[PATCH mptcp-next v2] mptcp: update userspace pm subflows info

Geliang Tang posted 1 patch 1 year, 2 months ago
Patches applied successfully (tree, apply log)
git fetch https://github.com/multipath-tcp/mptcp_net-next tags/patchew/d262f6beb6899f0a79acbeb1d3c8c855e95e098a.1673599496.git.geliang.tang@suse.com
Maintainers: Mat Martineau <mathew.j.martineau@linux.intel.com>, Matthieu Baerts <matthieu.baerts@tessares.net>, "David S. Miller" <davem@davemloft.net>, Eric Dumazet <edumazet@google.com>, Jakub Kicinski <kuba@kernel.org>, Paolo Abeni <pabeni@redhat.com>
net/mptcp/pm_userspace.c | 7 +++++++
1 file changed, 7 insertions(+)
[PATCH mptcp-next v2] mptcp: update userspace pm subflows info
Posted by Geliang Tang 1 year, 2 months ago
Increase pm subflows counter when userspace pm creates a new subflow,
and decrease the counter when it closes a subflow.

Closes: https://github.com/multipath-tcp/mptcp_net-next/issues/329
Signed-off-by: Geliang Tang <geliang.tang@suse.com>
---
v2:
 - hold pm locks.
---
 net/mptcp/pm_userspace.c | 7 +++++++
 1 file changed, 7 insertions(+)

diff --git a/net/mptcp/pm_userspace.c b/net/mptcp/pm_userspace.c
index a02d3cbf2a1b..9f7fc0b0d960 100644
--- a/net/mptcp/pm_userspace.c
+++ b/net/mptcp/pm_userspace.c
@@ -302,6 +302,10 @@ int mptcp_nl_cmd_sf_create(struct sk_buff *skb, struct genl_info *info)
 		goto create_err;
 	}
 
+	spin_lock_bh(&msk->pm.lock);
+	msk->pm.subflows++;
+	spin_unlock_bh(&msk->pm.lock);
+
 	lock_sock(sk);
 
 	err = __mptcp_subflow_connect(sk, &addr_l, &addr_r);
@@ -424,6 +428,9 @@ int mptcp_nl_cmd_sf_destroy(struct sk_buff *skb, struct genl_info *info)
 		mptcp_subflow_shutdown(sk, ssk, RCV_SHUTDOWN | SEND_SHUTDOWN);
 		mptcp_close_ssk(sk, ssk, subflow);
 		MPTCP_INC_STATS(sock_net(sk), MPTCP_MIB_RMSUBFLOW);
+		spin_lock_bh(&msk->pm.lock);
+		msk->pm.subflows--;
+		spin_unlock_bh(&msk->pm.lock);
 		err = 0;
 	} else {
 		err = -ESRCH;
-- 
2.35.3
Re: [PATCH mptcp-next v2] mptcp: update userspace pm subflows info
Posted by Matthieu Baerts 1 year, 2 months ago
Hi Geliang,

On 13/01/2023 09:45, Geliang Tang wrote:
> Increase pm subflows counter when userspace pm creates a new subflow,
> and decrease the counter when it closes a subflow.
> 
> Closes: https://github.com/multipath-tcp/mptcp_net-next/issues/329

(sorry for the delay)

Thank you, the patch looks good to me! But I don't think it is enough to
close the ticket 329.

I think there are other fields visible in `MPTCP_INFO` but only updated
by the in-kernel PM. e.g.

- add_addr_signaled
- add_addr_accepted
- local_addr_used

I guess these counters should be modified as well (in different
commits), no?

Also, I don't know if we should consider this as a "new feature" for
net-next: to me, it looks like we forgot to update the counters and it
is not really a new feature to update them but more a bug-fix. So maybe
we can target -net? I'm open to suggestions on that point :)

Not directly related but for the issue #330 -- 'subflows' counters not
incremented for the server side -- it could be considered as a bug fix
as well depending on how it is implemented: if it is backported, we
would have the same behaviour with all kernels. Or we add new field(s)
in MPTCP_INFO but this is maybe not needed.

