* [PATCH mptcp-next 0/7] fullmesh flag setting support
@ 2022-01-10 3:29 Geliang Tang
2022-01-10 3:29 ` [PATCH mptcp-next 1/7] mptcp: fix removing ids bitmap Geliang Tang
` (6 more replies)
0 siblings, 7 replies; 13+ messages in thread
From: Geliang Tang @ 2022-01-10 3:29 UTC (permalink / raw)
To: mptcp; +Cc: Geliang Tang
Added the fullmesh flag setting support:
# pm_nl_ctl set 10.0.1.1 flags backup
# pm_nl_ctl clear 10.0.1.1 flags backup
# pm_nl_ctl set 10.0.1.1 flags fullmesh
# pm_nl_ctl clear 10.0.1.1 flags fullmesh
Geliang Tang (7):
mptcp: fix removing ids bitmap
mptcp: add clear_flags in pm_netlink
selftests: mptcp: add clear_flags in pm_nl_ctl
selftests: mptcp: add clear_flags in mptcp_join.sh
mptcp: set fullmesh flag in pm_netlink
selftests: mptcp: set fullmesh flag in pm_nl_ctl
selftests: mptcp: add fullmesh setting tests
include/uapi/linux/mptcp.h | 1 +
net/mptcp/pm_netlink.c | 45 ++++++++++++++-----
.../testing/selftests/net/mptcp/mptcp_join.sh | 42 ++++++++++++++---
tools/testing/selftests/net/mptcp/pm_nl_ctl.c | 16 ++++---
4 files changed, 82 insertions(+), 22 deletions(-)
--
2.31.1
^ permalink raw reply [flat|nested] 13+ messages in thread
* [PATCH mptcp-next 1/7] mptcp: fix removing ids bitmap
2022-01-10 3:29 [PATCH mptcp-next 0/7] fullmesh flag setting support Geliang Tang
@ 2022-01-10 3:29 ` Geliang Tang
2022-01-10 11:40 ` Paolo Abeni
2022-01-10 3:30 ` [PATCH mptcp-next 2/7] mptcp: add clear_flags in pm_netlink Geliang Tang
` (5 subsequent siblings)
6 siblings, 1 reply; 13+ messages in thread
From: Geliang Tang @ 2022-01-10 3:29 UTC (permalink / raw)
To: mptcp; +Cc: Geliang Tang
In mptcp_pm_nl_rm_addr_or_subflow(), the bit of rm_list->ids[i] in the
id_avail_bitmap should be set, not rm_list->ids[1]. This patch fixed it.
Fixes: 86e39e04482b ("mptcp: keep track of local endpoint still available for each msk")
Signed-off-by: Geliang Tang <geliang.tang@suse.com>
---
net/mptcp/pm_netlink.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/net/mptcp/pm_netlink.c b/net/mptcp/pm_netlink.c
index f12effa71942..47ad00d01cf2 100644
--- a/net/mptcp/pm_netlink.c
+++ b/net/mptcp/pm_netlink.c
@@ -777,7 +777,7 @@ static void mptcp_pm_nl_rm_addr_or_subflow(struct mptcp_sock *msk,
removed = true;
__MPTCP_INC_STATS(sock_net(sk), rm_type);
}
- __set_bit(rm_list->ids[1], msk->pm.id_avail_bitmap);
+ __set_bit(rm_list->ids[i], msk->pm.id_avail_bitmap);
if (!removed)
continue;
--
2.31.1
^ permalink raw reply related [flat|nested] 13+ messages in thread
* [PATCH mptcp-next 2/7] mptcp: add clear_flags in pm_netlink
2022-01-10 3:29 [PATCH mptcp-next 0/7] fullmesh flag setting support Geliang Tang
2022-01-10 3:29 ` [PATCH mptcp-next 1/7] mptcp: fix removing ids bitmap Geliang Tang
@ 2022-01-10 3:30 ` Geliang Tang
2022-01-10 11:38 ` Paolo Abeni
2022-01-10 3:30 ` [PATCH mptcp-next 3/7] selftests: mptcp: add clear_flags in pm_nl_ctl Geliang Tang
` (4 subsequent siblings)
6 siblings, 1 reply; 13+ messages in thread
From: Geliang Tang @ 2022-01-10 3:30 UTC (permalink / raw)
To: mptcp; +Cc: Geliang Tang
Splite set_flags() into two parts, set_flags() and clear_flags(), make it
easy to add new flags to set or clear.
This patch added a new PM command MPTCP_PM_CMD_CLEAR_FLAGS, and a new
function mptcp_nl_cmd_clear_flags().
