[PATCH bpf-next 0/5] net/smc: Introduce smc_hs_ctrl

[PATCH bpf-next 0/5] net/smc: Introduce smc_hs_ctrl

[D. Wythe]

This patch aims to introduce BPF injection capabilities for SMC and
includes a self-test to ensure code stability.

Since the SMC protocol isn't ideal for every situation, especially
short-lived ones, most applications can't guarantee the absence of
such scenarios. Consequently, applications may need specific strategies
to decide whether to use SMC. For example, an application might limit SMC
usage to certain IP addresses or ports.

To maintain the principle of transparent replacement, we want applications
to remain unaffected even if they need specific SMC strategies. In other
words, they should not require recompilation of their code.

Additionally, we need to ensure the scalability of strategy implementation.
While using socket options or sysctl might be straightforward, it could
complicate future expansions.

Fortunately, BPF addresses these concerns effectively. Users can write
their own strategies in eBPF to determine whether to use SMC, and they can
easily modify those strategies in the future.

This is a rework of the series from [1]. Changes since [1] are limited to
the SMC parts:

1. Rename smc_ops to smc_hs_ctrl and change interface name.
2. Squash SMC patches, removing standalone non-BPF hook capability.
3. Fix typos

[1]: https://lore.kernel.org/bpf/20250123015942.94810-1-alibuda@linux.alibaba.com/#t

D. Wythe (5):
  bpf: export necessary sympols for modules with struct_ops
  net/smc: fix UAF on smcsk after smc_listen_out()
  net/smc: bpf: Introduce generic hook for handshake flow
  libbpf: fix error when st-prefix_ops and ops from differ btf
  bpf/selftests: add selftest for bpf_smc_hs_ctrl

 include/net/netns/smc.h                       |   3 +
 include/net/smc.h                             |  53 +++
 kernel/bpf/bpf_struct_ops.c                   |   2 +
 kernel/bpf/syscall.c                          |   1 +
 net/ipv4/tcp_output.c                         |  18 +-
 net/smc/Kconfig                               |  12 +
 net/smc/Makefile                              |   1 +
 net/smc/af_smc.c                              |  14 +-
 net/smc/smc_hs_bpf.c                          | 131 ++++++
 net/smc/smc_hs_bpf.h                          |  31 ++
 net/smc/smc_sysctl.c                          |  90 ++++
 tools/lib/bpf/libbpf.c                        |  37 +-
 tools/testing/selftests/bpf/config            |   4 +
 .../selftests/bpf/prog_tests/test_bpf_smc.c   | 396 ++++++++++++++++++
 tools/testing/selftests/bpf/progs/bpf_smc.c   | 117 ++++++
 15 files changed, 886 insertions(+), 24 deletions(-)
 create mode 100644 net/smc/smc_hs_bpf.c
 create mode 100644 net/smc/smc_hs_bpf.h
 create mode 100644 tools/testing/selftests/bpf/prog_tests/test_bpf_smc.c
 create mode 100644 tools/testing/selftests/bpf/progs/bpf_smc.c

-- 
2.45.0

[PATCH bpf-next 1/5] bpf: export necessary sympols for modules with struct_ops

[D. Wythe]

Exports three necessary symbols for implementing struct_ops with
tristate subsystem.

To hold or release refcnt of struct_ops refcnt by inline funcs
bpf_try_module_get and bpf_module_put which use bpf_struct_ops_get(put)
conditionally.

And to copy obj name from one to the other with effective checks by
bpf_obj_name_cpy.

Signed-off-by: D. Wythe <alibuda@linux.alibaba.com>
---
 kernel/bpf/bpf_struct_ops.c | 2 ++
 kernel/bpf/syscall.c        | 1 +
 2 files changed, 3 insertions(+)

diff --git a/kernel/bpf/bpf_struct_ops.c b/kernel/bpf/bpf_struct_ops.c
index 687a3e9c76f5..1b35c5f0562c 100644
--- a/kernel/bpf/bpf_struct_ops.c
+++ b/kernel/bpf/bpf_struct_ops.c
@@ -1162,6 +1162,7 @@ bool bpf_struct_ops_get(const void *kdata)
 	map = __bpf_map_inc_not_zero(&st_map->map, false);
 	return !IS_ERR(map);
 }
+EXPORT_SYMBOL_GPL(bpf_struct_ops_get);
 
 void bpf_struct_ops_put(const void *kdata)
 {
@@ -1173,6 +1174,7 @@ void bpf_struct_ops_put(const void *kdata)
 
 	bpf_map_put(&st_map->map);
 }
+EXPORT_SYMBOL_GPL(bpf_struct_ops_put);
 
 static bool bpf_struct_ops_valid_to_reg(struct bpf_map *map)
 {
diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c
index e63039817af3..60b4922e5e15 100644
--- a/kernel/bpf/syscall.c
+++ b/kernel/bpf/syscall.c
@@ -1201,6 +1201,7 @@ int bpf_obj_name_cpy(char *dst, const char *src, unsigned int size)
 
 	return src - orig_src;
 }
+EXPORT_SYMBOL_GPL(bpf_obj_name_cpy);
 
 int map_check_no_btf(const struct bpf_map *map,
 		     const struct btf *btf,
-- 
2.45.0

Re: [PATCH bpf-next 1/5] bpf: export necessary sympols for modules with struct_ops

[Alexandra Winter]

typo in commit message title? s/sympols/symbols/g

Re: [PATCH bpf-next 1/5] bpf: export necessary sympols for modules with struct_ops

[D. Wythe]

On Thu, Jul 31, 2025 at 02:53:08PM +0200, Alexandra Winter wrote:
> typo in commit message title? s/sympols/symbols/g

What a mistake...  I will fix it in the next series.

Thanks,
D. Wythe

> 
> 
> 

[PATCH bpf-next 2/5] net/smc: fix UAF on smcsk after smc_listen_out()

[D. Wythe]

BPF CI testing report a UAF issue:

  [   16.446633] BUG: kernel NULL pointer dereference, address: 000000000000003  0
  [   16.447134] #PF: supervisor read access in kernel mod  e
  [   16.447516] #PF: error_code(0x0000) - not-present pag  e
  [   16.447878] PGD 0 P4D   0
  [   16.448063] Oops: Oops: 0000 [#1] PREEMPT SMP NOPT  I
  [   16.448409] CPU: 0 UID: 0 PID: 9 Comm: kworker/0:1 Tainted: G           OE      6.13.0-rc3-g89e8a75fda73-dirty #4  2
  [   16.449124] Tainted: [O]=OOT_MODULE, [E]=UNSIGNED_MODUL  E
  [   16.449502] Hardware name: QEMU Ubuntu 24.04 PC (i440FX + PIIX, 1996), BIOS 1.16.3-debian-1.16.3-2 04/01/201  4
  [   16.450201] Workqueue: smc_hs_wq smc_listen_wor  k
  [   16.450531] RIP: 0010:smc_listen_work+0xc02/0x159  0
  [   16.452158] RSP: 0018:ffffb5ab40053d98 EFLAGS: 0001024  6
  [   16.452526] RAX: 0000000000000001 RBX: 0000000000000002 RCX: 000000000000030  0
  [   16.452994] RDX: 0000000000000280 RSI: 00003513840053f0 RDI: 000000000000000  0
  [   16.453492] RBP: ffffa097808e3800 R08: ffffa09782dba1e0 R09: 000000000000000  5
  [   16.453987] R10: 0000000000000000 R11: 0000000000000000 R12: ffffa0978274640  0
  [   16.454497] R13: 0000000000000000 R14: 0000000000000000 R15: ffffa09782d4092  0
  [   16.454996] FS:  0000000000000000(0000) GS:ffffa097bbc00000(0000) knlGS:000000000000000  0
  [   16.455557] CS:  0010 DS: 0000 ES: 0000 CR0: 000000008005003  3
  [   16.455961] CR2: 0000000000000030 CR3: 0000000102788004 CR4: 0000000000770ef  0
  [   16.456459] PKRU: 5555555  4
  [   16.456654] Call Trace  :
  [   16.456832]  <TASK  >
  [   16.456989]  ? __die+0x23/0x7  0
  [   16.457215]  ? page_fault_oops+0x180/0x4c  0
  [   16.457508]  ? __lock_acquire+0x3e6/0x249  0
  [   16.457801]  ? exc_page_fault+0x68/0x20  0
  [   16.458080]  ? asm_exc_page_fault+0x26/0x3  0
  [   16.458389]  ? smc_listen_work+0xc02/0x159  0
  [   16.458689]  ? smc_listen_work+0xc02/0x159  0
  [   16.458987]  ? lock_is_held_type+0x8f/0x10  0
  [   16.459284]  process_one_work+0x1ea/0x6d  0
  [   16.459570]  worker_thread+0x1c3/0x38  0
  [   16.459839]  ? __pfx_worker_thread+0x10/0x1  0
  [   16.460144]  kthread+0xe0/0x11  0
  [   16.460372]  ? __pfx_kthread+0x10/0x1  0
  [   16.460640]  ret_from_fork+0x31/0x5  0
  [   16.460896]  ? __pfx_kthread+0x10/0x1  0
  [   16.461166]  ret_from_fork_asm+0x1a/0x3  0
  [   16.461453]  </TASK  >
  [   16.461616] Modules linked in: bpf_testmod(OE) [last unloaded: bpf_testmod(OE)  ]
  [   16.462134] CR2: 000000000000003  0
  [   16.462380] ---[ end trace 0000000000000000 ]---
  [   16.462710] RIP: 0010:smc_listen_work+0xc02/0x1590

