patches for DPDK stable branches
 help / color / mirror / Atom feed
* [dpdk-stable] [PATCH 06/14] eal: mp: end the multiprocess thread during cleanup
       [not found] <20200104013341.19809-1-stephen@networkplumber.org>
@ 2020-01-04  1:33 ` Stephen Hemminger
  2020-04-27 15:37   ` [dpdk-stable] [dpdk-dev] " Burakov, Anatoly
  2020-01-04  1:33 ` [dpdk-stable] [PATCH 08/14] eal: vfio: cleanup the mp sync handle Stephen Hemminger
                   ` (3 subsequent siblings)
  4 siblings, 1 reply; 13+ messages in thread
From: Stephen Hemminger @ 2020-01-04  1:33 UTC (permalink / raw)
  To: dev; +Cc: Stephen Hemminger, qi.z.zhang, stable

When rte_eal_cleanup is called, all control threads should exit.
For the mp thread, this best handled by closing the mp_socket
and letting the thread see that.

This also fixes potential problems where the mp_socket gets
another hard error, and the thread runs away repeating itself
by reading the same error.

Fixes: 85d6815fa6d0 ("eal: close multi-process socket during cleanup")
Cc: qi.z.zhang@intel.com
Cc: stable@dpdk.org
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
---
 lib/librte_eal/common/eal_common_proc.c | 17 ++++++++++++-----
 1 file changed, 12 insertions(+), 5 deletions(-)

diff --git a/lib/librte_eal/common/eal_common_proc.c b/lib/librte_eal/common/eal_common_proc.c
index 935e8fefeba8..f369d8bf6dd8 100644
--- a/lib/librte_eal/common/eal_common_proc.c
+++ b/lib/librte_eal/common/eal_common_proc.c
@@ -276,8 +276,17 @@ read_msg(struct mp_msg_internal *m, struct sockaddr_un *s)
 	msgh.msg_control = control;
 	msgh.msg_controllen = sizeof(control);
 
+retry:
 	msglen = recvmsg(mp_fd, &msgh, 0);
+
+	/* zero length message means socket was closed */
+	if (msglen == 0)
+		return 0;
+
 	if (msglen < 0) {
+		if (errno == EINTR)
+			goto retry;
+
 		RTE_LOG(ERR, EAL, "recvmsg failed, %s\n", strerror(errno));
 		return -1;
 	}
@@ -305,7 +314,7 @@ read_msg(struct mp_msg_internal *m, struct sockaddr_un *s)
 		RTE_LOG(ERR, EAL, "invalid received data length\n");
 		return -1;
 	}
-	return 0;
+	return msglen;
 }
 
 static void
@@ -376,10 +385,8 @@ mp_handle(void *arg __rte_unused)
 	struct mp_msg_internal msg;
 	struct sockaddr_un sa;
 
-	while (1) {
-		if (read_msg(&msg, &sa) == 0)
-			process_msg(&msg, &sa);
-	}
+	while (read_msg(&msg, &sa) > 0)
+		process_msg(&msg, &sa);
 
 	return NULL;
 }
-- 
2.20.1


^ permalink raw reply	[flat|nested] 13+ messages in thread

* [dpdk-stable] [PATCH 08/14] eal: vfio: cleanup the mp sync handle
       [not found] <20200104013341.19809-1-stephen@networkplumber.org>
  2020-01-04  1:33 ` [dpdk-stable] [PATCH 06/14] eal: mp: end the multiprocess thread during cleanup Stephen Hemminger
@ 2020-01-04  1:33 ` Stephen Hemminger
  2020-04-27 12:12   ` Burakov, Anatoly
  2020-01-04  1:33 ` [dpdk-stable] [PATCH 10/14] tap: close netlink socket on device close Stephen Hemminger
                   ` (2 subsequent siblings)
  4 siblings, 1 reply; 13+ messages in thread
From: Stephen Hemminger @ 2020-01-04  1:33 UTC (permalink / raw)
  To: dev; +Cc: Stephen Hemminger, anatoly.burakov, stable

When rte_eal_cleanup is called the rte_mp_action for VFIO
should be freed.