Signed-off-by: Geliang Tang <geliang.tang@suse.com>
---
include/uapi/linux/mptcp.h | 1 +
net/mptcp/pm_netlink.c | 36 ++++++++++++++++++++++++++----------
2 files changed, 27 insertions(+), 10 deletions(-)
diff --git a/include/uapi/linux/mptcp.h b/include/uapi/linux/mptcp.h
index f106a3941cdf..1ca9b13c2ed0 100644
--- a/include/uapi/linux/mptcp.h
+++ b/include/uapi/linux/mptcp.h
@@ -92,6 +92,7 @@ enum {
MPTCP_PM_CMD_SET_LIMITS,
MPTCP_PM_CMD_GET_LIMITS,
MPTCP_PM_CMD_SET_FLAGS,
+ MPTCP_PM_CMD_CLEAR_FLAGS,
__MPTCP_PM_CMD_AFTER_LAST
};
diff --git a/net/mptcp/pm_netlink.c b/net/mptcp/pm_netlink.c
index 47ad00d01cf2..da1bef34e8e6 100644
--- a/net/mptcp/pm_netlink.c
+++ b/net/mptcp/pm_netlink.c
@@ -1736,21 +1736,21 @@ static int mptcp_nl_addr_backup(struct net *net,
return ret;
}
-static int mptcp_nl_cmd_set_flags(struct sk_buff *skb, struct genl_info *info)
+static int __mptcp_nl_cmd_set_flags(struct sk_buff *skb,
+ struct genl_info *info,
+ int clear_flags)
{
struct mptcp_pm_addr_entry addr = { .addr = { .family = AF_UNSPEC }, }, *entry;
struct nlattr *attr = info->attrs[MPTCP_PM_ATTR_ADDR];
struct pm_nl_pernet *pernet = genl_info_pm_nl(info);
struct net *net = sock_net(skb->sk);
- u8 bkup = 0, lookup_by_id = 0;
+ u8 lookup_by_id = 0;
int ret;
ret = mptcp_pm_parse_addr(attr, info, false, &addr);
if (ret < 0)
return ret;
- if (addr.flags & MPTCP_PM_ADDR_FLAG_BACKUP)
- bkup = 1;
if (addr.addr.family == AF_UNSPEC) {
lookup_by_id = 1;
if (!addr.addr.id)
@@ -1760,18 +1760,29 @@ static int mptcp_nl_cmd_set_flags(struct sk_buff *skb, struct genl_info *info)
list_for_each_entry(entry, &pernet->local_addr_list, list) {
if ((!lookup_by_id && addresses_equal(&entry->addr, &addr.addr, true)) ||
(lookup_by_id && entry->addr.id == addr.addr.id)) {
- mptcp_nl_addr_backup(net, &entry->addr, bkup);
-
- if (bkup)
- entry->flags |= MPTCP_PM_ADDR_FLAG_BACKUP;
- else
- entry->flags &= ~MPTCP_PM_ADDR_FLAG_BACKUP;
+ if (addr.flags & MPTCP_PM_ADDR_FLAG_BACKUP) {
+ mptcp_nl_addr_backup(net, &entry->addr, !clear_flags);
+ if (clear_flags)
+ entry->flags &= ~MPTCP_PM_ADDR_FLAG_BACKUP;
+ else
+ entry->flags |= MPTCP_PM_ADDR_FLAG_BACKUP;
+ }
}
}
return 0;
}
+static int mptcp_nl_cmd_set_flags(struct sk_buff *skb, struct genl_info *info)
+{
+ return __mptcp_nl_cmd_set_flags(skb, info, 0);
+}
+
+static int mptcp_nl_cmd_clear_flags(struct sk_buff *skb, struct genl_info *info)
+{
+ return __mptcp_nl_cmd_set_flags(skb, info, 1);
+}
+
static void mptcp_nl_mcast_send(struct net *net, struct sk_buff *nlskb, gfp_t gfp)
{
genlmsg_multicast_netns(&mptcp_genl_family, net,
@@ -2074,6 +2085,11 @@ static const struct genl_small_ops mptcp_pm_ops[] = {
.doit = mptcp_nl_cmd_set_flags,
.flags = GENL_ADMIN_PERM,
},
+ {
+ .cmd = MPTCP_PM_CMD_CLEAR_FLAGS,
+ .doit = mptcp_nl_cmd_clear_flags,
+ .flags = GENL_ADMIN_PERM,
+ },
};
static struct genl_family mptcp_genl_family __ro_after_init = {
--
2.31.1
^ permalink raw reply related [flat|nested] 13+ messages in thread
* [PATCH mptcp-next 3/7] selftests: mptcp: add clear_flags in pm_nl_ctl
2022-01-10 3:29 [PATCH mptcp-next 0/7] fullmesh flag setting support Geliang Tang
2022-01-10 3:29 ` [PATCH mptcp-next 1/7] mptcp: fix removing ids bitmap Geliang Tang
2022-01-10 3:30 ` [PATCH mptcp-next 2/7] mptcp: add clear_flags in pm_netlink Geliang Tang
@ 2022-01-10 3:30 ` Geliang Tang
2022-01-10 3:30 ` [PATCH mptcp-next 4/7] selftests: mptcp: add clear_flags in mptcp_join.sh Geliang Tang
` (3 subsequent siblings)
6 siblings, 0 replies; 13+ messages in thread
From: Geliang Tang @ 2022-01-10 3:30 UTC (permalink / raw)
To: mptcp; +Cc: Geliang Tang
This patch added 'clear' command for pm_nl_ctl, to replace the command
'pm_nl_ctl set ip flags nobackup' with 'pm_nl_ctl clear ip flags backup'.
