From 6e02318eaea53eaafe628c4ffc254f57b2704561 Mon Sep 17 00:00:00 2001 From: Chaitanya Kulkarni Date: Mon, 17 Dec 2018 22:42:03 -0500 Subject: nvme: add support for the Write Zeroes command Allow write zeroes operations (REQ_OP_WRITE_ZEROES) on the block device, if the device supports an optional command bit set for write zeroes. Add support to setup write zeroes command. Set maximum possible write zeroes sectors in one write zeroes command according to nvme write zeroes command definition. This patch was posted as a part of block-write-zeroes support implementation (https://patchwork.kernel.org/patch/9454859/), but did not make into mainline kernel as it got reverted due to failure on the Linus's machine. In this patch in order to be more cautious, we use NVMe controller's maximum hardware sector size which is calculated based on the controller's MDTS (Maximum Data Transfer Size) field to calculate the maximum sectors for the write zeroes request. Signed-off-by: Chaitanya Kulkarni Reviewed-by: Martin K. Petersen [folded a fix from Keith Busch to properly respect NVME_QUIRK_DEALLOCATE_ZEROES] Signed-off-by: Christoph Hellwig --- drivers/nvme/host/core.c | 52 ++++++++++++++++++++++++++++++++++++++++++++++-- 1 file changed, 50 insertions(+), 2 deletions(-) diff --git a/drivers/nvme/host/core.c b/drivers/nvme/host/core.c index 150e49723c15..5c2d2f1e7261 100644 --- a/drivers/nvme/host/core.c +++ b/drivers/nvme/host/core.c @@ -611,6 +611,22 @@ static blk_status_t nvme_setup_discard(struct nvme_ns *ns, struct request *req, return BLK_STS_OK; } +static inline blk_status_t nvme_setup_write_zeroes(struct nvme_ns *ns, + struct request *req, struct nvme_command *cmnd) +{ + if (ns->ctrl->quirks & NVME_QUIRK_DEALLOCATE_ZEROES) + return nvme_setup_discard(ns, req, cmnd); + + cmnd->write_zeroes.opcode = nvme_cmd_write_zeroes; + cmnd->write_zeroes.nsid = cpu_to_le32(ns->head->ns_id); + cmnd->write_zeroes.slba = + cpu_to_le64(nvme_block_nr(ns, blk_rq_pos(req))); + cmnd->write_zeroes.length = + cpu_to_le16((blk_rq_bytes(req) >> ns->lba_shift) - 1); + cmnd->write_zeroes.control = 0; + return BLK_STS_OK; +} + static inline blk_status_t nvme_setup_rw(struct nvme_ns *ns, struct request *req, struct nvme_command *cmnd) { @@ -705,7 +721,8 @@ blk_status_t nvme_setup_cmd(struct nvme_ns *ns, struct request *req, nvme_setup_flush(ns, cmd); break; case REQ_OP_WRITE_ZEROES: - /* currently only aliased to deallocate for a few ctrls: */ + ret = nvme_setup_write_zeroes(ns, req, cmd); + break; case REQ_OP_DISCARD: ret = nvme_setup_discard(ns, req, cmd); break; @@ -1509,6 +1526,37 @@ static void nvme_config_discard(struct nvme_ns *ns) blk_queue_max_write_zeroes_sectors(queue, UINT_MAX); } +static inline void nvme_config_write_zeroes(struct nvme_ns *ns) +{ + u32 max_sectors; + unsigned short bs = 1 << ns->lba_shift; + + if (!(ns->ctrl->oncs & NVME_CTRL_ONCS_WRITE_ZEROES)) + return; + /* + * Even though NVMe spec explicitly states that MDTS is not + * applicable to the write-zeroes:- "The restriction does not apply to + * commands that do not transfer data between the host and the + * controller (e.g., Write Uncorrectable ro Write Zeroes command).". + * In order to be more cautious use controller's max_hw_sectors value + * to configure the maximum sectors for the write-zeroes which is + * configured based on the controller's MDTS field in the + * nvme_init_identify() if available. + */ + if (ns->ctrl->max_hw_sectors == UINT_MAX) + max_sectors = ((u32)(USHRT_MAX + 1) * bs) >> 9; + else + max_sectors = ((u32)(ns->ctrl->max_hw_sectors + 1) * bs) >> 9; + + blk_queue_max_write_zeroes_sectors(ns->queue, max_sectors); +} + +static inline void nvme_ns_config_oncs(struct nvme_ns *ns) +{ + nvme_config_discard(ns); + nvme_config_write_zeroes(ns); +} + static void nvme_report_ns_ids(struct nvme_ctrl *ctrl, unsigned int nsid, struct nvme_id_ns *id, struct nvme_ns_ids *ids) { @@ -1562,7 +1610,7 @@ static void nvme_update_disk_info(struct gendisk *disk, capacity = 0; set_capacity(disk, capacity); - nvme_config_discard(ns); + nvme_ns_config_oncs(ns); if (id->nsattr & (1 << 0)) set_disk_ro(disk, true); -- cgit v1.2.3 From 794a4cb3d2f7c105112f9b59525533f00ac06006 Mon Sep 17 00:00:00 2001 From: Sagi Grimberg Date: Tue, 1 Jan 2019 00:19:30 -0800 Subject: nvme: remove the .stop_ctrl callout It is used now just to flush error recovery and reconnect work items in the RDMA and TCP transports, which can simply be moved to the corresponding teardown routines. Signed-off-by: Sagi Grimberg Signed-off-by: Christoph Hellwig --- drivers/nvme/host/core.c | 2 -- drivers/nvme/host/nvme.h | 1 - drivers/nvme/host/rdma.c | 12 +++--------- drivers/nvme/host/tcp.c | 10 +++------- 4 files changed, 6 insertions(+), 19 deletions(-) diff --git a/drivers/nvme/host/core.c b/drivers/nvme/host/core.c index 5c2d2f1e7261..c365f0aa9433 100644 --- a/drivers/nvme/host/core.c +++ b/drivers/nvme/host/core.c @@ -3639,8 +3639,6 @@ void nvme_stop_ctrl(struct nvme_ctrl *ctrl) nvme_stop_keep_alive(ctrl); flush_work(&ctrl->async_event_work); cancel_work_sync(&ctrl->fw_act_work); - if (ctrl->ops->stop_ctrl) - ctrl->ops->stop_ctrl(ctrl); } EXPORT_SYMBOL_GPL(nvme_stop_ctrl); diff --git a/drivers/nvme/host/nvme.h b/drivers/nvme/host/nvme.h index ab961bdeea89..a0cc733c753e 100644 --- a/drivers/nvme/host/nvme.h +++ b/drivers/nvme/host/nvme.h @@ -363,7 +363,6 @@ struct nvme_ctrl_ops { void (*submit_async_event)(struct nvme_ctrl *ctrl); void (*delete_ctrl)(struct nvme_ctrl *ctrl); int (*get_address)(struct nvme_ctrl *ctrl, char *buf, int size); - void (*stop_ctrl)(struct nvme_ctrl *ctrl); }; #ifdef CONFIG_FAULT_INJECTION_DEBUG_FS diff --git a/drivers/nvme/host/rdma.c b/drivers/nvme/host/rdma.c index 52abc3a6de12..ac365366c2ec 100644 --- a/drivers/nvme/host/rdma.c +++ b/drivers/nvme/host/rdma.c @@ -942,14 +942,6 @@ static void nvme_rdma_teardown_io_queues(struct nvme_rdma_ctrl *ctrl, } } -static void nvme_rdma_stop_ctrl(struct nvme_ctrl *nctrl) -{ - struct nvme_rdma_ctrl *ctrl = to_rdma_ctrl(nctrl); - - cancel_work_sync(&ctrl->err_work); - cancel_delayed_work_sync(&ctrl->reconnect_work); -} - static void nvme_rdma_free_ctrl(struct nvme_ctrl *nctrl) { struct nvme_rdma_ctrl *ctrl = to_rdma_ctrl(nctrl); @@ -1854,6 +1846,9 @@ static const struct blk_mq_ops nvme_rdma_admin_mq_ops = { static void nvme_rdma_shutdown_ctrl(struct nvme_rdma_ctrl *ctrl, bool shutdown) { + cancel_work_sync(&ctrl->err_work); + cancel_delayed_work_sync(&ctrl->reconnect_work); + nvme_rdma_teardown_io_queues(ctrl, shutdown); if (shutdown) nvme_shutdown_ctrl(&ctrl->ctrl); @@ -1902,7 +1897,6 @@ static const struct nvme_ctrl_ops nvme_rdma_ctrl_ops = { .submit_async_event = nvme_rdma_submit_async_event, .delete_ctrl = nvme_rdma_delete_ctrl, .get_address = nvmf_get_address, - .stop_ctrl = nvme_rdma_stop_ctrl, }; /* diff --git a/drivers/nvme/host/tcp.c b/drivers/nvme/host/tcp.c index 5f0a00425242..208ee518af65 100644 --- a/drivers/nvme/host/tcp.c +++ b/drivers/nvme/host/tcp.c @@ -1822,6 +1822,9 @@ static void nvme_tcp_error_recovery_work(struct work_struct *work) static void nvme_tcp_teardown_ctrl(struct nvme_ctrl *ctrl, bool shutdown) { + cancel_work_sync(&to_tcp_ctrl(ctrl)->err_work); + cancel_delayed_work_sync(&to_tcp_ctrl(ctrl)->connect_work); + nvme_tcp_teardown_io_queues(ctrl, shutdown); if (shutdown) nvme_shutdown_ctrl(ctrl); @@ -1859,12 +1862,6 @@ out_fail: nvme_tcp_reconnect_or_remove(ctrl); } -static void nvme_tcp_stop_ctrl(struct nvme_ctrl *ctrl) -{ - cancel_work_sync(&to_tcp_ctrl(ctrl)->err_work); - cancel_delayed_work_sync(&to_tcp_ctrl(ctrl)->connect_work); -} - static void nvme_tcp_free_ctrl(struct nvme_ctrl *nctrl) { struct nvme_tcp_ctrl *ctrl = to_tcp_ctrl(nctrl); @@ -2115,7 +2112,6 @@ static const struct nvme_ctrl_ops nvme_tcp_ctrl_ops = { .submit_async_event = nvme_tcp_submit_async_event, .delete_ctrl = nvme_tcp_delete_ctrl, .get_address = nvmf_get_address, - .stop_ctrl = nvme_tcp_stop_ctrl, }; static bool -- cgit v1.2.3