Fixes: edf73dd33072 ("ipc: handle unsupported IPC in action register")
Cc: anatoly.burakov@intel.com
Cc: stable@dpdk.org
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
---
 lib/librte_eal/linux/eal/eal.c              | 3 +++
 lib/librte_eal/linux/eal/eal_vfio.h         | 1 +
 lib/librte_eal/linux/eal/eal_vfio_mp_sync.c | 8 ++++++++
 3 files changed, 12 insertions(+)

diff --git a/lib/librte_eal/linux/eal/eal.c b/lib/librte_eal/linux/eal/eal.c
index eb95f4f0c317..9ad81378f23c 100644
--- a/lib/librte_eal/linux/eal/eal.c
+++ b/lib/librte_eal/linux/eal/eal.c
@@ -1338,6 +1338,9 @@ rte_eal_cleanup(void)
 	}
 
 	rte_service_finalize();
+#ifdef VFIO_PRESENT
+	vfio_mp_sync_cleanup();
+#endif
 	rte_eal_intr_cleanup();
 	rte_eal_alarm_cleanup();
 	rte_mp_channel_cleanup();
diff --git a/lib/librte_eal/linux/eal/eal_vfio.h b/lib/librte_eal/linux/eal/eal_vfio.h
index cb2d35fb1206..bf7408a897a7 100644
--- a/lib/librte_eal/linux/eal/eal_vfio.h
+++ b/lib/librte_eal/linux/eal/eal_vfio.h
@@ -132,6 +132,7 @@ int
 vfio_has_supported_extensions(int vfio_container_fd);
 
 int vfio_mp_sync_setup(void);
+void vfio_mp_sync_cleanup(void);
 
 #define EAL_VFIO_MP "eal_vfio_mp_sync"
 
diff --git a/lib/librte_eal/linux/eal/eal_vfio_mp_sync.c b/lib/librte_eal/linux/eal/eal_vfio_mp_sync.c
index 5f2a5fc1d94e..b8ae9c65892e 100644
--- a/lib/librte_eal/linux/eal/eal_vfio_mp_sync.c
+++ b/lib/librte_eal/linux/eal/eal_vfio_mp_sync.c
@@ -120,4 +120,12 @@ vfio_mp_sync_setup(void)
 	return 0;
 }
 
+void
+vfio_mp_sync_cleanup(void)
+{
+	if (rte_eal_process_type() != RTE_PROC_PRIMARY)
+		return;
+
+	rte_mp_action_unregister(EAL_VFIO_MP);
+}
 #endif
-- 
2.20.1


^ permalink raw reply	[flat|nested] 13+ messages in thread

* [dpdk-stable] [PATCH 10/14] tap: close netlink socket on device close
       [not found] <20200104013341.19809-1-stephen@networkplumber.org>
  2020-01-04  1:33 ` [dpdk-stable] [PATCH 06/14] eal: mp: end the multiprocess thread during cleanup Stephen Hemminger
  2020-01-04  1:33 ` [dpdk-stable] [PATCH 08/14] eal: vfio: cleanup the mp sync handle Stephen Hemminger
@ 2020-01-04  1:33 ` Stephen Hemminger
  2020-04-25 15:52   ` [dpdk-stable] [dpdk-dev] " David Marchand
       [not found] ` <20200428231415.17985-1-stephen@networkplumber.org>
       [not found] ` <20200428235827.15383-1-stephen@networkplumber.org>
  4 siblings, 1 reply; 13+ messages in thread
From: Stephen Hemminger @ 2020-01-04  1:33 UTC (permalink / raw)
  To: dev; +Cc: Stephen Hemminger, pascal.mazon, stable

The netlink socket for flow creation was left open and never
closed.

Fixes: bf7b7f437b49 ("net/tap: create netdevice during probing")
Cc: pascal.mazon@6wind.com
Cc: stable@dpdk.org
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
---
 drivers/net/tap/rte_eth_tap.c | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/drivers/net/tap/rte_eth_tap.c b/drivers/net/tap/rte_eth_tap.c
index a13d8d50d7d7..d293dd8eeed5 100644
--- a/drivers/net/tap/rte_eth_tap.c
+++ b/drivers/net/tap/rte_eth_tap.c
@@ -1035,6 +1035,11 @@ tap_dev_close(struct rte_eth_dev *dev)
 				&internals->remote_initial_flags);
 	}
 
+	if (internals->nlsk_fd != -1) {
+		close(internals->nlsk_fd);
+		internals->nlsk_fd = -1;
+	}
+
 	if (internals->ka_fd != -1) {
 		close(internals->ka_fd);
 		internals->ka_fd = -1;
@@ -2406,7 +2411,7 @@ rte_pmd_tap_remove(struct rte_vdev_device *dev)
 	TAP_LOG(DEBUG, "Closing %s Ethernet device on numa %u",
 		tuntap_types[internals->type], rte_socket_id());
 
-	if (internals->nlsk_fd) {
+	if (internals->nlsk_fd != -1) {
 		tap_flow_flush(eth_dev, NULL);
 		tap_flow_implicit_flush(internals, NULL);
 		tap_nl_final(internals->nlsk_fd);
-- 
2.20.1