Signed-off-by: Geliang Tang <geliang.tang@suse.com>
---
tools/testing/selftests/net/mptcp/pm_nl_ctl.c | 12 ++++++++----
1 file changed, 8 insertions(+), 4 deletions(-)
diff --git a/tools/testing/selftests/net/mptcp/pm_nl_ctl.c b/tools/testing/selftests/net/mptcp/pm_nl_ctl.c
index 354784512748..85ca37e4b34c 100644
--- a/tools/testing/selftests/net/mptcp/pm_nl_ctl.c
+++ b/tools/testing/selftests/net/mptcp/pm_nl_ctl.c
@@ -28,7 +28,8 @@ static void syntax(char *argv[])
fprintf(stderr, "\tadd [flags signal|subflow|backup|fullmesh] [id <nr>] [dev <name>] <ip>\n");
fprintf(stderr, "\tdel <id> [<ip>]\n");
fprintf(stderr, "\tget <id>\n");
- fprintf(stderr, "\tset <ip> [flags backup|nobackup]\n");
+ fprintf(stderr, "\tset <ip> [flags backup]\n");
+ fprintf(stderr, "\tclear <ip> [flags backup]\n");
fprintf(stderr, "\tflush\n");
fprintf(stderr, "\tdump\n");
fprintf(stderr, "\tlimits [<rcv addr max> <subflow max>]\n");
@@ -647,7 +648,7 @@ int get_set_limits(int fd, int pm_family, int argc, char *argv[])
return 0;
}
-int set_flags(int fd, int pm_family, int argc, char *argv[])
+int set_flags(int fd, int pm_family, int clear_flags, int argc, char *argv[])
{
char data[NLMSG_ALIGN(sizeof(struct nlmsghdr)) +
NLMSG_ALIGN(sizeof(struct genlmsghdr)) +
@@ -662,7 +663,8 @@ int set_flags(int fd, int pm_family, int argc, char *argv[])
memset(data, 0, sizeof(data));
nh = (void *)data;
- off = init_genl_req(data, pm_family, MPTCP_PM_CMD_SET_FLAGS,
+ off = init_genl_req(data, pm_family,
+ clear_flags ? MPTCP_PM_CMD_CLEAR_FLAGS : MPTCP_PM_CMD_SET_FLAGS,
MPTCP_PM_VER);
if (argc < 3)
@@ -755,7 +757,9 @@ int main(int argc, char *argv[])
else if (!strcmp(argv[1], "limits"))
return get_set_limits(fd, pm_family, argc, argv);
else if (!strcmp(argv[1], "set"))
- return set_flags(fd, pm_family, argc, argv);
+ return set_flags(fd, pm_family, 0, argc, argv);
+ else if (!strcmp(argv[1], "clear"))
+ return set_flags(fd, pm_family, 1, argc, argv);
fprintf(stderr, "unknown sub-command: %s", argv[1]);
syntax(argv);
--
2.31.1
^ permalink raw reply related [flat|nested] 13+ messages in thread
* [PATCH mptcp-next 4/7] selftests: mptcp: add clear_flags in mptcp_join.sh
2022-01-10 3:29 [PATCH mptcp-next 0/7] fullmesh flag setting support Geliang Tang
` (2 preceding siblings ...)
2022-01-10 3:30 ` [PATCH mptcp-next 3/7] selftests: mptcp: add clear_flags in pm_nl_ctl Geliang Tang
@ 2022-01-10 3:30 ` Geliang Tang
2022-01-10 3:30 ` [PATCH mptcp-next 5/7] mptcp: set fullmesh flag in pm_netlink Geliang Tang
` (2 subsequent siblings)
6 siblings, 0 replies; 13+ messages in thread
From: Geliang Tang @ 2022-01-10 3:30 UTC (permalink / raw)
To: mptcp; +Cc: Geliang Tang
This patch changed the command 'pm_nl_ctl set $addr flags nobackup' to
'pm_nl_ctl clear $addr flags backup', and renamed bkup to setflags.
