Re: [PATCH] Kernel: Block non-UID-0 profiling completely

["Peter Müller" <peter.mueller@ipfire.org>]

[-- Attachment #1: Type: text/plain, Size: 1972 bytes --]

Hello Arne,
hello Michael,

thanks for your replies.

Convinced there is no way around rebasing the kernel patch, I will do so and submit a
second version within the next few days. :-)

All the best,
Peter Müller


> Such .orig files are created by patch if a hunk not apply without fuzzing.
> rebase the patch to the current kernel source should help.
> 
> Arne
> 
> Am 2022-02-04 17:56, schrieb Michael Tremer:
>> Hello,
>>
>>> On 4 Feb 2022, at 13:45, Peter Müller <peter.mueller(a)ipfire.org> wrote:
>>>
>>> Hello Michael,
>>>
>>> [...]
>>>
>>>>>> #lib/modules/KVER-ipfire/build/include/linux/perf_event.h
>>>>>> +#lib/modules/KVER-ipfire/build/include/linux/perf_event.h.orig
>>>>>
>>>>> We should not install any files like this.
>>>>
>>>> Hm. I wonder where it came from. Perhaps I forgot a "./make.sh clean" after the first attempt.
>>>>
>>>> Will check and report back.
>>>
>>> after running a clean build, this file stays present. It's attached to this mail.
>>>
>>> It is odd to see this happening, as the additional "patch" command introduced by this
>>> patch does not differ from those already present in the kernel's LFS file, except for
>>> the patch file itself.
>>
>> If the patch doesn’t match exactly, patch will create backup files so
>> that you can fix anything manually if things broke.
>>
>> What you can do is either to rediff the patch that it applies cleanly,
>> or you can use this option:
>>
>>        --no-backup-if-mismatch
>>
>>           Do not back up a file if the patch does not match the file
>>           exactly and if backups are not otherwise requested.  This is
>>           the default if patch  is conforming to POSIX.
>>
>> This should however be the default.
>>
>> -Michael
>>
>>>
>>> Can you give me any hint on where to look next?
>>>
>>> Thanks, and best regards,
>>> Peter Müller<perf_event.h.orig>
> 
> 
> 
> 
>