nvmet: add error log page support

index f90576862736..090ce8f69ef2 100644
@@ -1257,12 +1257,12 @@ static int nvme_user_cmd(struct nvme_ctrl *ctrl, struct nvme_ns *ns,
 	c.common.nsid = cpu_to_le32(cmd.nsid);
 	c.common.cdw2[0] = cpu_to_le32(cmd.cdw2);
 	c.common.cdw2[1] = cpu_to_le32(cmd.cdw3);
-	c.common.cdw10[0] = cpu_to_le32(cmd.cdw10);
-	c.common.cdw10[1] = cpu_to_le32(cmd.cdw11);
-	c.common.cdw10[2] = cpu_to_le32(cmd.cdw12);
-	c.common.cdw10[3] = cpu_to_le32(cmd.cdw13);
-	c.common.cdw10[4] = cpu_to_le32(cmd.cdw14);
-	c.common.cdw10[5] = cpu_to_le32(cmd.cdw15);
+	c.common.cdw10 = cpu_to_le32(cmd.cdw10);
+	c.common.cdw11 = cpu_to_le32(cmd.cdw11);
+	c.common.cdw12 = cpu_to_le32(cmd.cdw12);
+	c.common.cdw13 = cpu_to_le32(cmd.cdw13);
+	c.common.cdw14 = cpu_to_le32(cmd.cdw14);
+	c.common.cdw15 = cpu_to_le32(cmd.cdw15);
 
 	if (cmd.timeout_ms)
 		timeout = msecs_to_jiffies(cmd.timeout_ms);

@@ -1625,7 +1625,7 @@ static int nvme_pr_command(struct block_device *bdev, u32 cdw10,
 	memset(&c, 0, sizeof(c));
 	c.common.opcode = op;
 	c.common.nsid = cpu_to_le32(ns->head->ns_id);
-	c.common.cdw10[0] = cpu_to_le32(cdw10);
+	c.common.cdw10 = cpu_to_le32(cdw10);
 
 	ret = nvme_submit_sync_cmd(ns->queue, &c, data, 16);
 	nvme_put_ns_from_disk(head, srcu_idx);

@@ -1699,8 +1699,8 @@ int nvme_sec_submit(void *data, u16 spsp, u8 secp, void *buffer, size_t len,
 	else
 		cmd.common.opcode = nvme_admin_security_recv;
 	cmd.common.nsid = 0;
-	cmd.common.cdw10[0] = cpu_to_le32(((u32)secp) << 24 | ((u32)spsp) << 8);
-	cmd.common.cdw10[1] = cpu_to_le32(len);
+	cmd.common.cdw10 = cpu_to_le32(((u32)secp) << 24 | ((u32)spsp) << 8);
+	cmd.common.cdw11 = cpu_to_le32(len);
 
 	return __nvme_submit_sync_cmd(ctrl->admin_q, &cmd, NULL, buffer, len,
 				      ADMIN_TIMEOUT, NVME_QID_ANY, 1, 0);

index 196d5bd56718..1978deb6fcc7 100644
@@ -115,8 +115,8 @@ TRACE_EVENT(nvme_setup_cmd,
 		__entry->nsid = le32_to_cpu(cmd->common.nsid);
 		__entry->metadata = le64_to_cpu(cmd->common.metadata);
 		__assign_disk_name(__entry->disk, req->rq_disk);
-		memcpy(__entry->cdw10, cmd->common.cdw10,
-		       sizeof(__entry->cdw10));
+		memcpy(__entry->cdw10, &cmd->common.cdw10,
+			6 * sizeof(__entry->cdw10));
 	    ),
 	    TP_printk("nvme%d: %sqid=%d, cmdid=%u, nsid=%u, flags=0x%x, meta=0x%llx, cmd=(%s %s)",
 		      __entry->ctrl_id, __print_disk_name(__entry->disk),

index 753515fc8028..721b041a6b3b 100644
@@ -557,7 +557,7 @@ static u16 nvmet_write_protect_flush_sync(struct nvmet_req *req)
 
 static u16 nvmet_set_feat_write_protect(struct nvmet_req *req)
 {
-	u32 write_protect = le32_to_cpu(req->cmd->common.cdw10[1]);
+	u32 write_protect = le32_to_cpu(req->cmd->common.cdw11);
 	struct nvmet_subsys *subsys = req->sq->ctrl->subsys;
 	u16 status = NVME_SC_FEATURE_NOT_CHANGEABLE;
 

@@ -589,7 +589,7 @@ static u16 nvmet_set_feat_write_protect(struct nvmet_req *req)
 
 u16 nvmet_set_feat_kato(struct nvmet_req *req)
 {
-	u32 val32 = le32_to_cpu(req->cmd->common.cdw10[1]);
+	u32 val32 = le32_to_cpu(req->cmd->common.cdw11);
 
 	req->sq->ctrl->kato = DIV_ROUND_UP(val32, 1000);
 

@@ -600,7 +600,7 @@ u16 nvmet_set_feat_kato(struct nvmet_req *req)
 
 u16 nvmet_set_feat_async_event(struct nvmet_req *req, u32 mask)
 {
-	u32 val32 = le32_to_cpu(req->cmd->common.cdw10[1]);
+	u32 val32 = le32_to_cpu(req->cmd->common.cdw11);
 
 	if (val32 & ~mask)
 		return NVME_SC_INVALID_FIELD | NVME_SC_DNR;

