[PATCH AUTOSEL 5.4 15/23] vhost scsi: fix cmd completion race

[PATCH AUTOSEL 5.4 15/23] vhost scsi: fix cmd completion race

[Sasha Levin]

From: Mike Christie <michael.christie@oracle.com>

[ Upstream commit 47a3565e8bb14ec48a75b48daf57aa830e2691f8 ]

We might not do the final se_cmd put from vhost_scsi_complete_cmd_work.
When the last put happens a little later then we could race where
vhost_scsi_complete_cmd_work does vhost_signal, the guest runs and sends
more IO, and vhost_scsi_handle_vq runs but does not find any free cmds.

This patch has us delay completing the cmd until the last lio core ref
is dropped. We then know that once we signal to the guest that the cmd
is completed that if it queues a new command it will find a free cmd.

Signed-off-by: Mike Christie <michael.christie@oracle.com>
Reviewed-by: Maurizio Lombardi <mlombard@redhat.com>
Link: https://lore.kernel.org/r/1604986403-4931-4-git-send-email-michael.christie@oracle.com
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Acked-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 drivers/vhost/scsi.c | 42 +++++++++++++++---------------------------
 1 file changed, 15 insertions(+), 27 deletions(-)

diff --git a/drivers/vhost/scsi.c b/drivers/vhost/scsi.c
index f63f84a257256..98c484149ac7f 100644
--- a/drivers/vhost/scsi.c
+++ b/drivers/vhost/scsi.c
@@ -320,7 +320,7 @@ static u32 vhost_scsi_tpg_get_inst_index(struct se_portal_group *se_tpg)
 	return 1;
 }
 
-static void vhost_scsi_release_cmd(struct se_cmd *se_cmd)
+static void vhost_scsi_release_cmd_res(struct se_cmd *se_cmd)
 {
 	struct vhost_scsi_cmd *tv_cmd = container_of(se_cmd,
 				struct vhost_scsi_cmd, tvc_se_cmd);
@@ -340,6 +340,16 @@ static void vhost_scsi_release_cmd(struct se_cmd *se_cmd)
 	target_free_tag(se_sess, se_cmd);
 }
 
+static void vhost_scsi_release_cmd(struct se_cmd *se_cmd)
+{
+	struct vhost_scsi_cmd *cmd = container_of(se_cmd,
+					struct vhost_scsi_cmd, tvc_se_cmd);
+	struct vhost_scsi *vs = cmd->tvc_vhost;
+
+	llist_add(&cmd->tvc_completion_list, &vs->vs_completion_list);
+	vhost_work_queue(&vs->dev, &vs->vs_completion_work);
+}
+
 static u32 vhost_scsi_sess_get_index(struct se_session *se_sess)
 {
 	return 0;
@@ -362,28 +372,15 @@ static int vhost_scsi_get_cmd_state(struct se_cmd *se_cmd)
 	return 0;
 }
 
-static void vhost_scsi_complete_cmd(struct vhost_scsi_cmd *cmd)
-{
-	struct vhost_scsi *vs = cmd->tvc_vhost;
-
-	llist_add(&cmd->tvc_completion_list, &vs->vs_completion_list);
-
-	vhost_work_queue(&vs->dev, &vs->vs_completion_work);
-}
-
 static int vhost_scsi_queue_data_in(struct se_cmd *se_cmd)
 {
-	struct vhost_scsi_cmd *cmd = container_of(se_cmd,
-				struct vhost_scsi_cmd, tvc_se_cmd);
-	vhost_scsi_complete_cmd(cmd);
+	transport_generic_free_cmd(se_cmd, 0);
 	return 0;
 }
 
 static int vhost_scsi_queue_status(struct se_cmd *se_cmd)
 {
-	struct vhost_scsi_cmd *cmd = container_of(se_cmd,
-				struct vhost_scsi_cmd, tvc_se_cmd);
-	vhost_scsi_complete_cmd(cmd);
+	transport_generic_free_cmd(se_cmd, 0);
 	return 0;
 }
 
@@ -429,15 +426,6 @@ vhost_scsi_allocate_evt(struct vhost_scsi *vs,
 	return evt;
 }
 