Signed-off-by: Geliang Tang <geliang.tang@suse.com>
---
.../testing/selftests/net/mptcp/mptcp_join.sh | 18 ++++++++++++------
1 file changed, 12 insertions(+), 6 deletions(-)
diff --git a/tools/testing/selftests/net/mptcp/mptcp_join.sh b/tools/testing/selftests/net/mptcp/mptcp_join.sh
index e48ce23d2386..7dbd36c558cd 100755
--- a/tools/testing/selftests/net/mptcp/mptcp_join.sh
+++ b/tools/testing/selftests/net/mptcp/mptcp_join.sh
@@ -289,7 +289,7 @@ do_transfer()
addr_nr_ns1="$7"
addr_nr_ns2="$8"
speed="$9"
- bkup="${10}"
+ setflags="${10}"
port=$((10000+$TEST_COUNT))
TEST_COUNT=$((TEST_COUNT+1))
@@ -461,14 +461,20 @@ do_transfer()
fi
fi
- if [ ! -z $bkup ]; then
+ if [ ! -z $setflags ]; then
sleep 1
for netns in "$ns1" "$ns2"; do
dump=(`ip netns exec $netns ./pm_nl_ctl dump`)
if [ ${#dump[@]} -gt 0 ]; then
addr=${dump[${#dump[@]} - 1]}
- backup="ip netns exec $netns ./pm_nl_ctl set $addr flags $bkup"
- $backup
+ if [ $setflags = "backup" ]; then
+ cmd=set
+ flags=backup
+ elif [ $setflags = "nobackup" ]; then
+ cmd=clear
+ flags=backup
+ fi
+ ip netns exec $netns ./pm_nl_ctl $cmd $addr flags $flags
fi
done
fi
@@ -545,7 +551,7 @@ run_tests()
addr_nr_ns1="${5:-0}"
addr_nr_ns2="${6:-0}"
speed="${7:-fast}"
- bkup="${8:-""}"
+ setflags="${8:-""}"
lret=0
oldin=""
@@ -574,7 +580,7 @@ run_tests()
fi
do_transfer ${listener_ns} ${connector_ns} MPTCP MPTCP ${connect_addr} \
- ${test_linkfail} ${addr_nr_ns1} ${addr_nr_ns2} ${speed} ${bkup}
+ ${test_linkfail} ${addr_nr_ns1} ${addr_nr_ns2} ${speed} ${setflags}
lret=$?
}
--
2.31.1
^ permalink raw reply related [flat|nested] 13+ messages in thread
* [PATCH mptcp-next 5/7] mptcp: set fullmesh flag in pm_netlink
2022-01-10 3:29 [PATCH mptcp-next 0/7] fullmesh flag setting support Geliang Tang
` (3 preceding siblings ...)
2022-01-10 3:30 ` [PATCH mptcp-next 4/7] selftests: mptcp: add clear_flags in mptcp_join.sh Geliang Tang
@ 2022-01-10 3:30 ` Geliang Tang
2022-01-10 3:30 ` [PATCH mptcp-next 6/7] selftests: mptcp: set fullmesh flag in pm_nl_ctl Geliang Tang
2022-01-10 3:30 ` [PATCH mptcp-next 7/7] selftests: mptcp: add fullmesh setting tests Geliang Tang
6 siblings, 0 replies; 13+ messages in thread
From: Geliang Tang @ 2022-01-10 3:30 UTC (permalink / raw)
To: mptcp; +Cc: Geliang Tang
This patch added the fullmesh flag setting and clearing support in
pm_netlink.
Signed-off-by: Geliang Tang <geliang.tang@suse.com>
---
net/mptcp/pm_netlink.c | 7 +++++++
1 file changed, 7 insertions(+)
diff --git a/net/mptcp/pm_netlink.c b/net/mptcp/pm_netlink.c
index da1bef34e8e6..5fcf0f979579 100644
--- a/net/mptcp/pm_netlink.c
+++ b/net/mptcp/pm_netlink.c
@@ -1766,6 +1766,13 @@ static int __mptcp_nl_cmd_set_flags(struct sk_buff *skb,
entry->flags &= ~MPTCP_PM_ADDR_FLAG_BACKUP;
else
entry->flags |= MPTCP_PM_ADDR_FLAG_BACKUP;
+ } else if (addr.flags & MPTCP_PM_ADDR_FLAG_FULLMESH) {
+ mptcp_nl_remove_subflow_and_signal_addr(net, &entry->addr);
+ if (clear_flags)
+ entry->flags &= ~MPTCP_PM_ADDR_FLAG_FULLMESH;
+ else
+ entry->flags |= MPTCP_PM_ADDR_FLAG_FULLMESH;
+ mptcp_nl_add_subflow_or_signal_addr(net);
}
}
}
--
2.31.1
^ permalink raw reply related [flat|nested] 13+ messages in thread
* [PATCH mptcp-next 6/7] selftests: mptcp: set fullmesh flag in pm_nl_ctl
2022-01-10 3:29 [PATCH mptcp-next 0/7] fullmesh flag setting support Geliang Tang
` (4 preceding siblings ...)