@@ -614,7 +614,7 @@ u16 nvmet_set_feat_async_event(struct nvmet_req *req, u32 mask)
 static void nvmet_execute_set_features(struct nvmet_req *req)
 {
 	struct nvmet_subsys *subsys = req->sq->ctrl->subsys;
-	u32 cdw10 = le32_to_cpu(req->cmd->common.cdw10[0]);
+	u32 cdw10 = le32_to_cpu(req->cmd->common.cdw10);
 	u16 status = 0;
 
 	switch (cdw10 & 0xff) {

@@ -675,7 +675,7 @@ void nvmet_get_feat_async_event(struct nvmet_req *req)
 static void nvmet_execute_get_features(struct nvmet_req *req)
 {
 	struct nvmet_subsys *subsys = req->sq->ctrl->subsys;
-	u32 cdw10 = le32_to_cpu(req->cmd->common.cdw10[0]);
+	u32 cdw10 = le32_to_cpu(req->cmd->common.cdw10);
 	u16 status = 0;
 
 	switch (cdw10 & 0xff) {

@@ -715,7 +715,7 @@ static void nvmet_execute_get_features(struct nvmet_req *req)
 		break;
 	case NVME_FEAT_HOST_ID:
 		/* need 128-bit host identifier flag */
-		if (!(req->cmd->common.cdw10[1] & cpu_to_le32(1 << 0))) {
+		if (!(req->cmd->common.cdw11 & cpu_to_le32(1 << 0))) {
 			status = NVME_SC_INVALID_FIELD | NVME_SC_DNR;
 			break;
 		}

index 4d8757ae8210..e1bb254671de 100644
@@ -247,7 +247,7 @@ static void nvmet_execute_identify_disc_ctrl(struct nvmet_req *req)
 
 static void nvmet_execute_disc_set_features(struct nvmet_req *req)
 {
-	u32 cdw10 = le32_to_cpu(req->cmd->common.cdw10[0]);
+	u32 cdw10 = le32_to_cpu(req->cmd->common.cdw10);
 	u16 stat;
 
 	switch (cdw10 & 0xff) {

@@ -268,7 +268,7 @@ static void nvmet_execute_disc_set_features(struct nvmet_req *req)
 
 static void nvmet_execute_disc_get_features(struct nvmet_req *req)
 {
-	u32 cdw10 = le32_to_cpu(req->cmd->common.cdw10[0]);
+	u32 cdw10 = le32_to_cpu(req->cmd->common.cdw10);
 	u16 stat = 0;
 
 	switch (cdw10 & 0xff) {

index 89df51ee5bdf..dafee1af4829 100644
@@ -349,7 +349,7 @@ struct nvmet_async_event {
 
 static inline void nvmet_clear_aen_bit(struct nvmet_req *req, u32 bn)
 {
-	int rae = le32_to_cpu(req->cmd->common.cdw10[0]) & 1 << 15;
+	int rae = le32_to_cpu(req->cmd->common.cdw10) & 1 << 15;
 
 	if (!rae)
 		clear_bit(bn, &req->sq->ctrl->aen_masked);

index 4d7907e3771e..b94fe8fadc4f 100644
@@ -662,7 +662,12 @@ struct nvme_common_command {
 	__le32			cdw2[2];
 	__le64			metadata;
 	union nvme_data_ptr	dptr;
-	__le32			cdw10[6];
+	__le32			cdw10;
+	__le32			cdw11;
+	__le32			cdw12;
+	__le32			cdw13;
+	__le32			cdw14;
+	__le32			cdw15;
 };
 
 struct nvme_rw_command {

index f145fc0220d6..b759c25c89c8 100644
@@ -937,9 +937,9 @@ static int nvme_nvm_user_vcmd(struct nvme_ns *ns, int admin,
 	/* cdw11-12 */
 	c.ph_rw.length = cpu_to_le16(vcmd.nppas);
 	c.ph_rw.control  = cpu_to_le16(vcmd.control);
-	c.common.cdw10[3] = cpu_to_le32(vcmd.cdw13);
-	c.common.cdw10[4] = cpu_to_le32(vcmd.cdw14);
-	c.common.cdw10[5] = cpu_to_le32(vcmd.cdw15);
+	c.common.cdw13 = cpu_to_le32(vcmd.cdw13);
+	c.common.cdw14 = cpu_to_le32(vcmd.cdw14);
+	c.common.cdw15 = cpu_to_le32(vcmd.cdw15);
 
 	if (vcmd.timeout_ms)
 		timeout = msecs_to_jiffies(vcmd.timeout_ms);

index b94fe8fadc4f..bbcc83886899 100644
@@ -1168,6 +1168,20 @@ struct nvme_command {
 	};
 };
 
+struct nvme_error_slot {
+	__le64		error_count;
+	__le16		sqid;
+	__le16		cmdid;
+	__le16		status_field;
+	__le16		param_error_location;
+	__le64		lba;
+	__le32		nsid;
+	__u8		vs;
+	__u8		resv[3];
+	__le64		cs;
+	__u8		resv2[24];
+};
+
 static inline bool nvme_is_write(struct nvme_command *cmd)
 {
 	/*

index e468100b9211..32dca1b50c21 100644
@@ -769,6 +769,8 @@ bool nvmet_req_init(struct nvmet_req *req, struct nvmet_cq *cq,
 	req->rsp->status = 0;
 	req->rsp->sq_head = 0;
 	req->ns = NULL;
+	req->error_loc = -1;
+	req->error_slba = 0;
 
 	/* no support for fused commands yet */
 	if (unlikely(flags & (NVME_CMD_FUSE_FIRST | NVME_CMD_FUSE_SECOND))) {

@@ -1174,6 +1176,9 @@ u16 nvmet_alloc_ctrl(const char *subsysnqn, const char *hostnqn,
 	/* keep-alive timeout in seconds */
 	ctrl->kato = DIV_ROUND_UP(kato, 1000);
 