-static void vhost_scsi_free_cmd(struct vhost_scsi_cmd *cmd)
-{
-	struct se_cmd *se_cmd = &cmd->tvc_se_cmd;
-
-	/* TODO locking against target/backend threads? */
-	transport_generic_free_cmd(se_cmd, 0);
-
-}
-
 static int vhost_scsi_check_stop_free(struct se_cmd *se_cmd)
 {
 	return target_put_sess_cmd(se_cmd);
@@ -556,7 +544,7 @@ static void vhost_scsi_complete_cmd_work(struct vhost_work *work)
 		} else
 			pr_err("Faulted on virtio_scsi_cmd_resp\n");
 
-		vhost_scsi_free_cmd(cmd);
+		vhost_scsi_release_cmd_res(se_cmd);
 	}
 
 	vq = -1;
@@ -1088,7 +1076,7 @@ vhost_scsi_handle_vq(struct vhost_scsi *vs, struct vhost_virtqueue *vq)
 						      &prot_iter, exp_data_len,
 						      &data_iter))) {
 				vq_err(vq, "Failed to map iov to sgl\n");
-				vhost_scsi_release_cmd(&cmd->tvc_se_cmd);
+				vhost_scsi_release_cmd_res(&cmd->tvc_se_cmd);
 				goto err;
 			}
 		}
-- 
2.27.0

[PATCH AUTOSEL 5.4 16/23] vhost scsi: add lun parser helper

[Sasha Levin]

From: Mike Christie <michael.christie@oracle.com>

[ Upstream commit 18f1becb6948cd411fd01968a0a54af63732e73c ]

Move code to parse lun from req's lun_buf to helper, so tmf code
can use it in the next patch.

Signed-off-by: Mike Christie <michael.christie@oracle.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Acked-by: Jason Wang <jasowang@redhat.com>
Link: https://lore.kernel.org/r/1604986403-4931-5-git-send-email-michael.christie@oracle.com
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Acked-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 drivers/vhost/scsi.c | 9 +++++++--
 1 file changed, 7 insertions(+), 2 deletions(-)

diff --git a/drivers/vhost/scsi.c b/drivers/vhost/scsi.c
index 98c484149ac7f..6b816b3a65ea7 100644
--- a/drivers/vhost/scsi.c
+++ b/drivers/vhost/scsi.c
@@ -895,6 +895,11 @@ vhost_scsi_get_req(struct vhost_virtqueue *vq, struct vhost_scsi_ctx *vc,
 	return ret;
 }
 
+static u16 vhost_buf_to_lun(u8 *lun_buf)
+{
+	return ((lun_buf[2] << 8) | lun_buf[3]) & 0x3FFF;
+}
+
 static void
 vhost_scsi_handle_vq(struct vhost_scsi *vs, struct vhost_virtqueue *vq)
 {
@@ -1033,12 +1038,12 @@ vhost_scsi_handle_vq(struct vhost_scsi *vs, struct vhost_virtqueue *vq)
 			tag = vhost64_to_cpu(vq, v_req_pi.tag);
 			task_attr = v_req_pi.task_attr;
 			cdb = &v_req_pi.cdb[0];
-			lun = ((v_req_pi.lun[2] << 8) | v_req_pi.lun[3]) & 0x3FFF;
+			lun = vhost_buf_to_lun(v_req_pi.lun);
 		} else {
 			tag = vhost64_to_cpu(vq, v_req.tag);
 			task_attr = v_req.task_attr;
 			cdb = &v_req.cdb[0];
-			lun = ((v_req.lun[2] << 8) | v_req.lun[3]) & 0x3FFF;
+			lun = vhost_buf_to_lun(v_req.lun);
 		}
 		/*
 		 * Check that the received CDB size does not exceeded our
-- 
2.27.0

[PATCH AUTOSEL 5.4 17/23] vhost scsi: Add support for LUN resets.

[Sasha Levin]

From: Mike Christie <michael.christie@oracle.com>

[ Upstream commit efd838fec17bd8756da852a435800a7e6281bfbc ]

In newer versions of virtio-scsi we just reset the timer when an a
command times out, so TMFs are never sent for the cmd time out case.
However, in older kernels and for the TMF inject cases, we can still get
resets and we end up just failing immediately so the guest might see the
device get offlined and IO errors.

For the older kernel cases, we want the same end result as the
modern virtio-scsi driver where we let the lower levels fire their error
handling and handle the problem. And at the upper levels we want to
wait. This patch ties the LUN reset handling into the LIO TMF code which
will just wait for outstanding commands to complete like we are doing in
the modern virtio-scsi case.

