mptcp: support for MPC prio exchange
Subsystems (3)
Paolo Abeni
maintainerv2 cover mptcp: support for MPC prio exchange
This series introduces the ability to control the MPC subflow for the in kernel PM.
patch 1 && 2 refactor a bit the endpoint lookup and MP_PRIO generation code, to simplify the next patch.
patch 3 introduces the new feature quering the backup status of the endpoint used by the MPC subflow and eventually generating the needed MP_PRIO option.
Patch 4 is a related cleanup.
v1 -> v2:
- rebased, and omitted old patch 1 which should go through -net
Paolo Abeni (5):
mptcp: introduce and use mptcp_pm_send_ack()
mptcp: address lookup improvements
mptcp: allow the in kernel PM to set MPC subflow priority
mptcp: more accurate MPC endpoint tracking
selftests: mptcp: add MPC backup tests
net/mptcp/pm_netlink.c | 128 ++++++++++-------- net/mptcp/protocol.c | 2 +- net/mptcp/protocol.h | 2 +- .../testing/selftests/net/mptcp/mptcp_join.sh | 30 ++++ 4 files changed, 104 insertions(+), 58 deletions(-)
Signature · 1 line
2.35.3
Paolo Abeni
maintainerPATCH v2 1/5 mptcp: introduce and use mptcp_pm_send_ack()
The in-kernel PM has a bit of duplicate code related to ack generation. Create a new helper factoring out the PM-specific needs and use it in a couple of places.
As a bonus, mptcp_subflow_send_ack() is not used anymore outside its own compilation unit and can become static.
--- net/mptcp/pm_netlink.c | 56 +++++++++++++++++++++++++----------------- net/mptcp/protocol.c | 2 +- net/mptcp/protocol.h | 1 - 3 files changed, 35 insertions(+), 24 deletions(-)
net/mptcp/pm_netlink.c +34 -22
@@ -463,6 +463,37 @@ static unsigned int fill_remote_addresses_vec(struct mptcp_sock *msk, bool fullmreturn i; } +static void __mptcp_pm_send_ack (struct mptcp_sock *msk,struct mptcp_subflow_context *subflow, +bool prio,bool backup) +{ +struct sock *ssk =mptcp_subflow_tcp_sock (subflow); +bool slow; + +pr_debug ("send ack for %s" , + prio ?"mp_prio" : (mptcp_pm_should_add_signal (msk) ?"add_addr" :"rm_addr" )); + + slow =lock_sock_fast (ssk); +if (prio) { +if (subflow->backup != backup) + msk->last_snd =NULL ; + + subflow->send_mp_prio =1 ; + subflow->backup = backup; + subflow->request_bkup = backup; + } + +__mptcp_subflow_send_ack (ssk); +unlock_sock_fast (ssk, slow); +} + +static void mptcp_pm_send_ack (struct mptcp_sock *msk,struct mptcp_subflow_context *subflow, +bool prio,bool backup) +{ +spin_unlock_bh (&msk->pm.lock); +__mptcp_pm_send_ack (msk, subflow, prio, backup); +spin_lock_bh (&msk->pm.lock); +} +static struct mptcp_pm_addr_entry *__lookup_addr_by_id (struct pm_nl_pernet *pernet,unsigned int id) {
@@ -705,16 +736,8 @@ void mptcp_pm_nl_addr_send_ack(struct mptcp_sock *msk)return ; subflow =list_first_entry_or_null (&msk->conn_list,typeof (*subflow), node); -if (subflow) { -struct sock *ssk =mptcp_subflow_tcp_sock (subflow); - -spin_unlock_bh (&msk->pm.lock); -pr_debug ("send ack for %s" , -mptcp_pm_should_add_signal (msk) ?"add_addr" :"rm_addr" ); - -mptcp_subflow_send_ack (ssk); -spin_lock_bh (&msk->pm.lock); - } +if (subflow) +mptcp_pm_send_ack (msk, subflow, false, false); } int mptcp_pm_nl_mp_prio_send_ack(struct mptcp_sock *msk,
@@ -729,7 +752,6 @@ int mptcp_pm_nl_mp_prio_send_ack(struct mptcp_sock *msk,mptcp_for_each_subflow (msk, subflow) {struct sock *ssk =mptcp_subflow_tcp_sock (subflow);struct mptcp_addr_info local, remote; -bool slow;local_address ((struct sock_common *)ssk, &local);if (!mptcp_addresses_equal (&local, addr, addr->port))