The direct cause of this issue is that after smc_listen_out_connected(),
newclcsock->sk may be NULL since it will releases the smcsk. Therefore,
if the application closes the socket immediately after accept,
newclcsock->sk can be NULL. A possible execution order could be as
follows:

smc_listen_work                                 | userspace
-----------------------------------------------------------------
lock_sock(sk)                                   |
smc_listen_out_connected()                      |
| \- smc_listen_out                             |
|    | \- release_sock                          |
     | |- sk->sk_data_ready()                   |
                                                | fd = accept();
                                                | close(fd);
                                                |  \- socket->sk = NULL;
/* newclcsock->sk is NULL now */
SMC_STAT_SERV_SUCC_INC(sock_net(newclcsock->sk))

Since smc_listen_out_connected() will not fail, simply swapping the order
of the code can easily fix this issue.

Fixes: 3b2dec2603d5 ("net/smc: restructure client and server code in af_smc")
Signed-off-by: D. Wythe <alibuda@linux.alibaba.com>
Reviewed-by: Guangguan Wang <guangguan.wang@linux.alibaba.com>
---
 net/smc/af_smc.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/net/smc/af_smc.c b/net/smc/af_smc.c
index 1882bab8e00e..dc72ff353813 100644
--- a/net/smc/af_smc.c
+++ b/net/smc/af_smc.c
@@ -2568,8 +2568,9 @@ static void smc_listen_work(struct work_struct *work)
 			goto out_decl;
 	}
 
-	smc_listen_out_connected(new_smc);
 	SMC_STAT_SERV_SUCC_INC(sock_net(newclcsock->sk), ini);
+	/* smc_listen_out() will release smcsk */
+	smc_listen_out_connected(new_smc);
 	goto out_free;
 
 out_unlock:
-- 
2.45.0

Re: [PATCH bpf-next 2/5] net/smc: fix UAF on smcsk after smc_listen_out()

[Alexandra Winter]

On 31.07.25 10:42, D. Wythe wrote:
> BPF CI testing report a UAF issue:
> 
[..]
> 
> Fixes: 3b2dec2603d5 ("net/smc: restructure client and server code in af_smc")
> Signed-off-by: D. Wythe <alibuda@linux.alibaba.com>
> Reviewed-by: Guangguan Wang <guangguan.wang@linux.alibaba.com>
> ---
>  net/smc/af_smc.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/net/smc/af_smc.c b/net/smc/af_smc.c
> index 1882bab8e00e..dc72ff353813 100644
> --- a/net/smc/af_smc.c
> +++ b/net/smc/af_smc.c
> @@ -2568,8 +2568,9 @@ static void smc_listen_work(struct work_struct *work)
>  			goto out_decl;
>  	}
>  
> -	smc_listen_out_connected(new_smc);
>  	SMC_STAT_SERV_SUCC_INC(sock_net(newclcsock->sk), ini);
> +	/* smc_listen_out() will release smcsk */
> +	smc_listen_out_connected(new_smc);
>  	goto out_free;
>  
>  out_unlock:


As this is a problem fix, you could send it directly to 'net'
instead of including it to this series.

Reviewed-by: Alexandra Winter <wintera@linux.ibm.com>

Re: [PATCH bpf-next 2/5] net/smc: fix UAF on smcsk after smc_listen_out()

[D. Wythe]

On Thu, Jul 31, 2025 at 02:57:31PM +0200, Alexandra Winter wrote:
> 
> 
> On 31.07.25 10:42, D. Wythe wrote:
> > BPF CI testing report a UAF issue:
> > 
> [..]
> > 
> > Fixes: 3b2dec2603d5 ("net/smc: restructure client and server code in af_smc")
> > Signed-off-by: D. Wythe <alibuda@linux.alibaba.com>
> > Reviewed-by: Guangguan Wang <guangguan.wang@linux.alibaba.com>
> > ---
> >  net/smc/af_smc.c | 3 ++-
> >  1 file changed, 2 insertions(+), 1 deletion(-)
> > 
> > diff --git a/net/smc/af_smc.c b/net/smc/af_smc.c
> > index 1882bab8e00e..dc72ff353813 100644
> > --- a/net/smc/af_smc.c
> > +++ b/net/smc/af_smc.c
> > @@ -2568,8 +2568,9 @@ static void smc_listen_work(struct work_struct *work)
> >  			goto out_decl;
> >  	}
> >  
> > -	smc_listen_out_connected(new_smc);
> >  	SMC_STAT_SERV_SUCC_INC(sock_net(newclcsock->sk), ini);
> > +	/* smc_listen_out() will release smcsk */
> > +	smc_listen_out_connected(new_smc);
> >  	goto out_free;
> >  
> >  out_unlock:
> 
> 
> As this is a problem fix, you could send it directly to 'net'
> instead of including it to this series.
>

Hi Alexandra,

Yes, it should be sent to net. But the problem is that if I don't carry
this patch, the BPF CI test will always crash. Maybe I should send a
copy to both net and bpf-next? Do you have any suggestions?

Best wishes,
D. Wythe


> Reviewed-by: Alexandra Winter <wintera@linux.ibm.com>

Re: [PATCH bpf-next 2/5] net/smc: fix UAF on smcsk after smc_listen_out()

[Alexandra Winter]

On 11.08.25 03:54, D. Wythe wrote:
> On Thu, Jul 31, 2025 at 02:57:31PM +0200, Alexandra Winter wrote:
>>
>>
>> On 31.07.25 10:42, D. Wythe wrote:
>>> BPF CI testing report a UAF issue:
>>>
>> [..]
>>
>> As this is a problem fix, you could send it directly to 'net'
>> instead of including it to this series.
>>
> 
> Hi Alexandra,
> 
> Yes, it should be sent to net. But the problem is that if I don't carry
> this patch, the BPF CI test will always crash. Maybe I should send a
> copy to both net and bpf-next? Do you have any suggestions?
> 
> Best wishes,
> D. Wythe

I do not have any experience with bpf-next. But typically patches
to 'net' are taken after one or two days, if there are no issues.
I'd assume they are then picked to net-next and bpf-next(?) almost instantly.
Then you would not need it in your bpf series anymore.

Sending a patch to two different mailing lists in parallel sounds like
a bad idea to me.

Re: [PATCH bpf-next 2/5] net/smc: fix UAF on smcsk after smc_listen_out()

[Jakub Kicinski]

On Mon, 11 Aug 2025 11:24:50 +0200 Alexandra Winter wrote:
> > Yes, it should be sent to net. But the problem is that if I don't carry
> > this patch, the BPF CI test will always crash. Maybe I should send a
> > copy to both net and bpf-next? Do you have any suggestions?
> 
> I do not have any experience with bpf-next. But typically patches
> to 'net' are taken after one or two days, if there are no issues.
> I'd assume they are then picked to net-next and bpf-next(?) almost instantly.
> Then you would not need it in your bpf series anymore.

AFAIU the patches which land in net will make it to -next trees after
respective PR with fixes. So


 patch -> 
          net ->
                 [next Thu] Linus ->
                                      [same day] net-next
                                      [at some point] bpf PR ->
                                                                 Linux
                                                                        -> bpf-next

What gets applied to net should be in net-next in a week,
and most -next trees within 2 weeks.

Re: [PATCH bpf-next 2/5] net/smc: fix UAF on smcsk after smc_listen_out()

[D. Wythe]

On Mon, Aug 11, 2025 at 08:33:56AM -0700, Jakub Kicinski wrote:
> On Mon, 11 Aug 2025 11:24:50 +0200 Alexandra Winter wrote:
> > > Yes, it should be sent to net. But the problem is that if I don't carry
> > > this patch, the BPF CI test will always crash. Maybe I should send a
> > > copy to both net and bpf-next? Do you have any suggestions?
> > 
> > I do not have any experience with bpf-next. But typically patches
> > to 'net' are taken after one or two days, if there are no issues.
> > I'd assume they are then picked to net-next and bpf-next(?) almost instantly.
> > Then you would not need it in your bpf series anymore.
> 
> AFAIU the patches which land in net will make it to -next trees after
> respective PR with fixes. So
> 
> 
>  patch -> 
>           net ->
>                  [next Thu] Linus ->
>                                       [same day] net-next
>                                       [at some point] bpf PR ->
>                                                                  Linux
>                                                                         -> bpf-next
> 
> What gets applied to net should be in net-next in a week,
> and most -next trees within 2 weeks.

