[PATCH AUTOSEL 5.10 27/43] nvmet-tcp: fix a race condition between release_queue and io_work

[PATCH AUTOSEL 5.10 27/43] nvmet-tcp: fix a race condition between release_queue and io_work

[Sasha Levin]

From: Maurizio Lombardi <mlombard@redhat.com>

[ Upstream commit a208fc56721775987c1b86e20d86d7e0d017c0b2 ]

If the initiator executes a reset controller operation while
performing I/O, the target kernel will crash because of a race condition
between release_queue and io_work;
nvmet_tcp_uninit_data_in_cmds() may be executed while io_work
is running, calling flush_work() was not sufficient to
prevent this because io_work could requeue itself.

Fix this bug by using cancel_work_sync() to prevent io_work
from requeuing itself and set rcv_state to NVMET_TCP_RECV_ERR to
make sure we don't receive any more data from the socket.

Signed-off-by: Maurizio Lombardi <mlombard@redhat.com>
Reviewed-by: Keith Busch <kbusch@kernel.org>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Reviewed-by: John Meneghini <jmeneghi@redhat.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 drivers/nvme/target/tcp.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/drivers/nvme/target/tcp.c b/drivers/nvme/target/tcp.c
index 1251fd6e92780..f726964b56555 100644
--- a/drivers/nvme/target/tcp.c
+++ b/drivers/nvme/target/tcp.c
@@ -1407,7 +1407,9 @@ static void nvmet_tcp_release_queue_work(struct work_struct *w)
 	mutex_unlock(&nvmet_tcp_queue_mutex);
 
 	nvmet_tcp_restore_socket_callbacks(queue);
-	flush_work(&queue->io_work);
+	cancel_work_sync(&queue->io_work);
+	/* stop accepting incoming data */
+	queue->rcv_state = NVMET_TCP_RECV_ERR;
 
 	nvmet_tcp_uninit_data_in_cmds(queue);
 	nvmet_sq_destroy(&queue->nvme_sq);
-- 
2.33.0

[PATCH AUTOSEL 5.10 28/43] nvmet-tcp: add an helper to free the cmd buffers

[Sasha Levin]

From: Maurizio Lombardi <mlombard@redhat.com>

[ Upstream commit 69b85e1f1d1d1e49601ec3e85d2031188657cca2 ]

Makes the code easier to read and to debug.

Sets the freed pointers to NULL, it will be useful
when destroying the queues to understand if the commands'
buffers have been released already or not.

Signed-off-by: Maurizio Lombardi <mlombard@redhat.com>
Reviewed-by: Keith Busch <kbusch@kernel.org>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Reviewed-by: John Meneghini <jmeneghi@redhat.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 drivers/nvme/target/tcp.c | 28 +++++++++++++++++++---------
 1 file changed, 19 insertions(+), 9 deletions(-)

diff --git a/drivers/nvme/target/tcp.c b/drivers/nvme/target/tcp.c
index f726964b56555..f1b9b0c3a3a63 100644
--- a/drivers/nvme/target/tcp.c
+++ b/drivers/nvme/target/tcp.c
@@ -154,6 +154,8 @@ static struct workqueue_struct *nvmet_tcp_wq;
 static const struct nvmet_fabrics_ops nvmet_tcp_ops;
 static void nvmet_tcp_free_cmd(struct nvmet_tcp_cmd *c);
 static void nvmet_tcp_finish_cmd(struct nvmet_tcp_cmd *cmd);
+static void nvmet_tcp_free_cmd_buffers(struct nvmet_tcp_cmd *cmd);
+static void nvmet_tcp_unmap_pdu_iovec(struct nvmet_tcp_cmd *cmd);
 
 static inline u16 nvmet_tcp_cmd_tag(struct nvmet_tcp_queue *queue,
 		struct nvmet_tcp_cmd *cmd)
@@ -285,6 +287,16 @@ static int nvmet_tcp_check_ddgst(struct nvmet_tcp_queue *queue, void *pdu)
 	return 0;
 }
 