@@ -741,17 +763,7 @@ int mptcp_pm_nl_mp_prio_send_ack(struct mptcp_sock *msk,continue ; } - slow =lock_sock_fast (ssk); -if (subflow->backup != bkup) - msk->last_snd =NULL ; - subflow->backup = bkup; - subflow->send_mp_prio =1 ; - subflow->request_bkup = bkup; - -pr_debug ("send ack for mp_prio" ); -__mptcp_subflow_send_ack (ssk); -unlock_sock_fast (ssk, slow); - +__mptcp_pm_send_ack (msk, subflow, true, bkup);return 0 ; }
net/mptcp/protocol.c +1 -1
@@ -508,7 +508,7 @@ void __mptcp_subflow_send_ack(struct sock *ssk)tcp_send_ack (ssk); } -void mptcp_subflow_send_ack (struct sock *ssk) +static void mptcp_subflow_send_ack (struct sock *ssk) {bool slow;
net/mptcp/protocol.h +0 -1
@@ -610,7 +610,6 @@ void mptcp_subflow_shutdown(struct sock *sk, struct sock *ssk, int how);void mptcp_close_ssk (struct sock *sk,struct sock *ssk,struct mptcp_subflow_context *subflow);void __mptcp_subflow_send_ack (struct sock *ssk); -void mptcp_subflow_send_ack (struct sock *ssk);void mptcp_subflow_reset (struct sock *ssk);void mptcp_subflow_queue_clean (struct sock *ssk);void mptcp_sock_graft (struct sock *sk,struct socket *parent);
Signature · 1 line
2.35.3
Paolo Abeni
maintainerPATCH v2 2/5 mptcp: address lookup improvements
In reply to Paolo AbeniWhen looking-up a socket address in the endpoint list, we must prefer port-based matches over address only match.
Ensure that port-based endpoints are listed first, using head insertion for them. Additionally be sure that only port-based endpoints carry a non zero port number.
--- net/mptcp/pm_netlink.c | 15 ++++++++++----- 1 file changed, 10 insertions(+), 5 deletions(-)
net/mptcp/pm_netlink.c +10 -5
@@ -413,7 +413,7 @@ static bool lookup_address_in_vec(const struct mptcp_addr_info *addrs, unsignedint i;for (i =0 ; i < nr; i++) { -if (mptcp_addresses_equal (&addrs[i], addr, addr->port)) +if (addrs[i].id == addr->id)return true; }
@@ -449,7 +449,8 @@ static unsigned int fill_remote_addresses_vec(struct mptcp_sock *msk, bool fullmmptcp_for_each_subflow (msk, subflow) { ssk =mptcp_subflow_tcp_sock (subflow);remote_address ((struct sock_common *)ssk, &addrs[i]); -if (deny_id0 &&mptcp_addresses_equal (&addrs[i], &remote, false)) + addrs[i].id = subflow->remote_id; +if (deny_id0 && !addrs[i].id)continue ; if (!lookup_address_in_vec(addrs, i, &addrs[i]) &&
@@ -919,10 +920,11 @@ static int mptcp_pm_nl_append_new_local_addr(struct pm_nl_pernet *pernet,/* do not insert duplicate address, differentiate on port only * singled addresses */ +if (!address_use_port (entry)) + entry->addr.port =0 ;list_for_each_entry (cur, &pernet->local_addr_list, list) {if (mptcp_addresses_equal (&cur->addr, &entry->addr, -address_use_port (entry) && -address_use_port (cur))) { + cur->addr.port || entry->addr.port)) {/* allow replacing the exiting endpoint only if such * endpoint is an implicit one and the user-space * did not provide an endpoint id
@@ -968,7 +970,10 @@ static int mptcp_pm_nl_append_new_local_addr(struct pm_nl_pernet *pernet, } pernet->addrs++; -list_add_tail_rcu (&entry->list, &pernet->local_addr_list); +if (!entry->addr.port) +list_add_tail_rcu (&entry->list, &pernet->local_addr_list); +else +list_add_rcu (&entry->list, &pernet->local_addr_list); ret = entry->addr.id; out:
Signature · 1 line
2.35.3
Paolo Abeni
maintainerPATCH v2 3/5 mptcp: allow the in kernel PM to set MPC subflow priority
In reply to Paolo AbeniAny local endpoints configured on the address matching the MPC subflow are currently ignored.