That's very helpful.

Based on those infomations, I will send this prerequisite patch to the net
tree first. I'll then wait for about two weeks for it to propagate to
bpf-next before submitting the rest of my BPF series.

Best regards,
D. Wythe

[PATCH bpf-next 3/5] net/smc: bpf: Introduce generic hook for handshake flow

[D. Wythe]

The introduction of IPPROTO_SMC enables eBPF programs to determine
whether to use SMC based on the context of socket creation, such as
network namespaces, PID and comm name, etc.

As a subsequent enhancement, to introduce a new generic hook that
allows decisions on whether to use SMC or not at runtime, including
but not limited to local/remote IP address or ports.

User can write their own implememtion via bpf_struct_ops now to choose
whether to use SMC or not before TCP 3rd handshake to be comleted.

Signed-off-by: D. Wythe <alibuda@linux.alibaba.com>
---
 include/net/netns/smc.h |   3 +
 include/net/smc.h       |  53 ++++++++++++++++
 net/ipv4/tcp_output.c   |  18 ++++--
 net/smc/Kconfig         |  12 ++++
 net/smc/Makefile        |   1 +
 net/smc/af_smc.c        |  11 ++++
 net/smc/smc_hs_bpf.c    | 131 ++++++++++++++++++++++++++++++++++++++++
 net/smc/smc_hs_bpf.h    |  31 ++++++++++
 net/smc/smc_sysctl.c    |  90 +++++++++++++++++++++++++++
 9 files changed, 346 insertions(+), 4 deletions(-)
 create mode 100644 net/smc/smc_hs_bpf.c
 create mode 100644 net/smc/smc_hs_bpf.h

diff --git a/include/net/netns/smc.h b/include/net/netns/smc.h
index fc752a50f91b..2af1a2f0035e 100644
--- a/include/net/netns/smc.h
+++ b/include/net/netns/smc.h
@@ -17,6 +17,9 @@ struct netns_smc {
 #ifdef CONFIG_SYSCTL
 	struct ctl_table_header		*smc_hdr;
 #endif
+#if IS_ENABLED(CONFIG_SMC_HS_CTRL_BPF)
+	struct smc_hs_ctrl __rcu		*hs_ctrl;
+#endif /* CONFIG_SMC_HS_CTRL_BPF */
 	unsigned int			sysctl_autocorking_size;
 	unsigned int			sysctl_smcr_buf_type;
 	int				sysctl_smcr_testlink_time;
diff --git a/include/net/smc.h b/include/net/smc.h
index db84e4e35080..c808d5df3006 100644
--- a/include/net/smc.h
+++ b/include/net/smc.h
@@ -18,6 +18,8 @@
 #include "linux/ism.h"
 
 struct sock;
+struct tcp_sock;
+struct inet_request_sock;
 
 #define SMC_MAX_PNETID_LEN	16	/* Max. length of PNET id */
 
@@ -97,4 +99,55 @@ struct smcd_dev {
 	u8 going_away : 1;
 };
 
+#define SMC_HS_CTRL_NAME_MAX 16
+
+enum {
+	/* ops can be inherit from init_net */
+	SMC_HS_CTRL_FLAG_INHERITABLE = 0x1,
+
+	SMC_HS_CTRL_ALL_FLAGS = SMC_HS_CTRL_FLAG_INHERITABLE,
+};
+
+struct smc_hs_ctrl {
+	/* private */
+
+	struct list_head list;
+	struct module *owner;
+
+	/* public */
+
+	/* unique name */
+	char name[SMC_HS_CTRL_NAME_MAX];
+	int flags;
+
+	/* Invoked before computing SMC option for SYN packets.
+	 * We can control whether to set SMC options by returning various value.
+	 * Return 0 to disable SMC, or return any other value to enable it.
+	 */
+	int (*syn_option)(struct tcp_sock *tp);
+
+	/* Invoked before Set up SMC options for SYN-ACK packets
+	 * We can control whether to respond SMC options by returning various
+	 * value. Return 0 to disable SMC, or return any other value to enable
+	 * it.
+	 */
+	int (*synack_option)(const struct tcp_sock *tp,
+			     struct inet_request_sock *ireq);
+};
+
+#if IS_ENABLED(CONFIG_SMC_HS_CTRL_BPF)
+#define smc_call_hsbpf(init_val, sk, func, ...) ({	\
+	typeof(init_val) __ret = (init_val);		\
+	struct smc_hs_ctrl *ctrl;			\
+	rcu_read_lock();				\
+	ctrl = READ_ONCE(sock_net(sk)->smc.hs_ctrl);	\
+	if (ctrl && ctrl->func)				\
+		__ret = ctrl->func(__VA_ARGS__);	\
+	rcu_read_unlock();				\
+	__ret;						\
+})
+#else
+#define smc_call_hsbpf(init_val, ...) (init_val)
+#endif /* CONFIG_SMC_HS_CTRL_BPF */
+
 #endif	/* _SMC_H */
diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c
index 3ac8d2d17e1f..d3c2bdc725df 100644
--- a/net/ipv4/tcp_output.c
+++ b/net/ipv4/tcp_output.c
@@ -40,6 +40,7 @@
 #include <net/tcp.h>
 #include <net/mptcp.h>
 #include <net/proto_memory.h>
+#include <net/smc.h>
 
 #include <linux/compiler.h>
 #include <linux/gfp.h>
@@ -764,14 +765,18 @@ static void tcp_options_write(struct tcphdr *th, struct tcp_sock *tp,
 	mptcp_options_write(th, ptr, tp, opts);
 }
 
-static void smc_set_option(const struct tcp_sock *tp,
+static void smc_set_option(struct tcp_sock *tp,
 			   struct tcp_out_options *opts,
 			   unsigned int *remaining)
 {
 #if IS_ENABLED(CONFIG_SMC)
+	struct sock *sk = &tp->inet_conn.icsk_inet.sk;
 	if (static_branch_unlikely(&tcp_have_smc)) {
 		if (tp->syn_smc) {
-			if (*remaining >= TCPOLEN_EXP_SMC_BASE_ALIGNED) {
+			tp->syn_smc = !!smc_call_hsbpf(1, sk, syn_option, tp);
+			/* re-check syn_smc */
+			if (tp->syn_smc &&
+			    *remaining >= TCPOLEN_EXP_SMC_BASE_ALIGNED) {
 				opts->options |= OPTION_SMC;
 				*remaining -= TCPOLEN_EXP_SMC_BASE_ALIGNED;
 			}
@@ -781,14 +786,19 @@ static void smc_set_option(const struct tcp_sock *tp,
 }
 
 static void smc_set_option_cond(const struct tcp_sock *tp,
-				const struct inet_request_sock *ireq,
+				struct inet_request_sock *ireq,
 				struct tcp_out_options *opts,
 				unsigned int *remaining)
 {
 #if IS_ENABLED(CONFIG_SMC)
+	const struct sock *sk = &tp->inet_conn.icsk_inet.sk;
 	if (static_branch_unlikely(&tcp_have_smc)) {
 		if (tp->syn_smc && ireq->smc_ok) {
-			if (*remaining >= TCPOLEN_EXP_SMC_BASE_ALIGNED) {
+			ireq->smc_ok = !!smc_call_hsbpf(1, sk, synack_option,
+							 tp, ireq);
+			/* re-check smc_ok */
+			if (ireq->smc_ok &&
+			    *remaining >= TCPOLEN_EXP_SMC_BASE_ALIGNED) {
 				opts->options |= OPTION_SMC;
 				*remaining -= TCPOLEN_EXP_SMC_BASE_ALIGNED;
 			}
diff --git a/net/smc/Kconfig b/net/smc/Kconfig
index ba5e6a2dd2fd..90408afa6119 100644
--- a/net/smc/Kconfig
+++ b/net/smc/Kconfig
@@ -33,3 +33,15 @@ config SMC_LO
 	  of architecture or hardware.
 