Note: I did not handle the ABORT case to keep this simple. For ABORTs
LIO just waits on the cmd like how it does for the RESET case. If
an ABORT fails, the guest OS ends up escalating to LUN RESET, so in
the end we get the same behavior where we wait on the outstanding
cmds.

Signed-off-by: Mike Christie <michael.christie@oracle.com>
Link: https://lore.kernel.org/r/1604986403-4931-6-git-send-email-michael.christie@oracle.com
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Acked-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 drivers/vhost/scsi.c | 147 +++++++++++++++++++++++++++++++++++++++----
 1 file changed, 134 insertions(+), 13 deletions(-)

diff --git a/drivers/vhost/scsi.c b/drivers/vhost/scsi.c
index 6b816b3a65ea7..e37362fd2e935 100644
--- a/drivers/vhost/scsi.c
+++ b/drivers/vhost/scsi.c
@@ -140,6 +140,7 @@ struct vhost_scsi_tpg {
 	struct se_portal_group se_tpg;
 	/* Pointer back to vhost_scsi, protected by tv_tpg_mutex */
 	struct vhost_scsi *vhost_scsi;
+	struct list_head tmf_queue;
 };
 
 struct vhost_scsi_tport {
@@ -209,6 +210,20 @@ struct vhost_scsi {
 	int vs_events_nr; /* num of pending events, protected by vq->mutex */
 };
 
+struct vhost_scsi_tmf {
+	struct vhost_work vwork;
+	struct vhost_scsi_tpg *tpg;
+	struct vhost_scsi *vhost;
+	struct vhost_scsi_virtqueue *svq;
+	struct list_head queue_entry;
+
+	struct se_cmd se_cmd;
+	struct vhost_scsi_inflight *inflight;
+	struct iovec resp_iov;
+	int in_iovs;
+	int vq_desc;
+};
+
 /*
  * Context for processing request and control queue operations.
  */
@@ -340,14 +355,32 @@ static void vhost_scsi_release_cmd_res(struct se_cmd *se_cmd)
 	target_free_tag(se_sess, se_cmd);
 }
 
+static void vhost_scsi_release_tmf_res(struct vhost_scsi_tmf *tmf)
+{
+	struct vhost_scsi_tpg *tpg = tmf->tpg;
+	struct vhost_scsi_inflight *inflight = tmf->inflight;
+
+	mutex_lock(&tpg->tv_tpg_mutex);
+	list_add_tail(&tpg->tmf_queue, &tmf->queue_entry);
+	mutex_unlock(&tpg->tv_tpg_mutex);
+	vhost_scsi_put_inflight(inflight);
+}
+
 static void vhost_scsi_release_cmd(struct se_cmd *se_cmd)
 {
-	struct vhost_scsi_cmd *cmd = container_of(se_cmd,
+	if (se_cmd->se_cmd_flags & SCF_SCSI_TMR_CDB) {
+		struct vhost_scsi_tmf *tmf = container_of(se_cmd,
+					struct vhost_scsi_tmf, se_cmd);
+
+		vhost_work_queue(&tmf->vhost->dev, &tmf->vwork);
+	} else {
+		struct vhost_scsi_cmd *cmd = container_of(se_cmd,
 					struct vhost_scsi_cmd, tvc_se_cmd);
-	struct vhost_scsi *vs = cmd->tvc_vhost;
+		struct vhost_scsi *vs = cmd->tvc_vhost;
 
-	llist_add(&cmd->tvc_completion_list, &vs->vs_completion_list);
-	vhost_work_queue(&vs->dev, &vs->vs_completion_work);
+		llist_add(&cmd->tvc_completion_list, &vs->vs_completion_list);
+		vhost_work_queue(&vs->dev, &vs->vs_completion_work);
+	}
 }
 
 static u32 vhost_scsi_sess_get_index(struct se_session *se_sess)
@@ -386,7 +419,10 @@ static int vhost_scsi_queue_status(struct se_cmd *se_cmd)
 
 static void vhost_scsi_queue_tm_rsp(struct se_cmd *se_cmd)
 {
-	return;
+	struct vhost_scsi_tmf *tmf = container_of(se_cmd, struct vhost_scsi_tmf,
+						  se_cmd);
+
+	transport_generic_free_cmd(&tmf->se_cmd, 0);
 }
 
