From: Geliang Tang <tanggeliang@kylinos.cn>

This patch defines the original in-kernel netlink path manager as a
new struct mptcp_pm_ops named "mptcp_pm_kernel", and register it in
mptcp_pm_kernel_register().

This mptcp_pm_ops will be skipped in mptcp_pm_unregister().

Signed-off-by: Geliang Tang <tanggeliang@kylinos.cn>
---
 net/mptcp/pm.c        |  4 ++++
 net/mptcp/pm_kernel.c | 26 ++++++++++++++++++++++++++
 net/mptcp/protocol.h  |  3 +++
 3 files changed, 33 insertions(+)

diff --git a/net/mptcp/pm.c b/net/mptcp/pm.c
index a2b210873b23..28ea8bdaa8b0 100644
--- a/net/mptcp/pm.c
+++ b/net/mptcp/pm.c
@@ -1076,6 +1076,10 @@ int mptcp_pm_register(struct mptcp_pm_ops *pm)
 
 void mptcp_pm_unregister(struct mptcp_pm_ops *pm)
 {
+	/* skip unregistering the default path manager */
+	if (pm == &mptcp_pm_kernel)
+		return;
+
 	spin_lock(&mptcp_pm_list_lock);
 	list_del_rcu(&pm->list);
 	spin_unlock(&mptcp_pm_list_lock);
diff --git a/net/mptcp/pm_kernel.c b/net/mptcp/pm_kernel.c
index 806a9b5b3c07..e6a1aef738a8 100644
--- a/net/mptcp/pm_kernel.c
+++ b/net/mptcp/pm_kernel.c
@@ -1398,8 +1398,34 @@ static struct pernet_operations mptcp_pm_pernet_ops = {
 	.size = sizeof(struct pm_nl_pernet),
 };
 
+static void mptcp_pm_nl_initialize(struct mptcp_sock *msk)
+{
+	bool subflows_allowed = !!mptcp_pm_get_subflows_max(msk);
+	struct mptcp_pm_data *pm = &msk->pm;
+
+	/* pm->work_pending must be only be set to 'true' when
+	 * pm is the default path manager
+	 */
+	WRITE_ONCE(pm->work_pending,
+		   (!!mptcp_pm_get_local_addr_max(msk) &&
+		    subflows_allowed) ||
+		   !!mptcp_pm_get_add_addr_signal_max(msk));
+	WRITE_ONCE(pm->accept_addr,
+		   !!mptcp_pm_get_add_addr_accept_max(msk) &&
+		   subflows_allowed);
+	WRITE_ONCE(pm->accept_subflow, subflows_allowed);
+}
+
+struct mptcp_pm_ops mptcp_pm_kernel = {
+	.init			= mptcp_pm_nl_initialize,
+	.name			= "kernel",
+	.owner			= THIS_MODULE,
+};
+
 void __init mptcp_pm_kernel_register(void)
 {
 	if (register_pernet_subsys(&mptcp_pm_pernet_ops) < 0)
 		panic("Failed to register MPTCP PM pernet subsystem.\n");
+
+	mptcp_pm_register(&mptcp_pm_kernel);
 }
diff --git a/net/mptcp/protocol.h b/net/mptcp/protocol.h
index 246b44db9775..f700cb55bf49 100644
--- a/net/mptcp/protocol.h
+++ b/net/mptcp/protocol.h
@@ -1050,6 +1050,9 @@ int mptcp_pm_remove_addr(struct mptcp_sock *msk, const struct mptcp_rm_list *rm_
 void mptcp_pm_remove_addr_entry(struct mptcp_sock *msk,
 				struct mptcp_pm_addr_entry *entry);
 
+/* the default path manager, used in mptcp_pm_unregister */
+extern struct mptcp_pm_ops mptcp_pm_kernel;
+
 struct mptcp_pm_ops *mptcp_pm_find(const char *name);
 int mptcp_pm_validate(struct mptcp_pm_ops *pm);
 int mptcp_pm_register(struct mptcp_pm_ops *pm);
-- 
2.43.0

Hi Geliang,

On 04/03/2025 12:40, Geliang Tang wrote:
> From: Geliang Tang <tanggeliang@kylinos.cn>
> 
> This patch defines the original in-kernel netlink path manager as a
> new struct mptcp_pm_ops named "mptcp_pm_kernel", and register it in
> mptcp_pm_kernel_register().
> 
> This mptcp_pm_ops will be skipped in mptcp_pm_unregister().
> 
> Signed-off-by: Geliang Tang <tanggeliang@kylinos.cn>
> ---
>  net/mptcp/pm.c        |  4 ++++
>  net/mptcp/pm_kernel.c | 26 ++++++++++++++++++++++++++
>  net/mptcp/protocol.h  |  3 +++
>  3 files changed, 33 insertions(+)
> 
> diff --git a/net/mptcp/pm.c b/net/mptcp/pm.c
> index a2b210873b23..28ea8bdaa8b0 100644
> --- a/net/mptcp/pm.c
> +++ b/net/mptcp/pm.c
> @@ -1076,6 +1076,10 @@ int mptcp_pm_register(struct mptcp_pm_ops *pm)
>  
>  void mptcp_pm_unregister(struct mptcp_pm_ops *pm)
>  {
> +	/* skip unregistering the default path manager */

Please see my questions from v7: why this skip?

When looking at this, I can understand that we don't want to unregister
built-in modules and the default one, but:

- When are we going to that? mptcp_pm_unregister() is still unused in
this series.

