bpf.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Andrii Nakryiko <andrii.nakryiko@gmail.com>
To: Kumar Kartikeya Dwivedi <memxor@gmail.com>
Cc: bpf <bpf@vger.kernel.org>, "Toke Høiland-Jørgensen" <toke@redhat.com>
Subject: Re: [PATCH bpf] samples: bpf: Fix vmlinux.h generation for XDP samples
Date: Mon, 27 Sep 2021 22:28:01 -0700	[thread overview]
Message-ID: <CAEf4BzYOedEAXN0vCj0+seThLwkwv=e83KWDKkym7B9cE36yAQ@mail.gmail.com> (raw)
In-Reply-To: <20210928045131.psoz6gjb3q3suxc6@apollo.localdomain>

On Mon, Sep 27, 2021 at 9:51 PM Kumar Kartikeya Dwivedi
<memxor@gmail.com> wrote:
>
> On Tue, Sep 28, 2021 at 10:06:09AM IST, Andrii Nakryiko wrote:
> > On Mon, Sep 27, 2021 at 9:16 PM Kumar Kartikeya Dwivedi
> > <memxor@gmail.com> wrote:
> > >
> > > Generate vmlinux.h only from the in-tree vmlinux, and remove enum
> > > declarations that would cause a build failure in case of version
> > > mismatches.
> > >
> > > There are now two options when building the samples:
> > > 1. Compile the kernel to use in-tree vmlinux for vmlinux.h
> > > 2. Override VMLINUX_BTF for samples using something like this:
> > >    make VMLINUX_BTF=/sys/kernel/btf/vmlinux -C samples/bpf
> > >
> > > This change was tested with relative builds, e.g. cases like:
> > >  * make O=build -C samples/bpf
> > >  * make KBUILD_OUTPUT=build -C samples/bpf
> > >  * make -C samples/bpf
> > >  * cd samples/bpf && make
> > >
> > > When a suitable VMLINUX_BTF is not found, the following message is
> > > printed:
> > > /home/kkd/src/linux/samples/bpf/Makefile:333: *** Cannot find a vmlinux
> > > for VMLINUX_BTF at any of "  ./vmlinux", build the kernel or set
> > > VMLINUX_BTF variable.  Stop.
> > >
> > > Cc: Toke Høiland-Jørgensen <toke@redhat.com>
> > > Fixes: 384b6b3bbf0d (samples: bpf: Add vmlinux.h generation support)
> > > Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
> > > ---
> > >  samples/bpf/Makefile                     | 13 ++++++-------
> > >  samples/bpf/xdp_redirect_map_multi.bpf.c |  5 -----
> > >  2 files changed, 6 insertions(+), 12 deletions(-)
> > >
> > > diff --git a/samples/bpf/Makefile b/samples/bpf/Makefile
> > > index 4dc20be5fb96..a05130e91403 100644
> > > --- a/samples/bpf/Makefile
> > > +++ b/samples/bpf/Makefile
> > > @@ -324,15 +324,9 @@ $(obj)/hbm_edt_kern.o: $(src)/hbm.h $(src)/hbm_kern.h
> > >
> > >  VMLINUX_BTF_PATHS ?= $(if $(O),$(O)/vmlinux)                           \
> > >                      $(if $(KBUILD_OUTPUT),$(KBUILD_OUTPUT)/vmlinux)    \
> > > -                    ../../../../vmlinux                                \
> > > -                    /sys/kernel/btf/vmlinux                            \
> > > -                    /boot/vmlinux-$(shell uname -r)
> > > +                    ./vmlinux
> >
> > isn't this relative to samples/bpf subdirectory, so shouldn't it be
> > ../../vmlinux?
> >
>
> I was confused about this too; but it executes this when you call make:
>
> # Trick to allow make to be run from this directory
> all:
>         $(MAKE) -C ../../ M=$(CURDIR) BPF_SAMPLES_PATH=$(CURDIR)
>
> so it fails if the path isn't relative to kernel source root.

Hm... what if we do $(abspath ./vmlinux) here (and probably same for
$(O) and $(KBUILD_OUTPUT) just in case those are relative as well,
would that do the right thing here? The way it is right now it's
extremely confusing (and obviously ../../../../vmlinux never worked
and we never knew that).

>
> > >  VMLINUX_BTF ?= $(abspath $(firstword $(wildcard $(VMLINUX_BTF_PATHS))))
> > >
> > > -ifeq ($(VMLINUX_BTF),)
> > > -$(error Cannot find a vmlinux for VMLINUX_BTF at any of "$(VMLINUX_BTF_PATHS)")
> > > -endif
> > > -
> >
> > [...]
>
> --
> Kartikeya

  reply	other threads:[~2021-09-28  5:28 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-09-28  4:13 [PATCH bpf] samples: bpf: Fix vmlinux.h generation for XDP samples Kumar Kartikeya Dwivedi
2021-09-28  4:36 ` Andrii Nakryiko
2021-09-28  4:51   ` Kumar Kartikeya Dwivedi
2021-09-28  5:28     ` Andrii Nakryiko [this message]
2021-09-28  5:44       ` Kumar Kartikeya Dwivedi

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to='CAEf4BzYOedEAXN0vCj0+seThLwkwv=e83KWDKkym7B9cE36yAQ@mail.gmail.com' \
    --to=andrii.nakryiko@gmail.com \
    --cc=bpf@vger.kernel.org \
    --cc=memxor@gmail.com \
    --cc=toke@redhat.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).