+	ctrl->err_counter = 0;
+	spin_lock_init(&ctrl->error_lock);
+
 	nvmet_start_keep_alive_timer(ctrl);
 
 	mutex_lock(&subsys->lock);

index dafee1af4829..61b3cc415905 100644
@@ -202,6 +202,10 @@ struct nvmet_ctrl {
 
 	struct device		*p2p_client;
 	struct radix_tree_root	p2p_ns_map;
+
+	spinlock_t		error_lock;
+	u64			err_counter;
+	struct nvme_error_slot	slots[NVMET_ERROR_LOG_SLOTS];
 };
 
 struct nvmet_subsys {

@@ -317,6 +321,8 @@ struct nvmet_req {
 
 	struct pci_dev		*p2p_dev;
 	struct device		*p2p_client;
+	u16			error_loc;
+	u64			error_slba;
 };
 
 extern struct workqueue_struct *buffered_io_wq;

index 32dca1b50c21..b6493753070b 100644
@@ -611,14 +611,44 @@ static void nvmet_update_sq_head(struct nvmet_req *req)
 	req->rsp->sq_head = cpu_to_le16(req->sq->sqhd & 0x0000FFFF);
 }
 
+static void nvmet_set_error(struct nvmet_req *req, u16 status)
+{
+	struct nvmet_ctrl *ctrl = req->sq->ctrl;
+	struct nvme_error_slot *new_error_slot;
+	unsigned long flags;
+
+	req->rsp->status = cpu_to_le16(status << 1);
+
+	if (!ctrl || req->error_loc == -1)
+		return;
+
+	spin_lock_irqsave(&ctrl->error_lock, flags);
+	ctrl->err_counter++;
+	new_error_slot =
+		&ctrl->slots[ctrl->err_counter % NVMET_ERROR_LOG_SLOTS];
+
+	new_error_slot->error_count = cpu_to_le64(ctrl->err_counter);
+	new_error_slot->sqid = cpu_to_le16(req->sq->qid);
+	new_error_slot->cmdid = cpu_to_le16(req->cmd->common.command_id);
+	new_error_slot->status_field = cpu_to_le16(status << 1);
+	new_error_slot->param_error_location = cpu_to_le16(req->error_loc);
+	new_error_slot->lba = cpu_to_le64(req->error_slba);
+	new_error_slot->nsid = cpu_to_le32(req->cmd->common.nsid);
+	spin_unlock_irqrestore(&ctrl->error_lock, flags);
+
+	/* set the more bit for this request */
+	req->rsp->status |= cpu_to_le16(1 << 14);
+}
+
 static void __nvmet_req_complete(struct nvmet_req *req, u16 status)
 {
 	if (!req->sq->sqhd_disabled)
 		nvmet_update_sq_head(req);
 	req->rsp->sq_id = cpu_to_le16(req->sq->qid);
 	req->rsp->command_id = req->cmd->common.command_id;
+
 	if (unlikely(status))
-		nvmet_set_status(req, status);
+		nvmet_set_error(req, status);
 	if (req->ns)
 		nvmet_put_namespace(req->ns);
 	req->ops->queue_response(req);

index 61b3cc415905..a2a7fd69b66a 100644
@@ -327,11 +327,6 @@ struct nvmet_req {
 
 extern struct workqueue_struct *buffered_io_wq;
 
-static inline void nvmet_set_status(struct nvmet_req *req, u16 status)
-{
-	req->rsp->status = cpu_to_le16(status << 1);
-}
-
 static inline void nvmet_set_result(struct nvmet_req *req, u32 result)
 {
 	req->rsp->result.u32 = cpu_to_le32(result);

index ba0ab5481cc5..cc81d0231587 100644
@@ -677,7 +677,7 @@ static void nvmet_set_error(struct nvmet_req *req, u16 status)
 	new_error_slot->status_field = cpu_to_le16(status << 1);
 	new_error_slot->param_error_location = cpu_to_le16(req->error_loc);
 	new_error_slot->lba = cpu_to_le64(req->error_slba);
-	new_error_slot->nsid = cpu_to_le32(req->cmd->common.nsid);
+	new_error_slot->nsid = req->cmd->common.nsid;
 	spin_unlock_irqrestore(&ctrl->error_lock, flags);
 
 	/* set the more bit for this request */

index b6493753070b..4177df6e0d70 100644
@@ -51,22 +51,28 @@ static struct nvmet_subsys *nvmet_find_get_subsys(struct nvmet_port *port,
 u16 nvmet_copy_to_sgl(struct nvmet_req *req, off_t off, const void *buf,
 		size_t len)
 {
-	if (sg_pcopy_from_buffer(req->sg, req->sg_cnt, buf, len, off) != len)
+	if (sg_pcopy_from_buffer(req->sg, req->sg_cnt, buf, len, off) != len) {
+		req->error_loc = offsetof(struct nvme_common_command, dptr);
 		return NVME_SC_SGL_INVALID_DATA | NVME_SC_DNR;
+	}
 	return 0;
 }
 
 u16 nvmet_copy_from_sgl(struct nvmet_req *req, off_t off, void *buf, size_t len)
 {
-	if (sg_pcopy_to_buffer(req->sg, req->sg_cnt, buf, len, off) != len)
+	if (sg_pcopy_to_buffer(req->sg, req->sg_cnt, buf, len, off) != len) {
+		req->error_loc = offsetof(struct nvme_common_command, dptr);
 		return NVME_SC_SGL_INVALID_DATA | NVME_SC_DNR;
+	}
 	return 0;
 }
 
 u16 nvmet_zero_sgl(struct nvmet_req *req, off_t off, size_t len)
 {
-	if (sg_zero_buffer(req->sg, req->sg_cnt, len, off) != len)
+	if (sg_zero_buffer(req->sg, req->sg_cnt, len, off) != len) {
+		req->error_loc = offsetof(struct nvme_common_command, dptr);
 		return NVME_SC_SGL_INVALID_DATA | NVME_SC_DNR;
+	}
 	return 0;
 }
 

