Re: [PATCH bpf-next v3] bpftool: use only nftw for file tree parsing

On Mon, 20 Jul 2020 at 01:13, Quentin Monnet [off-list ref] wrote:

On 17/07/2020 23:55, Tony Ambardar wrote:

quoted

The bpftool sources include code to walk file trees, but use multiple
frameworks to do so: nftw and fts. While nftw conforms to POSIX/SUSv3 and
is widely available, fts is not conformant and less common, especially on
non-glibc systems. The inconsistent framework usage hampers maintenance
and portability of bpftool, in particular for embedded systems.

Standardize code usage by rewriting one fts-based function to use nftw and
clean up some related function warnings by extending use of "const char *"
arguments. This change helps in building bpftool against musl for OpenWrt.

[...]

quoted

 int build_pinned_obj_table(struct pinned_obj_table *tab,
                         enum bpf_obj_type type)
 {

[...]

quoted

      while ((mntent = getmntent(mntfile))) {

[...]

quoted

-             while ((ftse = fts_read(fts))) {

[...]

quoted

+             if (nftw(path, do_build_table_cb, nopenfd, flags) == -1)
+                     break;

Sorry I missed that on the previous reviews; but I think a simple break
out of the loop changes the previous behaviour, we should instead
"return -1" from build_pinned_obj_table() if nftw() returns -1, as we
were doing so far.

Thanks for the catch. Sorry, that's totally my fault: this was meant
to be an "err = nftw(...)" with a "return err" on the common exit
path, similar to the rest of the patch. Let me fix and resend.

Kind regards,
Tony

Looks good otherwise.

quoted

      }
      fclose(mntfile);
      return 0;