2022-01-10 3:30 ` [PATCH mptcp-next 5/7] mptcp: set fullmesh flag in pm_netlink Geliang Tang
@ 2022-01-10 3:30 ` Geliang Tang
2022-01-10 3:30 ` [PATCH mptcp-next 7/7] selftests: mptcp: add fullmesh setting tests Geliang Tang
6 siblings, 0 replies; 13+ messages in thread
From: Geliang Tang @ 2022-01-10 3:30 UTC (permalink / raw)
To: mptcp; +Cc: Geliang Tang
This patch added the fullmesh flag setting and clearing support in
pm_nl_ctl:
# pm_nl_ctl set ip flags fullmesh
# pm_nl_ctl clear ip flags fullmesh
Signed-off-by: Geliang Tang <geliang.tang@suse.com>
---
tools/testing/selftests/net/mptcp/pm_nl_ctl.c | 8 +++++---
1 file changed, 5 insertions(+), 3 deletions(-)
diff --git a/tools/testing/selftests/net/mptcp/pm_nl_ctl.c b/tools/testing/selftests/net/mptcp/pm_nl_ctl.c
index 85ca37e4b34c..396d53d8fcec 100644
--- a/tools/testing/selftests/net/mptcp/pm_nl_ctl.c
+++ b/tools/testing/selftests/net/mptcp/pm_nl_ctl.c
@@ -28,8 +28,8 @@ static void syntax(char *argv[])
fprintf(stderr, "\tadd [flags signal|subflow|backup|fullmesh] [id <nr>] [dev <name>] <ip>\n");
fprintf(stderr, "\tdel <id> [<ip>]\n");
fprintf(stderr, "\tget <id>\n");
- fprintf(stderr, "\tset <ip> [flags backup]\n");
- fprintf(stderr, "\tclear <ip> [flags backup]\n");
+ fprintf(stderr, "\tset <ip> [flags backup|fullmesh]\n");
+ fprintf(stderr, "\tclear <ip> [flags backup|fullmesh]\n");
fprintf(stderr, "\tflush\n");
fprintf(stderr, "\tdump\n");
fprintf(stderr, "\tlimits [<rcv addr max> <subflow max>]\n");
@@ -711,7 +711,9 @@ int set_flags(int fd, int pm_family, int clear_flags, int argc, char *argv[])
str = NULL) {
if (!strcmp(tok, "backup"))
flags |= MPTCP_PM_ADDR_FLAG_BACKUP;
- else if (strcmp(tok, "nobackup"))
+ else if (!strcmp(tok, "fullmesh"))
+ flags |= MPTCP_PM_ADDR_FLAG_FULLMESH;
+ else
error(1, errno,
"unknown flag %s", argv[arg]);
}
--
2.31.1
^ permalink raw reply related [flat|nested] 13+ messages in thread
* [PATCH mptcp-next 7/7] selftests: mptcp: add fullmesh setting tests
2022-01-10 3:29 [PATCH mptcp-next 0/7] fullmesh flag setting support Geliang Tang
` (5 preceding siblings ...)
2022-01-10 3:30 ` [PATCH mptcp-next 6/7] selftests: mptcp: set fullmesh flag in pm_nl_ctl Geliang Tang
@ 2022-01-10 3:30 ` Geliang Tang
6 siblings, 0 replies; 13+ messages in thread
From: Geliang Tang @ 2022-01-10 3:30 UTC (permalink / raw)
To: mptcp; +Cc: Geliang Tang
This patch added the fullmesh setting and clearing selftests in
mptcp_join.sh.
Signed-off-by: Geliang Tang <geliang.tang@suse.com>
---
.../testing/selftests/net/mptcp/mptcp_join.sh | 24 +++++++++++++++++++
1 file changed, 24 insertions(+)
diff --git a/tools/testing/selftests/net/mptcp/mptcp_join.sh b/tools/testing/selftests/net/mptcp/mptcp_join.sh
index 7dbd36c558cd..8c4ee9092e83 100755
--- a/tools/testing/selftests/net/mptcp/mptcp_join.sh
+++ b/tools/testing/selftests/net/mptcp/mptcp_join.sh
@@ -473,6 +473,12 @@ do_transfer()
elif [ $setflags = "nobackup" ]; then
cmd=clear
flags=backup
+ elif [ $setflags = "fullmesh" ]; then
+ cmd=set
+ flags=fullmesh
+ elif [ $setflags = "nofullmesh" ]; then
+ cmd=clear
+ flags=fullmesh
fi
ip netns exec $netns ./pm_nl_ctl $cmd $addr flags $flags
fi
@@ -1927,6 +1933,24 @@ fullmesh_tests()
run_tests $ns1 $ns2 10.0.1.1 0 0 fullmesh_2 slow
chk_join_nr "fullmesh test 1x2, limited" 4 4 4
chk_add_nr 1 1
+
+ # fullmesh flag setting test
+ reset
+ ip netns exec $ns1 ./pm_nl_ctl limits 4 4
+ ip netns exec $ns1 ./pm_nl_ctl add 10.0.2.1 flags subflow
+ ip netns exec $ns2 ./pm_nl_ctl limits 4 4
+ run_tests $ns1 $ns2 10.0.1.1 0 0 1 slow fullmesh
+ chk_join_nr "fullmesh flag setting test" 2 2 2
+ chk_rm_nr 1 1
+
+ # fullmesh flag clearing test
+ reset
+ ip netns exec $ns1 ./pm_nl_ctl limits 4 4
+ ip netns exec $ns1 ./pm_nl_ctl add 10.0.2.1 flags subflow,fullmesh
+ ip netns exec $ns2 ./pm_nl_ctl limits 4 4
+ run_tests $ns1 $ns2 10.0.1.1 0 0 fullmesh_1 slow nofullmesh
+ chk_join_nr "fullmesh flag clearing test" 2 2 2
+ chk_rm_nr 1 1
}
userspace_tests()
--
2.31.1
^ permalink raw reply related [flat|nested] 13+ messages in thread
* Re: [PATCH mptcp-next 2/7] mptcp: add clear_flags in pm_netlink
2022-01-10 3:30 ` [PATCH mptcp-next 2/7] mptcp: add clear_flags in pm_netlink Geliang Tang
@ 2022-01-10 11:38 ` Paolo Abeni
2022-01-10 12:35 ` Geliang Tang
0 siblings, 1 reply; 13+ messages in thread
From: Paolo Abeni @ 2022-01-10 11:38 UTC (permalink / raw)
To: Geliang Tang, mptcp
Hello,
On Mon, 2022-01-10 at 11:30 +0800, Geliang Tang wrote:
> Splite set_flags() into two parts, set_flags() and clear_flags(), make it
> easy to add new flags to set or clear.
