netdev - Re: [PATCH v2 bpf-next 2/2] btf: expose API to work with raw btf data

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Tue, 5 Feb 2019 21:46:14 -0800
From:   Andrii Nakryiko <andrii.nakryiko@...il.com>
To:     Alexei Starovoitov <alexei.starovoitov@...il.com>
Cc:     Andrii Nakryiko <andriin@...com>, Song Liu <songliubraving@...com>,
        Yonghong Song <yhs@...com>, Alexei Starovoitov <ast@...com>,
        Martin Lau <kafai@...com>, netdev@...r.kernel.org,
        Daniel Borkmann <daniel@...earbox.net>
Subject: Re: [PATCH v2 bpf-next 2/2] btf: expose API to work with raw btf data

On Tue, Feb 5, 2019 at 7:07 PM Alexei Starovoitov
<alexei.starovoitov@...il.com> wrote:
>
> On Tue, Feb 05, 2019 at 04:29:49PM -0800, Andrii Nakryiko wrote:
> > This patch exposes two new APIs btf__get_raw_data_size() and
> > btf__get_raw_data() that allows to get a copy of raw BTF data out of
> > struct btf. This is useful for external programs that need to manipulate
> > raw data, e.g., pahole using btf__dedup() to deduplicate BTF type info
> > and then writing it back to file.
> >
> > Signed-off-by: Andrii Nakryiko <andriin@...com>
> > Acked-by: Song Liu <songliubraving@...com>
> > ---
> >  tools/lib/bpf/btf.c      | 10 ++++++++++
> >  tools/lib/bpf/btf.h      |  2 ++
> >  tools/lib/bpf/libbpf.map |  2 ++
> >  3 files changed, 14 insertions(+)
> >
> > diff --git a/tools/lib/bpf/btf.c b/tools/lib/bpf/btf.c
> > index 1c2ba7182400..34bfb3641aac 100644
> > --- a/tools/lib/bpf/btf.c
> > +++ b/tools/lib/bpf/btf.c
> > @@ -437,6 +437,16 @@ int btf__fd(const struct btf *btf)
> >       return btf->fd;
> >  }
> >
> > +__u32 btf__get_raw_data_size(const struct btf *btf)
> > +{
> > +     return btf->data_size;
> > +}
> > +
> > +void btf__get_raw_data(const struct btf *btf, char *data)
> > +{
> > +     memcpy(data, btf->data, btf->data_size);
> > +}
>
> I cannot think of any other way to use this api,
> but to call btf__get_raw_data_size() first,
> then malloc that much memory and then call btf__get_raw_data()
> to store btf into it.
>
> If so, may be api should be single call that allocates, copies,
> and returns pointer to new mem and its size?
> Probably less error prone?
>

I don't have strong preference, but providing pointer to allocated memory
seems more flexible and allows more clever/optimal use of memory from caller
side. E.g., instead of doing two mallocs, you can imagine doing something
like:

int max_size = max(btf__get_raw_data_size(btf),
                   btf_ext__get_raw_data_size(btf_ext));
char *m = malloc(max_size);
btf__get_raw_data(btf, m);
dump_btf_section_to_file(m, some_file);
btf_ext__get_raw_data(btf_ext, m);
dump_btf_ext_section_to_file(m, some_file);
free(m);

Also, pointer to memory could be mmap()'ed file, for instance. In general,
for a library it might be a good thing to not be prescriptive as to how one
gets that piece of memory.

If those examples are not convincing enough, I'm happy to go with single
btf__get_raw_data() call doing allocation and returning pointer.