Re: [PATCH net-next 7/7] net: mctp: Add bind lookup test

[Matt Johnston <matt@codeconstruct.com.au>]

Hi Alexandre,

Thanks for the report.

On Mon, 2025-08-11 at 13:35 +0200, Alexandre Ghiti wrote:

> I'm hitting different issues with this test.
> 
> With a simple riscv defconfig, I get the following warning a bit after 
> the test succeeds:
> 
> [    5.991567] WARNING: CPU: 1 PID: 0 at net/netlink/af_netlink.c:402 
> netlink_sock_destruct+0x3a/0x6a
> [    5.992039] Modules linked in:

I can reproduce it with the um kunit runner if I enable kasan. The skb_pkt
shouldn't be freed, I've submitted a fix for that.

Cheers,
Matt

> [    5.992384] CPU: 1 UID: 0 PID: 0 Comm: swapper/1 Tainted: 
> G                 N  6.17.0-rc1-00005-g1692753ec9e8-dirty #40 NONE
> [    5.992502] Tainted: [N]=TEST
> [    5.992521] Hardware name: riscv-virtio,qemu (DT)
> [    5.992604] epc : netlink_sock_destruct+0x3a/0x6a
> [    5.992643]  ra : netlink_sock_destruct+0x18/0x6a
> [    5.992669] epc : ffffffff809bdf26 ra : ffffffff809bdf04 sp : 
> ff2000000000bd90
> [    5.992685]  gp : ffffffff81a1b458 tp : ff600000802ce000 t0 : 
> 0000000000000040
> [    5.992700]  t1 : 0000000000000000 t2 : 0000000080150012 s0 : 
> ff2000000000bdb0
> [    5.992715]  s1 : ff600000825a0800 a0 : ff600000825a08a8 a1 : 
> 0000000000000068
> [    5.992729]  a2 : 0000000000000008 a3 : 000000000000000a a4 : 
> 0000000000000000
> [    5.992743]  a5 : ffffffffffffce00 a6 : ffffffffffffffff a7 : 
> ffffffff81a20810
> [    5.992757]  s2 : ff600000825a0800 s3 : ffffffff81a7c040 s4 : 
> ff2000000000be60
> [    5.992770]  s5 : 0000000000000000 s6 : 000000000000000a s7 : 
> 0000000000000000
> [    5.992784]  s8 : ff600003fff1e580 s9 : 0000000000000001 s10: 
> ffffffff81890a80
> [    5.992798]  s11: ff600000825a0c28 t3 : 0000000000000002 t4 : 
> 0000000000000402
> [    5.992811]  t5 : ff60000080069168 t6 : ff60000080069170
> [    5.992824] status: 0000000200000120 badaddr: ffffffff809bdf26 cause: 
> 0000000000000003
> [    5.992925] [<ffffffff809bdf26>] netlink_sock_destruct+0x3a/0x6a
> [    5.993015] [<ffffffff809369b8>] __sk_destruct+0x22/0x1aa
> [    5.993034] [<ffffffff809385b0>] sk_destruct+0x46/0x50
> [    5.993049] [<ffffffff8093860a>] __sk_free+0x50/0xc6
> [    5.993062] [<ffffffff809386aa>] sk_free+0x2a/0x46
> [    5.993077] [<ffffffff809bdb08>] deferred_put_nlk_sk+0x42/0x62
> [    5.993089] [<ffffffff800a5b80>] rcu_core+0x1ba/0x5c8
> [    5.993106] [<ffffffff800a5f9a>] rcu_core_si+0xc/0x14
> [    5.993119] [<ffffffff8002d8de>] handle_softirqs+0x110/0x2e8
> [    5.993133] [<ffffffff8002dbd2>] __irq_exit_rcu+0xd0/0xfa
> [    5.993144] [<ffffffff8002ddbe>] irq_exit_rcu+0xc/0x14
> [    5.993156] [<ffffffff80b438e0>] handle_riscv_irq+0x64/0x74
> [    5.993177] [<ffffffff80b4f88a>] call_on_irq_stack+0x32/0x40
> 
> The following diff fixes it (completely unsure this is the right fix 