>
> This patch added a new PM command MPTCP_PM_CMD_CLEAR_FLAGS, and a new
> function mptcp_nl_cmd_clear_flags().
>
> Signed-off-by: Geliang Tang <geliang.tang@suse.com>
> ---
> include/uapi/linux/mptcp.h | 1 +
> net/mptcp/pm_netlink.c | 36 ++++++++++++++++++++++++++----------
> 2 files changed, 27 insertions(+), 10 deletions(-)
>
> diff --git a/include/uapi/linux/mptcp.h b/include/uapi/linux/mptcp.h
> index f106a3941cdf..1ca9b13c2ed0 100644
> --- a/include/uapi/linux/mptcp.h
> +++ b/include/uapi/linux/mptcp.h
> @@ -92,6 +92,7 @@ enum {
> MPTCP_PM_CMD_SET_LIMITS,
> MPTCP_PM_CMD_GET_LIMITS,
> MPTCP_PM_CMD_SET_FLAGS,
> + MPTCP_PM_CMD_CLEAR_FLAGS,
>
> __MPTCP_PM_CMD_AFTER_LAST
> };
I'm unsure if this has already been discussed in some mtg I missed. If
so, I'm sorry for the late feedback :(
Looking at the existing netlink interfaces, I *think* the preferred way
of handling this would be adding instead a couple of MPTCP address
flags:
MPTCP_PM_ADDR_FLAG_NOBACKUP,
MPTCP_PM_ADDR_FLAG_NOFULLMESH
As e.g. IFA_F_NOPREFIXROUTE, IFA_F_NODAD,...
The above should require a similar amount of code and will make it
clear that we can't/wan't to flip the SUBFLOW or SIGNAL flags.
WDYT?
Thanks!
Paolo
^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: [PATCH mptcp-next 1/7] mptcp: fix removing ids bitmap
2022-01-10 3:29 ` [PATCH mptcp-next 1/7] mptcp: fix removing ids bitmap Geliang Tang
@ 2022-01-10 11:40 ` Paolo Abeni
2022-01-10 12:33 ` Geliang Tang
0 siblings, 1 reply; 13+ messages in thread
From: Paolo Abeni @ 2022-01-10 11:40 UTC (permalink / raw)
To: Geliang Tang, mptcp
On Mon, 2022-01-10 at 11:29 +0800, Geliang Tang wrote:
> In mptcp_pm_nl_rm_addr_or_subflow(), the bit of rm_list->ids[i] in the
> id_avail_bitmap should be set, not rm_list->ids[1]. This patch fixed it.
>
> Fixes: 86e39e04482b ("mptcp: keep track of local endpoint still available for each msk")
> Signed-off-by: Geliang Tang <geliang.tang@suse.com>
> ---
> net/mptcp/pm_netlink.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/net/mptcp/pm_netlink.c b/net/mptcp/pm_netlink.c
> index f12effa71942..47ad00d01cf2 100644
> --- a/net/mptcp/pm_netlink.c
> +++ b/net/mptcp/pm_netlink.c
> @@ -777,7 +777,7 @@ static void mptcp_pm_nl_rm_addr_or_subflow(struct mptcp_sock *msk,
> removed = true;
> __MPTCP_INC_STATS(sock_net(sk), rm_type);
> }
> - __set_bit(rm_list->ids[1], msk->pm.id_avail_bitmap);
> + __set_bit(rm_list->ids[i], msk->pm.id_avail_bitmap);
> if (!removed)
> continue;
>
LGTM!
Acked-by: Paolo Abeni <pabeni@redhat.com>
I think this one could be spooled independently from the other patches,
so it could land on -net early.
WDYT?
Thanks!
