net: Convert to skb_dstref_steal and skb_dstref_restore

[PATCH net-next v2 0/7] net: Convert to skb_dstref_steal and skb_dstref_restore

Posted by Stanislav Fomichev 1 month, 2 weeks ago

To diagnose and prevent issues similar to [0], emit warning
(CONFIG_DEBUG_NET) from skb_dst_set and skb_dst_set_noref when
overwriting non-null reference-counted entry. Two new helpers
are added to handle special cases where the entry needs to be
reset and restored: skb_dstref_steal/skb_dstref_restore. The bulk of
the patches in the series converts manual _skb_refst manipulations
to these new helpers.

0: https://lore.kernel.org/netdev/20250723224625.1340224-1-sdf@fomichev.me/T/#u

v2:
- rename to skb_dstref_steal/skb_dstref_restore (Jakub)
- fix kdoc (Jakub)

Stanislav Fomichev (7):
  net: Add skb_dstref_steal and skb_dstref_restore
  xfrm: Switch to skb_dstref_steal to clear dst_entry
  netfilter: Switch to skb_dstref_steal to clear dst_entry
  net: Switch to skb_dstref_steal/skb_dstref_restore for ip_route_input
    callers
  staging: octeon: Convert to skb_dst_drop
  chtls: Convert to skb_dst_reset
  net: Add skb_dst_check_unset

 .../chelsio/inline_crypto/chtls/chtls_cm.c    | 10 ++---
 .../chelsio/inline_crypto/chtls/chtls_cm.h    |  4 +-
 .../chelsio/inline_crypto/chtls/chtls_io.c    |  2 +-
 drivers/staging/octeon/ethernet-tx.c          |  3 +-
 include/linux/skbuff.h                        | 41 +++++++++++++++++++
 net/ipv4/icmp.c                               |  7 ++--
 net/ipv4/ip_options.c                         |  5 +--
 net/ipv4/netfilter.c                          |  5 ++-
 net/ipv6/netfilter.c                          |  5 ++-
 net/xfrm/xfrm_policy.c                        | 10 ++++-
 10 files changed, 72 insertions(+), 20 deletions(-)

-- 
2.50.1

[syzbot ci] Re: net: Convert to skb_dstref_steal and skb_dstref_restore

Posted by syzbot ci 1 month, 2 weeks ago

syzbot ci has tested the following series

[v2] net: Convert to skb_dstref_steal and skb_dstref_restore
https://lore.kernel.org/all/20250818154032.3173645-1-sdf@fomichev.me
* [PATCH net-next v2 1/7] net: Add skb_dstref_steal and skb_dstref_restore
* [PATCH net-next v2 2/7] xfrm: Switch to skb_dstref_steal to clear dst_entry
* [PATCH net-next v2 3/7] netfilter: Switch to skb_dstref_steal to clear dst_entry
* [PATCH net-next v2 4/7] net: Switch to skb_dstref_steal/skb_dstref_restore for ip_route_input callers
* [PATCH net-next v2 5/7] staging: octeon: Convert to skb_dst_drop
* [PATCH net-next v2 6/7] chtls: Convert to skb_dst_reset
* [PATCH net-next v2 7/7] net: Add skb_dst_check_unset

and found the following issue:
WARNING in nf_reject_fill_skb_dst

Full report is available here:
https://ci.syzbot.org/series/74fec874-bca1-42a5-bb58-f3f49b95e348

***

WARNING in nf_reject_fill_skb_dst

tree:      net-next
URL:       https://kernel.googlesource.com/pub/scm/linux/kernel/git/netdev/net-next.git
base:      8159572936392b514e404bab2d60e47cebd5acfe
arch:      amd64
compiler:  Debian clang version 20.1.7 (++20250616065708+6146a88f6049-1~exp1~20250616065826.132), Debian LLD 20.1.7
config:    https://ci.syzbot.org/builds/10161afb-c9c1-4272-851c-1ded15995879/config
C repro:   https://ci.syzbot.org/findings/7f681f6a-9952-49a0-8533-0bc185291f81/c_repro
syz repro: https://ci.syzbot.org/findings/7f681f6a-9952-49a0-8533-0bc185291f81/syz_repro

