Message-ID: <61de09de-b988-3097-05a8-fd6053b9288a@gmail.com>
Date:   Thu, 4 Aug 2022 11:05:18 +0300
From:   Tariq Toukan <ttoukan.linux@...il.com>
To:     Jakub Kicinski <kuba@...nel.org>
Cc:     netdev@...r.kernel.org, edumazet@...gle.com, pabeni@...hat.com,
        borisp@...dia.com, john.fastabend@...il.com, maximmi@...dia.com,
        tariqt@...dia.com, vfedorenko@...ek.ru,
        Ran Rozenstein <ranro@...dia.com>,
        "gal@...dia.com" <gal@...dia.com>,
        "David S. Miller" <davem@...emloft.net>
Subject: Re: [PATCH net-next v3 7/7] tls: rx: do not use the standard
 strparser



On 8/4/2022 4:24 AM, Jakub Kicinski wrote:
> On Tue, 2 Aug 2022 17:54:01 +0300 Tariq Toukan wrote:
>>    [  407.589886] RIP: 0010:tls_device_decrypted+0x7a/0x2e0
> 
> Sorry, got distracted yesterday. This?
> 
> --->8--------------------
> tls: rx: device: bound the frag walk
> 
> We can't do skb_walk_frags() on the input skbs, because
> the input skbs is really just a pointer to the tcp read
> queue. We need to bound the "is decrypted" check by the
> amount of data in the message.
> 
> Note that the walk in tls_device_reencrypt() is after a
> CoW so the skb there is safe to walk. Actually in the
> current implementation it can't have frags at all, but
> whatever, maybe one day it will.
> 
> Reported-by: Tariq Toukan <tariqt@...dia.com>
> Fixes: 84c61fe1a75b ("tls: rx: do not use the standard strparser")
> Signed-off-by: Jakub Kicinski <kuba@...nel.org>
> ---
>   net/tls/tls_device.c | 8 +++++++-
>   1 file changed, 7 insertions(+), 1 deletion(-)
> 
> diff --git a/net/tls/tls_device.c b/net/tls/tls_device.c
> index e3e6cf75aa03..6ed41474bdf8 100644
> --- a/net/tls/tls_device.c
> +++ b/net/tls/tls_device.c
> @@ -984,11 +984,17 @@ int tls_device_decrypted(struct sock *sk, struct tls_context *tls_ctx)
>   	int is_decrypted = skb->decrypted;
>   	int is_encrypted = !is_decrypted;
>   	struct sk_buff *skb_iter;
> +	int left;
>   
> +	left = rxm->full_len - skb->len;
>   	/* Check if all the data is decrypted already */
> -	skb_walk_frags(skb, skb_iter) {
> +	skb_iter = skb_shinfo(skb)->frag_list;
> +	while (skb_iter && left > 0) {
>   		is_decrypted &= skb_iter->decrypted;
>   		is_encrypted &= !skb_iter->decrypted;
> +
> +		left -= skb_iter->len;
> +		skb_iter = skb_iter->next;
>   	}
>   
>   	trace_tls_device_decrypted(sk, tcp_sk(sk)->copied_seq - rxm->full_len,

Now we see a different trace:

------------[ cut here ]------------
WARNING: CPU: 4 PID: 45887 at net/tls/tls_strp.c:53 
tls_strp_msg_make_copy+0x10e/0x120
Modules linked in: sch_netem iptable_raw bonding rdma_ucm nf_tables 
ib_ipoib ib_umad ip_gre ip6_gre gre ip6_tunnel tunnel6 ipip tunnel4 
mlx5_vfio_pci vfio_pci_core vfio_virqfd vfio_iommu_type1 vfio geneve 
mlx5_ib ib_uverbs mlx5_core openvswitch nsh xt_conntrack xt_MASQUERADE 
nf_conntrack_netlink nfnetlink xt_addrtype iptable_nat nf_nat 
br_netfilter rpcrdma ib_iser libiscsi scsi_transport_iscsi rdma_cm iw_cm 
ib_cm overlay ib_core fuse [last unloaded: ib_uverbs]
CPU: 4 PID: 45887 Comm: iperf Not tainted 
5.19.0-rc8_net_next_mlx5_18607aa #1
Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 
rel-1.13.0-0-gf21b5a4aeb02-prebuilt.qemu.org 04/01/2014
RIP: 0010:tls_strp_msg_make_copy+0x10e/0x120
Code: 41 c7 47 44 00 00 00 00 48 8b 44 24 08 65 48 2b 04 25 28 00 00 00 
75 16 48 83 c4 10 4c 89 f8 5b 5d 41 5c 41 5d 41 5e 41 5f c3 <0f> 0b eb 
a5 e8 79 0a 28 00 66 0f 1f 84 00 00 00 00 00 0f 1f 44 00
RSP: 0018:ffff88810574fb28 EFLAGS: 00010282
RAX: 00000000fffffff2 RBX: ffff8881136b8cd0 RCX: 0000000000000872
RDX: ffff8881c4bb6000 RSI: 0000000000004855 RDI: ffff88810c4ab500
RBP: 0000000000004855 R08: 00000000000042f0 R09: ffff88810c4aa000
R10: 0000000000000001 R11: 0000000000001000 R12: 0000160000000000
R13: 0000000000000001 R14: ffff8881ada4f8d0 R15: ffff88810c4aa000
FS:  00007f5efe4fe700(0000) GS:ffff88846fa00000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00007f5ef4004000 CR3: 00000001ee214005 CR4: 0000000000370ea0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
Call Trace:
  <TASK>
  tls_rx_one_record+0x1f2/0x310
  tls_sw_recvmsg+0x327/0x8d0
  inet6_recvmsg+0x62/0x220
  ____sys_recvmsg+0x109/0x120
  ? _copy_from_user+0x44/0x80
  ? iovec_from_user+0x4a/0x150
  ___sys_recvmsg+0xa4/0xe0
  ? find_held_lock+0x2b/0x80
  ? __fget_files+0xb9/0x190
  ? __fget_files+0xd3/0x190
  __sys_recvmsg+0x4e/0x90
  do_syscall_64+0x3d/0x90
  entry_SYSCALL_64_after_hwframe+0x46/0xb0
RIP: 0033:0x7f5effe086dd
Code: 28 89 54 24 1c 48 89 74 24 10 89 7c 24 08 e8 5a ef ff ff 8b 54 24 
1c 48 8b 74 24 10 41 89 c0 8b 7c 24 08 b8 2f 00 00 00 0f 05 <48> 3d 00 
f0 ff ff 77 33 44 89 c7 48 89 44 24 08 e8 8e ef ff ff 48
RSP: 002b:00007f5efe4fda70 EFLAGS: 00000293 ORIG_RAX: 000000000000002f
RAX: ffffffffffffffda RBX: 00007f5ef4022be0 RCX: 00007f5effe086dd
RDX: 0000000000000000 RSI: 00007f5efe4fdad0 RDI: 0000000000000004
RBP: 0000000000004130 R08: 0000000000000000 R09: 00007f5efe4fdc38
R10: 0000000000000000 R11: 0000000000000293 R12: 00007f5ef40284e3
R13: 00007f5efe4fe660 R14: 00007f5efe4fdb58 R15: 00007f5ef4020ba0
  </TASK>
irq event stamp: 2999
hardirqs last  enabled at (3017): [<ffffffff811daf27>] 
__up_console_sem+0x67/0x70
hardirqs last disabled at (3030): [<ffffffff811daf0c>] 
__up_console_sem+0x4c/0x70
softirqs last  enabled at (2680): [<ffffffff81170d47>] 
irq_exit_rcu+0x97/0xd0
softirqs last disabled at (2671): [<ffffffff81170d47>] 
irq_exit_rcu+0x97/0xd0
---[ end trace 0000000000000000 ]---

Powered by blists - more mailing lists