Merge git://git.kernel.org/pub/scm/linux/kernel/git/nab/target-pending

Pull SCSI target fixes from Nicholas Bellinger:
 "Here are the outstanding target fixes queued up for v3.12-rc4 code.

  The highlights include:

   - Make vhost/scsi tag percpu_ida_alloc() use GFP_ATOMIC
   - Allow sess_cmd_map allocation failure fallback to use vzalloc
   - Fix COMPARE_AND_WRITE se_cmd->data_length bug with FILEIO backends
   - Fixes for COMPARE_AND_WRITE callback recursive failure OOPs + non
     zero scsi_status bug
   - Make iscsi-target do acknowledgement tag release from RX context
   - Setup iscsi-target with extra (cmdsn_depth / 2) percpu_ida tags

  Also included is a iscsi-target patch CC'ed for v3.10+ that avoids
  legacy wait_for_task=true release during fast-past StatSN
  acknowledgement, and two other SRP target related patches that address
  long-standing issues that are CC'ed for v3.3+.

  Extra thanks to Thomas Glanzmann for his testing feedback with
  COMPARE_AND_WRITE + EXTENDED_COPY VAAI logic"

* git://git.kernel.org/pub/scm/linux/kernel/git/nab/target-pending:
  iscsi-target; Allow an extra tag_num / 2 number of percpu_ida tags
  iscsi-target: Perform release of acknowledged tags from RX context
  iscsi-target: Only perform wait_for_tasks when performing shutdown
  target: Fail on non zero scsi_status in compare_and_write_callback
  target: Fix recursive COMPARE_AND_WRITE callback failure
  target: Reset data_length for COMPARE_AND_WRITE to NoLB * block_size
  ib_srpt: always set response for task management
  target: Fall back to vzalloc upon ->sess_cmd_map kzalloc failure
  vhost/scsi: Use GFP_ATOMIC with percpu_ida_alloc for obtaining tag
  ib_srpt: Destroy cm_id before destroying QP.
  target: Fix xop->dbl assignment in target_xcopy_parse_segdesc_02

Author: torvalds

drivers/infiniband/ulp/srpt/ib_srpt.c

@@ -1588,7 +1588,7 @@ static int srpt_build_tskmgmt_rsp(struct srpt_rdma_ch *ch,
 	int resp_data_len;
 	int resp_len;
 
-	resp_data_len = (rsp_code == SRP_TSK_MGMT_SUCCESS) ? 0 : 4;
+	resp_data_len = 4;
 	resp_len = sizeof(*srp_rsp) + resp_data_len;
 
 	srp_rsp = ioctx->ioctx.buf;
@@ -1600,11 +1600,9 @@ static int srpt_build_tskmgmt_rsp(struct srpt_rdma_ch *ch,
 				    + atomic_xchg(&ch->req_lim_delta, 0));
 	srp_rsp->tag = tag;
 
-	if (rsp_code != SRP_TSK_MGMT_SUCCESS) {
-		srp_rsp->flags |= SRP_RSP_FLAG_RSPVALID;
-		srp_rsp->resp_data_len = cpu_to_be32(resp_data_len);
-		srp_rsp->data[3] = rsp_code;
-	}
+	srp_rsp->flags |= SRP_RSP_FLAG_RSPVALID;
+	srp_rsp->resp_data_len = cpu_to_be32(resp_data_len);
+	srp_rsp->data[3] = rsp_code;
 
 	return resp_len;
 }
@@ -2358,6 +2356,8 @@ static void srpt_release_channel_work(struct work_struct *w)
 	transport_deregister_session(se_sess);
 	ch->sess = NULL;
 
+	ib_destroy_cm_id(ch->cm_id);
+
 	srpt_destroy_ch_ib(ch);
 
 	srpt_free_ioctx_ring((struct srpt_ioctx **)ch->ioctx_ring,
@@ -2368,8 +2368,6 @@ static void srpt_release_channel_work(struct work_struct *w)
 	list_del(&ch->list);
 	spin_unlock_irq(&sdev->spinlock);
 
-	ib_destroy_cm_id(ch->cm_id);
-
 	if (ch->release_done)
 		complete(ch->release_done);
 

drivers/target/iscsi/iscsi_target.c

@@ -753,27 +753,32 @@ static void iscsit_unmap_iovec(struct iscsi_cmd *cmd)
 
 static void iscsit_ack_from_expstatsn(struct iscsi_conn *conn, u32 exp_statsn)
 {
-	struct iscsi_cmd *cmd;
+	LIST_HEAD(ack_list);
+	struct iscsi_cmd *cmd, *cmd_p;
 
 	conn->exp_statsn = exp_statsn;
 
 	if (conn->sess->sess_ops->RDMAExtensions)
 		return;
 