 	  if unsure, say N.
+
+config SMC_HS_CTRL_BPF
+	bool "Generic eBPF hook for SMC handshake flow"
+	depends on SMC && BPF_SYSCALL
+	default n
+	help
+	  SMC_HS_CTRL_BPF enables support to register generic eBPF hook for SMC
+	  handshake flow, which offer much greater flexibility in modifying the behavior
+	  of the SMC protocol stack compared to a complete kernel-based approach. Select
+	  this option if you want filtring the handshake process via eBPF programs.
+
+	  if unsure, say N.
\ No newline at end of file
diff --git a/net/smc/Makefile b/net/smc/Makefile
index 60f1c87d5212..a3ac276cdd39 100644
--- a/net/smc/Makefile
+++ b/net/smc/Makefile
@@ -7,3 +7,4 @@ smc-y += smc_cdc.o smc_tx.o smc_rx.o smc_close.o smc_ism.o smc_netlink.o smc_sta
 smc-y += smc_tracepoint.o smc_inet.o
 smc-$(CONFIG_SYSCTL) += smc_sysctl.o
 smc-$(CONFIG_SMC_LO) += smc_loopback.o
+smc-$(CONFIG_SMC_HS_CTRL_BPF) += smc_hs_bpf.o
\ No newline at end of file
diff --git a/net/smc/af_smc.c b/net/smc/af_smc.c
index dc72ff353813..4973c408b80c 100644
--- a/net/smc/af_smc.c
+++ b/net/smc/af_smc.c
@@ -59,6 +59,7 @@
 #include "smc_sysctl.h"
 #include "smc_loopback.h"
 #include "smc_inet.h"
+#include "smc_hs_bpf.h"
 
 static DEFINE_MUTEX(smc_server_lgr_pending);	/* serialize link group
 						 * creation on server
@@ -3610,8 +3611,18 @@ static int __init smc_init(void)
 		pr_err("%s: smc_inet_init fails with %d\n", __func__, rc);
 		goto out_ulp;
 	}
+
+	rc = bpf_smc_hs_ctrl_init();
+	if (rc) {
+		pr_err("%s: bpf_smc_hs_ctrl_init fails with %d\n", __func__,
+		       rc);
+		goto out_inet;
+	}
+
 	static_branch_enable(&tcp_have_smc);
 	return 0;
+out_inet:
+	smc_inet_exit();
 out_ulp:
 	tcp_unregister_ulp(&smc_ulp_ops);
 out_lo:
diff --git a/net/smc/smc_hs_bpf.c b/net/smc/smc_hs_bpf.c
new file mode 100644
index 000000000000..d5a07532cf78
--- /dev/null
+++ b/net/smc/smc_hs_bpf.c
@@ -0,0 +1,131 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ *  Shared Memory Communications over RDMA (SMC-R) and RoCE
+ *
+ *  Generic hook for SMC subsystem.
+ *
+ *  Copyright IBM Corp. 2016
+ *  Copyright (c) 2024, Alibaba Inc.
+ *
+ *  Author: D. Wythe <alibuda@linux.alibaba.com>
+ */
+
+#include <linux/bpf_verifier.h>
+#include <linux/bpf.h>
+#include <linux/btf.h>
+#include <linux/rculist.h>
+
+#include "smc_hs_bpf.h"
+
+static DEFINE_SPINLOCK(smc_hs_ctrl_list_lock);
+static LIST_HEAD(smc_hs_ctrl_list);
+
+static int smc_hs_ctrl_reg(struct smc_hs_ctrl *ctrl)
+{
+	int ret = 0;
+
+	spin_lock(&smc_hs_ctrl_list_lock);
+	/* already exist or duplicate name */
+	if (smc_hs_ctrl_find_by_name(ctrl->name))
+		ret = -EEXIST;
+	else
+		list_add_tail_rcu(&ctrl->list, &smc_hs_ctrl_list);
+	spin_unlock(&smc_hs_ctrl_list_lock);
+	return ret;
+}
+
+static void smc_hs_ctrl_unreg(struct smc_hs_ctrl *ctrl)
+{
+	spin_lock(&smc_hs_ctrl_list_lock);
+	list_del_rcu(&ctrl->list);
+	spin_unlock(&smc_hs_ctrl_list_lock);
+
+	/* Ensure that all readers to complete */
+	synchronize_rcu();
+}
+
+struct smc_hs_ctrl *smc_hs_ctrl_find_by_name(const char *name)
+{
+	struct smc_hs_ctrl *ctrl;
+
+	list_for_each_entry_rcu(ctrl, &smc_hs_ctrl_list, list) {
+		if (strcmp(ctrl->name, name) == 0)
+			return ctrl;
+	}
+	return NULL;
+}
+
+static int __bpf_smc_stub_set_tcp_option(struct tcp_sock *tp) { return 1; }
+static int __bpf_smc_stub_set_tcp_option_cond(const struct tcp_sock *tp,
+					      struct inet_request_sock *ireq)
+{
+	return 1;
+}
+
+static struct smc_hs_ctrl __bpf_smc_hs_ctrl = {
+	.syn_option	= __bpf_smc_stub_set_tcp_option,
+	.synack_option	= __bpf_smc_stub_set_tcp_option_cond,
+};
+
+static int smc_bpf_hs_ctrl_init(struct btf *btf) { return 0; }
+
+static int smc_bpf_hs_ctrl_reg(void *kdata, struct bpf_link *link)
+{
+	return smc_hs_ctrl_reg(kdata);
+}
+
+static void smc_bpf_hs_ctrl_unreg(void *kdata, struct bpf_link *link)
+{
+	smc_hs_ctrl_unreg(kdata);
+}
+
+static int smc_bpf_hs_ctrl_init_member(const struct btf_type *t,
+				       const struct btf_member *member,
+				       void *kdata, const void *udata)
+{
+	const struct smc_hs_ctrl *u_ctrl;
+	struct smc_hs_ctrl *k_ctrl;
+	u32 moff;
+
+	u_ctrl = (const struct smc_hs_ctrl *)udata;
+	k_ctrl = (struct smc_hs_ctrl *)kdata;
+
+	moff = __btf_member_bit_offset(t, member) / 8;
+	switch (moff) {
+	case offsetof(struct smc_hs_ctrl, name):
+		if (bpf_obj_name_cpy(k_ctrl->name, u_ctrl->name,
+				     sizeof(u_ctrl->name)) <= 0)
+			return -EINVAL;
+		return 1;
+	case offsetof(struct smc_hs_ctrl, flags):
+		if (u_ctrl->flags & ~SMC_HS_CTRL_ALL_FLAGS)
+			return -EINVAL;
+		k_ctrl->flags = u_ctrl->flags;
+		return 1;
+	default:
+		break;
+	}
+
+	return 0;
+}
+
+static const struct bpf_verifier_ops smc_bpf_verifier_ops = {
+	.get_func_proto		= bpf_base_func_proto,
+	.is_valid_access	= bpf_tracing_btf_ctx_access,
+};
+
+static struct bpf_struct_ops bpf_smc_hs_ctrl_ops = {
+	.name		= "smc_hs_ctrl",
+	.init		= smc_bpf_hs_ctrl_init,
+	.reg		= smc_bpf_hs_ctrl_reg,
+	.unreg		= smc_bpf_hs_ctrl_unreg,
+	.cfi_stubs	= &__bpf_smc_hs_ctrl,
+	.verifier_ops	= &smc_bpf_verifier_ops,
+	.init_member	= smc_bpf_hs_ctrl_init_member,
+	.owner		= THIS_MODULE,
+};
+
+int bpf_smc_hs_ctrl_init(void)
+{
+	return register_bpf_struct_ops(&bpf_smc_hs_ctrl_ops, smc_hs_ctrl);
+}
diff --git a/net/smc/smc_hs_bpf.h b/net/smc/smc_hs_bpf.h
new file mode 100644
index 000000000000..d6ca3377f682
--- /dev/null
+++ b/net/smc/smc_hs_bpf.h
@@ -0,0 +1,31 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ *  Shared Memory Communications over RDMA (SMC-R) and RoCE
+ *
+ *  Generic hook for SMC subsystem.
+ *
+ *  Copyright IBM Corp. 2016
+ *  Copyright (c) 2024, Alibaba Inc.
+ *
+ *  Author: D. Wythe <alibuda@linux.alibaba.com>
+ */
+
+#ifndef __SMC_HS_CTRL
+#define __SMC_HS_CTRL
+
+#include <net/smc.h>
+
+/* Find hs_ctrl by the target name, which required to be a c-string.
+ * Return NULL if no such ctrl was found,otherwise, return a valid ctrl.
+ *
+ * Note: Caller MUST ensure it's was invoked under rcu_read_lock.
+ */
+struct smc_hs_ctrl *smc_hs_ctrl_find_by_name(const char *name);
+
+#if IS_ENABLED(CONFIG_SMC_HS_CTRL_BPF)
+int bpf_smc_hs_ctrl_init(void);
+#else
+static inline int bpf_smc_hs_ctrl_init(void) { return 0; }
+#endif /* CONFIG_SMC_HS_CTRL_BPF */
+
+#endif /* __SMC_HS_CTRL */
diff --git a/net/smc/smc_sysctl.c b/net/smc/smc_sysctl.c
index 2fab6456f765..9b15f5f49039 100644
--- a/net/smc/smc_sysctl.c
+++ b/net/smc/smc_sysctl.c
@@ -18,6 +18,7 @@
 #include "smc_core.h"
 #include "smc_llc.h"
 #include "smc_sysctl.h"
+#include "smc_hs_bpf.h"
 
 static int min_sndbuf = SMC_BUF_MIN_SIZE;
 static int min_rcvbuf = SMC_BUF_MIN_SIZE;
@@ -30,6 +31,69 @@ static int links_per_lgr_max = SMC_LINKS_ADD_LNK_MAX;
 static int conns_per_lgr_min = SMC_CONN_PER_LGR_MIN;
 static int conns_per_lgr_max = SMC_CONN_PER_LGR_MAX;
 
