Re: [PATCH v2 bpf-next 1/9] libbpf: name internal functions consistently

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Tue, Jan 2, 2024 at 11:01 AM Andrii Nakryiko <andrii@xxxxxxxxxx> wrote:
>
> For a while now all new internal libbpf functions stopped using
> <obj>__<method>() naming, which was historically used both for public
> APIs and all the helper functions that can be thought of as "methods" of
> libbpf "objects" (bpf_object, bpf_map, bpf_program, etc).

I don't think this shift to single underscore was discussed before.
I could have missed it. Personally I was under the impression that
we're still doing double for methods.
This convention came from perf, since back then
libbpf was part of it and perf is using double everywhere.
For external api-s and for internal functions.
I feel we should continue doing double for existing objects.
This rename feels too churn-y.

At the same time I agree that a public function looking different from
internal is a good thing to have.
We have LIBBPF_API that is used in the headers.
Maybe we should start using something similar in .c files
than there will be no confusion.

Not a strong opinion.

Eduard,
what's your take?





[Index of Archives]     [Linux Samsung SoC]     [Linux Rockchip SoC]     [Linux Actions SoC]     [Linux for Synopsys ARC Processors]     [Linux NFS]     [Linux NILFS]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]


  Powered by Linux