- Why would we want to unregister the userspace PM as well?

It makes sense to have an exception for the default one, but it feels
like we should simply not try to unregister the in-kernel ones. In other
words, there is probably no need to have such exceptions because
mptcp_pm_unregister() should never be called with the built-in PMs. In
this case, maybe we could add a WARN_ON_ONCE()?

  if (WARN_ON_ONCE(pm == &mptcp_pm_kernel))
       return;

(or something else if we need to catch the userspace PM as well, e.g.
pm->built_in, but that should not be needed)


> +	if (pm == &mptcp_pm_kernel)
> +		return;
> +
>  	spin_lock(&mptcp_pm_list_lock);
>  	list_del_rcu(&pm->list);
>  	spin_unlock(&mptcp_pm_list_lock);
> diff --git a/net/mptcp/pm_kernel.c b/net/mptcp/pm_kernel.c
> index 806a9b5b3c07..e6a1aef738a8 100644
> --- a/net/mptcp/pm_kernel.c
> +++ b/net/mptcp/pm_kernel.c
> @@ -1398,8 +1398,34 @@ static struct pernet_operations mptcp_pm_pernet_ops = {
>  	.size = sizeof(struct pm_nl_pernet),
>  };
>  
> +static void mptcp_pm_nl_initialize(struct mptcp_sock *msk)
> +{
> +	bool subflows_allowed = !!mptcp_pm_get_subflows_max(msk);
> +	struct mptcp_pm_data *pm = &msk->pm;
> +
> +	/* pm->work_pending must be only be set to 'true' when
> +	 * pm is the default path manager
> +	 */
> +	WRITE_ONCE(pm->work_pending,
> +		   (!!mptcp_pm_get_local_addr_max(msk) &&
> +		    subflows_allowed) ||
> +		   !!mptcp_pm_get_add_addr_signal_max(msk));
> +	WRITE_ONCE(pm->accept_addr,
> +		   !!mptcp_pm_get_add_addr_accept_max(msk) &&
> +		   subflows_allowed);
> +	WRITE_ONCE(pm->accept_subflow, subflows_allowed);

It might feel clearer to add this helper in patch 8 ("mptcp: pm:
initialize and release mptcp_pm_ops"), to understand you are moving
existing code here.

If you do that, then maybe better to squash the existing patches 6
("mptcp: pm: in-kernel: register mptcp_pm_kernel") and 7 ("mptcp: pm:
userspace: register mptcp_pm_userspace"), no?

  mptcp: pm: register in-kernel and userspace PM


> +}
> +
> +struct mptcp_pm_ops mptcp_pm_kernel = {
> +	.init			= mptcp_pm_nl_initialize,
> +	.name			= "kernel",
> +	.owner			= THIS_MODULE,
> +};
> +
>  void __init mptcp_pm_kernel_register(void)
>  {
>  	if (register_pernet_subsys(&mptcp_pm_pernet_ops) < 0)
>  		panic("Failed to register MPTCP PM pernet subsystem.\n");
> +
> +	mptcp_pm_register(&mptcp_pm_kernel);
>  }
> diff --git a/net/mptcp/protocol.h b/net/mptcp/protocol.h
> index 246b44db9775..f700cb55bf49 100644
> --- a/net/mptcp/protocol.h
> +++ b/net/mptcp/protocol.h
> @@ -1050,6 +1050,9 @@ int mptcp_pm_remove_addr(struct mptcp_sock *msk, const struct mptcp_rm_list *rm_
>  void mptcp_pm_remove_addr_entry(struct mptcp_sock *msk,
>  				struct mptcp_pm_addr_entry *entry);
>  
> +/* the default path manager, used in mptcp_pm_unregister */
(to be adapted if it is no longer used there. Or: mptcp_pm_initialize)

Or maybe better: it could be exported in patch 8 ("mptcp: pm: initialize
and release mptcp_pm_ops").

> +extern struct mptcp_pm_ops mptcp_pm_kernel;
> +
>  struct mptcp_pm_ops *mptcp_pm_find(const char *name);
>  int mptcp_pm_validate(struct mptcp_pm_ops *pm);
>  int mptcp_pm_register(struct mptcp_pm_ops *pm);

Cheers,
Matt
-- 
Sponsored by the NGI0 Core fund.

Hi Matt,

Thanks for the review.

On Wed, 2025-03-05 at 12:51 +0100, Matthieu Baerts wrote:
> Hi Geliang,
> 
> On 04/03/2025 12:40, Geliang Tang wrote:
> > From: Geliang Tang <tanggeliang@kylinos.cn>
> > 
> > This patch defines the original in-kernel netlink path manager as a
> > new struct mptcp_pm_ops named "mptcp_pm_kernel", and register it in
> > mptcp_pm_kernel_register().
> > 
> > This mptcp_pm_ops will be skipped in mptcp_pm_unregister().
> > 
> > Signed-off-by: Geliang Tang <tanggeliang@kylinos.cn>
> > ---
> >  net/mptcp/pm.c        |  4 ++++
> >  net/mptcp/pm_kernel.c | 26 ++++++++++++++++++++++++++
> >  net/mptcp/protocol.h  |  3 +++
> >  3 files changed, 33 insertions(+)
> > 
> > diff --git a/net/mptcp/pm.c b/net/mptcp/pm.c
> > index a2b210873b23..28ea8bdaa8b0 100644
> > --- a/net/mptcp/pm.c
> > +++ b/net/mptcp/pm.c
> > @@ -1076,6 +1076,10 @@ int mptcp_pm_register(struct mptcp_pm_ops
> > *pm)
> >  
> >  void mptcp_pm_unregister(struct mptcp_pm_ops *pm)
> >  {
> > +	/* skip unregistering the default path manager */
> 
> Please see my questions from v7: why this skip?

mptcp_pm_kernel is the default pm, skip it to ensure that there's
always a valid path manager available.