+#if IS_ENABLED(CONFIG_SMC_HS_CTRL_BPF)
+static int smc_net_replace_smc_hs_ctrl(struct net *net, const char *name)
+{
+	struct smc_hs_ctrl *ctrl = NULL;
+
+	rcu_read_lock();
+	/* null or empty name ask to clear current ctrl */
+	if (name && name[0]) {
+		ctrl = smc_hs_ctrl_find_by_name(name);
+		if (!ctrl) {
+			rcu_read_unlock();
+			return -EINVAL;
+		}
+		/* no change, just return */
+		if (ctrl == rcu_dereference(net->smc.hs_ctrl)) {
+			rcu_read_unlock();
+			return 0;
+		}
+		if (!bpf_try_module_get(ctrl, ctrl->owner)) {
+			rcu_read_unlock();
+			return -EBUSY;
+		}
+	}
+	/* xhcg old ctrl with the new one atomically */
+	ctrl = xchg(&net->smc.hs_ctrl, ctrl);
+	/* release old ctrl */
+	if (ctrl)
+		bpf_module_put(ctrl, ctrl->owner);
+
+	rcu_read_unlock();
+	return 0;
+}
+
+static int proc_smc_hs_ctrl(const struct ctl_table *ctl, int write,
+			    void *buffer, size_t *lenp, loff_t *ppos)
+{
+	struct net *net = container_of(ctl->data, struct net, smc.hs_ctrl);
+	char val[SMC_HS_CTRL_NAME_MAX];
+	const struct ctl_table tbl = {
+		.data = val,
+		.maxlen = SMC_HS_CTRL_NAME_MAX,
+	};
+	struct smc_hs_ctrl *ctrl;
+	int ret;
+
+	rcu_read_lock();
+	ctrl = rcu_dereference(net->smc.hs_ctrl);
+	if (ctrl)
+		memcpy(val, ctrl->name, sizeof(ctrl->name));
+	else
+		val[0] = '\0';
+	rcu_read_unlock();
+
+	ret = proc_dostring(&tbl, write, buffer, lenp, ppos);
+	if (ret)
+		return ret;
+
+	if (write)
+		ret = smc_net_replace_smc_hs_ctrl(net, val);
+	return ret;
+}
+#endif /* CONFIG_SMC_HS_CTRL_BPF */
+
 static struct ctl_table smc_table[] = {
 	{
 		.procname       = "autocorking_size",
@@ -99,6 +163,15 @@ static struct ctl_table smc_table[] = {
 		.extra1		= SYSCTL_ZERO,
 		.extra2		= SYSCTL_ONE,
 	},
+#if IS_ENABLED(CONFIG_SMC_HS_CTRL_BPF)
+	{
+		.procname	= "hs_ctrl",
+		.data		= &init_net.smc.hs_ctrl,
+		.mode		= 0644,
+		.maxlen		= SMC_HS_CTRL_NAME_MAX,
+		.proc_handler	= proc_smc_hs_ctrl,
+	},
+#endif /* CONFIG_SMC_HS_CTRL_BPF */
 };
 
 int __net_init smc_sysctl_net_init(struct net *net)
@@ -109,6 +182,16 @@ int __net_init smc_sysctl_net_init(struct net *net)
 	table = smc_table;
 	if (!net_eq(net, &init_net)) {
 		int i;
+#if IS_ENABLED(CONFIG_SMC_HS_CTRL_BPF)
+		struct smc_hs_ctrl *ctrl;
+
+		rcu_read_lock();
+		ctrl = rcu_dereference(init_net.smc.hs_ctrl);
+		if (ctrl && ctrl->flags & SMC_HS_CTRL_FLAG_INHERITABLE &&
+		    bpf_try_module_get(ctrl, ctrl->owner))
+			rcu_assign_pointer(net->smc.hs_ctrl, ctrl);
+		rcu_read_unlock();
+#endif /* CONFIG_SMC_HS_CTRL_BPF */
 
 		table = kmemdup(table, sizeof(smc_table), GFP_KERNEL);
 		if (!table)
@@ -139,6 +222,9 @@ int __net_init smc_sysctl_net_init(struct net *net)
 	if (!net_eq(net, &init_net))
 		kfree(table);
 err_alloc:
+#if IS_ENABLED(CONFIG_SMC_HS_CTRL_BPF)
+	smc_net_replace_smc_hs_ctrl(net, NULL);
+#endif /* CONFIG_SMC_HS_CTRL_BPF */
 	return -ENOMEM;
 }
 
@@ -148,6 +234,10 @@ void __net_exit smc_sysctl_net_exit(struct net *net)
 
 	table = net->smc.smc_hdr->ctl_table_arg;
 	unregister_net_sysctl_table(net->smc.smc_hdr);
+#if IS_ENABLED(CONFIG_SMC_HS_CTRL_BPF)
+	smc_net_replace_smc_hs_ctrl(net, NULL);
+#endif /* CONFIG_SMC_HS_CTRL_BPF */
+
 	if (!net_eq(net, &init_net))
 		kfree(table);
 }
-- 
2.45.0

Re: [PATCH bpf-next 3/5] net/smc: bpf: Introduce generic hook for handshake flow

[kernel test robot]

Hi Wythe,

kernel test robot noticed the following build warnings:

[auto build test WARNING on bpf-next/master]

url:    https://github.com/intel-lab-lkp/linux/commits/D-Wythe/bpf-export-necessary-sympols-for-modules-with-struct_ops/20250731-164431
base:   https://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf-next.git master
patch link:    https://lore.kernel.org/r/20250731084240.86550-4-alibuda%40linux.alibaba.com
patch subject: [PATCH bpf-next 3/5] net/smc: bpf: Introduce generic hook for handshake flow
config: arm-randconfig-003-20250801 (https://download.01.org/0day-ci/archive/20250801/202508010316.wuSPjSOr-lkp@intel.com/config)
compiler: arm-linux-gnueabi-gcc (GCC) 10.5.0
reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20250801/202508010316.wuSPjSOr-lkp@intel.com/reproduce)

If you fix the issue in a separate patch/commit (i.e. not just a new version of
the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <lkp@intel.com>
| Closes: https://lore.kernel.org/oe-kbuild-all/202508010316.wuSPjSOr-lkp@intel.com/

All warnings (new ones prefixed by >>):

   net/ipv4/tcp_output.c: In function 'smc_set_option':
>> net/ipv4/tcp_output.c:773:15: warning: unused variable 'sk' [-Wunused-variable]
     773 |  struct sock *sk = &tp->inet_conn.icsk_inet.sk;
         |               ^~
   net/ipv4/tcp_output.c: In function 'smc_set_option_cond':
   net/ipv4/tcp_output.c:794:21: warning: unused variable 'sk' [-Wunused-variable]
     794 |  const struct sock *sk = &tp->inet_conn.icsk_inet.sk;
         |                     ^~


vim +/sk +773 net/ipv4/tcp_output.c

   767	
   768	static void smc_set_option(struct tcp_sock *tp,
   769				   struct tcp_out_options *opts,
   770				   unsigned int *remaining)
   771	{
   772	#if IS_ENABLED(CONFIG_SMC)
 > 773		struct sock *sk = &tp->inet_conn.icsk_inet.sk;
   774		if (static_branch_unlikely(&tcp_have_smc)) {
   775			if (tp->syn_smc) {
   776				tp->syn_smc = !!smc_call_hsbpf(1, sk, syn_option, tp);
   777				/* re-check syn_smc */
   778				if (tp->syn_smc &&
   779				    *remaining >= TCPOLEN_EXP_SMC_BASE_ALIGNED) {
   780					opts->options |= OPTION_SMC;
   781					*remaining -= TCPOLEN_EXP_SMC_BASE_ALIGNED;
   782				}
   783			}
   784		}
   785	#endif
   786	}
   787	

-- 
0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests/wiki

[PATCH bpf-next 4/5] libbpf: fix error when st-prefix_ops and ops from differ btf

[D. Wythe]

When a struct_ops named xxx_ops was registered by a module, and
it will be used in both built-in modules and the module itself,
so that the btf_type of xxx_ops will be present in btf_vmlinux
instead of in btf_mod, which means that the btf_type of
bpf_struct_ops_xxx_ops and xxx_ops will not be in the same btf.

