) destroy real hw queues However, there may be race between #3 and #4, because blk_mq_complete_request() may run q->mq_ops->complete(rq) remotelly and asynchronously, and ->complete(rq) may be run after #4. This patch introduces blk_mq_complete_request_sync() for fixing the above race. Cc: Sagi Grimberg Cc: Bart Van Assche Cc: James Smart Cc: linux-nvme@lists.infradead.org Reviewed-by: Keith Busch Reviewed-by: Christoph Hellwig Signed-off-by: Ming Lei Signed-off-by: Jens Axboe --- block/blk-mq.c | 7 +++++++ include/linux/blk-mq.h | 1 + 2 files changed, 8 insertions(+) (limited to 'include') diff --git a/block/blk-mq.c b/block/blk-mq.c index a9354835cf51..9516304a38ee 100644 --- a/block/blk-mq.c +++ b/block/blk-mq.c @@ -654,6 +654,13 @@ bool blk_mq_complete_request(struct request *rq) } EXPORT_SYMBOL(blk_mq_complete_request); +void blk_mq_complete_request_sync(struct request *rq) +{ + WRITE_ONCE(rq->state, MQ_RQ_COMPLETE); + rq->q->mq_ops->complete(rq); +} +EXPORT_SYMBOL_GPL(blk_mq_complete_request_sync); + int blk_mq_request_started(struct request *rq) { return blk_mq_rq_state(rq) != MQ_RQ_IDLE; diff --git a/include/linux/blk-mq.h b/include/linux/blk-mq.h index cb2aa7ecafff..db29928de467 100644 --- a/include/linux/blk-mq.h +++ b/include/linux/blk-mq.h @@ -302,6 +302,7 @@ void blk_mq_requeue_request(struct request *rq, bool kick_requeue_list); void blk_mq_kick_requeue_list(struct request_queue *q); void blk_mq_delay_kick_requeue_list(struct request_queue *q, unsigned long msecs); bool blk_mq_complete_request(struct request *rq); +void blk_mq_complete_request_sync(struct request *rq); bool blk_mq_bio_list_merge(struct request_queue *q, struct list_head *list, struct bio *bio); bool blk_mq_queue_stopped(struct request_queue *q); -- cgit v1.2.3 From d808b7f759b50acf0784ce6230ffa63e12ef465d Mon Sep 17 00:00:00 2001 From: Keith Busch Date: Tue, 9 Apr 2019 10:03:59 -0600 Subject: nvmet: fix discover log page when offsets are used The nvme target hadn't been taking the Get Log Page offset parameter into consideration, and so has been returning corrupted log pages when offsets are used. Since many tools, including nvme-cli, split the log request to 4k, we've been breaking discovery log responses when more than 3 subsystems exist. Fix the returned data by internally generating the entire discovery log page and copying only the requested bytes into the user buffer. The command log page offset type has been modified to a native __le64 to make it easier to extract the value from a command. Signed-off-by: Keith Busch Tested-by: Minwoo Im Reviewed-by: Chaitanya Kulkarni Reviewed-by: Hannes Reinecke Reviewed-by: James Smart Signed-off-by: Christoph Hellwig --- drivers/nvme/target/admin-cmd.c | 5 +++ drivers/nvme/target/discovery.c | 68 +++++++++++++++++++++++++++-------------- drivers/nvme/target/nvmet.h | 1 + include/linux/nvme.h | 9 ++++-- 4 files changed, 58 insertions(+), 25 deletions(-) (limited to 'include') diff --git a/drivers/nvme/target/admin-cmd.c b/drivers/nvme/target/admin-cmd.c index 76250181fee0..9f72d515fc4b 100644 --- a/drivers/nvme/target/admin-cmd.c +++ b/drivers/nvme/target/admin-cmd.c @@ -24,6 +24,11 @@ u32 nvmet_get_log_page_len(struct nvme_command *cmd) return len; } +u64 nvmet_get_log_page_offset(struct nvme_command *cmd) +{ + return le64_to_cpu(cmd->get_log_page.lpo); +} + static void nvmet_execute_get_log_page_noop(struct nvmet_req *req) { nvmet_req_complete(req, nvmet_zero_sgl(req, 0, req->data_len)); diff --git a/drivers/nvme/target/discovery.c b/drivers/nvme/target/discovery.c index c872b47a88f3..33ed95e72d6b 100644 --- a/drivers/nvme/target/discovery.c +++ b/drivers/nvme/target/discovery.c @@ -131,54 +131,76 @@ static void nvmet_set_disc_traddr(struct nvmet_req *req, struct nvmet_port *port