Paolo
^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: [PATCH mptcp-next 1/7] mptcp: fix removing ids bitmap
2022-01-10 11:40 ` Paolo Abeni
@ 2022-01-10 12:33 ` Geliang Tang
0 siblings, 0 replies; 13+ messages in thread
From: Geliang Tang @ 2022-01-10 12:33 UTC (permalink / raw)
To: Paolo Abeni, Matthieu Baerts; +Cc: mptcp
On Mon, Jan 10, 2022 at 12:40:14PM +0100, Paolo Abeni wrote:
> On Mon, 2022-01-10 at 11:29 +0800, Geliang Tang wrote:
> > In mptcp_pm_nl_rm_addr_or_subflow(), the bit of rm_list->ids[i] in the
> > id_avail_bitmap should be set, not rm_list->ids[1]. This patch fixed it.
> >
> > Fixes: 86e39e04482b ("mptcp: keep track of local endpoint still available for each msk")
> > Signed-off-by: Geliang Tang <geliang.tang@suse.com>
> > ---
> > net/mptcp/pm_netlink.c | 2 +-
> > 1 file changed, 1 insertion(+), 1 deletion(-)
> >
> > diff --git a/net/mptcp/pm_netlink.c b/net/mptcp/pm_netlink.c
> > index f12effa71942..47ad00d01cf2 100644
> > --- a/net/mptcp/pm_netlink.c
> > +++ b/net/mptcp/pm_netlink.c
> > @@ -777,7 +777,7 @@ static void mptcp_pm_nl_rm_addr_or_subflow(struct mptcp_sock *msk,
> > removed = true;
> > __MPTCP_INC_STATS(sock_net(sk), rm_type);
> > }
> > - __set_bit(rm_list->ids[1], msk->pm.id_avail_bitmap);
> > + __set_bit(rm_list->ids[i], msk->pm.id_avail_bitmap);
> > if (!removed)
> > continue;
> >
>
> LGTM!
>
> Acked-by: Paolo Abeni <pabeni@redhat.com>
>
> I think this one could be spooled independently from the other patches,
> so it could land on -net early.
>
> WDYT?
Great!
Matt, could you please help me to update the subject of this patch to
"mptcp: fix removing ids bitmap setting" when merging it?
Thanks,
-Geliang
>
> Thanks!
>
> Paolo
>
^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: [PATCH mptcp-next 2/7] mptcp: add clear_flags in pm_netlink
2022-01-10 11:38 ` Paolo Abeni
@ 2022-01-10 12:35 ` Geliang Tang
2022-01-10 15:59 ` Paolo Abeni
0 siblings, 1 reply; 13+ messages in thread
From: Geliang Tang @ 2022-01-10 12:35 UTC (permalink / raw)
To: Paolo Abeni; +Cc: mptcp
On Mon, Jan 10, 2022 at 12:38:14PM +0100, Paolo Abeni wrote:
> Hello,
>
> On Mon, 2022-01-10 at 11:30 +0800, Geliang Tang wrote:
> > Splite set_flags() into two parts, set_flags() and clear_flags(), make it
> > easy to add new flags to set or clear.
> >
> > This patch added a new PM command MPTCP_PM_CMD_CLEAR_FLAGS, and a new
> > function mptcp_nl_cmd_clear_flags().
> >
> > Signed-off-by: Geliang Tang <geliang.tang@suse.com>
> > ---
> > include/uapi/linux/mptcp.h | 1 +
> > net/mptcp/pm_netlink.c | 36 ++++++++++++++++++++++++++----------
> > 2 files changed, 27 insertions(+), 10 deletions(-)
> >
> > diff --git a/include/uapi/linux/mptcp.h b/include/uapi/linux/mptcp.h
> > index f106a3941cdf..1ca9b13c2ed0 100644
> > --- a/include/uapi/linux/mptcp.h
> > +++ b/include/uapi/linux/mptcp.h
> > @@ -92,6 +92,7 @@ enum {
> > MPTCP_PM_CMD_SET_LIMITS,
> > MPTCP_PM_CMD_GET_LIMITS,
> > MPTCP_PM_CMD_SET_FLAGS,
> > + MPTCP_PM_CMD_CLEAR_FLAGS,
> >
> > __MPTCP_PM_CMD_AFTER_LAST
> > };
>
> I'm unsure if this has already been discussed in some mtg I missed. If
> so, I'm sorry for the late feedback :(
>
> Looking at the existing netlink interfaces, I *think* the preferred way
> of handling this would be adding instead a couple of MPTCP address
> flags:
>
> MPTCP_PM_ADDR_FLAG_NOBACKUP,
> MPTCP_PM_ADDR_FLAG_NOFULLMESH
>
> As e.g. IFA_F_NOPREFIXROUTE, IFA_F_NODAD,...
>
> The above should require a similar amount of code and will make it
> clear that we can't/wan't to flip the SUBFLOW or SIGNAL flags.
>
> WDYT?
Agree, will update in v2.
Thanks,
-Geliang
>
> Thanks!
