[PATCH bpf 0/2] tcp_bpf: Fix the sk_mem_uncharge logic in tcp_bpf_sendmsg

[PATCH bpf 0/2] tcp_bpf: Fix the sk_mem_uncharge logic in tcp_bpf_sendmsg

[zijianzhang]

From: Zijian Zhang <zijianzhang@bytedance.com>

When apply_bytes are not zero, sk_mem_uncharge for __SK_REDIRECT and
__SK_DROP in tcp_bpf_sendmsg has some problem. Added a selftest to trigger
the memory accounting WARNING, and fixed the sk_mem_uncharge logic in
tcp_bpf_sendmsg

Zijian Zhang (2):
  selftests/bpf: Add apply_bytes test to test_txmsg_redir_wait_sndmem in
    test_sockmap
  tcp_bpf: Fix the sk_mem_uncharge logic in tcp_bpf_sendmsg

 net/ipv4/tcp_bpf.c                         | 11 ++++-------
 tools/testing/selftests/bpf/test_sockmap.c |  6 +++++-
 2 files changed, 9 insertions(+), 8 deletions(-)

-- 
2.20.1

[PATCH bpf 1/2] selftests/bpf: Add apply_bytes test to test_txmsg_redir_wait_sndmem in test_sockmap

[zijianzhang]

From: Zijian Zhang <zijianzhang@bytedance.com>

Add this to more comprehensively test the socket memory accounting logic
in the __SK_REDIRECT and __SK_DROP cases of tcp_bpf_sendmsg. We don't have
test when apply_bytes are not zero in test_txmsg_redir_wait_sndmem.
test_send_large has opt->rate=2, it will invoke sendmsg two times.
Specifically, the first sendmsg will trigger the case where the ret value
of tcp_bpf_sendmsg_redir is less than 0; while the second sendmsg happens
after the 3 seconds timeout, and it will trigger __SK_DROP because socket
c2 has been removed from the sockmap/hash.

Signed-off-by: Zijian Zhang <zijianzhang@bytedance.com>
---
 tools/testing/selftests/bpf/test_sockmap.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)

diff --git a/tools/testing/selftests/bpf/test_sockmap.c b/tools/testing/selftests/bpf/test_sockmap.c
index 3e02d7267de8..446f7cca56dc 100644
--- a/tools/testing/selftests/bpf/test_sockmap.c
+++ b/tools/testing/selftests/bpf/test_sockmap.c
@@ -1485,8 +1485,12 @@ static void test_txmsg_redir(int cgrp, struct sockmap_options *opt)
 
 static void test_txmsg_redir_wait_sndmem(int cgrp, struct sockmap_options *opt)
 {
-	txmsg_redir = 1;
 	opt->tx_wait_mem = true;
+	txmsg_redir = 1;
+	test_send_large(opt, cgrp);
+
+	txmsg_redir = 1;
+	txmsg_apply = 4097;
 	test_send_large(opt, cgrp);
 	opt->tx_wait_mem = false;
 }
-- 
2.20.1

[PATCH bpf 2/2] tcp_bpf: Fix the sk_mem_uncharge logic in tcp_bpf_sendmsg

[zijianzhang]

From: Zijian Zhang <zijianzhang@bytedance.com>

The current sk memory accounting logic in __SK_REDIRECT is pre-uncharging
tosend bytes, which is either msg->sg.size or a smaller value apply_bytes.
Potential problems with this strategy are as follows:
- If the actual sent bytes are smaller than tosend, we need to charge some
bytes back, as in line 487, which is okay but seems not clean.
- When tosend is set to apply_bytes, as in line 417, and (ret < 0), we may
miss uncharging (msg->sg.size - apply_bytes) bytes.

415 tosend = msg->sg.size;
416 if (psock->apply_bytes && psock->apply_bytes < tosend)
417   tosend = psock->apply_bytes;
...
443 sk_msg_return(sk, msg, tosend);
444 release_sock(sk);
446 origsize = msg->sg.size;
447 ret = tcp_bpf_sendmsg_redir(sk_redir, redir_ingress,
448                             msg, tosend, flags);
449 sent = origsize - msg->sg.size;
...
454 lock_sock(sk);
455 if (unlikely(ret < 0)) {
456   int free = sk_msg_free_nocharge(sk, msg);
458   if (!cork)
459     *copied -= free;
460 }
...
487 if (eval == __SK_REDIRECT)
488   sk_mem_charge(sk, tosend - sent);