+static void nvmet_tcp_free_cmd_buffers(struct nvmet_tcp_cmd *cmd)
+{
+	WARN_ON(unlikely(cmd->nr_mapped > 0));
+
+	kfree(cmd->iov);
+	sgl_free(cmd->req.sg);
+	cmd->iov = NULL;
+	cmd->req.sg = NULL;
+}
+
 static void nvmet_tcp_unmap_pdu_iovec(struct nvmet_tcp_cmd *cmd)
 {
 	struct scatterlist *sg;
@@ -294,6 +306,8 @@ static void nvmet_tcp_unmap_pdu_iovec(struct nvmet_tcp_cmd *cmd)
 
 	for (i = 0; i < cmd->nr_mapped; i++)
 		kunmap(sg_page(&sg[i]));
+
+	cmd->nr_mapped = 0;
 }
 
 static void nvmet_tcp_map_pdu_iovec(struct nvmet_tcp_cmd *cmd)
@@ -375,7 +389,7 @@ static int nvmet_tcp_map_data(struct nvmet_tcp_cmd *cmd)
 
 	return 0;
 err:
-	sgl_free(cmd->req.sg);
+	nvmet_tcp_free_cmd_buffers(cmd);
 	return NVME_SC_INTERNAL;
 }
 
@@ -620,10 +634,8 @@ static int nvmet_try_send_data(struct nvmet_tcp_cmd *cmd, bool last_in_batch)
 		}
 	}
 
-	if (queue->nvme_sq.sqhd_disabled) {
-		kfree(cmd->iov);
-		sgl_free(cmd->req.sg);
-	}
+	if (queue->nvme_sq.sqhd_disabled)
+		nvmet_tcp_free_cmd_buffers(cmd);
 
 	return 1;
 
@@ -652,8 +664,7 @@ static int nvmet_try_send_response(struct nvmet_tcp_cmd *cmd,
 	if (left)
 		return -EAGAIN;
 
-	kfree(cmd->iov);
-	sgl_free(cmd->req.sg);
+	nvmet_tcp_free_cmd_buffers(cmd);
 	cmd->queue->snd_cmd = NULL;
 	nvmet_tcp_put_cmd(cmd);
 	return 1;
@@ -1376,8 +1387,7 @@ static void nvmet_tcp_finish_cmd(struct nvmet_tcp_cmd *cmd)
 {
 	nvmet_req_uninit(&cmd->req);
 	nvmet_tcp_unmap_pdu_iovec(cmd);
-	kfree(cmd->iov);
-	sgl_free(cmd->req.sg);
+	nvmet_tcp_free_cmd_buffers(cmd);
 }
 
 static void nvmet_tcp_uninit_data_in_cmds(struct nvmet_tcp_queue *queue)
-- 
2.33.0

[PATCH AUTOSEL 5.10 29/43] nvmet-tcp: fix memory leak when performing a controller reset

[Sasha Levin]

From: Maurizio Lombardi <mlombard@redhat.com>

[ Upstream commit af21250bb503a02e705b461886321e394b300524 ]

If a reset controller is executed while the initiator
is performing some I/O the driver may leak the memory allocated
for the commands' iovec.

Make sure that nvmet_tcp_uninit_data_in_cmds() releases
all the memory.

Signed-off-by: Maurizio Lombardi <mlombard@redhat.com>
Reviewed-by: Keith Busch <kbusch@kernel.org>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Reviewed-by: John Meneghini <jmeneghi@redhat.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 drivers/nvme/target/tcp.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/drivers/nvme/target/tcp.c b/drivers/nvme/target/tcp.c
index f1b9b0c3a3a63..8d669db40251b 100644
--- a/drivers/nvme/target/tcp.c
+++ b/drivers/nvme/target/tcp.c
@@ -1397,7 +1397,10 @@ static void nvmet_tcp_uninit_data_in_cmds(struct nvmet_tcp_queue *queue)
 
