Re: [PATCH bpf-next 3/3] bpftool: print the comm of the process that loaded the program

[Stanislav Fomichev <sdf@fomichev.me>]

On 10/11, Martin Lau wrote:
> On Fri, Oct 11, 2019 at 09:21:24AM -0700, Stanislav Fomichev wrote:
> > Print recently added created_by_comm along the existing created_by_uid.
> > 
> > Example with loop1.o (loaded via bpftool):
> > 4: raw_tracepoint  name nested_loops  tag b9472b3ff5753ef2  gpl
> >         loaded_at 2019-10-10T13:38:18-0700  uid 0  comm bpftool
> >         xlated 264B  jited 152B  memlock 4096B
> >         btf_id 3
> Hopefully CAP_BPF may avoid uid 0 in the future.
Yeah, but this also requires creating a user with CAP_BPF and running
a daemon under this user.

> What will be in "comm" for the python bcc script?
I guess it will be "python". But at least you get a signal that it's
not some other system daemon :-)

> > 
> > Signed-off-by: Stanislav Fomichev <sdf@google.com>
> > ---
> >  tools/bpf/bpftool/prog.c | 4 +++-
> >  1 file changed, 3 insertions(+), 1 deletion(-)
> > 
> > diff --git a/tools/bpf/bpftool/prog.c b/tools/bpf/bpftool/prog.c
> > index 27da96a797ab..400771a942d7 100644
> > --- a/tools/bpf/bpftool/prog.c
> > +++ b/tools/bpf/bpftool/prog.c
> > @@ -296,7 +296,9 @@ static void print_prog_plain(struct bpf_prog_info *info, int fd)
> >  		print_boot_time(info->load_time, buf, sizeof(buf));
> >  
> >  		/* Piggy back on load_time, since 0 uid is a valid one */
> > -		printf("\tloaded_at %s  uid %u\n", buf, info->created_by_uid);
> > +		printf("\tloaded_at %s  uid %u  comm %s\n", buf,
> > +		       info->created_by_uid,
> > +		       info->created_by_comm);
> >  	}
> >  
> >  	printf("\txlated %uB", info->xlated_prog_len);
> > -- 
> > 2.23.0.700.g56cf767bdb-goog
> >