@@ -769,14 +775,20 @@ static u16 nvmet_parse_io_cmd(struct nvmet_req *req)
 		return ret;
 
 	req->ns = nvmet_find_namespace(req->sq->ctrl, cmd->rw.nsid);
-	if (unlikely(!req->ns))
+	if (unlikely(!req->ns)) {
+		req->error_loc = offsetof(struct nvme_common_command, nsid);
 		return NVME_SC_INVALID_NS | NVME_SC_DNR;
+	}
 	ret = nvmet_check_ana_state(req->port, req->ns);
-	if (unlikely(ret))
+	if (unlikely(ret)) {
+		req->error_loc = offsetof(struct nvme_common_command, nsid);
 		return ret;
+	}
 	ret = nvmet_io_cmd_check_access(req);
-	if (unlikely(ret))
+	if (unlikely(ret)) {
+		req->error_loc = offsetof(struct nvme_common_command, nsid);
 		return ret;
+	}
 
 	if (req->ns->file)
 		return nvmet_file_parse_io_cmd(req);

@@ -804,6 +816,7 @@ bool nvmet_req_init(struct nvmet_req *req, struct nvmet_cq *cq,
 
 	/* no support for fused commands yet */
 	if (unlikely(flags & (NVME_CMD_FUSE_FIRST | NVME_CMD_FUSE_SECOND))) {
+		req->error_loc = offsetof(struct nvme_common_command, flags);
 		status = NVME_SC_INVALID_FIELD | NVME_SC_DNR;
 		goto fail;
 	}

@@ -814,6 +827,7 @@ bool nvmet_req_init(struct nvmet_req *req, struct nvmet_cq *cq,
 	 * byte aligned.
 	 */
 	if (unlikely((flags & NVME_CMD_SGL_ALL) != NVME_CMD_SGL_METABUF)) {
+		req->error_loc = offsetof(struct nvme_common_command, flags);
 		status = NVME_SC_INVALID_FIELD | NVME_SC_DNR;
 		goto fail;
 	}

@@ -859,9 +873,10 @@ EXPORT_SYMBOL_GPL(nvmet_req_uninit);
 
 void nvmet_req_execute(struct nvmet_req *req)
 {
-	if (unlikely(req->data_len != req->transfer_len))
+	if (unlikely(req->data_len != req->transfer_len)) {
+		req->error_loc = offsetof(struct nvme_common_command, dptr);
 		nvmet_req_complete(req, NVME_SC_SGL_INVALID_DATA | NVME_SC_DNR);
-	else
+	} else
 		req->execute(req);
 }
 EXPORT_SYMBOL_GPL(nvmet_req_execute);

index c1cb2ed5531c..8a70c58d4246 100644
@@ -44,13 +44,69 @@ void nvmet_bdev_ns_disable(struct nvmet_ns *ns)
 	}
 }
 
+static u16 blk_to_nvme_status(struct nvmet_req *req, blk_status_t blk_sts)
+{
+	u16 status = NVME_SC_SUCCESS;
+
+	if (likely(blk_sts == BLK_STS_OK))
+		return status;
+	/*
+	 * Right now there exists M : 1 mapping between block layer error
+	 * to the NVMe status code (see nvme_error_status()). For consistency,
+	 * when we reverse map we use most appropriate NVMe Status code from
+	 * the group of the NVMe staus codes used in the nvme_error_status().
+	 */
+	switch (blk_sts) {
+	case BLK_STS_NOSPC:
+		status = NVME_SC_CAP_EXCEEDED | NVME_SC_DNR;
+		req->error_loc = offsetof(struct nvme_rw_command, length);
+		break;
+	case BLK_STS_TARGET:
+		status = NVME_SC_LBA_RANGE | NVME_SC_DNR;
+		req->error_loc = offsetof(struct nvme_rw_command, slba);
+		break;
+	case BLK_STS_NOTSUPP:
+		req->error_loc = offsetof(struct nvme_common_command, opcode);
+		switch (req->cmd->common.opcode) {
+		case nvme_cmd_dsm:
+		case nvme_cmd_write_zeroes:
+			status = NVME_SC_ONCS_NOT_SUPPORTED | NVME_SC_DNR;
+			break;
+		default:
+			status = NVME_SC_INVALID_OPCODE | NVME_SC_DNR;
+		}
+		break;
+	case BLK_STS_MEDIUM:
+		status = NVME_SC_ACCESS_DENIED;
+		req->error_loc = offsetof(struct nvme_rw_command, nsid);
+		break;
+	case BLK_STS_IOERR:
+		/* fallthru */
+	default:
+		status = NVME_SC_INTERNAL | NVME_SC_DNR;
+		req->error_loc = offsetof(struct nvme_common_command, opcode);
+	}
+
+	switch (req->cmd->common.opcode) {
+	case nvme_cmd_read:
+	case nvme_cmd_write:
+		req->error_slba = le64_to_cpu(req->cmd->rw.slba);
+		break;
+	case nvme_cmd_write_zeroes:
+		req->error_slba =
+			le64_to_cpu(req->cmd->write_zeroes.slba);
+		break;
+	default:
+		req->error_slba = 0;
+	}
+	return status;
+}
+
 static void nvmet_bio_done(struct bio *bio)
 {
 	struct nvmet_req *req = bio->bi_private;
 
-	nvmet_req_complete(req,
-		bio->bi_status ? NVME_SC_INTERNAL | NVME_SC_DNR : 0);
-
+	nvmet_req_complete(req, blk_to_nvme_status(req, bio->bi_status));
 	if (bio != &req->b.inline_bio)
 		bio_put(bio);
 }