Cheers,
Matt
-- 
Tessares | Belgium | Hybrid Access Solutions
www.tessares.net
Re: [PATCH mptcp-next v2] mptcp: update userspace pm subflows info
Posted by Geliang Tang 1 year, 2 months ago
On Wed, Jan 25, 2023 at 07:13:31PM +0100, Matthieu Baerts wrote:
> Hi Geliang,
> 
> On 13/01/2023 09:45, Geliang Tang wrote:
> > Increase pm subflows counter when userspace pm creates a new subflow,
> > and decrease the counter when it closes a subflow.
> > 
> > Closes: https://github.com/multipath-tcp/mptcp_net-next/issues/329
> 
> (sorry for the delay)
> 
> Thank you, the patch looks good to me! But I don't think it is enough to
> close the ticket 329.
> 
> I think there are other fields visible in `MPTCP_INFO` but only updated
> by the in-kernel PM. e.g.
> 
> - add_addr_signaled
> - add_addr_accepted
> - local_addr_used
> 
> I guess these counters should be modified as well (in different
> commits), no?

Yes, I'll modify these counters in v3.

> 
> Also, I don't know if we should consider this as a "new feature" for
> net-next: to me, it looks like we forgot to update the counters and it
> is not really a new feature to update them but more a bug-fix. So maybe
> we can target -net? I'm open to suggestions on that point :)

Sure. I'll update this in v3.

> 
> Not directly related but for the issue #330 -- 'subflows' counters not
> incremented for the server side -- it could be considered as a bug fix
> as well depending on how it is implemented: if it is backported, we
> would have the same behaviour with all kernels. Or we add new field(s)
> in MPTCP_INFO but this is maybe not needed.

OK, I'll address #330 in v3 too.

Thanks,
-Geliang

> 
> Cheers,
> Matt
> -- 
> Tessares | Belgium | Hybrid Access Solutions
> www.tessares.net
Re: [PATCH mptcp-next v2] mptcp: update userspace pm subflows info
Posted by Matthieu Baerts 1 year, 2 months ago
Hi Geliang,

On 28/01/2023 03:44, Geliang Tang wrote:
> On Wed, Jan 25, 2023 at 07:13:31PM +0100, Matthieu Baerts wrote:
>> Hi Geliang,
>>
>> On 13/01/2023 09:45, Geliang Tang wrote:
>>> Increase pm subflows counter when userspace pm creates a new subflow,
>>> and decrease the counter when it closes a subflow.
>>>
>>> Closes: https://github.com/multipath-tcp/mptcp_net-next/issues/329
>>
>> (sorry for the delay)
>>
>> Thank you, the patch looks good to me! But I don't think it is enough to
>> close the ticket 329.
>>
>> I think there are other fields visible in `MPTCP_INFO` but only updated
>> by the in-kernel PM. e.g.
>>
>> - add_addr_signaled
>> - add_addr_accepted
>> - local_addr_used
>>
>> I guess these counters should be modified as well (in different
>> commits), no?
> 
> Yes, I'll modify these counters in v3.
> 
>>
>> Also, I don't know if we should consider this as a "new feature" for
>> net-next: to me, it looks like we forgot to update the counters and it
>> is not really a new feature to update them but more a bug-fix. So maybe
>> we can target -net? I'm open to suggestions on that point :)
> 
> Sure. I'll update this in v3.
> 
>>
>> Not directly related but for the issue #330 -- 'subflows' counters not
>> incremented for the server side -- it could be considered as a bug fix
>> as well depending on how it is implemented: if it is backported, we
>> would have the same behaviour with all kernels. Or we add new field(s)
>> in MPTCP_INFO but this is maybe not needed.
> 
> OK, I'll address #330 in v3 too.

Great, thank you!

Happy New Year!

Cheers,
Matt
-- 
Tessares | Belgium | Hybrid Access Solutions
www.tessares.net