Here are four possible case:

+--------+---------------+-------------+---------------------------------+
|        | st_ops_xxx_ops| xxx_ops     |                                 |
+--------+---------------+-------------+---------------------------------+
| case 0 | btf_vmlinux   | bft_vmlinux | be used and reg only in vmlinux |
+--------+---------------+-------------+---------------------------------+
| case 1 | btf_vmlinux   | bpf_mod     | INVALID                         |
+--------+---------------+-------------+---------------------------------+
| case 2 | btf_mod       | btf_vmlinux | reg in mod but be used both in  |
|        |               |             | vmlinux and mod.                |
+--------+---------------+-------------+---------------------------------+
| case 3 | btf_mod       | btf_mod     | be used and reg only in mod     |
+--------+---------------+-------------+---------------------------------+

At present, cases 0, 1, and 3 can be correctly identified, because
st_ops_xxx_ops is searched from the same btf with xxx_ops. In order to
handle case 2 correctly without affecting other cases, we cannot simply
change the search method for st_ops_xxx_ops from find_btf_by_prefix_kind()
to find_ksym_btf_id(), because in this way, case 1 will not be
recognized anymore.

To address the issue, we always look for st_ops_xxx_ops first,
figure out the btf, and then look for xxx_ops with the very btf to avoid
such issue.

Fixes: 590a00888250 ("bpf: libbpf: Add STRUCT_OPS support")
Signed-off-by: D. Wythe <alibuda@linux.alibaba.com>
Acked-by: Andrii Nakryiko <andrii@kernel.org>
---
 tools/lib/bpf/libbpf.c | 37 ++++++++++++++++++-------------------
 1 file changed, 18 insertions(+), 19 deletions(-)

diff --git a/tools/lib/bpf/libbpf.c b/tools/lib/bpf/libbpf.c
index e067cb5776bd..3cbb6d7aadb1 100644
--- a/tools/lib/bpf/libbpf.c
+++ b/tools/lib/bpf/libbpf.c
@@ -1013,35 +1013,34 @@ find_struct_ops_kern_types(struct bpf_object *obj, const char *tname_raw,
 	const struct btf_member *kern_data_member;
 	struct btf *btf = NULL;
 	__s32 kern_vtype_id, kern_type_id;
-	char tname[256];
+	char tname[256], stname[256];
 	__u32 i;
 
 	snprintf(tname, sizeof(tname), "%.*s",
 		 (int)bpf_core_essential_name_len(tname_raw), tname_raw);
 
-	kern_type_id = find_ksym_btf_id(obj, tname, BTF_KIND_STRUCT,
-					&btf, mod_btf);
-	if (kern_type_id < 0) {
-		pr_warn("struct_ops init_kern: struct %s is not found in kernel BTF\n",
-			tname);
-		return kern_type_id;
-	}
-	kern_type = btf__type_by_id(btf, kern_type_id);
+	snprintf(stname, sizeof(stname), "%s%.*s", STRUCT_OPS_VALUE_PREFIX,
+		 (int)strlen(tname), tname);
 
-	/* Find the corresponding "map_value" type that will be used
-	 * in map_update(BPF_MAP_TYPE_STRUCT_OPS).  For example,
-	 * find "struct bpf_struct_ops_tcp_congestion_ops" from the
-	 * btf_vmlinux.
+	/* Look for the corresponding "map_value" type that will be used
+	 * in map_update(BPF_MAP_TYPE_STRUCT_OPS) first, figure out the btf
+	 * and the mod_btf.
+	 * For example, find "struct bpf_struct_ops_tcp_congestion_ops".
 	 */
-	kern_vtype_id = find_btf_by_prefix_kind(btf, STRUCT_OPS_VALUE_PREFIX,
-						tname, BTF_KIND_STRUCT);
+	kern_vtype_id = find_ksym_btf_id(obj, stname, BTF_KIND_STRUCT, &btf, mod_btf);
 	if (kern_vtype_id < 0) {
-		pr_warn("struct_ops init_kern: struct %s%s is not found in kernel BTF\n",
-			STRUCT_OPS_VALUE_PREFIX, tname);
+		pr_warn("struct_ops init_kern: struct %s is not found in kernel BTF\n", stname);
 		return kern_vtype_id;
 	}
 	kern_vtype = btf__type_by_id(btf, kern_vtype_id);
 
+	kern_type_id = btf__find_by_name_kind(btf, tname, BTF_KIND_STRUCT);
+	if (kern_type_id < 0) {
+		pr_warn("struct_ops init_kern: struct %s is not found in kernel BTF\n", tname);
+		return kern_type_id;
+	}
+	kern_type = btf__type_by_id(btf, kern_type_id);
+
 	/* Find "struct tcp_congestion_ops" from
 	 * struct bpf_struct_ops_tcp_congestion_ops {
 	 *	[ ... ]
@@ -1054,8 +1053,8 @@ find_struct_ops_kern_types(struct bpf_object *obj, const char *tname_raw,
 			break;
 	}
 	if (i == btf_vlen(kern_vtype)) {
-		pr_warn("struct_ops init_kern: struct %s data is not found in struct %s%s\n",
-			tname, STRUCT_OPS_VALUE_PREFIX, tname);
+		pr_warn("struct_ops init_kern: struct %s data is not found in struct %s\n",
+			tname, stname);
 		return -EINVAL;
 	}
 
-- 
2.45.0

[PATCH bpf-next 5/5] bpf/selftests: add selftest for bpf_smc_hs_ctrl

[D. Wythe]

This tests introduces a tiny smc_hs_ctrl for filtering SMC connections
based on IP pairs, and also adds a realistic topology model to verify it.

Also, we can only use SMC loopback under CI test, so an additional
configuration needs to be enabled.

Follow the steps below to run this test.

make -C tools/testing/selftests/bpf
cd tools/testing/selftests/bpf
sudo ./test_progs -t smc

Results shows:
Summary: 1/1 PASSED, 0 SKIPPED, 0 FAILED

Signed-off-by: D. Wythe <alibuda@linux.alibaba.com>
Tested-by: Saket Kumar Bhaskar <skb99@linux.ibm.com>
Reviewed-by: Zhu Yanjun <yanjun.zhu@linux.dev>
---
 tools/testing/selftests/bpf/config            |   4 +
 .../selftests/bpf/prog_tests/test_bpf_smc.c   | 396 ++++++++++++++++++
 tools/testing/selftests/bpf/progs/bpf_smc.c   | 117 ++++++
 3 files changed, 517 insertions(+)
 create mode 100644 tools/testing/selftests/bpf/prog_tests/test_bpf_smc.c
 create mode 100644 tools/testing/selftests/bpf/progs/bpf_smc.c