@@ -137,18 +193,21 @@ u16 nvmet_bdev_flush(struct nvmet_req *req)
 	return 0;
 }
 
-static u16 nvmet_bdev_discard_range(struct nvmet_ns *ns,
+static u16 nvmet_bdev_discard_range(struct nvmet_req *req,
 		struct nvme_dsm_range *range, struct bio **bio)
 {
+	struct nvmet_ns *ns = req->ns;
 	int ret;
 
 	ret = __blkdev_issue_discard(ns->bdev,
 			le64_to_cpu(range->slba) << (ns->blksize_shift - 9),
 			le32_to_cpu(range->nlb) << (ns->blksize_shift - 9),
 			GFP_KERNEL, 0, bio);
-	if (ret && ret != -EOPNOTSUPP)
-		return NVME_SC_INTERNAL | NVME_SC_DNR;
-	return 0;
+
+	if (ret)
+		req->error_slba = le64_to_cpu(range->slba);
+
+	return blk_to_nvme_status(req, errno_to_blk_status(ret));
 }
 
 static void nvmet_bdev_execute_discard(struct nvmet_req *req)

@@ -164,7 +223,7 @@ static void nvmet_bdev_execute_discard(struct nvmet_req *req)
 		if (status)
 			break;
 
-		status = nvmet_bdev_discard_range(req->ns, &range, &bio);
+		status = nvmet_bdev_discard_range(req, &range, &bio);
 		if (status)
 			break;
 	}

@@ -205,16 +264,16 @@ static void nvmet_bdev_execute_write_zeroes(struct nvmet_req *req)
 	u16 status = NVME_SC_SUCCESS;
 	sector_t sector;
 	sector_t nr_sector;
+	int ret;
 
 	sector = le64_to_cpu(write_zeroes->slba) <<
 		(req->ns->blksize_shift - 9);
 	nr_sector = (((sector_t)le16_to_cpu(write_zeroes->length) + 1) <<
 		(req->ns->blksize_shift - 9));
 
-	if (__blkdev_issue_zeroout(req->ns->bdev, sector, nr_sector,
-				GFP_KERNEL, &bio, 0))
-		status = NVME_SC_INTERNAL | NVME_SC_DNR;
-
+	ret = __blkdev_issue_zeroout(req->ns->bdev, sector, nr_sector,
+			GFP_KERNEL, &bio, 0);
+	status = blk_to_nvme_status(req, errno_to_blk_status(ret));
 	if (bio) {
 		bio->bi_private = req;
 		bio->bi_end_io = nvmet_bio_done;

@@ -249,6 +308,7 @@ u16 nvmet_bdev_parse_io_cmd(struct nvmet_req *req)
 	default:
 		pr_err("unhandled cmd %d on qid %d\n", cmd->common.opcode,
 		       req->sq->qid);
+		req->error_loc = offsetof(struct nvme_common_command, opcode);
 		return NVME_SC_INVALID_OPCODE | NVME_SC_DNR;
 	}
 }

index 4177df6e0d70..ba0ab5481cc5 100644
@@ -45,6 +45,44 @@ u32 nvmet_ana_group_enabled[NVMET_MAX_ANAGRPS + 1];
 u64 nvmet_ana_chgcnt;
 DECLARE_RWSEM(nvmet_ana_sem);
 
+inline u16 errno_to_nvme_status(struct nvmet_req *req, int errno)
+{
+	u16 status;
+
+	switch (errno) {
+	case -ENOSPC:
+		req->error_loc = offsetof(struct nvme_rw_command, length);
+		status = NVME_SC_CAP_EXCEEDED | NVME_SC_DNR;
+		break;
+	case -EREMOTEIO:
+		req->error_loc = offsetof(struct nvme_rw_command, slba);
+		status = NVME_SC_LBA_RANGE | NVME_SC_DNR;
+		break;
+	case -EOPNOTSUPP:
+		req->error_loc = offsetof(struct nvme_common_command, opcode);
+		switch (req->cmd->common.opcode) {
+		case nvme_cmd_dsm:
+		case nvme_cmd_write_zeroes:
+			status = NVME_SC_ONCS_NOT_SUPPORTED | NVME_SC_DNR;
+			break;
+		default:
+			status = NVME_SC_INVALID_OPCODE | NVME_SC_DNR;
+		}
+		break;
+	case -ENODATA:
+		req->error_loc = offsetof(struct nvme_rw_command, nsid);
+		status = NVME_SC_ACCESS_DENIED;
+		break;
+	case -EIO:
+		/* FALLTHRU */
+	default:
+		req->error_loc = offsetof(struct nvme_common_command, opcode);
+		status = NVME_SC_INTERNAL | NVME_SC_DNR;
+	}
+
+	return status;
+}
+
 static struct nvmet_subsys *nvmet_find_get_subsys(struct nvmet_port *port,
 		const char *subsysnqn);
 

index 12eaa8ddc248..517522305e5c 100644
@@ -112,6 +112,7 @@ static ssize_t nvmet_file_submit_bvec(struct nvmet_req *req, loff_t pos,
 static void nvmet_file_io_done(struct kiocb *iocb, long ret, long ret2)
 {
 	struct nvmet_req *req = container_of(iocb, struct nvmet_req, f.iocb);
+	u16 status = NVME_SC_SUCCESS;
 
 	if (req->f.bvec != req->inline_bvec) {
 		if (likely(req->f.mpool_alloc == false))

@@ -120,8 +121,9 @@ static void nvmet_file_io_done(struct kiocb *iocb, long ret, long ret2)
 			mempool_free(req->f.bvec, req->ns->bvec_pool);
 	}
 