When running the selftest test_txmsg_redir_wait_sndmem with txmsg_apply,
the following warning will be reported,
------------[ cut here ]------------
WARNING: CPU: 6 PID: 57 at net/ipv4/af_inet.c:156 inet_sock_destruct+0x190/0x1a0
Modules linked in:
CPU: 6 UID: 0 PID: 57 Comm: kworker/6:0 Not tainted 6.12.0-rc1.bm.1-amd64+ #43
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.12.0-1 04/01/2014
Workqueue: events sk_psock_destroy
RIP: 0010:inet_sock_destruct+0x190/0x1a0
RSP: 0018:ffffad0a8021fe08 EFLAGS: 00010206
RAX: 0000000000000011 RBX: ffff9aab4475b900 RCX: ffff9aab481a0800
RDX: 0000000000000303 RSI: 0000000000000011 RDI: ffff9aab4475b900
RBP: ffff9aab4475b990 R08: 0000000000000000 R09: ffff9aab40050ec0
R10: 0000000000000000 R11: ffff9aae6fdb1d01 R12: ffff9aab49c60400
R13: ffff9aab49c60598 R14: ffff9aab49c60598 R15: dead000000000100
FS:  0000000000000000(0000) GS:ffff9aae6fd80000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00007ffec7e47bd8 CR3: 00000001a1a1c004 CR4: 0000000000770ef0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
PKRU: 55555554
Call Trace:
<TASK>
? __warn+0x89/0x130
? inet_sock_destruct+0x190/0x1a0
? report_bug+0xfc/0x1e0
? handle_bug+0x5c/0xa0
? exc_invalid_op+0x17/0x70
? asm_exc_invalid_op+0x1a/0x20
? inet_sock_destruct+0x190/0x1a0
__sk_destruct+0x25/0x220
sk_psock_destroy+0x2b2/0x310
process_scheduled_works+0xa3/0x3e0
worker_thread+0x117/0x240
? __pfx_worker_thread+0x10/0x10
kthread+0xcf/0x100
? __pfx_kthread+0x10/0x10
ret_from_fork+0x31/0x40
? __pfx_kthread+0x10/0x10
ret_from_fork_asm+0x1a/0x30
</TASK>
---[ end trace 0000000000000000 ]---

In __SK_REDIRECT, a more concise way is delaying the uncharging after sent
bytes are finalized, and uncharge this value. When (ret < 0), we shall
invoke sk_msg_free.

Same thing happens in case __SK_DROP, when tosend is set to apply_bytes,
we may miss uncharging (msg->sg.size - apply_bytes) bytes. The same
warning will be reported in selftest.

468 case __SK_DROP:
469 default:
470 sk_msg_free_partial(sk, msg, tosend);
471 sk_msg_apply_bytes(psock, tosend);
472 *copied -= (tosend + delta);
473 return -EACCES;

So instead of sk_msg_free_partial we can do sk_msg_free here.

Fixes: 604326b41a6f ("bpf, sockmap: convert to generic sk_msg interface")
Fixes: 8ec95b94716a ("bpf, sockmap: Fix the sk->sk_forward_alloc warning of sk_stream_kill_queues")
Signed-off-by: Zijian Zhang <zijianzhang@bytedance.com>
---
 net/ipv4/tcp_bpf.c | 11 ++++-------
 1 file changed, 4 insertions(+), 7 deletions(-)

diff --git a/net/ipv4/tcp_bpf.c b/net/ipv4/tcp_bpf.c
index e7658c5d6b79..7b49bf0afbac 100644
--- a/net/ipv4/tcp_bpf.c
+++ b/net/ipv4/tcp_bpf.c
@@ -440,7 +440,6 @@ static int tcp_bpf_send_verdict(struct sock *sk, struct sk_psock *psock,
 			cork = true;
 			psock->cork = NULL;
 		}
-		sk_msg_return(sk, msg, tosend);
 		release_sock(sk);
 
 		origsize = msg->sg.size;
@@ -452,8 +451,9 @@ static int tcp_bpf_send_verdict(struct sock *sk, struct sk_psock *psock,
 			sock_put(sk_redir);
 
 		lock_sock(sk);
+		sk_mem_uncharge(sk, sent);
 		if (unlikely(ret < 0)) {
-			int free = sk_msg_free_nocharge(sk, msg);
+			int free = sk_msg_free(sk, msg);
 
 			if (!cork)
 				*copied -= free;
@@ -467,7 +467,7 @@ static int tcp_bpf_send_verdict(struct sock *sk, struct sk_psock *psock,
 		break;
 	case __SK_DROP:
 	default:
-		sk_msg_free_partial(sk, msg, tosend);
+		sk_msg_free(sk, msg);
 		sk_msg_apply_bytes(psock, tosend);
 		*copied -= (tosend + delta);
 		return -EACCES;
@@ -483,11 +483,8 @@ static int tcp_bpf_send_verdict(struct sock *sk, struct sk_psock *psock,
 		}
 		if (msg &&
 		    msg->sg.data[msg->sg.start].page_link &&
-		    msg->sg.data[msg->sg.start].length) {
-			if (eval == __SK_REDIRECT)
-				sk_mem_charge(sk, tosend - sent);
+		    msg->sg.data[msg->sg.start].length)
 			goto more_data;
-		}
 	}
 	return ret;
 }
-- 
2.20.1