memcpy(traddr, port->disc_addr.traddr, NVMF_TRADDR_SIZE); } +static size_t discovery_log_entries(struct nvmet_req *req) +{ + struct nvmet_ctrl *ctrl = req->sq->ctrl; + struct nvmet_subsys_link *p; + struct nvmet_port *r; + size_t entries = 0; + + list_for_each_entry(p, &req->port->subsystems, entry) { + if (!nvmet_host_allowed(p->subsys, ctrl->hostnqn)) + continue; + entries++; + } + list_for_each_entry(r, &req->port->referrals, entry) + entries++; + return entries; +} + static void nvmet_execute_get_disc_log_page(struct nvmet_req *req) { const int entry_size = sizeof(struct nvmf_disc_rsp_page_entry); struct nvmet_ctrl *ctrl = req->sq->ctrl; struct nvmf_disc_rsp_page_hdr *hdr; + u64 offset = nvmet_get_log_page_offset(req->cmd); size_t data_len = nvmet_get_log_page_len(req->cmd); - size_t alloc_len = max(data_len, sizeof(*hdr)); - int residual_len = data_len - sizeof(*hdr); + size_t alloc_len; struct nvmet_subsys_link *p; struct nvmet_port *r; u32 numrec = 0; u16 status = 0; + void *buffer; + + /* Spec requires dword aligned offsets */ + if (offset & 0x3) { + status = NVME_SC_INVALID_FIELD | NVME_SC_DNR; + goto out; + } /* * Make sure we're passing at least a buffer of response header size. * If host provided data len is less than the header size, only the * number of bytes requested by host will be sent to host. */ - hdr = kzalloc(alloc_len, GFP_KERNEL); - if (!hdr) { + down_read(&nvmet_config_sem); + alloc_len = sizeof(*hdr) + entry_size * discovery_log_entries(req); + buffer = kzalloc(alloc_len, GFP_KERNEL); + if (!buffer) { + up_read(&nvmet_config_sem); status = NVME_SC_INTERNAL; goto out; } - down_read(&nvmet_config_sem); + hdr = buffer; list_for_each_entry(p, &req->port->subsystems, entry) { + char traddr[NVMF_TRADDR_SIZE]; + if (!nvmet_host_allowed(p->subsys, ctrl->hostnqn)) continue; - if (residual_len >= entry_size) { - char traddr[NVMF_TRADDR_SIZE]; - - nvmet_set_disc_traddr(req, req->port, traddr); - nvmet_format_discovery_entry(hdr, req->port, - p->subsys->subsysnqn, traddr, - NVME_NQN_NVME, numrec); - residual_len -= entry_size; - } + + nvmet_set_disc_traddr(req, req->port, traddr); + nvmet_format_discovery_entry(hdr, req->port, + p->subsys->subsysnqn, traddr, + NVME_NQN_NVME, numrec); numrec++; } list_for_each_entry(r, &req->port->referrals, entry) { - if (residual_len >= entry_size) { - nvmet_format_discovery_entry(hdr, r, - NVME_DISC_SUBSYS_NAME, - r->disc_addr.traddr, - NVME_NQN_DISC, numrec); - residual_len -= entry_size; - } + nvmet_format_discovery_entry(hdr, r, + NVME_DISC_SUBSYS_NAME, + r->disc_addr.traddr, + NVME_NQN_DISC, numrec); numrec++; } @@ -190,8 +212,8 @@ static void nvmet_execute_get_disc_log_page(struct nvmet_req *req) up_read(&nvmet_config_sem); - status = nvmet_copy_to_sgl(req, 0, hdr, data_len); - kfree(hdr); + status = nvmet_copy_to_sgl(req, 0, buffer + offset, data_len); + kfree(buffer); out: nvmet_req_complete(req, status); } diff --git a/drivers/nvme/target/nvmet.h b/drivers/nvme/target/nvmet.h index 51e49efd7849..1653d19b187f 100644 --- a/drivers/nvme/target/nvmet.h +++ b/drivers/nvme/target/nvmet.h @@ -428,6 +428,7 @@ u16 nvmet_copy_from_sgl(struct nvmet_req *req, off_t off, void *buf, u16 nvmet_zero_sgl(struct nvmet_req *req, off_t off, size_t len); u32 nvmet_get_log_page_len(struct nvme_command *cmd); +u64 nvmet_get_log_page_offset(struct nvme_command *cmd); extern struct list_head *nvmet_ports; void nvmet_port_disc_changed(struct nvmet_port *port, diff --git a/include/linux/nvme.h b/include/linux/nvme.h index baa49e6a23cc..c40720cb59ac 100644 --- a/include/linux/nvme.h +++ b/include/linux/nvme.h @@ -967,8 +967,13 @@ struct nvme_get_log_page_command { __le16 numdl; __le16 numdu; __u16 rsvd11; - __le32 lpol; - __le32 lpou; + union { + struct { + __le32 lpol; + __le32 lpou; + }; + __le64 lpo; + }; __u32 rsvd14[2]; }; -- cgit v1.2.3