 	for (i = 0; i < queue->nr_cmds; i++, cmd++) {
 		if (nvmet_tcp_need_data_in(cmd))
-			nvmet_tcp_finish_cmd(cmd);
+			nvmet_req_uninit(&cmd->req);
+
+		nvmet_tcp_unmap_pdu_iovec(cmd);
+		nvmet_tcp_free_cmd_buffers(cmd);
 	}
 
 	if (!queue->nr_cmds && nvmet_tcp_need_data_in(&queue->connect)) {
-- 
2.33.0

[PATCH AUTOSEL 5.10 30/43] nvme-tcp: validate R2T PDU in nvme_tcp_handle_r2t()

[Sasha Levin]

From: Varun Prakash <varun@chelsio.com>

[ Upstream commit 1d3ef9c3a39e04be31155c27ebf80342350c3abf ]

If maxh2cdata < r2t_length then driver will form multiple
H2CData PDUs, validate R2T PDU in nvme_tcp_handle_r2t() to
reuse nvme_tcp_setup_h2c_data_pdu().

Also set req->state to NVME_TCP_SEND_H2C_PDU in
nvme_tcp_setup_h2c_data_pdu().

Signed-off-by: Varun Prakash <varun@chelsio.com>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 drivers/nvme/host/tcp.c | 55 ++++++++++++++++++-----------------------
 1 file changed, 24 insertions(+), 31 deletions(-)

diff --git a/drivers/nvme/host/tcp.c b/drivers/nvme/host/tcp.c
index e99d439894187..c8efa98192a4f 100644
--- a/drivers/nvme/host/tcp.c
+++ b/drivers/nvme/host/tcp.c
@@ -563,7 +563,7 @@ static int nvme_tcp_handle_comp(struct nvme_tcp_queue *queue,
 	return ret;
 }
 
-static int nvme_tcp_setup_h2c_data_pdu(struct nvme_tcp_request *req,
+static void nvme_tcp_setup_h2c_data_pdu(struct nvme_tcp_request *req,
 		struct nvme_tcp_r2t_pdu *pdu)
 {
 	struct nvme_tcp_data_pdu *data = req->pdu;
@@ -572,32 +572,11 @@ static int nvme_tcp_setup_h2c_data_pdu(struct nvme_tcp_request *req,
 	u8 hdgst = nvme_tcp_hdgst_len(queue);
 	u8 ddgst = nvme_tcp_ddgst_len(queue);
 
+	req->state = NVME_TCP_SEND_H2C_PDU;
+	req->offset = 0;
 	req->pdu_len = le32_to_cpu(pdu->r2t_length);
 	req->pdu_sent = 0;
 
-	if (unlikely(!req->pdu_len)) {
-		dev_err(queue->ctrl->ctrl.device,
-			"req %d r2t len is %u, probably a bug...\n",
-			rq->tag, req->pdu_len);
-		return -EPROTO;
-	}
-
-	if (unlikely(req->data_sent + req->pdu_len > req->data_len)) {
-		dev_err(queue->ctrl->ctrl.device,
-			"req %d r2t len %u exceeded data len %u (%zu sent)\n",
-			rq->tag, req->pdu_len, req->data_len,
-			req->data_sent);
-		return -EPROTO;
-	}
-
-	if (unlikely(le32_to_cpu(pdu->r2t_offset) < req->data_sent)) {
-		dev_err(queue->ctrl->ctrl.device,
-			"req %d unexpected r2t offset %u (expected %zu)\n",
-			rq->tag, le32_to_cpu(pdu->r2t_offset),
-			req->data_sent);
-		return -EPROTO;
-	}
-
 	memset(data, 0, sizeof(*data));
 	data->hdr.type = nvme_tcp_h2c_data;
 	data->hdr.flags = NVME_TCP_F_DATA_LAST;
@@ -613,7 +592,6 @@ static int nvme_tcp_setup_h2c_data_pdu(struct nvme_tcp_request *req,
 	data->command_id = nvme_cid(rq);
 	data->data_offset = pdu->r2t_offset;
 	data->data_length = cpu_to_le32(req->pdu_len);
-	return 0;
 }
 