^ permalink raw reply	[flat|nested] 13+ messages in thread

* Re: [dpdk-stable] [dpdk-dev] [PATCH 10/14] tap: close netlink socket on device close
  2020-01-04  1:33 ` [dpdk-stable] [PATCH 10/14] tap: close netlink socket on device close Stephen Hemminger
@ 2020-04-25 15:52   ` David Marchand
  0 siblings, 0 replies; 13+ messages in thread
From: David Marchand @ 2020-04-25 15:52 UTC (permalink / raw)
  To: Stephen Hemminger; +Cc: dev, dpdk stable

On Sat, Jan 4, 2020 at 2:35 AM Stephen Hemminger
<stephen@networkplumber.org> wrote:
>
> The netlink socket for flow creation was left open and never
> closed.
>
> Fixes: bf7b7f437b49 ("net/tap: create netdevice during probing")
> Cc: pascal.mazon@6wind.com
> Cc: stable@dpdk.org
> Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>

Afaics, superseded by the "fixes for tap" series recently merged.
http://patchwork.dpdk.org/cover/68602/


-- 
David Marchand


^ permalink raw reply	[flat|nested] 13+ messages in thread

* Re: [dpdk-stable] [PATCH 08/14] eal: vfio: cleanup the mp sync handle
  2020-01-04  1:33 ` [dpdk-stable] [PATCH 08/14] eal: vfio: cleanup the mp sync handle Stephen Hemminger
@ 2020-04-27 12:12   ` Burakov, Anatoly
  0 siblings, 0 replies; 13+ messages in thread
From: Burakov, Anatoly @ 2020-04-27 12:12 UTC (permalink / raw)
  To: Stephen Hemminger, dev; +Cc: stable

On 04-Jan-20 1:33 AM, Stephen Hemminger wrote:
> When rte_eal_cleanup is called the rte_mp_action for VFIO
> should be freed.
> 
> Fixes: edf73dd33072 ("ipc: handle unsupported IPC in action register")
> Cc: anatoly.burakov@intel.com
> Cc: stable@dpdk.org
> Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
> ---

Acked-by: Anatoly Burakov <anatoly.burakov@intel.com>

-- 
Thanks,
Anatoly

^ permalink raw reply	[flat|nested] 13+ messages in thread

* Re: [dpdk-stable] [dpdk-dev] [PATCH 06/14] eal: mp: end the multiprocess thread during cleanup
  2020-01-04  1:33 ` [dpdk-stable] [PATCH 06/14] eal: mp: end the multiprocess thread during cleanup Stephen Hemminger
@ 2020-04-27 15:37   ` Burakov, Anatoly
  0 siblings, 0 replies; 13+ messages in thread
From: Burakov, Anatoly @ 2020-04-27 15:37 UTC (permalink / raw)
  To: Stephen Hemminger, dev; +Cc: qi.z.zhang, stable

On 04-Jan-20 1:33 AM, Stephen Hemminger wrote:
> When rte_eal_cleanup is called, all control threads should exit.
> For the mp thread, this best handled by closing the mp_socket
> and letting the thread see that.
> 
> This also fixes potential problems where the mp_socket gets
> another hard error, and the thread runs away repeating itself
> by reading the same error.
> 
> Fixes: 85d6815fa6d0 ("eal: close multi-process socket during cleanup")
> Cc: qi.z.zhang@intel.com
> Cc: stable@dpdk.org
> Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
> ---

Acked-by: Anatoly Burakov <anatoly.burakov@intel.com>

-- 
Thanks,
Anatoly

^ permalink raw reply	[flat|nested] 13+ messages in thread

* [dpdk-stable] [PATCH v2 5/8] eal: mp: end the multiprocess thread during cleanup
       [not found] ` <20200428231415.17985-1-stephen@networkplumber.org>
@ 2020-04-28 23:14   ` Stephen Hemminger
  2020-04-28 23:14   ` [dpdk-stable] [PATCH v2 6/8] eal: vfio: cleanup the mp sync handle Stephen Hemminger
  1 sibling, 0 replies; 13+ messages in thread
From: Stephen Hemminger @ 2020-04-28 23:14 UTC (permalink / raw)
  To: dev; +Cc: Stephen Hemminger, qi.z.zhang, stable, Anatoly Burakov

When rte_eal_cleanup is called, all control threads should exit.
For the mp thread, this best handled by closing the mp_socket
and letting the thread see that.

This also fixes potential problems where the mp_socket gets
another hard error, and the thread runs away repeating itself
by reading the same error.

