[Bug] "WARNING in should_fail_ex" in Linux kernel v6.14

[Bug] "WARNING in should_fail_ex" in Linux kernel v6.14

[John]

Dear Linux Kernel Maintainers,

I hope this message finds you well.

I am writing to report a potential vulnerability I encountered during
testing of the Linux Kernel version v6.14.

Git Commit: 38fec10eb60d687e30c8c6b5420d86e8149f7557 (tag: v6.14)

Bug Location: net/ipv4/ipmr.c:440 ipmr_free_table net/ipv4/ipmr.c:440

Bug report: https://pastebin.com/xkfF5DBt

Complete log: https://pastebin.com/uCfqY4D8

Entire kernel config: https://pastebin.com/MRWGr3nv

Root Cause Analysis:
The kernel warning is triggered in ipmr_free_table() at
net/ipv4/ipmr.c:440, where the multicast routing table (mr_table) is
being freed during network namespace exit (ipmr_rules_exit).
The warning indicates that the multicast forwarding cache count
(mfc_cache_cnt) is non-zero, implying that resources were not
correctly cleaned up prior to netns teardown.
This suggests a possible bug in reference counting or teardown logic
of the IPv4 multicast routing infrastructure.
Additionally, the environment is running with fail_usercopy fault
injection enabled, which deliberately triggers copy-from-user failures
to test kernel error paths.
While these failures are expected, they may interact with multicast
socket setup/teardown paths, exacerbating latent issues.

At present, I have not yet obtained a minimal reproducer for this
issue. However, I am actively working on reproducing it, and I will
promptly share any additional findings or a working reproducer as soon
as it becomes available.

Thank you very much for your time and attention to this matter. I
truly appreciate the efforts of the Linux kernel community.

Best regards,
John

Re: [Bug] "WARNING in should_fail_ex" in Linux kernel v6.14

[Kuniyuki Iwashima]

From: John <john.cs.hey@gmail.com>
Date: Wed, 21 May 2025 22:13:15 +0800
> Dear Linux Kernel Maintainers,
> 
> I hope this message finds you well.
> 
> I am writing to report a potential vulnerability I encountered during
> testing of the Linux Kernel version v6.14.
> 
> Git Commit: 38fec10eb60d687e30c8c6b5420d86e8149f7557 (tag: v6.14)
> 
> Bug Location: net/ipv4/ipmr.c:440 ipmr_free_table net/ipv4/ipmr.c:440
> 
> Bug report: https://pastebin.com/xkfF5DBt
> 
> Complete log: https://pastebin.com/uCfqY4D8
> 
> Entire kernel config: https://pastebin.com/MRWGr3nv
> 
> Root Cause Analysis:
> The kernel warning is triggered in ipmr_free_table() at
> net/ipv4/ipmr.c:440, where the multicast routing table (mr_table) is
> being freed during network namespace exit (ipmr_rules_exit).
> The warning indicates that the multicast forwarding cache count
> (mfc_cache_cnt) is non-zero, implying that resources were not
> correctly cleaned up prior to netns teardown.
> This suggests a possible bug in reference counting or teardown logic
> of the IPv4 multicast routing infrastructure.
> Additionally, the environment is running with fail_usercopy fault
> injection enabled, which deliberately triggers copy-from-user failures
> to test kernel error paths.
> While these failures are expected, they may interact with multicast
> socket setup/teardown paths, exacerbating latent issues.

I think this was fixed by

commit c46286fdd6aa1d0e33c245bcffe9ff2428a777bd
Author: Paolo Abeni <pabeni@redhat.com>
Date:   Thu May 15 18:49:26 2025 +0200

    mr: consolidate the ipmr_can_free_table() checks.