Specifically, setting a backup flag on them has no effect on the first subflow, as the MPC handshake can't carry such info.
This change refactors the MPC endpoint id accounting to additionally fetch the priority info from the relevant endpoint and eventually trigger the MP_PRIO handshake as needed.
As a result, the MPC subflow now switches to backup priority after that the MPTCP socket is fully established, according to the local endpoint configuration.
--- net/mptcp/pm_netlink.c | 37 +++++++++++++++---------------------- 1 file changed, 15 insertions(+), 22 deletions(-)
net/mptcp/pm_netlink.c +15 -22
@@ -514,30 +514,14 @@ __lookup_addr(struct pm_nl_pernet *pernet, const struct mptcp_addr_info *info,struct mptcp_pm_addr_entry *entry;list_for_each_entry (entry, &pernet->local_addr_list, list) { -if ((!lookup_by_id &&mptcp_addresses_equal (&entry->addr, info, true)) || +if ((!lookup_by_id && +mptcp_addresses_equal (&entry->addr, info, entry->addr.port)) || (lookup_by_id && entry->addr.id == info->id))return entry; }return NULL ; } -static int -lookup_id_by_addr (const struct pm_nl_pernet *pernet,const struct mptcp_addr_info *addr) -{ -const struct mptcp_pm_addr_entry *entry; -int ret =-1 ; - -rcu_read_lock (); -list_for_each_entry (entry, &pernet->local_addr_list, list) { -if (mptcp_addresses_equal (&entry->addr, addr, entry->addr.port)) { - ret = entry->addr.id; -break ; - } - } -rcu_read_unlock (); -return ret; -} -static void mptcp_pm_create_subflow_or_signal_addr (struct mptcp_sock *msk) {struct sock *sk = (struct sock *)msk;
@@ -555,13 +539,22 @@ static void mptcp_pm_create_subflow_or_signal_addr(struct mptcp_sock *msk)/* do lazy endpoint usage accounting for the MPC subflows */ if (unlikely (!(msk->pm.status &BIT (MPTCP_PM_MPC_ENDPOINT_ACCOUNTED ))) && msk->first) { +struct mptcp_subflow_context *subflow =mptcp_subflow_ctx (msk->first); +struct mptcp_pm_addr_entry *entry;struct mptcp_addr_info mpc_addr; -int mpc_id; +bool backup = false;local_address ((struct sock_common *)msk->first, &mpc_addr); - mpc_id =lookup_id_by_addr (pernet, &mpc_addr); -if (mpc_id >=0 ) -__clear_bit (mpc_id, msk->pm.id_avail_bitmap); +rcu_read_lock (); + entry =__lookup_addr (pernet, &mpc_addr, false); +if (entry) { +__clear_bit (entry->addr.id, msk->pm.id_avail_bitmap); + backup = !!(entry->flags &MPTCP_PM_ADDR_FLAG_BACKUP ); + } +rcu_read_unlock (); + +if (backup) +mptcp_pm_send_ack (msk, subflow, true, backup); msk->pm.status |=BIT (MPTCP_PM_MPC_ENDPOINT_ACCOUNTED ); }
Signature · 1 line
2.35.3
Paolo Abeni
maintainerPATCH v2 4/5 mptcp: more accurate MPC endpoint tracking
In reply to Paolo AbeniCurrently the id accounting for the ID 0 subflow is not correct:
at creation time we mark (correctly) as unavailable the endpoint id corresponding the MPC subflow source address, while at subflow removal time set as available the id 0.
With this change we track explicitly the endpoint id corresponding to the MPC subflow so that we can mark it as available at removal time. Additionally this allow deleting the initial subflow via the NL PM specifying the corresponding endpoint id.