 	spin_lock_bh(&conn->cmd_lock);
-	list_for_each_entry(cmd, &conn->conn_cmd_list, i_conn_node) {
+	list_for_each_entry_safe(cmd, cmd_p, &conn->conn_cmd_list, i_conn_node) {
 		spin_lock(&cmd->istate_lock);
 		if ((cmd->i_state == ISTATE_SENT_STATUS) &&
 		    iscsi_sna_lt(cmd->stat_sn, exp_statsn)) {
 			cmd->i_state = ISTATE_REMOVE;
 			spin_unlock(&cmd->istate_lock);
-			iscsit_add_cmd_to_immediate_queue(cmd, conn,
-						cmd->i_state);
+			list_move_tail(&cmd->i_conn_node, &ack_list);
 			continue;
 		}
 		spin_unlock(&cmd->istate_lock);
 	}
 	spin_unlock_bh(&conn->cmd_lock);
+
+	list_for_each_entry_safe(cmd, cmd_p, &ack_list, i_conn_node) {
+		list_del(&cmd->i_conn_node);
+		iscsit_free_cmd(cmd, false);
+	}
 }
 
 static int iscsit_allocate_iovecs(struct iscsi_cmd *cmd)

drivers/target/iscsi/iscsi_target_nego.c

@@ -1192,7 +1192,7 @@ int iscsi_target_locate_portal(
 	 */
 alloc_tags:
 	tag_num = max_t(u32, ISCSIT_MIN_TAGS, queue_depth);
-	tag_num += ISCSIT_EXTRA_TAGS;
+	tag_num += (tag_num / 2) + ISCSIT_EXTRA_TAGS;
 	tag_size = sizeof(struct iscsi_cmd) + conn->conn_transport->priv_size;
 
 	ret = transport_alloc_session_tags(sess->se_sess, tag_num, tag_size);

drivers/target/iscsi/iscsi_target_util.c

@@ -736,7 +736,7 @@ void iscsit_free_cmd(struct iscsi_cmd *cmd, bool shutdown)
 		 * Fallthrough
 		 */
 	case ISCSI_OP_SCSI_TMFUNC:
-		rc = transport_generic_free_cmd(&cmd->se_cmd, 1);
+		rc = transport_generic_free_cmd(&cmd->se_cmd, shutdown);
 		if (!rc && shutdown && se_cmd && se_cmd->se_sess) {
 			__iscsit_free_cmd(cmd, true, shutdown);
 			target_put_sess_cmd(se_cmd->se_sess, se_cmd);
@@ -752,7 +752,7 @@ void iscsit_free_cmd(struct iscsi_cmd *cmd, bool shutdown)
 			se_cmd = &cmd->se_cmd;
 			__iscsit_free_cmd(cmd, true, shutdown);
 
-			rc = transport_generic_free_cmd(&cmd->se_cmd, 1);
+			rc = transport_generic_free_cmd(&cmd->se_cmd, shutdown);
 			if (!rc && shutdown && se_cmd->se_sess) {
 				__iscsit_free_cmd(cmd, true, shutdown);
 				target_put_sess_cmd(se_cmd->se_sess, se_cmd);

drivers/target/target_core_sbc.c

@@ -349,7 +349,16 @@ static sense_reason_t compare_and_write_post(struct se_cmd *cmd)
 {
 	struct se_device *dev = cmd->se_dev;
 
-	cmd->se_cmd_flags |= SCF_COMPARE_AND_WRITE_POST;
+	/*
+	 * Only set SCF_COMPARE_AND_WRITE_POST to force a response fall-through
+	 * within target_complete_ok_work() if the command was successfully
+	 * sent to the backend driver.
+	 */
+	spin_lock_irq(&cmd->t_state_lock);
+	if ((cmd->transport_state & CMD_T_SENT) && !cmd->scsi_status)
+		cmd->se_cmd_flags |= SCF_COMPARE_AND_WRITE_POST;
+	spin_unlock_irq(&cmd->t_state_lock);
+
 	/*
 	 * Unlock ->caw_sem originally obtained during sbc_compare_and_write()
 	 * before the original READ I/O submission.
@@ -363,7 +372,7 @@ static sense_reason_t compare_and_write_callback(struct se_cmd *cmd)
 {
 	struct se_device *dev = cmd->se_dev;
 	struct scatterlist *write_sg = NULL, *sg;
-	unsigned char *buf, *addr;
+	unsigned char *buf = NULL, *addr;
 	struct sg_mapping_iter m;
 	unsigned int offset = 0, len;
 	unsigned int nlbas = cmd->t_task_nolb;
@@ -378,6 +387,15 @@ static sense_reason_t compare_and_write_callback(struct se_cmd *cmd)
 	 */
 	if (!cmd->t_data_sg || !cmd->t_bidi_data_sg)
 		return TCM_NO_SENSE;
+	/*
+	 * Immediately exit + release dev->caw_sem if command has already
+	 * been failed with a non-zero SCSI status.
+	 */
+	if (cmd->scsi_status) {
+		pr_err("compare_and_write_callback: non zero scsi_status:"
+			" 0x%02x\n", cmd->scsi_status);
+		goto out;
+	}
 