-	nvmet_req_complete(req, ret != req->data_len ?
-			NVME_SC_INTERNAL | NVME_SC_DNR : 0);
+	if (unlikely(ret != req->data_len))
+		status = errno_to_nvme_status(req, ret);
+	nvmet_req_complete(req, status);
 }
 
 static bool nvmet_file_execute_io(struct nvmet_req *req, int ki_flags)

@@ -140,7 +142,7 @@ static bool nvmet_file_execute_io(struct nvmet_req *req, int ki_flags)
 
 	pos = le64_to_cpu(req->cmd->rw.slba) << req->ns->blksize_shift;
 	if (unlikely(pos + req->data_len > req->ns->size)) {
-		nvmet_req_complete(req, NVME_SC_LBA_RANGE | NVME_SC_DNR);
+		nvmet_req_complete(req, errno_to_nvme_status(req, -ENOSPC));
 		return true;
 	}
 

@@ -254,9 +256,7 @@ static void nvmet_file_execute_rw(struct nvmet_req *req)
 
 u16 nvmet_file_flush(struct nvmet_req *req)
 {
-	if (vfs_fsync(req->ns->file, 1) < 0)
-		return NVME_SC_INTERNAL | NVME_SC_DNR;
-	return 0;
+	return errno_to_nvme_status(req, vfs_fsync(req->ns->file, 1));
 }
 
 static void nvmet_file_flush_work(struct work_struct *w)

@@ -277,30 +277,34 @@ static void nvmet_file_execute_discard(struct nvmet_req *req)
 	int mode = FALLOC_FL_PUNCH_HOLE | FALLOC_FL_KEEP_SIZE;
 	struct nvme_dsm_range range;
 	loff_t offset, len;
-	u16 ret;
+	u16 status = 0;
+	int ret;
 	int i;
 
 	for (i = 0; i <= le32_to_cpu(req->cmd->dsm.nr); i++) {
-		ret = nvmet_copy_from_sgl(req, i * sizeof(range), &range,
+		status = nvmet_copy_from_sgl(req, i * sizeof(range), &range,
 					sizeof(range));
-		if (ret)
+		if (status)
 			break;
 
 		offset = le64_to_cpu(range.slba) << req->ns->blksize_shift;
 		len = le32_to_cpu(range.nlb);
 		len <<= req->ns->blksize_shift;
 		if (offset + len > req->ns->size) {
-			ret = NVME_SC_LBA_RANGE | NVME_SC_DNR;
+			req->error_slba = le64_to_cpu(range.slba);
+			status = errno_to_nvme_status(req, -ENOSPC);
 			break;
 		}
 
-		if (vfs_fallocate(req->ns->file, mode, offset, len)) {
-			ret = NVME_SC_INTERNAL | NVME_SC_DNR;
+		ret = vfs_fallocate(req->ns->file, mode, offset, len);
+		if (ret) {
+			req->error_slba = le64_to_cpu(range.slba);
+			status = errno_to_nvme_status(req, ret);
 			break;
 		}
 	}
 
-	nvmet_req_complete(req, ret);
+	nvmet_req_complete(req, status);
 }
 
 static void nvmet_file_dsm_work(struct work_struct *w)

@@ -340,12 +344,12 @@ static void nvmet_file_write_zeroes_work(struct work_struct *w)
 			req->ns->blksize_shift);
 
 	if (unlikely(offset + len > req->ns->size)) {
-		nvmet_req_complete(req, NVME_SC_LBA_RANGE | NVME_SC_DNR);
+		nvmet_req_complete(req, errno_to_nvme_status(req, -ENOSPC));
 		return;
 	}
 
 	ret = vfs_fallocate(req->ns->file, mode, offset, len);
-	nvmet_req_complete(req, ret < 0 ? NVME_SC_INTERNAL | NVME_SC_DNR : 0);
+	nvmet_req_complete(req, ret < 0 ? errno_to_nvme_status(req, ret) : 0);
 }
 
 static void nvmet_file_execute_write_zeroes(struct nvmet_req *req)

@@ -380,6 +384,7 @@ u16 nvmet_file_parse_io_cmd(struct nvmet_req *req)
 	default:
 		pr_err("unhandled cmd for file ns %d on qid %d\n",
 				cmd->common.opcode, req->sq->qid);
+		req->error_loc = offsetof(struct nvme_common_command, opcode);
 		return NVME_SC_INVALID_OPCODE | NVME_SC_DNR;
 	}
 }

index a2a7fd69b66a..3b5f0bcaf3e8 100644
@@ -494,4 +494,6 @@ static inline u32 nvmet_rw_len(struct nvmet_req *req)
 	return ((u32)le16_to_cpu(req->cmd->rw.length) + 1) <<
 			req->ns->blksize_shift;
 }
+
+u16 errno_to_nvme_status(struct nvmet_req *req, int errno);
 #endif /* _NVMET_H */

index e1bb254671de..d2cb71a0b419 100644
@@ -259,6 +259,8 @@ static void nvmet_execute_disc_set_features(struct nvmet_req *req)
 						  NVMET_DISC_AEN_CFG_OPTIONAL);
 		break;
 	default:
+		req->error_loc =
+			offsetof(struct nvme_common_command, cdw10);
 		stat = NVME_SC_INVALID_FIELD | NVME_SC_DNR;
 		break;
 	}

@@ -279,6 +281,8 @@ static void nvmet_execute_disc_get_features(struct nvmet_req *req)
 		nvmet_get_feat_async_event(req);
 		break;
 	default:
+		req->error_loc =
+			offsetof(struct nvme_common_command, cdw10);
 		stat = NVME_SC_INVALID_FIELD | NVME_SC_DNR;
 		break;
 	}