------------[ cut here ]------------
WARNING: CPU: 1 PID: 5862 at ./include/linux/skbuff.h:1165 skb_dst_check_unset include/linux/skbuff.h:1164 [inline]
WARNING: CPU: 1 PID: 5862 at ./include/linux/skbuff.h:1165 skb_dst_set include/linux/skbuff.h:1211 [inline]
WARNING: CPU: 1 PID: 5862 at ./include/linux/skbuff.h:1165 nf_reject_fill_skb_dst+0x2a4/0x330 net/ipv4/netfilter/nf_reject_ipv4.c:234
Modules linked in:
CPU: 1 UID: 0 PID: 5862 Comm: kworker/u8:2 Not tainted 6.17.0-rc1-syzkaller-00207-g815957293639-dirty #0 PREEMPT(full) 
Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.16.2-debian-1.16.2-1 04/01/2014
Workqueue: ipv6_addrconf addrconf_dad_work
RIP: 0010:skb_dst_check_unset include/linux/skbuff.h:1164 [inline]
RIP: 0010:skb_dst_set include/linux/skbuff.h:1211 [inline]
RIP: 0010:nf_reject_fill_skb_dst+0x2a4/0x330 net/ipv4/netfilter/nf_reject_ipv4.c:234
Code: 8b 0d 60 75 8b 08 48 3b 8c 24 e0 00 00 00 75 5d 48 8d 65 d8 5b 41 5c 41 5d 41 5e 41 5f 5d e9 03 91 67 01 cc e8 ad d0 aa f7 90 <0f> 0b 90 e9 38 ff ff ff 44 89 f9 80 e1 07 fe c1 38 c1 0f 8c 2b fe
RSP: 0018:ffffc900001e0360 EFLAGS: 00010246
RAX: ffffffff8a14dae3 RBX: ffff88810f943b00 RCX: ffff888109618000
RDX: 0000000000000100 RSI: 0000000000000000 RDI: 0000000000000000
RBP: ffffc900001e0490 R08: ffffffff8fa37e37 R09: 1ffffffff1f46fc6
R10: dffffc0000000000 R11: fffffbfff1f46fc7 R12: ffff88810ec56101
R13: dffffc0000000001 R14: 1ffff9200003c070 R15: 0000000000000000
FS:  0000000000000000(0000) GS:ffff8881a3c1b000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 0000000002f6a520 CR3: 0000000027716000 CR4: 00000000000006f0
Call Trace:
 <IRQ>
 nf_send_unreach+0x17b/0x6e0 net/ipv4/netfilter/nf_reject_ipv4.c:325
 nft_reject_inet_eval+0x4bc/0x690 net/netfilter/nft_reject_inet.c:27
 expr_call_ops_eval net/netfilter/nf_tables_core.c:237 [inline]
 nft_do_chain+0x40c/0x1920 net/netfilter/nf_tables_core.c:285
 nft_do_chain_inet+0x25d/0x340 net/netfilter/nft_chain_filter.c:161
 nf_hook_entry_hookfn include/linux/netfilter.h:158 [inline]
 nf_hook_slow+0xc5/0x220 net/netfilter/core.c:623
 nf_hook include/linux/netfilter.h:273 [inline]
 NF_HOOK+0x206/0x3a0 include/linux/netfilter.h:316
 __netif_receive_skb_one_core net/core/dev.c:5979 [inline]
 __netif_receive_skb+0x143/0x380 net/core/dev.c:6092
 process_backlog+0x60e/0x14f0 net/core/dev.c:6444
 __napi_poll+0xc7/0x360 net/core/dev.c:7494
 napi_poll net/core/dev.c:7557 [inline]
 net_rx_action+0x707/0xe30 net/core/dev.c:7684
 handle_softirqs+0x286/0x870 kernel/softirq.c:579
 do_softirq+0xec/0x180 kernel/softirq.c:480
 </IRQ>
 <TASK>
 __local_bh_enable_ip+0x17d/0x1c0 kernel/softirq.c:407
 local_bh_enable include/linux/bottom_half.h:33 [inline]
 rcu_read_unlock_bh include/linux/rcupdate.h:910 [inline]
 __dev_queue_xmit+0x1d79/0x3b50 net/core/dev.c:4740
 neigh_output include/net/neighbour.h:547 [inline]
 ip6_finish_output2+0x11fe/0x16a0 net/ipv6/ip6_output.c:141
 NF_HOOK include/linux/netfilter.h:318 [inline]
 ndisc_send_skb+0xb96/0x1470 net/ipv6/ndisc.c:512
 ndisc_send_ns+0xcb/0x150 net/ipv6/ndisc.c:670
 addrconf_dad_work+0xaae/0x14b0 net/ipv6/addrconf.c:4282
 process_one_work kernel/workqueue.c:3236 [inline]
 process_scheduled_works+0xae1/0x17b0 kernel/workqueue.c:3319
 worker_thread+0x8a0/0xda0 kernel/workqueue.c:3400
 kthread+0x711/0x8a0 kernel/kthread.c:463
 ret_from_fork+0x3fc/0x770 arch/x86/kernel/process.c:148
 ret_from_fork_asm+0x1a/0x30 arch/x86/entry/entry_64.S:245
 </TASK>


***

If these findings have caused you to resend the series or submit a
separate fix, please add the following tag to your commit message:
Tested-by: syzbot@syzkaller.appspotmail.com

---
This report is generated by a bot. It may contain errors.
syzbot ci engineers can be reached at syzkaller@googlegroups.com.

