Re: [trace:trace/core 16/16] WARNING: modpost: vmlinux: section mismatch in reference: trace_printk_init_buffers.part.0+0xc8 (section: .text.unlikely) -> initcall_level_names (section: .init.data)

Woradorn Laodhanadhaworn <[email protected]> Fri, 31 Jul 2026 09:31:36 +0700
Newsgroups dev.linux.lists.oe-kbuild-all
Message-ID <[email protected]>
On 30/7/2569 BE 20:28, Steven Rostedt wrote:
> On Thu, 30 Jul 2026 14:04:20 +0700
> Woradorn Laodhanadhaworn <[email protected]> wrote:
> 
>> I reproduced this locally with commit
>> `c33330b2ec170dd38b3284a72bc6a07a49cc9a65`, and it also reports:
>>
>> WARNING: modpost: vmlinux: section mismatch in reference:
>> trace_printk_init_buffers.part.0+0xc8 (section: .text.unlikely) ->
>> initcall_level_names (section: .init.data)
>>
>> Then I tried the same configuration and reproduction steps with the parent
>> commit, `b6ccdca8d11c28338776811f0b9024d92c4bedb4`. It doesn't report anything
>> related to `modpost: vmlinux: section mismatch in reference:
>> trace_printk_init_buffers.part`.
> 
> Did you do a "make mrproper" between both builds?
> 

Yes, I did, but with a different approach. I removed the build_dir and 
recreated it.

>>
>> So it looks like this commit introduced the warning. I'm not sure where to look
>> yet, but I'll dig into `modpost.c` to understand the root cause. If you have any 
>> suggestions on where I should start looking, I'd really appreciate them.
>>
>> Also, I'd like to ask how urgent this issue is. It might take me some time to
>> find the root cause.
> 
> It's only a warning with the W=1 build. There's many other warnings there
> too, so I don't think it's high priority.
> 

Understood, thanks.

> Thanks,
> 
> -- Steve


Thanks,
Woradorn