Fixes: 85d6815fa6d0 ("eal: close multi-process socket during cleanup")
Cc: qi.z.zhang@intel.com
Cc: stable@dpdk.org
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Acked-by: Anatoly Burakov <anatoly.burakov@intel.com>
---
 lib/librte_eal/common/eal_common_proc.c | 17 ++++++++++++-----
 1 file changed, 12 insertions(+), 5 deletions(-)

diff --git a/lib/librte_eal/common/eal_common_proc.c b/lib/librte_eal/common/eal_common_proc.c
index 935e8fefeba8..f369d8bf6dd8 100644
--- a/lib/librte_eal/common/eal_common_proc.c
+++ b/lib/librte_eal/common/eal_common_proc.c
@@ -276,8 +276,17 @@ read_msg(struct mp_msg_internal *m, struct sockaddr_un *s)
 	msgh.msg_control = control;
 	msgh.msg_controllen = sizeof(control);
 
+retry:
 	msglen = recvmsg(mp_fd, &msgh, 0);
+
+	/* zero length message means socket was closed */
+	if (msglen == 0)
+		return 0;
+
 	if (msglen < 0) {
+		if (errno == EINTR)
+			goto retry;
+
 		RTE_LOG(ERR, EAL, "recvmsg failed, %s\n", strerror(errno));
 		return -1;
 	}
@@ -305,7 +314,7 @@ read_msg(struct mp_msg_internal *m, struct sockaddr_un *s)
 		RTE_LOG(ERR, EAL, "invalid received data length\n");
 		return -1;
 	}
-	return 0;
+	return msglen;
 }
 
 static void
@@ -376,10 +385,8 @@ mp_handle(void *arg __rte_unused)
 	struct mp_msg_internal msg;
 	struct sockaddr_un sa;
 
-	while (1) {
-		if (read_msg(&msg, &sa) == 0)
-			process_msg(&msg, &sa);
-	}
+	while (read_msg(&msg, &sa) > 0)
+		process_msg(&msg, &sa);
 
 	return NULL;
 }
-- 
2.20.1


^ permalink raw reply	[flat|nested] 13+ messages in thread

* [dpdk-stable] [PATCH v2 6/8] eal: vfio: cleanup the mp sync handle
       [not found] ` <20200428231415.17985-1-stephen@networkplumber.org>
  2020-04-28 23:14   ` [dpdk-stable] [PATCH v2 5/8] eal: mp: end the multiprocess thread during cleanup Stephen Hemminger
@ 2020-04-28 23:14   ` Stephen Hemminger
  1 sibling, 0 replies; 13+ messages in thread
From: Stephen Hemminger @ 2020-04-28 23:14 UTC (permalink / raw)
  To: dev; +Cc: Stephen Hemminger, anatoly.burakov, stable

When rte_eal_cleanup is called the rte_mp_action for VFIO
should be freed.

Fixes: edf73dd33072 ("ipc: handle unsupported IPC in action register")
Cc: anatoly.burakov@intel.com
Cc: stable@dpdk.org
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Acked-by: Anatoly Burakov <anatoly.burakov@intel.com>
---
 lib/librte_eal/linux/eal.c              | 3 +++
 lib/librte_eal/linux/eal_vfio.h         | 1 +
 lib/librte_eal/linux/eal_vfio_mp_sync.c | 8 ++++++++
 3 files changed, 12 insertions(+)

diff --git a/lib/librte_eal/linux/eal.c b/lib/librte_eal/linux/eal.c
index 27b768b15c4a..7c56dbf49508 100644
--- a/lib/librte_eal/linux/eal.c
+++ b/lib/librte_eal/linux/eal.c
@@ -1346,6 +1346,9 @@ rte_eal_cleanup(void)
 	}
 
 	rte_service_finalize();
+#ifdef VFIO_PRESENT
+	vfio_mp_sync_cleanup();
+#endif
 	rte_eal_alarm_cleanup();
 	rte_mp_channel_cleanup();
 	rte_trace_save();
diff --git a/lib/librte_eal/linux/eal_vfio.h b/lib/librte_eal/linux/eal_vfio.h
index cb2d35fb1206..bf7408a897a7 100644
--- a/lib/librte_eal/linux/eal_vfio.h
+++ b/lib/librte_eal/linux/eal_vfio.h
@@ -132,6 +132,7 @@ int
 vfio_has_supported_extensions(int vfio_container_fd);
 
 int vfio_mp_sync_setup(void);
+void vfio_mp_sync_cleanup(void);
 
 #define EAL_VFIO_MP "eal_vfio_mp_sync"
 
diff --git a/lib/librte_eal/linux/eal_vfio_mp_sync.c b/lib/librte_eal/linux/eal_vfio_mp_sync.c
index 5f2a5fc1d94e..b8ae9c65892e 100644
--- a/lib/librte_eal/linux/eal_vfio_mp_sync.c
+++ b/lib/librte_eal/linux/eal_vfio_mp_sync.c
@@ -120,4 +120,12 @@ vfio_mp_sync_setup(void)
 	return 0;
 }
 
