On Fri, Nov 04, 2022 at 12:39:55AM +0530, Kumar Kartikeya Dwivedi wrote:
> -
> - for (i = 0; i < nr_off; i++) {
> + rec->cnt = 0;
> + for (i = 0; i < cnt; i++) {
> const struct btf_type *t;
> s32 id;
>
> @@ -3500,28 +3499,24 @@ struct bpf_map_value_off *btf_parse_kptrs(const struct btf *btf,
> ret = -EINVAL;
> goto end_mod;
> }
> - tab->off[i].kptr.dtor = (void *)addr;
> + rec->fields[i].kptr.dtor = (void *)addr;
> }
>
> - tab->off[i].offset = info_arr[i].off;
> - tab->off[i].type = info_arr[i].type;
> - tab->off[i].kptr.btf_id = id;
> - tab->off[i].kptr.btf = kernel_btf;
> - tab->off[i].kptr.module = mod;
> + rec->fields[i].offset = info_arr[i].off;
> + rec->fields[i].type = info_arr[i].type;
> + rec->fields[i].kptr.btf_id = id;
> + rec->fields[i].kptr.btf = kernel_btf;
> + rec->fields[i].kptr.module = mod;
> + rec->cnt++;
> }
> - tab->nr_off = nr_off;
> - return tab;
> + rec->cnt = cnt;
> + return rec;
This is weird. You also undo this assignment in the next patch.
What is the point of rec->cnt = 0;
followed by rec->cnt++
just to be overwritten with rec->cnt = cnt;
??