> though):
> 
> diff --git a/net/mctp/test/route-test.c b/net/mctp/test/route-test.c
> index fb6b46a952cb4..03f2f76c97f75 100644
> --- a/net/mctp/test/route-test.c
> +++ b/net/mctp/test/route-test.c
> @@ -1585,7 +1585,6 @@ static void mctp_test_bind_lookup(struct kunit *test)
>          }
> 
>   cleanup:
> -       kfree_skb(skb_sock);
>          kfree_skb(skb_pkt);
> 
>          /* Drop all binds */
> 
> 
> But then with a more consequent config attached, even with the previous 
> fix, I'm hitting:
> 
> [  137.494285] kernel BUG at mm/slub.c:563!
> [  137.494365] Kernel BUG [#2]
> [  137.494388] Modules linked in:
> [  137.494449] CPU: 4 UID: 0 PID: 4699 Comm: kunit_try_catch Tainted: 
> G      D W        N  6.17.0-rc1+ #16 PREEMPT(voluntary)
> [  137.494500] Tainted: [D]=DIE, [W]=WARN, [N]=TEST
> [  137.494514] Hardware name: riscv-virtio,qemu (DT)
> [  137.494534] epc : kmem_cache_free+0x334/0x39e
> [  137.494588]  ra : kmem_cache_free+0x76/0x39e
> [  137.494617] epc : ffffffff802e8ec4 ra : ffffffff802e8c06 sp : 
> ff2000000b6cbb60
> [  137.494637]  gp : ffffffff82ac7b70 tp : ff60000087cc1f00 t0 : 
> ff60000087ba6820
> [  137.494657]  t1 : 0000000000000000 t2 : 0000000000000000 s0 : 
> ff2000000b6cbbb0
> [  137.494676]  s1 : ffffffff80c34db8 a0 : ff1c0000021ee900 a1 : 
> ff60000087ba6680
> [  137.494695]  a2 : ff60000087ba6680 a3 : ffffffff82e4a858 a4 : 
> ff600003fff21d80
> [  137.494715]  a5 : ff60000087ba67e0 a6 : 000000000002dc04 a7 : 
> 0000000000000000
> [  137.494733]  s2 : ff60000087ba6680 s3 : ff600000802cc900 s4 : 
> ff1c0000021ee900
> [  137.494754]  s5 : 0000000000000002 s6 : ff600000833d9b00 s7 : 
> ff60000087b55a80
> [  137.494772]  s8 : 0000000000000000 s9 : ff2000000b6cbd18 s10: 
> ffffffff81e29a90
> [  137.494791]  s11: 000000000000060e t3 : 0000000000000000 t4 : 
> 0000000000000000
> [  137.494810]  t5 : 0000000000000000 t6 : ff60000087ba6685
> [  137.494826] status: 0000000200000120 badaddr: ffffffff802e8ec4 cause: 
> 0000000000000003
> [  137.494849] [<ffffffff802e8ec4>] kmem_cache_free+0x334/0x39e
> [  137.494883] [<ffffffff80c34db8>] skb_free_head+0xca/0x112
> [  137.494922] [<ffffffff80c37bba>] skb_release_data+0xe2/0x112
> [  137.494952] [<ffffffff80c3a202>] sk_skb_reason_drop+0x40/0x124
> [  137.494984] [<ffffffff80e57afe>] mctp_test_bind_lookup+0x15e/0x3f4
> [  137.495019] [<ffffffff80623454>] kunit_try_run_case+0x4c/0x13e
> [  137.495051] [<ffffffff80625a10>] 
> kunit_generic_run_threadfn_adapter+0x1a/0x34
> [  137.495080] [<ffffffff80064a9c>] kthread+0xea/0x1c2
> [  137.495111] [<ffffffff8001b42e>] ret_from_fork_kernel+0x10/0x162
> [  137.495139] [<ffffffff80e9d192>] ret_from_fork_kernel_asm+0x16/0x18
> [  137.495199] Code: 80a3 00e7 5097 ffd5 80e7 9880 9002 b715 ec56 e85a 
> (9002) 9b61
> [  137.495760] ---[ end trace 0000000000000000 ]---
> 
> Let me know if I can do anything to help.
> 
> Thanks,
> 
> Alex