Re: [syzbot ci] Re: net: Convert to skb_dstref_steal and skb_dstref_restore

Posted by Jakub Kicinski 1 month, 2 weeks ago

On Tue, 19 Aug 2025 08:41:36 -0700 syzbot ci wrote:
> syzbot ci has tested the following series
> 
> [v2] net: Convert to skb_dstref_steal and skb_dstref_restore
> https://lore.kernel.org/all/20250818154032.3173645-1-sdf@fomichev.me
> * [PATCH net-next v2 1/7] net: Add skb_dstref_steal and skb_dstref_restore
> * [PATCH net-next v2 2/7] xfrm: Switch to skb_dstref_steal to clear dst_entry
> * [PATCH net-next v2 3/7] netfilter: Switch to skb_dstref_steal to clear dst_entry
> * [PATCH net-next v2 4/7] net: Switch to skb_dstref_steal/skb_dstref_restore for ip_route_input callers
> * [PATCH net-next v2 5/7] staging: octeon: Convert to skb_dst_drop
> * [PATCH net-next v2 6/7] chtls: Convert to skb_dst_reset
> * [PATCH net-next v2 7/7] net: Add skb_dst_check_unset

> ***
> 
> If these findings have caused you to resend the series or submit a
> separate fix, please add the following tag to your commit message:
> Tested-by: syzbot@syzkaller.appspotmail.com

Hi Aleksandr!

Could we do something about this Tested-by: tag?
Since the syzbot CI reports are sent in reply to a series patchwork and
other tooling will think that syzbot is sending it's Tested-by tag for
this series.

In some cases we know that the issues found are unrelated, or rather
expect them to be fixed separately.

Could we perhaps indent the tag with a couple of spaces? Not 100% sure
but I _think_ most tools will match the tags only from start of line.

Re: [syzbot ci] Re: net: Convert to skb_dstref_steal and skb_dstref_restore

Posted by Aleksandr Nogikh 1 month, 2 weeks ago

Hi Jakub,

Thanks for the heads-up!
I didn't even think that this tag could pose problems for other
automatic tools :)

On Wed, Aug 20, 2025 at 2:58 AM Jakub Kicinski <kuba@kernel.org> wrote:
>
> On Tue, 19 Aug 2025 08:41:36 -0700 syzbot ci wrote:
> > syzbot ci has tested the following series
> >
> > [v2] net: Convert to skb_dstref_steal and skb_dstref_restore
> > https://lore.kernel.org/all/20250818154032.3173645-1-sdf@fomichev.me
> > * [PATCH net-next v2 1/7] net: Add skb_dstref_steal and skb_dstref_restore
> > * [PATCH net-next v2 2/7] xfrm: Switch to skb_dstref_steal to clear dst_entry
> > * [PATCH net-next v2 3/7] netfilter: Switch to skb_dstref_steal to clear dst_entry
> > * [PATCH net-next v2 4/7] net: Switch to skb_dstref_steal/skb_dstref_restore for ip_route_input callers
> > * [PATCH net-next v2 5/7] staging: octeon: Convert to skb_dst_drop
> > * [PATCH net-next v2 6/7] chtls: Convert to skb_dst_reset
> > * [PATCH net-next v2 7/7] net: Add skb_dst_check_unset
>
> > ***
> >
> > If these findings have caused you to resend the series or submit a
> > separate fix, please add the following tag to your commit message:
> > Tested-by: syzbot@syzkaller.appspotmail.com
>
> Hi Aleksandr!
>
> Could we do something about this Tested-by: tag?
> Since the syzbot CI reports are sent in reply to a series patchwork and
> other tooling will think that syzbot is sending it's Tested-by tag for
> this series.
>
> In some cases we know that the issues found are unrelated, or rather
> expect them to be fixed separately.

FWIW if you notice the reported issues that are completely unrelated,
please let me know.

>
> Could we perhaps indent the tag with a couple of spaces? Not 100% sure
> but I _think_ most tools will match the tags only from start of line.

Sure, that sounds like a very simple solution.
I've adjusted the email template - now there are several leading
whitespaces on the tag line. I hope it will help (otherwise we'll see
what else can be done).

-- 
Aleksandr

Re: [syzbot ci] Re: net: Convert to skb_dstref_steal and skb_dstref_restore

Posted by Jakub Kicinski 1 month, 2 weeks ago

On Wed, 20 Aug 2025 12:45:52 +0200 Aleksandr Nogikh wrote:
> > Could we do something about this Tested-by: tag?
> > Since the syzbot CI reports are sent in reply to a series patchwork and
> > other tooling will think that syzbot is sending it's Tested-by tag for
> > this series.
> >
> > In some cases we know that the issues found are unrelated, or rather
> > expect them to be fixed separately.  
> 
> FWIW if you notice the reported issues that are completely unrelated,
> please let me know.
> 
> > Could we perhaps indent the tag with a couple of spaces? Not 100% sure
> > but I _think_ most tools will match the tags only from start of line.  
> 
> Sure, that sounds like a very simple solution.
> I've adjusted the email template - now there are several leading
> whitespaces on the tag line. I hope it will help (otherwise we'll see
> what else can be done).

Looks like we have received a report with your adjustment in place for:
https://lore.kernel.org/all/20250820092925.2115372-1-jackzxcui1989@163.com/
I can confirm that it fixes the tag propagation for our tooling. 
Thank you!