Re: [PATCH v3 bpf-next 3/3] libbpf: auto-generate list of BPF helper definitions

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

 



On Sun, Oct 6, 2019 at 7:24 PM Andrii Nakryiko
<andrii.nakryiko@xxxxxxxxx> wrote:
>
> On Sun, Oct 6, 2019 at 5:32 PM Alexei Starovoitov
> <alexei.starovoitov@xxxxxxxxx> wrote:
> >
> > On Sun, Oct 6, 2019 at 5:13 PM Andrii Nakryiko
> > <andrii.nakryiko@xxxxxxxxx> wrote:
> > >
> > > On Sun, Oct 6, 2019 at 4:56 PM Alexei Starovoitov
> > > <alexei.starovoitov@xxxxxxxxx> wrote:
> > > >
> > > > On Sat, Oct 5, 2019 at 10:46 PM Andrii Nakryiko <andriin@xxxxxx> wrote:
> > > > >
> > > > > Get rid of list of BPF helpers in bpf_helpers.h (irony...) and
> > > > > auto-generate it into bpf_helpers_defs.h, which is now included from
> > > > > bpf_helpers.h.
> > > > >
> > > > > Suggested-by: Alexei Starovoitov <ast@xxxxxx>
> > > > > Signed-off-by: Andrii Nakryiko <andriin@xxxxxx>
> > > > > ---
> > > > >  tools/lib/bpf/.gitignore    |   1 +
> > > > >  tools/lib/bpf/Makefile      |  10 +-
> > > > >  tools/lib/bpf/bpf_helpers.h | 264 +-----------------------------------
> > > > >  3 files changed, 10 insertions(+), 265 deletions(-)
> > > >
> > > > This patch doesn't apply to bpf-next.
> > >
> > > Yes, it has to be applied on top of bpf_helpers.h move patch set. I
> > > can bundle them together and re-submit as one patch set, but I don't
> > > think there were any remaining issues besides the one solved in this
> > > patch set (independence from any specific bpf.h UAPI), so that one can
> > > be applied as is.
> >
> > It looks to me that auto-gen of bpf helpers set is ready,
> > whereas move is till being debated.
> > I also would like to test autogen-ed .h in my environment first
> > before we move things around.
>
> Alright, will post v4 based on master with bpf_helpers.h still in selftests/bpf

Posted v4 w/ completely different Makefile change. For libbpf it's
going to be the one from v3 of this patch set.

But I'm not sure what debate you mean for bpf_helpers.h move. The only
contentious issue was bpf_helpers.h depending on BPF_FUNC_xxx enum
values, which is solved/bypassed by this auto-generation approach. So
if we are landing auto-generation of helpers, there is nothing for me
to address for bpf_helpers.h move.

Keep in mind, that depending on order of applying this and bpf_helpers
move patchsets, it will be either:

1. Apply bpf_helpers.h move patchset
2. Apply v3 of this patch set

Or:

1. Apply v4 of this patch set
2. I'll have to rebase bpf_helpers.h move patchset after that (it
probably won't apply cleanly)
3. Follow-up patches to undo selftests/bpf Makefile change and re-do
libbpf's Makefile change.

The latter one is the same end result with more work for me (part of
which I already did for v4, some part still needs to be done).



[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