 static void vhost_scsi_aborted_task(struct se_cmd *se_cmd)
@@ -1117,9 +1153,9 @@ vhost_scsi_handle_vq(struct vhost_scsi *vs, struct vhost_virtqueue *vq)
 }
 
 static void
-vhost_scsi_send_tmf_reject(struct vhost_scsi *vs,
-			   struct vhost_virtqueue *vq,
-			   struct vhost_scsi_ctx *vc)
+vhost_scsi_send_tmf_resp(struct vhost_scsi *vs, struct vhost_virtqueue *vq,
+			 int in_iovs, int vq_desc, struct iovec *resp_iov,
+			 int tmf_resp_code)
 {
 	struct virtio_scsi_ctrl_tmf_resp rsp;
 	struct iov_iter iov_iter;
@@ -1127,17 +1163,87 @@ vhost_scsi_send_tmf_reject(struct vhost_scsi *vs,
 
 	pr_debug("%s\n", __func__);
 	memset(&rsp, 0, sizeof(rsp));
-	rsp.response = VIRTIO_SCSI_S_FUNCTION_REJECTED;
+	rsp.response = tmf_resp_code;
 
-	iov_iter_init(&iov_iter, READ, &vq->iov[vc->out], vc->in, sizeof(rsp));
+	iov_iter_init(&iov_iter, READ, resp_iov, in_iovs, sizeof(rsp));
 
 	ret = copy_to_iter(&rsp, sizeof(rsp), &iov_iter);
 	if (likely(ret == sizeof(rsp)))
-		vhost_add_used_and_signal(&vs->dev, vq, vc->head, 0);
+		vhost_add_used_and_signal(&vs->dev, vq, vq_desc, 0);
 	else
 		pr_err("Faulted on virtio_scsi_ctrl_tmf_resp\n");
 }
 
+static void vhost_scsi_tmf_resp_work(struct vhost_work *work)
+{
+	struct vhost_scsi_tmf *tmf = container_of(work, struct vhost_scsi_tmf,
+						  vwork);
+	int resp_code;
+
+	if (tmf->se_cmd.se_tmr_req->response == TMR_FUNCTION_COMPLETE)
+		resp_code = VIRTIO_SCSI_S_FUNCTION_SUCCEEDED;
+	else
+		resp_code = VIRTIO_SCSI_S_FUNCTION_REJECTED;
+
+	vhost_scsi_send_tmf_resp(tmf->vhost, &tmf->svq->vq, tmf->in_iovs,
+				 tmf->vq_desc, &tmf->resp_iov, resp_code);
+	vhost_scsi_release_tmf_res(tmf);
+}
+
+static void
+vhost_scsi_handle_tmf(struct vhost_scsi *vs, struct vhost_scsi_tpg *tpg,
+		      struct vhost_virtqueue *vq,
+		      struct virtio_scsi_ctrl_tmf_req *vtmf,
+		      struct vhost_scsi_ctx *vc)
+{
+	struct vhost_scsi_virtqueue *svq = container_of(vq,
+					struct vhost_scsi_virtqueue, vq);
+	struct vhost_scsi_tmf *tmf;
+
+	if (vhost32_to_cpu(vq, vtmf->subtype) !=
+	    VIRTIO_SCSI_T_TMF_LOGICAL_UNIT_RESET)
+		goto send_reject;
+
+	if (!tpg->tpg_nexus || !tpg->tpg_nexus->tvn_se_sess) {
+		pr_err("Unable to locate active struct vhost_scsi_nexus for LUN RESET.\n");
+		goto send_reject;
+	}
+
+	mutex_lock(&tpg->tv_tpg_mutex);
+	if (list_empty(&tpg->tmf_queue)) {
+		pr_err("Missing reserve TMF. Could not handle LUN RESET.\n");
+		mutex_unlock(&tpg->tv_tpg_mutex);
+		goto send_reject;
+	}
+
+	tmf = list_first_entry(&tpg->tmf_queue, struct vhost_scsi_tmf,
+			       queue_entry);
+	list_del_init(&tmf->queue_entry);
+	mutex_unlock(&tpg->tv_tpg_mutex);
+
+	tmf->tpg = tpg;
+	tmf->vhost = vs;
+	tmf->svq = svq;
+	tmf->resp_iov = vq->iov[vc->out];
+	tmf->vq_desc = vc->head;
+	tmf->in_iovs = vc->in;
+	tmf->inflight = vhost_scsi_get_inflight(vq);
+
+	if (target_submit_tmr(&tmf->se_cmd, tpg->tpg_nexus->tvn_se_sess, NULL,
+			      vhost_buf_to_lun(vtmf->lun), NULL,
+			      TMR_LUN_RESET, GFP_KERNEL, 0,
+			      TARGET_SCF_ACK_KREF) < 0) {
+		vhost_scsi_release_tmf_res(tmf);
+		goto send_reject;
+	}
+
+	return;
+
+send_reject:
+	vhost_scsi_send_tmf_resp(vs, vq, vc->in, vc->head, &vq->iov[vc->out],
+				 VIRTIO_SCSI_S_FUNCTION_REJECTED);
+}
+
 static void
 vhost_scsi_send_an_resp(struct vhost_scsi *vs,
 			struct vhost_virtqueue *vq,
@@ -1163,6 +1269,7 @@ vhost_scsi_send_an_resp(struct vhost_scsi *vs,
 static void
 vhost_scsi_ctl_handle_vq(struct vhost_scsi *vs, struct vhost_virtqueue *vq)
 {
+	struct vhost_scsi_tpg *tpg;
 	union {
 		__virtio32 type;
 		struct virtio_scsi_ctrl_an_req an;
@@ -1244,12 +1351,12 @@ vhost_scsi_ctl_handle_vq(struct vhost_scsi *vs, struct vhost_virtqueue *vq)
 		vc.req += typ_size;
 		vc.req_size -= typ_size;
 