 	buf = kzalloc(cmd->data_length, GFP_KERNEL);
 	if (!buf) {
@@ -508,6 +526,12 @@ sbc_compare_and_write(struct se_cmd *cmd)
 		cmd->transport_complete_callback = NULL;
 		return TCM_LOGICAL_UNIT_COMMUNICATION_FAILURE;
 	}
+	/*
+	 * Reset cmd->data_length to individual block_size in order to not
+	 * confuse backend drivers that depend on this value matching the
+	 * size of the I/O being submitted.
+	 */
+	cmd->data_length = cmd->t_task_nolb * dev->dev_attrib.block_size;
 
 	ret = cmd->execute_rw(cmd, cmd->t_bidi_data_sg, cmd->t_bidi_data_nents,
 			      DMA_FROM_DEVICE);

drivers/target/target_core_transport.c

@@ -236,17 +236,24 @@ int transport_alloc_session_tags(struct se_session *se_sess,
 {
 	int rc;
 
-	se_sess->sess_cmd_map = kzalloc(tag_num * tag_size, GFP_KERNEL);
+	se_sess->sess_cmd_map = kzalloc(tag_num * tag_size,
+					GFP_KERNEL | __GFP_NOWARN | __GFP_REPEAT);
 	if (!se_sess->sess_cmd_map) {
-		pr_err("Unable to allocate se_sess->sess_cmd_map\n");
-		return -ENOMEM;
+		se_sess->sess_cmd_map = vzalloc(tag_num * tag_size);
+		if (!se_sess->sess_cmd_map) {
+			pr_err("Unable to allocate se_sess->sess_cmd_map\n");
+			return -ENOMEM;
+		}
 	}
 
 	rc = percpu_ida_init(&se_sess->sess_tag_pool, tag_num);
 	if (rc < 0) {
 		pr_err("Unable to init se_sess->sess_tag_pool,"
 			" tag_num: %u\n", tag_num);
-		kfree(se_sess->sess_cmd_map);
+		if (is_vmalloc_addr(se_sess->sess_cmd_map))
+			vfree(se_sess->sess_cmd_map);
+		else
+			kfree(se_sess->sess_cmd_map);
 		se_sess->sess_cmd_map = NULL;
 		return -ENOMEM;
 	}
@@ -412,7 +419,10 @@ void transport_free_session(struct se_session *se_sess)
 {
 	if (se_sess->sess_cmd_map) {
 		percpu_ida_destroy(&se_sess->sess_tag_pool);
-		kfree(se_sess->sess_cmd_map);
+		if (is_vmalloc_addr(se_sess->sess_cmd_map))
+			vfree(se_sess->sess_cmd_map);
+		else
+			kfree(se_sess->sess_cmd_map);
 	}
 	kmem_cache_free(se_sess_cache, se_sess);
 }

drivers/target/target_core_xcopy.c

@@ -298,8 +298,8 @@ static int target_xcopy_parse_segdesc_02(struct se_cmd *se_cmd, struct xcopy_op
 		(unsigned long long)xop->dst_lba);
 
 	if (dc != 0) {
-		xop->dbl = (desc[29] << 16) & 0xff;
-		xop->dbl |= (desc[30] << 8) & 0xff;
+		xop->dbl = (desc[29] & 0xff) << 16;
+		xop->dbl |= (desc[30] & 0xff) << 8;
 		xop->dbl |= desc[31] & 0xff;
 
 		pr_debug("XCOPY seg desc 0x02: DC=1 w/ dbl: %u\n", xop->dbl);

drivers/vhost/scsi.c

@@ -728,7 +728,12 @@ vhost_scsi_get_tag(struct vhost_virtqueue *vq,
 	}
 	se_sess = tv_nexus->tvn_se_sess;
 
-	tag = percpu_ida_alloc(&se_sess->sess_tag_pool, GFP_KERNEL);
+	tag = percpu_ida_alloc(&se_sess->sess_tag_pool, GFP_ATOMIC);
+	if (tag < 0) {
+		pr_err("Unable to obtain tag for tcm_vhost_cmd\n");
+		return ERR_PTR(-ENOMEM);
+	}
+
 	cmd = &((struct tcm_vhost_cmd *)se_sess->sess_cmd_map)[tag];
 	sg = cmd->tvc_sgl;
 	pages = cmd->tvc_upages;