 static int nvme_tcp_handle_r2t(struct nvme_tcp_queue *queue,
@@ -621,7 +599,7 @@ static int nvme_tcp_handle_r2t(struct nvme_tcp_queue *queue,
 {
 	struct nvme_tcp_request *req;
 	struct request *rq;
-	int ret;
+	u32 r2t_length = le32_to_cpu(pdu->r2t_length);
 
 	rq = nvme_find_rq(nvme_tcp_tagset(queue), pdu->command_id);
 	if (!rq) {
@@ -632,13 +610,28 @@ static int nvme_tcp_handle_r2t(struct nvme_tcp_queue *queue,
 	}
 	req = blk_mq_rq_to_pdu(rq);
 
-	ret = nvme_tcp_setup_h2c_data_pdu(req, pdu);
-	if (unlikely(ret))
-		return ret;
+	if (unlikely(!r2t_length)) {
+		dev_err(queue->ctrl->ctrl.device,
+			"req %d r2t len is %u, probably a bug...\n",
+			rq->tag, r2t_length);
+		return -EPROTO;
+	}
 
-	req->state = NVME_TCP_SEND_H2C_PDU;
-	req->offset = 0;
+	if (unlikely(req->data_sent + r2t_length > req->data_len)) {
+		dev_err(queue->ctrl->ctrl.device,
+			"req %d r2t len %u exceeded data len %u (%zu sent)\n",
+			rq->tag, r2t_length, req->data_len, req->data_sent);
+		return -EPROTO;
+	}
+
+	if (unlikely(le32_to_cpu(pdu->r2t_offset) < req->data_sent)) {
+		dev_err(queue->ctrl->ctrl.device,
+			"req %d unexpected r2t offset %u (expected %zu)\n",
+			rq->tag, le32_to_cpu(pdu->r2t_offset), req->data_sent);
+		return -EPROTO;
+	}
 
+	nvme_tcp_setup_h2c_data_pdu(req, pdu);
 	nvme_tcp_queue_request(req, false, true);
 
 	return 0;
-- 
2.33.0

[PATCH AUTOSEL 5.10 31/43] nvme-tcp: fix memory leak when freeing a queue

[Sasha Levin]

From: Maurizio Lombardi <mlombard@redhat.com>

[ Upstream commit a5053c92b3db71c3f7f9f13934ca620632828d06 ]

Release the page frag cache when tearing down the io queues

Signed-off-by: Maurizio Lombardi <mlombard@redhat.com>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Reviewed-by: John Meneghini <jmeneghi@redhat.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 drivers/nvme/host/tcp.c | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/drivers/nvme/host/tcp.c b/drivers/nvme/host/tcp.c
index c8efa98192a4f..aaad985175d60 100644
--- a/drivers/nvme/host/tcp.c
+++ b/drivers/nvme/host/tcp.c
@@ -1212,6 +1212,7 @@ static int nvme_tcp_alloc_async_req(struct nvme_tcp_ctrl *ctrl)
 
 static void nvme_tcp_free_queue(struct nvme_ctrl *nctrl, int qid)
 {
+	struct page *page;
 	struct nvme_tcp_ctrl *ctrl = to_tcp_ctrl(nctrl);
 	struct nvme_tcp_queue *queue = &ctrl->queues[qid];
 
@@ -1221,6 +1222,11 @@ static void nvme_tcp_free_queue(struct nvme_ctrl *nctrl, int qid)
 	if (queue->hdr_digest || queue->data_digest)
 		nvme_tcp_free_crypto(queue);
 
+	if (queue->pf_cache.va) {
+		page = virt_to_head_page(queue->pf_cache.va);
+		__page_frag_cache_drain(page, queue->pf_cache.pagecnt_bias);
+		queue->pf_cache.va = NULL;
+	}
 	sock_release(queue->sock);
 	kfree(queue->pdu);
 	mutex_destroy(&queue->queue_lock);
-- 
2.33.0

[PATCH AUTOSEL 5.10 32/43] nvme-pci: add NO APST quirk for Kioxia device

[Sasha Levin]

From: Enzo Matsumiya <ematsumiya@suse.de>

[ Upstream commit 5a6254d55e2a9f7919ead8580d7aa0c7a382b26a ]

This particular Kioxia device times out and aborts I/O during any load,
but it's more easily observable with discards (fstrim).