--- net/mptcp/pm_netlink.c | 20 +++++++++++++------- net/mptcp/protocol.h | 1 + 2 files changed, 14 insertions(+), 7 deletions(-)
net/mptcp/pm_netlink.c +13 -7
@@ -549,6 +549,7 @@ static void mptcp_pm_create_subflow_or_signal_addr(struct mptcp_sock *msk) entry =__lookup_addr (pernet, &mpc_addr, false);if (entry) {__clear_bit (entry->addr.id, msk->pm.id_avail_bitmap); + msk->mpc_endpoint_id = entry->addr.id; backup = !!(entry->flags &MPTCP_PM_ADDR_FLAG_BACKUP ); }rcu_read_unlock ();
@@ -764,6 +765,11 @@ int mptcp_pm_nl_mp_prio_send_ack(struct mptcp_sock *msk,return -EINVAL ; } +static bool mptcp_local_id_match (const struct mptcp_sock *msk,u8 local_id,u8 id) +{ +return local_id == id || (!local_id && msk->mpc_endpoint_id == id); +} +static void mptcp_pm_nl_rm_addr_or_subflow (struct mptcp_sock *msk,const struct mptcp_rm_list *rm_list,enum linux_mptcp_mib_field rm_type)
@@ -787,6 +793,7 @@ static void mptcp_pm_nl_rm_addr_or_subflow(struct mptcp_sock *msk,return ;for (i =0 ; i < rm_list->nr; i++) { +u8 rm_id = rm_list->ids[i];bool removed = false;list_for_each_entry_safe (subflow, tmp, &msk->conn_list, node) {
@@ -794,15 +801,14 @@ static void mptcp_pm_nl_rm_addr_or_subflow(struct mptcp_sock *msk,int how =RCV_SHUTDOWN |SEND_SHUTDOWN ;u8 id = subflow->local_id; -if (rm_type ==MPTCP_MIB_RMADDR ) - id = subflow->remote_id; - -if (rm_list->ids[i] != id) +if (rm_type ==MPTCP_MIB_RMADDR && subflow->remote_id != rm_id) +continue ; +if (rm_type ==MPTCP_MIB_RMSUBFLOW && !mptcp_local_id_match (msk, id, rm_id))continue ; -pr_debug (" -> %s rm_list_ids[%d]=%u local_id=%u remote_id=%u" , +pr_debug (" -> %s rm_list_ids[%d]=%u local_id=%u remote_id=%u mpc_id=%u" , rm_type ==MPTCP_MIB_RMADDR ?"address" :"subflow" , - i, rm_list->ids[i], subflow->local_id, subflow->remote_id); + i, rm_id, subflow->local_id, subflow->remote_id, msk->mpc_endpoint_id);spin_unlock_bh (&msk->pm.lock);mptcp_subflow_shutdown (sk, ssk, how);
@@ -814,7 +820,7 @@ static void mptcp_pm_nl_rm_addr_or_subflow(struct mptcp_sock *msk,__MPTCP_INC_STATS (sock_net (sk), rm_type); }if (rm_type ==MPTCP_MIB_RMSUBFLOW ) -__set_bit (rm_list->ids[i], msk->pm.id_avail_bitmap); +__set_bit (rm_id ? rm_id : msk->mpc_endpoint_id, msk->pm.id_avail_bitmap);if (!removed)continue ;
net/mptcp/protocol.h +1 -0
@@ -282,6 +282,7 @@ struct mptcp_sock {bool use_64bit_ack;/* Set when we received a 64-bit DSN */ bool csum_enabled;bool allow_infinite_fallback; +u8 mpc_endpoint_id; u8 recvmsg_inq:1, cork:1, nodelay:1;
Signature · 1 line
2.35.3
Paolo Abeni
maintainerPATCH v2 5/5 selftests: mptcp: add MPC backup tests
In reply to Paolo AbeniAdd a couple of test-cases covering the newly introduced features - priority update for the MPC subflow.