+void
+vfio_mp_sync_cleanup(void)
+{
+	if (rte_eal_process_type() != RTE_PROC_PRIMARY)
+		return;
+
+	rte_mp_action_unregister(EAL_VFIO_MP);
+}
 #endif
-- 
2.20.1


^ permalink raw reply	[flat|nested] 13+ messages in thread

* [dpdk-stable] [PATCH v3 5/8] eal: mp: end the multiprocess thread during cleanup
       [not found] ` <20200428235827.15383-1-stephen@networkplumber.org>
@ 2020-04-28 23:58   ` Stephen Hemminger
  2020-04-28 23:58   ` [dpdk-stable] [PATCH v3 6/8] eal: vfio: cleanup the mp sync handle Stephen Hemminger
                     ` (3 subsequent siblings)
  4 siblings, 0 replies; 13+ messages in thread
From: Stephen Hemminger @ 2020-04-28 23:58 UTC (permalink / raw)
  To: dev; +Cc: Stephen Hemminger, qi.z.zhang, stable, Anatoly Burakov

When rte_eal_cleanup is called, all control threads should exit.
For the mp thread, this best handled by closing the mp_socket
and letting the thread see that.

This also fixes potential problems where the mp_socket gets
another hard error, and the thread runs away repeating itself
by reading the same error.

Fixes: 85d6815fa6d0 ("eal: close multi-process socket during cleanup")
Cc: qi.z.zhang@intel.com
Cc: stable@dpdk.org
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Acked-by: Anatoly Burakov <anatoly.burakov@intel.com>
---
 lib/librte_eal/common/eal_common_proc.c | 17 ++++++++++++-----
 1 file changed, 12 insertions(+), 5 deletions(-)

diff --git a/lib/librte_eal/common/eal_common_proc.c b/lib/librte_eal/common/eal_common_proc.c
index 935e8fefeba8..f369d8bf6dd8 100644
--- a/lib/librte_eal/common/eal_common_proc.c
+++ b/lib/librte_eal/common/eal_common_proc.c
@@ -276,8 +276,17 @@ read_msg(struct mp_msg_internal *m, struct sockaddr_un *s)
 	msgh.msg_control = control;
 	msgh.msg_controllen = sizeof(control);
 
+retry:
 	msglen = recvmsg(mp_fd, &msgh, 0);
+
+	/* zero length message means socket was closed */
+	if (msglen == 0)
+		return 0;
+
 	if (msglen < 0) {
+		if (errno == EINTR)
+			goto retry;
+
 		RTE_LOG(ERR, EAL, "recvmsg failed, %s\n", strerror(errno));
 		return -1;
 	}
@@ -305,7 +314,7 @@ read_msg(struct mp_msg_internal *m, struct sockaddr_un *s)
 		RTE_LOG(ERR, EAL, "invalid received data length\n");
 		return -1;
 	}
-	return 0;
+	return msglen;
 }
 
 static void
@@ -376,10 +385,8 @@ mp_handle(void *arg __rte_unused)
 	struct mp_msg_internal msg;
 	struct sockaddr_un sa;
 
-	while (1) {
-		if (read_msg(&msg, &sa) == 0)
-			process_msg(&msg, &sa);
-	}
+	while (read_msg(&msg, &sa) > 0)
+		process_msg(&msg, &sa);
 
 	return NULL;
 }
-- 
2.20.1


^ permalink raw reply	[flat|nested] 13+ messages in thread

* [dpdk-stable] [PATCH v3 6/8] eal: vfio: cleanup the mp sync handle
       [not found] ` <20200428235827.15383-1-stephen@networkplumber.org>
  2020-04-28 23:58   ` [dpdk-stable] [PATCH v3 5/8] eal: mp: end the multiprocess thread during cleanup Stephen Hemminger
@ 2020-04-28 23:58   ` Stephen Hemminger
       [not found]   ` <20211113002824.338343-1-stephen@networkplumber.org>
                     ` (2 subsequent siblings)
  4 siblings, 0 replies; 13+ messages in thread
From: Stephen Hemminger @ 2020-04-28 23:58 UTC (permalink / raw)
  To: dev; +Cc: Stephen Hemminger, anatoly.burakov, stable

When rte_eal_cleanup is called the rte_mp_action for VFIO
should be freed.