> 
> When looking at this, I can understand that we don't want to
> unregister
> built-in modules and the default one, but:
> 
> - When are we going to that? mptcp_pm_unregister() is still unused in
> this series.

mptcp_pm_unregister is not used in this set, but will be invoked
in .unreg of struct bpf_struct_ops.

> 
> - Why would we want to unregister the userspace PM as well?

The default one is mptcp_pm_kernel, not mptcp_pm_userspace, we set it
in mptcp_pm_ops_init when the input pm_ops is invalid.

> 
> It makes sense to have an exception for the default one, but it feels
> like we should simply not try to unregister the in-kernel ones. In
> other
> words, there is probably no need to have such exceptions because
> mptcp_pm_unregister() should never be called with the built-in PMs.
> In
> this case, maybe we could add a WARN_ON_ONCE()?
> 
>   if (WARN_ON_ONCE(pm == &mptcp_pm_kernel))
>        return;

Added this in v10.

> 
> (or something else if we need to catch the userspace PM as well, e.g.
> pm->built_in, but that should not be needed)
> 
> 
> > +	if (pm == &mptcp_pm_kernel)
> > +		return;
> > +
> >  	spin_lock(&mptcp_pm_list_lock);
> >  	list_del_rcu(&pm->list);
> >  	spin_unlock(&mptcp_pm_list_lock);
> > diff --git a/net/mptcp/pm_kernel.c b/net/mptcp/pm_kernel.c
> > index 806a9b5b3c07..e6a1aef738a8 100644
> > --- a/net/mptcp/pm_kernel.c
> > +++ b/net/mptcp/pm_kernel.c
> > @@ -1398,8 +1398,34 @@ static struct pernet_operations
> > mptcp_pm_pernet_ops = {
> >  	.size = sizeof(struct pm_nl_pernet),
> >  };
> >  
> > +static void mptcp_pm_nl_initialize(struct mptcp_sock *msk)
> > +{
> > +	bool subflows_allowed = !!mptcp_pm_get_subflows_max(msk);
> > +	struct mptcp_pm_data *pm = &msk->pm;
> > +
> > +	/* pm->work_pending must be only be set to 'true' when
> > +	 * pm is the default path manager
> > +	 */
> > +	WRITE_ONCE(pm->work_pending,
> > +		   (!!mptcp_pm_get_local_addr_max(msk) &&
> > +		    subflows_allowed) ||
> > +		   !!mptcp_pm_get_add_addr_signal_max(msk));
> > +	WRITE_ONCE(pm->accept_addr,
> > +		   !!mptcp_pm_get_add_addr_accept_max(msk) &&
> > +		   subflows_allowed);
> > +	WRITE_ONCE(pm->accept_subflow, subflows_allowed);
> 
> It might feel clearer to add this helper in patch 8 ("mptcp: pm:
> initialize and release mptcp_pm_ops"), to understand you are moving
> existing code here.
> 
> If you do that, then maybe better to squash the existing patches 6
> ("mptcp: pm: in-kernel: register mptcp_pm_kernel") and 7 ("mptcp: pm:
> userspace: register mptcp_pm_userspace"), no?
> 
>   mptcp: pm: register in-kernel and userspace PM

Done.

> 
> 
> > +}
> > +
> > +struct mptcp_pm_ops mptcp_pm_kernel = {
> > +	.init			= mptcp_pm_nl_initialize,
> > +	.name			= "kernel",
> > +	.owner			= THIS_MODULE,
> > +};
> > +
> >  void __init mptcp_pm_kernel_register(void)
> >  {
> >  	if (register_pernet_subsys(&mptcp_pm_pernet_ops) < 0)
> >  		panic("Failed to register MPTCP PM pernet
> > subsystem.\n");
> > +
> > +	mptcp_pm_register(&mptcp_pm_kernel);
> >  }
> > diff --git a/net/mptcp/protocol.h b/net/mptcp/protocol.h
> > index 246b44db9775..f700cb55bf49 100644
> > --- a/net/mptcp/protocol.h
> > +++ b/net/mptcp/protocol.h
> > @@ -1050,6 +1050,9 @@ int mptcp_pm_remove_addr(struct mptcp_sock
> > *msk, const struct mptcp_rm_list *rm_
> >  void mptcp_pm_remove_addr_entry(struct mptcp_sock *msk,
> >  				struct mptcp_pm_addr_entry
> > *entry);
> >  
> > +/* the default path manager, used in mptcp_pm_unregister */
> (to be adapted if it is no longer used there. Or:
> mptcp_pm_initialize)
> 
> Or maybe better: it could be exported in patch 8 ("mptcp: pm:
> initialize
> and release mptcp_pm_ops").

Done.

