[PATCH] net: change netdev_unregister_timeout_secs min value to 1

netdev_unregister_timeout_secs=0 can lead to printing the
"waiting for dev to become free" message every jiffy.
This is too frequent and unnecessary.
Set the min value to 1 second.

Signed-off-by: Dmitry Vyukov <dvyukov@google.com>
Suggested-by: Eric Dumazet <edumazet@google.com>
Fixes: 5aa3afe107d9 ("net: make unregister netdev warning timeout configurable")
Cc: netdev@vger.kernel.org
Cc: linux-kernel@vger.kernel.org
---
 Documentation/admin-guide/sysctl/net.rst | 2 +-
 net/core/sysctl_net_core.c               | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/Documentation/admin-guide/sysctl/net.rst b/Documentation/admin-guide/sysctl/net.rst
index 2090bfc69aa50..c941b214e0b7f 100644
--- a/Documentation/admin-guide/sysctl/net.rst
+++ b/Documentation/admin-guide/sysctl/net.rst

@@ -320,7 +320,7 @@ waiting for a network device refcount to drop to 0 during device
 unregistration. A lower value may be useful during bisection to detect
 a leaked reference faster. A larger value may be useful to prevent false
 warnings on slow/loaded systems.
-Default value is 10, minimum 0, maximum 3600.
+Default value is 10, minimum 1, maximum 3600.
 
 optmem_max
 ----------

diff --git a/net/core/sysctl_net_core.c b/net/core/sysctl_net_core.c
index d84c8a1b280e2..c8496c1142c9d 100644
--- a/net/core/sysctl_net_core.c
+++ b/net/core/sysctl_net_core.c

@@ -577,7 +577,7 @@ static struct ctl_table net_core_table[] = {
 		.maxlen		= sizeof(unsigned int),
 		.mode		= 0644,
 		.proc_handler	= proc_dointvec_minmax,
-		.extra1		= SYSCTL_ZERO,
+		.extra1		= SYSCTL_ONE,
 		.extra2		= &int_3600,
 	},
 	{ }

base-commit: 84c7f6c33f42a12eb036ebf0f0e3670799304120
-- 
2.31.0.291.g576ba9dcdaf-goog

On 3/25/21 11:31 AM, Dmitry Vyukov wrote:

netdev_unregister_timeout_secs=0 can lead to printing the
"waiting for dev to become free" message every jiffy.
This is too frequent and unnecessary.
Set the min value to 1 second.

Signed-off-by: Dmitry Vyukov <dvyukov@google.com>
Suggested-by: Eric Dumazet <edumazet@google.com>
Fixes: 5aa3afe107d9 ("net: make unregister netdev warning timeout configurable")
Cc: netdev@vger.kernel.org
Cc: linux-kernel@vger.kernel.org
---

Please respin your patch, and fix the merge issue [1]

For networking patches it is customary to tell if its for net or net-next tree.

[1]

diff --git a/net/core/dev.c b/net/core/dev.c
index 4bb6dcdbed8b856c03dc4af8b7fafe08984e803f..7bb00b8b86c6494c033cf57460f96ff3adebe081 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c

@@ -10431,7 +10431,7 @@ static void netdev_wait_allrefs(struct net_device *dev)
 
                refcnt = netdev_refcnt_read(dev);
 
-               if (refcnt &&
+               if (refcnt != 1 &&
                    time_after(jiffies, warning_time +
                               netdev_unregister_timeout_secs * HZ)) {
                        pr_emerg("unregister_netdevice: waiting for %s to become free. Usage count = %d\n",

On Thu, Mar 25, 2021 at 3:34 PM Eric Dumazet [off-list ref] wrote:

On 3/25/21 11:31 AM, Dmitry Vyukov wrote:

quoted

netdev_unregister_timeout_secs=0 can lead to printing the
"waiting for dev to become free" message every jiffy.
This is too frequent and unnecessary.
Set the min value to 1 second.

Signed-off-by: Dmitry Vyukov <dvyukov@google.com>
Suggested-by: Eric Dumazet <edumazet@google.com>
Fixes: 5aa3afe107d9 ("net: make unregister netdev warning timeout configurable")
Cc: netdev@vger.kernel.org
Cc: linux-kernel@vger.kernel.org
---

Please respin your patch, and fix the merge issue [1]

Is net-next rebuilt and rebased? Do I send v4 of the whole change?
I cannot base it on net-next now, because net-next already includes
most of it... so what should I use as base then?

On 3/25/21 3:38 PM, Dmitry Vyukov wrote:

On Thu, Mar 25, 2021 at 3:34 PM Eric Dumazet [off-list ref] wrote:

quoted

On 3/25/21 11:31 AM, Dmitry Vyukov wrote:

quoted

netdev_unregister_timeout_secs=0 can lead to printing the
"waiting for dev to become free" message every jiffy.
This is too frequent and unnecessary.
Set the min value to 1 second.

Signed-off-by: Dmitry Vyukov <dvyukov@google.com>
Suggested-by: Eric Dumazet <edumazet@google.com>
Fixes: 5aa3afe107d9 ("net: make unregister netdev warning timeout configurable")
Cc: netdev@vger.kernel.org
Cc: linux-kernel@vger.kernel.org
---

Please respin your patch, and fix the merge issue [1]

Is net-next rebuilt and rebased? Do I send v4 of the whole change?
I cannot base it on net-next now, because net-next already includes
most of it... so what should I use as base then?

quoted

For networking patches it is customary to tell if its for net or net-next tree.

[1]

diff --git a/net/core/dev.c b/net/core/dev.c
index 4bb6dcdbed8b856c03dc4af8b7fafe08984e803f..7bb00b8b86c6494c033cf57460f96ff3adebe081 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c

@@ -10431,7 +10431,7 @@ static void netdev_wait_allrefs(struct net_device *dev)

                refcnt = netdev_refcnt_read(dev);

-               if (refcnt &&
+               if (refcnt != 1 &&
                    time_after(jiffies, warning_time +
                               netdev_unregister_timeout_secs * HZ)) {
                        pr_emerg("unregister_netdevice: waiting for %s to become free. Usage count = %d\n",

Please include my fix into your patch.

Send a V2, based on current net-next.

net-next is never rebased, we have to fix the bug by adding a fix on top of it.

On Thu, Mar 25, 2021 at 3:43 PM Eric Dumazet [off-list ref] wrote:

On 3/25/21 3:38 PM, Dmitry Vyukov wrote:

quoted

On Thu, Mar 25, 2021 at 3:34 PM Eric Dumazet [off-list ref] wrote:

quoted

On 3/25/21 11:31 AM, Dmitry Vyukov wrote:

quoted

netdev_unregister_timeout_secs=0 can lead to printing the
"waiting for dev to become free" message every jiffy.
This is too frequent and unnecessary.
Set the min value to 1 second.

Signed-off-by: Dmitry Vyukov <dvyukov@google.com>
Suggested-by: Eric Dumazet <edumazet@google.com>
Fixes: 5aa3afe107d9 ("net: make unregister netdev warning timeout configurable")
Cc: netdev@vger.kernel.org
Cc: linux-kernel@vger.kernel.org
---

Please respin your patch, and fix the merge issue [1]

Is net-next rebuilt and rebased? Do I send v4 of the whole change?
I cannot base it on net-next now, because net-next already includes
most of it... so what should I use as base then?

quoted

For networking patches it is customary to tell if its for net or net-next tree.

[1]

diff --git a/net/core/dev.c b/net/core/dev.c
index 4bb6dcdbed8b856c03dc4af8b7fafe08984e803f..7bb00b8b86c6494c033cf57460f96ff3adebe081 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c

@@ -10431,7 +10431,7 @@ static void netdev_wait_allrefs(struct net_device *dev)

                refcnt = netdev_refcnt_read(dev);

-               if (refcnt &&
+               if (refcnt != 1 &&
                    time_after(jiffies, warning_time +
                               netdev_unregister_timeout_secs * HZ)) {
                        pr_emerg("unregister_netdevice: waiting for %s to become free. Usage count = %d\n",

Please include my fix into your patch.

Send a V2, based on current net-next.

net-next is never rebased, we have to fix the bug by adding a fix on top of it.

Ah, got it. Mailed:
[PATCH net-next v2] net: change netdev_unregister_timeout_secs min value to 1