-		ret = vhost_scsi_get_req(vq, &vc, NULL);
+		ret = vhost_scsi_get_req(vq, &vc, &tpg);
 		if (ret)
 			goto err;
 
 		if (v_req.type == VIRTIO_SCSI_T_TMF)
-			vhost_scsi_send_tmf_reject(vs, vq, &vc);
+			vhost_scsi_handle_tmf(vs, tpg, vq, &v_req.tmf, &vc);
 		else
 			vhost_scsi_send_an_resp(vs, vq, &vc);
 err:
@@ -1814,11 +1921,19 @@ static int vhost_scsi_port_link(struct se_portal_group *se_tpg,
 {
 	struct vhost_scsi_tpg *tpg = container_of(se_tpg,
 				struct vhost_scsi_tpg, se_tpg);
+	struct vhost_scsi_tmf *tmf;
+
+	tmf = kzalloc(sizeof(*tmf), GFP_KERNEL);
+	if (!tmf)
+		return -ENOMEM;
+	INIT_LIST_HEAD(&tmf->queue_entry);
+	vhost_work_init(&tmf->vwork, vhost_scsi_tmf_resp_work);
 
 	mutex_lock(&vhost_scsi_mutex);
 
 	mutex_lock(&tpg->tv_tpg_mutex);
 	tpg->tv_tpg_port_count++;
+	list_add_tail(&tmf->queue_entry, &tpg->tmf_queue);
 	mutex_unlock(&tpg->tv_tpg_mutex);
 
 	vhost_scsi_hotplug(tpg, lun);
@@ -1833,11 +1948,16 @@ static void vhost_scsi_port_unlink(struct se_portal_group *se_tpg,
 {
 	struct vhost_scsi_tpg *tpg = container_of(se_tpg,
 				struct vhost_scsi_tpg, se_tpg);
+	struct vhost_scsi_tmf *tmf;
 
 	mutex_lock(&vhost_scsi_mutex);
 
 	mutex_lock(&tpg->tv_tpg_mutex);
 	tpg->tv_tpg_port_count--;
+	tmf = list_first_entry(&tpg->tmf_queue, struct vhost_scsi_tmf,
+			       queue_entry);
+	list_del(&tmf->queue_entry);
+	kfree(tmf);
 	mutex_unlock(&tpg->tv_tpg_mutex);
 
 	vhost_scsi_hotunplug(tpg, lun);
@@ -2158,6 +2278,7 @@ vhost_scsi_make_tpg(struct se_wwn *wwn, const char *name)
 	}
 	mutex_init(&tpg->tv_tpg_mutex);
 	INIT_LIST_HEAD(&tpg->tv_tpg_list);
+	INIT_LIST_HEAD(&tpg->tmf_queue);
 	tpg->tport = tport;
 	tpg->tport_tpgt = tpgt;
 
-- 
2.27.0