@@ -293,6 +297,8 @@ u16 nvmet_parse_discovery_cmd(struct nvmet_req *req)
 	if (unlikely(!(req->sq->ctrl->csts & NVME_CSTS_RDY))) {
 		pr_err("got cmd %d while not ready\n",
 		       cmd->common.opcode);
+		req->error_loc =
+			offsetof(struct nvme_common_command, opcode);
 		return NVME_SC_INVALID_OPCODE | NVME_SC_DNR;
 	}
 

@@ -323,6 +329,8 @@ u16 nvmet_parse_discovery_cmd(struct nvmet_req *req)
 		default:
 			pr_err("unsupported get_log_page lid %d\n",
 			       cmd->get_log_page.lid);
+			req->error_loc =
+				offsetof(struct nvme_get_log_page_command, lid);
 		return NVME_SC_INVALID_OPCODE | NVME_SC_DNR;
 		}
 	case nvme_admin_identify:

@@ -335,10 +343,12 @@ u16 nvmet_parse_discovery_cmd(struct nvmet_req *req)
 		default:
 			pr_err("unsupported identify cns %d\n",
 			       cmd->identify.cns);
+			req->error_loc = offsetof(struct nvme_identify, cns);
 			return NVME_SC_INVALID_OPCODE | NVME_SC_DNR;
 		}
 	default:
 		pr_err("unhandled cmd %d\n", cmd->common.opcode);
+		req->error_loc = offsetof(struct nvme_common_command, opcode);
 		return NVME_SC_INVALID_OPCODE | NVME_SC_DNR;
 	}
 

105 lines not on this page · raw message

index 932242e27b03..b15c6915fb40 100644
@@ -629,8 +629,11 @@ static u16 nvmet_rdma_map_sgl_inline(struct nvmet_rdma_rsp *rsp)
 	u64 off = le64_to_cpu(sgl->addr);
 	u32 len = le32_to_cpu(sgl->length);
 
-	if (!nvme_is_write(rsp->req.cmd))
+	if (!nvme_is_write(rsp->req.cmd)) {
+		rsp->req.error_loc =
+			offsetof(struct nvme_common_command, opcode);
 		return NVME_SC_INVALID_FIELD | NVME_SC_DNR;
+	}
 
 	if (off + len > rsp->queue->dev->inline_data_size) {
 		pr_err("invalid inline data offset!\n");

@@ -695,6 +698,8 @@ static u16 nvmet_rdma_map_sgl(struct nvmet_rdma_rsp *rsp)
 			return nvmet_rdma_map_sgl_inline(rsp);
 		default:
 			pr_err("invalid SGL subtype: %#x\n", sgl->type);
+			rsp->req.error_loc =
+				offsetof(struct nvme_common_command, dptr);
 			return NVME_SC_INVALID_FIELD | NVME_SC_DNR;
 		}
 	case NVME_KEY_SGL_FMT_DATA_DESC:

@@ -705,10 +710,13 @@ static u16 nvmet_rdma_map_sgl(struct nvmet_rdma_rsp *rsp)
 			return nvmet_rdma_map_sgl_keyed(rsp, sgl, false);
 		default:
 			pr_err("invalid SGL subtype: %#x\n", sgl->type);
+			rsp->req.error_loc =
+				offsetof(struct nvme_common_command, dptr);
 			return NVME_SC_INVALID_FIELD | NVME_SC_DNR;
 		}
 	default:
 		pr_err("invalid SGL type: %#x\n", sgl->type);
+		rsp->req.error_loc = offsetof(struct nvme_common_command, dptr);
 		return NVME_SC_SGL_INVALID_TYPE | NVME_SC_DNR;
 	}
 }

index 721b041a6b3b..fa62db7a5e9e 100644
@@ -47,6 +47,7 @@ static u16 nvmet_get_smart_log_nsid(struct nvmet_req *req,
 	if (!ns) {
 		pr_err("Could not find namespace id : %d\n",
 				le32_to_cpu(req->cmd->get_log_page.nsid));
+		req->error_loc = offsetof(struct nvme_rw_command, nsid);
 		return NVME_SC_INVALID_NS;
 	}
 

@@ -380,6 +381,7 @@ static void nvmet_execute_identify_ns(struct nvmet_req *req)
 	u16 status = 0;
 
 	if (le32_to_cpu(req->cmd->identify.nsid) == NVME_NSID_ALL) {
+		req->error_loc = offsetof(struct nvme_identify, nsid);
 		status = NVME_SC_INVALID_NS | NVME_SC_DNR;
 		goto out;
 	}

@@ -500,6 +502,7 @@ static void nvmet_execute_identify_desclist(struct nvmet_req *req)
 
 	ns = nvmet_find_namespace(req->sq->ctrl, req->cmd->identify.nsid);
 	if (!ns) {
+		req->error_loc = offsetof(struct nvme_identify, nsid);
 		status = NVME_SC_INVALID_NS | NVME_SC_DNR;
 		goto out;
 	}

@@ -562,8 +565,10 @@ static u16 nvmet_set_feat_write_protect(struct nvmet_req *req)
 	u16 status = NVME_SC_FEATURE_NOT_CHANGEABLE;
 
 	req->ns = nvmet_find_namespace(req->sq->ctrl, req->cmd->rw.nsid);
-	if (unlikely(!req->ns))
+	if (unlikely(!req->ns)) {
+		req->error_loc = offsetof(struct nvme_common_command, nsid);
 		return status;
+	}
 
 	mutex_lock(&subsys->lock);
 	switch (write_protect) {

@@ -602,8 +607,10 @@ u16 nvmet_set_feat_async_event(struct nvmet_req *req, u32 mask)
 {
 	u32 val32 = le32_to_cpu(req->cmd->common.cdw11);
 