RE: [PATCH bpf 1/2] selftests/bpf: Add apply_bytes test to test_txmsg_redir_wait_sndmem in test_sockmap

[John Fastabend]

zijianzhang@ wrote:
> From: Zijian Zhang <zijianzhang@bytedance.com>
> 
> Add this to more comprehensively test the socket memory accounting logic
> in the __SK_REDIRECT and __SK_DROP cases of tcp_bpf_sendmsg. We don't have
> test when apply_bytes are not zero in test_txmsg_redir_wait_sndmem.
> test_send_large has opt->rate=2, it will invoke sendmsg two times.
> Specifically, the first sendmsg will trigger the case where the ret value
> of tcp_bpf_sendmsg_redir is less than 0; while the second sendmsg happens
> after the 3 seconds timeout, and it will trigger __SK_DROP because socket
> c2 has been removed from the sockmap/hash.
> 
> Signed-off-by: Zijian Zhang <zijianzhang@bytedance.com>
> ---

Thanks, sorry for the rather long delay.

Acked-by: John Fastabend <john.fastabend@gmail.com>

RE: [PATCH bpf 2/2] tcp_bpf: Fix the sk_mem_uncharge logic in tcp_bpf_sendmsg

[John Fastabend]

zijianzhang@ wrote:
> From: Zijian Zhang <zijianzhang@bytedance.com>
> 
> The current sk memory accounting logic in __SK_REDIRECT is pre-uncharging
> tosend bytes, which is either msg->sg.size or a smaller value apply_bytes.
> Potential problems with this strategy are as follows:
> - If the actual sent bytes are smaller than tosend, we need to charge some
> bytes back, as in line 487, which is okay but seems not clean.
> - When tosend is set to apply_bytes, as in line 417, and (ret < 0), we may
> miss uncharging (msg->sg.size - apply_bytes) bytes.
> 
> 415 tosend = msg->sg.size;
> 416 if (psock->apply_bytes && psock->apply_bytes < tosend)
> 417   tosend = psock->apply_bytes;
> ...
> 443 sk_msg_return(sk, msg, tosend);
> 444 release_sock(sk);
> 446 origsize = msg->sg.size;
> 447 ret = tcp_bpf_sendmsg_redir(sk_redir, redir_ingress,
> 448                             msg, tosend, flags);
> 449 sent = origsize - msg->sg.size;
> ...
> 454 lock_sock(sk);
> 455 if (unlikely(ret < 0)) {
> 456   int free = sk_msg_free_nocharge(sk, msg);
> 458   if (!cork)
> 459     *copied -= free;
> 460 }
> ...
> 487 if (eval == __SK_REDIRECT)
> 488   sk_mem_charge(sk, tosend - sent);
> 
> When running the selftest test_txmsg_redir_wait_sndmem with txmsg_apply,
> the following warning will be reported,
> ------------[ cut here ]------------
> WARNING: CPU: 6 PID: 57 at net/ipv4/af_inet.c:156 inet_sock_destruct+0x190/0x1a0
> Modules linked in:
> CPU: 6 UID: 0 PID: 57 Comm: kworker/6:0 Not tainted 6.12.0-rc1.bm.1-amd64+ #43
> Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.12.0-1 04/01/2014
> Workqueue: events sk_psock_destroy
> RIP: 0010:inet_sock_destruct+0x190/0x1a0
> RSP: 0018:ffffad0a8021fe08 EFLAGS: 00010206
> RAX: 0000000000000011 RBX: ffff9aab4475b900 RCX: ffff9aab481a0800
> RDX: 0000000000000303 RSI: 0000000000000011 RDI: ffff9aab4475b900
> RBP: ffff9aab4475b990 R08: 0000000000000000 R09: ffff9aab40050ec0
> R10: 0000000000000000 R11: ffff9aae6fdb1d01 R12: ffff9aab49c60400
> R13: ffff9aab49c60598 R14: ffff9aab49c60598 R15: dead000000000100
> FS:  0000000000000000(0000) GS:ffff9aae6fd80000(0000) knlGS:0000000000000000
> CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> CR2: 00007ffec7e47bd8 CR3: 00000001a1a1c004 CR4: 0000000000770ef0
> DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
> DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
> PKRU: 55555554
> Call Trace:
> <TASK>
> ? __warn+0x89/0x130
> ? inet_sock_destruct+0x190/0x1a0
> ? report_bug+0xfc/0x1e0
> ? handle_bug+0x5c/0xa0
> ? exc_invalid_op+0x17/0x70
> ? asm_exc_invalid_op+0x1a/0x20
> ? inet_sock_destruct+0x190/0x1a0
> __sk_destruct+0x25/0x220
> sk_psock_destroy+0x2b2/0x310
> process_scheduled_works+0xa3/0x3e0
> worker_thread+0x117/0x240
> ? __pfx_worker_thread+0x10/0x10
> kthread+0xcf/0x100
> ? __pfx_kthread+0x10/0x10
> ret_from_fork+0x31/0x40
> ? __pfx_kthread+0x10/0x10
> ret_from_fork_asm+0x1a/0x30
> </TASK>
> ---[ end trace 0000000000000000 ]---
> 
> In __SK_REDIRECT, a more concise way is delaying the uncharging after sent
> bytes are finalized, and uncharge this value. When (ret < 0), we shall
> invoke sk_msg_free.
> 
> Same thing happens in case __SK_DROP, when tosend is set to apply_bytes,
> we may miss uncharging (msg->sg.size - apply_bytes) bytes. The same
> warning will be reported in selftest.
> 
> 468 case __SK_DROP:
> 469 default:
> 470 sk_msg_free_partial(sk, msg, tosend);
> 471 sk_msg_apply_bytes(psock, tosend);
> 472 *copied -= (tosend + delta);
> 473 return -EACCES;
> 
> So instead of sk_msg_free_partial we can do sk_msg_free here.
> 
> Fixes: 604326b41a6f ("bpf, sockmap: convert to generic sk_msg interface")
> Fixes: 8ec95b94716a ("bpf, sockmap: Fix the sk->sk_forward_alloc warning of sk_stream_kill_queues")
> Signed-off-by: Zijian Zhang <zijianzhang@bytedance.com>
> ---