Thanks,
-Geliang

> 
> > +extern struct mptcp_pm_ops mptcp_pm_kernel;
> > +
> >  struct mptcp_pm_ops *mptcp_pm_find(const char *name);
> >  int mptcp_pm_validate(struct mptcp_pm_ops *pm);
> >  int mptcp_pm_register(struct mptcp_pm_ops *pm);
> 
> Cheers,
> Matt

Hi Geliang,

Thank you for the reply!

On 06/03/2025 12:09, Geliang Tang wrote:
> On Wed, 2025-03-05 at 12:51 +0100, Matthieu Baerts wrote:
>> On 04/03/2025 12:40, Geliang Tang wrote:

(...)

>> When looking at this, I can understand that we don't want to
>> unregister
>> built-in modules and the default one, but:
>>
>> - When are we going to that? mptcp_pm_unregister() is still unused in
>> this series.
> 
> mptcp_pm_unregister is not used in this set, but will be invoked
> in .unreg of struct bpf_struct_ops.

OK, that's clearer. So mptcp_pm_unregister() will only be called for the
BPF PM, and then not with the kernel/userspace PMs.

Cheers,
Matt
-- 
Sponsored by the NGI0 Core fund.

Hi Matt,

On Tue, 2025-03-04 at 19:40 +0800, Geliang Tang wrote:
> From: Geliang Tang <tanggeliang@kylinos.cn>
> 
> This patch defines the original in-kernel netlink path manager as a
> new struct mptcp_pm_ops named "mptcp_pm_kernel", and register it in
> mptcp_pm_kernel_register().
> 
> This mptcp_pm_ops will be skipped in mptcp_pm_unregister().
> 
> Signed-off-by: Geliang Tang <tanggeliang@kylinos.cn>
> ---
>  net/mptcp/pm.c        |  4 ++++
>  net/mptcp/pm_kernel.c | 26 ++++++++++++++++++++++++++
>  net/mptcp/protocol.h  |  3 +++
>  3 files changed, 33 insertions(+)
> 
> diff --git a/net/mptcp/pm.c b/net/mptcp/pm.c
> index a2b210873b23..28ea8bdaa8b0 100644
> --- a/net/mptcp/pm.c
> +++ b/net/mptcp/pm.c
> @@ -1076,6 +1076,10 @@ int mptcp_pm_register(struct mptcp_pm_ops *pm)
>  
>  void mptcp_pm_unregister(struct mptcp_pm_ops *pm)
>  {
> +	/* skip unregistering the default path manager */
> +	if (pm == &mptcp_pm_kernel)
> +		return;
> +
>  	spin_lock(&mptcp_pm_list_lock);
>  	list_del_rcu(&pm->list);
>  	spin_unlock(&mptcp_pm_list_lock);
> diff --git a/net/mptcp/pm_kernel.c b/net/mptcp/pm_kernel.c
> index 806a9b5b3c07..e6a1aef738a8 100644
> --- a/net/mptcp/pm_kernel.c
> +++ b/net/mptcp/pm_kernel.c
> @@ -1398,8 +1398,34 @@ static struct pernet_operations
> mptcp_pm_pernet_ops = {
>  	.size = sizeof(struct pm_nl_pernet),
>  };
>  
> +static void mptcp_pm_nl_initialize(struct mptcp_sock *msk)
> +{
> +	bool subflows_allowed = !!mptcp_pm_get_subflows_max(msk);
> +	struct mptcp_pm_data *pm = &msk->pm;
> +
> +	/* pm->work_pending must be only be set to 'true' when
> +	 * pm is the default path manager
> +	 */
> +	WRITE_ONCE(pm->work_pending,
> +		   (!!mptcp_pm_get_local_addr_max(msk) &&
> +		    subflows_allowed) ||
> +		   !!mptcp_pm_get_add_addr_signal_max(msk));
> +	WRITE_ONCE(pm->accept_addr,
> +		   !!mptcp_pm_get_add_addr_accept_max(msk) &&
> +		   subflows_allowed);
> +	WRITE_ONCE(pm->accept_subflow, subflows_allowed);
> +}
> +
> +struct mptcp_pm_ops mptcp_pm_kernel = {
> +	.init			= mptcp_pm_nl_initialize,

A better name would be mptcp_pm_nl_init, but this is already used in
pm_netlink.c. I would like to rename mptcp_pm_nl_init in pm_netlink.c
to mptcp_pm_genl_init, do you think this is a good idea?

If so, is it better to squash this renaming into "mptcp: pm: split
netlink and in-kernel init" or make it a separate patch?

Thanks,
-Geliang

> +	.name			= "kernel",
> +	.owner			= THIS_MODULE,
> +};
> +
>  void __init mptcp_pm_kernel_register(void)
>  {
>  	if (register_pernet_subsys(&mptcp_pm_pernet_ops) < 0)
>  		panic("Failed to register MPTCP PM pernet
> subsystem.\n");
> +
> +	mptcp_pm_register(&mptcp_pm_kernel);
>  }
> diff --git a/net/mptcp/protocol.h b/net/mptcp/protocol.h
> index 246b44db9775..f700cb55bf49 100644
> --- a/net/mptcp/protocol.h
> +++ b/net/mptcp/protocol.h
> @@ -1050,6 +1050,9 @@ int mptcp_pm_remove_addr(struct mptcp_sock
> *msk, const struct mptcp_rm_list *rm_
>  void mptcp_pm_remove_addr_entry(struct mptcp_sock *msk,
>  				struct mptcp_pm_addr_entry *entry);
>  
> +/* the default path manager, used in mptcp_pm_unregister */
> +extern struct mptcp_pm_ops mptcp_pm_kernel;
> +
>  struct mptcp_pm_ops *mptcp_pm_find(const char *name);
>  int mptcp_pm_validate(struct mptcp_pm_ops *pm);
>  int mptcp_pm_register(struct mptcp_pm_ops *pm);