The device gets to a state that is also not possible to use
"nvme set-feature" to disable APST.
Booting with nvme_core.default_ps_max_latency=0 solves the issue.

We had a dozen or so of these devices behaving this same way in
customer environments.

Signed-off-by: Enzo Matsumiya <ematsumiya@suse.de>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 drivers/nvme/host/core.c | 14 ++++++++++++++
 1 file changed, 14 insertions(+)

diff --git a/drivers/nvme/host/core.c b/drivers/nvme/host/core.c
index 99b5152482fe4..c76f3a7fb1d18 100644
--- a/drivers/nvme/host/core.c
+++ b/drivers/nvme/host/core.c
@@ -2663,6 +2663,20 @@ static const struct nvme_core_quirk_entry core_quirks[] = {
 		.vid = 0x14a4,
 		.fr = "22301111",
 		.quirks = NVME_QUIRK_SIMPLE_SUSPEND,
+	},
+	{
+		/*
+		 * This Kioxia CD6-V Series / HPE PE8030 device times out and
+		 * aborts I/O during any load, but more easily reproducible
+		 * with discards (fstrim).
+		 *
+		 * The device is left in a state where it is also not possible
+		 * to use "nvme set-feature" to disable APST, but booting with
+		 * nvme_core.default_ps_max_latency=0 works.
+		 */
+		.vid = 0x1e0f,
+		.mn = "KCD6XVUL6T40",
+		.quirks = NVME_QUIRK_NO_APST,
 	}
 };
 
-- 
2.33.0

[PATCH AUTOSEL 5.10 33/43] nvme: fix write zeroes pi

[Sasha Levin]

From: Klaus Jensen <k.jensen@samsung.com>

[ Upstream commit 00b33cf3da726757aef636365bb52e9536434e9a ]

Write Zeroes sets PRACT when block integrity is enabled (as it should),
but neglects to also set the reftag which is expected by reads. This
causes protection errors on reads.

Fix this by setting the reftag for type 1 and 2 (for type 3, reads will
not check the reftag).

Signed-off-by: Klaus Jensen <k.jensen@samsung.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 drivers/nvme/host/core.c | 15 ++++++++++++---
 1 file changed, 12 insertions(+), 3 deletions(-)

diff --git a/drivers/nvme/host/core.c b/drivers/nvme/host/core.c
index c76f3a7fb1d18..9b57a9db8f1dc 100644
--- a/drivers/nvme/host/core.c
+++ b/drivers/nvme/host/core.c
@@ -750,10 +750,19 @@ static inline blk_status_t nvme_setup_write_zeroes(struct nvme_ns *ns,
 		cpu_to_le64(nvme_sect_to_lba(ns, blk_rq_pos(req)));
 	cmnd->write_zeroes.length =
 		cpu_to_le16((blk_rq_bytes(req) >> ns->lba_shift) - 1);
-	if (nvme_ns_has_pi(ns))
+
+	if (nvme_ns_has_pi(ns)) {
 		cmnd->write_zeroes.control = cpu_to_le16(NVME_RW_PRINFO_PRACT);
-	else
-		cmnd->write_zeroes.control = 0;
+
+		switch (ns->pi_type) {
+		case NVME_NS_DPS_PI_TYPE1:
+		case NVME_NS_DPS_PI_TYPE2:
+			cmnd->write_zeroes.reftag =
+				cpu_to_le32(t10_pi_ref_tag(req));
+			break;
+		}
+	}
+
 	return BLK_STS_OK;
 }
 
-- 
2.33.0