Agree thanks.

Acked-by: John Fastabend <john.fastabend@gmail.com>

RE: [PATCH bpf 0/2] tcp_bpf: Fix the sk_mem_uncharge logic in tcp_bpf_sendmsg

[John Fastabend]

zijianzhang@ wrote:
> From: Zijian Zhang <zijianzhang@bytedance.com>
> 
> When apply_bytes are not zero, sk_mem_uncharge for __SK_REDIRECT and
> __SK_DROP in tcp_bpf_sendmsg has some problem. Added a selftest to trigger
> the memory accounting WARNING, and fixed the sk_mem_uncharge logic in
> tcp_bpf_sendmsg
> 
> Zijian Zhang (2):
>   selftests/bpf: Add apply_bytes test to test_txmsg_redir_wait_sndmem in
>     test_sockmap
>   tcp_bpf: Fix the sk_mem_uncharge logic in tcp_bpf_sendmsg

I would probably prefer the patches with the fix first than the selftest
just to avoid tripping up any bisect.

But patches look good. Thanks and sorry for the delay again I was
travelling and OOO for a bit.

> 
>  net/ipv4/tcp_bpf.c                         | 11 ++++-------
>  tools/testing/selftests/bpf/test_sockmap.c |  6 +++++-
>  2 files changed, 9 insertions(+), 8 deletions(-)
> 
> -- 
> 2.20.1
> 

Re: [PATCH bpf 0/2] tcp_bpf: Fix the sk_mem_uncharge logic in tcp_bpf_sendmsg

[patchwork-bot+netdevbpf]

Hello:

This series was applied to bpf/bpf.git (master)
by Daniel Borkmann <daniel@iogearbox.net>:

On Wed, 16 Oct 2024 23:48:36 +0000 you wrote:
> From: Zijian Zhang <zijianzhang@bytedance.com>
> 
> When apply_bytes are not zero, sk_mem_uncharge for __SK_REDIRECT and
> __SK_DROP in tcp_bpf_sendmsg has some problem. Added a selftest to trigger
> the memory accounting WARNING, and fixed the sk_mem_uncharge logic in
> tcp_bpf_sendmsg
> 
> [...]

Here is the summary with links:
  - [bpf,1/2] selftests/bpf: Add apply_bytes test to test_txmsg_redir_wait_sndmem in test_sockmap
    https://git.kernel.org/bpf/bpf/c/3448ad23b34e
  - [bpf,2/2] tcp_bpf: Fix the sk_mem_uncharge logic in tcp_bpf_sendmsg
    https://git.kernel.org/bpf/bpf/c/ca70b8baf2bd

You are awesome, thank you!
-- 
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html