diff --git a/tools/testing/selftests/bpf/config b/tools/testing/selftests/bpf/config
index 7247833fe623..257ac8537212 100644
--- a/tools/testing/selftests/bpf/config
+++ b/tools/testing/selftests/bpf/config
@@ -119,3 +119,7 @@ CONFIG_XDP_SOCKETS=y
 CONFIG_XFRM_INTERFACE=y
 CONFIG_TCP_CONG_DCTCP=y
 CONFIG_TCP_CONG_BBR=y
+CONFIG_INFINIBAND=y
+CONFIG_SMC=y
+CONFIG_SMC_HS_CTRL_BPF=y
+CONFIG_SMC_LO=y
\ No newline at end of file
diff --git a/tools/testing/selftests/bpf/prog_tests/test_bpf_smc.c b/tools/testing/selftests/bpf/prog_tests/test_bpf_smc.c
new file mode 100644
index 000000000000..200658d28793
--- /dev/null
+++ b/tools/testing/selftests/bpf/prog_tests/test_bpf_smc.c
@@ -0,0 +1,396 @@
+// SPDX-License-Identifier: GPL-2.0
+#include <test_progs.h>
+#include <linux/genetlink.h>
+#include "network_helpers.h"
+#include "bpf_smc.skel.h"
+
+#ifndef IPPROTO_SMC
+#define IPPROTO_SMC 256
+#endif
+
+#define CLIENT_IP			"127.0.0.1"
+#define SERVER_IP			"127.0.1.0"
+#define SERVER_IP_VIA_RISK_PATH	"127.0.2.0"
+
+#define SERVICE_1	11234
+#define SERVICE_2	22345
+#define SERVICE_3	33456
+
+#define TEST_NS	"bpf_smc_netns"
+
+static struct netns_obj *test_netns;
+
+struct smc_strat_ip_key {
+	__u32  sip;
+	__u32  dip;
+};
+
+struct smc_strat_ip_value {
+	__u8	mode;
+};
+
+#if defined(__s390x__)
+/* s390x has default seid  */
+static bool setup_ueid(void) { return true; }
+static void cleanup_ueid(void) {}
+#else
+enum {
+	SMC_NETLINK_ADD_UEID = 10,
+	SMC_NETLINK_REMOVE_UEID
+};
+
+enum {
+	SMC_NLA_EID_TABLE_UNSPEC,
+	SMC_NLA_EID_TABLE_ENTRY,    /* string */
+};
+
+struct msgtemplate {
+	struct nlmsghdr n;
+	struct genlmsghdr g;
+	char buf[1024];
+};
+
+#define GENLMSG_DATA(glh)	((void *)(NLMSG_DATA(glh) + GENL_HDRLEN))
+#define GENLMSG_PAYLOAD(glh)	(NLMSG_PAYLOAD(glh, 0) - GENL_HDRLEN)
+#define NLA_DATA(na)		((void *)((char *)(na) + NLA_HDRLEN))
+#define NLA_PAYLOAD(len)	((len) - NLA_HDRLEN)
+
+#define SMC_GENL_FAMILY_NAME	"SMC_GEN_NETLINK"
+#define SMC_BPFTEST_UEID	"SMC-BPFTEST-UEID"
+
+static uint16_t smc_nl_family_id = -1;
+
+static int send_cmd(int fd, __u16 nlmsg_type, __u32 nlmsg_pid,
+		    __u16 nlmsg_flags, __u8 genl_cmd, __u16 nla_type,
+		    void *nla_data, int nla_len)
+{
+	struct nlattr *na;
+	struct sockaddr_nl nladdr;
+	int r, buflen;
+	char *buf;
+
+	struct msgtemplate msg = {0};
+
+	msg.n.nlmsg_len = NLMSG_LENGTH(GENL_HDRLEN);
+	msg.n.nlmsg_type = nlmsg_type;
+	msg.n.nlmsg_flags = nlmsg_flags;
+	msg.n.nlmsg_seq = 0;
+	msg.n.nlmsg_pid = nlmsg_pid;
+	msg.g.cmd = genl_cmd;
+	msg.g.version = 1;
+	na = (struct nlattr *)GENLMSG_DATA(&msg);
+	na->nla_type = nla_type;
+	na->nla_len = nla_len + 1 + NLA_HDRLEN;
+	memcpy(NLA_DATA(na), nla_data, nla_len);
+	msg.n.nlmsg_len += NLMSG_ALIGN(na->nla_len);
+
+	buf = (char *)&msg;
+	buflen = msg.n.nlmsg_len;
+	memset(&nladdr, 0, sizeof(nladdr));
+	nladdr.nl_family = AF_NETLINK;
+
+	while ((r = sendto(fd, buf, buflen, 0, (struct sockaddr *)&nladdr,
+			   sizeof(nladdr))) < buflen) {
+		if (r > 0) {
+			buf += r;
+			buflen -= r;
+		} else if (errno != EAGAIN) {
+			return -1;
+		}
+	}
+	return 0;
+}
+
+static bool get_smc_nl_family_id(void)
+{
+	struct sockaddr_nl nl_src;
+	struct msgtemplate msg;
+	struct nlattr *nl;
+	int fd, ret;
+	pid_t pid;
+
+	fd = socket(AF_NETLINK, SOCK_RAW, NETLINK_GENERIC);
+	if (!ASSERT_OK_FD(fd, "nl_family socket"))
+		return false;
+
+	pid = getpid();
+
+	memset(&nl_src, 0, sizeof(nl_src));
+	nl_src.nl_family = AF_NETLINK;
+	nl_src.nl_pid = pid;
+
+	ret = bind(fd, (struct sockaddr *)&nl_src, sizeof(nl_src));
+	if (!ASSERT_OK(ret, "nl_family bind"))
+		goto fail;
+
+	ret = send_cmd(fd, GENL_ID_CTRL, pid,
+		       NLM_F_REQUEST, CTRL_CMD_GETFAMILY,
+		       CTRL_ATTR_FAMILY_NAME, (void *)SMC_GENL_FAMILY_NAME,
+		       strlen(SMC_GENL_FAMILY_NAME));
+	if (!ASSERT_OK(ret, "nl_family query"))
+		goto fail;
+
+	ret = recv(fd, &msg, sizeof(msg), 0);
+	if (!ASSERT_FALSE(msg.n.nlmsg_type == NLMSG_ERROR || ret < 0 ||
+			  !NLMSG_OK(&msg.n, ret), "nl_family response"))
+		goto fail;
+
+	nl = (struct nlattr *)GENLMSG_DATA(&msg);
+	nl = (struct nlattr *)((char *)nl + NLA_ALIGN(nl->nla_len));
+	if (!ASSERT_EQ(nl->nla_type, CTRL_ATTR_FAMILY_ID, "nl_family nla type"))
+		goto fail;
+
+	smc_nl_family_id = *(uint16_t *)NLA_DATA(nl);
+	close(fd);
+	return true;
+fail:
+	close(fd);
+	return false;
+}
+
+static bool smc_ueid(int op)
+{
+	struct sockaddr_nl nl_src;
+	struct msgtemplate msg;
+	struct nlmsgerr *err;
+	char test_ueid[32];
+	int fd, ret;
+	pid_t pid;
+
+	/* UEID required */
+	memset(test_ueid, '\x20', sizeof(test_ueid));
+	memcpy(test_ueid, SMC_BPFTEST_UEID, strlen(SMC_BPFTEST_UEID));
+	fd = socket(AF_NETLINK, SOCK_RAW, NETLINK_GENERIC);
+	if (!ASSERT_OK_FD(fd, "ueid socket"))
+		return false;
+
+	pid = getpid();
+	memset(&nl_src, 0, sizeof(nl_src));
+	nl_src.nl_family = AF_NETLINK;
+	nl_src.nl_pid = pid;
+
+	ret = bind(fd, (struct sockaddr *)&nl_src, sizeof(nl_src));
+	if (!ASSERT_OK(ret, "ueid bind"))
+		goto fail;
+
+	ret = send_cmd(fd, smc_nl_family_id, pid,
+		       NLM_F_REQUEST | NLM_F_ACK, op, SMC_NLA_EID_TABLE_ENTRY,
+		       (void *)test_ueid, sizeof(test_ueid));
+	if (!ASSERT_OK(ret, "ueid cmd"))
+		goto fail;
+
+	ret = recv(fd, &msg, sizeof(msg), 0);
+	if (!ASSERT_FALSE(ret < 0 ||
+			  !NLMSG_OK(&msg.n, ret), "ueid response"))
+		goto fail;
+
+	if (msg.n.nlmsg_type == NLMSG_ERROR) {
+		err = NLMSG_DATA(&msg);
+		switch (op) {
+		case SMC_NETLINK_REMOVE_UEID:
+			if (!ASSERT_FALSE((err->error && err->error != -ENOENT),
+					  "ueid remove"))
+				goto fail;
+			break;
+		case SMC_NETLINK_ADD_UEID:
+			if (!ASSERT_OK(err->error, "ueid add"))
+				goto fail;
+			break;
+		default:
+			break;
+		}
+	}
+	close(fd);
+	return true;
+fail:
+	close(fd);
+	return false;
+}
+
+static bool setup_ueid(void)
+{
+	/* get smc nl id */
+	if (!get_smc_nl_family_id())
+		return false;
+	/* clear old ueid for bpftest */
+	smc_ueid(SMC_NETLINK_REMOVE_UEID);
+	/* smc-loopback required ueid */
+	return smc_ueid(SMC_NETLINK_ADD_UEID);
+}
+
+static void cleanup_ueid(void)
+{
+	smc_ueid(SMC_NETLINK_REMOVE_UEID);
+}
+#endif /* __s390x__ */
+
+static bool setup_netns(void)
+{
+	test_netns = netns_new(TEST_NS, true);
+	if (!ASSERT_OK_PTR(test_netns, "open net namespace"))
+		goto fail_netns;
+
+	if (!ASSERT_OK(system("ip addr add 127.0.1.0/8 dev lo"),
+		       "add server node"))
+		goto fail_ip;
+
+	if (!ASSERT_OK(system("ip addr add 127.0.2.0/8 dev lo"),
+		       "server via risk path"))
+		goto fail_ip;
+
+	return true;
+fail_ip:
+	netns_free(test_netns);
+fail_netns:
+	return false;
+}
+
+static void cleanup_netns(void)
+{
+	netns_free(test_netns);
+	remove_netns(TEST_NS);
+}
+
+static bool setup_smc(void)
+{
+	if (!setup_ueid())
+		return false;
+
+	if (!setup_netns())
+		goto fail_netns;
+