Fixes: edf73dd33072 ("ipc: handle unsupported IPC in action register")
Cc: anatoly.burakov@intel.com
Cc: stable@dpdk.org
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Acked-by: Anatoly Burakov <anatoly.burakov@intel.com>
---
 lib/librte_eal/linux/eal.c              | 3 +++
 lib/librte_eal/linux/eal_vfio.h         | 1 +
 lib/librte_eal/linux/eal_vfio_mp_sync.c | 8 ++++++++
 3 files changed, 12 insertions(+)

diff --git a/lib/librte_eal/linux/eal.c b/lib/librte_eal/linux/eal.c
index 27b768b15c4a..7c56dbf49508 100644
--- a/lib/librte_eal/linux/eal.c
+++ b/lib/librte_eal/linux/eal.c
@@ -1346,6 +1346,9 @@ rte_eal_cleanup(void)
 	}
 
 	rte_service_finalize();
+#ifdef VFIO_PRESENT
+	vfio_mp_sync_cleanup();
+#endif
 	rte_eal_alarm_cleanup();
 	rte_mp_channel_cleanup();
 	rte_trace_save();
diff --git a/lib/librte_eal/linux/eal_vfio.h b/lib/librte_eal/linux/eal_vfio.h
index cb2d35fb1206..bf7408a897a7 100644
--- a/lib/librte_eal/linux/eal_vfio.h
+++ b/lib/librte_eal/linux/eal_vfio.h
@@ -132,6 +132,7 @@ int
 vfio_has_supported_extensions(int vfio_container_fd);
 
 int vfio_mp_sync_setup(void);
+void vfio_mp_sync_cleanup(void);
 
 #define EAL_VFIO_MP "eal_vfio_mp_sync"
 
diff --git a/lib/librte_eal/linux/eal_vfio_mp_sync.c b/lib/librte_eal/linux/eal_vfio_mp_sync.c
index 5f2a5fc1d94e..b8ae9c65892e 100644
--- a/lib/librte_eal/linux/eal_vfio_mp_sync.c
+++ b/lib/librte_eal/linux/eal_vfio_mp_sync.c
@@ -120,4 +120,12 @@ vfio_mp_sync_setup(void)
 	return 0;
 }
 
+void
+vfio_mp_sync_cleanup(void)
+{
+	if (rte_eal_process_type() != RTE_PROC_PRIMARY)
+		return;
+
+	rte_mp_action_unregister(EAL_VFIO_MP);
+}
 #endif
-- 
2.20.1


^ permalink raw reply	[flat|nested] 13+ messages in thread

* [PATCH v4 3/5] eal: vfio: cleanup the mp sync handle
       [not found]   ` <20211113002824.338343-1-stephen@networkplumber.org>
@ 2021-11-13  0:28     ` Stephen Hemminger
  0 siblings, 0 replies; 13+ messages in thread
From: Stephen Hemminger @ 2021-11-13  0:28 UTC (permalink / raw)
  To: dev; +Cc: Stephen Hemminger, anatoly.burakov, stable

When rte_eal_cleanup is called the rte_mp_action for VFIO
should be freed.

Fixes: edf73dd33072 ("ipc: handle unsupported IPC in action register")
Cc: anatoly.burakov@intel.com
Cc: stable@dpdk.org
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Acked-by: Anatoly Burakov <anatoly.burakov@intel.com>
---
 lib/eal/linux/eal.c              | 4 ++++
 lib/eal/linux/eal_vfio.h         | 1 +
 lib/eal/linux/eal_vfio_mp_sync.c | 8 ++++++++
 3 files changed, 13 insertions(+)

diff --git a/lib/eal/linux/eal.c b/lib/eal/linux/eal.c
index b67030b2792e..ad8f130b509d 100644
--- a/lib/eal/linux/eal.c
+++ b/lib/eal/linux/eal.c
@@ -1362,7 +1362,11 @@ rte_eal_cleanup(void)
 
 	if (rte_eal_process_type() == RTE_PROC_PRIMARY)
 		rte_memseg_walk(mark_freeable, NULL);
+
 	rte_service_finalize();
+#ifdef VFIO_PRESENT
+	vfio_mp_sync_cleanup();
+#endif
 	rte_mp_channel_cleanup();
 	/* after this point, any DPDK pointers will become dangling */
 	rte_eal_memory_detach();
diff --git a/lib/eal/linux/eal_vfio.h b/lib/eal/linux/eal_vfio.h
index 6ebaca6a0cbf..921ee595388c 100644
--- a/lib/eal/linux/eal_vfio.h
+++ b/lib/eal/linux/eal_vfio.h
@@ -133,6 +133,7 @@ int
 vfio_has_supported_extensions(int vfio_container_fd);
 
 int vfio_mp_sync_setup(void);
+void vfio_mp_sync_cleanup(void);
 
 #define EAL_VFIO_MP "eal_vfio_mp_sync"
 
diff --git a/lib/eal/linux/eal_vfio_mp_sync.c b/lib/eal/linux/eal_vfio_mp_sync.c
index a2accfab3a38..d12bbaee648b 100644
--- a/lib/eal/linux/eal_vfio_mp_sync.c
+++ b/lib/eal/linux/eal_vfio_mp_sync.c
@@ -120,4 +120,12 @@ vfio_mp_sync_setup(void)
 	return 0;
 }
 
+void
+vfio_mp_sync_cleanup(void)
+{
+	if (rte_eal_process_type() != RTE_PROC_PRIMARY)
+		return;
+
+	rte_mp_action_unregister(EAL_VFIO_MP);
+}
 #endif
-- 
2.30.2


^ permalink raw reply	[flat|nested] 13+ messages in thread

* [PATCH v5 3/5] eal: vfio: cleanup the mp sync handle
       [not found]   ` <20211113033209.341027-1-stephen@networkplumber.org>
@ 2021-11-13  3:32     ` Stephen Hemminger
  0 siblings, 0 replies; 13+ messages in thread
From: Stephen Hemminger @ 2021-11-13  3:32 UTC (permalink / raw)
  To: dev; +Cc: Stephen Hemminger, anatoly.burakov, stable

When rte_eal_cleanup is called the rte_mp_action for VFIO
should be freed.

Fixes: edf73dd33072 ("ipc: handle unsupported IPC in action register")
Cc: anatoly.burakov@intel.com
Cc: stable@dpdk.org
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Acked-by: Anatoly Burakov <anatoly.burakov@intel.com>
---
 lib/eal/linux/eal.c              | 4 ++++
 lib/eal/linux/eal_vfio.h         | 1 +
 lib/eal/linux/eal_vfio_mp_sync.c | 8 ++++++++
 3 files changed, 13 insertions(+)

diff --git a/lib/eal/linux/eal.c b/lib/eal/linux/eal.c
index b67030b2792e..ad8f130b509d 100644
--- a/lib/eal/linux/eal.c
+++ b/lib/eal/linux/eal.c
@@ -1362,7 +1362,11 @@ rte_eal_cleanup(void)
 
 	if (rte_eal_process_type() == RTE_PROC_PRIMARY)
 		rte_memseg_walk(mark_freeable, NULL);
+
 	rte_service_finalize();
+#ifdef VFIO_PRESENT
+	vfio_mp_sync_cleanup();
+#endif
 	rte_mp_channel_cleanup();
 	/* after this point, any DPDK pointers will become dangling */
 	rte_eal_memory_detach();
diff --git a/lib/eal/linux/eal_vfio.h b/lib/eal/linux/eal_vfio.h
index 6ebaca6a0cbf..921ee595388c 100644
--- a/lib/eal/linux/eal_vfio.h
+++ b/lib/eal/linux/eal_vfio.h
@@ -133,6 +133,7 @@ int
 vfio_has_supported_extensions(int vfio_container_fd);
 
 int vfio_mp_sync_setup(void);
+void vfio_mp_sync_cleanup(void);
 
 #define EAL_VFIO_MP "eal_vfio_mp_sync"
 
diff --git a/lib/eal/linux/eal_vfio_mp_sync.c b/lib/eal/linux/eal_vfio_mp_sync.c
index a2accfab3a38..d12bbaee648b 100644
--- a/lib/eal/linux/eal_vfio_mp_sync.c
+++ b/lib/eal/linux/eal_vfio_mp_sync.c
@@ -120,4 +120,12 @@ vfio_mp_sync_setup(void)
 	return 0;
 }
 
+void
+vfio_mp_sync_cleanup(void)
+{
+	if (rte_eal_process_type() != RTE_PROC_PRIMARY)
+		return;
+
+	rte_mp_action_unregister(EAL_VFIO_MP);
+}
 #endif
-- 
2.30.2


^ permalink raw reply	[flat|nested] 13+ messages in thread

* [PATCH v6 3/5] eal: vfio: cleanup the mp sync handle
       [not found]   ` <20211113172257.6543-1-stephen@networkplumber.org>
@ 2021-11-13 17:22     ` Stephen Hemminger
  0 siblings, 0 replies; 13+ messages in thread
From: Stephen Hemminger @ 2021-11-13 17:22 UTC (permalink / raw)
  To: dev; +Cc: Stephen Hemminger, anatoly.burakov, stable

When rte_eal_cleanup is called the rte_mp_action for VFIO
should be freed.

