[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <CAP-5=fW=v9oFr1+B_JcEe+kG60wB2W0dtKkD7Fmkh8KSKdu7Aw@mail.gmail.com>
Date: Fri, 6 Oct 2023 13:22:05 -0700
From: Ian Rogers <irogers@...gle.com>
To: Andrii Nakryiko <andrii.nakryiko@...il.com>
Cc: Quentin Monnet <quentin@...valent.com>,
Alexei Starovoitov <ast@...nel.org>,
Daniel Borkmann <daniel@...earbox.net>,
Andrii Nakryiko <andrii@...nel.org>,
Martin KaFai Lau <martin.lau@...ux.dev>,
Song Liu <song@...nel.org>,
Yonghong Song <yonghong.song@...ux.dev>,
John Fastabend <john.fastabend@...il.com>,
KP Singh <kpsingh@...nel.org>,
Stanislav Fomichev <sdf@...gle.com>,
Hao Luo <haoluo@...gle.com>, Jiri Olsa <jolsa@...nel.org>,
bpf@...r.kernel.org, linux-kernel@...r.kernel.org,
Alan Maguire <alan.maguire@...cle.com>
Subject: Re: [PATCH v5 1/2] bpftool: Align output skeleton ELF code
On Fri, Oct 6, 2023 at 1:20 PM Andrii Nakryiko
<andrii.nakryiko@...il.com> wrote:
>
> On Thu, Oct 5, 2023 at 3:03 PM Ian Rogers <irogers@...gle.com> wrote:
> >
> > libbpf accesses the ELF data requiring at least 8 byte alignment,
> > however, the data is generated into a C string that doesn't guarantee
> > alignment. Fix this by assigning to an aligned char array. Use sizeof
> > on the array, less one for the \0 terminator, rather than generating a
> > constant.
> >
> > Fixes: a6cc6b34b93e ("bpftool: Provide a helper method for accessing skeleton's embedded ELF data")
> > Signed-off-by: Ian Rogers <irogers@...gle.com>
> > Acked-by: Quentin Monnet <quentin@...valent.com>
> > Reviewed-by: Alan Maguire <alan.maguire@...cle.com>
> > ---
> > tools/bpf/bpftool/gen.c | 15 +++++++++------
> > 1 file changed, 9 insertions(+), 6 deletions(-)
> >
> > diff --git a/tools/bpf/bpftool/gen.c b/tools/bpf/bpftool/gen.c
> > index 2883660d6b67..b8ebcee9bc56 100644
> > --- a/tools/bpf/bpftool/gen.c
> > +++ b/tools/bpf/bpftool/gen.c
> > @@ -1209,7 +1209,7 @@ static int do_skeleton(int argc, char **argv)
> > codegen("\
> > \n\
> > \n\
> > - s->data = (void *)%2$s__elf_bytes(&s->data_sz); \n\
> > + s->data = (void *)%1$s__elf_bytes(&s->data_sz); \n\
>
> Seems like you based this on top of bpf tree, can you please rebase
> onto bpf-next, it has a small change here and I can't apply it cleanly
> anymore. Other than that it looks good. Thanks!
Sure thing, sorry for so many V-s, we'll get there in the end.
Thanks,
Ian
> > \n\
> > obj->skeleton = s; \n\
> > return 0; \n\
> > @@ -1218,12 +1218,12 @@ static int do_skeleton(int argc, char **argv)
> > return err; \n\
> > } \n\
> > \n\
> > - static inline const void *%2$s__elf_bytes(size_t *sz) \n\
> > + static inline const void *%1$s__elf_bytes(size_t *sz) \n\
> > { \n\
> > - *sz = %1$d; \n\
> > - return (const void *)\"\\ \n\
> > - "
> > - , file_sz, obj_name);
> > + static const char data[] __attribute__((__aligned__(8))) = \"\\\n\
> > + ",
> > + obj_name
> > + );
> >
> > /* embed contents of BPF object file */
> > print_hex(obj_data, file_sz);
> > @@ -1231,6 +1231,9 @@ static int do_skeleton(int argc, char **argv)
> > codegen("\
> > \n\
> > \"; \n\
> > + \n\
> > + *sz = sizeof(data) - 1; \n\
> > + return (const void *)data; \n\
> > } \n\
> > \n\
> > #ifdef __cplusplus \n\
> > --
> > 2.42.0.609.gbb76f46606-goog
> >
Powered by blists - more mailing lists