>
> Paolo
>
^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: [PATCH mptcp-next 2/7] mptcp: add clear_flags in pm_netlink
2022-01-10 12:35 ` Geliang Tang
@ 2022-01-10 15:59 ` Paolo Abeni
0 siblings, 0 replies; 13+ messages in thread
From: Paolo Abeni @ 2022-01-10 15:59 UTC (permalink / raw)
To: Geliang Tang; +Cc: mptcp
On Mon, 2022-01-10 at 20:35 +0800, Geliang Tang wrote:
> On Mon, Jan 10, 2022 at 12:38:14PM +0100, Paolo Abeni wrote:
> > Hello,
> >
> > On Mon, 2022-01-10 at 11:30 +0800, Geliang Tang wrote:
> > > Splite set_flags() into two parts, set_flags() and clear_flags(), make it
> > > easy to add new flags to set or clear.
> > >
> > > This patch added a new PM command MPTCP_PM_CMD_CLEAR_FLAGS, and a new
> > > function mptcp_nl_cmd_clear_flags().
> > >
> > > Signed-off-by: Geliang Tang <geliang.tang@suse.com>
> > > ---
> > > include/uapi/linux/mptcp.h | 1 +
> > > net/mptcp/pm_netlink.c | 36 ++++++++++++++++++++++++++----------
> > > 2 files changed, 27 insertions(+), 10 deletions(-)
> > >
> > > diff --git a/include/uapi/linux/mptcp.h b/include/uapi/linux/mptcp.h
> > > index f106a3941cdf..1ca9b13c2ed0 100644
> > > --- a/include/uapi/linux/mptcp.h
> > > +++ b/include/uapi/linux/mptcp.h
> > > @@ -92,6 +92,7 @@ enum {
> > > MPTCP_PM_CMD_SET_LIMITS,
> > > MPTCP_PM_CMD_GET_LIMITS,
> > > MPTCP_PM_CMD_SET_FLAGS,
> > > + MPTCP_PM_CMD_CLEAR_FLAGS,
> > >
> > > __MPTCP_PM_CMD_AFTER_LAST
> > > };
> >
> > I'm unsure if this has already been discussed in some mtg I missed. If
> > so, I'm sorry for the late feedback :(
> >
> > Looking at the existing netlink interfaces, I *think* the preferred way
> > of handling this would be adding instead a couple of MPTCP address
> > flags:
> >
> > MPTCP_PM_ADDR_FLAG_NOBACKUP,
> > MPTCP_PM_ADDR_FLAG_NOFULLMESH
> >
> > As e.g. IFA_F_NOPREFIXROUTE, IFA_F_NODAD,...
> >
> > The above should require a similar amount of code and will make it
> > clear that we can't/wan't to flip the SUBFLOW or SIGNAL flags.
> >
> > WDYT?
>
> Agree, will update in v2.
[ mostly C&P from IRC ]
I'm sorry for being self-contradictory on this point. Think again about
the above even the
MPTCP_PM_ADDR_FLAG_NOBACKUP,
MPTCP_PM_ADDR_FLAG_NOFULLMESH
thing does look great.
I'm wondering if we could simply extend the current
MPTCP_PM_ADDR_FLAG_BACKUP handling to the MPTCP_PM_ADDR_FLAG_FULLMESH
flag? e.g. set or clear the corresponding entry->flags bit according to
the provided address flag attribute.
Cheers,
Paolo
^ permalink raw reply [flat|nested] 13+ messages in thread
end of thread, other threads:[~2022-01-10 15:59 UTC | newest]
Thread overview: 13+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2022-01-10 3:29 [PATCH mptcp-next 0/7] fullmesh flag setting support Geliang Tang
2022-01-10 3:29 ` [PATCH mptcp-next 1/7] mptcp: fix removing ids bitmap Geliang Tang
2022-01-10 11:40 ` Paolo Abeni
2022-01-10 12:33 ` Geliang Tang
2022-01-10 3:30 ` [PATCH mptcp-next 2/7] mptcp: add clear_flags in pm_netlink Geliang Tang
2022-01-10 11:38 ` Paolo Abeni
2022-01-10 12:35 ` Geliang Tang
2022-01-10 15:59 ` Paolo Abeni
2022-01-10 3:30 ` [PATCH mptcp-next 3/7] selftests: mptcp: add clear_flags in pm_nl_ctl Geliang Tang
2022-01-10 3:30 ` [PATCH mptcp-next 4/7] selftests: mptcp: add clear_flags in mptcp_join.sh Geliang Tang
2022-01-10 3:30 ` [PATCH mptcp-next 5/7] mptcp: set fullmesh flag in pm_netlink Geliang Tang
2022-01-10 3:30 ` [PATCH mptcp-next 6/7] selftests: mptcp: set fullmesh flag in pm_nl_ctl Geliang Tang
2022-01-10 3:30 ` [PATCH mptcp-next 7/7] selftests: mptcp: add fullmesh setting tests Geliang Tang
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox