Re: [PATCH bpf-next 2/2] samples: bpf: refactor perf_event user program with libbpf bpf_link

["Daniel T. Lee" <danieltimlee@gmail.com>]

On Wed, Mar 11, 2020 at 6:34 AM John Fastabend <john.fastabend@gmail.com> wrote:
>
> Daniel T. Lee wrote:
> > The bpf_program__attach of libbpf(using bpf_link) is much more intuitive
> > than the previous method using ioctl.
> >
> > bpf_program__attach_perf_event manages the enable of perf_event and
> > attach of BPF programs to it, so there's no neeed to do this
> > directly with ioctl.
> >
> > In addition, bpf_link provides consistency in the use of API because it
> > allows disable (detach, destroy) for multiple events to be treated as
> > one bpf_link__destroy.
> >
> > This commit refactors samples that attach the bpf program to perf_event
> > by using libbbpf instead of ioctl. Also the bpf_load in the samples were
> > removed and migrated to use libbbpf API.
> >
> > Signed-off-by: Daniel T. Lee <danieltimlee@gmail.com>
> > ---
>
> [...]
>
> >
> >  int main(int argc, char **argv)
> >  {
> > +     int prog_fd, *pmu_fd, opt, freq = DEFAULT_FREQ, secs = DEFAULT_SECS;
> > +     struct bpf_program *prog;
> > +     struct bpf_object *obj;
> > +     struct bpf_link **link;
> >       char filename[256];
> > -     int *pmu_fd, opt, freq = DEFAULT_FREQ, secs = DEFAULT_SECS;
> >
> >       /* process arguments */
> >       while ((opt = getopt(argc, argv, "F:h")) != -1) {
> > @@ -165,36 +170,47 @@ int main(int argc, char **argv)
> >       /* create perf FDs for each CPU */
> >       nr_cpus = sysconf(_SC_NPROCESSORS_CONF);
> >       pmu_fd = malloc(nr_cpus * sizeof(int));
> > -     if (pmu_fd == NULL) {
> > -             fprintf(stderr, "ERROR: malloc of pmu_fd\n");
> > +     link = malloc(nr_cpus * sizeof(struct bpf_link *));
> > +     if (pmu_fd == NULL || link == NULL) {
> > +             fprintf(stderr, "ERROR: malloc of pmu_fd/link\n");
> >               return 1;
> >       }
> >
> >       /* load BPF program */
> >       snprintf(filename, sizeof(filename), "%s_kern.o", argv[0]);
> > -     if (load_bpf_file(filename)) {
> > +     if (bpf_prog_load(filename, BPF_PROG_TYPE_PERF_EVENT, &obj, &prog_fd)) {
> >               fprintf(stderr, "ERROR: loading BPF program (errno %d):\n",
> >                       errno);
> > -             if (strcmp(bpf_log_buf, "") == 0)
> > -                     fprintf(stderr, "Try: ulimit -l unlimited\n");
> > -             else
> > -                     fprintf(stderr, "%s", bpf_log_buf);
> >               return 1;
> >       }
> > +
> > +     prog = bpf_program__next(NULL, obj);
> > +     if (!prog) {
> > +             printf("finding a prog in obj file failed\n");
> > +             return 1;
> > +     }
> > +
> > +     map_fd = bpf_object__find_map_fd_by_name(obj, "ip_map");
> > +     if (map_fd < 0) {
> > +             printf("finding a ip_map map in obj file failed\n");
> > +             return 1;
> > +     }
> > +
> >       signal(SIGINT, int_exit);
> >       signal(SIGTERM, int_exit);
> >
> >       /* do sampling */
> >       printf("Sampling at %d Hertz for %d seconds. Ctrl-C also ends.\n",
> >              freq, secs);
> > -     if (sampling_start(pmu_fd, freq) != 0)
> > +     if (sampling_start(pmu_fd, freq, prog, link) != 0)
> >               return 1;
> >       sleep(secs);
> > -     sampling_end(pmu_fd);
> > +     sampling_end(link);
> >       free(pmu_fd);
> > +     free(link);
>
> Not really a problem with this patch but on error we don't free() memory but
> then on normal exit there is a free() its a bit inconsistent. How about adding
> free on errors as well?

I think you're right.
I'll add free() on errors to keep it consistent.
Will apply feedback right away!

>
> >
> >       /* output sample counts */
> > -     print_ip_map(map_fd[0]);
> > +     print_ip_map(map_fd);
> >
> >       return 0;
> >  }
>
> [...]
>
> >  static void print_ksym(__u64 addr)
> > @@ -137,6 +136,7 @@ static inline int generate_load(void)
> >  static void test_perf_event_all_cpu(struct perf_event_attr *attr)
> >  {
> >       int nr_cpus = sysconf(_SC_NPROCESSORS_CONF);
> > +     struct bpf_link **link = malloc(nr_cpus * sizeof(struct bpf_link *));
>
> need to check if its null? Its not going to be very friendly to segfault
> later. Or maybe I'm missing the check.
>

Also, checking whether it is null will be more safe.
I'll apply and send next version patch.

> >       int *pmu_fd = malloc(nr_cpus * sizeof(int));
> >       int i, error = 0;
> >
> > @@ -151,8 +151,12 @@ static void test_perf_event_all_cpu(struct perf_event_attr *attr)
> >                       error = 1;
> >                       goto all_cpu_err;
> >               }
> > -             assert(ioctl(pmu_fd[i], PERF_EVENT_IOC_SET_BPF, prog_fd[0]) == 0);
> > -             assert(ioctl(pmu_fd[i], PERF_EVENT_IOC_ENABLE) == 0);
> > +             link[i] = bpf_program__attach_perf_event(prog, pmu_fd[i]);
> > +             if (link[i] < 0) {
> > +                     printf("bpf_program__attach_perf_event failed\n");
> > +                     error = 1;
> > +                     goto all_cpu_err;
> > +             }
> >       }
> >
> >       if (generate_load() < 0) {
> > @@ -161,11 +165,11 @@ static void test_perf_event_all_cpu(struct perf_event_attr *attr)
> >       }
> >       print_stacks();
> >  all_cpu_err:
> > -     for (i--; i >= 0; i--) {
> > -             ioctl(pmu_fd[i], PERF_EVENT_IOC_DISABLE);
> > -             close(pmu_fd[i]);
> > -     }
> > +     for (i--; i >= 0; i--)
> > +             bpf_link__destroy(link[i]);
> > +
> >       free(pmu_fd);
> > +     free(link);
> >       if (error)
> >               int_exit(0);
> >  }
>
> Thanks,
> John

Thank you for your time and effort for the review.

Best,
Daniel