+	return true;
+fail_netns:
+	cleanup_ueid();
+	return false;
+}
+
+static int set_client_addr_cb(int fd, void *opts)
+{
+	const char *src = (const char *)opts;
+	struct sockaddr_in localaddr;
+
+	localaddr.sin_family = AF_INET;
+	localaddr.sin_port = htons(0);
+	localaddr.sin_addr.s_addr = inet_addr(src);
+	return !ASSERT_OK(bind(fd, &localaddr, sizeof(localaddr)), "client bind");
+}
+
+static void run_link(const char *src, const char *dst, int port)
+{
+	struct network_helper_opts opts = {0};
+	int server, client;
+
+	server = start_server_str(AF_INET, SOCK_STREAM, dst, port, NULL);
+	if (!ASSERT_OK_FD(server, "start service_1"))
+		return;
+
+	opts.proto = IPPROTO_TCP;
+	opts.post_socket_cb = set_client_addr_cb;
+	opts.cb_opts = (void *)src;
+
+	client = connect_to_fd_opts(server, &opts);
+	if (!ASSERT_OK_FD(client, "start connect"))
+		goto fail_client;
+
+	close(client);
+fail_client:
+	close(server);
+}
+
+static void block_link(int map_fd, const char *src, const char *dst)
+{
+	struct smc_strat_ip_value val = { .mode = /* block */ 0 };
+	struct smc_strat_ip_key key = {
+		.sip = inet_addr(src),
+		.dip = inet_addr(dst),
+	};
+
+	bpf_map_update_elem(map_fd, &key, &val, BPF_ANY);
+}
+
+/*
+ * This test describes a real-life service topology as follows:
+ *
+ *                             +-------------> service_1
+ *            link 1           |                     |
+ *   +--------------------> server                   |  link 2
+ *   |                         |                     V
+ *   |                         +-------------> service_2
+ *   |        link 3
+ *  client -------------------> server_via_unsafe_path -> service_3
+ *
+ * Among them,
+ * 1. link-1 is very suitable for using SMC.
+ * 2. link-2 is not suitable for using SMC, because the mode of this link is
+ *    kind of short-link services.
+ * 3. link-3 is also not suitable for using SMC, because the RDMA link is
+ *    unavailable and needs to go through a long timeout before it can fallback
+ *    to TCP.
+ * To achieve this goal, we use a customized SMC ip strategy via smc_hs_ctrl.
+ */
+static void test_topo(void)
+{
+	struct bpf_smc *skel;
+	int rc, map_fd;
+
+	skel = bpf_smc__open_and_load();
+	if (!ASSERT_OK_PTR(skel, "bpf_smc__open_and_load"))
+		return;
+
+	rc = bpf_smc__attach(skel);
+	if (!ASSERT_OK(rc, "bpf_smc__attach"))
+		goto fail;
+
+	map_fd = bpf_map__fd(skel->maps.smc_strats_ip);
+	if (!ASSERT_OK_FD(map_fd, "bpf_map__fd"))
+		goto fail;
+
+	/* Mock the process of transparent replacement, since we will modify
+	 * protocol to ipproto_smc accropding to it via
+	 * fmod_ret/update_socket_protocol.
+	 */
+	system("sysctl -w net.smc.hs_ctrl=linkcheck");
+
+	/* Configure ip strat */
+	block_link(map_fd, CLIENT_IP, SERVER_IP_VIA_RISK_PATH);
+	block_link(map_fd, SERVER_IP, SERVER_IP);
+
+	/* should go with smc */
+	run_link(CLIENT_IP, SERVER_IP, SERVICE_1);
+	/* should go with smc fallback */
+	run_link(SERVER_IP, SERVER_IP, SERVICE_2);
+
+	ASSERT_EQ(skel->bss->smc_cnt, 2, "smc count");
+	ASSERT_EQ(skel->bss->fallback_cnt, 1, "fallback count");
+
+	/* should go with smc */
+	run_link(CLIENT_IP, SERVER_IP, SERVICE_2);
+
+	ASSERT_EQ(skel->bss->smc_cnt, 3, "smc count");
+	ASSERT_EQ(skel->bss->fallback_cnt, 1, "fallback count");
+
+	/* should go with smc fallback */
+	run_link(CLIENT_IP, SERVER_IP_VIA_RISK_PATH, SERVICE_3);
+
+	ASSERT_EQ(skel->bss->smc_cnt, 4, "smc count");
+	ASSERT_EQ(skel->bss->fallback_cnt, 2, "fallback count");
+
+fail:
+	bpf_smc__destroy(skel);
+}
+
+void test_bpf_smc(void)
+{
+	if (!setup_smc()) {
+		printf("setup for smc test failed, test SKIP:\n");
+		test__skip();
+		return;
+	}
+
+	if (test__start_subtest("topo"))
+		test_topo();
+
+	cleanup_ueid();
+	cleanup_netns();
+}
diff --git a/tools/testing/selftests/bpf/progs/bpf_smc.c b/tools/testing/selftests/bpf/progs/bpf_smc.c
new file mode 100644
index 000000000000..d6abc3b9e6c9
--- /dev/null
+++ b/tools/testing/selftests/bpf/progs/bpf_smc.c
@@ -0,0 +1,117 @@
+// SPDX-License-Identifier: GPL-2.0
+
+#include "vmlinux.h"
+
+#include <bpf/bpf_helpers.h>
+#include <bpf/bpf_tracing.h>
+#include "bpf_tracing_net.h"
+
+char _license[] SEC("license") = "GPL";
+
+enum {
+	BPF_SMC_LISTEN	= 10,
+};
+
+struct smc_sock___local {
+	struct sock sk;
+	struct smc_sock *listen_smc;
+	bool use_fallback;
+} __attribute__((preserve_access_index));
+
+int smc_cnt = 0;
+int fallback_cnt = 0;
+
+SEC("fentry/smc_release")
+int BPF_PROG(bpf_smc_release, struct socket *sock)
+{
+	/* only count from one side (client) */
+	if (sock->sk->__sk_common.skc_state == BPF_SMC_LISTEN)
+		return 0;
+	smc_cnt++;
+	return 0;
+}
+
+SEC("fentry/smc_switch_to_fallback")
+int BPF_PROG(bpf_smc_switch_to_fallback, struct smc_sock___local *smc)
+{
+	/* only count from one side (client) */
+	if (smc && !smc->listen_smc)
+		fallback_cnt++;
+	return 0;
+}
+
+/* go with default value if no strat was found */
+bool default_ip_strat_value = true;
+
+struct smc_strat_ip_key {
+	__u32	sip;
+	__u32	dip;
+};
+
+struct smc_strat_ip_value {
+	__u8	mode;
+};
+
+struct {
+	__uint(type, BPF_MAP_TYPE_HASH);
+	__uint(key_size, sizeof(struct smc_strat_ip_key));
+	__uint(value_size, sizeof(struct smc_strat_ip_value));
+	__uint(max_entries, 128);
+	__uint(map_flags, BPF_F_NO_PREALLOC);
+} smc_strats_ip SEC(".maps");
+
+static bool smc_check(__u32 src, __u32 dst)
+{
+	struct smc_strat_ip_value *value;
+	struct smc_strat_ip_key key = {
+		.sip = src,
+		.dip = dst,
+	};
+
+	value = bpf_map_lookup_elem(&smc_strats_ip, &key);
+	return value ? value->mode : default_ip_strat_value;
+}
+
+SEC("fmod_ret/update_socket_protocol")
+int BPF_PROG(smc_run, int family, int type, int protocol)
+{
+	struct task_struct *task;
+
+	if (family != AF_INET && family != AF_INET6)
+		return protocol;
+
+	if ((type & 0xf) != SOCK_STREAM)
+		return protocol;
+
+	if (protocol != 0 && protocol != IPPROTO_TCP)
+		return protocol;
+
+	task = bpf_get_current_task_btf();
+	/* Prevent from affecting other tests */
+	if (!task || !task->nsproxy->net_ns->smc.hs_ctrl)
+		return protocol;
+
+	return IPPROTO_SMC;
+}
+
+SEC("struct_ops/bpf_smc_set_tcp_option_cond")
+int BPF_PROG(bpf_smc_set_tcp_option_cond, const struct tcp_sock *tp,
+	     struct inet_request_sock *ireq)
+{
+	return smc_check(ireq->req.__req_common.skc_daddr,
+			 ireq->req.__req_common.skc_rcv_saddr);
+}
+
+SEC("struct_ops/bpf_smc_set_tcp_option")
+int BPF_PROG(bpf_smc_set_tcp_option, struct tcp_sock *tp)
+{
+	return smc_check(tp->inet_conn.icsk_inet.sk.__sk_common.skc_rcv_saddr,
+			 tp->inet_conn.icsk_inet.sk.__sk_common.skc_daddr);
+}
+
+SEC(".struct_ops.link")
+struct smc_hs_ctrl  linkcheck = {
+	.name		= "linkcheck",
+	.syn_option	= (void *)bpf_smc_set_tcp_option,
+	.synack_option	= (void *)bpf_smc_set_tcp_option_cond,
+};
-- 
2.45.0