-	if (val32 & ~mask)
+	if (val32 & ~mask) {
+		req->error_loc = offsetof(struct nvme_common_command, cdw11);
 		return NVME_SC_INVALID_FIELD | NVME_SC_DNR;
+	}
 
 	WRITE_ONCE(req->sq->ctrl->aen_enabled, val32);
 	nvmet_set_result(req, val32);

@@ -635,6 +642,7 @@ static void nvmet_execute_set_features(struct nvmet_req *req)
 		status = nvmet_set_feat_write_protect(req);
 		break;
 	default:
+		req->error_loc = offsetof(struct nvme_common_command, cdw10);
 		status = NVME_SC_INVALID_FIELD | NVME_SC_DNR;
 		break;
 	}

@@ -648,9 +656,10 @@ static u16 nvmet_get_feat_write_protect(struct nvmet_req *req)
 	u32 result;
 
 	req->ns = nvmet_find_namespace(req->sq->ctrl, req->cmd->common.nsid);
-	if (!req->ns)
+	if (!req->ns)  {
+		req->error_loc = offsetof(struct nvme_common_command, nsid);
 		return NVME_SC_INVALID_NS | NVME_SC_DNR;
-
+	}
 	mutex_lock(&subsys->lock);
 	if (req->ns->readonly == true)
 		result = NVME_NS_WRITE_PROTECT;

@@ -716,6 +725,8 @@ static void nvmet_execute_get_features(struct nvmet_req *req)
 	case NVME_FEAT_HOST_ID:
 		/* need 128-bit host identifier flag */
 		if (!(req->cmd->common.cdw11 & cpu_to_le32(1 << 0))) {
+			req->error_loc =
+				offsetof(struct nvme_common_command, cdw11);
 			status = NVME_SC_INVALID_FIELD | NVME_SC_DNR;
 			break;
 		}

@@ -727,6 +738,8 @@ static void nvmet_execute_get_features(struct nvmet_req *req)
 		status = nvmet_get_feat_write_protect(req);
 		break;
 	default:
+		req->error_loc =
+			offsetof(struct nvme_common_command, cdw10);
 		status = NVME_SC_INVALID_FIELD | NVME_SC_DNR;
 		break;
 	}

@@ -848,5 +861,6 @@ u16 nvmet_parse_admin_cmd(struct nvmet_req *req)
 
 	pr_err("unhandled cmd %d on qid %d\n", cmd->common.opcode,
 	       req->sq->qid);
+	req->error_loc = offsetof(struct nvme_common_command, opcode);
 	return NVME_SC_INVALID_OPCODE | NVME_SC_DNR;
 }

index fa62db7a5e9e..00956b4106a8 100644
@@ -37,6 +37,34 @@ static void nvmet_execute_get_log_page_noop(struct nvmet_req *req)
 	nvmet_req_complete(req, nvmet_zero_sgl(req, 0, req->data_len));
 }
 
+static void nvmet_execute_get_log_page_error(struct nvmet_req *req)
+{
+	struct nvmet_ctrl *ctrl = req->sq->ctrl;
+	u16 status = NVME_SC_SUCCESS;
+	unsigned long flags;
+	off_t offset = 0;
+	u64 slot;
+	u64 i;
+
+	spin_lock_irqsave(&ctrl->error_lock, flags);
+	slot = ctrl->err_counter % NVMET_ERROR_LOG_SLOTS;
+
+	for (i = 0; i < NVMET_ERROR_LOG_SLOTS; i++) {
+		status = nvmet_copy_to_sgl(req, offset, &ctrl->slots[slot],
+				sizeof(struct nvme_error_slot));
+		if (status)
+			break;
+
+		if (slot == 0)
+			slot = NVMET_ERROR_LOG_SLOTS - 1;
+		else
+			slot--;
+		offset += sizeof(struct nvme_error_slot);
+	}
+	spin_unlock_irqrestore(&ctrl->error_lock, flags);
+	nvmet_req_complete(req, status);
+}
+
 static u16 nvmet_get_smart_log_nsid(struct nvmet_req *req,
 		struct nvme_smart_log *slog)
 {

@@ -789,13 +817,7 @@ u16 nvmet_parse_admin_cmd(struct nvmet_req *req)
 
 		switch (cmd->get_log_page.lid) {
 		case NVME_LOG_ERROR:
-			/*
-			 * We currently never set the More bit in the status
-			 * field, so all error log entries are invalid and can
-			 * be zeroed out.  This is called a minum viable
-			 * implementation (TM) of this mandatory log page.
-			 */
-			req->execute = nvmet_execute_get_log_page_noop;
+			req->execute = nvmet_execute_get_log_page_error;
 			return 0;
 		case NVME_LOG_SMART:
 			req->execute = nvmet_execute_get_log_page_smart;

index 00956b4106a8..11baeb14c388 100644
@@ -135,6 +135,7 @@ static void nvmet_execute_get_log_page_smart(struct nvmet_req *req)
 {
 	struct nvme_smart_log *log;
 	u16 status = NVME_SC_INTERNAL;
+	unsigned long flags;
 
 	if (req->data_len != sizeof(*log))
 		goto out;

@@ -150,6 +151,11 @@ static void nvmet_execute_get_log_page_smart(struct nvmet_req *req)
 	if (status)
 		goto out_free_log;
 
+	spin_lock_irqsave(&req->sq->ctrl->error_lock, flags);
+	put_unaligned_le64(req->sq->ctrl->err_counter,
+			&log->num_err_log_entries);
+	spin_unlock_irqrestore(&req->sq->ctrl->error_lock, flags);
+
 	status = nvmet_copy_to_sgl(req, 0, log, sizeof(*log));
 out_free_log:
 	kfree(log);