Hi Geliang,

On 05/03/2025 02:35, Geliang Tang wrote:
> Hi Matt,
> 
> On Tue, 2025-03-04 at 19:40 +0800, Geliang Tang wrote:
>> From: Geliang Tang <tanggeliang@kylinos.cn>
>>
>> This patch defines the original in-kernel netlink path manager as a
>> new struct mptcp_pm_ops named "mptcp_pm_kernel", and register it in
>> mptcp_pm_kernel_register().
>>
>> This mptcp_pm_ops will be skipped in mptcp_pm_unregister().
>>
>> Signed-off-by: Geliang Tang <tanggeliang@kylinos.cn>
>> ---
>>  net/mptcp/pm.c        |  4 ++++
>>  net/mptcp/pm_kernel.c | 26 ++++++++++++++++++++++++++
>>  net/mptcp/protocol.h  |  3 +++
>>  3 files changed, 33 insertions(+)
>>
>> diff --git a/net/mptcp/pm.c b/net/mptcp/pm.c
>> index a2b210873b23..28ea8bdaa8b0 100644
>> --- a/net/mptcp/pm.c
>> +++ b/net/mptcp/pm.c
>> @@ -1076,6 +1076,10 @@ int mptcp_pm_register(struct mptcp_pm_ops *pm)
>>  
>>  void mptcp_pm_unregister(struct mptcp_pm_ops *pm)
>>  {
>> +	/* skip unregistering the default path manager */
>> +	if (pm == &mptcp_pm_kernel)
>> +		return;
>> +
>>  	spin_lock(&mptcp_pm_list_lock);
>>  	list_del_rcu(&pm->list);
>>  	spin_unlock(&mptcp_pm_list_lock);
>> diff --git a/net/mptcp/pm_kernel.c b/net/mptcp/pm_kernel.c
>> index 806a9b5b3c07..e6a1aef738a8 100644
>> --- a/net/mptcp/pm_kernel.c
>> +++ b/net/mptcp/pm_kernel.c
>> @@ -1398,8 +1398,34 @@ static struct pernet_operations
>> mptcp_pm_pernet_ops = {
>>  	.size = sizeof(struct pm_nl_pernet),
>>  };
>>  
>> +static void mptcp_pm_nl_initialize(struct mptcp_sock *msk)
>> +{
>> +	bool subflows_allowed = !!mptcp_pm_get_subflows_max(msk);
>> +	struct mptcp_pm_data *pm = &msk->pm;
>> +
>> +	/* pm->work_pending must be only be set to 'true' when
>> +	 * pm is the default path manager
>> +	 */
>> +	WRITE_ONCE(pm->work_pending,
>> +		   (!!mptcp_pm_get_local_addr_max(msk) &&
>> +		    subflows_allowed) ||
>> +		   !!mptcp_pm_get_add_addr_signal_max(msk));
>> +	WRITE_ONCE(pm->accept_addr,
>> +		   !!mptcp_pm_get_add_addr_accept_max(msk) &&
>> +		   subflows_allowed);
>> +	WRITE_ONCE(pm->accept_subflow, subflows_allowed);
>> +}
>> +
>> +struct mptcp_pm_ops mptcp_pm_kernel = {
>> +	.init			= mptcp_pm_nl_initialize,
> 
> A better name would be mptcp_pm_nl_init, but this is already used in
> pm_netlink.c. I would like to rename mptcp_pm_nl_init in pm_netlink.c
> to mptcp_pm_genl_init, do you think this is a good idea?
> 
> If so, is it better to squash this renaming into "mptcp: pm: split
> netlink and in-kernel init" or make it a separate patch?

What about calling the one here mptcp_pm_kernel_init()?

Cheers,
Matt
-- 
Sponsored by the NGI0 Core fund.

On Wed, 2025-03-05 at 10:11 +0100, Matthieu Baerts wrote:
> Hi Geliang,
> 
> On 05/03/2025 02:35, Geliang Tang wrote:
> > Hi Matt,
> > 
> > On Tue, 2025-03-04 at 19:40 +0800, Geliang Tang wrote:
> > > From: Geliang Tang <tanggeliang@kylinos.cn>
> > > 
> > > This patch defines the original in-kernel netlink path manager as
> > > a
> > > new struct mptcp_pm_ops named "mptcp_pm_kernel", and register it
> > > in
> > > mptcp_pm_kernel_register().
> > > 
> > > This mptcp_pm_ops will be skipped in mptcp_pm_unregister().
> > > 
> > > Signed-off-by: Geliang Tang <tanggeliang@kylinos.cn>
> > > ---
> > >  net/mptcp/pm.c        |  4 ++++
> > >  net/mptcp/pm_kernel.c | 26 ++++++++++++++++++++++++++
> > >  net/mptcp/protocol.h  |  3 +++
> > >  3 files changed, 33 insertions(+)
> > > 
> > > diff --git a/net/mptcp/pm.c b/net/mptcp/pm.c
> > > index a2b210873b23..28ea8bdaa8b0 100644
> > > --- a/net/mptcp/pm.c
> > > +++ b/net/mptcp/pm.c
> > > @@ -1076,6 +1076,10 @@ int mptcp_pm_register(struct mptcp_pm_ops
> > > *pm)
> > >  
> > >  void mptcp_pm_unregister(struct mptcp_pm_ops *pm)
> > >  {
> > > +	/* skip unregistering the default path manager */
> > > +	if (pm == &mptcp_pm_kernel)
> > > +		return;
> > > +
> > >  	spin_lock(&mptcp_pm_list_lock);
> > >  	list_del_rcu(&pm->list);
> > >  	spin_unlock(&mptcp_pm_list_lock);
> > > diff --git a/net/mptcp/pm_kernel.c b/net/mptcp/pm_kernel.c
> > > index 806a9b5b3c07..e6a1aef738a8 100644
> > > --- a/net/mptcp/pm_kernel.c
> > > +++ b/net/mptcp/pm_kernel.c
> > > @@ -1398,8 +1398,34 @@ static struct pernet_operations
> > > mptcp_pm_pernet_ops = {
> > >  	.size = sizeof(struct pm_nl_pernet),
> > >  };
> > >  
> > > +static void mptcp_pm_nl_initialize(struct mptcp_sock *msk)
> > > +{
> > > +	bool subflows_allowed =
> > > !!mptcp_pm_get_subflows_max(msk);
> > > +	struct mptcp_pm_data *pm = &msk->pm;
> > > +
> > > +	/* pm->work_pending must be only be set to 'true' when
> > > +	 * pm is the default path manager
> > > +	 */
> > > +	WRITE_ONCE(pm->work_pending,
> > > +		   (!!mptcp_pm_get_local_addr_max(msk) &&
> > > +		    subflows_allowed) ||
> > > +		   !!mptcp_pm_get_add_addr_signal_max(msk));
> > > +	WRITE_ONCE(pm->accept_addr,
> > > +		   !!mptcp_pm_get_add_addr_accept_max(msk) &&
> > > +		   subflows_allowed);
> > > +	WRITE_ONCE(pm->accept_subflow, subflows_allowed);
> > > +}
> > > +
> > > +struct mptcp_pm_ops mptcp_pm_kernel = {
> > > +	.init			= mptcp_pm_nl_initialize,
> > 
> > A better name would be mptcp_pm_nl_init, but this is already used
> > in
> > pm_netlink.c. I would like to rename mptcp_pm_nl_init in
> > pm_netlink.c
> > to mptcp_pm_genl_init, do you think this is a good idea?
> > 
> > If so, is it better to squash this renaming into "mptcp: pm: split
> > netlink and in-kernel init" or make it a separate patch?
> 
> What about calling the one here mptcp_pm_kernel_init()?

mptcp_pm_kernel_init is not good, because other functions start with
mptcp_pm_nl_:

struct mptcp_pm_ops mptcp_pm_kernel = {
        .get_local_id           = mptcp_pm_nl_get_local_id,
        .get_priority           = mptcp_pm_nl_is_backup,
        .established            = mptcp_pm_nl_fully_established,
        .subflow_established    = mptcp_pm_nl_subflow_established,
        .add_addr_echo          = mptcp_pm_nl_add_addr_echo,
        .add_addr_received      = mptcp_pm_nl_add_addr_received,
        .rm_addr_received       = mptcp_pm_nl_rm_addr_received,
        .rm_subflow_received    = mptcp_pm_nl_rm_subflow_received,
        .add_addr               = mptcp_pm_nl_add_addr,
        .del_addr               = mptcp_pm_nl_del_addr,
        .flush_addrs            = mptcp_pm_nl_flush_addrs,
        .set_priority           = mptcp_pm_nl_set_priority,
        .init                   = mptcp_pm_nl_init,
        .name                   = "kernel",
        .owner                  = THIS_MODULE,
};

Thanks,
-Geliang

> 
> Cheers,
> Matt

Hi Geliang,

On 05/03/2025 10:14, Geliang Tang wrote:
> On Wed, 2025-03-05 at 10:11 +0100, Matthieu Baerts wrote:
>> Hi Geliang,
>>
>> On 05/03/2025 02:35, Geliang Tang wrote:
>>> Hi Matt,
>>>
>>> On Tue, 2025-03-04 at 19:40 +0800, Geliang Tang wrote:
>>>> From: Geliang Tang <tanggeliang@kylinos.cn>
>>>>
>>>> This patch defines the original in-kernel netlink path manager as
>>>> a
>>>> new struct mptcp_pm_ops named "mptcp_pm_kernel", and register it
>>>> in
>>>> mptcp_pm_kernel_register().
>>>>
>>>> This mptcp_pm_ops will be skipped in mptcp_pm_unregister().
>>>>
>>>> Signed-off-by: Geliang Tang <tanggeliang@kylinos.cn>
>>>> ---
>>>>  net/mptcp/pm.c        |  4 ++++
>>>>  net/mptcp/pm_kernel.c | 26 ++++++++++++++++++++++++++
>>>>  net/mptcp/protocol.h  |  3 +++
>>>>  3 files changed, 33 insertions(+)
>>>>
>>>> diff --git a/net/mptcp/pm.c b/net/mptcp/pm.c
>>>> index a2b210873b23..28ea8bdaa8b0 100644
>>>> --- a/net/mptcp/pm.c
>>>> +++ b/net/mptcp/pm.c
>>>> @@ -1076,6 +1076,10 @@ int mptcp_pm_register(struct mptcp_pm_ops
>>>> *pm)
>>>>  
>>>>  void mptcp_pm_unregister(struct mptcp_pm_ops *pm)
>>>>  {
>>>> +	/* skip unregistering the default path manager */
>>>> +	if (pm == &mptcp_pm_kernel)
>>>> +		return;
>>>> +
>>>>  	spin_lock(&mptcp_pm_list_lock);
>>>>  	list_del_rcu(&pm->list);
>>>>  	spin_unlock(&mptcp_pm_list_lock);
>>>> diff --git a/net/mptcp/pm_kernel.c b/net/mptcp/pm_kernel.c
>>>> index 806a9b5b3c07..e6a1aef738a8 100644
>>>> --- a/net/mptcp/pm_kernel.c
>>>> +++ b/net/mptcp/pm_kernel.c
>>>> @@ -1398,8 +1398,34 @@ static struct pernet_operations
>>>> mptcp_pm_pernet_ops = {
>>>>  	.size = sizeof(struct pm_nl_pernet),
>>>>  };
>>>>  
>>>> +static void mptcp_pm_nl_initialize(struct mptcp_sock *msk)
>>>> +{
>>>> +	bool subflows_allowed =
>>>> !!mptcp_pm_get_subflows_max(msk);
>>>> +	struct mptcp_pm_data *pm = &msk->pm;
>>>> +
>>>> +	/* pm->work_pending must be only be set to 'true' when
>>>> +	 * pm is the default path manager
>>>> +	 */
>>>> +	WRITE_ONCE(pm->work_pending,
>>>> +		   (!!mptcp_pm_get_local_addr_max(msk) &&
>>>> +		    subflows_allowed) ||
>>>> +		   !!mptcp_pm_get_add_addr_signal_max(msk));
>>>> +	WRITE_ONCE(pm->accept_addr,
>>>> +		   !!mptcp_pm_get_add_addr_accept_max(msk) &&
>>>> +		   subflows_allowed);
>>>> +	WRITE_ONCE(pm->accept_subflow, subflows_allowed);
>>>> +}
>>>> +
>>>> +struct mptcp_pm_ops mptcp_pm_kernel = {
>>>> +	.init			= mptcp_pm_nl_initialize,
>>>
>>> A better name would be mptcp_pm_nl_init, but this is already used
>>> in
>>> pm_netlink.c. I would like to rename mptcp_pm_nl_init in
>>> pm_netlink.c
>>> to mptcp_pm_genl_init, do you think this is a good idea?
>>>
>>> If so, is it better to squash this renaming into "mptcp: pm: split
>>> netlink and in-kernel init" or make it a separate patch?
>>
>> What about calling the one here mptcp_pm_kernel_init()?
> 
> mptcp_pm_kernel_init is not good, because other functions start with
> mptcp_pm_nl_:
> 
> struct mptcp_pm_ops mptcp_pm_kernel = {
>         .get_local_id           = mptcp_pm_nl_get_local_id,
>         .get_priority           = mptcp_pm_nl_is_backup,
>         .established            = mptcp_pm_nl_fully_established,
>         .subflow_established    = mptcp_pm_nl_subflow_established,
>         .add_addr_echo          = mptcp_pm_nl_add_addr_echo,
>         .add_addr_received      = mptcp_pm_nl_add_addr_received,
>         .rm_addr_received       = mptcp_pm_nl_rm_addr_received,
>         .rm_subflow_received    = mptcp_pm_nl_rm_subflow_received,
>         .add_addr               = mptcp_pm_nl_add_addr,
>         .del_addr               = mptcp_pm_nl_del_addr,
>         .flush_addrs            = mptcp_pm_nl_flush_addrs,
>         .set_priority           = mptcp_pm_nl_set_priority,
>         .init                   = mptcp_pm_nl_init,
>         .name                   = "kernel",
>         .owner                  = THIS_MODULE,
> };

What about switching to the "mptcp_pm_kernel_" prefix when switching to
the new ops? For some of them, I guess there will be modifications
around the declaration of the function because 'static' will be added, no?

The "mptcp_pm_nl_" was making sense before because everything was in
pm_netlink.c. But now with the split, it might be good to take this
opportunity to rename the functions here to clearly mention it is from
the kernel PM, no?

Cheers,
Matt
-- 
Sponsored by the NGI0 Core fund.

On Wed, 2025-03-05 at 10:22 +0100, Matthieu Baerts wrote:
> Hi Geliang,
> 
> On 05/03/2025 10:14, Geliang Tang wrote:
> > On Wed, 2025-03-05 at 10:11 +0100, Matthieu Baerts wrote:
> > > Hi Geliang,
> > > 
> > > On 05/03/2025 02:35, Geliang Tang wrote:
> > > > Hi Matt,
> > > > 
> > > > On Tue, 2025-03-04 at 19:40 +0800, Geliang Tang wrote:
> > > > > From: Geliang Tang <tanggeliang@kylinos.cn>
> > > > > 
> > > > > This patch defines the original in-kernel netlink path
> > > > > manager as
> > > > > a
> > > > > new struct mptcp_pm_ops named "mptcp_pm_kernel", and register
> > > > > it
> > > > > in
> > > > > mptcp_pm_kernel_register().
> > > > > 
> > > > > This mptcp_pm_ops will be skipped in mptcp_pm_unregister().
> > > > > 
> > > > > Signed-off-by: Geliang Tang <tanggeliang@kylinos.cn>
> > > > > ---
> > > > >  net/mptcp/pm.c        |  4 ++++
> > > > >  net/mptcp/pm_kernel.c | 26 ++++++++++++++++++++++++++
> > > > >  net/mptcp/protocol.h  |  3 +++
> > > > >  3 files changed, 33 insertions(+)
> > > > > 
> > > > > diff --git a/net/mptcp/pm.c b/net/mptcp/pm.c
> > > > > index a2b210873b23..28ea8bdaa8b0 100644
> > > > > --- a/net/mptcp/pm.c
> > > > > +++ b/net/mptcp/pm.c
> > > > > @@ -1076,6 +1076,10 @@ int mptcp_pm_register(struct
> > > > > mptcp_pm_ops
> > > > > *pm)
> > > > >  
> > > > >  void mptcp_pm_unregister(struct mptcp_pm_ops *pm)
> > > > >  {
> > > > > +	/* skip unregistering the default path manager */
> > > > > +	if (pm == &mptcp_pm_kernel)
> > > > > +		return;
> > > > > +
> > > > >  	spin_lock(&mptcp_pm_list_lock);
> > > > >  	list_del_rcu(&pm->list);
> > > > >  	spin_unlock(&mptcp_pm_list_lock);
> > > > > diff --git a/net/mptcp/pm_kernel.c b/net/mptcp/pm_kernel.c
> > > > > index 806a9b5b3c07..e6a1aef738a8 100644
> > > > > --- a/net/mptcp/pm_kernel.c
> > > > > +++ b/net/mptcp/pm_kernel.c
> > > > > @@ -1398,8 +1398,34 @@ static struct pernet_operations
> > > > > mptcp_pm_pernet_ops = {
> > > > >  	.size = sizeof(struct pm_nl_pernet),
> > > > >  };
> > > > >  
> > > > > +static void mptcp_pm_nl_initialize(struct mptcp_sock *msk)
> > > > > +{
> > > > > +	bool subflows_allowed =
> > > > > !!mptcp_pm_get_subflows_max(msk);
> > > > > +	struct mptcp_pm_data *pm = &msk->pm;
> > > > > +
> > > > > +	/* pm->work_pending must be only be set to 'true'
> > > > > when
> > > > > +	 * pm is the default path manager
> > > > > +	 */
> > > > > +	WRITE_ONCE(pm->work_pending,
> > > > > +		   (!!mptcp_pm_get_local_addr_max(msk) &&
> > > > > +		    subflows_allowed) ||
> > > > > +		   !!mptcp_pm_get_add_addr_signal_max(msk));
> > > > > +	WRITE_ONCE(pm->accept_addr,
> > > > > +		   !!mptcp_pm_get_add_addr_accept_max(msk)
> > > > > &&
> > > > > +		   subflows_allowed);
> > > > > +	WRITE_ONCE(pm->accept_subflow, subflows_allowed);
> > > > > +}
> > > > > +
> > > > > +struct mptcp_pm_ops mptcp_pm_kernel = {
> > > > > +	.init			= mptcp_pm_nl_initialize,
> > > > 
> > > > A better name would be mptcp_pm_nl_init, but this is already
> > > > used
> > > > in
> > > > pm_netlink.c. I would like to rename mptcp_pm_nl_init in
> > > > pm_netlink.c
> > > > to mptcp_pm_genl_init, do you think this is a good idea?
> > > > 
> > > > If so, is it better to squash this renaming into "mptcp: pm:
> > > > split
> > > > netlink and in-kernel init" or make it a separate patch?
> > > 
> > > What about calling the one here mptcp_pm_kernel_init()?
> > 
> > mptcp_pm_kernel_init is not good, because other functions start
> > with
> > mptcp_pm_nl_:
> > 
> > struct mptcp_pm_ops mptcp_pm_kernel = {
> >         .get_local_id           = mptcp_pm_nl_get_local_id,
> >         .get_priority           = mptcp_pm_nl_is_backup,
> >         .established            = mptcp_pm_nl_fully_established,
> >         .subflow_established    = mptcp_pm_nl_subflow_established,
> >         .add_addr_echo          = mptcp_pm_nl_add_addr_echo,
> >         .add_addr_received      = mptcp_pm_nl_add_addr_received,
> >         .rm_addr_received       = mptcp_pm_nl_rm_addr_received,
> >         .rm_subflow_received    = mptcp_pm_nl_rm_subflow_received,
> >         .add_addr               = mptcp_pm_nl_add_addr,
> >         .del_addr               = mptcp_pm_nl_del_addr,
> >         .flush_addrs            = mptcp_pm_nl_flush_addrs,
> >         .set_priority           = mptcp_pm_nl_set_priority,
> >         .init                   = mptcp_pm_nl_init,
> >         .name                   = "kernel",
> >         .owner                  = THIS_MODULE,
> > };
> 
> What about switching to the "mptcp_pm_kernel_" prefix when switching
> to
> the new ops? For some of them, I guess there will be modifications
> around the declaration of the function because 'static' will be
> added, no?
> 
> The "mptcp_pm_nl_" was making sense before because everything was in
> pm_netlink.c. But now with the split, it might be good to take this
> opportunity to rename the functions here to clearly mention it is
> from
> the kernel PM, no?

Sure, I can do that in v9.

Thanks for your suggestion.

-Geliang

> 
> Cheers,
> Matt