Fixes: edf73dd33072 ("ipc: handle unsupported IPC in action register")
Cc: anatoly.burakov@intel.com
Cc: stable@dpdk.org
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Acked-by: Anatoly Burakov <anatoly.burakov@intel.com>
---
 lib/eal/linux/eal.c              | 4 ++++
 lib/eal/linux/eal_vfio.h         | 1 +
 lib/eal/linux/eal_vfio_mp_sync.c | 8 ++++++++
 3 files changed, 13 insertions(+)

diff --git a/lib/eal/linux/eal.c b/lib/eal/linux/eal.c
index b67030b2792e..ad8f130b509d 100644
--- a/lib/eal/linux/eal.c
+++ b/lib/eal/linux/eal.c
@@ -1362,7 +1362,11 @@ rte_eal_cleanup(void)
 
 	if (rte_eal_process_type() == RTE_PROC_PRIMARY)
 		rte_memseg_walk(mark_freeable, NULL);
+
 	rte_service_finalize();
+#ifdef VFIO_PRESENT
+	vfio_mp_sync_cleanup();
+#endif
 	rte_mp_channel_cleanup();
 	/* after this point, any DPDK pointers will become dangling */
 	rte_eal_memory_detach();
diff --git a/lib/eal/linux/eal_vfio.h b/lib/eal/linux/eal_vfio.h
index 6ebaca6a0cbf..921ee595388c 100644
--- a/lib/eal/linux/eal_vfio.h
+++ b/lib/eal/linux/eal_vfio.h
@@ -133,6 +133,7 @@ int
 vfio_has_supported_extensions(int vfio_container_fd);
 
 int vfio_mp_sync_setup(void);
+void vfio_mp_sync_cleanup(void);
 
 #define EAL_VFIO_MP "eal_vfio_mp_sync"
 
diff --git a/lib/eal/linux/eal_vfio_mp_sync.c b/lib/eal/linux/eal_vfio_mp_sync.c
index a2accfab3a38..d12bbaee648b 100644
--- a/lib/eal/linux/eal_vfio_mp_sync.c
+++ b/lib/eal/linux/eal_vfio_mp_sync.c
@@ -120,4 +120,12 @@ vfio_mp_sync_setup(void)
 	return 0;
 }
 
+void
+vfio_mp_sync_cleanup(void)
+{
+	if (rte_eal_process_type() != RTE_PROC_PRIMARY)
+		return;
+
+	rte_mp_action_unregister(EAL_VFIO_MP);
+}
 #endif
-- 
2.30.2


^ permalink raw reply	[flat|nested] 13+ messages in thread

end of thread, other threads:[~2021-11-13 17:23 UTC | newest]

Thread overview: 13+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
     [not found] <20200104013341.19809-1-stephen@networkplumber.org>
2020-01-04  1:33 ` [dpdk-stable] [PATCH 06/14] eal: mp: end the multiprocess thread during cleanup Stephen Hemminger
2020-04-27 15:37   ` [dpdk-stable] [dpdk-dev] " Burakov, Anatoly
2020-01-04  1:33 ` [dpdk-stable] [PATCH 08/14] eal: vfio: cleanup the mp sync handle Stephen Hemminger
2020-04-27 12:12   ` Burakov, Anatoly
2020-01-04  1:33 ` [dpdk-stable] [PATCH 10/14] tap: close netlink socket on device close Stephen Hemminger
2020-04-25 15:52   ` [dpdk-stable] [dpdk-dev] " David Marchand
     [not found] ` <20200428231415.17985-1-stephen@networkplumber.org>
2020-04-28 23:14   ` [dpdk-stable] [PATCH v2 5/8] eal: mp: end the multiprocess thread during cleanup Stephen Hemminger
2020-04-28 23:14   ` [dpdk-stable] [PATCH v2 6/8] eal: vfio: cleanup the mp sync handle Stephen Hemminger
     [not found] ` <20200428235827.15383-1-stephen@networkplumber.org>
2020-04-28 23:58   ` [dpdk-stable] [PATCH v3 5/8] eal: mp: end the multiprocess thread during cleanup Stephen Hemminger
2020-04-28 23:58   ` [dpdk-stable] [PATCH v3 6/8] eal: vfio: cleanup the mp sync handle Stephen Hemminger
     [not found]   ` <20211113002824.338343-1-stephen@networkplumber.org>
2021-11-13  0:28     ` [PATCH v4 3/5] " Stephen Hemminger
     [not found]   ` <20211113033209.341027-1-stephen@networkplumber.org>
2021-11-13  3:32     ` [PATCH v5 " Stephen Hemminger
     [not found]   ` <20211113172257.6543-1-stephen@networkplumber.org>
2021-11-13 17:22     ` [PATCH v6 " Stephen Hemminger

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).