--- .../testing/selftests/net/mptcp/mptcp_join.sh | 30 +++++++++++++++++++ 1 file changed, 30 insertions(+)
tools/testing/selftests/net/mptcp/mptcp_join.sh +30 -0
@@ -2428,6 +2428,36 @@ backup_tests() chk_add_nr 1 1 chk_prio_nr 1 1 fi + + if reset "mpc backup"; then + pm_nl_add_endpoint $ns2 10.0.1.2 flags subflow,backup + run_tests $ns1 $ns2 10.0.1.1 0 0 0 slow + chk_join_nr 0 0 0 + chk_prio_nr 0 1 + fi + + if reset "mpc backup both sides"; then + pm_nl_add_endpoint $ns1 10.0.1.1 flags subflow,backup + pm_nl_add_endpoint $ns2 10.0.1.2 flags subflow,backup + run_tests $ns1 $ns2 10.0.1.1 0 0 0 slow + chk_join_nr 0 0 0 + chk_prio_nr 1 1 + fi + + if reset "mpc switch to backup"; then + pm_nl_add_endpoint $ns2 10.0.1.2 flags subflow + run_tests $ns1 $ns2 10.0.1.1 0 0 0 slow backup + chk_join_nr 0 0 0 + chk_prio_nr 0 1 + fi + + if reset "mpc switch to backup both sides"; then + pm_nl_add_endpoint $ns1 10.0.1.1 flags subflow + pm_nl_add_endpoint $ns2 10.0.1.2 flags subflow + run_tests $ns1 $ns2 10.0.1.1 0 0 0 slow backup + chk_join_nr 0 0 0 + chk_prio_nr 1 1 + fi } add_addr_ports_tests()
Signature · 1 line
2.35.3
Quoted 6 lines · Paolo Abeni
This series introduces the ability to control the MPC subflow for the in kernel PM.
patch 1 && 2 refactor a bit the endpoint lookup and
MP_PRIOgeneration code, to simplify the next patch.
Quoted 9 lines
patch 3 introduces the new feature quering the backup status of the endpoint used by the MPC subflow and eventually generating the needed
MP_PRIOoption.Patch 4 is a related cleanup.
v1 -> v2:
- rebased, and omitted old patch 1 which should go through -net
Thanks Paolo, v2 looks good to me.
Quoted 7 lines
Paolo Abeni (5):
mptcp: introduce and usemptcp_pm_send_ack()
mptcp: address lookup improvements
mptcp: allow the in kernel PM to set MPC subflow priority
mptcp: more accurate MPC endpoint tracking
selftests: mptcp: add MPC backup tests
Quoted 11 lines
net/mptcp/pm_netlink.c | 128 ++++++++++-------- net/mptcp/protocol.c | 2 +- net/mptcp/protocol.h | 2 +- .../testing/selftests/net/mptcp/mptcp_join.sh | 30 ++++ 4 files changed, 104 insertions(+), 58 deletions(-)Signature · 1 line
2.35.3
--
Mat Martineau
Intel
Hi Paolo, Mat,
Quoted 11 lines · Paolo Abeni
This series introduces the ability to control the MPC subflow for the in kernel PM.
patch 1 && 2 refactor a bit the endpoint lookup and
MP_PRIOgeneration code, to simplify the next patch.patch 3 introduces the new feature quering the backup status of the endpoint used by the MPC subflow and eventually generating the needed
MP_PRIOoption.Patch 4 is a related cleanup.
Thank you for the patches and the reviews!
Now in our tree (feat. for net-next) with Mat's RvB tag and without 2 small issues reported by checkpatch.
New patches for t/upstream:
- 063213e73149: mptcp: introduce and use mptcp_pm_send_ack()
- 69bc499efccb: mptcp: address lookup improvements
- 0417a959d55e: mptcp: allow the in kernel PM to set MPC subflow priority
- 084ba8b4dc1a: mptcp: more accurate MPC endpoint tracking
- 267566754670: selftests: mptcp: add MPC backup tests
- Results: c47b82c7b9aa..7d868133e0d9 (export)
Builds and tests are now in progress:
https://cirrus-ci.com/github/multipath-tcp/mptcp_net-next/export/20220706T170628
https://github.com/multipath-tcp/mptcp_net-next/actions/workflows/build-validation.yml?query=branch:export
Cheers,
Matt
Signature · 2 lines
Tessares | Belgium | Hybrid Access Solutions www.tessares.net